The CD28 molecule was described by Hansen in 1980 as a T-cell surface antigen. A study from 2005 proved that the extracellular fragment of CD28 shows a structure characteristic of the immunoglobulin superfamily [27]. CD28 is a 44 kDa glycoprotein with a homodimer structure. The structure is composed of a V-shaped extracellular domain with disulfide bonds and a MYPPPY motif, which is required for B7 ligand binding, as well as a membrane part and a cytoplasmic domain composed of 41 amino acids [28]. The gene encoding the molecule is 300 kb in size and is located on chromosome 2 in the region of the q33 strand [29]. The molecule is constitutively present in 30–50% of human CD8+ T cells and 95–100% of CD4+ T cells, with 60,000 molecules per cell [30]. CD28 expression increases after binding to the TCR/CD3 complex. When B7 is bound to CD28 during T-cell activation, the mRNA level of the CD28 molecule is decreased, as is surface expression. This inhibitory regulation is transient and lasts up to 48 h. It prevents the restimulation of T cells, which affects the extent and duration of the immune response [31].

CD28 plays a role in various T-cell activities, such as reorganizing the cell structure, generating signaling molecules like cytokines and chemokines, and facilitating internal biochemical processes like phosphorylation, transcriptional signaling, and metabolism. These processes are crucial for T-cell growth and specialization. Activation of CD28 receptors induces changes in T cells at epigenetic, transcriptional, and post-translational levels. One notable effect of CD28 co-stimulation is its control over several functions in T cells, including activating cytokine genes. One specific gene encodes IL-2, a cytokine that impacts T-cell survival, growth, and differentiation. When CD28 co-stimulation is absent, IL-2 production decreases, making T cells unresponsive to stimulation [32]. Additionally, CD28 activation leads to a type of protein modification called arginine methylation in multiple proteins [24].

The formation of an immune synapse initiates the activation of lymphocytes. Within it, there is a concentration of TCR receptors, and their connection to MHC complexes is made possible by the interaction of adhesion molecules. It is now known that the signal from CD28 enhances the initial fusion of T cells and APCs. The CD28 molecule participates, thus, in cellular adhesion, proving its high significance for the early stages of activation. By regulating IL-2 expression, CD28 indirectly affects T cell proliferation [33].

Signaling from the TCR complex does not always stimulate cell activation but can redirect the cell into a state of anergy or even programmed cell death. For apoptosis to occur, a large amount of antigen on the T cell and the absence of pro-inflammatory factors, including anti-apoptotic cytokines (IL-2, IL-7, and IL-15), are required [34]. Apoptosis induced by the interactions between Fas (CD95) and its ligand FasL (CD95L) is crucial for the elimination of T cells in the final phase of the immune response. This type of apoptosis is called activation-induced cell death (AICD). The CD28 molecule protects the cell from Fas-induced apoptosis as the signaling pathway involving the kinases Akt and PI3K (phosphoinositide 3-kinase) is activated. Activated Akt (also known as protein kinase B) inhibits the progression of apoptosis by suppressing the recruitment of caspase 8, which is crucial in cell death [35]. CD28 also promotes lymphocyte survival by upregulating Bcl-xL expression and reducing cell membrane permeability [36].

Lymphocyte maturation is a several-step process, with only 10% of properly selected cells reaching the periphery. Generally, the maturation process can be distinguished by an early phase, positive and negative selection [37]. It has been proven that the CD28 molecule is important during positive selection when double-positive lymphocytes have CD4 and CD8 antigens. This selection verifies whether rearrangements of genes encoding α and β subunits have resulted in the emergence of TCRs capable of recognizing cellular antigens. In the process of positive selection, the CD28 molecule plays a significant role. It has been shown in an experimental model that the absence of CD28/B7 costimulation leads to increased selection in the thymus. This is because the CD28 molecule inhibits the differentiation of mature single-positive T cells by suppressing their selection during differentiation in the thymus [38]. In contrast, negative selection (clonal deletion) leads to the removal of cells that recognize their own MHC class antigens with increased affinity, which could result in autoreactivity [38]. Two distinct pathways are likely involved in negative selection in vitro. The first is CD28-dependent for interactions with low binding capacity, and the second is CD28-independent for interactions with high affinity [39]. Thymocytes primarily need signals that inhibit their apoptosis to survive. Signaling from the CD28 molecule enhances Bcl-XL expression, which in turn prevents Fas-induced cell death [38]. Despite this thorough double-selection process, not all autoreactive lymphocytes are eliminated, as not all existing autoantigens will be presented to lymphocytes in the thymus. Therefore, mechanisms that determine peripheral tolerance are necessary. For example, T-cell activity can be inhibited by increased expression of the CTLA-4 antigen, which transmits an inhibitory signal [40].

