Localization of CD26/DPPIV in nucleus and its nuclear translocation enhanced by anti-CD26 monoclonal antibody with anti-tumor effect

We assessed the subcellular localization of CD26 by biochemical analysis. Karpas 299 cells and JMN cells, with endogenous CD26 expression, were fractionated into membrane, cytoplasmic, or nuclear fractions. By immunoblotting, CD26 was detected as a 110 kDa band in the nuclear fraction, as well as in the membrane and cytoplasmic fractions (Fig. 1A and 1B). CD26 expression was also observed in the nuclear, membrane and cytoplasmic fractions of J/CD26 cells, but not in any fractions of J/Wt cells (known to be CD26-negative cells at both the protein and mRNA level) [11] (Fig. 1C). Mesothelioma cell line JMN relatively exhibits higher CD26 level in the nuclear fraction than that in two T cell lines, Karpas 299 and J/CD26 cells. The difference may be due to difference in cell lineages between mesothelioma cell and T lymphocyte. To confirm the 110 kDa band observed in immunoblotting was indeed CD26, CD26 was immunoprecipitated from each fraction of the JMN cells with the murine CD26 mAb, 1F7, and detected by immunoblotting with a CD26 pAb. The 110 kDa band was confirmed as the CD26 protein. However, no NLS sequence of CD26, which is defined as cluster of a series of basic amino acids, is found by PSORT II program that predicts subcellular localization http://​www.​psort.​org. CD26 may be required to associate with other partner molecules to move to the nucleus.

Immunoelectron microscopic analysis of Karpas 299 and JMN cells showed immunogold labeled-CD26 on the cell surface side of the plasma membrane and in the lysosome, as previously reported [9, 10] (Fig. 2A). In addition, immunogold particles were diffusely observed in the nucleus of the Karpas 299 cells (Fig. 2A), while in the JMN cells, the immunogold particles were clustered in areas of high electron density (such as the chromatin structure) in the nucleus (Fig. 2B).

The CD26 in nucleus was assessed for DPPIV enzyme activity. As a result, specific enzyme activity was detected in the nuclear fraction (and the cytosolic fraction) of both the J/CD26 cells and the Karpas 299 cells (Fig. 2C and 2D). DPPIV activity in the nuclear fraction was also observed following fractionation by an alternative method (Qproteome cell compartment kit, Qiagen) using J/CD26 cells and the JMN mesothelioma cells (data not shown). This DPPIV activity in the nuclear and cytosolic fraction was inhibited by treatment with diprotin A, an inhibitor of the DPPIV enzyme (Fig. 2C and 2D). These data suggest that CD26/DPPIV functions as a catalytic enzyme against, as yet, unidentified molecules in the nucleus.

We examined whether a change in CD26 localization into the nucleus could be initiated by treatment with the CD26 mAb, 1F7, since 1F7 treatment was shown to induce internalization of CD26 from the plasma membrane into the cytoplasm of CD26+ T-cells [7]. CD26 levels in the nuclear fraction of J/CD26 cells were found to be at a peak within one hour of treatment with 1F7, but not 5F8 or IgG₁ (Fig. 3A). Accordingly, CD26 levels in the membrane fraction were decreased within one hour of 1F7 treatment (Fig. 3B). The augmentation of CD26 in the nuclear fraction following 1F7 treatment may be an extreme early event because the increased localization of CD26 in nuclear fraction was also observed at the zero hour time point. The increased amount of CD26 at zero time point seems to depend on some reactions during extra consumed time indispensable for experimental procedure. On the other hand, the CD26 mAb, 5F8, which recognizes a distinct epitope to 1F7, did not influence the nuclear localization of CD26. 1F7 reportedly inhibits the cellular proliferation of T-cells in vitro, while no such inhibitory effect has been observed with 5F8 [5, 8]. Thus, the translocation of CD26 into the nucleus, following 1F7 treatment, may result in the inhibition of cell growth in an epitope-specific manner.

