| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
J. Biol. Chem., Vol. 279, Issue 17, 17607-17616, April 23, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
¶
From the
Institute of Immunology, Rikshospitalet University Hospital, N-0027 Oslo, Norway, Protein Research Group, Department of Biochemistry and Molecular Biology, University of Southern Denmark, DK-5230 Odense M, Denmark, and ||Institute of Organic Chemistry, University of Tübingen, D-72076 Tübingen, Germany
Received for publication, September 14, 2003 , and in revised form, January 26, 2004.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
(
-glutamyl)lysine bonds within their target proteins (1, 2). In this Ca2+-dependent transamidation reaction, an acyl residue, derived from the -carboxamide group of a peptide-bound glutamine (acyl donor), is transferred to an appropriate primary amine, most commonly the -amino group of lysine residues (acyl acceptor) (3). A catalytic triad involving cysteine 277, histidine 335, and aspartic acid 358 forms the active site of TG2 (4). In the first step, the side chain of a glutamine residue forms a thioester with the active site cysteine, and ammonia is released (acylation). In the following transamidation step, the activated acyl group is transferred to the acyl acceptor amine, forming an isopeptide bond (deacylation). Alternatively, hydrolysis of the thioester bond leads to deamidation, converting the glutamine into a glutamic acid residue.
The food-sensitive enteropathy celiac disease (CD) is a chronic inflammatory disorder with a multifactorial etiology (5, 6). In patients with CD, ingestion of wheat gluten and related proteins of barley and rye induces mucosal lymphocyte infiltration and villous atrophy. CD demonstrates a strong genetic association with the genes encoding for HLA-DQ2 and -DQ8 (5). Gluten-reactive CD4+ T cells have been isolated from the small intestine of CD patients and are almost exclusively restricted by either of these HLA molecules (7, 8). Activation of such T cells is likely to be a critical event in the disease development (5). The gluten-reactive T cells of the celiac lesion predominantly recognize modified gluten peptides in which distinct glutamines have been deamidated and thereby converted to glutamic acid residues (9). Accumulating evidence suggests that this deamidation and formation of gluten-derived T cell epitopes in the intestinal mucosa is mediated by TG2 (10–12). Although the transamidation reaction has a higher rate than the deamidation reaction at neutral pH (3), the rate of the deamidation reaction is increased at lower pH (13), such as in the proximal small intestine (14). The sequence specificity of TG2 toward acyl donor substrates was recently elucidated (13, 15), and studies on the affinity of TG2 for various gliadin peptides suggested that TG2 is directly involved in the selection of gluten T cell epitopes (13).
Patients with active CD have autoantibodies to TG2 (16), which are highly disease-specific and whose formation is dependent on ingestion of gluten. Notably, TG2 itself was described to be involved in covalent complex formation with gluten-derived peptides in the presence of calcium (6, 17). Those TG2-gliadin peptide complexes have been suggested to act as hapten-carrier complexes, which could explain the typical antibody response against TG2 in untreated CD patients (18). In this model the complexes are taken up after binding to the immunoglobulin receptor of TG2-specific B cells. Subsequently the gliadin peptide cargo is presented and recognized by gliadin-specific T cells, which give help to the B cells to secrete TG2-specific antibodies. However, little is known about the formation and the molecular nature of these TG2-gluten peptide complexes.
The aim of the current study was to investigate the complex formation between TG2 and two immunodominant gliadin-derived T cell epitopes. Because these peptides contain glutamine but no lysine residues, TG2 will act as an acyl acceptor in these experiments. Our experiments allowed quantitation of the number of gliadin peptides covalently bound to TG2 under different conditions and demonstrated the chemical nature of this linkage. After proteolytic digestion of these complexes, the sites of the modified lysine residues within TG2 were identified by mass spectrometry.
