- Email: [email protected]
Lab-on-valve: a useful tool in biochemical analysis
Trends
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
Lab-on-valve: a useful tool in biochemical analysis M.D. Luque de Castro, J. Ruiz-Jime´nez, J.A. Pe´rez-Serradilla Despite the short life-time of lab-on-valve (LOV) devices – the third generation of dynamic approaches invented by J. Ruzicka – they have proved to be of great interest in biochemical analysis. LOV works automatically with sequential injection to move micro or sub-micro volumes of liquids, gases and/ or beads in a ‘‘digital’’ manner by stopping, reversing, and accelerating flow rates in a way that is of interest in sample preparation – especially modes such as bead-injection spectrometry, micro affinity chromatography or inline LOV–bioreactor – and when LOV is coupled to high-resolution equipment, such as capillary electrophoresis and chromatography. LOV-based equipment has proved to be useful in handing ll volumes (e.g., in cell-culture and antibody studies, to assess the metabolic regime of living cells; and, in DNA assays, to detect single-stranded nucleic acid sequences). LOV therefore offers a promising way ahead for in-valve biochemical steps, which can be expanded by coupling to separation units (e.g., dialyzers, gas diffusers, pervaporators, and novel liquid–liquid extractors), which facilitate automatic interference removal and preconcentration in complex samples with low concentrations of analytes. ª 2008 Elsevier Ltd. All rights reserved. Keywords: Bead injection; Biochemical analysis; Coupling; Lab-on-valve; Sample preparation
1. Introduction M.D. Luque de Castro*, J. Ruiz-Jime´nez, J.A. Pe´rez-Serradilla Department of Analytical Chemistry, Annex Marie Curie Building, Campus of Rabanales, University of Co´rdoba, E-14071 Co´rdoba, Spain
*
Corresponding author. Tel./Fax: +34 957 218 615; E-mail: [email protected]
118
Automation and miniaturization of solution-based analysis are essential to make them fast and efficient for routine and research tasks in biochemical laboratories. Ideally, analytical equipment should be versatile, capable of accommodating a wide variety of assays without the need for system reconfiguration, and compatible with a wide range of detectors [1]. Since the introduction of flow injection (FI) in 1975 [2], thousands of scientific publications [3] have demonstrated its feasibility for automation, which has evolved through two new generations [4–6]: sequential injection (SI) [7], introduced in the 1990s and referred to as the second-generation FI; and, laboratory-on-valve (lab-on-valve, LOV), a recent continuous methodology, referred to as third-generation FI
and also introduced by Ruzicka [8], which works in SI mode and has downscaled reagent-based analysis to ml and sub-ml levels. During the past decade, miniaturization has gained increased interest in several analytical fields. Taking into account that the ultimate goal of miniaturization of reagent-based analysis is to decrease use of materials consumed and waste generated [8], its importance in the biochemical field is crucial, as both biochemical reagents and samples are usually scant and/or expensive. Thus, reducing consumption of expensive (or scant) sample and/or reagents through different approaches to miniaturization is obviously one of the most effective strategies to reduce analytical costs. A key way to downscale is to replace continuous flow by programmable flow, which allows moving liquids, gases and/or beads in a ‘‘digital’’ fashion, by stopping, reversing, and accelerating flow rates [5]. This principle of SI [4] has been downscaled, integrated into the LOV platform and used for miniaturization of reagentbased analysis, immunoassays, bioligandinteraction analysis, and ion-exchange and affinity chromatography [9], among others. A LOV platform consists of: (1) a transparent, monolithic structure made of Perspex; (2) a multiposition valve as the main component of the structure; and, (3) a propulsion unit, usually a syringe pump, characteristic of SI, to circulate the required liquids through the system, which is overall referred to as l-SI–LOV. Fig. 1 shows a scheme of a LOV manifold, which also includes details of the valve itself and the ways the flow cell can
0165-9936/$ - see front matter ª 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.trac.2008.01.004
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
Figure 1. (A) Overall configuration of l-SI–LOV with integrated UV–Vis absorption detection; (B) Details of the LOV platform; (C) Other detection modes (with permission of Prof. J. Ruzicka).
be integrated into the platform, depending on the detection technique. Integration of the multipurpose flow-cell, in which light measurements are carried out through optical fibers, means that the system is very versatile in measurements, although LOV can also be used with non-integrated cells for optical or electroanalytical measurements. In addition, an adequate selection of auxiliary units is essential in order to achieve the required configuration. Auxiliary units more frequently used are: one [8,10,11] or two [12–15] syringe pumps; an auxiliary six-port
Trends
selection valve [11,16,17]; one or two holding coils [12,18]; an auxiliary peristaltic pump [5,8,19–21]; a mixing coil [5,22]; and, T connections [12,19]. These units confer on LOV the high versatility, which, together with other characteristics, makes it an attractive tool in the biochemical field. LOV allows miniaturization of the flow channel; thus, sample and reagent volumes are downscaled to the 10– 20-ll range, while waste production is typically 0.1–0.2 ml per assay [1]. These systems are therefore positioned between traditional flow techniques that operate at the ml scale and the more futuristic designs of the micro total analysis systems (lTAS) concept, which is supposed to work at the nl scale. (The SI–LOV manifold is also known as a meso-fluidic system – ‘‘meso’’ from its capability to manipulate fluid between ‘‘micro’’ and ‘‘macro’’. Despite of this, most authors refer to SI–LOV manifolds to as micro-fluidic systems). Similarly to microchips, LOV must be assisted by external units for proper circulation of samples and reagents. By operating within the ll range with channel diameters about 0.5 mm, l-SI–LOV systems are compatible with real-life samples that often contain particles that can clog the lm-sized channels of nl-scale devices. The choice of LOV channel dimensions has two significant advantages: (1) a large volume-to-surface ratio, which minimizes the unwanted adsorption on channel walls that may result in carry-over; and, (2) even more importantly, this channel geometry allows LOV devices to be used as platforms for the bead-injection (BI) technique [1,5]. The main goal of a LOV system is sample preparation, which can involve sample dilution, common analyte preconcentration and derivatization or steps such as those based on BI. We discuss below these goals, the coupling of LOV approaches to high-resolution equipment and very different detectors, and salient applications in biochemical analysis. An approach quite similar to LOV that can sometimes be found in the literature and confused with it is the labat-valve (LAV), introduced as an alternative cost-effective lTAS device. In it, instead of replacing a stator plate of a multiposition selection valve by a perfectly machined piece, as in LOV, sample processing and detection unit(s) in LAV are attached or plugged onto port(s) of a commercial conventional multiposition selection valve without taking apart any component of such a valve [23].
