Clinical Chemistry 55:9 Molecular Diagnostics and Genetics 1686–1693 (2009)

Aptamer-Based Regionally Protected PCR for Detection Jun Sheng Lin1* and Kenneth P. McNatty2

BACKGROUND: DNA are single-stranded nu- In 1990, 2 laboratories independently developed the cleotide sequences that bind specifically to target mol- technique of selecting nucleic acid–based ligands ecules. By combining the advantages of PCR for ampli- against target molecules from a pool of random se- fying specific DNA sequences and technology, quences. Ellington and Szostak named the process “in we have developed a new strategy to detect target mol- vitro selection” and coined the term “aptamer” (from ecules such as . the Latin, aptus, meaning “fitting”) (1), and Tuerk and Gold used the term “SELEX” (systematic evolution of METHODS: Ovine follicle-stimulating hormone ␣ sub- ligands by exponential enrichment) to describe the unit (oFSH␣) was used as the model protein to gener- process (2). Aptamers are actually oligonucleic acids, ate a specific DNA aptamer via an in vitro evolutionary either single-stranded DNA (ssDNA)3 or RNA, that process. A targeted regional-mapping approach and a fold into secondary and tertiary structures and bind target-capturing were used to identify the bind- with high affinity and specificity to different target ing region on the aptamer molecule. In the detection molecules, such as proteins. assay, referred to as “aptamer-based regionally pro- In the postgenomics era, there is an increasing de- tected PCR” (ARP-PCR), the aptamer was allowed to mand for the detection of a wide range of proteins for bind to the target protein in solution before digestion basic research, drug discovery, and clinical applica- with DNase I. The region of the aptamer bound to the tions, together with a need for increasingly lower de- target was protected from DNase I cleavage. The target- tection limits. The ability shared by antibodies and binding region of the aptamer protected from the en- aptamers to bind specifically to their target molecules zymatic treatment was then amplified by the PCR. makes them good candidates for molecular diagnos- tics. Aptamers have the advantages of ease of discovery, RESULTS: Aptamers against oFSH␣ were generated. Six thermal stability, and low cost, making them very at- sequences of 20 selected aptamer clones were identical. tractive for targeted protein assays (3). Indeed, re- This aptamer sequence was divided into 4 regions ac- search into aptamer-based methods for diagnostics is cording to the aptamer’s secondary structure. From ex- currently the subject of considerable attention (4, 5). amination of the target-binding ability of each region, Notably, 2 new methods based on affinity PCR, prox- we determined the specific binding region, for which imity ligation (6) and exonuclease-protection assay primers were designed. With the aptamer and primers (7), that incorporate aptamer binding and signal am- ␣ to detect oFSH by means of the ARP-PCR method, we plification by the PCR have achieved very low limits of were able to detect the target protein at concentrations Ϫ14 quantification for protein detection. The key point of as low as 10 mol/L. the affinity-based PCR approach is the ability to distin- guish aptamer–protein complexes from unbound CONCLUSIONS: Combining the use of a DNA aptamer aptamers before signal amplification. Both of these with the PCR is a potentially useful analytic tool for methods rely on an extra ligation reaction that occurs detection of proteins at low concentrations. preferentially on the aptamer–protein complex. Con- © 2009 American Association for Clinical Chemistry sequently, the ligation products are chemically differ- entiated from unbound aptamers, giving rise to an am-

1 Centre for Reproduction and Genomics, AgResearch, Invermay, Mosgiel, New 3 Nonstandard abbreviations: ssDNA, single-stranded DNA; ARP-PCR, aptamer- Zealand; 2 School of Biological Sciences, Victoria University of Wellington, based regionally protected PCR; oFSH␣, ovine follicle-stimulating hormone ␣ Wellington, New Zealand. subunit; Pf, forward primer; Pr, reverse primer; Pfb, Pf primer with a biotin * Address correspondence to this author at: Centre for Reproduction and Genom- group at the 5Ј end; Prph, Pr primer modified with a phosphate group at the 5Ј ics, AgResearch, Invermay, Private Bag 50034, Mosgiel 9053, New Zealand. Fax end; Pdr2, reverse diagnostic primer; roFSH␣␤, recombinant oFSH containing ϩ64-3-4893739; e-mail [email protected]. both ␣ and ␤ subunits; Ova2, highly purified sheep pituitary gland extract of Received March 24, 2009; accepted June 18, 2009. native oFSH; dNTP, deoxynucleoside triphosphate; dITP, deoxyinosine triphos- Previously published online at DOI: 10.1373/clinchem.2009.127266 phate; dsDNA, double-stranded DNA; SMB, streptavidin magnetic beads; Hy- pox, serum from hypophysectomized sheep devoid of FSH.