|
EXPERIMENTAL PROCEDURES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
I, corresponding to 9-gliadin-(57–68)), the deamidated derivative 9-gliadin-(57–68) E65 (QLQPFPQPELPY, defined as IE), the N-terminal-biotinylated analogues of peptides I and IE (B-I and B-IE, respectively), the N-terminal fluorescein-labeled analogue of peptide I (F-I), and peptide PQPQLPYPQPQLPY (defined as II, corresponding to 2-gliadin-(62–75)). Peptides were prepared by solid-phase peptide synthesis on a robotic system (Syro MultiSynTech, Bochum, Germany) using Fmoc/O-tert-butyl chemistry and 2-chlorotrityl resin (Senn Chemicals AG, Dielsdorf, Switzerland) (19). A 5-fold molar excess of Fmoc-L-amino acids was used for coupling. For biotinylation and fluorescein labeling of peptides, biotin or carboxyfluorescein (3 eq each, both from Sigma) was coupled to the free N-terminal amino group of the resin-bound, protected peptides using diisopropylcarbodiimide (5eq) as coupling reagent. The identity of all peptides was confirmed by electrospray mass spectrometry, and purity was analyzed by reverse phase high performance liquid chromatography. The 125I-radiolabeling at the C-terminal Tyr residue of peptide I was performed using the chloramines T method (20). Production of Tissue Transglutaminase—Human TG2 was expressed as a GST fusion protein in Escherichia coli using the vector construct as described (21). Thus, "TG2" in this paper stands for the GST-TG2 fusion protein. The protein concentration of the GST-TG2 preparation was determined by the Bio-Rad protein assay, a modified Bradford assay. To obtain data as accurate as possible, the sequence homolog guinea pig TG2 (Sigma) was used as a standard.
Detection of TG2-Peptide Complexes by SDS-PAGE—Biotinylated peptides B-I and B-IE were incubated at concentrations indicated with 1 µM TG2 at 37 °C in phosphate-buffered saline, 2 mM CaCl2, pH 7.4, for 2 h. In one sample, TG2 was inactivated by incubation with 10 mM iodoacetamide for 15 min at room temperature, before incubation with peptide. The samples were added to Laemmli buffer containing 1% 
-mercaptoethanol and 2% SDS, heated at 97 °C for 3 min, and subjected to SDS-PAGE separation with 0.1% SDS in the running buffer. The proteins were blotted onto a nitrocellulose membrane (OPTITRAN BA-S 85, 0.45 µm, Schleicher & Schuell) by using 25 mM Tris, 192 mM glycine, 20% methanol. After blocking of unspecific binding (5 mM EDTA, 2% dry milk powder in 50 mM Tris/HCl, 150 mM NaCl, 0.1% Nonidet P-40, pH 8.0, for 20 min), the membrane was incubated with streptavidin-horseradish peroxidase (Southern Biotechnology, 1:10,000) for 60 min and washed by 50 mM Tris/HCl, 150 mM NaCl, 0.1% Nonidet P-40, pH 8.0 (4 times x 10 min). Fluorescent reagents (Amersham Biosciences) were added, and autofluorescence was detected by exposure to a film (Scientific Imagin film, Kodak X-OmatTM AR).
Mini Spin Columns—In the kinetics study 2.5 µM TG2 was incubated at 37 °C in phosphate-buffered saline, 2 mM CaCl2 with minute amounts of radiolabeled 125I-I or 125I-IE together with 100 µM unlabeled I or IE, respectively. After incubation, the peptide-TG2 complexes were separated from unbound peptides on Sephadex G50 Superfine (Amersham Biosciences) spin columns equilibrated with phosphate-buffered saline, 0.5% Nonidet P-40, and 0.1% NaN3, which was used as the running buffer throughout the experiment. Preparation of the spin columns and separations were performed as described (22). Just before separation columns (1-ml pipette tips closed by small cotton plugs) were packed with Sephadex G50 Superfine (swelled in the running buffer) and spun semi-dry (1500 rpm, 593 x g for 5 min, Sigma 4–10P, rotor 11140), resulting in a 650-µl gel bed volume. Samples (15 µl) were gently pipetted on top of the gel matrix, and columns were spun by an adapted protocol (linear acceleration 0–3500 rpm (2150 x g) within 8 min followed by 3500 rpm for 2 min). Radioactivity in the void volume and in the columns was counted by a -counter (Wallac, Turku, Finland). The counts were converted to an average number of gliadin peptides bound per TG2 molecule as follows: number of peptides/TG2 = [void count/(column count + void count)] x [mol of peptide/mol of TG2].
