The rapid, reliable detection of pathogenic bacteria is imperative in many different industries, of which food and agriculture, healthcare, environmental monitoring, and bio-defense are the four main players [1]. With recent devastating outbreaks of Salmonella and Escherichia coli in the United States, the food industry is largely concerned with the detection of pathogenic bacteria in agricultural products and processed foods. The presence of pathogenic bacteria can cost the food industry and consumers many millions of dollars every year due to food recalls, and is estimated to cause over 30,000 hospitalizations and over 1,000 deaths each year in the United States [2]. In the healthcare sector, approximately 25% of physician visits are caused by infectious diseases, many due to pathogenic agents. The ongoing evolution of microbes due to changing ecological, environmental, and human demographical factors necessitates improvements in the readiness of healthy and emergency service providers to respond to threats through effective surveillance, treatment, and control measures [3]. The development of a fast and sensitive platform for the detection of pathogens in human blood and waste samples is required in order to implement a quick and effective response to an outbreak. In the environmental monitoring arena, considerable attention is given to the evaluation of microbial cells in water and environment quality control, as well as for the study of microorganisms evolution and populations, for example in bio-waste composting substrates and their communities [4]. In the area of biodefense, biological agents are considered far more difficult to detect and defend against than chemical agents, and with bioterrorism now an issue of serious concern, the technology to counter a potential incident needs to be in place. To date, a multitude of reviews on micro total analysis systems for nucleic acid-based detection and microchip pathogen detection methods have been published [1,2,5-19], demonstrating great interest in the development of this field. A comprehensive literature survey was carried out for this present paper, and due to the immense amount of literature related to pathogenic detection, our study focuses primarily on rapid portable systems for the nucleic acid-based detection of bacterial pathogens.

Despite vast improvements in modern-day pathogen detection techniques, the tried and true culturing and plating method still remains the standard method of detection. This technique involves culturing and measuring the growth of individual viable microorganisms using either non-selective media, such as trypticase soy agar, or selective media specially formulated to detect a particular bacterial species. At lower detection levels, however, this method would require a lengthy pre-enrichment step to increase the numbers of viable target bacteria before detection could be conducted. Detection is mainly through enumeration by ocular inspection, which leads to sources of uncertainty due to human variations in sampling and measurements [20]. Due to the low throughput, time consuming and labor intensive process of colony enumeration, this method exhibits low potential for integration and miniaturization into micro total analysis systems. Though several bacterial colony counters have been proposed to automate and standardize this process [21-23], culturing remains a time-consuming process and the required high-quality imaging equipment and software are expensive and cumbersome for integration into a portable system. The recent push for reliable, rapid detection techniques is prompting researchers to explore alternative methods, particularly for detection of bacteria with slower generation times such as the gastroenteritis-causing Campylobacter species, which require a minimum of 3 – 4 days for full-confirmation [24].

Nucleic acid-based methods in pathogen detection are promising in their rapid results, high specificity, and low detection limits of up to, in theory, a single cell. Developed in the mid-1980s, nucleic acid-based technology quickly achieved widespread use in the field of pathogen detection, with a particular focus in polymerase chain reaction (PCR) assays that were developed to detect virtually every clinically relevant bacterial pathogen [1]. In the past decades, our understanding of DNA has grown considerably, with currently 788 fully sequenced microbial genomes [25]. The versatility of nucleic acid-based methods allowed for the design of specific probe sequences, typically on the order of 10 to 30 base pairs in length, to target antibiotic resistance genes as well as for sub-typing of bacteria. DNA is an excellent vehicle for signal transduction due to its characteristic negative charge, and in addition to the typical optical and mechanical measurements, pathogen sensors are often designed to quantify hybridization events between analytes and probe DNA based upon electrical measurements as well. Since these methods target nucleic acids, however, they do not indicate the viability of the target pathogen, so care must be exercised when performing these tests. On the other hand, there are situations where the detection of hibernating or non-viable pathogens is favorable, particularly when aiming to detect unculturable cells [26-28] or to quantify antibiotic effectiveness [29-31]. Nucleic acid-based techniques have a higher sensitivity, therefore requiring a higher level of quality control to prevent contamination, elevating the importance of effective sample preparation to a critical step for successful detection. Consideration of contamination, inhibitors in the specimen sample, and DNA degradation due to unfavorable conditions must be accounted for in the sensor design to help reduce the incidence of false positive or false negative results.

