A key suppressor role has recently been ascribed to the natural CD4+CD25+ regulatory T cells (Treg), the removal of which leads to the development of autoimmune disease and aggravated pathogen-induced inflammation in otherwise normal hosts. The repertoire of antigen specificities of Treg is as broad as that of naïve T cells, recognizing both self and non-self antigens, enabling Treg to control a broad range of immune responses. Although widely acknowledged to play a role in the maintenance of self-tolerance, recent studies indicate that Treg can be activated and expanded against bacterial, viral and parasite antigens in vivo. Such pathogen-specific Treg can prevent infection-induced immunopathology but may also increase the load of infection and prolong pathogen persistence by suppressing protective immune responses. This review discusses the role of Treg in the prevention of exaggerated inflammation favoring chronicity in bacterial or fungal infections and latency in viral infections. Special attention is given to the role of Treg in the modulation of gastric inflammation induced by Helicobacter pylori infection. Findings in both experimentally infected mice and humans with natural infection indicate that Treg are important in protecting the H. pylori-infected host against excessive gastric inflammation and disease symptoms but on the negative side promote bacterial colonization at the gastric and duodenal mucosa which may increase the risk in H. pylori-infected individuals to develop duodenal ulcers.
A major mechanism for self/non-self discrimination by the immune system and establishment of self tolerance is the clonal deletion of self reactive T and B cells exposed to self antigens during development in the thymus . The deletion mechanism is not complete however, and potentially hazardous self-reactive lymphocytes are present in the periphery of normal individuals. It has become increasingly evident that active suppression of self-reactive T cells by regulatory T cells takes place in the periphery of normal individuals avoiding the onset of harmful autoimmunity. Several phenotypically distinct subsets of suppressor-regulatory T cells have been described based on one or more surface marker antigens and/or cytokine production profiles; for e.g., the natural CD4+CD25+ T regulatory cells (Treg) , the IL-10 secreting Tr1 cells  and the TGF-β secreting Th3 cells , which functionally both in vitro and in vivo have been shown to suppress the proliferation and cytokine secretion of effector T cells.
However, an accumulating body of evidence now indicates that the most important among these regulatory T cells are the unique lineage of thymus-derived CD4+CD25+ T cells referred to as natural regulatory T cells (Treg) . Treg constitutively co-express several cell surface markers (Table 1) that allows for their discrimination and subsequent isolation for carrying out functional studies . In vivo, Treg were first identified based on their active suppression of self-reactive T cells in normal mice. Removal of thymus at day 3 of life in mice or transfer of T cells depleted of Treg to immune deficient animals resulted in wide spread autoimmune disease. Co-transfer of Treg together with the pathogenic cells prevented autoimmune disease proving the unequivocal role of Treg in the prevention of harmful immune reactivity to self in normal mice [2,7]. Recent studies suggest that Treg are also important in the induction of other regulatory T cells such as Tr1 and Th3 cells (see below). Treg have also been purified and characterized from the peripheral blood [8,9], thymus [9,10] and the cord blood  of humans. Interestingly, cord blood Treg have reduced suppressive efficacy compared to Treg isolated from peripheral blood suggesting that further differentiation and expansion of the Treg may take place in the periphery .
Table 1. Cell surface and intracellular markers constitutively expressed by thymus derived natural Treg
2Characteristics and functions of Treg in vitro and in vivo
In vitro Treg from both mice and humans are anergic to stimulation via their TCR and further inhibit CD4+CD25+ T cell responses to anti-CD3 stimulation [12,13]. Suppression is cell contact dependant and occurs only when Treg are activated through their T cell receptor (TCR) . Several studies have shown that, Treg fail to suppress if they are fixed with formaldehyde before activation or if they are cultured with an antigen other than their specificity [8,12]. The suppression of CD25+ responses by Treg can be overcome by addition of IL-2 in culture or by enhancing endogenous IL-2 production by the addition of anti-CD28 antibody [12,14].
