2432

DOI 10.1002/pmic.201400215

Proteomics 2014, 14, 2432–2436

TECHNICAL BRIEF

Parallel barcoding of antibodies for DNA-assisted proteomics Mahya Dezfouli1 , Sanja Vickovic1 , Maria Jesus Iglesias2 , Jochen M. Schwenk2 and Afshin Ahmadian1 1

Science for Life Laboratory, Division of Gene Technology, School of Biotechnology, Royal Institute of Technology (KTH), Solna, Sweden 2 Science for Life Laboratory, Affinity Proteomics, Division of Proteomics and Nanobiotechnology, School of Biotechnology, Royal Institute of Technology (KTH), Solna, Sweden

DNA-assisted proteomics technologies enable ultra-sensitive measurements in multiplex format using DNA-barcoded affinity reagents. Although numerous antibodies are available, nowadays targeting nearly the complete human proteome, the majority is not accessible at the quantity, concentration, or purity recommended for most bio-conjugation protocols. Here, we introduce a magnetic bead-assisted DNA-barcoding approach, applicable for several antibodies in parallel, as well as reducing required reagents quantities up to a thousand-fold. The success of DNA-barcoding and retained functionality of antibodies were demonstrated in sandwich immunoassays and standard quantitative Immuno-PCR assays. Specific DNA-barcoding of antibodies for multiplex applications was presented on suspension bead arrays with read-out on a massively parallel sequencing platform in a procedure denoted Immuno-Sequencing. Conclusively, human plasma samples were analyzed to indicate the functionality of barcoded antibodies in intended proteomics applications.

Received: May 14, 2014 Revised: August 27, 2014 Accepted: September 24, 2014

Keywords: Antibody / Bio-conjugation / DNA-assisted proteomics / DNA-barcoding / Massively parallel sequencing / Technology



Additional supporting information may be found in the online version of this article at the publisher’s web-site

Correspondence: Dr. Afshin Ahmadian, Science for Life Laboratory, Division of Gene Technology, School of Biotechnology, Royal Institute of Technology (KTH), SE-171 65 Solna, Sweden E-mail: [email protected]

to compensate for the above-mentioned limitations and combine affinity-based proteomics with genomics toolbox is coupling affinity-reagents such as antibodies to short oligonucleotides known as DNA-barcodes [5]. The coupled DNA molecules can be used for signal amplification via PCR and multiplex target identification by means of hybridization probes or on massively parallel sequencing (MPS) platforms [6–8]. The significant improvements in read-out and strong forthcoming potentials using DNA-assisted proteomics initiated a rapid development of various methods such as quantitative immuno-PCR (qIPCR) [9], immunorolling-circle-amplification [10], proximity-ligation-assay [11], oligonucleotide-linked immuno-sorbent assay [12], and integrated-blood-barcode-chips [13]. Nevertheless, challenges in generation of a vast set of DNA-barcoded antibodies for high-throughput assay designs is considered as a key

Abbreviations: IPCR, immuno-PCR; MPS, massively parallel sequencing

Colour Online: See the article online to view Figs. 1–3 in colour.

Current affinity-based proteomics is aiming at ultra-sensitive and multiplex measurements using high-throughput technologies for analyzing many targets simultaneously [1, 2]. Although optical reporter molecules for labeling and multiplex identification of a pool of targets are available, existing labeling reagents provide a limited number of distinguishable reporters [3]. Moreover, in contrast to genomics area, there is a lack of target amplification technique in proteomics. Hence, a competent signal enhancement is required for low abundant targets, particularly in a complex sample with a broad dynamic range [4]. A promising approach

