Imaging Ca2+ Responses During Shigella Infection of Epithelial Cells

Summary

Here, we present protocols to visualize calcium (Ca²⁺) responses elicited by HeLa cells infected by Shigella. By optimizing the parameters of bacterial infection and imaging with Ca²⁺ fluorescent probes, atypical global and local Ca²⁺ signals induced by bacteria over a large range of infection kinetics are characterized.

Abstract

Ca²⁺ is a ubiquitous ion involved in all known cellular processes. While global Ca²⁺ responses may affect cell fate, local variations in free Ca²⁺ cytosolic concentrations, linked to release from internal stores or an influx through plasma membrane channels, regulate cortical cell processes. Pathogens that adhere to or invade host cells trigger a reorganization of the actin cytoskeleton underlying the host plasma membrane, which likely affects both global and local Ca²⁺ signaling. Because these events may occur at low frequencies in a pseudo-stochastic manner over extended kinetics, the analysis of Ca²⁺ signals induced by pathogens raises major technical challenges that need to be addressed.

Here, we report protocols for the detection of global and local Ca²⁺ signals upon a Shigella infection of epithelial cells. In these protocols, artefacts linked to a prolonged exposure and photodamage associated with the excitation of Ca²⁺ fluorescent probes are troubleshot by stringently controlling the acquisition parameters over defined time periods during a Shigella invasion. Procedures are implemented to rigorously analyze the amplitude and frequency of global cytosolic Ca²⁺ signals during extended infection kinetics using the chemical probe Fluo-4.

Introduction

Ca²⁺ regulates all known cell processes, including cytoskeletal reorganization, inflammatory responses, and cell death pathways related to host-pathogen interactions^1,2,3. Under physiological conditions, basal cytosolic Ca²⁺ concentrations are low, in the hundreds of nM range, but can be subjected to transient increases upon agonist stimulation. These variations often show oscillatory behavior through the action of pumps and channels at the plasma and endoplasmic reticulum membranes. The pattern of these oscillations is characterized by the period, duration, and amplitude of Ca²⁺ increases, and is decrypted by cells which, in turn, trigger specific responses in what is known as the Ca²⁺ code^4,5. A sustained increase in the cytosolic Ca²⁺ concentration under pathological conditions may lead to cell death associated with the permeabilization of mitochondrial membranes and the release of pro-apoptotic or necrotic factors^6,7.

Shigella, the causative agent of bacillary dysentery, invades epithelial cells by injecting effectors into host cells using a type III secretion system (T3SS)^8,9. A Shigella invasion of host cells is associated with local and global Ca²⁺ signals elicited by the T3SS. As for pore-forming toxins, the T3SS translocon that inserts into host cell membranes and is required for the injection of T3SS effectors is likely responsible for the activation of PLC and the inositol (1, 4, 5) trisphosphate (InsP3)-dependent Ca²⁺ release. The combination of the localized PLC stimulation and the accumulation of polymerized actin at sites of a Shigella invasion result in an atypically long-lasting InsP3-dependent Ca²⁺ release¹⁰. The type III effector IpgD, a phosphatidyl 4,5 bisphosphate (PIP2)-4-phosphatase, limits the local amount of PIP2, thereby controlling the quantity of available substrate for PLC to generate InsP3, which contributes to the confinement of local Ca²⁺ responses at bacterial invasion sites^11,12. These local Ca²⁺ responses likely contribute to the actin polymerization at Shigella invasion sites¹⁰. Global Ca²⁺ responses that are also elicited by Shigella, however, are dispensable for the bacterial invasion process but trigger the opening of connexin hemichannels at the plasma membrane and the release of ATP in the extracellular compartment. Released ATP acting in a paracrine manner, in turn, stimulates Ca²⁺ oscillatory responses in cells next to the infected cell. IpgD is also responsible for shaping global Ca²⁺ responses into erratic isolated responses with slow dynamics. Eventually, upon a prolonged bacterial infection, IpgD leads to the inhibition of InsP3-mediated Ca²⁺ signals. Through its interference with Ca²⁺ signaling, IpgD delays a Ca²⁺-dependent calpain activation leading to the disassembly of focal adhesion structures and the premature detachment of infected cells¹³.

