US007569396B1 (12) United States Patent (10) Patent No.: US 7,569,396 B1 Graham et al. (45) Date of Patent: Aug. 4, 2009 (54) CAFFEINE DETECTION USING 4,376, 110 A 3, 1983 David et al. INTERNALLY REFERENCED COMPETITIVE 4,703,017 A 10/1987 Campbell et al. ASSAYS 4,740,468 A 4/1988 Weng et al. 4,775,636 A 10, 1988 Moeremans et al. (75) Inventors: Anafior Q. Graham, Saratoga, CA 4,837,168 A 6/1989 de Jaeger et al. (US); Carralee Hathaway, Saratoga, 4,843,000 A 6, 1989 Litman et al. CA (US); Mark S. Geisberg, Arcadia, 4,849,338 A 7, 1989 Litman et al. CA (US) 4,855.240 A 8, 1989 Rosenstein et al. 4,861,711 A 8, 1989 Friesen et al. (73) Assignee: PurpleCow LLC, Redwood City, CA 4,916,056 A 4/1990 Brown, III et al. (US) 4.954,452 A 9, 1990 Yost et al. 5, 120,643 A 6/1992 Ching et al. (*) Notice: Subject to any disclaimer, the term of this 5,141,875 A 8, 1992 Kelton et al. patent is extended or adjusted under 35 5,212,060 A 5, 1993 Maddox U.S.C. 154(b) by 169 days. (21) Appl. No.: 11/530,232 (Continued) (22) Filed: Sep. 8, 2006 FOREIGN PATENT DOCUMENTS (51) Int. Cl EP O291194 A1 11F1988 GOIN 33/553 (2006.01) (52) U.S. Cl. ......................... 436/514; 436/20: 436/518; (Continued) 436/525; 436/810; 436/816:436/901; 435/7.1; 435/7.91; 435/7.92; 435/7.93; 435/970; 435/973; Primary Examiner Bao-Thuy L. Nguyen 435/975 (74) Attorney, Agent, or Firm Fenwick & West LLP (58) Field of Classification Search ................... 436/20, 436/514,518, 525,810,816,901; 435/7.1, (57) ABSTRACT 435/7.91, 7.92, 7.93, 970,973,975 See application file for complete search history. Methods, compositions, and apparatus for detecting the pres (56) References Cited ence of caffeine in a liquid sample are provided. In certain embodiments, an internally referenced competitive assay U.S. PATENT DOCUMENTS allows a very precise determination of a threshold value of caffeine for use in semiquantitative types of ligand-receptor 3,646,346 A 2, 1972 Catt 3,857,931 A 12/1974 Hager assays. By using a detection means that participates in two 3.996,345 A 12/1976 Ullman et al. assays, sensitivity is doubled in the maximum sensitivity 4,062,733. A 12/1977 Edwards et al. range and the range can be adjusted to match the predicted 4,104,029 A 8, 1978 Maier, Jr. concentration range of an analyte. This format and the mate 4,168,146 A 9, 1979 Grubb et al. rials described herein allow the assay to complete within three 4,181,636 A 1, 1980 Fischer minutes. In addition, this format accommodates common 4.264,766 A 4, 1981 Fischer attributes of liquid samples for detecting caffeine, such as the 4,313,734 A 2, 1982 Leuvering inclusion of milk or Sugar in a coffee-type beverage. 4,366,241 A 12/1982 Tom et al. 4,373.932 A 2f1983 Gribnau et al. 6 Claims, 10 Drawing Sheets da 250 230 235 105 (8 . 110 115 O& 122 205 210 x^ 125 215 . 9 . 250 . US 7,569,396 B1 Page 2 U.S. PATENT DOCUMENTS 6,368,875 B1 4/2002 Geisberg 6,461,873 B1 10/2002 Catania et al. 5,229,073. A 7, 1993 Luo et al. 6,500,665 B2 12/2002 Deegan 5,252.459 A 10, 1993 Tarcha et al. 6,557.484 B1 5/2003 Engelman 5,451,504 A 9/1995 Fitzpatricket al. 6,649,418 B1 1 1/2003 Geisberg 5,451,507 A 9, 1995 Skold et al. 6,699,722 B2 3/2004 Bauer et al. 5,559,041 A 9/1996 Kang et al. 6,797.481 B1 9/2004 Ullman et al. 5,591,645 A 1/1997 Rosenstein 5,610,072 A 3/1997 Scherlet al. FOREIGN PATENT DOCUMENTS 5,824,554 A 10/1998 McKay EP O323605 A2 7, 1989 6,001,658 A 12/1999 Fredrickson WO WO92,12428 A1 7, 1992 6,103,536 A 8/2000 Geisberg WO WO94/01774 A1 1/1994 6,153,147 A 1 1/2000 Craig WO WO95/O1775 A1 1, 1995 6,287,875 B1 9/2001 Geisberg WO WO96,27795 A1 9, 1996 U.S. Patent Aug. 4, 2009 Sheet 1 of 10 US 7,569,396 B1 110 FIGURE 1A Time FIGURE 1C FIGURE 1D U.S. Patent Aug. 4, 2009 Sheet 2 of 10 US 7,569,396 B1 FIGURE 2A easY 250 230 235 105 - 110 115 O 122 205 210 125 215 FIGURE 2B 9 250 Y O {) FIGURE 2C —o- d —o- Y O FIGURE 2D g U.S. Patent Aug. 4, 2009 Sheet 3 of 10 US 7,569,396 B1 FIGURE 3A 220 235 210 125 FIGURE 3B d Y O al FIGURE 3C p X U.S. Patent Aug. 4, 2009 Sheet 4 of 10 US 7,569,396 B1 110 I 205 21 O FIGURE 4A 110 I 205 210 110 205 210 FIGURE 4C 110 205 21 O U.S. Patent Aug. 4, 2009 Sheet 5 of 10 US 