2. Sample-preparation steps The LOV manifold uses a universal hardware configuration to control the units that form part of the overall manifold (syringe pump, additional valves, detector) and http://www.elsevier.com/locate/trac
119
Trends
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
all possible steps, the development of which requires only changes in the software protocol and choice of reagents. An overall LOV procedure can include one or several of the following steps: (1) sample introduction into the LOV manifold with the aid of a syringe or a peristaltic pump via a flowthrough port that can be connected to an autosampler to process vessels for continuous monitoring or even inserted manually by Eppendorf pipettes (when samples are valuable or scant); (2) development of derivatization, preconcentration, dilution, or individual separation steps; and, (3) detection of the analytes by a LOV-integrated detector or by coupling the LOV manifold with other separation systems, such as capillary electrophoresis (CE), high-performance liquid chromatography (HPLC), or with mass spectrometry (MS) or atomic spectrometric detectors. The whole procedure is controlled by a single software program that makes possible total automation, system control and data treatment. 2.1. Sample dilution A simple preparation step, such as sample dilution, can easily be implemented by a LOV manifold in a very reproducible way by inserting a large volume of carrier, serving as a spacer, following sample aspiration. After appropriate dispersion, most of the analyte–spacer zone is discarded to waste leaving a diluted portion at the tail of the sample zone to be mixed with the reagents. Very efficient dilution can also be achieved by reversing the flow, as many times as required, in a holding coil containing a low ratio of sample volume to carrier volume. Automatic sample dilution is commonly required to monitor fermentations where the concentration changes within a wide range [22]. 2.2. Derivatization reactions The use of LOV with derivatization overcomes shortcomings, such as lack of reproducibility due to poor control of the reaction kinetics, through precise control of the working conditions concerning mixing degree, reaction time, and product delivery in an automated and timely manner [24]. SI of sample and reagent zones into a holding coil results in the formation of a complex concentration gradient that offers a continuous variation of sample–reagent concentrations and volume ratios [8]. Two approaches are used with this aim [24]: two-zone mixing and sandwich mixing. In the former, more frequently used, a sample zone is injected into the holding coil followed by a reagent zone; in the latter, a sample zone is injected between two reagent zones with the same or different composition. Injection of the sample zone first, followed by the reagent zone(s) in the se-
120
http://www.elsevier.com/locate/trac
quence required by the chemistry involved has been shown to be the preferable sequence [7,25]. To promote good sample–reagent zone overlap and mixing, an additional spacer zone of carrier solution can be aspirated into the holding coil. This zone increases the distance traveled and thus dispersion within the sample– reagent zone. Sample–reagent(s) mixing is promoted by flow-reversal transport at the detector cell, where monitoring can be performed with or without stopping the flow and using absorption [26,27], fluorescence [24,28] or chemiluminescence molecular detectors [29] (see detail in Fig. 1). A weakness of the LOV design, which is detrimental to derivatization, is the absence of temperature control, a desirable feature for kinetic-based measurements. However, this drawback has been conveniently eliminated by placing the LOV manifold into an air thermostatted incubator that allows the temperature to be elevated and maintained [11]. Also, when the LOV is fabricated from Plexiglas, use of organic solvents has to be avoided. Biocatalyzed reactions can be considered as a special kind of derivatization reaction the development of which in a LOV manifold provides unique advantages, particularly for relatively low reaction rates. In order to increase the analysis frequency, holding coils can be used to halt the sample–reagent mixtures in them, isolated from the rest of the system, while previous runs are monitored in the flow cell [10]. By choosing a large volume of spacer, memory effects between runs are prevented and the flow cell is well rinsed before monitoring the next run [22]. 2.3. Bead injection and associated modes BI, the most interesting mode of l-SI–LOV [1], involves suspended beads, which can be used as carriers for reactive groups or reagents. In the simplest case (i.e. suspended beads without reactive groups or reagents), the material used has to be endowed with good capability for retention of a wide variety of analytes and potential for use as a renewable sorbent. It is worth stressing that there are some requirements to be fulfilled by the solid-phase material for straightforward handling in the LOV structure. They must be perfectly spherical, uniform in size distribution and water-wettable to prevent a prompt settlement into the LOV cavities [13]. To this end, a commercially available divinylbenzene-co-Nvinylpyrrolidone reversed-phase copolymeric sorbent that accomplishes the three demands can be selected [30]. Reagents linked to the beads are commercially available as such or the linkage can be developed in the laboratory [31]. In the former case, there are different commercial microcarrier beads: Sephadex used for the linkage of reagents [11]; Sepharose 4B [8,32–34]; agarose [35]; and Aminolink [31], which provide optimal
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
bead support for bioligand immobilization; Cytopore is the most usual for cell linkage [11,36]. The bead suspension can be injected into the flow channel and remain in it or be transported to the detector for in situ monitoring of the reaction(s) that take place when the beads are trapped, perfused by the sample, buffers and/or auxiliary reagent, and then discharged. All these steps are a function of the flow rate (i.e. high flow rates up to 500 ll/s are used for fill up and discharge beads, moderate flow rates of 20 ll/s for bead transport, and low flow rates between 1 and 5 ll/s are needed for bead packing and long-term stability monitoring, if retained in the flow cell). Chemical reactions taking place at bead surfaces can be monitored in real time, either directly on the solid phase or by monitoring the eluted liquid phase (where the beads are retained before they reach the flow cell). Fig. 2A shows the experimental set-up, in which Port 1 is used to discard the beads after analysis. A detail of the flow cell can be seen in Fig. 2B [1,6]. In the l-BI spectrometry mode, the reaction is monitored at the bead surface. In this case, the spectral
Trends
baseline must be established prior to application of the sample mixture to the bead column [32] as the reagentcoated beads can absorb light at the monitoring wavelengths. The main limitation of l-BI spectrometry comes from the flow-cell configuration, where a significant portion of target molecules remains outside the optical path. In addition, the light-path geometry must be designed to allow sufficient light to penetrate the bead layer [37]. The BI approach has some characteristics that make it particularly useful in biochemical analysis, namely: (1) very small sample and reagent volumes; (2) minimum sample contamination and/or evaporation and prudent biohazard handling; and, (3) easy adaptation to a variety of analytical steps using the same configuration and simple changes of the software program [11]. The target analyte or complex can be recovered by simple chemical dissociation from the beads for secondary analysis, if required [33]. Microscale affinity chromatography uses BI for assembly, perfusion, discharge, and renewal of a micro-column
Figure 2. l-SI–LOV configured for bead-injection spectrometry. (A) System set-up. (B) Flow-cell configuration (plug ‘‘a’’ focuses the carrier stream into the center of the packed beads while plug ‘‘b’’ helps retain the beads within the optical path (with permission of Prof. J. Ruzicka).
http://www.elsevier.com/locate/trac
121
Trends
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
that is coupled on-line to a LOV module [21] (see Fig. 3). Beads furnished with an immobilized bioligand selectively capture target biomolecules from the sample; then, the composition of the mobile phase is changed and the biomolecules released for quantification. The most common eluent is an inorganic acid solution [21]. The main differences between traditional affinity chromatography and l-affinity chromatography are the volume scale (ml vs ll) and the mode of operation as the l-column is automatically packed, perfused, eluted and discarded as a function of the flow rate. The main limitation in l-affinity chromatography is that only species eluted from the column can be monitored, and not those remaining in the column, thus creating serious problems whenever target or matrix species bind irreversibly to the column, as the column capacity diminishes gradually. Even worse, if the composition of the eluent is changed and species irreversibly captured in previous runs are unexpectedly eluted, ‘‘ghost’’ or carry-over peaks appear [37]. These shortcomings are circumvented if the stationary phase instead of the eluate is interrogated by the detector as is the case with l-BI spectrometry [37]. The automated discharge of the beads makes elution of the target analytes unnecessary. [21,38]. Thanks to a longer light path and the absence of light scattering, l-affinity chromatography is more sensitive than l-BI spectrometry. This physical phenomenon increases both imprecision and errors by excess.
3.1. LOV–CE coupling Connection between a LOV module and a CE system has been made through two quite similar interfaces both based on that from Kuban, Karlberg and Fang [17,24]. As shown in Fig. 4A, one of them uses the detection cell of the LOV manifold as interface, and a two-way isolation valve placed in the outlet port of the detection cell for hydrodynamic sample injection, refreshing the CE buffer, or automated capillary flushing–reconditioning. A bare platinum wire was used as the cathode placed in a container filled with the CE buffer. The anode is housed in a PEEK tube and connected away from the LOV manifold. The capillary was enclosed in a PEEK tube and the capillary window was adjusted to the light path of the two optical fibers in the LOV manifold. CE separation was driven by a high-voltage power supply, with the cathode side away from the LOV manifold and the whole system controlled by the personal computer that controls the LOV system [24]. The other type of interface (Fig. 4B) is made of two Tpieces connected through a short piece of tube, one of them used as the capillary and the other both for the electrode (grounded) and for outlet through waste [17]. These interfaces enable either hydrodynamic or electro-
3. Coupling Despite the good results obtained by integrated affinity or monolithic separations, the resolution power characteristic of HPLC [30] and CE [17,29] cannot be achieved by LOV. Nevertheless, LOV manifolds can play a key role in sample preparation prior to high-resolution equipment.
Figure 3. Typical configuration of a LOV manifold for l-affinity chromatography. CB, Carrier buffer; D, Detector; E, Eluent; HC, Holding coil; S, Sample; SC, Separation column; SP, Syringe pump; W, Waste.
122
http://www.elsevier.com/locate/trac
Figure 4. Connection between a LOV module and CE equipment. (A) Using the LOV detection cell as interface. (B) Using two Tpieces connected through a short piece of tubing as interface. C, Capillary; CB, Carrier buffer; CEBR, Capillary electrophoresis buffer reservoir; CED, Capillary electrophoresis detector; DC, Detection cell; EB, Electrophoretic buffer; HC, Holding coil; IV, Isolation valve; R, Reagent; S, Sample; SP, Syringe pump; SV, Selection valve; W, Waste.
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
Trends
kinetic modes to be implemented. The injection process is thus more reproducible and fully automated. The LOV–CE coupling constitutes an adequate tool for real-time studies thanks to the high separation efficiency and short analysis time characteristic of CE. The low sensitivity characteristic of absorbance measurements in CE [39] can be circumvented by derivatization prior to CE injection [40]. The formation of a product with high molar absorptivity can be developed automatically in the LOV manifold. 3.2. LOV–LC coupling The coupling of a LOV manifold to a liquid chromatograph allows chemical compatibility between the sample (or the solution resulting from a previous step) and the initial LC gradient conditions. Band-broadening effects in conventional on-line SPE-based sample processors, caused by low eluting strength of the initial mobile phase, are thus circumvented [30]. The LOV manifold can be directly interfaced with the injection valve of the liquid chromatograph by a PTFE tube with a diameter smaller than that of the former; meanwhile the loop size of the LC injection valve should be large enough to contain the volume treated by the LOV manifold, which, in addition, can also be used as pumping system (the syringe pump) and injection system (the valve itself) in SI chromatography. In this case, a short (25 mm) monolithic separation column is placed in-line between the multi-position valve of the LOV and an external flow cell. The column used in SI chromatography can be coupled to a low-pressure system with a high resolving power but an extremely low flow resistance [35]. The performance of the monolithic phase is equivalent to a typical C-18 HPLC column. The main features of SI chromatography are economy and speed of gradient elution achieved without a gradient device [35].
