Streamlined circular proximity ligation assay provides PNAS PLUS high stringency and compatibility with low-affinity

Roxana Jalilia,b, Joe Horeckaa,1, James R. Swartzb, Ronald W. Davisa,2, and Henrik H. J. Perssona,2

aStanford Genome Technology Center, Stanford University School of Medicine, Palo Alto, CA 94304; and bDepartment of Chemical Engineering, Stanford University, Stanford, CA 94305

Contributed by Ronald W. Davis, December 7, 2017 (sent for review October 20, 2017; reviewed by Simon Fredriksson and S. Patrick Walton) Proximity ligation assay (PLA) is a powerful tool for quantitative probes are individually conjugated to short ssDNA molecules to detection of biomarkers in biological fluids and tissues. form proximity probes that carry either a phosphorylated 5′ end Here, we present the circular proximity ligation assay (c-PLA), a or a 3′-hydroxyl group. When the probe pairs subsequently bind highly specific protein detection method that outperforms tradi- to their cognate target analyte in solution, the associated DNA tional PLA in stringency, ease of use, and compatibility with low- strands are brought into close proximity and aligned by hybrid- affinity reagents. In c-PLA, two proximity probes bind to an ization to a third bridging oligonucleotide. The free DNA ends analyte, providing a scaffolding that positions two free oligonu- are ligated, forming a new DNA sequence that is amplified and cleotides such that they can be ligated into a circular DNA quantified using qPCR (Fig. 1A). We have previously showed molecule. This assay format stabilizes proximity probe that PLA provides femtomolar sensitivity and a wide dynamic complexes and enhances stringency by reducing the probability of range over five orders of magnitude, consuming as little as 1 μL random background ligation events. Circle formation also in- of sample (12). One key benefit of PLA is that it addresses the creases selectivity, since the uncircularized DNA can be removed widespread problem of cross-reactivity in -based protein enzymatically. We compare this method with traditional PLA on detection. Potential signals from cross-reactive antibodies are several biomarkers and show that the higher stringency for c-PLA eliminated by tailoring the DNA sequences, such that ligation improves reproducibility and enhances sensitivity in both buffer only takes place when cognate proximity probes are bound to the and human plasma. The limit of detection ranges from femtomolar target analyte. We have applied this technology to detect more to nanomolar concentrations for both methods. Kinetic analyses than 20 different biomarkers in clinical plasma samples using using surface plasmon resonance (SPR) and biolayer interferome- panels of seven multiplex PLA reactions (13, 14). Others have try (BLI) reveal that the variation in limit of detection is due to the expanded the multiplexing capability even further (15). variation in antibody affinity and that c-PLA outperforms tradi- One of the limitations of affinity-based immunoassays, in- tional PLA for low-affinity antibodies. The lower background cluding PLA, is the so-called hook effect, in which the signal

signal can be used to increase proximity probe concentration BIOCHEMISTRY while maintaining a high signal-to-noise ratio, thereby enabling decreases at high antigen concentrations, resulting in incorrectly low the use of low-affinity reagents in a homogeneous assay format. signals or even false negatives (16). The hook effect becomes pre- We anticipate that the advantages of c-PLA will be useful in a dominant when the analyte concentration exceeds the concentration variety of clinical protein detection applications where high-affinity reagents are lacking. Significance

proximity ligation assay | antibody affinity | kinetic analysis | Quantitative detection of protein biomarkers over a wide immuno-PCR | qPCR concentration range from minute amounts of blood is essential for clinical diagnostics. Proximity ligation assay combines an- uantitative detection of protein biomarkers in biological tibody–oligo conjugates, enzymatic ligation, and PCR amplifi- Qfluids is essential for diagnosis, monitoring, and personal- cation into a sensitive method for quantitative protein ized treatment of disease. Despite considerable progress in re- detection from small volumes. This report describes a stream- cent years, the clinical use of validated proteomic biomarkers lined and more stringent assay format that takes advantage of remains limited (1). The benchmark for affinity-based protein DNA circle formation to remove unwanted DNA molecules. – measurements is defined by the ELISA where affinity ligands Kinetic analysis of antibody antigen interactions shows that (e.g., antibodies) are used in a sandwich format to detect and variation in assay performance between various biomarkers is quantify the protein of interest (2, 3). ELISA generally involves an effect of antibody quality. We show that this assay format several steps starting with sample incubation, where the target enables compatibility with low-affinity reagents, a major limi- analyte is captured on a surface precoated with primary anti- tation for most protein quantitation methods, while improving bodies, followed by washing steps and recognition with a sec- sensitivity and reproducibility. ondary antibody, which facilitates detection with a colorimetric, Author contributions: R.J., J.H., J.R.S., R.W.D., and H.H.J.P. designed research; R.J., J.H., fluorescent, or luminescent label. ELISA offers reasonable sen- and H.H.J.P. performed research; R.J., R.W.D., and H.H.J.P. analyzed data; and R.J. and sitivity, but it requires a large sample volume, has limited dynamic H.H.J.P. wrote the paper. range, and frequently suffers from false positives due to nonspecific Reviewers: S.F., Genagon Therapeutics; and S.P.W., Michigan State University. binding (4, 5). These limitations interfere with the discovery and Conflict of interest statement: S.F. owns stock in a company (Olink AB) with patents on the validation of novel biomarker candidates that have the potential to core technology described, "Proximity Ligation Assay." enable early diagnosis and regular molecular monitoring of disease. Published under the PNAS license. Immunoassays combined with nucleic acid-based amplification 1Deceased October 20, 2017. and detection have facilitated new approaches and have ex- 2To whom correspondence may be addressed. Email: [email protected] or hpersson@ tended the analytical sensitivity beyond that achievable with stanford.edu. – ELISA (6 9). One particularly promising approach is the prox- This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. imity ligation assay (PLA) (10, 11). In PLA, pairs of affinity 1073/pnas.1718283115/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1718283115 PNAS | Published online January 16, 2018 | E925–E933 Downloaded by guest on September 25, 2021 A

Ligation Termination Preamplification Detection qPCR 1 Oligo Uracil Excision

B

Ligation Termination Preamplification Detection X qPCR 2 Oligos Exonuclease P2 P1

Fig. 1. Schematic representation of t-PLA and c-PLA. (A) t-PLA detects using pairs of antibody–DNA conjugates (red and blue), which are brought into close proximity on binding to target analyte. The addition of a bridge oligonucleotide and DNA ligase enables ligation of the antibody-tethered oli- gonucleotides to form a new DNA sequence. Ligation is terminated by selective degradation of the bridge oligonucleotide. The newly formed ligation product is subsequently preamplified followed by quantification using qPCR. (B) In c-PLA, the antibody-tethered oligonucleotides act as bridges for two li- gation events between free oligonucleotides, resulting in the formation of a circular ligation product. The addition of an extra oligonucleotide increases stringency compared with t-PLA; it lowers the probability of random background ligation events, since four components must assemble in the absence ofthe target analyte to generate an independent circular ligation product. Circle formation also allows exonuclease treatment, which terminates ligation and reduces background by degrading all uncircularized DNA. The reduction in background also simplifies the workflow by eliminating the need for pre- amplification. Circular ligation products are quantified by qPCR using primer sites spanning the newly formed junctions (P1 and P2).

