Medical Device Link.

IVD Technology Magazine
IVDT Article Index

Originally Published March 2000

The place of gold in rapid tests

John Chandler, Tracey Gurmin, and Nicola Robinson

The demand for rapid membrane-based tests encompasses applications in a wide variety of fields (see box, below). The number of such potential applications will likely increase as these low-cost alternatives to expensive instrumented methods of testing become more sensitive and more specific.

This article focuses on the importance of a critical component of such tests—the detection label—and how its manufacture can affect the performance and reliability of the whole test system. Specifically, the article will address the advantages of using gold conjugates in this capacity.

What Is a Rapid Test?

A rapid test is an inexpensive, disposable, membrane-based assay that provides visual evidence of the presence of an analyte in a liquid sample. Such tests can be formatted either as freestanding dipsticks or as devices enclosed within plastic housings. Typically, as little as 200 µl of liquid sample is required to perform the test, which is usually complete within 2–5 minutes. In clinical assays, the sample may be urine, blood, serum, saliva, or other body fluids. In nonclinical tests, the sample may be a small volume of solution prepared from soil, dust, plants, or food, and similarly applied directly to the membrane test strip. No instrumentation is required to perform such tests, which can be used in clinics, laboratories, field locations, and the home—often by inexperienced personnel.

A large flask of gold conjugate. Photo by Hugh Burden, courtesy BB International

Rapid tests come in two forms—lateral flow and flow-through. The lateral-flow format is by far the most common because it is easier to manufacture and use. Although the same principles apply to both formats, the lateral-flow test is discussed here.

The base substrate of a rapid test is typically a nitrocellulose strip onto which is immobilized a capture binding protein, usually an antibody or antigen (see Figure 1). A pad (often glass fiber) containing dried conjugate is attached to the membrane strip. For the majority of currently available tests, this conjugate pad contains gold particles adsorbed with antibodies or antigens specific to the analyte being detected. A sample pad, usually paper, is attached to the conjugate pad. When applied to the sample pad, the liquid sample migrates by capillary diffusion through the conjugate pad, rehydrating the gold conjugate and allowing the interaction of the sample analyte with the conjugate. The complex of gold conjugate and analyte then moves onto the membrane strip and migrates towards the capture binding protein, where it becomes immobilized and produces a distinct signal in the form of a sharp red line. A second line, a control, may also be formed on the membrane by excess gold conjugate, indicating the test is complete.

Figure 1. Construction of a lateral-flow rapid test.

By definition, rapid tests should provide results in a short time, preferably minutes. Such tests must be convenient, accurate, reliable, inexpensive, disposable, and foolproof. They must also be easily and unambiguously interpreted, even by users without experience. From the manufacturers' point of view they should carry a large added value and be easily marketed worldwide to users who may be either experienced or inexperienced in the use of such tests. (see sidebar, below).

Rapid tests are versatile. By switching the antibodies and making small adjustments to the chemistry of the strip format, the same test design can be used for many applications.

Why Use Gold?

Early rapid tests used colored latex to form the visual signal, and some current versions continue to use this method. Latex was originally, and still is, the prime labeling method used in agglutination tests. This is because of its predisposition to agglutinate in the presence of binding components. For rapid tests, in which stability of the conjugate is critical for avoiding false positives, this predisposition to agglutinate can become a major problem.

Because of their greater potential stability, gold labels were introduced into membrane-based rapid tests in the late 1980s. Gold particles of any accurately defined size can be manufactured reproducibly under the appropriate manufacturing conditions. Different sizes may be used for different applications. Its superior stability, sensitivity, and precision and reproducibility of manufacture make gold suitable for use in rapid tests. Gold is essentially inert and forms almost perfectly spherical particles when properly manufactured. Proteins bind to the surfaces of these gold particles with enormous strength when correctly coupled, thus providing a high degree of long-term stability in both liquid and dried forms. Also, when accurately stabilized during manufacture, nonspecific interaction of gold conjugates can be reduced to zero.

