Here, we describe a simple confocal imaging method to visualize the in situ localization of cells secreting the cytokine Interferon gamma in murine secondary lymphoid organs. This protocol can be extended for the visualization of other cytokines in diverse tissues.
Cytokines are small proteins secreted by cells, mediating cell-cell communications that are crucial for effective immune responses. One characteristic of cytokines is their pleiotropism, as they are produced by and can affect a multitude of cell types. As such, it is important to understand not only which cells are producing cytokines, but also in which environment they do so, in order to define more specific therapeutics. Here, we describe a method to visualize cytokine production in situ following bacterial infection. This technique relies on imaging cytokine-producing cells in their native environment by confocal microscopy. To do so, tissue sections are stained for markers of multiple cell types together with a cytokine stain. Key to this method, cytokine secretion is blocked directly in vivo before harvesting the tissue of interest, allowing for detection of the cytokine that accumulated inside the producing cells. The advantages of this method are multiple. First, the microenvironment in which cytokines are produced is preserved, which could ultimately inform on the signals required for cytokine production and the cells affected by those cytokines. In addition, this method gives an indication of the location of the cytokine production in vivo, as it does not rely on artificial in vitro re-stimulation of the producing cells. However, it is not possible to simultaneously analyze cytokine downstream signaling in cells that receive the cytokine. Similarly, the cytokine signals observed correspond only to the time-window during which cytokine secretion was blocked. While we describe the visualization of the cytokine Interferon (IFN) gamma in the spleen following mouse infection by the intracellular bacteria Listeria monocytogenes, this method could potentially be adapted to the visualization of any cytokine in most organs.
Orchestrating an efficient immune response against a pathogen requires complex integration of signals displayed by a variety of immune cells that are often dispersed among the organism. In order to communicate, these cells produce small soluble proteins with multiple biological functions that act as immunomodulators named cytokines. Cytokines control cell recruitment, activation and proliferation and hence are known to be key players in the promotion of immune responses1. Effective immune responses require cytokines to be released in a very organized spatiotemporal pattern connecting specific cells to induce specific signals. Therefore, it is crucial to study cytokine production and its signaling in situ, taking into account the microenvironment in which cytokines are produced.
Listeria monocytogenes (L. monocytogenes) is a Gram-positive intracellular bacterium used as a prime model to study immune responses to intracellular pathogens in mice. One cytokine, IFN gamma (IFNγ) is produced rapidly, within 24 h following L. monocytogenes infection. It is necessary for pathogen clearance, as mice knocked out for IFNγ are highly susceptible to L. monocytogenes infection2. IFNγ is pleiotropic and produced by multiple cells following infection3. While IFNγ produced by natural killer (NK) cells is required for direct anti-bacterial activity4, IFNγ from other sources have been shown to have other functions. Indeed, we and others recently found that IFNγ produced by CD8+ T cells has a specific function in directly regulating T cell differentiation5,6,7. As such, understanding which cells produce IFNγ (and in which microenvironment) is crucial to dissect its function.
The most common technique to study cytokine production relies on intracellular cytokine staining analyzed by flow cytometry. This method allows the simultaneous detection of multiple cytokines combined with cell surface markers within a single sample, providing an extremely useful tool to study cytokine production. However, using the aforementioned technique implies losing any spatial information. In addition, cytokine detection often relies on in vitro re-stimulation to enable cytokine detection. As such, the capacity of a given cell to produce a cytokine is analyzed, and it does not necessarily correlate with actual cytokine secretion in situ. Other methods use reporter mice for which fluorescent protein expression correlates with cytokine transcription and allows for visualization on a single-cell level8. Although this method can track cytokine transcription in situ, there are a limited number of cytokine-reporter mice available. In addition, transcription, translation and secretion can sometimes be unlinked, and fluorescent proteins have a different half-life than the cytokine they report, making this method sometimes not adequate for in situ cytokine visualization.
Here, we describe a method to visualize in situ cytokine production by confocal microscopy at single cell resolution. This technique enables the visualization of the cellular source and surrounding niche within the tissue. This protocol specifically describes the visualization of IFNγ production in the spleen of L. monocytogenes infected mice, focusing here on IFNγ production by NK cells and antigen specific CD8+ T cells. However, it can be extended and adapted to the characterization of any cytokine production in the context of other situations where cytokines are produced such as infection, inflammation or autoimmune diseases, as long the targeted cytokine can be retained in cells by intracellular protein transport inhibitor.
All experiments involving mice were in agreement with the UK Scientific Procedures Act of 1986.
