Up-converting reporters for biological and other assays using laser excitation techniques

Page 1 — bibliographic record

Page 2

OTHER PUBLICATIONS Rich and Pinnow (1972), "Exploring the ultimate efficiency Leif and Vallarino (1991), "Rare-earth chelates as fluores in infrared-to-visible converting phosphors activated with cent markers in cell separation and analysis,” in Cell Sepa Er and sensitized with Yb,” J. Appl. Phys. 43:2357-2365. ration Science and Technology, Amer: Chen. Soc., Chapt. 3, Schindele and Renzoni (1990), "Ultra fluors: new fluoro pp. 41-58. phores for immunological applications." J. Clin. Immun. Lenth and MacFarlane (1992), Lasers, Optics & Photonics 13:182-186.

News, 3:8-15. Seveus et al. (1992), "Time-resolved fluorescence imaging Louge et al. (1991), "Optical fiber measurements of particle of europium chelate label in immunohistochemistry and in velocity using laser-induced phosphorescence." Applied situ hybridization.” Cytometry 13:329-338. Optics 30:1976-1981. Silversmith et al. 1986, "Green infrared-pumped erbium Lovgren et al. (1992), "Detection of lanthanide chelates by upconversion laser." J. Opt, Soc. Am. A3, p128, pdp12. time-resolved fluorescence," in Nonistopic DNA Probe Smart et al. (1991), "CW room temperature upconversion Techniques, pp. 227-261, L.J. Kricka, ed., Academic Press. lasing at blue, green and red wavelengths in infrared Manashirov et al. (Jan. 23, 1989), "Effect of the purity of -pumped Pr'-doped fluoride fibre.” Electronics Letters initial substances on luminescence intensity of erbium in 27:1307-1309.

anti-stroke luminophores." Chemical Abstracts, 110:457, Soini and Kojola (1983), "Time-resolved fluorometer for Abstract No. 30750B. lanthanide cacehlates-a new generation of nonisotopic McFarlane (1989), "Dual wavelength visible upconversion immunoassays." Clin. Chem. 29/1:65-68. laser." Appl. Phys. Letts. 54:2301-2302. Soules and Hoffman (1981), "Luminescent materials (phos McFarlane (1988), "Violet CW neodymium upconversion phors)," in Encyclopedia of Chemical Technology. Third laser." Appl. Phys. Lett. 52:1300-1302. Edition, vol. 14, pp. 527-545. Mukkala et al. (1989), "The synthesis and use of activated

N-benzyl derivatives of diethylenetriaminetetraacetic acids: Tiffany (1986), "Fluorometry, nephelometry, and turbidim alternative reagents for labelling of antibodies with metal etry," in Ttextbook of Clinical Chemistry, Tietz, ed., W.B. ions." Anal. Bio. 176:319-325. Saunders Co. pp. 78-90.

Nguyen et al. (1989), "Blue-green (450-nm) upconversion Voller (1978), "The enzyme linked immunosorbent assay Tm":YLF laser." Applied Optics 28:3553-3555. (ELISA)," in Diagnostic Horizons, vol. 2, No. 1, pp. 1-7. Eichstein et al. (1988), "Laser-excited time-resolved solid Xu and hemmila (1992), "Co-fluorescence enhancement phase fluoroimmunoassays with the new europium chelate system based on pivaloyltrifluoroacetone and yttrium for the 4.7-bis(chlorosulfophenyl)-1,10-phenanthroline-2,9-di simulatneous detection of europium, terbium, samarium and carboxylic acid as label.” Anal. Chem. 60:1069-1074. dysprosium." Anal. Chinica Acta. 256:9-16.

Page 3

Drawing sheet — no readable text.

Page 4

Drawing sheet — no readable text.

Page 5

Drawing sheet — no readable text.

Page 6

Drawing sheet — no readable text.

Page 7

Drawing sheet — no readable text.

Page 8

Drawing sheet — no readable text.

Page 9

Drawing sheet — no readable text.

Page 10

Drawing sheet — no readable text.

Page 11

Drawing sheet — no readable text.

Page 12

Drawing sheet — no readable text.

Page 13

Drawing sheet — no readable text.

Page 14

Drawing sheet — no readable text.

Page 15

Drawing sheet — no readable text.

Page 16

Drawing sheet — no readable text.

Page 17

Drawing sheet — no readable text.

Page 18

Drawing sheet — no readable text.

Page 19

Drawing sheet — no readable text.

Page 20

Drawing sheet — no readable text.

Page 21

Drawing sheet — no readable text.

Page 22

Drawing sheet — no readable text.

Page 23

Drawing sheet — no readable text.

Page 24

Visible Light Out = Negative Sample

Infrared

Cooted n Fiber Opti Phosphor Iber UpTIC

Target linked

Step 1: Mix Antigen Step 2: Interogate with Cooted Phosphors Antibody Cooted Probe Competitive Homogeneous Assay

Page 25

ADD ANTIGEN

COATED

PHOSPHOR

PHOSPHOR

ANTIGEN

EMITTED LIGHT

S OF LOW NO CAPTURE INTENSITY ATSURFACE

FOCAL (HIGH

INTENSITY

EXCITATION)

EXCITATIONLIGHT

OF LOW INTENSITY

CAPTURE

ATSURFACE

FOCAL (HIGH

NTENSITY

EXCITATION LIGHT

OF LOW INTENSITY

COMPETITIVE HOMOGENOUS

ANTIGEN CAPTURE ASSAY

Page 26

Phosphor particles phosphor particles In the presence of analyte, Sharply convergent the antibody coated excitation light beam phosphors will form a large enough complex to drop to the bottom of the well

Homogeneous Immunoprecipitation ASSOy

Page 27

Drawing sheet — no readable text.

Page 28

Drawing sheet — no readable text.

Page 29

Drawing sheet — no readable text.

Page 30

Drawing sheet — no readable text.

Page 31

Drawing sheet — no readable text.

Page 32

FIG. 31A F3 10 x 10 Gridded Array of 980 nm Diode Losers

- Aqueous

Sample Flow

YZY YAZ77 10 x 10 Gridded Array of Photodiode Detectors

individual

Diode Laser in Array

Aqueous

Support Motrix a sea

10-25 um Polymer film

Overloy Used as individuo Photodiode Capture Surface Detector in Array

Coniugated Antigen Capture Probe Bonded to

PhoSchor

FIG 51B

Page 33

Drawing sheet — no readable text.

Page 34

UP-CONVERTING REPORTERS FOR signal, and accurate measurement with precision radiometric BIOLOGICAL AND OTHER ASSAYS USING instruments (scintillation and gamma counters) or with LASER EXCITATION TECHNIQUES inexpensive and sensitive autoradiographic techniques. However, radioisotopic labels also have several disadvantages, such as: potential health hazards, difficulty in

CROSS REFERENCE TO RELATED disposal, special licensing requirements, and instability APPLICATION (radioactive decay and radiolysis). Further, the fact that This application is a division of application Ser. No. radioisotopic labels typically do not produce a strong (i.e., 08/416,023 filed Mar. 30, 1995, which is a continuation-in non-Cerenkov) signal in the ultraviolet, infrared, or visible part of application Ser. No. 08/381,006 (now abandoned), portions of the electromagnetic spectrum makes radioiso filed Jan. 30, 1995, which is a continuation of Ser. No. topes generally unsuitable as labels for applications, such as 07/946,068 filed Sep. 14, 1992 (now abandoned). This microscopy, image spectroscopy, and flow cytometry, that application claims foreign priority benefits under Title 35, employ optical methods for detection. United States Code, Section 119, to PCT/US93/08712 filed For these and other reasons, the fields of clinical Sep. 14, 1993. Each of these applications is specifically 15 chemistry, water and air monitoring, and biomedical incorporated herein by reference. research have sought alternative detectable labels that do not require radioisotopes. Examples of such non-radioactive

BACKGROUND OF THE INVENTION labels include: (1) enzymes that catalyze conversion of a chromogenic substrate to an insoluble, colored product (e.g.,

The invention relates generally to detectable labels and alkaline compositions useful in assay methods for detecting soluble, peroxidase)phosphatase, or catalyze a

B-galactosidase, horseradish reaction that yields a fluorescent or suspended, or particulate substances or analytes such as luminescent product (e.g., luciferase) proteins, carbohydrates, nucleic acids, bacteria, viruses, and (1990) Anal. Chem, 62: 2258; Durrant, I. (Beck and Koster eukaryotic cells and more specifically relates to composi 297; Analytical Applications of Bioluminescence 346: (1990) Nature and tions and methods that include luminescent (phosphorescent 25 Chemiluminescence (1984) Kricka et al. (Eds.) Academic or fluorescent) labels. Press, London), and (2) direct fluorescent labels (e.g., fluo Methods for detecting specific macromolecular species, rescein isothiocyanate, rhodamine, Cascade blue), which such as proteins, drugs, and polynucleotides, have proven to absorb electromagnetic energy in a particular absorption be very valuable analytical techniques in biology and wavelength spectrum and subsequently emit visible light at medicine, particularly for characterizing the molecular com one or more longer (i.e., less energetic) wavelengths. position of normal and abnormal tissue samples and genetic 30 Using enzymes and phosphorescent/fluorescent or colo material. Many different types of such detection methods are widely used in biomedical research and clinical laboratory rimetric detectable labels offers the significant advantage of signal amplification, since a single enzyme molecule typi medicine. Examples of such detection methods include: cally has a persistent capacity to catalyze the transformation immunoassays, immunochemical staining for microscopy, 35 of a chromogenic substrate into detectable product. With fluorescence-activated cell sorting (FACS), nucleic acid appropriate reaction conditions and incubation time, a single hybridization, water sampling, air sampling, and others. enzyme molecule can produce a large amount of product, Typically, a detection method employs at least one ana and hence yield considerable signal amplification. However, lytical reagent that binds to a specific target macromolecular detection methods that employ enzymes as labels disadvan species and produces a detectable signal. These analytical tageously require additional procedures and reagents in reagents typically have two components: (1) a probe order to provide a proper concentration of substrate under macromolecule, for example, an antibody or conditions suitable for the production and detection of the oligonucleotide, that can bind a target macromolecule with colored product. Further, detection methods that rely on a high degree of specificity and affinity, and (2) a detectable enzyme labels typically require prolonged time intervals for label, such as a radioisotope or covalently-linked fluorescent 45 generating detectable quantities of product, and also gener dye molecule. In general, the binding properties of the probe ate an insoluble product that is not attached to the probe macromolecule define the specificity of the detection molecule.

method, and the detectability of the associated label deter An additional disadvantage of enzyme labels is the dif mines the sensitivity of the detection method. The sensitivity ficulty of detecting multiple target species with enzyme of detection is in turn related to both the type of label labeled probes. It is problematic to optimize reaction con employed and the quality and type of equipment available to ditions and development time(s) for two or more discrete detect it.

enzyme label species and, moreover, there is often consid

For example, radioimmunoassays (RIA) have been erable spectral overlap in the chromophore end products among the most sensitive and specific analytical methods which makes discrimination of the reaction products diffi used for detecting and quantitating biological macromol 55 cult.

ecules. Radioimmunoassay techniques have been used to Fluorescent labels do not offer the signal amplification detect and measure minute quantities of specific analytes, advantage of enzyme labels, nonetheless, fluorescent labels such as polypeptides, drugs, steroid hormones, possess significant advantages which have resulted in their polynucleotides, metabolites, and tumor markers, in biologi widespread adoption in immunocytochemistry. Fluorescent cal samples. Radioimmunoassay methods employ immuno labels typically are small organic dye molecules, such as globulins labeled with one or more radioisotopes as the analytical reagent. Radiation (or, B, or y) produced by decay fluorescein, Texas Red, or rhodamine, which can be readily conjugated to probe molecules, such as immunoglobulins or of the attached radioisotope label serves as the signal which Staph. aureus Protein A. The fluorescent molecules can be detected and quantitated by various radiometric (fluorophores) can be detected by illumination with light of methods. 65 an appropriate excitation frequency and the resultant spec Radioisotopic labels possess several advantages, such as: tral emissions can be detected by electro-optical sensors or very high sensitivity of detection, very low background light microscopy.

Page 35

A wide variety of fluorescent dyes are available and offer region is not well-suited for detection by the human eye, a selection of excitation and emission spectra. It is possible hampering the use of phycobiliprotein and cyanine labels in to select fluorophores having emission spectra that are optical fluorescence microscopy, (2) cyanines, sufficiently different so as to permit multitarget detection and phycobiliproteins, and the coupled accessory molecules discrimination with multiple probes, wherein each probe (e.g., Azure A) are organic molecules susceptible to pho species is linked to a different fluorophore. Because the tobleaching and undergoing undesirable chemical interac spectra of fluorophores can be discriminated on the basis of tions with other reagents, and (3) emitted radiation is down both narrow band excitation and selective detection of converted, i.e., of longer wavelength(s) than the absorbed emission spectra, two or more distinct target species can be excitation radiation. For example, Azure A absorbs at 632 detected and resolved (Titus et al. (1982) J. Immunol. 10 nm and emits at 645 nm, and allophycocyanin absorbs at 645 Methods 50: 193; Nederlof et al. (1989) Cytometry 10:20; nm and emits at 655 nm, and therefore autofluorescence and Ploem, J. S. (1971) Ann. NY Acad. Sci. 177: 414). background noise from scattered excitation light is not Unfortunately, detection methods which employ fluores eliminated.

cent labels are of limited sensitivity for a variety of reasons. Another alternative class of fluorophore that has been First, with conventional fluorophores it is difficult to dis 15 proposed are the down-converting luminescent lanthanide criminate specific fluorescent signals from nonspecific back chelates (Soini and Lovgren (1987) CRC Crit. Rev. Anal. ground signals. Most common fluorophores are aromatic Chem. 18: 105; Leif et al. (1977) Clin. Chem. 23: 1492; organic molecules which have broad absorption and emis Soini and Hemmila (1979) Clin. Chem. 25:353; Seveus et sion spectra, with the emission maximum red-shifted al. (1992) Cytometry 13:329). Down-converting lanthanide 50-100 nm to a longer wavelength than the excitation (i.e., 20 chelates are inorganic phosphors which possess a large absorption) wavelength. Typically, both the absorption and downward Stokes shift (i.e., emission maxima is typically at emission bands are located in the UV/visible portion of the least 100 nm greater than absorption maxima) which aids in spectrum. Further, the lifetime of the fluorescence emission the discrimination of signal from scattered excitation light. is usually short, on the order of 1 to 100 ns. Unfortunately, Lanthanide phosphors possess emission lifetimes that are these general characteristics of organic dye fluorescence are 25 sufficiently long (i.e., greater than 1 us) to permit their use also applicable to background signals which are contributed in time-gated detection methods which can reduce, but not by other reagents (e.g., fixative or serum), or autofluores totally eliminate, noise caused by shorter-lived autofluores cence or the sample itself (Jongkind et al. (1982) Ep, Cell cence and scattered excitation light. Further, lanthanide Res. 138: 409; Aubin, J. E. (1979).J. Histochem. Cytochem. phosphors possess narrow-band emission, which facilitates 27: 36). Autofluorescence of optical lenses and reflected 30 wavelength discrimination against background noise and excitation light are additional sources of background noise scattered excitation light, particularly when a laser excita in the visible spectrum (Beverloo et al. (1991) Cytometry 11: tion source is utilized (Reichstein et al. (1988) Anal. Chem. 784; Beverloo et al. (1992) Cytometry 13:561). Therefore, 60: 1069). Recently, enzyme-amplified lanthanide lumines the limit of detection of specific fluorescent signal from cence using down-converting lanthanide chelates has been typical fluorophores is limited by the significant background 35 proposed as a fluorescent labeling technique (Evangelista et noise contributed by nonspecific fluorescence and reflected al. (1991) Anal. Biochem. 197: 213; Gudgin-Templeton et al. excitation light. (1991) Clin Chem. 37: 1506).

A second problem of organic dye fluorophores that limits Until recently, down-converting lanthanide phosphors sensitivity is photolytic decomposition of the dye molecule have had the significant disadvantage that their quantum (i.e., photobleaching). Thus, even in situations where back efficiency in aqueous (oxygenated) solutions is so low as to ground noise is relatively low, it is often not possible to render them unsuitable for cytochemical staining. Beverloo integrate a weak fluorescent signal over a long detection et al. (op.cit.) have described a particular down-converting time, since the dye molecules decompose as a function of lanthanide phosphor (yttrium oxysulfide activated with incident irradiation in the UV and near-UV bands. europium) that produces a signal in aqueous solutions which However, because fluorescent labels are attractive for 45 can be detected by time-resolved methods. Seveus et al. various applications, several alternative fluorophores having (op.cit.) have used down-converting europium chelates in conjunction with time-resolved fluorescence microscopy to advantageous properties for sensitive detection have been proposed. One approach has been to employ organic dyes reject the signal from prompt fluorescence and thereby comprising a phycobiliprotein acceptor molecule dye that reduce autofluorescence. Tanke et al. (U.S. Pat. No. 5,043, 265) report down-converting phosphor particles as labels for emits in the far red or near infrared region of the spectrum 50 immunoglobulins where nonspecific fluorescent noise is reduced. Phycobilip and polynucleotides. roteins are used in conjunction with accessory molecules However, the down-converting lanthanide phosphor of that effect a large Stokes shift via energy transfer mecha Beverloo et al. and the europium chelate of Seveus et al. nisms (U.S. Pat. No. 4666,862; Oiet al. (1982) J. Cell. Biol. require excitation wavelength maxima that are in the ultra 93: 891). Phycobiliprotein labels reduce the degree of spec 55 violet range, and thus produce significant sample autofluo tral overlap between excitation frequencies and emission rescence and background noise (e.g., serum and/or fixative frequencies. An alternative approach has been to use cyanine fluorescence, excitation light scattering and refraction, etc.) dyes which absorb in the yellow or redregion and emit in the that must be rejected (e.g., by filters or time-gated signal red or far red where autofluorescence is reduced (Mujumbar rejection). Further, excitation with ultraviolet irradiation et al. (1989) Cytometry 10: 11). damages nucleic acids and other biological macromolecules, However, with both the phycobiliproteins and the cyanine posing serious problems for immunocytochemical applica dyes the emission frequencies are red-shifted (i.e., frequency tions where it is desirable to preserve the viability of living downshifted) and emission lifetimes are short, therefore cells and retain cellular structures (e.g., FACS, cyto background autofluorescence is not completely eliminated architectural microscopy).

as a noise source. More importantly perhaps, phycobilipro 65 Laser scanning fluorescence microscopy has been used teins and cyanine dyes possess several distinct disadvan for two-photon excitation of a UV-excitable fluorescent tages: (1) emission in the red, far red, and near infrared organic dye, Hoechst 33258, using a stream of strongly

Page 36

focused laser pulses (Denk et al. (1990) Science 248: 73). reporter's excitation band and a detector that is sensitive to The organic fluorphore used by Denk et al. was significantly at least some wavelengths in the reporter's emission band. photobleached by the intense, highly focused laser light The laser light is preferably focused to a small region in the during the course of imaging. Motsenbocker et al. (EP 476 sample, and light emanating from that region is collected 556) describes a method to increase luminol chemilumines and directed to the detector. An electrical signal representing cence by adding a dye catalyst that absorbs long wavelength the intensity of light in the emission band provides a radiation (deep red light) and subsequently reacts with measure of the amount of reporter present. Depending on the molecular oxygen to generate an oxidant which can itself detector's spectral response, it may be necessary to provide react with luminol and produce oxidized luminol which a filter to block the excitation light.

emits blue light. Gavrilovic (U.S. Pat. No. 5,166,948) dis 10 Simultaneous detection of multiple reporters is possible, closes a method and apparatus for optical pumping of at least where the reporters have different excitation bands or infrared pump light to a visible or ultraviolet emission light different emission bands. Where the excitation bands differ. having a wavelength shorter than the pump light (i.e., multiple laser diodes emitting at respective appropriate up-converted emission). wavelengths are combined using a wavelength division Thus, there exists a significant need in the art for labels 15 multiplexer or other suitable techniques, such as frequency and detection methods that permit sensitive optical and/or labeling, frequency modulation, and lock-in detector device. spectroscopic detection of specific label signal(s) with If the emission bands are different (whether or not the essentially total rejection of nonspecific background noise, excitation bands are different), light in the different emission and which are compatible with intact viable cells and bands is separated and sent to multiple detectors. If the aqueous or airborne environments. emission bands overlap, a single detector may be used, but The references discussed herein are provided solely for time other detection techniques are used. One example is to use their disclosure prior to the filing date of the present appli emitting multiplexing techniques so that only one reporter is cation. Nothing herein is to be construed as an admission diodes canatbea given time. Alternatively, the different laser modulated at different characteristic frequen that the inventors are not entitled to antedate such disclosure 25 cies and lock-in detection performed.

by virtue of prior invention.

