Anal. Chem. 2010, 82, 9138–9140
Letters to Analytical Chemistry Sensitive Detection of Bacterial DNA by Magnetic Nanoparticles Jenny Go¨ransson,†,‡,§ Teresa Zarda´n Go´mez De La Torre,† Mattias Stro¨mberg,† Camilla Russell,‡ Peter Svedlindh,† Maria Strømme,*,† and Mats Nilsson*,‡ Ångstro¨m Laboratory, Department of Engineering Sciences, Uppsala University, Box 534, SE-751 21 Uppsala, Sweden, and Rudbeck Laboratory, Department of Genetics and Pathology, Uppsala University, SE-751 85 Uppsala, Sweden This work presents sensitive detection of bacterial genomic DNA using a magnetic nanoparticle-based substratefree method. For the first time, such a method is employed for detection of a clinically relevant analyte by implementing a solid-phase-based molecular probing and amplification protocol that can be executed in 80 min. The molecular detection and amplification protocol is presented and verified on samples containing purified genomic DNA from Escherichia coli cells, showing that as few as 50 bacteria can be detected. This study moves the use of volume-amplified magnetic nanoparticles one step further toward rapid, sensitive, and selective infectious diagnostics. Infectious diagnosis of diseases is generally carried out by cultivation of bacteria, which is a tedious and time-consuming process. Fungal and bacterial pathogens are cultivated and further typed by biochemical testing or more advanced molecular methods like the polymerase chain reaction (PCR), which can also be used for viral typing.1-5 The analysis time of cultivation strategies is very long, and the response back to the patient is, therefore, much delayed. PCR is a quick and sensitive method but demands a lot from the user and is sensitive to contaminations.6 Furthermore, PCR is expensive and, therefore, restrictively used. Inexpensive methods for diagnosis are of great importance to clinics with constrained budgets. Assays with read-out techniques * To whom correspondence should be addressed. E-mail: maria.stromme@ angstrom.uu.se (M. Strømme);
[email protected] (M. Nilsson). † Department of Engineering Sciences. ‡ Department of Genetics and Pathology. § Current address: Q-linea AB, Dag Hammarskjo ¨lds va¨g 58A, SE-751 83 Uppsala, Sweden. (1) Saiki, R. K.; Scharf, S.; Faloona, F.; Mullis, K. B.; Horn, G. T.; Erlich, H. A.; Arnheim, N. Science 1985, 230, 1350–1354. (2) Cho, S. N.; Brennan, P. J. Tuberculosis (Edinb) 2007, 87 (Suppl 1), S14– 17. (3) Fenn, J. P.; Segal, H.; Barland, B.; Denton, D.; Whisenant, J.; Chun, H.; Christofferson, K.; Hamilton, L.; Carroll, K. J. Clin. Microbiol. 1994, 32, 1184–1187. (4) Haynes, K. A.; Westerneng, T. J. J. Med. Microbiol. 1996, 44, 390–396. (5) Mahony, J. B. Clin. Microbiol. Rev. 2008, 21, 716–747. (6) Lo, Y. M.; Chan, K. C. Methods Mol. Biol. 2006, 336, 11–18.
