Anal. Chem. 2001, 73, 1403-1407
Articles
Detection of Biotin in Individual Sea Urchin Oocytes Using a Bioluminescence Binding Assay Agatha Feltus,† Anne L. Grosvenor,‡ Richard C. Conover,† Kimberly W. Anderson,*,‡ and Sylvia Daunert*,†
Departments of Chemistry and Pharmaceutical Sciences, Department of Chemical and Materials Engineering, University of Kentucky, Lexington, Kentucky 40506
The ability to detect biomolecules in single cells is important in order to fully understand the processes by which many biochemical events occur. To that end, we have developed a bioluminescence binding assay capable of measuring the intracellular biotin content of individual cells. The assay depends on competition between an aequorin-biotin conjugate (AEQ-biotin) and free biotin within the oocytes for binding sites on the protein avidin. The assay is performed by microinjecting each component into the oocytes and following the resulting bioluminescence within the oocyte upon triggering of aequorin. Results obtained using sea urchin oocytes show that the assay performed within the cells behaves in a manner consistent with assay theory. Using the assay, the individual biotin content of the oocytes is an average of ∼20 amol. To our knowledge, this is the first reported multicomponent binding assay to be performed inside an intact single cell. INTRODUCTION The ability to detect biomolecules in small volumes, that is, single cells, is important for a number of reasons. To be costeffective, high-throughput screening for new drug candidates requires the use of minimal sample sizes. In even smaller samples, single-cell analysis is a method by which information can be gained about the responses of individual cells within a given population. It has been known for a long time that individual cells respond differently to external stimuli.1,2 Most current methods of analysis depend on measuring the signal from a large number of cells, perhaps 104-106. Under such conditions, the response of the individual cell will be averaged out. The ability to examine the individual cell allows the scientist to relate the effects of age, toxic exposure, and developmental differences to the chemical composition of the cell. In that respect, as a first step, methods capable of determining such biomolecules in single cells must be developed. * To whom correspondence should be addressed. † Departments of Chemistry and Pharmaceutical Sciences. ‡ Department of Chemical and Materials Engineering. (1) Campbell, K. Chemiluminescence; Ellis Horwood: Chichester, England, 1988. (2) Ploem, J. S. In Applications of Fluorescence in the Medical Sciences; Taylor, D. L., Waggoner, A. S., Lanni, F., Murphy, R. F., Birge, R. R., Eds.; Alan R. Liss: New York, 1986; Chapter 13. 10.1021/ac001258a CCC: $20.00 Published on Web 02/28/2001
© 2001 American Chemical Society
In this work, we discuss the development of a bioluminescence binding assay for the vitamin biotin. Many current methods of single cell analysis depend on measuring molecules that are either natively fluorescent (i.e., proteins, NADH, fluorescent enzymatic products, etc.)3-10 or electroactive (i.e., neurotransmitters).11-19 The means of assaying these compounds is direct. To quantify biomolecules that are natively neither fluorescent nor electroactive, other detection methods must be developed. These methods must be sufficiently sensitive to detect the analyte of interest in a single cell while having enough selectivity to distinguish between the analyte and the many possible interfering agents present in such a complex sample. Path-length-dependent optical detection (absorbance) has low sensitivity due to the short path lengths involved, but it has been used for the detection of nitrate and nitrite in individual neurons.20 Mass spectrometry, although an excellent source of qualitative information,21,22 requires an internal standard. (3) Chang, H.