Detection methods and detection apparatus of the present

SUMMARY OF THE INVENTION invention enable the ultrasensitive detection of up-converting phosphors and up-converting organic dyes by

The present invention provides labels, detection methods, exploiting what is essentially the total absence of back and detection apparatus which permit ultrasensitive detec 30 ground noise (e.g., autofluorescence, serum/fixative tion of cells, biological macromolecules, and other analytes, fluorescence, which can be used for multiple target detection and target characteristicsexcitation of light scatter) that are advantageous up-converting labels. Some embodiments discrimination. The up-converting labels of the invention of the invention utilize time-gated detection and/or permit essentially total rejection of non-specific background wavelength-gated detection for optimizing detection autofluorescence and are characterized by excitation and 35 sensitivity, discriminating multiple samples, and/or detect emitted wavelengths that are typically in the infrared or ing multiple probes on a single sample. Phase-sensitive visible portions of the spectrum, respectively, and thus avoid detection can also be used to provide discrimination between the potentially damaging effects of ultraviolet radiation. The signal(s) attributable to an up-converting phosphor and up-converting labels of the invention convert long background noise (e.g. autofluorescence) which has a dif wavelength excitation radiation (e.g., near-IR) to emitted ferent phase shift.

radiation at about one-half to one-third the wavelength of the excitation wavelength. Since background fluorescence in the alsoUp-converting organic dyes, such as red-absorbing dyes, can be used in an alternate embodiment that converts visible range is negligible if near-IR excitation wavelengths the photons absorbed by the dye into a transient voltage that are used, the use of up-converting labels provides essentially can be measured using electrodes and conventional elec background-free detection of signal. 45 tronic circuitry. After having undergone two-photon absorp In brief, the invention provides the use of luminescent tion the dye is ionized by additional photons from the light materials that are capable of multiphoton excitation and source (e.g., a laser) leading to short-lived molecular ions have upshifted emission spectra. In one embodiment of the whose presence can be detected and quantified by measuring invention, up-converting phosphors (i.e., which absorb mul the transient photoconductivity following the excitation intra tiple photons in a low frequency band and emit in a higher diation. In this embodiment, resonant multiphoton ioniza frequency band) are used as labels which can be linked to tion is used to provide a quantitative measurement of the one or more probes, such as an immunoglobulin, number and/or concentration of dye molecules in a sample. polynucleotide, streptavidin, Protein A, receptor ligand, or Furthermore, essentially all photoions formed in the irradi other probe molecule. In an another embodiment, ated sample contribute to the signal, whereas photons are up-converting organic dyes serve as the label. The organic 55 emitted isotopically and only a fraction can be collected dye labels and phosphor labels of the invention are highly using optics. Measurement of the transient photocurrent compatible with automated diagnostic testing, microscopic effectively imaging applications, and coded particle detection, among tronic signaltransfers that is the conversion of photons into an elec readily measured with relatively simple many other applications. and inexpensive sensors such as electrodes. The nature of the invention provides considerable flex In some embodiments, the present invention utilizes one ibility in the apparatus for carrying out the methods. As a or more optical laser sources for generating excitation general matter, the excitation source may be any convenient illumination of one or more discrete frequency(ies). In light source, including inexpensive near-infrared laser certain variations of the invention, laser irradiation of an diodes or light-emitting diodes (LEDs), and the detector up-converting label can modify the immediate molecular may be any convenient detector, such as a photodiode. In the 65 environment through laser-induced photochemical pro case of a single reporter, the apparatus includes a laser diode cesses involving either direct absorption or energy transfer; capable of emitting light at one or more wavelengths in the such spatially-controlled deposition of energy can be used to

Page 37

produce localized damage and/or to probe the chemical tor the progress of ligand-binding reactions, polynucleotide environment of a defined location. In such embodiments, the hybridization reactions, including hybridization kinetics and up-converting label can preferably act as a photophysical thermodynamic stability of hybridized polynucleotides. catalyst. The invention also provides methods, up-converting The invention provides methods for producing targeted labels, and compositions of labeled binding reagents for damage (e.g., catalysis) in chemical or biological materials, performing fluorescence-activated cell sorting (FACS) by wherein a probe is employed to localize a linked flow cytometry using excitation radiation that is in the up-converting label to a position near a targeted biological infrared portion of the spectrum and does not significantly structure that is bound by the probe. The localized damage cells. This provides a significant advantage over up-converting label is excited by one or more excitation 10 present FACS methods which rely on excitation illumination wavelengths and emits at a shorter wavelength which may in the ultraviolet portion of the spectrum, including wave be directly cytotoxic or genotoxic (e.g., by producing free lengths radicals such as superoxide, and/or by generating thymine damage which cells.

are known to produce DNA lesions and thymine dimers), or which may induce a local photolytic The invention also provides compositions comprising at chemical reaction to produce reactive chemical species in 5 least one fluorescent organic dye molecule attached to an the immediate vicinity of the label, and hence in the vicinity inorganic up-converting phosphor. The fluorescent organic of the targeted biological material. Thus, targeting probes dye molecule is selected from the group consisting of: labeled with one or more up-converting labels (e.g., an rhodamines, cyanines, xanthenes, acridines, oxazines, up-converting inorganic phosphor) may be used to produce porphyrins, and phthalocyanines, targeted damage to biological structures, such as cells, complexed with a heavy metal. Theand may optionally be tissues, neoplasms, vasculature, or other anatomical or his may be adsorbed to the inorganic up-convertingorganic fluorescent dye phosphor tological structures. crystal and/or may be covalently attached to a coated Embodiments of the present invention also include inorganic up-converting phosphor, a derivatized vitrocer up-converting phosphors which can also be excited by an amic up-converting phosphor, or a microencapsulated inor electron beam or other beam of energetic radiation of 25 ganic up-converting phosphor. Frequently, covalent conju sufficient energy and are cathodoluminescent. Such gation between the up-converting inorganic phosphor electron-stimulated labels afford novel advantages in elimi particles and proteins (e.g., avidin, immunoglobulin) can be nating background in ultrasensitive biomolecule detection methods. Typically, stimulation of the up-converting phos accomplished with heterobifunctional crosslinkers. phor with at least two electrons is employed to generate a 30 BRIEF DESCRIPTION OF THE DRAWINGS visible-light or UV band emission.

The invention also provides for the simultaneous detec FIG. 1 is an optical and electronic block diagram illus tion of multiple target species by exploiting the multiphoton trating representative apparatus for performing diagnostics excitation and subsequent background-free fluorescence on a sample according to the present invention; detection of several up-converting phosphors or 35 FIG. 2A shows apparatus for implementing phase sensi up-converting dyes. In one embodiment, several phosphors/ tive detection in the context of a single channel; dyes are selected which have overlapping absorption bands FIG. 2B shows apparatus where first and second laser which allow simultaneous excitation at one wavelength (or diodes are modulated by signals from waveform generators; in a narrow bandwidth), but which vary in emission char FIG. 3 shows apparatus for performing gated detection; acteristics such that each probe-label species is endowed FIG. 4 shows an apparatus for performing diagnostics on with a distinguishable fluorescent "fingerprint.” By using a sample various methods and devices, the presence and concentra centered atusing A1 first and second reporters excitation bands and A2, respectively, and having overlapping tion of each of the phosphors or dyes can be determined. emission bands near A:

The invention also provides biochemical assay methods 45 FIG. 5A, SB, SC show schematically energy state transi for determining the presence and concentration of one or tions in multi-photon excitation schemes. more analytes, typically in solution. The assay methods employ compositions of probes labeled with up-converting FIG. 6 shows a miniaturized instrument using a hand-held phosphors and/or up-converting dyes and apparatus for probe;

magnetically and/or optically trapping particles that com FIG. 7A shows the use of a charge-coupled device (CCD) prise the analyte and the labeled probe. In one embodiment, array used to detect emissions from a large plurality of a sandwich assay is performed, wherein an immobilized binding sites;

probe, immobilized on a particle, binds to a predetermined FIG.7B shows the CCD array used in conjunction with a analyte, producing an immobilization of the bound analyte lens array;

on the particle; a second probe, labeled with an FIG. 8 shows an embodiment using optical trapping; up-converting label can then bind to the bound analyte to 55 FIG. 9 shows schematically dye coating and encapsula produce a bound sandwich complex containing an tion of an up-converting phosphor particle;

up-converting label bound to a particle. By combining different probe-label combinations, particles of various ingFIG. 10 shows schematically an apparatus for determin sizes, colors, and/or shapes with distinct immobilized probe properties of velocity particle a target;

and hydrodynamic or aerodynamic (s), and/or various excitation wavelengths, it is possible to FIG. 11 is a phosphor emission spectrum of sodium perform multiple assays essentially simultaneously or con temporaneously. This multiplex advantage affords detection yttrium fluoride-ytterbium/erbium up-converting phosphor and quantitation of multiple analyte species in a single with an excitation laser source at a wavelength maximum of sample. The assay methods are also useful for monitoring 977.2 mm; emission maximum is about 541.0 nmi; the progress of a reaction, such as a physical, chemical, 65 FIG. 12 is an excitation scan of the sodium yttrium biochemical, or immunological reaction, including binding fluoride-ytterbium/erbium phosphor excitation spectrum, reactions. For example, the invention may be used to moni with emission collection window set at 541.0 nmi;

Page 38

FIG. 13 is a time-decay measurement of the phosphor FIG. 31B is a detailed view of a small section of the luminescence at 541 nm after termination of excitation device shown in FIG. 31A.

illumination for sodium yttrium fluoride-ytterbium/erbium; FIG. 32 is an emission spectrum for up-conversion from FIG. 14 shows the phosphor emission intensity as a neodymium chelated in EDTA.

function of excitation illumination intensity for a sodium 5 yttrium fluoride-ytterbium/erbium phosphor; DESCRIPTION OF SPECIFIC EMBODIMENTS FIG. 15 shows effective single-photon phosphorescence Definitions cross-section for 0.3 um particles of Na(YosYbo. Eros)F Unless defined otherwise, all technical and scientific following excitation with 200 W/cm at 970 nm. terms used herein have the same meaning as commonly FIG. 16 shows size-dependence of phosphorescence O understood by one of ordinary skill in the art to which this cross-section for Na(YoYboEroos)f particles. invention belongs. Although any methods and materials FIG. 17A shows a fluorescence scan of an up-converting similar or equivalent to those described herein can be used phosphor reporter in Hepes-buffered saline induced by exci in the practice or testing of the present invention, the tation with a 970-nm laser source;

15 preferred methods and materials are described. For purposes

FIG. 17B shows a fluorescence spectrum scan of an of the present invention, the following terms are defined up-converting phosphor reporter coated with streptavidin in below.

Hepes-buffered saline induced by excitation with a 970-nm As used herein, "label" refers to a chemical substituent laser source; that produces, under appropriate excitation conditions, a FIG. 18A shows an excitation spectrum scan of an detectable optical signal. The optical signal produced by an up-converting phosphor reporter in Hepes-buffered saline excited label is typically electromagnetic radiation in the with monochromatic detection of emission at 54.1 nm, near-infrared, visible, or ultraviolet portions of the spectrum. FIG. 18B shows an excitation spectrum scan of an The labels of the invention are up-converting labels, which up-converting phosphor reporter coated with streptavidin in means that the chemical substituent absorbs at least two Hepes-buffered saline with monochromatic detection of 25 photons at an excitation frequency and subsequently emits emission at 541 nm, electromagnetic energy at an emission frequency higher than FIG. 19 shows the integrated signal obtained from the excitation frequency. Thus, there is generally a signifi samples of (YosYbooseros)OS showing the relationship cant Stokes shift between the original excitation frequency between phosphor concentration and up-converted signal; and the final emission frequency. A label is generally FIG. 20 shows schematically one embodiment of an 30 attached to a probe to serve as a reporter that indicates the sandwich immunoassay for detecting an analyte in a solution presence and/or location of probe. The invention encom by binding the analyte (e.g. an antigen target) to a biotiny passes organic and inorganic up-converting labels, but pref lated antibody and to an immobilized antibody, wherein the erably employs up-converting inorganic lanthanide phos analyte forms a sandwich complex immobilized on a solid phors as labels. Thus, a typical label of the invention is a substrate superparamagnetic microbead; and 35 submicron-size up-converting lanthanide phosphor particle. FIG.21 shows schematically detection and discrimination The label can alternatively comprise a lanthanide ion in a of two cell surface antigens with specific antibodies labeled chelate or cage compound.

with two phosphors with distinct phosphorescence charac As used herein, a "probe" refers to a binding component teristics. which binds preferentially to one or more targets (e.g., FIG. 22 shows a schematic of an apparatus for phase antigenic epitopes, polynucleotide sequences, macromo sensitive detection. lecular receptors) with an affinity sufficient to permit dis FIG. 23 show a schematic of a competitive homogeneous crimination of labeled probe bound to target from nonspe assay using phosphors as labels and fiber optic illumination cifically bound labeled probe (i.e., background). Generally, at a capture surface. the probe-target binding is a non-covalent interaction with a FIG. 24 show a schematic of a competitive homogeneous 45 binding affinity (K) of at least about 1x10''', preferably antigen capture assay using phosphors as labels and a with at least about 1x10'M', and more preferably with an convergent illumination beam focused on the capture sur affinity of at least about 1x10'M' or greater. Antibodies face. typically have a binding affinity for cognate antigen of about FIG. 25 shows a schematic of a homogeneous immuno 1x10'M' or more. For example but not limitation, probes of the invention include: antibodies, polypeptide hormones, precipitation assay using phosphors as labels and a conver polynucleotides, gent illumination beam focused on the capture surface tein A, receptorstreptavidin, Staphlyococcus aureus pro ligands (e.g., steroid or polypeptide wherein the capture surface collects immunoprecipitates.

FIG. 26 shows a block diagram of one embodiment of hormones), leucine zipper polypeptides, lectins, antigens (polypeptide, carbohydrate, nucleic acid, and hapten apparatus for carrying out the present invention on a sample 55 epitopes), and others.

using a microscope.

FIG. 27 is a block diagram of a microtiter plate reader for probe" referherein,

As used

a "probe-label conjugate" and a "labeled combination comprising a label attached to use with the present invention.

a probe. In certain embodiments,

FIG. 28 is an illustration of the data for upconverting substituent may be attached to a probe. more than one label phosphors in three test wells. Alternatively, in FIG. 29 is a schematic view of a second embodiment of some embodiments more than one probe may be attached to a hand-held probe for carrying out the present invention. a label (e.g., multiple antibody molecules may be attached to a submicron-size inorganic up-converting phosphor bead).

FIG. 30 illustrates a three channel configuration using Various attachment chemistries can be employed to link a interference filters. label to a probe, including, but not limited to, the formation FIG. 31A is an illustration of an embodiment of the 65 of: covalent bonds, hydrogen bonds, ionic bonds, electro invention in which a diode laser array F1 and a detector static interactions, and surface tension (phase boundary) array F2 are combined in a single device. interactions. Attachment of label can also involve incorpo

Page 39

ration of the label into or onto microspheres, microparticles, probe polynucleotide may be a portion of a cDNA corre immunobeads, and superparamagnetic magnetic beads sponding to a particular mRNA sequence, a portion of a (Polysciences, Inc., Warrington, Pa.; Bangs Laboratories, genomic clone, a synthetic oligonucleotide having sufficient Inc. 979 Keystone Way, Carmel, Ind. 46032). For example, sequence homology to a known target sequence (e.g. a inorganic up-converting phosphor particles can be encapsu telomere repeat TTAGGG or an Alu repetitive sequence) for lated in microspheres that are composed of polymer material specific hybridization, a transcribed RNA (e.g., from an SP6 that is essentially transparent or translucent in the wave cloning vector insert), or a polyamide nucleic acid (Nielsen length range(s) of the excitation and emitted electromagnetic et al. (1991) Science 254: 1497). Various target polynucle radiation (U.S. Pat. No. 5.132.242, incorporated herein by otides may be detected by hybridization of a labeled probe reference). Such microspheres can be functionalized by 10 polynucleotide to the target sequence(s). For example but surface derivatization with one or more reactive groups (e.g., not limitation, target polynucleotides may be: genomic carboxylate, amino, hydroxylate, or polyacrolein) for cova sequences (e.g., structural genes, chromosomal repeated lent attachment to a probe, such as a protein. Probe-label sequences, regulatory sequences, etc.), RNA (e.g., mRNA, conjugates can also comprise a phosphor chelate. hnRNA, rRNA, etc.), pathogen sequences (e.g. viral or As used herein, the term "target” and "target analyte” 15 mycoplasmal DNA or RNA sequences), or transgene refer to the object(s) that is/are assayed for by the methods sequences.

of the invention. For example but not limitation, targets can "Specific hybridization” is defined herein as the formation comprise polypeptides (e.g., h0H, insulin, albumin), glyco of hybrids between a probe polynucleotide and a target proteins (e.g., immunoglobulins, thrombomodulin, polynucleotide, wherein the probe polynucleotide preferen y-glutamyltranspeptidase; Goodspeed et al. (1989) Gene 76: 20 tially hybridizes to the target DNA such that, for example, at 1), lipoproteins, viruses, microorganisms (e.g., pathogenic least one discrete band can be identified on a Southern blot bacteria, yeasts), polynucleotides (e.g., cellular genomic of DNA prepared from eukaryotic cells that contain the DNA, RNA in a fixed histological specimen for in situ target polynucleotide sequence, and/or a probe polynucle hybridization. DNA or RNA immobilized on a nylon or otide in an intact nucleus localizes to a discrete chromo nitrocellulose membrane, viral DNA or RNA in a tissue or 25 somal location characteristic of a unique or repetitive biological fluid), and pharmaceuticals (i.e., prescribed or sequence. In some instances, a target sequence may be over-the-counter drugs listed in the Physicians Drug Refer present in more than one target polynucleotide species (e.g., ence and/or Merck Manual, or illegal substances such as a particular target sequence may occur in multiple members intoxicants or anabolic steroids). of a gene family or in a known repetitive sequence). It is As used herein, the term "antibody" refers to a protein evident that optimal hybridization conditions will vary consisting of one or more polypeptides substantially depending upon the sequence composition and length(s) of encoded by immunoglobulin genes. The recognized immu the targeting polynucleotide(s) and target(s), and the experi noglobulin genes include the kappa, lambda, alpha, gamma mental method selected by the practitioner. Various guide (IgG, IgG, IgG, IgG), delta, epsilon and mu constant lines may be used to select appropriate hybridization con region genes, as well as the myriad immunoglobulin vari 35 ditions (see, Maniatis et al., Molecular Cloning: A able region genes. Full-length immunoglobulin "light Laboratory Manual (1989), 2nd Ed., Cold Spring Harbor, chains” (about 25Kd or 214 amino acids) are encoded by a N.Y. and Berger and Kimmel, Methods in Enzymology, variable region gene at the NH2-terminus (about 110 amino Volume 152 Guide to Molecular Cloning Techniques (1987), acids) and a kappa or lambda constant region gene at the Academic Press, Inc., San Diego, Calif., Dunn et al. (1989) COOH-terminus. Full-length immunoglobulin "heavy J. Biol. Chem. 264: 13057 and Goodspeed et al. (1989) Gene chains” (about 50 Kd or 446 amino acids), are similarly 76: 1.

encoded by a variable region gene (about 116 amino acids) As used herein, the term "label excitation wavelength" and one of the other aforementioned constant region genes, refers to an electromagnetic radiation wavelength that, when e.g., gamma (encoding about 330 amino acids). One form of absorbed by an up-converting label, produces a detectable immunoglobulin constitutes the basic structural unit of an 45 fluorescent emission from the up-converting label, wherein antibody. This form is a tetramer and consists of two the fluorescent emission is of a shorter wavelength (i.e., identical pairs of immunoglobulin chains, each pair having higher frequency radiation) that the label excitation wave one light and one heavy. chain. In each pair, the light and length. As used herein, the term "label emission wave heavy chain variable regions are together responsible for length” refers to a wavelength that is emitted from an binding to an antigen, and the constant regions are respon up-converting label subsequent to, or contemporaneously sible for the antibody effector functions. in addition to with, illumination of the up-converting label with one or antibodies, immunoaiobulins may exist in a variety of other more excitation wavelengths; label emission wavelengths of forms including, for example, Fv, Fab, and F(ab'), as well up-converting labels are shorter (i.e., higher frequency as bifunctional hybrid antibodies (e.g., Lanzavecchia et al., radiation) than the corresponding excitation wavelengths. Eur, J. Immunol. 17, 105 (1987)) and in single chains (e.g., 55 Both label excitation wavelengths and label emission wave Huston et al., Proc. Natl. Acad. Sci. U.S.A., 85,5879-5883 lengths are characteristic to individual up-converting label (1988) and Bird et al., Science, 242, 423-426 (1988)). (See, species, and are readily determined by performing simple generally, Hood et al., "Immunology", Benjamin, N.Y., 2nd excitation and emission scans.

ed. (1984), and Hunkapiller and Hood, Nature, 323, 15-16 Invention Overview (1986)). Thus, not all immunoglobulins are antibodies. (See, The subject invention encompasses fluorescent labels that U.S. Ser. No. 07/634.278, which is incorporated herein by are excited by an excitation wavelength and subsequently reference, and Co et al. (1991) Proc. Natl. Acad. Sci. emit electromagnetic radiation at up-shifted frequencies (U.S.A.) 88: 2869, which is incorporated herein by (i.e., at higher frequencies than the excitation radiation). reference). In accordance with the present invention, labels compris As used herein, "probe polynucleotide" refers to a poly 65 ing up-converting inorganic phosphors and/or up-converting nucleotide that specifically hybridizes to a predetermined organic dyes are provided for various applications. The target polynucleotide. For example but not limitation, a up-converting labels of the invention may be attached to one

Page 40

or more probe(s) to serve as a reporter (i.e., a detectable invention. Such conversion efficiency data may be obtained marker) of the location of the probe(s). The up-converting from avaliable sources e.g., handbooks and published labels can be attached to various probes, such as antibodies, references) or may be obtained by generating a standard streptavidin, protein A, polypeptide ligands of cellular ization curve measuring quantum conversion efficiency as a receptors, polynucleotide probes, drugs, antigens, toxins, 5 function of particle size. In some applications, such as those and others. Attachment of the up-converting label to the requiring highly sensitive detection of small phosphor probe can be accomplished using various linkage particles, infrared laser diodes are preferably selected as an chemistries, depending upon the nature of the specific probe. excitation source.

