Yang-2012-Detection and Quanti

Detection and Quantification of Bacterial Autofluorescence at the Single-Cell Level by a Laboratory-Built High-Sensitivity Flow Cytometer

Lingling Yang,Yingxing Zhou,Shaobin Zhu,Tianxun Huang,Lina Wu,and Xiaomei Yan*

The Key Laboratory of Analytical Science,The Key Laboratory for Chemical Biology of Fujian Province,Department of Chemical Biology,College of Chemistry and Chemical Engineering,Xiamen University,Xiamen Fujian361005,China

*Supporting Information

distribution was generated by analyzing thousands of

found that bacterial autofluorescence can vary from80to

relatively large cell-to-cell variation in autofluorescence

ells of most organisms exhibit a natural intrinsic fluorescence,commonly called“autofluorescence”,due to the existence of numerous intracellular constituents including aromatic amino acids,NAD(P)H,flavins,and lipofuscins.1Because these endogenous molecules have specific excitation and emission wavelengths,characteristic autofluor-escence spectra associated with different species or strains have been used for bacterial detection,discrimination,and identification.2?6Although cellular autofluorescence can be exploited to distinguish biological material from inanimate particles,it can affect the sensitivity of fluorescence microscopic or flow cytometric assays by interfering with or even precluding the detection of low-level specific fluorescence.1,7?9 Bacteria possess well-studied genetics,high transformation efficiencies,simple cultivation requirements,and the ability to rapidly and inexpensively grow to a high density.10Augmented by the availability of an increasingly large number of cloning vectors and mutant host strains,the Gram-negative bacterium Escherichia coli(E.coli)remains the most attractive system for gene expression and regulation studies.11Recent advances in techniques for single-cell analysis have revealed that isogenic bacterial clonal populations exhibit substantial phenotypic variation.12?16More than80%of the genes in E.coli express fewer than100copies of their protein products per cell.17 Quantitative analyses of the E.coli proteome and transcriptome in single cells with single-molecule sensitivity showed that for approximately90%of the essential proteins,an average of10or more copies per cell are present,whereas some nonessential proteins are less abundant.18Therefore,autofluorescence from endogenous components presents a great challenge in detecting low-abundance proteins in a single bacterium.Xie and co-workers developed single-molecule fluorescence microscopy methods to track single protein expression events in vivo by physical segregation of the fluorescent target and autofluor-escent components.19,20Dovichi et al.reported the application of capillary electrophoresis to separate green fluorescence protein(GFP)from cellular autofluorescent components to reduce background.21

GFP,β-galactosidase,luciferase,and tetracysteine?biarsen-ical systems are the most widely used gene expression reporters that enable sensitive fluorescent detection.19,21?23Their fluorescent products all share remarkably similar spectral characteristics with fluorescein isothiocyanate(FITC).There-fore,when single bacteria are analyzed by flow cytometry or fluorescence microscopy,intracellular autofluorescence in the green region of the visible spectrum(500?560nm)will

Received:October21,2011

Accepted:December17,2011

Published:January6,2012

generate background fluorescence in the FITC channel.This will unavoidably interfere with the quantification or visual-ization of low abundance gene expression or FITC-labeled probes.Because the detection limit for the analyte of interest is determined by the fluctuation in the background fluorescence from autofluorescent cellular components,21it is important to quantify bacterial autofluorescence at the single-cell level to provide guidance for studies that aim to detect low level gene expression or low copy number biomolecules.For example, with a bacterial strain expressing one type of receptor proteins on the surface,if bacterial autofluorescence can contribute102?103FITC equivalents,histograms of bacterial populations binding102FITC-labeled ligand molecules per cell cannot be resolved from the background.The autofluorescence for eukaryotic cells has been studied intensively by flow cytometry and has been reported to be5000?50000FITC equivalents per cell depending on cell size,cell type,and the culture media.7,8,24However,applications of flow cytometry in bacterial analysis have been mainly limited to highly fluorescent signals.25,26Significant improvement in sensitivity is required to quantitatively measure the autofluorescence of single bacterial cells.

