www.advmat.dewww.MaterialsViews.com

CO

MM

UN

ICATIO

Fan Zhang , * Qihui Shi , Yichi Zhang , Yifeng Shi , Kunlun Ding , Dongyuan Zhao , and Galen D. Stucky *

Fluorescence Upconversion Microbarcodes for Multiplexed Biological Detection: Nucleic Acid Encoding

N

Bioanalysis, disease diagnosis, and the development of thera-peutic treatments all require real-time information from multi ple targets such as proteins and genes for the global defi -nition of biosystem function. Multiplexed assays and target-pattern recognition that complement advances in genomics and proteomics are necessary to achieve this goal. [ 1 , 2 ] Today the burden of this challenge is being borne by multiplexed planar arrays. [ 3 ] Although planar arrays are having a major impact on high-density screening, the quality of the results and the speed at which they can be obtained are severely limited by the prop-erties of the planar surface.

Multiplexed suspension arrays of encoded microspheres offer advantages over planar arrays that include fl exibility in target selection, fast binding kinetics, and well controlled binding conditions. [ 1 , 2 ] Until now, this has primarily been done with both organic dye molecules [ 4–9 ] and quantum dots nano-crystals (NCs) [ 10–17 ] that have been embedded into or attached to external surface of polymer or silica microbeads for capacity spectral coding. However, for the dye-doped microbarcodes, there are only a limited number of spectrally well-resolved dyes that do not also interfere with commonly used biological labels. [ 1 ] Moreover, measurements of intensities and their ratios are intrinsically diffi cult, which limits the number of levels of dye incorporation to give distinguishable beads. Multiple excitation lasers are required if the dyes have different excita-tion wavelengths, which increases the cost of the decoding instrumentation. Furthermore the spectra overlap with radia-tive and/or nonradiative energy transfer may complicate the code. [ 1 ] There are signifi cant advantages to the use of quantum dot doped microbarcodes over those that are doped with con-ventional fl uorescent dyes. Quantum dots are relatively more photo stable and have narrower emission linewidths, but are generally made of toxic materials (e.g., CdS, CdSe, CdTe). Fur-thermore the problems associated with differentiating codes

© 2011 WILEY-VCH Verlag GAdv. Mater. 2011, 23, 3775–3779

Prof. F. Zhang , Dr. Q. H. Shi , Prof. D. Y. Zhao Department of Chemistry,and Laboratory of Advanced MaterialsFudan UniversityShanghai 200433, P R China E-mail: zhang_fan@fudan.edu.cn Prof. G. D. Stucky , Dr. Y. C. Zhang , Dr. Y. F. Shi , Dr. K. L. Ding Department of Chemistry and BiochemistryUniversity of California Santa BarbaraCA 93106-9510, USAE-mail: stucky@chem.ucsb.edu

DOI: 10.1002/adma.201101868

based on different amounts of the same quantum dots are similar to those for organic dyes. [ 1 ] In addition, the dyes and quantum dots must be compatible in the swelling solvent and the doping process must be reproducible; this becomes more diffi cult as the numbers and concentrations of the dyes and quantum dots increase. [ 1 ]

In addition, for multiplexed detection a reporter dye is required and the region of the spectrum that is occupied by its emission profi le is not available for encoding. But for the dye and quantum dot encoding materials referred to above, [ 1 , 4 , 10 ] optical interference with organic reporter tags that have excita-tions in the visible or UV region is still a major problem. In order to minimize this spectral interference, it is necessary to separate the coding and target signals to as great an extent as possible. This disadvantage limits the different color use of the target molecules and the code number. In order to advance spa-tial and temporal capabilities of multiplexed biological detec-tion for the development of modern gene technology and the medical sciences, the formulation of more appropriate lumi-nescent encoding species remains a challenging and crucially needed task.