 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.proteomics-journal.com

Proteomics 2014, 14, 2432–2436

bottleneck for widespread application of these approaches [14]. In addition to being lab-intensive if done in a large scale, an apparent limitation of current DNA-antibody bio-conjugation techniques is the requirement of purified antibody and DNA in relative large amounts (nmoles) (Solulink, Antibody-oligonucleotide all-in-one conjugation kit, available at: http://store.solulink.com/collections/antibody-oligonu cleotide-products (accessed May 5, 2014); Pierce Protein Biology Products, sulfo-SMCC, available at: http://www. piercenet.com/product/sulfo-smcc (accessed May 5, 2014). To date, a vast antibody repertoire is available such as the Human Protein Atlas (HPA), with ca. 22 000 validated antibodies that cover over 82% of the human protein coding genes [15]. However, the requirement for large antibody quantities in current bio-conjugation procedures results in limiting the DNA-assisted proteomics to a restricted antibody collection, which are available in sufficient amounts, high concentration, and desired purity. This fact dramatically lowers the number of targets that can be further analyzed. On the other hand, applying large quantities of DNA molecules might lead to undesirable cross-contamination between samples and could cause increased background signals due to amplification of remaining free DNA [16], which ultimately leads to sensitivity and multiplicity failures. Therefore, performing bio-conjugation at low reagent quantities in a high-throughput and automated format would enhance the possibilities to apply DNA-assisted proteomics on a considerably larger target set and increase the opportunities for development of novel promising technologies. Recently, we have introduced a bead-assisted labeling procedure and showed efficient biotinylation and dye-labeling of antibodies at very low quantities (pmoles) [17]. By performing the buffer-exchange, purification, and concentration of antibodies in a single step on solid-support, we avoid repeated antibody loss; hence the protocol enables a feasible decrease in the required starting material. Moreover, chemical reactions done on solid-phase show an efficiency enhancement by more effective diffusion and increased local concentration of immobilized molecules, therefore allowing for the use of lower reagent amounts [18]. Here, based on the bead-assisted labeling concept, we present covalent DNAantibody bio-conjugation at low reagent amounts. This work presents up to a thousand-fold reduction in reagents quantities compared to the minimum requirements of in-solution coupling procedures. In addition, the process is introduced as a semi-automated system in 96-well plate format enabling high-throughput antibody barcoding, important for multiplex assay designs. The bead-assisted labeling system is applied in two subsequent phases, i.e. to modify antibodies with a chemical linker and to perform the DNA-barcoding reaction on modified antibodies (Fig. 1). As previously noted, DNA-assisted proteomics methods demand thorough removal of excess DNA-barcodes prior the detection step. Therefore, an efficient purification of magnetic beads from excess of DNA molecules is fundamental. To optimize the second phase of  C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

2433

Figure 1. Schematic view of the bead-assisted DNA-antibody bio-conjugation. (A) Antibodies are mixed and incubated with protein-A coated magnetic-beads and (B) are purified from unbound contaminants/supplements. (C) The cross-linker molecules are added and react via amine-reactive N-hydroxysuccinimide-ester-groups to antibodies as well as the protein-A. (D) After removal of excess cross-linkers, functionalized antibodies are recaptured on fresh beads. The thiol-modified oligonucleotides are added and the reaction proceeds overnight. (E) The beads are thoroughly washed from excess DNA molecules and (F) the purified barcoded antibodies are eluted by lowering the pH, followed by pH adjustment and storage at 4⬚C. All steps are performed automated on a robotic workstation.

the procedure, the microspheres were incubated with DNAbarcodes overnight and washing was performed using five different candidate buffers with variable salt and detergent compositions (Supporting Information Methods 1.1). The washing supernatants were subsequently tested for the presence and quantity of remaining DNA on a quantitative PCR thermocycler (qPCR), where PBS-T (PBS pH 7.4, 0.1% Tween 20) was selected as the most efficient (Supporting Information Fig. 1). Using the optimized system, one microgram (6 pmoles) of polyclonal rabbit IgG antibody was coupled to a 50 bp thiol-modified oligonucleotide (20 pmoles) using a hetero-bi-functional amine-to-sulfhydryl cross-linker (Supporting Information Methods 1.2). The bio-conjugates were investigated on qPCR for carrying DNA and the analysis showed an approximate ratio of 1.2 estimated DNA molecules over antibody, comparable to common in-solution approaches using the recommended high reagent amounts. Subsequently, 18 HPA antibodies (Supporting Information Tables 1) were barcoded with specific DNA molecules (Supporting Information Table 2) and quality-assured for www.proteomics-journal.com