of proximity probes (>>1 nM), and it effectively determines the although detailed kinetic characterization of antibodies used in upper quantification limit (17). The proximity probe concen- PLA and its relationship to assay performance remain unexplored. tration can potentially be increased but must be carefully bal- We suggest that a better understanding of how antibody–antigen anced against a deteriorating signal-to-noise ratio from random affinity and kinetic parameters affect PLA performance has great background ligation events when probes and the bridge oligo value for assay development and optimization efforts. come together by chance. The hook effect can ordinarily be Here, we describe a circular proximity ligation assay (c-PLA), avoided by sample dilution; however, dilution alters binding where in contrast to traditional proximity ligation assay (t-PLA), equilibrium. This does not present a problem for high-affinity the proximity probes are used as bridges that enable the con- interactions, but it may prevent low-affinity antibodies from nection of two free oligonucleotides via dual ligation events, binding to their target analytes, thereby limiting the effective- resulting in the formation of a circle (Fig. 1B). The addition of an ness of PLA to high-affinity capture agents (18). While in- extra oligonucleotide decreases the probability of random compatibility with low-affinity antibodies is a major drawback background ligation events due to stringency in assay design. We for PLA, it is not restricted to this assay alone. The availability define stringency as the minimization of background ligation, of high-affinity antibodies or other capture reagents is a general since four molecules (two proximity probes and two free con- limitation for all sandwich immunoassays, because the sensi- nector oligonucleotides) must come together in the absence of tivity is ultimately determined by the quality of the reagents target antigen to form a circle. In addition, circle formation has used (19). selective advantages, as uncircularized DNA can be removed by Numerous approaches have been developed to improve PLA a simple exonuclease treatment (32), streamlining the workflow performance, including concentration of minute amounts of by eliminating preamplification before qPCR. As a result, our c- analyte on a solid phase before ligation (10, 18, 20), use of PLA is much more straightforward than not only PLA but also, multivalent proximity probes (21), addition of multiple affinity most existing protein detection methods and can be performed in probes (22, 23), special design of asymmetric bridge oligos (17), a single reaction tube with a tiny sample volume (2 μL). The inclusion of protecting oligonucleotides prehybridized to prox- assay format utilizes the same affinity probes as in t-PLA, which imity probes to reduce background ligation events (22, 24), and enable a direct performance comparison between the two the use of novel amplification schemes (25–28). Some of the methods. We apply both assays to six biomarkers and show equal amplification schemes entail enzymatic manipulations, where the or better performance for c-PLA with improvements in pre- ligation products are released from antibodies and converted cision. We also provide kinetic binding analysis of the antibodies into circles used for isothermal rolling circle amplification (26, used and find a direct correlation between antibody affinity and 29–31). Many of these approaches have improved assay re- assay performance. We show that kinetic information can be producibility, although precision still remains a challenge for used to guide assay development: the enhanced stringency in c- adaptation into clinical diagnostics. The relationship between af- PLA is utilized to increase affinity probe concentration while finity and assay sensitivity has been described by Gullberg et al. maintaining a low probability of background ligation events— (11). They showed that the dose–response curve and limit of de- thereby improving signal-to-background ratio. Finally, we show tection correlated with modeled equilibrium affinity constants, that c-PLA is compatible with complex mixtures, such as human

E926 | www.pnas.org/cgi/doi/10.1073/pnas.1718283115 Jalili et al. Downloaded by guest on September 25, 2021 plasma, and provides a path for adaptation with low-affinity events facilitated by target-bound proximity probes. The uracil PNAS PLUS antibodies as well as alternative capture agents (33, 34) with- excision step is replaced with exonuclease treatment, which has out the need for preconcentration on solid phase. the selective advantage of enriching for circularized DNA (32). Consequently, the background is dramatically decreased as all Results and Discussion uncircularized nucleic acids are degraded as opposed to only the c-PLA Workflow. The workflows for both the t-PLA and the c-PLA bridge oligo in t-PLA. This allows for omission of the pre- are shown in Fig. 1. Both assays utilize the same set of proximity amplification before qPCR analysis without any loss in signal-to- probes that are prepared by bioconjugation of polyclonal anti- noise ratio or reproducibility. bodies with amine-modified oligonucleotides through aromatic hydrazone chemistry (details are provided in SI Methods). Assay Characteristics. We designed the c-PLA from existing t-PLA Polyclonal antibodies are cost-efficient, as a single batch of an- probes that have been optimized for minimization of heterodu- tibodies is divided into two portions and coupled to oligonucle- plex formation (12). This allowed us to compare the two methods otides that are terminated by either a phosphorylated 5′ end or a using the same set of proximity probes. For c-PLA, we relaxed the 3′ end. This produces two heterogeneous mixtures of proximity ligation conditions to account for the requirement of two ligation probes against several different epitopes of the same target an- events and to accommodate the efficiency of Ampligase at higher tigen. During sample incubation, the antibody portions of these temperature. Ligation was consequently performed for 30 min at proximity probes bind to distinct epitopes of the target analyte. 45 °C instead of 15 min at 30 °C. A comparable modification of This is followed by a ligation step, in which a solution containing t-PLA did not yield any improvements. DNA ligase is added to the incubation mixture. In t-PLA, this A direct comparison between t-PLA and c-PLA for six bio- ligation mixture contains a third bridge oligonucleotide com- markers is shown in Fig. 2, and the results are summarized in plementary to the ends of the proximity probes, thereby facili- Table 1. These experiments were performed with three bi- tating ligation to form a new DNA sequence. Ligation is ological replicates and quantified in triplicate qPCR experi- terminated by the addition of a uracil excision mixture, which ments, generating nine data points for each concentration. For selectively degrades the bridge oligonucleotide. The ligation VEGF, a dose-dependent response across nearly five orders of mixture is subsequently preamplified across the newly formed magnitude is shown for both methods (Fig. 2A). The difference DNA junction to increase signal-to-background ratio and re- of almost four orders of magnitude in counts (estimated number producibility (12, 35) before quantification by qPCR. In c-PLA, of ligated molecules) generated by the two methods is due to ligation products are not formed by a direct junction of proximity both the elimination of preamplification and the enhanced probes but rather, by the formation of a circle when two free stringency for c-PLA. For comparison, we performed pre- connector oligonucleotides are joined in two distinct ligation amplification in c-PLA and found that the omission of this step

A B C 1011 1011 1011 1010 1010 1010 BIOCHEMISTRY 109 109 109 108 108 108 107 107 107 106 106 106 105 105 105 104 104 104 Counts [-] Counts [-] Counts [-] 103 103 103 2 VEGF 2 GDNF 2 IL-6 10 t-PLA 10 t-PLA 10 t-PLA 1 1 1 10 c-PLA 10 c-PLA 10 c-PLA 100 100 100 10-14 10-13 10-12 10-11 10-10 10-9 10-8 10-14 10-13 10-12 10-11 10-10 10-9 10-8 10-14 10-13 10-12 10-11 10-10 10-9 10-8 Ag Concentration [M] Ag Concentration [M] Ag Concentration [M] D E F 109 109 109 108 108 108 107 107 107 106 106 106 105 105 105 104 104 104 Counts [-] Counts [-] Counts [-] 3 3 3 10 MIF 10 TNF 10 IGF-II 2 2 t-PLA 2 t-PLA 10 t-PLA 10 10 c-PLA c-PLA c-PLA 101 101 101 10-11 10-10 10-9 10-8 10-7 10-11 10-10 10-9 10-8 10-7 10-6 10-11 10-10 10-9 10-8 10-7 10-6 Ag Concentration [M] Ag Concentration [M] Ag Concentration [M]

Fig. 2. Dose–response curves of t-PLA and c-PLA for detection of VEGF (A), GDNF (B), IL-6 (C), MIF (D), TNF-α (E), and IGF-II (F). The x axis displays antigen (Ag) concentration, and the y axis displays an estimated number of ligated molecules. The enhanced stringency for c-PLA is shown by a lower number of counts because of the rigor imposed by circle formation, background reduction through exonuclease treatment, elimination of preamplification, and tailored qPCR primer sites. Error bars denote 1 SD (n = 9), and the dashed lines denote limit of detection, defined as the mean signal of a blank sample +3 SD.

