Jun 5, 2006 - Primary Structure of Microbial Transglutaminase from. Streptoverticillium sp. Strain s-8112â. (Received for publication, November 20, 1992, and ...
Vol. 268, No. 16, Issue of June 5, pp. 11565-11572.1993 Printed in U.S. A.
THEJOURNALOF BIOLOGICAL CHEMISTRY
0 1993 by The American Society for Biochemistry and Molecular Biology, Inc.
Primary Structure of Microbial Transglutaminase from Streptoverticillium sp. Strain s-8112” (Received for publication, November 20, 1992, and in revised form, February 3, 1993)
Toshiya Kanaji, Hiroshi Ozaki, Toshifumi Takao, Hideo KawajiriS, Hiroyuki IdeS, Masao MotokiS, and Yasutsugu Shimonishi4 From the Institute for Protein Research, Osaka University, Suita, Osaka 565, Japan and the $Food Research and Development Laboratories, Ajinomoto Co., Inc., Suzuki-cho, Kawasaki-ku, Kawasaki 210, Japan
of TGases have recently been determined by analyses of the proteinsthemselvesand/ortheircDNAs (9-18). These TGases show low degrees of homology in their amino acid sequences but great similarity in certain regions, especially around the predicted active site Cys residue(Fig. 6). The latter suggests relatedness in the origin of the active site structures of TGases. Moreover, a common origin of these active site structures and those of thiol proteases has been suggested (6). The putative Ca2+ binding regions of TGases have also been suggested to show sequence homology (14), but details of the mechanism of activation of these enzyme by Caz+ are unknown. Some of us (19, 20) recently isolated a TGase from the culture supernatant of a microorganism, Streptouerticillium sp. s-8112. This enzyme is the first TGase to be obtained from asource other than a mammal. We areinterested in the structuralrelationship of this enzyme withmammalian TGases, because the activityof this enzyme is Ca2+-independent, whereas those of mammalian TGases are Ca2+-dependent. More recently, another TGasewas isolated from the acellular slime mold, Physarum polycephalum,and its catalytic activity was found tobe Ca*+-dependent(21). Therefore, in this study to obtain information on the reaction mechanisms and structural and functional relationshipsof mammalian TGases and the microbial TGase, we determined thecomplete amino acid sequence of the TGase produced by Streptouerticillium sp. s-8112 and compared it with those of mammalian TGases Transglutaminases (protein-g1utamine:amine y-glutamyl- including factor XIIIa andguinea pig liver TGase. We found transferase, EC 2.3.2.13) catalyze a n acyl transfer reaction in little sequence homology of the microbial enzyme t o Ca2+the presence of Ca2+, forming amide bonds between the y- dependent TGases from other species, even in the predicted carboxyl groups of glutamine residues and the primary amino active site region consisting of the consensus sequence motif groups of a variety of amines or the t-aminogroups of lysine of thiol proteases. We found, however, that the hydrophobic residues in proteins (1, 2). TGases’ are widely distributed in environment of the catalytic site including a singleCys residue most tissues andbody fluids, including theliver, hair follicles, in the microbial TGase is similar to those of other TGases. epidermis, prostate, andblood, and are thought to involved be The resultsare discussed inrelationtothe evolutionary in diversephysiological functions, such as maintenance of origins of the microbial and mammalian TGases. gross forms of structures and limited degrees of extensibility (3, 4), but little is yetknown of their roles in thesephysiologEXPERIMENTALPROCEDURES ical processes. Of the TGases studied so far, factor XI11 and Materials-The crude TGase was purified from the culture superguinea pig liver TGase (type I1 TGase) are the best charac- natant of Streptoverticillium sp. s-8112 as described previously (19). terized (5-8). The amino acid sequences of a limited number Achromobacter lyticusprotease I and Staphylococcus aureus protease
The complete amino acid sequence of transglutaminase (EC 2.3.2.13) (TGase),which is producedby a microorganism, Streptoverticillium sp. strain s-8112, and catalyzes the acyl transfer reaction between ycarboxyamide groups of glutamine residues in proteins and various primary amines, has been established by a combination of fast atom bombardmentmass spectrometry and standard Edman degradation of peptide fragments produced by treatment of the TGase with various proteolytic enzymes and purified by a reversedphase high performance liquid chromatography. The TGase consists of 33 1 amino acid residues with a chemical molecular weight of 37,863, inagreement with the observed molecular weight (37,869.2 2 8.8) determined from its electrospray ionization mass spectrum. The sequence of the enzyme is very different from those of mammalianTGases represented by guinea pig liver enzyme. The enzyme contains a sole Cys residue, which is essential for its catalytic activity. Hydropathy analysis indicated that the secondary structure of the region around the active site Cys residue is similar to those of mammalianTGases. These results suggest that this microbial protein evolved by a different pathway from that ofmammalian TGases and acquired acyl transfer activity during the evolutional process.