A comparison of the benefits of different labels used to mark antibodies and antigens is shown in Table I.

Manufacture of Gold Colloids

When producing large volumes of gold conjugate, sophisticated processing methods are required to attain reproducible batch-to-batch manufacture and to avoid instability. The manufacture of 100 liters of high-quality gold conjugate requires the utmost care and attention in order to achieve a final stable and sensitive product. Electron microscopy must be used to provide quality control at each step of the manufacture.

During the past 20 years, manufacturers have introduced a variety of different methods for synthesizing gold colloids. The goal has been to obtain colloids of a monodisperse nature, which are of a controlled and uniform diameter. Although all of the production methods rely upon the reduction of tetrachloric acid (HAuCl₄) to form gold atoms, they vary considerably in the physical conditions, order of reagent addition, reducing agent used, and quality of the final colloid produced (size, shape, and coefficient of variance).

In general, all of the production methods use a reducer to donate electrons to the positively charged gold ions in solution and produce atomic gold. This is shown as follows:

Gold tetrachloric acid + reducer = gold colloid

HAuCl₄ + e^– = Au⁰

Commonly used reducers include sodium citrate, yellow phosphorus, sodium borohydride, and sodium thiocyanate. Figure 2 shows the process of reduction at the ionic level.

Figure 2. Reduction of gold ions to form gold particles.

Before the addition of reducer, there are 100% gold ions in solution. The ordinate of the graph indicates the progress from gold ions to gold atoms as the reducer is added. Immediately after the reducer is added, there is a sharp rise in gold atom content in the solution until this level reaches supersaturation. Aggregation then occurs, in a process called nucleation, to form central icosahedral gold cores of 11 atoms at nucleation sites. The formation of nucleation sites, in order to reduce the supersaturation of gold atoms in solution, occurs extremely quickly. Once this is achieved, the remaining gold atoms in solution continue to bind to the nucleation sites under an energy-reducing gradient until all atoms are removed from solution.

The number of nuclei formed initially determines how many particles finally grow in the solution. This number, in turn, depends on the amount of reducer added. A large amount of reducer produces a large number of nucleation sites and hence a large number of gold particles. Clearly, the larger the number of nucleation sites for a given amount of gold chloride in solution, the smaller will be the final size of each gold particle. Particle size is thus carefully controlled by the amount of reducer added. If manufacturing conditions are optimized, then all nucleation sites will be formed instantaneously and simultaneously, resulting in all gold particles growing to exactly the same size (monodispersal). This is very difficult to do. Most manufacturing methods do not achieve instantaneous reduction and formation of nucleation sites, resulting in uneven growth and a multidisperse colloid that is virtually unreproducible and results in a very unstable conjugate.

A gold colloid comprises a suspension of gold particles individually surrounded by a negative charge layer arising from the residual negative ions in solution (see Figure 3). This charge layer, called the zeta potential, provides the means for the gold particles to repel one another and to stay in suspension indefinitely. The zeta potential can be compressed or expanded depending on the total ionic concentration of the surrounding solution.

Figure 3. Colloidal gold particle surrounded by a double ionic layer.

High-quality gold colloids and custom-manufactured conjugates are readily available commercially. By purchasing these components from a reputable supplier—instead of attempting to process large volumes of gold themselves—manufacturers can save time and considerably reduce their manufacturing risks.

Good Gold, Bad Gold

The apparent ease with which gold colloids and conjugates can be manufactured has led to the commercial availability of many poor-quality, poorly characterized, and nonreproducible products. When incorporated into rapid tests, such products can lead to poor stability, sensitivity, and specificity. To prevent this, gold colloids should be evaluated ultrastructurally using a transmission electron microscope (TEM). Such an evaluation should enable the manufacturer to compare the diameter of the colloids to that of a calibrated standard and to obtain information about particle sphericity, particle irregularity, and overall variance in particle diameters.