1686 Detection of Protein by Aptamer-Based PCR

plifiable detection signal; however, potential false- GCAATCT-3Ј) from Invitrogen was used as an positive results can arise if ligation of the free aptamers initial pool. The 60-mer nucleotide portion of ran- takes place. Omitting the extra ligation reaction by ap- domized sequence (N60) was flanked by defined 18- plying rolling-circle amplification on a circular DNA nucleotide priming regions for PCR amplification. aptamer has been suggested (3). Alternatively, an The ssDNA pool was reconstituted in 10 mmol/L aptamer–protein complex could potentially be sepa- Tris-HCl, pH 7.5, diluted to 1 mmol/L, and stored at rated from unbound aptamers by means of capillary Ϫ20 °C until use. electrophoresis (8). We developed an assay that uses an aptamer-based Solid phase–bound target protein. Twenty pieces of ni- regionally protected PCR (ARP-PCR) method and in- trocellulose membrane (Hybond-C Extra; Amersham vestigated the suitability of the assay to detect a selected Biosciences/GE Healthcare) were cut into 2-mm squares, ␮ ␣ ␮ target protein, ovine follicle-stimulating hormone soaked with 50 g purified roFSH in 50 L PBS ␣ subunit (oFSH␣), at picomolar or femtomolar (2.7 mmol/L KCl, 1.5 mmol/L KH2PO4, 137 mmol/L concentrations. NaCl, and 8 mmol/L Na2HPO4, pH 7.2) containing 0.5 g/L sodium azide, air-dried, and stored at 4 °C until Materials and Methods use for selection. Selection of aptamer. Aptamers were generated by an in PRIMERS vitro evolutionary process that included several rounds We designed 2 conserved primers that flank a candi- of selection by target binding and mutation via mu- date aptamer for PCR amplification: forward primer tagenesis PCR. In the first round, 3 ␮L of the 1-mmol/L Ј Ј Pf (5 -ATACCAGCTTATTCAATA-3 ) and reverse initial ssDNA pool was blocked with 500 ␮L eastern Ј Ј primer Pr (5 -AGATTGCACTTACTATCA-3 ). We blocking solution [50 g/L nonfat milk powder in east- Ј also prepared a Pf primer with a biotin group at the 5 ern buffer containing 20 mmol/L Tris-HCl, pH 7.5, end (Pfb) and a Pr primer modified with a phosphate 100 mmol/L NaCl, 5 mmol/L KCl, 2 mmol/L MgCl , group at the 5Ј end (Prph). Once we determined the 2 1 mmol/L CaCl2, 10 g/L yeast RNA (Roche), and sequence of a binding region, we designed a pair of 1 mL/L NP-40] at room temperature for 1 h. Mean- primers to amplify the binding region via the PCR. In while, 1 piece of the Hybond-C Extra membrane this study, we designed a reverse diagnostic primer bound with roFSH␣ was separately blocked with Ј Ј (Pdr2) (5 -TCGCTCCATTAAATTCTC-3 ) to pair 500 ␮L eastern blocking solution. This membrane was with Pf for signal amplification via the PCR. then transferred into the blocked ssDNA solution and incubated for1hat4°Cformaximum selection. The PREPARATION OF TARGET PROTEIN incubated membrane was washed 3 times (10 min ␣ ␤ FSH consists of 2 glycosylated subunits, and . For each) with eastern buffer at room temperature to re- ␣ ␣ this study, we prepared recombinant oFSH (roFSH ) move excess ssDNA. The ssDNA bound with target as described previously to generate the aptamer (9). protein on the membrane was eluted at 80 °C for 5 min ␣ ␤ ␣␤ roFSH containing both and subunits (roFSH ) with 30 ␮L of Milli-Q–purified water (Millipore) as the was also produced with transfected insect cells (High template for the next round of PCR. The conditions for Five™; Invitrogen NZ) with the expressed protein se- successive rounds were the same as for the first round ␣␤ creted into serum-free culture medium. The roFSH except that the initial ssDNA pool was replaced by the was demonstrated to be biologically active by in vitro succeeding candidate aptamer preparation (see below). bioassay (9). A highly purified sheep pituitary gland All processes were performed at room temperature extract of native oFSH (Ova2) was a gift of Dr. Lloyd (Fig. 1). Moore (AgResearch Invermay, Mosgiel, New Zea- land). This material was comparable in biological ac- Mutagenesis PCR. Random mutations in DNA frag- tivity to the highly purified oFSH RP2 standard pro- ments were generated with mixtures of deoxynucleo- duced by the National Hormone and Pituitary side triphosphates (dNTPs) and deoxyinosine triphos- Program (USA), as assessed by both RIA and in vitro phate (dITP) in the PCR (10). The PCR reaction bioassay (9). Both the roFSH␣␤ and Ova2 were used contained 10 ␮L of the eluted templates, 1 ␮mol/L each for characterization of the aptamer that had been gen- of primers Pfb and Prph, 10 mmol/L Tris-HCl, pH 8.6, ␣ erated against the roFSH . 50 mmol/L KCl, 5 mmol/L MgCl2, 0.1 mmol/L MnCl2, 200 ␮mol/L dITP (Roche), a 20-␮mol/L rotation of 1 GENERATION AND CHARACTERIZATION OF APTAMER of the 4 dNTPs (dGTP, dATP, dCTP, and dTTP; Roche) at each round, 200 ␮mol/L of the other 3 Initial ssDNA pool. A 96-base ssDNA oligomer (5Ј- dNTPs, and 3 U Taq DNA polymerase (Roche) in a ATACCAGCTTATTCAATA-N60-TGATAGTAAGT- total volume of 50 ␮L (10). The temperature regimen