For example but not limitation, microcrystalline Although the properties of the up-converting phosphors up-converting lanthanide phosphor particles may be coated O will be described in detail in a later section, it is useful to with a polycarboxylic acid (e.g. AdditionXW 330, Hoechst, outline the basic mechanisms involved. Up-conversion has Frankfurt, Germany) during milling and various proteins been found to occur in certain materials containing rare (e.g., immunoglobulin, streptavidin or protein A) can be earth ions in certain crystal materials. For example, ytter physically adsorbed to the surface of the phosphor particle bium and erbium act as an activator couple in a phosphor (Beverloo et al. (1991) op.cit., which is incorporated herein 15 host material such as barium-yttrium-fluoride. The ytter by reference). Alternatively, various inorganic phosphor bium ions act as the absorber, and transfer energy non coating techniques can be employed including, but not radiatively to excite the erbium ions. The emission is thus limited to: spray drying, plasma deposition, and derivatiza characteristic of the erbium ion's energy levels. tion with functional groups (e.g., -COOH, -NH2, Up-Converting Microcrystalline Phosphors -CONH2) attached by a silane coupling agent to -SiOH 20 Although the invention can be practiced with a variety of moieties coated on the phosphor particle or incorporated into up-converting inorganic phosphors, it is believed that the a vitroceramic phosphor particle comprising silicon oxide(s) preferred embodiment(s) employ one or more phosphors and up-converting phosphor compositions. Vitroceramic derived from one of several different phosphor host phosphor particles can be aminated with, for example, materials, each doped with at least one activator couple. aminopropyltriethoxysilane for the purpose of attaching 25 Suitable phosphor host materials include: sodium yttrium amino groups to the vitroceramic surface on linker fluoride (NaYF), lanthanum fluoride (LaF), lanthanum molecules, however other omega-functionalized silanes can oxysulfide, yttrium oxysulfide, yttrium fluoride (YF). be substituted to attach alternative functional groups. yttrium gallate, yttrium aluminum garnet, gadolinium fluo Probes, such as proteins or polynucleotides may then be ride (GdF), barium yttrium fluoride (BaYF BaYF), and directly attached to the vitroceramic phosphor by covalent 30 gadolinium oxysulfide. Suitable activator couples are linkage, for example through siloxane bonds or through selected from: ytterbium/erbium, ytterbium/thulium, and carbon-carbon bonds to linker molecules (e.g., organofunc ytterbium?holmium. Other activator couples suitable for tional silylating agents) that are covalently bonded to or up-conversion may also be used. By combination of these adsorbed to the surface of a phosphor particle. Covalent host materials with the activator couples, at least three conjugation between the up-converting inorganic phosphor 35 phosphors with at least three different emission spectra (red, particles and proteins (e.g., avidin, immunogiobulin) can be green, and blue visible light) are provided. Generally, the accomplished with homobifunctional, or preferably absorber is ytterbium and the emitting center can be selected heterobifunctional, crosslinkers. For example, surface slian from: erbium, holmium, terbium, and thulium; however, ization of the phosphors with tri(ethoxy)thiopropyl silane other up-converting phosphors of the invention may contain leaves a phosphor surface with a thiol functionality to which other absorbers and/or emitters. The molar ratio of absorber: a protein (e.g., antibody) or any compound containing a emitting center is typically at least about 1:1, more usually primary amine can be grafted using conventional

N-succinimidyl(4-iodoacetyl)aminobenzoate (SIAB) chem at least about 3:1 to 5:1, preferably at least about 8:1 to 10:1, more preferably at least about 11:1 to 20:1, and typically less istry (Weltman et al. (1983). Other silanization and cross than about 250:1, usually less than about 100:1, and more linking methods compatible with the inorganic phosphors 45 usually less than about 50:1 to 25:1, although various ratios may be used at the discretion of the practitioner. may be selected by the practitioner on the basis of desired Microcrystalline up-converting phosphor particles are characteristics (e.g., chemical properties, manufacturing typically smaller than about 2 microns in diameter, prefer efficiency, absorption cross-section, excitation and emission ably less than about 1 micron in diameter (i.e., submicron), wavelengths, quantum efficiency, or other considerations). and more preferably are 0.1 to 0.3 microns or less in 50 The ratio(s) chosen will generally also depend upon the diameter. It is generally most preferred that the phosphor particular absorber-emitter couple(s) selected, and can be particles are as Small as possible while retaining sufficient calculated from reference values in accordance with the quantum conversion efficiency to produce a detectable sig desired characteristics.

nal; however, for any particular application, the size of the The optimum ratio of absorber (e.g., ytterbium) to the phosphor particle(s) to be used should be selected at the 55 emitting center (e.g., erbium, thulium, or holmium) varies, discretion of the practitioner. For instance, some applica depending upon the specific absorberdemitter couple. For tions (e.g., detection of a non-abundant cell surface antigen) example, the absorber:emitter ratio for Yb:Er couples is may require a highly sensitive phosphor label that need not typically in the range of about20:1 to about 100:1, whereas be small but must have high conversion efficiency and/or the absorber:emitter ratio for Yb:Tm and Yb:Ho couples is absorption cross-section, while other applications (e.g., typically in the range of about 500:1 to about 2000:1. These detection of an abundant nuclear antigen in a permeablized different ratios are attributable to the different matching cell) may require a very small phosphor particle that can energy levels of the Er, Tm, or Ho with respect to the Yb readily diffuse and penetrate subcellular structures, but level in the crystal. For most applications, up-converting which need not have high conversion efficiency. Therefore, phosphors may conveniently comprise about 10-30% Yb the optimal size of inorganic phosphor particle is application 65 and either: about 1-2% Er, about 0.1-0.05% Ho, or about dependent and is selected by the practitioner on the basis of 0.1-0.05% Tm, although other formulations may be quantum efficiency data for the various phosphors of the employed.

Page 41

Some embodiments of the invention employ inorganic the quantum conversion efficiency is low, intense laser phosphors that are optimally excited by infrared radiation of illumination can increase signal and decrease detection about 950 to 1000 nm, preferably about 960 to 980 nm. For times. Alternatively, some applications of the invention may example but not limitation, a microcrystalline inorganic quantumphosphor require compositions that have inherently low conversion efficiencies (e.g., low doping levels of phosphor of the formula YF:YbooBroo exhibits a lumi activator nescence intensity maximum at an excitation wavelength of istics (e.g.,couple), but which have other desirable character manufacturing efficiency, ease of derivatization, about 980 nm. Inorganic phosphors of the invention typi etc.); such low efficiency up-converting cally have emission maxima that are in the visible range. For erably excited with laser illumination atphosphorsa are pref frequency at or example, specific activator couples have characteristic emis near (i.e., within about 25 to 75 nm) an absorption maximum sion spectra: ytterbium-erbium couples have emission 10 of the material. The fact that no other light is generated in the maxima in the red or green portions of the visible spectrum, system other than from the up-converting phosphor allows depending upon the phosphor host; ytterbium-holmium for extremely sensitive signal detection, particularly when couples generally emit maximally in the green portion, intense laser illumination is used as the source of excitation ytterbium-thulium typically have an emission maximum in radiation. Thus, the unique property of up-conversion of the blue range, and ytterbiumterbium usually emit maxi photon energy by up-converting phosphors makes possible

mally in the green range. For example, YosYbo. Broof the detection of very small particles of microcrystalline emits maximally in the green portion of the spectrum. inorganic phosphors. For practical implementation of phos Although up-converting inorganic phosphor crystals of phors as ultrasensitive reporters, particularly as intracellular various formulae are suitable for use in the invention, the following formulae, provided for example and not to limit 20 as small as practicable that reporters, it is essential the grain size of the phosphor be (typically less than about 0.3 to 0.1 the invention, are generally suitable: pum), for which laser-excited up-converting phosphors are Na(YYbEF: x is 0.7 to 09, y is 0.09 to 0.29, and z is 0.05 to well-suited.

0.01; For example, various phosphor material compositions 25 capable of up-conversion are suitable for use in the inven

Na(Y.YbHoF: x is 0.7 to 09, y is 0.0995 to 0.2995, and z is tion are shown in Table I.

Na(Y.YbTm)F: x is 0.7 to 09, y is 0.0995 to 0.2995, and z is TABLE I

Phosphor Material Compositions

(YYbBr)OS: x is 0.7 to 09, y is 0.05 to 0.12; z is 0.05 to Host Material Absorber on Emitter on Color

(YossybooBros)O is a relatively efficient up-converting phos Oxysulfides (OS) phor material. YOS Ytterbium Erbium Green

For exemplification, but not to limit the invention, 35 LaOS Ytterbium Holmium Green ytterbium(Yb)-erbium(Er)-doped yttrium oxysulfides lumi Oxyhalides (OX) nesce in the green after excitation at 950 nm. These are YOF Ytterbium Thulium Blue non-linear phosphors, in that the ytterbium acts as an YOCl, Yterbium Terbium Green "antenna" (absorber) for two 950 nm photons and transfers Fluorides (F) its energy to erbium which acts as an emitter (activator). The critical grain size of the phosphor is given by the quantum YF, Ytterbium Erbium Red yield for green emission and the doping level of both Yb and GdF, Ytterbium Erbium Green

Er, which is generally in the range of about 1 to 10 percent, NaYF, Ytterbium Thulium Blue more usually in the range of about 2 to 5 percent. A typical BaYF. Ytterbium Thuium Blue Yb:Er phosphor crystal comprises about 10-30% Yb and 45 BaYF Ytterbiun Terbium Green about 1-2% Er. Thus, a phosphor grain containing several Gallates (Ga.0) thousand formula units ensures the emission of at least one YGaO Ytterbiun Erbium Red or more photons during a typical laser irradiation time. YGasol Ytterbium Erbium Green However, the nonlinear relationship between absorption and Silicates (SisO) emission indicates that intense illumination at the excitation 50

wavelength(s) may be necessary to obtain satisfactory signal YSiO, Ytterbium Thulium Blue in embodiments employing very small phosphor particles (i.e., less than about 0.3 m). Additionally, it is usually desirable to increase the doping levels of activatorfemitter In addition to the materials shown in Table I and variations couples for producing very small phosphor particles so as to 55 thereof, aluminates, phosphates, and vanadates can be suit maximize quantum conversion efficiency. able phosphor host materials. In general, when silicates are Inorganic microcrystalline phosphors with rare earth acti used as a host material, the conversion efficiency is rela vators generally have narrow absorption and line emission tively low. In certain uses, hybrid up-converting phosphor spectra. The line emission spectra are due to f-f transitions crystals may be made (e.g., combining one or more host material and/or one or more absorber ion and/or one or more within the rare earth ion. These are shielded internal tran sitions which result in narrow line emission. emitter ion).

In certain applications, such as where highly sensitive Exemplary up-converting phosphors excited at about 980 detection is required, intense illumination can be provided nm include, but are not limited to: YosYbosBroo)P; by commercially available sources, such as infrared laser Yo..s 7Ybo.13T mo.o.o.1)F3; Yo soYbo.19s Hoo.oo2)F3; sources (e.g., continuous wave (CW) or pulsed semiconduc 65 Gido.soYbo.1s Ero.oz.) F3; Gido.87Ybo.13T mooo..)F3; tor laser diodes). For example, in applications where the Gido...so Ybo.19s Hoo-oo:2)Fa; Yossyboos Eroos)2O2S: microcrystalline phosphor particle must be very small and Yo.87Ybo.13Tmolool)2O2S: Yo so Ybo.198Hoo-oo:2)2O2S:

Page 42

Gdoss Ybooseroos)2O2S: Gidos, Ybo.13Tmooo)2O2S: faceting of the spherical particles which can generate aggre GdosoyborgsHoooo2)2O2S. gate formation. Faceting can be substantially reduced by Exemplary up-converting phosphors excited at about converting the small spherical particles of the oxide or 1500 nm include, but are not limited to: YoFro)OS; hydroxy carbonate precursor to the oxysulfide phase by Preparation of Inorganico Phosphor Labels

Gdooroos)2O2S. including a polysulfide flux for annealing. Using this technique,

Techniques and methods for manufacture of inorganic 0.4 um diameter highly efficient oxysulfide particles in the 0.3 to phosphors has been described in the art. Up-converting water. Frequently,range were prepared as a dispersion in sonication can be used to produce a phosphor crystals can be manufactured by those of ordinary monodisperse mixture of discrete spherical particles. After skill in the art by various published methods, including but 10 fractionation and coating, these particles can be used as not limited to the following: Yocom et al. (1971) Metallur. up-converting reporters. Furthermore, this general prepara gical Transactions 2: 763; Kano et al. (1972) J. Electro chem, Soc., p. 1561;Wittke et al. (1972).J. Appl. Physics 43: tive procedure is suitable for preparing much smaller phos 595; Van Uitert et al. (1969) Mat. Res. Bull. 4: 381; which phor particles (e.g., 0.1 pm diameter or smaller), which may are incorporated herein by reference. Other references which be advantageous for various assay formats. may be referred to are: Jouart J. P. and Mary G. (1990) J. 15 Frequently, such as with phosphors having an oxysulfide Luminescence 46:39; McPherson G. L. and Meyerson S. L. host material, the phosphor particles are preferably dis (1991) Chem. Phys. Lett. (April) p. 325; Oomen et al. (1990) persed in a polar solvent, such as acetone or DMSO and the J. Luminescence 46:353; NIH and Rand SC (1991) Optics like, to generate a substantially monodisperse emulsion Lett, 16 (September); McFarlane R. A. (1991) Optics Lett. (e.g., for a stock solution). Aliquots of the monodisperse 16 (September); Koch et al. (1990) Appl. Phys. Lett. 56: stock solution may be further diluted into an aqueous 1083; Silversmith et al. (1987) Appl. Phys. Lett. 51: 1977; solution (e.g., a solution of avidin in buffered water or Lenth W. and McFarlane R. M. (1990).J. Luminescence 45: buffered saline).

346; Hirao et al. (1991) J. Non-crystalline Solids 135: 90; It was found that washing phosphors in acetone or DMSO McFarlane et al. (1988) Appl. Phys. Lett. 52: 1300, incor improved suspendability of inorganic phosphor particles in porated herein by reference). 25 water. In particular, the phosphor particles prepared with In general, inorganic phosphor particles are milled to a polysulfide flux are preferably resuspended and washed in desired average particle size and distribution by conven hot DMSO and heated for about an hour in a steam bath then tional milling methods known in the art, including milling in allowed to cool to room temperature under continuous a conventional barrel mill with zirconia and/or alumina balls agitation. The phosphor particles may be pre-washed with for periods of up to about 48 hours or longer. Phosphor 30 acetone (typically heated to boiling) prior to placing the particles used in binding assays are typically about 3.0 to particles in the DMSO. Hot DMSO-treated phosphors were 0.01 min diameter (or along the long axis if non-spherical), found to be reasonably hydrophilic and form stable suspen more usually about 2.0 to 0.1 m in size, and more conve sions. A Microfluidizer M (Microfluidics Corp.) can be used niently about 1.0 to 0.3 m in size, although phosphor to further improve the dispersion of particles in the mixture. particles larger or smaller than these dimensions may be 35 DMSO-phosphor suspensions can be easily mixed with preferred for certain embodiments. Phosphor particle size is water, preferably with small amounts of surfactant present. selected by the practitioner on the basis of the desired In general, polysaccharides (e.g., guar gum, xanthan gum, characteristics and in accordance with the guidelines pro gum arabic, alginate, guaiac gum) can be used to promote vided herein. Fractions having a particular particle size deaggregation of particles. In a variation, particles are range may be prepared by sedimentation, generally over an washed in hot DMSO and serially diluted into a 0.1% extended period (i.e., a day or more) with removal or the aqueous gum arabic solution, which appears to virtually desired size range fraction after the appropriate sedimenta eliminate water dispersion problems of phosphors. tion time. The sedimentation process may be monitored, Resuspended phosphors in organic solvent, such as such as with a Horiba Particle Analyzer. DMSO, are typically allowed to settle for a suitable period However, milling crystalline materials has several weak 45 (e.g., about 1-3 days), and the supernatant which is typically nesses. With milling, the particle morphology is not turbid is used for subsequent conjugation.

uniform, as milled particles result from random fracture of LudoxTM is a colloidal silica dispersion in water with a larger crystalline particles. Since the sensitivity of a detec Small amount of organic material (e.g., formaldehyde. tion assay using up-converting inorganic phosphors depends glycols) and a small amount of alkalimetal. LudoxTM and its on the ability to distinguish between bound and unbound 50 equivalents can be used to coat up-converting phosphor phosphor particles, it is preferable that the particles be of particles which can subsequently be fired to form a ceramic identical size and morphology. Size, weight, and morphol silica coating which cannot be removed from the phosphor ogy of up-converting microcrystalline phosphor particles particles, but which can be readily silanized with organo can affect the number of potential binding sites per particle functional silanes (containing thiol, primary amine, and and thus the potential strength of particle binding to reporter SS carboxylic acid functionalities) using standard silanization and/or analyte. Monodisperse submicron spherical particles chemistries (Arkles, B., in: Silicon Compounds: Register of uniform size can be generated by homogeneous precipi and Review; 5th Edition (1991); Anderson, R. G. Larson, G. tation reactions at high dilutions. For example, small yttrium L., and Smith, C., eds.; p. 59-64, Huls America, Piscataway, hydroxy carbonate particles are formed by the hydrolysis of N.J.).

urea in a dilute yttrium solution. Similarly, up-converting Phosphor particles can be coated or treated with surface inorganic phosphors can be prepared by homogeneous pre active agents (e.g., anionic surfactants such as Aerosol OT) cipitation reactions in dilute conditions. For example, during the milling process or after milling is completed. For (Yoss YboosBroos),0s was prepared as monodisperse example, particles may be coated with a polycarboxylic acid spherical particles in the submicron size range by precipi (e.g., Addition XW 330, Hoechst, Frankfurt, Germany or tation. 65 Tamol, see Beverloo et al. (1992) op.cit.) during milling to However, after precipitation it is typically necessary to produce a stable aqueous suspension of phosphor particles, anneal the oxide in air at about 1500° C., which can cause typically at about pH 6-8. The pH of an aqueous solution of

Page 43

phosphor particles can be adjusted by addition of a suitable binding reagents, either directly or indirectly, for use in buffer and titration with acid or base to the desired pH range. binding assays to detect and quantitate the presence of Depending upon the chemical nature of the coating, some analyte(s) in a sample. Binding reagents are labeled directly minor loss in conversion efficiency of the phosphor may by attachment to up-converting reporters (e.g., surface occur as a result of coating, however the power available in adsorption, covalent linkage). Binding reagents which can a laser excitation source can compensate for such reduction be directly labeled include, but are not limited to: primary in conversion efficiency and ensure adequate phosphor emis antibodies (i.e., which bind to a target analyte), secondary S.O. antibodies (i.e., which bind to a primary antibody or pros In general, preparation of inorganic phosphor particles thetic group, such as biotin or digoxygenin), Staphlococcus and linkage to binding reagents is performed essentially as 10 aureus Protein A, polynucleotides, streptavidin, and receptor described in Beverloo et al. (1992) op.cit., and Tanke U.S. ligands. Binding reagents can also be indirectly labeled; Pat. No. 5,043,265. Alternatively, a water-insoluble poly thus, a primary antibody (e.g., a rabbit anti-erb-B antibody) functional polymer which exhibits glass and melt transition can be indirectly labeled by noncovalent binding to a temperatures well above room temperature can be used to directly labeled second antibody (e.g., a goat anti-rabbit coat the up-converting phosphors in a nonaqueous medium.

For example, such polymer functionalities include: carboxy 15 antibody linked to an up-converting inorganic phosphor). lic acids (e.g., 5% acrylic acid/95% methyl acrylate conducted indetection

Quantitative of the analyte-probe complex may be copolymer), amine (e.g., 5% aminoethyl acrylate/95% assay for eachconjunction probe with proper calibration of the employed. A probe is conveniently methyl acrylate copolymer) reducible sulfonates (e.g., 5% sulfonated polystyrene), and aldehydes (e.g., polysaccharide detected under saturating excitation conditions using, for copolymers). The phosphor particles are coated with water 20 example, a laser source or focused photodiode source for insoluble polyfunctional polymers by coacervative encap excitation illumination.

sulation in nonaqueous media, washed, and transferred to a Specific binding assays are commonly divided into homo suitable aqueous buffer solution to conduct the heterobifunc geneous and heterogeneous assays. In a homogeneous assay, tional crosslinking to a protein (e.g., antibody) or polynucle the signal emitted by the bound labeled probe is different otide probe molecule. An advantage of using water 25 from the signal emitted by the unbound labeled probe, hence insoluble polymers is that the polymer microcapsule will not the two can be distinguished without the need for a physical migrate from the surface of the phosphor upon aging the separation step. In heterogeneous assays, the signal emitted encapsulated phosphors in an aqueous solution (i.e., improved reagent stability). Another advantage in using from the bound and unbound labeled probes is identical, hence the two must be physically separated in order to copolymers in which the encapsulating polymer is only 30 distinguish between them. The classical heterogeneous spe partially functionalized is that one can control the degree of cific binding assay is the radioimmunoassay (RIA) (Yallow et functionalization, and thus the number of biological probe al. (1978) Science 200: 1245, which is incorporated herein molecules which can be attached to a phosphor particle, on by reference). Other heterogeneous binding assays include average. Since the solubility and coacervative encapsulation the radioreceptor assay (Cuatrecasas et al. (1974) Ann. Rey. process will depend on the dominant nonfunctionalized 35 Biochem.

component of the copolymer, the functionalized copolymer Pat. No. 43: 109), the sandwich radioimmunoassay (U.S. 4,376,110, which is incorporated herein by ratio can be varied over a wide range to generate a range of potential crosslinking sites per phosphor, without having to reference), and the antibodyflectin sandwich assay (EP0166 substantially change the encapsulation process. 623, which is incorporated herein by reference). Heteroge A preferred functionalization method employs heterobi neous assays are usually preferred, and are generally more functional crosslinkers that can be made to link the biologi sensitive and reliable than homogeneous assays. cal macromolecule probe to the insoluble phosphor particle Whether a tissue extract is made or a biological fluid in three steps: (1) bind the crosslinker to the polymer coating sample is used, it is often desirable to dilute the sample in on the phosphor, (2) separate the unbound crosslinker from one or more diluents that do not substantially interfere with the coated phosphors, and (3) bind the biological macro 45 subsequent assay procedures. Generally, suitable diluents molecule to the washed, linked polymer-coated phosphor. are aqueous solutions containing a buffer system (e.g., 50 mM NaH2PO, or 5-100 mM Tris, pH4-pH10), non

This method prevents undesirable crosslinking interactions interfering between biological macromolecules and so reduces irrevers ionic species (5-500 mM KCl or NaCl, or ible aggregation as described by Tanke et al. Examples of sucrose), and optionally a nonionic detergent such as Tween. suitable heterobifunctional crosslinkers, polymer coating When the sample to be analyzed is affixed to a solid support, functionalities, and linkable biological macromolecules 50 itsupport is usually desirable to wash the sample and the solid with diluent prior to contacting with probe. The include, but are not limited to:

sample, either straight or diluted, is then analyzed for the diagnostic analyte.