4. In-line bioreactors There are two operational modes of bioreactors depending on whether cells or enzymes are used. Cells are commonly linked to beads and packed in microcolumns [16,36], while enzymes are used in a stirred micro-reactor attached to Channel 1 of the LOV (see Fig. 1A) [41]. These arrangements permit the bioreactor volume be separated from the detection volume of the flow cell. Separation is very useful for cell studies in which contact between the reagent and cells is undesirable. Perturbations of cells caused by reagent changes in pH, osmotic pressure and toxicity are thus avoided. To include cell bioreactors in a LOV manifold, a loop connects Port 3 of the LOV and the flow cell, as shown in Fig. 5A, thus allowing an aliquot of beads to be delivered
Figure 5. In-line coupling of a LOV manifold and a bioreactor (packed bead reactor), using a loop (A) or a multi-position valve (B). (C) Solution bioreactor. ASP, Auxiliary syringe pump; BR, Bead reservoir; BiR, Bioreactor; CB, Carrier buffer; DC, Detection cell; E, Eluent; EN, Enzyme; FO, Fiber optics; HC, Holding coil; I, Inhibitor; L, Loop; MPV, Multi-position valve R, Reagent; S, Sample; SP, Syringe pump; W, Waste; WB, Waste beads.
through that port and captured in a microcolumn behind the bead-retention plug–nozzle directly upstream from the flow cell. Once the microcolumn is packed, it can be perfused by the auxiliary syringe pump through the loop with the resulting perfusate entering the flow cell through the plug–nozzle. Inside the flow cell, the perfusate is mixed with reagent for analyte detection [36]. This manifold can be modified by connecting a multiposition valve (MPV) to Port 2 and deleting the loop connected to Port 3 (see Fig. 5B). The MPV enables the components of a cell study (i.e. cells-on-beads, buffer and inhibitors) be grouped around the MPV, leaving the LOV ports available for delivery of reagents [16]. http://www.elsevier.com/locate/trac
123
Trends
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
The in-line connection of an enzymatic l–bioreactor to a LOV manifold is performed as in Fig. 5C. The insertion of the optic fiber in the micro-reactor converts it into a measurement cell. By adjusting the distance between the two fiber-optic probes, the light path of the optical cell can be set. The total volume of the l-bioreactor is 1 ml and a minimum volume of 350 ll is required in order to fully insert the optic fibers. Two syringe pumps are required: one of them (primary syringe pump) is used to sequentially aspirate enzyme, reagent, inhibitor or buffer into the holding coil and subsequently deliver the stacked zones into the l-reactor by flow reversal. At the end of each measuring cycle, the other syringe pump (auxiliary syringe pump), connected to the other channel of the flow-through port, is used to empty the l-reactor [41]. This manifold allows enzyme kinetics and inhibition studies to be carried out under well-controlled conditions. While maintaining advantages of flow-mode l-SI–LOV, such as automation and real-time measurement, the stirred batch-mode l-reactor ensures thorough mixing, which makes it possible to measure reactant concentrations, essential for the determination of kinetic constants.
5. Detection modes Any type of detection can be implemented in dealing with meso-fluidic, LOV devices. Thus, detection has been integrated in the more recent LOV models through a multi-purpose flow cell incorporated at Port 2 in Fig. 1A. This cell is furnished with optical fibers connected to an external light source and a detection device to facilitate real-time monitoring of the reactions taking place inside the cell. The fiber-optic probes can be in different positions and configurations to obtain values of absorbance, fluorescence and even absorbance + fluorescence if a third fiber-optic probe is used (see Fig. 1) [42]. A Z-type multipurpose flow-through cell has been connected to a port of the LOV unit to monitor the light emission from a chemiluminescent reaction with a side-on photomultiplier tube located close to the flow cell [29]. The most usual detector in LOV is an Ocean Optics spectrometer, which monitors a wide range of the spectrum – from UV to IR [42]. Other detectors, such as charge-coupled detectors (CCDs) or intensified CCDs, can also be used. Electrochemical detection has recently been applied to biochemical–LOV analysis [43]. Potentiometric detection has been successfully used for the determination of chloride in water. In this case, the simple LOV flow-through electrode system consists of two simple laboratory-made Ag/AgCl electrodes plugged into a port of the flow cell [23]. Other detectors successfully used in conjunction LOV manifolds have been electrothermal atomic absorption 124
http://www.elsevier.com/locate/trac
spectrometers and inductively coupled plasmamass spectrometers [6,23]. These instruments have not yet been used in biochemical analysis, despite the presence of metal in the structures of key biological compounds making their application in this field very promising. A mass spectrometer with the electrospray ionization source also acting as interface to the LOV system provides high capability in biochemical analysis. PEEK tube (0.127-mm i.d.) and a three-way solenoid valve were necessary to connect the LOV system to the MS instrument [18].
6. Applications The SI–LOV format has been applied in the biochemical field with different aims, namely: 6.1. Cell culture The use of a BI–LOV approach for monitoring cell cultures has some advantages in comparison with classical methods, such as: (1) a large number of cells can be interrogated automatically within a small packed column in a sterile enclosed environment; (2) the amount of cells can be easily changed; (3) the whole system can be introduced into an incubator to be kept at constant temperature; and, (4) the cell culture can be replaced in a reproducible manner at any time during a series of experimental runs. These advantages, together with others obtained from the detection system, make the BI–LOV approach a promising alternative for real-time determination of lactate [16], glucose [16,22,36], ammonia, glycerol, and free iron [22], thus providing a means to assess the metabolic regimen of living cells. Monitoring extracellular acidification rates, which is important for the study of cellular activities, can also be carried out by this option [11]. This approach has recently been applied for in situ degradation of H2O2 by living cells using voltammetric monitoring [43]. 6.2. Antibody studies Research focused on real-time monitoring of IgG interactions with protein-coated beads [8,32] implemented in SI–LOV can obtain a great deal of information about antibody–protein interactions and can determine the target analyte (antibody or protein) [21,37]. Nonradioactive ELISA-based methods using beads coated with streptavidin in SI–LOV format have been applied to the determination of GAD65 antibodies involved in type 1 diabetes mellitus [33]. The immobilization protocols of IgG and other proteins have been carried out by l-SI– LOV working in the BI mode [31].
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
Affinity chromatography is being widely used in the purification of antibody production, in pharmaceutical and biotech companies, which require that the injected antibody is completely eluted and the column completely regenerated [11,44]. l-BI spectrometry may also be used to monitor carry-over on the column and to optimize elution conditions in order to ensure complete recovery from the column. 6.3. DNA assays SI–LOV is a feasible alternative to conventional methods for nucleic-acid determination, as shown by its successful application to DNA determination using photometric [27] or laser-induced fluorescence detection [28], based on decoloration of a Crystal Violet solution and substantial fluorescence enhancement of ethidium bromide, respectively. The BI mode of SI–LOV has also been applied to the detection of single-strand nucleic-acid sequences [34]. 6.4. Enzymatic assays lSI–LOV also enables monitoring of initial reaction rates and determination of reactant concentrations in a reaction mixture; examples of the latter are the determination of glucose, ethanol or glycerol [10,45]. The approach is also able to determine both kinetic constants for enzymatic or inhibition reactions, and the type of inhibition (competitive, uncompetitive, or mixed) [41]. Another interesting application is the determination of savinase protease activity, based on the enzymatic degradation of an artificial fluorogenic substrate covalently attached to a tripeptide. As the fluorophore is released from the tripeptide by proteolytic action, fluorescence increase can be measured [8]. 6.5. Other applications The SI–LOV–CE hyphenated approach allows automated monitoring of insulin, proinsulin and c-peptide levels, which are of great interest in the clinical field. The target proteins are sampled and derivatized with fluorescamine in the LOV platform. Individual separation–detection is carried out by a CE-epiluminescence fluorescence microscope [24]. Studies of different conformations of myoglobin caused by denaturation have been conducted using an SI–LOV– CE approach by mixing the analyte with variable concentrations of sodium dodecyl sulfate as denaturing agent in the LOV module. The SI–LOV–LC coupling has been applied to the determination of non-steroidal anti-inflammatory drugs and lipid regulators at the trace level [30]; meanwhile SI–LOV has been used for automated affinity capture and release of biotin-containing conjugates on immobilized streptavidin [18] and for simultaneous measurements of multiple-ligand affinities to proteins immobilized on
Trends
beads [38] – dissociation rate constants were calculated in both cases. l-SI–LOV has also been applied to the determination of pharmaceutical residues (ketoprofen, naproxen, bezafibrate, diclofenac, ibuprofen) and one metabolite (salicylic acid) in surface water, urban wastewater, and urine [46], and to the chemiluminescent determination of tetracycline [29]. In the last case, a bismuthate–immobilized microcolumn was incorporated in one port of the LOV for in situ oxidation of KBr and generation of bromine as oxidant for the bromine–hydrogen peroxide– tetracycline chemiluminescent reaction.
