A growing body of evidence suggests that several members of the complement family interact with CD4+ T cells through interaction with cell surface and intracellular receptors modulating activation and subset differentiation.
T cell activation and differentiation
Activation by CD46 has been shown to profoundly affect T cell activation. Co-stimulation by CD3/CD46 transduces signals resulting in a strong proliferative response of activated T cells [
46‐
48], and differentiation into Th1 cells characterized by IFNγ production [
49]. As IL-2 accumulates in the surroundings of the cells, a contraction phase occurs, with decreased IFNγ production but enhanced secretion of IL-10. The release of high levels of IL-10 and low IFNγ leads to suppression of proliferation of bystander T cells; hence, CD46-costimulated T cells acquire a phenotype of so-called type I regulatory T cells (Tr1) [
50]. CD46 co-stimulation modulates expression of Notch family members at the surface of activated T cells, while on resting T cells, CD46 binds to the Notch family member Jagged1, inhibiting both Notch signaling and CD46 activation. Activation of CD46 leads to CD46 downregulation, freeing Jagged 1 to interact with Notch [
51]. Interestingly, Notch regulates IL-10 secretion in Th1 cells [
52]. Hence, CD46 may provide the link between the Notch and complement cascades to regulate T cell differentiation and IL-10 production.
Enzymatic processing of CD46 occurs upon T cell activation, leading to the shedding of its extracellular domain and subsequent cleavage of its two cytoplasmic tails, Cyt1 and Cyt2. The two tails exhibit antagonistic roles in the control of inflammation, in both huCD46-transgenic mice (mice only express CD46 on testis) [
53] and in human T cells in vitro [
54,
55]. Processing of CD46 is required for IL-10 production and allows T cell activation but also T cell termination [
54,
55]. Both tails contain a nuclear localization signal (NLS) sequence, and indeed, cleaved tails translocate to the nucleus likely controlling target genes [
56]. CD46 expression on activated T cells is tightly controlled [
57‐
60]. Vitamin D, for example, which is known to promote immune regulation, triggers the Th1 to Tr1 switch and favors CD46 cleavage [
60]. Expression of CD46 on activated T cells is also under control of a crosstalk with CD28 [
61]. Regulation of the splicing of the Cyt1 exon have been recently reported [
62]. A few interactors of CD46 tails (albeit only with Cyt1) have been identified, leading to the discovery of the role of CD46 in controlling autophagy (via interaction with GOPC) [
63] and cell polarity via binding to DLG4 [
64,
65]. Cyt1 interacts with the kinase SPAK, and this is required for IL-10 production [
49]. CD46 also interacts with alpha-E-catenin in human CD4+ T cells and knockdown of α-E-catenin impaired CD46 downregulation suggesting a role for α-E-catenin in CD46-controlled expression [
51]. CD46 is recruited to lipid rafts [
66] and T cell polarization is affected by CD46 ligation [
47,
65]. CD46 ligation before specific T cell activation can prevent subsequent TCR signaling by recruiting the lipid rafts and preventing the immune synapse formation [
65], while co-ligation with the TCR promotes its recruitment to the immune synapse (S. Ni Choileain, J. Hay, et al., submitted). Together, these studies highlight the key role of CD46 signaling on controlling T cell differentiation and function.
Other members of the RCA family also control T cell activation. Similar to CD46, CD55 affects co-stimulation of human T cells by binding to CD97, promoting T cell activation and Tr1 cell differentiation [
67]. In mice, it was observed that the lack of CD55 promotes T cell activation, suggesting an inhibitory role of CD55 in T cell effector function [
68,
69]. A role for CD59 in T cell co-stimulation has also been reported [
70] and shown to couple signaling events to the TCR [
71]. Moreover, recent studies have highlighted an immunomodulatory role of CD59 in T cells suggesting that its ligation may also potentially induce a regulatory T cell subset [
72].
