Polarized light scattered from small particles -- spheres, fibers, dust, defects and biomaterial -- carries information about the optical and geometrical properties and even more esoteric properties of the scatterer. This highly developed, powerful, but under used technique is sensitive enough to detect a 0.5 nanometer radius change in a 1.0 micron diameter quartz fiber. The complete set of 16 polarized light scattering signals from any scatterer form a 16- element Mueller matrix, S_ij, which is the signature of the scatterer's status quo. Changes in cell morphology, internal arrangement and metabolic activity cause changes in radius, refractive index and absorption and therefore changes in the polarization signals S_ij which act as the probe. We demonstrate how the S_ij can type human red blood cells, do bacteria autopsies, distinguish between different types of white blood cells, bacteria, spores, and study cell life cycle and degradation as biomaterial undergoes biologically significant change.

A miniaturized, handheld biosensor for identification of hazardous biowarfare agents with high specificity is being developed. An innovative biological recognition system based on bacteriophage displayed peptide receptors will be utilized in conjunction with the miniature biosensor technology being developed. A bacteriophage library has been constructed to provide the artificial receptors. The library can contain millions of bacteriophage with randomly displayed peptide sequences in the phage outer protein coat which act as binding sites for the agents of interest. This library will be used to 'bio-pan' for phages that bind to a number of toxins and infectious agents and can, thus, provide an endless supply of low cost, reliable, specific, and stable artificial receptors. The biosensor instrument will utilize evanescent wave, planar waveguide, far-red dyes, diode laser and miniature circuit technologies for performance and portability.

Currently, there are four general approaches to microbiological analysis, i.e., the detection, identification and quantification of micro-organisms: (1) Traditional culturing and staining procedures, metabolic fermentations and visual morphological characteristics; (2) Immunological approaches employing microbe-specific antibodies; (3) Biotechnical techniques employing DNA probes and related genetic engineering methods; and (4) Physical measurement techniques based on the biophysical properties of micro-organisms. This paper describes an instrumentation development in the fourth of the above categories, physical measurement, that uses a combination of fluorometric and light scatter spectra to detect and identify micro-organisms at the species level. A major advantage of this approach is the rapid turnaround possible in medical diagnostic or water testing applications. Fluorometric spectra serve to define the biochemical characteristics of the microbe, and light scatter spectra the size and shape morphology. Together, the two spectra define a 'fingerprint' for each species of microbe for detection, identification and quantification purposes. A prototype instrument has been developed and tested under NASA sponsorship based on fluorometric spectra alone. This instrument demonstrated identification and quantification capabilities at the species level. The paper reports on test results using this instrument, and the benefits of employing a combination of fluorometric and light scatter spectra.

Recently, we have developed a biosensor for zinc based on the very tight binding of this metal by the enzyme carbonic anhydrase, which requires Zn(II) for catalysis. We were able to transduce the binding of the metal as a change in fluorescence intensity or lifetime by use of a colored inhibitor whose metal-dependent binding permits fluorescence resonance energy transfer (Forster transfer) to occur. We have extended this concept to include other metals and other analytes which may be bound in the native (or mutant) enzyme active site with a concomitant color change; the color change is transduced as a change in energy transfer efficiency. We have also recently demonstrated a similar approach, wherein the presence of a metal ion in the binding site is transduced as a change in fluorescence anisotropy. Results in cuvettes and with fiber optic sensors are shown.

Immunoassays based upon evanescent wave interactions are finding increased biosensing application. In these devices, the evanescent tail associated with total internal reflection of an incident beam at the substrate/solution interface provides sensitivity for surface-bound protein over bulk molecules, allowing homogeneous assays and real-time measurement of binding dynamics. Among such systems are surface plasmon resonance sensors and a resonant mirror device. Several research groups are also developing fluorescent fiberoptic or planar waveguide sensors for biomedical applications. We describe a second-generation planar waveguide fluoroimmunoassay system being developed in our laboratory which uses a molded polystyrene sensor. The 633-nm beam from a laser diode is focused into the 500 micrometer- thick planar waveguide by an integral lens. Antibodies to the desired analyte (hCG) are immobilized on the waveguide surface and fluorescence from bound analyte/tracer antibodies in a sandwich format is imaged onto the detector. The geometry of the waveguide allows several zones to be detected, providing the capability for on-sensor calibration. This sensor has shown picomolar sensitivity for the detection of hCG.

In order to maximize the applications of advanced optical techniques for immunoassay it is critical that one can analyze multiple analytes simultaneously. One method of creating a multiple analyte sensor is to pattern antibodies against ligands of interest onto distinct regions of a single waveguide for fluorescence immunoassay. To achieve protein patterning, we are using a photolabile protected biotin with the caging moiety MeNPOC, (methyl nitropiprionyloxy carbonyl). The biotin molecules within a given region are selectively deprotected by exposure to ultraviolet light and subsequently bound to streptavidin. Incubation with a biotinylated antibody results in a functionalized region on the surface. This paper characterizes the method for immobilizing caged biotin onto the wave guide surface. Two surface biotinylation methods were examined silane coupling via aminpropyl triethoxy silane to a biotin-MeNPOC ester, and adsorption of biotin-MeNPOC conjugated bovine serum albumin. Using an I-125 label, protein surface densities have been determined for streptavidin bound to protected and deprotected surfaces. In addition, the duration of ultra-violet light exposure was evaluated to assess the ultimate effect on bound protein. The ability of an antibody bound within a patterned region to detect its corresponding analyte was determined.