7. Future outlook The LOV approach has opened a promising avenue for in-valve immunoassay and bio-chemical analysis via BI. Nevertheless, the high potential of LOV for sample preparation at the l-scale is even more important. Thus, LOV-based microscale dialysis and non-chromatographic miniaturized separation–preconcentration may be excellent alternatives to conventional analysis. The determination of organic molecules of clinical interest can be one of the areas to benefit from these approaches. The use of LOV manifold and in-situ fluorescence detection could be a new alternative to separate and to purify nucleic acids and other species of biological interest. Multi-reagent assays could be conducted by connecting the central port of the LOV to a second syringe pump. In this way, mixing of multiple zones previously stacked into holding coils could be achieved, merged by simultaneous pumping through the central port. The label dilution technique can be used to evaluate receptor binding, avidin–biotin interactions or any binding between two molecules, one of which can be immobilized on the bead surface. This should provide a very useful tool for the study of diseases such as diabetes. The immobilization of reagents in beads may be focused on the improvement of the way in which the reagent – usually a protein – distribution between beads and supernatant can be monitored simultaneously, and how the reaction mixture can be prepared automatically. For the first goal, a LOV with two in-series flowcells is required to monitor the bead layer and the protein content of the supernatant solution at the first and second flow-cell, respectively. The LOV–l-reactor system could also be used to study different enzymes with photometric detection, to perform automated non-enzymatic kinetic assays, and to carry out automated l-scale titrations. Also the LOV–CE coupling might be applied to real-time monitoring of both cell metabolism (to test the functional response with respect to repeated stimulation) and cell health. New research in LOV–LC coupling should be focused on: http://www.elsevier.com/locate/trac
125
Trends
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
(a) further miniaturization of the entire analytical setup by hyphenation of the BI–LOV approach with l-LC; and, (b) the synthesis of selective materials fulfilling the stringent demands for their straightforward handling within the LOV conduits. Pending goals on integration of separation devices in the LOV system are l-dialyzers, gas-diffusers, pervaporators or liquid–liquid extractors. These devices could provide macromolecule-free solutions, volatile analyte removal from the sample matrix or solvent exchange and preconcentration, respectively. Liquid–liquid extraction (e.g., without phase separation or with supported liquid membranes) also offers a wide range of possibilities for a judicious choice according to the (bio)chemical system under study.
Acknowledgements SpainÕs Ministry of Science and Technology is gratefully acknowledged for financial support (Project No. CTQ2006-01614). J. Ruiz-Jime´nez and J.A. Pe´rez-Serradilla are also grateful to SpainÕs Ministry of Education for award of their research training fellowships. References [1] P. Solich, M. Polasek, J. Klimundova, J. Ruzicka, Trends Anal. Chem. 22 (2003) 116. [2] J. Ruzicka, E.H. Hansen, Anal. Chim. Acta 78 (1975) 145. [3] E.H. Hansen, Flow injection bibliography (http://www.flowinjection.com/). [4] E.H. Hansen, J.H. Wang, Anal. Chim. Acta 467 (2002) 3. [5] J. Wang, E.H. Hansen, Trends Anal. Chem. 22 (2003) 225. [6] E.H. Hansen, Talanta 64 (2004) 1076. [7] J. Ruzicka, G.D. Marshall, Anal. Chim. Acta 237 (1990) 329. [8] J. Ruzicka, Analyst (Cambridge, U. K.) 125 (2000) 1053. [9] J. Ruzicka, Sequential Injection for Biomolecular Assays, CD-ROM Tutorial, www.flowinjection.com (2004). [10] Y. Chen, J. Ruzicka, Analyst (Cambridge, U. K.) 129 (2004) 597. [11] H.A. Erxleben, M.K. Manion, D.M. Hockenbery, L. Scampavia, J. Ruzicka, Analyst (Cambridge, U. K.) 129 (2004) 205. [12] W. Wang, J.H. Wang, Z.L. Fang, Anal. Chem. 77 (2005) 5396. [13] J. Wang, E.H. Hansen, M. Miro´, Anal. Chim. Acta 499 (2003) 139. [14] X. Long, E.H. Hansen, M. Miro´, Talanta 66 (2005) 1326. [15] X. Long, R. Chomchoei, P. Gala, E.H. Hansen, Anal. Chim. Acta 523 (2004) 279.
126
http://www.elsevier.com/locate/trac
[16] C.M. Schulz, L. Scampavia, J. Ruzicka, Analyst (Cambridge, U. K.) 127 (2002) 1583. [17] S. Kulka, G. Quinta´s, B. Lendl, Analyst (Cambridge, U. K.) 131 (2006) 739. [18] Y. Ogata, L. Scampavia, J. Ruzicka, C.R. Scott, M.H. Gelb, F. Turecek, Anal. Chem. 74 (2002) 4702. [19] C.H. Wu, J. Ruzicka, Analyst (Cambridge, U. K.) 126 (2001) 1947. [20] J. Jakmunee, L. Pathimapornlert, S.K. Hartwell, K. Grudpan, Analyst (Cambridge, U. K.) 130 (2005) 299. [21] H. Erxleben, J. Ruzicka, Analyst (Cambridge, U. K.) 130 (2005) 469. [22] C.H. Wu, L. Scampavia, J. Ruzicka, B. Zamost, Analyst (Cambridge, U. K.) 126 (2001) 291. [23] K. Grudpan, Talanta 64 (2004) 1084. [24] C.H. Wu, L. Scampavia, J. Ruzicka, Analyst (Cambridge, U. K.) 128 (2003) 1123. [25] J. Ruzicka Micro total analysis systems 2000, In: A. Berg, W. Olthuis, P. Bergveld (Editors), Kluwer Academic Publishers, Dodrecht, The Netherlands, 2000, p. 1. [26] M.D. Krom, Analyst (Cambridge, U. K.) 105 (1980) 305. [27] X. Chen, J. Wang, Z. Fang, Talanta 67 (2005) 227. [28] X. Chen, W. Wang, J. Wang, Analyst (Cambridge, U. K.) 130 (2005) 1240. [29] M. Yang, Y. Xu, J.H. Wang, Anal. Chem. 78 (2006) 5900. [30] J.B. Quintana, M. Miro´, J.M. Estela, V. Cerda´, Anal. Chem. 78 (2006) 2832. [31] J. Ruzicka, A.D. Carroll, I. Lahdesmaki, Analyst (Cambridge, U. K.) 131 (2006) 799. [32] A.D. Carroll, L. Scampavia, J. Ruzicka, Analyst (Cambridge, U. K.) 127 (2002) 1228. [33] A.D. Carroll, L. Scampavia, D. Luo, A. Lernmark, J. Ruzicka, Analyst (Cambridge, U. K.) 128 (2003) 1157. [34] K.A. Edwards, A.J. Baeumner, Anal. Chem. 78 (2006) 1958. [35] M. Miro´, S. Jonczyk, J. Wang, E.H. Hansen, J. Anal. At. Spectrom. 18 (2003) 89. [36] C.M. Schulz, J. Ruzicka, Analyst (Cambridge, U. K.) 127 (2002) 1293. [37] Y. Gutzman, A.D. Carroll, J. Ruzicka, Analyst (Cambridge, U. K.) 131 (2006) 809. [38] Y. Ogata, L. Scampavia, T.L. Carter, E. Fan, F. Turecek, Anal. Biochem. 331 (2004) 161. [39] C.H. Wu, L. Scampavia, J. Ruzicka, Analyst (Cambridge, U. K.) 127 (2002) 898. [40] X. Long, M. Miro´, E.H. Hansen, Anal. Chem. 77 (2005) 6032. [41] Y. Chen, A.D. Carroll, L. Scampavia, J. Ruzicka, Anal. Sci. 22 (2006) 9. [42] MicroSIA, FIAlab-3200, and FIAlab-3500 Operation Manual, FIAlab Instrument, Inc., 2005. [43] I. La¨hdesma¨ki, Y.K. Park, A.D. Carroll, M. Decuir, J. Ruzicka, Analyst (Cambridge, U. K.) 132 (2007) 811. [44] D.S. Hage, Clin. Chem. 45 (1999) 593. [45] J. Wang, Anal. Bioanal. Chem. 381 (2005) 809. [46] J. Wang, E.H. Hansen, Trends Anal. Chem. 22 (2003) 836.