Unevenly shaped particles occur when the nucleation and growth rate during the reduction process have been uncontrolled (see Figure 4). Particles may be seen to be of different sizes and shapes (oval, triangular, oblong, rhomboid, etc.). These particles will not coat evenly with protein during conjugation and will not evenly repel one another in solution. This can affect the color, sensitivity, specificity, and stability of the final product. A mere 5% of odd-shaped particles can influence a test result, making it completely nonreproducible. The signal formed by such "bad gold" is usually seen as having a bluish or purple color compared with the normal cherry red that typifies a well made 40-nm monodisperse colloid. Although the darker color may be easier to see against the white membrane of a test system, it indicates an inherent instability that is much more likely to give false results. Even worse, uneven particle shape and size produce very uneven protein coating, which leads to long-term clustering and aggregation of the conjugate. Such changes may occur within days or even hours of storage of the conjugate in solution.

Figure 4. Good gold colloids and bad (unstable) gold colloids.

Even drying the conjugate immediately onto the solid-phase substrate does not entirely overcome this problem. During the drying process, surface proteins that are not securely attached because of particle shapes, can easily become detached, yielding false negatives and high background.

One of the most common problems encountered in rapid-test performance is failure of the gold conjugate to release at the correct speed and with integrity from the glass-fiber conjugate pad. This failing is often the result of poorly coated gold particles being directly exposed to the glass-fiber material and not being able to release. This inevitably means that surface proteins have either become permanently attached to the fiber matrix or have left the gold particles and are floating free. With well-prepared gold conjugates, starting with evenly coated monodisperse and spherical particles, such dangers are greatly reduced.

Although TEM examination is the only true way to determine the quality of colloid, a quick method for determining whether a colloid or conjugate contains fused, aggregated, or heterogeneous particles, or a mixed-size population, is to examine its color visually. A good 40-nm colloid (the most frequently used size for diagnostic applications) should be cherry red. If the colloid or conjugate appears purple, it is likely to be of poor quality and unstable.

Choice of Gold Particle Size

The signal is generated on the test strip by the accumulation of gold particles at the test or control line. These particles must be large enough to be seen. The greater the particle size, the easier it is to see an accumulation of these particles. For example, particles of 1 nm diameter would be virtually impossible to see, no matter how many had accumulated, because 1-nm particles do not have the bright red color of the larger sizes. It is not until particles reach 20 nm that a worthwhile signal can be seen. Steric hindrance becomes a problem as the particles increase in size (see Figure 5). For example, if an IgG molecule (160,000 daltons) is just 8 nm in length, only approximately 4 nm of this will extend from the surface of a gold particle. Particles of 100 nm will tend to dwarf the small surface molecules and make it difficult for them to interact with specific proteins. In addition, the larger the particle size, the fewer of them can be contained in a given volume of solution.

Figure 5. Choice of gold particle size for optimized signal.

This trade-off between required visibility and steric hindrance dictates that, for most immunoassay applications, the optimum particle size is 40 nm. In some cases where steric hindrance is a greater problem (e.g., for smaller antigens), particles of 20 nm are preferred. Larger particles may be preferred when a darker red color is desired or when the lower curvature of the particle surface would improve the molecular interaction between the antibody and antigen. Experiments should be performed to determine which particle size gives the highest sensitivity with the lowest background and greatest stability.

Manufacturing Gold Conjugates

Manufacturers should understand the chemical and physical processes that permit a permanent coupling of their protein to the gold colloid. Armed with this knowledge, it is then simple to proceed from small test batches to large-scale manufacturing without compromising sensitivity, stability, specificity, or reproducibility. Simply throwing a conjugate together, while cheap and easy, will often lead to the formation of clusters. Viewed through a TEM, each cluster is a group of four or more gold conjugate particles. The presence of such clusters generally indicates an unstable product whose performance will change over time. Although initially appearing more sensitive than an unclustered product, such tests will soon begin to show stability and specificity problems.

