Anal Bioanal Chem (2014) 406:3315–3322 DOI 10.1007/s00216-013-7597-3
RESEARCH PAPER
Nanoparticle-encapsulated vis- and NIR-emissive fluorophores with different fluorescence decay kinetics for lifetime multiplexing Katrin Hoffmann & Thomas Behnke & Markus Grabolle & Ute Resch-Genger
Received: 7 August 2013 / Revised: 16 December 2013 / Accepted: 21 December 2013 / Published online: 16 January 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Bioanalytical, clinical, and security applications increasingly require simple, efficient, and versatile strategies to measure an ever increasing number of analytes or events in parallel in a broad variety of detection formats as well as in conjunction with chromatographic separation techniques or flow cytometry. An attractive alternative to common optical multiplexing and encoding methods utilizing spectral multiplexing/color encoding and intensity encoding is lifetime multiplexing, which relies on the discrimination between different fluorescent reporters based on their fluorescence decay kinetics. Here, we propose a platform of surfacefunctionalizable polymeric nanoparticles stained with fluorophores differing in their fluorescence lifetimes as a new multiplexing and encoding approach. Proof-of-concept measurements with different sets of lifetime-encoded polystyrene nanoparticles are presented, obtained via staining of preformed particles with visible (vis)- and near-infrared (NIR)-emissive organic dyes, which display very similar absorption and emission spectra to enable excitation and detection at the same wavelengths, yet sufficiently different fluorescence decay kinetics in suspension, thereby minimizing instrumentation costs. Data analysis was performed with a linear combination approach in the lifetime domain. Our results and first cell experiments with these reporter sets underline the suitability of our multiplexing strategy for the discrimination between and the quantification of different labels. This simple and versatile concept can be extended to all types of fluorophores, thereby expanding the accessible time scale, and can be used, e.g., for the design of labels and Published in the topical collection Multiplex Platforms in Diagnostics and Bioanalytics with guest editors Günter Peine and Günther Proll. K. Hoffmann : T. Behnke : M. Grabolle : U. Resch-Genger (*) BAM Federal Institute for Materials Research and Testing, Richard-Willstaetter-Str. 11, 12489 Berlin, Germany e-mail: [email protected]
targeted probes for fluorescence assays and molecular imaging, cellular imaging studies, and barcoding applications, also in conjunction with spectral and intensity encoding. Keywords Fluorescent label . Multiplexing . Optical encoding . Lifetime multiplexing . Fluorescence lifetime imaging FLIM . Nanoparticles
Introduction The simultaneous measurement of multiple analytes or events in parallel, termed multiplexing, has become increasingly important for the growing demands of molecular biology, medical diagnostics, and drug discovery as well as current security concerns [1–4]. The detection, monitoring, and quantification of multiple targets in a single measurement can reduce, e.g., the amount of reagents, consumables, and sample as well as the time of analysis and can decrease sampling errors. Often, this is the only way to solve the complex problems imposed on analytical techniques by modern medicine. At present, planar microarrays and bead-based assays, which are often read out with a flow cytometer or a fluorescence microscope, present the most striking examples for high-throughput, multiplexed assays, used for, e.g., DNA sequencing, biomarker analysis, disease diagnosis, and drug discovery [5, 6]. To enable the readout of a maximum number of measurable species, most multiplexing applications combine local resolution, provided by the spatial position on an array or in a microfluidic channel or by separation in a flow or on a gel, with detection methods. Amongst the most popular readout methods are fluorescence techniques due to their comparable ease of use, high sensitivity, relatively inexpensive instrumentation, and high information content providing several analyte-specific quantities simultaneously [7–9]. At
3316
the core of all fluorescence-based methods are fluorescent reporters, used as labels, sensors, or encoding reagents [10, 11]. Typically, fluorescence-based multiplexing schemes rely on spectral multiplexing (color or wavelength encoding mode), i.e., the discrimination between different emission colors, utilizing dyes that can be excited at the same wavelength with a single excitation light source and distinguished by their different emission spectra. This requires either different detectors, one for each integrally measured emission wavelength region, or a spectrally resolving detection system [5]. Suitable candidates for this application are semiconductor quantum dots with their very broad absorption spectra and their narrow emission spectra [10, 11] or lanthanide labels with their very narrow emission bands and large Stokes shifts, which, however, require typically distinct excitation wavelengths >400 nm [12]. In any case, the number of spectrally distinct codes is limited by spectral crosstalk in excitation and/ or emission and energy transfer between fluorophores with overlapping absorption and emission spectra. Bead-based multiplexed analysis like the suspension microparticle assays from Luminex and the microparticle platform from Illumina for DNA sequencing rely on single or typically, on dual wavelength intensity pattern-encoded particles with different surface chemistries, i.e., different ligands [13]. The encoded beads are commonly fabricated by incorporation of different dyes with distinguishable ratios with precise concentration control, rendering their preparation challenging and requiring comparable photostabilities of all encoding reagents. An attractive option can present fluorescence lifetime multiplexing, i.e., the discrimination between fluorophores based on their different fluorescence decay kinetics that are insensitive to variations in excitation light intensity and dye concentration, in contrast to (relative) fluorescence intensities [7, 14, 15]. Hence, fluorescence lifetime measurements are increasingly being utilized for imaging and sensing applications and recently also for multiplexed detection and barcoding [7, 16, 17]. Lifetime multiplexing and encoding require fluorescent reporters with sufficiently different lifetimes, which should be excitable at the same wavelength and detectable within the same spectral window to minimize instrumentation costs and as prerequisite for the eventually desired combination with color multiplexing/encoding to increase the number of distinguishable codes. Up to now, the vast majority of lifetime-based sensing and multiplexing schemes exploit molecular or nanocrystalline fluorophores, the fluorescence lifetimes and decay kinetics of which are sensitive to their microenvironment and in the case of the latter, also to their surface chemistry/ligand shell. Although mainly molecular fluorophores are the reporter of choice for lifetime-based sensing and discrimination, this environment sensitivity can hamper many other multiplexing and barcoding applications, where environment-insensitive codes are desired.
K. Hoffmann et al.
The need for efficient multiplexing and encoding strategies motivated us to develop a new approach to lifetime multiplexing using different fluorescent reporters including organic dyes and quantum dots and a new concept for data analysis pursuing a simple type of “pattern matching” in the time domain [7]. This decomposition procedure is independent of the actual shape of the decay curves, thus enabling also the analysis of multi-exponential and nonexponential decays. To assess the versatility of this concept, we studied the suitability of this approach for different sets of fluorescent nanoparticles obtained by staining of preformed carboxylated polystyrene (PS) nanoparticles (NP, PSNP) with organic dyes absorbing and emitting in the visible (vis) and near infrared (NIR), which can be only distinguished by their different fluorescence lifetimes. Dye encapsulation elegantly circumvents possible interactions between dyes and their surrounding matrix, thereby rendering the decay kinetics of each particle independent of its location and should enable the prediction of label decay kinetics from time-resolved fluorescence measurements with the reporters in suspension. Moreover, it allows the use of hydrophobic dyes [18, 19] and the design of labels containing dye combinations [20], and can enhance dye brightness and stability [21–25], which is especially beneficial for NIR dyes. Also, it can reduce toxicity concerns, e.g., for cellular imaging [26]. More important, it enables the subsequent covalent functionalization of these particles with, e.g., different target-specific ligands like antibodies, peptides, or oligonucleotides using always the same and established bioconjugation protocols [20, 27]. Hence, such an encapsulation strategy can be very advantageous for the design of an easily fabricated platform of fluorescent reporters for lifetimebased fluorescence imaging or lifetime encoding. Here, we present a proof-of-concept study using two sets of lifetime-encoded fluorescent particle reporters, one for the vis and one for the NIR region. In this respect, we studied the potential of fluorescence lifetime-based analysis using our linear combination approach previously validated with organic dyes and quantum dots to distinguish between differently stained PSNP in suspension as well as in fibroblast cells incubated with different mixtures of these particles employing fluorescence lifetime imaging microscopy (FLIM). Our intention was the development of a simple and versatile multiplexing approach exploiting the analyte-specific fluorescence decay behavior and which can be applied to complex systems like biological systems. Our aim here was to demonstrate the feasibility of a new approach for distinguishing samples containing different fractions of lifetime-encoded nanoparticles. The comparably simple linear combination fitting procedure of the resulting multicomponent fluorescence decay curves has proven to be robust and suitable for use even in complex systems like living cells. These measurements present the first step towards a new platform of lifetime-encoded particles with different surface chemistries like nanoparticles equipped with
Nanoparticle-based lifetime multiplexing in living cells
different ligands like monoclonal antibodies binding different targets for multiplexed biomarker analysis in cellular studies, which can be also easily expanded utilizing fluorophores with longer lifetimes.
