Medical Device Link.

Originally Published IVD Technology April 2003

Assay Development

A novel amplification technology holds promise for increasing ELISA sensitivity and adding quantitative capability to lateral-flow tests.

Tina Meyer Hansen

Figure 1. A nondextran fluorescent conjugate (a) with the fluorophors conjugated directly to the antibody is susceptible to quenching of the dye, but a dextran fluorescent conjugate (b), in which the fluorophors are conjugated via a dextran chain, displays little or no quenching (Click to Enlarge).

A new high-performance dextran-based amplification technology allows for covalent coupling of dyes that act both as fluorescent and as visible absorbent dyes. The combined capabilities of this technology and high-performance fluorophors make possible the development of new lateral-flow test devices that not only give end-users simple yes-or-no results, but also provide quantitative readouts when used with a basic fluorometer. Having both visible and fluorescent capabilities in the same test opens a completely new range of value-added applications while retaining the handling simplicity of the dipstick format. Moreover, the technology can be used to increase the sensitivity of microtiter plate assays.

This article discusses the potential of the new technology for use in such applications after presenting some results of experimental trials that demonstrate the robustness and flexibility of the technology for possible quantitative applications.

The experimental study used a Cy cyanine dye from Amersham Biosciences AB (Uppsala, Sweden); specifically, the monofunctional Cy5 fluorophore whose absorption maximum is near 650 nm and fluorescence maximum is near 670 nm.^1,2 When the fluorescent Cy5 was conjugated to a dextran chain by means of the technology developed by Amdex a/s (Jyllinge, Denmark), immunoassay sensitivity and performance were observed to be higher than with standard nondextran Cy5 conjugates in which the Cy5 normally is conjugated directly to the protein. One explanation for this superiority might be reduced quenching of the dye; because the fluorophors are spread over the entire dextran, the average distance between the individual dye molecules is consequently greater. Another explanation is that the dextran accommodates more dye per binding molecule (hapten).

Use of dextran-Cy conjugates in enzyme-linked immunosorbent assay (ELISA) tests was observed to lower detection limits by as much as 100 times and to produce up to eightfold increases in signal intensity. In lateral-flow tests, the visual signal of the dextran-Cy5 conjugates was similar to that of other known visual conjugates such as gold and latex, but when the fluorescent signal of the dextran-Cy5 conjugates was read directly off the nitrocellulose strip with a fluorometer, the threshold improved by 20 to 30 times. Different batches of nitrocellulose membranes and different batches of the dextran-Cy5 conjugates were examined to make sure that the direct fluorescence reading was reproducible.

Dextran Amplification Technology

Figure 2. A comparison of the fluorescence of three dextran-Cy5 conjugates with different loads of Cy5 per dextran and no binding molecule attached, as measured in a microwell plate fluorometer. Dilutions were performed to allow measurement of fluorescence at the same absorbance level for each conjugate. The numbers in the x-axis represent the molar ratio between Cy5 and dextran (Click to Enlarge).

Amdex dextran conjugates are based on a patented chemistry involving a hydrophilic sugar-chain backbone to which hundreds of specific binding molecules and label molecules can be covalently coupled.3–5 The molecules are linked to the dextran by means of a reactive group such as divinylsulfone (DVS). Dextran sized between molecular weight (Mw) 20,000 and 2,000,000 atomic mass units (amu) is typically used. On each dextran can be as many as 1000 reactive DVS groups, depending on the size of the dextran and the degree of activation. Many of the reactive groups are involved in the cross-binding of the dextran chains (making a mask structure), but others are used to link label molecules and binding molecules. Hundreds of molecules can be attached to each dextran, but the total number typically is kept below 100 in order to ensure good cross-binding of the dextrans. A high dextran-to-dextran cross-binding gives higher signal intensity per direct binding event than a lower cross-binding.