Recent evidence suggests that in contrast to the in vitro observations, Treg are not anergic to stimulation via their TCR in vivo and can proliferate as extensively as naïve T cells in response to immunization [15,16]. Interestingly, in spite of antigen specific proliferation in vivo, Treg failed to produce any IL-2 or effector cytokines and do not upregulate CD40L. Thus, after immunization while CD4+CD25+ T cells expand and acquire effector T cell phenotype, Treg expand but do not adopt a phenotype consistent with the provision of T cell help to CD8+ cells or B cells . In addition, in vivo expanded Treg retain their suppressive function in vitro .
Murine scurfy and human IPEX (immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome) are X-linked recessive disorders of immune regulation. Efforts to identify genetic defects in IPEX patients or scurfy mice have revealed mutations in the Foxp3 gene (FOXP3 in humans) encoding a novel and highly conserved protein called scurfin . Several studies in both mice and humans have shown that the Foxp3 gene is constitutively and specifically expressed in natural Treg and play an indispensable role in their development and furthermore, the ectopic expression of the Foxp3 gene in CD4+CD25+T cells resulted in cells with a Treg phenotype and function . However, there are still some conflicting reports as to whether the expression of Foxp3 gene is exclusive to Treg or could induced regulatory T cells that have acquired suppressor function also express Foxp3. In a recent study, Apostolou and Von Boehmer  have shown in TCR transgenic mice specific for the hemagglutinin peptide that CD4+CD25+ induced suppressor T cells express the Foxp3 gene transcript. Thus, consensus in this area of research is that even though Foxp3 could be expressed by suppressor T cells that are not “natural” thymus derived Treg; it still confers on the T cells a suppressor phenotype.
The mechanism of suppression by Treg has been a subject of intense investigation in several laboratories these past few years. Studies both in humans and animal models suggest that Treg exert their effect through a cell contact dependent mechanism. However considering their low frequency (5–10% of the CD4+ T cells), it seems unlikely that this is the main mechanism leading to both local and systemic suppression. Through some elegant experiments Dieckmann et al.  and Jonuleit et al.  have shown that suppression by Treg in humans could take place in two steps. The first step is cell contact dependent; perhaps mediated through cell surface bound TGF-β, and involves the “education” of CD4+ T helper cells to become induced suppressor cells (Tsup). In the next step Tsup cells then exert their suppressive effects on effector T cells in a cell contact independent manner through the secretion of cytokines IL-10 and TGF-β. In addition, it has also been shown in humans that Treg in the peripheral blood can be separated based on their surface expression of homing receptors, α4β7 (mucosal homing) and α4β1 (peripheral homing). Both populations of sorted Treg were able to confer suppressive capacity on conventional CD4+ T cells converting them to Tsup cells . Interestingly, α4β7 sorted Treg, induced IL-10 producing (Tr1 like) Tsup cells and α4β1 sorted Treg induced TGF-β producing (Th3 like) Tsup cells, respectively. Thus, in healthy individuals, circulating peripheral Treg inducing TGF-β secreting Th3 like Tsup cells could play a central role in preventing autoimmunity, and the mucosa-homing Treg inducing IL-10 secreting Tr1 like Tsup cells could play a more important role in the prevention of intestinal inflammation including IBD .
3The role of Treg in the modulation of pathogen induced inflammation
The recognition of an important role for regulatory T cells in the suppression of pathogen induced inflammatory responses has just started to emerge. Recent evidence suggests that the activation of regulatory T cells including both Treg and Tr1 cells might result in decreased pathological responses and prolonged persistence of infection as a mechanism for the maintenance of pathogen specific immunologic memory. In a study by Mc Guirk et al.  Tr1 clones specific for filamentous hemagglutinin and pertactin could be isolated from the lungs of mice chronically infected with Bordetella pertussis. These Tr1 clones produced high levels of IL-10 but not IL-4 or IFN-γ and suppressed Th1 responses in vitro to B. pertussis or an unrelated pathogen. A role for Treg cells in the modulation of immune responses to pathogens has also been described in two different murine models of infection caused by Leishmania major and Pneumocystis carini. In both models it was shown that the depletion of Treg resulted in a reduction of the infection load but at the cost of more severe inflammation [26,27]. In addition, Belkaid et al.  in their study made an important observation that, the development of a memory immune response to cutaneous Leishmania major infection in mice was dependant on the presence Treg. Thus, although Treg down regulate inflammation and promote persistence of the pathogen, the memory response generated due to antigen persistence is crucial for protection against a second encounter with the same pathogen.