Modern advances in micro- and nanofabrication technology have led to the development of a wide range of nucleic-acid based biosensors that capitalize on the new capabilities of microfluidic technologies and micro total analysis systems in order to reduce reagent and power consumption, enhance analytical performance, and enable portability. These lab-on-a-chip devices incorporate multiple laboratory processes in a semi-automated and miniaturized format. Many of these technologies have been extensively studied [32], successfully commercialized, and are currently widely used in clinical and research laboratories. Nevertheless, portable biosensors systems for point-of-care diagnostics and on-site field testing are still in the infancy stage. Current portable systems tend to be costly and require additional resources as well as skilled operators, therefore rendering the technology unsuitable for point-of-care testing, especially in resource-poor regions such as Africa, Asia, and Latin America that would benefit the most from the development of these platforms [33,34]. Performance of a biosensor platform in the third-world is challenged by the absence or scarcity of trained workers, electricity, equipped laboratories, transportation, and refrigerated storage [34]. Specific areas that need to be addressed during further development include sample pre-treatment, long term storage of reagents, ease of use, and costs [35]. Point-of-care biosensor systems, particularly those utilizing disposable cartridges, must direct some attention towards the development of environmentally-friendly chemicals and materials [34]. Though multiple sensors and assays have been developed for lab-on-a-chip nucleic acid-based detection, few systems have successfully integrated all the necessary sample preparation, sample handling, and detection components into a single automated, portable platform with raw-sample-to-result capabilities. An overview of the translation of traditional microbiological techniques into microfluidic technology is represented in Figure 1.

Proposed by Manz et al. [36] in the early 1990s, micro total analysis systems (μTAS) are integrated miniaturized platforms composed of multi-step sample preparation and detection systems on a single chip that has raw-sample-to-result capabilities – the quintessential “lab-on-a-chip” concept. μTAS systems have experienced rapid growth and development since the completion of the human genome project. The driving force for miniaturization has always been improvement in performance. At the microscale, faster, higher-throughput analysis using parallel systems can be achieved due to a combination of larger surface-to-volume ratios, reduced separation times, shorter diffusion paths, and more efficient reactions. This points not only to the potential for low costs associated with reduced reagent consumption, but also to the ability to analyze smaller samples that were previously insufficient in size. In terms of DNA detection, polymerase chain reaction (PCR) and various other sensing schemes have been successfully carried using integrated microfluidic systems [32]. DNA sequencing and genotyping have been achieved through advances in microchannels technology and capillary array electrophoresis. Complete nucleic acid-based analysis involves complex processes, such as cell concentrating and capture, cell lysis, nucleic acid purification, amplification, and final detection.

The current selection of nucleic acid-based biosensors with target detection of a highly specific DNA signature dictates the need for simple and effective methods of obtaining high-quality DNA. For the majority of biosensing applications, the starting samples consist of tissue, blood, environmental, or food samples [81] and need to undergo careful sample preparation for sensitive detection due to trace or low-abundance species. Although many of the assays based upon polymerase chain reaction (PCR) are fairly robust, a variety of contaminants can inhibit amplification and diminish the success of such analytical instruments. In order to circumvent this problem, target cells must first be extracted and purified from a raw sample through a variety of cell separation and capture techniques. Cell concentrators increase the concentration of microorganisms through gentle means, so as to preserve specific activity or viability, and are important to help increase the sensitivity and strength of the final detection signal [82]. Also, raw samples taken from blood, soil, water, or food are often large in volume for microfluidic analysis, and this discrepancy in volumes makes concentration necessary due to time constraints and the need for rapid detection. The volume analyzed in a typical microscale pathogen detection device ranges from a couple picoliters to, at most, a few microliters. Cell separation is important for separating target cells from contaminants in the raw sample. The three main techniques for cell manipulation involve the use of magnetic, electrokinetic, and mechanical principles.