Jalili et al. PNAS | Published online January 16, 2018 | E927 Downloaded by guest on September 25, 2021 Table 1. Comparison of limit of detection and dynamic range for six biomarkers measured by c-PLA and t-PLA c-PLA t-PLA

Analyte Function Data in figure Limit of detection Dynamic range CVs, % Limit of detection Dynamic range CVs, %

VEGF Growth factor Fig. 2A <10 fM <5 15.6 10 fM <5 29.3 GDNF Cell survival Fig. 2B 15 fM <5 12.4 10 fM 5 30.8 IL-6 Inflammation Fig. 2C 100 fM <4 28.2 150 fM <4 48.5 MIF Innate immunity Fig. 2D 3 pM 3 10.4 5 pM 3 39.9 TNF-α Cell signaling Fig. 2E 20 pM 3 19.4 70 pM <3 44.2 IGF-II Inflammation Fig. 2F 1 nM 2 17.0 3 nM <2 24.3

accounts for approximately three orders of magnitude difference reported limit of detection using various definitions (12, 15). in signal, whereas stringency in assay design accounts for the Here, we strictly define limit of detection as the mean signal remainder (Fig. S2). Preamplification was originally introduced corresponding to a blank sample plus three SDs (40). Discrep- to improve signal-to-background ratio and precision, although ancies in reported limit of detection can be affected by the use of t-PLA is still impeded by relatively high coefficients of variation alternative definitions, differences in protocol, or batch-to-batch (CVs), an obstacle for standard use in clinical diagnostics. Var- variation in proximity probe conjugation, as affinity probes with- iation in t-PLA has traditionally been addressed by addition of out conjugated oligonucleotides may bind to target analytes and internal controls and normalization of data, allowing for relative interfere with the ability to facilitate ligation. One benefit of using biomarker profiling (13, 15). Background signals for PLA are the same batch of proximity probes for both traditional and cir- generated from random ligation events, nonspecific binding of cular proximity ligation is that such effects are eliminated, making two cognate proximity probes, or nucleic acid amplification ar- comparison between the two methods more straightforward. tifacts. In c-PLA, random ligation events are minimized by the Furthermore, keeping the oligonucleotide sequences constant for stringency in assay design, as four molecules and two ligation all analytes eliminates any variation that may arise from sequence events are required for signal generation. Nonspecific binding specificity, albeit that this limits the possibility of multiplexing for effects are addressed by the addition of excess bulk IgG mole- the purpose of this study. The similarities in performance be- cules from the same species as the proximity probes (12). Nucleic tween the two methods indicate that the variation in limit of acid amplification artifacts are minimized in c-PLA, as dual li- detection between analytes is inherently an effect of the quality of gation events result in two new DNA sequences at the junction the proximity probes and their ability to enable antibody–antigen sites. We tested multiple primers targeting different sites on the interactions. This motivated us to further investigate the kinetic newly formed circular DNA and found that a primer pair properties of these interactions. spanning both junctions produced the lowest background. Exo- nuclease treatment in c-PLA also reduces background, which Kinetic Analysis of Affinity Reagents. Knowledge about the kinetics enhances signal-to-noise ratio and assay performance. This im- of binding between antibodies and is important for all proves precision within the linear dynamic range as shown by a immunoassays (3, 19). The wide range in limit of detection decrease in the average CVs from 29% for t-PLA to below 16% among the analytes studied in this report led us to further ex- for c-PLA (Table 1). PLA precision in general is limited by the plore the relationship between kinetics and assay performance. qPCR readout, and variation at low counts can be decreased A previous report showed good correlation in t-PLA between further with digital quantification methods (31, 36). limit of detection and theoretical affinity but did not determine We compared c-PLA with t-PLA for five additional bio- equilibrium dissociation constants (Kd) for the antibodies used markers: glial cell line-derived neurotrophic factor (GDNF), IL- (11). Furthermore, no analyses have been reported on either 6, macrophage migration inhibitory factor (MIF), TNF-α, and association rate constants, ka (on rates), or dissociation rate insulin-like growth factor II (IGF-II) (Fig. 2 B–F and raw data constants, kd (off rates) and their significance in PLAs. To ad- provided in SI Methods). We found that the higher stringency for dress this, we used surface plasmon resonance (SPR) (41) to c-PLA yielded improvements in precision and signal-to- determine Kd values as well as on and off rates for the antibody– background ratios and equal or better assay performance in antigen interactions investigated. Kinetic parameters are listed in terms of limit of detection and dynamic range for all analytes Table 2, and SPR sensorgrams for individual analytes fitted to a except GDNF. Average CVs across the linear range were all 1:1 binding model are provided in SI Methods. The results are below 20% for c-PLA, with the exception of IL-6 (28%), while also summarized in an isoaffinity graph (Fig. 3A), where the two CVs for t-PLA varied from 24 to 48%. Reproducibility is gen- rate constants are plotted against each other to provide Kd values erally better at higher concentrations, while precision deterio- along diagonal isoaffinity lines (42). The on rates varied 30-fold − − − − rates at low concentrations. This represents a common problem from 4.9 × 105 M 1 s 1 for IGF-II to 1.6 × 107 M 1 s 1 for − − for PLA and other highly sensitive methods for protein quanti- VEGF. Off rates varied from 1.6 × 10 4 s 1 for TNF-α to ex- tation (37–39). ceptionally slow off rates below the sensitivity range of the Bia- − − Notable observations are that the limit of detection for the core T200 instrument (1.0 × 10 5 s 1) for both GDNF and IL-6. analytes varies more than five orders of magnitude from low The software used to fit the data still provided off rates beyond femtomolar to nanomolar concentration and that the dynamic the limit of the instrument (results are in SI Methods), but be- range is confined by the limit of detection on the lower end and cause of uncertainty in the accuracy, we limited off rates to no −5 −1 the proximity probe concentration on the upper end, where the lower than 1.0 × 10 s in the calculation of Kd. The resulting hook effect starts to interfere. Consequently, IGF-II, which ex- Kd values spanned more than two orders of magnitude, rang- − − hibits the worst limit of detection (1 nM) among the analytes that ing from 2.8 × 10 10 MforTNF-α to below 7.1 × 10 13 M we tested, also displays a restricted dynamic range of about two for GDNF. orders of magnitude. The assay performance for the same bio- Because some of the off rates determined by SPR were found markers in t-PLA generally follows the same trend as found for to be below the sensitivity range of the Biacore instrument, we c-PLA, albeit with a slightly higher limit of detection. We note performed complementary analysis using biolayer interferometry that previous proximity ligation studies of the same analytes have (BLI) (43). These measurements corroborated the SPR results;

