CD8+ T cells play an important role in controlling pathogenic infections and are therefore key players in the immune response. It has been shown that among other factors CD4+ T cells can shape the magnitude as well as the quality of primary and/or secondary CD8+ T-cell responses. However, due to the complexity and the differences among diverse immunization or infection models, the overall requirement, the time points, as well as the specific mechanism(s) of CD4+ T-cell help may differ substantially. Here, we summarize current knowledge about the differential requirement of CD4+ T-cell help in promoting primary CD8+ T-cell responses as well as establishing functional memory CD8+ T cells in various experimental settings.
A number of different parameters influence, by virtue of their strength and composition, CD8+ T-cell activation; they subsequently also shape the size and the phenotypical and functional properties of the resultant memory CD8+ T-cell pool. These parameters include antigen-specific T-cell precursor frequencies [], the strength of the T-cell receptor interaction with peptide–MHC complexes, and the signals provided by co-stimulatory receptors, as well as innate immune system derived inflammatory cytokines [[2, 3]]. Among the factors that modulate the activation of dendritic cells (DCs), the cells that are the main inducers of CD8+ T-cell responses, is the help provided by CD4+ T cells. CD4+ T-cell engagement of DCs promotes the upregulation of certain co-stimulatory molecules (such as CD80 and CD86) on, as well as the release of pro-inflammatory cytokines such as IL-12 by, DCs. Thus, in many defined experimental settings, T helper cells have been implicated in the expansion and survival of CD8+ T cells during the primary response, and have a key role in establishing long-lived, functionally robust memory CD8+ T-cell responses [[4-7]]. The concept of T-cell help for CD8+ T-cell responses was further supported by the finding that chemokines secreted by activated CD4+ T helper cells can play a key role in the recruitment of naïve antigen-specific CD8+ T cells to antigen-bearing antigen presenting cells (APCs) in secondary lymphoid organs [] or to sites of infection []. Moreover, in some experimental settings CD4+ T cells were proposed to directly interact with CD8+ T cells, thereby promoting their activation and expansion []. Hence, the overall contribution, as well as the mechanism, of helper CD4+ T cells in the initiation of primary and secondary CD8+ T-cell responses is more complex than initially believed and appears to strongly depend on the specific biology of the immunization/infection system.
Primary CD8+ T-cell responses — do they require T-cell help?
CD8+ T-cell priming with noninflammatory agents — T-cell help is necessary
The initial evidence that T helper cells condition the ability of DCs to prime CD8+ T-cell responses was provided by Bennett et al.,  showing that priming of ovalbumin-specific CD8+ T cells requires that both CD4+ and CD8+ T-cell subsets recognize their antigen on the same DC (cognate T-cell help). In accordance with this finding, several subsequent studies showed that after in vivo priming with noninfectious agents (such as minor histocompatibility antigens, tumor antigens or protein antigens), CD4+ T-cell help is essential for the stimulation of a measurable primary CD8+ T-cell response [[12-14]]. In these settings, T-cell help is thought to mediate the activation of APCs via a mechanism that involves CD40/CD40L interaction between CD4+ T cells and APCs, a process which is referred to as DC “licensing”, Hence, it was believed that, exclusively, immunizations with noninflammatory agents require T-cell help due to a lack of “danger signals,” which in turn would promote activation of DCs and thereby replace the need for T-cell help (Table 1).
