Analysis of phagocytosis induced TNF-α production of murine macrophage using traditional flow cytometry in a mixed population

Introduction
Phagocytosis is one of the most ancient immune defenses and is vastly conserved across evolution. It is critical not only in pathogen defence, but also in the clearance of dead or dying cells. Traditionally, phagocytosis was studied using light or fluorescence microscopy to quantify particle internalization, but this technique is meticulous, slow, and usually generates statistically weak data. By utilizing flow cytometry, fluorescent target particles associated (both surface bound and internalized) with tagged effector cells can be analyzed at high speed to generate large statistically robust datasets. In addition, other downstream effects of phagocytosis can be analyzed in a multi-parametric manner using other fluorescent markers and channels. Here in we demonstrate the use of traditional flow cytometry to analyze the production of TNF-α by murine macrophages in relation to bacterial internalization.

Materials and Methods

Intracellular staining of RAW macrophages following phagocytosis of Eschericheri coli.
Cultured murine macrophage were incubated with GFP-E. coli according to MMI Lab 3 protocol. Briefly, 1x10⁶ RAW murine macrophage and Jurkat T-cells were seeded into complete DMEM with GFP expressing DH5α E. coli at a 21.1:1 target to effector ratio. Tubes were incubated with Golgiplug (BD) at 37°C and 5% CO₂. Samples were then spun down at 200 x g and washed twice with PBS before blocking with FACS buffer (2% serum) and incubated with rat anti-mouse αCD45-PE IgG (Biolegend) for 30 minutes, washed twice and resuspended in Cytofix buffer and held at room temperature for 20 minutes. Following PBS wash, cells were resuspended in permeablization buffer (0.01% Triton X-100) and held at room temp for 10 minutes before incubation with rat anti mouse αTNFα IgG-APC (Biolegend) for 30 minutes at 4°C. Samples were washed with permeablization buffer and spun at 326 x g twice before being stored overnight at 4°C in PBS then being run on a BD FACS Canto II. Small debris and potential clumps were gated out using gating techniques described in MMI 590 Lab 1.

Results and Discussion

Antibody staining allows for high resolution of gating on desired cell populations.
Although forward and side scatter can sometimes be used to separate out morphologically distinct cells, side scatter profiles often overlap and prevent analysis of a population of interest. In this experiment, RAW macrophage and Jurkat T-cells are seen to overlap quite heavily in SSC-FSC plots (Figure 1a). The use of fluorescently tagged αCD-45 antibodies allows a much higher level of resolution when separating these two populations. Although CD-45 is a pan- leukocyte marker, this particular antibody is against murine CD-45 and does not cross-react with human CD-45. Thus the antibody stains RAW cells, which are of murine origin, but not human Jurkat cells and allows for downstream analysis of each population respectively (Figure 1b).

Figure 1. Analysis of distinct cell types in a mixed population using surface markers. Co-incubated RAW murine macrophage and Jurkat human T-cells were acquired on a BD FACS Canto II. a) Forward and side scatter profile of the mixed population P1 gate (RAW cells orange, Jurkat cells Green based on P4/P5 gates) b) Total single cell events were gated based on fluorescence of anti-mouse CD45 antibody conjugated to a PE fluorochrome.

Raw macrophage exposed to E. coli produce TNFα regardless of bacterial internalization.
Macrophage are professional phagocytes, and tend to be some of the first responders in infection which recruit other cells to the site and activate them via chemokine/cytokine production. Thus, as expected, our RAW macrophages responded quickly when exposed to E. coli, being roughly 15% and 22% phagocytic by 30 and 60 minutes respectively (Figure 2, GFP+). This rate seems low in relation to total TNFα producing, APC+ RAW cells (85% and 95% at 30 and 60 minutes respectively), but may be due to paracrine action of cytokine production by phagocytic cells. Though this scenario seems unlikely since the addition of golgiplug should prevent the secretion of any produced cytokines. Most likely, these non phagocytic TNF producing cells are being activated through bacterial PAMPs present in the culture media binding through TLRs. A non-stimulated control could also be used to see if TNF production is constitutive in this cell line.

 Since Jurkat cells are lymphocytes, which are not known to be phagocytic except for in specific species, we expect them to not internalize any bacteria.  As expected, the vast majority of the Jurkat cells were negative for bacterial internalization. The very small proportions of GFP+ Jurkat cells (1.7% and 2.5% at 30 and 60 minutes respectively) could be due to a bacterial cell stuck to the outside of a Jurkat cell, which would likely not change the area enough to be removed as a clump using a height to surface area gate. This is where traditional flow cytometry falters, as it cannot distinguish associated fluorescence from true internalization. The use of imaging flow cytometry could eliminate these events and greatly increase confidence in the data.

 CD4+ T helper cells have been known to produce TNFα in macrophage co-cultures. Since Jurkat cell lines are CD4+ derived, we expected some level of TNFα production, but virtually no TNF staining was seen in these cell types. This could be explained by murine macrophages not being able to stimulate human T cells. Alternatively, there may be TNF production, but since we used an anti-mouse TNF antibody, there may simply be no cross-reactivity and thus the cells appear to be TNF(-). A repeated experiment using anti-human antibodies may be useful in deciphering these results.