While Ca²⁺ signals are involved in critical aspects of pathogenesis, the use of a microorganism raises a number of technical challenges that are not encountered in classical agonist studies. The protocols described here use the commonly used fluorescent Ca²⁺ chemical indicator Fluo-4 that we engineered to characterize local Ca²⁺ signals during a Shigella infection. Steps critical for the detection of these signals are discussed, as well as procedures implemented for their quantitative analysis that is required to characterize the role of bacterial effectors in Ca²⁺ signaling.

Protocol

1. Preparations

1. Bacteria preparation
   1. Plate the bacteria—Shigella wild-type strain expressing the AfaE adhesin (M90T-AfaE)—onto a trypticase soy (TCS) agar plate containing 0.01% Congo red (CR) and incubate them for 18 h at 37 °C.
      Note: To increase their reproducibility, the Shigella plates obtained in step 1.1.1 are stored at 4 °C and should be used within a week, because all colonies on CR medium will eventually turn red over time.
   2. Inoculate TCS broth pre-culture by picking 3 red colonies from the streak plates.
      Note: An inoculation with 3 colonies is performed to limit the probability of picking a single clone whose virulence plasmid has undergone a genetic recombination.
   3. Grow liquid bacterial cultures in a shaking incubator for 16 h at 37 °C at 200 rpm. Add ampicillin at a final concentration of 75 mg/mL. Antibiotic concentrations should not exceed 3x the minimal inhibitory concentration, as an excess of antibiotic may lead to the loss of the large virulence plasmid.
      Note: The maintenance of the Shigella virulence plasmid should be verified by plating the bacterial cultures on CR plates supplemented with appropriate antibiotic concentrations. The presence of white colonies indicates plasmid loss.
   4. Inoculate the TCS culture with the bacterial pre-culture at a 1:100 dilution. Incubate it at 37 °C in a shaking incubator for 2 h at 200 rpm. Ensure that the optical density at 600 nm (OD_600nm) is 0.2 - 0.4.
   5. Centrifuge the bacterial culture for 2 min at 13,000 x g at 21 °C and resuspend the pellet in an equivalent volume of EM medium (120 mM NaCl, 7 mM KCl, 1.8 mM CaCl₂, 0.8 mM MgCl₂, 5 mM glucose, and 25 mM HEPES, pH = 7.3).
   6. Dilute the bacterial suspension in an EM buffer at a final OD_600nm of 0.1 and use it immediately or store it at 21 °C and use it within the next 60 min.
2. Cell preparation
   1. Culture HeLa cells in DMEM with 1 g/L of glucose supplemented with 10% fetal calf serum (FCS) and grow them at 37 °C with 10% CO₂. Grow TC-7 grown in DMEM with 4.5 g/L of glucose, containing 10% FCS and non-essential amino acids in a 37 °C incubator containing 10% CO₂.
   2. For cell maintenance, split the cells regularly to avoid a growth-contact inhibition linked to confluent states; this is critical for a high Ca²⁺ probe loading/transfection efficiency.
   3. Plate HeLa cells onto sterile 25-mm diameter circular coverslips in 6-well plates at a density of 3 x 10⁵ cells/well the day before the experiment for the HeLa cells.
   4. For TC-7 cells, trypsinize and count cells. Seed them at 4 x 10⁵ cells/well and incubate them 5 - 7 days at 37 °C in a 10% CO₂-incubator before the experiments to allow for a polarization.
      1. Refresh the cell culture medium every day until the day of the experiment.
         Note: Glass-bottom 35-mm diameter disposable chambers may also be used as an alternative to the coverslips-containing wells. If the latter are used, temperature control through a heated stage or objective is not sufficient, and a thermostat incubation chamber mounted on the microscope stage is required.
3. The day of the experiment, remove the medium and wash the cells 3x with 1 mL EM medium at 21 °C.
4. Load the cells with 3 mM Fluo-4-AM¹⁴ in an EM buffer for 30 min at 21 °C.
   Note: While the loading of sub-confluent HeLa cells does not raise major problems, the loading of polarized TC-7 cells is often heterogeneous and poorly efficient. For these cells, we found that the loading efficiency is enhanced by the addition of the dispersing agent—pluronic acid—at a final concentration of 0.1% and of the anion-transporter inhibitor bromosulfophtalein at a final concentration of 20 µM to the Fluo-4-AM-containing EM. The use of ratiometric chemical probes or genetically encoded-FRET probes requiring dual-acquisition is not compatible with the speed of acquisition required for the imaging of local Ca²⁺ responses.
5. Wash the samples 3x with EM buffer and further incubate them in 1 mL of EM buffer at 21 °C to allow for the hydrolysis of the AM moiety. Use Fluo-4 loaded cells immediately or within the next 90 min (maintained at 21 °C).