In the present work, we show that the problems noted above can be resolved by the development of upconversion optically encoded microbeads. This encoded bead technology is based on the optical properties of fl uoride-based rare earth (RE) ion-doped upconversion nanocrystals (UPNCs), which are excited in the infrared region instead of the UV and visible region to give emission in the visible domain. [ 18–20 , 28 , 29 ] In parallel, we have also recently considered the use of RE oxide doped nano-particles for immunoassay applications. [ 30 ] However, upconver-sion effi ciency in fl uoride is relatively high due to the lower phonon energy of their lattices compared to the oxide host. The fl uoride UPNCs show a high photostability and low toxicity, and their multicolor optical properties with narrower emission band compared to oxide host can be tuned fi nely by variation of lanthanide dopants. [ 21–38 ] In particular, it is extremely impor-tant to note that there is no optical cross talk between the upconversion optical code and any reporter dyes. The single wavelength 980-nm continuous wave (CW) laser used to excite the upconversion materials does not excite dyes that absorb in the visible and UV regions and, conversely, the upconversion materials are not excited by the visible lasers used to excite the organic dyes. All these favorable properties contribute to the promise of a great potential of UPNCs in the multiplexed detection.

The UPNCs in this work are hexagonal phase NaYF 4 :Yb,Ho,Tm@NaYF 4 core/shell nanocrystals (NCs),

mbH & Co. KGaA, Weinheim 3775wileyonlinelibrary.com

377

www.advmat.dewww.MaterialsViews.com

CO

MM

UN

ICATI

ON

Figure 1 . TEM (a) and HRTEM (b) images of the NaYF 4 :Yb,Ho,Tm@NaYF 4 core/shell nano-crystals. (c,d) Confocal upconverted luminescent image of individual UCNPs, laser power den-sity is about 8 × 10 6 W/cm 2 . (e) The time trace of emission intensity from a single UPNCs under continuous laser illumination for more than 1 h, suggesting the durable photostability of the UPNCs. (f) The zoom-in time trace and histogram of emission intensity, showing non-blinking behavior of a single UPNCs. (g) The long time trace of emission intensity from three individual UPNCs under continuous laser illumination for more than 6 h, suggesting the photostability behavior of the UPNCs.

a composition optimized for effi cient upconversion and multicolor encoding. UPNCs were synthesized according to modifi ed versions of the literature methods. [ 26 ] TEM images demonstrate that the mean particle size is ∼ 22 nm ( Figure 1 a). High-resolution images show that the NCs con-sist of crystalline domain (Figure 1b). The highly crystal-line hexagonal phase of the UPNCs is confi rmed by XRD (Figure S1a). EDX analyses reveal that the NCs have a molar ratio of 1.0Na/1.0Y/4.0F, demonstrating the formation of stoichiometric NaYF 4 (Figure S1b).

In order to determine the photostability of the core/shell UPNCs, a dilute sample of NaYF 4 :Yb,Ho,Tm@NaYF 4 UCNPs was dispersed on a silicon nitride membrane and imaged in a sample-scanning confocal optical microscopy while exciting with a tightly focused 980-nm CW laser. [ 21 ] An upconverted luminescent image of the sample shows homo-geneous and randomly distributed diffraction-limited spots on the membrane (Figure 1c,d). In addition, as confi rmed by the 1:1 correspondence of the optical and AFM images

6 © 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weiwileyonlinelibrary.com

(Figure S2), particles #1 to #3 could be unambiguously assigned as single UPNCs (Figure 1 c). The luminescence time traces for particle #1 have been studied under continuous 20 mW 980 nm CW at power density of 8 × 10 6 W/cm 2 . Notably, the NaYF 4 :Yb,Ho,Tm@NaYF 4 UCNP investi-gated here did not exhibit photobleaching or photodamage for the samples studied after 1 h of continuous laser irradiation (Figure 1 e). Furthermore, no photoblinking could be determined when the bin time for each data point in emission intensity was reduced to 1 ms (Figure 1 f). The intensity at single-particle level was preserved even after 6 h of continuous irradiation by a 980 nm CW laser, indicating excellent stability of NaYF 4 :Yb,Ho,Tm@NaYF 4 UCNPs against photobleaching (Figure 1 g).