E928 | www.pnas.org/cgi/doi/10.1073/pnas.1718283115 Jalili et al. Downloaded by guest on September 25, 2021 Table 2. Summary of kinetic analysis data for six biomarkers using SPR PNAS PLUS Association rate Dissociation rate Equilibrium dissociation −1 −1 −1 Analyte constant ka,M s constant kd,s constant Kd,M

GDNF 1.4 E+07 <1.0 E−05 <7.1 E−13 IL-6 2.3 E+06 <1.0 E−05 <4.3 E−12 VEGF 1.6 E+07 5.6 E−05 3.6 E−12 IGF-II 4.9 E+05 2.5 E−05 5.1 E−11 MIF 1.1 E+06 9.4 E−05 8.3 E−11 TNF-α 5.7 E+05 1.6 E−04 2.8 E−10

off rates for GDNF and IL-6 were also beyond the sensitivity of their own individual kinetic characteristics. Due to this hetero- the Fortebio Octet RED instrument. Kinetic data for BLI geneity, kinetic properties of polyclonal antibody are inherently measurements, binding curves for individual analytes fitted to a difficult to characterize with precision, and the derived kinetic 1:1 binding model, and an isoaffinity chart for all analytes are constants and affinities are considered an average for the dif- provided in SI Methods. Kd values determined by BLI varied ferent subpopulations existing within a batch of antisera. One − more than three orders of magnitude, ranging from 6.4 × 10 9 M prospective effect is that extended incubation time between − for TNF-α to below 2.5 × 10 12 M for GDNF and IL-6. These sample and proximity probes will allow continued exchanges that values are approximately an order of magnitude higher than the progress toward higher-affinity interactions. corresponding SPR data, largely due to differences in the de- termination of on rates. SPR is a flow cell-based method with a Using Kinetics Data to Improve c-PLA Performance. The isoaffinity 3D dextran matrix, whereas BLI utilizes planar fiberoptic sensors graph in Fig. 3A includes the kinetic information for all anti- in a well-plate format. It has previously been reported that the body–antigen interactions investigated, revealing a clear differ- BLI system underestimates fast on rates due to mass transfer entiation in affinities. The antibodies with the highest affinity are limitations, which is consistent with our findings (44). The nu- located in the upper left corner, as they are characterized by low meric values of kinetic constants determined by surface-based off rates and high on rates that result in low Kd values. GDNF, methods are likely to differ from solution-based values, which VEGF, and IL-6 antibodies have high affinity, with Kd values are presumably more applicable to homogenous PLA. None- below 5 pM. Accordingly, these analytes provide subpicomolar theless, the relative rankings between the two methods display limits of detection in c-PLA and wide dynamic ranges of at least good agreement and are important indicators of comparative four orders of magnitude as seen in dose–response curves (Fig. antibody quality. 3B: dose–response curves are derived from the same batches of All proximity probes described in this work originate from antibodies and antigens as those used in Fig. 3A but are not affinity-purified polyclonal antibodies. These antibodies are completely the same batches as those used in Fig. 2). The

mixtures derived from different cell lineages, producing anti- remaining analytes, MIF, TNF-α, and IGF-II, are clustered to- BIOCHEMISTRY bodies that recognize distinct epitopes of the antigen each with gether, with Kd values above 50 pM. Consequently, c-PLA for

A B 108 106

5 107 10

104 106 103 [1/Ms] GDNF

a GDNF k IL-6 Counts [-] IL6 105 VEGF 102 VEGF IGF-II IGF-II MIF MIF 1 TNF 10 TNF 104 10-8 10-7 10-6 10-5 10-4 10-3 10-2 10-1 10-14 10-13 10-12 10-11 10-10 10-9 10-8 10-7 10-6 k [1/s] Ag Concentration [M] d

Fig. 3. Isoaffinity analysis and corresponding c-PLA performance for six biomarkers. (A) Kinetic analysis reveals two distinct groups for Kd values: one group with high affinity (single-digit picomolar) and another group with affinity above 50 pM. (B) The differences in antibody affinities are directly reflected in the

c-PLA dose–response curves, where analytes with low Kd values display limits of detection in the subpicomolar range, while the other group exhibits limits of detection in the midpicomolar range or higher.

Jalili et al. PNAS | Published online January 16, 2018 | E929 Downloaded by guest on September 25, 2021 − − − these analytes does not perform nearly as well, exhibiting limits from 10 4 to below 10 5 s 1, values that are comparable with or of detection in the picomolar range (or higher) and narrower lower than the off rates for the antibody–antigen interactions dynamic ranges that are constrained by the lower affinity of the described in this work (45, 46). The calculations above highlight antibodies. This comparison shows that the inherent difference not only the benefits of high-affinity probes in PLAs but also, the in sensitivity and dynamic range among the six analytes is not importance of slow off rates or other means to demote dissocia- entirely dependent on the assay design but rather, is caused by tion, especially when it entails large-volume additions for ligation. the quality of the affinity reagents used. We note that the only The large-volume addition associated with ligation was origi- biomarker for which c-PLA did not improve assay performance nally introduced to minimize background ligation in t-PLA (12). compared with t-PLA is GDNF, the analyte with the highest Background ligation events in t-PLA occur when affinity probes affinity. The improvements in limit of detection for c-PLA over and bridge oligo are randomly brought into close proximity. For t-PLA are also larger for the lowest-affinity interactions (Tables c-PLA, the likelihood of background ligation is lower, as this 1 and 2), indicating a trend toward increased benefits of c-PLA complex requires an additional connector oligo and two ligation when only low-affinity reagents are available. events before producing a detectable signal. This results in less The detailed information obtained by kinetic analysis is ben- background and a better signal-to-noise ratio. Additional mod- eficial for assay development, as it suggests modifications that eling studies provided in SI Methods (Fig. S6E) also indicate that may result in improved signals. We used the kinetic constants to higher proximity probe concentrations result in more complex predict the number of complexes formed at equilibrium during formation, which may increase signal-to-noise ratios even fur- sample incubation and the new equilibrium that is established ther. This is analogous to experimental findings on solid phase, when the volume is increased by addition of the ligation mixture. where higher densities of surface-bound capture antibodies have Details are provided in SI Methods. The calculations suggest that been shown to improve both limit of detection and dynamic ∼50% of the antigens present in solution are captured in a range (47). Consequently, the combination of using a more complex at equilibrium during sample incubation when the Kd stringent assay format, like c-PLA, with increased proximity − value of an interaction equals 1 × 10 10 M (comparable with probe concentrations should improve assay performance with − − MIF and TNF-α; 8.3 × 10 11 and 2.8 × 10 10 M, respectively) low-affinity antibodies. (Fig. S6B). However, the subsequent addition of ligation mixture triggers complex dissociation until a new equilibrium is estab- Stringency of c-PLA Improves Signal-to-Background Ratio and Limit −10 lished. Modeling using Kd = 1 × 10 M indicates that, at the new of Detection. Based on simulations described above, we tested the equilibrium, essentially no antigens would remain in a complex, impact of increased proximity probe concentration for both t- resulting in no circle formation and consequently, no detectable PLA and c-PLA. We chose to study this using the TNF-α assay, as qPCR signal (Fig. S6C). However, the assay still functions for it displayed the highest Kd and kd, providing a good example of analytes with these Kd values, and the discrepancy is explained by low-affinity reagents with fast off rates. We performed both assay suppressed dissociation as predicted by the low off rates. The formats with the standard affinity probe concentration (1×= complex half-life (t1/2) is defined as ln2/kd, indicating that, for 0.25 nM) as well as a 10-fold increase. The results are shown in − − both MIF and TNF-α (which have off rates of ∼1 × 10 4 s 1), it Fig. 4A and show a consistently higher signal-to-background ratio would take roughly 2 h for one-half of the complexes to dissociate. for c-PLA compared with t-PLA. The increase in signal-to- This is a sufficient duration to avoid significant losses during a background ratio between the 10× and 1× affinity probe cases 30-min ligation step (or 15 min in the case of t-PLA). In addition, for the two methods is an expansion of our previous findings that these calculations do not consider any added advantages that the probe concentration can be increased for low-affinity interactions two circle-forming connector oligos have over t-PLA in retarding (12), although it is a complex subject that must be carefully bal- proximity probe diffusion away from the analyte, thereby facili- anced between antibody affinity and a higher probability of ran- tating antibody rebinding. DNA duplexes can be remarkably stable dom background ligation (48). The greater signal-to-noise ratio for as exemplified by our SPR measurements, which use DNA- c-PLA is explained by a reduction in background noise as seen in directed immobilization of the capture agents. Previous studies Fig. 4B. This is a consequence of rigorous assay design in combi- have determined the off rates for short oligonucleotides to range nation with exonuclease treatment, which effectively eliminates the