Clinical Chemistry 55:9 (2009) 1687 Fig. 1. Schematic representation of the in vitro evolutionary processes for DNA aptamer generation. The aptamers that bound to the target protein were subjected to selection and mutation before characterization by sequencing. for the PCR was as follows: 95 °C for 3 min and 40 gel by SDS-PAGE under reducing conditions and cycles of 95 °C for 20 s, 53 °C for 20 s, and 72 °C for 10 s. then electrophoretically transferred to a To optimize the product, we reduced the number of membrane (Hybond-C; Amersham Pharmacia Bio- PCR cycles by 3 for every successive evolution round. tech/GE Healthcare) in an alkaline transfer buffer (20 mL/L methanol, 25 mmol/L Tris, 190 mmol/L Preparation of candidate aptamer. The double-stranded glycine) at 4 °C. We also used dot blotting with either PCR products were separated by agarose gel electro- antibody or aptamer to detect folded protein without phoresis. The band of amplified fragments with the ex- consideration of molecular weight. Protein samples pected size was excised, purified with a QIAquick Gel were applied directly to a nitrocellulose membrane. Extraction Kit (Qiagen), and eluted with 50 ␮L Milli- The membranes were then blocked for1hatroom Q–purified water at pH 8.5 for each PCR reaction. The temperature with either western blocking solution 5Ј-phosphorylated strand was removed from the puri- (50 g/L nonfat milk powder in western buffer contain- fied double-stranded DNA (dsDNA) by enzymatic ing 20 mmol/L Tris-HCl, pH 7.5, 150 mmol/L NaCl, digestion at 37 °C for2hinareaction containing 5 U ␭ and 1 mL/L NP-40) followed by incubation with anti- exonuclease (New England Biolabs), 67 mmol/L body to oFSH␣ (9) in western blocking solution at glycine-KOH (pH 9.4), 2.5 mmol/L MgCl , and 50 2 room temperature for 2 h, or eastern blocking solution mg/L BSA (Sigma–Aldrich). The reaction was then followed by incubation with biotinylated aptamer in stopped, and the remaining strand that had been elon- eastern blocking solution at room temperature for 2 h. gated from the biotinylated primer was extracted with Nonbound reagent (either primary antibody or biotin- chloroform. The presence of the generated ssDNA was ylated aptamer) was removed by washing 3 times monitored by UV spectrophotometry, and this prepa- (10 min each) with the corresponding western or east- ration was used as the candidate aptamer preparation ern buffer. The washed membranes were incubated for for the next selection round. From evolution round 10 1 h at room temperature with either antirabbit IgG– onwards, we also used the candidate aptamer prepara- peroxidase conjugate (Sigma–Aldrich) diluted with tion for dot blotting to monitor the selection of 2000 volumes of western blocking solution or with aptamer. streptavidin–peroxidase polymer (Sigma–Aldrich) di- Western, eastern, and dot blotting. We used western luted with 10 000 volumes of eastern buffer. Labeled blotting to detect protein with an antibody and used bands were visualized with the Immun-Star HRP eastern blotting to detect protein with an aptamer (11). chemiluminescence kit (Bio-Rad Laboratories), pho- Both methods relate the results to the corresponding tographed with a Molecular Imager Gel Doc XR system molecular weight of the protein. Protein samples were (Bio-Rad), and analyzed with Quantity One analysis denatured and separated on a 150-g/L polyacrylamide software (version 4.6; Bio-Rad). The densities were