Magnetic manipulation techniques typically use magnetic particles that can selectively attach to cells of interest through the use of antibodies and other linking chemistries, and use magnetic field gradients to capture the bead-cell conjugates, as illustrated in Figure 3. Investigations into continuous flow separations [83,84] and matrix-based manipulations [85] using magnetic capture have been promising. E. coli has been shown to be magnetically separated from PBS and whole blood on an integrated microfluidic device consisting of a chaotic mixer, incubation channel, and a capture channel [86]. The magnetic method is clean, versatile, and non-invasive, and with advances in magnetic bead materials and chemical modification techniques, the technique has the potential to become increasingly efficient and easily integrated into a portable system [18].

Cell manipulation using dielectrophoresis (DEP) takes advantage of the intrinsic dielectric properties of cells and their response to electric fields, and has been extensively studied on microscale devices [87]. A DEP chip fabricated from acrylic has been reported by Huang and colleageues to separate B. cereus, E. coli, and L. monocytogenes from blood [88]. Using DEP microchip technology, live cells have been shown to be separated from dead cells through differences in cellular dielectric properties at differing states of viability [89]. In addition, single-cell trapping has been achieved using DEP in conjunction with laser-trapping forces [90]. Mechanical cell separations have been achieved using microfilters [91], microwells [92,93], and surface-modified microchannels [94,95]. Microbial cells have been concentrated using size-dependent filter-based microfluidic devices [91], which are typically rapid and highly efficient, though lacking in selectivity. For portable point-of-care devices, low cost and power consumption is necessary, without sacrificing on sensitivity, and magnetic bead-based separations have shown considerable promise in this area.

Upon cell capture and isolation from the raw sample, cell lysis is necessary to release the nucleic acids for further analysis. Among the various lysis methods, chemical lysis is most common. Chemical lysis can be easily incorporated into an integrated microfluidic design with methods such as on-chip mixing of captured cells with sodium dodecyl sulfate or guanidinium thiocyanate [96] and hydroxide electro-generation-induced cell poration and lysis [97]. Unlike mammalian cells, the efficient lysis of certain bacteria for DNA extraction can be more challenging. Gram-negative bacteria is commonly treated with alkaline buffers or guanadinium thiocyanate, whereas gram-positive bacteria is more difficult to lyse often requiring multistep methods, though heating in the presence of chelating resins, sometimes with beads, has been shown to be effective [98]. Heat-based techniques, such as freeze-thaw or freeze-boil methods [99] are also available, and pulsed laser irradiation of carboxyl-terminated magnetic beads [100] has been reported for on-chip pathogenic DNA extraction. However, most thermal methods are seldom employed due to likelihood of denaturation due to high heat. Electrical pulsing methods have also been incorporated into microfluidic chips to electroporate cells [101]. Mechanical disruption methods, such as sonication, release cellular components into solution but often require more energy but have been demonstrated in microscale devices [102,103]. High-frequency sonication uses piezoelectric materials to generate pressure waves that disrupt cell membranes, and though effective against hard-to-lyse cells, this method generates considerable amounts of unwanted heat and free radicals [98].

The traditional method of purifying DNA is performed via proteinase K digestion in the presence of detergents, phenol-chloroform extraction, and concentration by alcohol precipitation [98]. One of the most common modern techniques for DNA purification is through chemical lysis followed by capture using silica-based resins. DNA in chaotropic salt-containing buffers such as those containing guanidinium or sodium iodide salts, preferentially bind to silica surfaces, whereas other macromolecules such as proteins and lipids remain free in solution [104,105]. These unwanted components are traditionally removed using centrifugation and alcohol washing steps using commercially-available kits. However, the fact that they are usually based upon particulate matrices presents challenges to integration onto μTAS devices. While incorporation of silica-based resins into a microfluidic device has been reported [105,106], new innovative silica pillar arrays (see Figure 4) have also been investigated for microscale DNA purification [48] which circumvent the problems associated with filling channels with binding matrices after microfabrication.

