Lipopolysaccharide-Activated CD4⁺CD25⁺ T Regulatory Cells Inhibit Neutrophil Function and Promote Their Apoptosis and Death¹

The innate immune system recognizes the microbial environment via pattern recognition receptors such as TLRs that can be activated by a range of pathogen-associated molecules including LPS, lipoproteins, lipoteichoic acid, peptidoglycans, CpG-containing DNA, and flagellin (1). Polymorphonuclear neutrophils (PMNs)³ are the most abundant type of leukocytes in humans. They possess a wide arsenal of antimicrobial effector functions and the ability to initiate inflammatory response. PMN activation is initiated upon the recognition of opsonized particles or the direct identification of microbial products via pattern recognition receptors. Neutrophils express all known TLRs with the exception of TLR3 (2). Ligation of neutrophil TLRs, especially TLR2 and TLR4, triggers signal transduction pathways that contribute to the prolongation of cell survival, the augmentation of cell adhesion, the production of reactive oxygen intermediates (ROIs), chemokines, and cytokines, and the enhancement of degranulation and phagocytosis (3, 4). Thus, neutrophil activation and survival should be tightly regulated, as excessive release of cytotoxic and inflammatory substances can cause bystander tissue damage. In addition, immunomodulatory agents released by PMNs can coordinate early responses of monocytes, macrophages, dendritic cells, and lymphocytes, providing a link between the innate and acquired immune responses (5).

Recently, TLRs have been shown on mice and human CD4⁺CD25⁺ T regulatory (Treg) cells (6, 7, 8, 9). The reports suggest that the stimulation of CD4⁺CD25⁺ Treg cells via TLRs could modulate their function. For example, mice lacking TLR2 have decreased numbers of peripheral CD4⁺CD25⁺ Treg cells (10), and LPS, a TLR4 ligand, has been shown to enhance the suppressive activity of CD4⁺CD25⁺ Treg cells in mice (6). Moreover, CD4⁺CD25⁺ Treg cells were necessary to control the increased TLR4- and TLR2-induced inflammatory cytokine production by innate immune cells after burn injury in mice (11). In addition, CD4⁺CD25⁺ Treg cells are known to control the immune responses to a wide range of pathogens (12). Thus, interactions between CD4⁺CD25⁺ Treg cells, innate immune cells, and microorganisms seem to be very probable. Actually, very recent reports have shown that CD4⁺CD25⁺ Treg cells inhibit the proinflammatory activities of PMNs in mice and modulate dendritic cell activation (13, 14).

We suggest that during early immune response CD4⁺CD25⁺ Treg cells could be attracted to the inflammatory environment, exposed to LPS, and interact with immune cells. We investigated whether CD4⁺CD25⁺ Treg cells can directly modulate PMN function, and how the presence of LPS affects these interactions. Our study demonstrates that CD4⁺CD25⁺ Treg cells inhibit PMN function and reverse LPS-induced survival of neutrophils. These findings suggest that CD4⁺CD25⁺ Treg cells may have an important role in the direct control of innate immune responses.

Following approval of the Polish Mother’s Hospital Institute Ethical Committee, venous blood from healthy donors was collected in heparin. PBMCs were isolated by density centrifugation over Lymphoprep (Axis-Shield). Human CD4⁺CD25⁺ and CD4⁺CD25⁻ T cells were purified using the Treg cell isolation kit (Miltenyi Biotec) and the AutoMACS cell sorter (Miltenyi Biotec). First, CD4⁺ T cells were purified by negative selection. In the second step, CD25⁺ cells were isolated by positive selection with anti-CD25 microbeads (5 μl per 10⁷ cells) using a double-column procedure. The flow-through from CD25⁺ selection was used for the following purification of CD25⁻ cells by the addition of 20 μl of anti-CD25 microbeads to 10⁷ cells and passage over a MS column (Miltenyi Biotec). Purity of both cell populations assessed by two-color flow cytometry was >95%.