1688 Clinical Chemistry 55:9 (2009) Detection of Protein by Aptamer-Based PCR

measured, when appropriate, with Scion Image soft- ware (Scion Corporation, Frederick, MD).

Aptamer clonal screening. We amplified the selected aptamers from evolution round 12 by high-fidelity PCR. The conditions were the same as for mutagenesis PCR except that dITP was omitted. Platinum Taq DNA polymerase (Invitrogen) and nonmodified primers (Pf and Pr) were used for 35 PCR cycles. The PCR products were purified with the QIAquick PCR Purification Kit (Qiagen) and ligated with the cloning vector pGEM-T Easy (Promega). This product was designated the pGEM-T Easy aptamer. The Escherichia coli DH5␣ strain was transformed with the pGEM-T Easy aptamer and selected with ampicillin. We confirmed the inser- tion by high-fidelity PCR with individual selected col- onies as template and primers Pfb and Prph. The PCR product of each insert was separated by agarose and purified with a QIAquick Gel Ex- traction Kit, with 50 ␮L Milli-Q–purified water, pH 8, being used in the elution step. The purified dsDNA was used for ssDNA preparation and subsequently screened by dot blotting (both as described above). The sequence of each positive clone identified by dot - ting was determined by DNA sequencing.

Targeting regional mapping. Each secondary structure Fig. 2. Recognition of the candidate aptamer to tar- of the sequenced aptamers was predicted with a DNA- get protein characterized by dot blotting and Eastern folding program (12). We selected different regions of blotting. the aptamer on the basis of the predicted secondary (A), Dot blot for roFSH␣ incubated with a buffer without stem–loop structure and then synthesized these re- (panel 1) and with (panel 2) the candidates from evolution Ј gions with biotin at their 5 ends (Invitrogen). We im- round 11. (B), Eastern blot for roFSH␣ (ra) detected by the mobilized 100 ng of purified roFSH␣ and 3 ␮L of cul- candidates from evolution round 12 (lane 1). Pituitary oFSH ture media of roFSH␣␤ on nitrocellulose membranes (Ova2) was detected by the stem–loop region A62 (lane 2). to map the target-binding region of the aptamer by dot One single band of glycosylated native oFSH␣ (na) was blotting. PBS, 100 ng BSA, and 3 ␮L of untransfected recognized, whereas the corresponding position of the ␤ null-cell culture media were used as negative controls. subunit (nb) contained in Ova2 showed no signal.

