CD26 is able to provide a costimulatory signal to T cells [17, 22, 54, 72, 75]. Co-stimulation can be defined as a ‘stimulus necessary for optimal T cell activation delivered jointly to T cells along with the TCR signal’. The costimulatory characteristics of CD26 are that it augments T-cell responses to foreign antigen, initiates signal transduction, increases cytokine secretion and proliferation, upregulates activation markers such as CD25, CD71 and CD69, induces differentiation into effector cells, and enhances provision of help to B cells and cytotoxic T lymphocytes (CTL) [27, 77]. Certain anti-CD26 monoclonal antibodies (MoAbs) induce T cells or CD26 transfected Jurkat cells to increase proliferation and IL-2 production. Moreover, mitogen- and antigen-induced proliferation can be inhibited in vitro by some anti-CD26 antibodies [25, 78]. Cytotoxicity in CTL has been shown to be triggered by antibodies to CD26 [2, 79]. The association of CD26 with other molecules on the cell surface, such as CD45 and ADA, points to mechanisms by which CD26 may mediate signal transduction pathways [52, 80]. CD26 has been shown to induce tyrosine phosphorylation of kinases in these pathways [80–82].
CD26+ cells produce IL-2 regardless of the coexpression of CD4, whereas CD26−/CD4− cells produce insignificant amounts of IL-2 . Levels of IL-2 production correlate with CD26 expression in mitogen-stimulated rabbit T cells . Moreover, studies on sorted populations found that the CD26+ subset shows a 2–6-fold greater T-cell proliferation than the CD26− subset in response to tetanus toxoid (TT) . CD26 has been shown to be required for antigen induced proliferation in the absence of exogenous IL-2 . These studies indicate that CD26 is a marker for T cells producing IL-2.
Plana et al. (1991)  demonstrated that the anti-CD26 MoAb 134–2C2 increases proliferation of purified T cells either stimulated by phorbol 12-myristate 13-acetate (PMA) at 5 ng/ml or in the presence of exogenous IL-2 (20 U/ml). In addition, the IL-2 secretion is increased, as is the expression of mRNA of IL-2, interferon (IFN)-γ and the IL-2R α and β chains (CD25 and CD122). T-cell proliferation is inhibited when the IL-2R is blocked with an antibody to CD122 in this situation, suggesting that activation via CD26 modifies the IL-2/IL-2R autocrine proliferation pathway . Concordantly, the MoAb CB.1 enhances the IL-2 production by CD4+ and CD8+ T cells in the presence of antigen-presenting cells (APC) and proliferation is inhibited by anti-CD122 . To confirm the role of the autocrine pathway in CD26 mediated T-cell proliferation, Dang et al. have shown that a solid phase immobilized anti-CD26 moAb (1F7) with submitogenic doses of anti-CD3 enhances 3H-thymidine uptake by CD4+ cells . Furthermore, this increase in CD4+ T-cell proliferation was associated with a marked increase in the expression of both IL-2 and IL-2R and proliferation was completely inhibited by an anti-CD122 blocking MoAb. Similar results were obtained when anti-CD2 rather than anti-CD3 was used. CD26 also has a comitogenic effect on CD3 induced IL-2 production by Jurkat cells transfected with cell surface CD26 [81, 86]. In soluble form, the anti-CD26 MoAb 1F7 inhibits T-cell proliferation induced by TT, but in a solid phase immobilized form enhances PMA, CD3 or CD2-mediated proliferation [19, 25]. Similarly, the MoAb AC7 enhances T-cell proliferation stimulated by PMA only in the presence of APC and induces enhanced proliferation only when coimmobilized with anti-CD3 MoAb OKT3 . Anti-CD26 MoAb TA5.9 or Ta1 together with a subthreshold level of anti-CD3 MoAb are able to induce expression of activation markers, such as CD69, CD25 (IL-2R α chain) and CD71 on CD4+ and CD8+ T cells . This is an important indicator of costimulatory ability.
CD26 has a role in the development of effector functions by CD8+ T cells. CD26-triggered cytotoxic activity of CTL requires the presence of Fc receptors on target cells . Anti-CD26 MoAb Ta1 or CB.1 can activate T-cell proliferation and trigger cytotoxicity and granule exocytosis in CTL clones only in the presence of exogenous IL-2 and APC with high density Fc receptors [1, 2, 79]. Others found that cytotoxic activity could be generated in CD8+ T cells costimulated with coimmobilized anti-CD3 and anti-CD26 without APC .