Materials and methods Materials Carboxyl-functionalized polystyrene nanoparticles with diameters of 25 and 100 nm were purchased from Kisker Biotech. The vis dyes Nile Red (7-diethylamino-3,4benzophenoxazin-2-on; Lambda Physics) and Lumogen F305 (BASF) [28] and the NIR dyes Itrybe [29] (Otava Chemicals) and Squaraine Sq730 (SETA BioMedicals) [30] were encapsulated in the PSNP without further purification. The PSNP were sonicated and stained with these dyes using a previously described staining procedure [18, 19, 31]. Tetrahydrofuran (THF) was of UV-spectroscopic grade and purchased from Sigma-Aldrich. Cell culture medium Dulbecco’s Modified Eagle Medium (DMEM), fetal calf serum (FCS), and phosphate-buffered saline (PBS) were purchased from Biochrom AG. Ste ady-state fluores cence spe ctrosc opy and microscopy Absorption spectra were recorded on a Cary 5000 spectrometer (Varian). Fluorescence emission spectra of Itrybe and Sq730 were measured with a previously described spectrofluorometer 8100 (Spectronics Instruments) in a standard 0°/90° measurement geometry [32]. Fluorescence measurements of Nile Red and F305 were carried out on a Perkin Elmer LS50B fluorometer in a 0°/90° standard geometry. All absorption and fluorescence measurements were performed in standard 1-cm quartz cuvettes (Hellma) at room temperature. Fluorescence microscopic images were recorded with a confocal laser scanning microscope Fluoview 1000 (OLYMPUS GmbH) using a HeNe-Laser (10 mW) as excitation source (excitation wavelength of 633 nm), which was reflected by a dichroic mirror (DM 488/543/633) and focused onto the sample through an Olympus Planapochromat ×60, N.A.=1.35 oil immersion objective. The resulting fluorescence emission was recollected by the same objective and detected in the wavelength region between 685 and 760 nm. Lifetime measurements Fluorescence lifetimes of aqueous suspensions (0.1 w%) of particles doped with Itrybe/Sq730 in standard quartz cuvettes (Hellma) were measured with a FluoTime 200 spectrometer (Picoquant GmbH) in a 0°/90° excitation-emission geometry using magic angle conditions (excitation polarizer set to 0° and emission polarizer set to 54.7°) and time-correlated
3317
single-photon counting (TCSPC). A LDH-D-C-635 diode (pulse repetition rate of 20 MHz) was used for excitation at 633 nm. The emission was detected at 750 nm (spectral bandwidth of 11 nm). Data acquisition and analysis were done with a PMA 182 detector and the Picoharp 300 TCSPC PCboard using the software SymphoTime. Fluorescence lifetimes of aqueous suspensions (0.1 w%) of particles doped with Nile Red/F305 in standard quartz cuvettes (Hellma) were measured with a custom-made fluorescence lifetime spectrometer at the group of Prof. W. Retting and Dr. W. Weigel (HU-Berlin) in a 0°/90° excitationemission geometry using time-correlated single-photon counting (TCSPC). A 560 nm diode (Horiba Jobin Yvon/IBH, pulse repetition rate of 1 MHz) was employed as excitation light source. The emission was detected >600 nm (interference filter) with a microchannel plate photomultiplier (R 1564 U-01, Hamamatsu). Data accumulation and analysis was done with a multichannel analyzer (MCDLAP, Fast Comtec) using the software Globals Unlimited. FLIM of cells incubated with fluorescent particles was performed with an OLYMPUS FV 1000 confocal laser scanning microscope, upgraded with a FLIM-FCS Upgrade Kit (Picoquant GmbH) using a 640-nm laser diode (LDH-c-640, PDL800-D, pulse width 700 nm, display a high degree of spectral overlap in emission, rendering their discrimination in the intensity domain very difficult. All PSNP-encapsulated dyes display multiexponential decay kinetics in aqueous suspension as to be expected for organic dyes in a heterogeneous environment as given by the PS matrix. Nile Red-stained PSNP show an amplitudeweighted lifetime τA of 3.7 ns, and F305-stained PSNP have a τA of 4.8 ns. Although the lifetimes of the differently stained vis-PSNPs do not vary by the desired factor of two, we assumed that these fluorescence lifetimes to be sufficiently different for our goal to differentiate between both types of vis reporters with fluorescence measurements in the time domain. In the case of the NIR-emissive PSNP, we observe a larger difference in fluorescence lifetime, with PSNP-encapsulated Itrybe showing a τA of 0.8 ns and PSNP-encapsulated Sq730, a τA of 3.0 ns, respectively. Lifetime multiplexing in the vis and data analysis Although our encapsulation approach yielded sets of vis- and NIRemissive PSNP with complex decay kinetics, previous studies with mixtures of quantum dots and organic dyes, all revealing multi-exponential decay kinetics, demonstrated that even under such conditions, label discrimination and quantification are possible [7]. In our concept of data analysis, the measured decay curves are described as linear combinations of the decay curves of the individual components present in the reporter mixture. A prerequisite for lifetime-based multiplexed analysis of two or more fluorophores with this procedure is the proper characterization of the fluorescence decay kinetics of the individual reporters. For a two component system with the components A and B, the photon numbers NA and NB detected for components A an B, and the overall number of photons emitted by the reporter mixture NO at a specific time t are given by Eq. 1. The a and b in Eq. 1 represent the fractions of the each single component, contributing to the overall measured fluorescence decay. These fractions can be obtained via a fit of NO(t) of the
Nanoparticle-based lifetime multiplexing in living cells
3319
Fluorescence
Absorbance
1.1
1.0
Detection 0.8
0.6
0.8 0.6
0.4 0.4 0.2
Excitation
0.8
Absorbance
Excitation
1.0
0.9
1.0
Absorbance
1.0
Fluorescence (norm.)
1.2
Fluorescence
Detection 0.8
0.7 0.6
0.6
0.5 0.4
0.4
0.3 0.2
Fluorescence (norm.)
Absorbance 1.4
0.2
0.2 0.1
0.0
0.0 400
450
500
550
600
650
700
0.0 400
750
500
600
700
800
900
0.0 1000
Wavelength/ nm
Wavelength/ nm
Fig. 1 Absorption (dashed lines) and emission spectra (solid lines) of vis and NIR-emissive dyes, encapsulated in 100-nm-sized PSNP, in aqueous suspension (0.1 w%). Left panel: Excitation of Nile Red (black)
and F305 (red) was at 540 nm. Right panel: Itrybe (black) and Sq730 (red) were excited at 633 nm
measured decay curve Nexp(t), thereby minimizing χ2 according to Eq. 2.
leads to an attenuation of the excitation light in the illuminated and the detected volume in the center of the cuvette and, thus, to a reduction of the overall number of detected photons due to inner filter effects. To account for these effect, the transmission at the excitation wavelength in suspensions of each type of dye-loaded particle were measured individually, and the fractions of each dye obtained from our linear combination approach were subsequently multiplied with the transmissions in the center of the cuvette. The good agreement between the applied fractions of Nile Red- and F305-stained PSNP and the fractions calculated via our linear combination approach is demonstrated exemplarily in Fig. 2b, showing the measured and calculated decay curves of a 1:1 mixture of particles stained with Nile Red and F305 and, as a summary, in Table 1, revealing the results from measurements of different particle mixtures. Obviously, with our simple data evaluation procedure, it is possible to distinguish the five different particle reporter ratios up to a ratio of Nile Red- to F305-stained PSNP of 1:5. Although the mean lifetimes of the single dyes display a difference of only 30 %,
N O ðt Þ ¼ aN A ðt Þ þ bN B ðt Þ X N O ðt Þ−N exp ðt Þ χ ¼ N exp ðt Þ t 2
ð1Þ
2 ð2Þ
This simple decomposition procedure is independent of the actual decay kinetics of the individual components, provided that both decay profiles differ sufficiently to allow a reliable decomposition in the individual components, i.e., contributions from A and B. No complex curve fitting, e.g., a multiexponential decay analysis, is required. Moreover instrumentspecific contributions can be neglected. Especially, no deconvolution with the instrument response function is necessary [7]. To study the general feasibility of our differently stained PSNP and our simple data evaluation procedure for lifetime multiplexing, the fluorescence decay kinetics of the single vis reporters and different mixtures of the vis reporters were measured in aqueous suspension (PSNP concentration of 0.1 w%). The mass ratios of both vis reporters in the PSNP mixtures analyzed, i.e., 5:1, 2:1, 1:1, 1:2, and 1:5, follow from Table 1. The results of these measurements are summarized in Fig. 2. To compensate for the different fluorescence intensities of the single components, the decay curves of both components were area-normalized yielding identical total photon numbers for components A and B. Furthermore, the different absorbances of the dye-loaded particles at the excitation wavelength, resulting in different numbers of emitted photons, had to be considered. An increasing fraction of the component with a higher absorbance
Table 1 Applied (theoretical) mass ratios of Nile Red- and F305-stained PSNP in the reporter mixtures, corresponding volume/particle fractions, results obtained via decomposition of the measured decay curves using a linear combination of the decay curves of the single bead reporters, and deviation (relative difference) between the applied and measured fractions of the encapsulated dyes Mass ratios
1/1
2/1
1/2
5/1
1/5
Fraction Nile Red Calculated fraction Nile Red Calculated fraction F305 Difference of calculated and applied Nile Red fraction (%)