Table 1. T-cell help requirement during primary CD8+ T-cell responses
CD8+ T-cell responses upon infection: differential requirement of T-cell help
In accordance with the “licensing model,” many pathogenic infections (such as lymphocytic choriomeningitis virus (LCMV), VSV, Ectromelia virus, and HIV) induce strong CD8+ T-cell responses in the absence of T-cell help (Table 1) [[4, 33, 34]], most likely due to their ability to directly activate APCs via pattern recognition receptors (PRRs) []. Considering that CD4+ T cells modulate various aspects of the CD8+ T-cell response, this simplified model was challenged by the observation that primary CD8+ T-cell responses to several pathogens such as adenovirus [], influenza virus [], herpes simplex virus (HSV) [[22, 23]], and vaccinia virus [[26, 27]] were compromised in the absence of T-cell help. These findings raised the question of why certain pathogens differ from others in their ability to generate CD4+ T-cell help-independent CD8+ T-cell responses; possible explanations will be provided in the following section “What renders certain infections T-cell help dependent?” However, there are even reports using the same infection model documenting discrepant results on the CD4+ T-cell dependence of CD8+ T-cell responses. For instance, primary CD8+ T-cell responses were shown in some reports to depend on CD4+ T cells during infection with vaccinia virus [[26, 27]], while other reports did not find such dependence [[28, 29]]. When carefully comparing the experimental conditions used in these studies, apparent differences included: (i) the dose of the virus inoculum (with higher doses leading to T-cell help-independent CD8+ T-cell responses), (ii) the use of different vaccinia virus recombinants which might vary in their virulence, and (iii) the concomitant transfer of virus-specific, TCR-transgenic CD8+ T cells, thereby increasing the precursor frequency of the responding CD8+ T cells. Indeed, higher-than-physiological precursor frequencies of responding CD8+ T cells abolish the CD4+ T-cell dependence during vaccinia virus infection (unpublished data MW and AO), most likely due to the CD8+ T cell-mediated maturation of DCs [[36, 37]].
In some reports CD4+ T cells (or CD4+ Treg cells) were also shown to influence the immunodominance of CD8+ T-cell responses, such as during DNA immunization or RSV infection [[38, 39]]. In contrast, the absence of CD4+ T cells did not affect the CD8+ T-cell response hierarchy during influenza virus infection [].
Besides affecting the size of the CD8+ T-cell response, CD4+ T cells have also been implicated in shaping the phenotypic and functional properties of CD8+ T cells. The absence of CD4+ T-cells during infection with Listeria monocytogenes resulted in impaired effector memory (CD127+ CD62L−) CD8+ T-cell differentiation [] and the absence of CD4+ T cells during LCMV infection prevented the development of central memory (CD44+ CD62L+) CD8+ T cells []. However, whether such phenotypic alterations are directly inferred by the absence of CD4+ T cells is often unclear, since it should be kept in mind that studying CD8+ T-cell responses in the absence of T-cell help might be problematic in some instances (in particular in the context of replicating infections), where CD4+ T cells might be critically involved in controlling pathogen levels and hence antigen load. It is well known that the level and duration of exposure to antigen critically influences the phenotype and functionality of CD8+ T cells, with longer antigen exposure and higher levels of antigen favoring effector cell differentiation at the expense of memory CD8+ T-cell differentiation [[43, 44]].
In this context it should also be considered that different CD4+ T-cell-deficient models are used to study the requirement of T-cell help, such as CD4– or MHC class II-deficient mice or active depletion of CD4+ T cells using a specific antibody. The caveat of the latter approach is that besides T helper cells, T regulatory (Treg) cells are also depleted and hence it might be difficult to dissect the contributions of classical T-cell help from those of Treg cells in shaping CD8+ T-cell responses.
What renders certain infections T-cell help dependent?
As mentioned earlier, it is conceivable that PRR ligands of microbial pathogens directly activate DCs and thereby might compensate for the requirement of T-cell help []. However, as all viral or bacterial pathogens bear PRR ligands, such as LPS, CpG DNA, dsRNA, ssRNA, lipoproteins, flagellin, etc. that can trigger inflammatory responses and thereby mediate the activation of DCs, it remains unclear which PRR–PAMP (where PAMP is pathogen-associated molecular pattern) interactions render microbial infections T-cell help dependent or independent.