PMNs were isolated in parallel in LPS-free conditions using Polymorphprep (Axis-Shield) centrifugation of freshly drawn blood. The PMN population was assessed by flow cytometry and shown to be >95% pure. After isolation, PMNs were maintained in culture medium and kept at 37°C with 5% CO₂ until the completion of CD4⁺CD25⁺ and CD4⁺CD25⁻ T cell purification (∼5 h). Afterward, PMNs were incubated with CD4⁺CD25⁺ or CD4⁺CD25⁻ T cells in decreasing amounts, starting at a 100:1 ratio in RPMI 1640 supplemented with 10% FCS (Sigma-Aldrich), 2 mM l-glutamine (Invitrogen Life Technologies), and 1% penicillin/streptomycin (Sigma-Aldrich) in 12 × 75-mm sterile (nonpyrogenic) polystyrene round-bottom tubes (BD Biosciences). Ultrapure LPS from Escherichia coli serotype R515 (Alexis Biochemicals) was added directly to the cocultures of PMNs with CD4⁺ T cells. In some cases, before use in different experiments the purified CD4⁺CD25⁺ and CD4⁺CD25⁻ T cells were preincubated with LPS and then cultured with freshly isolated autologous PMNs. Viability of the PMNs was determined directly after the purification process and 5 and 18 h after culturing with CD4⁺CD25⁺/CD4⁺CD25⁻ T cells by means of trypan blue exclusion.

For immunostaining, the conjugated Abs used (BD-Pharmingen) were CD11b-PE (clone D12.11), CD62L-FITC (clone SK11), and CD16-FITC (clone NKP15i). The respective mouse, rat, or goat isotype controls were used. For TLR4 surface staining, unconjugated anti-TLR4 Abs (clone HTA125; Alexis Biochemicals) were used as the primary Abs and FITC-conjugated goat anti-mouse as the secondary Abs (DakoCytomation). Unconjugated anti-IL-10 (clone 23738.111; R&D Systems), anti-TGF-β (clone 141322; R&D Systems), and anti-TLR4 (clone HTA125; Alexis) were used for neutralization experiments. Anti-CD3 (clone HIT3a; BD-Pharmingen) and anti-CD28 (clone CD28.2; BD-Pharmingen) were used for polyclonal activation of T cells.

Surface expression of various markers was determined using FACScalibur flow cytometer (BD Biosciences) and BD CellQuest analysis software. For immunofluorescence staining, cells were incubated for 30 min with optimal dilution of each Ab.

TLR4 expression on the cell surface was detected on freshly isolated and LPS-stimulated CD4⁺CD25⁺/CD4⁺CD25⁻ T cells. Cells were harvested and stained with anti-TLR4 Ab. Then, FITC-conjugated goat anti-mouse secondary Ab and isotype control were added in parallel (FITC-conjugated mouse IgG2a). Cells were fixed with 1% paraformaldehyde and analyzed by flow cytometry.

CD11b, CD62L, and CD16 expression on PMNs was measured after incubation with CD4⁺ T cells. Samples were centrifuged, washed in ice-cold PBS without Ca²⁺ and Mg²⁺ and stained with appropriate amounts of Abs. After 30 min of incubation, samples were washed, fixed with 1% paraformaldehyde, and analyzed by flow cytometry. PMNs were identified and gated on the side scatter (SSC)/forward scatter (FSC) dot plot.

FOXP 3 expression was determined with a human Treg flow kit (BioLegend) that used the conjugated Abs CD4-PE-Cy5, CD25-PE, FOXP3-Alexa Fluor 488, and mouse isotype control IgG1-Alexa Fluor 488. Cell staining was performed according to the manufacturer’s instructions.

The binding of ANXV-FITC to phosphatidylserine was used as a sensitive measure of neutrophil apoptosis. Additional staining with propidium iodide (PI) enabled us to distinguish between the early and late stages of apoptosis, as ANXV binds to both types of cells. After incubation of neutrophils with CD4⁺CD25⁺ and CD4⁺CD25⁻ T cells for 5–16 h, samples (100 μl) were washed twice in ice cold PBS without Ca²⁺/Mg²⁺ and incubated with ANXV-FITC and PI (BD Pharmingen) according to the manufacturer’s instructions. Samples were analyzed by flow cytometry. PMNs were identified and gated on the SSC/FSC dot plot and analyzed by flow cytometry.

IL-1β, IL-6, IL-8, IL-10, and TNF-α were measured in supernatants of 18-h cultures of PMN with CD4⁺CD25⁺ or CD4⁺CD25⁻ T cells using a human inflammation cytometric bead array (CBA) kit and CellQuest software (BD Biosciences). Briefly, 50 μl of supernatants were mixed with 50 μl of PE-conjugated cytokine capture beads. After 3 h of incubation, samples were washed, fixed with 1% paraformaldehyde, and analyzed by flow cytometry.