The minimal detectable concentration (MDC), the smallest analyte concentration an immunoassay can reliably measure, is a fundamental property of an assay. Different interpretations of 'the smallest concentration an immunoassay can reliably measure' have led to different mathematical formulations of the MDC definition. We interpret this concept to mean the smallest analyte concentration the immunoassay may report as greater than the zero dose with high probability, say 0.95. Using Bayes' theorem we have developed a new MDC definition and shown that each of the current definitions approximates some component of the new one. We extend our paradigm for computing the MDC and derive a statistical framework for analyzing uncertainty in small analyte concentrations. We compute for any immunoassay measurement the probability that it exceeds the zero dose, its 95% confidence interval, its coefficient-of-variation (CV) and the probability that it is greater than any previous measurement. We illustrate the framework in a study of theoretical data based on our previous analyses of the Abbott microparticle capture enzyme immunoassay assay (MEIA) for prostate- specific antigen (PSA). The paradigm provides a sounder conceptual and computational approach for measuring reliably low concentrations of biological substances and for defining a positive result for screening and diagnostic tests.

There is no closed form method for the evaluation of the precision of parameter estimates for nonlinear regression problems. There are, however, a number of approximate methods which require different restrictive assumptions and require different amounts of computer time. While none of these estimates are exactly correct, some provide substantially better results than others. This work discusses some of these methods with specific application to frequency domain fluorescence lifetime measurements. A significant conclusion being that the commonly used asymptotic standard errors do not provide reasonable estimates of the precision of parameter values for nonlinear regression problems.

Analytical instrument systems present assemblies of components which each introduces its own contribution to the overall system precision. Combining realistic models of detector response with a Beer's Law model for absorbance (A) provides a prediction for the overall system uncertainty. For the constant detector uncertainty model, reasonable values of sample parameters lead to substantial increases in the absorbance at minimum relative concentration error (RCE), the value of the RCE and the limits of the range of A values for best RCE. Similar effects follow from the square root detector model. In both detector models these effects are increased by concentration uncertainty. Assay chemistry factors dominate the system RCE at low A, making high precision difficult at low A.

The opportunity to monitor the onset and progression of disease may be enabled by the smart implementation of biomedical optical engineering approaches to monitor tissue biochemistry. Uncorrected, these biochemical disturbances manifest themselves in microscopic structural changes which lead to gross pathophysiology and symptoms by which the disease is outwardly identified. Biomedical optical engineering techniques offer the opportunity to detect these biochemical changes and provide diagnostic information at earlier stages in the disease process, enabling greater efficacy of therapeutic intervention. In this paper, we concentrate on the fluorescence and phosphorescence lifetime spectroscopy due to advantages these techniques have to offer in tissues. The development of fluorescent and phosphorescent dyes which excite and re-emit in the near-infrared wavelength region promises the capacity for non- invasive biochemical sensing in tissues. Fluorescence intensity and fluorescence lifetime spectroscopies are established methods by which a fluorophore can provide sensing in dilute, non-scattering samples. However, fluorescence intensity or fluorescence lifetime spectroscopy in tissues or other scattering media is a complex problem. In order to extract the intrinsic fluorescence intensity for identification of the fluorophore concentration and yield, a priori information about tissue absorption and scattering must be obtained or assumed. Yet in tissues, the optical properties of absorption and scattering are highly variable. Nonetheless when successful, fluorescence intensity spectroscopy enables determination of the product of fluorescent yield and fluorophore concentration. In contrast to fluorescence intensity spectroscopy, fluorescence-lifetime tissue spectroscopy offers the ability to directly determine metabolite concentration independently of the concentration of fluorophore, whether it is endogenous or exogenous. Instead of monitoring the fluorescent intensity due to the re- emission process, the 'lifetime' or stability of the photon-activated fluorophore is measured. The lifetime of the activated state is defined as the mean time between absorption of the excitation photon and re-emission of a fluorescent photon. Typically, endogenous fluorophores have lifetimes on the order of nanoseconds while exogenous compounds have lifetimes ranging from sub nanosecond to milliseconds.