In a nondextran conjugate, the label molecule normally is attached directly to the protein. This limits the number of label molecules per protein. When a dextran chain is used as a linker, the ratio of label to binding molecules can be varied to suit the application. If a high degree of signal per binding event is required, this can be obtained by binding a large quantity of fluorophors and only a few binding molecules to each dextran chain.

Typical haptens are DNA or proteins such as monoclonal and polyclonal antibodies, streptavidin, protein A, protein G, and antigens (drugs or hormones) for competitive assays.

Typical label molecules are enzymes such as alkaline phosphatase and peroxidase; absorbance dyes such as Rhodamine, Texas Red, Alexa dyes, and Cy5; and fluorescent dyes such as phycoerythrin, FITC, Texas Red, Alexa dyes, and Cy5. Many dyes, such as the Cy5 selected for the experiments described in this article, are both absorbent and fluorescent dyes.
However, numerous other types of molecules could be linked to the dextran, depending on the type of assay and its requirements. The dextran conjugates can be used in all sorts of assays, among them sandwich assays, direct-binding assays, competitive-inhibition assays, and serological assays. Here, only assays using Cy5 to illustrate the technology concept are discussed.

Reduced Quenching

Figure 3. A comparison of the performance of streptavidin/dextran-Cy5 and nondextran streptavidin/Cy5 in ELISA testing. The green line indicates the highest background level measured (Click to Enlarge).

The difference in structure between a standard nondextran fluorescent conjugate and a dextran fluorescent conjugate is significant (see Figure 1). In nondextran conjugates, the Cy molecules are conjugated directly to the protein hapten and, normally, the interlabel distance does not vary. But in conjugates formed using the new dextran technology, the average distance between individual Cy molecules depends on the amount of Cy per dextran and the size of the dextran. Conjugating Cy5 to a dextran chain thus results in an average distance between Cy molecules that is longer, with consequent diminished intermolecular quenching than in nondextran conjugates.⁶

The reduction in quenching can be demonstrated by loading the dextran with different amounts of Cy5 molecules, as was done in a simple experiment (see Figure 2). The assumption was that the average distance between Cy5 molecules on a dextran with a low load of Cy5 would be longer than on a dextran with a higher load of Cy5.

In the experiment, dextrans carrying different loads of Cy5—125, 190, and 250 molecules per dextran—were diluted so as to allow measuring the fluorescence of the conjugates at the same absorbance, that is, with the same amount of dye per milliliter. These were simple twofold dilutions, with no binding molecule attached.

Figure 4. A comparison of the performance of goat antirabbit/dextran-Cy5 and nondextran goat antirabbit/Cy5 in ELISA testing. The green line indicates the highest background level measured (Click to Enlarge).

When the fluorescence of these conjugates was measured in a microwell plate fluorometer, the difference was obvious. Some of the fluorescence was turned off as the Cy5 load increased, most likely because of intermolecular quenching. As shown in the figure, the conjugate with 125 Cy5 molecules per dextran had the highest fluorescence, with the dextran carrying 190 Cy5 molecules always displaying higher fluorescence than the one with 250, at all molar ratios of Cy5 to dextran. Thus, fluorescence seems clearly to be affected by the amount of Cy5 per dextran.

Higher Sensitivity in ELISA Tests

Following conjugation with binding molecules, the dextran-Cy5 conjugates can be used in a variety of immunoassays.^7,8 ELISA tests with streptavidin/dextran-Cy5 and goat antirabbit/dextran-Cy5 conjugates were conducted in the investigation reported here.

The performances of a streptavidin/dextran-Cy5 conjugate and a fluorescent nondextran streptavidin/ Cy5 conjugate were compared (see Figure 3), as were the performances of a fluorescent goat antirabbit/ dextran-Cy5 conjugate and a nondextran goat antirabbit/Cy5 conjugate (see Figure 4). The superior sensitivity of the dextran-Cy5 conjugates displayed in these tests may be explained by the boosting effect of the higher ratio of label molecules to binding molecules, but the minimized quenching could also have contributed to the improved signal.

