- Email: [email protected]
Antigen detection immunoassay using dipsticks and colloidal dyes
57
Journal of lmmunological Methods, 140 (1991) 57-65 © 1991 Elsevier Science Publishers B.V. 0022-1759/91/$03.50 ADONIS 002217599100193E JIM05948
Antigen detection immunoassay using dipsticks and colloidal dyes K. S n o w d e n a n d M. H o m m e l Liverpool School of Tropical Medicine, Liverpool, U.K. (Received 21 December 1990, revised received 19 February 1991, accepted 24 February 1991)
A dipstick colloidal dye immunoassay (DIA) for multiple antigen detection is described. The test combines the concepts of double antibody (Ab) sandwich ELISA, dot blotting, and colloidal particle-linked Abs to produce a dipstick test for multiple antigen (Ag) detection. Dipsticks prepared from Ab coated nitrocellulose membrane mounted on acetate strips served as the assay capture matrix. Abs absorbed to colored dye particles from a family of commercially available textile dyes ( D y e / A b reagent) served as Ag detecting reagents. DIA and enzyme labelled dot blot assays showed similar Ag detection limits down to a sensitivity of 10 n g / m l . In a pilot study, an assay designed to detect species-specific IgG for use in mosquito bloodmeal identification demonstrated the feasibility of the technique. Experiments comparing bloodmeal analysis of mosquitoes using DIA and ELISA methods showed 100% agreement. This DIA method provides an inexpensive, simple, robust test for multiple Ag detection without instrumentation suitable for a wide variety of field applications. Key words: Immunoassay; Colloidal dye; Dipstick; Antigen
Introduction
A number of immunological test methods for antigen detection have been developed in recent years for a variety of diseases, with the sandwich
Correspondence to: K. Snowden, Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L3 5QA, U.K. Abbreviations: Ag, antigen; Ab, antibody; ELISA, enzymelinked immunosorbent assay; NC, nitrocellulose membrane; Dye/Ab, colloidal dye-antibody reagent; Mab, monoclonal antibody; PBS, phosphate-buffered saline; PBSM, PBS plus 2% powdered milk (w/v); PBST, PBS plus 0.05% Tween 20; BSA, bovine serum albumin; HRP, horseradish peroxidase enzyme; pl, isoelectric point.
type ELISA test achieving wide popularity for its sensitivity and specificity. While the ELISA test has been used for a number of research and commercial purposes, it has distinct drawbacks which make it unsuitable for use in the field and in laboratories where no refrigeration or expensive spectrophotometric equipment are available. In such situations, a simple, inexpensive colorimetric assay using robust reagents and no instrumentation could have many diagnostic applications. A number of modifications of the ELISA test have been described in efforts to produce a more field-ready assay format. Dot blotting, a modification of the ELISA assay using a nitrocellulose (NC) membrane as a test matrix, is becoming widely used in simple qualitative research applications (Pappas et al., 1983; Stott, 1989). Colloidal
58 gold labelled Abs are also being used in dot blot assays to avoid use of the sometimes problematic enzyme labelled detecting Abs (Hsu, 1984; Moeremans et al., 1984). Simplified test formats using single test cassettes, cards or dipsticks have been developed (e.g., Rota-Cube test, Difco Laboratories, Detroit, MI; Test Pack hCG assay, Abbott Diagnostica, Wiesbaden, F.R.G) instead of a polystyrene ELISA plate. Dipstick based assays have been designed as a field diagnostic tool (Gezahegan, 1989) and have already been commercially exploited for isotyping monoclonal antibody supernatants (INNO LIA mouse MAB isotyping kit, Innogenetics, Antwerp, Belgium). A new technique for differential antigen (Ag) detection using a combination of these simple technologies is described here. The test utilizes the concepts of sandwich ELISA, dot blot assays, and colloidal particle-linked Abs to produce an inexpensive, robust NC based dipstick test for multiple Ag detection. In this test, dipsticks prepared from Ab coated N C membrane mounted on acetate strips served as the assay capture matrix and Abs absorbed to coloured dye particles ( D y e / A b reagent) served as detecting reagents. Coloured colloidal dye particles, from a family of commercially available textile dyes ('disperse dyes'), have been shown to be a good alternative to colloidal gold particles as Ab carriers. In a pilot study, a dipstick colloidal dye immunoassay (DIA) designed to detect speciesspecific immunoglobulins was used to demonstrate the feasibility of the technique. Speciesspecific polyclonal rabbit antisera against human or chicken IgG were used as capture Abs on the dipsticks and the same antisera were absorbed to red or blue dye particles to serve as the colored detecting reagents. The test has immediate applications as a qualitative technique for insect bloodmeal identification. Experiments comparing bloodmeal analysis of mosquitoes using DIA and ELISA methods showed 100% agreement and 100% accuracy. The method could potentially be applied to qualitative antigen detection tests for a variety of human or veterinary diseases either in clinical practice or in the field where simple technology is required.
Materials and methods
Disperse dyes Ten samples of various textile disperse dyes of red, blue, pink or yellow hues were obtained as gifts from commercial sources (Disperse Dye sources: Samaron dyes Red HBSL, Brilliant Yellow H10GF, Brilliant Pink 5B, Blue FBLN 200 from Hoechst, F.R.G.; Palanil dyes Luminous Red G, Red BF, Brilliant Blue BGF 200, Dark Blue 3RT, Yellow 3GE 200, Luminous Yellow G from BASF United Kingdom). Dye particle suspensions were prepared using a washing/centrifugation procedure similar to that previously described in the literature (Gribnau et al., 1983). Briefly, a 5% suspension of each colloidal dye in water (w/v) was washed four times by centrifuging 20,000 x g for 30 min and resuspending in the same volume of water. A final low speed centrifugation (125 x g, 30 rain) removed aggregated colloidal particles. The supernatant was decanted and 0.01% thimersol was added as a preservative to produce the stock colloidal dye particle suspension. The optimum wavelength (AYkmax) for detection and quantitation of each dye was determined using a visible spectrophotometric scan (Pye Unicam SP8-500 U V / / V I S spectrophotometer, Philips). A small volume of each dye particle suspension was solubilized in ethanol for the spectral scan (i.e., Samaron Red HBSL Xmax=510 nm, Palanil Dark Blue 3RT ~ma×=620 nm, Samaron Brilliant Yellow H10F ?~max= 445 nm, etc.). The concentration of each dye was then standardized in water so that a 1/500 dilution resulted in a spectrophotometric absorbance value of 1 at the appropriate wavelength (A~ . . . . 1 cm = 1). Concentrations of the dye used in the D y e / A b reagent were expressed as multiples of this concentration (i.e., A~kma x = 5 has five times the dye concentration as the suspension of AXmax =1).