There is extensive evidence that infectious agents have developed specific evasion strategies to downregulate inflammation and/or costimulatory molecules, which might be linked to their T-cell help dependence. In support of this notion it was shown that vaccinia virus encodes for soluble receptors, which decrease Type-I interferon (IFN) production and thereby limit inflammation [[46, 47]]. Similarly, HSV has been described to modulate the level of costimulatory molecules expressed on both T cells and DCs []. Moreover, influenza virus, in contrast to HSV, is capable of eliciting CD4+ T-cell help-independent CD8+ T-cell priming, presumably due to its ability to upregulate CD40L on DCs. Furthermore, the pathogen-induced inflammatory milieu may also account for the ability to instruct T-cell help-independent CD8+ T-cell responses. We have previously shown that upon infection with vaccinia virus, IL-12 is induced in a T-cell help-dependent manner [], whereas certain other virus infections (e.g., LCMV) induce potent Type-I IFN responses at the expense of IL-12 production []. In contrast to IL-12 induction after vaccinia virus infection, Type-I IFN production after LCMV infection is T-cell help-independent and is therefore a candidate molecule driving T-cell help-independent CD8+ T-cell responses. In support of this hypothesis, we could show that differences in the T-cell help-dependence between various infections is chiefly influenced by the ability of a specific infectious agent to stimulate an early robust production of Type-I IFN []. There is also extensive evidence that signals via the IL-12 receptor or Type-I IFN receptor initiate a differentiation program which involves increased expression of numerous genes that encode for proteins important for clonal expansion and survival of both effector and memory cells [[51, 52]]. However, all of these studies cannot exclude a synergistic effect between direct Type-I IFN signaling on CD8+ T cells and additional signals provided by other cell types, as it has been reported that the immune stimulatory activity of Type-I IFN results at least partly from its ability to induce DC maturation [].
In conclusion, the current data suggest that in infections/ immunizations, which lead to robust CD4+ T cell-independent Type-I IFN production, CD4+ T-cell help is not required for primary CD8+ T-cell responses, as long as APC maturation is provided by the infecting agent or an adjuvant. In the case of infections/immunizations, which are associated with a predominant IL-12 response, the CD4+ T-cell dependence of primary CD8+ T-cell responses may relate to whether sufficient IL-12 production by the priming APC requires engagement with CD4+ T cells. Finally, in the case of “weak” immunogens, which do not by themselves promote the maturation of DCs, CD4+ T cells are required not only to induce inflammatory third signals for CD8+ T-cell activation but also to induce DC maturation (Fig. 1).
Functional memory CD8+ T cells in the absence of T-cell help?
The concept of T-cell help for CD8+ T-cell responses described earlier was profoundly expanded with the finding that T-cell help is required for the generation of long-lived, functional memory CD8+ T cells that can respond quickly to re-challenge with a pathogen [[28, 54, 55]]. In particular, studies using noninflammatory, cellular antigens showed that early primary CD8+ T-cell responses can in fact be T-cell help-independent—even in these noninflammatory conditions. In the absence of T-cell help during the first 3–4 days, functional effector CD8+ T cells were generated with respect to their ability to produce IFN-γ as well as IL-2, but they were unable to mount productive recall responses [[10, 56]]. Thus, although potent primary CD8+ T-cell responses can be induced in the absence of T-cell help in many viral or bacterial infections, it became clear the generation of proliferation-competent memory CD8+ T cells as well as their long-term maintenance is in many experimental systems dependent on CD4+ T-cell help (Table 2 and 3) [[28, 54, 56]]. Although the phenomenon of poor secondary expansion of “helpless” CD8+ T cells held true for many in vivo experimental systems [], there are also reports demonstrating that “helpless” CD8+ T cells are not necessarily impaired in their recall proliferation potential [[26, 30, 57]].
Table 2. T-cell help requirement during the priming phase for the generation of functional memory CD8+ T cells
Priming and challenge performed in T-cell help-deficient mice.
IL-2 could restore the proliferation defect in vitro
The intrinsic molecular program that instructs the recall proliferation defect of unhelped memory CD8+ T cells remains incompletely understood and several mechanistic pathways have been proposed. It was shown that elevated levels of T-bet in “helpless” LCMV-specific CD8+ T cells repress the transcription of IL-7Rα and thereby drive the differentiation of effector memory CD8+ T cells at the expense of central memory CD8+ T cells. Interestingly, deletion of T-bet restores the pool of central memory CD8+ T cells as well as their functional properties []. In addition, there is evidence that increased levels of TRAIL mRNA found in “helpless” memory CD8+ T cells account for their defective secondary expansion []. This finding was challenged by other studies showing that TRAIL deficiency is insufficient to overcome the defective functionality of “helpless” memory CD8+ T cells [[60, 61]], indicating that increased TRAIL expression in “helpless” CD8+ T cells does not fully account for their impaired phenotype and function.