Coating Heterobifunctional Biological In the general method of the invention, an analyte in a Functionality Crosslinker Macromolecule 55 sample is detected and quantified by contacting the sample carboxylate N-hydroxysuccinide Proteins (e.g., with a probe-label conjugate that specifically or preferen 1-ethyl-3-(3-dimethylamino Ab, avidin) tially binds to an analyte to form a bound complex, and then propyl)-carbodiimide (EDC) detecting the formation of bound complex, typically by primary amine N-5-azido-2-nitrobenzoyl All having 1 amine measuring the presence of label present in the bound com oxysuccimide (ANB-NOS) plexes. A probe-label conjugate can include a directly

aminobenzoate (SAB) labeled analyte-binding reagent (e.g., a primary antibody thiol (reduced N-succinimidyl (4-iodoacetyl) Proteins linked to an up-converting phosphor) and/or an indirectly sulfonate) aminobenzoate (SIAB) labeled analyte-binding reagent (e.g., a primary antibody that is detected by a labeled second antibody, or a biotiny

Binding Assays 65 lated polynucleotide that is detected by labeled streptavidin). Up-converting phosphors and up-converting organic dyes The bound complex(es) are typically isolated from unbound are used as reporters (i.e., detectable markers) to label probe-label conjugate(s) prior to detection of label, usually

Page 44

by incorporating at least one washing step, so as to remove usually performed by first incubating the sample with a background signal attributable to label present in unbound blocking or prehybridization solution, followed by incubat probe-label conjugate(s). Hence, it is usually desirable to ing the sample with probe under binding conditions for a incubate probe-label conjugate(s) with the analyte sample suitable binding period, followed by washing or otherwise under binding conditions for a suitable binding period. removing unbound probe, and finally by detecting the Binding conditions vary, depending upon the nature of the presence, quantity, and/or location of bound probe. The step probe-label conjugate, target analyte, and specific assay of detecting bound probe can be accomplished by detecting method. Thus, binding conditions will usually differ if the label, if the probe is directly labeled, or by incubating the probe is a polynucleotide used in an in situ hybridization, in bound complex(es) with a second binding reagent (e.g., a Northern or Southern blot, or in solution hybridization 10 streptavidin) that is labeled and which binds to the probe, assay. Binding conditions will also be different if the probe thus accomplishing indirect labeling of the probe. is an antibody used in an in situ histochemical staining Up-converting labels are attached to probe(s) or second method or a Western blot (Towbin et al. (1979) Proc. Natl. binding reagents that specifically or preferentially bind to Acad. Sci. (U.S.A.) 76: 4350, incorporated herein by probe(s) by any of the various methodologies discussed reference). In general, binding conditions are selected in 5 herein. Additionally, up-converting phosphor particles can accordance with the general binding methods known in the be encapsulated in microspheres and coated with a probe art. For example, but not for limitation, the following (e.g., a specific antigen or antibody) for use as a labeled binding conditions are provided for general guidance: probe in an immunodiagnostic assay or nucleic acid hybrid For antibody probes: ization assay to detect an analyte in a sample, such as the 10-200 mM Tris, pH 6-8; usually 100 mM Tris pH 7.5 20 presence of an antibody, virus, or antigen in a bloom serum 15-250 mM. NaCl; usually 150 mM NaCl sample, according to the method of Hari et al. (1990) 0.01-0.5 percent, by volume, Tween 20 Biotechniques 9: 342, which is incorporated herein by 1 percent bovine serum albumin reference. Microencapsulation of phosphor can be accom 4°-37° C.; usually 4° to 15° C. plished in several ways known in the art, including coating For polynucleotide probes: 25 the phosphor with a monomer solution and polymerizing the 3-10x SSC, pH 6-8; usually 5x SSC, pH 7.5 monomer to generate a polymer shell encasing the phosphor 0-50 percent deionized formamide particle. Phosphor particles embedded in a polymer coating, 1-10X Denhardt's solution such as a gel coating, can be functionalized (e.g., with amino 0-1 percent sodium dodecyl sulfate groups) for covalent attachment to a binding component. 30 Similarly, up-converting phosphor particles can be coated 10-200 pg/ml sheared denatured salmon sperm DNA with probe directly, either by surface adsorption, by multiple 20°-65°C., usually 37-45° C. for polynucleotide probes hydrogen bonding, by electrostatic interaction, by van der longer than 50 bp, usually 55-65 C. for shorter Waals binding, or by covalent linkage to a functional group oligonucleotide probes on a functionalized inorganic phosphor particle (e.g., a Additional examples of binding conditions for antibodies 35 vitroceramic phosphor), for example, by linking an amino and polynucleotides are provided in several sources, includ acid side-chain amine or carboxylate group of a probe ing: Maniatis et al., Molecular Cloning: A Laboratory protein to a carboxylate or amine group, respectively, on a Manual (1989), 2nd Ed., Cold Spring Harbor, N.Y. and functionalized phosphor particle.

Berger and Kimmel, Methods in Enzoymlogy, Volume 152 In certain embodiments, such as where stearic and/or Guide to Molecular Cloning Techniques (1987), Academic charge interference of a bulky up-converting phosphor par Press, Inc., San Diego, Calif.; Young and Davis (1983) Proc. ticle inhibits binding of the linked binding reagent to a Natl. Acad. Sci. (U.S.A.) 80: 1194, which are incorporated target, it is desirable to incorporate a molecular spacer herein by reference. When the probe is a receptor ligand, between the phosphor particle and the binding reagent. For such as IL-2, B-interferon, or other polypeptide hormones, example, a derivatized microencapsulated phosphor or vit cytokines, or lymphokines, suitable binding conditions gen 45 roceramic phosphor may be conjugated to a heterobifunc erally are those described in the art for performing the tional reagent having a -(CH2)- spacer, where n is respective receptor-ligand binding assay, usually an integer from about 2 to about 50, between Various examples of suitable binding conditions useful in terminal functional groups. Similarly, phosphors may be immunoassays and immunohistochemistry are discussed, directly derivatized with derivatizing agents (e.g., omega for example, in Harlow and Lane, Antibodies: A Laboratory 50 functionalized silanes) having long intramolecular spacer Manual, Cold Spring Harbor, N.Y. (1988), which is incor chains, wherein a functional group reactive with a desired porated herein by reference. In general, suitable binding binding reagent is separated from the surface of the phos conditions for immunological reactions include an aqueous phor by a spacer of usually at least about 15 A (i.e., the binding buffer containing a salt (e.g. 5-500 mM. NaCl or equivalent of about 10 -CH- straight-chain groups). In KCl), a buffer (e.g., Tris or phosphate buffer at pH 4-10), 55 some embodiments, labels are attached by spacer arms of and optionally a nonionic detergent (e.g., Tween). In some various lengths to reduce potential stearic hindrance. Mul embodiments, proteinase inhibitors or stabilizers may be tiple layers of spacer arms may also be used (e.g., multiple included. The binding reactions are conducted for a suitable layers of streptavidin-biotin linkages). binding period, which, for antibody reactions, are typically Multiple Analyte Detection at least about 1 to 5 minutes, preferably at least about 30 Since up-converting phosphors can be differentiated on minutes to several hours, although typically less than about the basis of the excitation and/or emission wavelength 24 hours, more preferably less than about a few hours or spectra, up-converting phosphors can be used to detect and less. Binding reactions (including washes) are typically discriminate multiple analyte targets, such as, for example, carried out a temperature range of about 0°C. to about 45° cell surface antigens or soluble macromolecules. C., preferably about 4° C. to about 20°-25° C. 65 For example, streptavidin, avidin, or another linker mac Binding assays, which include in situ hybridization, in romolecule (e.g., antidigoxigenin antibody) are attached, situ binding assays, and immunohistochemical staining, are respectively, to each of two different phosphors (for

Page 45

illustration, designated here as Phosphortf1 and Phosphorf2) different oligopeptides affixed to a solid support). One or which differ in their absorption and/or emission spectra so as more of the species of first binding component may bind to to facilitate discrimination of the two phosphors based on a particular analyte (e.g., a muscarinic receptor) in an absorption and/or emission wavelengths; e.g., one phosphor analyte solution that is in contact with the solid support. may emit in the blue and the other may emitin the green. For Binding of the analyte to one or more of the first binding example and not limitation, Na(YosoYbolisEroo)F emits component species may then be detected with a second predominantly in the green, and Na(Yo-YboTrnooo..)F binding component (e.g., an anti-muscarinic receptor emits predominantly in the blue, and thus these two phos antibody) labeled with an up-converting phosphor (either phors may be discriminated on the basis of their phospho directly or through a biotinylated secondary antibody). rescent emissions. Alternatively, two phosphors may pro 10 Solid substrates can be attached to a first binding com duce essentially similar emission spectra but may have ponent which can bind more than one distinct analyte (e.g., different excitation wavelengths which provide a basis for may be immunocrossreactive or polyspecific) and/or can be their discrimination in multiple analyte detection. A first attached to multiple first binding component species which binding component (e.g., an antibody) that binds specifically can bind multiple distinct analytes. Similarly, multiple sec to a first analyte species (e.g., a lymphocyte CD4 antigen) 15 ond binding component species with binding specificities and incorporates biotinyl moieties which may be bound by for particular analytes can be employed. When multiple streptavidin-Phosphorf1 conjugates can be used to quanti second binding component species are employed, it is typi tatively detect the presence of a first analyte in a sample cally desirable to label each second binding component (e.g., a serum sample) by measuring phosphorescence of species with a unique up-converting label that can be Phosphorf1 in analyte-binding component complexes. A distinguished on the basis of its absorption and/or emission second binding component (e.g., a probe polynucleotide) properties.

that binds specifically to a second analyte species (e.g., an It is possible to use different absorbers in combination HIV-1 sequence) and incorporates digoxygenin moieties with various emitters to produce a collection of phosphors (e.g., 11-UTP-digoxygenin) which may be bound by having several differentiable combinations of excitation and antidigoxigenin-Phosphorf2 conjugates can be used to 25 emission spectra. For example but not limitation, six differ quantitatively detect the presence of a second analyte in the entiable phosphors may be generated from two absorbers sample by measuring phosphorescence of Phosphorf2 in and three emitters. A first absorber, A, has an excitation analyte-binding component complexes. Thus, by simulta wavelength of A, a second absorber, A has an excitation neously or contemporaneously detecting the presence of wavelength of Wa, a first emitter, E, has an emission line multiple phosphor reporters having differentiable signal at W, a second emitter, E. has an emission line at , and characteristics, multiple analytes may be quantitatively a third emitter, E, has an emission line at A. The six detected in a single sample. phosphors may be differentiated and the signal from each Sandwich Binding Assays individually quantitated by illuminating the sample with an Up-converting phosphors labels can be used as reporters excitation wavelength and detecting separately the emit for sandwich binding assays (U.S. Pat. No. 4,376,110, which 35 ted radiation at Ae, We, and A and separately illuminat is incorporated herein by reference). For example, a mag ing the sample with A2 and detecting separately the netic bead, such as a superparamagnetic immunobead or emitted radiation at W, A, and W. Table II shows the functionalized magnetizable polymer particle (Polysciences, various absorber:emitter combinations and their excitation Inc., Warrington, Pa.), can serve as the solid substrate which and emission wavelengths.

has an immobilized first binding component (e.g., an antibody, a polynucleotide, or a lectin) that binds to a first TABLE I epitope (i.e., a binding locus: an antigenic determinant, Absorber:Emitter Combination Excitation A Emission A sugar moiety, chemical substituent, or nucleotide sequence) of an analyte. The analyte binds to the first binding com A1E1 Al AE ponent and also to a second binding component (e.g., an 45 A1:E2 AA1 AE2

antibody, a lectin, or a polynucleotide) which binds to a A2E1 AA2 AE1 second epitope of the analyte. Thus, the analyte bridges the A2E2 A2 E2 two binding components to form a sandwich complex which A2E3 AA2 E3 is immobilized with respect to the solid substrate. The second binding component typically has an attached or 50 Of course, additional absorber:emitter combinations are incorporated label, such as a biotinyl group which can be possible bound to a streptavidin-coated up-converting phosphor. labels. to provide more than six differentiable phosphor Alternatively, the second binding component can be linked It is also possible to utilize solid substrates of different directly to an up-converting phosphor, such as through a types which may be distinguished (e.g., by size, color, covalent linkage with a functionalized vitroceramic density, magnetic properties, shape, charge) so that a par

up-converting phosphor.

The sandwich complex comprises the first binding ticular species type of solid substrate is associated with a particular of first binding component.

component, an analyte, and the second binding component, For example and not limitation, the following three brief which is labeled, either directly or indirectly, with an examples are provided to explicate further possible appli up-converting reporter. The sandwich complex is thus cations of multiple analyte sandwich assay methods. immobilized on the solid substrate, although the solid sub strate itself may be mobile (e.g., a superparamagnetic bead Substrate Differentiation circulating in a sample slurry). The presence and amount of analyte(s) can be quantitatively measured by detecting the The following example describes the use of distinguish presence of up-converting reporter in sandwich complexes. 65 able substrate types to detect the presence of specific immu For example, a solid substrate may have a plurality of noglobulin idiotypes in a sample (e.g., a blood serum sample distinct species of first binding component (e.g., an array of taken from a patient) which can provide diagnostic infor

Page 46

mation about the immune status of a patient (e.g., is a patient isoform(s) having the Z epitope. If the emissions from two seroreactive with a particular antigen). phosphors are readily distinguishable, A and A may be Large superparamagnetic beads are conjugated to an identical. The standardized relative intensities of the two immunogenic Herpesvirus Type II envelope glycoprotein, phosphors provides a measure of the relative abundance of medium-sized superparamagnetic beads are conjugated to 5 the APP isoform(s) containing the Y or Z epitopes. HTV gp120 glycoprotein, and small superparamagnetic beads are conjugated to an immunogenic cytomegalovirus Phosphor and Substrate Differentiation envelope glycoprotein. A serum sample is taken from a The following example describes the use of differentiable patient and is incubated with a mixture of the superpara up-converting magnetic beads under binding conditions to permit specific 10 able substrate phosphorstypes to in conjunction with distinguish detect the presence and relative binding of immunoglobulins in the sample with the three abundance of particular T lymphocyte immobilized viral glycoprotein species. The superparamag blood sample taken from an individual. subpopulations Although

described netic beads are separated from the sample to remove non here with reference to detecting T cell subpopulations, specifically bound immunoglobulin and incubated with up-converting phosphor particles coated with Staphylococ 15 analyte multiplexing (i.e., detecting and/or characterizing multiple analytes in a sample by using various solid sub cus aureus Protein A, which binds to IgG, under binding strate types and/or up-converting phosphor labels) is conditions. Superparamagnetic beads having specifically bound IgG are thus labeled with the phosphor-Protein A believed to be a generally applicable method. Large superparamagnetic beads are conjugated to an conjugate. Large, medium, and small superparamagnetic anti-CD4 antibody, medium-sized superparamagnetic beads beads are then separately illuminated with phosphor exci 20 are conjugated tation electromagnetic radiation and time-gated emitted magneticbeads to anti-CD8 antibody, and small superpara are conjugated to an anti-CD28 antibody. An phosphorescence is detected. Background attributable to antibody that specifically binds to the CD2 antigen is labeled non-specific binding, if any, is determined and subtracted using internal standard beads (bovine serum albumin coated with an up-converting phosphor that has an excitation wave length A and emits in the red. An antibody that specifically superparamagnetic beads) and positive and negative control 25 binds to the CD45R antigenis labeled with an up-converting serum samples. The intensity of phosphorescence associated phosphor that has an excitation wavelength A and emits in with the large, medium, and small beads provides a measure of the amount of antibodies in the sample which are reactive the green. An antibody that specifically binds to the CDwó0 with the Herpesvirus Type II envelope glycoprotein, HTV antigen is labeled with an up-converting phosphor that has gp120 glycoprotein, and cytomegalovirus envelope 30 an excitation wavelength A and emits in the blue. glycoprotein, respectively. This information can be used to A blood (or serum, sputum, urine, feces, biopsy tissue, determine whether an individual patient has been infected etc.) sample is taken from a patient and is incubated with a with the HIV-1, human CMV, and/or Herpes SimplexType mixture of the superparamagnetic beads and phosphor I viruses. labeled antibodies under binding conditions to permit spe 35 cific binding of cells in the blood sample with the three

Phosphor Differentiation bead-immobilized antibody species and the three phosphor The following example describes the use of differentiable labeled antibody species. After antigen-antibody binding up-converting phosphors to detect the presence and relative occurs, the superparamagnetic beads are segregated and abundance of particular isoforms of human APP (amyloid examined, either sequentially or simultaneously, by illumi precursor protein) in a serum or brain biopsy sample. nation with A, 2 and A, and quantitative detection of red, Various isoforms of APP arise in the brain as a consequence green, and blue emissions, respectively. For example, the of alternative exon usage and/or alternative proteolytic pro intensity of A-induced red light emission associated with cessing pathways. Thus, although all APP isoforms may the large beads is a rough measure of the amount of cells share a common, hypothetical epitope (X), a particular APP having both CD4 and CD2 surface antigens and/or the isoform may have a unique epitope (Y), while another APP 45 relative abundance of those surface antigens (e.g., there may isoform has a unique epitope (Z). It is possible that the be very few CD4 cells that have CD2, but those few cells relative abundance of a particular APP isoform in a sample may have a large amount of CD2 antigen, and hence a large may be of predictive value or may be pathognomonic for CD2 phosphorescent signal). Similarly, the intensity of Alzheimer's Disease. A-induced green light associated with the large beads is a Superparamagnetic beads are conjugated to an antibody 50 rough measure of the amount of cells having both CD4 and that binds specifically to a common APP epitope (X) shared CD45R surface antigens and/or the relative abundance of by all isoforms. A specific antibody reactive with the unique those surface antigens in a sample.

Y epitope is labeled with Phosphor #1, which is excited by In this manner, an analyte sample, such as a blood sample, wavelength and emits in a wavelength spectrum centered can be "fingerprinted" for the presence and relative in the blue. A specific antibody reactive with the unique Z 55 distribution(s) (e.g., cosegregation and/or correlation) of epitope is labeled with Phosphor #2, which is excited by a various analyte species. Such an analyte fingerprint may be wavelength A and emits in a wavelength spectrum centered used for providing diagnostic or therapeutic information, for in the green. A sample containing APP isoforms is incubated example, as to measuring a patient's immune status or with the superparamagnetic beads and labeled specific anti measuring response to chemotherapy directed against a bodies under binding conditions. The superparamagnetic particular blood cell subset. Similar analyte fingerprints can beads are retrieved from the sample, either individually or in be used to type pathogenic organisms and viruses, as well as bulk. The beads are illuminated with wavelength and blue to order polynucleotide sequences for gene mapping and/or light emission is detected and measured, and illuminated sequencing.

with and green light emission is detected and measured. Superparamagnetic beads which can be differentiated The intensity of A-induced blue emission is a measure of 65 based on size, shape, color, or density can be magnetically the APP isoform(s) having the Yepitope, while the intensity trapped individually and scanned with appropriate excitation of the A-induced green emission is a measure of the APP illumination(s) and phosphor emission(s) characteristic of

Page 47

particular analytes detected. For example, a unitary detector damage to non-targets, and (2) employing hydrodynamic can simultaneously or contemporaneously trap the super focusing to pass cells (both targets and non-targets) single paramagnetic bead from a suspension, determine the bead file through an illumination zone (e.g., as in a fluorescence type (size, shape, and/or color), and scan for presence and activated cell sorter or the like). Thus, an up-converting abundance of particular phosphors (by illuminating with phosphor linked to an anti-CD8' antibody can be used to excitation wavelength(s) and detecting emitted selectively damage CD8" lymphocytes in a lymphocyte wavelengths). sample, where (1) the phosphor emits at a wavelength that By performing binding assays under dilute conditions is either directly cytotoxic and/or (2) the phosphor emits at wherein an average of one analyte or less (e.g., lymphocyte) a wavelength that produces reactive chemical species by is bound per microbead, it is possible totype cells individu 10 photocatalysis of a compound present in the sample (e.g., a ally (e.g., determine the abundance of CD45R on each sample can be doped with buckminsterfullerene). individual CD4 cell) and thus generate more precise lym Instead of using the emitted radiation directly for photo phocyte subpopulation definitions. catalytic action on tissue or tumors, an excited form of Biotinylated magnetic beads can also be used to monitor oxygen, so called singlet excited oxygen (OAg) can be the kinetics of binding streptavidin to phosphor particles 15 generated by energy transfer from a dye sensitizer to dis and/or to segregate or purify streptavidin-coated solved molecular oxygen. This scheme makes use of the up-converting phosphor particles from a reaction. Thus, tissue penetrating power of near-infrared radiation (red and streptavidin and up-converting phosphor particles are mixed ultrared region light, including 970 nm) which reaches the in a reaction vessel under binding conditions for forming inorganic up-converting phosphor. Two of the infrared pho streptavidin-coated phosphor particles. After a suitable bind 20 tons are converted either into a red, green, or blue photon ing period, unbound streptavidin may be removed (e.g., by depending on the absorption spectrum of the sensitizer dye. centrifugation wherein phosphor particles are collected as The dye is excited by the up-converted radiation into a triplet the pellet, unbound streptavidin in the supernatant is state which transfers its energy to a dissolved molecular decanted, and the pellet is resuspended), biotinylated mag oxygen molecule to yield an excited (singlet) oxygen mol netic beads are added to the remaining phosphor suspension ecule. The cytotoxic activity of singlet oxygen is well in binding conditions, and streptavidin-coated phosphor documented in photodynamic therapy and other biomedical particles are recovered bound to the biotinylated magnetic applications (see, Wagnieres et al. (19-21 Jan. 1990) Future beads. Directions and Applications of Photodynamic Therapy, pp. Photophysical Catalysis by Up-Converting Phosphors 249, SPTE Institutes for Advanced Optical Technologies, Other applications of the invention employ phosphors as 30 Society of Photo-Optical Instrumentation Engineers, Box a photophysical catalyst linked to a probe, where the radia 10, Bellingham, Wash. 98277; Pelegrin et al. (1991) Cancer tion emitted by the phosphor is used, typically in conjunc 67:2529; Wagnieres et al. (24-25 May 1991) Future Direc tion with a dye molecule, to produce localized intense tions and Applications of Photodynamic Therapy, pp. 219; electromagnetic radiation in an area adjacent to the probe for Folli et al. (17 Dec. 1991) Fluoresceine Clinique 4; Bra various purposes other than detection (e.g., cytotoxicity, 35 ichotte et al. (May 1991) ENT-Clinic, Lausanne, ionization of chemical species, mutagenesis, etc.). For Switzerland).