The observed enhanced localization of CD26 in the nucleus, and the concomitant decrease of CD26 in the membrane fraction, following 1F7 treatment, suggests that CD26 on the cell surface moves into the nucleus as a result of stimulation with 1F7. Thus, we biotin-labeled J/CD26 cell surface proteins and these were chased following 1F7 treatment. Biotin-labeled CD26 levels increased dramatically in the nuclear fraction within one hour following 1F7 treatment (Fig. 4A). In contrast with the nuclear fraction, biotin-labeled CD26 decreased in the membrane fraction within one hour following 1F7 treatment (Fig. 4A). Thus, cell surface CD26 appears to translocate into the nucleus following 1F7 treatment. In addition, it appeared that this translocation of CD26 into the nucleus from the cell surface was constant as biotin-CD26 was always observed in the nuclear fraction, before and after treatment with IgG₁ (Fig. 4A). Karpas 299 cells express CD26 endogenously, and our previous report has shown that cell cycle arrest in Karpas 299 cells related to p21 expression is induced following 1F7 treatment [6]. Therefore we investigated the effect of 1F7 on translocation of CD26 into nucleus of Karpas 299 cells. As a result, like J/CD26 cells, biotin-CD26 increased in the nuclear fraction of Karpas 299 cells following stimulation with 1F7, compared with IgG₁ (Fig. 4A and 4B). These results indicate that cell surface CD26 is constantly translocated into the nucleus, and 1F7 treatment enhances this translocation. On the other hand, in JMN cells, these effects of 1F7 were slight. We think this discrepancy may be explained by the difference of cell lineage and nuclear CD26 amount before the 1F7 treatment (Fig. 1A and 1B).

To confirm the augmentation of CD26 translocation into the nucleus by 1F7, we assessed whether biotin-labeled 1F7 could enhance this CD26 translocation in J/Wt and J/CD26 cells. CD26 bound to biotin-1F7 was detected in the nuclear fraction and the membrane fraction of J/CD26 cells, in contrast with biotin-IgG₁ (Fig. 4C, top panel). The biotin-1F7 enhanced CD26 translocation into the nuclear fraction from the membrane fraction (Fig. 4C, middle panel), confirming that the biotinylation of 1F7 did not influence the induction of CD26 translocation. These data show that the specific binding of 1F7 to a particular epitope on CD26 may induce the translocation of cell surface CD26 into the nucleus. However, the mechanism of this translocation of CD26 into nucleus, and the augmentation by 1F7 treatment, remains unknown.

Previous reports show that membrane-associated proteins, such as the epidermal growth factor receptor (EGFR) family [12‐14], fibroblast growth factor receptor (FGFR)-1 [15, 16], heparin-binding EGF-like growth factor (HB-EGF) [17], and CD40 [18], are shuttled from the plasma membrane into the nucleus. The translocation of these molecules into the nucleus can also be elevated by stimulation with, for example, specific ligands for the receptor, and the ectodomain shedding of proHB-HGF at the extracellular region on the plasma membrane. These membrane-associated molecules transmit specific signals from the cell surface to the nucleus after stimulation, and travel directly to the nucleus via internalization. The receptors that are transported into the nucleus have a common function as transcription regulators. There is no evidence that CD26 functions as transcriptional regulator, but a study which silenced CD26 using small interfering RNA showed an increase in basic FGF in prostate cancer cells in vitro [19]. Altered gene expression, mediated by CD26, may indicate a novel function of CD26 in the nucleus. DPPIV activity in the nucleus may be also implicated in the intra-nucleus dynamics by regulating the activation of unknown substrates of DPPIV in the nucleus.

These results permit us to speculate that nuclear events, such as transcription or cell growth, may be regulated by nuclear CD26.

Karpas 299 cells (derived from an anaplastic, large T-cell lymphoma) [6], JMN cells (derived from a mesothelioma) [4], Jurkat cells (J/Wt), and Jurkat cells containing a CD26 expression vector (J/CD26) [8], were cultured at 37°C in RPMI 1640 medium supplemented with 10% fetal bovine serum, 100 units/mL penicillin, and 100 μg/mL streptomycin.