In the peptide titration studies, 1 µM TG2 was incubated with different concentrations of peptide I for 30 min at 37 °C. Samples were either directly analyzed by spin column experiments or further treated with 10 mM hydroxylamine, pH 7.4, or 10 mM hydroxylamine, 10 mM EDTA, pH 8.9, before spin column separation.
Capillary Electrophoresis—To quantify deamidation of peptide I, 1 µM TG2 was incubated for 30 min with the fluorescein-labeled peptide F-I under conditions identical to those applied in spin column experiments. The deamidated product F-IE was separated and quantified by micellar electrokinetic chromatography and laser-induced fluorescence detection at 488 nm (Beckman MDQ capillary electrophoresis system) as described (13). Briefly, samples were injected by pressure (0.5 p.s.i., 5 s) into a fused-silica capillary (25 cm in length, 75 µm inner diameter) equilibrated by three rinsing steps using 100 mM sodium hydroxide, water, and electrophoresis buffer (64 mM sodium borate, 20 mM sodium dodecyl sulfate, pH 9.3) (20 p.s.i., 1.5 min in each step). Separations were performed at 22 kV at room temperature, and samples were run from the cathode to the anode. All experiments were performed at least in duplicate.
Chemical Modification of Lysine Residues within GST-TG2—An aliquot of GST-TG2 was dialyzed against reaction buffer (20 mM sodium phosphate, 150 mM NaCl, pH 7.2) supplemented with 10 mM EDTA (Micro-dialyzing system, Pierce). GST-TG2 (2.1 mg, 21 nmol) was dissolved in 700 µl of reaction buffer, and N-[-maleimidobutyryloxy]succinimide ester (Pierce; 80 µl of 10 mM, dissolved in 100% Me2SO) was added and incubated for 30 min. The sample was dialyzed against reaction buffer/10 mM EDTA and incubated with the control peptide CLRMKLPQPELPYPQPELPY (200 µl 1 mM) for 120 min. After the addition of 100 mM cysteine (in 0.1 M Tris/Cl, pH 7.0) for 30 min, the modified GST-TG2 was dialyzed against 10 mM NH4HCO3 and lyophilized. For digestion, modified GST-TG2 (30 pmol) was incubated with trypsin (1 µg, Promega) in 50 mM NH4HCO3 (65 µl total) for 24 h at 37 °C, and proteolytic fragments were analyzed by MALDI-TOF and ESI mass spectrometry.
In-gel Digestion—Protein bands corresponding to GST-TG2-peptide complexes (104 kDa) were excised from the Coomassie-stained polyacrylamide gel and cut into several pieces. The pieces were washed twice by acetonitrile/water 1:1 (v/v) (20 min each) and acetonitrile (20 min each) and dried in a SpeedVac. Proteins were reduced by swelling the gel particles in 10 mM dithiothreitol, 0.1 M NH4HCO3 (30 min, 56 °C). Excess liquid was removed, and thiol groups were alkylated by adding 55 mM iodoacetamide, 0.1 M NH4HCO3 (30 min at room temperature in the dark). Gel particles were washed and dried as described above. A digestion buffer composed of 12.5 ng/µl of trypsin (Promega, modified, sequencing grade) in 50 mM NH4HCO3 was added, and samples were kept on ice for 30 min to allow rehydration of the gel particles with limited autoproteolysis of trypsin. The remaining supernatant was replaced by 50 mM NH4HCO3, and tryptic digestion was carried out overnight at 37 °C.
