Transglutaminase 2 (TG2) is a ubiquitously expressed multifunctional member of the transglutaminase enzyme family. It has been implicated to have roles in many physiological and pathological processes such as differentiation, apoptosis, signal transduction, adhesion and migration, wound healing and inflammation. Previous studies revealed that TG2 has various intra- and extra-cellular interacting partners, which contribute to these processes. In the present study, we identified a molecular co-chaperone, DNAJA1, as a novel interacting partner of human TG2 using a GST pull-down assay and subsequent mass spectrometry analysis, and further confirmed this interaction via ELISA and surface plasmon resonance measurements. Interaction studies were also performed with domain variants of TG2 and results suggest that the catalytic core domain of TG2 is essential for the TG2–DNAJA1 interaction. Cross-linking activity was not essential for the interaction since DNAJA1 was also found to interact with the catalytically inactive form of TG2. Furthermore, we have showed that DNAJA1 interacts with the open form of TG2 and regulates its transamidation activity under both in vitro and in situ conditions. We also found that DNAJA1 is a glutamine donor substrate of TG2. Since DNAJA1 and TG2 are reported to regulate common pathological conditions such as neurodegenerative disorders and cancer, the findings in the present paper open up possibilities to explore molecular mechanisms behind TG2-regulated functions.
Transglutaminase 2 (TG2) is a widely expressed multifaceted enzyme with distinct biochemical activities that functions both inside and outside the cell . TG2 was primarily described with its transamidation activity, which results in post-translational amine incorporation into proteins or the formation of proteolytically resistant γ-glutamyl-ε-lysine isopeptide bonds between γ-carboxamide group of a protein-bound glutamine and ε-amino group of a peptide-bound lysine residues . TG2, as a transglutaminase, is regulated allosterically by calcium and GTP/GDP; that is, TG2 is found in open conformation and activated when bound to calcium, whereas GTP/GDP binding keeps the enzyme in closed conformation, which in turn results in its inactivation . In addition to its transamidating activity, TG2 acts as a G-protein, protein disulfide-isomerase , protein kinase  and DNA hydrolase , which distinguish TG2 from other members of the family. Besides regulating enzymatic activities, TG2 also has certain non-enzymatic roles such as functioning as an adaptor protein, a cell surface adhesion mediator  and forming protein scaffolds .
Since TG2 has such diverse catalytic activities and non-enzymatic functions, it regulates plethora of physiological and pathological conditions. In the intracellular environment, TG2 participates in signaling events  and thus regulates cell survival particularly in response to hypoxia , oxidative stress  and wound healing , whereas outside the cell TG2 modulates cell–ECM adhesion, cell migration and outside-in signaling, which is largely linked to its interaction with members of the integrin family and fibronectin . Furthermore, TG2 has been implicated in a wide range of pathological conditions such as tissue fibrosis, inflammation, cardiovascular and neurodegenerative diseases, cancer progression and metastasis .
It is amazing how a single protein can participate in such diverse functions, which has long been a matter of intense investigation. In our recent review article, we suggested several crucial factors such as the presence of compartment-specific interacting partners of TG2, short linear motifs (SLiMs) and intrinsically disordered regions (IDRs) in the TG2 sequence, which are most likely to contribute to its functional diversity [15,16]. It is well known that SLiMs and IDRs facilitate highly specific protein–protein interactions with moderate affinities; therefore, they are often exploited in signaling pathways [17–19]. Numerous SLiMs and 13 IDRs have been revealed in the TG2 structure. However, we could report only six known interacting partner-binding regions, namely integrins, syndecan-4, small ubiquitin-related modifier 1 (SUMO1), 14-3-3 protein, BAX/BAK and α1-adrenoreceptor, overlapping these regions, suggesting that probably there are a large number of interacting partners which need to be explored .
