Immuno-Quantitative Polymerase Chain Reaction for ...

JOURNAL OF IMMUNOASSAY & IMMUNOCHEMISTRY Vol. 25, No. 3, pp. 241-258, 2004

Immuno-Quantitative Polymerase Chain Reaction for Detection and Quantitation of Prion Protein Stéphanie Gofflot,1 Benaïssa El Moualij,1 Danièle Zorzi,1 Laurence Melen,1 Stefan Roels,2 Dominique Quatpers,3 Jacques Grassi,4 Emmanuel Vanopdenbosch,2 Ernst Heinen,1 and WUly Zorzi1'* 1

Service d'Histologie Humaine-Centre de Recherche sur les Protéines Prions (CRPP), Université de Liège-CHU, Tour de Pharmacie, Liège, Belgium 2 VAR-CERVA-CODA, Veterinary and Agrochemical Research Center, Department of Biocontrol-Section of Pathology, National Reference Laboratory for Veterinary TSE (Belgium & Luxemburg), Brussels, Belgium 3 CRPP-Laboratoire ULg BSE, Tour de Pharmacie, Liège, Belgium 4 CEA, Service de Pharmacologie et d'Immunologie, CEA Saclay, Gif sur Yvette, France

ABSTRACT Immuno-polymerase chain reaction (PCR) is an extremely sensitive detection method, combining the specificity of antibody detection and

*Correspondence: Willy Zorzi, Service d'Histologie Humaine-Centre de Recherche sur les Protéines Prions (CRPP), Université de Liège-CHU, Tour de Pharmacie, Avenue de l'Höpital l, B-4000 Liège, Belgium; E-mail: [email protected]. 241 DOI: 10.1081/IAS-200028044 Copyright © 2004 by Marcel Dekker, Inc.

1532-1819 (Print); 1532-4230 (Online) www.dekker.com S

s

MARCEL DEKKER, [NC. 270 Madison Avenue, New York, New York 10016

242

Gofflot et al. the sensitivity of PCR. We have developed an immuno-quantitative PCR (iqPCR), exploiting real-time PCR technology, in order to improve this immuno-detection method and make it quantitative. To illustrate the advantages of iqPCR, we have compared it with a conventional enzyme linked immuno sorbent assay (ELISA) technique in experiments aimed at detecting the cellular and the resistant form of prion protein in bovine brain extract. The iqPCR technique proved to be more sensitive than ELISA, so it could be a technique of choice for the diagnosis of infected animals both at an ante mortem and post-mortem stage. Key Words:

Immuno-quantitative PCR; Prion; ELISA.

INTRODUCTION

_j | s 1) = g b" ^

In biology today, methods for detecting organic molecules or infectious agents are diversified and very sensitive. Yet, the need to lower detection limits persists and, in many cases, quantitative estimates are required. We have devised a new method combining the advantages of two advanced techniques: high-specificity enzyme linked immuno sorbent assay (ELISA) based immuno-detection and highly sensitive, quantitative real-time PCR. We have called our new technique immuno-quantitative polymerase chain reaction (iqPCR) (Patent WO0131056). In this study, we have applied it to the detection and quantitation of both forms of the prion protein, comparing its performance with that of the Bio-Rad Platelia sandwich ELISA technique. This last technique has proven to be the most sensitive[1] and robust technique[2] that is available for the moment. This same technique has also been proven to be useful to detect preclinical cases of B SE in cattle.131 Prion diseases are fatal transmissible neurodegenerative disorders present in many countries worldwide. They are characterised by behavioral and locomotor changes, cerebral amyloid plaques, and spongiform degeneration of the brain. They affect a wide variety of animals, from rodents to primates. Animal diseases include transmissible mink disease, chronic wasting disease of mule deer, white tailed deer and elk, scrapie in sheep and goat, feline spongiform encephalopathy, and bovine spongiform encephalopathy (known as "mad cow disease"). Human prion diseases have been classified as CreutzfeldtJakob disease (CJD), Gerstmann-Straussler-Scheinker disease (GSS), fatal familial insomnia (FFI), and kuru. Variant CJD (vCJD) has made a more recent appearance; it involves the prion strain that causes bovine spongiform encephalopathy in cattle (for review see Ref.[4]). Recently, data have occurred that even the sporadic form of CJD could be linked to BSE.[5]