Mass Spectrometry—Matrix-assisted laser desorption ionization mass spectra were acquired on a Bruker Reflex II MALDI-TOF instrument (Bruker-Daltonik, Bremen, Germany). Electrospray ionization tandem mass spectra were recorded on a quadrupole time-of-flight mass spectrometer (Micromass, Manchester, England). Tryptic peptide mixtures were desalted and concentrated on Poros 20 R2 reverse-phase packing sorbent (Applied Biosystems) packed in GELoader tips (Eppendorf). For ESI mass spectrometry, samples were eluted into nanoelectrospray needles (Protana Engineering, Odense, Denmark) as previously described (23). For spraying, needles were typically held at 800 V toward a skimmer cone (40 V). In collision-induced dissociation experiments (collision gas argon, manifold pressure 
8 x 10-15 millibar, collision energy 32–40 eV) product ions were analyzed by the orthogonal TOF analyzer.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
I). The reaction was stopped by the addition of Laemmli buffer containing 1% -mercaptoethanol and a heating step, and samples were subjected to SDS-PAGE. Specific detection of the biotin moiety on the following blot resulted in a distinct band corresponding to the molecular weight of the GST-TG2 fusion protein (104 kDa) (Fig. 1). The intensity of the signal gradually increased with higher peptide concentrations, reaching a plateau at 150 µM. A very weak band was observed at 20 µM, whereas no signal was visible at 5 µM. Glutamine 65 in peptide I is known to be specifically deamidated by TG2 (24). Substitution of that glutamine for a glutamic acid residue (peptide B-IE) resulted in a complete loss of the 104-kDa-band even when B-IE was used at 150 µM. The same negative result was obtained when TG2 was treated with iodoacetamide (10 mM,15 min) before incubation with peptide, leading to an irreversible alkylation of cysteine residues in TG2 (data not shown).
|
I-complex formation was studied quantitatively by using a 125I-labeled derivative of I as a tracer peptide during incubation. Samples were loaded on a miniaturized gel filtration column, and free peptides were separated from generated TG2-peptide complexes and the unmodified enzyme. The high molecular weight components were detected exclusively in the void volume as shown by SDS-PAGE (Fig. 2A). The peptide fraction, however, was retained in the gel matrix, as proved by measuring the radioactivity in the void volume and the remaining column. After incubation of TG2 with peptide I, radioactivity in these two compartments was counted, and the molar amount of peptide in the void volume was calculated. Together with the known amount of TG2, the average number of peptides bound to a single TG2-molecule could be determined.
|
I, and the average number of peptides bound to TG2 was determined after different time intervals (Fig. 2B). Complex formation was found to occur rapidly. After 5 and 10 min, about 1.2 and 1.9 I peptides per TG2 molecule were calculated, respectively. A maximum of 2.8 peptides/TG2 was observed between 30 and 60 min of incubation time. Interestingly, a slight but continuous decrease of this ratio was found during further incubation, reaching a value of 2.1 peptides/TG2 after 240 min. No radioactivity was found in the void volume if TG2 was treated by iodoacetamide before incubation with peptide I (data not shown).
Subsequently, TG2 (1 µM) was incubated with a 0.5–250-fold molar excess of peptide I (Fig. 2C). An incubation time of 30 min was chosen to allow maximal complex formation (Fig. 2B). For an equimolar ratio, an average number of about 0.2 peptides/TG2 was found. The titration curve showed a maximal slope between 10 µM (1.1 peptides/TG2) and 60 µM (3.8 peptides/TG2). At a 40-fold molar excess, 3.0 peptides were determined to be bound to TG2, consistent with the result from Fig. 2B (2.8 peptides/TG2 for an identical incubation time and molar excess). The titration curve reached a plateau at 100–150-fold molar excess, where a maximal number of 4.2 peptides/TG2 was obtained.
Isopeptide and Thioester Bonds in TG2-Peptide Complexes— In the TG2-peptide complexes described so far (Figs. 1, 2B, and 2C), peptide I could be bound via a thioester bond to C-277 in the active site of the enzyme or via isopeptide bonds involving lysine resides of TG2. To quantify these two peptide fractions, TG2 (1 µM) was incubated with peptide I for 30 min followed by treatment with hydroxylamine at pH 7.4 or 8.9. The nucleophile hydroxylamine specifically cleaves thioester but not amide bonds.