Therefore, in the present study we have explored novel interacting partners of TG2 and investigated how they could regulate TG2-mediated functions. Besides some of the already known interacting partners, we could identify a co-chaperone, DNAJA1, as a novel interacting partner of TG2. The DNAJ family of proteins is the largest and most diverse family of co-chaperones that works in collaboration with chaperones heat-shock protein 70 (HSP70) and HSP90. However, they also include members which can work independently from these chaperones . DNAJ proteins are involved in several important cellular functions including the suppression of protein aggregation, folding of nascent and damaged proteins, translocation of proteins into cellular compartments and the targeting of proteins for degradation . Furthermore, they play important roles in several pathological processes such as neurodegenerative disorders [22,23] and cancer . In particular, DNAJA1 itself has been reported to participate in various pathological conditions such as autoimmune diseases , neurodegenerative diseases  and cancer [27,28].
Like DNAJA1, TG2 is also reported to regulate cancer cell migration , apoptosis  and neurological disorders . We speculated that DNAJA1 and TG2 might work in collaboration in regulating these cellular processes and pathological states. In the present study, we have performed several biochemical and biophysical analysis to confirm DNAJA1 and TG2 interactions. We used deletion mutants of TG2 to determine, which domains of TG2 were important for TG2 and DNAJA1 interactions. Furthermore, we have also explored the biological relevance of this interaction under cellular conditions and investigated whether DNAJA1 is a potential novel substrate of TG2.
Materials and methods
Expression, purification and analysis of recombinant proteins
Human TG2 was cloned into pET 30 EK/LIC (Novagen) and pTRIEX 4 Neo mammalian expression vectors (Novagen) as described previously . DNAJA1 cDNA was subcloned into pET 30 EK/LIC via ligation-independent cloning according to the manufacturer's protocol. TG2 was purified as described previously , and the same protocol was used to purify DNAJA1. For Western blots, CUB7402 antibody for TG2 (1:2000) (ThermoFisher), MA5-12745 antibody for DNAJA1 (1:1000) and secondary antibody, horseradish peroxidase (HRP)-conjugated antimouse IgG (Covalab) (1:10 000) were used. Full-length and domain-deleted variants of TG2 in pGEX 2T vector (GE Healthcare) were available in the laboratory . GST-tagged full-length TG2 was a valine-224 variant, whereas GST-tagged domain variants of TG2 had glycine instead of valine at the 224th position. The constructs were expressed in Rosetta 2 DE3 Escherichia coli cells (Novagen). The cells were induced at 25°C with 100 µM isopropyl-β-d-thiogalactoside for 6 h. Cell lysis was performed as described above using binding buffer (50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 1 mM DTT, 1 mM EDTA and 5% glycerol) and the supernatant was incubated with Pierce Glutathione Superflow Agarose Resin (Thermo Scientific) for 2 h at 4°C. After two washes with binding buffer (without glycerol), fractions were collected with the binding buffer containing 10 mM reduced l-glutathione (Sigma) and analyzed via SDS/PAGE.
General cell culture reagents were purchased from Sigma, unless otherwise stated. Human embryonic kidney (HEK) 293T AD cells (Agilent Technologies, USA) were cultured in Dulbecco's modified Eagle's medium (DMEM) and NB4 cells (DSMZ GmbH) were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (Sigma and GIBCO® Life Technologies, respectively), l-glutamine and penicillin/streptomycin antibiotics. Cells were grown in a 5% CO2-containing humidified atmosphere at 37°C.
The generation of stable cell lines of HEK 293T AD cells overexpressing human TG2 was carried out as described previously . For the down-regulation of DNAJA1, TG2-overexpressing HEK 293T AD cells were transfected with DNAJA1-specific Silencer Select Pre-designed siRNA (Ambion) using Lipofectamine® 2000 (Invitrogen) and scrambled RNA-transfected cells were used as a control. NB4 cells were differentiated for 72 h by adding 1 µM all-trans-retinoic acid (ATRA; Sigma, R2625) to express endogenous TG2 . These differentiated cells were used for experimental analysis.
