Covalent Attachment of Oligonucleotides to Cellulose Acetate Membranes

Abstract

During the last decade, DNA has become an increasingly important biomolecule in several areas. DNA technology has found many applications, e.g., in forensic science, environmental studies, diagnosis and archeometry. DNA microarrays and DNA biosensors applying the principle of immobilization of oligonucleotide on solid supports are used in these areas. DNA immobilization can be performed by adsorption and covalent attachment. In this study cellulose acetate was used as a solid support for oligonucleotide immobilization. Cellulose acetate was activated with 1,1′-Carbonyldiimidazole (CDI) and then coupled with 1,6-hexanediamine (HDA) as a linker. A hexadecadesoxy oligonucleotide was also activated by 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) and immobilized on the membrane by coupling via amino groups. The effects of various parameters on the immobilization oligonucleotide were investigated.

• 1,1′-Carbonyldiimidazole (CDI)
• Cellulose acetate membrane
• Covalent attachment
• Oligonucleotide immobilization

Introduction

Deoxyribonucleic acid is one of the most important molecules in living organisms and can be regarded as a naturally occurring and highly specific functional biopolymer. DNA diagnostic methods have become indispensable tools in molecular biology and biotechnology studies. The detection of specific base sequences in human, viral, and bacterial nucleic acids is becoming increasingly important in several areas, with applications ranging from the detection of disease-causing and food-contaminating organisms to forensic and environmental research (Pividori et al., [2000]).

Conventional methods for the analysis of specific gene sequences are based on direct sequencing or DNA hybridization methods. The various hybridization methods are based on immobilization of DNA onto a solid phase, typically nitrocellulose or nylon membrane. This immobilization has the disadvantage that the DNA molecules are bound noncovalently and they are attached to the solid phase at multiple sites. Immobilization can be realized by passive adsorption, by UV light or by covalent attachment of modified DNA molecules (Henke et al., [1997]; Niemeyer et al., [1999]; Rogers et al., [1999]; Sen-fang et al., [1998]). Adsorption is the simplest method to attach biomacromolecules to surfaces, since no reagents or special molecular modifications are required. Covalent attachment, on the other hand, is the most ideal method for DNA immobilization since hybridization can take place without being removed from the substrate. This type of attachment was first described by Gilham, where DNA molecules were activated by carbodiimide and attached to cellulose paper via the 5′-end terminal phosphate group (Gilham, [1974]). Carbodiimide mediated binding of DNA to dextran supports, latex particles, controlled porous glass, magnetic polystyrene beads were also described (Ghosh and Musso, [1987]; Rasmussen et al., [1991]). The simplicity and convenience of chemical methods have led to the attachment of DNA to solid supports through stable covalent linkages.

Solid carriers for covalent attachment are variable and easy to use so that immobilization of DNA on different kinds of supports has become an important tool in recent years (Moser et al., [1997]). The advantages of surface-bound DNA include facile purification, conservation of material and reagents, reduction of interference between oligonucleotides and facilitated sample handling. Ideal support characteristics are surface flatness and homogeneity, control of surface properties, thermal and chemical stability, reproducibility, and amenability to DNA immobilization (Strother et al., [2000]).

Nucleic acid layers combined with electrochemical transducers produce DNA biosensors. New biosensor technologies have been intensively investigated because of their potential for rapid, low-cost DNA testing. These sensors have been based on electrochemical, optical, mass sensitive, and acoustic wave transducers (Junhuı et al., [1997]; Marazza et al., [1999]; Walsh et al., [2001]). These technologies relied on the immobilization of ssDNA onto different supports that are combined with different physicochemical transducers.

In this study, cellulose acetate membrane was used as a carrier matrix for the immobilization of oligonucleotide cellulose acetate, a natural polymer which have many advantageous properties such as: easy preparation, biocompatibility towards all biomolecules, both hydrophilic and hydrophobic properties, and solubility in a wide variety of solvents (Garnier et al., [1998]). The aim of this study is the optimization of membrane characteristics and reaction conditions for the immobilization of oligonucleotide on cellulose acetate membrane by attachment via 5′-phosphate terminal groups.

Materials and Methods

Materials

Cellulose acetate (acetyl content approx. 40%), 2,4,6-trinitrobenzene sulfonate (TNBS)[5% v/v aqeous solution], 1,1′-carbonyldiimidazole (CDI), 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (EDC), 1-methylimidazole (MeIm₇), 1,6-hexanediamine, and boric acid were obtained from Sigma Chemical Co. (St. Louis, CA), the oligonucleotide 5′-ACGCATGCGTGGAACC-3′ was purchased from Genemed Synthesis Inc. (San Francisco, CA).

