Medical Device Link.

Originally Published IVD Technology September 2001

Membrane immobilization of nucleic acids, Part 2: Probe Attachment Techniques

Kevin D. Jones

The best techniques ensure control of probe orientation but are time-consuming and difficult to integrate into automated production processes.

Kevin D. Jones, PhD, is the manager for diagnostic technology at Whatman International Ltd. (Maidstone, Kent, UK). He can be reached via e-mail at kjones@whatman.co.uk.

The first in this series of articles looked at membrane substrates that can be employed in nucleic acid tests (IVD Technology, July/August 2001, page 50). This second installment considers techniques for binding nucleic acid probes to these substrates in preparation for hybridization (see Figure 1).

Figure 1. Cleaned DNA associated with fibers after washing with FTA reagent.

Many membranes recommended for nucleic acid immobilization are supplied with detailed manufacturer's instructions. These instructions should always be followed carefully as a first step in determining optimal binding conditions. It should be remembered that many linkage techniques result in uncontrolled addition to the membrane. The orientation of the probe may then be unfavorable and limit the potential reaction. Directed adsorption, where the immobilized nucleic acid is end-linked, will therefore be preferable in most cases.

Physical Adsorption

Physical adsorption of biomolecules to a solid surface is a process of great interest to the IVD industry. The physical adsorption of proteins to membranes is the basis of immunochromatographic assays.¹ A great deal of research on the physical adsorption of proteins onto a solid phase has been conducted. However, there have been fewer studies published on nucleic acid binding. It should be borne in mind, however, that even if the final mode of attachment is via a chemical link, the initial separation and attraction steps whereby the nucleic acid comes into proximity with the reactive groups are based on physical adsorptive processes.² The chemical structures of the individual bases are responsible for any physical or chemical binding activity (see Figure 2).

The precise mechanism underlying the adsorption of a biomolecule onto a solid phase is unclear. It has been observed that practically any material will interact with almost any surface.³ However, the level of interaction between material and surface can vary significantly. The interaction is probably a five-stage affair, consisting of the following.³

• Transport of the molecule to the surface.
• Adsorption to the surface.
• Rearrangement of the adsorbed molecule.
• Potential desorption of the adsorbed molecule.
• Transport of the desorbed molecule away from the surface.

Although this scheme implies that the potential for desorption is inherent, the binding is practically irreversible provided the molecule is of sufficient size.⁴ This is thought to be due to the large number of binding sites (termed Z) that are present. Although any one binding site may dissociate from the surface at any time, the effect of a large Z value is that the molecule as a whole will remain bound. This correlation between molecular weight and binding has been widely reported for proteins, and the same effect has been reported for nucleic acids. Although the binding observed is not as strong as seen with proteins, large nucleic acids do show good binding properties to membranes.

Figure 2. Bases that are found in DNA and RNA.

Nucleic acids can be immobilized on nitrocellulose membranes by simply air-drying or baking the membrane. Air-drying typically involves exposure for 2–8 hours. The alternative is oven-drying at 80°C for 2 hours.⁵ If nitrocellulose membranes are to be baked, care must be taken to avoid combustion. Drying and baking are believed to result in the nucleic acids becoming attached to the membrane by hydrophobic interaction, although the exact nature of the binding is not well understood. This method of physical adsorption has been routinely used for many years.

Some nucleic acids, particularly DNA, large RNA, and polymerase chain reaction products, will bind very efficiently upon drying. Unfortunately, success depends on a number of factors that will vary among samples. Problems with the attachment of small oligonucleotide probes (typically fewer than 50 bases) have frequently been reported. Because an oligonucleotide probe is often used as the capture material for an assay, the binding of the oligonucleotide is key to the functioning of the whole assay. Nitrocellulose, with its relatively low binding capacity and low physical strength, may thus be less suitable for use as a membrane for nucleic acid testing, according to some. However, it has been widely used in commercial assays, and the use of relatively straightforward techniques (described in the next section) can enhance its binding capability.