ROI production was measured by luminol-ECL using an MLX microtiter plate luminometer (Dynex). CD4⁺CD25⁺ Treg or CD4⁺CD25⁻ T cells were first incubated in the presence or absence of LPS for 24 h. Afterward, they were cocultured for 45 min with freshly isolated autologous PMN with or without 200 ng/ml PMA (Sigma-Aldrich). In some experiments, PMNs were incubated with CD4⁺CD25⁺ or CD4⁺CD25⁻ T cells in the presence or absence of LPS, and after 18 h of incubation the PMNs were restimulated with PMA.

Isolated CD4⁺CD25⁺ Treg cells and CD4⁺CD25⁻ cells (250 × 10⁴) were suspended in 30 μl of PBS, and total RNA was extracted using the RNeasy mini kit (Qiagen) according to manufacturer’s protocol. RT-PCR were performed in FlexiGene (Techne) thermocycler using the OneStep RT-PCR kit (Qiagen) with 7 μl of total RNA solution and specific TLR4 primers (forward primer, 5′-CTgCAATggATCAAggACCA-3′; reverse primer, 5′-TCCCACTCCAggTAAgTgTT-3′). RT-PCR products were analyzed with 1.7% agarose gel electrophoresis in TAE buffer (40 mM Tris acetate and 1 mM EDTA) and visualized with ethidium bromide staining. For densitometry analysis, the intensity of the bands were measured by the Kodak 1D Image Analysis software with Kodak camera DS40-D2120 (Kodak) and normalized with GAPDH intensity.

The arithmetic means and SD values were calculated for all parameters in at least four independent experiments. Statistical analysis of differences in all of the data was done using the one-way ANOVA test. Scheffe’s test (as a post hoc test) was used for multiple comparisons when statistical significances were identified in the ANOVA test. Statistical significance was set at p < 0.05.

We investigated the presence of TLR4 on CD4⁺ T cells at the protein and mRNA levels using flow cytometry and semiquantitative RT-PCR. Flow cytometry demonstrated TLR4 extracellular fluorescence on freshly isolated CD4⁺CD25⁺ Treg and CD4⁺CD25⁻ T cells in every donor (Fig. 1 A). CD4⁺CD25⁺ Treg cells consistently expressed higher levels of surface TLR4 (p < 0.01) than their CD4⁺CD25⁻ counterparts (Fig. 1A). RT-PCR analysis confirmed that both CD4⁺CD25⁺ Treg and CD4⁺CD25⁻ T cells express TLR4 mRNA (Fig. 1B).

FOXP3 is specifically expressed at high levels in CD4⁺CD25⁺ Treg cells, and this transcriptional factor is essential for their development and function (15, 16). We therefore attempted to determine whether TLR4 stimulation by LPS had an impact on FOXP3 protein expression in Treg cells. CD4⁺CD25⁺ Treg cells were incubated with LPS at different concentrations or soluble anti-CD3/CD28 mAbs for 24 h, and FOXP3 expression was measured at two time points (Fig. 2). Following LPS stimulation (0.1 ng/ml), FOXP3 expression decreased consistently by ∼2-fold after 5 h of exposition and remained at a similar level until the end of the incubation time. Surprisingly, the addition of LPS at higher concentrations did not result in a more considerable decrease of FOXP3. TCR activation of Treg with anti-CD3/CD28 mAbs (5 μg/ml) resulted in a noticeable reduction of intracellular FOXP3 staining, which is consistent with previous studies (8, 17).

We investigated early and late changes in PMNs cocultured with CD4⁺CD25⁺ Treg cell in 45-min and 18-h incubations. As both CD4⁺ T cells and PMNs express TLR4, we used ultrapure LPS to define a probable effect of CD4⁺CD25⁺ Treg on PMNs during TLR4 stimulation. As neutrophils are the most numerous leukocytes in blood as well as at the site of acute immune response to Gram-negative bacteria, we cultured PMNs and T cells at a ratio of 100:1 (equivalent to conditions in peripheral blood) and at 0:1, which may reflect physiological proportions of PMNs and Treg cells at the site of inflammation.

