Medical Device Link.

Originally Published IVD Technology May 2004

Detection Technologies

The luminescence of the reaction product of the PS-atto substrate by Lumigen Inc. (Southfield, MI) with horseradish peroxidase.

Horseradish peroxidase (HRP) is one of the two most widely used enzyme labels in medical diagnostics and research applications, the other being alkaline phosphatase (AP).¹ HRP is applied often in immunoassays and nucleic acid hybridization assays, in part because of the ready availability of peroxidase-conjugated antibodies to haptens such as biotin, fluorescein isothiocyanate, and digoxigenin. An automated analyzer that employs detection based on HRP-catalyzed chemiluminescence, the Vitros immunodiagnostic system from Ortho-Clinical Diagnostics Inc. (Raritan, NJ), is in commercial use. The smaller size of HRP relative to AP means that its use in conjugates should be less prone to nonspecific binding to solid surfaces. A smaller label enzyme makes interference with immunological recognition by enzyme-antibody conjugates less likely as well.
Despite HRP’s advantages as a label, relatively few chemiluminescent detection reagents for the enzyme exist. New reagents exhibiting improved sensitivity, a wide dynamic range, speed, and user convenience would help to capitalize on the benefits of HRP as a label enzyme.

Figure 1. A reaction scheme depicting the chemiluminescent oxidation reaction of luminol catalyzed by HRP (click to enlarge).

Most reagents in current use incorporate the cyclic hydrazide luminol as the substrate, along with an enhancer (see Figure 1).² A new class of chemiluminescent reagents for peroxidase detection was described previously.² These reagents, including PS-1 and PS-3 from Lumigen Inc. (Southfield, MI), contain an acridan substrate that undergoes enzymatic oxidation to yield a chemiluminescent acridinium ester. The acridinium ester reacts with the hydrogen peroxide present in the reagent to form an unstable dioxetanone intermediate (see Figure 2). An enhancer incorporated into the reagent formulation accelerates and prolongs the enzymatic oxidation of the acridan substrate.

The term substrate is used loosely in regard to the action of HRP on a detection reagent. The actual mechanism is thought to be more complex than typical enzyme-substrate reactions.

Figure 2. A reaction scheme depicting the chemiluminescent oxidation reaction of the acridan compound PS-3 (Lumigen Inc., Southfield, MI) catalyzed by HRP (click to enlarge).

The only species to react directly with the enzyme in its resting state is hydrogen peroxide supplied either in free form or as a complex such as urea peroxide, a perborate salt, or a percarboxylate salt. Reaction of HRP with a source of hydrogen peroxide results in reduction of peroxide to hydroxide and oxidation of the enzyme to two higher-oxidation-state intermediates. The oxidized intermediates oxidize an enhancer in a one-electron process, the enhancer serving as an electron relay. The oxidized enhancer, in turn, oxidizes the so-called substrate which, upon oxidation, reacts with hydrogen peroxide or molecular oxygen to form the chemiluminescent species. The enhancer is reduced back to its native state in the process, and is ready to mediate further oxidation of the acridan.

Several substances are known to function as enhancers of peroxidase-mediated oxidation reactions. Numerous compounds having a phenol group are effective, as are aromatic amines, certain heterocycles, in particular benzothiazoles, and arylboronic acids.^4,5 Compounds in the last group probably work by means of conversion to a phenol through reaction in situ with hydrogen peroxide.

The effectiveness of an enhancer appears to be based principally on its oxidation potential under the reaction conditions and its relation to the oxidation potential of the detectable substrate.⁶ Promotion of peroxidase activity by enhancers applies equally to chromogenic, fluorogenic, and luminogenic reaction processes, since it is the electron transfer process upstream of the final detectable reaction that is critical. Newly discovered peroxidase-catalyzed reactions must nevertheless be carefully optimized with regard to the selection and concentration of enhancers.

Figure 3. Structures of the PS-atto (a) and TMA-6 (b) substrate compounds (click to enlarge).

Chemiluminescent reagents for detection of peroxidase are carefully optimized multicomponent solutions. Typical formulations consist of a source of hydrogen peroxide, the substrate, one or more enhancers, surfactants, and other proprietary components in a buffer. They must invariably be supplied as two components.