GST pull-down assay
GST pull-down experiment was performed with differentiated NB4 cells. NB4 cells were collected and washed in phosphate-buffered saline (PBS) and subsequently resuspended in RIPA buffer (50 mM Tris/HCl, 150 mM NaCl, 1 mM DTT, 1 mM EDTA, 1 mM PMSF, 0.5% Triton and protease inhibitor cocktail). The cells were lysed for 30 min at 4°C and then centrifuged at 10 000 g for 20 min. The clear lysate was estimated via the Bradford assay. Glutathione resin was incubated with 1 mg/ml cell lysate and 100 μg of purified recombinant GST-tagged TG2 for 1 h. An equal amount of GST was used as a control. Later, the resins were washed four or five times with 1 ml of RIPA buffer and then boiled in Laemmli sample buffer and separated by SDS/PAGE. DNAJA1 was detected with anti-DNAJA1 antibody. The GST pull-down experiment was also repeated with recombinant purified GST-TG2 and DNAJA1. The equal amount of GST-TG2/GST and DNAJA1 (100 μg) was combined in a single Eppendorf tube with glutathione beads and left for interaction for 1 h at 4°C. The rest of the procedure was the same as described above.
Mass spectrometry (LC–MS/MS)
GST pull-down assay samples were given for mass spectrometry analysis. Only those protein bands which were unique to TG2 pulled down samples compared with GST control, were excised from the SDS/PAGE gel. The gel slices were in-gel digested with trypsin. During digestion, a reduction was first performed using DTT followed by alkylation with iodoacetamide. The overnight trypsinization was performed using stabilized MS grade bovine trypsin (AB SCIEX) at room temperature and the digested peptides were extracted and lyophilized. The peptides were redissolved in 10 µl of 1% formic acid and 4 µl of sample was used for LC–MS/MS analysis. The 4000 Q TRAP (AB SCIEX)–nanoHPLC (Bruker) LC–MS/MS system was used for data acquisition. The acquired LC–MS/MS data were used for protein identification with the help of ProteinPilot v4.0 (ABSciex) search engine searching the SwissProt database and using the Biological modification table included in ProteinPilot v4.0. The proteomic analyses were done in the Proteomics Core Facility, University of Debrecen.
Non-denaturing polyacrylamide gel electrophoresis
Recombinant His-TG2 and active site mutant of His-TG2 (TG2 C277S) were incubated in reaction buffers (50 mM Tris/HCl, pH 7.4, 150 mM NaCl and 0.1% Tween 20) including EDTA, CaCl2, CaCl2 + Z-DON (zedira) or GTP for overnight at 4°C. Non-denaturing electrophoresis was carried out in an 8% polyacrylamide gel in 25 mM Tris/HCl and 192 mM glycine-including buffer, pH 8.3, for 2 h at 4°C at 40 mA, and different conformers of TG2 were visualized by Page Blue protein staining solution (Thermo Scientific).
Interaction of His-DNAJA1 with full-length GST-TG2 and domain-deleted variants of TG2 (GST-TG2Δβ-barrel1 and GST-TG2Δβ-barrel2, GST-TG2ΔCAT, GST-TG2Δβ-sandwich and GST-CAT) was analyzed via ELISA according to the standard protocol as described previously . Briefly, TG2 and domain variants (1 μg) diluted in coating buffer (50 mM Tris/HCl, pH 7.5, 150 mM NaCl, 5 mM EGTA and 5 mM EDTA) were used to coat the wells of the Maxisorp (Nunc) microtiter plate overnight at 4°C and wells were blocked with 5% (w/v) milk powder in TBS-T (0.05 M Tris/HCl pH 7.5, 0.15 M NaCl, 0.01 M EDTA and 0.1% Tween 20) for 1 h at room temperature (RT). The plate was then incubated with 0.5 μg of DNAJA1 in TBS-T including 5 mM CaCl2 for 1 h at RT, and binding was detected with an anti-DNAJA1 monoclonal antibody diluted (1:1000) in TBS-T for 1 h at RT. After three washes, wells were incubated with HRP-conjugated anti-mouse IgG (1:5000) in TBS-T and reaction was detected by adding 3,3′,5,5′-tetramethylbenzidine and measuring the absorbance at 450 nm. GST-coated wells were used to measure non-specific binding and subtracted from the values observed in the TG2- and domain variant-coated wells. Assays were carried out in triplicates.