The membrane most commonly used in the laboratory for nucleic acid binding is nylon. Nylon can be surface-modified during manufacture; positively charged groups introduced in production are able to enhance separation from the liquid phase by means of electrostatic interaction with the phosphate backbone of the nucleic acid. When the nucleic acid is dried, a proportion of the thymine residues may cross-link to amine groups on the surface of the membrane. This cross-linking can be furthered by exposure to UV radiation (see below).⁶ Whatever the actual surface chemistry, however, it has been proposed that the binding is still predominantly hydrophobic in nature.⁷

A key problem with physical adsorption is that the material immobilized on the surface has undergone rearrangement steps and thus may not be in a suitable conformation to interact with any incoming sequence. While this may not be consequential for long sequences of nucleic acid, as the nucleic acid chain length decreases, the potential loss of activity may become a significant burden. The observable deformation of structure, as some of the bases are immobilized onto the surface rather than remaining free to interact in solution, may stop any hybridization event from occurring.

Enhanced Physical Adsorption

From a production standpoint, physical adsorption is a very simple and robust technique, and therefore useful.⁸ The materials to be immobilized can be applied in a continuous process, and linkage is effected by a simple and well-controlled procedure (drying). Alternative techniques are generally either batch processes or involve complex and time-consuming steps.

If the problems evident with physical adsorption could be solved—that is, if the process were enhanced to provide a stronger or directed binding—physical adsorption could well become the technique of choice for large-scale manufacture of nucleic acid diagnostic tests. Several techniques have been proposed for enhancing the performance of physical adsorption to improve its utility.

One suggestion is to attach a poly-T tail to the probe. This is expected to increase binding to a number of substrates, including nylon and nitrocellulose. The hydrophobic tail would add preferentially to the membrane, providing directed adsorption as well as stronger adhesion. The probe is oriented correctly for hybridization: the sequence of interest is away from the membrane surface and hence free to interact in solution.

Figure 3. The structure of linker reagents: homobifunctional (a); heterobifunctional (b); and photoactivatable (c).

Typically, a poly-T would be between 50 and 100 bases in length. Increasing the length of the poly-T tail would not significantly improve the strength of the interaction, but it would add considerably to the cost of the probe. It has been observed that the use of a poly-T tail enhances the observed binding of a nucleic acid onto a membrane, and it would be possible for the sequence of interest to bind onto the membrane surface and no longer be accessible for reaction. This is a valid point. In practice, however, the large increase in the number of nucleic acids bound would more than compensate for the small number of cases where the sequence of interest is bound flat.

Baking oligonucleotide probes onto the membrane could improve the level of binding observed. However, this process is time-consuming and would probably result in the probe being immobilized flat onto the membrane surface. The consequent loss of freedom could inhibit hybridization significantly and thus seriously limit the utility of the technique.

Another proposed performance-enhancement technique is charge attraction, which can be achieved either through modification of the membrane surface during manufacture or by postproduction coating with a charged species such as poly-L-lysine. The process has been demonstrated to work with polyvinylidene fluoride (PVDF) or nylon membranes and glass slides. Charge modification has been widely used, particularly with glass slides in array fabrication. Many early procedures relied upon coating the glass slides with poly-L-lysine, but newer procedures produce higher binding levels than observed with poly-L-lysine. Most of these newer formulations are proprietary and are either covered by patents or not disclosed in the public domain.⁹

Most of the enhanced physical adsorption techniques not only increase binding efficiency (especially for very short probes) but also, significantly, result in a probe that is oriented so that the sequence of interest is projected away from the solid surface. When directed in this manner, the sequence of interest is free to interact with incoming sample. The possible improvement in selectivity and sensitivity holds great potential, especially when shorter probes are being used.