A B C 108 1010

9 10000 TNF 7 10 c-PLA 1X 10 -10 10 M TNF (Signal) 8 c-PLA 10X 10 106 t-PLA 1X 7 1000 10 t-PLA 10X 5 0 M TNF (Noise) 10 106 104 105 100 4 103 10 Counts [-] Counts Counts [-] Counts 103 TNF 102 t-PLA 10X 10 2 10 t-PLA 1X 1 c-PLA 10X 10 101

Signal-to-Background Ratio [-] Signal-to-Background c-PLA 1X 1 100 100 10-11 10-10 10-9 10-8 10-7 10-6 -- 0 9 8 7 X X X X ttt -11 0- 0- - 1 0 1 0 1 1 0 A A 1 10-1 x1 PL LA1 PL LA Ag Concentration [M] 10 M 5 M c- -P t- -P Ag Concentration [M] c t

Fig. 4. Comparison of PLA methods at different probe concentrations for TNF-α.(A) c-PLA results offer a larger signal-to-background ratio than t-PLA. (B) Individual components for 100 pM signal-to-noise ratio. The greater signal-to-noise ratio for c-PLA is a consequence of higher stringency in c-PLA, which produces lower overall signals and larger differences between positive signal (100 pM) and negative background noise. (C) Dose–response curves showing that the higher signal-to-background ratios result in a more than 10-fold improvement in limit of detection (7 pM) for c-PLA when the probe concentrations are increased 10-fold compared with t-PLA (1×). Error bars denote 1 SD (n = 9), and the dashed lines denote limit of detection, defined as the mean signal of a blank sample +3 SD. Ag, antigen.

E930 | www.pnas.org/cgi/doi/10.1073/pnas.1718283115 Jalili et al. Downloaded by guest on September 25, 2021 background noise. Note that the higher stringency of c-PLA also assay format without the need for preamplification. We per- PNAS PLUS lowers the signal but that the stability of circle-forming complexes formed additional experiments with VEGF in chicken plasma (in combined with even lower background noise more than compen- which human VEGF is absent) to verify that the changes in limit sates for this decline. The higher signal-to-background ratio for c- of detection and dynamic range are not impeded by the assay PLA 10× improves limit of detection about one order of magni- format but are rather an effect of naturally occurring background tude compared with the original t-PLA (Fig. 4C). Additional ex- levels of VEGF in human plasma (Fig. 5A). In chicken plasma, periments suggest that the probe concentration can be increased we achieved femtomolar limit of detection and more than four even further for low-affinity interactions, although it must be ac- orders of magnitude dynamic range, which more closely reflect companied by an increase in the amount of connector oligonu- our original measurements in buffer solution. Hence, both PLA cleotides to ensure efficient circle formation. methods are compatible with complex mixtures, like human plasma, with additional benefits in assay performance for c-PLA. c-PLA Effectively Detects Proteins in Human Plasma. The results The advantages in assay performance of c-PLA over t-PLA for described thus far were all from buffer solutions, as we sought to TNF-α (a low-affinity interaction) also persist in human plasma investigate the relationship between affinity and assay perfor- (Fig. 5B). The limit of detection for c-PLA is about 10 pM for mance. Affinity analysis using SPR is not amenable to complex both 1× and 10× proximity probe concentrations, with lower CVs mixtures, like plasma, and although BLI is less sensitive to matrix for the higher probe concentration (8% for 10× vs. 16% for 1×), effects (43), both methods require pure ligands for accurate presumably due to more efficient capture of antigens at the determination of kinetic constants. The correlation between ki- higher concentration. The limit of detection for t-PLA is ap- netic constants and assay performance is established in Fig. 3 and proximately an order of magnitude higher compared with c-PLA, provides valuable information for assay development. However, which reduces the dynamic range accordingly. CVs remain at assay performance is known to differ greatly between different around 30% across the linear range for both proximity probe matrices, and assay compatibility with complex mixtures is es- concentrations. In contrast to c-PLA, a 10-fold increase in sential for diagnostic and clinical relevance (1). proximity probe concentration for t-PLA did not improve sen- We have previously shown that multiplexed t-PLA enables sitivity (limit of detection about 700 pM), underscoring the quantitation of biomarkers in plasma samples (12–14). To de- challenges when using this method with low-affinity antibodies. termine the performance of c-PLA in a complex matrix, we The ability to increase probe concentration in c-PLA also par- tested two analytes: VEGF as an example of a high-affinity in- tially addresses the hook effect, as it results in a later onset of teraction (low Kd and slow off rates) and TNF-α as an example of proximity probe saturation. Furthermore, statistical analysis of a low-affinity interaction (high Kd and fast off rates). VEGF was the CVs for TNF-α in plasma reveals that the lower CVs for c- effectively detected in human plasma by c-PLA down to physi- PLA 10× are significantly different from CVs for c-PLA 1× as ological single-digit picomolar concentrations (12, 49) (Fig. 5A). well as t-PLA 1× and 10× (P < 0.001). The difference in CVs Limit of detection is increased 100-fold from femtomolar con- between c-PLA 1× and t-PLA 1× is also significant (P < 0.001). centration in buffer to 1 pM for c-PLA and 2 pM for t-PLA in This is encouraging, as the generally accepted precision re- human plasma. Dynamic range is reduced accordingly by two quirement for clinical immunoassays is below 20% (40, 50). orders of magnitude for both methods. Average CVs were lower These findings indicate that the increased stringency of c-PLA BIOCHEMISTRY for c-PLA than t-PLA (at 11 and 22%, respectively), which is provides advantages over t-PLA in terms of ease of use and consistent with our findings in buffer solutions. These results sensitivity, while reducing variation. Most importantly, the op- confirm that the c-PLA provides less variation in a simplified portunity to improve the performance of low-affinity antibodies