-T.; Yeung, E. S. Anal. Chem. 1995, 67, 1079-1083. (4) Lillard, S. J.; Yeung, E. S.; Lautamo, R. M.; Mao, D. T. J. Chromatogr. A 1995, 718, 397-404. (5) Lillard, S. J.; Yeung, E. S. J. Chromatogr. B Biomed. Appl. 1996, 687, 363369. (6) Lillard, S. J.; Yeung, E. S.; McCloskey, M. A. Anal. Chem. 1996, 68, 28972904. (7) Lillard, S. J.; Yeung, E. S. J. Neurosci. Methods 1997, 75, 103-109. (8) Tong, W.; Yeung, E. S. J. Chromatogr. B Biomed. Appl. 1997, 689, 321325. (9) Tong, W.; Yeung, E. S. J. Chromatogr. B Biomed. Appl. 1996, 685, 35-40. (10) Tong, W.; Yeung, E. S. J. Neurosci. Methods 1997, 76, 193-201. (11) Olefirowicz, T. M.; Ewing, A. G. Anal. Chem. 1990, 62, 1872-1876. (12) Wightman, R. M.; Jankowski, J. A.; Kennedy, R. T.; Kawagoe, K. T.; Schroeder, T. J.; Leszczyszyn, D. J.; Near, J. A.; Diliberto, E. J., Jr.; Viveros, O. H. Proc. Natl. Acad. Sci. U.S.A. 1991, 88, 10754-10758. (13) Kennedy, R. T.; Jorgenson, J. W. Anal. Chem. 1989, 61, 436-441. (14) Clark, R. A.; Ewing, A. G. Mol. Neurobiol. 1997, 15, 1-16. (15) Chen, G.; Ewing, A. G. Crit. Rev. Neurobiol. 1997, 11, 59-90. (16) Chen, G.; Gutman, D. A.; Zerby, S. E.; Ewing, A. G. Brain Res. 1996, 733, 119-124. (17) Pihel, K.; Hsieh, S.; Jorgenson, J. W.; Wightman, R. M. Anal. Chem. 1995, 67, 4514-21. (18) Cooper, B. R.; Wightman, R. M.; Jorgenson, J. W. J. Chromatogr. B Biomed. Appl. 1994, 653, 25-34. (19) Cooper, B. R.; Jankowski, J. A.; Leszczyszyn, D. J.; Wightman, R. M.; Jorgenson, J. W. Anal. Chem. 1992, 64, 691-694. (20) Cruz, L.; Moroz, L. L.; Gillette, R.; Sweedler, J. V. J. Neurochem. 1997, 69, 110-5. (21) Whittal, R. M.; Keller, B. O.; Li, L. Anal. Chem. 1998, 70, 5344-7. (22) Rubakhin, S. S.; Garden, R. W.; Fuller, R. R.; Sweedler, J. V. Nat. Biotechnol. 2000, 18, 172-5.
Analytical Chemistry, Vol. 73, No. 7, April 1, 2001 1403
Binding assays, on the other hand, have excellent selectivity for the analyte of interest as well as excellent detection limits when using a suitable label.23,24 Aequorin is a photoprotein native to the jellyfish Aequorea victoria that can be detected at levels as low as 10-21 mol.25,26 For this reason, it has been extensively used both in our laboratory and by others in the development of bioluminescence binding assays for various analytes,27-34 including the vitamin biotin.26,28 The biotin assay depends on competition between a biotinaequorin conjugate (AEQ-biotin) and free biotin for biotin binding sites on the protein avidin. Binding of avidin to AEQ-biotin causes inhibition of the bioluminescence signal emitted by the photoprotein. If, however, biotin successfully competes with AEQ-biotin and occupies some avidin binding sites, then there will be fewer bound molecules of AEQ-biotin and the amount of inhibition will be correspondingly less. When originally developed on a large scale (volumes of 400 µL), the detection limit for biotin was 10-18 mol.26 Recently, this assay has successfully been scaled down by a factor of 106 to volumes of 350 pL,35 which allowed us to demonstrate the principle of using these types of assays for the detection of biomolecules in volumes approaching those of certain single cells. To demonstrate that these bioluminescence assays can be employed in the detection of biomolecules in single cells, we employed the biotin assay for the detection of biotin in individual sea urchin oocytes (volume ≈ 525 pL). Our results indicate that the assay is capable of detecting biotin in single oocytes and that each oocyte does, indeed, contain a different amount of biotin. In addition, it shows that the assay behaves as expected in this rarified environment. EXPERIMENTAL PROTOCOL Apparatus. Experiments were performed using an inverted Zeiss Axiovert 35 microscope (Thornwood, NY) equipped with a 20× objective (NA ) 0.75). Reagents