The development of high sensitivity flow cytometer (HSFCM)has been pioneered by Los Alamos National Laboratory.27?29By reducing the probe volume to subpicoliter and extending the transit time of each molecule/particle passing through the focused laser beam to milliseconds,rapid sizing of individual DNA fragments and detection of single phycoerythrin(PE)molecules have been demonstrated via fluorescence detection.30?32The measurement of spherical polystyrene particles as small as176nm was first reported by Dovichi et https://www.360docs.net/doc/1012432703.html,ing light scatter detection within the sheath flow cuvette.33In2009,we developed a HSFCM that enables the simultaneous detection of light scatter and fluorescence signals as nanoparticles flow through the laser beam individually.34The fluorescence channel can detect single PE molecules with signal-to-noise ratio(S/N)of17,and the side scatter channel can detect single polystyrene nanoparticles of 100nm with S/N of25.One distinctive advantage of our HSFCM is that correlation between the light scatter and fluorescence signals can be captured.On the HSFCM,absolute and simultaneous quantification of specific pathogenic strain and total bacterial cells has been demonstrated with dual fluorescence channels.35Currently,the side scatter channel can detect single gold nanoparticles as small as24nm,and the S/N ratio for analyzing single polystyrene nanoparticles of100nm has been improved to104.36,37In the present study,the ultrasensitivity of the HSFCM is exploited in the detection and quantification of bacterial autofluorescence at the single-cell level.In particular,statistic distribution can be generated rapidly by analyzing thousands of bacterial cells in1?2min.■EXPERIMENTAL SECTION

Reagents and Chemicals.Nonfluorescent polystyrene beads were purchased from Spherotech(Libertyville,IL). These particles had reported diameters of530±24nm and 1270±75nm and were obtained in concentrations of6.1×1011and4.4×1010particles/mL,respectively.Yellow-green fluorescent FluoSpheres beads of24±2,36±3,110±5,and 200±9nm with excitation/emission maxima of505/515nm were purchased from Molecular Probes/Invitrogen(Carlsbad, CA).The manufacturer-reported concentrations were2.6×1015,2.0×1015,2.7×1013,and4.5×1012particles/mL for the 24,36,110,and200nm fluorescent nanospheres,respectively. The manufacturer-reported FITC equivalents for the above four sizes of nanospheres were180,350,7400,and110000 molecules per sphere,respectively.The nanospheres were diluted to108?109particles/mL with ultrapure water containing0.05%Tween20to prevent aggregation.All of the diluted sphere solutions were stored at4°C,and sonication was performed immediately prior to analysis.Sodium dithionite was purchased from Sinopharm Chemical Reagent Co.,Ltd. (Shanghai,China).Fluorescein isothiocyanate(FITC)and all other chemicals for buffer preparation were obtained from Sigma(St.Louis,MO).Distilled,deionized water supplied by a Milli-Q RG unit(Millipore,Bedford,MA)was used in the preparation of buffer solutions.

Instrumentation.The optical setup of the laboratory-built HSFCM was tailored to enable the simultaneous detection of side scatter(SS)and green fluorescence(FL)signals(Figure S1).Briefly,a solid-state488nm continuous-wave laser (Newport Corp.,Irvine,CA)was used as the excitation source.

A laser excitation power(measured after mirror reflection)of

4.5mW was used in the present study.The0.7mm laser output beam was focused to an～8.9μm diameter spot(1/e2) by an achromatic doublet lens onto the hydrodynamically focused sample stream inside a250μm square quartz flow channel(NSG Precision Cells,Farmingdale,NY).The light emitted from individual bacterial cells was collected by an aspheric lens and then directed by a dichroic beam splitter (FF500-Di02,Semrock Inc.,Rochester,NY)into two light paths for side scatter and green fluorescence detection, respectively.The reflected side scatter light was directly detected by a R928photomultiplier tube(PMT,Hamamatsu, Japan).The transmitted fluorescence passed through a Raman edge filter(LP03-488RS,Semrock)and a bandpass filter(FF0-520/35,Semrock)and was detected by a separate R928PMT. The output signals from the PMT detectors were routed to a National Instruments DAQ card(PCI-6221,Austin,TX)and recorded at a sampling rate of10kHz.A program written in LabVIEW was used to perform data acquisition and processing. The data were processed as previously described.34Burst height and/or burst area distribution histograms for both the fluorescence and side scattering can be generated dynamically. Mean value and standard deviation can be calculated directly with the HSFCM data processing program.Sample solutions were delivered pneumatically via a precise pressure regulator, and ultrapure water served as the sheath fluid via gravity feeding.The mean transit time of E.coli O157:H7passing through the focused laser beam individually was～1.29±0.16 ms(Figure S2,Supporting Information),corresponding to a linear sample flow rate of6.9mm/s in the flow cell.The sheath flow rate was measured to be～25μL/min by increment method.The sample volumetric flow rate was measured to be ～20nL/min by weighing the sample vial containing pure water before and after running at a consistent pressure for a predetermined time period.Defined by the overlap of the focused laser spot(～8.9μm in diameter)and the sample stream(～7.8μm in diameter),the detection volume was calculated to be～0.43pL.Based on Poisson statistics,when the concentration of bacteria was～108/mL,the probability that two bacterial cells coincidently fall inside the detection volume was0.09%,which was negligible.For each bacterial sample,60s of data acquisition time was used.