A B 1010 1010 109 109 108 108 107 107 106 106 105 105 104 104

Counts [-] 3 Counts [-] 3 10 VEGF in Plasma 10 TNF in Human Plasma 102 t-PLA Human 102 t-PLA 10X t-PLA Chicken t-PLA 1X 101 c-PLA Human 101 c-PLA 10X c-PLA Chicken c-PLA 1X 100 100 10-14 10-13 10-12 10-11 10-10 10-9 10-8 10-11 10-10 10-9 10-8 10-7 10-6 Ag Concentration [M] Ag Concentration [M]

Fig. 5. Performance of PLAs in human plasma. (A) Dose–response curves for VEGF detection show assay compatibility for both PLA methods in human and chicken plasma. The difference in limit of detection between the two matrices is attributed to endogenous VEGF levels in human plasma that are absent in chicken plasma. (B) Dose–response curves for TNF-α detection in human plasma showing improvement in assay performance for c-PLA over t-PLA. A 10-fold increase in probe concentration (10×=2.5 nM) improves reproducibility for c-PLA, while there is no improvement for t-PLA. Error bars denote 1 SD (n = 9), and the dashed lines denote limit of detection, defined as the mean signal of a blank sample +3 SD. Ag, antigen.

Jalili et al. PNAS | Published online January 16, 2018 | E931 Downloaded by guest on September 25, 2021 by simply increasing their concentration in c-PLA is maintained Antibody–oligonucleotide conjugates were analyzed on a 2100 Agilent in complex matrices, like plasma and serum. Bioanalyzer using the protein 80 kit under reducing conditions following the manufacturer’s protocol. The estimated yield for all conjugates varied be- Conclusions tween 0.65 and 1.35 oligonucleotides per antibody. Final antibody–oligo- In this report, we present a highly specific, sensitive, and con- nucleotide concentration was determined using a Bradford protein assay (Bio-Rad) according to the manufacturer’s specification. Antibody–oligo venient assay for quantitative protein detection. Our method probes were diluted to 5 nM in 1× PBS supplemented with 2 mM EDTA builds on the existing PLA but also benefits from the formation (Thermo Fisher Scientific), 0.10% BSA, and 0.02% NaN3 and were stored at 4 °C. of circular DNA molecules. PLAs combined with circle forma- tion have traditionally been used for rolling circle amplification Proximity Probe Target Incubation. Two microliters of sample [diluted in either and have shown impressive results for in situ detection and 1× PBS with 0.1% BSA or neat plasma from chicken or human (Sigma- digital quantification. We show that the circle formation in- Aldrich)] was added to 2 μL of proximity probe mix and incubated for 2 h creases reproducibility by minimizing noise across the linear at 37 °C to establish complex formation between the target analyte and dynamic range. This is facilitated by the inclusion of exonuclease antibodies in proximity probes. The combined 4 μL incubation mixture treatment, which hydrolyzes noncircular DNA. This design also contained a final concentration of 250 pM for each proximity probe and was simplifies the workflow, as it eliminates the need for pre- supplemented with 0.35 mg/mL polyadenylic acid potassium salt, 1% BSA, 0.1% Triton X-100, 0.05% IgG, 0.01% aprotinin, 1 mM PMSF, and 4 mM amplification before qPCR quantification without any loss in EDTA (Thermo Fisher Scientific) in 0.375× PBS. assay performance. We compare both methods using six com- monly used biomarkers and show equal or better performance Ligation Step for c-PLA. One hundred twenty microliters of ligation mixture for c-PLA. was added to each 4-μL sample and incubated for 30 min at 45 °C. The li- Our kinetic analysis of the required antibody–antigen inter- gation mixture contained 100 nM each of the two circle-forming connector actions, an essential precaution rarely taken during assay de- oligos (SI Methods), 0.025 U/μL Ampligase (Epicentre), 0.5 mM NAD, 1 mM velopment for protein detection, shows a direct correlation DTT (Sigma-Aldrich), 0.01% Triton X-100, and 0.01% BSA in 20 mM Tris, between kinetic constants and assay performance. We establish pH 8.4, 10 mM MgCl2, and 50 mM KCl (Thermo Fisher Scientific). Ligation the value of kinetic information to screen for suitable antibodies was terminated by adding 10 μL of exonuclease mixture and incubated for before proximity probe conjugation and to optimize proximity 30 min at 37 °C followed by heat inactivation of exonuclease enzymes for μ probe concentrations as well as assay procedures. Our work thus 20 min at 80 °C. Exonuclease mixture contained 2 U/ L exonuclease I (New England Biolabs) and 2 U/μL exonuclease III (New England Biolabs) in 1× indicates that the regular use of kinetic analysis will be highly NEBuffer 1 containing 10 mM Bis·Tris·Propane·HCl, pH 7.0, 10 mM MgCl2, beneficial in improving assay performance across a wide variety and 1 mM DTT (New England Biolabs). of analytes. The advantages of c-PLA over t-PLA for low-affinity inter- Ligation Step for t-PLA. One hundred twenty microliters of ligation mixture actions are aided by suppressed dissociation provided by longer was added to each 4-μL sample and incubated for 15 min at 30 °C. The li- DNA duplexes. We show that the additional proofreading step gation mixture contained 100 nM t-PLA bridge oligo, 0.025 U/μL Ampligase facilitated by the extra oligo in c-PLA can be exploited to in- (Epicentre), 0.25 mM NAD, 10 mM DTT, 0.02% Triton X-100, and 0.01% BSA crease proximity probe concentration without a concomitant in 20 mM Tris, pH 8.4, 1.5 mM MgCl2, and 50 mM KCl (Thermo Fisher Sci- increase in background ligation. This improves the signal-to- entific). Ligation was terminated by adding 2 μL of stop ligation mixture followed by a 5-min incubation at room temperature. The stop ligation noise ratio and limit of detection compared with t-PLA and μ enables the use of low-affinity reagents. This is a major benefit, mixture contained 0.125 U/ L Uracil-DNA Excision Mix (Epicentre) in 20 mM Tris, pH 8.4, and 50 mM KCl (Thermo Fisher Scientific). as low-affinity antibodies and variation in antibody specificity often contribute to systematic errors in research and clinical Preamplification Step for t-PLA. Twenty-five microliters of the terminated diagnostics (51). We also show that the benefits of c-PLA over t- ligation mixture was added to 25 μL preamplification mixture and cycled PLA persist in complex media, such as human plasma. This will with the following conditions: 95 °C for 3 min (1 cycle), 95 °C for 30 s and facilitate applications when high-affinity reagents are not avail- 60 °C for 4 min (13 cycles), and a final hold at 4 °C. The preamplification able, allowing not only more analytes to be detected but also, mixture contained 20 nM preamplification primers, 0.06 U/μL Platinum Taq greater sensitivity for protein detection. These advantages are DNA Polymerase (Thermo Fisher Scientific), and 1.6 mM dNTP Mix (Agilent strengthened by the ease of application of our assay as well as its Technologies) in 20 mM Tris, pH 8.4, 6 mM MgCl2, and 50 mM KCl (Thermo amenability to automation and a variety of quantification Fisher Scientific). After preamplification, the product was diluted 10-fold in methods (digital, fluorescent, electrical, and more). The forma- 10 mM Tris, pH 8, 0.1 mM EDTA buffer and stored at 4 °C until qPCR quantification. tion of circular DNA also supports detection using rolling circle amplification and other isothermal amplification methods that qPCR Quantification and Data Analysis. Two microliters of either exonuclease- treated ligation product (for c-PLA) or diluted preamplification product (for can be accomplished on low-cost point-of-care devices (52, 53). t-PLA) was added to qPCR master mix to a final volume of 10 μL containing We anticipate that an assay of this versatility and performance 400 nM primers and 1× Power SYBR Green PCR Master Mix (Thermo Fisher will help pave the way for much broader adoption of protein Scientific). Samples were quantified using real-time qPCR (ABI 7900 HT) with quantification into clinical and even resource-poor settings. the following thermal cycling conditions: 95 °C for 10 min (1 cycle) and 95 °C for 15 s and 60 °C for 60 s (40 cycles). Cycle threshold (Ct) values were con- Methods verted to an estimated number of ligated molecules using the formula − × + Materials. Affinity-purified polyclonal antibodies and antigen targets were 10 0.301 Ct 11.439 as previously described (13). Experiments were performed from R&D Systems. Product numbers are provided in SI Methods. All oligo- with three biological replicates that were quantified in triplicate qPCR ex- nucleotides were from Integrated DNA Technologies. Sequences are listed periments, generating nine data points per concentration analyzed. Limit of in SI Methods and were kept the same for all analytes to allow for an un- detection was derived relative to the blank sample, where the dose– biased comparison. Oligonucleotides used for proximity probe conjugation response curve and the dashed line (mean signal of a blank sample + 3SDs) were HPLC purified and designed to minimize probe–probe heteroduplex intersect (40). Statistical analysis was performed using bootstrapping (ran- formation. All other reagents were from Sigma-Aldrich unless otherwise dom sampling with replacement on the independent measurements of bi- indicated. ological replicates 1,000 times).