0.50 0.55 0.45 5
0.67 0.65 0.35 −2
0.33 0.36 0.64 3
0.83 0.76 0.24 −7
0.17 0.28 0.72 11
3320
K. Hoffmann et al.
Normalized intensity
A -3
10
Nile Red 5:1 2:1 1:1 1:2 1:5 F305
-5
10 0
2
4
6
Fraction of PSNP-Itrybe (simulation)
PSNP-Sq730 100 %
10-4
Normalized intensity
10-3
B
10-3
10-4
10-5
10-6
10-5
10
-6
Deviation
Normalized intensity
10-2
PSNP-Itrybe 100 % 0
2
4
6
8 10 12 14 16 18 20
0.02 0.00 -0.02
0
10-5
0.02 0.01 0.00 -0.01 -0.02
0
5
Time/ns
Fig. 3 A Fluorescence decay curves of Itrybe- and Sq730-loaded PSNP (black curves) and different mixtures of both reporters (green curves) in aqueous suspension (PSNP concentration of 0.1 w%). B Measured fluorescence decay (black) and calculated decay curve obtained via linear combination (red) of an exemplarily shown 1:1 mixture of PSNP-Itrybe and PSNP-Sq730 in aqueous suspension. The deviation between the
5
10
15 20 Time/ ns
25
30
FLIM of living cells To demonstrate the application potential of our lifetime-encoded PSNP reporters in conjunction with our new concept of data analysis, we exemplarily chose cellular imaging. In this respect, we incubated 3T3 fibroblast cells with similar mixtures of Itrybe- and Sq730-stained PSNP reporters as used for our proof-of-concept measurements in suspension. Figure 4a shows the FLIM image of a living 3T3 cell, incubated with a 5:1 mixture of 25-nm-sized Itrybestained PSNP and 100-nm-sized Sq730-stained PSNP. Different particle sizes used here provide the possibility to roughly track particle uptake and localization in the cells without the need for PSNP functionalization with a targeted ligand. Details on FLIM imaging with NIR-emitting particle reporters were recently presented by us, albeit in a different context and with a different type of data analysis [16, 17]. Time-resolved fluorescence measurements with cells containing mixtures of both reporters, each of which already displaying complex decays which can be fitted only with two or three exponentials, lead to an enhanced complexity of
Multiplexing in the NIR spectral region To underline the general feasibility of our multiplexing approach, we also studied different mixtures of Itrybe- and Sq730-stained PSNP in aqueous suspension (PSNP concentration of 0.1 w%). Similarly as for the vis reporter pair, both types of NIR-emitting nanoparticles were mixed in mass ratios of 5:1, 2:1, 1:1, 1:2, and 1:5, and the fluorescence lifetimes of the single components and the different mixtures were subsequently measured. The accordingly obtained and calculated decay curves are summarized in Fig. 3. Comparison of the reporter mass ratios calculated via our linear combination approach with the actually applied mass ratios of the different NIR PSNP shown in Fig. 3c highlights the generally good agreement of both values which deviate at maximum by 16 %.
A
10-4
8 10 12 14 16 18 20 Time/ ns
the calculated and applied ratios of the Nile Red/F305 pair deviated by only maximum 11 %. This demonstrates that even slight differences in the mean lifetime τA of both encoding dyes principally enable lifetime multiplexing.
10-2
B 10-3
10-6
Deviation
-4
10
10
15
Time/ ns
20
25
Deviation/ %
Normalized intensity
Fig. 2 A Measured fluorescence decay curves of different mixtures of Nile Red- and F305-stained PSNP in aqueous suspension (PSNP concentration of 0.1 w%). B Measured (black) and calculated (red) decay curves of an exemplarily shown 1:1 mixture of Nile Red- and F305-stained PSNP. The corresponding deviation is shown in the bottom panel
1.0
C
0.8 0.6 0.4 0.2 0.0 10 0 -10 -20
0.0
0.2
0.4
0.6
0.8
1.0
Fraction of PSNP-Itrybe (theory)
measured and calculated data is depicted in the bottom panel. C Decomposition results obtained for the different mixtures of PSNP-Itrybe (100 nm sized) and PSNP-Sq730 (100 nm sized). The relative deviation from the applied mass ratios of both reporters is shown in the bottom panel
10
fraction of PSNP-Itrybe (simulation)
1.0
1
deviation/ %
A
3321
normalized intensity (simulation)
Nanoparticle-based lifetime multiplexing in living cells
0.0 20 10 0 -10
B 1000
100 PSNP-Itrybe/PSNP-Sq730 1/5
PSNP-Itrybe/PSNP-Sq730 5/1
0
5
10
15
time/ ns
20
25
C 0.8
0.6
0.4
0.2
0.0
0.2
0.4
0.6
0.8
1.0
fraction of PSNP-Itrybe (theory)
Fig. 4 A FLIM image of a living 3T3 cell, incubated with a 5:1 mixture of 25-nm-sized Itrybe-stained PSNP and of 100-nm-sized Sq730-stained PSNP. The color code of the image, calculated from three fluorescence lifetimes [17] depicts the average lifetimes. B Simulated fluorescence
decay curves of a 3T3 cells exposed to mixtures of differently sized and stained PSNP. C Decomposition results obtained for 3T3 fibroblast cells incubated with different mixtures of reporter PSNP. Deviations from the expected theoretical values are shown in the bottom panel
the measured decays and complicates data analysis following conventional fitting routines [17]. This complexity can be reduced with our simple, yet efficient linear combination approach as displayed in Fig. 4b showing the FLIM-based, simulated lifetime traces of individual cells incubated with different reporter mixtures. A systematic decrease in the decay profiles of the simulated curves was found which corresponds to the increase in the fraction of Itrybe-loaded PSNP with their shorter lifetimes. Figure 4c correlates the different particle fractions in the binary incubation mixtures and the results of the decomposition of the measured decay curves calculated via linear combination. The relative deviation from the expected values as determined from the applied reporter ratios are shown in the bottom panel. Despite the complexity of this problem, the comparatively small deviations, which amount to maximum 18 %, suggest the principal suitability of our approach also for lifetime multiplexing in living cells. In all cases, the decomposition results for the fraction of 25-nm-sized PSNP-Itrybe are systematically higher than the theoretical values. Whether this overestimation of the faster decaying component can be attributed to, e.g., scattering effects of the biological matrix or differences in the uptake behavior of 3T3 cells for the differently sized PSNP reporters, i.e., to a less efficient uptake of 100-nm-sized Sq730-doped PSNP, is currently being investigated.
a single excitation light source and a single detector, can be distinguished and quantified in solution and in living cells using a simple linear combination model that does not require complex multi-exponential decay analysis. With this procedure, even mixtures of nanoscale reporters varying in their amplitude-weighted mean lifetimes by only 30 % could be easily discriminated with relative deviations of 11 %. Moreover, fluorescence lifetime imaging studies with living mouse fibroblast cells incubated with mixtures of a set of NIR-fluorescent nanoparticle reporters with different decay kinetics demonstrated the principal suitability of our approach for cellular imaging. Most likely, this simple and versatile concept can be extended to all types of fluorophores, thereby expanding the accessible time scale from the nanosecond time domain to the microsecond and millisecond regime, as well as to different encapsulating matrices. Prerequisites are fluorescence decay kinetics of encapsulated dyes which are not affected by the immediate particle environment. We currently assess the full potential of this approach for the design of labels and targeted probes for fluorescence assays and molecular imaging and cellular imaging studies as well as barcoding applications, also in conjunction with spectral and intensity coding modes and for differently sized particles. In this respect, also new methods for data acquisition and decay analysis are being developed that will enable the very fast distinction between different lifetime pattern-encoded beads with a minimum number of photons collected. One ultimate goal here is the fabrication of surface functionalized beads containing single encoding dyes or dye mixtures for the generation of lifetime barcodes for different surface chemistries which could be utilized for bead-based assays in conjunction with suitable methods for very fast data acquisition and reliable decay analysis enabling the accurate distinction between different decay patterns with a minimum number of photons collected.
