Activation of CD4 and CD8 T cell receptors and regulatory T cells in response to human proteins

Introduction

The stability and plasticity of various T lymphocyte subgroups parallel their various physiological functions and suggest selection mechanisms within the context of T cell maturation (Shah & Zúñiga-Pflücker, 2014). The various T helper subsets (i.e., Th1, Th2, and Th17) and regulatory T cells (T-reg cells) are interconvertible. Over time, T helper cells can differentiate into other cell types of this group via the activation of specific transcription factors. Indeed, the T cell receptor (TCR) spectrum of various T helper subsets is plastic, and such plasticity in terms of TCR spectra and of the antigens recognized results in various subgroups of T helper cells. Notably, the expression of CD4 and/or CD8 co-receptors—and thus differentiation into one of these two cell types—is fixed once it has occurred (Shah & Zúñiga-Pflücker, 2014).

The respective TCR spectra of T cells are fixed into a CD4+ TCR spectrum and/or a CD8+ TCR spectrum through stable differentiation into CD4+ and/or CD8+ T cells. This differentiation occurs at an extremely early stage during development through positive and negative selection of the respective T cells during thymic T cell development (Shah & Zúñiga-Pflücker, 2014). Although the plasticity of T helper cell subgroups allows increased variability in T cell populations later in life, fixation of CD4 and CD8 TCR spectra (Shah & Zúñiga-Pflücker, 2014) leads to permanent changes throughout the lifetime of an organism.

Several authors have described immunological memory as nothing more than “leftover” cells from former immunizations (Zinkernagel et al., 1996; Zinkernagel, 2003; Zinkernagel & Hengartner, 2006). Therefore, improved knowledge of CD4+ and CD8+ T cell activation could lead to a detailed understanding of both immunological memory as a “leftovers” of specific activated CD4+ and CD8+ T cells and the immune reactions that occur later in life.

Methods

Results

Figure 1 presents the extracellular staining results for CD69, a surface marker for activated T cells. Significantly smaller percentages of activated T cells were observed after whole-blood samples were incubated overnight with low concentrations (2 µg/ml) of albumin or insulin, indicating that T cell inhibition occurred in these experiments. In contrast, the percentage of activated T cells increased after overnight incubation with 2 µg/ml of β-actin or GAPDH. Notably, incubating whole blood with a high concentration of albumin or insulin (2 mg/ml) led to a slight increase in the percentages of activated T cells (Figs. 1C and 1D). The albumin incubation experiments were also repeated using even higher concentrations of albumin (5 and 10 mg/ml), with no relevant differences in the results compared with those with 2 mg/ml albumin.

1. Tube 1: incubation with Concanavalin A (as a positive control experiment)

2. Tube 2: untreated (as a negative control experiment, background)

3. Tube 3: incubated with human albumin

4. Tube 4: incubated with human β-actin

5. Tube 5: incubated with human GAPDH

6. Tube 6: incubated with human insulin

7. Tube 7: untreated (as a negative control experiment, no difference to tube 2).

Figure 2 presents the second activation marker HLA-DR+ on CD4+ and on CD8+ T cells. Measurements were performed after three days (72 h) of incubation of cultured T cells with antigens as above and additional co-stimulation with CD28.

Figure 3 depicts the intracellular staining results for the activation marker IFγ after whole-blood samples were incubated with the aforementioned proteins. Similar to the trend for surface markers, the levels of intracellular IFγ indicated that incubation with β-actin or GAPDH led to significantly stronger T cell activation than incubation with albumin or insulin. Again, however, incubating the samples with a high concentration of albumin or insulin (2 mg/ml) increased the percentage of cytokine-positive T cells (Figs. 3C and 3D).

Figure 4 depicts the intracellular staining results for the intracellular activation marker interleukin-2 (IL-2) after T cell culture incubation with the aforementioned proteins for 72 h (3 days).