Cells were incubated in medium with/without the CD26 mAbs, 1F7 or 5F8 (2 μg/mL), or control mouse IgG₁ (Dako Cytomation, Denmark). After appropriate incubation, cells were harvested and washed with PBS. Membrane, cytoplasm and nuclear fractions were extracted using the Qproteome cell compartment kit, and according to the manufacturer's instructions (Qiagen, Hilden, Germany). The membrane fraction contains endosomes and membrane-compartmented organelles, such as mitochondria and the ER, as well as cell surface plasma membrane. Protein concentrations of each fraction were determined using the BCA protein assay kit (Pierce Biotechnology, Rockford, IL, USA). Equal amounts of protein were subjected to 10% SDS-PAGE and transferred to a polyvinylidene fluoride (PVDF) membrane. The membranes were probed with the following antibodies: anti-CD26 polyclonal antibody (pAb) (R&D systems, Mineapolis, MN, USA), anti-Na⁺/K⁺ATPase mAb (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) as a plasma membrane marker, anti-culreticulin mAb (BD Pharmingen, San Diego, CA) as an ER marker, anti-calpain-1/2 mAb (Calbiochem, La Jolla, CA) as a cytoplasmic marker, anti-lamin A/C mAb (Santa Cruz) as a nuclear marker, and anti-nucleostemin Ab (Qiagen) as a nuclear marker. Signals were revealed by enhanced chemiluminescence (ECL). Relative CD26 amounts for representative experiment were quantitatively analyzed with the Image Quant 350 software (GE healthcare, UK) and represented as percentage of CD26 expression in each fraction.

Karpas 299 cells and JMN cells were fixed on ice overnight in 0.1 M cacodylate buffer (0.1% glutaraldehyde and 4% paraformaldehyde, pH 7.4). The cells were dehydrated by two, 5-min incubations in 50, 70, 95, and 100% dimethylformamide in water. Cell pellets were incubated in dimethylformamide/lowicryl (1:1) for 30 min at room temperature. Sections (8 nm) were cut and mounted on copper mesh with 150 grids, incubated with a CD26 pAb (Santa Cruz) overnight, washed four times with PBS, and labeled for 60 min with a secondary, 15-nm gold-immunolabeled, anti-rabbit antibody. Sections were washed with 2% uranyl acetate, followed by 4% lead citrate and visualized with an electron microscope.

Nuclear and cytosol fractions were prepared as described [15, 20], with some modifications. Briefly, cells ware washed with PBS and resuspended in 400 μL of ice-cold buffer containing 10 mM HEPES (pH 7.4), 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, and protease inhibitor (Roche, Mannheim, Germany). The cells were allowed to swell on ice for 15 min, at which time 25 μL of 10% NP-40 was added, and the cells vortexed vigorously for 10 s. Nuclei were pelleted by centrifugation, washed twice with ice-cold buffer containing NP-40, and resuspended in 150 μL nuclear extract buffer containing 20 mM HEPES (pH 7.4), 0.4 M NaCl, 1 mM EDTA, 1 mM EGTA and protease inhibitor (Roche). Equal amounts (5 μg) of fractionated proteins were seeded onto 96-well plates and incubated in the presence or absence of diprotin A (an inhibitor of the DPPIV enzyme) (200 μM; Sigma, St. Louis, MI, USA), and the DPPIV substrate Gly-Pro-aminoluciferin (DPPIV-Glo protease assay; Promega, Madison, WI), and gently mixed. Plates were incubated for 30 min in the buffer system optimized for DPPIV and luciferase activity. Luciferase activity was assessed with a luminometer.

Cells were washed twice with PBS containing 1 mM CaCl₂ and 0.5 mM MgCl₂ (PBS-CM), at 4°C, and incubated for 12 min at 4°C in 1.0 mg/mL N-hydroxysulfosuccinimide (sulfo-NHS)-Biotin (Pierce) dissolved in PBS-CM. To quench unreacted biotin, cells were washed three times with PBS-CM plus 50 mM glycine and washed twice with PBS [21, 22]. After incubation with the indicated antibodies, immunoprecipitation was performed using the CD26 mAb, 1F7 (2 μg), with equal amounts of protein from each fraction, overnight at 4°C. The antibody was bound directly to Protein A Sepharose beads for 1 h at 4°C. The beads were washed four times with NET-2 buffer [23] and the bound proteins eluted by directly adding SDS loading buffer to the beads.

1F7 and control mouse IgG₁ were biotinylated using a biotin labeling kit (Roche). Cells were incubated with culture medium containing the biotin-labeled antibodies for 1 h at 37°C. After washing with PBS, the cells were subcellular fractionated. Equal amounts of protein were precipitated with immobilized NeutrAvidin (Pierce) overnight at 4°C. The beads were washed four times with NET-2 buffer, and precipitated, biotinylated proteins were eluted by adding SDS sampling buffer.