9138
Analytical Chemistry, Vol. 82, No. 22, November 15, 2010
based on optics (like quantitative PCR7) usually require advanced and expensive equipment. An alternative to optical read-out is based on magnetic field measurements. Magnetic biosensors have gained ground in recent years, because of the high physical and chemical stability and the potentially low-cost of production of magnetic particles, also denoted magnetic nanobeads. The Brownian relaxation principle constitutes a substrate-free biosensor method, where suspended magnetic nanobeads exhibiting Brownian relaxation behavior are equipped with probes for recognition of specific target molecules.8,9 Binding of target molecules to the probes causes a hydrodynamic size increase of the magnetic nanobeads. This brings on a decrease of the Brownian relaxation frequency, defined by the position of the peak in the imaginary part (m′′) of the complex magnetization (m ) m′ - im′′) spectrum, of the beads since this frequency is inversely proportional to the hydrodynamic volume of the beads.10 A frequency shift, thus, confirms the presence of target molecules. By representing the target molecules as micrometersized DNA coils, the relaxation frequency of the magnetic nanobeads is drastically decreased. The concentration of target molecules can then be monitored as a corresponding decrease of the amplitude of the Brownian relaxation peak of free beads. The volume-amplified magnetic nanobead detection assay (VAM-NDA) uses this strategy.10 The DNA coils are produced in a series of reactions, starting with target recognition using the padlock probe technology.11 Padlock probes are oligonucleotides of about 70-90 nucleotides, whose ends have been designed to hybridize head-to-tail on a target sequence (Figure 1I). The ends of a correctly matched probe can be joined by the action of a DNA ligase and are thereby transformed into a DNA circle in a highly specific way.12 Reacted probes constitute suitable templates for rolling circle amplification (RCA), which is a method for linear polymerization (7) Ishiguro, T.; Saitoh, J.; Yawata, H.; Yamagishi, H.; Iwasaki, S.; Mitoma, Y. Anal. Biochem. 1995, 229, 207–213. (8) Connolly, J.; St Pierre, T. G. J. Magn. Magn. Mater. 2001, 225, 156–160. (9) Astalan, A. P.; Ahrentorp, F.; Johansson, C.; Larsson, K.; Krozer, A. Biosens. Bioelectron. 2004, 19, 945–951. (10) Stro ¨mberg, M.; Go ¨ransson, J.; Gunnarsson, K.; Nilsson, M.; Svedlindh, P.; Strømme, M. Nano Lett. 2008, 8, 816–821. (11) Nilsson, M.; Malmgren, H.; Samiotaki, M.; Kwiatkowski, M.; Chowdhary, B. P.; Landegren, U. Science 1994, 265, 2085–2088. 10.1021/ac102133e 2010 American Chemical Society Published on Web 10/22/2010
Figure 1. Overview of the molecular reactions, as well as magnetic labeling and read-out used in the assay. The detailed molecular procedure (I-Vb) and magnetic read-out (VI-VII) are described in the text.
creating long single-stranded products that collapse into micrometersized coils of DNA.13-16 Earlier studies describing the VAM-NDA method have mainly served to characterize features of the assay, (12) Conze, T.; Shetye, A.; Tanaka, Y.; Gu, J. J.; Larsson, C.; Göransson, J.; Tavoosidana, G.; Söderberg, O.; Nilsson, M.; Landegren, U. Annu. Rev. Anal. Chem. 2009, 2, 215–239. (13) Liu, D. Y.; Daubendiek, S. L.; Zillman, M. A.; Ryan, K.; Kool, E. T. J. Am. Chem. Soc. 1996, 118, 1587–1594.
such as the specificity in the signal generation, as well as to optimize the performance of the assay.10,17-19 In this study, the VAM-NDA has for the first time been applied for detection of a clinically relevant analyte (bacterial genomic DNA) by implement(14) Blab, G. A.; Schmidt, T.; Nilsson, M. Anal. Chem. 2004, 76, 495–498. (15) Lizardi, P. M.; Huang, X. H.; Zhu, Z. R.; Bray-Ward, P.; Thomas, D. C.; Ward, D. C. Nat. Genet. 1998, 19, 225–232. (16) Fire, A.; Xu, S. Q. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 4641–4645.
Analytical Chemistry, Vol. 82, No. 22, November 15, 2010
9139
ing a solid-phase-based molecular probing and amplification protocol suitable for microfluidic integration. A sample containing purified genomic DNA from Escherichia coli cells is initially fragmented and denatured by a short incubation at high temperature. Thereafter, the sample is subjected to the molecular detection and amplification protocol outlined in Figure 1I-V. (Panel I) The sample is mixed with padlock and capture probes, which hybridize the target DNA, and a ligase enables correctly matched padlock probes to be circularized. (Panel II) The biotinylated capture probes are, thereafter, allowed to couple to streptavidin-coated Dynabead particles, and nonbound excess padlock probes are washed away. (Panel III) The DNA circles are amplified with RCA, and (Panel IV) the concatemer DNA coils are then digested to monomers. Replication oligonucleotides are hybridized to the single-stranded DNA coils, and a restriction endonuclease enzyme cuts at a defined