example, an antibody that specifically binds to a cell surface In this application the up-converting phosphoris mixed or antigen, such as a CD8 antigen on a CD8 lymphocyte, may laced with a sensitizing dye such as methylene blue, rose be used as a probe linked to a up-converting phosphor to bengal or phthalocyanine derivatives, such as localize the phosphor to CD8 lymphocytes. A sample Zn-phthalocyanine. In the first and third case a red-emitting containing CD8 lymphocytes can be incubated with the phosphor is used, whereas for rose bengal a green-emitting anti-CD8 probe-phosphor conjugate and irradiated with an phosphor is best suited. The phthalocyanine derivatives are excitation wavelength (e.g., from an infrared laser diode), ideally suited for this purpose because of their total insolu resulting in emission of up-shifted photons (i.e., higher bility in aqueous or biological solutions. These dyes there frequency electromagnetic radiation) in the vicinity of CD8" 45 fore stay in close proximity to the emitters so that the lymphocytes to which the anti-CD8" probe-phosphor con specificity of the cell surface-reporter/probe/dye complex jugate has bound. The emitted radiation may be of a wave becomes the limiting factor. In this case, specialized com length that is directly mutagenic and/or cytotoxic (e.g., binations of reporter/probe?idye formulations preferably in ultraviolet radiation that can lead to formation of thymine the 0.1 to 0.3-micron size range must be synthesized in order dimers, 760–765 nm light is also believed to produce to enable efficient energy transfer: first, up-converted radia chromosomal damage) or may be of a wavelength that can tion is absorbed by the dye as completely as possible; and cause a photolytic decomposition of a chemical present in second, the dye excited energy (triplet state) is transferred to the environment, leading to local formation of reactive dissolved molecular oxygen. Both processes are very effi species that may damage adjacent cells (e.g., photodecom cient if the absorption spectrum of the sensitizer dye is position of buckminsterfullerene, Co. to Css and C, may 55 matched to the up-converted radiation. produce free radicals that may cause lipid peroxidation of This scheme presents a step beyond the traditional pho cell membranes). todynamic therapy methods in that the red light can be used Since phosphor-emitted radiation is isotropic, it is gener both for tracking and diagnostic as well as for therapeutic ally desirable to physically separate targets (e.g., CD8" purposes after up-converting thus necessitating only one lymphocytes) from non-targets (e.g., CD8 lymphocytes) (infrared) light source at about 1000 nm. Afurther advantage prior to excitation irradiation, so that undesirable damage to is the greaterrange within biological samples of the infrared non-targets by isotropic emission(s) (i.e., "secondary radiation compared to other known photodynamic therapy damage") is avoided. Physical separation may be accom excitation schemes (750-850 nm).

plished by various means, including but not limited to: (1) For embodiments employing up-converting phosphors as performing excitation irradiation on a dilute suspension of 65 photophysical catalysts, it is generally desirable that: (1) the target and non-target cells, wherein the mean distance sepa wavelength(s) of the excitation radiation do not produce rating individual cells is sufficient to reduce secondary significant photocatalysis of the substrate compound, (2) the

Page 48

wavelength(s) of the excication radiation are not directly (Silversmith et al. (1986) J. Opt. Soc. Am. A3: 128, and cytotoxic or mutagenic, and (3) the emitted radiation is Macfarlane et al. (1989) Appl. Phys. Lett. 54: 2301), neody directly cytotoxic and/or is of an appropriate wavelength to mium (Macfarlane et al. (1988) Appl. Phsy. Lett. 52: 1300), produce a biologically effective amount of photodecompo thulium (Nguyen et al. (1989) Appl. Opt. 28: 3553 and sition of a substrate compound (e.g., buckminsterfullerene, Allain et al. (1990a) Electron. Lett. 226: 166), holmium psoralen, compounds containing azide substituents or other (Allain et al. (1990b) Electron. Lett. 26:261), and praseody photoactivated groups). Alternatively, histidine side chains mium (Smart et al. (1991) Electron. Lett. 27: 1307). Other of polypeptides can be oxidized by light in the presence of up-conversion laser schemes that rely on energy transfer, dye sensitizers, such as methylene blue or rose bengal energy pooling, cross relaxation, or avalanche absorption are (Proteins Structures and Molecular Principles, (1984) O not appropriate for up-converting chelates because they rely Creighton (ed.). W. H. Freeman and Company, New York; on energy transfer between ions. These processes are introduction to Protein Structure, (1991), C. Branden and J. described by Auzel (1973) Proc. IEEE 61: 758 and Lenth Tooze, Garland Publishing, New York, N.Y., which are and Macfarlane (March 1992) Optics and Photonic News 3: incorporated herein by reference). Thus, for example, 8. Energy transfer can be efficient in a crystalline host up-converting phosphors linked to anti-CD8 antibodies can 15 containing many rare earth ions, but not in a solution where be used as photophysical catalysts to produce selective, the concentration of ions is low and the phonon structure is localized damage to CD8" lymphocytes. In accordance with less constrained.

the invention, essentially any antibody can be linked to an appropriate up-converting phosphor, either directly or by forInup-conversion certain cases, these schemes may not function as well in chelates. For example, certain of the conjugation to protein A which may then bind the immu 20 schemes have been demonstrated using crystalline host noglobulin. Thus, the up-converting photophysical catalysts materials at very low temperatures, and may not function as of the invention may be used to targetessentially any desired well at room temperature in a chelate. Schemes that do not antigen or cell type that can be distinguished by the presence involve intermediate relaxation such as that of Smart et al. of an identified antigen. have advantages in chelates because they can be excited Up-converting Chelates 25 more effectively with pulsed sources. Higher peak powers Certain applications require small reporters. For example, can be obtained from diode lasers when they are operated in the transport, ability to stay in suspension, the bonding a pulsed mode. The higher peak powers lead to more dynamics, and the tendency toward removal by microphages efficient up-conversion due to the nonlinear dependence on may be improved for smaller reporters. However, the excitation power.

reduced sensitivity available with smaller reporters must 30 Up-Converting Organic Dyes also be considered, Similar to the up-converting inorganic phosphor reporters One type of small up-converting inorganic phosphor we propose to use "molecular" labels whose fluorescence consists of rare earth ions in chelates. The use of lanthanide will be detected by optoelectronic means. Infrared or red chelates as reporters has been developed for biological light is exciting the probe-reporter complex bound to a assays as described on pages 6 and 7 of this application. This 35 target, after which light is emitted at shorter wavelengths prior use of lanthanide chelates involved down-conversion. with respect to the illuminating source. This up-converted That is, the emission light is at a wavelength which is longer light is free of scattered light from the source or autofluo than the excitation wavelength. rescence by virtue of its higher energy. Furthermore, autof Rare earth chelates may be used as up-converting report luorescence is greatly reduced by virtue of the excitation in ers through stepwise excitation such as shown in FIG.5a, or the infrared or red spectral range. The light source is a pump in FIG. 5b (except that all levels would be in the same ion). laser whose pump pulses are short in order to achieve high Energy transfer from a sensitizer ion to an activator ion powers and low energy in order to enable non-linear optical cannot be used in the case of a single rare earth ion. processes in the dye. The goal is to excite the second excited Chelates suitable for use as up-converting phosphors singlet state (S) in a dye with aps pulse from a tunable dye include ethylenediaminetetraacetic acid (EDTA), dipicolinic 45 laser using two red or infrared photons. After pumping the acid (DPA), diethylenetriaminetetraacetic acid (DTTA). S. state the dye relaxes within a few ps to the fluorescing diethylenetriaminepentaacetic acid (DTPA), tetraazacy state (S) which can be detected by optoelectronic means. clotetradecanetetraacetic acid (TETA), as well as antibiotics, The goal of reaching the S. state using two photons enables natural chelating proteins, phthalocyanines, and cryptates. one to take advantage of the increasing two-photon cross Methods for preparation of lanthanide chelates and their use 50 sections as one approaches the S. state using two-photon in biological assays are described in the literature (Mukkala absorption. The non-resonant two-photon absorption cross et al. (1989) Anal. Biochem. 176: 319, Hemmila et al. (1984) sections are on the order of 10 to 10 cms, whereas the Anal. Biochem. 137: 335, Soini and Kojola (1983) Clin. cross sections corresponding to S absorption are larger by Chem. 29: 65, Nonisotopic DNA Probe Techniques (1992) two to three orders of magnitude. A few specific examples Kricka (Ed.) Academic Press, New York, as well as the 55 will be mentioned: in general cyanines, xanthenes, references on page 6 of this application). Up-conversion rhodamines, acridines and oxazines are well suited for this phosphor reporters can also consist of rare earth ions inside purpose. Blue dyes can also be used, but the excitation cage compounds such as fullerene materials following the wavelength will be in the red. Rhodamine can be excited at procedures described by Bethune et al. (1993) Nature 366: 650 to 700 nm using two photons, and fluorescence is 123 and references therein. expected around 555 nm. Many IR dyes such as IR-140, Suitable ions for up-conversion in chelates include IR-132 and R-125 can be excited at 1060 nm using two erbium, neodymium, thulium, holmium, and praseodymium. photons of the Nd:YAG fundamental, and fluorescence is Other candidate ions include the other lanthanide elements, expected in the 850 to 950 nm range. An example of a blue the actinide elements, and other metal elements. Stepwise dye is BBQ excited at 480 nm to reach the S. state at 240 excitation schemes suitable for up-conversion in lanthanide nm, and fluorescence is expected at 390 nm. Many of these chelates are described in the literature on up-conversion dyes are only slightly soluble in aqueous solution and are lasers. Examples include up-conversion in erbium either polar in nature (cyanines) or have polar substituents.

Page 49

Depending on the nature of the probe, no or only minimal about 25-100 nm of excitation maxima wavelengths and attachment chemistry needs to be undertaken because of the thus are potential sources of background noise from scat abundance of functional groups on the dye chromophore. tered excitation illumination. Excitation illumination may Several companies sell entire lines of dyes: examples are also be multiplexed and/or collimated; for example, beams KODAK Exciton and Lambda Physik. The scientific foun of various discrete frequencies from multiple coherent dations of two-photon laser excitation in organic dye mol sources (e.g., lasers) can be collimated and multiplexed ecules have been treated in a few experimental papers: A. using an array of dichroic mirrors. In this way, samples Penzkofer and W. Leupacher, Optical and Quantum Elec containing multiple phosphor species having different exci tronics 19 (1987). 327-349; C. H. Chen and M. P. McCann, tation wavelength bands can be illuminated at their excita Optics Commun. 63 (1987), 335; J. P. Hermann and J. 10 tion frequencies simultaneously. Illumination may be con Duculing, Optics Commun. 6 (1972), 101; B. Foucault and J. tinuous or pulsed, or may combine continuous wave (CW) P. Hermann, Optics Commun. 15 (1975), 412; Shichun Li and pulsed illumination where multiple illumination beams and C. Y. She, Optica Acta 29 (1982), 281-287; D. J. are multiplexed (e.g., a pulsed beam is multiplexed with a Bradley, M. H. R. Hutchinson and H. Koetser, Proc. R. Soc. CW beam), permitting signal discrimination between phos Lond. A 329 (1972), 105-119. 15 phorescence induced by the CW source and phosphores Resonant Multiphoton Ionization cence induced by the pulsed source, thus allowing the At very high laser intensities the up-converting organic discrimination of multiple phosphor species having similar dyes are induced to absorb an additional exciting photon in emission spectra but different excitation spectra. For the field of focussed laser radiation. At those high laser example but not limitation, commercially available gallium intensities the fluorescence is suppressed in favor of absorp arsenide laser diodes can be used as an illumination source tion of an additional photon. This process usually brings the for providing near-infrared light. organic dye molecules above the ionization limit in solution The ability to use infrared excitation for stimulating and they stabilize by emitting an electron into the solvent up-converting phosphors provides several advantages. First, shell. The result of this three-photon interaction is a molecu inexpensive IR and near-IR diode lasers can be used for lar ion and an attached or solvated electron. When this 25 sustained high-intensity excitation illumination, particularly charge separation is taking place in an electric field, the in IR wavelength bands which are not absorbed by water. charges drift and generate a voltage that can be detected in This level of high-intensity illumination would not be suit an extremely sensitive manner. This amounts to the mea able for use with conventional labels, such as ordinary surement of the transient conductivity in the solvent system fluorescent dyes (e.g., FTTC), since high-intensity UV or and is usually more sensitive than light detection. The 30 visible radiation produces extensive photobleaching of the disadvantage of this method is that it necessitates electrodes label and, potentially, damage to the sample. The ability to that sense the moving charges. In that sense it is not as use higher illumination intensities without photobleaching non-invasive a method as light detection. On the other hand or sample damage translates into larger potential signals, and it bypasses the conversion of light into a photoelectric signal hence more sensitive assays.

which represents an enormous advantage. Every optical 35 The compatibility of up-converting labels with the use of system has a restricted viewing angle that reduces efficiency, diode lasers as illumination sources provide other distinct whereas photoionization "senses" always close to 100% of advantages over lamp sources and most other laser sources. the charges generated. Effectively, the non-linear interaction First, diode laser intensity can be modulated directly through of the laser field converts every excited organic dye mol modulation of the drive current. This allows modulation of ecule into an electric pulse at sufficiently high field inten the light for time-gated or phase-sensitive detection sities that can be routinely achieved using commercial laser techniques, which afford sensitivity enhancement without sources. Specific examples are the excitation of Rhodamine the use of an additional modulator. Modulators require around 650 to 700 nm, or BBQ excitation around 480 nm. high-voltage circuitry and expensive crystals, adding both Organic dyes absorbing in the red have to absorb two cost and additional size to apparatus. The laser diode or additional photons after being excited into S thus making 45 light-emitting diode may be pulsed through direct current the whole process a four-photon excitation process, which is modulation. Second, laser illumination sources provide illu slower than a three-photon non-linear process. There may, mination that is exceptionally monochromatic and can be however, be circumstances where such a four-photon pro tightly focused on very small spot sizes, which provides cess is desirable. advantages in signal-to-noise ratio and sensitivity due to Detection Apparatus 50 reduced background light outside of the desired excitation Detection and quantitation of inorganic up-converting spectral region and illuminated volume. Adiode laser affords phosphor(s) is generally/accomplished by: (1) illuminating a these significant advantages without the additional expense sample suspected of containing up-converting phosphors and size of other conventional or laser sources. with electromagnetic radiation at an excitation wavelength, Detection and quantitation of phosphorescent radiation and (2) detecting phosphorescent radiation at one or more 55 from excited up-converting phosphors can be accomplished emission wavelength band(s). by a variety of means. Various means of detecting phospho Illumination of the sample is produced by exposing the rescent emission(s) can be employed, including but not sample to electromagnetic radiation produced by at least one limited to: photomultiplier devices, avalanche photodiode, excitation source. various excitation sources may be used, charge-coupled devices (CCD), CID devices, photographic including infrared laser diodes and incandescent filaments, film emulsion, photochemical reactions yielding detectable as well as other suitable sources. Optical filters which have products, and visual observation (e.g., fluorescent light high transmissibility in the excitation wavelength range(s) microscopy). If the reporters are organic dyes, resonant and low transmissibility in one or more undesirable wave multiphoton ionization can be sensed using electrostatic length band(s) can be employed to filter out undesirable position-sensitive detectors. Detection can employ time wavelengths from the source illumination. Undesirable 65 gated and/or frequency-gated light collection for rejection of wavelength ranges generally include those wavelengths that residual background noise. Time-gated detection is gener produce detectable sample autofluoresence and/or are within ally desirable, as it provides a method for recording long

Page 50

lived emission(s) after termination of illumination; thus, other focusing mechanism 27, and encounters the sample. signal(s) attributable to phosphorescence or delayed fluo Light emitted by the phosphor reporters is collected by a lens rescence of up-converting phosphoris recorded, while short 30, and components in the two emission bands are separated lived autofluoresence and scattered illumination light, if any, by a suitable separation element 32 and directed to the is rejected. Time-gated detection can be produced either by 5 respective detectors.

specified periodic mechanical blocking by a rotating blade There are a number of possible regimes for driving the (i.e., mechanical chopper) or through electronic means laser diodes and detecting the emitted light in the different wherein prompt signals (i.e., occurring within about 0.1 to wavelength bands. This is shown generically as a control 0.3 us of termination of illumination) are rejected (e.g., an electronics block 35 communicating with the laser diodes electronic-controlled, solid-state optical shutter such as 10 andthe detectors. The particular timing and other characteristics Pockel's or Kerr cells). Up-converting phosphors and of nection control electronics will be described below in con with specific embodiments.

up-converting delayed fluorescent dyes typically have emis There may be a plurality of reporters having distinct sion lifetimes of approximately a few milliseconds (perhaps emission as much as 10 ms, but typically on the order of 1 ms), case, the bands but a common excitation band. In such a whereas background noise usually decays within about 100 15 single laser diode.would system include multiple detectors for a

Similarly, there may be a plurality of ns. Therefore, when using a pulsed excitation source, it is reporters having distinct excitation bands but a common generally desirable to use time-gated detection to reject emission band. In such a case, the system would include prompt signals. multiple laser diodes for a single detector, and would use Since up-converting phosphors are not subject to time multiplexing techniques or the like to separate the photobleaching, very weak emitted phosphor signals can be 20 wavelengths.

collected and integrated over very long detection times Light from the two sources is shown as being combined (continuous illumination or multiple pulsed illumination) to so as to be focused at a single location by a common increase sensitivity of detection. Such time integration can focusing mechanism. This is not necessary, even if it is be electronic or chemical (e.g., photographic film). When desired to illuminate the same region of the sample. non-infrared photographic film is used as a means for 25 Similarly, the collection need not be via a single collection detecting weak emitted signals, up-converting reporters pro mechanism. If it is necessary to preserve all the light, the vide the advantage as compared to down-converting phos combination and separation elements can include a wave phors that the excitation source(s) typically provide illumi length division multiplexer and a demultiplexer using dich nation in a wavelength range (e.g., infrared and near roic filters. If loss can be tolerated, 50% beam splitters and infrared) that does not produce significant exposure of the 30 filters can be used.

film (i.e., is similar to a darkroom safelight). Thus, The schematic shows the light passing through the sample up-converting phosphors can be used as convenient ultra and being detected in line. As a general matter, the emission sensitive labels for immunohistochemical staining and/or in from the phosphor reporters is generally isotropic, and it situ hybridization in conjunction with fluorescence micros may be preferred to collect light at an angle from the copy using an infrared source (e.g., a infrared laser diode) 35 direction of the incident light to avoid background from the and photographic film (e.g., Kodak Ektachrome) for signal excitation source. However, since the excitation and the and image detection of visible range luminescence (with or emission bands are widely separated, such background is without an infrared-blocking filter). unlikely to be an issue in most cases. Rather, other consid Instrumentation Overview erations may dictate other geometries. For example, it may The basic purpose of the instrumentation is to expose the 40 be desired to detect light traveling back along the path of the up-converting phosphor particles of an assay sample to incident radiation so that certain elements in the optical train near-infrared (NIR) light and to measure the amount of are shared between the excitation and the detection paths. visible light that is emitted. A typical type of instrument with shared elements is a FIG. 1 is an optical and electronic block diagram illus microscope where the objective is used to focus the excita trating representative apparatus 10 for performing diagnos- 45 tion radiation on the sample and collect the emitted radia tics on a sample 15 according to the present invention. The tion. A potentially advantageous variation on such a con invention may be carried out with one or a plurality of figuration makes use of the phenomenon of optical trapping. reporters. For purposes of illustration, the apparatus shows In a situation where the reporter is bound to a small bead, it a system wherein two diagnostics are performed on a single may be possible to trap the bead in the region near the beam sample in which two phosphor reporters are used. The first 50 focus. The same source, or a different source, can be used to reporter has an excitation band centered at A and an excite the reporter. The use of an infrared diode laser to trap emission band centered at 'while the second reporter has small particles is described in Sato et al., "Optical trapping respective excitation and emission bands centered at and of small particles using a 1.3 um compact InCaAsP laser." '. Since the reporters of the present invention rely on Optics Letters, Vol. 16, No. 5 (Mar. 1, 1991), incorporated multiphoton excitation, wavelengths A and A are longer 55 herein by reference.

than wavelengths 'and'. The former are typically in the Specific Detection Techniques near infrared and the latter in the visible. As outlined above, multichannel detection uses optical A pair of light sources 2001) and 2002), which may be devices such as filters or dichroic beam splitters where the laser diodes or light-emitting diodes (LEDs), provide light at emission bands of the phosphor reporters are sufficiently the desired excitation wavelengths, while respective detec- 60 separated. Similarly, it was pointed out that multiple report tors 22(1) and 22(2), which may be photodiodes, detect light ers having a common emission band could be detected using at the desired emission wavelengths. The emitted radiation electronic techniques. These electronic techniques will be is related to the incident flux by a power law, so efficiency described below in connection with multiple sources. can be maximized by having the incident beam sharply However, the techniques will be first described in the context focused on the sample. To this end, light from the two 65 of a single channel. The techniques are useful in this context sources is combined to a single path by a suitable combi since there are sources of background that are in the same nation element 25, is focused to a small region by a lens or wavelength range as the signal sought to be measured.