The hydroxylamine treatment led to a slight decrease of the total number of peptides bound to TG2, as shown by spin column analysis (Fig. 2D). The observed difference between both graphs corresponds to peptides bound via thioester bonds to TG2. At peptide concentrations between 45 and 135 µM, 1 µmol of TG2 carried about 0.5 to 0.7 µmol of peptide I linked via thioester bonds. With decreasing concentrations of peptide I the relative amount of thioester-bound peptides continually increased (Fig. 2E). After incubation with 1 µM peptide I, hydroxylamine treatment removed about 50% of the bound radioactivity, indicating that peptide I is bound in similar amounts via thioester and isopeptide linkage. Thus, both thioester and isopeptide bond formation take place over a broad concentration range.
To address whether the thiol to amine exchange reaction for the formation of isopeptide bonded complexes is intermolecular, we studied the backward reaction by adding extra TG2 to isolated TG2-peptide I complexes in a 5-fold molar excess. This, however, did not cause an accelerated decay of the reaction shown in Fig. 2B as it would be expected from an inter-molecular thiol to amine exchange reaction.
Complex Formation Versus Deamidation—To quantify the deamidation reaction taking place concurrently to complex formation, samples in which peptide I was replaced by its fluorescein-labeled derivative F-I were incubated under identical conditions. Samples were analyzed by micellar electrokinetic chromatography. Incubation for 30 min with 200 µM F-I resulted in 37 µM F-IE (Fig. 2C), which still increased at longer incubation times (data not shown). However, at lower peptide concentrations the ratio between TG2-bound peptides (both thioester and isopeptide-linked) and deamidated peptides clearly increased. At 10 µM, about 1.1 µM peptides were detected in the complex, and 4.3 µM were deamidated. At 1 µM, the amounts of covalently bound (0.2 µM) and deamidated peptides (0.33 µM) were almost identical. Under those conditions half of the bound peptides are linked via thioesters (Fig. 2E), meaning that only 0.1 µM (19%) of all converted peptides (0.53 µM) were isopeptide-linked. With increasing peptide concentrations, however, isopeptide bond formation became a rather infrequent event compared with thioester hydrolysis.
Identification of Acyl Acceptor Sites within TG2 by Mass Spectrometry—Next we identified which lysine residues in the GST-TG2 fusion protein are involved in isopeptide bond formation to the targeted glutamine residue in peptide I. After incubation of TG2 (1 µM) with peptide I at molar ratios of 1:1, 1:10, 1:50, and 1:150, complexes were separated by SDS-PAGE. The band corresponding to TG2-peptide complexes and unmodified TG2 (these are not separated under the conditions used) was cut out from the Coomassie-stained gel and subjected to in-gel digestion by trypsin. Digests were analyzed by MALDI-TOF mass spectrometry in the reflector and linear mode, and spectra were searched for masses characteristic for tryptic fragments from GST-TG2 cross-linked to peptide I. Theoretically expected masses were calculated by excluding tryptic cleavage C-terminal of a lysine residue involved in isopeptide bond formation and by allowing one missed cleavage site in the GST-TG2-derived peptide. In the MALDI-TOF mass spectrum obtained after incubating peptide I with TG2 in a 50-fold molar excess, six masses characteristic for the cross-linked tryptic fragments I–VI were found (Fig. 3, Table I). Because peptide I carries an N-terminal glutamine residue, all six signals are preceded by a typical -17-Da signal due to pyroglutamate formation of that residue.