* The costs of publication
of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. I To whom correspondence should be addressed Institute for Protein Research, Osaka University, Yamadaoka 3-2, Suita, Osaka 565, Japan. The abbreviations used are: TGase,transglutaminase; HPLC, high performance liquid chromatography; FAB, fast atom bombardment; ESI, electrospray ionization; TPCK, L-1-tosylamido-2-phenylethyl chloromethyl ketone; PTH, phenylthiohydantoin.
V8 were purchased from Wako Pure Chemical Industries (Tokyo, Japan) and Miles Laboratories (Elkhart, IN), respectively. Arginylendopeptidase, TPCK-treated trypsin, and endopeptidase Asp-N were products of Takara Shuzo Co. (Kyoto, Japan), Worthington Biochemical Corp. (Freehold, NJ), and Boehringer Mannheim, respectively. CNBr was purchased from Nacalai Tesque (Kyoto, Japan). YMC R-ODs-5 andCosmosil 5C18P-300 columns were obtained from Yamamura Chemical Laboratory Co. and Nacalai Tesque (Kyoto, Japan), respectively. H,180 (97 atom %) was a product of Commissariat a L’Energie Atomique (CEA). Protein Modification-The purified TGase of Streptouerticillium sp. s-8112 (3.7mg)was dissolved in a solution (pH 8.5) of 6 M
11565
Streptoverticillium sp. Transglutaminase
11566
guanidine. HC1, 0.1 M Tris, and 0.2% (w/v) EDTA and treated with dithiothreitol at 20-fold molar excess over the concentration of protein. The mixture was stood for 4 hat 40 "C, then treated with 2-fold molar excess of monoiodoacetic acid over dithiothreitol for 1 hour at room temperature in the dark. The solution was dialyzed against 0.1 M acetic acid and lyophilized. Enzymatic Digestions-The S-carboxymethylated enzyme (18 nmol) was dissolved in 0.1 M Tris.HC1 (pH9.5) containing 4 M urea and treated with A. lyticus protease I at an enzyme/substrate molar ratio of 1/400 at 37 "C for 6 h. One-fifth of the digest was submitted directly to HPLC to separate the digested peptides, as described below. Part of the remaining digest was dissolved in 1% NH,HCO, and treated further at 37 "C with S. aureus protease V8 for 12-18 h or TPCK-treated trypsin for 6 hat anenzyme/substrate molar ratios of 1/50-100. Other portions of the digest with A. lyticus protease I were dissolved in 0.1 M Tris. HCl (pH 9.5) containing 2 M urea or in 0.1 M Tris.HC1 (pH 9.0) and digested by arginyl-endoproteinase a t an enzyme/substrate molar ratio of 1/50 at 37 "C for 13 h. Another portion was dissolved in 0.1 M ammonium acetate (pH 6.0) containing 10% (v/v) acetonitrile and digested with endoproteinase Asp-N at an enzyme/substrate molar ratio of 1/50 at 37 "C for 6 h. These digested peptides were separated by HPLC asdescribed below. CNBr Degradation of S-Carboxymethylated TGase-S-Carboxymethyl-TGase (20 nmol) was dissolved in 70% formic acid and treated with CNBr in 300-fold molar excess over Met residues a t 4 "C for 36 h in the dark. The reaction was stopped with lyophilization, and the residue was treated with 1%NH,HCOa a t 40 "C for 4 h. The supernatant was separated from the precipitate by centrifugation. Purification of Peptide Mixtures-Peptides derived by enzymatic digestion or CNBr degradation were separated by HPLC on a reversed-phase column. The HPLC system consisted of two model 510 pumps, a model680 gradient controller, model440 and 441 UV detectors, and a model 820 MAXIMA chromatography work station (Waters, Milford, MA). A Cosmosil 5CI8 P-300 column (4.6 X 150 mm) was used for purification of CNBr peptides of S-carboxymethylTGase. Other enzymatic digests and CNBr cleaved peptides were separated on a YMC R-ODs-5 column (4.6X 250 mm). Peptideswere eluted from columns at 30 "C with a linear gradient of increasing acetonitrile concentration in 0.05% trifluoroacetic acid. The eluates were monitored for absorbance values at 214 and 280 nm. Isolated peptides were dried under reduced pressure, and their amino acid compositions, mass values, and N-terminal sequences were analyzed. ESI and FAB Mass Spectra-The ESI mass spectrum of the purified TGase from Streptouerticillium sp. s-8112 was measured using a Finigan triple quadrupole mass spectrometer equipped with an ESI source. The solution of the TGase (20 pmol/pl) in HCOOH:H20:CH30H (1:50:50) was infused into the ion source at a flow rate of 1 pl min" for 10 min. FAB mass spectra were recorded with a JEOL double-focusing mass spectrometer (JMS-HX100)with an FAB ion source and a data processor (JMA-DA-5000)as described (22). Amino Acid and N-terminal Sequence Analyses-Sampleswere hydrolyzed in 5.7 N HCl a t 110 "C for 24-72 h using a Waters PicoTag'"work station. The hydrolysates were concentrated under reduced pressure, dissolved in 0.02 N HC1, and analyzed in a Hitachi L-8500 amino acid analyzer. N-terminal sequences of peptides were analyzed in an Applied Biosystems model 477A sequenator equipped with a 120A PTH-analyzer. PTH-S-carboxymethyl-Cys was eluted between PTH-Thr and PTH-Gln. C-terminal Analysis-The C-terminal peptide of the purified TGase was identified by the procedure of Rose et al. as described in Ref. 23with some modifications. S-Carboxymethyl-TGaselyophilized with A. Lytc'cus protease I a t a molar ratio of 400:l in 97% "0atomic H2180 was dissolvedin H2180and digested at 37 "C for 7 h. Peptides generated by these treatments were purified by HPLC as described above and subjected to FAB mass analyses. RESULTS