Until now, most studies have focused on the role of Treg in the suppression of CD4+ T cell responses against infectious agents, but there is recent evidence indicating that Treg are also capable of suppressing effector CD8+ T cell responses. Kursar et al.  reported in a murine model of intracellular Listeria monocytogenes infection that the CD8+ responses to the pathogen is under the control of Treg. This observation was rather serendipitous as they first observed that depleting CD4+ T cells in L. monocytogenes infected mice enhanced the CD8+ T cell responses specifically to listeriolysin antigen. Further investigation revealed that depletion of CD4+ T cells in vivo indeed affected the numbers of CD4+CD25+ expressing Treg and overcame the suppressive effects of Treg on the expansion of effector CD8+ T cell responses. It is interesting to note that similar depletion of Treg (using anti-CD25 antibody) resulted in only modest elevation in the CD8+ T cell responses in these mice in response to L. monocytogenes infection. The authors hypothesize that although depletion of CD25+ T cells targets specifically Treg, the depletion achieved was only 75% compared to 90% depletion achieved with the CD4+ antibody. Alternatively, one could speculate an additional role for CD4+CD25+ regulatory T cells, such as Tr1 or Th3 cells, in the mice treated with anti-CD25 antibody, on the suppression of CD8+ T cell responses, which to our knowledge has not yet been investigated.
In addition to intracellular infections, Treg are also known to play a role in the control of CD8+ T cell responses to viral infections. A study by Suvas et al.  showed that the depletion of Treg in vivo before infection with Herpes simplex virus-1 (HSV-1) resulted in significantly enhanced CD8+ T cell responses to the immunodominant peptide. Interestingly in their study Treg isolated from chronically HSV-1 infected mice had enhanced suppressive activity against CD8+ T cell proliferative responses to the immunodominant HSV peptide in vitro compared to Treg isolated from uninfected naive mice. The authors propose that chronic viral infection leads to the induction and expansion of Treg in vivo that in turn inhibit antiviral immune responses. Whether depletion of Treg during chronic viral infections could a beneficial immunological intervention to the host is still a matter of debate. Further studies need to be carried out on the effects of Treg depletion and the side effects to the host before it could be considered as a treatment approach in humans.
How do Treg selectively suppress autoimmune responses or excessive antimicrobial responses, but still allow the development of protective responses against invading pathogenic microbes? This question was answered to some extent by a study by Caramalho et al.  showing that Treg in normal mice selectively express Toll-like receptor (TLR) 4, 5, 7 and 8, while TLR 1, 2 and 6 appear to be more broadly expressed on CD4+ T cells. As a consequence of expressing TLR-4, Treg respond in vitro to lipopolysaccharide (LPS) stimulation, which elicits proliferation, enhanced survival and also increased suppressive capacity. However, it is also known that LPS stimulation of dendritic cells (DC) triggers their maturation leading to increased MHC expression and induction of co stimulatory molecules such as CD80 and CD86. Thus it seems evident that stimulation through TLR has different effects on the immune responses, (i) to trigger DC maturation and augment T cell mediated adaptive immunity and (ii) to activate Treg and thereby down-regulate immune responses. Although these effects are apparently conflicting, the concentration of LPS required for evoking the proliferation of Treg is several orders of magnitude higher than the concentration required for in vitro activation of DC . It is therefore likely that upon gram-negative bacterial infection, LPS stimulation of TLR on DC will stimulate maturation of DC thereby inducing the activation and expansion and differentiation of microbe specific naïve T cells together with LPS-specific B cell antibody responses. If a large enough amounts of LPS is present, e.g., as a result of successful attack of the pathogen by antibody and complement, Treg having a higher threshold for activation will respond to LPS through TLR and increase their suppressive activity and thereby prevent local or systemic immunopathology.