|
|
ItoVI-I was verified by nano-ESI MS/MS. All six products were observed as 3- or 4-fold charged ions in the ESI MS spectrum (Table I; insets in Fig. 4). Fig. 4, A and B, shows the MS/MS spectrum of the triply charged ions 962.5 and 705.7, corresponding to products II-I and VI-I, respectively. MS/MS was performed by collision-induced dissociation (CID) of selected parent ions in the second quadrupole of the quadrupole time-of-flight mass spectrometer. The identity of both the tryptic TG2-derived peptides II and VI and the gliadin peptide I could be clearly elucidated from the generated characteristic b- and y-type fragment ions (25) (found as single and double charged daughter ions). Results were further proved by the series of counter parting b*- and y*-ions, which describe fragments containing the isopeptide bond. Fragmentation within peptide I was preferentially observed N-terminal to proline, as expected in ESI CID MS/MS, and the C-terminal fragment PY was the major signal in all MS/MS spectra recorded. These MS/MS spectra also unambiguously confirmed the site at which the cross-linking reaction had taken place. Of 4 glutamine residues in peptide I, only Gln-65 is cross-linked to lysine 562 and lysine 677 of the TG2-derived peptides II and VI, respectively. By increasing the collision energy, weak signals deriving from fragmentation of the isopeptide bond itself were observed. Under these conditions multiple fragmentation in the backbone of peptide I were induced as well (Fig. 4). Identified fragment ions in the MS/MS spectra of tryptic, peptides I–VI cross-linked to peptide I, are summarized in Fig. 5, left side.
|
|
II) using identical conditions. Analysis of the tryptic in-gel digest by MALDI-TOF MS and nano-ESI MS/MS confirmed the results obtained for peptide I. Expected masses of tryptic peptides I–VI cross-linked to peptide II were observed, both in MALDI-TOF- and ESI MS spectra. By MS/MS analysis characteristic daughter ions for cross-linked peptides I, II, III, V, and VI were detected, whereas no clear signals were observed for cross-linked peptide IV (presumably due to the low signal intensity of the parent ion) or any other TG2-derived peptide (Table I; Fig. 5, right side).
Hierarchy of Lysine Residues Involved in Complex Formation—Cross-linking within TG2-peptide complexes generated at four different TG2/peptide molar ratios (1:1, 1:10, 1:50, and 1:150) was analyzed as described. Comparing the results obtained from MALDI-TOF and nano-ESI mass spectrometry for complexes made with both the I and II gliadin peptides, a clear hierarchy in the participation of the identified lysine residues in the cross-linking reaction was observed (Table II). At a 1:1 ratio, a low signal was observed only for the crosslinked product of tryptic peptide III. This signal got more intensive at a 1:10 ratio, and also peaks for cross-linked peptides II and IV appeared. By increasing the TG2/peptide ratio to 1:50 and 1:150, signals for all six cross-linked peptides were observed, although with different intensities. However, no additional lysine residues within the GST-TG2 fusion protein were found to be involved in isopeptide bond formation when a 150-fold molar excess of peptides was applied during complex formation.
|
-maleimidobutyryloxy]succinimide ester (800 nmol) and subsequently incubated with the control peptide CLRMKLPQPELPYPQPELPY (200 nmol). The modified fusion protein was digested by trypsin, which also cleaves the control peptide C-terminal of Arg-3 and Lys-5. Thus, accessible lysine residues are modified by a carboxy-arginyl-leucyl-cysteinyl-succinimidebutyryl moiety, resulting in a typical mass shift of 555.66 atomic mass units of the corresponding tryptic fragment. The digest was analyzed by MALDI-TOF MS and nano-ESI MS/MS, and 10 lysine residues were identified as modified. Within TG2, Lys-74, Lys-414, Lys-425, and Lys-649 were found as acylated (TG2 sequence given by gi4759228). In the GST protein Lys-27, Lys-45, Lys-113, Lys-125, Lys-191, and Lys-218 were identified as modified (GST sequence given by gi595726). Location of Lysine Residues Involved in Isopeptide Bond Formation—As seen from the three-dimensional structure of TG2 (Protein Date Bank file pdb1fau.ent.Z in the Brookhaven Data Bank), the six lysine residues participating in enzymatic isopeptide bond formation to gliadin peptides are not randomly distributed (Fig. 6). Lys-590, Lys-600, Lys-649, and Lys-677 are all located in the C-terminal domain 4, and Lys-444 and Lys-562 are situated in domain 2. Interestingly, Lys-677 is located at the interface of domain 2 and 4, rather buried in the interior of the protein than exposed at the surface. Also, Lys-562 appears to be partly buried, whereas the other four lysine residues are clearly exposed to the surface of the protein.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
I and II).