Surface plasmon resonance measurements
Surface plasmon resonance (SPR) measurements were performed in a Biacore 3000 instrument (Biacore, Uppsala, Sweden). Binding assays were performed at 25°C. Anti-GST antibody (GE Healthcare) was immobilized on sensor chip CM5 (BR-1000-12; Biacore™, GE Healthcare) using the amine coupling method as recommended by the manufacturer's protocol. On the sensor chip surfaces, full-length GST-TG2 and GST-tagged domain variants of TG2 (GST-TG2Δβ-barrel1, GST-TG2Δβ-barrel2, GST-TG2ΔCAT, GST-TG2Δβ-sandwich and GST-CAT) were immobilized in running buffer containing 50 mM Tris/HCl, pH 7.5, 150 mM NaCl and 1 mM EDTA, and recombinant GST was used as a reference. The immobilization level of GST-fusion proteins was 500–1000 RU according to their molecular mass. The various concentrations of DNAJA1 protein were injected over the different TG2 variant-coated surfaces and control for 7 min at a 5 μl/min flow rate. The evaluation of the sensorgrams was carried out with BIAevaluation v3.1 software.
TG2-overexpressing HEK 293T AD cells were cultured on glass coverslips and fixed with 4% paraformaldehyde for 15 min at 25°C. After that, the cells were treated with NH4Cl/PBS for 10 min at 25°C to quench free aldehyde groups and permeabilized with 0.1% Triton X-100 for 10 min at 25°C. After washing with PBS-T (PBS, pH 7.4, with 0.1% Tween 20), they were blocked first with 5% goat serum in PBS-T for 30 min and with 5% milk powder for 1 h at 25°C. The coverslips were then incubated with anti-TG2 (1:500, polyclonal rabbit IgG, Santa Cruz Biotechnology) and anti-DNAJA1 antibody (1:500) in 5% goat serum in PBS-T for 2 h at 25°C. They were blocked once more with 5% goat serum for 10 min and treated with secondary goat antibodies coupled with Alexa Fluor® 488 (anti-mouse IgG) and Alexa Fluor® 568 (antirabbit IgG; 1:5000) in 5% goat serum. The nuclei were stained by DRAQ5 (Thermo Scientific.) (1:1000). Finally, the coverslips were mounted with DAPCO/Mowiol (1:50) and visualized by confocal microscopy (Olympus FluoView FV1000).
TG2 activity assays
The measurement of BPA incorporation into surface-bound N,N-dimethylated casein (DMC) was performed as described previously by Slaughter et al.  with some modifications . In case of the kinetic TG2 activity assay, monodansylcadaverine (Dansyl-Cd, a fluorometric labeled cadaverine; Sigma) incorporation into DMC was monitored. The reaction mixture comprised 50 mM Tris/HCl, pH 7.5, 0.2 mM Dansyl-Cd, 16 μM DMC, 1 mM DTT, 5 mM CaCl2 and 100 nM His-TG2 with or without 3 μM DNAJA1. The reaction mixture was incubated at 37°C for 60 min. The increase in flourescence intensity was followed using the BioTek Synergy H1 Multi-Mode Reader (BioTek US; excitation/emission: 360/530 nm). For precise determination of Km and Kcat values, Dansyl-Cd fluorescence standard curve was prepared.
Analysis of DNAJA1 as a substrate of TG2
To determine whether DNAJA1 is a substrate of TG2, a Maxisorp (Nunc) microtiter plate was coated with 0.2–1 μg of DNAJA1 and BPA incorporation into DNAJA1 was measured as described above in TG2 activity assay. In another confirmation experiment described by Ruoppolo et al. , recombinant His-TG2 (0.5 μg) was incubated with various concentrations of DNAJA1 (0.1–1 μg) in reaction buffer including 0.1 M Tris/HCl, pH 7.5, 1 mM BPA as a acyl donor, 1 mM DTT and 5 mM CaCl2 for 1 h at 37°C. The reaction products were directly analyzed by immunoblotting, using streptavidin–peroxidase and the monoclonal antibodies against TG2 and DNAJA1.
Experiments were repeated three times (as stated in the figure legends) with three parallels and the data reported are as means ± SEM from the representative experiment. Statistical significance was determined by two-tailed paired Student's t-test (parametric) by using GraphPad Prism version 5.0. The value of P ≤ 0.05 was considered significant.