In addition, working solutions must be prepared daily because of the limited storage stability of the substrate/enhancer/peroxide mixture. Slow reaction among these components in the absence of peroxidase can lead to unacceptably high reagent blanks and consumption of the reactants. The need to prepare fresh working solutions complicates the use of these reagents, especially in automated analytical systems.

Now, however, a robust new family of chemiluminescent detection reagents for HRP has been developed by Lumigen. Two of these reagents have been commercialized: PS-atto for detection in solution assays and TMA-6 for blotting applications. The most significant property of these reagents is their ability to be stored in a single container as ready-to-use solutions. Storage of the working solution for at least 3 weeks at room temperature and up to 4 months at 4°C has been seen to cause no degradation of performance.

Figure 4. A reaction scheme depicting the chemiluminescent oxidation reaction of PS-atto catalyzed by HRP (click to enlarge).

The new reagents, which are based on an entirely new type of substrate, exhibit a linear response to peroxidase over five orders of magnitude. Measurement of plateau light intensity 6 minutes after addition of HRP to solutions of the reagents resulted in a linear calibration curve over five orders of magnitude of enzyme concentration. An additional characteristic is that chemiluminescence emission develops extremely rapidly by comparison with other substrates, reaching peak intensity in solution assays in less than 1 minute. An investigation into the capabilities of these reagents is reported in this article.

Experimental Setup

The PS-atto working reagent was prepared by combining two solutions in a 1:1 ratio. HRP from Biozyme Laboratories International Ltd. (San Diego) was diluted in water in a series of ten-fold dilutions. The time course of light emission from reaction of PS-atto with HRP was assessed by reacting duplicate 100-µl aliquots of the working reagent with 3.5 ¥ 10–16 mol of HRP at 25°C. Light intensity was recorded in either a Turner Designs Inc. (Sunnyvale, CA) TD 20e luminometer provided with a neutral density filter to attenuate light intensity or a Thermo Labsystems (Franklin, MA) Luminoskan microplate luminometer. All chemiluminescence measurements were performed at room temperature except where specifically noted.

Figure 5. The time course of chemiluminescence emitted by the reaction of PS-atto and TMA-6 with HRP at room temperature (click to enlarge).

The relationship of chemiluminescence peak intensity to the quantity of HRP was measured by reacting triplicate 100-µl aliquots of the working solutions of PS-atto and TMA-6 individually with 10 µl of the HRP solutions prepared in the range 1.4 ¥ 10–16 to 1.4 ¥ 10–21 mol. Light generation ensued upon mixing and reached maximum intensity rapidly. Light intensity was recorded 6 minutes after mixing in either a TD 20e luminometer or a Luminoskan 96-well plate luminometer. (Intensity measurements given in this article and its figures and tables are in relative light units [RLU].)

Storage stability of PS-atto and TMA-6 were evaluated over a 1-year period. Samples of the working solution and of the separate constituent solutions were stored at 4°C, room temperature, and 40°C in brown opaque polyethylene bottles. Aliquots were withdrawn at periodic intervals and analyzed for background luminescence and peak luminescence produced with 3.5 ¥ 10–16 mol of HRP at 25°C. Percent remaining activity was determined by comparison with the control samples stored at 4°C. The latter were shown independently to exhibit no change in signal intensity over more than a year of storage as separated component solutions.

Table I. The relationship of chemiluminescence intensity to amount of HRP using PS-atto or TMA-6 for chemiluminescent detection (click to enlarge).

A Western blot assay of beta-galactosidase as the test antigen and HRP-antibody conjugate was conducted using nitrocellulose membrane and TMA-6 chemiluminescent detection, with beta-galactosidase standards in the range of 5000 down to 5 pg. Blots were incubated briefly with the reagent, excess reagent was drained off, and the damp blots were imaged with x-ray film over the course of 2 hours.