were injected using a Narishige IM-300 air-pressure-driven microinjector (Sea Cliff, NY). Micropipets were pulled immediately before use on a Flaming/ Brown micropipet puller (Sutter Instrument Co., Novato, CA) using borosilicate glass capillaries (Sutter catalog no. BF 100-78(23) Gosling, J. P. Clin. Chem. 1990, 36, 1408-1427. (24) Ishikawa, E. In Immunochemical Assays and Biotechnology for the 1990s; Nakamura, R. M., Kasahara, Y., Rechnitz, G. A., Eds.; American Society for Microbiology: Washington, D. C., 1992; pp 183-203. (25) Casadei, J.; Powell, M. J.; Kenten, J. H. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 2047-2051. (26) Witkowski, A.; Ramanathan, S.; Daunert, S. Anal. Chem. 1994, 66, 18371840. (27) Erikaku, T.; Zenno, S.; Inouye, S. Biochem. Biophys. Res. Commun. 1991, 174, 1331-1336. (28) Feltus, A.; Ramanathan, S.; Daunert, S. Anal. Biochem. 1997, 254, 62-68. (29) Galvan, B.; Christopoulous, T. K. Anal. Chem. 1996, 68, 3545-3550. (30) Smith, D. F.; Stults, N. L.; Mercer, W. D. Am. Biotech. Lab. 1995, April, 17-18. (31) Stults, N. L.; Stocks, N. F.; Rivera, H.; Gray, J.; McCann, R. O.; O’Kane, D.; Cummings, R. D.; Cormier, M. J.; Smith, D. F. Biochemistry 1992, 31, 14331442. (32) Ramanathan, S.; Lewis, J. C.; Kindy, M. S.; Daunert, S. Anal. Chim. Acta 1998, 369, 181-188. (33) Zatta, P. F.; Nyame, K.; Cormier, M. J.; Mattox, S. A.; Prieto, P. A.; Smith, D. F.; Cummings, R. D. Anal. Biochem. 1991, 194, 185-91. (34) Zenno, S.; Inouye, S. Biochem. Biophys. Res. Commun. 1990, 171, 169174. (35) Grosvenor, A. L.; Feltus, A.; Conover, R. C.; Anderson, K. W.; Daunert, S. Anal. Chem. 2000, 72, 2590-2594.
1404
Analytical Chemistry, Vol. 73, No. 7, April 1, 2001
15) having an outside diameter of 1.0 mm and an inside diameter of 0.78 mm. Volume determination experiments are described in detail in ref 36. Briefly, injection volume was determined by measuring the diameter of the bubble created upon injection of the aqueous reagent into microscope immersion oil. Equipment used for injection volume determination included a Hamamatsu SIT video camera (Bridgewater, NJ), a Panasonic model AG 6500 VCR (Seaucus, NJ), a Sony black and white monitor (Tokyo, Japan), and an IPM image shearing monitor (San Diego, CA). Unless otherwise stated, an injection is 100 pL. Bioluminescence intensity was quantified by integrating the light-emission peak and subtracting the background signal. Light collection was performed through a modified Rm-L microscope photometry system, which included a single-emission photomultiplier tube from Photon Technology International (South Brunswick, NJ). Light emission data were collected, and peaks were integrated using Felix software, also from Photon Technology International. All of the buffers and solutions were prepared using distilled deionized water (Milli-Q Water Purification System, Millipore, Bedford, MA). All of the reagents were of highest quality possible. AEQ-biotin (aequorin with an average of 2.6 molecules biotin attached per aequorin molecule) was purchased from Molecular Probes (Eugene, OR). Biotin, avidin, calcium chloride, and all of the buffer salts and additives were purchased from Sigma (St. Louis, MO). All of the assay reagents were prepared in 50 mM Tris-HCl, pH 7.8, containing 0.15 M NaCl, 0.1% (w/v) NaN3, 20 mM Na2EDTA, 10% (v/v) glycerol, and 0.1% BSA. They were filtered through a 0.2-µm pore-size syringe filter before being loaded into the micropipets. Oocyte Collection. Oocytes were collected from the sea urchin Arbacia punctulata by standard methods. Briefly, individual sea urchins were received live from Gulf Specimen Marine Laboratories (Panacea, FL) and acclimated in a marine aquarium for a day before spawning. An indiviual sea urchin was removed from the tank and placed over a small beaker containing artificial seawater (425 mM