Figure 2 shows the fl uorescence spectra of the NaYF 4 :Yb,Ho,Tm@NaYF 4 UPNCs upon a single 980 nm optical excitation. By varying the rare earth composition and the doping amount, UPNCs with unique spectral signatures can be created. Since the multiple narrow band rare earth emit-ters display far less spectral overlap than quantum dots or organic dyes, it is possible to very accurately measure their relative fl uorescent intensity. This allows the reso-lution of a very large number of ratiometric optical codes or optical signatures. We dem-onstrated that code fi ne-tuning can be alter-natively achieved via the three-component dopant UPNC system (NaYF 4 : Yb/Ho/Tm) (Figure S3) in a multi-photon emission process (Figure 2 A–C). The co-doped NaYF 4 : 20%Yb/0.2%Tm system exhibits a blue color emission, resulting from 1 D 2 → 3 F 4 and 1 G 4 → 3 H 6 . [ 25 , 28 , 29 ] By adding a second

emitter (Ho 3 + ) with different concentrations to the system, the relative intensity ratio of the dual emissions can be precisely controlled. For example, variations in the Ho 3 + concentration (0.2–2.0%) lead to prominent changes in the green (( 2 H 11/2 , 4 S 3/2 ) → 4 I 15/2 ) and red ( 4 F 9/2 → 4 I 15/2 ) spectral region. [ 25 ] Consequently, the adjustable balance of emission intensities enables the tri-doped particle system to display tunable color output from blue to green (Figure 2 D). The fl uorescent inten-sity of the various emissions within a nanocrystal relative to one emission acting as an internal standard gives n optical codes for n colors. The number of codes increases exponen-tially when multiple wavelengths and multiple intensities are used at the same time: [ 1 , 9 , 19 ] n intensity levels with m colors generate ( n m −1) unique codes. Figure 2 E shows a multicolor two-photon confocal image under 980-nm excitation obtained from a mixture of NaYF 4 :Yb,Ho,Tm@NaYF 4 UPNC-tagged beads spread on a glass surface. Remarkably, all the coded beads are observed, and their emission colors are clearly dis-tinguishable. The codes of UPNC-tagged beads depend on the

nheim Adv. Mater. 2011, 23, 3775–3779

www.advmat.dewww.MaterialsViews.com

CO

MM

UN

ICATIO

N

Figure 2 . (A–C) Fluorescence signals obtained from the NaYF 4 : Yb/Ho/Tm (20/0.2/2.0 (i), 1.5 (ii), 1.0 (iii), 0.5 (iv) and 0.2 mol% (v)) upconversion nanocrystals encoded beads. Both absolute intensities and relative intensity ratios at different wavelengths are used for coding purposes; for example (365), (687) and (677) are distinguishable codes. (D) Ten dis-tinguishable emission colors of upconversion nanocrystals excited with 980 nm NIR laser, (a) NaYF 4 : Yb/Tm (20/0.2 mol%), (b–f) NaYF 4 : Yb/Ho/Tm (20/0.2/0.2, 0.5, 1.0, 1.5, 2.0 mol%), (g–j) NaYF 4 : Yb/Ho (20, 30, 40, 60/2.0 mol%). (E) Two-photon Confocal luminescence images of a mixture of NaYF 4 : Yb/Tm, NaYF 4 : Yb/Ho and NaYF 4 : Yb/Ho/Tm nanocrystals-tagged beads. (F) Relationship between the blue, green and red emission intensity of a single bead and the number of embedded UPNCs.

UPNCs. Only one type of UPNC was embedded in the bead to realize a microbarcode. The uniformity and reproducibility of the tagged beads were analyzed by examining the variations of single-bead signals and by plotting histograms for each of the 10 intensity levels. As shown in Figure 2 F, the small spreads in the measured blue, green and red emission intensity of a single bead indicate a high level of bead uniformity. The rela-tive standard deviations are about 5–7% at low-intensity levels, and decrease to about 2–4% at high-intensity levels. A major source of the errors appears to be the intrinsic variation in bead size (2–4% in diameter), as dictated by the emulsion polymerization procedure.