Cell Culturing and Growth Conditions.E.coli TG1, Vibrio harveyi,and Vibrio alginolyticus were laboratory-kept

reagents.Nontoxigenic E.coli O157:H7(NCTC 12900),Pseudomonas aeruginosa ,Bacillus subtilis ,Micrococcus lysodeikti-cus ,and Staphylococcus aureus were purchased from the China Center for Type Culture Collection.These eight bacterial strains were grown overnight in Luria ?Bertani (LB)media (10g of tryptone,5g of yeast extract,and 5g of NaCl per liter)at 37°C in baffled flasks with rotary aeration for approximately 16h until they reached the stationary phase.The harvested cells were washed three times in normal saline solution (0.9%sodium chloride)and then stored at 4°C for future use.All bacterial stock samples were diluted to a concentration of ～108cells/mL prior to HSFCM analysis.■RESULTS AND DISCUSSION Detection of Bacterial Autofluorescence.A schematic diagram of the laboratory-built HSFCM is shown in (Figure S1,Supporting Information).The bacteria were passed through the tightly focused laser beam individually via hydrodynamic focusing.Both the side scatter (SS)and fluorescence (FL)signals were detected simultaneously using two PMTs.Figures 1a1and 1a2are typical side scatter and fluorescence burst traces obtained for unstained E.coli O157:H7.Distinct light scattering peaks were clearly detected on the side scatter channel with an achieved peak height of 1043±398mV (derived from 60s of data).Green fluorescence bursts generated from each individual E.coli O157:H7cell were easily distinguishable from the background noise and had a peak height of 71±20mV (derived from 60s of data).Because the bursts detected on the side scatter and green fluorescence channels correlated well in the time frame,the signal measured on the green fluorescence channel could be ascribed to bacterial autofluorescence or potential crosstalk from the strong light scattering signal.In order to identify the true origin of the fluorescence burst,nonfluorescent poly-styrene beads of two sizes were analyzed on the HSFCM in parallel.As shown in Figure 1b1,when 530nm polystyrene beads were measured on the HSFCM,the peak height of the side scatter bursts (1611±193mV)was stronger than that of the unstained E.coli O157:H7.Meanwhile,there were no substantial bursts observed on the green fluorescence channel (Figure 1b2).These data suggested that the green fluorescence signal detected for the unstained E.coli O157:H7originated from autofluorescence rather than the leakage of light scattering.To verify this conclusion,polystyrene beads of 1.27μm were analyzed.As we can see from Figure 1c1and 1c2,the peak height of the side scatter bursts (8632±533mV)was about 8.3-fold stronger than that of the unstained E.coli O157:H7,yet the corresponding peak height detected on the green fluorescence channel (26±5mV)due to the crosstalk of the side scatter signal was much weaker than that of the

unstained E.coli O157:H7.These results confirmed that (1)the laboratory-built HSFCM can detect bacterial autofluorescence

at the single-cell level unambiguously,(2)the signal detected on the green fluorescence channel for unstained E.coli O157:H7wholly originated from bacterial autofluorescence,and (3)when the side scatter signal is extremely strong,leakage to the green fluorescence channel can occur,but this effect is negligible for bacterial samples.