The build-up of the sequence was performed as follows. Initially, the intact enzyme was subjected to ESI mass spectrometry to confirm its molecular weight, which had been deduced from the mobility of the enzyme on SDS-polyacrylamide gel electrophoresis under non-reduced and reduced conditions (19) and to obtain a more accurate value for the molecular weight. The observed mass (37,869.2 f 8.8) in Fig. 1 was consistent with the apparent value determined previ-
ously by SDS-polyacrylamide gel electrophoresis (19). This and the amino acid composition (Table I) showed that the protein contains 1residue of cysteine. The thiol group of this cysteine is free, because the hydrolysate of the protein carboxymethylated under non-reduced conditions gave -0.9 mol of carboxymethyl-Cys on amino acid analysis. Direct Edman degradation of the intact TGase provided the N-terminal21amino acid residues. The S-carboxymethylTGase was then treated with A . lyticus protease I. The HPLC profile of the resulting peptides (named AP1-AP18) is shown in Fig. 2. The amino acid compositions, yields, and observed mass values of these 18 peptides are given in Table 11. FAB mass spectrometry was usedto measure the molecular weights of peptide fragments obtained from the protein, to confirm ambiguous amino acid residues in the sequences of peptide fragments. Among these peptides (APl-AP18), 12 peptides (AP3-APE, AP14, and AP17) were completely sequenced and four peptides (AP13, A P E , AP16, and AP18) were sequenced with the aid of the sequences of their subfragments. The remaining two peptides (API and AP2) consisted of only 2 amino acid residues each including 1 mol of lysine and were not sequenced. The shoulder peaks (AP4', AP9', and AP14') in Fig. 2, eluted immediately after AP4, AP9, and AP14, respectively, were examined by amino acid, FAB mass, and N-terminal sequence analyses (data not shown). These peptides hadthe same sequences as thoseof AP4, AP9, and AP14, respectively, except for having Asp-Gly instead of Asn-Gly in their sequences. This conversion might be an artifact, because on HPLC these shoulder peaks were minor to those of the corresponding peptides. These peptides accounted almost completely for the amino acid composition and the observed molecular weight of the protein. The S-carboxymethyl-TGase was treated with CNBr and the resulting peptides were separated by reversed-phase HPLC, as shown in Fig. 3. The amino acid compositions and yields of six major peptides, named CB1-CB6, are given in Table 111. Peptide CB6 was not found on HPLC (Fig. 3) but was isolated from the insoluble fraction of the S-carboxymethyl-TGase treated with CNBr. CB1 was identified as the N-terminal peptide of TGase by amino acid and FAB mass analyses. The other CNBr peptides (CB2-CB6) were partly sequenced by examining their amino acid compositions, molecular weights, and partial amino acid sequences of their subfragments, which were prepared by treatment with arginylendopeptidase, S. aureus protease V8, or endoproteinase AspN (complete data not shown). The overlapping of the sequences of these CNBrpeptides with those of the A P peptides provided the complete amino acid sequence of the TGase of Streptoverticillium sp. s-8112, as shown in Fig. 4. Forfurther confirmation of the C-terminal amino acid residue of the TGase, S-carboxymethyl-TGase was hydrolyzed with A . lyticus protease I in 97% H2180in this experiment instead of 50% H2"0 used in an experiment of Rose et al. (23). By this treatment l80should be incorporated into newly generated carboxyl groups but not into the original Cterminal carboxyl group of the TGase. Thus the values for the signals of these peptides were expected to be different from those of these peptides generated in normal water. The FAB mass spectra of the digests shown in Fig. 5 indicate that APll did not incorporate an "0 atomduring enzymatic hydrolysis, but other peptides incorporated one or two "0 atoms. Thus clearly APll was the C-terminal peptide of the TGase. The TGase from Streptoverticillium sp. s-8112 consists of 331 amino acid residues in a single polypeptide chain,as shown in Fig. 4. There aretwo potential N-glycosylation sites
d
Streptoverticillium sp. Transglutaminme
I
11567
1
342.5
-I -1
971.9
I
FIG. 1. ESI mass spectrum of intact TGase. Peaks on the spectrum are
3.2
the multiply charged ion signals with mass values ( [ M ,+ nH]"+/n)(where M, is the molecular weight of the protein, n is the charge of the ion, and nH is the mass of protons added to the protein molecule) (24). From each observed value, the average molecular weight of the TGase was calculated to be 37,869.2
1184.5
(+ 8.8 S.D.).