Page 51

FIG. 2A shows an apparatus for implementing phase photon process, there is no requirement that the two photons sensitive detection in the context of a single channel. Cor have the same energy. Rather, it is only necessary that the responding reference numerals are used for elements corre band. total energy of the two photons fall within the excitation sponding to those in earlier described figures. In this context, pensiveThus, since it is relatively straightforward and inex control electronics 35 comprises a waveform generator 37 there aretomore provide different wavelengths with laser diodes, possible combinations, i.e., more possible and a frequency mixer 40. Waveform generator 37 drives choices of total excitation laser diode 2001) at a frequency f, and provides a signal at in the choice of rare earthenergy. This allows more latitude f to the frequency mixer. The frequency mixer also receives excitation steps need not ions for up-converters since the rely on energy transfer coinci the signal from detector 22(1) and a phase control input dences involving a single photon signal. This circuitry provides additional background dis 10 possible to achieve direct stepwiseenergy. Further, it may be excitation of the emitting crimination because the background has a much shorter ion (the erbium ion in the example outlined above) without lifetime than the signal sought to be measured (nanoseconds using energy transfer from another absorbing ion (the ytter or microseconds compared to milliseconds). This causes the bium ion in the example) while taking advantage of resonant signal and background to have different phases (although enhancement of intermediate levels. Additionally, the use of they are both modulated at the characteristic frequency of 15 different wavelengths for a single reporter can provide the waveform generator). For a discussion of the lifetime additional options for excitation-dependent multiplexing dependent phase shift, see Demtroder, Laser Spectroscopy, and background discrimination techniques. Springer-Verlag, New York, 1988, pp. 557-559, incorpo Multiple wavelength excitation of a single phosphor may rated herein by reference). The phase input signal is con occur in a number of ways, as shown in FIGS. 5A through trolled to maximize the signal and discriminate against the 20 5C.shown

Two lasers may cause stepwise excitation of a single ion, background. This background discrimination differs from as in FIG. 5A. A first laser stimulates excitation from level 1 to level 2, and a second laser stimulates excitation that typical for phase sensitive detection where the signal is from level 2 to level 3, at which level emission occurs. modulated and the background is not. Discrimination against unmodulated background is also beneficial here, Single shown ion excitation can also occur using energy transferas in FIG. 5B. In this case, a first laser stimulates leading to two types of discrimination. 25

Because the signal relies on two-photon excitation, it is from levelfrom excitation level 1 to level 2, energy transfer occurs 2 to level 3, and a second laser stimulates possible to use two modulated laser diodes and to detect the excitation from level 3 to level 4. In a variation of the latter signal at the sum or difference of the modulation frequen process, levels 1 and 2 can be in a first ion (i.e., a sensitizer cies. FIG. 2B shows such an arrangement where first and ion) and levels 3 and 4 in a second ion (i.e., activator ion) second laser diodes 2001) and 2001)" (emitting at the same 30 as shown in FIG. 5C.

wavelength, or possibly different wavelengths) are modu In a stepwise excitation scheme shown in FIG. 5A, energy lated by signals from waveform generators 37a and 37b transfer is not required, and thus information on the polar operating at respective frequencies f, andf. The waveform ization of the excitation lasers may be preserved and cause generator output signals are communicated to a first fre polarization of the emitted radiation. In this case, depolar quency mixer 42, and a signal at f-f is communicated to 35 ization of the light may allow for enhanced discrimination a second frequency mixer 45. The signal from detector 22(1) between signal and background noise. and a phase input signal are also communicated to frequency For the multi-ion multi-laser excitation scheme shown in mixer 45. FIG. 5C, there may be several phosphors that share a FIG. 3 shows apparatus for performing gated detection. common excitation wavelength. In this case, discrimination Since the background is shorter-lived than the signal, delay between different phosphors may be performed on the basis ing the detection allows improved discrimination. To this of different emission wavelengths and/or through time end, the laser diode is driven by a pulse generator 50, a gated, frequency-modulated, and/or phase-sensitive detec delayed output of which is used to enable a gated integrator tion utilizing modulation of the excitation wavelength(s). or other gated analyzer 55. Specific Instrument Embodiments

FIG. 4 shows an apparatus for performing diagnostics on 45 FIG. 6 is a schematic view snowing the optical train of a a sample using first and second, reporters having excitation particular embodiment of apparatus for carrying out the bands centered at A and A2, and having overlapping emis present invention on a sample using a hand-held probe. This sion bands near . The sample is irradiated by light from embodiment takes the form of a miniaturized instrument laser diodes 2001) and 2002) as discussed above in connec comprising a housing 75 (shown in phantom), a hand-held tion with FIG. 1. First and second waveform generators probe 80, with a fiber optic connecting cable 82. The optical 37(1) and 37(2) drive the laser diodes at respective frequen and electronics components are located within the housing. cies f and f, and further provide signals at f, and f, to For purposes of illustration, the optical components of a respective frequency mixers 60(1) and 60(2). The signal 3-channel system are shown. The sample may contain up to from detector 22(3) is communicated to both frequency three reporters having distinct emission bands, for example, mixers, which also receive respective phase input signals. 55 in the blue, green, and red portions of the visible spectrum. Thus, frequency mixer 60(1) provides an output signal It is also assumed that the reporters have distinct excitation corresponding to the amount of emitted light modulated at bands in the near infrared.

frequency f, which provides a measure of the presence of The output beams from three laser diodes 85a-c are the first reporter in the sample. Similarly, frequency mixer communicated through graded index (GRN) lenses 87a-c, 60(2) provides an output signal corresponding to the amount focused onto the ends of respective fiber segments 88a-c of emitted light modulated at frequency f, which provides and coupled into a single fiber 90 by a directional coupler 92 a measure of the presence of the second reporter in the or other suitable device. The light emerging from the end of sample. fiber 90 is collimated by a GRIN lens 95, passes through a The use of two different wavelengths was discussed above dichroic beam splitter 97, and is refocused by a GRIN lens in the context of two reporters having different excitation 100 onto the end of fiber optic cable 82. The beam splitter bands. However, the discussion is germane to a single is assumed to pass the infrared radiation from the laser reporter situation as well. Since the excitation is a two diodes but reflect visible light.

Page 52

Hand-held probe 80 includes a handpiece 102, an internal herein by reference. The particular array discrete contains GRIN lens 105, and a frustoconical alignment tip 110. The 1024 discrete elements in a 1.28 cmx1.28 cm area. light emerging from fiber 82 is focused by GRIN lens 105 The embodiment of FIG. 7A shows the peptide array in at a focus point 115 that is slightly beyond alignment tip 110. intimate contact with the CCD array. Indeed it may be The alignment tip is brought into proximity with the test tube possible to deposit the peptides directly on the passivation holding the sample so that focus point 115 is in the sample. layer without a separate substrate. However, there may be It is assumed that the test tube is transmissive to the laser situations where spatially separated arrays are preferred radiation. FIG.7B shows an embodiment where the peptide array and A portion of the light emanating from the region of focus the CCD array are separated. An array of lenses 165 collect point 115 in the sample is collected by GRIN lens 105. 10 the light from respective binding sites and focus it on focused into fiber 82, collimated by GRIN lens 100, and respective detector elements. This arrangement facilitates reflected at dichroic beam splitter 97. This light may contain the use of filters to the extent that other techniques for wavelengths in up to the three emission bands. Optical filters rejecting the excitation radiation are not used. 120a-c direct the particular components to respective pho Optical trapping may be used to transiently immobilize a todetectors 125a-c. Aparticular filter arrangement is shown 15 sample particle for determination of the presence or absence where each filter reflects light in a respective emission band, of phosphor on the particle. Conveniently, the wavelength but other arrangements would be used if, for example, one range used to trap sample particles may be essentially or more of the filters were bandpass filters for the emission identical to an excitation wavelength range for the bands. up-converting phosphor(s) selected, so that optical trapping The control electronics are not shown, but could incor 20 and excitation illumination is performed with the same porate the time-multiplexed or heterodyne techniques dis source. FIG. 3 shows a block diagram of an apparatus used cussed above. Such techniques would be necessary, for for single-beam gradient force trapping of small particles. example, if the emission bands were not distinct. FIG. 26 is a block diagram of one embodiment of appa FIG. 7A is a schematic of an embodiment of the invention ratus for carrying out the present invention on a sample in which a charge coupled device (CCD) imaging array 150 25 using a microscope. In this embodiment a standard micro is used as a detector in combination with a two dimensional scope is modified to accept infrared scanning optics and array 152 of peptides or other biologically active species image processing electronics. A suitable microscope for deposited on a glass or plastic substrate. The CCD array has modification is the Zeiss model CLSM-10. a number of individually addressable photosensitive detec The microscope is fitted with a HeNe laser A1 for visible tor elements 155 with an overlying passivation layer 157 30 imaging and an argon laser A2 for both visible and UV while the peptide array has a number of individual binding imaging. Both lasers are mounted internally and are indi sites 160. The probe containing the phosphor would be vidually selectable through a series of motorized shutters reaction specific to one or more of the elements in this A3. The upconverting phosphors are excited with an exter peptide array and would therefore become physically nally mounted IR laser diode. In the preferred embodiment, attached to those elements and only those elements. The two IR laser diodes A4 and A5, operating at two different IR peptide array is shown as having a one-to-one geometric wavelengths, are coupled to the microscope thereby allow relation to the imaging array in which one pixel corresponds ing multiple phosphor reporters to be identified. Laser to each element in the peptide array. However, it is also diodes A4 and A5 are individually selectable using motor possible to have larger peptide elements that cover a group ized shutters A6. When an IR beam is selected, it is routed of detector elements should such be necessary. through the microscope's galvanometrically controlled Various of the techniques described above can be used to scanning mirrors A7 which scan the beam in a raster fashion. enable the detector array to distinguish the emissions of the The beam passes through the objective lens (not shown) phosphor from the infrared laser stimulation. onto a sample A8 and is reflected back through the objective First, it is possible to use a phosphor that responds to IR lens to a set of galvanometrically controlled receiving mir stimulation beyond the sensitivity range of the detector 45 rors A7. Receiving mirrors A7 reflect the light onto pinhole array. An example of such a phosphor would be Gadolinium optics A9. If the confocal mode is selected, pinhole A9 limits oxysulfide: 10% Erbium. This phosphor is stimulated by the detected image to the light collected from the focal 1.5-micron radiation and emits at 960 nm and 520 mm. The plane. The light is imaged on a photomultiplier tube (PMT) detector array is insensitive to 1.5-micron radiation but is A10. The thickness of the focal plane is proportional to the sensitive to the up-converted radiation. 50 size of the pinhole. The scanning speed is chosen such that Further, since the phosphor emission is relatively slow in sufficient signal intensity is received at the PMTA10. In the rise and fall time it could be time resolved from a pulsed preferred embodiment of this apparatus, a 20 micrometer laser stimulation source by the CCD detector array. The diameter pinhole is used which results in a depth of field of decay time for the upconversion process is a variable about 1 micrometer. If the confocal mode is not selected, the dependent on the particular emitting transition and the 55 beam is deflected around pinhole optics A9 directly to PMT phosphor host; however, it is normally in the range 500 us A10.

seconds to 10 ms. This is very slow compared to the laser Once the optical signal is converted into an electronic one, excitation pulse and the capability of the detector array. a standard, composite video signal can be developed and The techniques for fabricating the CCD array are well displayed as an image on a television monitor A11. The known since CCD imaging arrays have been commercially image can be manipulated and enhanced through standard available for many years. A variety of such devices can be image processing software. in the preferred embodiment of obtained from David Sarnoff Research Center, Princeton, this apparatus the software runs on an IBM 486 PCA12. The N.J. software can be used to perform averaging, filtering, edge The techniques for fabricating the peptide array are detection and overlaying the images received from each of described in a paper by Fodor et al., "Light-Directed, 65 the different light sources.

Spatially Addressable Parallel Chemical Synthesis.” In the confocal mode, it is possible to reconstruct a 3 Science, Vol. 251. pp. 767-773 (Feb. 15, 1991), incorporated dimensional view of sample A8. The reconstruction is

Page 53

formed by stepping through sample A8 at small intervals, is an illustration of the data for upconverting phosphors in making an image of the sample at each interval. The multiple three test wells.

sequential images are transferred to an external graphics FIG. 29 is a schematic view of a second embodiment of machine (not shown) for reconstruction of the sample in 3 a hand-held probe for carrying out the present invention. dimensions. These 3-D images can then be rotated to give This embodiment is comprised of a housing D1 and a different perspectives of the data sets, leading to a better capillary wick D2. Within housing D1 is a diode excitation understanding of the samples. laser D3, a lens assembly D4, a photodiode detector D5, and FIG. 27 is a block diagram of a microtiter plate reader for a battery supply D6. A display D7 mounted to one surface use with the present invention. Within a light-tight test of housing D1 communicates the results of the test to the chamber B1 is a near IR laser excitation source B2, a 10 user. In the preferred embodiment, laser D3 operates in the photomultiplier tube (PMT) detector B3, and a sample assay 960-980 nanometer range.

plate B4. In the preferred embodiment of this apparatus, In use, wick D2 wicks up a portion of a sample fluid D8 assay plate B4 is a Terasaki HLA plate. This plate is which is suspected of containing the target antigens. Target preferred due to its small tapered sample wells which tend antigens bind to the antibodies present at a capture surface to concentrate the sample material into a relatively small 15 D9. Capture surface D9 is positioned at the focal point of target area. The target area in this configuration is still larger source D3. The target antigens can be labeled with than the diameter of the laser beam. Furthermore, it is phosphor-antibody conjugates either before or after capture. possible that the distribution of the assay material across the thistheconfiguration

In preferred embodiment wick D2 is formed of glass. In capture surface D9 is prepared simply by bottom of the well is not even. Because of these two factors, simply aiming the laser at the center of the bottom well filling the inside of the capillary with a bubble containing the surface is unlikely to provide accurate readings. There are antibodies of interest. By silanizing the inner surface with several approaches that can be used to circumvent this organofunctional silanes, conventional chemistries can be problem. The first approach is to defocus the laser beam used to covalently link the antibodies or other biological sufficiently to allow a larger amount of the target area to be macromolecule(s) to the inner tube wall at the site of the interrogated. However, depending upon the output of laser 25 liquid bubble. The surface energy of the capillary is also B2, defocussing the beam may lower The sensitivity of the easy to modify by silanization, which will help prevent apparatus to an unacceptable level. Another approach is to nonspecific reagent and antigen adherence to the walls of the raster scan the laser beam across the bottom of target well. tube.

A third and preferred approach is to simply automate the In the preferred embodiment of this apparatus, the lower scanning and data collection system. 30 portion of wick D2 is impregnated with upconverting phos Light from laser B2 passes through a filter B5 and is phors that are conjugated to the target analytes or a cross focussed by a lens B6 onto an individual sample well of reactive epitope for the capture probe. In use, the phosphor assay plate B4. Plate B4 is mounted on a pair of translators conjugates chromatograph towards capture surface D9 as B7 which allow positioning in the horizontal and vertical sample fluid D8 is drawn up wick D2. As phosphors directions. In the present configuration translators B7 allow 35 accumulate at capture surface D9, they will begin to emit approximately 2.5 centimeters of travel; sufficient to address visible light upon excitation by diode laser D3. The visible 3 sample wells in each direction. Translators B7 are con The light emitted by the phosphors is detected by detector D5. trolled by an x-y controller B8. Controller B8 allows for output of detector D5 is displayed on display D7. The either manual or computerized control. amount of upconverted light reaching the detector is directly A sample well on plate B4, when containing upconverting proportional to the concentration of labeled target antigen phosphors, will emit visible light which is collected by a lens captured at the capture surface.

B9, passed through a filter B10, and focussed through a lens The apparatus of FIG. 29 can be designed to simulta B11 and a shutter B12 onto PMT B3. PMT B3 outputs a neously detect more than one target antigen. FIG. 30 illus current which is measured by a picoammeter B13. The PMT 45 trates a three channel configuration using interference filters. signal is proportional to the phosphor emission intensity. In this configuration capillary wick D2 is placed at the focus Shutter B12, controlled by a shutter driver B14, provides of a small parabolic reflector D10 capable of collecting exposure protection to PMTB3, thereby preventing damage approximately half of the emitted phosphorescence. The which may result from exposure to very intense light beam from diode laser D3 is directed onto capillary wick D2 sources. Furthermore, overexposure of PMT B3 to light at capture surface D9 along a direction perpendicular to the causes high dark currents which require several hours to optical axis of reflector D10. Phosphorescent light from decrease. PMT B3 is cooled for lower dark current and capture surface D9 is collected and collimated by mirror noise. Associated with the PMT cooler is a water-cooled D10, directed through a notch filter D11 to reject the pump power supply B15. A power supply B16 supplies high light, and onto three detectors D5 using three dichroic beam voltage to PMT B3. splitters D12. The reflectance bands of dichroic beamsplit When the apparatus is operated in a computerized mode, 55 ters D12 are matched to the emission bands of the three a computer B17 regulates controller B8 through an interface phosphors used in the detection process. box B18. Picoammeter B13 can also be connected to com In an alternate embodiment of this apparatus, dichroic puter B17, thereby allowing automated data acquisition to beamsplitters D12 could be replaced with three bandpass be performed. The data acquisition procedure moves trans filters used in the transmission mode. By placing the three lator B7 in the x direction to a first position at which location filters on a rotation wheel, a single detector D5 could be a specified number of current readings are taken and the used. Another alternative is to use a diffraction grating and average is calculated. Translator B7 then moves sample B4 a linear detector array to obtain an actual emission spectrum. FIG. 31A is an illustration of an embodiment of the a predetermined distance in the x direction to a new location where new data is collected. During this process, the data is invention in which a diode laser array F1 and a detector plotted in order to provide the user with an immediate visual 65 array F2 are combined in a single device. In the preferred evaluation. After the scan is completed, the data can be embodiment, arrays F1 and F2 are fabricated on a pair of saved or further data processing can be performed. FIG. 28 silicon chips F3 with array dimensions of approximately 1

Page 54

square centimeter. FIG. 31B is a detailed view of a small cell surface protein bind to the labeled binding reagent, section of the device shown in FIG. 31A. Overlaying whereas cells lacking the cell surface protein do not sub detector array F2 is a polymer film F4 of approximately 10 stantially bind to the labeled binding reagent. The suspended to 25 micrometers thickness which is used as the capture cells are passed across a sample detector under conditions surface. Arrays F1 and F2 are separated by a spacer F5. wherein only about one individual cell is present in a sample Array F1 is comprised of Fabrey-Perot diode lasers, pref detection zone at a time. A source, typically an IR laser, erably tuned to 980 nanometers. Lasers of this type are illuminates each cell and a detector, typically a photomul easily fabricated in gridded array patterns using conven tiplier or photodiode, detects emitted radiation. The detector tional photolithography techniques. Each individual laser in controls gating of the cell in the detection zone into one of array F1 has a columnar beam designed to strike only the 10 asignal(s) plurality of sample collection regions on the basis of the detected. A general description of FACS apparatus adjacent portion of capture surface F4. The required power and methods in provided in U.S. Pat. Nos. 4,172,227; density of the individual lasers is dependent upon the efficiencies of the phosphors being used as well as the 4,347,935; 4,661,913; 4,667,830; 5,093.234; 5,094.940; and 5.144.224, incorporated herein by reference. It is preferred required detection efficiency. The detectors comprising array that up-converting phosphors used as labels for FACS meth F2 are chosen to have an extremely low sensitivity in the 15 ods have wavelength region in which laser array F1 operates. If range(s)) excitation which do range(s) (and preferably also emission not damage cells or genetic material;

additional discrimination between the excitation and emis generally, radiation in the far red, and infrared ranges are sion wavelengths is required, a cutoff filter can be used, preferred for excitation. It is believed that radiation in the preferably incorporated directly into capture surface F3. range of 200 nm to 400 nm should be avoided, where Upconverting phosphors F6 are conjugated by any of a possible, and the wavelength range 760 nm to 765 nm may variety of conventional biochemical crosslinking chemis be avoided in applications where maintenance of viable cells tries to antibody, nucleic acid probes, or other biological is desired.

macromolecules (e.g., carbohydrates, lectins, streptavidin, Additional Variations

MHC complexes), as well as to biological or chemical There are several apparatus design issues relating to the antigens (F7). Bonded to overlay F3 is a grid array F8 of 25 unique excitation and emission characteristics of upconvert complementary probes or antigens which are bound to ing phosphors which must be considered when using capture surface F3 using the same crosslinking chemistries. up-converting phosphors with flow cytometry. The first In use, a sample fluid F9 flows between arrays F1 and F2, issue is the time required to reach maximum emission target probes or antigens are captured by grid array F8 and intensity. Since upconversion is a two photon process, excited by laser array F1, and the emissions detected by 30 upconverting phosphor emission is time delayed approxi detector array F2. mately 100 microseconds. The phosphor must remain within Typically, the upconverting phosphors to be used with this the excitation beam for this period of time regardless of the apparatus are approximately 0.1 to 0.5 micrometers. Since flow rate. Therefore given a flow rate between 1 and 10 the size of the individual phosphor particles is of the order meters per second with a channel width of 70 to 200 of the excitation wavelength, the power of the emission from 35 micrometers, the length of the excitation beam must be the phosphors can be approximated by: between 100 and 1000 micrometers. Given that the phos P-fND1 phor emissions saturate at an excitation intensity of about where f is the phosphorescence efficiency (generally less must 200 watts per square centimeter, the laser source typically than or equal to 10 cm Wumparticle'), N is the achievehave a power between 0.01 and 400 milliwatts to phosphor saturation. This implies that multiple laser number of phosphor particles in the light path, D is the diodes may be required to obtain maximum phosphores diameter of the phosphor particles, and I is the power cence at the fastest flow rates.