Covalent TG2-peptide complexes have been described previously (6, 16), but their molecular structure has remained elusive. Incubation of TG2 in the presence of calcium with the known substrate peptide B-I resulted in complex formation between the peptide and TG2 (Fig. 1). The pretreatment of those samples and the presence of SDS indicate a covalent linkage between the biotinylated peptide and TG2. These covalent complexes might derive from thioester formation between Cys-277 in the active site of TG2 and the side chain of Gln-65 in B-I. Alternatively, TG2 itself could act as an acyl acceptor molecule, resulting in isopeptide bond formation.
The acyl acceptor sites in TG2 were identified by mass spectrometric analysis of tryptic digests of the complexes. By MALDI-TOF MS, nano-ESI MS, and nano-ESI MS/MS, six lysine residues were identified to be involved in isopeptide bond formation to Gln-65 in peptide I (Table I). With peptide II, evidence for involvement of at least five of these six lysine residues was obtained. The number of identified acyl acceptor sites for different peptide/TG2 ratios were slightly higher than expected from the data obtained from spin column experiments (Fig. 2, C and D). The latter results provide only the average number of peptides bound via isopeptide linkage per TG2. However, these peptides could be statistically attached to different lysine residues in different TG2 molecules. The maximum of 3.5 peptides bound per TG2 molecule via isopeptide bonds corresponds to only six lysine residues of TG2 identified as acyl acceptors, demonstrating that isopeptide bond formation is rather specific. Moreover, all modified lysine residues reside in the TG2 moiety. No lysine residues in the fused GST-protein were found to be involved in isopeptide bond formation, although GST has been described as a substrate of TG2 (26).
The identified acyl acceptor (lysine) residues in TG2 are remote from Cys-277 in the active site. Thus, the thiol to amine exchange reaction resulting in the isopeptide-bonded complexes is either intermolecular or TG2 may undergo a major conformational change upon binding to peptide I. With the multiple lysine residues targeted, we believe the former is more likely. However, experiments in which extra TG2 was added to complexes after removing the free peptide did not lead to accelerated decay of the reaction depicted in Fig. 2B as could be expected if the reaction is intermolecular (27). The reason may be that this backward reaction is too slow to be detected under our experimental conditions. Moreover, we observed that hydroxylamine-treated and separated complexes were very stabile over 15 h (data not shown), speaking against a conformational change as the mechanism involved in the thiol to amine exchange reaction.
The enzymatic specificity for the targeted acyl acceptor residues is supported by data generated by chemical modification of lysine residues within GST-TG2. After coupling a control peptide via a succinimide ester and maleimide chemistry, 10 lysine residues were identified as modified, 6 of them in the GST moiety. Only one of those lysines involved in enzymatic cross-linking was modified in these experiments (Lys-649), underlining the specific selection of these acyl acceptor sites. In addition, the observed hierarchy within the targeted lysine residues demonstrates the enzymatic specificity (Table II). The findings might reflect different accessibilities of the targeted lysine residues to the acyl-loaded active site of a second TG2 molecule in an intermolecular acyl transfer reaction or might come from differences in the nucleophilicity between the lysine residues.
In four of the six cross-linked TG2-derived peptides (II–V), the targeted lysine residue is followed by another positively charged amino acid in position +2. In peptide VI an arginine residue is located in position +3. Peptide I deviates from the others because it carries proline in position +2, which induces a kink in the polypeptide backbone. A hydrophobic residue precedes all targeted lysine residues. Possibly these findings may reflect a consensus sequence of preferred acyl acceptor sites.
For some tryptic GST-TG2-derived peptides cross-linked to the I or II peptides, masses above 4000 atomic mass units would, at least theoretically, be expected. These products would be difficult to observe by analyzing the heterogeneous tryptic digest of the complexes by MALDI-TOF MS in the reflector mode. They might be detected by ESI MS as multiple charged ions; however, dependent on their ability to be ionized, signals might also be suppressed. To search more reliably for crosslinked molecules of higher molecular weight within the digests, MALDI-TOF mass spectra were recorded in the linear mode, and furthermore, samples were analyzed by LC-ESI MS/MS experiments, separating the compounds before entering the mass spectrometer. No additional cross-linked products were, however, observed by either of these methods.