NaCl, 9 mM KCl, 9.3 mM CaCl2, 25.5 mM MgSO4, 23 mM MgCl2, 2 mM NaHCO3, pH 8.0). Injecting the oral cavity of the urchin with approximately 2 mL of 0.5 M KCl, caused expulsion of its gametes. Because male urchins will expel sperm rather than oocytes, each urchin was isolated over its own beaker to prevent possible mixing of sperm and oocytes and fertilization (it is difficult to determine the sex of the urchin before gamete release). After the shedding was complete, the purplecolored oocytes (sperm are milky white) were collected and examined under a dissecting microscope to remove any other particles with a small pipet. The oocytes were then allowed to settle, and the supernatant removed. The oocytes were resuspended in an equal volume of calcium-free artificial seawater before analysis. Oocytes to be analyzed were chosen on the basis of size. Each oocyte in this study had an approximate diameter of 100 µm for a volume of 525 pL. Injection of Reagents into Isolated Oocytes. A droplet of oocyte suspension was placed on a microscope cover slip into which picoliter-volume microvials had been fabricated. Reference (36) Grosvenor, A. L.; Crofcheck, C. L.; Anderson, K. W.; Scott, D. L.; Daunert, S. Anal. Chem. 1997, 69, 3115-3118.
Figure 1. Photograph of an individual sea urchin oocyte and micropipet inside a picoliter-volume microvial. The microvials were formed by laser ablation on a glass microscope slide. The vial dimensions were 100 µm in diameter and 45 µm in depth (total volume is 350 pL). For aesthetic reasons, a smaller oocyte (diameter 70 µm) is shown than was used in our assays (100 µm oocytes). With the 100 µm oocytes, the oocyte sits atop the vial, with the majority of the oocyte extending above the vial.
37 describes the manufacture of the microvials. Briefly, the vials are created by drilling into the coverslip with a KrF excimer laser. The vials thus created are cylindrical in shape with a diameter of 100 µm and a depth of 45 µm, resulting in a total volume of 350 pL. A micropipet was used to deposit an oocyte in its own vial by shifting the oocyte along the surface of the coverslip. The sides of the vial immobilized the oocyte, enabling the membrane to be pierced by the micropipet (see Figure 1). The oocyte was then injected with 100 pL of either calcium (100 mM) or calcium mixed with a 5 × 10-5 M avidin solution. An injection was considered successful if the membrane of the oocyte expanded. Next, a micropipet containing AEQ-biotin (5.8 × 10-7 M) was calibrated to inject 100 pL. The pipet pierced the cell, and the system was prepared for light collection. The AEQ-biotin solution was injected into the oocyte after light-collection began. Several cells injected with either calcium or calcium/avidin were analyzed. Determination of Effect of Spiking on Oocytes. To determine the effect that spiking of the oocyte with biotin would have, we injected oocytes (n ) 6 for each biotin concentration) with a mixture of AEQ-biotin (5.8 × 10-7 M) and biotin (1 × 10-4, 1 × 10-8, or 1 × 10-9 M) before injecting the calcium/avidin solution (100 mM CaCl2 and 5 × 10-5 M avidin). Into a separate set of oocytes (n ) 6), the same injections were made, except that the first micropipet contained only 100 mM CaCl2. After the signal was determined for each oocyte, the percent inhibition of the signal from the AEQ-biotin conjugate was determined for each oocyte. This was done by taking the signal with avidin and subtracting the average signal from the oocytes in the absence of avidin. The resulting number was then divided by the average signal from the oocytes without avidin. Control experiments were also performed in microvials containing artificial seawater but no oocyte. (37) Crofcheck, C. L.; Grosvenor, A. L.; Anderson, K. W.; Lumpp, J. K.; Scott, D. L.; Daunert, S. Anal. Chem. 1997, 69, 4768-4772.