Target-capture assay. Streptavidin magnetic beads (SMB) (New England Biolabs) coated with synthesized DETECTION OF TARGET PROTEIN BY ARP-PCR biotinylated ssDNA of the selected regions were used to Samples (1 ␮L each) containing Ova2 at concentra- capture the target protein in solution. We coated the Ϫ8 Ϫ17 ␮ tions of 10 to 10 mol/L were incubated at room SMB (160 g) with 80 pmol of each biotinylated can- temperature for 1 h with the selected aptamer probe (1 didate region according to the manufacturer’s instruc- ␮Lof10Ϫ7 mol/L) in 2 ␮L eastern buffer. Serum from tions. The beads were picked up with a PickPen (Bio- hypophysectomized sheep devoid of FSH (Hypox) as ␮ Nobile) and transferred into 100 L eastern buffer assessed by RIA (a gift from Norma Hudson, Ag- containing 300 ng Ova2. The binding reaction was con- Research, Wallaceville Animal Research Centre, New ducted at room temperature for 1 h. The SMB–region– Zealand) and BSA were used as negative controls. Be- Ova2 complexes were transferred with a PickPen and cause the activity of DNase I for dsDNA is approxi- washed 3 times with eastern buffer, resuspended in 10 mately 500 times higher than for ssDNA (13),itwas ␮L PBS, and loaded onto an SDS-PAGE gel for subse- helpful to use primers Pdr2 and Pr and make both se- quent western blotting to determine the capturing abil- quences double-stranded by complementary binding ity of each biotinylated ssDNA region. to the aptamer probes to ensure an efficient cleanup of

Clinical Chemistry 55:9 (2009) 1689 Results

GENERATION AND DETERMINATION OF APTAMER SEQUENCES We generated the aptamers by an in vitro evolutionary process with a series of rounds of selection and mutation (Fig. 1). Dot blotting was performed to monitor the pres- ence of the candidate aptamers in the ssDNA preparation from the 10th round onward. A positive signal was de- tected at the 11th round of evolution (Fig. 2A). The can- didate aptamer generated from the 12th round was ana- lyzed by eastern blotting, and a positive band of the expected molecular weight (approximately 16 kDa) was detected (lane 1 of Fig. 2B). The results of both dot and eastern blotting suggested the generation of polyclonal aptamers capable of recognizing the target protein under nonreducing and reducing conditions. Furthermore, the individual sequence of the monoclonal aptamer was de- termined by DNA sequencing of the positive clones, which were screened by dot blotting (data not shown). Plasmid preparations of 20 positive colonies were se- quenced. The sequence data revealed 11 different se- quences with homologies ranging from 29% to 98%. The most abundant sequence, shared by 6 of the 20 clones and named A6, was chosen for further investigation (see Table 1 in the Data Supplement that accompanies the on- Fig. 3. Secondary structure of the aptamer A6 pre- line version of this article at http://www.clinchem.org/ dicted with a DNA-folding program [Zuker (12)]. content/vol55/issue9). Two regions of the stem–loop were designated A61 and A62. A diagnostic 41-mer segment was identified between TARGETING REGIONAL MAPPING primers Pf and Pdr2. A 87-mer between Pf and Pr was used The DNA-folding program (12) predicted a potential as a control. A63 included both A61 and A62. A64 was secondary structure of aptamer A6 (Fig. 3) consisting located in the 3Ј region of A6. of 2 stem–loop regions on the chosen aptamer se- quence. To characterize the target-binding features of A6, we synthesized biotinylated oligonucleotides cor- responding to the 2 stem–loop regions and named them A61 (5Ј-biotin-ATGGAGCGACATGGAGTC GGTCA-3Ј) and A62 (5Ј-biotin-CTTATTCAATACG the unbound sequences. Therefore, we added 0.2 pmol GGGGAGAATT-3Ј). We also synthesized another each of Pdr2 and Pr before adding freshly prepared sequence, A63 (5Ј-biotin-CTTATTCAATACGGGGG 100 nmol MnCl2 and 25 U DNase I (Roche); the total Ј ␮ AGAATTTAATGGAGCGACATGGAGTCGGTCA- 3 ), reaction volume was adjusted to 10 L with 20 mmol/L that covered both the A61 and A62 regions. The A61, Tris-HCl, pH 7.5. Unbound sequences were digested at A62, and A63 oligonucleotides were used to evaluate 37 °C for 2 h and then heated at 80 °C for 10 min to stop their reactions with roFSH␣ and roFSH␣␤ by dot blot- the reaction. The bound protein was detected by am- ting. The results showed that the synthesized DNA plifying the protected binding region by high-fidelity fragments of the different aptamer regions exhibited PCR. The conditions were as described in the aptamer- different capacities. The signal from the A62 blot was screening section of the methodology except that in substantially stronger than that from the A61 blot (Fig. this case 3 primers (1 ␮mol/L each of Pf and Pdr2, and 4A) with either roFSH␣ or roFSH␣␤. The A63 se- 1.5 ␮mol/L of Pr) and 1 ␮L template from the 10-␮L quence also showed a clearly positive signal, likely be- digestion reaction were used. We then analyzed 20 ␮L cause the A62 region is included in the A63 sequence. from each PCR reaction by electrophoresis on 30 g/L Weak binding of A61 may also make a minor contri- agarose gels and subsequent staining with SYBR Green bution to the A63 signal, which is approximately equal (Molecular Probes/Invitrogen). The image was imme- to the signal shown in the A61 dot (Fig. 4A). The map- diately visualized and photographed under UV light ping results suggested that the A62 fragment was the with the Molecular Imager Gel Doc XR system. main region on the A6 molecule binding to roFSH␣.