Results

The PS-atto detection reagent consists of the PS-atto substrate compound (see Figure 3a) in a buffered solution containing a phenolic enhancer, peroxide, and other proprietary components for optimum analytical performance. TMA-6 contains a peroxidase substrate compound (see Figure 3b) in a similar formulation.

Figure 6. The ELISA for TSH using components of the Cobas Core immunoassay kit (Roche). Luminescent detection employed PS-atto. Secondary antibody conjugate was diluted twofold to improve signal relative to the background. Colorimetric assay data come from the package insert (click to enlarge).

Oxidation of the PS-atto substrate with HRP and peroxide is believed to produce a dioxetane intermediate that subsequently fragments to produce an acridone in the excited state (see Figure 4). The light emitted by the reaction spans the region of approximately 400–550 nm, with maxima at 430 and 445 nm. The spectrum matches the fluorescence spectrum of N-benzylacridone.

Oxidation of TMA-6 with HRP involves a similar mechanism, the emitter being N-phenylacridone. Chemiluminescence maxima are nearly identical to those of PS-atto.

Reaction of either the PS-atto or TMA-6 reagent with HRP produced chemiluminescence emission immediately upon mixing. Maximum intensity during the reaction was reached after several seconds at room temperature and maintained a constant intensity for several minutes (see Figure 5). Plateau chemiluminescence intensity from reaction of HRP with PS-atto was linearly related to the amount of enzyme over at least four orders of magnitude as shown in Table I. The slope of the curve was described by the equation 

y = 0.931x + 18.689, r2 = 0.999.

A similar dynamic range of HRP quantification was measured when the reactions were run at 25° and 37°C using PS-atto. (These data are not shown.) The dynamic range of measurement with TMA-6 did not extend to as low a level as seen with PS-atto under the conditions employed.

Figure 7. The percentage capacity for activity of PS-atto working solutions that remained after various periods of storage at 4°C or room temperature (a), and the percentage capacity for activity of TMA-6 working solutions that remained after various periods of storage at 4°C (b). Activity is expressed in relation to a control sample freshly prepared from components stored individually at 4°C (click to enlarge).

The extreme detection sensitivity allowed the development of highly sensitive enzyme immunoassays. The Cobas Core commercial ELISA kit for thyroid-stimulating hormone (TSH) produced by Roche (Basel, Switzerland), which uses an antibody-HRP conjugate with colorimetric detection, was modified for chemiluminescent detection with PS-atto. The colorimetric assay quotes a detection limit of 0.05 mIU/L; the lowest standard supplied was 1 mIU/L. The assay was adapted for chemiluminescent detection by diluting the conjugate twofold and substituting PS-atto for the detection reagent. Standards were diluted to 0.003 mIU/L.

This chemiluminescent assay achieved a measurement of 0.003 mIU/L (see Figure 6). The signal for the lowest calibrator, 20.5 ± 2.3 RLU, substantially exceeded the blank at 10.2 ± 2.3 RLU, so the calculated detection limit of the assay is actually somewhat lower. This value does not represent an effort to fully optimize the assay, however; analytical sensitivity is adversely affected when chemiluminescence from reagents bound to the surface of polymer beads in a manual test-tube format is measured.

To assess the reagent stability of working solutions of PS-atto and TMA-6, the solutions were prepared by combining 50-µl aliquots of the two components and storing the mixtures at 4°C and at room temperature. Chemiluminescent assays were run with 3.5 ¥ 10–16 mol of HRP at 25°C. Those measurements were compared with the values obtained with fresh reagent prepared from the components stored individually at 4°C. The results revealed no deterioration of reagent performance on storage up to 3 weeks at room temperature. PS-atto was unchanged after 4 months at 4°C, and TMA-6 was unchanged after 9 months at that temperature (see Figure 7).

In additional experiments, the solution component containing the peroxidase substrate was incubated at room temperature and at 40°C for several months. Aliquots were combined with the second component and the resulting working solution assayed as described above. The results were compared with those obtained using a solution freshly prepared from components stored separately at 4°C. No detectable change in signal was observed in working solutions of either reagent prepared from substrate stored separately up to 5 months at 40°C or 1 year at room temperature.