The study of intracellular trafficking using labeled molecules has been aided by the development of the cyanine fluorochromes, which are easily coupled, very soluble, resist photobleaching, and fluoresce at far-red wavelengths where background fluorescence is minimal. We have used Cy3-, Cy5-, and Cy5.5-labeled antibodies, antigen-binding fragments, and specifically binding single-stranded oligonucleotides to follow expression and trafficking of nucleolin, the most abundant protein of the nucleolus. Nucleolin shuttles between the nucleolus and the cytoplasm, and is also expressed on the cell surface, allowing us to test our techniques at all three cellular sites. Differentially cyanine-labeled non-specific antibodies were used to control for non-specific binding. Similarly, the differentially labeled non-binding strand of the cloned oligonucleotide served as a control. The multimode microscope allowed us to follow both rapid and slow redistributions of labeled ligands in the same study. We also performed 3-D reconstructions of nucleolin distribution in cells using rapid acquisition and deconvolution. Microinjection of labeled ligands was used to follow intracellular distribution, while incubation of whole cells with antibody and antigen-binding fragments was used to study uptake. To unambiguously define trafficking, and eliminate the possibility of interference by cross-reactive proteins, we transfected mouse renal cell carcinoma cells that express cell surface nucleolin with human nucleolin. We used microinjection and cell surface staining with Cy3- or Cy5- labeled monoclonal antibody D3 (specific for human nucleolin) to assess the cellular distribution of the human protein. Several clones expressed human nucleolin on their surfaces and showed high levels of transport of the human protein into the mouse nucleus and nucleolus. This distribution roughly parallels that of mouse nucleolin as determined by labeled polyclonal antibody. We have used these engineered transfectants to determine whether the cell surface-expressed xenogeneic nucleolin can serve as a target for antibodies in vivo.

Because of the small sample sizes that are typically inserted onto the separation column in capillary electrophoresis (1-100 nL), ultrasensitive detection strategies are required. The common detection approach used in CE is laser-induced fluorescence with He-Cd, Ar or Kr ion laser excitation. We are developing a detector system which utilizes solid-state diode lasers and avalanche photodiodes to produce a low-cost, durable and ultrasensitive fluorescence detector for CE applications. Along these lines, we have prepared some labeling dyes which readily conjugate to primary bioamines and show absorption and emission properties in the near-IR allowing low-level analyses of these target analytes in complex sample matrices. Our discussion will focus on the properties of diode lasers and avalanche diodes for fluorescence detection in CE applications. In addition, we discuss the characteristics of these near-IR dyes and tagging dyes synthesized in our laboratory for the covalent labeling of bioamines and their use in CE. The specific bioanalysis examples that we present utilizing near-IR fluorescence detection for CE are amino acids separations. In addition, we also discuss the ability to do time-resolved fluorescence measurements during CE for peak identification purposes.

Recent investigations by our group have demonstrated that near-infrared spectra collected from lysed blood solutions can be used to create clinically useful partial least squares (PLS) models for pH with standard errors of prediction below 0.05 pH units for a pH range of 1 (6.8 to 7.8). Further work was performed in order to discern the primary source of pH information in the spectra. Results from these experiments are presented using spectral data acquired over the spectral range of 1300 nm to 2500 nm from plasma, lysed blood and amino acids solutions. Data were analyzed by principal component analysis (PCA) and loading vectors were compared. Experiments were designed to eliminate possible correlation between pH and other components in the system in order to ensure variations in the spectral data were due to hydrogen ion changes only. Results indicate that variations in the spectral characteristics of histidine mimic those seen in lysed blood, but not those seen in plasma, suggesting that histidine residues from hemoglobin are providing the necessary variation for pH modeling in the lysed blood solutions.

AquaLite^R is a direct, bioluminescent label capable of detecting attomol levels of analyte in clinical immunoassays and assays for the quantitative measurement of nucleic acids. Bioluminescent immunoassays (BIAs) require no radioisotopes and avoid complex fluorescent measurements and many of the variables of indirect enzyme immunoassays (EIAs). AquaLite, a recombinant form of the photoprotein aequorin from a bioluminescent jellyfish, is coupled directly to antibodies to prepare bioluminescent conjugates for assay development. When the AquaLite-antibody complex is exposed to a solution containing calcium ions, a flash of blue light ((lambda) _max equals 469 nm) is generated. The light signal is measured in commercially available luminometers that simultaneously inject a calcium solution and detect subattomol photoprotein levies in either test tubes or microtiter plates. Immunometric or 'sandwich' type assays are available for the quantitative measurement of human endocrine hormones and nucleic acids. The AquaLite TSH assay can detect 1 attomol of thyroid stimulating hormone (TSH) in 0.2 mL of human serum and is a useful clinical tool for diagnosing hyperthyroid patients. AquaLite-based nucleic acid detection permits quantifying attomol levels of specific nucleic acid markers and represents possible solution to the difficult problem of quantifying the targets of nucleic acid amplification methods.