Dye / Ab reagent Commercially available IgG fractions or affinity purified antisera from several sources specific for either human or chicken IgG were linked to several of the colloidal particle dyes using modifications of a previously described technique (Gribnau et al., 1983). Several buffer systems, dye
59 concentrations and antibody concentrations were used in various combinations in a simple absorption procedure to determine the optimum conditions for D y e / A b reagent production. A 10 mM sodium phosphate buffer with p H ranges from 4.5 to 10.5 and with 2.7 mM to 0.15 M NaC1 concentrations were used to determine the optimum dye and Ab adsorbing conditions. Various concentrations of dye particles from A = 5 to A = 25 were used in initial studies attempting to link the dyes to antibodies in concentrations ranging from 5 to 100 /~g/ml protein concentration. Optimal D y e / A b suspensions which gave strongly coloured positive dots without increased background colour used dyes at a final concentration of A~kma x = 10 and affinity purified antibody protein concentrations of 10 /tg/ml. The optimum D y e / A b absorbing conditions were 10 mM PO 4 buffers with 10 mM or less NaC1 at pH 7.0-7.5. To prepare the D y e / A b reagents used in subsequent assays, appropriate volumes of dye and Ab were incubated in a final concentration of 10 mM phosphate 2.7 mM NaC1 buffer pH 7.4 (V volume) for 1 h to allow for Ab absorption to colloidal dye particles. The solution was then spiked with 1I/5 volume of a 30% bovine serum albumin (BSA) solution in 5 mM NaC1 pH 7.4 and incubated for another hour to stabilize the dye particle surfaces. The D y e / A b reagents were centrifuged at 12,000 x g and the pellet was resuspended in a 33.3 mM phosphate, 0.125 M NaC1 solution p H 7.4, containing 5% BSA and 0.01% thiomersol. If the D y e / A b reagent was used individually it was resuspended to a final concentration of A ~ k m a x = 10. If different D y e / A b reagents were pooled to produce a suspension detecting more than one antigen, each component had a final concentration of A~ma x = 10. D y e / A b reagents could be stored at 4 ° C for 7-10 days, lyophilized or frozen to - 2 0 ° C. Several compounds including polyethylene glycol, polyvinyl pyrrolidone, dextran, and gelatin were used unsuccessfully in preliminary experiments as stabilizing agents for the D y e / A b reagents instead of BSA.
Dipsticks Dipsticks 1 X 8 cm in size were prepared by adhering N C membrane to clear acetate sheets using double sided sellotape. The strips were de-
signed with various numbers of spaces on N C for species-specific antisera dots and with a free acetate end for labelling. Dipsticks used various brands of commercially available Rb antichicken IgG or Rb antihuman IgG antiserum as a capture layer by allowing 3/~1 dots of a 1 / 1 0 0 - 1 / 4 0 0 0 dilution or 0 . 5 - 2 0 / ~ g / m l protein (depending on the antiserum) to dry on the NC strips. Strips could be used immediately, stored dry at room temperature or 4 ° C, or blocked with PBSM (see below) and then stored dry. Preliminary experiments conducted with nylon (Zetaprobe, Bio-Rad, Hertfordshire, U.K.), polycarbonate (Nuclepore, Pleasanton, CA), PVFD ( I M M O B I L O N P, Millipore, Bedford, MA) or reinforced N C (Biotrace, Gelman Corp., Ann Arbor, MI) membranes did not prove more sensitive or specific than N C membrane (Schleicher and Schuell, Dassel, F.R.G.) which was used for the remainder of the experimental protocols (Penny, 1989; Stott, 1989).
Blocking agents After the application of Ab dots, the remaining protein binding sites on the NC surface were blocked with powdered milk, BSA, casein or Tween 20 (Sigma) as 1% or 2% solutions in phosphate buffered saline p H 7.4 (PBS) using 10-60 min incubation times. All solutions except Tween 20 were effective in blocking non-specific binding. Subsequent tests used PBSM with 30 minute incubations for N C surface blocking.
Dipstick test method To standardize the DIA, dipsticks prepared with dots of various brands of antisera were assayed using pooled normal sera from chicken or human diluted 1/1000 to 1/40,000 in PBS as antigens. D y e / A b reagents using various combinations of dye particles and antisera were used to detect the presence of Ag in order to select the most effective combinations of dye and antisera. To test the sensitivity and specificity of the assay, replicate dipsticks were used to compare the DIA and traditional sandwich dot blots using H R P enzyme labelled antisera as a detecting reagent. Methods used in one typical experiment are described in detail. Dipsticks were prepared with Rb antichicken IgG dots (affinity purified Rb antichicken IgG, 2.2 m g / m l , Sigma, St. Louis, MO) in
60 concentrations ranging from 0.5 to 20 /~g/ml Ab and blocked with PBSM as previously described. Seven purified chicken IgG standards (Sigma, St. Louis, MO) diluted from 1 0 / t g / m l to 10 n g / m l in PBS were used as antigens. The dipsticks were incubated in the various Ag dilutions for 1 h and washed for 15 s under gently running tap water. Sets of dipsticks were placed in either of the D y e / A b detecting reagents (Palanil R e d / R b anti-chicken IgG or Samaron B l u e / R b antichicken IgG reagents A = 10, 10 ~ g / m l Ab protein) for 1 h. Dipsticks were then briefly rinsed in tap water, and dots of the appropriate red or blue color and location were observed. For comparative purposes, one set of dipsticks was incubated in commercially available H R P labelled affinity purified Rb anti-chicken antisera (Sigma, St. Louis, MO) at the recommended 1/1000 dilution in PBS for 1 h. After washing for 15 s in tapwater, brown coloured dots were detected using a 3,3'-diaminobenzidine substrate (0.5 m g / m l ) with 0.01% H202 in PBS. Similar experiments using Rb antihuman IgG reagents were conducted. Mosquito bloodmeal detection To compare the DIA and ELISA tests, and to use mixed D y e / A b detecting reagents in a practical test, a mosquito bloodmeal detection assay was developed. A blind study compared the DIA with a previously described sandwich ELISA test for bloodmeal analysis (Service et al., 1986) for 50 mosquitoes. Laboratory raised Aedes aegypti mosquitoes were fed on chicken, human or guinea pig bloodmeals or were unfed. Smears were prepared by squashing individual mosquitoes on Whatman no. 1 filter paper and were stored desiccated at 4 o C until use. Each mosquito smear from the filter paper was eluted in 5 ml PBS for at least 1 h at 2 2 ° C with occasional agitation to produce the mosquito eluate used for both DIA and ELISA. Both assays used the same Rb anti-chicken IgG and Rb anti-human IgG (affinity purified Rb anti-human IgG, 1.6 m g / m l , Jackson Immunoresearch Laboratories, West Grove, PA) antisera. PBS eluates from the same 50 laboratory raised mosquitoes were used as Ag sources in both tests. E L I S A test method Polystyrene ELISA plates (Nunc brand, Gibco, Paisley, Scotland) were coated with Rb anti-
chicken IgG or Rb anti-human IgG antisera in carbonate buffer, p H 9.6, overnight, washed with PBST, and blocked with PBSM for 2 h. After washing again with PBST, 100/~1 of the mosquito eluates were added to the wells for 1 h. After further PBST washes, commercially available H R P labelled affinity purified Rb anti-chicken IgG (Sigma, St. Louis, MO) and H R P labelled affinity purified Rb anti-human IgG (ICN Immunobiologicals, Lisle, IL) reagents were used as detecting antibodies at the recommended 1/1000 dilutions in the ELISA. Positive mosquito bloodmeals were detected visually using an o-phenylene diamine substrate (0.3 m g / m l ) with 0.01% H202 in 0.1 M citrate/phosphate buffer p H 5.0. The protocol of the DIA for differentiating between different host sources of mosquito bloodmeal was similar to that described above (Fig. 1). Dipsticks were prepared using 3 /xl dots of Rb anti-chicken IgG and Rb anti-human IgG at 1/500 dilutions and blocked with PBSM. Normal chicken or human serum dilutions of 1/5000 were used as positive controls. The purple colouted detecting reagent was a combination including Palanil Red BF d y e / R b anti-chicken IgG and Samaron Blue d y e / R b anti-human IgG in equal volumes (each dye final concentration A~ m a x ---" 1 0 , each Ab final concentration 10/~g/ml). Therefore, two possible antigens were potentially detected simultaneously. Dipsticks were then briefly rinsed in tap water, and dots of the appropriate colour and location were observed on the dipstick identifying the correct species of mosquito bloodmeal or serum (i.e., red dot = chicken bloodmeal, blue dot = human bloodmeal). All reagent and dipstick production, DIA and ELISA assays were carried out at room temperature.