A key requirement for a good RE code is low spectral inter-ference (excitation and/or emission) with the fl uorescent dyes commonly used in bioassays (e.g., FITC, Cy3, Cy5, and Texas red). [ 18 ] Overlap of spectral emission requires deconvolution of the fl uorescence emission from the organic tag and the RE code and compromises the sensitivity of the fl uorescence assays. The background contribution of the RE was evaluated according to a previously reported methodology. [ 18 ] A laser-based microarray scanner that was optimized to be maximally sensitive to FITC, Cy3, Cy5, and Texas Red was used for detec-tion. The experiment revealed that the glass slide coated by the Yb 3 + , Tm 3 + and Ho 3 + doped nanocrystals has backgrounds that are equal to or lower than that of a conventional glass slide.

In order to demonstrate the use of upconversion opti-cally encoded microbeads for biological multiplexed assays, a model DNA hybridization system was designed using oligo

© 2011 WILEY-VCH Verlag GmbH & Co. KGaA, WeinheimAdv. Mater. 2011, 23, 3775–3779

probes and triple-color UPNC encoded beads. [ 9 , 16 ] Target DNA mole cules #2–#4 are directly labeled with three different fl uorescent dyes ( Figure 3 b,c). Optical spectroscopy at the single-bead level yields both the coding and the target signals. The coding signals identify the DNA sequence, whereas the target signal indicates the pres-ence and the abundance of that sequence. Target DNA #2–#4 were labeled by Cascade Blue, FAM (Fluorescein), and Alexa Fluor 633, respectively. Figure 3 shows the assay results of one mismatched and three com-plementary oligos hybridized to triple-color encoded beads. The code 365 corresponds to the oligo probe 5′-TCA AGG CTC AGT TCG AAT GCA CCA TA-3′. No analyte fl uo-rescence was detected when control oligos (noncomplemen-tary sequences) were used for hybridization (Figure 3 a,e,f), which confi rms a high degree of sequence specifi city and

a low level of nonspecifi c adsorption. Analyte fl uorescence signals were observed only in the presence of complemen-tary targets (Figure 3 b–c,d,h,j,l). In particular, it is extremely important to note that there is no optical cross talk between the upconversion optical code and any reporter dyes. The mid-range IR radiation used to excite the upconversion mate-rials does not excite dyes which absorb in the visible region (Figure 3 e,g,i,k) and, conversely, the upconversion materials are not excited by the visible lasers used to excite the organic dyes (Figure 3 f,h,j,l).

In conclusion, fl uoride-based RE-doped upconversion microbarcodes have been successfully developed for mul-tiplexed signaling and nuclei acid encoding. This work, together with our studies of the use of oxide-based RE-doped nanobarcode for immunoassay applications, [ 30 ] demonstrates that rare earth barcodes can be easily assembled and labeled with biomolecules and possess a large optical encoding band-width capability. In comparison to the downconversion target materials (organic dye or quantum dots), the upconversion encoded materials provide important advantages. Because there is no optical cross talk between the upconversion optical code and any reporter dyes under different excitation con-dition, the target labels can be selected in a wide emission range. Furthermore, the number of codes can be increased because the code emission range has been widened greatly. This kind of novel barcode material can be used for rapid and sensitive analysis of antigens and nucleic acids, which would have many potential applications in clinical, food and environ-ment detection.

3777wileyonlinelibrary.com

3778

www.advmat.dewww.MaterialsViews.com

CO

MM

UN

ICATI

ON

Figure 3 . Schematic illustration and detection results of DNA hybridization assays using multicolor upconversion nanocrystals encoded beads. (a) Fluorescence signals obtained from a single bead with the code 365 (corresponding to probe 5′-TCA AGG CTC AGT TCG AAT GCA CCA TA-3′), after exposure to control DNA sequence Target #1 (3′-TGA TTC TCA AT GTC CCT GGA ACA GA-cascade Blue-5′). (b) Fluorescence signals of a single bead with the code 365 (same as in panel (a)), after hybridization with its target #2 5′-cascade Blue-TAT GGT GCA TTC GAA CTG AGC CTT GA-3′. (c) Fluorescence signals of a single bead with the code 677 (corresponding to probe 5′-CCG TAC AAG CAT GGA ACG GCT TTT AC-3′), after hybridi-zation with its target #3 5′-FAM-GTA AAA GCC GTT CCA TGC TTG TAC GG-3′. (d)) Fluorescence signals of a single bead with the code 634 (corre-sponding to probe 5′-TAC TCA GTA GCG ACA CAT GGT TCG AC-3′), after hybridization with its target #4 5′-Alexa Fluor®633-GCT GAA CCA TGT GTC GCT ACT GAG TA-3′. Confocal luminescence images of beads with the code 365, 365, 677 and 634, after exposure to control DNA sequence Target#1 (e, f), Target#2 (g, h), Target#3 (i, j) and Target#4 (k, l) under visible (f, h, j, l) and NIR (e, g, i, k) excitation. (The scale bar is 20 μ m).