Preparation of Activated Cellulose Acetate Membranes

Glass plates (76 × 26 mm) were cleaned with 65% nitric acid and then washed 2–3 times with deionized water (Garnier et al., [1998]). Cellulose acetate was activated as described below. Briefly, cellulose acetate (1 g) was dissolved in 15 mL dioxane and then added a 1:1 molar excess of CDI, and the reaction mixture was stirred for 2 h at room temperature. The mixture was then poured onto glass plates as a thin film and the excess of CDI was washed from the membrane with deionized water (Bethell et al., [1979]; Önal and Telefoncu, [2003]).

For the determination of activation efficiency the activated cellulose acetate membranes were reacted with glycine (in MeIm₇ buffer) and unreacted glycine was estimated by a modified TNBS assay (Mastuı et al., [1996]). Briefly, a 25 µL sample (washing solutions) and 500 µL of a borate buffer (pH 9.1) were mixed and then 25 µL of 0.1 M TNBS in 0.1 M Na₂HPO₄ was added. After incubation at 37°C for 20 min, 2 mL of 0.1 M Na₂SO₄ in a 0.2 M NaH₂PO₄ solution was added and the absorbance at 416 nm was measured.

Coupling of Linker and Glycine to Cellulose Acetate Membranes

1,6-Hexanediamine (HDA) was used as a linker exposing free amino groups on cellulose acetate membranes. The activated cellulose acetate membranes were treated with HDA solutions (in MeIm₇ buffer) at dark approx. 10 h at room temperature. The excess of HDA was washed off with deionized water. A modified TNBS assay was used to determine the amount of amino groups on the membrane.

Activated cellulose acetate membranes were treated with different amounts of glycine and HDA in order to obtain cellulose acetate membranes of varying polyionic character. For the determination of bound glycine on the membranes, it was calculated indirectly by the determination of glycine in washings by using the ninhydrin test (Robyt and White, [1990]). HDA and glycine can be assayed both with TNBS, and only glycine can be determined with the ninhydrin assay. Briefly, Reagent A (2 g ninhydrin and 60 mg hydrindantine dihydrate in 10 mL methanol, the mixture becomes clear in an ultrasonic bath after 10–15 min and can be used after 8 h) and Reagent B (acetate buffer pH 5.5) were prepared. These two reagents were mixed 1:9(v/v) before use. 1 mL sample was treated with 2 mL of ninhydrin reagent A + B. The mixtures were warmed to 100°C in a water bath for 10–20 min. After cooling, the absorbance was measured at 570 nm.

Activation and Immobilization of Oligonucleotide

Oligonucleotide molecules were activated by 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC) via their 5′-phosphate terminals (Ghosh and Musso, [1987]). Briefly, 3 pmol of oligonucleotide was added 0.75 mL of 0.15 M EDC in 0.1 M MeIm₇ buffer and the resulting preparation was stirred with the linker bounded cellulose acetate membranes for 30 min at room temperature in dark. The efficiency of the oligonucleotide immobilization was measured spectrophotometrically at 280/260 nm (Holme and Peck, [1998]).

Results and Discussion

Properties of Cellulose Acetate Membranes

Cellulose acetate is a natural polymer possessing free hydroxyl groups, which are esterified with acetic acid to form di- and triacetates. Cellulose acetate membranes are rigid, thick, and have optimal pore sizes (Ruckstein and Guo, [2001]). Cellulose acetate does not dissolve in water; consequently it does not hydrolyze.

To optimize the preparation of cellulose acetate membranes, different cellulose acetate concentrations were tested (4–10%). Glass plates were covered with varying cellulose acetate solutions and the resulting membranes dried first at room temperature and then at 70°C. The dry weight of membrane, density of cellulose acetate (d_ca = 1.176 g/mL) and the area of glass plates were used to calculate the thickness of membranes.

As a result, 8% cellulose acetate membrane proved to be best for further studies. Its thickness was 0.27 µm. Cellulose acetate membranes were not fragile in the 4–10% range.

Preparation of Activated Cellulose Acetate Membranes and Oligonucleotide Immobilization

Carbonyldiimidazole (CDI), a product of the reaction between phosgene and imidazole, and related heterocyclic carbonylating reagents could be used to convert free hydroxyl groups of polysaccharide gels and other hydroxyl copolymer supports into activated imidazolyl-carbamate groups that on reaction with N-nucleophiles yield N-alkyl carbonates. An important advantage of the 1,1′-carbonyl diimidazole method compared to the standard CNBr procedure is the absence of any additional charged groups introduced by the functional groups of the activation reagent during both activation or coupling steps (Milton and Hearn, [1987]).