ROI production is an important antimicrobial response of neutrophils. We tested ROI generation of PMNs incubated with CD4⁺CD25⁺ Treg or CD4⁺CD25⁻ T cells. First, CD4⁺CD25⁺ Treg or CD4⁺CD25⁻ T cells were incubated in the presence or absence of LPS for 24 h. After preincubation, they were washed twice in PBS and cocultured for 45 min with freshly isolated autologous neutrophils. At the end of the incubation, PMA was added to the samples. ROI generation was measured immediately using luminol-dependent chemiluminescence. CD4⁺ T cells preincubated with LPS significantly influenced ROI production by PMN; CD4⁺CD25⁺ Treg cells inhibited ROI generation, whereas CD4⁺CD25⁻ T cells notably increased ROI production. When PMA was added to the samples, the different influence of the CD4⁺CD25⁺ Treg and CD4⁺CD25⁻ T cells was even more noticeable (Fig. 3 A). We also examined ROI production in samples in which LPS was added directly to the mixture of freshly isolated PMNs and CD4⁺CD25⁺ Treg cells. After 18 h of incubation, neutrophils were restimulated with PMA, and chemiluminescence was measured immediately. In this case, the inhibition of ROI generation triggered by CD4⁺CD25⁺ Treg cells was noted in both LPS-treated and untreated cells, although the presence of LPS augmented the suppression effect (Fig. 3B). We did not demonstrate any significant influence of CD4⁺CD25⁺ Treg/CD4⁺CD25⁻ T cells on PMN ROI production when cells were cocultured at a ratio of 100:1 (data not shown).

CD11b and CD62L expression is a sensitive marker of early PMN activation. We therefore examined whether the incubation of neutrophils with CD4⁺CD25⁺ Treg and CD4⁺CD25⁻ T cells was accompanied by up-regulation of CD11b and L-selectin shedding. When neutrophils were incubated with CD4⁺CD25⁺ Treg cells at a ratio of 10:1, the dramatic increase of CD11b expression was noted (Fig. 4). In contrast, CD4⁺CD25⁻ T cells had no effect on CD11b expression. Changes in CD11b expression were accompanied by CD62L shedding (Fig. 4). LPS-triggered CD11b up- regulation and L-selectin shedding in PMNs were noted independently of the presence or absence of CD4⁺CD25⁺ Treg/CD4⁺CD25⁻ T cells in analyzed samples. When neutrophils were incubated with CD4⁺ T cell populations at a ratio of 100:1, no significant changes were noted.

Next we examined the effect of CD4⁺CD25⁺ Treg cells on cytokine production by PMNs. CD4⁺CD25⁺ Treg or CD4⁺CD25⁻ T cells were incubated together with neutrophils in the presence or absence of 100 ng/ml LPS. After 18 h the amounts of IL-1β, IL-6, IL-8, IL-10 and TNF-α were determined (Fig. 5). In contrast to CD4⁺CD25⁻ T cells, CD4⁺CD25⁺ Treg cells did not generate any detectable amounts of IL-6, IL-8, or TNF-α. We failed to detect the presence of IL-1β or IL-10 in any sample. The addition of CD4⁺CD25⁺ Treg cells to PMN suspension resulted in a significant decrease of IL-6, IL-8, and TNF-α production. In contrast, in cocultures of CD4⁺CD25⁻ T cells with PMNs we detected increased levels of cytokines, probably as a result of adding up the amounts of cytokines generated by PMNs and T effector cells. Inhibition of cytokine production was enhanced following the activation of Treg cells with LPS or anti-CD3/CD28 mAbs (Fig. 6). Treatment of CD4⁺CD25⁺ Treg cells with anti-TLR4 blocking mAbs before the incubation with PMNs diminished the suppression capacity of Treg cells. We also investigated whether the inhibition of cytokine production was dependent on the release of IL-10 and TGF-β by CD4⁺CD25⁺ Treg cells. Although IL-10 and TGF-β were undetectable in 18-h culture supernatants, both anti-IL-10 and anti-TGF-β neutralizing mAbs added to the LPS-treated samples affected cytokine production by neutrophils and attenuated the inhibitory effect of CD4⁺CD25⁺ Treg cells (Fig. 6). Interestingly, IL-10 and TGF-β blockade failed to affect the inhibitory ability of anti-CD3/CD28-stimulated Treg cells.