Calibration Curve for Biotin. A calibration curve for biotin was constructed by injecting the assay components into an empty microvial instead of into an oocyte. After the oocyte suspension was placed on the coverslip, some microvials remained empty. Into these microvials, two micropipets were moved, one containing AEQ-biotin (5.8 × 10-7 M) premixed with various concentrations of biotin (10-4-10-9 M) and a second containing either the calcium/avidin solution described earlier (100 mM CaCl2 and 5 × 10-5 M avidin) or 100 mM CaCl2. The AEQ-biotin/biotin was injected first, followed by the calcium or calcium/avidin solution, while measuring the bioluminescence signal. After the signal was integrated, the amount of inhibition was determined as described above. Three or four replicates were used for each concentration of biotin. Determination of Biotin in Single Cells. To determine the intracellular biotin content of individual oocytes, we injected oocytes (n ) 6) with AEQ-biotin (5.8 × 10-7 M) before injecting the calcium/avidin solution (100 mM CaCl2 and 5 × 10-5 M avidin). A series of oocytes were also injected with calcium buffer not containing avidin. After the bioluminescence signal had been collected and integrated, the percent inhibition was calculated, and the concentration of biotin present was determined from the calibration curve. RESULTS AND DISCUSSION In typical binding assays, which use volumes in the hundreds of microliters, optimization requires several steps. These include changing the amount of labeled ligand and binder, incubation time, buffer composition and pH, etc. Although all of these parameters need to be considered in the development of picoliter-volume assays, there are several other challenges that must also be overcome. For example, buffer evaporation, which plays a practically insignificant part in large-scale assays, assumes a much larger role in picoliter volumes. Lengthy incubations must, therefore, be shortened, because decreasing buffer volumes cause irreproAnalytical Chemistry, Vol. 73, No. 7, April 1, 2001
1405
ducible changes in ionic strength and pH. In addition, the high surface area-to-volume ratio in the microvials can lead to high levels of protein adsorption, which results in a loss of signal. These problems can be overcome through the addition of glycerol and BSA to the assay buffers to prevent evaporation and adsorption, respectively. As the volume of the vial decreases, diffusion times are also decreased and mixing of the reagents occurs almost instantaneously. When this assay was originally scaled down to picoliter volumes, there were two incubation steps in the assay: one after biotin and avidin were mixed and another after the AEQ-biotin was added (a third injection added Ca2+).35 It was found that one incubation step could be eliminated by adding Ca2+ with the biotin and avidin then adding AEQ-biotin after incubation, thus reducing the total amount of time the assay components spend in the picovial. When attempting to perform an assay on individual oocytes, the largest challenge is the evaluation of assay performance. The main problem stems from the major hypothesis of single cell analysis: that is, that each cell has a biochemical composition that is different from its neighbors. If this is true, when the assay is performed on individual cells, the amount of bioluminescence from each cell will be different. How then do we evaluate the assay in terms of reproducibility and response characteristics? A series of experiments were performed to answer this question and to test hypotheses consistent with the theory that each oocyte would contain a different amount of biotin. Injection of Assay Components into Isolated Oocytes. One important factor in the performance of an assay in such small volumes is the amount of error that is introduced during the delivery of the assay components. Previous results have indicated that the single largest source of error in these assays lies in microinjection of the reagents.35-37 For this reason, we developed a protocol minimizing the number of injections necessary to perform the assay. This involves first injecting the AEQ-biotin (and free biotin, if needed) from one micropipet, followed by injection of the avidin mixed with Ca2+ from a second micropipet. When injecting into the oocytes (Figure 1 shows an individual oocyte and micropipet inside