1. Adoptive Transfer of Antigenic-specific CD8+ T Cells in Mice
2. Listeria monocytogenes Infection
3. Treatment with Brefeldin A (BFA) to Block Cytokine Secretion
4. Harvesting the Spleen
5. Fixation of the Spleen with Paraformaldehyde (PFA)
6. Freezing and Sectioning
7. Immunofluorescent Staining
8. Imaging and Analysis
IFNγ produced within the first 24 h after Listeria monocytogenes infection is critical to control the spread of this pathogen. Using this protocol, we can visualize not only which cells are producing IFNγ but also whether they are located in a specific microenvironment. To help us delineate the architecture of the spleen, we labeled cells known to have particular location within the spleen. The marker F4/80 labels all macrophages and highlights the red pulp. The marker B220 labels B cells and highlights B cell follicles surrounding the T cell zone. The marker CD169 labels marginal zone macrophages, surrounding the white pulp (Figure 1). Most OTI cells, whether they express IFNγ or not, are present in the white pulp and as such, all images are those of the white pulp, unless indicated.
One critical step in this protocol is the use of BFA to inhibit cytokine secretion. Indeed, the detection of IFNγ by NK cells was greatly impaired when mice were not treated with BFA (Figure 2). Using our protocol, we could find that at least two cell types produce IFNγ 24 h after infection—NK cells and antigen-specific CD8+ T cells (Figure 3)—similarly to what has been found previously by flow cytometry3.
In situ imaging of IFNγ producing cells revealed that IFNγ production is not spread throughout the spleen, but concentrated into discreet areas (Figure 4). Indeed, we found that T cells were activated throughout the spleen (highlighted by T cell clustering), and this did not necessarily correlate with IFNγ production. One likely explanation is that IFNγ production is restricted to the location of the infected cells15,16, and T cell activation—represented by clustering—can be supported by both infected (IFNγ positive) and non-infected (IFNγ negative) antigen-presenting cells. Other stains will be required to pinpoint the exact location and get an indication of the mechanism restricting IFNγ production to this area and its relationship to antigen transfer. Interestingly, we found that activated, clustered, antigen-specific T cells are located throughout the white pulp of the spleen but they produce IFNγ only in regions where NK cells are coexisting with them (Figure 5). As such, the presence of NK cells delineates a specific microenvironment in the white pulp, in which clustered T cells produce IFNγ as opposed to clustered T cells in the other part of the white pulp. This suggests that T cell activation is not sufficient to dictate IFNγ production at this time point.
Another interesting feature highlighted by our protocol is the different sub-cellular localization of IFNγ in NK versus CD8+ T cells5. As shown in Figure 6, while IFNγ localization in NK cells is diffused in the cytosol, CD8+ T cells often recruit IFNγ towards another T cell.
Figure 1: Markers highlighting the spleen architecture. Mice were infected with 2 x 104 CFU LM-OVA and euthanized 24 h post infection. Spleen was explanted and processed as described in the protocol. (A) Sections were stained for NK cells (anti-NCR1 followed by anti-goat IgG-FITC; green), OTI-RFP cells (red) and macrophages (anti-F4/80-APC; magenta). RP = Red Pulp; WP = White Pulp. Scale bar = 200 µm. (B) Sections were stained for B cells (anti-B220-Pacific Blue; Blue), OTI-GFP cells (GFP signal shown in red) and marginal zone macrophages (anti-CD169-Alexa647; magenta). RP = Red Pulp; BF = B cell follicle; TZ = T cell zone. Scale bar = 50 µm. This is a representative image of 3 independent experiments (N = 4). Please click here to view a larger version of this figure.
Figure 2: BFA treatment allows for the detection of intracellular IFNγ in situ. Nγ production is restricted to specific areas in the sple Mice were infected with 2 x 104 CFU LM-OVA and treated with BFA (A) or left untreated (B) after 18 h. Mice were euthanized 24 h post infection. Spleen was explanted and processed as described in the protocol. Sections were stained for NK cells (anti-NCR1 followed by anti-goat IgG-FITC; green), OTI-RFP cells (red) and IFNγ (anti-IFNγ-BV421; cyan). Scale bar = 5 µm. This is a representative image of NK cell-rich areas from 3 independent experiments (N = 3). Please click here to view a larger version of this figure.
Figure 3: IFNγ producing cells in the spleen. Mice were infected with 2 x 104 CFU LM-OVA when indicated and treated with BFA after 18 h. Mice were euthanized 24 h post infection. Spleen was explanted and processed as described in the protocol. Sections were stained for NK cells (anti-NCR1 followed by anti-goat IgG-FITC; green), OTI-RFP cells (red) and IFNγ (anti-IFNγ-BV421; cyan). (A) Representative image of a spleen from an un-infected naïve mouse to demonstrate absence of IFNγ non-specific staining. White lines delineate the white pulp. WP = White pulp; RP = Red pulp. (B) Representative image of the white pulp from the spleen of a mouse infected by LM-OVA, showing invasion of NK cells to the white pulp and production of IFNγ by NK cells, OTI cells and non-labelled cells. Images are representative from 4 independent experiments (N = 4). Scale bars = 70 µm (A); and 20 µm (B). Please click here to view a larger version of this figure.