UV Irradiation

UV cross-linking is still one of the simplest ways to ensure covalent binding of a nylon membrane to a nucleic acid probe. The use of UV irradiation to fix nucleic acid probes to nylon has been well documented. The linkage proceeds predominantly through thymine residues (and through other nucleotides to a lesser extent), which react with amine groups present on the nylon membrane when they are activated.^10,11 Some of these bases therefore become covalently linked to the membrane surface and are unavailable for hybridization. Thus, UV linkage can destroy subsequent hybridization if the membrane is overexposed.

The difficulty in successfully exposing a membrane to UV radiation lies in determination of the correct dose. Too little exposure results in inefficient cross-linking, whereas too much will reduce hybridization efficiency. Many membrane manufacturers provide guidelines with their products. In such cases, it is important to deliver an accurately measured dose of radiation rather than using an uncalibrated transilluminator and guesswork. Although cross-linking radiation levels will vary according to the type of probe used and with trade-offs between immobilization and hybridization efficiencies, typical conditions are 1.2–1.6 kJ/m² for damp membranes and 0.12–0.16 kJ/m² for totally dry membranes.⁵

One key problem associated with the UV irradiation of membranes is their moisture content. Water absorbs UV irradiation, so a variation in the drying process could result in a higher or lower level of cross-linking than anticipated. For critical applications, the use of linkage through the thymine bases cannot be recommended as it is very hard to control the degree of cross-linking. Perhaps the major exception is when UV irradiation is practiced. When the probe contains a poly-T tail, the variation observed would be insignificant in this case.

Photoactivatable compounds, either on the membrane or on the probe, are an advance in cross-linking technology.¹² Use of the right photoactive compound will provide the probe with the correct orientation and freedom. Photoactive chemicals can be used to introduce either haptens (e.g., photobiotin)¹³ or reactive moieties (e.g., succinimidyl-6-[4'-azido-2'-nitrophenylamino] hexanoate).¹⁴ Other photoactive compounds besides these examples have been proposed or used; however, most have seen little application owing to their often uncontrollable reaction upon exposure to UV radiation.

Microparticle Conjugates

The covalent linkage or physical adsorption of a nucleic acid onto a microparticle is a very efficient and controllable way to immobilize a nucleic acid, as discussed above. The available surface chemistries on beads, and the methods for linking aminated or thiolated probes to the beads, have been well reported elsewhere.¹⁵ And the technique of spraying the particles onto membranes to effect entrapment is well known. It is possible to produce general lines (as in the "boulders in a stream" approach) when the particles used are 30–50% of the pore size of the cast membrane used to trap them. If the beads are used with a track-etched membrane (TEM), however, then bead capture processes can be employed. The use of fluorescent labels can be optimized by means of dyed membranes that cut out nonspecific signals and thus provide a significantly enhanced signal-to-noise ratio.¹⁶

Haptens

Figure 4. Structure of the chemical modifiers used to introduce thiol or amino groups to the 5' end of DNA or RNA.

Covalent Linkage

Covalent linkage is an increasingly popular technique for immobilizing nucleic acid probes onto a range of surfaces. There are a number of journals whose focus is the attachment of nucleic acids to solid surfaces via traditional chemistry. Of the various linking chemistries that can be used, the most common are amine, carbonyl, carboxyl, and thiol. Covalent linkage ensures that the orientation of the probe is controlled, but the chemistry can be time-consuming and expensive (see Figure 3).

This immobilization technique is most efficient when reaction takes place through the end groups. The mode of attachment can range from very simple—for example, a Schiff's base reaction where an aldehyde group reacts with an amine followed by reduction—to sophisticated, with specifically designed linkers representing the latter.

A controlled-chemistry linkage makes possible controlled attachment of the nucleic acid via a terminal group. A spacer of the correct size can easily be introduced to ensure that the immobilized acid can move freely to hybridize with any added probe. If attachment were to proceed via the bases, the ability of the immobilized nucleic acid to hybridize would be significantly restricted. The linking groups are used to introduce active end groups (see Figure 4).