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
Lab-on-valve: a useful tool in biochemical analysis M.D. Luque de Castro, J. Ruiz-Jime´nez, J.A. Pe´rez-Serradilla Despite the short life-time of lab-on-valve (LOV) devices – the third generation of dynamic approaches invented by J. Ruzicka – they have proved to be of great interest in biochemical analysis. LOV works automatically with sequential injection to move micro or sub-micro volumes of liquids, gases and/ or beads in a ‘‘digital’’ manner by stopping, reversing, and accelerating flow rates in a way that is of interest in sample preparation – especially modes such as bead-injection spectrometry, micro affinity chromatography or inline LOV–bioreactor – and when LOV is coupled to high-resolution equipment, such as capillary electrophoresis and chromatography. LOV-based equipment has proved to be useful in handing ll volumes (e.g., in cell-culture and antibody studies, to assess the metabolic regime of living cells; and, in DNA assays, to detect single-stranded nucleic acid sequences). LOV therefore offers a promising way ahead for in-valve biochemical steps, which can be expanded by coupling to separation units (e.g., dialyzers, gas diffusers, pervaporators, and novel liquid–liquid extractors), which facilitate automatic interference removal and preconcentration in complex samples with low concentrations of analytes. ª 2008 Elsevier Ltd. All rights reserved. Keywords: Bead injection; Biochemical analysis; Coupling; Lab-on-valve; Sample preparation
1. Introduction M.D. Luque de Castro*, J. Ruiz-Jime´nez, J.A. Pe´rez-Serradilla Department of Analytical Chemistry, Annex Marie Curie Building, Campus of Rabanales, University of Co´rdoba, E-14071 Co´rdoba, Spain
*
Corresponding author. Tel./Fax: +34 957 218 615; E-mail: [email protected]
118
Automation and miniaturization of solution-based analysis are essential to make them fast and efficient for routine and research tasks in biochemical laboratories. Ideally, analytical equipment should be versatile, capable of accommodating a wide variety of assays without the need for system reconfiguration, and compatible with a wide range of detectors [1]. Since the introduction of flow injection (FI) in 1975 [2], thousands of scientific publications [3] have demonstrated its feasibility for automation, which has evolved through two new generations [4–6]: sequential injection (SI) [7], introduced in the 1990s and referred to as the second-generation FI; and, laboratory-on-valve (lab-on-valve, LOV), a recent continuous methodology, referred to as third-generation FI
and also introduced by Ruzicka [8], which works in SI mode and has downscaled reagent-based analysis to ml and sub-ml levels. During the past decade, miniaturization has gained increased interest in several analytical fields. Taking into account that the ultimate goal of miniaturization of reagent-based analysis is to decrease use of materials consumed and waste generated [8], its importance in the biochemical field is crucial, as both biochemical reagents and samples are usually scant and/or expensive. Thus, reducing consumption of expensive (or scant) sample and/or reagents through different approaches to miniaturization is obviously one of the most effective strategies to reduce analytical costs. A key way to downscale is to replace continuous flow by programmable flow, which allows moving liquids, gases and/or beads in a ‘‘digital’’ fashion, by stopping, reversing, and accelerating flow rates [5]. This principle of SI [4] has been downscaled, integrated into the LOV platform and used for miniaturization of reagentbased analysis, immunoassays, bioligandinteraction analysis, and ion-exchange and affinity chromatography [9], among others. A LOV platform consists of: (1) a transparent, monolithic structure made of Perspex; (2) a multiposition valve as the main component of the structure; and, (3) a propulsion unit, usually a syringe pump, characteristic of SI, to circulate the required liquids through the system, which is overall referred to as l-SI–LOV. Fig. 1 shows a scheme of a LOV manifold, which also includes details of the valve itself and the ways the flow cell can
0165-9936/$ - see front matter ª 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.trac.2008.01.004
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
Figure 1. (A) Overall configuration of l-SI–LOV with integrated UV–Vis absorption detection; (B) Details of the LOV platform; (C) Other detection modes (with permission of Prof. J. Ruzicka).
be integrated into the platform, depending on the detection technique. Integration of the multipurpose flow-cell, in which light measurements are carried out through optical fibers, means that the system is very versatile in measurements, although LOV can also be used with non-integrated cells for optical or electroanalytical measurements. In addition, an adequate selection of auxiliary units is essential in order to achieve the required configuration. Auxiliary units more frequently used are: one [8,10,11] or two [12–15] syringe pumps; an auxiliary six-port
Trends
selection valve [11,16,17]; one or two holding coils [12,18]; an auxiliary peristaltic pump [5,8,19–21]; a mixing coil [5,22]; and, T connections [12,19]. These units confer on LOV the high versatility, which, together with other characteristics, makes it an attractive tool in the biochemical field. LOV allows miniaturization of the flow channel; thus, sample and reagent volumes are downscaled to the 10– 20-ll range, while waste production is typically 0.1–0.2 ml per assay [1]. These systems are therefore positioned between traditional flow techniques that operate at the ml scale and the more futuristic designs of the micro total analysis systems (lTAS) concept, which is supposed to work at the nl scale. (The SI–LOV manifold is also known as a meso-fluidic system – ‘‘meso’’ from its capability to manipulate fluid between ‘‘micro’’ and ‘‘macro’’. Despite of this, most authors refer to SI–LOV manifolds to as micro-fluidic systems). Similarly to microchips, LOV must be assisted by external units for proper circulation of samples and reagents. By operating within the ll range with channel diameters about 0.5 mm, l-SI–LOV systems are compatible with real-life samples that often contain particles that can clog the lm-sized channels of nl-scale devices. The choice of LOV channel dimensions has two significant advantages: (1) a large volume-to-surface ratio, which minimizes the unwanted adsorption on channel walls that may result in carry-over; and, (2) even more importantly, this channel geometry allows LOV devices to be used as platforms for the bead-injection (BI) technique [1,5]. The main goal of a LOV system is sample preparation, which can involve sample dilution, common analyte preconcentration and derivatization or steps such as those based on BI. We discuss below these goals, the coupling of LOV approaches to high-resolution equipment and very different detectors, and salient applications in biochemical analysis. An approach quite similar to LOV that can sometimes be found in the literature and confused with it is the labat-valve (LAV), introduced as an alternative cost-effective lTAS device. In it, instead of replacing a stator plate of a multiposition selection valve by a perfectly machined piece, as in LOV, sample processing and detection unit(s) in LAV are attached or plugged onto port(s) of a commercial conventional multiposition selection valve without taking apart any component of such a valve [23].