a picoliter-volume microvial), we found that there was insufficient internal Ca2+ to effectively trigger the bioluminescence reaction of aequorin, because injecting only AEQ-biotin into the cells elicits no photon emission (data not shown). This is due to either sequestration of the internal calcium stores by the oocyte or the presence of EDTA in the assay buffer. Assuming there is no compound inside the oocytes that significantly affects the bioluminescence of the AEQ-biotin, injecting each cell with only AEQ-biotin and Ca2+ should give a relatively constant bioluminescence signal. If, however, avidin were also injected, the assay would actually be performed, and the bioluminescence signal would, therefore, reflect the presence of intracellular biotin by giving a much more varied response from each cell. Figure 2 demonstrates that this is, indeed, the case. In Figure 2a, when only AEQ-biotin is injected into the oocyte, the signal across the oocytes is fairly constant (mean ) 2038 ( 271, 13.3% CV). However, in Figure 2b, in which avidin is also present, the variation in the signal is much larger (mean ) 1456 ( 502, 34.4% CV). It should be noted that these assays were performed using the same number of injections by premixing the avidin and Ca2+ in the second micropipet. The effect, therefore, is not due 1406 Analytical Chemistry, Vol. 73, No. 7, April 1, 2001
Figure 2. Injection of assay components into isolated oocytes: (A) oocytes were injected first with 100 mM CaCl2, then from a separate micropipet with 5.8 × 10-7 M AEQ-biotin/1 × 10-9 M biotin; (B) oocytes were injected first with a 100 mM CaCl2/5 × 10-6 M avidin solution, then from a separate micropipet with 5.8 × 10-7 M AEQbiotin/1 × 10-9 M biotin. The variation in signal intensities was greater when using the avidin solution than when using just calcium, even though the number of injections was the same.
to compounding injection errors. Effect of Spiking Oocytes with Different Concentrations of Biotin. Our next experiment investigated how spiking the oocytes with different concentrations of biotin would affect the signals obtained. Our hypothesis was that each oocyte contains a small amount of biotin that varies greatly from cell to cell. Therefore, if we spike the cell with a similar amount of biotin, the average signal from the oocytes will be significantly different from a control experiment in which the assay is performed following exactly the same protocol in a picoliter vial that does not contain an oocyte. However, if we spike the cells with a very large excess of biotin, the signal should be the same between the control and the spiked oocyte, because the amount of biotin in the spike should overwhelm the endogenous amount in the oocyte. In Table 1, this is indeed the phenomenon that occurs. When using spikes of 10-9 M or 10-8 M, the signal obtained from the oocyte is significantly different (p < 0.05) than that of a control sample containing no oocyte. However, when the oocyte is spiked with 10-4 M biotin, there is no difference between the control sample and the oocyte. It is also evident from Table 1 that the variance in signal is greater within the oocytes than in the control samples. This indicates two things: first that the endogenous amount of biotin within a population of oocytes is on the order of 10-8-10-9 M and second, that the oocytes do have a large variation in the amount of biotin they contain. Table 1 is given in percent inhibition in order to normalize for any variations in injection due to changing micropipets during the performance of the assay. Estimation of Biotin Content in Individual Oocytes. To determine the amount of biotin present in each oocyte, it was first
Table 1. Effect of Spiking Oocytes with Different Concentrations of Biotina
biotin (M)
jx
1 × 10-4 1 × 10-8 1 × 10-9
19.0 28.6 36.0
control SD 1.1 1.7 1.3
n
jx
4 3 4
20.0 49.9 16.9
oocyte SD
n
singnificant differenceb
13.0 49.2 8.1
5 6 6
no yes yes
a Data are given as percent inhibition. Oocytes were first injected with a 100 mM CaCl2/5 × 10-6 M avidin solution, then from a separate micropipet with 5.8 × 10-7 M AEQ-biotin spiked with varying concentrations of biotin. The amount of inhibition seen is calculated from the amount of luminescence from the AEQ-biotin without adding any avidin. b Student’s t-test (two-tailed) was used to test for significant differences (p < 0.05).