Figure 5 presents the results of T-reg cell measurements performed by staining for CD4+, CD25 +, and FoxP3. The percentage of T-reg cells increased significantly following overnight incubation with albumin or insulin, whereas no significant increase was detected following overnight incubation with β-actin or GAPDH. Therefore, the percentage of T-reg cells and activated T cells, as measured by the expression of surface and intracellular markers, showed differential responses to the four proteins: albumin and insulin increased the percentage of T-reg cells but not activated T cells, whereas β-actin and GAPDH increased the percentage of activated T cells but not the percentage of T-reg cells. Incubating whole blood with a high concentration of albumin or insulin (2 mg/ml) resulted in almost the same percentage of T-reg cells as observed when the cells were incubated with a low concentration (2 µg/ml) of these proteins.

Figure 6 illustrates the gating strategy. First, T cells are identified and gated using the CD3 vs. side scatter diagram, and then activation was detected. This procedure was performed by drawing CD69 vs. CD4 and vs. CD8 using two separate diagrams. Using these diagrams, activated cells were identified as double-positive T cells (CD69+ CD4+ and/or CD69+ CD8+).

Figures 7–9 show the kinetics of the cell activations and/or of the cell differentiation.

Table 1 shows the expression-level of markers and numbers of T cells and Tregs before incubation period (expression level at time-point zero).

An intracellular cytokine assay (FastImmune) was used to determine the downstream effects of exposure to intracellular proteins by examining the effects on intracellular cytokine production after incubating human whole blood with human proteins (Figs. 3A and 3B). In addition, the expressions of IFN-γ and IL-2 as activation markers of the corresponding T cells were confirmed.

T cell activation was significantly stronger when whole blood was incubated with the intracellular protein β-actin or GAPDH than when incubated with extracellular albumin or insulin. Under physiological conditions, extracellular proteins are normally present in the extracellular space and thus are generally ignored by the immune system. Therefore, incubation with albumin or insulin leads to a much weaker activation than incubation with β-actin or GAPDH.

Incubation with high concentrations of albumin or insulin resulted in high intracellular concentrations of the proteins, leading to an equal level of activation compared with β-actin or GAPDH. This activation was observed for CD69+ T cells, HLA-DR + T cells, and cytokine-positive T cells (IFg and IL-2). In addition, the influence of these proteins on the formation of T-reg cells (Fig. 3) was investigated, and a trend opposite to that of CD4+ and/or CD8 + effector T cells was observed. In particular, T-reg cells were strongly induced by albumin or insulin, which explains why exposure to albumin or insulin (extracellular proteins) suppressed CD4+ and CD8+ T cell activation: incubation with albumin or insulin suppressed T cell activation via the formation of T-reg cells.

The addition of a small amount of a human protein resulted in increased extracellular concentrations of that protein, and T cell activation was inhibited by the induction of T-reg cells after exposure to albumin or insulin (extracellular proteins). In addition, incubation with a high concentration of albumin or insulin led to significant induction of T-reg cells nearly equivalent to that following exposure to a low concentration. Conversely, when GAPDH or β-actin (both intracellular proteins) was extracellularly added at both high and low concentrations, T cells recognized these proteins as non-self-molecules, resulting in both activation at the T cell surface (CD69 and HLA-DR) and cytokine expression (IFγ and IL-2).

Discussion

Both surface marker staining for CD4+ and CD8+ T lymphocyte activation (CD69+ and HLA-DR+) and intracellular cytokine detection (IFg and IL-2) showed that pre-incubation with albumin or insulin resulted in a low percentage of activated T lymphocytes. However, contrasting results were obtained when the T cells were pre-incubated with human GAPDH or β-actin. Pre-treatment with these proteins resulted in a marked, statistically significant increase in the percentage of activated T lymphocytes, as assessed by the percentages of lymphocytes positive for CD69 (extracellular activation marker) CD4+ and/or CD8+ T lymphocytes, HLA-DR positive CD4+ and/or CD8 + lymphocytes, and/or T lymphocytes positive for intracellular cytokines (IFγ and/or IL-2).