•c

u MARCEL DEKKER, INC. 270 Madison Avenue, New York, New York 10016

iqPCR for Detection and Quantitation of Prion Protein

243

Infectious forms of prion diseases are caused by infectious particles that resist both mild proteinase K digestion and inactivation by nucleic-acidmodifying procedures. The particles represent an abnormal isoform of a host-encoded glycoprotein (PrPc). Prusiner called this unconventional transmissible agent "prion" for "proteinaceous infectious particle" (PrP), to highlight that it is not a virus.[6] The resistant form of the prion protein (PrP*168) has a secondary structure rich in /3 sheets. A peculiar property of PrP1*68 is its partial resistance to protease (e.g. proteinase K) treatment. Such treatment leaves an undigested 27-30kDa carboxy-terminal fragment which is insoluble in non-denaturing detergents and tends to aggregate and form scrapie-associated-fibrils (SAFs).[7] At the present time, there are very few works describing methods suitable for measuring PrP1*68 in live animals.[8>9] So far, these approaches have not demonstrated their capacity to diagnose TSEs in routine conditions. The tests currently used for the screening of cattle carcasses are only post-mortem diagnosis tests. They consist of classical or confirmatory tests, including histopathology, immunohistopathology, electron microscopical SAF detection and Western blotting. A second, and more recent group include the "rapid tests." These are all antibody based—sandwich ELISA from Bio-Rad, direct ELISA from Enfer Scientific, Western blotting tests of Prionics, ELISA luminescence immuno assay (Prionics) and conformation-dependent immunoassay (CDI, InPro).'11 New approaches for sample preparation and concentration could be considered like a filter retention assay to exploit proteinase K (PK) resistance and detergent insoluble aggregated state.[10] Others focused on the low quantity detection of infectious particle, trying to reproduce the conversion by improvement of contact by sonication (protein misfolding cyclic amplification).[11] Immuno-PCR has been proposed as a potentially sensitive tooi for diagnosing prion diseasesJ121 This is why our laboratory has developed and patented the "immuno-quantitative polymerase chain reaction," a new, highly sensitive technique applicable to the detection and quantitation of prion proteins.[13] Immuno-PCR was developed in the 1990s by Sano et al.[14] One of its advantages is its applicability in many fields, including diverse areas of medicine and environmental studies. It has been used to detect cytokines hke IL-18[15] and TNF,[16~18] micro-organisms like Clostridiumll9] and Salmonella[20] and their toxins,[21] viruses like bovine herpesvirus,[22! and other proteins such as angiotensinogen.'231 Most studies using Standard immuno-PCR have involved electrophoretic analysis on agarose gels. But vascular endothelial growth factor (VEGF) has been detected by the real-time immuno-PCR by means of a TaqMan probe.[24] In the last example, the reagents were incubated all together in a polystyrene plate and then placed in a polypropylene plate for PCR. Our iqPCR technique measures amplification

244

Gofflot et al.