Conclusion and outlook Based upon the incorporation of hydrophobic dyes into differently sized carboxylated polystyrene nanoparticles via a simple staining procedure, we could obtain a platform of surface-functionalizable nanoscale reporters with emission in the vis and NIR which reveals different fluorescence decay profiles. These reporters, which can be read out with
3322 Acknowledgments We gratefully acknowledge the financial support from the Federal Ministry of Economics and Technology (BMWI-22/06 and BMWI-13/09). We would like to thank D. Drescher and J. Kneipp (BAM 1.0) for designing and performing the cell culture experiments. We also express our gratitude to Picoquant GmbH for support with the FLIM setup.
References 1. Nolan JP, Mandy F (2006) Multiplexed and microparticle-based analyses: quantitative tools for the large-scale analysis of biological systems. Cytometry Part A 69A(5):318–325 2. Ugozzoli LA (2004) Multiplex assays with fluorescent microbead readout: a powerful tool for mutation detection. Clin Chem 50(11): 1963–1965 3. Shepard JRE (2006) Polychromatic microarrays: simultaneous multicolor array hybridization of eight samples. Anal Chem 78(8):3589–3597 4. Evans M, Sewter C, Hill E (2003) An encoded particle array tool for multiplex bioassays. Assay Drug Dev Technol 1(1):199–207 5. De Rosa SC, Brenchley JM, Roederer M (2003) Beyond six colors: a new era in flow cytometry. Nat Med 9(1):112–117 6. Lieberwirth U, Arden-Jacob J, Drexhage KH, Herten DP, Muller R, Neumann M, Schulz A, Siebert S, Sagner G, Klingel S, Sauer M, Wolfrum J (1998) Multiplex dye DNA sequencing in capillary gel electrophoresis by diode laser-based time-resolved fluorescence detection. Anal Chem 70(22):4771–4779 7. Grabolle M, Kapusta P, Nann T, Shu X, Ziegler J, Resch-Genger U (2009) Fluorescence lifetime multiplexing with nanocrystals and organic labels. Anal Chem 81(18):7807–7813 8. Song EQ, Hu J, Wen CY, Tian ZQ, Yu X, Zhang ZL, Shi YB, Pang DW (2011) Fluorescent-magnetic-biotargeting multifunctional nanobioprobes for detecting and isolating multiple types of tumor cells. ACS Nano 5(2):761–770 9. Lakowicz JR (2006) Principles of fluorescence spectroscopy. Principles of fluorescence spectroscopy, 3rd edn. Springer, New York 10. Resch-Genger U, Grabolle M, Cavaliere-Jaricot S, Nitschke R, Nann T (2008) Quantum dots versus organic dyes as fluorescent labels. Nat Methods 5(9):763–775 11. Hotzer B, Medintz IL, Hildebrandt N (2012) Fluorescence in nanobiotechnology: sophisticated fluorophores for novel applications. Small 8(15):2297–2326 12. Hemmila I, Laitala V (2005) Progress in lanthanides as luminescent probes. J Fluoresc 15(4):529–542 13. Battersby BJ, Trau M (2007) Optically encoded particles and their applications in multiplexed biomedical assays. Aust J Chem 60(5): 343–353 14. Raymond SB, Boas DA, Bacskai BJ, Kumar ATN (2010) Lifetimebased tomographic multiplexing. J Biomed Opt 15(4):046011 15. Ehlert O, Thomann R, Darbandi M, Nann T (2008) A four-color colloidal multiplexing nanoparticle system. ACS Nano 2(1):120–124 16. Hoffmann K, Behnke T, Drescher D, Kneipp J, ReschGenger U (2011) Lifetime-based discrimination between spectrally matching vis and NIR emitting particle labels and probes. In: Proceedings of SPIE, 2011, single molecule spectroscopy and imaging IV 17. Hoffmann K, Behnke T, Drescher D, Kneipp J, Resch-Genger U (2013) Near-infrared-emitting nanoparticles for lifetime-based multiplexed analysis and imaging of living cells. ACS Nano 7(8):6674–6684 18. Behnke T, Würth C, Hoffmann K, Hubner M, Panne U, Resch-Genger U (2011) Encapsulation of hydrophobic dyes in polystyrene microand nanoparticles via swelling procedures. J Fluoresc 21(3):937–944 19. Behnke T, Würth C, Laux EM, Hoffmann K, Resch-Genger U (2012) Simple strategies towards bright polymer particles via one-step staining procedures. Dyes Pigments 94(2):247–257
K. Hoffmann et al. 20. Napp J, Behnke T, Fischer L, Wurth C, Wottawa M, Katschinski DM, Alves F, Resch-Genger U, Schaferling M (2011) Targeted luminescent near-infrared polymer-nanoprobes for in vivo imaging of tumor hypoxia. Anal Chem 83(23):9039–9046 21. Seydack M (2005) Nanoparticle labels in immunosensing using optical detection methods. Biosens Bioelectron 20(12):2454–2469 22. Burns A, Ow H, Wiesner U (2006) Fluorescent core-shell silica nanoparticles: towards “lab on a particle” architectures for nanobiotechnology. Chem Soc Rev 35(11):1028–1042 23. Yan JL, Estevez MC, Smith JE, Wang KM, He XX, Wang L, Tan WH (2007) Dye-doped nanoparticles for bioanalysis. Nano Today 2(3): 44–50 24. Wolfbeis OS (2005) Materials for fluorescence-based optical chemical sensors. J Mater Chem 15(27–28):2657–2669 25. Sharma P, Brown S, Walter G, Santra S, Moudgil B (2006) Nanoparticles for bioimaging. Adv Colloid Interface Sci 123(126): 471–485 26. Miletto I, Gilardino A, Zamburlin P, Dalmazzo S, Lovisolo D, Caputo G, Viscardi G, Martra G (2010) Highly bright and photostable cyanine dye-doped silica nanoparticles for optical imaging: photophysical characterization and cell tests. Dyes Pigments 84(1):121–127 27. Behnke T, Mathejczyk JE, Brehm R, Würth C, Gomes FR, Dullin C, Napp J, Alves F, Resch-Genger U (2013) Target-specific nanoparticles containing a broad band emissive NIR dye for the sensitive detection and characterization of tumor development. Biomaterials 34(1):160–170 28. BASF Dispersions & Pigments North America (2013) http://www. dispersions-pigments.basf.us/p02/USWeb-Internet/pigments/en_GB/ content/microsites/pigmentsdispersions/products/Lumogen. Accessed 6 Jan 2014 29. Behnke T, Hoffmann A, Hoffmann K, Resch-Genger U (2010) DE Patent DE102008040513 Verwendung einer langwellig emittierenden Cyaninverbindung als NIR-Fluoreszenzstandard und Kit zur Kalibrierung von Photolumineszenzmesssystemen 30. Patsenker LD, Tatarets AL, Povrozin YA, Terpetschnig EA (2011) Long-wavelength fluorescence lifetime labels. Bioanal Rev 3:115– 137 31. Laux EM, Behnke T, Hoffmann K, Resch-Genger U (2012) Keeping particles brilliant—simple methods for the determination of the dye content of fluorophore-loaded polymeric particles. Anal Methods 4(6):1759–1768 32. Resch-Genger U, Pfeifer D, Monte C, Pilz W, Hoffmann A, Spieles M, Rurack K, Hollandt J, Taubert D, Schonenberger B, Nording P (2005) Traceability in fluorometry: part II. Spectral fluorescence standards. J Fluoresc 15(3):315–336 33. Krämer B, Koberling F, Tannert A, Korte T, Hermann A (2005) Fluorescence lifetime imaging (FLIM) based analysis of lipid organization in hepatocytes using the MicroTime 200. PicoQuant, Berlin 34. PicoQuant (2008) Picoquant application note SymphoTime software; http://www.picoquant.de/technotes/appnote_spt_demo_workspace.pdf. Accessed 8 Oct 2013 35. Berezin MY, Akers WJ, Guo K, Fischer GM, Daltrozzo E, Zumbusch A, Achilefu S (2009) Long fluorescence lifetime molecular probes based on near infrared pyrrolopyrrole cyanine fluorophores for in vivo imaging. Biophys J 97(9):L22–L24 36. Nothdurft R, Sarder P, Bloch S, Culver J, Achilefu S (2012) Fluorescence lifetime imaging microscopy using near-infrared contrast agents. J Microsc 247(2):202–207 37. Berezin MY, Achilefu S (2010) Fluorescence lifetime measurements and biological imaging. Chem Rev 110(5):2641–2684 38. Abbasi AZ, Amin F, Niebling T, Friede S, Ochs M, Carregal-Romero S, Montenegro JM, Gil PR, Heimbrodt W, Parak WJ (2011) How colloidal nanoparticles could facilitate multiplexed measurements of different analytes with analyte-sensitive organic fluorophores. ACS Nano 5(1):21–25
RESEARCH PAPER
Nanoparticle-encapsulated vis- and NIR-emissive fluorophores with different fluorescence decay kinetics for lifetime multiplexing Katrin Hoffmann & Thomas Behnke & Markus Grabolle & Ute Resch-Genger
Received: 7 August 2013 / Revised: 16 December 2013 / Accepted: 21 December 2013 / Published online: 16 January 2014 # Springer-Verlag Berlin Heidelberg 2014
Abstract Bioanalytical, clinical, and security applications increasingly require simple, efficient, and versatile strategies to measure an ever increasing number of analytes or events in parallel in a broad variety of detection formats as well as in conjunction with chromatographic separation techniques or flow cytometry. An attractive alternative to common optical multiplexing and encoding methods utilizing spectral multiplexing/color encoding and intensity encoding is lifetime multiplexing, which relies on the discrimination between different fluorescent reporters based on their fluorescence decay kinetics. Here, we propose a platform of surfacefunctionalizable polymeric nanoparticles stained with fluorophores differing in their fluorescence lifetimes as a new multiplexing and encoding approach. Proof-of-concept measurements with different sets of lifetime-encoded polystyrene nanoparticles are presented, obtained via staining of preformed particles with visible (vis)- and near-infrared (NIR)-emissive organic dyes, which display very similar absorption and emission spectra to enable excitation and detection at the same wavelengths, yet sufficiently different fluorescence decay kinetics in suspension, thereby minimizing instrumentation costs. Data analysis was performed with a linear combination approach in the lifetime domain. Our results and first cell experiments with these reporter sets underline the suitability of our multiplexing strategy for the discrimination between and the quantification of different labels. This simple and versatile concept can be extended to all types of fluorophores, thereby expanding the accessible time scale, and can be used, e.g., for the design of labels and Published in the topical collection Multiplex Platforms in Diagnostics and Bioanalytics with guest editors Günter Peine and Günther Proll. K. Hoffmann : T. Behnke : M. Grabolle : U. Resch-Genger (*) BAM Federal Institute for Materials Research and Testing, Richard-Willstaetter-Str. 11, 12489 Berlin, Germany e-mail: [email protected]