Investigation of the Origin of Bacterial Autofluor-escence.Because our HSFCM uses a 488nm laser excitation source and a 520/35bandpass filter for fluorescence detection,the autofluorescence detected in the green region may have originated from flavins,the ubiquitous coenzymes crucial to the metabolism of most organisms.Flavins comprise a category of molecules that include riboflavin (RF,vitamin B 2)and its derivatives flavin adenine dinucleotide (FAD)and flavin mononucleotide (FMN).1,38The oxidized form of flavins can be efficiently excited at 450?490nm with fluorescence emission at 500?560nm,whereas the reduced state exhibits no fluorescence.Scheme 1depicts the reversible redox property of flavins and the accompanying fluorescence changes between the oxidized and reduced forms.To investigate the origin of bacterial autofluorescence in the green region,sodium dithionite,a powerful reducing agent,39was used to treat the bacterial samples.The autofluorescence signals measured before and after dithionite reduction and upon oxidation by air re-exposure were compared.Because sodium dithionite is relatively unstable and tends to rapidly decompose in water,crystals of sodium dithionite were directly added to the bacterial sample to make a final concentration of 1%.After 2min of incubation under the anaerobic condition with nitrogen

Figure 1.(a1,b1,c1)Side scatter (SS)burst traces of E.coli O157:H7,530nm polystyrene beads,and 1.27μm polystyrene beads,respectively.(a2,b2,c2)Green fluorescence (FL)burst traces for the above three samples,respectively.The concentrations of the 530nm and 1.27μm beads were 3×108and 2.2×108particles/mL,respectively.The concentration of E.coli O157:H7was ～108cells/mL.The data were binned into 100μs intervals.

purging,the bacterial sample was analyzed immediately on the HSFCM.Figure 2a1and 2a2shows the typical side scatter and green fluorescence burst traces for E.coli O157:H7.The peak heights on both channels were 1079±410mV and 73±22mV,respectively,which is consistent with our findings in Figure 1a1and 1a2.When the bacteria were treated with 1%sodium dithionite for 2min,the green fluorescence peaks completely disappeared (Figure 2b2),and when the E.coli O157:H7sample was re-exposed to air for 60min,green fluorescence signals recovered by approximately 80%(57±18mV)(Figure 2c2)in accordance with the redox properties of flavins.The incomplete restoration of the green fluorescence signal could be ascribed to the inefficient oxidation of air exposure.This redox experiment comprising dithionite reduction and air re-exposure was also performed with a gram-positive bacterial strain Micrococcus lysodeikticus (M.lysodeikticus ),and similar results were obtained (Figure S3,Supporting Information).The dithionite reduction experiment validated that the bacterial autofluorescence detected by the HSFCM in the green region comes from endogenous flavins.It is worth noting that dithionite reduction markedly increased the side scatter signals.For example,the peak height increased from 1079±410mV (Figure 2a1)for untreated E.coli O157:H7to 2607±1069mV (Figure 2b1).The enhanced side scatter signals for E.coli O157:H7and M.lysodeikticus upon sodium dithionite treat-ment may be explained as a result of the production of gases such as H 2S and SO 2within the bacteria.39,40The side scatter and fluorescence burst area distribution histograms of the three samples described above are plotted in Figure 2d1and 2d2,respectively,for a statistically meaningful https://www.360docs.net/doc/1012432703.html,ing monodisperse fluorescent polystyrene beads of 200nm (110000FITC equivalents)in diameter,burst area variations in side scatter (1676±184)and fluorescence (35392±4955)detection were measured to have coefficient of variations (CVs)of 11%and 14%,respectively,on the HSFCM (Figure S4,Supporting Information).Because the mean burst area of E.coli O157:H7side scatter is about 2.3-fold stronger than that of the 200nm fluorescent nanospheres (3782vs 1676),the large spread of bacterial side scatter distribution (CV of 35%)could be mainly attributed to the natural phenotypic diversity of the bacterial clonal population and the different angles taken by the rod-shaped bacterial cells when passing through the inter-rogating laser beam one-by-one in a stream.For the broadened distribution of bacterial autofluorescence (CV of 34%for E.coli O157:H7),photon counting statistics in light-limited samples 41could be an equally plausible explanation besides the phenotypic variations.