Results
The DIA is a simple, robust assay that compares favourably with dot blot and ELISA assays using enzyme labelled Abs. Once the dipstick is prepared, the assay uses only two simple steps, first incubating in an Ag containing solution followed by the colloidal D y e / A b suspension detecting reagent (Fig. 1). Qualitative positive Ag detection for multiple Ags is simultaneously demon-
61 IgG reagents. No crossreactivity was detected between human and chicken IgG at the highest Ag concentrations. To illustrate the durability of the DIA reagents, Fig. 2B shows the same assay using dipsticks that have been stored for 15 days at room temperature and with the same Palanil R e d / R b anti-chicken IgG detecting reagent that has been frozen and thawed. To demonstrate the use of the DIA, a simple mosquito bloodmeal identification test was designed and compared with a previously described ELISA assay. Table I indicates the results of the bloodmeal identification comparison between the DIA test and a previously developed ELISA assay. Both the tests were in agreement and were 100% correct in identifying human or chicken bloodmeals using 50 laboratory raised mosquitoes.
strated directly by the presence of different coloured dots at the appropriate locations on the dipstick. Fig. 2A illustrates the sensitivity and optimization of the DIA assay by comparing replicate dipsticks using a typical sandwich dot blot system with HRP-labelled detecting Abs. Dipsticks showed similar sensitivities indicated by clearly visible red, brown or blue dots using two different D y e / A b combinations (Palanil Red B F / R b anti-chicken IgG or Samaron Blue/Rb anti-chicken IgG) or enzyme labelled antisera. The Ag detection limits were as sensitive as 10 ng/ml using purified chicken IgG as an antigen source and either the DIA reagents or enzyme labelled detecting Abs when using dipstick capture dots of affinity purified Ab as dilute as 5/~g/ml protein. Similar results were obtained using anti-human
(9
YRbantihumanigo
v
:
v
!
, OR
,~
R b antihuman igG
.~
"~
C!~~G
~
enIgG
v
O:
AND
/
~
~
,~(~
( ~ H u m a n IgG
O
®
®
Fig. 1. Assay protocol and schematic diagram of dipstick design for differentiating between different host sources of mosquito bloodmeals. 1: dipstick dotted with species-specificantisera as a capture layer; 2: non-specificprotein binding to NC blockedwith PBSM and tap water rinsed; 3: dipstick incubated in Ag containing PBS solution (IgG standard or mosquitobloodmeaieluate) and tap water rinsed; 4: dipstick incubated in multiple Ag detecting reagent (Palanil Red/Rb anti-chicken IgG and Sarnaron Blue/Rb anti-human IgG) and tap water rinsed; 5: dipsticks observed for red or blue dots at appropriate locations indicating positive Ag detection results.
62 N o cross-reactivity was detected with guinea pig fed mosquitoes. Fig. 3 shows dipsticks recognizing human or chicken bloodmeals using a pooled purple suspension containing Palanil R e d / R b antichicken IgG and Samaron B l u e / R b anti-human detecting reagents. 50 dipsticks (for 50 mosquitoes) were processed in a total volume of 5 ml of D y e / A b reagent without significant loss of colour intensity in the positive dots.
Discussion The D I A combines a number of previously described techniques to produce a new, sensitive, inexpensive, robust test utilizing simple technology. Since the technique requires no instrumentation for qualitative multiple Ag detection, it has many potential field applications. The production of the detecting D y e / A b reagents is a simple absorption procedure using common buffer solutions. The selection of appropriate disperse dyes can only be made by screening an assortment of dyes since each coloured colloidal dye suspension has a unique Ab binding capacity. Ten samples of different disperse dyes were evaluated for their suitability for use by absorbing several different Abs specific for either human or chicken IgG to each dye and comparing the intensity of positive coloured dots on the dipstick. Of the ten dyes screened, Palanil Red BF (BASF, U.K.), Samaron Red HBSL, Samaron Brilliant Pink and Samaron Blue FBLN (Hoechst, F.R.G), consistently produced effective D y e / A b reagents using several goat or Rb source antisera. Several yellow dyes labelled adequately with Ab, although positive yellow dots on the white NC background were sometimes difficult to visualize. These observations agreed with the previously published comment that disperse dyes suitable for D I A reagents
TABLE I BLOODMEAL IDENTIFICATION OF FRESHLY FED A E G Y P T I MOSQUITOES USING DIA AND ELISA ASSAYS AEDES
Both methods were 100% correct in identifying human or chicken bloodmeals. no. of mosquitoes
Host species of bloodmeal
no. of bloodmeals identifiedby:
20 20 5 5
Human Chicken Guinea pig Unfed
DIA 20 20 0 0
ELISA 20 20 0 0
50
can only be found by screening and that selection of a particular dye depended on the Ab preparation (Gribnau et al., 1983). If textile dyes were mixed to produce orange, green or purple hues before labelling, the Abs sometimes bound in differing proportions to each coloured dye so that the hue of the coloured dot on the dipstick did not necessarily match the hue of the D y e / A b suspension. Since a wide variety of coloured disperse dyes are commercially available it is probably most appropriate to select the desired colour of dye particle initially rather than to attempt to mix colours to generate new hues for D I A reagent production. The preparation and use of D y e / A b reagents is similar to and even simpler than Ab linked colloidal gold particle probes (DeMay, 1983). Adequate Dye-Ab absorption, as judged by the production of appropriate positive dots on a dipstick, occurred at a variety of pHs ranging from 6.0 to 8.5. The ionic strengths of the absorption buffers were more critical than pH. Exceeding a salt con-
Fig. 2.A: sensitivity and optimization of DIA. Replicate dipsticks were dotted with various dilutions of capture antibody (Rb anti-chicken IgG, 0.5 to 20 #g/ml), blocked with PBSM, and incubated in various Ag concentrations (10 #g/ml to 10 ng/ml) of purified chicken IgG standard. Ag detection is indicated by red dots on dipsticks A-G using Palanil Red/Rb anti-chicken IgG, brown dots on dipsticks H-N using HRP labelled Rb antichicken IgG, and blue dots on dipsticks O-U using Samaron Blue/Rb antichicken IgG. Dipsticks, G, N, and O are PBS controls. B: durabihty of DIA. Assaywas conducted as in A using stored dipsticks and Palanil Red/Rb antichicken IgG reagent that had been frozen and thawed. Fig. 3. Bloodmeal detection DIA. Dipsticks indicate positive identification of chicken (A) and human (B) bloodmeals from laboratory fed mosquitoes.