Five different CDI concentrations were used for optimization of cellulose acetate:CDI ratio (0.4; 0.6; 0.8; 1.0; 1.2) where the linker concentration was held constant as 0.01 µmol/mL. The results are summarized in Table 1. The activation level of cellulose acetate membranes were determined by TNBS assay. Oligonucleotide was activated by EDC in MeIm₇ buffer by stirring at room temperature in dark. The membranes were immersed into activated oligonucleotide solutions so that immobilization yields could be estimated and the immobilization yield were estimated as previously described.

Table 1. Relationship between activation levels of cellulose acetate membranes (CA) and yield of oligonucleotide immobilization

mmol CDI/g Cellulose acetate	Bounded CDI (%)	Immobilization yield (%)
0.4	3.87	50.0
0.6	5.31	90.0
0.8	4.93	91.5
1.0	8.72	92.5
1.2	5.16	87.0

As it is seen in Table 1, ratios of CDI higher than 0.4 do not affect oligonucleotide immobilization considerably. According to these results a ratio of 1 mmol CDI/g CA was chosen for further experiments.

The Effect of Linker Concentration on Oligonucleotide Immobilization

For the estimation of optimal HDA/cm² membrane surface, HDA:CA various ratios were studied. TNBS assays were performed to determine the optimal linker/cm² membrane ratio for HDA concentrations in the range of 0.1–0.001 µmol/mL. In this study 8% cellulose acetate and 1 mmol CDI/g CA was used for activation at room temperature.

The results showed that at 4°C the coupling reaction has not gone to completion yet and 0.01 µmol HDA/g CA is optimal for oligonucleotide immobilization (Table 2).

Table 2. The coupling of HDA to activated cellulose acetate membranes as a function of its concentration and resulting oligonucleotide immobilization yields

	HDA coupling efficiency (%)		Immobilization yield (%)
µ mol HDA/g CA	R.T.	4°C	R.T.	4°C
0.1	47.5	17.5	41.2	5.25
0.01	49.0	29	47.9	23.2
0.001	49.5	20.5	45.4	6.97

R.T.: Room temperature.

The Effect of Oligonucleotide Concentration on its Immobilization

In order to find the appropriate oligonucleotide concentration for immobilization, different amounts of oligonucleotides between the range of 1–10 pmol/mL were studied at room temperature with 1 mmol CDI/g CA and 0.01 µmol/mL HDA/g CA (Table 3).

Table 3. Oligonucleotide immobilization as a function of oligonucleotide concentration.a

Oligonucleotide concentration (pmol/mL)	Oligonucleotide immobilization yield (%)
1	30
3	91
5	90
10	47

^aAt room temperature.

The Effect of Linker (HDA) and Glycine Concentrations on the Immobilization of Oligonucleotide

In order to obtain the best hybridization results, optimal membrane charge characteristics are very important. DNA is a polyanionic biomolecule whereas the membrane is pronounced polycationic because of its HDA linker's charge. Coupling, in addition, of an excess of glycine, however, would render the membrane more polyanionic. Therefore, optimization studies for the optimal charge characteristics are required.

1,6-Hexanediamine:Glycine ratios were studied to optimize oligonucleotide immobilization. Prior to oligonucleotide immobilization the respective amounts of bound glycine were estimated using the ninhydrin assay. Table 4 shows that HDA can couple to the membrane faster than glycine. On the other hand, glycine can also couple in the presence of HDA. Our aim was to determine an acceptable ratio of both ligands since hybridization is best. In the case of the glycine concentration increases the percentage of oligonucleotide immobilization is diminished (Table 4); however, this result may not necessarily be negative, since the bound oligonucleotide can possibly be presumed to possess an optimal conformation towards hybridization.

Table 4. Deoxyribonucleic acid immobilization on activated cellulose acetate membrane with various HDA:Gly ratios

HDA:Gly	Bounded glycine (µg/g membrane)	Oligonucleotide immobilization yield (%)
1:2	496	46
1:5	609	43
1:10	942	30

To get the best hybridization ratio when the oligonucleotide is bound to the support in a more rectangular orientation due to charge rejection by the membrane's more polyanionic surface characteristics using glycine (amino acid) is an alternative and new method. These polyanionic membranes can be used as a membrane of DNA biosensors.

Acknowledgments

This work was supported by Ege University Research Fund Project 200FEN051.