600
1000
1 12b0
1403.4
I
1400
1600
l8bO
m/z TABLEI Amino acid composition of the TGase from Streptoverticillium sp.
APIE
S-8112 Amino acid
Analysis"
Sequenceb
residuesfmolecule
26
ASP 45.2 Asn 21 14 Thr 13.7' 27 Ser 25.7' Glu Gln 10 21 Pro 21.6 Gly 26 Ala 1/2cys 0.3 1 Val 17 Met 5.0 6 Iled 5.1 5 Leud 12.3 12 TYr 15 Phe 15.7 16 17.9 18 LYs His 8.0 8 9 Trp Arg Total 331 "Values are shown as mol/mol Ala, which is calculated as 26.0 mmol in the 24-h hydrolysate. Values are from the sequence data. Values were estimated by extrapolating values for the 24-, 48-, and 72-h hydrolysates to time zero. The 72-h hydrolysate was adopted. e -, not determined.
at AsnZe2,AsnZg7,but these residues were not glycosylated because no saccharidewas detectable. The chemical molecular weight calculated from the sequence is 37,863, which is in good agreement with the mass value (37,869.2 t 8.8) determined from the ESI mass spectrum. The TGase has 1 Cys residue with a free thiol group, which is essential for enzy-
FIG. 2. HPLC profile of A. lytic- protease I peptides derived from carboxymethyl-TGase. An HPLC column (4.6 X 250 mm) was equilibrated with 0.05% trifluoroacetic acid aqueous solution containing 5% acetonitrile and developed at 30 "C with increase of acetonitrile concentration (0.5%/min) after holding for 3 min with the initial solvent. The flow rate was 1.0 ml/min.
matic activity (data not shown), indicating that the TGase is a member of the thiol proteasefamily. DISCUSSION
To obtain information of the structural relationship between TGases from a microbial origin and mammalian species, we determined the complete amino acidsequence of TGase from Streptouerticillium sp. strain s-8112. Recent molecular cloning of cDNAs encoding TGases from various mammalian organs, tissues, and body fluids has resulted in the finding that mammalian TGases can be classified as type I and typeTI. Both types havetwo characteristic features; their active site region includes a Cys residue, and their activation
11568
Transglutaminme sp. Streptoverticillium TABLEI1
Amino acid compositims of A. lytieus protease I peptides derived from CM-TGase Values in parentheses are from the seauence data. ~
AP4
AP3 Amino acid AP2
AP1
AP7
AP8
2.00 (2) 0.97 (1) 0.99 (1)
1.04 (1)
AP9
residueslmolecule
CMC" Asx Thr Ser Glx G~Y Ala Val Met Ile Leu TYr Phe LYS His
Trp Arg Pro Total Position Yield' nmol 13.06 % 71.6 m/z Found Calculated AP18 Amino AP17acid AP16
1.02 (1)
2.95 (3)
1.18 (1)
1.08 (1)
1.00 (1) 1.19 (1) 2.03 (2)
1.05 (1) 3.07 (3)
0.92 (1) 1.00 (1) 0.55 (1)
0.96 (1)
0.93 (1) 1.95 (2) 1.10 (1)
1.00 (1) 0.93 (1)
1.03 0.96 0.95 2.97 3.05 1.96 0.99
(1) (1) (1) (3) (3) (2) (1)
0.72 (1) 0.47 (1) 0.97 (1) 0.76 (1) 1.00 (1)
1.00 (1)
1.00 (1)
1.00 (1)
0.98 (1) 1.00 (1)
1.00 (1)
0.65 (1) 1.00 (1) (1)
-b
0.96 (1)
1.82 (2) 0.87 (1)
1.00 (1) 1.01 (1) - (1) 1.96 (2)
0.96 (1)
1.03 (1) 1.00 (1) 3.76 (4) 0.95 (1)
2
2
4
13
7
6
8
12
19
215-216
326-327
92-95
182-194
115-121
195-200
318-325
38-49
96-114
7.17 39.3
8.10 44.4
15.26 83.6
18.10 99.2
14.96 82.0
9.47 51.9 303.2 303.2 APlO AP15
246.0 246.2 AP14APllAP13
503.1 509.3 .""