Figure 2. Macrophage exposed to bacteria produce TNFα regardless of bacterial internalization. Mixed RAW macrophage and Jurkat T cells were incubated for 30 or 60 minutes with 22.1:1 GFP E. coli. Populations gated based on anti-mouse CD-45 staining. Samples were also stained with anti-mouse anti-TNFα antibody conjugated to APC. Population proportions are noted in the corner of each respective gate.

Flow cytometric analysis of cultured T-cell viability via Annexin V and propidium iodide staining

Introduction
Cell viability is a critical factor during any experiment in flow cytometry. The proportions of dead or dying cells can vary greatly due to growth state of in vitro culture, time spent ex vivo, exposure to reagents and protocol procedures in addition to myriad other factors. Not only can dying cells lack the functional responses that may be being assessed in the experiment, but they also generally have increased non-specific binding due to sugar and nucleic acid based ‘stickiness’ which may lead to false positives. In addition, necrotic cells can also induce cell activation in samples and may lead to skewed results. Fortunately, dead or dying cell proportions can be analyzed using viability stains to assess impact on results, or to gate them out and analyze only live cells. Herein we demonstrate the use of an annexin V and propidium iodide assay to distinguish live from necrotic and apoptotic Jurkat T cells.

Materials and Methods

Lymphocyte staining with annexin V and propidium iodide.
Jurkat T lymphocytes were cultured and processed following the MMI 590 Lab 2 protocol to be run on a BD FACS Canto II flow cytometer. Briefly, cultured cells were spun down at 200 x g for 5 minutes and resuspended in fresh complete RPMI media. 1x10⁶ cells were seeded into FACS tubes with varying concentrations of ethanol and incubated at 37°C, 5% CO₂ for 40 minutes. One 0% ethanol tube was held on ice for 2 hours prior to incubation. Cells were then washed twice and resuspended in 1x annexin V binding buffer. Annexin V-FITC and propidium iodide were then added to each tube and incubated for 20 minutes in the dark at room temperature. Extra tubes incubated with 25% ethanol were stained with Annexin V or propidium iodide only for compensation controls. Fluorescence signals were compensated then experimental samples were analyzed on a BD FACS Canto II, gating out debris and doublets using the gating strategy described in MMI 490 Lab 1. Extra tubes incubated with 25% ethanol were stained with Annexin V or propidium iodide only for compensation controls.

Results and Discussion

Annexin V and propidium iodide can be used to distinguish specific states of cell death.
Cell viability is important to any experiment, as it can greatly impact results. Fortunately, this can easily analyzed using flow cytometry with an Annexin V- propidium iodide stain. Phosphatidylerine (PS) is a phospholipid anchored into the cytoplasmic side of the plasma membrane of healthy cells. When apoptosis is triggered, flippases translocate PS to the outer membrane surface where it acts as an apoptotic signal for macrophages to clear the cell. Annexin V binds PS with anti-body like affinity and is used in this experiment to distinguish cells undergoing apoptosis by high FITC and low propidium iodide (PI) staining (Figure 1, Q4). Propidium iodide is non-membrane permeable and will stain any cells with compromised membrane integrity, namely necrotic cells (Figure 1, Q1). Because Annexin V is a large protein, it generally cannot pass to the cytoplasm without a large amount of membrane damage. Thus double positive (Figure 1, Q2) cells are either late apoptotic and have extracellular facing PS with membrane degradation to allow PI staining, or they are late necrotic and have so much membrane damage that Annexin V can enter the cell to bind cytoplasmic facing PS. Herein, this assay is able to discern viable cells (Figure 1, Q3) from apoptotic, necrotic, and late necrotic/apoptotic populations. These two latter populations could be distinguished by use of a TUNEL kit to look for apoptosis triggered DNA fragmentation. The downside of Annexin-PI staining being that annexin requires a specific calcium concentration to bind PS, and may not be compatible with buffers needed for other fluorescent stains. This could be remedied by the use of an array of commercially available live/dead stains.

Figure 1. Annexin V and propidium iodide stains can distinguish live and dead cell populations. FACS plot of Jurkat T-cells incubated in 5% ethanol at 37°C in 5% CO2­­ ­for 40 minutes then stained with Annexin V-FITC and propidium iodide. Samples were acquired on a BD FACS Canto II. Percent cell population indicated in the corner of each gate.