A good conjugate is one where the protein is adsorbed onto the surface of the gold. Unlike latex, proteins are only adsorbed passively onto the surface; covalent coupling is not performed. However, hormones, drugs, and other small molecules (<10 kD) must first be coupled to a larger carrier protein, typically BSA, before passive conjugation can be performed. This is because there are not enough points of contact between the molecule and the gold surface. Steric hindrance would in any case prohibit the small molecule from rising above the zeta potential. The carrier should be inert, in that it should have no effect on detector protein activity.

Figure 6. Binding forces between an antibody and a gold particle.

The antibody will be firmly attached to the gold by the Fc region, leaving the Fab region protruding through the double-ionic layer surrounding the gold particle and able to bind the analyte (see Figure 6). There must be no excess of labeling antibody present. Such excess will cause unlabeled antibody to compete with the labeled antibody, resulting in the presence of false negatives, and may affect the long-term stability of the conjugate. Optimum binding of proteins to gold occurs at, or close to, the isoelectric point.

Adsorption of protein to gold occurs within a matter of seconds through the follwing mechanisms:

• Initial charge attraction of the negative gold particle to the positively charged amino acids within the protein (e.g., lysine).

• Hydrophobic adsorption of the protein to the particle surface, through certain amino acid residues including tryptophan.

• Dative binding between sulphur residues on the protein (from cysteine residues) and the gold particle.

Stabilizing the Conjugate

Following the conjugation of the specific protein, the conjugate must then be stabilized with a suitable agent. BSA, gelatin, PEG (Carbowax), or casein are commonly used. The purpose of the stabilizer is twofold. First, it reduces nonspecific interactions by blocking any sites on the colloidal surface that are not occupied by the specific protein. Second, it helps provide a more-stable suspension.

Once conjugated, the gold is adjusted to the desired concentration, and suspended in a suitable buffer. This buffer should confer stability to the liquid conjugate. Commercial manufacturers will often use a low-molarity buffer so that the consumer can easily resuspend the conjugate in the buffer of his or her choice. Generally, high-salt buffers and surfactants should be omitted from the final storage buffer as they may cause damage by hydrolysis or displacement of the antibody. Preservatives containing sulphur residues or mercury will cause the conjugate to collapse. The most common preservative used for gold conjugates is 0.1% sodium azide.

Expertly manufactured gold colloid has a virtually infinite shelf life. By calculating how much binding agent can be adsorbed onto the surfaces of the gold particles, it is possible to minimize batch-to-batch variation, and the resulting conjugate has a longer shelf life than antibody alone. Correctly dried onto solid-phase components, these gold-labeled proteins remain stable for many years.

Manufacturing Bulk Quantities of Gold Conjugates

Although technical and scientific sources describe the simple small-volume manufacture of gold colloids and conjugates, the commercial methods used for large-volume manufacture (i.e., 100-liter batches) remain proprietary. Given that the reduction process of gold chloride to gold colloid must take place in microseconds, the accomplishment of this in volumes of 100 liters or more requires very sophisticated techniques that do not follow the same rules as small-volume production.

Because these techniques are known only to specialists in the gold conjugate industry, whose products are commercially available, many manufacturers choose not to attempt the difficult task of producing large volumes of gold colloids in-house. Instead, such rapid-test manufacturers are partnering with gold specialists, thereby reducing their overhead costs and risk of failure, while also making it possible to get their product to market as quickly as possible.

Such partnership arrangements typically begin with the manufacturer conducting in-house experiments using small volumes of commercially available colloids in order to determine which antibodies are most suitable for their application. Following that, some in-house conjugation experiments using larger volumes of colloids are conducted to determine whether scaling up by 10x and by 100x produces the same quality of conjugate. It is important at this stage to perform .rigorous stability trials of the gold conjugate at elevated temperatures and in a variety of buffers as well as in the dried-down state. Quality control should include checks on clustering (by TEM), optical density, sensitivity, and specificity. Several batches should be made at these higher volumes to determine the reproducibility of manufacture before committing to rapid-test batch production.