2. Infection and Image Acquisition of Local Ca²⁺ Responses

1. Place the coverslip containing the Fluo-4 loaded cells in an imaging chamber or glass-bottom chamber. Wash the samples in the imaging chamber or glass-bottom disposable chamber 3x with EM buffer to remove compounds potentially resulting from cell lysis. Add 1 mL of EM buffer.
2. Place the chamber on an inverted fluorescence microscope stage heated at 33 °C.
   Note: This temperature is selected as a compromise between the optimal temperatures compatible with the Shigella invasion process and the slowing down of Ca²⁺ responses to visualize local responses. We use a 63X immersion oil objective (NA = 1.25) equipped with phase contrast rings.
3. Select a microscopy field and set up acquisition parameters, including exposure time and binning if necessary, to optimize the Fluo-4 fluorescent signal.
   Note: The major challenges of local Ca²⁺ imaging are the low amplitude and the small duration of the responses, requiring acquisition at least every 30 ms. Several commercially available back-illuminated EM-CCD or C-MOS cameras present the required sensitivity to detect local Ca²⁺ responses using Fluo-4. Detection sensitivity is also an important issue to limit photodamage or artefacts such as spontaneous responses linked to the fluorescent excitation of Fluo-4 at a high frequency. Here, an LED-based illumination of 470 nm, with a 480 ± 40 nm band-pass excitation filter, a 505-nm dichroic filter, and a 527 ± 30 nm band-pass emission used at 5% of its maximal intensity combined with a 1.0 neutral density filter were used to minimize these problems under a stream of acquisitions of limited durations. An analysis of local Ca²⁺ signals by batches of 120 s streams was routinely performed. Under these low illumination conditions, fluorophore photobleaching was not observed during the 2 min-stream acquisitions. Successive acquisitions on the same sample can be performed, provided different fields are used. As a general rule, a first acquisition stream is performed on a control field in the absence of bacteria stimulation to ensure the absence of spontaneous responses. If spontaneous responses are observed, the samples are washed with EM buffer until no responses are observed.
4. To add bacteria to the sample, remove 500 µL of EM buffer from the chamber and add 500 µL of the bacterial suspension prepared in step 1.1.6 to obtain a final OD_600nm of 0.05, with the appropriate care to avoid moving the selected microscopy field; the volumes ensure a proper mixing of the bacteria and a homogeneous distribution of the bacteria onto the cell samples.
5. Perform an acquisition upon the bacterial addition. Alternatively, perform an acquisition 10 min following the bacterial addition, which corresponds to the time required for bacteria to sediment onto cells under the conditions used. At the end of the acquisition stream, acquire a phase-contrast image of the selected field to visualize the bacteria contacting the cells and the membrane ruffles associated with bacterial invasion sites.
6. Repeat the acquisition procedure as in step 2.5, at intervals that are amenable to the duration of the image acquisitions, files savings, and selection of a new field to cover the whole process.
   Note: For Shigella, we found that acquisition streams every 5 min for 20 min, following the 10 min bacterial sedimentation, allowed to cover the majority of the invasion events.
7. At the end of the acquisition procedure, add a final concentration of 2 µM of the Ca²⁺ ionophore ionomycin samples to determine the maximal amplitude of the Ca²⁺ signals. Acquire images every 3 - 5 s until signal stabilization, usually for less than 10 min.
8. Follow by adding Ca²⁺ chelator EGTA at a final concentration of 10 mM to determine the fluorescence signal in the absence of Ca²⁺. Acquire images every 3 - 5 s until signal stabilization, usually for less than 10 min.
   Note: Because of the relatively slow dynamics of the global Ca²⁺ variations, perform these acquisitions every 3 - 5 s (not in streaming mode), allowing for Ca²⁺ imaging on the same microscopic field during the successive treatments.