Albumin and insulin are both physiological serum proteins located in the extracellular space and serum; therefore, they should not be able to provoke an immune response or activate T lymphocytes. The present study hypothesized that albumin and insulin contain immunomodulatory sequences. Indeed, immunomodulatory sequences have already been reported in the Fab and Fc termini of immunoglobulins (T-regitope sequences) (De Groot et al., 2008). These or similar sequences may also be present in other serum proteins, including albumin and/or insulin (De Groot et al., 2008), which would explain the immunomodulatory effects of both these proteins on T cell activation.

In contrast, under physiological conditions, the intracellular cytoplasmic proteins GAPDH and β-actin are not found in the serum nor extracellular space. The results of the present study indicate that these proteins are able to induce the differentiation and activation of human T lymphocytes, specifically CD4+ and CD8+, an observation that was experimentally confirmed at both the surface marker and intracellular cytokine levels. Accordingly, this activation of human CD4+ and CD8+ T lymphocytes should be effective and should promote relevant immune responses.

The findings here clearly demonstrate that some physiological intracellular proteins have the potential to activate relevant percentages of CD4+ and CD8+ T lymphocytes, in turn promoting surface marker and intracellular cytokine expression and subsequent activation of the immune system. In contrast, albumin and insulin, which are both physiological extracellular proteins, do not possess the potential to activate T cells, at least under physiologically low concentrations. However, they can induce the formation of T-reg cells. These findings are consistent with common immunological research findings (De Groot et al., 2008; Kaufmann, 1993) and with the characteristics of the evaluated proteins.

T cell activation is closely related to and necessary for the activation of B cells (Katz & Benacerraf, 1972). Several authors have described the occurrence of autoreactive activated T cells (Pauken et al., 2015; Chiorean, Mahler & Sitaru, 2014; Serre, Fazilleau & Guerder, 2015) in healthy subjects (Pauken et al., 2015) and in diverse diseases (Chiorean, Mahler & Sitaru, 2014) and disease models (Serre, Fazilleau & Guerder, 2015).

Physiologically speaking, extracellular proteins should not lead to the activation of endogenous T cells in vivo, and the corresponding sequences of TCRs on CD4+ and CD8+ T cells should be absent due to negative selection. However, although negative selection eliminates autoreactive T cells in the thymus, it does not completely eliminate all autoreactive T cells (Pauken et al., 2015; Chiorean, Mahler & Sitaru, 2014; Serre, Fazilleau & Guerder, 2015). In addition, counter-regulatory mechanisms lead to reduced activation of T lymphocytes when extracellular proteins are added. Alternatively, CD4+ cells may detect the above-mentioned substances, leading directly to reduced activation of both CD4+ and CD8+ T cells. In the present study, results differed when GAPDH and/or β-actin were added: because both these proteins are intracellular, they were able to activate both CD4+ and CD8 + T lymphocytes, even at low concentrations.

The parallel activation of both CD4+ and CD8+ T lymphocytes by GAPDH or β-actin can be explained by the concept of cross-presentation: both intracellular proteins are phagocytosed and presented via MHC-I and MHC-II receptors, which can lead to the activation of both CD4+ and CD8+ T lymphocytes or cross-presentation (Rock, 2006). However, different results were obtained when high concentrations of albumin or insulin were added (Figs. 1C and 1D); namely, a dramatic increase in CD8+ T lymphocyte activation, even though CD4+ cells were not activated or only marginally activated.

The present study had hypothesized that T cell activation by intracellular proteins, including β-actin or GAPDH, would also depend on the amount of protein added and that an increased amount would enhance T cell activation. However, this hypothesis could not be confirmed using β-actin or GAPDH. In fact, reduced T cell activation was observed when cells were incubated with high concentrations of β-actin or GAPDH. This effect may be explained by osmotic cellular gradients.