The streptavidin assay depicted in Figure 3 was carried out on microtiter plates that had been coated with goat antirabbit overnight (20 µg per milliliter of sample), then incubated for 1 hour with biotinylated rabbit IgG dilutions ranging from 1250 ng/ml down to 0.08 ng/ml. The streptavidin conjugates were added and the plates incubated for 1 hour at room temperature. The plates were read in a microtiter plate fluorometer (630–670 nm). The background controls were prepared by omitting the goat antirabbit in the first layer, by omitting the biotinylated IgG in the second layer, or by omitting the conjugate addition in the third layer.

The assay with the goat antirabbit conjugates (Figure 4) was carried out on microtiter plates that had been coated with goat antirabbit overnight (10 µg per milliliter of sample), then incubated for 1 hour with rabbit IgG dilutions ranging from 10,000 ng/ml down to 40 ng/ml. The goat antirabbit conjugates were added and the plates incubated for 1 hour at room temperature. These plates were read in the same microtiter plate fluorometer. Background controls for this assay were prepared by omitting the goat antirabbit in the first layer, by omitting the rabbit IgG in the second layer, or by omitting the conjugate addition in the third layer.

Figure 5. A fluorometric graph illustrating the low-detection-area fluorescence of biotinylated IgG using streptavidin/dextran-Cy5 in the experimental ELISA test. The green line indicates the background level (Click to Enlarge).

The greatest increase in the ratio of signal to noise is observed in the low detection range. Therefore, the dextran-Cy5 conjugates appear to be very useful for maximizing the sensitivity of a fluorescent assay. Use of these conjugates makes possible a lower detection limit than is achievable with established Cy5 conjugates already considered sensitive.
The biotinylated IgG used in the streptavidin ELISA test has a low detection range (see Figure 5). The figure reveals that the detection limit of the dextran-Cy5 conjugate is significantly lower than that of the nondextran streptavidin/Cy5 conjugate—by a factor of about 100.

Dextran-Cy5 conjugates can thus be seen to have utility in ELISA testing, an inherently quantitative assay. They can be used in other applications as well, for instance, in lateral-flow tests for point-of-care (POC) testing. Experiments employing dextran-Cy5 conjugates in nitrocellulose-based lateral-flow tests also were conducted.

New Applications in POC Testing 

Lateral-flow technology is often used in point-of-care (POC) and over-the-counter (OTC) dipstick tests. Most of these applications are designed to yield a simple yes-or-no answer. Only a few of the commercially available lateral-flow tests are quantitative. One of the main issues with such tests is that normally it is hard to obtain a good quantitative reading of the visible line or dot on the membrane.

The simple yes or no is sufficient in many tests, but for others a quantitative result in addition could be very useful. For instance, consider POC tests for heart disease markers. Yes or no with respect to a threshold value is enough to indicate whether the level of such a marker in the blood is too high, but if more-detailed information is desirable or necessary, then a blood sample has to be sent to a laboratory for further testing. A reliable quantitative POC test, in which a specific measurement could be read directly on the dipstick, would save time and money in this application.

Dextran-Cy5 conjugates can be used in such a test. These conjugates give a line that can be read by eye to ascertain the simple yes-or-no result, but at the same time, the fluorescent signal on the test line can be read by fluorometer to produce a detailed quantitative answer.

Figure 6. A simple lateral-flow test in which streptavidin/dextran-Cy5 binds to biotinylated IgG on the membrane and no antigen is used. Biotinylated IgG is applied on the test line and a control antibody on the control line. Binding of the streptavidin/dextran-Cy5 conjugate gives the blue lines observed in the flow test (Click to Enlarge).