The presence of CD26+ T cells enhances the Ig production by human peripheral B cells stimulated with pokeweed mitogen , suggesting that CD26 may have a role in potentiating T-helper cell functions. These evidences indicate that CD26 has an extensive range of costimulatory properties. However, such data needs the caveat that antibody cross-linking is a strong signal. Antibody-mediated cross-linking of CD2 generates signals significantly greater than those triggered by its physiological ligands, rat CD48 and human CD58 . Thus, it is possible that MoAbs to CD26 generate signals that are greater than physiological.
CD26-dependent signal transduction
The mechanism by which signals initiated by CD26 on the cell surface are transduced to the interior of the cell are not fully characterized (Fig. 1). The signalling pathway initiated by CD26 involves intracellular calcium mobilization  and precedes increased proliferation . Cross-linking CD26 with anti-CD26 MoAb 1F7 and a secondary antibody also results in increased tyrosine phosphorylation of a variety of proteins known to be substrates of T-cell activation . These proteins are identical to those phosphorylated following cross-linking of anti-CD3 antibodies by a secondary antibody, namely p56lck, p59fyn, ZAP-70, c-Cbl, phospholipase C-γ, and MAP kinase. Thus, CD26 mediated signals involve many of the same substrates as the TCR signal. Inhibition of src PTKs, such as p56lck, by herbimycin A prevents tyrosine phosphorylation of these substrates, further confirming the role of these kinases in CD26 mediated signalling .
Figure 1. A model of CD26/DPP IV action at the T-cell surface. The ADA binding activity of CD26 contributes to the control of adenosine levels at the T-cell surface. Our models of residues 133–766 of CD26/DPP IV (Fig. 2) and of human ADA  are depicted here. The peptidase activity of CD26 requires dimerisation [16, 91–93] but the dimerization site is unknown. The CD26-triggered signal transduction pathway is CD3 zeta chain dependent  and includes phosphorylation of the PTKs p56(Lck), p59(Fyn) and zeta associated PTK of 70 000 kDa (ZAP-70), and of MAP kinase, Cbl, and phospholipase Cγ[81, 82], suggesting that such signals are transmitted to the nucleus via the TCR/CD3 pathway. The diagrams and models of molecules are not to scale.
Download figure to PowerPoint
There is evidence for a functional relationship between CD26 and the TCR/CD3 complex. Studies of cloned cytotoxic T cells and NK cells have shown that an anti-CD26 MoAb, CB.1, can trigger cytotoxicity in CD3+ cells, but not in CD3− cells . Furthermore, modulation of CD3 leads to transient refractoriness, during which the cells are unresponsive to triggering by anti-CD26 . In contrast, modulation of CD26 enhances T-cell proliferation and does not lead to such refractoriness .
Anti-CD3 plus anti-CD26 MoAb 1F7 can modulate the CD3/TCR complex and inhibit the proliferative response of T cells, whereas anti-CD3-mediated T-cell proliferation is increased following CD26 modulation . These findings indicate that cell-surface expression of the TCR/CD3 is required for a CD26 function. However, CD26 is not comodulated with the TCR/CD3 following incubation with anti-CD3 MoAb, indicating that a strong physical linkage between CD26 and TCR/CD3 is unlikely [1, 19]. The ζ chain of CD3 is essential, but not sufficient, for inducing IL-2 production via CD26 . PMA-induced T-cell proliferation is enhanced by anti-CD26 MoAb in the absence of anti-CD3, suggesting the involvement of protein kinase C in the CD26 mediated signal. The link is probably not direct, as anti-CD26 MoAb are not comitogenic with ionomycin . In summary, it may be concluded that while CD26 mediated signalling requires the expression of the TCR and a ζ chain of CD3 and involves many of the same substrates as TCR-mediated signalling, this association is not due to a direct physical connection.
CD26 has been shown to coprecipitate with the 180 kDa isoform of CD45. This isoform is expressed preferentially on CD45RO+‘memory’cells, corresponding with the distribution of CD26 . An interaction between CD26 and CD45 could account for increases in activity of PTKs, phosphorylation of CD3 and calcium mobilization following anti-CD26 modulation.