2434

M. Dezfouli et al.

Proteomics 2014, 14, 2432–2436

Figure 2. Quality-assurance of bio-conjugates and confirmation of retained binding functionality. (A) qPCR analysis of the bio-conjugates eluted from solid-phase (two subsequent eluates) as well as the first supernatants before purification and last washing samples are shown. Results demonstrate significantly higher DNA-amplicon concentration in barcoded antibody samples compared to respective negative controls (i.e. similar bio-conjugation procedure with absence of cross-linker molecules). In addition, no detectable DNA is observed in the last washing samples before elution. Blank experiment is analysis on pure water. (B) Results from qIPCR assays on a dilution series of target antigens, using ten different barcoded antibodies, illustrate that DNA-amplicon concentrations correlate (R2 = 0.93) with antigen quantities. (C) Schematic workflow of Immuno-sequecing (I-Seq). Step-1: Each detection antibody is barcoded with a unique DNA-sequence and a homogeneous pool of barcoded antibodies is used to target a number of antigens in multiplex. Unbound antibodies (excess bio-conjugates or antibodies with no respective target) are washed and Step-2: a library is prepared for sequencing, using universal adaptors. Step-3: The DNA-barcodes are sequenced on a MPS platform to be decoded and the reads-per-barcode are counted to quantify target molecules. (D) The summary of I-Seq data from all antibodies is illustrated as box-plot. Variable target amounts that address high (H), medium (M) and low (L) quantities are used and data shows the correlation of signals to the antigen amounts used in the assays. For detailed values refer to Supporting Information Table 4 and Supporting Information Fig. 4.

successful bio-conjugation and purification from excess DNA molecules (Supporting Information Methods 1.3). The results demonstrated bio-conjugation of all antibodies by significantly (Student’s t-test, p < 0.01) higher signals from bio-conjugates (1024-folds) than the respective negative controls (i.e. performing bio-conjugation on nonmodified antibodies); also confirming no detectable excess of DNA in washing supernatants after final purification (Fig. 2A). With the purpose of demonstrating retained functionality of barcoded antibodies for binding their specific targets, bio-conjugates were applied in simplex qIPCR assays using protein fragments of target epitopes bound to magnetic beads (Supporting Information Methods 1.4 and Supporting Information Fig. 2). Results showed significantly (p < 0.01) higher amplicon concentration of the test assays compared to the negative controls (i.e. noncoated beads). In addition, signals

 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

from a dilution series of ten different target antigens (decreasing number of coated beads) were analyzed and showed proportional amplicon concentrations (Pearson correlation, R2 = 0.93) to the amount of antigen used in the assays (Fig. 2B and Supporting Information Fig. 3). This confirms the retained binding functionality of antibodies following the bioconjugation process. To investigate the possibility of unambiguous target identification in multiplex and possibility of read-out via parallel sequencing of the DNA-barcodes, a model system on suspension bead arrays was applied (Supporting Information Methods 1.4–1.5). Variable amounts of target antigens were simulated by loading different numbers of antigen-coated beads to the wells to test high, medium, and low antigen quantities in different combinations and in replicates (Supporting Information Table 3). Each experiment was specified

www.proteomics-journal.com

Proteomics 2014, 14, 2432–2436

2435

Figure 3. Evaluation of barcoded antibodies in sandwich assays and for protein detection in human blood plasma. (A–C) Dilution series of target antigens were analysed using antibody pairs for ANG, CCL16 and MMP-9 by sandwich immunoassays. The signals obtained from on-bead detection (x-axis; Luminex) were compared to data generated by qIPCR (y-axis). The correlation between the two detection systems confirms the DNA-based read-out. (D–F) The barcoded sandwich assays were further applied to detect three target proteins in a dilution series of normal human plasma samples. The relation of the signal to plasma proportion is shown and the dynamic ranges are indicated (red line) by using a linear model fitted for each antigen between upper saturation and lower background levels.