Western blot assays of protein antigens with peroxidase-conjugated antibodies and TMA-6 chemiluminescent detection were performed with a beta-galactosidase antigen test system. The amounts of antigen blotted ranged from 5 pg to 5 ng. Comparisons were made on ECL+ nitrocellulose membranes (Amersham Biosciences; Amersham, UK) with Lumigen PS-3 and the most sensitive commercial luminol reagents. Detection limits with TMA-6 were as low as or lower than the most sensitive reagents in current use. Background luminescence was controlled well enough to allow detection of low-picogram quantities of antigen. All levels of antigen were detectable at time points from 10 minutes to 2 hours (see Figure 8).

Figure 8. Images of Western blotted beta-galactosidase detected by chemiluminescence with TMA-6 using 5-second exposures. The blots demonstrate picogram sensitivity over a 2-hour period; A = 5000 pg, B = 1000 pg, C = 180 pg, D = 30 pg, and E = 5 pg. Time points depicted are (a) 11 minutes, and (b) 120 minutes (click to enlarge).

A further measure of the long-term stability of the TMA-6 working solution was provided by serial Western blot assays carried out over a 1-year period. The performance comparison in the Western blot protocol described above using fresh versus 1-year-old TMA-6 working solutions is depicted (see Figure 9). The results show no diminution in blotting assay sensitivity.

Discussion

Lumigen PS-atto and TMA-6 are new reagents for chemiluminescent detection of peroxidase enzymes and peroxidase conjugates. The working reagent solution for each of these reagents is prepared by combining two aqueous solutions in a 1:1 ratio and contains the proprietary substrate, the peroxide, and an enhancer in buffer solution. Reaction of a peroxidase enzyme with the reagents leads to very rapid generation of peak light intensity. Unlike luminol-based reagents in current use, these involve essentially no buildup time to peak emission.

The development of a plateau intensity meant that, in the experiment, precise timing of light measurement was unnecessary for determining the dependence of light intensity on enzyme quantity. The light intensity was sufficient to enable detection of 10–20 mol of HRP in a 6-minute assay with PS-atto and 10–19 mol with TMA-6 using a limit-of-detection criterion of a signal-blank more than twice the standard deviation of the blank (see Table I). A log-log plot of light intensity versus moles of HRP over the experimental range of 10–16–10–20 mol exhibited a high degree of linearity. It is expected that analytical precision and dynamic range would be improved were data to be collected from a larger number of replicates on an instrument with a wider dynamic range of measurement.

Figure 9. Images comparing Western blotted beta-galactosidase detected by chemiluminescence with TMA-6 working solution that was (a) freshly prepared and (b) stored for 1 year at 4°C. They were both obtained after a 9-minute incubation of the membrane in TMA-6 and a 5-second exposure. (A = 5000 pg, B = 1000 pg, C = 180 pg, D = 30 pg, and E = 5 pg) (click to enlarge).

Although detection of peroxidase with TMA-6 in solution assays provides a sensitivity and working range nearly equivalent to those realized with PS-atto, the higher absolute light intensities with PS-atto for a given quantity of enzyme magnify somewhat the effects of nonspecific binding on blotting membranes. This can have the effect in some cases of limiting sensitivity in blotting assays. Moreover, the lower light intensity produced by TMA-6 allows the use of a higher concentration of substrate in the working reagent. This, in turn, minimizes substrate depletion by high levels of bound enzyme during extended incubation or exposure.

Reliable comparisons of detection reagents in membrane-based blotting assays are notoriously difficult. As is discussed in more detail below, quantitative measurement on blotting membranes, especially by densitometry on x-ray films, is imprecise. To infer effects of the detection reagent alone from this type of measurement requires caution. A further difficulty is presented by the influence of antibody quantities and blocking conditions on apparent sensitivity. These parameters must be optimized empirically in a blotting assay. Which parameters are optimal appears to depend on the choice of detection reagent as well as the protein analyte and other factors. Anyone intending to compare detection reagents, then, is faced with the choice of standardizing all assay variables, in which case some tests will be performed under nonoptimal conditions, or else performing a technically unequal comparison using individually optimized conditions.