2. Sample-preparation steps The LOV manifold uses a universal hardware configuration to control the units that form part of the overall manifold (syringe pump, additional valves, detector) and http://www.elsevier.com/locate/trac
119
Trends
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
all possible steps, the development of which requires only changes in the software protocol and choice of reagents. An overall LOV procedure can include one or several of the following steps: (1) sample introduction into the LOV manifold with the aid of a syringe or a peristaltic pump via a flowthrough port that can be connected to an autosampler to process vessels for continuous monitoring or even inserted manually by Eppendorf pipettes (when samples are valuable or scant); (2) development of derivatization, preconcentration, dilution, or individual separation steps; and, (3) detection of the analytes by a LOV-integrated detector or by coupling the LOV manifold with other separation systems, such as capillary electrophoresis (CE), high-performance liquid chromatography (HPLC), or with mass spectrometry (MS) or atomic spectrometric detectors. The whole procedure is controlled by a single software program that makes possible total automation, system control and data treatment. 2.1. Sample dilution A simple preparation step, such as sample dilution, can easily be implemented by a LOV manifold in a very reproducible way by inserting a large volume of carrier, serving as a spacer, following sample aspiration. After appropriate dispersion, most of the analyte–spacer zone is discarded to waste leaving a diluted portion at the tail of the sample zone to be mixed with the reagents. Very efficient dilution can also be achieved by reversing the flow, as many times as required, in a holding coil containing a low ratio of sample volume to carrier volume. Automatic sample dilution is commonly required to monitor fermentations where the concentration changes within a wide range [22]. 2.2. Derivatization reactions The use of LOV with derivatization overcomes shortcomings, such as lack of reproducibility due to poor control of the reaction kinetics, through precise control of the working conditions concerning mixing degree, reaction time, and product delivery in an automated and timely manner [24]. SI of sample and reagent zones into a holding coil results in the formation of a complex concentration gradient that offers a continuous variation of sample–reagent concentrations and volume ratios [8]. Two approaches are used with this aim [24]: two-zone mixing and sandwich mixing. In the former, more frequently used, a sample zone is injected into the holding coil followed by a reagent zone; in the latter, a sample zone is injected between two reagent zones with the same or different composition. Injection of the sample zone first, followed by the reagent zone(s) in the se-
120
http://www.elsevier.com/locate/trac
quence required by the chemistry involved has been shown to be the preferable sequence [7,25]. To promote good sample–reagent zone overlap and mixing, an additional spacer zone of carrier solution can be aspirated into the holding coil. This zone increases the distance traveled and thus dispersion within the sample– reagent zone. Sample–reagent(s) mixing is promoted by flow-reversal transport at the detector cell, where monitoring can be performed with or without stopping the flow and using absorption [26,27], fluorescence [24,28] or chemiluminescence molecular detectors [29] (see detail in Fig. 1). A weakness of the LOV design, which is detrimental to derivatization, is the absence of temperature control, a desirable feature for kinetic-based measurements. However, this drawback has been conveniently eliminated by placing the LOV manifold into an air thermostatted incubator that allows the temperature to be elevated and maintained [11]. Also, when the LOV is fabricated from Plexiglas, use of organic solvents has to be avoided. Biocatalyzed reactions can be considered as a special kind of derivatization reaction the development of which in a LOV manifold provides unique advantages, particularly for relatively low reaction rates. In order to increase the analysis frequency, holding coils can be used to halt the sample–reagent mixtures in them, isolated from the rest of the system, while previous runs are monitored in the flow cell [10]. By choosing a large volume of spacer, memory effects between runs are prevented and the flow cell is well rinsed before monitoring the next run [22]. 2.3. Bead injection and associated modes BI, the most interesting mode of l-SI–LOV [1], involves suspended beads, which can be used as carriers for reactive groups or reagents. In the simplest case (i.e. suspended beads without reactive groups or reagents), the material used has to be endowed with good capability for retention of a wide variety of analytes and potential for use as a renewable sorbent. It is worth stressing that there are some requirements to be fulfilled by the solid-phase material for straightforward handling in the LOV structure. They must be perfectly spherical, uniform in size distribution and water-wettable to prevent a prompt settlement into the LOV cavities [13]. To this end, a commercially available divinylbenzene-co-Nvinylpyrrolidone reversed-phase copolymeric sorbent that accomplishes the three demands can be selected [30]. Reagents linked to the beads are commercially available as such or the linkage can be developed in the laboratory [31]. In the former case, there are different commercial microcarrier beads: Sephadex used for the linkage of reagents [11]; Sepharose 4B [8,32–34]; agarose [35]; and Aminolink [31], which provide optimal
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
bead support for bioligand immobilization; Cytopore is the most usual for cell linkage [11,36]. The bead suspension can be injected into the flow channel and remain in it or be transported to the detector for in situ monitoring of the reaction(s) that take place when the beads are trapped, perfused by the sample, buffers and/or auxiliary reagent, and then discharged. All these steps are a function of the flow rate (i.e. high flow rates up to 500 ll/s are used for fill up and discharge beads, moderate flow rates of 20 ll/s for bead transport, and low flow rates between 1 and 5 ll/s are needed for bead packing and long-term stability monitoring, if retained in the flow cell). Chemical reactions taking place at bead surfaces can be monitored in real time, either directly on the solid phase or by monitoring the eluted liquid phase (where the beads are retained before they reach the flow cell). Fig. 2A shows the experimental set-up, in which Port 1 is used to discard the beads after analysis. A detail of the flow cell can be seen in Fig. 2B [1,6]. In the l-BI spectrometry mode, the reaction is monitored at the bead surface. In this case, the spectral
Trends
baseline must be established prior to application of the sample mixture to the bead column [32] as the reagentcoated beads can absorb light at the monitoring wavelengths. The main limitation of l-BI spectrometry comes from the flow-cell configuration, where a significant portion of target molecules remains outside the optical path. In addition, the light-path geometry must be designed to allow sufficient light to penetrate the bead layer [37]. The BI approach has some characteristics that make it particularly useful in biochemical analysis, namely: (1) very small sample and reagent volumes; (2) minimum sample contamination and/or evaporation and prudent biohazard handling; and, (3) easy adaptation to a variety of analytical steps using the same configuration and simple changes of the software program [11]. The target analyte or complex can be recovered by simple chemical dissociation from the beads for secondary analysis, if required [33]. Microscale affinity chromatography uses BI for assembly, perfusion, discharge, and renewal of a micro-column
Figure 2. l-SI–LOV configured for bead-injection spectrometry. (A) System set-up. (B) Flow-cell configuration (plug ‘‘a’’ focuses the carrier stream into the center of the packed beads while plug ‘‘b’’ helps retain the beads within the optical path (with permission of Prof. J. Ruzicka).
http://www.elsevier.com/locate/trac
121
Trends
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
that is coupled on-line to a LOV module [21] (see Fig. 3). Beads furnished with an immobilized bioligand selectively capture target biomolecules from the sample; then, the composition of the mobile phase is changed and the biomolecules released for quantification. The most common eluent is an inorganic acid solution [21]. The main differences between traditional affinity chromatography and l-affinity chromatography are the volume scale (ml vs ll) and the mode of operation as the l-column is automatically packed, perfused, eluted and discarded as a function of the flow rate. The main limitation in l-affinity chromatography is that only species eluted from the column can be monitored, and not those remaining in the column, thus creating serious problems whenever target or matrix species bind irreversibly to the column, as the column capacity diminishes gradually. Even worse, if the composition of the eluent is changed and species irreversibly captured in previous runs are unexpectedly eluted, ‘‘ghost’’ or carry-over peaks appear [37]. These shortcomings are circumvented if the stationary phase instead of the eluate is interrogated by the detector as is the case with l-BI spectrometry [37]. The automated discharge of the beads makes elution of the target analytes unnecessary. [21,38]. Thanks to a longer light path and the absence of light scattering, l-affinity chromatography is more sensitive than l-BI spectrometry. This physical phenomenon increases both imprecision and errors by excess.
3.1. LOV–CE coupling Connection between a LOV module and a CE system has been made through two quite similar interfaces both based on that from Kuban, Karlberg and Fang [17,24]. As shown in Fig. 4A, one of them uses the detection cell of the LOV manifold as interface, and a two-way isolation valve placed in the outlet port of the detection cell for hydrodynamic sample injection, refreshing the CE buffer, or automated capillary flushing–reconditioning. A bare platinum wire was used as the cathode placed in a container filled with the CE buffer. The anode is housed in a PEEK tube and connected away from the LOV manifold. The capillary was enclosed in a PEEK tube and the capillary window was adjusted to the light path of the two optical fibers in the LOV manifold. CE separation was driven by a high-voltage power supply, with the cathode side away from the LOV manifold and the whole system controlled by the personal computer that controls the LOV system [24]. The other type of interface (Fig. 4B) is made of two Tpieces connected through a short piece of tube, one of them used as the capillary and the other both for the electrode (grounded) and for outlet through waste [17]. These interfaces enable either hydrodynamic or electro-
3. Coupling Despite the good results obtained by integrated affinity or monolithic separations, the resolution power characteristic of HPLC [30] and CE [17,29] cannot be achieved by LOV. Nevertheless, LOV manifolds can play a key role in sample preparation prior to high-resolution equipment.
Figure 3. Typical configuration of a LOV manifold for l-affinity chromatography. CB, Carrier buffer; D, Detector; E, Eluent; HC, Holding coil; S, Sample; SC, Separation column; SP, Syringe pump; W, Waste.
122
http://www.elsevier.com/locate/trac
Figure 4. Connection between a LOV module and CE equipment. (A) Using the LOV detection cell as interface. (B) Using two Tpieces connected through a short piece of tubing as interface. C, Capillary; CB, Carrier buffer; CEBR, Capillary electrophoresis buffer reservoir; CED, Capillary electrophoresis detector; DC, Detection cell; EB, Electrophoretic buffer; HC, Holding coil; IV, Isolation valve; R, Reagent; S, Sample; SP, Syringe pump; SV, Selection valve; W, Waste.
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
Trends
kinetic modes to be implemented. The injection process is thus more reproducible and fully automated. The LOV–CE coupling constitutes an adequate tool for real-time studies thanks to the high separation efficiency and short analysis time characteristic of CE. The low sensitivity characteristic of absorbance measurements in CE [39] can be circumvented by derivatization prior to CE injection [40]. The formation of a product with high molar absorptivity can be developed automatically in the LOV manifold. 3.2. LOV–LC coupling The coupling of a LOV manifold to a liquid chromatograph allows chemical compatibility between the sample (or the solution resulting from a previous step) and the initial LC gradient conditions. Band-broadening effects in conventional on-line SPE-based sample processors, caused by low eluting strength of the initial mobile phase, are thus circumvented [30]. The LOV manifold can be directly interfaced with the injection valve of the liquid chromatograph by a PTFE tube with a diameter smaller than that of the former; meanwhile the loop size of the LC injection valve should be large enough to contain the volume treated by the LOV manifold, which, in addition, can also be used as pumping system (the syringe pump) and injection system (the valve itself) in SI chromatography. In this case, a short (25 mm) monolithic separation column is placed in-line between the multi-position valve of the LOV and an external flow cell. The column used in SI chromatography can be coupled to a low-pressure system with a high resolving power but an extremely low flow resistance [35]. The performance of the monolithic phase is equivalent to a typical C-18 HPLC column. The main features of SI chromatography are economy and speed of gradient elution achieved without a gradient device [35].