by a distinct index for parallel sequencing on a bench-top MPS platform in a process we called Immuno-Sequencing (Fig. 2C). The sequencing resulted in a total data of over 585 Mbp (an average of 145 235 reads/experiment). The data showed significantly (p < 0.01) elevated read-counts from bioconjugates targeting the present antigens compared to the antibodies with no respective targets used as internal-controls. In addition, the sequencing data showed increased values, when higher antigen quantities were present with a correlation (R2 = 0.97) to the antigen amount (Fig. 2D and Supporting Information Table 4 and Supporting Information Fig. 4).

 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Conclusively, the possibility of applying the bio-conjugates in sandwich assays using complex plasma samples was assessed (Supporting Information Methods 1.6). Three antibody pairs, targeting human proteins found in blood plasma (ANG, CCL-16, and MMP-9) were used (Supporting Information Table 5). The capture antibodies were coupled to magnetic beads and aliquots of detection antibodies were DNA-barcoded for read-out through qIPCR. First, the sandwich immunoassay was performed using recombinant full-length proteins as targets, with parallel read-out on Luminex instrument as well as on qPCR. A side-by-side

www.proteomics-journal.com

2436

M. Dezfouli et al.

comparison showed a supportive confirmation between the two read-out systems, both showing the antigen concentration dependent detection (Fig. 3A–C). Next, linearity of the signals were assessed via a dilution series of normal human plasma (Fig. 3D–F) by assuming a signal linear relation between upper saturation and lower background levels (average R2 = 0.97 for linear range; average CV = 3.7% over all data points). Data showed a distinct range of linearity for each target antigen. Lastly, to show the applicability of bio-conjugates in a complex sample context, a standard spike-and-recovery assay was performed in PBS buffer (supplemented with 1% BSA) versus plasma. The assay revealed an average recovery of 113% in plasma compared to controls (Supporting Information Table 6). Taken together, this work demonstrates the applicability of its bio-conjugation technique in high-throughput antibody barcoding with minimal reagent requirements. It shows an accurate multiplex read-out through DNA-sequencing as a promising additional method from genomics toolbox. The achieved low background signals in qPCR (comparable to pure water) were ensured by efficient purification of bioconjugates from excess DNA-barcodes, in order to allow for improving sensitivity and precision of follow-up detection systems. In conclusion, the growing number of specific antibodies that are being produced in large-scale projects and the developments within the genomics toolbox toward higher throughput, lower cost, more user-friendly bench-top devices pave the way for an integration of the presented work into promising technologies of high-throughput and ultra-sensitive affinityproteomics. This work was supported by ProNova (VINN Excellence Centre for Protein Technology) and VinnCARDIO from VINNOVA Swedish Governmental Agency for Innovation Systems, the Knut and Alice Wallenberg Foundation and grants from Science for Life Laboratory. The authors would like to thank the staff of the Human Protein Atlas for material contribution and gratefully acknowledge Eni Andersson and Cecilia Mattsson from Biobank Profiling at Science for Life Laboratory for their assistance in this work. The authors have declared no conflict of interest.

References [1] Chandra, H., Reddy, P. J., Srivastava, S., Protein microarrays and novel detection platforms. Expert. Rev. Proteomic 2011, 8, 61–79. [2] Landegren, U., Vanelid, J., Hammond, M., Nong, R. Y. et al., Opportunities for sensitive plasma proteome analysis. Anal. Chem. 2012, 84, 1824–1830. [3] Flanigon, J., Kamali-Moghaddam, M., Burbulis, I., Annink, C. et al., Multiplex protein detection with DNA readout via mass spectrometry. New Biotechnol. 2013, 30, 153–158.