63
Ill
~
0.5 1
10
~ 2o /~g/ml ng/mi 2A
~g/ml
ng/ml
#g/ml
ng/ml
Ag CONCENTRATION
,O,
0.5 1
~ 10
~ 20 3
~tg/ml 2B
ng/ml
Ag CONCENTRATION
B
64 centration of 15 mM before or during dye particle-Ab absorption resulted in the disruption of the colloidal suspension and irreversible settling out of the dye particles. A pH of 7.2-7.4 was used in the POa/NaC1 buffer because of its physiologic value in maintaining Abs at optimum conditions. The importance of pH during D y e / A b absorption needs to be further explored if Mabs are used instead of affinity purified polyvalent Abs. If this colloidal particle/Ab system resembles Ab labelled colloidal gold particle production, the p I of the Mab should be considered since the absorption of the Ab to the gold particle is optimized near the p I of the Mab (DeMay, 1983). While BSA and PEG have been used to stabilize g o l d / A b suspensions, only BSA was effectively used in preliminary experiments with the colloidal D y e / A b reagent. Textile disperse dyes are produced widely in the textile chemistry industry and are typically marketed in kilogram quantities. Consequently costs for the production of DIA reagents are negligible compared with colloidal gold products. The D y e / A b reagent production has been previously described and utilized in sandwich type assays detecting human chorionic gonadotropin or human placental lactogen (Gribnau et al., 1982, 1983). The assay used Ab coated microtitre plates to capture purified hormone Ag. The Ag was detected by the D y e / A b detecting reagent and colour was produced in the microtitre wells by solubilizing the bound dye using ethanol. The optical density of reaction wells was read in a similar manner to ELISA tests. To our knowledge this is the first report of the use of this D y e / A b reagent for direct colour visualization on a solid matrix such as nitrocellulose. The test described here is a simpler, faster qualitative Ag detection test using the D y e / A b reagent in a different manner than previously described. Dipsticks which had been dotted with capture Ab, blocked with PBSM, dried and stored at 22 ° C maintained reactivity for at least 3 months (data not shown). The D y e / A b reagents maintained reactivity for at least 1 week when stored at 4 ° C as a liquid suspension. Unlike gold-labelled reagents, they could also be frozen to - 2 0 ° C or lyophilized for longterm storage. The D y e / A b
reagents maintained their colloidal suspension and reactivity after freezing and thawing. Lyophilized samples maintained their reactivity when reconstituted, but occasionally did not completely redisperse so that the optical densities of the reagents were slightly decreased. Replicate dipsticks comparing D y e / A b and enzyme labelled detecting reagents consistently gave similar Ag detection sensitivities as low as 10 n g / m l indicating that the DIA method gives results comparable to typical sandwich dot blot systems. Several brands of antisera were screened for their sensitivity and lack of cross-reactivity between IgG from human and chicken. Antisera concentrations as low as 10 /Lg/ml produced effective D y e / A b reagents if affinity purified antisera were used. Not surprisingly, if whole serum or the IgG fraction of antisera were used, protein concentrations up to 100/~g/ml were required to produce effective D y e / A b suspensions. In some cases when whole serum or the IgG fraction of antisera were used as capture Ab dots, some crossreactivity between human and chicken IgG detection occurred with both the DIA and the H R P labelled detecting reagents. Predictably, affinity purified IgG fractions of antisera to human or chicken IgG lacked cross-reactivity problems and provided the best results in DIA, enzyme dot blot and ELISA comparative assays. To demonstrate the use of the DIA, a simple mosquito bloodmeal identification test was designed and compared with a previously described ELISA assay. In this test comparison, the DIA was faster, much easier to perform and used fewer Ab reagents than the ELISA method. It had the added advantage that two or more Ags (chicken or human IgG) were potentially detected using a single test, rather than performing an ELISA test to detect each of the potential species providing bloodmeals. The dipstick also serves as a simple permanent test record. The DIA provides a simple robust multiple Ag detection system with a wide range of possible applications. The assay needs to be further developed by adapting other sandwich antigen capture systems to this method, and needs to be tested under field rather than laboratory conditions.
65
Acknowledgements We are grateful to Dr. M.W. Service for providing bloodfed mosquitoes and the Wolfson Foundation for funding K.S.
References DeMay, J. (1983) Colloidal gold probes in immunocytochemistry In: J.M. Polak and S. Van Noorden (Eds.), Immunocytochemistry, Practical Applications in Pathology and Biology. Wright PSG, Bristol, pp. 82-112. Gezahegan, T. (1989) Development of a dipstick method for mosquito blood-meal identification and comparison with ELISA for potential field use. MSc Thesis, Liverpool School of Tropical Medicine, Liverpool, U.K. Gribnau, T., Roeles, F., v.d. Biezen, J., Leuvering, J. and Schuurs, A. (1982) The application of colloidal dye particles as label in immunoassays: disperse(d) dye immunoassay ("DIA") In: T.C.J. Gribnau, J. Visser and R.J.F. Nivard (Eds.), Affinity Chromatography and Related Techniques. Elsevier, Amsterdam, pp. 411-424.
Gribnau, T., van Sommeren, A. and Van Dinther, F. (1983) DIA-disperse dye immunoassay. In: I.M. Chaiken, M. Wilchek and I. Parikh (Eds.), Affinity Chromatography and Biological Recognition. Academic Press, Orlando, FL, pp. 375-380. Hsu, Y. (1984) Immunogold for detection of antigen on nitrocellulose paper. Analyt. Biochem. 142, 221-225. Moeremans, M., Daneels, G., VanDijck, A., Langanger, G. and DeMey, J. (1984) Sensitive visualization of antigen-antibody reactions in dot and blot immune overlay assays with immunogold and immunogold/silver staining. J. Immunol. Methods 74, 353-360. Penney, C.L., Gauldie, J., Evelegh, M., Penney, M.T., Chong, D. and Horsewood, P. (1989) Polycarbonate membranes: a novel surface for solid-phase determination with utility in field format serological assays. J. Immunol. Methods 123, 185-192. Service, M.W., Voller, A. and Bidwell, D.E. (1986) The enzyme-linked immunosorbent assay (ELISA) test for the identification of blood-meals of haematophagous insects. Bull. Ent. Res. 76, 321-330. Stott, D.I. (1989) Immunoblotting and dot blotting. J. Immunol. Methods. 119, 153-187.
