Biotechnology Letters

, Volume 30, Issue 6, pp 1025–1029

A chemically modified glass surface that facilitates transglutaminase-mediated protein immobilization


  • Yusuke Tanaka
    • Department of Applied Chemistry, Graduate School of EngineeringKyushu University
  • Satoshi Doi
    • Department of Applied Chemistry, Graduate School of EngineeringKyushu University
    • Department of Applied Chemistry, Graduate School of EngineeringKyushu University
    • Center for Future ChemistryKyushu University
  • Noriyuki Kawata
    • New Products & Buisiness Development DepartmentNippon Sheet Glass Co., Ltd.
  • Shinji Kamiya
    • New Products & Buisiness Development DepartmentNippon Sheet Glass Co., Ltd.
  • Kenichi Nakama
    • New Products & Buisiness Development DepartmentNippon Sheet Glass Co., Ltd.
  • Masahiro Goto
    • Department of Applied Chemistry, Graduate School of EngineeringKyushu University
    • Center for Future ChemistryKyushu University
Original Research Paper

DOI: 10.1007/s10529-008-9656-y

Cite this article as:
Tanaka, Y., Doi, S., Kamiya, N. et al. Biotechnol Lett (2008) 30: 1025. doi:10.1007/s10529-008-9656-y


An amino-modified glass surface for enzymatic protein immobilization by microbial transglutaminase (MTG) was developed. Diamine substrates with secondary amino groups in the linker moiety, like triethylenetetramine (TETA), exhibited at most a 2-fold higher reactivity in the MTG-catalyzed reaction compared to those with the alkyl linker. A 96-well glass plate was subsequently modified with selected diamine substrates. Validation of the modified surface by enzymatic immobilization of enhanced green fluorescent protein tagged with a glutamine donor-substrate peptide (LLQG) of MTG revealed that the protein loading onto the TETA-modified glass surface was approximately 15-fold higher than that on the unmodified one.


96-Well glass platePeptide tagProtein immobilizationSite-specific protein modificationTransglutaminase


Protein microarrays have facilitated the high-throughput analysis of complicated biological events, such as protein–protein, antigen–antibody, and protein–small molecule interactions, and have been recognized as a promising format in bioscience and biotechnology (MacBeath et al. 2000; Phizicky et al. 2003). Although many protein immobilization techniques have been developed (Rusmini et al. 2007), functional protein immobilization has been a major challenge because of the difficulty in maintaining the conformation and biological activity of the immobilized proteins.

To achieve robust immobilization of proteins under protein-friendly conditions, we have focused on the capability of microbial transglutaminase (MTG) for use in site-specific protein cross-linking reactions. MTG is an enzyme that catalyzes an acyl transfer reaction between the γ-carboxyamide group of peptide- or protein-bound Gln residues and a variety of primary amines including the ε-amino group of Lys residues with a loss of ammonia (Yokoyama et al. 2004). We have recently demonstrated the potential of MTG in the peptide tag-directing covalent and site-specific immobilization of proteins onto a polystyrene surface physically coated with a substrate protein of MTG, and found that a substantial quantity of proteins can be immobilized on a casein-adsorbed polystyrene surface (Kamiya et al. 2005; Tanaka et al. 2007; Kamiya et al. 2007).

Here, we report on enzymatic protein immobilization on an amino-modified glass surface. Since MTG accepts a wide range of primary amines, we have developed a glass surface displaying a primary amine for MTG-mediated protein immobilization. For amine substrates to be displayed on a 96-well glass plate, we found a diamine substrate for which MTG has a high catalytic activity. In order to evaluate the immobilization efficiency using fluorescence measurements, enhanced green fluorescent protein (EGFP) was chosen as a base and its recombinant protein with the extension of a Leu-Leu-Gln-Gly sequence (LLQG-tag) at the C-terminus (Tanaka et al. 2007) was employed as a model protein. We found that the MTG-mediated immobilization of the recombinant EGFP on an amino-modified glass surface was possible, and the immobilization efficiency was highly dependent on the type of amine substrates deposited on the surface.

Materials and methods

Materials and apparatus

MTG was provided by Ajinomoto Co., Inc. (Japan). A 96-well amino-coated glass plate was provided by Nippon Sheet Glass Co., Ltd. (Japan). Specific reagents used were as follows: N-benzyloxycarbonyl-l-Gln-Gly (Z-QG) and diethylenediamine (Sigma-Aldrich, USA), ethylenediamine (Kishida Chemicals, Japan), triethylenediamine and hexamethylenediamine (Wako, Japan), and octamethylenediamine (Tokyo Kasei, Japan). The other reagents were of commercially available analytical grade. The HPLC apparatus used was a LC-10AT (Shimadzu, Japan) with an Inertsil ODS-3 column (GL Sciences Inc., Japan). The fluorescent intensity on the glass surface was measured using a Molecular Imager FX Pro (Bio-Rad, USA).