Screening for TG2 interacting proteins in differentiated NB4 cells
TG2-interacting proteins play a very crucial role in modulating TG2 functions. We have focused on identifying novel interacting partners of TG2 and used the human promyelocytic leukemia cell line (NB4). NB4 cells upon differentiation with ATRA express a very high level of TG2, which leads to a massive up-regulation of several thousand genes involved in many physiological and pathological processes. Knocking down TG2 in ATRA-treated NB4 cells leads to down-regulation of the above reported genes . Therefore, NB4 cells were considered an excellent cell model to explore TG2-interacting partners and the related functions they modulate. GST pull-down experiments and subsequent mass spectrometry analysis of differentiated NB4 cells led to the identification of human glutathione transferase Pi 1 (hGSTP1), tubulin-α, histone H2A and DNAJA1 as potential binding partners of TG2 (Figure 1A). Since all of the above proteins, except DNAJA1, have already been reported to interact either with TG2 or with a substrate of TG2, DNAJA1 could be considered as a novel interacting protein. Mass spectrometry identified peptides of DNAJA1 with 90% confidence. We confirmed the specific interaction by using anti-DNAJA1 antibody in a Western blot on GST pull-down samples (Figure 1B). Furthermore, we expressed and purified recombinant TG2 and DNAJA1, and performed GST pull-down experiments to confirm their direct interactions (Figure 1C).
TG2–DNAJA1 interaction via GST pull-down assay and mass spectrometry.
TG2 directly interacts with DNAJA1 mainly through its catalytic domain
ELISA and SPR measurements were carried out to verify physical interaction of TG2 with DNAJA1. Interaction of DNAJA1 with domain variants of TG2 has also been investigated to determine the DNAJA1-binding domain of TG2. The results shown in Figure 2 demonstrate that DNAJA1 interacts with full-length TG2 and domain variants of TG2 including GST-TG2Δβ-barrel2, GST-CAT (catalytic domain), GST-TG2Δβ-sandwich, GST-TG2Δβ-barrel1 and GST-TG2ΔCAT. Dissociation constants (Kd) for these interactions were calculated from three independent experiments and given in Figure 2D. We observed the highest binding affinity for DNAJA1 with GST-TG2Δβ-barrel2 (which contains the β-sandwich, the catalytic and the β-barrel 1 domains) with a dissociation constant of 5.87 × 10−8 μM, whereas the GST-TG2ΔCAT variant showed the least interaction with a dissociation constant of 5.3 × 10−7 μM. There is a strong interaction of DNAJA1 with the GST-CAT domain itself (Kd: 6.885 × 10−8 μM), which is almost the same as its interaction with full-length TG2 (Kd: 6.095 × 10−8 μM). These results suggest that the catalytic domain of TG2 is required for TG2–DNAJA1 interaction and the β-sandwich domain together with β-barrel 1 domain stabilizes this interaction.
Confirmation of TG2–DNAJA1 interaction and identification of DNAJA1-binding domains of TG2 via ELISA and SPR measurements.
DNAJA1 interacts with the open conformer of TG2
As mentioned before, TG2 can adopt mainly two distinct conformations depending on the type of binding effectors. In agreement with earlier reports, electrophoresis under non-denaturing conditions revealed that recombinant human TG2 produced in E. coli adopts an open conformation in the presence of Ca2+, whereas GTP induces a closed conformation (Figure 3A). Since TG2 undergoes self-cross-linking in the presence of calcium, we have treated TG2 with ZDON (referred to as iTG2), which binds irreversibly to the cysteine present in the active site of TG2 thereby inhibiting its cross-linking activity (lane 3, Figure 3A). We also checked the TG2 conformer in the presence of EDTA, which was used to remove residual calcium from the bacterial expression system. However, instead of observing a closed conformation, the majority of TG2 existed in an open conformation in the presence of EDTA (Figure 3A). This is most likely because TG2 binds very strongly to the calcium present in the bacterial cell lysate and, by mere addition of EDTA, was not enough to chelate/dissociate this bound calcium from TG2. Also, there are probably some effectors derived from the expression system that help in keeping the TG2 in the open conformation and was not completely removed during the purification and by the addition of EDTA either. To explore whether DNAJA1 binds to the open or closed form of TG2, we performed interaction studies via ELISA in the presence of aforementioned effectors. As indicated in Figure 3B, TG2–DNAJA1 interaction was significantly higher in the presence of EDTA and Ca2+ in combination with Z-DON compared with GTP-including conditions, which suggest that DNAJA1 interacts mainly with the open conformer of TG2. DNAJA1 also showed interaction with the transamidation-inactive mutant of TG2 (TG2 C277S), which indicates that interaction is not dependent on cross-linking activity of TG2.