Proximity Probe Conjugation. Antibody–oligonucleotide conjugation was Kinetic Analysis Using SPR. SPR analysis was performed at 25 °C using a Bia- performed using hydrazone chemistry (Solulink) followed by purification core T200 instrument and a Biotin CAPture Kit (GE Healthcare) (Fig. S3A). All according to the manufacturer’s protocol (Fig. S1). One polyclonal antibody antibodies were biotinylated using the EZ-Link NHS-PEO4-Biotin kit (Thermo batch was divided into two portions and coupled to amine-modified oligo- Fisher Scientific) according to the manufacturer’s recommendation using a nucleotides containing either a free phosphorylated 5′ end or a free 3′ end. 1:1 mol ratio of biotin to antibody. Excess biotin was removed using Zeba

E932 | www.pnas.org/cgi/doi/10.1073/pnas.1718283115 Jalili et al. Downloaded by guest on September 25, 2021 spin columns to avoid interference with the binding of biotinylated anti- immobilization was tailored to a binding level of ∼3.6 nm by varying the PNAS PLUS bodies to the surface of the sensor chip. Antibody immobilization was tai- antibody loading time. All measurements were conducted at 30 °C in 200 μL lored to a level that resulted in a maximum analyte binding capacity (Rmax) total with constant agitation. Association was measured for 120 s, and dis- – of 25 50 RU/s. Analytes were diluted in 0.01 M Hepes, pH 7.4, 0.15 M NaCl, 3 sociation was measured for 1,200 s. Sensors were regenerated using a mM EDTA, 0.05% vol/vol Surfactant P20 (GE Healthcare) buffer and passed 10 mM glycine solution at pH 1.7. For each analyte, three replicates covering over the chip at a rate of 30 μL/min. Association was measured for 120 s, and full kinetics cycles were performed and resulted in identical binding curves dissociation was measured for 1,800 s. Sensors were regenerated using a across all cycles. Rate constants (ka and kd) were determined using data solution containing 6 M Guanidine-HCl and 0.25 M NaOH. Rate constants (ka analysis software, version 8.2 (Fortebio) with a 1:1 model and global fitting and kd) were determined using BIAevaluation software, version 4.1 (Biacore) with a 1:1 model and global fitting of at least five concentrations in twofold of at least five concentrations in threefold dilution series ranging from dilution series ranging from 200 to 0 nM (Fig. S4 and Table S1). Affinity 500 to 0 nM. Affinity constants (Kd) were subsequently derived from the

constants (Kd) were subsequently derived from the ratio of ka and kd. ratio of ka and kd. The results of the kinetic analysis by BLI are provided in Fig. S5 and Table S2. Kinetic Analysis Using BLI. BLI analysis was performed on a Fortebio Octet RED instrument using protein G biosensors (Fortebio) (Fig. S3B). All antibodies ACKNOWLEDGMENTS. We thank Andrew Hill, Terence Hui, Raeka Aiyar, and analytes were diluted in assay buffer (0.1 mg/mL BSA and 0.13% Triton and Wenzhong Xiao for valuable discussions. This project was supported

X-100 in 10 mM Tris, pH 7.5, 1 mM CaCl2, and 150 mM NaCl). Antibody by NIH Grant HG000205.