targeted probes for fluorescence assays and molecular imaging, cellular imaging studies, and barcoding applications, also in conjunction with spectral and intensity encoding. Keywords Fluorescent label . Multiplexing . Optical encoding . Lifetime multiplexing . Fluorescence lifetime imaging FLIM . Nanoparticles
Introduction The simultaneous measurement of multiple analytes or events in parallel, termed multiplexing, has become increasingly important for the growing demands of molecular biology, medical diagnostics, and drug discovery as well as current security concerns [1–4]. The detection, monitoring, and quantification of multiple targets in a single measurement can reduce, e.g., the amount of reagents, consumables, and sample as well as the time of analysis and can decrease sampling errors. Often, this is the only way to solve the complex problems imposed on analytical techniques by modern medicine. At present, planar microarrays and bead-based assays, which are often read out with a flow cytometer or a fluorescence microscope, present the most striking examples for high-throughput, multiplexed assays, used for, e.g., DNA sequencing, biomarker analysis, disease diagnosis, and drug discovery [5, 6]. To enable the readout of a maximum number of measurable species, most multiplexing applications combine local resolution, provided by the spatial position on an array or in a microfluidic channel or by separation in a flow or on a gel, with detection methods. Amongst the most popular readout methods are fluorescence techniques due to their comparable ease of use, high sensitivity, relatively inexpensive instrumentation, and high information content providing several analyte-specific quantities simultaneously [7–9]. At
3316
the core of all fluorescence-based methods are fluorescent reporters, used as labels, sensors, or encoding reagents [10, 11]. Typically, fluorescence-based multiplexing schemes rely on spectral multiplexing (color or wavelength encoding mode), i.e., the discrimination between different emission colors, utilizing dyes that can be excited at the same wavelength with a single excitation light source and distinguished by their different emission spectra. This requires either different detectors, one for each integrally measured emission wavelength region, or a spectrally resolving detection system [5]. Suitable candidates for this application are semiconductor quantum dots with their very broad absorption spectra and their narrow emission spectra [10, 11] or lanthanide labels with their very narrow emission bands and large Stokes shifts, which, however, require typically distinct excitation wavelengths >400 nm [12]. In any case, the number of spectrally distinct codes is limited by spectral crosstalk in excitation and/ or emission and energy transfer between fluorophores with overlapping absorption and emission spectra. Bead-based multiplexed analysis like the suspension microparticle assays from Luminex and the microparticle platform from Illumina for DNA sequencing rely on single or typically, on dual wavelength intensity pattern-encoded particles with different surface chemistries, i.e., different ligands [13]. The encoded beads are commonly fabricated by incorporation of different dyes with distinguishable ratios with precise concentration control, rendering their preparation challenging and requiring comparable photostabilities of all encoding reagents. An attractive option can present fluorescence lifetime multiplexing, i.e., the discrimination between fluorophores based on their different fluorescence decay kinetics that are insensitive to variations in excitation light intensity and dye concentration, in contrast to (relative) fluorescence intensities [7, 14, 15]. Hence, fluorescence lifetime measurements are increasingly being utilized for imaging and sensing applications and recently also for multiplexed detection and barcoding [7, 16, 17]. Lifetime multiplexing and encoding require fluorescent reporters with sufficiently different lifetimes, which should be excitable at the same wavelength and detectable within the same spectral window to minimize instrumentation costs and as prerequisite for the eventually desired combination with color multiplexing/encoding to increase the number of distinguishable codes. Up to now, the vast majority of lifetime-based sensing and multiplexing schemes exploit molecular or nanocrystalline fluorophores, the fluorescence lifetimes and decay kinetics of which are sensitive to their microenvironment and in the case of the latter, also to their surface chemistry/ligand shell. Although mainly molecular fluorophores are the reporter of choice for lifetime-based sensing and discrimination, this environment sensitivity can hamper many other multiplexing and barcoding applications, where environment-insensitive codes are desired.
K. Hoffmann et al.
The need for efficient multiplexing and encoding strategies motivated us to develop a new approach to lifetime multiplexing using different fluorescent reporters including organic dyes and quantum dots and a new concept for data analysis pursuing a simple type of “pattern matching” in the time domain [7]. This decomposition procedure is independent of the actual shape of the decay curves, thus enabling also the analysis of multi-exponential and nonexponential decays. To assess the versatility of this concept, we studied the suitability of this approach for different sets of fluorescent nanoparticles obtained by staining of preformed carboxylated polystyrene (PS) nanoparticles (NP, PSNP) with organic dyes absorbing and emitting in the visible (vis) and near infrared (NIR), which can be only distinguished by their different fluorescence lifetimes. Dye encapsulation elegantly circumvents possible interactions between dyes and their surrounding matrix, thereby rendering the decay kinetics of each particle independent of its location and should enable the prediction of label decay kinetics from time-resolved fluorescence measurements with the reporters in suspension. Moreover, it allows the use of hydrophobic dyes [18, 19] and the design of labels containing dye combinations [20], and can enhance dye brightness and stability [21–25], which is especially beneficial for NIR dyes. Also, it can reduce toxicity concerns, e.g., for cellular imaging [26]. More important, it enables the subsequent covalent functionalization of these particles with, e.g., different target-specific ligands like antibodies, peptides, or oligonucleotides using always the same and established bioconjugation protocols [20, 27]. Hence, such an encapsulation strategy can be very advantageous for the design of an easily fabricated platform of fluorescent reporters for lifetimebased fluorescence imaging or lifetime encoding. Here, we present a proof-of-concept study using two sets of lifetime-encoded fluorescent particle reporters, one for the vis and one for the NIR region. In this respect, we studied the potential of fluorescence lifetime-based analysis using our linear combination approach previously validated with organic dyes and quantum dots to distinguish between differently stained PSNP in suspension as well as in fibroblast cells incubated with different mixtures of these particles employing fluorescence lifetime imaging microscopy (FLIM). Our intention was the development of a simple and versatile multiplexing approach exploiting the analyte-specific fluorescence decay behavior and which can be applied to complex systems like biological systems. Our aim here was to demonstrate the feasibility of a new approach for distinguishing samples containing different fractions of lifetime-encoded nanoparticles. The comparably simple linear combination fitting procedure of the resulting multicomponent fluorescence decay curves has proven to be robust and suitable for use even in complex systems like living cells. These measurements present the first step towards a new platform of lifetime-encoded particles with different surface chemistries like nanoparticles equipped with
Nanoparticle-based lifetime multiplexing in living cells
different ligands like monoclonal antibodies binding different targets for multiplexed biomarker analysis in cellular studies, which can be also easily expanded utilizing fluorophores with longer lifetimes.
