Differential ligand binding/trafficking for distinct CTLA-4 fates: is it an expandable mechanism?

In a recent issue of Nature Immunology, Kennedy et al. reported that CTLA-4 (CD152) differentially targets two distinct ligands, CD80 and CD86, on antigen-presenting cells (APCs) for destruction or recycling following transendocytosis (TE), resulting in separate fates for CTLA-4 on T_reg cells and thereby regulating autoimmunity. Mechanistically, whereas CTLA-4 tightly binds CD80 for TE and causes ubiquitination and trafficking to late endosomes and lysosomes for degradation, CTLA-4 weakly binds CD86 for TE, resulting in dissociation under acidic conditions (pH < 6) and recycling of CTLA-4 to the plasma membrane for further CD86 removal. Therefore, CD86 targeting by CTLA-4 regulates T-cell function in autoimmunity [1] and presumably immune surveillance of cancer cells (Fig. 1).

Fig. 1.

The different fates of transendocytosed CTLA-4 are determined via distinct binding to the ligand CD80 or CD86 and subcellular trafficking. CTLA-4 binding to dimeric CD80 or monomeric CD86 causes transendocytosis of immune-inhibitory CTLA-4, leading to a mechanism that depletes CD80 and CD86 from the surfaces of antigen-expressing cells, such as antigen-presenting cells (APCs). Following transendocytosis, CD80–CTLA-4 binding is insensitive to the acidic pH in the early endosome, whereas CD86–CTLA-4 binding is sensitive enough for dissociation. This distinct basis allows the sorting of CTLA-4 to either degradative (CD80-bound) organelles (late endosomes/lysosomes) or recycling (CD86-bound) organelles (endosomes and plasma membranes). Through such sorting, endocytosed CTLA-4 is recycled to the cell surface, a process that occurs by virtue of the vesicle chaperone endosomal lipopolysaccharide-responsive and beige-like anchor protein (LRBA) and the GTPase Rab11, whereas CTLA-4 targets tetraspanin Lamp3⁺ and GTPase Rab7⁺ late endosomes and later lysosomes for the degradation of CTLA-4. Other molecules or ligands, such as cell‒cell adhesion molecules, on antigen-expressing cells may have adapted to undergo distinct mechanisms such as cis- or trans-binding and trafficking to different subcellular compartments during the regulation of diverse immune responses, although such expandable possibilities need more aggressive exploration

Autoimmunity and cancer progression occur as immune cells lose their ability to distinguish healthy self-cells or control the outgrowth of malignant cells. Immune dysfunction results from limited antigen recognition, immune suppression, or functional remodeling of immune cells [2]. The inflammatory properties of autoimmune diseases are similar to those of early-stage cancer but different from those of advanced cancer. Therefore, the incidence of various cancers is elevated in patients with preexisting autoimmune diseases [3].

T-cell–mediated immunity can be regulated by the CD28–CTLA-4 system, which comprises a network of positive and negative costimulatory signals, the integration of which modulates T-cell responses such as activation, proliferation, and cytotoxicity [4]. CTLA-4 mediates the inhibition of T-cell responses by competing with the costimulatory receptor CD28 for CD80 and CD86, depleting the ligands from the surfaces of cells such as APCs and dendritic cells (DCs) through TE, and delivering inhibitory signals via its intracellular tail [4]. Although the roles of CD80 and CD86 are considered redundant, Kennedy et al. revealed a substantial difference in the fate of CTLA-4. CTLA-4 binds strongly to dimeric CD80 independent of intracellular pH, whereas it binds weakly to monomeric CD86 in a pH-sensitive manner. This results in differential subcellular trafficking either to late endosomes/lysosomes (degradative fate) or to the plasma membrane (recycling fate). Recycled CTLA-4 captures and removes additional CD86 on APCs to establish immune tolerance.