Journal of lmmunological Methods, 140 (1991) 57-65 © 1991 Elsevier Science Publishers B.V. 0022-1759/91/$03.50 ADONIS 002217599100193E JIM05948
Antigen detection immunoassay using dipsticks and colloidal dyes K. S n o w d e n a n d M. H o m m e l Liverpool School of Tropical Medicine, Liverpool, U.K. (Received 21 December 1990, revised received 19 February 1991, accepted 24 February 1991)
A dipstick colloidal dye immunoassay (DIA) for multiple antigen detection is described. The test combines the concepts of double antibody (Ab) sandwich ELISA, dot blotting, and colloidal particle-linked Abs to produce a dipstick test for multiple antigen (Ag) detection. Dipsticks prepared from Ab coated nitrocellulose membrane mounted on acetate strips served as the assay capture matrix. Abs absorbed to colored dye particles from a family of commercially available textile dyes ( D y e / A b reagent) served as Ag detecting reagents. DIA and enzyme labelled dot blot assays showed similar Ag detection limits down to a sensitivity of 10 n g / m l . In a pilot study, an assay designed to detect species-specific IgG for use in mosquito bloodmeal identification demonstrated the feasibility of the technique. Experiments comparing bloodmeal analysis of mosquitoes using DIA and ELISA methods showed 100% agreement. This DIA method provides an inexpensive, simple, robust test for multiple Ag detection without instrumentation suitable for a wide variety of field applications. Key words: Immunoassay; Colloidal dye; Dipstick; Antigen
Introduction
A number of immunological test methods for antigen detection have been developed in recent years for a variety of diseases, with the sandwich
Correspondence to: K. Snowden, Liverpool School of Tropical Medicine, Pembroke Place, Liverpool L3 5QA, U.K. Abbreviations: Ag, antigen; Ab, antibody; ELISA, enzymelinked immunosorbent assay; NC, nitrocellulose membrane; Dye/Ab, colloidal dye-antibody reagent; Mab, monoclonal antibody; PBS, phosphate-buffered saline; PBSM, PBS plus 2% powdered milk (w/v); PBST, PBS plus 0.05% Tween 20; BSA, bovine serum albumin; HRP, horseradish peroxidase enzyme; pl, isoelectric point.
type ELISA test achieving wide popularity for its sensitivity and specificity. While the ELISA test has been used for a number of research and commercial purposes, it has distinct drawbacks which make it unsuitable for use in the field and in laboratories where no refrigeration or expensive spectrophotometric equipment are available. In such situations, a simple, inexpensive colorimetric assay using robust reagents and no instrumentation could have many diagnostic applications. A number of modifications of the ELISA test have been described in efforts to produce a more field-ready assay format. Dot blotting, a modification of the ELISA assay using a nitrocellulose (NC) membrane as a test matrix, is becoming widely used in simple qualitative research applications (Pappas et al., 1983; Stott, 1989). Colloidal
58 gold labelled Abs are also being used in dot blot assays to avoid use of the sometimes problematic enzyme labelled detecting Abs (Hsu, 1984; Moeremans et al., 1984). Simplified test formats using single test cassettes, cards or dipsticks have been developed (e.g., Rota-Cube test, Difco Laboratories, Detroit, MI; Test Pack hCG assay, Abbott Diagnostica, Wiesbaden, F.R.G) instead of a polystyrene ELISA plate. Dipstick based assays have been designed as a field diagnostic tool (Gezahegan, 1989) and have already been commercially exploited for isotyping monoclonal antibody supernatants (INNO LIA mouse MAB isotyping kit, Innogenetics, Antwerp, Belgium). A new technique for differential antigen (Ag) detection using a combination of these simple technologies is described here. The test utilizes the concepts of sandwich ELISA, dot blot assays, and colloidal particle-linked Abs to produce an inexpensive, robust NC based dipstick test for multiple Ag detection. In this test, dipsticks prepared from Ab coated N C membrane mounted on acetate strips served as the assay capture matrix and Abs absorbed to coloured dye particles ( D y e / A b reagent) served as detecting reagents. Coloured colloidal dye particles, from a family of commercially available textile dyes ('disperse dyes'), have been shown to be a good alternative to colloidal gold particles as Ab carriers. In a pilot study, a dipstick colloidal dye immunoassay (DIA) designed to detect speciesspecific immunoglobulins was used to demonstrate the feasibility of the technique. Speciesspecific polyclonal rabbit antisera against human or chicken IgG were used as capture Abs on the dipsticks and the same antisera were absorbed to red or blue dye particles to serve as the colored detecting reagents. The test has immediate applications as a qualitative technique for insect bloodmeal identification. Experiments comparing bloodmeal analysis of mosquitoes using DIA and ELISA methods showed 100% agreement and 100% accuracy. The method could potentially be applied to qualitative antigen detection tests for a variety of human or veterinary diseases either in clinical practice or in the field where simple technology is required.
Materials and methods
Disperse dyes Ten samples of various textile disperse dyes of red, blue, pink or yellow hues were obtained as gifts from commercial sources (Disperse Dye sources: Samaron dyes Red HBSL, Brilliant Yellow H10GF, Brilliant Pink 5B, Blue FBLN 200 from Hoechst, F.R.G.; Palanil dyes Luminous Red G, Red BF, Brilliant Blue BGF 200, Dark Blue 3RT, Yellow 3GE 200, Luminous Yellow G from BASF United Kingdom). Dye particle suspensions were prepared using a washing/centrifugation procedure similar to that previously described in the literature (Gribnau et al., 1983). Briefly, a 5% suspension of each colloidal dye in water (w/v) was washed four times by centrifuging 20,000 x g for 30 min and resuspending in the same volume of water. A final low speed centrifugation (125 x g, 30 rain) removed aggregated colloidal particles. The supernatant was decanted and 0.01% thimersol was added as a preservative to produce the stock colloidal dye particle suspension. The optimum wavelength (AYkmax) for detection and quantitation of each dye was determined using a visible spectrophotometric scan (Pye Unicam SP8-500 U V / / V I S spectrophotometer, Philips). A small volume of each dye particle suspension was solubilized in ethanol for the spectral scan (i.e., Samaron Red HBSL Xmax=510 nm, Palanil Dark Blue 3RT ~ma×=620 nm, Samaron Brilliant Yellow H10F ?~max= 445 nm, etc.). The concentration of each dye was then standardized in water so that a 1/500 dilution resulted in a spectrophotometric absorbance value of 1 at the appropriate wavelength (A~ . . . . 1 cm = 1). Concentrations of the dye used in the D y e / A b reagent were expressed as multiples of this concentration (i.e., A~kma x = 5 has five times the dye concentration as the suspension of AXmax =1).