Materials and methods Materials Carboxyl-functionalized polystyrene nanoparticles with diameters of 25 and 100 nm were purchased from Kisker Biotech. The vis dyes Nile Red (7-diethylamino-3,4benzophenoxazin-2-on; Lambda Physics) and Lumogen F305 (BASF) [28] and the NIR dyes Itrybe [29] (Otava Chemicals) and Squaraine Sq730 (SETA BioMedicals) [30] were encapsulated in the PSNP without further purification. The PSNP were sonicated and stained with these dyes using a previously described staining procedure [18, 19, 31]. Tetrahydrofuran (THF) was of UV-spectroscopic grade and purchased from Sigma-Aldrich. Cell culture medium Dulbecco’s Modified Eagle Medium (DMEM), fetal calf serum (FCS), and phosphate-buffered saline (PBS) were purchased from Biochrom AG. Ste ady-state fluores cence spe ctrosc opy and microscopy Absorption spectra were recorded on a Cary 5000 spectrometer (Varian). Fluorescence emission spectra of Itrybe and Sq730 were measured with a previously described spectrofluorometer 8100 (Spectronics Instruments) in a standard 0°/90° measurement geometry [32]. Fluorescence measurements of Nile Red and F305 were carried out on a Perkin Elmer LS50B fluorometer in a 0°/90° standard geometry. All absorption and fluorescence measurements were performed in standard 1-cm quartz cuvettes (Hellma) at room temperature. Fluorescence microscopic images were recorded with a confocal laser scanning microscope Fluoview 1000 (OLYMPUS GmbH) using a HeNe-Laser (10 mW) as excitation source (excitation wavelength of 633 nm), which was reflected by a dichroic mirror (DM 488/543/633) and focused onto the sample through an Olympus Planapochromat ×60, N.A.=1.35 oil immersion objective. The resulting fluorescence emission was recollected by the same objective and detected in the wavelength region between 685 and 760 nm. Lifetime measurements Fluorescence lifetimes of aqueous suspensions (0.1 w%) of particles doped with Itrybe/Sq730 in standard quartz cuvettes (Hellma) were measured with a FluoTime 200 spectrometer (Picoquant GmbH) in a 0°/90° excitation-emission geometry using magic angle conditions (excitation polarizer set to 0° and emission polarizer set to 54.7°) and time-correlated
3317
single-photon counting (TCSPC). A LDH-D-C-635 diode (pulse repetition rate of 20 MHz) was used for excitation at 633 nm. The emission was detected at 750 nm (spectral bandwidth of 11 nm). Data acquisition and analysis were done with a PMA 182 detector and the Picoharp 300 TCSPC PCboard using the software SymphoTime. Fluorescence lifetimes of aqueous suspensions (0.1 w%) of particles doped with Nile Red/F305 in standard quartz cuvettes (Hellma) were measured with a custom-made fluorescence lifetime spectrometer at the group of Prof. W. Retting and Dr. W. Weigel (HU-Berlin) in a 0°/90° excitationemission geometry using time-correlated single-photon counting (TCSPC). A 560 nm diode (Horiba Jobin Yvon/IBH, pulse repetition rate of 1 MHz) was employed as excitation light source. The emission was detected >600 nm (interference filter) with a microchannel plate photomultiplier (R 1564 U-01, Hamamatsu). Data accumulation and analysis was done with a multichannel analyzer (MCDLAP, Fast Comtec) using the software Globals Unlimited. FLIM of cells incubated with fluorescent particles was performed with an OLYMPUS FV 1000 confocal laser scanning microscope, upgraded with a FLIM-FCS Upgrade Kit (Picoquant GmbH) using a 640-nm laser diode (LDH-c-640, PDL800-D, pulse width 700 nm, display a high degree of spectral overlap in emission, rendering their discrimination in the intensity domain very difficult. All PSNP-encapsulated dyes display multiexponential decay kinetics in aqueous suspension as to be expected for organic dyes in a heterogeneous environment as given by the PS matrix. Nile Red-stained PSNP show an amplitudeweighted lifetime τA of 3.7 ns, and F305-stained PSNP have a τA of 4.8 ns. Although the lifetimes of the differently stained vis-PSNPs do not vary by the desired factor of two, we assumed that these fluorescence lifetimes to be sufficiently different for our goal to differentiate between both types of vis reporters with fluorescence measurements in the time domain. In the case of the NIR-emissive PSNP, we observe a larger difference in fluorescence lifetime, with PSNP-encapsulated Itrybe showing a τA of 0.8 ns and PSNP-encapsulated Sq730, a τA of 3.0 ns, respectively. Lifetime multiplexing in the vis and data analysis Although our encapsulation approach yielded sets of vis- and NIRemissive PSNP with complex decay kinetics, previous studies with mixtures of quantum dots and organic dyes, all revealing multi-exponential decay kinetics, demonstrated that even under such conditions, label discrimination and quantification are possible [7]. In our concept of data analysis, the measured decay curves are described as linear combinations of the decay curves of the individual components present in the reporter mixture. A prerequisite for lifetime-based multiplexed analysis of two or more fluorophores with this procedure is the proper characterization of the fluorescence decay kinetics of the individual reporters. For a two component system with the components A and B, the photon numbers NA and NB detected for components A an B, and the overall number of photons emitted by the reporter mixture NO at a specific time t are given by Eq. 1. The a and b in Eq. 1 represent the fractions of the each single component, contributing to the overall measured fluorescence decay. These fractions can be obtained via a fit of NO(t) of the
Nanoparticle-based lifetime multiplexing in living cells
3319
Fluorescence
Absorbance
1.1
1.0
Detection 0.8
0.6
0.8 0.6
0.4 0.4 0.2
Excitation
0.8
Absorbance
Excitation
1.0
0.9
1.0
Absorbance
1.0
Fluorescence (norm.)
1.2
Fluorescence
Detection 0.8
0.7 0.6
0.6
0.5 0.4
0.4
0.3 0.2
Fluorescence (norm.)
Absorbance 1.4
0.2
0.2 0.1
0.0
0.0 400
450
500
550
600
650
700
0.0 400
750
500
600
700
800
900
0.0 1000
Wavelength/ nm
Wavelength/ nm
Fig. 1 Absorption (dashed lines) and emission spectra (solid lines) of vis and NIR-emissive dyes, encapsulated in 100-nm-sized PSNP, in aqueous suspension (0.1 w%). Left panel: Excitation of Nile Red (black)
and F305 (red) was at 540 nm. Right panel: Itrybe (black) and Sq730 (red) were excited at 633 nm
measured decay curve Nexp(t), thereby minimizing χ2 according to Eq. 2.
leads to an attenuation of the excitation light in the illuminated and the detected volume in the center of the cuvette and, thus, to a reduction of the overall number of detected photons due to inner filter effects. To account for these effect, the transmission at the excitation wavelength in suspensions of each type of dye-loaded particle were measured individually, and the fractions of each dye obtained from our linear combination approach were subsequently multiplied with the transmissions in the center of the cuvette. The good agreement between the applied fractions of Nile Red- and F305-stained PSNP and the fractions calculated via our linear combination approach is demonstrated exemplarily in Fig. 2b, showing the measured and calculated decay curves of a 1:1 mixture of particles stained with Nile Red and F305 and, as a summary, in Table 1, revealing the results from measurements of different particle mixtures. Obviously, with our simple data evaluation procedure, it is possible to distinguish the five different particle reporter ratios up to a ratio of Nile Red- to F305-stained PSNP of 1:5. Although the mean lifetimes of the single dyes display a difference of only 30 %,
N O ðt Þ ¼ aN A ðt Þ þ bN B ðt Þ X N O ðt Þ−N exp ðt Þ χ ¼ N exp ðt Þ t 2
ð1Þ
2 ð2Þ
This simple decomposition procedure is independent of the actual decay kinetics of the individual components, provided that both decay profiles differ sufficiently to allow a reliable decomposition in the individual components, i.e., contributions from A and B. No complex curve fitting, e.g., a multiexponential decay analysis, is required. Moreover instrumentspecific contributions can be neglected. Especially, no deconvolution with the instrument response function is necessary [7]. To study the general feasibility of our differently stained PSNP and our simple data evaluation procedure for lifetime multiplexing, the fluorescence decay kinetics of the single vis reporters and different mixtures of the vis reporters were measured in aqueous suspension (PSNP concentration of 0.1 w%). The mass ratios of both vis reporters in the PSNP mixtures analyzed, i.e., 5:1, 2:1, 1:1, 1:2, and 1:5, follow from Table 1. The results of these measurements are summarized in Fig. 2. To compensate for the different fluorescence intensities of the single components, the decay curves of both components were area-normalized yielding identical total photon numbers for components A and B. Furthermore, the different absorbances of the dye-loaded particles at the excitation wavelength, resulting in different numbers of emitted photons, had to be considered. An increasing fraction of the component with a higher absorbance
Table 1 Applied (theoretical) mass ratios of Nile Red- and F305-stained PSNP in the reporter mixtures, corresponding volume/particle fractions, results obtained via decomposition of the measured decay curves using a linear combination of the decay curves of the single bead reporters, and deviation (relative difference) between the applied and measured fractions of the encapsulated dyes Mass ratios
1/1
2/1
1/2
5/1
1/5
Fraction Nile Red Calculated fraction Nile Red Calculated fraction F305 Difference of calculated and applied Nile Red fraction (%)