1498.2 1497.7
883.1 883.4
680.3 680.3
16.46 90.2 1559.6 1559.8
918.1 918.4
~~~~
~
9.55 52.3 2117.0 2117.1 ~
AP12
-
residuesfmolecule
CMC" Asx Thr Ser Glx G~Y Ala Val Met Ile Leu 5 r Phe LYS His Trp Arg
Pro
Total Position Yield' nmol % m/z
2.06 (2) 3.47 1.02 1.04 1.00
(4) (1) (1) (1)
1.01 (1) 1.00 (1)
2.00 1.79 1.78 1.06 3.01 1.96 1.63 0.90
(2) (2) (2) (1) (3) (2) (2) (1)
0.99 (1) 1.42 (2) 0.94 (1) 1.00 (1) 0.96 (1) - (1) 0.97 (1) 14
-
(1)
1.00 (1) 3.74 (4) - (1)
2.85 (3) 2.03 (2)
2.90 (3)
29
30
53
23
42
153-181
122-151
217-269
295-317
50-91
11.63 63.8
11.88 65.1
14.89 81.6
16.91 92.7
0.98 (1) 25
37
17.45 95.6
15.79 86.5
13.35 73.2
Found 1607.8 4290.1 Calculated 2793.3 487.2 1607.7
1.00 (1)
2.70 (3) 0.88 (1) 1.36 (1) 1.00 (1) 0.90 (1)
4
270-294
487.0
2793.1
0.85 (1) 4.14 (4) 2.64 (3) 2.66 (3) 7.61 (8) 3.15 (3) 2.01 (2) 2.71 (3) 0.79 (1)
2.95 (3) 0.70 (1) 1.99 (2) 1.00 (1)
0.74 (1) 328-331
(3) (1) (2) (1) (2) (1) (1)
5.30 (5)
4.64 (5) 6.18 (6)
201-214
2.95 1.06 1.87 1.22 2.23 1.17 0.78
5.95 (6) 1.00 (1) 2.69 (3) 2.05 (2) 1.08 (1) 3.92 (4)
6.82 (7) 1.85 (2) 1.73 (2) 1.99 (2) 1.01 (1) 1.98 (2) 2.47 (3) 0.85 (1) 0.87 (I) 1.01 (1) 2.46 (3)
1-37 13.85 75.9
-
3253.8 3253.6
1.86 (2) 3.86 (4) 1.70 (1) 4.79 (5) 1.93 (2) 0.63 (1)
3404.1 3404.7
8.62 (9) 2.85 (3) 2.63 (3) 3.03 (3) 6.70 (7) 4.84 (5) 1.88 (2) 0.82 (1) 0.98 (1) 0.85 (1) 1.64 (2) 3.67 (4) 1.00 (1)
1.46 (2) 2.89 (3) 1.00 (1)
1.91 (2) 1.41 (2) 2.75 (3) 1.00 (1)
- (1) 3.75 (4) 6.02 (6)
- (1) 1.99 (2) 1.14 (1)
- (2) 2.81 (3) 0.96 (1)
-
5747.7
1.57 (2)
2768.2 2768.3
13.27 72.7 5034.6d 5034.5
S-Carboxylmethylcysteine. not determined. The yields of peptides were calculated based on the amounts of amino acids determined quantitatively in the acid hydrolysates of the peptides, which were isolated by HPLC from the digest of S-carboxymethylated-TGase,as shown in Fig. 2. *This value was obtained as an average mass.
-.
depends on a Caz+-bindingsequence. The sequences of type I and type I1 TGases are very similar in these two functional domains, butthe overall sequence homologies the of two types are not very high. These similarities of the conserved sequences evenin type I and typeI1 suggest that these sequences
are closely related to the function of the TGases. Comparison of amino acid sequences indicated no overall structural relationship of the TGase of Streptouerticillium sp. with those of mammalian type I and type I1 TGases except in the regions around the single Cys residue. Recently, another TGasefrom
11569
Streptoverticillium sp. Transglutaminme
FIG.3. HPLC profile of CNBr peptides derived from carboxymethyl-TGase. Chromatographic conditions were as for Fig.2.
1 .oo
0.00
3.00
2.00 x IO* m i n u t e s
TABLE111 Amino acid compositions of CNBr peptides derived from reduced and carboxymethylated TGase Values in parentheses are from sequence data. Amino acid
CMC" Asx
Thr Ser Hse Glx GlY Ala
Val Met
CB1
CB2
CB3
3.76 (4) 0.95 (1) 1.05 (1) -b (1) 1.43 (1)
5.14 (5) 1.02 (1) 4.60 (5) - (1) 1.54 (1) 4.17 (4) 4.00 (4) 1.86 (2)
13.98 (14) 1.11 (1) 4.65 (5) - (1) 5.38 (5) 3.81 (3) 7.00 (7)
1.00 (1) 1.10 (1)
Ile
Leu TYr Phe
1.03 (1)
LYS
235-288
His Trp Arg Pro 41 Total 134-193 Position 194-234 Yield'
- (0) 1.73 (2) 2.62 (3) 16
5.95 (6) 1.77 (2) 2.16 (2) 3.10 (3) 2.04 (2) - (0) 5.83 (6) 2.93 (3) 60
117
CB5
7.72 (8) 3.07 (4) 4.22 (4) - (1) 3.12 (3) 6.72 (7) 5.00 (5) 2.19 (2)
4.40 (5) 1.63 (2) 3.12 (4) - (0) 3.09 (3) 3.00 (3) 2.00 (2) 2.34 (3)
1.00 (1) 1.29 (1) 1.92 (2) 2.82 (3) 2.17 (1) 2.42 (3) - (2) 3.25 (3) 3.76 (4) 54
1.38 (2) 2.26 (3) 2.37 (3) 3.08 (4) 1.03 (1) - (3) 2.03 (2) 2.13 (3) 43
0.81 (1) 11.58 (11) 4.60 (5) 7.04 (8) - (2) 17.81 (20) 8.86 (9) 7.00 (7) 8.10 (9) 1.13 (1) 2.94 (3) 4.85 (6) 5.81 (6) 5.72 (6) 1.68 (1) - (3) 12.03 (14) 5.23 (5)
1-16
nmol 4.28 27.7 21.4 % S-Carboxymethylcysteine.