Figure 4: IFNγ production is restricted to specific areas in the spleen following LM-OVA infection. Mice were infected with 2 x 104 CFU LM-OVA and treated with BFA after 18 h. Mice were euthanized 24 h post infection. Spleen was explanted and processed as described in the protocol. All sections were stained for B cells (B220-Pacific Blue Ab, Blue) and IFNγ (anti-IFNγ-biotin followed by streptavidin-PE; cyan). OTI-GFP cells (GFP signal shown in red). Cyan lines correspond to areas of high IFNγ production. Those are representative images of 4 independent experiments (N = 4). (A) Sections were stained for marginal zone macrophages (anti-CD169-Alexa 647, magenta). Scale bar = 50 µm. (B) Sections were stained for all macrophages (F4/80). Scale bar = 100 µm. Please click here to view a larger version of this figure.
Figure 5: IFNγ production by activated OTI cells occurs in a specific microenvironment. Mice were infected with 2 x 104 CFU LM-OVA and treated with BFA after 18 h. Mice were euthanized 24 h post infection. Spleens were explanted and processed as described in the protocol. Sections were stained for NK cells (anti-NCR1 followed by anti-goat IgG-FITC; green), OTI-RFP cells (red) and IFNγ (anti-IFNγ-BV421; cyan). Green and red lines highlight NK and OTI cell zones, respectively. White arrow indicate examples of T cell clusters not producing IFNγ. Green arrows examples of T cell clusters producing IFNγ. Scale bar = 100 µm. This is a representative image of four independent experiments (N = 4). Please click here to view a larger version of this figure.
Figure 6: Sub-cellular localization of IFNγ in NK cells and T cells. Mice were infected with 2 x 104 CFU LM-OVA and treated with BFA after 18 h. Mice were euthanized 24 h post infection. Spleen was explanted and processed as described in the protocol. All sections were stained for IFNγ (anti-IFNγ-BV421; cyan). White lines delineate cell edges and white arrows shows directionality of secretion. This is a representative image of two independent experiments (N = 5). (A)- OTI-RFP cells are shown in red. Scale bar = 5 µm. (B) Sections were stained for NK cells (anti-NCR1 followed by anti-goat IgG-FITC; green. Scale bar = 2 µm. Please click here to view a larger version of this figure.
In this manuscript, we present a method to visualize IFNγ production in the spleen following L. monocytogenes infection in mice. This protocol is simple and can be adapted to other tissues and cytokine triggers, but the following aspects have to be considered. Cells often rapidly secrete the cytokines they produce, and cytokines are rapidly picked up by neighboring cells. It is as such difficult to detect cytokines in situ. A common method to rapidly re-initiate cytokine production is to re-stimulate the cells ex vivo followed by cytokine detection in the media by enzyme-linked immunosorbent assay. In this context, any information about the spatial localization of the cytokine-producing cells is lost. In addition, cytokine production following re-stimulation does not necessarily reflect whether cytokines are actually produced and secreted in vivo, but rather indicates the capacity of a given cell population to produce cytokines. Therefore, both methods will provide different information and one should consider which information is most valuable for their experiment.
In order to detect intracellular cytokines, our method uses an intracellular protein transport inhibitor to trap cytokines inside cells and increase signal detection. However, it is important to note that these inhibitors affect the normal transportation of proteins from the endothelial reticulum (RE) to the Golgi apparatus and to the secretory vesicle impairing their release, which could cause toxicity. As a consequence, BFA, or other inhibitor, should be used for a short period of time, typically no more than a few hours. Hence, it is important to find the right balance between the inhibitor dose and time of treatment in order to optimize the level of cytokines trapped inside the cell without causing serious cytotoxic effects. These variables can differ between cytokines and the route of administration for the BFA. In our infection model, the BFA was administrated intraperitoneally in order to provide a rapid systemic dispersion, but it can also be delivered intravenously.
The most commonly used intracellular protein transport inhibitors are BFA, used here, and monensin (MN). These inhibitors are often used indistinctly to accumulate and study cytokine production but they have slight differences in their mechanisms of action. MN inhibits transportation of proteins within the Golgi apparatus hence accumulating proteins in the Golgi17 while BFA prevents coatomer protein complex-I recruitment, inhibiting the retrograde movement of proteins to the endoplasmic reticulum (ER) and thereby promoting accumulation of cytokines in the ER18. As such, choosing the best intracellular protein transport inhibitor will depend on different factors, such as the cytokine to be detected. For example, it has been shown in lipopolysaccharide-induced intracellular staining of monocytes that BFA is more efficient to measure the cytokines IL-1β, IL-6 and TNF than MN19.