Separation of open and closed conformers of TG2 and determination of DNAJA1-binding conformer of TG2.
DNAJA1 facilitates and stabilizes in vitro cross-linking activity of TG2
Among several activities of TG2, the most prominent is its transamidase activity that has been implicated in the progression of many diseases [29,37]. To explore the significance of the DNAJA1 and TG2 interaction, we first wished to determine whether DNAJA1 could modulate the cross-linking activity of TG2. As we can see in Figure 4A, BPA incorporation into surface-bound glutamine donor substrate DMC by TG2 in an endpoint assay was higher in the presence of DNAJA1 when compared with control. An increase in the cross-linking reaction was observed on increasing the DNAJA1 concentration. This indicated that DNAJA1 either modulated or stabilized the cross-linking activity of TG2 in vitro. Kinetic measurement of dansylcadaverine incorporation into DMC by TG2 in the presence of DNAJA1 in a fluid phase system did not show any effect on TG2 cross-linking activity when the measurements were taken up to 30 min. Nevertheless, after 30–60 min an increase in transamidation activity of TG2 was observed (Figure 4B,C). The kinetic parameters of TG2 for DMC substrate in the presence and absence of DNAJA1 were estimated by Michaelis–Menten and Lineweaver–Burk plots. Km was calculated as 17.6 μM, Vmax as 8.9 µM/min and Kcat as 90 min−1 for TG2 without DNAJA1. In the presence of DNAJA1, we observed a decrease in the Km value, which was calculated as 13 μM, Vmax as 8 µM/min and Kcat value as 78 min−1. These results suggest that DNAJA1 increases the substrate affinity of TG2 thereby increasing its enzymatic activity as observed in the endpoint activity measurements and later half of the kinetic activity measurements.
The role of DNAJA1 on cross-linking activity of TG2 via in vitro transamidation assays.
Down-regulation of DNAJA1 results in increased BPA incorporation in TG2-overexpressing HEK cells
To explore how DNAJA1 influenced the cross-linking activity of TG2 in cells, we further performed cellular experiments wherein we used BPA, a cell-permeant amine substrate for transglutaminases, to determine the in situ TG2 activity in DNAJA1-down-regulated HEK 293T AD cells stably transfected with human TG2 (HEK-TG2 cells). DNAJA1 siRNA were used to knockdown the expression of DNAJA1 and scrambled RNA-transfected and - untransfected cells as control. Down-regulation of DNAJA1 was demonstrated via Western blotting (Figure 5A). Both ELISA and Western blot were performed to visualize the cross-linking reaction initiated by adding Ca2+-ionophore to the cells. The results indicated surprisingly that there was a significant increase in cross-linking activity of TG2 in cells with down-regulated DNAJA1 compared with the controls (Figure 5A,B). We also used TG2-untransfected HEK 293T AD cells and there was no cross-linking activity in these cells (Supplementary Figure S1). In our previous study, it was also demonstrated that 100 μM TG2 active site inhibitor Z-DON could inhibit the cross-linking activity in HEK-TG2 cells confirming the TG2-specific reaction . These results suggest that DNAJA1 has the ability to regulate the cross-linking activity of TG2 in the cells. To ascertain whether down-regulation of DNAJA1 has any impact on the expression of TG2 in HEK-TG2 cells, Western blotting (Figure 5A) was carried out and there was no difference in the amount of TG2 protein. To explore the changes in the gene expression levels of other the transglutaminases upon DNAJA1 down-regulation, we performed QPCR analysis. None of the other protein members of the transglutaminase family were expressed in HEK-TG2 cells either with or without down-regulation of DNAJ1 (Supplementary Figure S2).
The effect of DNAJA1 on in situ cross-linking activity of TG2.