A Quantitative Lateral-Flow Test. A very basic lateral-flow test was devised to illustrate test-line intensity when streptavidin/dextran-Cy5 conjugates are used (see Figure 6). The signal in the control line is based on the binding of the conjugate to a control antibody, and the signal in the test line is based on the binding of the conjugate directly to a biotinylated IgG. In trials, this test was used without any antigens, and only the binding between the streptavidin/dextran-Cy5 conjugate and a biotinylated IgG was examined. The test was kept highly simple in order to illustrate the various components of the system, such as the functionality of the conjugate and the consistency of the quantitative reading taken directly from the membrane.

Use in Visual Assays. Lateral-flow tests of this type can easily be employed as visual assays. A benefit of using the dextran-Cy5 conjugates is that the amount of Cy5 per dextran can be optimized for the particular application; for instance, the number of dye molecules per binding event can be increased. The same holds true for the binding molecule.

In a qualitative test requiring only a yes-or-no reading, the dextran can be loaded with a large quantity of Cy5 to give an intense blue line without concern for the quenching effect of the dye. Quenching effects are important only in fluorescence assays.

Compared with well-known lateral-flow conjugates such as gold, dextran-Cy5 conjugates produce test and control lines of about the same intensity. In some assays the sensitivity is slightly higher than for gold conjugates, but in other assays it is the same or a bit lower. No advantages in using dextran-Cy5 rather than gold conjugates are obvious when only the yes-or-no test reading is considered. However, when the fluorescence features of dextran-Cy5 come into play along with its visual ones, the advantages of such a conjugate become apparent, as will be described.

Figure 7. Results of lateral-flow tests performed using a protein A/dextran-Cy5 conjugate (a) and a protein A/gold conjugate (b). In the test line on the nitrocellulose, IgG has been applied in dilutions ranging from 2.0 to 0.06 mg/ml (Click to Enlarge).

The visual signals generated by a protein A/dextran-Cy5 conjugate in lateral-flow tests on a nitrocellulose strip can be compared with those produced by a protein A/gold conjugate (see Figure 7). IgG was applied directly to the nitrocellulose in a range of dilutions extending from 2 to 0.06 mg/ml. In this very simple type of test, the conjugates bind directly to the IgG on the membrane without any use of antigen.

Use in Fluorescence Assays. Such tests can be read for fluorescent signal directly on the nitrocellulose membrane when a dextran-Cy5 conjugate is used. Protein A/dextran-Cy5 conjugates, for example, have been tested in serological assays measuring IgG to certain antigens in serum samples. Such tests are reported to be working very well and are expected to be available on the market soon.

Some tests were performed experimentally with a streptavidin/dextran-Cy5 conjugate and a streptavidin/gold conjugate (see Figure 8). Biotinylated IgG was applied to the nitrocellulose in a range of dilutions from 0.13 to 0.004 mg/ml, without antigens, and the conjugates were bound directly to the IgG. Fluorescence readings were made directly on the nitrocellulose strips by means of a prototype fluorometer; however, any type of fluorescent strip reader could have been used.

Figure 8. A graph (a) of fluorescence measurements from lateral-flow tests performed with streptavidin/dextran-Cy5 as in the strips immediately below (b). Fluorescence was measured directly on the nitrocellulose. The lateral-flow tests performed with streptavidin/gold conjugate are shown in comparison (c) (Click to Enlarge).

In this particular investigation, the lower thresholds were 0.07 mg/ml for streptavidin/gold conjugate and 0.016 mg/ml for streptavidin/dextran-Cy5. However, when the fluorescence in the test line for streptavidin/dextran-Cy5 was measured with a fluorometer, the lower threshold went from 0.016 to 0.004 mg/ml. The sensitivity of the test was thereby improved to become 15–20 times greater than that of the gold assay, simply by exploiting the fluorescence features of the dextran-Cy5 conjugates. The test employing the gold conjugate can only be a qualitative assay, but the one using the dextran-Cy5 conjugate offers application either as a simple qualitative test or, when the fluorescent signal intensity is read by a fluorometer, as a quantitative test.