1690 Clinical Chemistry 55:9 (2009) Detection of Protein by Aptamer-Based PCR

The ability of A62 to bind to the native oFSH ␣ subunit from Ova2 was also demonstrated under reducing con- ditions. Ova2 contains native oFSH ␣ and ␤ subunits. Under such conditions, the ␣ subunit of 22–23 kDa would be dissociated from the ␤ subunit of 23–26 kDa (9). Biotinylated A62 (20 pmol) recognized a single band of the glycosylated ␣ subunit without a signal from the ␤ subunit (lane 2 of Fig. 2B). Note that the molecular weight of the native oFSH␣ due to glycosyl- ation was higher than that of roFSH␣ expressed in E. coli with no posttranslational modification (lane 1 of Fig. 2B).

TARGET-CAPTURING ASSAY We further characterized aptamer A6 by examining each region that might react with Ova2 in solution. The remainder of the A6 sequence was synthesized and named A64 (5Ј-biotin-GGTCAGTTATGGCTTTGT GTGATAGTAAGTGCAATCTA-3Ј). SMB coated with the synthesized biotinylated region A61, A62, A63, or A64 were used to capture Ova2 in solution before de- tection by western blotting. The ability of each region to capture target protein was determined by measuring the density of the signal from the . The capacity of the A62 region to bind Ova2 was greater than that of A61, as had also been observed with the targeting regional mapping, whereas A64 has no detectable signal of binding to Ova2 (Fig. 4B). The A63 region again showed a clearly positive signal, likely because of the A62 region being included in the A63 sequence. On the basis of the A62 sequence, we designed reverse diagnostic primer Pdr2 (5Ј- TCGCTCCATTAAATTCTC-3Ј) to pair with Pf for amplification of the binding region by the PCR.