To explain the typical antibody response against TG2 in untreated CD patients by TG2-gliadin complexes (18), the deamidated peptide must be released after endocytosis of the complexes. This could be achieved by the hydrolysis of either the thioester or the isopeptide bonds. The results shown in Fig. 2B imply that such a release of peptides is possible. Studying TG2-I-complex formation at a 40-fold molar excess of peptide I over 4 h, a slight but continuous decrease of the complexes was observed after 30 min. After 240 min a release of 25% was measured. Complexes generated with such a 40-fold molar excess of peptide I carry about 27% of their peptides via thioester linkage (Fig. 2E). Hydrolysis of the isopeptide bonds would either require an intermolecular transfer of the acylpeptide from the modified lysine to Cys-277 or, in an intramolecular reaction, a major conformational change of TG2, as the targeted lysine residues are remote from the active site thiol group. The failure to accelerate cleavage by adding more TG2 to isolated complexes and the stability of hydroxylamine-treated and purified complexes suggest that release by hydrolysis of thioester bonds strongly dominates over cleavage of isopeptide bonds. Hydrolysis of the thioester bond represents the common deamidation reaction and is favored at slightly acidic pH (13), which would be found in early endosomes after uptake of the complexes.
To what extent these complexes are formed in vivo and their physiological relevance will be investigated in the future. Especially in case of an overexpression of TG2 and after a glutenrich meal, generation of those complexes might well take place. Studies on their uptake by TG2-specific B cells and on antigen presentation will shed further light on their immunological relevance for antibody production against TG2.
|
FOOTNOTES |
|---|
¶ To whom correspondence should be addressed: Institute of Immunology, University of Oslo, Rikshospitalet University Hospital, N-0027 Oslo, Norway. Tel.: 47-230-71373; Fax: 47-230-73510; E-mail: burkhard.fleckenstein{at}labmed.uio.no.
1 The abbreviations used are: TG2, tissue transglutaminase; CD, celiac disease; CID, collision-induced dissociation; ESI, electrospray ionization; GST, glutathione S-transferase; MALDI-TOF MS, matrix-assisted laser desorption ionization-time-of-flight mass spectrometry; Fmoc, N-(9-fluorenyl)methoxycarbonyl.
|
ACKNOWLEDGMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONEXPERIMENTAL PROCEDURESRESULTSDISCUSSIONREFERENCES |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  Z. Keresztessy, E. Csosz, J. Harsfalvi, K. Csomos, J. Gray, R. N. Lightowlers, J. H. Lakey, Z. Balajthy, and L. Fesus Phage display selection of efficient glutamine-donor substrate peptides for transglutaminase 2. Protein Sci., November 1, 2006; 15(11): 2466 - 2480. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D A van Heel and J West Recent advances in coeliac disease Gut, July 1, 2006; 55(7): 1037 - 1046. [Full Text] [PDF]
|
|
|
|
|
|
W Dieterich, B Esslinger, D Trapp, E Hahn, T Huff, W Seilmeier, H Wieser, and D Schuppan Cross linking to tissue transglutaminase and collagen favours gliadin toxicity in coeliac disease Gut, April 1, 2006; 55(4): 478 - 484. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Sblattero, F. Maurano, G. Mazzarella, M. Rossi, S. Auricchio, F. Florian, F. Ziberna, A. Tommasini, T. Not, A. Ventura, et al. Characterization of the Anti-Tissue Transglutaminase Antibody Response in Nonobese Diabetic Mice J. Immunol., May 1, 2005; 174(9): 5830 - 5836. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Schwertz, F. Kahlenberg, U. Sack, T. Richter, M. Stern, K. Conrad, K.-P. Zimmer, and T. Mothes Serologic Assay Based on Gliadin-Related Nonapeptides as a Highly Sensitive and Specific Diagnostic Aid in Celiac Disease Clin. Chem., December 1, 2004; 50(12): 2370 - 2375. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||