In humans, the ICOS gene is located on chromosome 2q33.2, contains 199 amino acids, and encodes for a protein known as an immune checkpoint protein. It was the third CD28 family member to be identified42. Unlike other family members, ICOS is expressed in already activated T cells and peripheral tissues [41].

In contrast to CD28, the expression of ICOS displays greater variability. ICOS expression is dependent on the activation of T cells through the TCR/CD3 complex. Following activation, ICOS expression remains present on recently activated CD4+ T cells, as well as on memory Th1 and Th2 cells. Besides the TCR signal, the cytokines IL-12 and IL-23 boost ICOS expression in T cells [42]. While it is not obligatory for CD28 to be co-engaged to induce ICOS expression, CD28 can enhance ICOS levels [43]. Although ICOS shows structural similarity to CD28 and CD152, it lacks the specific MYPPPY motif present in CD28 and CTLA-4, which is essential for interaction with CD80 and CD86 [44,45]. Therefore, a ligand for this molecule is ICOSL (B7H, B7RP-1, and CD275) expressed on APCs [46].

The pathway involving ICOS and ICOSL offers a crucial co-stimulatory indication, supporting the expansion of T cells and primarily ensuring their survival. Furthermore, ICOS manages the formation and reaction of T follicular helper (Tfh), Th1, Th2, and Th17 cells and contributes to sustaining the equilibrium of memory effector T cells and regulatory T cells (Tregs) [47]. Indeed, ICOS has a more pronounced capability to initiate the PI3K/Akt pathway and activate the subsequent MAPK cascade when compared to CD28 [48].

The CTLA4 gene is located within strand 2q33 on chromosome 2 in humans. It consists of 233 amino acids [49]. CTLA-4 appears on the surface of activated T cells upon contact with antigen and inhibits further lymphocyte responses. Thus, CTLA-4 provides a negative feedback signal in the specific immune response, preventing it from expanding [50]. Ligands for CTLA-4 found on the surface of APC are CD80 and CD86 from the B7 family of ligands [51]. An unusual feature of the CTLA-4 dimer, rarely seen in other polymeric proteins, is that the ligand-binding sites are located not at the junction of two subunits, jointly forming a single binding site, but at a considerable distance from each other. This allows two ligands to be bound by a single CTLA-4 dimer and consequently allows polymeric CTLA-4 structures to be formed in the cell membrane, alternating with CD80/CD86 in a zip-like pattern [52]. The inhibitory role of CTLA-4, especially for CD28-induced signaling, is evident in mice possessing a defective, non-functional CTLA-4 gene; these animals die at 2–3 weeks of age due to uncontrolled lymphocyte division, leading to massive inflammation in most organs, which is known as lymphoproliferative syndrome [53,54].

The mechanisms of CTLA-4’s inhibitory action are based on specific molecular pathways. Both CD28 and CTLA-4 can bind the same ligands, but CTLA-4 has a higher affinity for CD80 and CD86, which results from the specific binding of dimers described above [55]. As a result, CD28 is displaced from the ligand complexes, and the lymphocyte-activating signal is attenuated [56].

Blocking of the TCR/CD3 complex-derived signal occurs due to the interaction of the intracellular CTLA-4 fragment with the amino acid motif YVKM with SHP-2 phosphatase (tyrosine phosphatase 2) indirectly through PI3K [57]. SHP-2 has the inhibitory properties of phospholipase C gamma 1 (PLCγ1) activated by CD3 [58]. In addition, TCR-mediated signaling can be inhibited by the phosphatase PP2A (protein phosphatase 2A), which likewise binds to CTLA-4 [59].

Inhibition of CD28 co-stimulatory signaling also occurs through PP2A, which can also inhibit CD28-enhanced IL-2 production by blocking Akt kinase [40]. It affects TCR-activated MAPK pathways by stimulating the JNK (c-Jun N-terminal kinase) pathway while inhibiting the ERK (extracellular signal-regulated kinase) pathway [60]. CTLA-4-derived signaling alters the expression of cell cycle control proteins. This is due to an early exit from the G1 phase, entry into the S phase, a prolonged S phase period, and inhibition induced by increased p27 (kip1) expression. The lack of CTLA4 translates into an increase in T-cell proliferation through an increase in IL-2 secretion in the S and G2-M phases [61].

The PD-1 receptor is a protein encoded in humans by the PDCD1 gene, whose locus is 2q37.3. It consists of 288 amino acids [62]. It is a receptor that is expressed on T cells, B cells, monocytes/macrophages, dendritic cells (DCs), and, although at a minimal level, in natural killer T (NKT) cells [63]. When it binds to PD-L1 (B7H-1) and PD-L2 (B7H-2), as well as other members of the B7 family of ligands, it inhibits immune system stimulation [64].