DETECTION OF TARGET PROTEIN BY ARP-PCR Both the dot-blotting and western-blotting results re- vealed that A62 was the main target-binding region. Oli- gonucleotide extensions from both ends of the binding region could therefore be designed without affecting aptamer-folding or -binding ability. In the case of the aptamer A6, it happened to already have the extensions (Fig. 3). Therefore, the full length of A6 was used as a probe to bind to Ova2 in solution. The unbound se- Fig. 4. Mapping targeting regions of the aptamer A6. quence of the probe and all free aptamers in solution were (A), Ability of the different regions of A6 to recognize removed by DNase I digestion. The aptamer region roFSH␣ and roFSH␣␤ was analyzed by dot blotting. The bound with protein, which was anticipated to be pro- target-binding regions are listed at the top, and the pro- tected from DNase I cleavage (14), was amplified specifi- teins are indicated on the left. PBS, BSA, and untransfected cally by the PCR. The 3 primers (Pf, Pdr2, and Pr, cover- cell culture medium (null) were used as negative controls. ing positions 1–18, 41–24, and 87–70, respectively of (B), Characterization of target-binding capability of each the aptamer probe) were present in every individual region on A6 by western blotting. Ova2 in solution was PCR reaction. With primers Pf and Pdr2, a diagnostic captured by SMB coated with different regions of A6. The band of 41 bp was produced when the protein-binding antibody to oFSH␣ was used as primary antibody. region was protected from the complete DNase I diges- tion [lane 4 of Fig. 5A and the first 3 lanes (indicated

Clinical Chemistry 55:9 (2009) 1691 Fig. 5. Ova2 detected by ARP-PCR using the aptamer A6. The 41-bp band was recognized as the diagnostic marker. (A), We used 1 ␮Lof10Ϫ8 mol/L Ova2 to bind with A6. The key reagents involved for each detection are indicated in the top panel. (B), Samples of 1 ␮L of Hypox containing decreasing molar concentrations of Ova2 were detected by ARP-PCR. The last lane contains the 25-bp ladder of molecular weight standards. Excess primer Pr (18 nucleotides) is at the bottom of each sample lane.

as 10Ϫ8,10Ϫ11,10Ϫ14 mol/L) of Fig. 5B]. With the visible background bands were produced in the reaction primers Pf and Pr, a band of 87 bp was produced when with Hypox (lanes 6 and 7 of Fig. 5A), BSA (lanes 8 and 9 the DNase I digestion was incomplete or even in the of Fig. 5A), or Hypox with added ␤ subunit of ovine lu- absence of DNase I digestion (lane 2 of Fig. 5A). BSA teinizing hormone (Bioscan Continental; data not (1 ng/␮L) and Hypox samples diluted with 8000 vol- shown). Serial dilutions of Ova2 in Hypox were prepared umes of PBS were used with A6 as negative controls. No for the assay thereafter. A concentration of Ova2 at 10Ϫ14

1692 Clinical Chemistry 55:9 (2009) Detection of Protein by Aptamer-Based PCR

mol/L could clearly be detected by the ARP-PCR assay tecting proteins by aptamer-based PCR has shown that (Fig. 5B). ARP-PCR can detect native oFSH at a concentration of 10Ϫ14 mol/L without evidence of false-positive readouts. Discussion Future work will focus on developing a nonligation method for quantitative analysis of protein by incorpora- The ARP-PCR technique we have described is a relatively tion of real-time PCR and aptamer technologies. straightforward approach for detecting proteins with high specificity and low limits of quantification. The key factor in this approach is to ensure that unbound aptamer is Author Contributions: All authors confirmed they have contributed to removed before the PCR amplification step. In the pres- the intellectual content of this paper and have met the following 3 re- ence of Mg2ϩ, DNase I makes single-strand nicks, and quirements: (a) significant contributions to the conception and design, Mn2ϩ maximally activates the activity of DNase I to cleave acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of dsDNA (15). Our ARP-PCR approach used oligonucleo- the published article. tides to form the complementary dsDNA portions on the aptamer probe. Importantly, the activity of DNase I on Authors’ Disclosures of Potential Conflicts of Interest: Upon manuscript submission, all authors completed the Disclosures of Poten- dsDNA is approximately 500 times higher than on ssDNA tial Conflict of Interest form. Potential conflicts of interest: (13). Therefore, with this approach only the target- binding region was retained after DNase I digestion in the Employment or Leadership: None declared. 2ϩ 2ϩ Consultant or Advisory Role: None declared. presence of Mg and Mn . It was important to con- Stock Ownership: None declared. firm that the amplified signal reflected only the target- Honoraria: None declared. binding region surviving from the enzymatic digestion. Research Funding: Foundation of Research, Science and Technol- Three primers (1 forward and 2 reverse) can be used for ogy of New Zealand. the ARP-PCR assay with an internal control. In our study, Expert Testimony: None declared. oFSH␣ was used as a model target, Pf was used as the Role of Sponsor: The funding organizations played no role in the forward primer, and Pdr2 and Pr were used as the reverse design of study, choice of enrolled patients, review and interpretation primers. We used the Pf and Pdr2 primers to amplify a of data, or preparation or approval of manuscript. 41-bp diagnostic fragment that included the target- Acknowledgments: We thank Jenny Juengel, Sue Galloway, and binding region. An 87-bp fragment could be amplified by Steve Lawrence for critical comments on this manuscript, Lloyd Moore for providing native oFSH and making available some highly priming with Pf and Pr as the internal control if DNase I purified native ovine luteinizing hormone, Andrea Western for the cleavage was found to be incomplete. Excess Pr was in- production of roFSH␣␤, and Norma Hudson for providing sera cluded to avoid false-positive results. The approach of de- from hypophysectomized sheep. References