Treg can also be considered to be harmful to the host as pathogens can take advantage of the suppressive effects of Treg and impair immune responses and eradication of the infection. For example patients with an ongoing malaria infection frequently show reduced immune responses not only to the malarial parasite but also to unrelated antigens suggesting that an active T cell suppressive mechanism operates during the course of malaria infection. In a mouse model of Plasmodium yoelii infection, Hisaeda et al.  addressed whether Treg are the mediators of immune suppression observed during malaria infection. Interestingly, wild type mice that were infected with a lethal dose of P. yoelii all died, while depletion of Treg resulted in rescue of 80% of the infected mice with resulting eradication of the parasite from the blood .
4The function of Treg in the regulation of Helicobacter pylori induced gastric inflammation
Helicobacter pylori, a spiral gram negative bacterium, colonizes the human stomach and duodenum and causes chronic inflammation, gastric atrophy and peptic ulcers in a sub-population of infected individuals . The infection is acquired during early childhood and is usually life-long. Although half the world's population is infected with this bacterium only approximately 10–15% of those colonized develop symptoms and the rest of the population is designated as “asymptomatic carriers”. It is attractive to speculate, supported by data from an experimental infection model that the activation of Treg in asymptomatic carriers keeps the pathology mild enough to avoid symptoms [34,35].
In our laboratory Lundgren et al.  have shown in H. pylori infected asymptomatic individuals, that the memory T cell responses to H. pylori antigens in the peripheral blood is under the control of Treg. Removal of Treg specifically from the memory T cell population increased the proliferative responses to H. pylori antigens and importantly, addition of Treg back to the memory T cells suppressed the H. pylori specific responses but failed to suppress responses to unrelated antigens. In addition, CD4+CD25high T cells (putative Treg) isolated from the gastric and duodenal mucosa of H. pylori infected asymptomatic carriers express the specific Treg marker FOXP3, supporting an important role for Treg in maintaining a balance between chronicity and development of symptoms at the site of infection .
We and others have previously shown in a mouse model of H. pylori infection that mucosal immunization with H. pylori antigen together with an adjuvant results in protection against H. pylori infection resulting in reduced bacterial loads but not complete eradication [37,38]. However the reduction in bacterial load achieved as a result of immunization is associated with inflammation in the gastric mucosa referred to post-immunization gastritis due to an expansion of H. pylori specific T cells that migrate to the gastric mucosa to perform their effector functions . However, a study by Garhart et al. , clearly showed that post-immunization gastritis resolves after a few months with low levels of bacteria remaining in the stomach.
To address the role of Treg in the modulation of immune response to H. pylori infection we used a H. pylori antigen specific in vitro co-culture assay, wherein Treg were mixed with CD4+CD25+ effector cells and antigen presenting cells (APC). Using this system, we have recently been able to test the hypothesis of a defective Treg function in H. pylori infected mice after immunization. Our results conclusively showed that Treg from naïve mice and less efficiently from infected mice suppressed the CD25+ effector T cell response to H. pylori antigens, while Treg isolated from the H. pylori immunized mice were the least efficient in suppressing H. pylori specific CD25+ effector T cell proliferation and cytokine secretion . We suggest based on our results, that post immunization gastritis in vivo could be attributed to a defective Treg function leading to uncontrolled effector T cell responses to H. pylori antigens, while suppression of H. pylori specific responses by Treg in naïve mice may enhance the susceptibility to infection.