of the material after each cycle, thanks to the fluorescence of an intercalating molecule (SYBRGreen). We consider only the value obtained in the exponential phase of the amplification curves. We use a multistep protocol where each element (antigen, antibody, streptavidin) is added separately as in an ELISA. In this paper, we compare two methods for detecting the cellular and the resistant form of the prion protein in bovine brain extract: the ELISA technique and iqPCR. We show that iqPCR is a powerful technique, more sensitive than a Standard ELISA. EXPERIMENT AL Antibodies The 12F10 antibody was raised against human recombinant PrP (A generous gift of Pr. Hunsmann, German Primate Center, Goettingen, Germany); it recognises the 144-152 amino acid (AA) sequence of human PrP (DYEDRYYRE).[25'26] The SAF34 antibody was produced in mice by injecting formic acid denaturated S AFs obtained from hamster infected brain. This antibody recognises the octo-repeat region located in the amino-terminal part of the protein (59-92 sequence of human PrP, provided by J. Grassi) (Table 1). Brain Extract and Recombinant Protein Bovine prion proteins were extracted by using a B SE purification kit from Bio-Rad (ref. 3551102). The negative samples assayed came from the obex area, whereas infectious tissue was from bovine cerebellum. Briefly, nervous tissue (+ 350 mg) was homogenised for 45 sec and 500 jjüL of this homogenate was PK treated 14 min at 37°C in buffer A. After stopping the reaction and addition of 500 |xL of buffer B, tubes were centrifuged for 5 min at 20,000 g. Immuno-detection was operated on resuspended pellet by heating 5min at 100°C in buffer Cl. The positive control was a recombinant cellular bovine prion protein (Prionics, Switzerland). Table 1. Description of antibodies used and epitopes recognised. Mouse monoclonal antibodies 12F10 SAF34

ol

©

MARCEL DEKKER, INC. 270 Madison Avenue, New York, New York 10016

Directed against

Recognised AA sequence

Human recombinant prion protein Denatured hamster SAFs

144-152 59-92


245

Techniques Used to Control the Brain Samples for TSE Both negative and positive samples derived from bovines of the routine TSE epidemie surveillance scheme.[27] They were checked using the BioRad Platelia™ rapid test and confirmed by histopathology, immunohistopathology, and electron microscopic detection of SAFs. The positive samples came from a 6-year-old cow with clinical signs suspected of BSE.

ELISA Classical ELISA The monoclonal antibody 12F10 was used as capture antibody; it was düuted in coating buffer (0.05 M NaHCO3 and 0.05 M Na2CO3, pH 9.4) and incubated in microtiter plates (50 |xL/well) overnight at 4°C. After washing with 300 |xL phosphate buffer saline (PBS) (0.1 M NaCl, 3mM KC1, 1.5mM KH2PO4, 8mM Na2HPO4 • 2H2O, pH 7.4), non-specific binding sites were saturated with PBS, 3% BSA for 2hr at room temperature (300 (jtL/well). After washing with PBS, solutions containing different concentrations of antigen (recombinant prion protein or bovine brain extract) were prepared in dilution buffer (PBS, 1% BSA) and incubated in wells for l hr at room temperature. The plate was washed with PBS and the biotinylated detection antibody SAF34 was then incubated l hr at room temperature. The wells were again washed three times with PBS, 0.1% Tween and three times with PBS, 1.5% BSA. Peroxidase-conjugated streptavidin (DakoCytomation) diluted l /7500 was added to each well and incubated for 30 min at room temperature. After five washes with PBS, peroxidase activity was measured by means of a chromogenic reaction: a solution containing 3,3',5,5'-tetramethyl benzidine (TMB) and H2O2 (v/v, BD PharMingen). The reaction was left to develop for 30 min in the dark at room temperature. It was stopped by the addition of 2 N H2SO4. The absorbance of the medium was determined at 450 nm with an automatic reader instrument. Bio-Rad Platelia™ BSE Detection Kit According to the manufacturer's instructions, pre-coated strips with an antiPrP monoclonal antibody were incubated with samples for 75 min at 37°C. After three washing cycles, the conjugate solution was distributed into the wells and incubated for 60 min at 4°C. Five washing cycles were performed and revelation solution was distributed and incubated 30 min in darkness. After stopping the reaction, determination of the optical density was performed at 450/620 nm.