Bearing in mind these caveats, TMA-6, incorporated in the ECL Advance Western blotting detection kit of Amersham Biosciences, has been compared with two luminol-based reagents and a reagent containing the acridan PS-3 for performance in a Western blot system.⁷ The TMA-6 detection reagent was shown to provide superior signal to background and to detect lower levels of protein than the other materials. Each reagent was used under individually optimized assay conditions.

Figure 10. A reaction scheme depicting the chemiluminescent oxidation reaction of Lumigen APS-5 catalyzed by AP or HRP (click to enlarge).

Chemiluminescent reagents for the highly sensitive detection of peroxidase enzymes and conjugates have been available commercially for several years. Most reagents in use constitute an enhanced luminol detection system. Examples include Amersham’s ECL, Super Signal from Pierce Biotechnology Inc. (Rockford, IL), and Lumiglo from KPL Inc. (Gaithersburg, MD), as well as reagents from Roche and Ortho-Clinical Diagnostics. Lumigen PS-1, the only previous nonluminol detection reagent for peroxidase, uses an acridan ester compound as the peroxidase substrate. All of these reagents are supplied by their manufacturers as two separate components that must be combined immediately before use in order to create a working solution. Both hydrolytic instability of the substrates in the working solution and premature reaction of the substrate with hydrogen peroxide in the mixed reagent are problems that prevent long-term storage of the solution.

The stated useful lifetimes of the working solutions for several commercial chemiluminescent peroxidase substrates were surveyed. They were found to range from “several hours” to 24 hours at room temperature. More-stable working reagent solutions offer the advantage of simplifying automated analyses.

The design of the PS-atto and TMA-6 substrates was prompted by an earlier finding that the chemiluminescent AP substrate Lumigen APS-5 also undergoes a chemiluminescent peroxidase-catalyzed reaction with peroxide (see Figure 10).⁸ Reaction of APS-5 with either enzyme generates intense chemiluminescence. Replacement of the phosphate group on the exocyclic double bond with a thioalkyl group results in a compound with greater hydrolytic stability.

More than 50 different candidate compounds in this structural class have been prepared and tested. The two compounds selected for commercial development, PS-atto for solution assays and TMA-6 for solid-phase assays, are structurally quite similar but have subtly different performance properties. Satisfactory water solubility is retained by virtue of an alkylsulfonate substituent built into each molecule. Enzymatic activity toward peroxidase is retained, as is the speed with which maximum light intensity is reached. Notable as well, nonenzymatic autoxidation of the vinyl sulfide moiety in the substrate by hydrogen peroxide is not observed, even in the final working solution.

Reagent stability of the working solutions of PS-atto and TMA-6, assessed by monitoring them for any change in chemiluminescence peak intensity after periods of storage, was markedly superior to that of other reagents in use. Working solutions of both reagents were stable at room temperature for as long as 3 weeks. Solutions could be stored at 4°C and used for several months, as shown in Figure 7. Results were controlled through comparison with freshly prepared working solutions. The component solution containing the peroxidase substrate demonstrated even longer storage stability. No detectable degradation of performance in solution assays occurred after 5 months at 40°C or 1 year at room temperature. This degree of hydrolytic stability far exceeds that of all other commercial chemiluminescent peroxidase substrates.

Chemiluminescent detection reagents for use in visualizing immobilized enzyme-antibody conjugates do not necessarily have to meet the same stringent standards of storage stability that, for example, reagents used on board an automated immunoassay analyzer must. Since analytical precision is rarely assessed formally in blotting techniques, measurement is often best viewed as semiquantitative. Key experimental variables are difficult to control: antibody quality, membrane properties, blocking protocols, and binding protocols. Moreover, densitometric analysis of x-ray film images is incapable of detecting subtle differences in two-dimensional light-intensity patterns. (Modern CCD camera imaging systems have, however, improved the ability to analyze blots somewhat.)

For these reasons, minor changes in the quality of detection reagents over time do not lead to measurable falloff in blotting-assay performance. In an assessment of the quality of stored TMA-6 working solutions in a functional blotting assay format, no difference in the results was obtained even when solutions were used in which some performance deterioration could be discerned in a more rigorous, instrument-based solution assay. Storage of TMA-6 working solutions up to a year at 4°C before use in detecting blotted proteins thus is acceptable practice.