MTG-catalyzed incorporation of a diamine into Z-QG

The reaction solution was composed of 20 mM Z-QG, 50 mM diamine in 100 mM Tris/HCl (pH 7.5). The reaction was initiated by the addition of 100 μl MTG (10 U/ml) into 900 μl of the reaction solution, and the mixture was incubated at 40°C. Aliquots of 10 μl of the reaction mixture solution were removed after 10, 20, 30, 60 and 120 min, and the enzymatic reaction was terminated by the addition of 90 μl 0.1% trifluoroacetic acid (TFA). The terminated reaction mixture was subjected to reverse-phase HPLC using an Inertsil ODS-3 column. The solvent system consisted of 0.1% TFA aqueous solution (A) and 0.1% TFA acetonitrile solution (B). The components of the terminated reaction mixture were eluted with a liner gradient of 10% B to 80% B at 1 ml/min over 30 min. The substrate and product elution was monitored at 258 nm. The conversion rate was calculated from the peak area of Z-QG using a Z-QG calibration curve.

Preparation of an LLQG-tagged recombinant protein

Recombinant EGFPs employed in this study are listed in Table 1. An EGFP mutant with a C-terminal extension of the LLQG-tag was prepared according to a previous report (Tanaka et al. 2007). The resultant recombinant protein is abbreviated as CQ-EGFP.
Table 1

Amino acid sequences of the recombinant proteins employed in this studya


Amino acid sequences of N- and C-terminal regions

Wild-type EGFP




aAmino acid sequences of N- and C-terminal regions are shown with one-letter codes. The substrate peptide sequence for MTG is underlined. Amino acids substituted by site-directed mutagenesis are shown in bold letters. The omitted regions have the same sequence as wild-type EGFP

Preparation of an amino-coated glass surface

A 96-well amino-coated glass plate was modified by a stepwise procedure based on a previous report (Lee et al. 2005). Firstly, 0.5 M succinic anhydride in DMF was added to each well of a 96-well amino-coated glass plate and the plate was left for 16 h with gentle shaking at room temperature. After washing with DMF three times, 0.8 M N, N′-dimethylcarbodiimide (DIC) and 0.8 M N-hydroxysuccinimide (NHS) in DMF were added to each well and the plate was left for 6 h with gentle shaking at room temperature. After washing with DMF three times, 1 M diamine in DMF was added to each well and the plate was left for 16 h with gentle shaking at room temperature. Finally, the wells were thoroughly washed with distilled water nine times and then dried with N2 gas.

Immobilization of CQ-EGFP and wild-type EGFP onto an amino-modified glass surface

A typical procedure for the immobilization of EGFPs onto the modified glass plate was carried out as follows. Aliquots, 90 μl, of a stock solution containing CQ-EGFP or wild-type EGFP in 5 mM Tris/HCl (pH 7.0) were placed in each well of the modified glass plate. Enzymatic immobilization was initiated by the addition of 10 μl of a stock solution of MTG. The fluorescence derived from the immobilized EGFPs on the surface was measured by adding 100 μl of 5 mM Tris/HCl (pH 7.0) in each well. Excitation was at 488 nm and the fluorescence was collected with a 530 nm band pass filter using a Molecular Imager FX Pro.

Results and discussion

Substrate specificity of MTG for diamines

The type of primary amines that are displayed on a glass surface are likely crucial for efficient MTG-mediated protein immobilization. Here we focused on diamines to facilitate the subsequent modification of the glass surface. Although a variety of primary amines are known as MTG substrates (Ohtsuka et al. 2000a; Yokoyama et al. 2004), little information was available for diamine substrates of MTG. Thus, the reactivity of diamines with different linker structure and the length, such as ethylenediamine (EDA), hexamethylenediamine (HMDA), octamethylenediamine (OMDA), diethylenetriamine (DETA), and triethylenetetramine (TETA) were tested as substrates in the MTG-catalyzed reaction. For the Gln-donor substrate, a typical transglutaminase substrate, Z-QG, was employed (Ohtsuka et al. 2000b).

Figure 1 depicts the time course of the incorporation of each diamine into Z-QG by MTG. Despite the presence of two reaction sites in the diamine molecule, the HPLC chromatogram gave only one major product peak up to 2 h, suggesting that the formation of a 1:1 Z-QG/diamine cross-linked product proceeded under the experimental conditions. All alkyldiamines (EDA, HMDA and OMDA) showed comparable initial activities, but OMDA exhibited a slightly higher conversion rate at 2 h compared with EDA and HMDA. The results prompted us to test other diamines with different linker structures between the terminal amino moieties. To this end, DETA and TETA, both with secondary amino groups between the terminal amino groups, were chosen for testing. Interestingly, MTG was more catalytically active towards these two diamines. In comparison, between TETA and OMDA, the former accelerated the reaction about 2-fold, and the reaction was almost finished in 1 h (Fig. 1).
Fig. 1