Dye / Ab reagent Commercially available IgG fractions or affinity purified antisera from several sources specific for either human or chicken IgG were linked to several of the colloidal particle dyes using modifications of a previously described technique (Gribnau et al., 1983). Several buffer systems, dye
59 concentrations and antibody concentrations were used in various combinations in a simple absorption procedure to determine the optimum conditions for D y e / A b reagent production. A 10 mM sodium phosphate buffer with p H ranges from 4.5 to 10.5 and with 2.7 mM to 0.15 M NaC1 concentrations were used to determine the optimum dye and Ab adsorbing conditions. Various concentrations of dye particles from A = 5 to A = 25 were used in initial studies attempting to link the dyes to antibodies in concentrations ranging from 5 to 100 /~g/ml protein concentration. Optimal D y e / A b suspensions which gave strongly coloured positive dots without increased background colour used dyes at a final concentration of A~kma x = 10 and affinity purified antibody protein concentrations of 10 /tg/ml. The optimum D y e / A b absorbing conditions were 10 mM PO 4 buffers with 10 mM or less NaC1 at pH 7.0-7.5. To prepare the D y e / A b reagents used in subsequent assays, appropriate volumes of dye and Ab were incubated in a final concentration of 10 mM phosphate 2.7 mM NaC1 buffer pH 7.4 (V volume) for 1 h to allow for Ab absorption to colloidal dye particles. The solution was then spiked with 1I/5 volume of a 30% bovine serum albumin (BSA) solution in 5 mM NaC1 pH 7.4 and incubated for another hour to stabilize the dye particle surfaces. The D y e / A b reagents were centrifuged at 12,000 x g and the pellet was resuspended in a 33.3 mM phosphate, 0.125 M NaC1 solution p H 7.4, containing 5% BSA and 0.01% thiomersol. If the D y e / A b reagent was used individually it was resuspended to a final concentration of A ~ k m a x = 10. If different D y e / A b reagents were pooled to produce a suspension detecting more than one antigen, each component had a final concentration of A~ma x = 10. D y e / A b reagents could be stored at 4 ° C for 7-10 days, lyophilized or frozen to - 2 0 ° C. Several compounds including polyethylene glycol, polyvinyl pyrrolidone, dextran, and gelatin were used unsuccessfully in preliminary experiments as stabilizing agents for the D y e / A b reagents instead of BSA.
Dipsticks Dipsticks 1 X 8 cm in size were prepared by adhering N C membrane to clear acetate sheets using double sided sellotape. The strips were de-
signed with various numbers of spaces on N C for species-specific antisera dots and with a free acetate end for labelling. Dipsticks used various brands of commercially available Rb antichicken IgG or Rb antihuman IgG antiserum as a capture layer by allowing 3/~1 dots of a 1 / 1 0 0 - 1 / 4 0 0 0 dilution or 0 . 5 - 2 0 / ~ g / m l protein (depending on the antiserum) to dry on the NC strips. Strips could be used immediately, stored dry at room temperature or 4 ° C, or blocked with PBSM (see below) and then stored dry. Preliminary experiments conducted with nylon (Zetaprobe, Bio-Rad, Hertfordshire, U.K.), polycarbonate (Nuclepore, Pleasanton, CA), PVFD ( I M M O B I L O N P, Millipore, Bedford, MA) or reinforced N C (Biotrace, Gelman Corp., Ann Arbor, MI) membranes did not prove more sensitive or specific than N C membrane (Schleicher and Schuell, Dassel, F.R.G.) which was used for the remainder of the experimental protocols (Penny, 1989; Stott, 1989).
Blocking agents After the application of Ab dots, the remaining protein binding sites on the NC surface were blocked with powdered milk, BSA, casein or Tween 20 (Sigma) as 1% or 2% solutions in phosphate buffered saline p H 7.4 (PBS) using 10-60 min incubation times. All solutions except Tween 20 were effective in blocking non-specific binding. Subsequent tests used PBSM with 30 minute incubations for N C surface blocking.
Dipstick test method To standardize the DIA, dipsticks prepared with dots of various brands of antisera were assayed using pooled normal sera from chicken or human diluted 1/1000 to 1/40,000 in PBS as antigens. D y e / A b reagents using various combinations of dye particles and antisera were used to detect the presence of Ag in order to select the most effective combinations of dye and antisera. To test the sensitivity and specificity of the assay, replicate dipsticks were used to compare the DIA and traditional sandwich dot blots using H R P enzyme labelled antisera as a detecting reagent. Methods used in one typical experiment are described in detail. Dipsticks were prepared with Rb antichicken IgG dots (affinity purified Rb antichicken IgG, 2.2 m g / m l , Sigma, St. Louis, MO) in
60 concentrations ranging from 0.5 to 20 /~g/ml Ab and blocked with PBSM as previously described. Seven purified chicken IgG standards (Sigma, St. Louis, MO) diluted from 1 0 / t g / m l to 10 n g / m l in PBS were used as antigens. The dipsticks were incubated in the various Ag dilutions for 1 h and washed for 15 s under gently running tap water. Sets of dipsticks were placed in either of the D y e / A b detecting reagents (Palanil R e d / R b anti-chicken IgG or Samaron B l u e / R b antichicken IgG reagents A = 10, 10 ~ g / m l Ab protein) for 1 h. Dipsticks were then briefly rinsed in tap water, and dots of the appropriate red or blue color and location were observed. For comparative purposes, one set of dipsticks was incubated in commercially available H R P labelled affinity purified Rb anti-chicken antisera (Sigma, St. Louis, MO) at the recommended 1/1000 dilution in PBS for 1 h. After washing for 15 s in tapwater, brown coloured dots were detected using a 3,3'-diaminobenzidine substrate (0.5 m g / m l ) with 0.01% H202 in PBS. Similar experiments using Rb antihuman IgG reagents were conducted. Mosquito bloodmeal detection To compare the DIA and ELISA tests, and to use mixed D y e / A b detecting reagents in a practical test, a mosquito bloodmeal detection assay was developed. A blind study compared the DIA with a previously described sandwich ELISA test for bloodmeal analysis (Service et al., 1986) for 50 mosquitoes. Laboratory raised Aedes aegypti mosquitoes were fed on chicken, human or guinea pig bloodmeals or were unfed. Smears were prepared by squashing individual mosquitoes on Whatman no. 1 filter paper and were stored desiccated at 4 o C until use. Each mosquito smear from the filter paper was eluted in 5 ml PBS for at least 1 h at 2 2 ° C with occasional agitation to produce the mosquito eluate used for both DIA and ELISA. Both assays used the same Rb anti-chicken IgG and Rb anti-human IgG (affinity purified Rb anti-human IgG, 1.6 m g / m l , Jackson Immunoresearch Laboratories, West Grove, PA) antisera. PBS eluates from the same 50 laboratory raised mosquitoes were used as Ag sources in both tests. E L I S A test method Polystyrene ELISA plates (Nunc brand, Gibco, Paisley, Scotland) were coated with Rb anti-
chicken IgG or Rb anti-human IgG antisera in carbonate buffer, p H 9.6, overnight, washed with PBST, and blocked with PBSM for 2 h. After washing again with PBST, 100/~1 of the mosquito eluates were added to the wells for 1 h. After further PBST washes, commercially available H R P labelled affinity purified Rb anti-chicken IgG (Sigma, St. Louis, MO) and H R P labelled affinity purified Rb anti-human IgG (ICN Immunobiologicals, Lisle, IL) reagents were used as detecting antibodies at the recommended 1/1000 dilutions in the ELISA. Positive mosquito bloodmeals were detected visually using an o-phenylene diamine substrate (0.3 m g / m l ) with 0.01% H202 in 0.1 M citrate/phosphate buffer p H 5.0. The protocol of the DIA for differentiating between different host sources of mosquito bloodmeal was similar to that described above (Fig. 1). Dipsticks were prepared using 3 /xl dots of Rb anti-chicken IgG and Rb anti-human IgG at 1/500 dilutions and blocked with PBSM. Normal chicken or human serum dilutions of 1/5000 were used as positive controls. The purple colouted detecting reagent was a combination including Palanil Red BF d y e / R b anti-chicken IgG and Samaron Blue d y e / R b anti-human IgG in equal volumes (each dye final concentration A~ m a x ---" 1 0 , each Ab final concentration 10/~g/ml). Therefore, two possible antigens were potentially detected simultaneously. Dipsticks were then briefly rinsed in tap water, and dots of the appropriate colour and location were observed on the dipstick identifying the correct species of mosquito bloodmeal or serum (i.e., red dot = chicken bloodmeal, blue dot = human bloodmeal). All reagent and dipstick production, DIA and ELISA assays were carried out at room temperature.