Covalent linking techniques are also quite appropriate for use with micro- or macroarrays. A vast number of different linking chemistries can be used, varying with the immobilization substrate, provided there is a suitable active group on the membrane surface.

Conclusion

Controlled addition to the membrane makes covalent linkage an appealing technique for immobilizing nucleic acid probes. This method of attachment is essentially a batch process, which makes it difficult to transfer to large-scale production and automated processes. The final part of this article will describe various methods of achieving covalent linkage and will conclude with a discussion of the manufacturability of the membrane-based systems reviewed.

References

1. KD Jones, "Troubleshooting Protein Binding in Nitrocellulose Membranes, Part 1: Principles," IVD Technology 5, no. 2 (1999): 32–41; available from Internet: http://www.devicelink.com/IVDT/archive/99/03/009.htm.

2. AH Dent and A Aslem, "Other Categories of Protein Coupling," in Bioconjugation: Protein Coupling Techniques for the Biomedical Sciences, ed. AH Dent and A Aslem (London: Macmillan, 1998), 504–569.

3. W Norde, "Adsorption of Proteins from Solution at the Solid-Liquid Interface," Advances in Colloid and Interface Science 25 (1986): 267–340.

4. AW Adamson, "The Solid-Liquid Interface: Adsorption From Solution," in Physical Chemistry of Surfaces, 5th ed. (Chichester, UK: Wiley, 1990), 421–459.

5. SA Nierzwicki-Bauer et al., "A Comparison of UV Cross-Linking and Vacuum Baking for Nucleic Acid Immobilisation and Retention," BioTechniques 9 (1990): 472–478.

6. JK Li, B Parker, and T Kowalin, "Rapid Alkaline Blot-Transfer of Viral dsRNAs," Analytical Biochemistry 163 (1987): 210–218.

7. JM Alter, "The Role of Membranes in Molecular Diagnostics," IVD Technology 4, no. 5 (1998): 43–50.

8. TC Tisone, "In-Line Manufacturing for Rapid-Flow Diagnostic Devices," IVD Technology 6, no. 3 (2000): 43–60; available from Internet: http://www.devicelink.com/IVDT/00/05/005.htm.

9. C Henke, "DNA-Chip Technologies, Part 2: State-of-the-Art and Competing Technologies," IVD Technology 4, no. 7 (1998): 35–44; available from Internet: http://www.devicelink.com/IVDT/98/11/010.htm.

10. GM Church and W Gilbert, "Genomic Sequencing," in Proceedings of the National Academy of Sciences USA 81 (1984): 1991–1995.

11. I Saito et al., "Photo Induced Reactions 133: Photo Chemical Ring Opening of Thymidine and Thymine in the Presence of Primary Amines," Tetrahedron Letter 22 (1981): 3265–3268.

12. JK Veilleux and LW Duran, "Covalent Immobilization of Biomolecules to Preactivated Surfaces," IVD Technology 2, no. 2 (1996): 26–31; available from Internet: http://www.devicelink.com/IVDT/96/03/005.htm.

13. AC Foster et al., "Non-Radioactive Hybridisation Probes Prepared by Chemical Labelling of DNA and RNA with a Novel Reagent, Photobiotin," Nucleic Acids Research 13 (1985): 745–761.

14. GH Keller and MM Manak, DNA Probes (London: Macmillan, 1989).

15. LB Bangs and MB Meza, "Microspheres, Part 2: Ligand Attachment and Test Formulation," IVD Technology 1, no. 4 (1995): 20–26; available from Internet: http://www.devicelink.com/IVDT/95/04/006.htm.

16. KD Jones, "Membrane Based Tests," in The Latex Course 2001 (San Diego, April 30–May 2, 2001).

17. M-C Shao et al., "Complex Neoglycoproteins," Methods in Immunology 184 (1990): 653–659.

Photo Courtesy Whatman Bioscience

Copyright ©2001 IVD Technology