3.06 59.5
0.67 (1) 1.04 (1) 1.43 (2) 2.12 (2) 3.91 (4) 1.11 (1) - (1) 2.92 (3) 2.80 (3)
CB6 CB4 residues/molecule
5.53 15.3
36.3
1.71 8.6
-, not determined. The yields of peptides were calculated based on the amounts of amino acids determined quantitatively in the acid hydrolysates of the as shown in Fig. 3. peptides, which were isolated by HPLC from the CNBr-treated S-carboxymethylated TGase, a microbial source (P.polycephalum) has been purified and identified as a Ca2+-dependent enzyme (21), indicating that in this enzyme both the catalytic site and the Ca2+-binding site are similar to those of mammalian TGases and that this enzyme is a different type from the TGase of Streptouerticillium sp., although no sequence information on the TGase from P. polycephalum is yet available. The amino acid sequence of the TGase of Streptoverticillium sp. also bears no significant similarity to any othersequence in current protein sequence data bases. The type I and I1 TGases characterized so far have high sequence homology to each other in the vicinity of their putative active Cys residue, e.g. a conserved sequence motif, Tyr-Gly-Gln-Cys-Trp, containingthe putative active site Cys residue of these enzymes to catalyze the acyl transfer reaction,
as shown in Fig. 6. Interestingly, the residues around this Cys residue are similar to those in a family of thiol proteinases represented by papain and cathepsin (Fig. 6). This Cys is concluded to play a role in catalysis in an acyl transfer reaction because erythrocyte band 4.2 protein (E),in which this Cys residue in the conserved region is replaced by Ala, shows no TGase activity. Streptoverticillium TGase catalyzes an acyl transfer reaction, but itscatalytic activity is lost when its sole Cys at position 64 is S-1,2-dicarboxyethylatedby treatment with N-ethylmaleimide (data not shown), suggesting that thisthiol group of Cysa is also essential for enzymatic activity and that the physicochemical environment around this Cys residue is similar to those of other TGases, although the amino acid sequence around Cysa is quite different from those of mammalian TGases.
11570
FIG. 4. Complete amino acid sequence of the TGase of Streptoverticillium sp. s-8112. Arrowheads indicateresiduesdetermined by Edman degradation. Amino acid residues are shown in single-letter code. RCM-T, reduced and S-carboxymethylated TGase; A P ,A . lyticus protease I peptides of RCM-T; AP-CB, CNBr peptides of AP; A P - T , tryptic peptides of AP; A P - V , S. aureus protease V8 peptides of AP CB, peptidesderived from reducedand Scarboxymethylated TGase by its treatment with CNBr; CB-D, endoproteinase Asp-N peptides of CB; CB-R, arginylendopeptidase peptides of CB. The Cys residue witha free thiolresidue is shown by an open star.
From secondary structural analysis (Fig. 7), the contentsof a-helix and @-sheet arepredicted to be 22 and 31%, respectively, which are not inconsistent with the results of circular dichroism analysis of this enzyme (a-helix 21% and @-sheet 43%) (25). This suggests that this enzyme hasa globular conformation, which is typical of secretory proteins. Hydropathy analysis of the sequence of the TGase of Streptouerticillium sp. (Fig. 7) indicated that this enzyme is rather hydrophilic as a whole, but with several hydrophobic regions interspersed along the sequence. These hydrophobic regions seem to be settled inhydrophilic areas on the surface of the enzyme molecule. The single cysteine, Cysa, which is essential for the catalytic activity, is predicted to be located in a @-turnconnecting a-helix and @-sheet structures (Fig. 7), as in mammalian TGases such as subunit a of blood coagulation factor XI11 and guinea pig liver TGase, as shown in Fig. 8. Thus, the secondary structural environment of the active site Cysa residue in the TGase of Streptouerticillium sp. is probably similar to those of the active Cys residues of other mammalian enzymes, even though the amino acid sequence in this region of the Streptouerticillium TGase has no significant homology to those of the mammalian enzymes.