Supporting Information Supporting Information is available from the Wiley Online Library or from the author.

Acknowledgements The work was supported by the Fudan Startup Foundation for Advanced Talents under award no. EYH1615071 and the National Science Foundation under award no. DMR 08-05148.

Received: May 19, 2011 Revised: June 8, 2011

Published online: July 18, 2011

[ 1 ] R. Wilson , A. R. Cossins , D. G. Spiller , Angew. Chem. Int. Ed. 2006 , 45 , 6104 .

[ 2 ] N. H. Finkel , X. H. Lou , C. Y. Wang , L. He , Anal. Chem. 2004 , 46 , 353A .

[ 3 ] R. S. Rao , S. R. Visuri , M. T. McBride , J. S. Albala , D. L. Mathews , M. A. Coleman , J. Lab. Med. 2003 , 27 , 182 .

© 2011 WILEY-VCH Verlag Gwileyonlinelibrary.com

[ 4 ] K. L. Kellar , J. P. Douglass , J. Immunol. Methods 2003 , 279 , 277 . [ 5 ] E. Morgan , R. Varro , H. Sepulveda , J. A. Ember , J. Apgar , J. Wilson ,

L. Lowe , R. Chen , L. Shivraj , A. Agadir , R. Campos , D. Ernst , A. Gaur , Clin. Immunol. 2004 , 110 , 252 .

[ 6 ] L. Wang , W. H. Tan , Nano. Lett. 2006 , 6 , 84 . [ 7 ] Y. Luo , Curr. Opin. Mol. Ther. 2005 , 7 , 251 . [ 8 ] P. O. Krutzik , G. P. Nolan , Nat. Methods 2006 , 3 , 361 . [ 9 ] M. Y. Han , X. H. Gao , J. Z. Su , S. Nie , Nat. Biotechnol. 2001 , 19 ,

631 . [ 10 ] P. O. K. Brien , S. S. Cummins , D. Darcy , A. Dearden , O Masala ,

N. L. Pickett , S. Ryley , A. J. Sutherland , Chem. Commun. 2003 , 2532 .

[ 11 ] Y. Li , E. C. Y. Liu , N. Pickett , P. J. Skabara , S. S. Cummins , S. Ryley , A. J. Sutherland , P. O’ Brien, J. Mater. Chem. 2005 , 15 , 1238 .

[ 12 ] H. X. Xu , M. Y. Sha , E. Y. Wong , J. Uphoff , Y. H. Xu , J. A. Treadway , A. Truong , E. OKBrien , S. Asquith , M. Stubbins , N. K. Spurr , E. H. Lai , W. Mahoney , Nucleic Acids Res. 2003 , 31 , e43 .

[ 13 ] J. J. A. Lee , S. Mardyani , A. Hung , A. Rhee , J. Klostranec , Y. Mu , D. Li , W. C. W. Chan , Adv. Mater. 2007 , 19 , 3113 .

[ 14 ] S. Rauf , A. Glidle , J. M. Cooper , Adv. Mater. 2009 , 21 , 4020 . [ 15 ] J. Yang , S. R. Dave , X. H. Gao , J. Am. Soc. Chem. 2008 , 130 , 5286 .

mbH & Co. KGaA, Weinheim Adv. Mater. 2011, 23, 3775–3779

www.advmat.dewww.MaterialsViews.com

CO

MM

UN

ICATIO

N

[ 16 ] W. C. W. Chan , D. J. Maxwell , X. H. Gao , R. E. Bailey , M. Y. Han , S. M. Nie , Curr. Opin. Biotechnol. 2002 , 13 , 40 .