1. Landegren U, et al. (2012) Opportunities for sensitive plasma proteome analysis. Anal 28. Lundberg M, Eriksson A, Tran B, Assarsson E, Fredriksson S (2011) Homogeneous Chem 84:1824–1830. antibody-based proximity extension assays provide sensitive and specific detection of 2. Kingsmore SF (2006) Multiplexed protein measurement: Technologies and applica- low-abundant proteins in human blood. Nucleic Acids Res 39:e102. tions of protein and antibody arrays. Nat Rev Drug Discov 5:310–320. 29. Ericsson O, et al. (2008) A dual-tag microarray platform for high-performance nucleic 3. Wild D, ed (2013) The Handbook (Elsevier, Oxford). acid and protein analyses. Nucleic Acids Res 36:e45. 4. Zhang H, Zhao Q, Li XF, Le XC (2007) Ultrasensitive assays for proteins. Analyst (Lond) 30. Gómez de la Torre TZ, et al. (2012) Sensitive detection of spores using volume- 132:724–737. amplified magnetic nanobeads. Small 8:2174–2177. 5. Spengler M, Adler M, Niemeyer CM (2015) Highly sensitive ligand-binding assays in 31. Ke R, Nong RY, Fredriksson S, Landegren U, Nilsson M (2013) Improving precision of pre-clinical and clinical applications: Immuno-PCR and other emerging techniques. proximity ligation assay by amplified single molecule detection. PLoS One 8:e69813. Analyst (Lond) 140:6175–6194. 32. Hardenbol P, et al. (2003) Multiplexed genotyping with sequence-tagged molecular 6. Sano T, Smith CL, Cantor CR (1992) Immuno-PCR: Very sensitive antigen detection by inversion probes. Nat Biotechnol 21:673–678. means of specific antibody-DNA conjugates. Science 258:120–122. 33. Gold L, et al. (2010) Aptamer-based multiplexed proteomic technology for biomarker 7. Nam J-M, Thaxton CS, Mirkin CA (2003) Nanoparticle-based bio-bar codes for the discovery. PLoS One 5:e15004. ultrasensitive detection of proteins. Science 301:1884–1886. 34. Das S, et al. (2015) A general synthetic approach for designing epitope targeted – 8. Schweitzer B, et al. (2002) Multiplexed protein profiling on microarrays by rolling- macrocyclic peptide ligands. Angew Chem Int Ed Engl 54:13219 13224. circle amplification. Nat Biotechnol 20:359–365. 35. Tsai CT, Robinson PV, Spencer CA, Bertozzi CR (2016) Ultrasensitive antibody de- – 9. Xie S, Walton SP (2010) Development of a dual-aptamer-based multiplex protein tection by Agglutination-PCR (ADAP). ACS Cent Sci 2:139 147. biosensor. Biosens Bioelectron 25:2663–2668. 36. Albayrak C, et al. (2016) Digital quantification of proteins and mRNA in single – 10. Fredriksson S, et al. (2002) Protein detection using proximity-dependent DNA ligation mammalian cells. Mol Cell 61:914 924. assays. Nat Biotechnol 20:473–477. 37. Darmanis S, et al. (2010) Sensitive plasma protein analysis by microparticle-based – 11. Gullberg M, et al. (2004) Cytokine detection by antibody-based proximity ligation. proximity ligation assays. Mol Cell Proteomics 9:327 335. Proc Natl Acad Sci USA 101:8420–8424. 38. Rissin DM, et al. (2010) Single-molecule enzyme-linked immunosorbent assay detects serum proteins at subfemtomolar concentrations. Nat Biotechnol 28:595–599. 12. Fredriksson S, et al. (2007) Multiplexed protein detection by proximity ligation for BIOCHEMISTRY 39. Todd J, et al. (2007) Ultrasensitive flow-based immunoassays using single-molecule cancer biomarker validation. Nat Methods 4:327–329. counting. Clin Chem 53:1990–1995. 13. Fredriksson S, et al. (2008) Multiplexed proximity ligation assays to profile putative 40. Armbruster DA, Pry T (2008) Limit of blank, limit of detection and limit of quanti- plasma biomarkers relevant to pancreatic and ovarian cancer. Clin Chem 54:582–589. tation. Clin Biochem Rev 29:S49–S52. 14. Chang ST, et al. (2009) Identification of a biomarker panel using a multiplex proximity 41. Cooper MA (2002) Optical biosensors in drug discovery. Nat Rev Drug Discov 1: ligation assay improves accuracy of pancreatic cancer diagnosis. J Transl Med 7:105. 515–528. 15. Lundberg M, et al. (2011) Multiplexed homogeneous proximity ligation assays for 42. Markgren P-O, et al. (2002) Relationships between structure and interaction kinetics high-throughput protein biomarker research in serological material. Mol Cell for HIV-1 protease inhibitors. J Med Chem 45:5430–5439. Proteomics 10:M110.004978. 43. Concepcion J, et al. (2009) Label-free detection of biomolecular interactions using 16. Tate J, Ward G (2004) Interferences in immunoassay. Clin Biochem Rev 25:105–120. BioLayer interferometry for kinetic characterization. Comb Chem High Throughput 17. Kim J, Hu J, Sollie RS, Easley CJ (2010) Improvement of sensitivity and dynamic range Screen 12:791–800. in proximity ligation assays by asymmetric connector hybridization. Anal Chem 82: 44. Estep P, et al. (2013) High throughput solution-based measurement of antibody- 6976–6982. antigen affinity and epitope binning. MAbs 5:270–278. 18. Zhu L, et al. (2006) A sensitive proximity ligation assay for active PSA. Biol Chem 387: 45. Gotoh M, Hasegawa Y, Shinohara Y, Shimizu M, Tosu M (1995) A new approach to 769–772. determine the effect of mismatches on kinetic parameters in DNA hybridization using 19. Xie S, Moya C, Bilgin B, Jayaraman A, Walton SP (2009) Emerging affinity-based an optical biosensor. DNA Res 2:285–293. – techniques in proteomics. Expert Rev Proteomics 6:573 583. 46. Liebermann T, Knoll W, Sluka P, Herrmann R (2000) Complement hybridization from 20. Castro-López V, Elizalde J, Pacek M, Hijona E, Bujanda L (2014) A simple and portable solution to surface-attached probe-oligonucleotides observed by surface-plasmon- α device for the quantification of TNF- in human plasma by means of on-chip magnetic field-enhanced spectroscopy. Colloids Surf A Physicochem Eng Asp – bead-based proximity ligation assay. Biosens Bioelectron 54:499 505. 169:337–350. 21. Pai S, Ellington AD, Levy M (2005) Proximity ligation assays with peptide conjugate 47. Fan R, et al. (2008) Integrated barcode chips for rapid, multiplexed analysis of pro- ‘ ’ burrs for the sensitive detection of spores. Nucleic Acids Res 33:e162. teins in microliter quantities of blood. Nat Biotechnol 26:1373–1378. 22. Schallmeiner E, et al. (2007) Sensitive protein detection via triple-binder proximity 48. Hansen MC, Nederby L, Henriksen MOB, Hansen M, Nyvold CG (2014) Sensitive ligand- – ligation assays. Nat Methods 4:135 137. based protein quantification using immuno-PCR: A critical review of single-probe and 23. Tavoosidana G, et al. (2011) Multiple recognition assay reveals prostasomes as proximity ligation assays. Biotechniques 56:217–228. promising plasma biomarkers for prostate cancer. Proc Natl Acad Sci USA 108: 49. Larsson A, Sköldenberg E, Ericson H (2002) Serum and plasma levels of FGF-2 and 8809–8814. VEGF in healthy blood donors. Angiogenesis 5:107–110. 24. Zhang H, Li X-F, Le XC (2012) Binding-induced DNA assembly and its application to 50. Andreasson U, et al. (2015) A practical guide to immunoassay method validation. yoctomole detection of proteins. Anal Chem 84:877–884. Front Neurol 6:179. 25. Di Giusto DA, Wlassoff WA, Gooding JJ, Messerle BA, King GC (2005) Proximity ex- 51. Bradbury A, Plückthun A (2015) Reproducibility: Standardize antibodies used in re- tension of circular DNA aptamers with real-time protein detection. Nucleic Acids Res search. Nature 518:27–29. 33:e64. 52. Shen F, et al. (2011) Digital isothermal quantification of nucleic acids via simultaneous 26. Jarvius J, et al. (2006) Digital quantification using amplified single-molecule de- chemical initiation of recombinase polymerase amplification reactions on SlipChip. tection. Nat Methods 3:725–727. Anal Chem 83:3533–3540. 27. Söderberg O, et al. (2006) Direct observation of individual endogenous protein 53. Yeh E-C, et al. (2017) Self-powered integrated microfluidic point-of-care low-cost complexes in situ by proximity ligation. Nat Methods 3:995–1000. enabling (SIMPLE) chip. Sci Adv 3:e1501645.

Jalili et al. PNAS | Published online January 16, 2018 | E933 Downloaded by guest on September 25, 2021