PMNs are short-lived immune cells, but their survival is prolonged upon activation. It has been widely reported that LPS potently delays PMN apoptosis (3, 18). Recent reports have also shown that CD4⁺CD25⁺ Treg cells can induce cell death via granzyme/perforin mechanisms (19, 20). We therefore decided to determine whether CD4⁺CD25⁺ Treg cells were able to accelerate PMN apoptosis and reverse the effect of LPS on PMNs. After isolation, neutrophils were maintained in culture medium and kept at 37°C with 5% CO₂ until the completion of CD4⁺CD25⁺ and CD4⁺CD25⁻ T cells purification (∼5 h). Then, CD4⁺CD25⁺ Treg or CD4⁺CD25⁻ T cells were mixed together with PMNs, stimulated with 100 ng/ml LPS or 5 μg/ml anti-CD3/CD28 mAbs, and cultured for a further 5 h. Over a 10-h period, ∼34% of neutrophils had undergone spontaneous apoptosis as determined by means of total ANXV, which was strongly prevented by LPS. We did not demonstrate significant differences in the ANXV⁺ PI⁻ compartment following incubation with T cells, but we found that PMNs had the greater percentage of ANXV⁺ PI⁺ cells when cocultured with Treg cells (Fig. 7). Neutrophil death was accelerated doubly in the presence of CD4⁺CD25⁺ Treg cells, which were stimulated with LPS or anti-CD3/CD28 mAbs. In contrast, CD4⁺CD25⁻ T cells were not efficacious in promoting PMN apoptosis. We also assessed neutrophil survival in a series of experiments when CD4⁺CD25⁺ Treg or CD4⁺CD25⁻ T cells were preincubated with LPS or anti-CD3/CD28 mAbs for 24 h and then cocultured with freshly isolated autologous PMNs for 5 h in the presence or absence of 100 ng/ml LPS. In this case we obtained similar results (Fig. 8).

To define the possible role of IL-10 and TGF-β in this process, anti-IL-10 and anti-TGF-β neutralizing mAbs were added to the prestimulated CD4⁺CD25⁺ Treg cells. We found that IL-10 and TGF-β blockade partly inhibited the acceleration of neutrophil death by LPS-stimulated Treg cells. In the same experiment, anti-IL-10 and anti-TGF-β neutralizing mAbs did not affect the effect of anti-CD3/CD28 stimulated Treg cells. In other experiments, PMNs were preincubated with LPS for 5 h before the coculturing with Treg cells. The preincubation of neutrophils with LPS made them completely insensitive to the proapoptotic effect of Treg (data not shown).

To confirm an effect of CD4⁺CD25⁺ Treg cells on neutrophil apoptosis, we measured CD16 expression, which decreases in apoptotic cells (21) on the PMN surface as shown by flow cytometry. As shown in Fig. 9, the rate of CD16 low-expressing cells increased doubly in the presence of CD4⁺CD25⁺ Treg cells. In parallel, PMN viability determined by trypan blue exclusion decreased in the presence of CD4⁺CD25⁺ Treg cells (Fig. 10).

Bacterial infections are accompanied by potent host reposes that are often followed by opposing anti-inflammatory effects. When immune homeostasis fails, increased inflammatory response against pathogens and their products can contribute to the development of severe pathologies. Because neutrophils are the most abundant type of leukocytes, their activation and life span should be precisely controlled. In this study we show that CD4⁺CD25⁺ Treg cells, but not CD4⁺ effector T cells, directly influence PMN function and survival and reverse the effect of LPS, the most common activator of innate immune response. When cultured for a short period of time in the absence of LPS, CD4⁺CD25⁺ Treg cells initially contributed to the up-regulation of CD11b expression on the PMN surface and the shedding of CD62L molecule, indicating neutrophil activation and possible interaction between PMNs and Treg cells. However, long-term incubations of PMNs and CD4⁺CD25⁺ Treg cells resulted in considerable inhibition of ROI and cytokine production and acceleration of PMN apoptosis. This effect was augmented when LPS was added to the cocultures or Treg cells preincubated with LPS or anti-CD3/CD28 mAbs. These experiments indicate that LPS enhances the suppressive capacity of Treg cells toward PMNs.

In a series of in vitro studies, it has been shown that CD4⁺CD25⁺ Treg cell suppressor function is dependent on the cell-cell contact mechanism rather than the secretion of TGF-β or IL-10 (22, 23, 24, 25). To test whether Treg cells inhibit PMN functions in a contact-dependent or cytokine-dependent manner, neutralizing anti-TGF-β and anti-IL-10 mAbs were added to the cultures. We found that Treg cells stimulated with LPS exert their inhibitory and proapoptotic actions through both IL-10 and TGF-β. Surprisingly, in those cases when Treg were prestimulated with CD3/CD28 mAbs, TGF-β and IL-10 neutralization did not reverse Treg inhibitory capacity. These results confirm previous evidence that Treg cells inhibit PMNs through both contact-dependent CTLA-4/B7-1 mechanism and IL-10 action (13).