If the only objective is to produce a qualitative test based on visual reading of the test line, then the difference between dextran-Cy5 conjugates and gold conjugates is small, and use of the former brings only limited improvement. However, the fluorescence capability of dextran-Cy5 creates the possibility of a test that is quantitative as well as qualitative, and at the same time more sensitive than other tests. This feature is useful for both OTC and POC applications.

With quantitative assays it is very important that the tests be reproducible over time. Therefore, the lot-to-lot consistency of the dextran-Cy5 conjugates and the nitrocellulose membranes was investigated (see Figures 9 and 10). The fluorescent signal on the membrane was measured for six different lots of streptavidin/dextran-Cy5 conjugates (Figure 9). Several different types of nitrocellulose membranes were studied, but only results for the HF135 membrane from Millipore Corp. (Bedford, MA) are shown in Figure 10. In these investigations, dilutions of biotinylated IgG from 0.13 to 0.004 mg/ml were applied to the membrane (no antigen was used), and the streptavidin conjugates were bound directly to the biotinylated IgG. All measurements of fluorescence were made directly on the nitrocellulose membrane using a fluorometer.

Conclusion

Figure 9. Lot-to-lot consistency of streptavidin/dextran-Cy5 conjugates, as shown by fluorescence measurements performed on lateral-flow tests using a fluorometer. Six different lots of the conjugates were examined (Click to Enlarge).

Dextran-Cy5 conjugates can be used to increase the sensitivity of an assay or to add features to a test. This creates opportunities for new applications in the POC/OTC market. For instance, patients could read out the qualitative yes-or-no answer themselves, and then the physician could perform the fluorescence reading to acquire more-detailed information. Another possibility is a disposable reader for patients with chronic diseases. The uses of these conjugates in lateral-flow testing are multiple.

Moreover, certain ELISA assays would benefit from having a lower sensitivity cutoff so that diseases could be detected at an earlier stage than is possible today. More-sensitive assays using dextran-Cy5 conjugates may make this a reality.

Tina Meyer Hansen is project manager at Amdex a/s (Jyllinge, Denmark) and is responsible for customer- specific development projects. She can be reached at tha@ amdex.dk.

References

1. RB Mujumdar et al., “Cyanine Dye Labeling Reagents: Sulfoindocyanine Succinimidyl Esters,” Bioconjugate Chemistry 4, no. 2 (1993): 105–111.

2. SR Mujumdar et al., “Cyanine-Labeling Reagent—Sulfobenzindocyanine Succinimidyl Esters,” Bioconjugate Chemistry 7, no. 3 (1996): 356–362.

Figure 9. Lot-to-lot consistency of streptavidin/dextran-Cy5 conjugates, as shown by fluorescence measurements performed on lateral-flow tests using a fluorometer. Six different lots of the conjugates were examined (Click to Enlarge).

3. Water-soluble, polymer-based reagents and conjugates comprising moieties derived from divinyl sulfone, U.S. Pat. 5,543,332 and European patents PCT/ DK92/00206 and WO 93/01498.

4. Use of nucleic acid bound to carrier macromolecules, U.S. Pat. 6,207,385 and European patents PCT/GB97/ 03160 and WO 98/22620.

5. A method for the preparation of water-soluble cross-linked conjugates, European patents PCT/DK99/00426 and WO 00/07019; U.S. patent pending.

6. L Brand and B Witholt, “Fluorescence Measurements,” Methods in Enzymology 11 (1967): 776–787.

7. I Hemmila, “Fluoroimmunoassays and Immunofluorometric Assays,” Clinical Chemistry 31, no. 3 (1985): 359–370.

8. CP Price and DJ Newman, Principles and Practice of Immunoassays (London: Macmillan Reference, 1997), 389–424. 

Copyright ©2003 IVD Technology