4. In-line bioreactors There are two operational modes of bioreactors depending on whether cells or enzymes are used. Cells are commonly linked to beads and packed in microcolumns [16,36], while enzymes are used in a stirred micro-reactor attached to Channel 1 of the LOV (see Fig. 1A) [41]. These arrangements permit the bioreactor volume be separated from the detection volume of the flow cell. Separation is very useful for cell studies in which contact between the reagent and cells is undesirable. Perturbations of cells caused by reagent changes in pH, osmotic pressure and toxicity are thus avoided. To include cell bioreactors in a LOV manifold, a loop connects Port 3 of the LOV and the flow cell, as shown in Fig. 5A, thus allowing an aliquot of beads to be delivered
Figure 5. In-line coupling of a LOV manifold and a bioreactor (packed bead reactor), using a loop (A) or a multi-position valve (B). (C) Solution bioreactor. ASP, Auxiliary syringe pump; BR, Bead reservoir; BiR, Bioreactor; CB, Carrier buffer; DC, Detection cell; E, Eluent; EN, Enzyme; FO, Fiber optics; HC, Holding coil; I, Inhibitor; L, Loop; MPV, Multi-position valve R, Reagent; S, Sample; SP, Syringe pump; W, Waste; WB, Waste beads.
through that port and captured in a microcolumn behind the bead-retention plug–nozzle directly upstream from the flow cell. Once the microcolumn is packed, it can be perfused by the auxiliary syringe pump through the loop with the resulting perfusate entering the flow cell through the plug–nozzle. Inside the flow cell, the perfusate is mixed with reagent for analyte detection [36]. This manifold can be modified by connecting a multiposition valve (MPV) to Port 2 and deleting the loop connected to Port 3 (see Fig. 5B). The MPV enables the components of a cell study (i.e. cells-on-beads, buffer and inhibitors) be grouped around the MPV, leaving the LOV ports available for delivery of reagents [16]. http://www.elsevier.com/locate/trac
123
Trends
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
The in-line connection of an enzymatic l–bioreactor to a LOV manifold is performed as in Fig. 5C. The insertion of the optic fiber in the micro-reactor converts it into a measurement cell. By adjusting the distance between the two fiber-optic probes, the light path of the optical cell can be set. The total volume of the l-bioreactor is 1 ml and a minimum volume of 350 ll is required in order to fully insert the optic fibers. Two syringe pumps are required: one of them (primary syringe pump) is used to sequentially aspirate enzyme, reagent, inhibitor or buffer into the holding coil and subsequently deliver the stacked zones into the l-reactor by flow reversal. At the end of each measuring cycle, the other syringe pump (auxiliary syringe pump), connected to the other channel of the flow-through port, is used to empty the l-reactor [41]. This manifold allows enzyme kinetics and inhibition studies to be carried out under well-controlled conditions. While maintaining advantages of flow-mode l-SI–LOV, such as automation and real-time measurement, the stirred batch-mode l-reactor ensures thorough mixing, which makes it possible to measure reactant concentrations, essential for the determination of kinetic constants.
5. Detection modes Any type of detection can be implemented in dealing with meso-fluidic, LOV devices. Thus, detection has been integrated in the more recent LOV models through a multi-purpose flow cell incorporated at Port 2 in Fig. 1A. This cell is furnished with optical fibers connected to an external light source and a detection device to facilitate real-time monitoring of the reactions taking place inside the cell. The fiber-optic probes can be in different positions and configurations to obtain values of absorbance, fluorescence and even absorbance + fluorescence if a third fiber-optic probe is used (see Fig. 1) [42]. A Z-type multipurpose flow-through cell has been connected to a port of the LOV unit to monitor the light emission from a chemiluminescent reaction with a side-on photomultiplier tube located close to the flow cell [29]. The most usual detector in LOV is an Ocean Optics spectrometer, which monitors a wide range of the spectrum – from UV to IR [42]. Other detectors, such as charge-coupled detectors (CCDs) or intensified CCDs, can also be used. Electrochemical detection has recently been applied to biochemical–LOV analysis [43]. Potentiometric detection has been successfully used for the determination of chloride in water. In this case, the simple LOV flow-through electrode system consists of two simple laboratory-made Ag/AgCl electrodes plugged into a port of the flow cell [23]. Other detectors successfully used in conjunction LOV manifolds have been electrothermal atomic absorption 124
http://www.elsevier.com/locate/trac
spectrometers and inductively coupled plasmamass spectrometers [6,23]. These instruments have not yet been used in biochemical analysis, despite the presence of metal in the structures of key biological compounds making their application in this field very promising. A mass spectrometer with the electrospray ionization source also acting as interface to the LOV system provides high capability in biochemical analysis. PEEK tube (0.127-mm i.d.) and a three-way solenoid valve were necessary to connect the LOV system to the MS instrument [18].
6. Applications The SI–LOV format has been applied in the biochemical field with different aims, namely: 6.1. Cell culture The use of a BI–LOV approach for monitoring cell cultures has some advantages in comparison with classical methods, such as: (1) a large number of cells can be interrogated automatically within a small packed column in a sterile enclosed environment; (2) the amount of cells can be easily changed; (3) the whole system can be introduced into an incubator to be kept at constant temperature; and, (4) the cell culture can be replaced in a reproducible manner at any time during a series of experimental runs. These advantages, together with others obtained from the detection system, make the BI–LOV approach a promising alternative for real-time determination of lactate [16], glucose [16,22,36], ammonia, glycerol, and free iron [22], thus providing a means to assess the metabolic regimen of living cells. Monitoring extracellular acidification rates, which is important for the study of cellular activities, can also be carried out by this option [11]. This approach has recently been applied for in situ degradation of H2O2 by living cells using voltammetric monitoring [43]. 6.2. Antibody studies Research focused on real-time monitoring of IgG interactions with protein-coated beads [8,32] implemented in SI–LOV can obtain a great deal of information about antibody–protein interactions and can determine the target analyte (antibody or protein) [21,37]. Nonradioactive ELISA-based methods using beads coated with streptavidin in SI–LOV format have been applied to the determination of GAD65 antibodies involved in type 1 diabetes mellitus [33]. The immobilization protocols of IgG and other proteins have been carried out by l-SI– LOV working in the BI mode [31].
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
Affinity chromatography is being widely used in the purification of antibody production, in pharmaceutical and biotech companies, which require that the injected antibody is completely eluted and the column completely regenerated [11,44]. l-BI spectrometry may also be used to monitor carry-over on the column and to optimize elution conditions in order to ensure complete recovery from the column. 6.3. DNA assays SI–LOV is a feasible alternative to conventional methods for nucleic-acid determination, as shown by its successful application to DNA determination using photometric [27] or laser-induced fluorescence detection [28], based on decoloration of a Crystal Violet solution and substantial fluorescence enhancement of ethidium bromide, respectively. The BI mode of SI–LOV has also been applied to the detection of single-strand nucleic-acid sequences [34]. 6.4. Enzymatic assays lSI–LOV also enables monitoring of initial reaction rates and determination of reactant concentrations in a reaction mixture; examples of the latter are the determination of glucose, ethanol or glycerol [10,45]. The approach is also able to determine both kinetic constants for enzymatic or inhibition reactions, and the type of inhibition (competitive, uncompetitive, or mixed) [41]. Another interesting application is the determination of savinase protease activity, based on the enzymatic degradation of an artificial fluorogenic substrate covalently attached to a tripeptide. As the fluorophore is released from the tripeptide by proteolytic action, fluorescence increase can be measured [8]. 6.5. Other applications The SI–LOV–CE hyphenated approach allows automated monitoring of insulin, proinsulin and c-peptide levels, which are of great interest in the clinical field. The target proteins are sampled and derivatized with fluorescamine in the LOV platform. Individual separation–detection is carried out by a CE-epiluminescence fluorescence microscope [24]. Studies of different conformations of myoglobin caused by denaturation have been conducted using an SI–LOV– CE approach by mixing the analyte with variable concentrations of sodium dodecyl sulfate as denaturing agent in the LOV module. The SI–LOV–LC coupling has been applied to the determination of non-steroidal anti-inflammatory drugs and lipid regulators at the trace level [30]; meanwhile SI–LOV has been used for automated affinity capture and release of biotin-containing conjugates on immobilized streptavidin [18] and for simultaneous measurements of multiple-ligand affinities to proteins immobilized on
Trends
beads [38] – dissociation rate constants were calculated in both cases. l-SI–LOV has also been applied to the determination of pharmaceutical residues (ketoprofen, naproxen, bezafibrate, diclofenac, ibuprofen) and one metabolite (salicylic acid) in surface water, urban wastewater, and urine [46], and to the chemiluminescent determination of tetracycline [29]. In the last case, a bismuthate–immobilized microcolumn was incorporated in one port of the LOV for in situ oxidation of KBr and generation of bromine as oxidant for the bromine–hydrogen peroxide– tetracycline chemiluminescent reaction.