Conclusion

PS-atto is the first chemiluminescent peroxidase substrate to offer the ability to be stored as a one-bottle ready-to-use reagent. Stability testing revealed no deterioration of reagent performance on storage of the working solution up to 3 weeks at room temperature or 4 months at 4°C. Additionally, the peroxidase substrate itself is stable for several months at room temperature. The extended storage stability of the working solution allows a single-container reagent to be employed for the chemiluminescent detection of peroxidase. Besides this obvious advantage of convenience, PS-atto can be used on automated analytical instruments without requiring the incorporation of additional pumping and mixing apparatus. The extreme sensitivity of peroxidase detection it enables allows quantification over four logs of peroxidase concentration with a detection limit conservatively estimated at 10 zmol in a 6-minute assay.

TMA-6, developed for use in blotting applications such as Western blotting, shares the advantages of PS-atto in terms of rapid signal generation and reagent stability. It allows highly sensitive detection in membrane-based blotting assays. Signal duration is extended on typical blotting membranes to provide sufficient time for optimization of imaging parameters.

References

1. RS Handley, H Akhavan-Tafti, and AP Schaap, “Chemiluminescent Detection in High Volume Ligand-Binder Assays,” Journal of Clinical Ligand Assay 20 (1998): 302–312.

2. EH Jansen, RH van den Berg, and G Zomer, “Horseradish Peroxidase as Label in Chemiluminescent Immunoassays,” in Luminescence Immunoassay and Molecular Applications, ed. K Van Dyke and R Van Dyke (Boca Raton, FL: CRC Press, 1990), 57–75.

3. H Akhavan-Tafti et al., “Lumigen PS: Chemiluminescent Detection of Horseradish Peroxidase by Enzymatic Generation of Acridinium Esters,” Clinical Chemistry 41 (1995): 1368–1369.

4. GH Thorpe and LJ Kricka, “Enhanced Chemiluminescent Assays for Horseradish Peroxidase: Characteristics and Applications,” in Bioluminescence and Chemiluminescence: New Perspectives, ed. J Schölmerich et al. (Chichester, UK: Wiley, 1987), 199–208.

5. LJ Kricka, M Cooper, and X Ji, “Synthesis and Characterization of 4-Iodophenylboronic Acid: A New Enhancer for the Horseradish Peroxidase–Catalyzed Chemiluminescence of Luminol,” Analytical Biochemistry 240 (1996): 119–125.

6. PA Easton et al., “Quantitative Model of the Enhancement of Peroxidase-Induced Luminol Luminescence,” Journal of the American Chemical Society 118 (1996): 6619–6624.

7. “Powerful Solutions for Western Blotting ECL Brochure” (Uppsala, Sweden: Amersham Biosciences, 2003) [accessed 9 April 2004] available from Internet: www5.amershambiosciences.com/applic/upp00738.nsf/vLookupDoc/224572186-J320/$file/ecl_4.pdf.

8. H Akhavan-Tafti et al., “Recent Advances in Chemiluminescent Enzyme Substrates,” in Bioluminescence and Chemiluminescence, ed. JF Case et al. (Singapore: World Scientific, 2000), 215–218. 

Hashem Akhavan-Tafti, PhD, is vice president and chief scientific officer; Wenhua Xie, Ph.D., and Renuka deSilva, PhD, are research scientists; William G. Cripps and Robert A. Eickholt are research chemists; Richard S. Handley, PhD, is director of intellectual property; Rhonda S. Linsky and Michael E. Mazelis are research chemists; and A. Paul Schaap, PhD, is president of Lumigen Inc. (Southfield, MI). The authors can be contacted at hat@lumigen.com  whx@lumigen.com, rds@lumigen.com, wgc@lumigen.com, rae@lumigen.com,  rsh@lumigen.com, rsl@lumigen.com, mem@lumigen.com, and aps@lumigen.com, respectively.

Copyright ©2004 IVD Technology