The induction of apoptosis and the death of PMNs after their contact with CD4⁺CD25⁺ Treg cells may be a crucial mechanism of PMN/Treg interactions. It was previously shown that neutrophil apoptosis is associated with an overall decrease in cellular functions such as chemotaxis, ROI and cytokine production, and phagocytosis (26, 27). That finding may explain the lower ROI and cytokine production by PMNs that were cultured with CD4⁺CD25⁺ Treg cells. Our studies revealed that Treg cells induced PI positivity in PMNs, which is a typical marker of cell death. We also demonstrated the decreased rate of viable neutrophils in coculture with Treg cells. This may result from perforin/granzyme cytotoxicity of CD4⁺CD25⁺ Treg cells against cells (19, 20). Interestingly, IL-10- and TGF-blocking Abs partly inhibited the acceleration of neutrophil death by LPS-stimulated Treg cells, which indicates that multiple mechanisms are involved in Treg cell-mediated suppression. When neutrophils were preincubated with LPS before the addition of Treg cells, they became absolutely insensitive to the proapoptotic effect of Treg cells. This observation suggests that Treg cells could inhibit only those cells that have not been activated yet.

On the strength of the data presented above, we propose a model, in which during an early immune response, recruited neutrophils are first exposed to LPS at the site of inflammation, resulting in the activation of their effector function. Treg cells appear later and become activated by TLR ligands. Although Treg cells fail to suppress PMNs that have already been activated, they could potently prevent the activation of newly appeared neutrophils. This mechanism allows an effective antimicrobial immune response that is crucial for the resolution of infection. Additionally, Treg cells control the timely removal of PMNs from affected site in later stages of inflammation and limit inappropriate inflammatory potential.

It is now increasingly appreciated that TLRs orchestrate most aspects of innate immune responses and indirectly control adaptive immunity through the activation of APCs (28). Recent reports have demonstrated the presence TLRs on mice and human T and B cells whose function can be directly modulated by the ligation of TLRs in the absence of APCs (6, 7, 8, 9, 29, 30, 31, 32). Because it remains controversial whether LPS, a major component of the Gram-negative bacterial wall, can modulate T cell function, we also assessed TLR4 expression in T lymphocytes. We have shown that both CD4⁺CD25⁺ Treg and CD4⁺CD25⁻ T cells express TLR4 at the protein and mRNA level. These data contrast with the previous reports showing that cell surface TLR4 is very low and often undetectable on unstimulated CD4⁺ cells (7, 8). This discrepancy may partially result from the type of analyzed CD4⁺ T cells, as Komai-Koma et al. (7) have investigated the cord blood CD⁺ T cells that are mostly naive. Next, we demonstrated that LPS and anti-CD3/CD28 activation of Treg cells contributed to comparable changes in FOXP3 protein expression. These findings suggest that TLR4 stimulation of CD4⁺CD25⁺ Treg cells results in their subsequent activation, which is reflected by the reduction of FOXP3 expression. Other authors have shown that proliferation and cytokine production of human T cells (conventional or regulatory) was not affected by purified LPS (7, 8) but that LPS slightly enhanced the suppression capacity of TCR-activated CD4⁺CD25⁺ Treg cells (8). It is likely that CD4⁺ T cells do not express all of the components of the LPS receptor, because they express TLR4 and MD2 but not CD14 (7, 8). This may be the reason for the hyporesponsiveness of CD4⁺ T cells to LPS. However, the soluble CD14 that is present in FCS has been shown to enable LPS response in some CD14-negative cells (33, 34). The presence of soluble CD14 has been demonstrated at the site of allergic inflammation (35), suggesting the possibility that LPS responsiveness of CD14-negative cells may be induced at the site of inflammation.

In conclusion, evidence that CD4⁺CD25⁺ Treg cells potently inhibit the response of neutrophils upon exposure to LPS clearly indicates that TLR4 signaling plays an important role in homeostasis. These findings provide a new insight into mechanisms of the regulatory interplay between the innate and adaptive immune cells and suggest a new role for Treg cells in the reduction of the potentially harmful effects of excessive inflammation.

We gratefully acknowledge helpful scientific discussion and the support of Zofia Su/lowska, Ma/lgorzata Banasik, Ewa G/lowacka, and Beata Jóźwiak.

The authors have no financial conflict of interest.