These findings provide evidence that the signal transduced by CD26 overlaps with the TCR/CD3 pathway. Both the TCR/CD3 and CD26 mediated signals result in activation and translocation of the transcription factor NF-κB . Such observations support a costimulatory role for CD26 in CD3 induced signal transduction.
The role of DPP IV activity in T-cell activation
CD26 has a DPP IV ecto-peptidase activity that catalyses the cleavage of N-terminal dipeptides from polypeptides with proline or, in order of decreasing efficiency, alanine, hydroxyproline, serine, glycine, valine or leucine at the second position from the amino terminus [30, 31, 37, 96]. Many natural substrates are known, including chemokines, neuropeptides, hormones and growth factors, whose activity is modulated by the cleavage [31, 36, 96–100]. Such cleavage generally results in the inactivation of biological activity and contributes to the regulatory mechanisms of many biological processes. Indeed, using CD26-deficient rodents it has been shown that the DPP IV activity is clearly important for digestion  and the control of blood glucose concentration . In addition, there is a clear role for DPP IV activity in T cell and monocyte migration [38, 70, 74, 103].
The importance of the DPP IV activity of CD26 in vivo is debated because inhibitor specificity is difficult to guarantee [10, 99]. Studies employing DPP IV enzyme inhibitors show that they inhibit T-cell proliferation and cytokine production [22, 24, 29, 104–107]. Studying both mitogen and antigen induced T-cell activation in vitro, it has been found that inhibitors of DPP IV impair DNA synthesis, Ig production and secretion, and production of IL-2 and IFN-γ. Similar results were observed in PWM-stimulated PBMC and U937-H cells, including a reduction in IL-6 and IL-1β production [24, 107]. In addition, competitive DPP IV inhibitors impair DNA synthesis in B cells  and downregulate proliferation, but not cytotoxicity, of IL-2 stimulated CD56+ NK cells. DPP IV inhibitors have been shown to suppress phosphorylation of p56lck, Ca2+ flux and activation of phosphoinositol-3 and the MAP kinases ERK1 and ERK2 , all known to be involved in T-cell activation. However, tyrosine phosphorylation of proteins that are unaffected by CD26-triggering, such as MAP kinase p38, are also induced by DPP IV inhibitors . Reinhold et al. (1997) have demonstrated that DPP IV inhibitors induce secretion of the inhibitory cytokine transforming growth factor-β1 in PWM-stimulated PBMC . These results, however, do not exclude effects of DPP IV inhibitors unrelated to their inhibition of the DPP IV activity of CD26 .
Many studies do not support a role for the DPP IV activity in in vitro assays of CD26 function. CD26-mediated triggering of cytotoxicity of activated T cells is not inhibited by ‘highly specific’ competitive and irreversible inhibitors . Both CD26+ and CD26− Jurkat cells have been found to be equally susceptible to the effects of DPP IV inhibition, indicating that suppression of T cell activation by DPP IV inhibitors could be due to effects on non-DPP IV enzymatic activities . The CD26/DPP IV deficient rat strain (DPP IV-negative Fischer344) exhibits no defects in in vitro responses to mitogen or antigen . Moreover, suppression of in vitro responses to mitogens or Mycobacterium tuberculosis antigen by the DPP IV inhibitor Lys(Z(NO2))-thiazolidide is equally effective in wild type and CD26-deficient Fischer344 rat cells . Interestingly, PBMC and kidney from CD26-deficient Fischer344 rats exhibit 50 and 1%, respectively, of the Ala-Pro hydrolysing (DPP IV-like) enzyme activity of wild-type Fischer344 rats , indicating that leucocytes possess important DPP IV-like enzyme activity(ies).
An approach to resolving conflicting results obtained with enzyme inhibitors is to produce mutant CD26 lacking DPP IV activity by mutating the active site serine. Most of 11 mutant enzyme-negative CD26 expressing TCR+ Jurkat clones were more easily triggered via CD26 than Jurkat cells transfected with wild-type CD26 . In a separate study, a Jurkat cell line that surface expressed DPP IV-negative CD26 produced significantly more IL-2 than untransfected but less than wild-type CD26 transfected cells, suggesting that DPP IV activity accounts for only some of the IL-2 inducing activity of CD26 [22, 27]. Most importantly, the costimulatory activity of CD26 is retained in mutants lacking most of the hydrolase domain . These data indicate that proteolytic activity is not a prerequisite for the T-cell activating or costimulating properties of CD26 in vitro.