In addition, this novel interpretation of the development of autoimmune diseases fits well with multiple observations described in the literature. For example, the hypothesis and results of the present study are in line with the recent finding that insulin often acts as an autoantigen in type 1 diabetes (Kent et al., 2005; Nakayama et al., 2005; Von Herrath, 2005; Wilson, 2005). In addition, the present study fits well with the observation that myocardial infarction can be accompanied and complicated by autoimmune reactions (Liao & Cheng, 2006).

Currently, the mosaic of autoimmunity is only partially understood (Gershwin, 2008). However, the expression of autoantibodies, which are often associated with autoimmune disease, increases in patients during early disease stages. Over the past 10 years, such autoantibodies have increasingly been demonstrated to have high diagnostic value (Gershwin, 2008). Notably, many of these autoantibodies are directed against intracellular molecules themselves, including ribonucleoproteins (RNPs) and double-stranded DNA (ds DNA) in systemic lupus patients (SLE-patients) and centromeres and type 1 topoisomerase in patients with scleroderma (Gershwin, 2008). Regardless, the mechanisms underlying this process remain poorly understood. The hypothesis and results of the present study could provide a possible explanation for such nonspecific, bystander-autoimmune phenomena.

Recently, cellular ubiquitination and proteolysis systems have been proposed as being closely correlated with the balance between immunity and tolerance (Gomez-Martin, Diaz-Zamudio & Alcocer-Varela, 2008).

CD4+ T cells are known to play an important role in the development of autoimmune diseases (Slavin et al., 2002), with their inhibition leading to the suppression of disease development (Slavin et al., 2002). Interestingly, transfected CD4+ T lymphocytes are used as the preferred vector in experimental gene therapy to treat autoimmune diseases in animals (Slavin et al., 2002).

Two well-accepted alternative hypotheses for the development of autoimmune diseases are the virus hypothesis and the molecular mimicry hypothesis. The viral hypothesis states that new, unknown viruses are responsible for the development of autoimmune diseases; however, this hypothesis remains unconfirmed in light of current knowledge (Denman & Rager-Zisman, 2004). Accordingly, future studies that attempt to detect unknown viruses should be performed, and the strong postulates of Koch may need to be changed (Denman & Rager-Zisman, 2004).

Cardiovascular disease has been shown to occur much more frequently in patients with rheumatoid arthritis and to more often lead to death in these patients than in healthy individuals (Sarzi-Puttini et al., 2005).

The present study has shown that, in principle, physiological autoreactive T lymphocytes are present even in healthy individuals. Several studies have shown the presence of a large repertoire of T lymphocytes oriented toward various organs and/or tissues in healthy mice (Sharma et al., 2007). If these autoimmune T lymphocytes expand, an autoimmune attack may occur (Sharma et al., 2007). In addition, several authors have reported the spontaneous development of autoimmune diseases in mice; these diseases target the skin, lungs, liver, and tail when T-reg cells (CD4+, CD25+, and Foxp3+ cells) are absent (Foxp3 protein-deficient scurfy mice Sharma et al., 2007; Davies, 2008; Goodnow et al., 2005). These results show that additional mechanisms in healthy individuals prevent autoimmune attacks, including important T-reg cells, which the present study investigated. However, the immune system is regulated by other control systems; indeed, multiple control systems appear to be necessary, particularly due to the possibility of normal, healthy individuals developing autoimmune diseases.

These mechanisms are both important and necessary for the accurate control of the self/non-self recognition process.

Over the past decade, several studies have reported that all immune-system components appear to be involved in the development and modulation of atherosclerosis (Bassi et al., 2009). In particular, pentraxins and antibodies appear to play important roles. Pentraxins are ancient multimer molecules that are also important to the recognition of self- and non-self structures and are functionally complementary to the cellular recognition of self- and non-self molecules (Bottazzi et al., 2006). As described above, these observations match the hypothesis outlined by the present study.