7. Future outlook The LOV approach has opened a promising avenue for in-valve immunoassay and bio-chemical analysis via BI. Nevertheless, the high potential of LOV for sample preparation at the l-scale is even more important. Thus, LOV-based microscale dialysis and non-chromatographic miniaturized separation–preconcentration may be excellent alternatives to conventional analysis. The determination of organic molecules of clinical interest can be one of the areas to benefit from these approaches. The use of LOV manifold and in-situ fluorescence detection could be a new alternative to separate and to purify nucleic acids and other species of biological interest. Multi-reagent assays could be conducted by connecting the central port of the LOV to a second syringe pump. In this way, mixing of multiple zones previously stacked into holding coils could be achieved, merged by simultaneous pumping through the central port. The label dilution technique can be used to evaluate receptor binding, avidin–biotin interactions or any binding between two molecules, one of which can be immobilized on the bead surface. This should provide a very useful tool for the study of diseases such as diabetes. The immobilization of reagents in beads may be focused on the improvement of the way in which the reagent – usually a protein – distribution between beads and supernatant can be monitored simultaneously, and how the reaction mixture can be prepared automatically. For the first goal, a LOV with two in-series flowcells is required to monitor the bead layer and the protein content of the supernatant solution at the first and second flow-cell, respectively. The LOV–l-reactor system could also be used to study different enzymes with photometric detection, to perform automated non-enzymatic kinetic assays, and to carry out automated l-scale titrations. Also the LOV–CE coupling might be applied to real-time monitoring of both cell metabolism (to test the functional response with respect to repeated stimulation) and cell health. New research in LOV–LC coupling should be focused on: http://www.elsevier.com/locate/trac
125
Trends
Trends in Analytical Chemistry, Vol. 27, No. 2, 2008
(a) further miniaturization of the entire analytical setup by hyphenation of the BI–LOV approach with l-LC; and, (b) the synthesis of selective materials fulfilling the stringent demands for their straightforward handling within the LOV conduits. Pending goals on integration of separation devices in the LOV system are l-dialyzers, gas-diffusers, pervaporators or liquid–liquid extractors. These devices could provide macromolecule-free solutions, volatile analyte removal from the sample matrix or solvent exchange and preconcentration, respectively. Liquid–liquid extraction (e.g., without phase separation or with supported liquid membranes) also offers a wide range of possibilities for a judicious choice according to the (bio)chemical system under study.
Acknowledgements SpainÕs Ministry of Science and Technology is gratefully acknowledged for financial support (Project No. CTQ2006-01614). J. Ruiz-Jime´nez and J.A. Pe´rez-Serradilla are also grateful to SpainÕs Ministry of Education for award of their research training fellowships. References [1] P. Solich, M. Polasek, J. Klimundova, J. Ruzicka, Trends Anal. Chem. 22 (2003) 116. [2] J. Ruzicka, E.H. Hansen, Anal. Chim. Acta 78 (1975) 145. [3] E.H. Hansen, Flow injection bibliography (http://www.flowinjection.com/). [4] E.H. Hansen, J.H. Wang, Anal. Chim. Acta 467 (2002) 3. [5] J. Wang, E.H. Hansen, Trends Anal. Chem. 22 (2003) 225. [6] E.H. Hansen, Talanta 64 (2004) 1076. [7] J. Ruzicka, G.D. Marshall, Anal. Chim. Acta 237 (1990) 329. [8] J. Ruzicka, Analyst (Cambridge, U. K.) 125 (2000) 1053. [9] J. Ruzicka, Sequential Injection for Biomolecular Assays, CD-ROM Tutorial, www.flowinjection.com (2004). [10] Y. Chen, J. Ruzicka, Analyst (Cambridge, U. K.) 129 (2004) 597. [11] H.A. Erxleben, M.K. Manion, D.M. Hockenbery, L. Scampavia, J. Ruzicka, Analyst (Cambridge, U. K.) 129 (2004) 205. [12] W. Wang, J.H. Wang, Z.L. Fang, Anal. Chem. 77 (2005) 5396. [13] J. Wang, E.H. Hansen, M. Miro´, Anal. Chim. Acta 499 (2003) 139. [14] X. Long, E.H. Hansen, M. Miro´, Talanta 66 (2005) 1326. [15] X. Long, R. Chomchoei, P. Gala, E.H. Hansen, Anal. Chim. Acta 523 (2004) 279.
126
http://www.elsevier.com/locate/trac
[16] C.M. Schulz, L. Scampavia, J. Ruzicka, Analyst (Cambridge, U. K.) 127 (2002) 1583. [17] S. Kulka, G. Quinta´s, B. Lendl, Analyst (Cambridge, U. K.) 131 (2006) 739. [18] Y. Ogata, L. Scampavia, J. Ruzicka, C.R. Scott, M.H. Gelb, F. Turecek, Anal. Chem. 74 (2002) 4702. [19] C.H. Wu, J. Ruzicka, Analyst (Cambridge, U. K.) 126 (2001) 1947. [20] J. Jakmunee, L. Pathimapornlert, S.K. Hartwell, K. Grudpan, Analyst (Cambridge, U. K.) 130 (2005) 299. [21] H. Erxleben, J. Ruzicka, Analyst (Cambridge, U. K.) 130 (2005) 469. [22] C.H. Wu, L. Scampavia, J. Ruzicka, B. Zamost, Analyst (Cambridge, U. K.) 126 (2001) 291. [23] K. Grudpan, Talanta 64 (2004) 1084. [24] C.H. Wu, L. Scampavia, J. Ruzicka, Analyst (Cambridge, U. K.) 128 (2003) 1123. [25] J. Ruzicka Micro total analysis systems 2000, In: A. Berg, W. Olthuis, P. Bergveld (Editors), Kluwer Academic Publishers, Dodrecht, The Netherlands, 2000, p. 1. [26] M.D. Krom, Analyst (Cambridge, U. K.) 105 (1980) 305. [27] X. Chen, J. Wang, Z. Fang, Talanta 67 (2005) 227. [28] X. Chen, W. Wang, J. Wang, Analyst (Cambridge, U. K.) 130 (2005) 1240. [29] M. Yang, Y. Xu, J.H. Wang, Anal. Chem. 78 (2006) 5900. [30] J.B. Quintana, M. Miro´, J.M. Estela, V. Cerda´, Anal. Chem. 78 (2006) 2832. [31] J. Ruzicka, A.D. Carroll, I. Lahdesmaki, Analyst (Cambridge, U. K.) 131 (2006) 799. [32] A.D. Carroll, L. Scampavia, J. Ruzicka, Analyst (Cambridge, U. K.) 127 (2002) 1228. [33] A.D. Carroll, L. Scampavia, D. Luo, A. Lernmark, J. Ruzicka, Analyst (Cambridge, U. K.) 128 (2003) 1157. [34] K.A. Edwards, A.J. Baeumner, Anal. Chem. 78 (2006) 1958. [35] M. Miro´, S. Jonczyk, J. Wang, E.H. Hansen, J. Anal. At. Spectrom. 18 (2003) 89. [36] C.M. Schulz, J. Ruzicka, Analyst (Cambridge, U. K.) 127 (2002) 1293. [37] Y. Gutzman, A.D. Carroll, J. Ruzicka, Analyst (Cambridge, U. K.) 131 (2006) 809. [38] Y. Ogata, L. Scampavia, T.L. Carter, E. Fan, F. Turecek, Anal. Biochem. 331 (2004) 161. [39] C.H. Wu, L. Scampavia, J. Ruzicka, Analyst (Cambridge, U. K.) 127 (2002) 898. [40] X. Long, M. Miro´, E.H. Hansen, Anal. Chem. 77 (2005) 6032. [41] Y. Chen, A.D. Carroll, L. Scampavia, J. Ruzicka, Anal. Sci. 22 (2006) 9. [42] MicroSIA, FIAlab-3200, and FIAlab-3500 Operation Manual, FIAlab Instrument, Inc., 2005. [43] I. La¨hdesma¨ki, Y.K. Park, A.D. Carroll, M. Decuir, J. Ruzicka, Analyst (Cambridge, U. K.) 132 (2007) 811. [44] D.S. Hage, Clin. Chem. 45 (1999) 593. [45] J. Wang, Anal. Bioanal. Chem. 381 (2005) 809. [46] J. Wang, E.H. Hansen, Trends Anal. Chem. 22 (2003) 836.