The reason for the highest efficacy of Treg isolated from naïve mice in the suppression of CD25+ effector T cell responses to H. pylori antigen is not clear. The mice used for these studies were specific pathogen free (SPF) but not Helicobacter free, and could thus have a low-grade infection in the intestine with naturally infecting Helicobacter sps such as H. hepaticus. Kullberg et al.  have recently shown using a H. hepaticus model of intestinal inflammation in IL-10+/+ mice that Treg are essential in the down-regulation of H. hepaticus driven intestinal inflammation. Thus, in naïve mice it is reasonable to assume that chronic H. hepaticus infection in the intestine expand Helicobacter specific Treg in vivo that cross-recognize H. pylori antigens and suppress effector T cell proliferation in vitro.
The site of Treg suppression of effector immune cells secreting pro-inflammatory cytokines and driving gastric inflammation is not yet known. H. pylori specific Treg may be activated in the draining lymph nodes of the gastrointestinal tract (mesenteric lymph nodes, MLN) and in turn inhibit DC activation and lifespan, preventing the further activation and development of effector T cells. Alternatively they may act locally in the gastric mucosa inhibiting T-effector cell function, particularly macrophage activation (Fig. 1). Given the potent pro-inflammatory milieu in the gastric mucosa of H. pylori infected individuals, it seems likely that Treg would have to work at multiple sites to perform their suppressor function to effectively down regulate H. pylori induced inflammation. In addition, it is possible in H. pylori induced inflammation in the stomach, that Treg specific for both H. pylori antigens and self antigens are activated  and can suppress effector T cell responses in a by-stander manner through the secretion of down-regulatory cytokines (Fig. 1).
However, as mentioned above the protective function of Treg against severe gastritis in H. pylori infection is not achieved without costs. As shown in our mouse studies , the protective effect of Treg against gastritis was associated with more extensive bacterial colonization. Evidence from H. pylori infected patients developing duodenal ulcers suggests, that the ulcers develop in the apparent presence of increased numbers of mucosal Treg and increased production of anti-inflammatory cytokines IL-10 and TGF-β in the duodenal epithelium [41,42]. In addition, the findings of increased numbers of H. pylori bacteria in duodenal biopsies of patients with duodenal ulcers compared to asymptomatic carriers  suggests that Treg in the duodenal mucosa may contribute to pathology by allowing more extensive local colonization of H. pylori bacteria which cause tissue damage by producing vacuolating cytotoxin and possibly other virulence factors. At the same time Treg in the gastric mucosa are important in the protection against symptomatic gastric inflammation and perhaps even progression to cancer.
The studies summarized above point to an equally important role for Treg in the down-regulation of immune responses to infection as has previously been shown for maintenance of self-tolerance and prevention of autoimmunity. Ideally, during an infection Treg could modulate the delicate balance between the need to mount an immune response that can effectively control the pathogen and the associated risk of a too vigorous immune response that would cause severe inflammation and disease. Expansion of the number of Treg or enhancement of their function may inhibit tissue damaging inflammation but at the cost of increasing the load of infection, and conversely their removal or inhibition may enhance host resistance to the infection.
Potentially, vaccines against infectious diseases may be designed to steer pathogen-specific Treg in the desired direction. Most, if not all infections trigger some extent of inflammation, as does efficient vaccination against pathogens. However, infection with H. pylori represents a situation where the balance between beneficial and harmful effects of the immune response seems particularly complex, including the possible role of Treg at different locations. This could have obvious bearing on the prospects for vaccine development against H. pylori, especially when a vaccine for use in infected individuals is being considered, and will require further studies in both experimental systems and in humans with natural H. pylori infection. Will it for instance be possible to design a vaccine that combines the ability to raise a strong protective immune response to a specific H. pylori antigen while at the same time expanding the number of Treg and/or inducing other regulatory T cells against different H. pylori antigens? And would such a combined T effector/Treg-inducing vaccine provide the desired result of increased pathogen elimination together with prevention of tissue damage? Whilst there are still many unanswered questions regarding the mechanism of suppression and antigen specificity, their potential as therapeutic targets in inflammatory diseases should encourage efforts to further characterize these cells.