Copalis^TM,1 (coupled particle light scattering) is an innovative particle-based homogeneous immunoassay technology that allows rapid, sensitive, simultaneous determination of multiple analytes in a single fluid sample. Particles, their aggregates and cells are differentiated on the basis of light scatter measurements as they flow across a finely focused elliptical beam produced by a semiconductor laser. Two distinct complementary test formats have been developed, one based on particle self-agglutination and the other based on binding of colloidal gold particles to latex micro-spheres. In the first format, particle self- agglutination gives rise to dimers and oligomers which are classified and enumerated from the light scatter histogram. The extent of agglutination determines the test result. The second format uses micron-sized latex micro-spheres as the carriers onto which varying amounts of colloidal gold particles are bound in sandwich or competitive immunoassay formats. The main characteristic of this type of immunoassay is the relative broadening of light scatter histogram peaks for particles coated with colloidal gold. The extent of broadening is used to determine the concentration of the analyte. With both formats, the use of latex particles of different sizes in the same reaction mixture allows multiple simultaneous assays. This technology yields an analytical sensitivity of 10^-13 M, an improvement by several orders of magnitude over other homogeneous assay technologies.

The Human Genome Project (HGP) is an international project to develop genetic, physical, and sequence-based maps of the human genome. Since the inception of the HGP it has been clear that substantially improved technology would be required to meet the scientific goals, particularly in order to acquire the complete sequence of the human genome, and that these technologies coupled with the information forthcoming from the project would have a dramatic effect on the way biomedical research is performed in the future. In this paper, we discuss the state-of-the-art for genomic DNA sequencing, technological challenges that remain, and the potential technological paths that could yield substantially improved genomic sequencing technology. The impact of the technology developed from the HGP is broad-reaching and a discussion of other research and medical applications that are leveraging HGP-derived DNA analysis technologies is included. The multidisciplinary approach to the development of new technologies that has been successful for the HGP provides a paradigm for facilitating new genomic approaches toward understanding the biological role of functional elements and systems within the cell, including those encoded within genomic DNA and their molecular products.

A practical capillary array electrophoresis sequencer has been constructed and tested for large scale DNA sequencing. This system employs a scanning confocal optical system to detect DNA fragments electrophoresed on a capillary array. By using the recently developed energy transfer fluorescent primers, sensitive four-color detection is facilitated while exciting with just a single laser wavelength (488 nm) from an argon ion laser. Software for facile reduction of the images to four-color trace files and for automated base-calling have also been developed. This system can detect up to 25 capillaries at a time and has a raw sequencing rate of approximately 6 kilobases/hour. Applications to mitochondrial D-loop DNA sequencing are presented as a demonstration.

A four-color multiple capillary DNA sequencer is used to determine the methylation pattern of double stranded DNA. The DNA sample is treated with bisulfite under conditions that convert cytosine to uracil. Methyl-cytosine is inert under these reaction conditions. After PCR amplification, the reaction products are subjected to a four-color fluorescent Sanger sequencing reaction. The sequence is then determined by use of capillary electrophoresis. Comparison of the sequence obtained after bisulfite treatment with the original sequence reveals that certain of the Cs in the original sequence are converted to Ts. This conversion occurs only if the original C was not methylated. Those Cs that are common to both sequences were methylated in the original sequence. Methylation patterns have been implicated in aging, developmental biology, and cancer; however, there has been no simple and rapid method for determining the methylation pattern in genomic DNA. The method described in this paper is quick, simple, and accurate, and demonstrates an exciting application of capillary electrophoresis DNA sequencing.

Nucleic acid electrophoretic separations in unentangled water-soluble polymers are rapid and applicable over the length range 500 bp to greater than 1.5 Mbp, depending upon the polymer matrix chosen. We have been able to achieve high resolution separations in 3 - 4 minutes, even in the Mbp size range. Pulsed field electrophoresis is necessary for chains longer than about 20 kbp and is beneficial above about 2 - 3 kbp. The separation speed is a consequence of several factors: (1) the low viscosity of an unentangled polymer solution. (2) the high electric fields usable in pulsed field electrophoresis. And (3) the use of short (ca. 10 cm) capillaries. Electrophoretic measurements are guided by video fluorescence microscopy, which is used to study the dynamics of the separation process. Analysis of image sequences demonstrates that the segmental motions of DNA, which generate the size-dependent mobilities, are quite similar to those previously observed in entangled linear polymers and cross-linked gels. We describe our most recent separations and the current state of understanding of DNA dynamics in unentangled polymer matrices.

The mapping of a disease locus to a specific chromosomal region is an important step in the eventual isolation and analysis of a disease causing gene. Conventional mapping methods analyze large multiplex families and/or smaller nuclear families to find linkage between the disease and a chromosome marker that maps to a known chromosomal region. This analysis is time consuming and tedious, typically requiring the determination of 30,000 genotypes or more. For appropriate populations, we have instead utilized pooled DNA samples for gene mapping which greatly reduces the amount of time necessary for an initial chromosomal screen. This technique assumes a common founder for the disease locus of interest and searches for a region of a chromosome shared between affected individuals. Our analysis involves the PCR amplification of short tandem repeat polymorphisms (STRP) to detect these shared regions. In order to reduce the cost of genotyping, we have designed unlabeled tailed PCR primers which, when combined with a labeled universal primer, provides for an alternative to synthesizing custom labeled primers. The STRP pattern is visualized with an infrared fluorescence based automated DNA sequencer and the patterns quantitated by densitometric analysis of the allele pattern. Differences in the distribution of alleles between pools of affected and unaffected individuals, including a reduction in the number of alleles in the affected pool, indicate the sharing of a region of a chromosome. We have found this method effective for markers 10 - 15 cM away from the disease locus for a recessive genetic disease.