Co-localization of TG2 with DNAJA1 in cytoplasm of TG2-overexpressing HEK 293T CA cells
To determine whether TG2 co-localizes with DNAJA1 in an intact mammalian cellular system, TG2-overexpressing HEK-TG2 cells were used since NB4 cells gave a very high background. Dual immunohistochemistry staining of HEK-TG2 cells was performed with anti-TG2 and anti-DNAJA1 antibodies. Both TG2 and DNAJA1 were observed to localize predominantly in cytosolic compartment with low expression in nuclei (Figure 6). Co-localization of TG2 and DNAJA1 in the cytosolic region indicates a potential association of the two molecules with physiological processes in vivo.
Immunoflourescent images of TG2 and DNAJA1 in TG2-overexpressing HEK cells.
DNAJA1 is a glutamine donor substrate of TG2
Since DNAJA1 was found to interact with catalytic core domain of TG2, we speculated that it can also serve as a TG2 substrate. To explore this hypothesis, DNAJA1 was incubated with the enzyme in the presence of an amine donor, BPA. Reaction products were analyzed by SDS/PAGE, immunoblotting and ELISA, and results showed that BPA was cross-linked to DNAJA1. The amount of incorporation was increased with higher concentrations of DNAJA1 (Figure 7A,B). Incorporation did not occur in the absence of TG2 or calcium, confirming that BPA incorporation into DNAJA1 is mediated by transamidating activity of TG2 (data not shown). We also performed the kinetic transamidation assay replacing dimethylated casein with DNAJA1 and monitored the incorporation of Dansyl-Cd into increasing concentrations of DNAJA1. We observed a linear increase in the amount of incorporated Dansyl-Cd at increasing concentrations of DNAJA1, confirming that it is a glutamine donor substrate of TG2 (Figure 7C).
TG2-mediated amine incorporation into DNAJA1.
TG2 shows broad substrate specificity, and it also non-enzymatically interacts with numerous proteins inside and outside the cells, which may account for its multiple biological functions. For this reason, we have chosen to focus on identification of novel interacting partners of human TG2 using the NB4 acute promyelocytic leukemia cell line since treating this cell line with ATRA results in enhanced expression of TG2 as well as massive changes in the expression of other genes . We identified a novel intracellular interacting partner of human TG2, which is a molecular heat-shock protein, namely DNAJA1. Previous reports suggest that there are few other heat-shock proteins that bind to TG2 and play important roles in various physiological and pathological processes in co-operation with TG2. For example, TG2 overexpression upon excitotoxic stress and thereby its interaction with Hsp20 leads to modulation of antiapoptotic function of Hsp20/27 complex and reduction in the activity of caspase 3, ultimately protecting the cells from apoptotic damage . TG2 was also shown to interact with HSP70 in HeLa and MDAMB231 breast carcinoma cell, which ultimately regulated cancer cell migration . We have used various biochemical and biophysical methods to confirm DNAJA1 and TG2 interactions.
GST pull-down experiments with NB4 cell lysate and subsequent Western blot analysis of pulled-down samples confirmed TG2 and DNAJA1 interactions. Moreover, we confirmed the direct interactions further by using recombinant TG2 and recombinant DNAJA1. TG2 adopts an open conformation in the presence of calcium and is catalytically active, whereas in the presence of GTP it is predominantly in the closed conformation and inactive. Our data in the present study demonstrate that DNAJA1 interacts mainly with open conformer of TG2. After we observed strong interaction of DNAJA1 with the open conformer of TG2, we could also observe a co-localization of TG2 and DNAJA1 in the cell cytoplasm. This observation reconfirms our previous proposition that TG2 can be present in the open conformation and in the active state in the cell cytoplasm [15,39], which is against the prevailing view that, under physiological conditions, TG2 cannot be active as a transamidase inside the cells owing to low Ca2+ and high GTP concentrations [40,41]. DNAJA1 was also found to interact with the active site mutant of TG2, thereby suggesting that transamidation activity of TG2 is not essential for the DNAJA1 and TG2 interaction.
When we used different constructs of TG2, each one lacking one or more domains of TG2, we observed that DNAJA1 interacted with TG2 domain variants with different affinities. DNAJA1 could interact with full-length TG2 as expected. The most significant observation was that TG2 showed the least interaction with the core domain-deleted variant. However, it interacted with the core domain alone, suggesting that the core domain of TG2 is the most important domain in this interaction and since the core domain alone has been shown to be catalytically inactive , this interaction also confirms that catalytic activity of TG2 is not required for TG2–DNAJA1 interaction.