246

Gofflot et al.

iqPCR iqPCR (Fig. 1) was identical to the ELISA up to the addition of the biotinylated detection antibody, except that it was carried out in polypropylene microplates (MicroAmp Optical 96-well Reaction Plate, Applied Biosystems). At each step, the incubation volume was 50 (xL/well as for the ELISA. After a 1-hr incubation at room temperature, the plate was washed three times with PBS, 0.1% Tween, and three times with PBS, 1.5% BSA. First, recombinant streptavidin (Roche) was pre-incubated with biotinylated reporter DNA (molar ratio 17 atomol/34 atomol) for 45 min at 4°C. Then, the resulting streptavidin-DNA complex was added to the wells and incubated for 30 min at room temperature. The plate was washed five times with PBS, 10 times with distilled water, and then subjected to PCR. Amplification was carried out in an ABI PRISM 7700 sequence detection system (Applied Biosystems) under the following conditions: 50% SYBR Green PCR master mix (Applied Biosystems), 0.3 (JiM of each primer, 38% water in a total volume of 50 (juL. The temperature program was as follows: Ten minutes at 95°C for activation of Taq polymerase. Forty cycles of denaturation at 95°C for 15 sec, 60°C for l min for the annealing and extension phases. SYBR Green is a dye that fluoresces when it is intercalated in doublestranded DNA. Amplification curves were analysed with sequence detection system software (Applied Biosystems) and the baseline was determined in

Biotin

Streptavidin Real Time PCR

Detection antibody PrP

Capture antibody Polypropylene microplate l l

Figure 1.


Diagram of an iqPCR assay.


247

the exponential phase to avoid background signals. The intersection between this threshold and the amplification curve, called cycle threshold (Ct), was plotted on a graph vs. the well position.

Reporter DNA The reporter DNA (nucleotide AX133313: http://www.ncbi.nlm.nih.gov) used is original in that it was built by association of two DNA fragments of eukaryotic and prokaryotic origin. At the 5' end are 105 base pairs (bp) of pHHL23 plasmid, containing an "ARS coding sequence." On the other side are 141 bp of the pCRü-TOPO plasmid. The reporter DNA was 5' end biotinylated by PCR. The sequence of the resulting biotin-labelled 246-bp reporter DNA was: 5'TATCCAATCCATTCCAGGCCCTGTCCAGGCCTCTGnTCACCCAGTTTATACCGTAGCTTGTGAATGTATATCCAGAAGCCTTGCAGGACATCTTCACTGAGGACAAGGGCGAATTCTGCAGATATCCATCACACTGGCGGCCGCTCGAGCATGCATCTAGAGGGCCCAATTCGCCCTATAGTGAGTCGTATTACAATTCACTGGCCGTCGTTTTACAACGTCGTGACTGGGAAAACCCTGGCGTT-3'. The primers were chosen for their hybridisation site, the forward primer annealing on the eukaryotic part and the reverse primer on the prokaryotic part. Finally, the test was optimised with regard to the proportions of streptavidin, biotinylated reporter DNA, and biotinylated detection antibody (data not shown).

Primers The primers were designed with the aid of Primer Express Software, version 1.0 (Applied Biosystems). The sequence of the forward primer was: 5'-AAGCCTTGCAGGACATCTTCA-3' and that of the reverse primer was 5'-GCCGCCAGTGTGAT GGATAT-3'. These primers raise a chimaeric amplicon of 67 bp.

RESULTS Prion Protein Detection by ELISA In the first step, we wanted to detect prion protein extracted from infected and non-infected bovine brains. At each step, the incubation volume was a 'S,

248

Gofflot et al.