Negative co-stimulatory signals conveyed by PD-1 and CTLA-4 serve distinct roles. While CTLA-4 modulates the priming of T cells in lymphoid organs, PD-1 primarily regulates inflammatory responses in peripheral tissues. Furthermore, unlike CTLA-4, PD-1 can hinder TCR- and CD28-triggered activation by enlisting inhibitory phosphatases like SHP-2, which dampens the initiation of PI3K activity [63]. Inhibition of T-cell activity occurs through the dephosphorylation of PI3K, which leads to the blocking of Akt activity and consequently impairs the energy metabolism of the cell [40].

Another effect induced by PD-1 is to block the binding of ZAP-70 kinase to the CD3ζ subunit by preventing the phosphorylation of both molecules. PD-1 consequently blocks TCR receptor-derived signal transduction. PD-1 also inhibits the activation of PKCθ and ERK kinase [65]. It has been proven that the inhibitory effect of PD-1 is reversible when CD28 stimulation occurs and STAT5-activating cytokines, for example, IL-2, IL-7, and IL-15, are released [66,67].

Interactions between PD-1 and its ligands transmit an inhibitory signal to T cells, which has the measurable effect of hampering the proliferation and production of cytokines by these cells [68]. However, studies show that stimulating PD-L1 can induce the conversion of naive CD4+ T cells to induced T regulatory cells (iTregs), which are critical mediators of peripheral tolerance that actively suppress the formation of effector T (Teff) cells [69]. In the absence of PD-1 signaling, PD-L1 and PD-L2 may provide a positive signal to T cells and stimulate their proliferation and cytokine production [70]. Teff cells, when activated, express both PD-1 and PD-L1. Also, higher PD-L1 expression is correlated with a higher Teff proliferation capacity [71].

The B- and T-lymphocyte attenuator (BTLA) is another CD28 family member functioning as a negative co-stimulatory receptor. It has been observed that BTLA is consistently expressed at low levels in various cell types, including naïve B and T cells, Tfh cells, macrophages, DCs, NKT cells, and natural killer (NK) cells. However, opinions are divided on the presence of the receptor on Tregs. While some argue that BTLA is found in Tregs [72,73], others deny it [74,75,76].

The BTLA gene is situated in the q13.2 region of chromosome 3 and comprises 5 exons, totaling 870 base pairs in length. The structure of BTLA bears resemblance to that of PD-1 and CTLA-4. BTLA is a type I transmembrane glycoprotein and a member of the immunoglobulin superfamily (IgSF), consisting of 289 amino acids [77]. This structure encompasses an extracellular domain, a transmembrane domain, and a cytoplasmic domain [78].

BTLA binds with the herpes virus entry mediator (HVEM), which does not belong to the classic B7 family. HVEM is a member of the tumor necrosis factor receptor (TNFR) superfamily that has been identified as a BTLA ligand [79]. Its binding with BTLA reduces cell activation, cytokine production, and proliferation [80]. Therefore, BTLA deficiencies in various experimental animal models promote the development of autoimmune diseases or worsen their course [81,82]. For example, BTLA-deficient mice presented more pronounced experimental autoimmune encephalomyelitis.

It should also be emphasized that HVEM can interact with BTLA in a cis or trans manner. HVEM and BTLA can be co-expressed on the same cells, forming a cis complex, or present on different cells, forming a trans interaction [83]. Both types of interaction inhibit T-cell activity, but in different mechanisms; the cis complex prevents interaction with other co-signaling molecules, while trans interaction inhibits nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB)-dependent cell activation.

The role of BTLA as a negative co-stimulatory receptor has been substantiated by studying mice lacking BTLA, which displayed increased susceptibility to autoimmune disorders [84]. Additionally, laboratory observations have demonstrated that anti-BTLA agonists convey inhibitory signals to T cells [85]. When BTLA is engaged, it hinders the activation of T cells mediated through CD3/CD28. BTLA transmits signals by recruiting the SHP-1 and SHP-2 phosphatases. The activation of SHP-1 and SHP-2 initiates a suppressive signal directed at TCR transduction, effectively halting T-cell activation. A notable distinction in signaling between BTLA and PD-1 is that BTLA brings in SHP-1, which is a more potent phosphatase for inhibiting TCR and CD28 signaling. In the literature, one can find information about bidirectional BTLA signaling. The GRB-2 motif recruits the PI3K protein subunit p85 and further stimulates the PI3K/Akt signaling pathway, which promotes B- and T-cell proliferation [86].

BTLA at the molecular level interacts with proteins involved in the production of IFN-γ (interferon gamma), IL-17RA (IL-17 receptor antagonist), or the development of B and T cells [87], as well as with bone morphogenetic protein (BMP), an intercellular signaling molecule responsible for growth and differentiation pathways. Studies have shown that BMP initiates the p38 MAPK pathway to produce INFs type I (INF-α and INF-β) [88]. If the DDX19B pathway relating to the response to viruses by the RIG-I-like receptors (RLR) is induced, there will be a negative regulation of INF production [89].

The list of CD28 receptor family members and their molecular interactions are presented in Table 1.