Recently, a mechanism has been described in which intracellular endogenous proteins are presented by MHC-II due to autophagy by antigen-presenting cells (Münz, 2012). This effect might be dependent on the exact physiological context and on the concentration of the respective intracellular protein in the plasma, as shown by the present study. The occurrence of this autophagy-driven presentation of intracellular peptides to MHC-II during infections and similar situations could explain the hygiene hypothesis. To date, autophagy has been described in connection with infection with the herpes simplex virus and influenza viruses (Münz, 2012). If no infection occurs in early childhood, then the MHC repertoires are different, and no cross-presentation occurs. However, if infections occur in early childhood, the intracellular antigen spectra might better display via the MHC-II pathway, and resistance to autoimmune diseases might increase.

Herein a strong division of the antigenic spectrum into an extracellular part, consisting of CD4 T cell receptors, and an intracellular part, consisting of CD8 T cell receptors is postulated.

With regard to the phenomenon of cross-presentation (presentation of exogenous proteins by the MHC-I receptor), the following must be said.

Scientific work in recent years has shown that cross-presentation is a limited, highly specialized, and precisely regulated process (Fehrens et al., 2014). The process is limited only to specific specialized cells (DC) (Fehrens et al., 2014) and specific organelles within these cells (Brode & Macary, 2004). In addition, many specialized enzymes are involved in precisely regulating the process (Brode & Macary, 2004).

Furthermore, as has been described, only antigens already bound to specific receptors (immunoglobulins and Fc receptors) can be taken up by these cells and then processed by cross-presentation (Platzer, Stout & Fiebiger, 2014).

However, if this is the case, then the phenomenon of cross-presentation does not in any way contradict the division of the antigen spectrum into an extracellular and intracellular spectrum, as described here, and does not preclude the representation of these divisions by the various T cell receptors and the various lymphocyte populations (CD4+ and CD8+ T lymphocytes).

On the contrary, if uptake and processing of an antigen is possible via cross-presentation only after binding to specific receptors (immunoglobulins and Fc receptors Platzer, Stout & Fiebiger, 2014), then that antigen has already been recognized as foreign, ensuing that the antigen is not a body-own-specific structure of the extracellular space. This strongly supports the correctness of the relationship postulated here.

In order to establish these findings it is necessary to test a line of further extracellular and intracellular proteins for their immune activation potential in future studies. In addition, species difference studies should also be performed in future research.

Conclusions

Overall, in addition to the location being important for immunization, as described herein, other mechanisms may be relevant to the development of autoimmune diseases, including molecular mimicry, viral infections, and redox reactions of antibodies, all of which may be responsible for the development of autoantibodies. However, the results of the present study suggest an additional mechanism; namely, the misguided formation of proteins that become antigens.

Several aspects of the experiments conducted in the present study have been reported in part by other researchers, although their research was conducted within different contexts. For instance, the existence of an autoreactive T cell repertoire, even in healthy individuals, has been reported by several authors (Kamate et al., 2007; Walker & Abbas, 2002; Munk et al., 1989; Zou et al., 2008; Vella et al., 2009). Their findings are consistent with those of the present study, and a hypothetical explanation for this phenomenon is provided herein. In addition, an autoreactive T cell response toward insulin was reported by Ito and Yang but only in patients with autoimmune diabetes (Ito et al., 1993; Yang et al., 2014). Lastly, induction of T-reg cells in response to a human protein—namely, intravenous immunoglobulin—was reported by Maddur et al. (2001) and Cousens et al. (2013).

The present study confirmed and extended their experimental results, placed them within the context of disease development, and provided an explanation. The present study is in line with the following novel proposal of autoimmune regulation: When self-molecules occur at the wrong place in an intact organism, they are recognized as non-self-molecules, leading to the activation of immune cells. Therefore, structural protein abnormalities, mutations, or errors, including those in signaling, are not necessary to induce immunization against the organism itself (self-immunization); indeed, a mere mislocalized protein is sufficient.

Supplemental Information