[ 17 ] B. J. Egner , S. Rana , H. Smith , N. Bouloc , J. G. Frey , W. S. Brocklesby , M. Bradley , Chem. Commun. 1997 , 735 .

[ 18 ] M. J. Dejneka , A. Streltsov , S. Pal , A. G. Frutos , K. Yost , P. K. Yuen , U. Muller , J. Lahiri , Proc. Natl. Acad. Sci. USA 2003 , 100 , 389 .

[ 19 ] R. C. Haushalter , US Patent 2009/0159510 A1 . [ 20 ] F. Auzel , Chem. Rev. 2004 , 104 , 139 . [ 21 ] S. W. Wu , G. Han , D. J. Milliron , S. Aloni , S. Altoe , D. V. Talapin ,

B. E. Cohen , P. J. Schuck , Proc. Natl. Acad. Sci. USA 2009 , 106 , 10917 .

[ 22 ] D. Van , F. Rijke , H. Zijlmans , S. Li , T. Vail , A. K. Raap , Nat. Bio-technol. 2001 , 19 , 273 .

[ 23 ] F. Wang , X. G. Liu , Chem. Soc. Rev. 2009 , 38 , 976 . [ 24 ] D. E. Achatz , A. Reham , O. S. Wolfbeis , Topics Curr. Chem. 2011 ,

300 , 29 . [ 25 ] F. Wang , X. G. Liu , J. Am. Chem. Soc. 2008 , 130 , 5642 . [ 26 ] G. S. Yi , G. M. Chow , Adv. Funct. Mater. 2006 , 16 , 2324 . [ 27 ] F. Zhang , Y. Wan , T. Ying , F. Q. Zhang , Y. F. Shi , S. H. Xie , Y. G. Li ,

L. Xu , B. Tu , D. Y. Zhao , Angew. Chem. Int. Ed. 2007 , 46 , 7976 .

© 2011 WILEY-VCH Verlag GmAdv. Mater. 2011, 23, 3775–3779

[ 28 ] O. Ehlert , R. Thomann , M. Darbandi , T. Nann , ACS Nano 2008 , 2 , 120 . [ 29 ] H. H. Gorris , R. Ali , S. M. Saleh , O. S. Wolfbeis , Adv. Mater 2011 , 23 ,

1652 . [ 30 ] F. Zhang , R. C. Haushalter , R. W. Haushalter , Y. F. Shi , Y. C. Zhang ,

K. L. Ding , D. Y. Zhao , G. D. Stucky , Small 2011 , in press. [ 31 ] P. Zhang , S. Rogelj , K. Nguyen , D. Wheeler , J. Am. Chem. Soc. 2008 ,

130 , 12420 . [ 32 ] F. Zhang , Y. F. Shi , G. B. Braun , Y. C. Zhang , X. H. Sun , N. O. Reich ,

D. Y. Zhao , G. D. Stucky , J. Am. Chem. Soc. 2010 , 132 , 2850 . [ 33 ] L. Wang , R. X. Yan , Z. Y. Hao , L. Wang , Y. D. Li , Angew. Chem. Int.

Ed. 2006 , 44 , 6054 . [ 34 ] S. Heer , K. Kompe , H. U. Gudel , M. Haase , Adv. Mater. 2004 , 16 ,

2102 . [ 35 ] H. X. Mai , Y. W. Zhang , R. Si , Z. G. Yan , C. H. Yan , J. Am. Chem. Soc.

2006 , 128 , 6426 . [ 36 ] J. C. Boyer , V. Cuccia , L. A. Cuccia , J. A. Capobianco , J. Am. Chem.

Soc. 2006 , 128 , 7444 . [ 37 ] J. Zhou , M. X. Yu , Y. Sun , X. Z. Zhang , X. J. Zhu , Z. H. Wu , D. M. Wu ,

F. Y. Li , Biomaterials 2011 , 32 , 1148–1156 . [ 38 ] S. H. Qian , Z. Q. Li , Y. Zhang , Nanotechnology 2008 , 19 , 255601 .

3779bH & Co. KGaA, Weinheim wileyonlinelibrary.com