0.50 0.55 0.45 5
0.67 0.65 0.35 −2
0.33 0.36 0.64 3
0.83 0.76 0.24 −7
0.17 0.28 0.72 11
3320
K. Hoffmann et al.
Normalized intensity
A -3
10
Nile Red 5:1 2:1 1:1 1:2 1:5 F305
-5
10 0
2
4
6
Fraction of PSNP-Itrybe (simulation)
PSNP-Sq730 100 %
10-4
Normalized intensity
10-3
B
10-3
10-4
10-5
10-6
10-5
10
-6
Deviation
Normalized intensity
10-2
PSNP-Itrybe 100 % 0
2
4
6
8 10 12 14 16 18 20
0.02 0.00 -0.02
0
10-5
0.02 0.01 0.00 -0.01 -0.02
0
5
Time/ns
Fig. 3 A Fluorescence decay curves of Itrybe- and Sq730-loaded PSNP (black curves) and different mixtures of both reporters (green curves) in aqueous suspension (PSNP concentration of 0.1 w%). B Measured fluorescence decay (black) and calculated decay curve obtained via linear combination (red) of an exemplarily shown 1:1 mixture of PSNP-Itrybe and PSNP-Sq730 in aqueous suspension. The deviation between the
5
10
15 20 Time/ ns
25
30
FLIM of living cells To demonstrate the application potential of our lifetime-encoded PSNP reporters in conjunction with our new concept of data analysis, we exemplarily chose cellular imaging. In this respect, we incubated 3T3 fibroblast cells with similar mixtures of Itrybe- and Sq730-stained PSNP reporters as used for our proof-of-concept measurements in suspension. Figure 4a shows the FLIM image of a living 3T3 cell, incubated with a 5:1 mixture of 25-nm-sized Itrybestained PSNP and 100-nm-sized Sq730-stained PSNP. Different particle sizes used here provide the possibility to roughly track particle uptake and localization in the cells without the need for PSNP functionalization with a targeted ligand. Details on FLIM imaging with NIR-emitting particle reporters were recently presented by us, albeit in a different context and with a different type of data analysis [16, 17]. Time-resolved fluorescence measurements with cells containing mixtures of both reporters, each of which already displaying complex decays which can be fitted only with two or three exponentials, lead to an enhanced complexity of
Multiplexing in the NIR spectral region To underline the general feasibility of our multiplexing approach, we also studied different mixtures of Itrybe- and Sq730-stained PSNP in aqueous suspension (PSNP concentration of 0.1 w%). Similarly as for the vis reporter pair, both types of NIR-emitting nanoparticles were mixed in mass ratios of 5:1, 2:1, 1:1, 1:2, and 1:5, and the fluorescence lifetimes of the single components and the different mixtures were subsequently measured. The accordingly obtained and calculated decay curves are summarized in Fig. 3. Comparison of the reporter mass ratios calculated via our linear combination approach with the actually applied mass ratios of the different NIR PSNP shown in Fig. 3c highlights the generally good agreement of both values which deviate at maximum by 16 %.
A
10-4
8 10 12 14 16 18 20 Time/ ns
the calculated and applied ratios of the Nile Red/F305 pair deviated by only maximum 11 %. This demonstrates that even slight differences in the mean lifetime τA of both encoding dyes principally enable lifetime multiplexing.
10-2
B 10-3
10-6
Deviation
-4
10
10
15
Time/ ns
20
25
Deviation/ %
Normalized intensity
Fig. 2 A Measured fluorescence decay curves of different mixtures of Nile Red- and F305-stained PSNP in aqueous suspension (PSNP concentration of 0.1 w%). B Measured (black) and calculated (red) decay curves of an exemplarily shown 1:1 mixture of Nile Red- and F305-stained PSNP. The corresponding deviation is shown in the bottom panel
1.0
C
0.8 0.6 0.4 0.2 0.0 10 0 -10 -20
0.0
0.2
0.4
0.6
0.8
1.0
Fraction of PSNP-Itrybe (theory)
measured and calculated data is depicted in the bottom panel. C Decomposition results obtained for the different mixtures of PSNP-Itrybe (100 nm sized) and PSNP-Sq730 (100 nm sized). The relative deviation from the applied mass ratios of both reporters is shown in the bottom panel
10
fraction of PSNP-Itrybe (simulation)
1.0
1
deviation/ %
A
3321
normalized intensity (simulation)
Nanoparticle-based lifetime multiplexing in living cells
0.0 20 10 0 -10
B 1000
100 PSNP-Itrybe/PSNP-Sq730 1/5
PSNP-Itrybe/PSNP-Sq730 5/1
0
5
10
15
time/ ns
20
25
C 0.8
0.6
0.4
0.2
0.0
0.2
0.4
0.6
0.8
1.0
fraction of PSNP-Itrybe (theory)
Fig. 4 A FLIM image of a living 3T3 cell, incubated with a 5:1 mixture of 25-nm-sized Itrybe-stained PSNP and of 100-nm-sized Sq730-stained PSNP. The color code of the image, calculated from three fluorescence lifetimes [17] depicts the average lifetimes. B Simulated fluorescence
decay curves of a 3T3 cells exposed to mixtures of differently sized and stained PSNP. C Decomposition results obtained for 3T3 fibroblast cells incubated with different mixtures of reporter PSNP. Deviations from the expected theoretical values are shown in the bottom panel
the measured decays and complicates data analysis following conventional fitting routines [17]. This complexity can be reduced with our simple, yet efficient linear combination approach as displayed in Fig. 4b showing the FLIM-based, simulated lifetime traces of individual cells incubated with different reporter mixtures. A systematic decrease in the decay profiles of the simulated curves was found which corresponds to the increase in the fraction of Itrybe-loaded PSNP with their shorter lifetimes. Figure 4c correlates the different particle fractions in the binary incubation mixtures and the results of the decomposition of the measured decay curves calculated via linear combination. The relative deviation from the expected values as determined from the applied reporter ratios are shown in the bottom panel. Despite the complexity of this problem, the comparatively small deviations, which amount to maximum 18 %, suggest the principal suitability of our approach also for lifetime multiplexing in living cells. In all cases, the decomposition results for the fraction of 25-nm-sized PSNP-Itrybe are systematically higher than the theoretical values. Whether this overestimation of the faster decaying component can be attributed to, e.g., scattering effects of the biological matrix or differences in the uptake behavior of 3T3 cells for the differently sized PSNP reporters, i.e., to a less efficient uptake of 100-nm-sized Sq730-doped PSNP, is currently being investigated.
a single excitation light source and a single detector, can be distinguished and quantified in solution and in living cells using a simple linear combination model that does not require complex multi-exponential decay analysis. With this procedure, even mixtures of nanoscale reporters varying in their amplitude-weighted mean lifetimes by only 30 % could be easily discriminated with relative deviations of 11 %. Moreover, fluorescence lifetime imaging studies with living mouse fibroblast cells incubated with mixtures of a set of NIR-fluorescent nanoparticle reporters with different decay kinetics demonstrated the principal suitability of our approach for cellular imaging. Most likely, this simple and versatile concept can be extended to all types of fluorophores, thereby expanding the accessible time scale from the nanosecond time domain to the microsecond and millisecond regime, as well as to different encapsulating matrices. Prerequisites are fluorescence decay kinetics of encapsulated dyes which are not affected by the immediate particle environment. We currently assess the full potential of this approach for the design of labels and targeted probes for fluorescence assays and molecular imaging and cellular imaging studies as well as barcoding applications, also in conjunction with spectral and intensity coding modes and for differently sized particles. In this respect, also new methods for data acquisition and decay analysis are being developed that will enable the very fast distinction between different lifetime pattern-encoded beads with a minimum number of photons collected. One ultimate goal here is the fabrication of surface functionalized beads containing single encoding dyes or dye mixtures for the generation of lifetime barcodes for different surface chemistries which could be utilized for bead-based assays in conjunction with suitable methods for very fast data acquisition and reliable decay analysis enabling the accurate distinction between different decay patterns with a minimum number of photons collected.
