From the textbooks, one can readily learn that CD4+ T-cells display a variety of helper functions, necessary for the mounting of an efficient adaptive immune response. Indeed, for about two decades, the study of CD4+ T-cells has been mainly focused on their polarization into T helper 1 or T helper 2 cells and the role of these subpopulations in cellular and humoral immunity, respectively; so that CD4+ T-cells have commonly been referred to as helper T-cells. Nonetheless, the tendency may be changing: with new subsets and new functions being ascribed to CD4+ T-cells, the emphasis on Th1/Th2 studies seems on the decline. Suppressor or regulatory CD4+ T-cells raise the strongest interest nowadays: their role in regulating T-cell proliferation comes out as being particularly important and the perspective to manipulate the activity of these cells may offer new hopes of improving immune based therapies (e.g. immune responses to vaccines).
In addition, over the past two decades, many mouse and human studies have reported the acquisition of lytic capacity by CD4+ T-cells [1–9]. However, these observations were usually restricted to cell lines and CD4+ T-cell clones generated by in vitro culture, and have therefore raised much disregard, sceptics arguing that such cytotoxic CD4+ T-cells represent an in vitro artefact and have no physiological role. Interestingly, isolated but repeated reports have recently described the presence of cytotoxic CD4+ T-cells detected directly from peripheral blood, i.e. in vivo, in various human pathologies like viral infections (HIV, CMV and EBV) [10,11], rheumatoid arthritis , ankylosing spondylitis  and B-cell chronic lymphocytic leukaemia . Healthy individuals seem to display few of these cells (on average no more than 2% of the whole CD4+ T-cell population), in contrast, up to 50% percent of the CD4+ T-cells in some HIV infected donors for instance can exhibit clear cytotoxic potential. Studies of clones and cell lines have suggested that CD4+ CTLs use the perforin-dependent cytotoxic mechanism, rather than the Fas-dependent pathway [8,15–17]. In keeping with these data, ex vivo (i.e. directly from blood) analysis of cytotoxic CD4+ T-cells indicate that they have lytic granules containing cytotoxic factors such as granzymes and perforin, and that their lytic activity is HLA class II restricted [10–14]. It is therefore important to appreciate that these cells are not merely an in vitro artefact.
CYTOTOXIC CD4+ T-CELLS AND DIFFERENTIATION
The existence of cytotoxic CD4+ T-cells in humans raises obvious questions concerning their nature and their role. There is no evidence to suggest that these CD4+ CTL are either derived from CD8+ T-cells or belong to some family of NK like cells or T regulatory cells. Instead, they seem to represent a subset of antigen experienced or memory cells, as indicated by their phenotype (CD11ahigh, CCR7–), as well as their reactivity for CMV antigens in particular, but also HIV antigens [10,11]. Interestingly, cytotoxic CD4+ T-cells are characterized by a loss of CD27 surface expression, as well as CD28, but gain of CD57, an overall phenotype that indicates an advanced stage of cellular differentiation. Increasing interest has recently been granted to the process of differentiation or post-thymic development of CD4+ T-cells, and several investigators seem to agree with a model of CD4+ T-cell differentiation characterized by a sequential down-regulation of CCR7, CD27 and then CD28, accompanied by changes of their functional characteristics [18–22] (Fig. 1). This makes an intriguing parallel with the process of CD8+ T-cell post-thymic development , all the more because, like in the case of CD8+ T-cells , virus specific CD4+ T-cells have been reported to exhibit distinct differentiation phenotypes in different infections [18,19,21,24,25]: Flu, HCV, EBV and HIV specific CD4+ T-cells are less differentiated than CMV specific CD4+ T-cells (Fig. 1). While the cytotoxic potential of CD8+ T-cells increases with differentiation, CD4+ T-cells gain such potential, with the acquisition of lytic granules with granzymes, as they lose CD27 expression, and acquisition of perforin at the CD28 negative stage, so that highly differentiated CD4+ T-cells become cytotoxic. It will be interesting to study the mechanisms of CD4+ T-cell cytolytic activity, for instance to see if this activity does not require the formation of a stable mature immunological synapse, but only of an early or lytic synapse with a low stimulation threshold, as this has recently been shown for CD8+ T-cells [26,27].