Our group is developing optical solid-state, near-IR (NIR) fluorescence detection systems for the analyses of DNA in restriction mapping and sequencing applications. Specifically, we are investigating a base-calling scheme using fluorescence lifetime discrimination in the NIR implementing intramolecular heavy-atom modified NIR fluorescence dyes which can be configured in a single lane, single fluor format. Results are presented concerning the ability to perform lifetime measurements on-line during capillary gel electrophoresis. Due to the high sensitivity associated with NIR fluorescence detection and small injection volumes in capillary electrophoresis, micro-reactor systems are also being developed which will potentially reduce the sample size for preparation of DNA sequencing ladders. The Sanger dideoxy-terminated fragments are prepared using solid-phase sequencing strategies in these micro-capillary reaction chambers. Our discussion focuses on the stability of the immobilized DNA template under typical sequencing conditions and the ability to sequence long templates using this strategy. Finally, we discuss our work on the preparation of nuclear staining dyes which show absorption and emission properties in the NIR for the low-level detection of restriction fragments. Fluorescence spectra of these dyes in the presence of dsDNAs is presented.

Early matrix assisted laser desorption/ionization (MALDI) studies of oligonucleotides elucidated many issues inherent to the analysis of nucleic acids. These studies demonstrated that fragmentation is an important issue and is responsible for some of the current limitations associated with the technique. Results in our laboratory and in others confirm that fragmentation is dependent on both oligonucleotide sequence and matrix composition. The proposed fragmentation pathway consists of nucleobase protonation inducing base loss followed by backbone cleavage at the 3' C-O bond on the corresponding deoxyribose. Using theoretical calculations, we explain why the four nucleobases display different fragmentation propensities. Additionally, if nucleobase protonation initiates fragmentation, then differences in fragmentation amounts when different matrices are used must be related to the various matrices' abilities to protonate the nucleobase. Currently, we are investigating the relationship between fragmentation propensity and matrix properties. Proton affinities of five common MALDI matrices have been measured and compared to fragmentation probabilities. By deepening our understanding of the fundamental chemistry of DNA fragmentation, we hope to be able to develop the MALDI technique into a powerful robust and versatile methodology for nucleic acid analysis.

DNA analysis based on template hybridization (or hybridization plus enzymatic processing) to an array of surface-bound oligonucleotides is well suited for high density, parallel, low cost and automatable processing. Direct fluorescence detection of labelled DNA provides the benefits of linearity, large dynamic range, multianalyte detection, processing simplicity and safe handling at reasonable cost. Molecular Tool has applied a proprietary enzymatic method of solid phase genotyping (Genetic Bit^TM Analysis or GBA^TM) to DNA processing in 96-well plates and glass microscope slides. Glass slides are an inexpensive, convenient format with relatively low fluorescence and the capability for microfabrication of miniature arrays of oligonucleotides. Detecting the fluor-labelled GBA^TM dideoxynucleotides requires a detection limit of approximately 100 molecules/micrometer ². Commercially available plate readers detect about 1000 molecules/micrometer ², and an experimental setup with an Ar laser and thermoelectrically-cooled CCD can detect approximately 1 order of magnitude less signal. The current limit is due to glass fluorescence, and data is presented describing experimental modifications and analysis to improve the performance. Dideoxynucleotides labelled with fluorescein, eosin, tetramethylrhodamine, Lissamine and Texas Red are being characterized, and photobleaching, quenching and indirect detection with fluorogenic substrates have been investigated.

The synthesis of fluorescent DNA base analogs that can replace the natural bases adenine, thymine and cytosine and their incorporation into synthetic oligodeoxynucleotides is described. The effect on the stability of such modified nucleotides like 2-aminopurine and some pteridine derivatives, is studied by thermal melting studies and comparison with the corresponding unaltered oligonucleotides. The fluorescence spectroscopic properties and several applications of these new fluorescent DNA probes are described in greater detail. Structural information on the conformation of special oligonucleotides like hairpins, junctions and bulged duplexes can be obtained from fluorescence lifetime and fluorescence depolarization data. For example the fluorescence lifetime pattern of 2-aminopurine is a sensitive indicator of DNA base pairing. As examples the structure of the oligonucleotide-linker junction in a synthetically linked oligonucleotide hairpin and the base-pairing of the first 'deoxyribozyme' are discussed. A third application uses doubly, i.e., donor and acceptor, labeled oligonucleotides to measure distances by fluorescence resonance energy transfer.