Different induction kinetics and biophysical properties and distinct expression of CD80 and CD86 on immune cells, such as monocytes, memory B cells, DCs, and T and T_reg cells, are known to exist [5]. Kennedy et al. observed differential intracellular trafficking of CTLA-4–CD80 or CTLA-4–CD86 using fluorescent protein–tagged CD80 or CD86 in human T_reg cells or other human cell lines and an in vitro TE assay. Combined microscopy and immunoblot analysis revealed that CTLA-4 binds CD80, undergoes TE, and is ubiquitinated and trafficked toward late endosomes with Lamp3/CD63 and lysosomes for degradation. In contrast, CTLA-4 binds CD86, undergoes TE, dissociates from CD86 in a pH-sensitive manner in acidic endosomes and is recycled to the cell surface by endosomal LRBA. These opposing, ligand-dependent CTLA-4 fates involve different binding affinities, pH sensitivities, and sorting toward late endosomes with Lamp3/CD63 and Rab11 or toward endosomes with LRBA. Recycling of CTLA-4 released from the CTLA-4–CD86 complex leads to increased CTLA-4 bioavailability to deplete ligands for inhibitory T_reg cell responses. Mutations in lysine residues in CTLA-4 that abolish ubiquitination and neutralization of intracellular pH using NH₄Cl increase ligand levels on APCs. Defects in T_reg cell function are linked to disruption of CD86-associated CTLA-4 recycling, regardless of normal CD80 activity, which can lead to failure of self-regulatory tolerance. Importantly, CTLA4 R70Q or R70W mutations in patients with autoimmune diseases result in selective loss of the ability to bind and promote CD86 TE, generating populations of APCs with robust T-cell proliferation, although binding to CD80 and downstream activity are retained. The significance of autoimmune pathologies that interfere with CD86 binding and TE in patients with CTLA-4 mutations is illustrated by impaired CTLA-4 recycling in patients with LRBA mutations linked to autoimmunity, lymphoproliferation, and humoral immune deficiency [6]. Although CTLA-4 is inducible on activated T cells, constitutive CTLA-4 expression on T_reg cells counteracts aberrant T-cell responses against self-proteins and sustains lymphocyte homeostasis and immune tolerance [7]. Such mechanisms of ligand binding/TE-mediated regulation of CTLA-4 bioavailability on the T_reg cell surface can facilitate the design of anticancer immunotherapies that function via selective depletion of T_reg cells using antibodies such as ipilimumab or tremelimumab and their pH-selective variants [8].

Despite these findings, questions remain. How and why is CD80, but not CD86, coupled to CTLA-4 ubiquitination and degradation? How is endocytosed ligand–CTLA-4 differentially sorted to Lamp3/Rab7⁺ late endosomes/lysosomes or LRBA⁺ endosomes? Can the distinct pH sensitivity explain all specific and differential CTLA-4 outcomes? Although the mechanisms described by Kennedy et al. broaden our understanding of the inhibitory signaling dynamics of immune checkpoints in autoimmunity and presumably cancer, the mechanistic characteristics of differential binding/trafficking leading to distinct fates of immune checkpoints may be generally applicable to more ligands, receptors, and membrane proteins. CD80 and CD86 are members of the immunoglobulin superfamily and play roles in cell‒cell recognition and adhesion. Other membrane receptors or proteins expressed on ligand-expressing cells (such as APCs, DCs, or cancer cells) or immune cells may also regulate binding, cis- or transendocytosis, and intracellular trafficking, leading to immune-related surface avidity and/or signaling activity (Fig. 1). Furthermore, crosstalk via cis-binding between CD80 and PD-L1 on APCs results in inhibition of PD-L1 and the immune regulatory functions of PD-1, leading to impaired PD-1-mediated T-cell suppression, although this cis-PD-L1/CD80 interaction does not interfere with CD80/CD28 or CD80/CTLA-4 binding [9]. It is likely that the effects of signaling involving CTLA-4/CD28 systems are more complex due to the diverse associations between components or neighbors.

Cell‒cell recognition involves adhesion via integrins, cadherin, immunoglobulin superfamily, syndecan, and tetraspanin, among others [10]. Immune-related binding of ligands to receptors and antigen presentation occur through cell‒cell adhesion via the engagement of membrane proteins from each side of the membrane. Although consensus sequences or conserved modules in molecules that bind immune checkpoints or regulate immune responses can be difficult to identify, analyses of evolutionary aspects of protein families, such as cell‒cell adhesion molecules, may reveal other linkages or relationships. We found that other membrane proteins that function as cell‒cell adhesion molecules, such as tetraspanins, can also bind immune checkpoints to modulate their surface avidity. Furthermore, binding between these molecules was observed in cell extracts with specific detergent-containing lysis buffers, indicating that the binding complexes are uniquely localized in intracellular compartments. Thus, differential binding between immune ligands and receptors, in addition to distinct subcellular trafficking patterns, appears to play crucial roles in mediating complex immune responses under diverse immunogenic challenges.

The results reported by Kennedy et al. demonstrate that CTLA-4 binds CD80 and CD86 with different affinities, pH sensitivities, and levels of subcellular trafficking, leading to distinct fates for CTLA-4, which affect the signaling required for immune tolerance.

Competing interests

The authors declare no competing interests.