Results
The DIA is a simple, robust assay that compares favourably with dot blot and ELISA assays using enzyme labelled Abs. Once the dipstick is prepared, the assay uses only two simple steps, first incubating in an Ag containing solution followed by the colloidal D y e / A b suspension detecting reagent (Fig. 1). Qualitative positive Ag detection for multiple Ags is simultaneously demon-
61 IgG reagents. No crossreactivity was detected between human and chicken IgG at the highest Ag concentrations. To illustrate the durability of the DIA reagents, Fig. 2B shows the same assay using dipsticks that have been stored for 15 days at room temperature and with the same Palanil R e d / R b anti-chicken IgG detecting reagent that has been frozen and thawed. To demonstrate the use of the DIA, a simple mosquito bloodmeal identification test was designed and compared with a previously described ELISA assay. Table I indicates the results of the bloodmeal identification comparison between the DIA test and a previously developed ELISA assay. Both the tests were in agreement and were 100% correct in identifying human or chicken bloodmeals using 50 laboratory raised mosquitoes.
strated directly by the presence of different coloured dots at the appropriate locations on the dipstick. Fig. 2A illustrates the sensitivity and optimization of the DIA assay by comparing replicate dipsticks using a typical sandwich dot blot system with HRP-labelled detecting Abs. Dipsticks showed similar sensitivities indicated by clearly visible red, brown or blue dots using two different D y e / A b combinations (Palanil Red B F / R b anti-chicken IgG or Samaron Blue/Rb anti-chicken IgG) or enzyme labelled antisera. The Ag detection limits were as sensitive as 10 ng/ml using purified chicken IgG as an antigen source and either the DIA reagents or enzyme labelled detecting Abs when using dipstick capture dots of affinity purified Ab as dilute as 5/~g/ml protein. Similar results were obtained using anti-human
(9
YRbantihumanigo
v
:
v
!
, OR
,~
R b antihuman igG
.~
"~
C!~~G
~
enIgG
v
O:
AND
/
~
~
,~(~
( ~ H u m a n IgG
O
®
®
Fig. 1. Assay protocol and schematic diagram of dipstick design for differentiating between different host sources of mosquito bloodmeals. 1: dipstick dotted with species-specificantisera as a capture layer; 2: non-specificprotein binding to NC blockedwith PBSM and tap water rinsed; 3: dipstick incubated in Ag containing PBS solution (IgG standard or mosquitobloodmeaieluate) and tap water rinsed; 4: dipstick incubated in multiple Ag detecting reagent (Palanil Red/Rb anti-chicken IgG and Sarnaron Blue/Rb anti-human IgG) and tap water rinsed; 5: dipsticks observed for red or blue dots at appropriate locations indicating positive Ag detection results.
62 N o cross-reactivity was detected with guinea pig fed mosquitoes. Fig. 3 shows dipsticks recognizing human or chicken bloodmeals using a pooled purple suspension containing Palanil R e d / R b antichicken IgG and Samaron B l u e / R b anti-human detecting reagents. 50 dipsticks (for 50 mosquitoes) were processed in a total volume of 5 ml of D y e / A b reagent without significant loss of colour intensity in the positive dots.
Discussion The D I A combines a number of previously described techniques to produce a new, sensitive, inexpensive, robust test utilizing simple technology. Since the technique requires no instrumentation for qualitative multiple Ag detection, it has many potential field applications. The production of the detecting D y e / A b reagents is a simple absorption procedure using common buffer solutions. The selection of appropriate disperse dyes can only be made by screening an assortment of dyes since each coloured colloidal dye suspension has a unique Ab binding capacity. Ten samples of different disperse dyes were evaluated for their suitability for use by absorbing several different Abs specific for either human or chicken IgG to each dye and comparing the intensity of positive coloured dots on the dipstick. Of the ten dyes screened, Palanil Red BF (BASF, U.K.), Samaron Red HBSL, Samaron Brilliant Pink and Samaron Blue FBLN (Hoechst, F.R.G), consistently produced effective D y e / A b reagents using several goat or Rb source antisera. Several yellow dyes labelled adequately with Ab, although positive yellow dots on the white NC background were sometimes difficult to visualize. These observations agreed with the previously published comment that disperse dyes suitable for D I A reagents
TABLE I BLOODMEAL IDENTIFICATION OF FRESHLY FED A E G Y P T I MOSQUITOES USING DIA AND ELISA ASSAYS AEDES
Both methods were 100% correct in identifying human or chicken bloodmeals. no. of mosquitoes
Host species of bloodmeal
no. of bloodmeals identifiedby:
20 20 5 5
Human Chicken Guinea pig Unfed
DIA 20 20 0 0
ELISA 20 20 0 0
50
can only be found by screening and that selection of a particular dye depended on the Ab preparation (Gribnau et al., 1983). If textile dyes were mixed to produce orange, green or purple hues before labelling, the Abs sometimes bound in differing proportions to each coloured dye so that the hue of the coloured dot on the dipstick did not necessarily match the hue of the D y e / A b suspension. Since a wide variety of coloured disperse dyes are commercially available it is probably most appropriate to select the desired colour of dye particle initially rather than to attempt to mix colours to generate new hues for D I A reagent production. The preparation and use of D y e / A b reagents is similar to and even simpler than Ab linked colloidal gold particle probes (DeMay, 1983). Adequate Dye-Ab absorption, as judged by the production of appropriate positive dots on a dipstick, occurred at a variety of pHs ranging from 6.0 to 8.5. The ionic strengths of the absorption buffers were more critical than pH. Exceeding a salt con-
Fig. 2.A: sensitivity and optimization of DIA. Replicate dipsticks were dotted with various dilutions of capture antibody (Rb anti-chicken IgG, 0.5 to 20 #g/ml), blocked with PBSM, and incubated in various Ag concentrations (10 #g/ml to 10 ng/ml) of purified chicken IgG standard. Ag detection is indicated by red dots on dipsticks A-G using Palanil Red/Rb anti-chicken IgG, brown dots on dipsticks H-N using HRP labelled Rb antichicken IgG, and blue dots on dipsticks O-U using Samaron Blue/Rb antichicken IgG. Dipsticks, G, N, and O are PBS controls. B: durabihty of DIA. Assaywas conducted as in A using stored dipsticks and Palanil Red/Rb antichicken IgG reagent that had been frozen and thawed. Fig. 3. Bloodmeal detection DIA. Dipsticks indicate positive identification of chicken (A) and human (B) bloodmeals from laboratory fed mosquitoes.