We have developed a fiber optic sensor for rapid and direct analysis of PCR-amplified DNA fragments with minimal sample processing and real-time data readout. To accomplish this, a novel DNA-recognition system was built onto the surface of fused silica fibers. DNA fragments, labeled with a fluorophore during amplification, are bound to and detected at the fiber surface by means of evanescent wave excitation/emission. Excess unincorporated fluorescent single-stranded oligonucleotide PCR primers make only a small contribution to the signal, as the modified fiber surface only efficiently binds double-stranded DNA with the proper PCR-incorporated terminal nucleotide sequence (5'-ATGACTCAT-3'). The surface- bound double-stranded DNA recognition element utilizes a genetically engineered dimeric sequence-specific DNA binding protein. Self-assembly into the proper conformation for binding DNA occurs by means of specific interactions of the active dimer with the F_c domains of a layer of IgG molecules (antibodies) covalently attached directly to the fiber surface. The modified fiber surface is regenerated between samples by stripping away bound DNA with high salt concentrations.

A second generation optical design for the ABI PRISM^TM 377 DNA sequencer enhances sensitivity and increases sequencing throughput by taking advantage of simultaneous four color detection using a spectrograph and CCD array. On-axis laser illumination using a small turning mirror to excite sample fluorescence replaces the Brewster's angle geometry of the first generation sequencer. The small turning mirror blocks direct reflection of laser light from the collection path. Alignment precision and mechanical stability must be sufficient to avoid image wander on the spectrograph slit which would add noise. The spectrograph slit is underfilled to maximize signal strength and reduce mechanical vibration sensitivity. The primary noise sources have been identified and minimized such that the instrument is shot noise limited. The instrument may be reconfigured through software for use of additional or different fluorescent dye labels as required by new genetic analysis applications. The overall result is a flexible instrument platform with increased throughput, improved sensitivity and longer sequences read.

Practical DNA sequencing in a capillary format requires sieving fluids that can be replaced after each sequencing experiment. Uncrosslinked water soluble polymers were used to prepare DNA sieving networks, both through simple entanglements of linear polymers, and by end- capping with hydrophobic tails; aqueous solutions of the end-capped polymers yield viscous networks held together by micellar crosslinks. Both networks are effective resolving media for DNA sequencing ladders, and flow under stress, which has enabled the design of an automated capillary DNA sequencing system composed of a gel pump interfaced to an electrophoresis apparatus. The sieving properties of these polymers are described.

Using homogeneous 5' nuclease assay (an energy transfer probe) as the reporting system, the progress of the polymerase chain reaction is monitored by laser induced fluorescence. Progress from cycle to cycle and within the extension plateau of a given cycle is measured in the exponential, linear and plateau phases. Quantitation of initial concentration is best done using exponential phase data. Information on the dynamics of PCR and enzyme kinetics are an important byproduct of the investigation.

A new scanning fluorescence detector (SCAFUD) was developed for high-throughput genotyping of short tandem repeat polymorphisms (STRPs). Fluorescent dyes are incorporated into relatively short DNA fragments via polymerase chain reaction (PCR) and are separated by electrophoresis in short, wide polyacrylamide gels (144 lanes with well to read distances of 14 cm). Excitation light from an argon laser with primary lines at 488 and 514 nm is introduced into the gel through a fiber optic cable, dichroic mirror, and 40X microscope objective. Emitted fluorescent light is collected confocally through a second fiber. The confocal head is translated across the bottom of the gel at 0.5 Hz. The detection unit utilizes dichroic mirrors and band pass filters to direct light with 10 - 20 nm bandwidths to four photomultiplier tubes (PMTs). PMT signals are independently amplified with variable gain and then sampled at a rate of 2500 points per scan using a computer based A/D board. LabView software (National Instruments) is used for instrument operation. Currently, three fluorescent dyes (Fam, Hex and Rox) are simultaneously detected with peak detection wavelengths of 543, 567, and 613 nm, respectively. The detection limit for fluorescein-labeled primers is about 100 attomoles. Planned SCAFUD upgrades include rearrangement of laser head geometry, use of additional excitation lasers for simultaneous detection of more dyes, and the use of detector arrays instead of individual PMTs. Extensive software has been written for automatic analysis of SCAFUD images. The software enables background subtraction, band identification, multiple- dye signal resolution, lane finding, band sizing and allele calling. Whole genome screens are currently underway to search for loci influencing such complex diseases as diabetes, asthma, and hypertension. Seven production SCAFUDs are currently in operation. Genotyping output for the coming year is projected to be about one million total genotypes (DNA samples X polymorphic markers) at a total cost of

The number of bases that can be read in a single run by a DNA sequencing instrument that detects fluorophore labeled DNA arriving at a 'finish-line' located a fixed distance from the starting wells is influenced by numerous parameters. Strategies for improving the length-of- read of a DNA sequencer can be based on quantitative models of the separation of DNA by gel electrophoresis. The dispersion function of the electrophoretic system -- the relationship between molecular contour length and time of arrival at the detector -- is useful in characterizing the performance of a DNA sequencer. We adapted analytical representations of dispersion functions, originally developed for snapshot imaging of DNA gels, (samples electrophoresed for constant time), to finish-line imaging, and demonstrated that a logistic- type function with non-integral exponent is required to describe the experimental data. We use this dispersion function to determine the resolution length and resolving power of a LI-COR DNA sequencing system and a custom built capillary gel electrophoresis system, and discuss the factors that presently limit the number of bases that can be determined reliably in a single sequencing run.