The putative calcium-binding domain of factor XIIIa (amino acid residues 468-479) has been suggested to be highly conserved in human, mouse, and guinea pig tissue TGases (type II-human, -mouse, and -guinea, respectively, in Fig. 6) (14). However, the TGase from Streptouerticillium sp. has no sequence homologous with the calcium-binding domains of other TGases. This is consistent with the fact that theTGase from Streptouerticillium sp. catalyzes a Ca2+-independentacyl transfer reaction, unlike the other TGases. Recently, Polakowska et al. (28) constructed aphylogenetic tree of TGases based on nucleotide and amino acid sequences. They showed that type I and type I1 TGases evolved from a common ancestral gene and diverged by subsequent gene duplication and acquisition and/or loss of a part of the sequence. Human erythrocyte membrane protein 4.2 came later into the type I1 family. Protein 4.2 is deduced to become inactive by substitution of the putative active site Cys. These data suggest that the Streptouerticillium TGase mayhave evolved along a different pathway from the Ca2+-dependent TGases and that the enzyme acquired catalytic activity by replacement of some amino acid residue(s) by the sole Cys or a cassette amino acid sequence in a similar environment in a
Streptoverticillium sp. Transglutaminme
IS90
1540
FIG. 5. FAB mass spectra of severa1 A. lyticus protease I peptides derived from carboxymethyl-TGase digested in Ha180. Arroros indicate the positions of unlabeled peptides.
11571
0
I5
9 2 0
5 3440
1000
800
N
"
"""_~""""""_____________________~~"-~""~""~""~"-
*
TGasea
W L S Y G C Y G V T F Y N S G Q ~ - P T N R
F a c t o r Xllla ( 5 )
V K Y G Q C Y V F A G V F N T F L K C L G
Type I - h u m a n ( 7 . 8 ) Type I-rat ( 7 ) T y p e I-rabbit ( 9 )
V P Y G Q C U V F A G V T T T V L R C L G
Type 1 1 - g u i n e a ( 6 ) Type 1 1 - h u m a n (10) T y p e 11-mouse ( 1 0 ) Type 11-bovine ( 1 1 )
V R Y G Q C W V
Band 1 . 2 ( 1 2 )
V Y D G Q A W V L A 4 V A C T Y L R C L G
Papain
G S C G S C F A F S A V V T I E G I
Cathepsin H
G A C G S C W T F S T T G A L E G A
F .A A . V A C T
V
L R C L G
-A
a : T G a s e f r o m Streptoverticillium s p . s - 8 1 1 2
FIG.6. Comparison of amino acid sequences in theputative active siteregions of the TGase of Streptoverticillium sp. and mammalian TGases. The sequences are from the following references: human factor XIIIa (Factor XIIIa) (5-71, human TGase-K (Type I-human) (9-12), rat keratinocyte TGase (Type I-rat) (9), rabbit cultured keratinocytes (TypeI-rabbit) (13), guinea pig liver TGase (Type 11-guinea) (B), human endothelial TGase (Type IIhuman) (14), mouse macrophage TGase (Type 11-mouse)(14),bovine endothelialTGase (Type 11-bouine) (E),and humanerythrocyte membrane protein (Band4.2) (16-18).
0
53
ea
1
158 ZOO Residue Number
253 358 308
FIG.7. Secondary structure and hydropathy plot of the amino acid sequence of the TGase from Streptoverticillium sp. The secondary structure was predicted by the method of Robson et al. (26). The symbols m ,m , and u indicate a-helix, 8-sheet, and p-turn, respectively. The hydropathic index was calculated at a span of 9 amino acid residues according to the method of Kyte and Doolittle (27).
n
protein with a quite different sequence from those of type I and type I1 TGases. Growing cells Streptouerticillium sp. secrete TGase intothe culture medium. The purified enzyme undoubtedly catalyzes the acylation of putrescine by glutamine under the in vitro reaction conditions used for monitoring the catalytic activity of TGases. Mammalian TGases function in network formation in tissuessuchas the erythrocyte membrane, semen coagulation, and cornification of epidermal cells. A similar function in the differentiation of vegetative plasmodia to spherules has been suggested for the TGase of P. polycephalum (29). However, the function of TGases in microorganisms such as Streptoverticillium sp. is unknown. The complete amino acid sequence of the microbial TGase determined in this work will allow production of a recombinant protein and study of the function and processing of the enzyme in
-3.5'"""""""" -40 -30 -20
-10
0
10
20
30
40
Relativeposition
FIG.8. Comparison of the hydropathy profile around the active site Cys (0 at relative position) at position 64 of the TGase (-) from Streptoverticillium sp. with those of mammalian TGases (guinea pig liver TGase, - - -; human factor XIIIa, - . .-) and papain, (- - -).