As there is no consensus on a strict T-cell help-dependent programming of proliferation-competent memory CD8+ T cells, it is likely that inherent differences in the experimental models account for the different outcomes. Thus, it is important to assess the T-cell help-dependence of (memory) CD8+ T-cell responses and the underlying mechanisms closely linked to the particular experimental system used.
Programming versus maintenance — what is the timing of T-cell help?
Based on the observation that T-cell help is critical for the functionality of memory CD8+ T cells, which are generated in response to many infections or immunizations, the exact timing that is involved in delivering help to CD8+ T cells is still controversial. Currently, there are two different models (programming versus maintenance) discussed. The first model proposes “imprinting” or programming of memory CD8+ T cells by CD4+ T cells during the first few days of initial activation [[56, 59]]. This model claims that a brief period of antigen stimulation in presence of CD4+ T cells is necessary to cause naïve CD8+ T cells to differentiate into effector cells that subsequently develop into long-lived protective memory CD8+ T cells. The second model suggests that the maintenance of CD8+ memory T cells requires continuous exposure to bystander CD4+ T cells far beyond the priming phase []. Instead of programming, CD4+ T cells seem to be required for the subsequent survival and maintenance of functional memory CD8+ T cells. The involvement of T-cell help in this system seems to be antigen nonspecific, however whether CD4+ T cells themselves are responsible for the production of factors necessary for the maintenance of memory CD8+ T cells, or other cells get instructed by CD4+ T cells to provide certain signals, needs to be further investigated []. A recent finding also points to a role for CD4+ T-cell help during the challenge phase []. Thus, it is likely that the nature of the challenge infection/immunization might be a crucial parameter in determining the T-cell help dependence of memory CD8+ T cells, a notion which we think should be carefully taken into consideration when comparing results from different experimental setups.
IL-2 – the key factor of T-cell help?
An important feature of T helper cells is the production of IL-2, since it was shown in various experimental settings that CD4+ T-cell derived IL-2 is a crucial mediator of T-cell help [[26, 65]]. Lately, there is also growing interest in the role of IL-2 in the differentiation process of CD8+ T cells in T-cell help-independent experimental settings. Although IL-2R deficient CD8+ T cells show only a modest impairment in their ability to make robust primary response upon infection with LCMV, IL-2 signaling during the priming seems to be required for the ability of the ensuing CD8+ memory cells to mount optimal secondary responses [[66, 67]]. More recent data further clarified these findings, showing that an early transient heterogeneity of CD25 expression on LCMV-specific CD8+ T cells directs them into different developmental programs, with increased CD25 expression, and hence increased sensitivity towards IL-2 signals, favoring effector cell differentiation at the expense of memory cell differentiation [[68, 69]]. Thus, although it remains unclear whether CD4+ T cells are the critical source of IL-2 in this process, these studies clearly indicate that the magnitude and duration of IL-2 signals can have a striking influence early on in CD8+ T-cell differentiation. In contrast to the data obtained with LCMV infection, the recall capacity of L. monocytogenes-specific memory cells was found to be independent of IL-2 signaling []. It should be mentioned that besides T-helper cells, DCs, and CD8+ T cells are also capable in producing IL-2. A recent study identified memory CD8+ T cells themselves as a crucial source for IL-2, with CD4+ T-cell help being important to facilitate the CD8+ T-cell autocrine IL-2 production []. Again, besides the overall requirement of T-cell help the underlying mechanism also seems to be defined by the nature of the stimuli.