sequence motif. (Panel V) The monomers are ligated and amplified by RCA to generate a new set of DNA coils. The cycle (Panels IV-V) is thereafter repeated but without capture probes and Dynabead particles. This way, the number of DNA coils is amplified. The strategy for doing this is denoted as circle-to-circle amplification (C2CA)20 and enhances the sensitivity of the assay. The procedure can be performed multiple times until a satisfying number of DNA coils have been formed. In this study, we chose to perform two cycles of C2CA (see Supporting Methods in the Supporting Information). The presence of the DNA coils is monitored by the hybridization of the oligonucleotide-functionalized magnetic nanobeads to the DNA coils (Panel VI), drastically increasing the hydrodynamic volume of the beads. In a superconducting quantum interference device (SQUID), this is detected as a large downward shift of the relaxation frequency of magnetic nanobeads hybridized to DNA coils (Panel VII and Figure S1 in the Supporting Information). As a proof-of-concept that the VAM-NDA approach can be used for a real clinical application, we used the described procedure to detect E. coli genomic DNA. A semiquantitative response was achieved with a limit of detection of at least 50 bacteria (Figure 2), applying a molecular probing and amplification procedure that sum up to about 80 min. This limit of detection is relevant for diagnosis of several infectious diseases such as urinary tract infection, diarrhea, and respiratory tract infections.21 Quantitative PCR (Q-PCR) is the conventional molecular technique for analysis of pathogen DNA. This method is, however, sensitive to contaminations and requires advanced and expensive equipment as well (17) Zarda´n Go´mez de la Torre, T.; Stro ¨mberg, M.; Russell, C.; Go ¨ransson, J.; Nilsson, M.; Svedlindh, P.; Strømme, M. J. Phys. Chem. B 2010, 114, 3707– 3713. (18) Stro ¨mberg, M.; Zarda´n Go´mez de la Torre, T.; Go ¨ransson, J.; Gunnarsson, K.; Nilsson, M.; Strømme, M.; Svedlindh, P. Biosens. Bioelectron. 2008, 24, 696–703. (19) Stro ¨mberg, M.; Zarda´n Go´mez de la Torre, T.; Go ¨ransson, J.; Gunnarsson, K.; Nilsson, M.; Svedlindh, P.; Strømme, M. Anal. Chem. 2009, 81, 3398– 3406. (20) Dahl, F.; Baner, J.; Gullberg, M.; Mendel-Hartvig, M.; Landegren, U.; Nilsson, M. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 4548–4553. (21) Todd, E. C. D.; Greig, J. D.; Bartleson, C. A.; Michaels, B. S. J. Food Prot. 2008, 71, 2582–2595. (22) Dalslet, B. T.; Damsgaard, C. D.; Donolato, M.; Strømme, M.; Stro¨mberg, M.; Svedlindh, P.; Hansen, M. F. Lab Chip 2010, DOI: 10.1039/c0lc00002g.
9140
Analytical Chemistry, Vol. 82, No. 22, November 15, 2010
Figure 2. Detection of genomic E. coli DNA diluted in steps of 1:10. The DNA was diluted and processed according to the description in Figure 1. The plot shows the estimated number of E. coli bacteria and the corresponding amplitude of two separate measurements of the free-bead relaxation peak at 37 °C of two independent samples (cf. Figure 1VII).
as trained personnel to interpret data. The VAM-NDA has a large potential to become automated. In contrast to Q-PCR, the assay is iso-thermal and requires no optical equipment. The current VAM-NDA setup includes the use of a SQUID for measuring the magnetic properties. It would be strongly advantageous if the SQUID could be replaced with a miniaturized on-chip magnetic field sensor capable of measuring the frequency-dependent magnetization of the magnetic nanobeads. We have recently demonstrated that AC susceptibility measurements can be conducted on a chip using the same type of magnetic beads as studied here.22 In this work, we showed that the AC susceptibility can be measured using on-chip magnetoresistive sensors operating at ambient conditions and have specifically demonstrated the feasibility of high sensitivity AC susceptibility measurements using the planar Hall effect (PHE) sensor chip. The use of such a labon-a-chip unit including a miniaturized magnetic field sensor with a size matching that of the investigated nanobead ensemble is expected to improve the sensitivity of the VAM-NDA assay even further. Finally, integration of the molecular reactions to a microfluidic format in the chip would enhance the degree of automation and reproducibility, which are parameters important to clinical applications. ACKNOWLEDGMENT The Swedish Research Council (VR) and the Knut and Alice Wallenberg Foundation (KAW) are acknowledged for their financial support. Maria Strømme and Mats Nilsson contributed equally to this work. Genomic DNA was kindly provided by Anna Zorzet at Department of Medical Biochemistry and Microbiology, Uppsala University. SUPPORTING INFORMATION AVAILABLE Supporting methods, supporting sequences, supporting table, supporting theory, Figure S1, and supporting references. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review August 13, 2010. Accepted October 14, 2010. AC102133E