Quantitative Measurement of Bacterial Autofluores-cence.After making sure that the HSFCM can explicitly detect the autofluorescence of a single bacterium,we proceeded to quantify bacterial autofluorescence in units of FITC equiv-alents.Three yellow-green fluorescent Fluospheres beads (Molecular Probes/Invitrogen)with different sizes and known fluorescein equivalents were analyzed on the HSFCM in parallel with bacterial samples to construct the calibration curve between the mean fluorescence burst area and the FITC equivalents per nanoparticles.The fluorescence burst area distribution histograms for 24nm (180FITC equiv),36nm (350FITC equiv),and 110nm (7400FITC equiv)Scheme 1.(a)Chemical Structures of FR,FMN,and FAD.(b)Reversible Redox Reaction of Flavins Figure 2.(a1,b1,c1)Side scatter burst traces of E.coli O157:H7,E.coli O157:H7treated with 1%sodium dithionite for 2min,and E.coli O157:H7exposed to air for 60min,respectively.(a2,b2,c2)Green fluorescence burst traces for the above

three samples,respectively.(d1,d2)Side scatter and fluorescence burst area distribution histograms for the above three samples,respectively.The concentration of E.coli O157:H7was ～108cells/

mL.

nanospheres are given in Figure 3.Clearly,signals of these three nanospheres were readily detectable on the green fluorescence channel with good burst area distribution histograms.The measured mean burst areas and standard deviations for the 24,36,and 110nm beads were 209±72,308±117,and 3556±1146,respectively.The calculated CVs were 34.4%,38.0%,and 32.2%for these three sizes of beads,respectively.The broadened distributions can be ascribed to the intrinsic heterogeneity of the beads.It is worth noting that for the dim fluorescence signals of 24and 36nm nanospheres,statistical variability in photon counting could contribute largely to the broader CVs.41When plotting the mean fluorescence burst areas associated with each bead population versus the manufacture assigned FITC equivalents for that population,we obtained a linear calibration curve with a squared correlation coefficient (R 2)of 0.9998.However,due to the limited availability of commercial fluorescent nanospheres in low FITC equivalents,there is a large gap between the 350and 7400FITC equivalents.Figure 4shows the bivariate dot plots of autofluorescence burst area versus side scatter burst area for eight different bacterial strains:Bacillus subtilis (B.subtilis ),E.coli TG1,E.coli O157:H7,M.lysodeikticus ,Pseudomonas aeruginosa (P.aeruginosa ),Vibrio alginolyticus (V.alginolyticus ),Vibrio harveyi (V.harveyi ),and Staphylococcus aureus (S.aureus ).As displayed in Table 1,the measured burst areas for bacterial autofluor-escence varied from 912±347to 152±52depending on the species.Because the mean burst area of bacterial autofluor-escence is much smaller than that of the 110nm fluorescent nanosphere,a two-point calibration curve constructed by 24and 36nm fluorescent beads was used to convert the burst area distribution histogram of bacterial autofluorescence to the distribution histogram of FITC equivalents per bacterial cell.The calculated FITC equivalents for all the eight bacterial strains are compiled in Table 1.As we can see from Table 1,the autofluorescence of a single bacterium can vary from 80?1400

Figure 3.(a,b,c)Fluorescence burst area distribution histograms derived from 60s of data for 24nm (180FITC equiv),36nm (350FITC equiv),and 110nm (7400FITC equiv)fluorescent nanospheres,respectively.

Figure 4.Bivariate dot plots of autofluorescence burst area versus side scatter burst area for eight different bacteria.(a ?h)B.subtilis ,E.coli TG1,E.coli O157:H7,M.lysodeikticus ,P.aeruginosa ,V.alginolyticus ,V.harveyi ,and S.aureus ,respectively.

Table 1.Quantification of Bacterial Autofluorescence at the Single-Cell Level Using the Laboratory-Built HSFCM nanosphere/

bacterium SS burst area ±σS FL burst area ±

σF N FITC ±σFITC

24nm 209±72180a 36nm 308±117350a 110nm 3556±11467400a B.subtilis 22215±7997912±3471387±546E.coli TG15158±1651525±157723±299E.coli O157:H73782±1324459±156609±315M.lysodeikticus 3133±1034374±101462±171P.aeruginosa 3001±1110319±99368±139V.alginolyticus 2844±1052215±84191±87V.harveyi 2702±973211±68182±85S.aureus 2047±839152±5281±61a Manufacturer ’s reported FITC equivalents.