The role of DNAJA1 on TG2 cross-linking activity was also investigated and we could observe an increase in the TG2 activity on increasing DNAJA1 concentrations in an endpoint assay. However, TG2 activity measurements via kinetic assay did not show any significant increase in enzymatic activity in the early phase of the kinetic reaction, but some increase was seen in the later stage, i.e. after 30–60 min. Statistical analysis on the kinetic assay confirmed that the increase in the activity in the later half was notsignificant. We also determined the kinetic parameters to understand how DNAJA1 regulated TG2 activity. The addition of DNAJA1 shifted the Km values of TG2 from 17.6 to 13 µM while Vmax remained unchanged. Kinetic parameters were calculated only for the first half of the reaction, i.e. until 30 min. During the later half, the reaction was non-linear; therefore, the kinetic parameters could not be calculated. Our results suggest that DNAJA1 stabilizes the TG2 active state conformation, thereby increasing its substrate specificity and its transamidation activity. Although in the early phase of the kinetic measurement, we did not observe any change in the activity, we could observe an increase in the activity in the later half, suggesting that DNAJA1 stabilizes the TG2 active state conformation for a longer time period. Surprisingly, the in situ BPA incorporation assay gave contrasting results compared with in vitro assays. We observed an increase in TG2 activity in DNAJA1-down-regulated HEK-TG2 cells compared with scrambled and untransfected controls. There might be several possible explanations for this result. One of the possible reasons could be that DNAJA1 interaction keeps the enzyme in a multiple protein complex which together masks the catalytic site of TG2 inside the cells thereby preventing catalysis. We have shown here that DNAJA1 binds to the catalytic domain of TG2, which corroborates with this observation. TG2 catalytic site becomes freely available for substrate binding on down-regulating DNAJA1; therefore, enhanced cross-linking activity is seen subsequently. Another possible reason could be that DNAJA1 might compete with other substrates of TG2 and limits its substrate specificity, since we also found that DNAJA1 is a glutamine donor substrate of TG2. In other words, we can conclude that compared with in vitro conditions, the effect of DNAJA1 on TG2 activity in cells shows difference most likely due to the presence of several other factors. As mentioned before, the cross-linking activity of TG2 plays significant roles in several important pathological processes and its in vivo inhibition has gained great importance for therapeutic treatments of human diseases. Our results therefore suggest that DNAJA1 could be considered an important protein target with therapeutic potential. Since we showed that the down-regulation of DNAJA1 improves the activity of TG2 in cells, it will be interesting to observe whether its overexpression inhibits TG2 activity in normal cells and in different cell models of diseases.
In the present study, we could successfully identify yet another interacting partner, which can interact with TG2 non-enzymatically as well as also functions as its substrate. Other proteins, which have been reported to have such dual functions, include fibronectin, BCR, angiocidin and retinoblastoma protein. As mentioned previously, DNAJ family of proteins and, particularly, DNAJA1 play important roles in certain types of cancer and neurodegeneration and since TG2 is also involved in these processes, it will be interesting to see whether DNAJA1 and TG2 interaction regulates these pathological conditions. Our future attempt will be to explore the significance of this interaction via using different cancer and neurological cell models.
bcl2 associated X protein (apoptosis regulator)
breakpoint cluster region protein
biotinamido pentyl amine
extra cellular matrix
human embryonic kidney
human glutathione transferase Pi 1
intrinsically disordered regions
phosphate-buffered saline with 0.1%
quantitative polymerase chain reaction
short linear motifs
surface plasmon resonance
This work was supported by the Hungarian Scientific Research Fund [OTKA NK 105046] and the European Union Framework Programme [7 TRANSPATH ITN 289964 and TRANSCOM-IAPP 251506].
The Authors declare that there are no competing interests associated with the manuscript.
We thank Dr. Éva Cso˝sz for mass spectrometry analysis and Dr. Róbert Király for providing GST-tagged TG2 and its domain variants. We would also like to thank Dr. Máte Demény, Dr. Róbert Király and Dr. Manoj Kumar for their their critical reading of the manuscript and their suggestions.
These authors contributed equally to this work.