Conclusion and outlook Based upon the incorporation of hydrophobic dyes into differently sized carboxylated polystyrene nanoparticles via a simple staining procedure, we could obtain a platform of surface-functionalizable nanoscale reporters with emission in the vis and NIR which reveals different fluorescence decay profiles. These reporters, which can be read out with
3322 Acknowledgments We gratefully acknowledge the financial support from the Federal Ministry of Economics and Technology (BMWI-22/06 and BMWI-13/09). We would like to thank D. Drescher and J. Kneipp (BAM 1.0) for designing and performing the cell culture experiments. We also express our gratitude to Picoquant GmbH for support with the FLIM setup.
References 1. Nolan JP, Mandy F (2006) Multiplexed and microparticle-based analyses: quantitative tools for the large-scale analysis of biological systems. Cytometry Part A 69A(5):318–325 2. Ugozzoli LA (2004) Multiplex assays with fluorescent microbead readout: a powerful tool for mutation detection. Clin Chem 50(11): 1963–1965 3. Shepard JRE (2006) Polychromatic microarrays: simultaneous multicolor array hybridization of eight samples. Anal Chem 78(8):3589–3597 4. Evans M, Sewter C, Hill E (2003) An encoded particle array tool for multiplex bioassays. Assay Drug Dev Technol 1(1):199–207 5. De Rosa SC, Brenchley JM, Roederer M (2003) Beyond six colors: a new era in flow cytometry. Nat Med 9(1):112–117 6. Lieberwirth U, Arden-Jacob J, Drexhage KH, Herten DP, Muller R, Neumann M, Schulz A, Siebert S, Sagner G, Klingel S, Sauer M, Wolfrum J (1998) Multiplex dye DNA sequencing in capillary gel electrophoresis by diode laser-based time-resolved fluorescence detection. Anal Chem 70(22):4771–4779 7. Grabolle M, Kapusta P, Nann T, Shu X, Ziegler J, Resch-Genger U (2009) Fluorescence lifetime multiplexing with nanocrystals and organic labels. Anal Chem 81(18):7807–7813 8. Song EQ, Hu J, Wen CY, Tian ZQ, Yu X, Zhang ZL, Shi YB, Pang DW (2011) Fluorescent-magnetic-biotargeting multifunctional nanobioprobes for detecting and isolating multiple types of tumor cells. ACS Nano 5(2):761–770 9. Lakowicz JR (2006) Principles of fluorescence spectroscopy. Principles of fluorescence spectroscopy, 3rd edn. Springer, New York 10. Resch-Genger U, Grabolle M, Cavaliere-Jaricot S, Nitschke R, Nann T (2008) Quantum dots versus organic dyes as fluorescent labels. Nat Methods 5(9):763–775 11. Hotzer B, Medintz IL, Hildebrandt N (2012) Fluorescence in nanobiotechnology: sophisticated fluorophores for novel applications. Small 8(15):2297–2326 12. Hemmila I, Laitala V (2005) Progress in lanthanides as luminescent probes. J Fluoresc 15(4):529–542 13. Battersby BJ, Trau M (2007) Optically encoded particles and their applications in multiplexed biomedical assays. Aust J Chem 60(5): 343–353 14. Raymond SB, Boas DA, Bacskai BJ, Kumar ATN (2010) Lifetimebased tomographic multiplexing. J Biomed Opt 15(4):046011 15. Ehlert O, Thomann R, Darbandi M, Nann T (2008) A four-color colloidal multiplexing nanoparticle system. ACS Nano 2(1):120–124 16. Hoffmann K, Behnke T, Drescher D, Kneipp J, ReschGenger U (2011) Lifetime-based discrimination between spectrally matching vis and NIR emitting particle labels and probes. In: Proceedings of SPIE, 2011, single molecule spectroscopy and imaging IV 17. Hoffmann K, Behnke T, Drescher D, Kneipp J, Resch-Genger U (2013) Near-infrared-emitting nanoparticles for lifetime-based multiplexed analysis and imaging of living cells. ACS Nano 7(8):6674–6684 18. Behnke T, Würth C, Hoffmann K, Hubner M, Panne U, Resch-Genger U (2011) Encapsulation of hydrophobic dyes in polystyrene microand nanoparticles via swelling procedures. J Fluoresc 21(3):937–944 19. Behnke T, Würth C, Laux EM, Hoffmann K, Resch-Genger U (2012) Simple strategies towards bright polymer particles via one-step staining procedures. Dyes Pigments 94(2):247–257
K. Hoffmann et al. 20. Napp J, Behnke T, Fischer L, Wurth C, Wottawa M, Katschinski DM, Alves F, Resch-Genger U, Schaferling M (2011) Targeted luminescent near-infrared polymer-nanoprobes for in vivo imaging of tumor hypoxia. Anal Chem 83(23):9039–9046 21. Seydack M (2005) Nanoparticle labels in immunosensing using optical detection methods. Biosens Bioelectron 20(12):2454–2469 22. Burns A, Ow H, Wiesner U (2006) Fluorescent core-shell silica nanoparticles: towards “lab on a particle” architectures for nanobiotechnology. Chem Soc Rev 35(11):1028–1042 23. Yan JL, Estevez MC, Smith JE, Wang KM, He XX, Wang L, Tan WH (2007) Dye-doped nanoparticles for bioanalysis. Nano Today 2(3): 44–50 24. Wolfbeis OS (2005) Materials for fluorescence-based optical chemical sensors. J Mater Chem 15(27–28):2657–2669 25. Sharma P, Brown S, Walter G, Santra S, Moudgil B (2006) Nanoparticles for bioimaging. Adv Colloid Interface Sci 123(126): 471–485 26. Miletto I, Gilardino A, Zamburlin P, Dalmazzo S, Lovisolo D, Caputo G, Viscardi G, Martra G (2010) Highly bright and photostable cyanine dye-doped silica nanoparticles for optical imaging: photophysical characterization and cell tests. Dyes Pigments 84(1):121–127 27. Behnke T, Mathejczyk JE, Brehm R, Würth C, Gomes FR, Dullin C, Napp J, Alves F, Resch-Genger U (2013) Target-specific nanoparticles containing a broad band emissive NIR dye for the sensitive detection and characterization of tumor development. Biomaterials 34(1):160–170 28. BASF Dispersions & Pigments North America (2013) http://www. dispersions-pigments.basf.us/p02/USWeb-Internet/pigments/en_GB/ content/microsites/pigmentsdispersions/products/Lumogen. Accessed 6 Jan 2014 29. Behnke T, Hoffmann A, Hoffmann K, Resch-Genger U (2010) DE Patent DE102008040513 Verwendung einer langwellig emittierenden Cyaninverbindung als NIR-Fluoreszenzstandard und Kit zur Kalibrierung von Photolumineszenzmesssystemen 30. Patsenker LD, Tatarets AL, Povrozin YA, Terpetschnig EA (2011) Long-wavelength fluorescence lifetime labels. Bioanal Rev 3:115– 137 31. Laux EM, Behnke T, Hoffmann K, Resch-Genger U (2012) Keeping particles brilliant—simple methods for the determination of the dye content of fluorophore-loaded polymeric particles. Anal Methods 4(6):1759–1768 32. Resch-Genger U, Pfeifer D, Monte C, Pilz W, Hoffmann A, Spieles M, Rurack K, Hollandt J, Taubert D, Schonenberger B, Nording P (2005) Traceability in fluorometry: part II. Spectral fluorescence standards. J Fluoresc 15(3):315–336 33. Krämer B, Koberling F, Tannert A, Korte T, Hermann A (2005) Fluorescence lifetime imaging (FLIM) based analysis of lipid organization in hepatocytes using the MicroTime 200. PicoQuant, Berlin 34. PicoQuant (2008) Picoquant application note SymphoTime software; http://www.picoquant.de/technotes/appnote_spt_demo_workspace.pdf. Accessed 8 Oct 2013 35. Berezin MY, Akers WJ, Guo K, Fischer GM, Daltrozzo E, Zumbusch A, Achilefu S (2009) Long fluorescence lifetime molecular probes based on near infrared pyrrolopyrrole cyanine fluorophores for in vivo imaging. Biophys J 97(9):L22–L24 36. Nothdurft R, Sarder P, Bloch S, Culver J, Achilefu S (2012) Fluorescence lifetime imaging microscopy using near-infrared contrast agents. J Microsc 247(2):202–207 37. Berezin MY, Achilefu S (2010) Fluorescence lifetime measurements and biological imaging. Chem Rev 110(5):2641–2684 38. Abbasi AZ, Amin F, Niebling T, Friede S, Ochs M, Carregal-Romero S, Montenegro JM, Gil PR, Heimbrodt W, Parak WJ (2011) How colloidal nanoparticles could facilitate multiplexed measurements of different analytes with analyte-sensitive organic fluorophores. ACS Nano 5(1):21–25