The presence of highly differentiated CD4+ CTL seems to correlate with conditions of strong or chronic activation like in infections with CMV, EBV, or HIV, or in cases of rheumatoid arthritis. Immune activation seems therefore to be a major driving factor of CD4+ T-cell differentiation, as it is actually known to be for CD8+ T-cells [28,29]. This may provide the explanation for the common observations of CD4+ CTL in cultured cell lines and clones, likely due to the in vitro conditions of prolonged stimulation and proliferation that drive further differentiation of CD4+ T-cells and the acquisition of lytic potential. Overall, these findings portray CD4+ CTLs as highly differentiated antigen experienced CD4+ T-cells driven to this stage through activation.
THE PHYSIOLOGICAL ROLE OF CYTOTOXIC CD4+ T-CELLS
Nonetheless, the identification and characterization of CD4+ CTLs in humans have not permitted to understand their role yet. The limitation of available tools (such as class II tetramers) and identified class II restricted epitopes renders the study of CD4+ T-cells more complex than it has been for CD8+ T-cells, so that it is currently only possible to speculate as regards the physiological role of CD4+ CTLs. Obviously, it is tempting to hypothesize that they may play a role in containing viruses which infect HLA class II expressing cells (e.g. EBV in B cells or HIV in activated CD4+ T-cells), or that they would preferentially target the class II antigen presentation pathway, in infections with viruses such as HIV, EBV and CMV, which can prevent normal MHC-class I expression in order to escape immune recognition by CD8+ T-cells. However, it may also be possible that these cells represent a by-product resulting from inflammation and elevated activation, exhibiting increasing characteristics of replicative senescence and demonstrating cytotoxic potential for reasons still unclear.
The identification of similar cells in the mouse would be an important advance, enabling the elaboration of diverse strategies to uncover their role. The study of CD8+ T-cells in mouse models has provided considerable information as regards the role of CD8+ T-cell subsets in protective immunity, although a complete analogy between human and mouse remains to be reached. In the present issue of Clinical and Experimental Immunology, Lyadova et al.  have identified CD27- CD4+ T-cells in the mouse, which may represent highly differentiated cells and play a particular role in the immune response as postulated by the authors. These cells produce increased levels of IFNγ and their presence in the lung is related to infection with M. bovis BCG or M. tuberculosis. It is likely that the inflammatory conditions induced by these pathogens have resulted in the differentiation of CD4+ T-cells. These cells may represent an equivalent to highly differentiated CD4+ T-cells in humans. In addition of their importance in the context of mycobacteria immune control, the findings by Lyadova et al. could be the first step towards a better understanding of the role of CD4+ T-cell differentiated subsets, in particular CD4+ CTL. It will be interesting to determine if murine CD27- CD4+ T-cells exhibit lytic granules and a cytotoxic potential, for instance using new tools available in mouse, such as the marker of granules and degranulation CD107a as well as the cytotoxic factor granzyme B . Depletion experiments or adoptive transfer experiments of CD27- CD4+ T-cells into infected mice (e.g. transgenic CD27 knockout mouse) may then shed light as regards the importance of these cells in controlling pathogens.
The role of CD4+ T-cells in immunity is complex as exemplified by their multiple functions. Evidence now shows that CD4+ CTL exist in vivo in humans, which challenges our ordinary view of CD4+ helper T-cells in contrast to CD8+ killer T-cells. What their role and its importance is in the immune response remain totally unclear. Nonetheless, the study of CD4+ T-cell subsets of differentiation and the identification of analogous CD4+ CTL in the mouse should provide the means to get some answers. When the existence of suppressor CD4+ T-cells was first reported, much skepticism was raised; 10 years later, these cells raise the strongest interest. Is the same fate awaiting cytotoxic CD4+ T-cells? The answer is yet to come.