1. Ellington AD, Szostak JW. In vitro selection of 6. Fredriksson S, Gullberg M, Jarvius J, Olsson C, tures of dNTP and dITP in PCR. Methods Mol Biol RNA molecules that bind specific ligands. Nature Pietras K, Gustafsdottir SM, et al. Protein detec- 1996;57:351–6. 1990;346:818–22. tion using proximity-dependent DNA ligation as- 11. Jayasena SD. Aptamers: an emerging class of 2. Tuerk C, Gold L. Systematic evolution of ligands by says. Nat Biotechnol 2002;20:473–7. molecules that rival antibodies in diagnostics. exponential enrichment: RNA ligands to bacteriophage 7. Wang XL, Li F, Su YH, Sun X, Li XB, Schluesener Clin Chem 1999;45:1628–50. T4 DNA polymerase. Science 1990;249:505–10. HJ, et al. Ultrasensitive detection of protein using 12. Zuker M. Mfold Web server for nucleic acid fold- 3. Di Giusto D, Wlassoff W, Gooding J, Messerle B, an aptamer-based exonuclease protection assay. ing and hybridization prediction. Nucleic Acids King G. Proximity extension of circular DNA Anal Chem 2004;76:5605–10. Res 2003;31:3406–15. aptamers with real-time protein detection. Nu- 8. Zhang H, Wang Z, Li X-F, Le XC. Ultrasensitive 13. Applied Biosystems/Ambion. DNase I de- cleic Acids Res 2005;33:e64. detection of proteins by amplification of affinity mystified. http://www.ambion.com/techlib/tips/ 4. Balamurugan S, Obubuafo A, Soper S, Spivak D. aptamers. Angew Chem Int Ed Engl 2006;45: dnase1demystified.html (Accessed June 2009). Surface immobilization methods for aptamer di- 1576–80. 14. Galas DJ, Schmitz A. DNAse footprinting: a sim- agnostic applications. Anal Bioanal Chem 2008; 9. Fidler AE, Lin JS, Lun S, Ng Chie W, Western A, ple method for the detection of protein-DNA 390:1009–21. Stent V, McNatty KP. Production of biologically binding specificity. Nucleic Acids Res 1978;5: 5. Mairal T, Ozalp V, Lozano Sanchez P, Mir M, active tethered ovine FSH␤␣ by the methylotro- 3157–70. Katakis I, O’Sullivan C. Aptamers: molecular tools phic yeast Pichia pastoris. J Mol Endocrinol 2003; 15. Roche. DNase I recombinant. https://www.roche- for analytical applications. Anal Bioanal Chem 30:213–25. applied-science.com/pack-insert/4536282a.pdf 2008;390:989–1007. 10. Kuipers OP. Random mutagenesis by using mix- (Accessed June 2009).

Clinical Chemistry 55:9 (2009) 1693