Advances in high throughput genotyping protocols over the past few years have been remarkable. Most protocols developed to increase the throughput of genotyping rely on fluorescent based technologies for data acquisition and capture. In general, the number of genotypes per day quoted for these protocols are the result of extrapolations based on ideal situations. Here we present our experience with respect to the day to day problems of high throughput genotyping. Our laboratory is currently working on several genetic mapping projects in both mouse and man. For example, we are looking at the genetic basis for susceptibility to rheumatoid arthritis in a local native American tribe as well as a mouse animal model for the same disease. The machines used to collect gel image data are two Li- Cor infrared DNA sequencers adapted for genotyping. During the evolution of these projects, we have addressed issues concerning the tracking and flow of information from the initial extraction of DNA to the calling of the genotypes. In particular, we have focused on designing methods that are efficient, cost effective and can be easily taught to the technical staff. Computer programs have been written that record gel specific information (e.g. ID information), archive data and capture genotypes in a simple point and click environment. Instrumentation was purchased to ease the repetitive nature of sample allocation, reagent disbursement and gel loading. Using this system, we can produce genotype data on 96 individuals for 20 loci (1920 genotypes) in one day. Solutions to the overall flow of information at each of these junctions are discussed.

We are currently developing a significantly improved gel electrophoresis and detection system that will allow more than an order of magnitude enhancement in the speed of DNA fragment analysis. This system is based upon the technique of horizontal ultrathin gel electrophoresis (HUGE) which employs denaturing polyacrylamide gels that are 75 microns thick. Because of the thinness of the gel, very high electric field strengths may be applied without deleterious thermal effects on resolution. Our proprietary fluorescence detector that scans the gel during electrophoresis allows for the simultaneous detection of up to four fluorophores. Because of the efficiency of the system of light collection, the gel can be scanned at speeds fast enough to generate high resolution gel images despite the high speed of separations. In addition, we are able to increase sample density by collecting 500 datapoints across the width of the gel. The resulting instrument has the capability to separate and resolve single-stranded DNA molecules that are between 25 and 300 bases in length from each of 60 lanes in less than 45 minutes. With the advent of 96 lane gels and attendant automation, this instrument will have the ability to analyze 18,432 genotypes per day.

Two instruments have been developed in our laboratory for the direct amino-acid sequencing of peptides. The first is a time-of-flight (TOF) mass spectrometer that utilizes a coaxial, curved-field reflectron which provides simultaneous focusing of product ins formed by post- source decay (PSD) and obviates the need for scanning or stepping the reflectron voltage. The second is a matrix-assisted laser desorption/ionization (MALDI) ion trap mass spectrometer (ITMS). In this instrument, MALDI ions are trapped efficiently by ramping the trapping rf field, and a linear mass scale is achieved by modulating the axial excitation voltage during mass scanning. Monoisotopic mass selection and excitation is accomplished using phase- modulated SWIFT (stored waveform inverse Fourier transform) techniques, while efficient CID (collision-induced dissociation) performance is achieved using pulsed introduction of Xe and other heavier gases. Both mass spectrometers have been designed as compact instruments for sequencing peptides in partially fractionated mixtures.

Database searching with tandem mass spectra has been demonstrated as a powerful method to correlate peptides and proteins to gene sequences for identification. In this study preliminary data is shown for extending the cross-correlation database searching method to multi-stage mass spectra of polypeptides. This method may provide a means to directly identify large polypeptides or proteins without the necessity of proteolytic hydrolysis to produce peptides amenable to tandem mass spectrometry.

Matrix-assisted laser desorption/ionization (MALDI) in conjunction with time-of-flight mass spectrometry is an effective technique for analysis of DNA oligomers. However, with increasing oligomer length, mass resolution and sensitivity degrade as the effects of prompt and metastable decay, adduct formation, and ion kinetic energy spreads become increasingly dominant. Attempts to reduce these factors have focused on the use of volatile liquids as matrices because these molecules have lower binding energies both to themselves and to the DNA and theoretically require less energy for desorption. We have investigated thin films of DNA in liquid matrices frozen on a liquid nitrogen-cooled sample stage. Several solvent systems have resulted in reproducible mass spectra of single-stranded DNA oligomers.

Methods have been developed for the rapid and sensitive mass spectrometric characterization of peptides. The approach uses bioreactive mass spectrometer probe tips which are capable of modifying proteins for analytical purpose. In the demonstrated case, enzymatic proteolysis is initiated upon application of analyte to the probe tips, time allowed for digestion, and the results analyzed using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. The probe tips have been used for proteolytic mapping and partial sequence determination of picomol quantities of peptide while maintaining analysis times of approximately 5 minutes.