FTIC equivalents depending on the https://www.360docs.net/doc/1012432703.html,pared to the 5000?50000FITC equivalents per cell observed for eukaryotic cells,bacterial autofluorescence is much weaker.Among the eight bacterial strains tested,B.subtilis and S.aureus exhibit the highest and lowest level of autofluorescence,equivalent to 1387and 81FITC molecules,respectively.The FITC equivalents per bacterial cell show a positive correlation with side scatter intensity,a reflection of cell dimension (Figure S5,Supporting Information).Nevertheless,the manufacturer ’s assignments of FITC equivalents per FluoSphere are nominal,owing to the uncertainty of the differences in responses of the reference beads compared to fluorescein,in instruments with different lasers,optics,and detectors.Therefore,without spectral response calibration of the instrument,the HSFCM-measured values FITC equivalents for bacterial autofluorescence will simply provide a relative reference owing to the different spectral properties among flavins,fluorescein,and the reference beads.■CONCLUSION We have developed a method for detecting and quantifying the very weak autofluorescence of a single bacterium in the green region of the spectrum by using a laboratory-built HSFCM.Dithionite reduction and air re-exposure experiments verified that the green autofluorescence mainly originates from endogenous flavins.Bacterial autofluorescence was quantified in units of FITC equivalents for the first time by using fluorescent nanospheres with known FITC equivalents as the quantitative calibration standards.Among the eight bacterial strains tested,it was found that bacterial autofluorescence can vary from 80to 1400FITC equivalents per cell,depending on the bacterial species,and a relatively large cell-to-cell variation in autofluorescence intensity was observed.Our reports provide a reference for the background signals that can be expected with bacteria,which is especially important when studying genes that are expressed at low levels.By detecting bacterial autofluor-escence at the single-cell level,HSFCM also will provide a simple,rapid,and reliable platform to detect and enumerate bacteria without fluorescent labeling.■ASSOCIATED CONTENT *Supporting Information Additional information as noted in the text.This material is available free of charge via the Internet at https://www.360docs.net/doc/1012432703.html,.■AUTHOR INFORMATION Corresponding Author *Phone:86-592-2184519.Fax:86-592-2189959.E-mail:xmyan@https://www.360docs.net/doc/1012432703.html,.■ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (20975087,90913015,21027010),the Program for New Century Excellent Talents in University (NCET-07-0729),Research Funds for the Doctoral Program of Higher Education of China (20090121120008,20090121110009),and the NFFTBS (No.J1030415),for which we are most grateful.■REFERENCES (1)Billinton,N.;Knight,A.W.

Anal.Biochem.2001,291,175?197.(2)Dufour,E.;Leblanc,L.FEMS Microbiol.Lett.2002,211,147?153.(3)Ammor,M.S.J.Fluoresc.2007,17,455?459.(4)Giana,H.E.;Silveira,L.;Zangaro,R.A.;Pacheco,M.T.T.J.Fluoresc.2003,13,489?493.(5)Leriche,F.;Bordessoules,A.;Fayolle,K.;Karoui,R.;Laval,K.;Leblanc,L.;Dufour,E.J.Microbiol.Methods 2004,59,33?41.(6)Patino,S.;Alamo,L.;Cimino,M.;Casart,Y.;Bartoli,F.;Garcia,M.J.;Salazar,L.J.Clin.Microbiol.2008,46,3296?3302.(7)Mosiman,V.L.;Patterson,B.K.;Canterero,L.;Goolsby,C.L.Cytometry 1997,30,151?156.(8)Sklar,L.A.Annu.Rev.Biophys.Biophys.Chem.1987,16,479?506.(9)Niswender,K.D.;Blackman,S.M.;Rohde,L.;Magnuson,M.A.;Piston,D.W.J.Microsc.1995,180,109?116.(10)Baneyx,F.Curr.Opin.Biotechnol.1999,10,411?421.(11)Terpe,K.Appl.Microbiol.Biotechnol.2006,72,211?222.(12)Elowitz,M.B.;Levine,A.J.;Siggia,E.D.;Swain,P.S.Science 2002,297,1183?1186.(13)Ozbudak,E.M.;Thattai,M.;Kurtser,I.;Grossman,A.D.;van