3. Analysis

1. Express cytosolic Ca²⁺ variations over time as the percentage of ΔF/F₀, where ΔF represents the variation of the average fluorescence intensity of the region of interest (ROI) and F₀ the baseline fluorescence in the same ROI, to which a non-relevant region of the field devoid of cells is subtracted.
   1. Remove the basal fluorescence levels for each cell in different fields, corresponding to baseline levels; these levels can be unambiguously determined due to the low frequency and relatively short duration of local Ca²⁺ responses in a given stream acquisition.
      Note: Quantify striking features of the responses related to the questions asked and the pathogenic model studied. Classically, various parameters are taken into account to analyze Ca²⁺ signals, including the frequency of cells showing local or global Ca²⁺ responses, as well as the duration, amplitude, and frequency of these responses. In the case of Shigella and local responses, another important aspect of the analysis is the spatial association of the responses with bacterial invasion sites.
2. To quantify the percentage of responsive cells showing global or local responses, draw ROIs in cells, in time-series corresponding to the variations of Fluo-4 intensity.
   1. Set strict parameters to identify local Ca²⁺ responses, such as an amplitude of three-fold above background variations of the average fluorescence intensity cell baseline, or a duration of at least 200 ms, to avoid scoring a false positive.
      Note: As a general rule, even small local Ca²⁺ responses can be distinguished from any background noise by their profile. The ROIs should be large enough to integrate enough fluorescence signals to allow for the detection of local variations. Depending on the intensity of the Fluo-4 detection, these ROIs may correspond to circles with diameters ranging from 2 - 5 mM.
   2. Measure the variations of the average Fluo-4 fluorescence intensity in 2 regions in a distinct area of the same cell, to determine whether the responses are local or global.
      Note: The analysis of a large number of cells is facilitated by the use of imaging software permitting the on-line screen display of the variations of the average fluorescence intensity over time for the selected regions. Commercially available imaging software has such an option, but this can also be performed using the Icy freeware, using the ROI Intensity Evolution plugin.
   3. Determine the percentage of cells showing local responses.
      Note: This percentage is calculated from the total number of cells analyzed in the microscopy field, which should be in sufficient numbers to support a statistically significance using a non-parametric test.
   4. Determine the duration of the local responses.
      Note: Local Ca²⁺ responses can be further distinguished according to their duration ranges (i.e., less than 500 ms, between 5 and 10 s, or above 10 s) and their association with Shigella invasion sites.
3. Quantify the amplitude of the responses. First, calibrate the maximal Ca²⁺ and cell background levels determined as in step 2.1.7. Since the Fluo-4 load may vary for each cell, perform these determinations for individual cells in the field.
   1. Express the amplitude of the responses as a relative percentage of the maximal response.
   2. Perform statistical testing using a normality test, followed by a parametric test to analyze potential differences in the amplitude of responses between samples.
   3. Determine the association of these responses relative to the bacterial invasion sites by their respective localization to the bacterial-induced membrane ruffles visualized in the corresponding phase contrast images.

Results

Shigella invasion is associated with atypical long-lasting local Ca²⁺ responses:

Following the protocol mentioned above, Fluo-4-loaded HeLa cells were challenged with WT Shigella and stream acquisitions were performed to analyze Ca²⁺ signals. A representative experiment is shown in Figure 1, with time-lapse images series of the fluorescence intensity of the Fluo-...

Discussion

This manuscript describes the protocol that we engineered to follow local Ca²⁺ signals during the relatively short kinetics of a Shigella invasion, as well as global Ca²⁺ responses during the extended kinetics of Shigella. Below, key issues can be found that need to be addressed to optimize the detection of Ca²⁺ signals while minimizing any interference with the biological processes.

Chemical vs. genetically encoded Ca²⁺ probe...

Materials

List of materials used in this article

Name	Company	Catalog Number	Comments
Fluo-4 AM	Invitrogen	F14201
Metamorph version 7.7	Universal Imaging
CoolLED illumination system pE-2	Roper Scientific
micro-dish 35 mm, high	IBIDI	81156
Trypticase Soy (TCS) broth	Thermofisher		B11768
TCS agar	Thermofisher		B11043
Congo red	Sigma-Aldrich		75768
M90T-AfaE	Sun et al. 2017	Shigella flexneri serotype V. expressing the AfaE adhesin
ipgD-AfaE	Sun et al. 2017	isogenic ipgD mutant strain expressing the AfaE adhesin