50 |xL/well. After coating of the wells with 12F10 (10 |xg/mL), unspecific sites were saturated with PBS and 3% BSA. During the incubation step, brain tissue was homogenised and treated with proteinase K for 14 min at 37°C. As positive control, we used recombinant bovine prion protein (lOOng/mL) without PK treatment to evaluate the test. The different samples were applied in successive dilutions to the wells and incubated for l hr. The detection limit for the recombinant bovine prion protein (defined here as the concentration giving an OD value three times as high as the one given by the controls) was between 6 and 12ng/mL [Fig. 2(A)]. It is worth noting that this does not correspond to the detection limit of the tests for detecting prPSens or denaturated PrP868, since it is not known if the conformation of the recombinant protein is equivalent to these two natural targets. Following a PK treatment, the resistant form of the prion protein was detected when the infected brain homogenate dilution was between 1/32 and 1/64. As expected, the low cellular prion protein signal observed in healthy brain disappeared following PK treatment [Fig. 2(B)].

Control for TSE of the Used Samples All samples were confirmed by histopathological (Fig- 3) and immunohistochemical tests (Fig. 4), that shows a profile characteristic of TSE and the presence of several SAFs during transmission electron microscopical examination.

Prion Protein Detection by Platelia™ BSE (Bio-Rad) In order to compare the sensitivity of our ELISA test with the Platelia BSE detection kit of Bio-Rad, we have shown (Table 2) that we reached the same sensitivity with both tests for the detection of the prion protein in infected brain.

_,

Prion Protein Detection by iqPCR

u>

u

s % = g s' l

In the iqPCR assays, we optimised various parameters such as the antibody, streptavidin, and DNA concentrations, so as to have a good signal-tobackground ratio. As shown in Fig. 5, the prion protein in healthy brain homogenate (A), as well as in the infected brain homogenate (B), was detected until a sample dilution of about 1/500 [Fig. 5, sample dilution (i)]. As a control the


249


A Recombinant bovine prion protein 3,000

1,500 f-

0,78

3,12

1,56

6,25

12,5

25

100

protein concentration (ng/ml)

B sandwich ELISA healthy brain + PK infected brain +• PK healthy farain

128x

64x

32x

16x

homogenate dilution

8x

Figure 2. Detection of prion protein by sandwich ELISA. (A) Bovine recombinant PrP was assayed at concentrations ranging from 0.78 to lOOng/mL. The proteins were captured with the 12F10 monoclonal antibody (10 |Ag/mL) and detected with biotinylated SAF34 (l fjLg/mL). The OD, measured at 450nm, is plotted against the antigen concentration. Error bars indicate the Standard deviations from the means of five replicate experiments. (B) Dilutions of homogenates from healthy or infected brain were assayed at dilutions ranging from 4 to 128 times. After completing the purification process, denaturated PrP was captured with the 12F10 monoclonal antibody (lOjtg/ ml) and detected with biotinylated SAF34 (l jj,g/mL). The OD, measured at 450nm, is plotted against the antigen concentration. Error bars indicate the maximum deviations observed for independent ELISA detection of antigens.

250

Gofflot et al.

Flgure 3. Histopathological examination of the ohex (D 1402): presence of pathognomonic intracytoplasmatic vacuoles in the neurons of the trigeminal tract (hematoxylin eosin staining x 160).

l

healthy brain was treated with PK [Fig. 5(A), right panel] and the signal was lost (Ct in the range of negative control without antigen). In comparison to the previous result (Table 2), iqPCR has been revealed to be about 10 times more sensitive. One negative control was a no-template control (NTC). Other negative controls were experiments without capture antibody, without antigen, without detection antibody, or without streptavidin-DNA complex (Fig. 5(a), (b), (c), and (d), respectively). The control (d) yielded a Ct value of about 40 and the others around 27-30. As we ran 40 PCR cycles, a Ct value of 40 means no amplification. The antigen-free control usually yielded a lower Ct value than the others. As positive control, we used bovine recombinant PrP. As shown by the Ct at Fig. 6, the detection limit was also improved in iqPCR (0.75 ng/mL) compared with ELISA results (6-12 ng/mL). It should be noted that we reach the same sensitivity factor with a recombinant prion protein as with a more complex medium like brain extract. This indicates that the reactivity of antibodies is the same in both experiment conditions.