Not only complement regulators but also complement receptors modulate T cell function and survival. In mice, C5aR and C3aR act as costimulatory signals for T cells and sustain naive T cell survival [
30]. C3aR is not expressed at the cell surface of naïve T cells but on lysosomes and is transported to the surface upon T cell activation [
73]. Defective cytokine secretion by T cells from
C3ar1
−/− and
C5ar1
−/− mice show that these receptors are also needed for effector function [
30]. In addition, these complement receptors control Treg function. nTregs express C3aR and C5aR, and triggering of these receptors inhibits Treg suppressive function by modulating Foxp3 expression [
74].
C3aR
−/−
and
C5aR
−/−
mice have increased levels of Foxp3 + Tregs [
75]. Similarly, blocking these receptors with C3aR and C5aR antagonists in human T cells induced suppressive human iTregs [
75].
Production of local and intracellular C3 and C5 in T cells
Until recently, it was thought that soluble complement components were mostly present in the plasma after production mainly by the liver. In recent years, a series of papers have demonstrated the local production of C3 and C5 fragments by adaptive immune cells. In mice, T cell-derived C3 activation is due to the formation of C3 convertase [
30]. In contrast, in resting human T cells, a recent report suggests that there is continuous production of C3 that is cleaved by cathepsin L (CTSL), giving rise to C3a and C3b [
3]. The resulting C3a binds to its C3aR at the surface of lysosomes and participates to T cell survival by regulation of mechanistic target of rapamycin (mTOR) activity [
3]. Upon TCR activation, both C3a and C3b are exported to the cell surface within minutes and bind to surface C3aR, also expressed at the surface after activation, and CD46, respectively, leading to T cell activation and Th1 differentiation. Uncontrolled cleavage results in hyperactivation of Th1 responses, such as that observed in a small number of patients with rheumatoid arthritis, and indeed, inhibition of CTSL normalized the response [
3].
Interestingly, CTSL has been linked to inflammatory responses and IL-17 production by controlling Th17 differentiation in both humans and mice [
76,
77], suggesting a broader role of cathepsins in regulating inflammatory conditions. This also suggests a role of the C3 and C5 fragments in modulating Th17 response.
The same group recently reported the intracellular cleavage of C5 in T cells upon co-stimulation by CD3/CD46 [
12]. Binding of cleaved C5a to intracellular C5R1 leads to activation of the NLRP3 inflammasome in T cells, and IL-1β production has been reported [
12]. The NLRP3-IL-1β axis participates in the activation of CD46 in T cells and IFNγ production. On the other hand, binding of C5a to C5aR2 that is expressed at the cell surface inhibits this pathway [
12]. Of note, we have failed to detect any IL-1β secreted upon CD46 costimulation of T cells from either healthy donors or patients with multiple sclerosis (A. Itchers, J. Killick, A. Astier, unpublished data), and the reasons for these discrepancies are unknown and may relate to the sensitivity of the assays used. Further research will allow clarification on the role of this pathway in T cell differentiation.
It is important to note that key differences between men and mice have been reported. Local production of C3 and C5 fragments has been demonstrated in mice, although this was thought to be due to increased C3 expression and extracellular cleavage by the C3 convertase [
78]. Moreover, although CTSL and C3 are present in murine T cells, normal Th1 differentiation occurred in T cells from CTSL knockout mice, suggesting that C3 fragments do not require CTSL processing in murine T cells, as opposed to human cells [
3]. Moreover, mice do not express CD46 (except for testis), implying that the interaction between C3b and CD46 is not required to generate a Th1 response. It would be interesting to assess the role of the C3b receptor and complement regulator Crry in this response, especially as Crry has been also shown to act as a costimulatory molecule for murine T cells [
79], similarly to CD46 engagement on human T cells [
46].
Importantly, although the role of intracellular complement activation has so far been only shown in human CD4 T cells, intracellular complement activation has been observed in several cell types by the authors, suggesting that this is a general biologic process necessary to overall cell function [
3], and the term “complosome” has recently been suggested [
80].