Integration of all the microfabricated components needed to perform DNA detection to achieve portable, automated raw-sample-to-result functionality is no easy task. Several groups have already begun to address this challenge, with the incorporation of micropumps, microvalves, micromixers, heaters, detectors, and other analytical components. Significant progress has been demonstrated in various types of platforms including but not limited to, capillary driven test strips, centrifugal microfluidic devices, droplet-based microfluidic platforms, and large-scale system integration platforms. Though true μTAS systems are primarily still in the laboratory development stage, partially integrated μTAS devices have been developed for commercial applications. Of the different types of μTAS systems, PCR microfluidics is among the most prevalent and has been integrated with on-chip sample preparation and capillary electrophoresis.

The functional integration of PCR and capillary electrophoresis on a single microchip was successfully integrated by Koh et al. for the detection and identification of two model bacteria, Escherichia coli O157 and Salmonella typhimurium [160]. Similar systems were fabricated in a variety of different materials, namely PMMA, polycarbonate, and PDMS [15]. More recently, DNA purification and real-time PCR were successfully integrated for the single-chip detection of Listeria monocytogenes by Cady et al. on a portable instrument shown in Figure 10(a) with on-board pumping, valving, thermal cycling, and detection functionalities. The single-chip is shown in Figure 10(b).

In this work, DNA purification was performed by running lysed cell samples through a channel arrayed with 10 μm silicon oxide pillars and PCR was conducted in a serpentine amplification channel using an external thermal cycler, and real-time detection was performed with LED-excitation of TaqMan and measurements using a miniaturized PMT [32,48]. Automated sample preparation PDMS chips have also been developed to isolate nucleic acids from small numbers of bacterial cells with all cell isolation, cell lysis, DNA and mRNA purification and recovery processes carried out on a single standalone nanoliter-volume chip [161,162]. A fully-integrated chip for immunomagnetic bead-based sample preparation, PCR, and DNA microarray detection has been developed for the detection of Eschericia coli K12 from rabbit blood samples [163]. Other fully-integrated DNA-based assays that have been shown to be capable of multiple pathogen detection make use of individual electrode surfaces immobilized with capture probes [164,165]. In one particular system, signal amplification is achieved via tagging of the target DNA with gold nanoparticles and detection is based on measuring the amount of subsequent electrocatalytic deposition of silver metal onto the nanoparticles [165]. Other integration formats include the innovative compact disk device, also known as LabCD [166], a commercial product which utilizes centrifugal forces for pumping of fluids through reservoirs, valves, mixing chambers, and heating chambers. The control of flow rates through the device is tuned via different disk spin speeds, and is capable of achieving sample preparation, DNA purification, and PCR amplification.

In the laboratory setting, fully integrated systems exist for DNA analysis of complex biological samples employing the concept of raw-sample-to-result. One such system employs thermally-actuated paraffin-based microvalves and electrochemical and thermopneumatic pumps to achieve sample preparation (magnetic bead-based cell capture, cell pre-concentration and purification, and cell lysis), PCR, DNA hybridization and electrochemical detection on a single device. It has been demonstrated to show detection of pathogenic bacteria from milliliters of whole blood samples. An integrated portable genetic analysis system for pathogen detection has been prototyped by Mathies et al. using rapid PCR amplification followed by capillary electrophoretic separation of labeled analyte and fluorescent detection, and has been demonstrated directly on E. coli and Staphylococcus aureus cells [167]. In the push for detection in smaller sample volumes, fully integrated nanoliter-volume systems have also been developed in recent years [168]. The commercially-available Cepheid GeneXpert^®(GX) system employs single-use microfluidic cartridges to integrate sample preparation, amplification, and detection. Utilized by the U.S. Postal Service for the detection of anthrax spores, the GX system has been shown to be both user-friendly and effective [169,170]. Other commercial μTAS systems for DNA analysis have also been developed by numerous microfluidic companies in a variety of formats and functionalities, including ACLARA BioSciences, Fluidigm Corporation, Affymetrix, Agilent Technologies, Alderon Biosciences, Roche Molecular Diagnostics, and Motorola Inc [15,171].