density of the excitation source. Another design issue is that associated with the detector. Since the emitted power scales as the square of the Since there is a considerable separation between the exci excitation intensity, diagnostics using upconverting phos tation and emission wavelengths of the upconverting phors perform better in a microassay format. Assuming a 45 phosphors, detection can be performed using a photomulti constant power output from the excitation source, the exci pler tube (PMT), a photodiode, or a CCD array. The phos tation power density increases proportionally with the phorescence decay time is long, with a decay half life of decrease in detection area, and the number of phosphor approximately 300 microseconds. The most sensitive particles in the light path decreases linearly with a decrease method of detection is to integrate the signal measured by in the detection area. Since the power of the light emitted 50 the PMT. However, 99 percent detection of the available from the phosphors scales with the square of the excitation phosphorescent signal requires that the phosphor remain in power density, but linearly with the number of phosphors, the sight path of the detector for 5 times the phosphores P will increase in inverse proportion to the detection area. cence decay half-life (i.e., 1.5 milliseconds). Assuming a Therefore, a 100x100 array will actually be 100 times more flow rate of 10 meters per second and a channel width of 200 sensitive than a 10x10 array. 55 micrometers, the PMT must be able to detect over a path Fluorescence-activated Cell Sorting length of 1.5 centimeters. This path length is also the The up-converting phosphors described herein can be required spacing between cells flowing through the used as phosphorescent labels in fluorescent cell sorting by cytometer, implying a maximum count rate of 667 cells per flow cytometry. Unlike conventional fluorescent dyes, second. It is, however, possible to sacrifice some detection up-converting phosphors possess the distinct advantage of sensitivity by reducing the detection path length, at least to not requiring excitation illumination in wavelength ranges that required to attain steady-state emission from the phos (e.g., UV) that damage genetic material and cells. Typically, phors. As long as a steady-state emission peak is reached by up-converting phosphor labels are attached to a binding the phosphor in the excitation window, the peak signal reagent, such as an antibody, that binds with high affinity and received by the PMT should be directly proportional to the specificity to a cell surface protein present on a subset of concentration of phosphors present. The nonphotobleaching cells in a population of cells in suspension. The phosphor 65 property of the phosphors makes this form of detection labeled binding component is contacted with the cell sus possible. The loss in detection sensitivity corresponding to a pension under binding conditions, so that cells having the 0.1 centimeter path length (versus a 1.5 centimeter path

Page 55

length) is approximately a factor of 3. Triggering the emis only a depiction of "side-on" excitation and detection is sion detector can be accomplished by observing the light shown, both side-on and end-on detection and excitation scattered by the cell as it passes through the excitation arrangements, or combinations, are possible. Reduction of SOCC. the CCD array intensity information by computer analysis In environments where absorption of the up-converted 5 will allow near-real time tracking of the particles in a phosphor radiation is high, the phosphor microparticles are dynamically evolving or living systems. Data analysis and coated with a fluorescent dye or combination of dyes, in reduction performed by the computer would include a selected proportions, which absorb at the up-converted fre convolution of the intrinsic decay of the phosphor emission, quency and subsequently re-radiate at other wavelengths. the number of pixels illuminated and their signal level, the Because the single-photon absorption cross-sections for orientation of the decaying signal on the array, and the these fluors are typically very high, only a thin layer is 10 intensity contributions from a blur circle from particles required for complete absorption of the phosphor emission. moving in and out of the focal plane of the array. In an This coat particle may then be encapsulated and coated in a end-onflow detection arrangement, the size of the blur circle suitable antigen or antibody receptor (e.g. microparticle). An would relate directly to how quickly the particle moves out example of this layering is depicted schematically in FIG. 9. 15 of focus, thereby allowing the velocity of the particle to be There exists a wide variety of fluorescent dyes with strong determined. One possible application would be monitoring absorption transitions in the visible, and their emission the chemistry and kinetics in a reaction column, covers the visible range and extends into the infrared. Most alternatively, the application of this method to flow cytom have fluorescent efficiencies of 10% or more. In this manner, etry may permit the resolution of cells on the basis of the emission wavelengths may be custom-tailored to pass hydrodynamic properties (size, shape, density). The method through the particle's environment, and optical interference 20 may also be useful for in vivo diagnostic applications (e.g., filters may again used to distinguish between excitation and blood perfusion rate).

emission wavelengths. If a relatively large wavelength “win Up-converting phosphor labels may also be used to sense dow" in the test medium exists, then the variety of emission the temperature in the region at which the up-converting wavelengths which may be coated on a single type of phosphor label is bound. Up-converting phosphor tempera phosphor is limited only by the number of available dyes and 25 ture measurement methods are described in Berthou H. and dye combinations. Discrimination between various reporters Jorgensen C. K. (October, 1990) Optics Lett, 15(19): 1100, is then readily carried out using the spectroscopic and incorporated herein by reference.

multiplexing techniques described herein. Thus, the number Although the present invention has been described in of probe/reporter "fingerprints" which may be devised and some detail by way of illustration for purposes of clarity of used in a heterogenous mixture of multiple targets is virtu understanding, it will be apparent that certain changes and ally unlimited. modifications may be practiced within the scope of the The principles described above may also be adapted to claims.

driving species-specific photocatalytic and photochemical The broad scope of this invention is best understood with reactions. In addition to spectroscopic selection, the long reference to the following examples, which are not intended emission decay times of the phosphors permit relatively 35 to limit the invention in any manner. slow reactions or series of reactions to take place within the EXPERMENTAL EXAMPLES emission following photoexposure. This is especially useful when the phosphor-catalyst or reactant) conjugate enters an Validation of Up-Converting Inorganic Phosphors as environment through which the excitation wavelength can Reporters not penetrate. This slow release also increases the probabil Up-converting phosphor particles comprising sodium ity that more targets will interact with the particle.

The unique decay rates of phosphor particles allow yttriumfluoride submicron size, doped with ytterbium-erbium were milled to fractionated by particle size, and coated dynamic studies as well. In a system where continuous exposure to the excitation source is not possible, or is with polycarboxylic acid. Na(YoYborooz.)F was cho sen for its high efficiency upon excitation in the range 940 invasive and thereby undesirable, pulsed excitation followed to 960 nm. ANd:Yagpumped dye laser/R dye combination by delayed fluorescence detection is necessary. After the 45 was phosphor reporter has been photoexcited, the subsequent aboveused to generate 8-ns to 10-ns duration pulses in the frequency range.

emission from the phosphor or phosphorldye conjugate particle lasts typically about a millisecond. In a dynamic milled The laser pulses were used to illuminate a suspension of environment, such as a static or flowing system with moving slides inphosphor particles in liquid and attached to glass situ. The suspension luminescence observed at right targets, the particle will emit a characteristically decaying 50 intensity of light as it travels relative to the excitation/ angles was monitored using a collection lens, a spatial filter detection apparatus. Combined with imaging optics appro in order to filter out scattered excitation light to the maxi mum possible extent, and a photomultiplier, vacuum priate to the scale of the system and the velocities within the photodiode, or simple solid state photodiode (depending on system, a CCD photoelectric sensor array will be used to detect the particle or particles movement across the array's 55 the light level observed).

field of view. The delayed emission of the phosphors, which The luminescent signal level was determined as a function is a well-characterized function of time, makes possible the of solution pH (range: 6-8), grain size, particle loading dynamic tracking of individual particle's positions, direc (ug/cm), and the nature of stabilizing anionic surfactant. tions and velocities, and optionally calculation of particle Signals were recorded both as a time integral from a boxcar size, density, and hydrodynamic conformation. As a particle integrator and from along RC time constant or as a transient moves, it exposes more elements of the array, but with signal using a transient digitizer in order to delineate the every-decreasing intensity. The more elements it exposes luminescence lifetime under particular experimental condi over a certain fraction of its decay time, the faster it is tions. In situ signals were also measured by laser scanning moving. Therefore, the integrated intensity pattern of a microscopy. FIG. 11 is a fluorescence scan of the phosphor particle's emission "track" collected by the array is directly 65 emission spectrum incident to excitation with a laser source related to the velocity of the particle. The particles may be at a wavelength maximum of 977.2 mm; emission maximum refreshed again at any time by the pulsed or chopped CW is about 541.0 nm. FIG. 12 is an excitation scan of the excitation source. FIG. 10 illustrates this scheme. Although phosphor excitation spectrum, with emission collection win

Page 56

dow set at 541.0 nmi; excitation maximum for the phosphor after evaporating the samples over the weekend at at the 541.0 nm, emission wavelength is approximately 110°-120° C. was noticeably yellow, but did phosphoresce about 977 nm. FIG. 13 is a time-decay measurement of the when tested with an IR diode laser.

phosphor luminescence at 541.0 nm after termination of Visual green light emanated from all serial dilutions down excitation illumination; maximal phosphorescence appears at approximately 400 us with a gradual decay to a lower, to 10 (i.e., 1.7 pg/mL or 23.6x10 particles/mL) in a 1 mL stable level of phosphorescence at about 1000 us. FIG. 14 polypropylene microfuge tube using a hand-held diode laser shows the phosphor emission intensity as a function of in a dark room. The 10' and 10 dilutions were visibly excitation illumination intensity; phosphorescence intensity cloudy. Either 1 pil of each serial dilution, or 0.1 pil of the increases with excitation intensity up to almost about 1000 O next higher dilution, were pipetted into a well on the Terisaki W/cm. plate. It was found that 1 pil fills the bottom of the well and Phosphorescence efficiencies of submicron 0.1 pil spreads along the edge of the well, but does not cover Na(YoYboBros.)F particles were measured. A Ti:sap the entire surface. Because of the statistical and pipetting phire laser was used as an excitation source and a spectro problems associated with small volumes with low particle photometer and photomultiplier was used as a detection concentrations, 2 to 4 replicates were prepared of each system. Two types of measurement were performed. The 5 dilution.

first was a direct measurement in which the absolute emis The well of a Terasaki plate holds a 10 ul sample volume. sion per particle for phosphor suspensions was measured in Assuming all the phosphor particles contained in this vol emission bands at 540 nm and 660 nm. The calibrated ume adhere to the bottom of the sample well, we can cross-sections are shown in FIG. 15, and size-dependence is estimate an equivalent detection sensitivity (Table III). It shown graphically in FIG. 16. This corresponded to a should be noted that 10 to 10'M is the normal range of phosphorescence cross-section of approximately 1x10' enzyme-linked surface assays.

cm for 0.3 um particles with excitation light at 975 nm and Control Sample Results an intensity of approximately 20 W/cm. The emission The control samples were scanned using a prototype efficiency of dry phosphor powder of about 25 um was also up-conversion fluorimeter device (David Sarnoff Research measured. On the basis of known values for the absorption 25 cross-section of Yb" in crystalline hosts (Lacovara et al. Center). The samples were scanned by moving the plate in (1991) Op. Lett. 16: 1089, incorporated herein by reference) relative to the focalusing 50 m increments, point a motorized X-Y positioning stage, of an infrared diode laser.

and the measured dependence of the phosphorescence emis The IR diode laser was operated at 63 mW (100 mA). The sion on particle size, a phosphorescence cross-section of beam was focused to 24x10 cm at the focal point. As the approximately 1x10 cm was found. The difference 30 between these two measurements may be due to a difference bottom of the sample well is about 1.4x10 cm (1365 m diameter), the beam covers less than 17% of the well bottom in phosphorescence efficiency between dry phosphor and aqueous suspensions, or due to absorption of multiply scat surface at any individual position. The well also has sloping tered photons in the dry phosphor. On the basis of either of side walls which widen from bottom to top of the sample these cross-section estimates, the cross-section is sufficiently 35 well and are also interrogated by a progressively divergent large to allow detection of single submicron phosphor par laser beam. Neglecting losses in the optics, the IR light ticles at moderate laser intensities. At laser intensities of intensity at the focal point (bottom of the sample well) was roughly 10 Wlcm, the phosphorescence scales as the laser approximately 26-27 Wlcm at 980 nm wavelength. A intensity to the 1.5 power. photomultipler tube (PMT) was used for detection of the Phosphor Particle Performance: Sensitivity of Detection visible (upconverted) light emitted from the sample. Since A series of Terasaki plates containing serial dilutions of the laser beam width was smaller than the surface area at the monodisperse 0.3 in up-converting phosphor particles con bottom of the sample well, the plate was aligned by visual sisting of (YosYbooseroos)2O2S were tested for inspection against the focal point of the diode laser so that up-conversion fluorescence under IR diode laser illumina the laser was centered in the middle well (C6 when reading tion in a prototype instrument. wells C5, C6 and C7, and D6 when reading wells D5, D6, The phosphor particles were prepared by settling in 45 and D7).

DMSO and were serially diluted into a 0.1% aqueous gum The PMT signal (amps) was recorded at each plate arabic solution. This appeared to completely eliminate any position and numerically integrated over the width of the water dispersion problems. The serial dilutions used are sample well (approximately 4000 pm). Several scans were listed in Table I. made at different positions in the 10 to 10' dilution sample 50 wells to determine the uniformity of the particle distribution.

TABLE The background signal was determined by integrating the Phosphor Phosphor Equivalent average dark field current of the PMT over a 4000 pm Loading Loading Detection distance, which yields an integrated background signal of Label (ng?well) (particles well) Sensitivity (M) 1x10a-m. The integration products of the samples wells

55 were scaled to this background signal, and are shown in FIG.

102 17 t 09 236,000 it 12,000 4 x 10 Immunodiagnostic Sample Detection

10-4 0.700.009 2,360 it 120 4 x 10-16 A series of IgG/anti-IgG samples for demonstrating the

106 0.0017 t 0.00009 23.6 t 1.2 4 x 10 capabilities of the up-converting phosphor reporters in a immunosorbant assay format was prepared. These samples consisted of six individual wells (positive samples) coated

The stock DMSO dispersion had a phosphor density of with antigen (mouse IgG) and bovine serum albumin (BSA), 1.70t0.09 mg/mL (at 95% confidence limits), determined and six wells coated with BSA alone (negative controls). gravimetrically by evaporating 4-1 mL samples. This trans 65 Nominal 0.3 pm (YosYbooseroos)2O2S phosphor particles lates to 23.6x10 particles/mL (assuming an average particle coated with goat anti-mouse IgG antibody (anti-IgG) were size of 0.3 m and particle density of 5.3 g/mL). The residue then used as the reporter-antibody conjugate.

Page 57

Six wells (C5, C6, C7, D5, D6, and D7) of a clear tically different at the 99.9% confidence level. The positive polystyrene Terasaki plate were coated with mouse IgG by samples emit on average 30,029.7 times more light than the incubating at 37° C. against 5 L of a 100 g/L mouse IgG negative controls.

solution in phosphate buffered saline (PBS). After 1 h, this Linkage of Phosphors to Biological Macromolecules solution was aspirated off and each sample well was washed In order to delineate further the parameters for with 10 L of 3% BSA in PBS. This was immediately up-converting phosphors as biochemical reporters, biologi aspirated off and replaced with 20 L of 3% BSA in PBS. cal linkers were attached to phosphor particles. Sodium Each sample well was post-coated with BSA by incubating yttrium fluoride-ytterbium/erbium phosphor particles were against the 20L of BSA/PBS solution for 1 h at 37° C. The coated with streptavidin. The excitation and emission spec post-coat solution was aspirated off and the plates stored at 10 tral properties of the phosphor alone and the phosphor 4° C. overnight. These wells were considered in positive coated with streptavidin were measured (FIGS. 17A, 17B, samples. The same six wells in a second Terasaki plate were 18A, and 18B) and both the uncoated and streptavidin prepared in an identical fashion, except they were not coated coated phosphors were almost identical in their absorption with mouse IgG. This second set of sample wells were and emission properties, indicating that the attachment of considered negative controls. 15 macromolecular linkers (e.g., proteins) have little if any Phosphor-Antibody Conjugate effect on the phosphorescent properties of the up-converting A solution of (Yossybooseroos)2O2S phosphor particles phosphor. The streptavidin-coated phosphors were then spe was prepared by suspending the dry phosphors into DMSO. cifically bound to biotinylated magnetic beads, demonstrat The initial particle density was approximately 10 particles/ ing the applicability of linker-conjugated inorganic phos mL as determined by counting the number of particles phors as reporters in biochemical assays, such as contained in the field of an optical microscope. It should be immunoassays, immunohistochemistry, nucleic acid noted that the 0.3 pm fundamental particle size was below hybridizations, and other assays. Magnetic bead technology the resolution limits of the microscope. This solution was allows for the easy separation of biotin-bound streptavidin allowed to settle undisturbed for 3 days. The supernatant, coated phosphor from a solution, and is particularly well which was turbid and presumably contained mostly mono 25 suited for sandwich assays wherein the magnetic bead is the disperse smaller particles was used for subsequent conjuga solid substrate.

tion. Advantageously, streptavidin-biotin chemistry is widely Goat anti-mouse IgG antibody (Ab) was conjugated (by used in a variety of biological assays, for which adsorption) onto the DMSOfractionated phosphor particles. up-converting phosphor reporters are suited. FIG. 20 shows This was done by mixing 200 L of the Absolution (in 0.1M 30 schematically, for example and not limitation, one embodi Tris-HCl, pH 7.2) with 100 L of the phosphor suspension ment of an immunoassay for detecting an analyte in a in DMSO. Several different Ab concentrations were tried in solution by binding the analyte (e.g., an antigen target) to a the range of 0.025 to 1 g/L. A concentration of 0.25 g/L biotinylated antibody, wherein the analyte forms a sandwich appeared to result in the most efficient coating (i.e., maxi complex immobilized on a solid substrate (e.g., a magnetic mum Ab utilization with a minimum of clumping of the 35 bead) by linking a first binding component bound directly to phosphor particles). The phosphors were equilibrated over the solid substrate to a second binding component (e.g., the night at room temperature with the Ab in this DMSO/Tris biotinylated antibody); a streptavidin-coated up-converting solution with gentle agitation. The resulting phosphor-Ab phosphor then binds specifically to the biotinylated antibody conjugates were centrifuged from this solution and resus in the sandwich and serves to report formation of the pended in a 3 g/mL BSA solution in PBS for post-coating. sandwich complex on the solid substrate (which is a measure The resulting BSA/PSA resuspension was used directly for of the analyte concmagnetic bead, it is solid substrate is a the assay, magnetic bead, it is readily removed from the sample The degree of Ab adsorption to the phosphors, and solution by magnetic separation and the amount of phosphor residual Ab activity, was determined by titrating the attached to the bead(s) in sandwich complex(es) are deter phosphor-bound Ab with a fluorescein isothiocyanate 45 mined by measuring specific up-converting phosphores (FITC) conjugated-mouse IgG. The resulting FTTC-labeled cence. Thus, sandwich complex phosphorescence provides a phosphors were passed through a Cyteron Absolute flow quantitative measure of analyte concentration. cytometer, which was also capable of measuring the relative Biotinylated polynucleotides are also conveniently used size of the particles. Two distinct size subpopulations were as hybridization probes, which can be bound by observed with about 65% of the counted particles appearing 50 streptavidin-coated up-converting phosphors to report as small, presumably monodisperse particles, and 35% being hybrid formation.

significantly larger, presumably aggregates. Only 60% of the Background Phosphorescence in Biological Samples smaller subpopulation appeared to have significant quanti Background signals were determined in two biological ties of active Ab (determined by FTTC fluorescence). Of the samples for determination of potential background in immu purported aggregates, about 90% appeared to contain active 55 noassays. Sputum and urine were used as samples in the Ab (by FTTC fluorescence). This suggests that less than 40% same apparatus as used for the phosphorescence sensitivity of the phosphor-Ab conjugates were of an appropriate size measurements (supra). No background levels were found (nominal 0.3 pm) and exhibited anti-mouse IgG activity. A above the systemnoise levels set by the photomultiplier dark similar fraction of phosphor-Ab conjugates (31%) were current. This noise level allows detection of signals from on active but carried a significantly larger phosphor reporter. the order of a few hundred particles/cm. This is close to a The PMT signal (amps) was recorded at each plate single particle in the detection volume of the system. position and numerically integrated over the width of the A photomultiplier is a preferred choice for a detector for sample well /approximately 4000 m). The average signals high sensitivity measurements of up-converting phosphors (with 95% confidence limits) are: since photomultipliers can be selected to produce high Average of Positive Samples=130x10+1.25x10'a-m 65 quantum efficiency at the up-converted (i.e., emitted) wave Average of Negative Controls=4.20x10-6.82x10 lengths and virtually no response in the range of the longer Ia-mThe positive samples and negative controls are statis excitation wavelengths.

Page 58

Detection of Cell Antigens with Phosphor-Labeled Antibod 17.7 mL of 0.1N NCl in 50 mL of deionized water, pH 8.3)

was prepared (Avidin stock). Another solution containing

Streptavidin is attached to the up-converting phosphor 1.7 mg of N-succinimidyl(4-iodoacetyl) aminobenzoate particles as described, supra. The mouse lymphoma cell line, (Pierce Chemical) in 1.2 mL of DMSO was prepared (SIAB EL-4, is probed with a hamster anti-CD3 antibody which stock). A quantity (10 L) of the SIAB stock was added to specifically binds to the 30 kD cell surface EL-4 CD3 T the 1.0 mL of Avidin stock and stirred at room temperature lymphocyte differentiation antigen. The primary hamster 30 minto allow the N-hydroxysuccimide ester of the SIAB antibody is then specifically bound by a biotinylated goat antihamster secondary antibody. The biotinylated secondary to react with primary amines on the avidin (Avidin-SIAB antibody is then detected with the streptavidin-phosphor stock).

conjugate. This type of multiple antibody attachment and O A20 mL scintilation vial was prepared containing 10 mL labeling is termed antibody layering. of borate buffer (pH 8.3). The following additions were then Addition of multiple layers (e.g., binding the primary made to this vial: 21.6 L of the avidin-SIAB stock solution hamster Ab with a goat-antihamster Ab, followed by binding followed by 1.5 mL of the phosphor stock. This reaction with a biotinylated rabbit-antigoat Ab) are used to increase mixture was stirred at room temperature in the dark over the distance separating the phosphor from the target. The 15 night to allow the SIAB activated avidin to react with the layering effect on signal intensity and target detection speci thiol groups present on the silanized phosphor surface and ficity is calibrated and optimized for the individual applica resulting in the covalent linkage of avidin to the phosphor tion by performing layer antibody layering from one layer particles.