Chronic viral infections — T-cell help is appreciated
In the context of chronic antigen exposure (such as in chronic viral infections), the presence of CD4+ T cells during the priming phase seems to play a critical role for the maintenance and functionality of CD8+ T-cell responses and recent reports indicate a pivotal role of IL-2 and IL-21 in this process [[66, 72-75]]. In contrast to acute/resolved infections where memory CD8+ T-cell maintenance is antigen-independent but dependent on the homeostatic cytokines IL-7 and IL-15 [[76-78]], the maintenance of CD8+ T cells during actively replicating chronic infections is strictly dependent on antigen presence and increased cell turnover [[79, 80]], which seems to be supported by T-helper cells and in particular by IL-21 secreted by CD4+ T cells in the context of chronic antigen encounter [[72-74, 81, 82]]. A recent report focusing on the requirement of CD4+ T-cell help during the memory recall response in the context of high antigen load and antigen persistence found memory LCMV-specific CD8+ T-cell responses more reliant on CD4+ T-cell help than naïve virus-specific CD8+ T-cell responses [].
In the case of persistent latent infections, such as CMV, which are associated with much lower antigen loads compared with those of actively replicating persistent viral infections, CD4+ T cells were shown to shape the virus-specific CD8+ T-cell responses. Murine CMV (MCMV) infection induces two distinct patterns of CD8+ T-cell responses. While CD8+ T cells with some specificities follow the classical expansion–contraction–memory kinetics usually observed during acute and resolved infections, CD8+ T cells with other specificities continue to expand and plateau at high frequencies which are maintained in an effector memory state during the entire course of the infection, collectively referred to as “memory CD8+ T-cell inflation” [[6, 84]]. Memory inflation is driven by recurrent exposure to CMV antigens [] and CD4+ T cells were essential in facilitating memory CD8+ T-cell inflation, which was likely mediated by their provision of IL-2 [[75, 86]].
In particular, during chronic viral infection CD4+ T cells might also influence CD8+ T-cell responses indirectly by altering the level of antigen load, which is perceived by the specific CD8+ T cells. In this line of reasoning, T-helper cells may influence CD8+ T-cell responses indirectly by interacting with B cells, thereby modulating virus-specific antibody production. Indeed, CD4+ T cells were recently shown to preferentially differentiate into follicular T helper (Tfh) cells in the context of chronic LCMV infection, thereby promoting LCMV-specific humoral immunity which resulted in eventual control of the infection []. Concomitant to the reduction in virus load, CD8+ T-cell responses were improved with respect to numbers and function []. Also during chronic LCMV infection, IL-6 has recently been identified to be a key molecule acting on CD4+ T cells during late stages of chronic infection []. Signals via the IL-6 receptor on CD4+ T cells drove their differentiation into Tfh cells in a BCL-6 dependent manner. Furthermore, increased numbers of Tfh cells were essential for germinal center formation, LCMV-specific antibody production and subsequent viral control.
Another CD4+ T-cell subset, which gains more and more interest in the context of chronic antigen exposure is the Treg cell subset. In particular, the ability of viruses to induce Treg cells, which subsequently suppress effector CD8+ T-cell responses appears to be a crucial viral escape mechanism [[89, 90]]. It was shown experimentally, that transient depletion of Treg cells during chronic Friend retrovirus infection is sufficient to reinvigorate virus-specific CD8+ T-cell responses, thereby decreasing virus load []. For more detailed information on the role of Treg cells in the context of host-microorganism interactions we would like to refer to an excellent review by Belkaid and Tarbell [].
Due to the complexity and the differences among the diverse immunization/infection models with respect to the antigen amounts, the nature of the inflammatory response present during the priming process of CD8+ T cells, the ability of the pathogen or adjuvant to induce DC maturation and the precursor frequencies of the responding CD8+ T cells, there are still unresolved controversies concerning the overall requirement of T-cell help, including the time points and mechanisms that are implicated in the delivery of help for CD8+ T-cell responses. Hence, further studies are needed focusing in particular on the molecular differences between helped and “helpless” memory CD8+ T cells, as well as on the mechanisms employed by CD4+ T cells to impact on the generation of potent effector CD8+ T cells and proliferation-competent memory CD8+ T cells, in the context of defined experimental models. In the future, such comparative studies are likely to reveal “public” and “private” patterns of the T-cell help (in-)dependence of CD8+ T-cell responses, which will be instrumental in tailoring T-cell based vaccines.