I ©

MARCEL DEKKER, INC. ' 270 Madison Avenue, New York, New York 10016


251

Figure 4. Immunohistopathological examination of the obex (D 1402): diffuse presence of positive staining under the form brown-orange (chromogen = DAB), deposition of the chromogen in the cytoplasm of the neurons and surrounding neuropil of the vagal nucleus. (x 160). DAB: 3,3'-diaminobenzidine tetrahydrochloride. antibody used: (polyclonal) rabbit anti-Prp-peptide serum: R524 (CIDC Lelystad, The Netherlands).

DISCUSSION iqPCR is an efficiënt and highly sensitive immuno-detection method. Antigen quantification is possible during amplification of a reporter DNA by the powerful tooi of the real-time PCR. The efficacy of the reaction is monitored by measuring the fluorescence due to a doublé strand intercalating molecule Uke SYBR Green, at each PCR cycle. The sensitivity is determined by measuring the Ct, i.e., the intersection between a baseline and the fluorescent amplification curve. When the Ct value is low, the reporter DNA is present in high amount and this translates the presence of a large amount of antigen-bound antibody. The first immuno-PCR tests were performed with a linker molecule used to link the antigen-antibody complex to the reporter DNA. This linker molecule, a streptavidin-protein A chimaera, could bind, on the one hand,

Gofflot et al.

252 Table 2. assay.

Sensitivity comparison of classical ELISA test and the Platelia Bio-Rad OD at 450 nm with Bio-Rad ELISA test

sample dilution

1 1/10 1/100

OD at 450 nm with classical ELISA

Infected

Healthy

Infected

Healthy

2.707 1.191 0.147

0.025 0.021 0.011

1.8 1.128 0.138

0.034 0.026 0.014

Note: The tahle shows the optica! density of ELISA tests of the healthy and the infected brain.

to biotinylated plasmid DNA and, on the other hand, to the Fe portion of unmunoglobulin. The amplified products were analysed on agarose gels by ethidium bromide staining.[14] Some modifications have been introduced because this linker was not commercially available. In ELISAs, moreover, there is a risk that the protein A portion will bind to the free Fe fragment of the capturing antibody, even if the latter is from a different species. So, instead of a streptavidinprotein A chimaera, a biotinylated antibody and an avidin-biotinylated DNA complex were then used.[28] Other researchers performed two successive incubations, one with free streptavidin and one with biotinylated DNA, to avoid saturation of streptavidin binding sites by biotinylated DNA molecules during pre-incubation of the complex.'291 Yet, it has been demonstrated that streptavidin, because of its high affinity for biotin, can bind to biotinylated compounds even when DN A-streptavidin complexes are preformed.[12) In our laboratory, we optimised the proportions of streptavidin, biotinylated reporter DNA, and detection antibody in the mixture (unpublished data). We have improved the sensitivity of immuno-detection by means of a DNA complex revealed and quantified by real-time quantitative PCR. Most published papers mentioning the sensitivity of immuno-PCR are based on estimates of amplification product amounts on agarose gels. In our method, we use fluorescence measurements to monitor PCR amplification during the exponential phase. Measurements are, thus, performed under optimal conditions, i.e., before any reagents are likely to be used up in the PCR mixture. In our prion protein immuno-detection experiments, the detection limit for recombinant PrPc was 6-12ng/mL by ELISA and 0.75ng/mL by iqPCR. The sensitivity of iqPCR was thus about 10-fold higher than that of the ELISA. For a complex mixture (bovine brain extract), the detection limit was between 1/32 and 1/64 for the ELISA and around 1/500 for

fN

©

MARCEL DEKKER, INC. ' 270 Madison Avenue, New York, New York 10016

253


45,000* 40,000^ ^35,000Q 30,000 "S 25,000* ^ 20,000 ,