7,569,396 B1 —505 -520 - 110 - 510 515 215 525 Flow Direction FIGURE 5A - 110 - NSN 2 Flow Direction FIGURE 5B U.S. Patent Aug. 4, 2009 Sheet 6 of 10 US 7,569,396 B1 V\ 605 FIGURE 6A 110 FIGURE 6B U.S. Patent Aug. 4, 2009 Sheet 7 of 10 US 7,569,396 B1 625 FIGURE 6C FIGURE 6D U.S. Patent Aug. 4, 2009 Sheet 8 of 10 US 7,569,396 B1 FEAEGEFA SA E FEWSESSEE FASSEFEF AAF GERAFES (SSEE FIGURE 7 U.S. Patent Aug. 4, 2009 Sheet 9 of 10 US 7,569,396 B1 Cross-Reactivity 140 120 ?h 100 O X 8O (v. d 60 -- Caffeine O -& Theophylline SS 40 Theobromine 20::::::::::::::::::::::: :- Guanine -:- Adenine O 6OO 3OO 150 75 37 18 9 4.5 2 1 0 Analyte Concentration (ug/ml) FIGURE 8 U.S. Patent Aug. 4, 2009 Sheet 10 of 10 US 7,569,396 B1 Dose-Response I IC50 50 - 672-1 SS O O 1 2.25 4.5 9. 18 37 75 150 3OO 600 Caffeine ug/ml (log) FIGURE 9A Dose-Response / IC50 100 : 50 - 672-2 SS 1O 1 OO 1OOO 1 O,OOO Caffeine uglml (log) FIGURE 9B US 7,569,396 B1 1. 2 CAFFEINE DETECTION USING a positive indication for the analyte at the least sensitive INTERNALLY REFERENCED COMPETITIVE portion of the analyte concentration versus signal intensity ASSAYS curve. Thus, these immunochromatographic test strips are to be interpreted as positive for analyte in the sample when no RELATED APPLICATIONS 5 signal is seen at the test line. As one skilled in the art would recognize, the precision of the determination of analyte con Not applicable. centration is compromised in Such an assay. Immunoassays. One flavor of ligand-receptor assay is an BACKGROUND OF THE INVENTION immunoassay. Various known formats exist for immunoas 10 1. Field of the Invention says, including immunochromatographic test strips for The invention relates generally to methods, apparatus and detecting Small molecule analytes. One format uses a com compositions, useful for detecting the presence of caffeine in petitive immunoassay, for which the result is revealed as two a liquid sample, and more specifically, for detecting the pres lines (negative result) or one line (positive result). Another ence of caffeine in a beverage. 15 format displays a single line as an indication of a positive 2. Description of the Related Art result. Drawbacks of these formats include a very low dose Competitive ligand-receptor assays. Ligand-receptor response ratio at the positive/negative cutoff concentration assays take advantage of the ability of bioreagents to identify for Some analytes, multiplicity of necessary reagents, high and quantify minute amounts of a wide range of Substances, cost of production, and uncertain adaptability to the concen also referred to as analytes, with a high degree of specificity 20 tration range of interest for some analytes. and sensitivity. Competitive ligand-receptor assays are one Caffeine detection assays. Immunoassays can be used to variant of ligand-receptor assays in general. In competitive detect various analytes, including assays using anti-caffeine ligand-receptor assays, analyte Substances in the sample antibodies to detect the presence of caffeine. Existing clinical compete with another Substance, for example a signal-pro laboratory analyses and test strip formats using traditional ducing Substance, for a limited number of binding sites on the 25 immunoassay techniques for caffeine provide varying results. counterpart of the ligand-receptor pair. After the binding has In the laboratory setting, the scientific literature includes taken place, the amount of other Substance bound to the methods such as electrometric determination in which a caf counterpart is detected by any of several means. The signal feine-specific electrode is prepared from a caffeine-picrylsul intensity of competitive ligand-receptor assay is in an inverse fonate ion-pair complex dissolved in octanol; fluorimetric relationship with the concentration of analyte present; thus, a 30 determination in which a buffered solution of caffeine is sample with no analyte will give a maximum signal intensity, oxidized with N-bromosuccinimide and then reacted with and a sample with a range of analyte concentration will pro dimethyl o-phenylenediamine followed by a fluorescence duce less than a maximum signal.