A sensitive and facile method is described to identify the glycosylation sites and site-specific heterogeneity in the carbohydrate portion of glycoproteins. In this procedure, the peptide backbone of the glycoprotein is cleaved enzymatically. The peptide/glycopeptide mixture is divided into three fractions. The first fraction is analyzed directly by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), while the other two aliquots are analyzed by MALDI-TOF MS after enzymatic release of the N-linked and N- and O-linked chains. Comparison of these mass spectra provides the molecular weight of each carbohydrate side chain and of the peptide to which it is attached. This information combined with the protein's amino acid sequence identifies the glycosylation sites and provides information concerning site-specific oligosaccharide heterogeneity. This approach is faster and simpler than procedures currently used for glycosylation site mapping and can be performed on as little as 10 picomoles of glycoprotein.

Matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS), because of its high sensitivity and relatively straightforward requirements for sample preparation, is contributing to the solution of structural problems in biology and to the development of therapeutic approaches through increased understanding of pharmacology and enhanced capabilities for quality control of pharmaceuticals. We are using a reflectron TOF- MS for the determination of molecular weights of individual compounds and the components of mixtures that are naturally occurring or are generated through enzymic digests, and employing the post-source decay mode to elucidate structural details. To maximize the sensitivity and information content of the spectra, varied matrices, derivative, and stepwise degradation procedures are being explored. Present studies include investigations of oligosaccharides, neutral glycolipids, gangliosides, glycoproteins, neuropeptides and proteins. Rules for fragmentation are being developed with model compounds and used for the structural elucidation of unknowns. When adequate sample amounts are available, the results are compared with low- and high-energy collision-induced decomposition spectra obtained with tandem MS in order to provide a data base for the correlation of spectral features and guidance in selection of approaches for scarce biological samples. Current projects include biophysical studies of glycoplipids, glycoproteins and oligosaccharides and investigations of the substance P receptor, transthyretin genetic variants and cisplatin-DNA interactions.

A general technique linking micro-scale affinity capture with matrix-assisted laser desorption/ionization (MALDI) mass spectrometric detection has been developed for the rapid, sensitive and accurate determination of antigens present in biological fluids. Strategies for the qualitative and quantitative determination of single and/or multiple antigens are presented.

A sensitive immunoprecipitation procedure suitable for matrix assisted laser desorption mass spectrometry (MALD-MS) has been developed for the analysis of proteins. The ultimate goal of this work is to develop a method suitable for analysis of proteins from biological fluids such as plasma, serum and expression media at physiologically relevant concentrations. Proteins are specifically precipitated from complex mixtures by immunoprecipitation with antibodies directed against portions of the protein of interest and the immunoprecipitates are analyzed by MALD-MS. Here we show that the method is applicable to two proteins, the 35 kDa glycoprotein, corticotropin releasing factor-binding protein (CRF-BP) and the 26 kDa homodimeric protein, activin. In order to elucidate the constraints for achieving our ultimate goal we report MALD-MS experiments and parallel analyses with gel electrophoresis and immunostaining. Using this MALD-MS immunoprecipitation procedure we have been able to successfully analyze samples of CRF-BP or activin in the presence or absence of serum.

Laser desorption mass spectrometry has been used for molecular diagnosis of cystic fibrosis. Both 3-base deletion and single-base point mutation have been successfully detected by clinical samples. This new detection method can possibly speed up the diagnosis by one order of magnitude in the future. It may become a new biotechnology technique for population screening of genetic disease.

In single molecule detection by laser induced fluorescence, a main problem is the low signal to noise ratio due to scattering of the exciting laser light. One common approach to solve this problem is the application of time resolved techniques. Here we present a high speed electronic (based on a pair of PC cards) specially suited for detection of TCSPC curves in a continuous flow system. The whole system works on two different time scales: a millisecond time scale (every millisecond a complete TCSPC curve is measured and stored) and a picosecond time scale (showing the fluorescence decay). The technique present here is of particular interest for applications such as fast DNA sequencing, where a distinction between the different bases solely by the decay times of the attached fluorescence labels is conceivable.

Recently, major advances have been reached in the fluorescence detection of small amounts of molecules in liquids, making possible even the detection of single molecules in liquid flows. Significant improvements of fluorescence detection techniques make single molecule detection feasible for many applications, especially in the field of molecular biology and genetics. For such techniques new compact and inexpensive lasers are desirable. laser diode systems are the most favorable candidates for such light sources. With the expanding number of available NIR-fluorescent dyes, the importance of cheap and reliable laser light sources above 630 nm will increase. But not only cw-laser systems are of growing interest. In a number of recent papers, the application of time-resolved fluorescence detection down to a single molecule level was shown to be of great use for further improving detection efficiency. Thus, one needs high- repetition rate pulsed laser diode systems with good time and optical performance, and detection electronics with high-speed and large data throughput. Here we present such a system, combining a pulsed diode laser system with excellent electrical and optical parameters, and a high speed electronic for time correlated single photon counting. This system is suitable for a broad range of applications in ultra sensitive fluorescence detection.