Figure 3. Standard curve for biotin. Microvials were first injected with a 100 mM CaCl2/5 × 10-6 M avidin solution then from a separate micropipet with 5.8 × 10-7 M AEQ-biotin spiked with varying concentrations of biotin. The results are plotted as percent inhibition ( S.E.M. (n ) 3-4).
necessary to construct a standard curve for biotin. This was accomplished by injecting the assay components into a picoliter vial in exactly the same way and using the same volumes that would be injected into the individual oocytes. The response for each biotin concentration is an average of at least three picoliter vials. As seen in Figure 3, the dose-response curve does not have a typically sigmoidal shape. In fact, it is linear over approximately 3 orders of magnitude. Finally, the assay was performed in individual oocytes. The percent inhibition from each oocyte was compared to those of the dose-response curve obtained from the control experiments in Figure 3. In these experiments, the inhibition from each oocyte was determined using the raw signal from each oocyte and the average bioluminescence signal from an equivalent number of oocytes injected with only AEQ-biotin and calcium. The amount
of biotin ranged from 12 to 32 amol per individual oocyte (mean ) 20.6 ( 7.0 amol). This is based on multiplying the concentration derived from the calibration curve by the volume of an oocyte (525 pL). This result appears to be within the expected range, because the only previous estimations of biotin content in cells have shown that they are below the single-digit attomole level in mammalian cells (volume ≈ 5 pL).38,39 Given that the volume of an oocyte is roughly 100 times larger than that of a typical mammalian cell, the concentration of biotin in mammalian cells and as determined here in the oocytes is equivalent. In conclusion, we have shown that a sensitive binding assay can be developed that is capable of detecting biomolecules in intact single cells. This assay is, to our knowledge, the first such one demonstrated. As a vitamin, biotin is a cofactor for necessary metabolic enzymes, and alterations in its content and the enzymes to which it is attached have been associated with colorectal cancer40 and various tumorigenic cell line,41 among several disease states. Because of its low concentration and the high level of characterization of the avidin-biotin system, biotin is an excellent model for studying the effect that toxins and other environmental/ clinical agents may have on single cells. In addition, we have shown that aequorin is well-suited to be a label for such assays. Moreover, we envision that the ability to perform such binding assays on single cells should lead to advances in many biochemical fields, once we are able to relate the concentration of a given biomolecule to the cell’s phenotype. ACKNOWLEDGMENT The authors acknowledge support from the National Institutes of Health and the Department of Energy. S.D. is a Cottrell Scholar and a Lilly Faculty Awardee. A.F. is supported by a National Science Foundation Graduate Fellowship, and both A.F. and A.L.G. were supported by an NSF-IGERT fellowship.
Received for review October 24, 2000. Accepted January 25, 2001. AC001258A (38) Chalifour, L. E.; Dakshinamurti, K. Biochem. Biophys. Res. Commun. 1982, 104, 1047-1053. (39) Dakshinamurti, K.; Mistry, S. P. J. Biol. Chem. 1963, 238, 294-296. (40) Cherbonnel-Lasserre, C. L.; Linares-Cruz, G.; Rigaut, J.-P.; Sabatier, L.; Dutrillaux, B. Int. J. Cancer 1997, 72, 768-775. (41) Bramwell, M. E.; Humm, S. M. Biochim. Biophys. Acta 1992, 1139, 115121.
Analytical Chemistry, Vol. 73, No. 7, April 1, 2001
1407