Oudenaarden,A.Nat.Genet.2002,31,69?73.(14)Kaern,M.;Elston,T.C.;Blake,W.J.;Collins,J.J.Nat.Rev.Genet.2005,6,451?464.(15)Locke,J.C.;Young,J.W.;Fontes,M.;Jimenez,M.J.;Elowitz,M.B.Science 2011.(16)Huang,B.;Wu,H.;Bhaya,D.;Grossman,A.;Granier,S.;Kobilka,B.K.;Zare,R.N.Science 2007,315,81?84.(17)Guptasarma,P.Bioessays 1995,17,987?997.(18)Taniguchi,Y.;Choi,P.J.;Li,G.W.;Chen,H.;Babu,M.;Hearn,J.;Emili,A.;Xie,X.S.Science 2010,329,533?538.(19)Cai,L.;Friedman,N.;Xie,X.S.Nature 2006,440,358?362.(20)Yu,J.;Xiao,J.;Ren,X.;Lao,K.;Xie,X.S.Science 2006,311,1600?1603.(21)Turner,E.H.;Lauterbach,K.;Pugsley,H.R.;Palmer,V.R.;Dovichi,N.J.Anal.Chem.2007,79,778?781.(22)Hoffmann,C.;Gaietta,G.;Zurn,A.;Adams,S.R.;Terrillon,S.;Ellisman,M.H.;Tsien,R.Y.;Lohse,M.J.Nat.Protoc.2010,5,1666?1677.(23)Russo-Marie,F.;Roederer,M.;Sager,B.;Herzenberg,L.A.;Kaiser,D.Proc.Natl.Acad.Sci.U.S.A.1993,90,8194?8198.(24)Kirby,J.A.;Ikuta,S.;Clark,K.;Proud,G.;Lennard,T.W.;Taylor,R.M.Immunology 1991,72,411?417.(25)Vives-Rego,J.;Lebaron,P.;Nebe-von Caron,G.FEMS

Microbiol.Rev.2000,24,429?448.(26)Tracy,B.P.;Gaida,S.M.;Papoutsakis,E.T.Curr.Opin.Biotechnol.2010,21,85?99.(27)Dovichi,N.J.;Martin,J.C.;Jett,J.H.;Keller,R.A.Science 1983,219,845?847.(28)Keller,R.A.;Ambrose,W.P.;Goodwin,P.M.;JETr,J.H.;

Martin,J.C.;Wu,M.Appl.Spectrosc.1996,50,12A ?32A.(29)Ambrose,W.P.;Goodwin,P.M.;Jett,J.H.;Van Orden,A.;Werner,J.H.;Keller,R.A.Chem.Rev.1999,99,2929?2956.(30)Petty,J.T.;Johnson,M.E.;Goodwin,P.M.;Martin,J.C.;Jett,J.H.;Keller,R.A.Anal.Chem.1995,67,1755?1761.(31)Yan,X.M.;Grace,W.K.;Yoshida,T.M.;Habbersett,R.C.;

Velappan,N.;Jett,J.H.;Keller,R.A.;Marrone,B.L.Anal.Chem.1999,71,5470?5480.(32)Habbersett,R.C.;Jett,J.H.Cytometry A 2004,60,125?134.(33)Zarrin,F.;Bornhop,D.J.;Dovichi,N.J.Anal.Chem.1987,59,854?860.(34)Yang,L.;Zhu,S.;Hang,W.;Wu,L.;Yan,X.Anal.Chem.2009,81,2555?2563.(35)Yang,L.;Wu,L.;Zhu,S.;Long,Y.;Hang,W.;Yan,X.Anal.Chem.2010,82,1109?1116.(36)Zhu,S.;Yang,L.;Long,Y.;Gao,M.;Huang,T.;Hang,W.;Yan,

X.J.Am.Chem.Soc.2010,132,12176?12178.(37)Zhu,S.B.;Wang,S.;Yang,L.L.;Huang,T.X.;Yan,X.M.Sci.

China Chem.2011,54,1244?1253.

(38)Rhee,H.W.;Choi,H.Y.;Han,K.;Hong,J.I.J.Am.Chem.Soc. 2007,129,4524?4525.

(39)Mayhew,S.G.;Massey,V.Biochim.Biophys.Acta1973,315, 181?190.

(40)Lambeth,D.O.;Palmer,G.J.Biol.Chem.1973,248,6095?6103.

(41)Hoffman,R.A.;Wood,J.C.Current Protocols in Cytometry; Wiley:New York,2007,Unit1.20.1-1.20.18.