-
Streptouerticillium sp. Transglutaminase
11572
Streptouerticillium sp. Furthermore, elucidation of the twoand three-dimensional structures of this enzyme may shed light on its catalytic functions common with those of mammalian enzymes. Acknowledgments-We are grateful to Dr. Hiroko Toda (Division of Protein Chemistry, Institute for Protein Research, Osaka University) for assistance in sequence analysis. We thank Amano Pharmaceutical Co. for providing information onmicrobial transglutaminase. We are also indebted to Aiko Kobatake and Yukiko Umemura for preparation of the manuscript. REFERENCES 1. Folk, J. E., and Finlayson, J. S. (1977)Adu. Protein Chem. 31, 1-333 2. Lorand, L.,.Losowsky, M. S., and Miloszewski, K. J. M. (1980)Prog. Henwstasts Throd. 6,245-290 3. Chung, S. L. (1972)Ann. N. Y. Acad. Sci. 202,240-255 4. Folk, J. E. (1980)Annu. Reu. Eiochem. 49,517-531 5. Ichinose, A:, Hendrickaon, L. E., Fujikawa, K., and Davie, E. W. (1986) Eiochemtstry 26,6900-6906 6. Takahashi, N.,Takahashi, Y., and Putnam, F. W. (1986)Pmc. NatL Acad. Sci. U. S. A. 83,8019-8023 7. GNndmEnn, U.,,Amann, E., Zettlmeissl, G., and Kupper, H. (1986)Proc. Natl. Acad. Sa. U. S. A. 63,8024-8028 8. Ikura K Nasu T Yokota, H., Tsuchiya, Y., Sasaki, R., and Chiba, H. (ldi)”siocheh&~ry 27,2898-2905 9. Phillips, M.A., Stewart, B. E., gin, Q., Chakravarty, R.,,Floyd, E. E., Jetten, A. M., and Rice, R. H. (1990)Proc. Natl. Acad. Scr U. S. A. 87, 9333-9337 10. Kim, H.C., Idler, W. W., Kim, I. G., Han, J. H., Chung, S., I., and Steiner, P. M. (1991)J. EioL Chem. 268,536-539
11. Polakowska, R. R., Herting, E., and Goldsmith, L. A. (1991)J. Invest. Dermatoi. 96,285-288 12. Yamanishi, K., Liew, F. M., Konishi, K., Yasuno, H., Doi, H.,Hirano, J., and Fukushima, S. (1991)Eiochem. Biophys. Res. Commun. 175, 90& 913 13. Floyd, E. E., and Jetten, A. M. (1989)Mol. Cell. Biol. 9,4846-4851 14. Gentile, V., Saydak, M., Chiocca, E. A., Akande, O., Birckbichler, P. J., Lee, K. N., Stein, J. P., and Davtes, P. J. A. (1991)J. Bml. Chern. 286, 478-483 15. Nakanishi, K., Nara, K., Hagiwara, H., Aoyama, Y., Ueno, H., and Hirose, S. (1991)Eur. J. Eiochem. 202.15-21 16. Korseren. C.. Lawler. J.. Lambert. S.. Soeicher,. D... and Cohen, C. M. (1990) P&. Nati Acad. Sci.‘ U.S. A. 87,‘6i3-617 17. Sung L. A., Chien S., Chang, L.-S., Lambert, K., Bliss, S. A., Bouhassira, E. E., Nagel, R. i..Schwartz. R. S., and Rybicki, A. C.(1990)Proc. Natl. Acad. Scir U. S. A. 87,955-959 Korsmen. C.. and Cohen. C. M. (1991) IS. . . Proc. Natl. Acad. Sci. U. S. A. 88, 4l3.lo-4844’ 19. Ando, H., Adachi, M., Umeda, K., Matsuura, A., Nonaka, M., Uchio, R., Tanaka H., and Motoki, M. (1989)Agrie. Biol. Chem. 63,2613-2617 20. Nonaka, k.,Tanaka, H., Okiyama,.A., Motoki, M., Ando, H., Umeda, K., and Matsuura, A. (1989)Agrre. BzoL Chem. 63,2619-2623 21. Klein, J. D., Guzman, E., and Kuehn, G. D. (1992)J. Bocteriol. 174,25992605 22. Takao T. Hitouji T. Shimonishi Y., Tanabe, T., Inouye, S., and Inouye, M.(1964)J. Bidl. dhem. 259,6105-6109 23. Rose, K., Savoy, L.-A., Simona, M., Offord, R. E., and Wingfeld, P. (1988) Bwchem, J. 260,253-259 24. Fenn, J. B.,,Mann, M., Meng, C. K., Wong, S. F., and Whitehouse, C. M. (1989)Sclence. 246,64-71 C. T., Wu, C.-S. C., and Yang, J. T.(1978)A d . Biochem. 91, 1325. Ch:ng, 01
26. Robson, B., and Suzuki, E. (1976)J. Mol. BWL 107,327-356 27. Kyte J. and Doolittle, R. F. (1982)J. Mol. Ewl. 157, 105-132 28. Polaio&ska, R. R., Eickbush, T., Fdciano, V., Razvi F., and Goldsmith, L. A. (1992)Proc. Natl. Acad. Sci. U. S. A. 89,2596-2605 29. Rush, H: P. (1969)Fed. Proc. 28, 1761-1779