Time course of Z-QG conversion with each diamine; open circles, ethylenediamine (EDA); closed squares, hexamethylenediamine (HMDA); closed triangles, octamethylenediamine (OMDA); open squares, diethylenetriamine (DETA); open triangles, triethylenetetramine (TETA). The reaction solution contained 20 mM Z-QG, 50 mM diamine, and 1 U MTG (pH 7.5) and was carried out at 40°C

We reasoned that the diamine substrates having the secondary amino-groups are better substrates for MTG because of the charge at the enzyme active site. The putative amine-substrate binding site of MTG consists of negatively charged amino acids (1D, 3D, 4D, 249E, 255D, and 300E) (Kashiwagi et al. 2002). Therefore, a positively charged amine substrate is preferred to a neutral amine. Thus, TETA is a more favorable substrate than OMDA, even though the molecular length of them would be comparable.

Immobilization of CQ-EGFP onto glass surfaces modified with different diamines

Based on the results in Fig. 1, we selected EDA, DETA, and TETA for glass surface modification. Including an unmodified base plate, four amino-modified glass plates were prepared (Fig. 2a). Then, CQ-EGFP (see Table 1) was immobilized onto the amino-modified glass surfaces by a MTG-catalyzed reaction (Fig. 2b). As controls, wild-type EGFP (Table 1) was also immobilized onto the TETA-modified plate using the same protocol. Figure 3a shows the fluorescent intensity of each plate derived from the immobilized EGFPs. The fluorescent intensity reflects the quantity of the immobilized protein retained on the plate. In the case of unmodified (base) and EDA-modified plates, although positive signals were obtained in the presence of MTG, comparable signals were also evident in the absence of MTG. On the other hand, the fluorescent intensity of CQ-EGFP immobilized onto DETA- and EDTA-modified plates was significantly higher than background. We assumed two possibilities to explain the phenomena.
Fig. 2

Scheme of MTG-mediated protein immobilization. (a) The surface modification procedure of amino-modified glass plates. (b) The MTG-mediated covalent and site-specific immobilization of CQ-EGFP onto the modified glass surface
Fig. 3

Comparison of fluorescent intensity of the immobilized EGFPs on glass plates modified with different diamines. (a) CQ-EGFP (10 μg/well) was immobilized onto the base, EDA-, DETA-, and TETA-modified plates without MTG (open bars) or with MTG (filled bars), respectively. The protein immobilization was conducted at 4°C for 6 h in the presence or absence of MTG (0.07 U/well). After the reaction, all the plates were washed six times each with TBST and 1 M NaCl, then subjected to fluorescent measurements. (b) As controls, wild-type EGFP was immobilized to the TETA-modified plate without MTG (open bar) or with MTG (filled bar). The error bars represent the standard deviation (n = 3). The background signal of the amino-modified plates was subtracted in each case

The first explanation considers the spatial requirement for MTG-mediated catalysis at the surface of the glass plate. Given that the modified diamines stand perpendicular to the surface, the molecular length from the surface to a terminal amine was estimated to be ca. 12.7 Å, 22.5 Å, 26.3 Å and 30.1 Å for the base plate, EDA-, DETA-, and TETA-modified plates, respectively. Therefore, in the MTG-mediated protein immobilization, a longer linker between the surface and the terminal primary amines was preferred because of the accessibility of a proteinaceous cross-linker (i.e., MTG) to reaction sites on the surface. The second explanation considers the intrinsic difference in reactivity of the diamines as was observed in Fig. 1. Although the steric component cannot be eliminated, the significant difference in the quantity of immobilized protein between the EDA- and DETA-modified plates was not predicted based on the small differences in their linker lengths (ca. 4 Å). This implies that the intrinsic reactivity of the immobilized substrate is very important to the overall reactivity of the enzyme. Overall, the TETA-modified surface is considered to be the most suitable to facilitate MTG-mediated protein immobilization among the diamines tested. Importantly, immobilization of wild-type EGFP on the TETA-modified plate (Fig. 3b) confirmed that without the LLQG-tag, the MTG-mediated protein immobilization did not proceed well. This supports the idea that EGFP was directly linked to the surface amino group through the substrate peptide tag of MTG (LLQG-tag).


A glass surface modified with MTG-reactive diamine substrates works well for MTG-mediated protein immobilization. Both enzymatic spatial requirements and substrate specificity are critical factors to consider in order to achieve direct protein immobilization by MTG on a modified glass surface. In preparing a solid support for protein immobilization, complicated procedures and specific regents are normally employed (Duckworth et al. 2006; Kwon et al. 2006; Watzke et al. 2006). Instead, as reported herein, simple chemical modifications of amino glass surfaces with diamines can have potential merit for MTG-mediated protein immobilization.


We are grateful to Ajinomoto Co., Inc. for providing the sample of MTG. The present work was supported by the Industrial Technology Research Grant Program in 2006 from the New Energy and Industrial Technology Development Organization (NEDO) of Japan (to N.K.).

Copyright information

© Springer Science+Business Media B.V. 2008