The development of a fast, sensitive, multiplexed, and easy to operate pathogen sensing systems will have global impacts on healthcare, agriculture, environmental monitoring, and bio-defense. Different strategies have been used in both research and commercial settings to develop nucleic acid-based sensors and lab-on-a-chip systems. With the need for portable, disposable DNA chips to replace traditional, expensive, and bulky instrumentation, applications for DNA-sensing devices are being rapidly driven towards rapid pathogen detection, DNA sequencing, and drug discovery. Miniaturization and reliability are the main challenges to widespread distribution of portable nucleic acid-based sensors with raw-sample-to-result functionality. One major bottleneck limiting the portability of nucleic acid-based microfluidic system lies in the difficulty of integrating sample preparation. In order to realize these handheld diagnostic systems, on-chip processing of raw samples and mastery of automated microfluidic control must be achieved. Enabling technologies discussed in this review will have a significant impact on the future development of handheld (point of care) nucleic acid-based detection systems. As mentioned in this review, dielectrophoretic (DEP) sample preparation, filtration-based separation, and immunomagnetic separation are all viable options for enriching target microorganisms from samples. Although DEP offers several advantages over other methods, such as its ability to distinguish between live and dead cells, this technology is more difficult to integrate into miniaturized systems. This is primarily due to complex electronic control architectures, and the incompatibility of this technique with heterogeneous sample matrices. At this time, filtration-based sample preparation, followed by immunomagnetic separation, is the most compatible purification technology with point of care systems. Passive methods, such as those offered by capillary forces, gravity, or creative topography are generally preferred due to lower power consumption, but battery-powered or hand-powered options can also provide a practical approach [34].

Following sample preparation, nucleic acid extraction/purification and nucleic acid detection must be addressed for the development of viable point of care detection systems. Multiple research groups have demonstrated that a microfluidic solid phase extraction approach towards nucleic extraction and purification is most compatible with miniaturized devices. The use of solid phase resins or microfabricated structures provides compatibility with microfluidic architectures and reduces the total volume of purified nucleic acid for subsequent detection. Detection technologies must also be optimized for point of care systems. As described, fluorescence-based detection methods, such as fluorogenic real-time PCR provide extremely high sensitivity. The complexity of the detection optics may limit the applicability towards miniaturized devices, but multiple research groups, including ours, have demonstrated optical detection in portable, low-power platforms. Competing methods, such as electrical detection may not provide the needed sensitivity for nucleic acid-based detection, but could offer less complex detection components. Of the currently used electrical techniques, impedance-based methods provide the highest sensitivity with the most information-rich output, and should be further investigated. Mechanical detection methods often require complex optical or electrical analysis (such as cantilever-based techniques), which make them no better suited to point of care applications than optical methods. From this standpoint, future point of care detection systems will mostly likely be based upon a microfluidic platform using solid phase extraction, PCR amplification, and a fluorescence-based optical readout. Concurrent development of high sensitivity electrical detection methods, such as transistor-based detection, may yield effective detection elements for far-term analytical systems.

To further decrease processing time and detection limits, further investigations into technologies with even higher specificity and sensitivity are needed. Nanotechnology will play a vital role in the development of new techniques for nucleic acid detection. The implementation μTAS systems will allow for easy standardization of methods necessary for the reliability and repeatability of results. However, the need for disposability imposes a limit on the size and cost of portable sensors, hence increasing the complexity of the technology may become too uneconomical for production. Much of the current published work, however, is promising, utilizing passive microfluidic components that are easily integrated into disposable devices. And, efforts to push towards non- or minimally instrumented diagnostic devices are in place [17].