This protocol involves the visualization of the cytokine in situ by confocal microscopy and therefore there are only a limited number of markers than can be used to study the cytokine- producing cells and their microenvironment. It is also necessary to consider that protein transport inhibitors such as BFA or MN disturb the normal expression of several proteins and hence their use when studying the simultaneous expression of certain activation cell surface markers has to be approached carefully. For example, BFA but not MN blocks the expression of CD69 in murine lymphocytes20. Despite this limitation, confocal imaging enables sub-cellular localization of cytokines, as well as the direction of the cytokine secretion within the cell. The data generated using this protocol suggest that NK cells tend to secrete IFN-y in a diffuse pattern while CD8+ T cells seem to direct the IFNγ secretion towards other CD8+ T cells that are in direct interaction with them5.
To conclude, this protocol is suitable to visualize a variety of cytokines in situ and identify producing cells and their microenvironment following many triggers such as infection or autoimmunity. The information obtained is instrumental to understand the importance of the in vivo spatial orchestration of different cell types and the cytokine they produce, necessary for an efficient immune response.
The authors have nothing to disclose.
We thank the Kennedy Institute Imaging Facility personnel for technical assistance with imaging. This work was supported by grants from the Kennedy Trust (to A.G.), and Biotechnology and Biological Sciences Research Council (BB/R015651/1 to A.G.).
Brefeldin A | Cambridge bioscience | CAY11861 | |||
Paraformaldehyde | Agar scientific | R1018 | |||
L-Lysin dihydrochloride | Sigma lifescience | L5751 | |||
Sodium meta-periodate | Thermo Scientific | 20504 | |||
D(+)-saccharose | VWR Chemicals | 27480.294 | |||
Precision wipes paper Kimtech science | Kimberly-Clark Professional | 75512 | |||
O.C.T. compound, mounting medium for cryotomy | VWR Chemicals | 361603E | |||
Fc block, purified anti-mouse CD16/32, clone 93 | Biolegend | 101302 | Antibody clone and Concentration used: 2.5 mg/ml | ||
Microscope slides – Superfrost Plus | VWR Chemicals | 631-0108 | |||
anti-CD169 – AF647 | Biolegend | 142407 | Antibody clone and Concentration used: clone 3D6.112 1.6 mg/ml Excitation wavelength: 650 Emission wavelength: 65 |
||
anti-F4/80 – APC | Biolegend | 123115 | Antibody clone and Concentration used: clone BM8 2.5 mg/ml Excitation wavelength: 650 Emission wavelength: 660 |
||
anti-B220 – PB | Biolegend | 103230 | Antibody clone and Concentration used: clone RA3-6B2 1.6 mg/ml Excitation wavelength: 410 Emission wavelength: 455 |
||
anti-IFNg – biotin | Biolegend | 505804 | Antibody clone and Concentration used: clone XMG1.2 5 mg/ml | ||
anti-IFNg – BV421 | Biolegend | 505829 | Antibody clone and Concentration used: clone XMG1.2 5 mg/ml Excitation wavelength: 405 Emission wavelength: 436 |
||
anti-Nkp46/NCRI | R&D Systems | AF2225 | Antibody clone and Concentration used: goat 2.5 mg/ml | ||
anti-goat IgG-FITC | Novusbio | NPp 1-74814 | Antibody clone and Concentration used: 1 mg/ml Excitation wavelength: 490 Emission wavelength: 525 |
||
Streptavidin – PE | Biolegend | 405203 | Antibody clone and Concentration used: 2.5 mg/ml Excitation wavelength: 565 Emission wavelength: 578 |
||
streptavidin – FITC | Biolegend | 405201 | Antibody clone and Concentration used: 2.5 mg/ml Excitation wavelength: 490 Emission wavelength: 525 |
||
Fluoromount G | SouthernBiotech | 0100-01 | |||
Cover glasses 22x40mm | Menzel-Glazer | 12352128 | |||
Liquid blocker super PAP PEN mini | Axxora | CAC-DAI-PAP-S-M | |||
Imaris – Microscopy Image Analysis Software | Bitplane | ||||
Confocal microscope – Olympus FV1200 Laser scanning microscope | Olympus | ||||
Cryostat – CM 1900 UV | Leica | ||||
Base mould disposable | Fisher Scientific UK Ltd | 11670990 | |||
PBS 1X | Life Technologies Ltd | 20012068 | |||
BHI Broth | VWR Brand | 303415ZA | |||
GFP | Excitation wavelength: 484 Emission wavelength: 507 |
||||
RFP | Excitation wavelength: 558 Emission wavelength: 583 |
||||
Insulin syringe, with needle, 29G | VWR International | BDAM324824 | |||
C57BL/6 wild type mice | Charles River |