63
Ill
~
0.5 1
10
~ 2o /~g/ml ng/mi 2A
~g/ml
ng/ml
#g/ml
ng/ml
Ag CONCENTRATION
,O,
0.5 1
~ 10
~ 20 3
~tg/ml 2B
ng/ml
Ag CONCENTRATION
B
64 centration of 15 mM before or during dye particle-Ab absorption resulted in the disruption of the colloidal suspension and irreversible settling out of the dye particles. A pH of 7.2-7.4 was used in the POa/NaC1 buffer because of its physiologic value in maintaining Abs at optimum conditions. The importance of pH during D y e / A b absorption needs to be further explored if Mabs are used instead of affinity purified polyvalent Abs. If this colloidal particle/Ab system resembles Ab labelled colloidal gold particle production, the p I of the Mab should be considered since the absorption of the Ab to the gold particle is optimized near the p I of the Mab (DeMay, 1983). While BSA and PEG have been used to stabilize g o l d / A b suspensions, only BSA was effectively used in preliminary experiments with the colloidal D y e / A b reagent. Textile disperse dyes are produced widely in the textile chemistry industry and are typically marketed in kilogram quantities. Consequently costs for the production of DIA reagents are negligible compared with colloidal gold products. The D y e / A b reagent production has been previously described and utilized in sandwich type assays detecting human chorionic gonadotropin or human placental lactogen (Gribnau et al., 1982, 1983). The assay used Ab coated microtitre plates to capture purified hormone Ag. The Ag was detected by the D y e / A b detecting reagent and colour was produced in the microtitre wells by solubilizing the bound dye using ethanol. The optical density of reaction wells was read in a similar manner to ELISA tests. To our knowledge this is the first report of the use of this D y e / A b reagent for direct colour visualization on a solid matrix such as nitrocellulose. The test described here is a simpler, faster qualitative Ag detection test using the D y e / A b reagent in a different manner than previously described. Dipsticks which had been dotted with capture Ab, blocked with PBSM, dried and stored at 22 ° C maintained reactivity for at least 3 months (data not shown). The D y e / A b reagents maintained reactivity for at least 1 week when stored at 4 ° C as a liquid suspension. Unlike gold-labelled reagents, they could also be frozen to - 2 0 ° C or lyophilized for longterm storage. The D y e / A b
reagents maintained their colloidal suspension and reactivity after freezing and thawing. Lyophilized samples maintained their reactivity when reconstituted, but occasionally did not completely redisperse so that the optical densities of the reagents were slightly decreased. Replicate dipsticks comparing D y e / A b and enzyme labelled detecting reagents consistently gave similar Ag detection sensitivities as low as 10 n g / m l indicating that the DIA method gives results comparable to typical sandwich dot blot systems. Several brands of antisera were screened for their sensitivity and lack of cross-reactivity between IgG from human and chicken. Antisera concentrations as low as 10 /Lg/ml produced effective D y e / A b reagents if affinity purified antisera were used. Not surprisingly, if whole serum or the IgG fraction of antisera were used, protein concentrations up to 100/~g/ml were required to produce effective D y e / A b suspensions. In some cases when whole serum or the IgG fraction of antisera were used as capture Ab dots, some crossreactivity between human and chicken IgG detection occurred with both the DIA and the H R P labelled detecting reagents. Predictably, affinity purified IgG fractions of antisera to human or chicken IgG lacked cross-reactivity problems and provided the best results in DIA, enzyme dot blot and ELISA comparative assays. To demonstrate the use of the DIA, a simple mosquito bloodmeal identification test was designed and compared with a previously described ELISA assay. In this test comparison, the DIA was faster, much easier to perform and used fewer Ab reagents than the ELISA method. It had the added advantage that two or more Ags (chicken or human IgG) were potentially detected using a single test, rather than performing an ELISA test to detect each of the potential species providing bloodmeals. The dipstick also serves as a simple permanent test record. The DIA provides a simple robust multiple Ag detection system with a wide range of possible applications. The assay needs to be further developed by adapting other sandwich antigen capture systems to this method, and needs to be tested under field rather than laboratory conditions.
65
Acknowledgements We are grateful to Dr. M.W. Service for providing bloodfed mosquitoes and the Wolfson Foundation for funding K.S.
References DeMay, J. (1983) Colloidal gold probes in immunocytochemistry In: J.M. Polak and S. Van Noorden (Eds.), Immunocytochemistry, Practical Applications in Pathology and Biology. Wright PSG, Bristol, pp. 82-112. Gezahegan, T. (1989) Development of a dipstick method for mosquito blood-meal identification and comparison with ELISA for potential field use. MSc Thesis, Liverpool School of Tropical Medicine, Liverpool, U.K. Gribnau, T., Roeles, F., v.d. Biezen, J., Leuvering, J. and Schuurs, A. (1982) The application of colloidal dye particles as label in immunoassays: disperse(d) dye immunoassay ("DIA") In: T.C.J. Gribnau, J. Visser and R.J.F. Nivard (Eds.), Affinity Chromatography and Related Techniques. Elsevier, Amsterdam, pp. 411-424.
Gribnau, T., van Sommeren, A. and Van Dinther, F. (1983) DIA-disperse dye immunoassay. In: I.M. Chaiken, M. Wilchek and I. Parikh (Eds.), Affinity Chromatography and Biological Recognition. Academic Press, Orlando, FL, pp. 375-380. Hsu, Y. (1984) Immunogold for detection of antigen on nitrocellulose paper. Analyt. Biochem. 142, 221-225. Moeremans, M., Daneels, G., VanDijck, A., Langanger, G. and DeMey, J. (1984) Sensitive visualization of antigen-antibody reactions in dot and blot immune overlay assays with immunogold and immunogold/silver staining. J. Immunol. Methods 74, 353-360. Penney, C.L., Gauldie, J., Evelegh, M., Penney, M.T., Chong, D. and Horsewood, P. (1989) Polycarbonate membranes: a novel surface for solid-phase determination with utility in field format serological assays. J. Immunol. Methods 123, 185-192. Service, M.W., Voller, A. and Bidwell, D.E. (1986) The enzyme-linked immunosorbent assay (ELISA) test for the identification of blood-meals of haematophagous insects. Bull. Ent. Res. 76, 321-330. Stott, D.I. (1989) Immunoblotting and dot blotting. J. Immunol. Methods. 119, 153-187.