Increasing concentrations of ethanol induces different states of cell death in cultured cells.
Experimental protocols can have varying effects on live cells, and it is important to know whether or not your cells are still viable. If a large proportion of cells are non- viable, it may be a clue that a protocol may have to be further optimized for those specific cells, or that a different assay should be used all together. In this experiment, we exposed Jurkat cells to increasing concentrations of ethanol to interesting effect. 5% ethanol exposure for 40 minutes showed little effect, with comparable cell viability to 0% controls (Figure 2). Incubation in 25% ethanol shifted almost the entire the population (93.2%) to double Annexin-PI positive, indicating a late necrotic/ apoptotic state. Interestingly 50% ethanol shifted most of this double positive population to PI positive or necrotic cells making up 22.8% and 74.4% of the cell population respectively. Perhaps, some property of 25% but not 50% ethanol allows for PS to reach the outer leaflet of the membrane, or destabilizes the membrane such that Annexin can enter. This experiment highlights the use of viability stains to assess the effects of protocols on cells, as 25% and 50% ethanol samples are useless for assaying live cells. In addition, although 5% ethanol protocol induced some cell death, viability stains allow the capacity to gate on only viable cells to garnish the most robust data possible.

Figure 2. Viability of Jurkat T-cells incubated with varying concentrations of ethanol. Cultured Jurkat T-cells were seeded into complete RPMI containing varying concentrations of ethanol prior to incubation at 37°C in 5% CO2­­ ­for 40 minutes. Ice shocked cells were held on ice for 2h before incubation with 0% ethanol. Cells were then washed and stained with Annexin V and propidium iodide prior to analysis on a BD FACS Canto II (n=1). Colors correspond populations scene in figure 1.

Light scatter properties of cultured murine macrophage and human T cells

Light scatter properties of cultured murine macrophage and human T cells.

Introduction
Flow cytometry has been a cornerstone in biological advancement, and in immunology in particular over the last 30 years. This technique achieves high levels of statistical robustness via high throughput analysis of individual cells using lasers, light scatter and fluorescent markers. Light scatter is a crucial component of flow cytometry that uses forward scatter to discern relative size and side scatter to measure relative internal complexity of individual cells. These properties can be used to gate out doublets, clumps and debris, identify distinct cell populations, or to look at any relative morphological changes within a population such as activation. Here in we will analyse the light scatter properties of Jurkat T cells and RAW macrophage cells using flow cytometry to understand how we can employ side and forward scatter to produce the best possible data in our future experiments.

Materials and Methods

Forward and side scatter analysis of in vitro cultured cell lines
RAW murine macrophage and Jurkat T lymphocytes were processed following the MMI 590 Lab 1 protocol to be run on a BD FACS Canto II flow cytometer. Briefly, cultured cells were spun down and resuspended in FACS Buffer. Cells were counted and due to low cell concentrations, 1.6x10⁵ RAW cells and 5.7x10⁵ Jurkat cells (in 1mL) were seeded into their respective FACS tube. The combined tube held  7.9x10⁴ and 2.9x10⁵. Combined results used in the data portion were obtained from Group 2 due to long sample acquisition times from low cell concentrations. FACS Canto II lasers were set at 330mV and 450mV for FSC and SSC respectively.

Results and Discussion

Light scatter can be employed to remove debris and doubletsCellular light scatter properties are an important tool when setting up any good flow cytometry experiment, as it can help to remove events that may skew results and thus the conclusions garnished from the experiment. The gating strategy used in this experiment (Figure 1a) displays this importance with 11.8% of total events (black dots) gated out in the P1 gate which is mostly debris of low size (forward scatter). The population remaining contains both the RAW and Jurkat cell populations within them, but there are still likely clumps and doublets that may skew data, especially in experiments looking for co-expression or other double-positive events since two single positive cells may appear as a double positive event if not gated out properly. To achieve this, cells with a low ratio of forward scatter area vs height are gated out (P2, red dots). When two cells are counted as a single event, their area to height ratio will be roughly twice that of a single cell event pulling the respective dot out of the bottom of the P2 gate and disregarding it from the dataset. In the mixed population, this removes an additional 12.4% of total events captured for a total of 24.2% of events removed with 52.8% and 20.7% of events removed for RAW and Jurkat only samples respectively (Figure 1b). Removing these events makes the dataset and thus and conclusions made more robust.

Figure 1. Light scatter profiles of RAW and Jurkat cell lines. a) Gating strategies on a mixed RAW and Jurkat cell population included excluding low side scatter and forward scatter events (P1 red and green). High FSC area to FSC height ratio events (red) were then excluded from this population to yield a population for final analysis (green). b) Forward and side scatter plots for post-gated RAW macrophage, Jurkat T- lymphocytes and a mixed population containing both cultures. Proportion of total captured events remaining for analysis is indicated at the top right of each plot.

Light scatter can be used to distinguish distinct cell populations
Light scatter can also be used to distinguish cell populations that are morphologically different by relative size measured by forward scatter, or relative internal complexity measured by side scatter.
We saw that the RAW macrophage sat, on average, at a lower forward scatter than the Jurkat cells indicating that they are on average smaller, although the Jurkat only population had proportion of cells that sat at the smaller range of RAW cells (figure 1b). These could be apoptotic cells and could possibly be gated out by a viability assay. Unfortunately, the use of some kind of fluorescent marker would be needed to separate these two cell populations in flow analysis because of their overlapping forward and side scatter profiles.