(primary antibody is biotinylated) to at least five layers and After the overnight incubation 1.0 mL of the reaction ascertaining the optimal number of layers for detecting CD3 mixture was centrifuged (1 min at 10,000 g) and the super on EL-4 cells. natant removed. The pellet was resuspended in 1.0 mL of FIG. 21 schematically portrays simultaneous detection of phosphate buffered saline (pH 7.2, Pierce) and centrifuged two EL-4 cell surface antigens using phosphors which can again to wash any uncongugated protein from the phos be distinguished on the basis of excitation and/or emission phors. This washing process was repeated. The washed spectra. Detection of both antigens in the scheme shown in 25 pellet was resuspended in 1.0 mL of phosphate buffered FIG. 21 uses a biotinylated terminal antibody which is saline and used directly in diagnostic assays as described conjugated to streptavidin-coated phosphor (#1 or 2) prior below.

to incubation with the Ab-layered sample. Thus, the Measurement Apparatus phosphor-antibody specificity is retained through the unusu A modified SLMAminco 48000 Fluorimeter was used to ally strong (Kp approx. 1x10"M") non-covalent bond measure the fluorescence spectrum from the phosphor between streptavidin and biotin which is pre-formed before samples. The modifications to this device consisted of incubation with the primary antibody-bound sample. Quan adding a laser diode (David Sarnoff CD-299R-FA #13) titation of each antigen is accomplished by detecting the which was input to the fluorimeter through port 3. The laser distinct signal(s) attributable to each individual phosphor diode emits at A=985.1 nm. Spectral data provided by the species. Phosphorescent signals can be distinguished on the 35 David Sarnoff Research Center also shows a small peak at basis of excitation spectrum, emission spectrum fluores 980.2 nm. This peak has 15% the intensity of the peak at 985 cence decay time, or a combination of these or other properties. A 5.08 cm focal length lens was used to collimate the FIG. 22 shows a schematic of an apparatus for phase diode laser beam. The power of the IR laser light was sensitive detection, which affords additional background measured as 6.1 mW at the cuvette location with a drive discrimination. The pulse or frequency mixer is set to pass current of 75 mA. The beam was not focused at the center the signal and discriminate against the background follow of the cuvette. This is true for the standard visible light from ing frequency calibration for maximum background rejec the fluorimeter excitation monochromator as well. The laser tion. diode beam is diverging as it enters the cuvette holder and Covalent Conjugation of Upconverting Phosphor Label to 45 is approximately 4mm (HOX2 mm (V) by the time it reaches Avidin the center of the cell, neglecting the changes in refractive An up converting ytrium-ytterbium-erbium index of the cell wall and the liquid. (YosYboossroos) oxysulfide (OS) phosphor was linked to Light emitted is scanned with a monochromator and avidin by the following procedure: detected by a photomultiplying tube (PMT) 90° from the Monodisperse upconverting phosphor particles were direction of the excitation light. The detection limits for the silanized with thiopropyltriethoxysilane (Huls) following modified SLM Aminco 48000 were determined by serial the procedure detailed by Arkles (in: Silicone Compounds: dilution to be 4x10M (240,000 phosphor particles per Register and Review, Hills America, pgs. 59-75, 1991). This mL) in PBS. Phosphor emission peaks in the spectrum were consisted of adding thiopropyltriethoxysilane (2 g) and 95% seen at wavelengths of 4062 nm, 434+2 nm,522+2 nm, and aq. ethanol (100 mL) to a 500 mL Erlenmeyer flask and 55 548+2 nm. The largest peak was at 548 nm. The intensity of stirred for 2 minutes. Approximately 8 mL of the 65 mg/mL the 548 nm peak was used to discriminate samples. phosphor suspension in DMSO was then added to the Linkage of Avidin-Phosphor Conjugate to Cell Surface mixture. This suspension was stirred for an additional 2 Marker minutes, then transferred to centrifuge tubes and centrifuged A lymphoblastoid cell line (Human Genetic Mutant Cell to separate the phosphor particles. The pellets were washed Repository #GM07092) was cultured in RPMI 1640 media twice with 95% aq, ethanol centrifuging each time. The containing 15% heat inactivated fetal calf serum. A suspen resulting particles were collected and dried overnight under sion of cells (10’ cells) was centrifuged and resuspended in vacuum at approximately 30° C. A quantity (127mg) of dry an equal volume of phosphate buffered saline (PBS) pH 7.4. silanized phosphors were resuspended in 1.5 mL of DMSO Cells were washed two times in PBS and resuspended to a (phosphor stock). final concentration of 5x10 cells/ml. These cells were then A solution containing 1.19 mg of avidin (Pierce) in 1.0 incubated with a mouse IgG1 monoclonal antibody to mL of borate buffer (954 mg sodium borate decahydrate and human 8-microglobulin, a Class I histocompatibility anti

Page 59

gen in polystyrene centrifuge tubes. The cells were immu a fresh tube, and up-converting phosphorescence was mea noprecipitated for 30 minutes at 4° C. with an antibody sured on the fluorimeter.

concentration of 10 g/ml. The cells were harvested by To scan for phosphor emission, the emission monochro centrifugation, washed twice in PBS, resuspended in PBS mator bandwidth was set to 8 nm and the spectra were and then aliquoted (250 L) into six fresh centrifuge tubes. scanned from 500 to 700 nm with a step size of 2 nm. Four of these samples received biotinylated goat anti-mouse Samples were also measured for FITC signal by exciting the IgG, while the remaining two received FTTC-labelled goat samples with 37 puM at A-490 nm with a 2 nm bandwidth. anti-mouse IgG. These immunoprecipitations were per Since the excitation wavelength (490 nm) and the emission formed at 4°C. for 30 minutes in volume of 400 L with a wavelength (514 nm) are very close for FTTC, higher final second antibody concentration of 20 g/ml. The cells O resolution was required to get separable signals than with were harvested and washed in PBS as above but were phosphor labelling. The intensity of the 490nm signal was resuspended in 50 L of blocking buffer (0.2% purified 240 Wlcm at the center of the well. FITC emission spectra casein in PBS, Tropix, Bedford, Mass.). The cell-antibody were scanned at 0.5 nm increments from 450 nm to 750 mm complexes were blocked in this solution for 30 minutes at with a 2 nm bandwidth on the emission monochromator. room temperature and then transferred to fresh tubes. 15 Sample 1 is the positive control and clearly yielded the A pre-blocked suspension (40 L) of either avidin highest emission signal. Sample 2 indicates that any non Phosphor conjugate, avidin-FITC, avidin, or unconjugated specific adsorption of the phosphors to the sample is limited Phosphor was added to four of the cell samples conjugated and is readily discriminated from signal attributable to with the biotinylated anti-mouse IgG (H&L). In addition, an avidin-conjugated phosphor and showing that avidin linked equal amount of pre-blocked avidin-Phosphorer unconju phosphors can specifically bind only when they are conju gated Phosphor was added to the remaining two cell samples gated with the probe, in this example through the biotin immunoprecipitated with the non-biotinylated FITC avidin linkage. Sample 3 is the negative control which labelled anti-mouse IgG (H&L). The avidin reporter conju contains no phosphors, only avidin. Sample 4 shows FITC gates or negative controls were pre-blocked as follows. conjugated avidin. Although FITC signals were observed on Avidin-Phosphor and Phosphor alone was diluted in block 25 the cell surface by laser microscopy, the signals were below ing buffer by adding 10 L of a 6.7 mg/ml suspension to a the level of detection on the fluorimeter for measurement of final volume of 100 L. Avidin-FTTC and the avidin alone FTC, and since there was no phosphor in the sample there controls were also diluted in blocking buffer by adding 27L was no significant phosphor signal. Samples 5 and 6 show of 2.5 mg/ml solution to a final volume of 100 L. These that FITC-conjugated primary antibodies can be detected reagents were blocked at room temperature for 3 hours with and that the presence of phosphor or avidin-phosphor does intermittent resuspension and then added to 50 L of cells not significantly disrupt binding of the primary antibody to labelled with biotinylated or non-biotinylated second anti its target antigen.

body. The avidin-biotin reactions were performed at room Linkage of Avidin-Phosphor Conjugate to DNA temperature for 30 minutes with occasional resuspension. Plasmid DNA (25 g) was nick translated in the presence The reactions were stopped by harvesting the cells by 35 of 20 mM dGTP, 20 mM dCTP, 20 mM biotin-14 dATP, 13 centrifugation and washing twice in blocking buffer. The mM dTTP, and 7 mM digoxigenin-11 duTP and purified by samples were resuspended in 100 L of blocking buffer and ethanol precipitation. The average size of the biotinylated, allowed to settle for 4-5 minutes. Slides for imaging were digoxygenin labelled fragments was estimated to be between prepared by pipetting 5 pull of settled cells from the bottom 200-300 nucleotides as estimated by gel electrophoresis. of the tube. Cells were imaged by confocal laser microscopy Approximately 20 pug DNA was immunoprecipitated for 1 under appropriate conditions to observe cell surface FTTC hour at 22° C. with 10 g/ml mouse monoclonal anti and upconverting phosphor signals. The observations are digoxigenin IgG1 solution (PBS) in a 200 L volume. An summarized in Table IV. equivalent reaction containing no DNA was also prepared. Each of the two samples were then aliquoted (50 L) into

TABLE TV 45 three fresh Eppendorf tubes.

The avidin-conjugates were blocked for 1 hour at room

Cell

Surface

Cell

Surface temperature by diluting 500 g of an avidin-phosphor

Type of goat Avidin Phosphor FITC suspension, unconjugated phosphor suspension, or avidin Tube anti-mouse IgG Conjugate Signal Signal solution in 300 L of blocking buffer. For each of the 50 samples (summarized below in Table V) 50 L of the

biotinylated biotinylated

Phosphor

m anti-digoxigenin conjugates was added to 150 L of pre 3 biotinylated Avidin -- m blocked avidin-conjugates or avidin and were incubated for 4. biotinylated Avidin-FTC - 30 minutes at room temperature. 5 FTC labeled Avidin-Phosphor -- Unbound avidin-conjugates were removed by resuspend 6 FTC abeled Phosphor - -- 55 ing 3x10' paramagnetic beads linked with sheep anti-mouse IgG (pre-blocked in blocking buffer). After incubation for 30

The remainder of the samples were used to resuspend minutes at room temperature with intermittent resuspension, paramagnetic, polystyrene beads bound with sheep anti the beads were separated on a magnetic rack and washed 4 mouse IgG. For each of the six samples, 3x10" beads were to 6 times in PBS. The antibody-DNA bound beads were prewashed with blocking buffer for 1 hour at room tempera then measured on the fluorimeter.

ture in Eppendorf tubes. The buffer was removed by aspi The samples were scanned from 500 to 700 nm with a ration while the tubes were in a magnetic rack. The magnetic bandwidth of 8 nm and step size of 2 nm. Each PMT value beads with anti-mouse IgG were allowed to bind to the reported (Table V) represents an average over 5 scans. antibody labelled cells for 1 hour at room temperature with Sample 1 is expected to provide the highest PMT signal intermittent resuspension. The magnetic beads were then 65 since biotinylated DNA is present and can bind to the collected on a magnetic rack, washed four times in blocking avidin-linked phosphors. Sample 2 indicates the level of buffer, resuspended in 100 L blocking buffer, transferred to nonspecific adsorption of the phosphors to the sample which

Page 60

is found to be insignificant since the PMT signal is observed of target present on the capturing surface. The target may be to be the same as that of the negative control (sample 4) linked directly to the capturing surface or may be immobi which contains no phosphors. Sample 3 is another control lized by interaction with a binding agent (e.g., specific and shows that the avidin-linked phosphors do not bind to antibody reactive with target, polynucleotide that binds the paramagnetic beads in the absence of DNA. Samples 5 target) that is itself linked to the capturing surface (such as and 6 show results of FTTC-labeled avidin used to validate in a sandwich immunoassay, for example). assay.

Detection of the phosphor bound to the capture surface is

TABLE W

effected using an excitation light that is focused from a low

intensity beam of large cross-section to a high intensity

Upconverting Phosphor Nucleic Acid Diagnostic Assay Results beam of small cross-section with the focal point of the beam being an or very near the capture surface. Focusing of the

PMTSignal PMTSignal excitation light is accomplished by transmission through

Sample DNA Reporter (W (a 546 mm) (V (a 514 nm) optical elements that have a very small focal length, such

that the beam diverges and becomes less intense, within a with digoxigenin linked

Phosphor short distance of the capture surface.

and biotin Since the intensity of the light emitted from the upcon 2 DNA labeled Silanized 10528 4.4022 verting phosphor labels is proportional to the excitation light with digoxigenin

Phosphor intensity raised to a power of two or greater, phosphors near and biotin 20 the focal point of the excitation source will emit significantly 3 No DNA Avidin 16302 3.5779 more light than those remaining in suspension in the sample linked away from the capture surface. Therefore, binding of upcon 4 DNA labeled

Phosphor

verting phosphor linked probes to the capture surface will with yield an increase in emitted light intensity measured from digoxigenin 25 the sample as a whole or as measured from a control sample and biotin in which phosphors do not bind to the capture surface. 5 DNA labelled with

Avidin 1.0484 8.4394 Emitted light intensity may be plotted as a function of target concentration using for standardization (calibration) a series and biotin of samples containing predetermined concentrations of tar 6. No DNA FTC- 10899 3.5779 get. The emitted light intensity from a test sample (unknown Avidin concentration of target) can be compared to the standard curve thus generated to determine the concentration of target.

Phosphor Downconversion Evaluation

A sample of the (Yossybooseroos).OS phosphors were Examples of suitable homogeneous assay formats scanned for the presence of a downconverted signal. This 35 include, hut are not limited to, immunodiagnostic sandwich was accomplished by exciting a sample of the monodisperse assays and antigen and/or antibody surface competition phosphors described above (4x10'M in DMSO) with 1.3 assays.

mW of monochromatic light at 350 nm with a 16 mm Homogeneous Assay Example 2 bandwidth for the excitation source. Detection was accom plished by scanning this sample from 350 to 800 nm with a Another embodiment allows for the accumulation of monochomator bandwidth of 8 nm. Scanning was performed upconverting phosphor linked probes at the detection sur in 2 nm increments. No downconversion was observed. face by the application of centrifugal or gravitational set Moreover, no downconversion was seen at the excitation tling. In this embodiment an upconverting phosphor is wavelengths cited by Tanke et al. (U.S. Pat. No. 5,043,265). linked to multiple probes. All the probes must bind to the Thus, the upconverting phosphors tested are unlike those 45 same target, although said binding can be accomplished at reported in Tanke et al. different locations (e.g., as antibody probes may target different epitopes on a single antigen). The multiprobe

HOMOGENEOUS ASSAYS phosphor can then be used to effect the aggregation of The multiphoton activation process characteristic of targets in solution or suspension in the sample. This aggre gation will result in the formation of a large insoluble upconverting phosphors can be exploited to produce assays 50 phosphor-probe-target complex that precipitates from solu that require no sample washing steps. Such diagnostic tion or suspension (FIG. 25). The aggregated complex assays that do not require the removal of unbound phosphor containing phosphors accumulates at a detection surface labels from the sample are herein termed homogeneous while nonaggregated material remains in solution or sus assays, and can also be termed pseudohomogeneous assays. 55 pension. Detection is accomplished as described in the Homogeneous Assay Example 1 above example using a sharply converging excitation beam. One embodiment of a homogeneous assay consists of the Evaluation of Up-converting Chelates use of an upconverting phosphor label linked to an appro Up-conversion has been performed in rare earth chelates priate probe (e.g., an antibody or DNA). The phosphor and rare earth salt solutions. Chelates of erbium and neody labeled probe specifically binds to a target (e.g., antigen or mium have been prepared with ethylenediaminetetraacetic nucleic acid) that is linked to a capturing surface. A suitable acid (EDTA) and dipicolinic acid (DPA). The erbium che capture surface can be the tip of a light carrying optical fiber lates were pumped using light near 793.5 nm from a (FIG. 23) or the bottom surface of a sample container (FIG. Ti:sapphire laser (the excitation scheme of Macfarlane 24). Upon incubation of the target-labelled capture surface 65 (1989) Appl. Phys. Let 54: 2301). This approach produced with the phosphor-labelled probe, phosphor particles will upconversion but not satisfactorily, which we attribute to accumulate at the capture surface as a function of the amount weak absorption for the first step due to the increase in

Page 61

SS 56 linewidth in the chelate over the low temperature crystal Na(YYbHoF: wherein x is 0.7 to 09, y is 0.0995 to 0.2995,

used for the up-conversion laser.

The neodymium chelates were excited with light near 580 Na(YYbTm)F: wherein x is 0.7 to 09, y is 0.0995 to 0.2995, nm from a Nd:YAG-pumped dye laser (following the exci and z is 0.0005 to 0.001; or tation scheme of Macfarlane et al. (1988) Appl. Phys. Lett 5 52: 1300). An emission spectrum for the emitted (Y.YbE)OS: whereinx is 0.7 to 09, y is 0.05 to 0.12; z is

up-converted light at 380 nm is shown in FIG. 32. We estimate the up-conversion cross section to be 8. A composition of claim 6, wherein the microcrystalline 10-275) cm2>> for this experiment. 10 up-converting phosphor has a formula selected from the We have also observed up-conversion in thulium acetate group consisting of:

hexahydrate and holmium chloride hexahydrate in solution following the excitation schemes of Allain et al. (1990) (YosoYbo.1s Ero.o.2Fs; (Yo.87Ybo. 13Tmoloo.1)Fa; Electron. Lett. 26: 166, and Allain et al. (1990) Electron. (YosoYbo.19s.Hoooo2)Fs;

Lett. 26:261, respectively. The salts were dissolved in heavy 5 (GdoosYbo.1s Eroo2)F3; (Gdos, Ybo.13Tmooo..)Fs; water, and excitation was performed using a krypton laser. (Gdosoyboisshoooo..)Fs;

Although the up-conversion was weak, the up-conversion (Yoss Ybooseroos)2O2S: (Yos, Ybo.13Tmolool)2O2S: should be improved if chelated compounds are used instead (YoosYbo.19shooooz2)O.S: of dissolved salts.

Although the present invention has been described in 20 (GIYbooBoo...O.S:

(Gdos, Ybo.3Tmoo)2O2S:

some detail by way of illustration for purposes of clarity of (GdoosYbogshoooo2)2O2S.

understanding, it will be apparent that certain changes and 9. A composition for diagnostic detection of an analyte, modifications may be practiced within the scope of the comprising an up-converting phosphor non-covalently claims. bound to antibodies, avidins, lectins, Staphylococcus aureus We claim: Protein A, antigens, polypeptides and polynucleotides.

1. A composition comprising an up-converting inorganic 10. Acomposition of claim 9 wherein the microcrystalline phosphor which converts excitation radiation to emission up-converting phosphor has the formula: radiation of a shorter wavelength and comprises at least one rare earth element in a host material and a probe selected from the group consisting of: antibodies, avidins, lectins, 30 Na(YYbBrF: wherein x is 0.7 to 09, y is 0.09 to 0.29, and z Staphylococcus aureus Protein A, antigens, polypeptides is 0.05 to 001;

and polynucleotides.

2. A composition according to claim 1, wherein the NaOYYbHo)F: wherein x is 0.7 to 09, y is 0.0995 to 0.2995,

up-converting inorganic phosphor comprises ytterbium and an emitter selected from erbium, holmium, thulium, and 35 terbium. NaCYYb,Tm)F: wherein x is 0.7 to 09, y is 0.0995 to 0.2995,

3. A composition according to claim 1, wherein the probe is attached to an up-converting inorganic phosphor by non (YYbE)OS: whereinx is 0.7 to 09, y is 0.05 to 0.12; z is covalent linkage. 0.05 to 0.12.

4. A composition according to claim 1, wherein the probe is streptavidin and the up-converting inorganic phosphor is 40 11. A composition of claim 9, wherein the microcrystal attached to the probe by noncovalent linkage. line up-converting phosphor has a formula selected from the 5. A composition of claim 4 further comprising a bioti group consisting of:

nylated magnetic bead, a streptavidin-coated magnetic bead, (Yo. soybo.1s Bro.oz.)F3; (Yos7Ybo. Tmo.o.o.)F3; an avidin-coated magnetic bead, or an immunoglobulin 45 (YosoYbo.19s.Hoooo...)Fs;

coated magnetic bead. (Gdo soYbo.1s Eroo2)Fa; (Gdos, Ybo.13Tmoloo.)Fa; 6. A composition for diagnostic detection of an analyte, (Gdosoybousshoooo...)F3;

comprising a microcrystalline up-convening phosphor (Yossybooseroos)2O2S: (YosYbo.13Tmoool)2O2S: covalently bound to antibodies, avidins, lectins, Staphylo (YosoYbo.19shooooz).O.S: coccus aureus Protein A. antigens, polypeptides and poly 50 nucleotides.

7. A composition of claim 6, wherein the microcrystalline

(Gdos, Ybo.13Tmool).O.S:

up-converting phosphor has the formula: (GdoosYboioshooooz)2O2S.

Na(YYbEF: whereinx is 0.7 to 09, y is 0.09 to 0.29, and z ck k k is sk