Possibilities for Quantitation

Most rapid tests are purely qualitative, designed only to indicate whether a particular biochemical component is present in the clinical sample above a detectable level. However, many other tests require an element of quantitative measurement, such as for monitoring the progression of a clinical condition or for determining the relative or absolute levels of particular analytes.

Such tests fall into three types. In the first, a number of built-in signals allow visual comparison of test results as the test is being performed. This type of test can provide only semiquantitative results, where the signal may fall into a preestablished negative-, low-, medium-, or high-result category.

In the second type of test, lines of capture reagent in the test zone are set to capture only a known quantity of analyte. Any excess analyte is captured by subsequent capture lines, thus producing a ladder of signal lines that exhibit a thermometer-type display on the strip.

In a third type of test, the signal generated by the test sample is read by a small, portable analyzer that converts the colored test line to a digital signal. A number of such generic test-strip readers have recently become commercially available.

Gold labels are especially suited to such quantitative or semiquantitative tests because the signals they generate are both accurate and highly reproducible. Signals are most often read by reflectance, with photodiodes producing numbers against a calibrated scale. Precision depends on both how accurately the test result can be read, and how reliably the test strip generates a reproducible signal. Because of the precision of gold labels, the latter can be achieved with greater confidence than with many other types of label.

Silver Enhancement of Gold Signals

Whether in the lateral-flow or flow-through format, rapid tests commonly offer sensitivities similar to or greater than ELISA kits. For many analytes, it is possible to achieve detection levels of 1 ng/ml or less. There are still many analytes, however, for which this direct-labeling system does not produce great enough sensitivity.

Sensitivity is limited by the ability of the user to visually detect the gold signal on a white membrane against a clear background. Quantitative instrumentation using light reflectance techniques usually does not achieve any greater sensitivity than the human eye. A valuable and promising technique for obtaining over 100x increased detection sensitivity is the use of gold labels enhanced with silver.¹ In this method, the initial (sometimes invisible) signal from the gold label is enhanced through the application of a silver solution to the test line. Where the particles used in the gold conjugate are small (~5 nm), greater overall sensitivity can be achieved by using this indirect method. Enhancement can be achieved without any extra steps by incorporating the silver reagents into the one-step test so that application of the liquid sample drives the entire reaction (see Figure 7). All reagents (both gold conjugates and silver enhancement salts) are dried onto the test strip and are solubilized by the sample alone.

Figure 7. Silver enhancement of gold signals in a lateral-flow rapid test.

This indirect enhancement technique is expected to result in analyte detection in the picogram/ml range. The dried reagents are stable and may be introduced into the test devices before assembly. Because of the increase in sensitivity, much lower sample volumes may be used than in current directly labeled test systems. With silver enhancement, it may be possible to develop one-step gold-based tests for analytes not previously detectable by rapid-test configurations.

Conclusion

Rapid tests have changed little since their introduction in the mid-1980s, when latex was commonly used as the conjugate. But since then, the greater demands of sensitivity, reliability, and reproducibility have made gold the conjugate of choice for the vast majority of rapid-test designs.

Well-made gold reagents have tremendous stability, in both the liquid and dried forms. Only the most rigorous approach to manufacturing and using gold colloids will be successful if the range of applications for gold-based rapid tests continues to require even greater sensitivity and specificity.

Reference

1. J Chandler, Assay device and method, 96901447.1. European patent application, British Biocell International.

John Chandler, PhD, is the managing director, and Tracey Gurmin and Nicola Robinson are part of the custom consultancy team at BBInternational (Cardiff, Wales, UK).