 C 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Proteomics 2014, 14, 2432–2436 [4] Spisak, S., Guttman, A., Biomedical applications of protein microarrays. Curr. Med. Chem. 2009, 16, 2806–2815. [5] Sano, T., Smith, C. L., Cantor, C. R., Immuno-PCR—very sensitive antigen-detection by means of specific antibody-DNA conjugates. Science 1992, 258, 120–122. [6] Hindson, B. J., Ness, K. D., Masquelier, D. A., Belgrader, P. et al., High-throughput droplet digital PCR system for absolute quantitation of DNA copy number. Anal. Chem. 2011, 83, 8604–8610. [7] Ullal, A. V., Peterson, V., Agasti, S. S., Tuang, S. et al., Cancer cell profiling by barcoding allows multiplexed protein analysis in fine-needle aspirates. Sci. Transl. Med. 2014, 6, 219ra219. [8] Darmanis, S., Nong, R. Y., Vanelid, J., Siegbahn, A. et al., ProteinSeq: high-performance proteomic analyses by proximity ligation and next generation sequencing. PloS one 2011, 6, e25583. [9] Niemeyer, C. M., Adler, M., Wacker, R., Detecting antigens by quantitative immuno-PCR. Nat. Protocols 2007, 2, 1918– 1930. [10] Akter, F., Mie, M., Kobatake, E., Immuno-rolling circle amplification using a multibinding fusion protein. Anal. Biochem. 2011, 416, 174–179. [11] Soderberg, O., Gullberg, M., Jarvius, M., Ridderstrale, K. et al., Direct observation of individual endogenous protein complexes in situ by proximity ligation. Nat. Methods 2006, 3, 995–1000. [12] Han, K. C., Ahn, D. R., Yang, E. G., An approach to multiplexing an immunosorbent assay with antibodyoligonucleotide conjugates. Bioconjugate Chem. 2010, 21, 2190–2196. [13] Fan, R., Vermesh, O., Srivastava, A., Yen, B. K. et al., Integrated barcode chips for rapid, multiplexed analysis of proteins in microliter quantities of blood. Nat. Biotechnol. 2008, 26, 1373–1378. [14] Kozlov, I. A., Melnyk, P. C., Stromsborg, K. E., Chee, M. S. et al., Efficient strategies for the conjugation of oligonucleotides to antibodies enabling highly sensitive protein detection. Biopolymers 2004, 73, 621–630. [15] Uhlens, M., Bjorling, E., Agaton, C., Szigyarto, C. A. et al., A human protein atlas for normal and cancer tissues based on antibody proteomics. Mol. Cell. Proteomics 2005, 4, 1920–1932. [16] Hajibabaei, M., DeWaard, J. R., Ivanova, N. V., Ratnasingham, S. et al., Critical factors for assembling a high volume of DNA barcodes. Philos. Trans. Roy. Soc. B 2005, 360, 1959–1967. [17] Dezfouli, M., Vickovic, S., Iglesias, M. J., Nilsson, P. et al., Magnetic bead assisted labeling of antibodies at nanogram scale. Proteomics 2014, 14, 14–18. [18] Sista, R. S., Eckhardt, A. E., Srinivasan, V., Pollack, M. G. et al., Heterogeneous immunoassays using magnetic beads on a digital microfluidic platform. Lab Chip 2008, 8, 2188–2196.

www.proteomics-journal.com

Parallel barcoding of antibodies for DNA-assisted proteomics.

DNA-assisted proteomics technologies enable ultra-sensitive measurements in multiplex format using DNA-barcoded affinity reagents. Although numerous a...
564KB Sizes 0 Downloads 5 Views