Plasmofluidics for Biosensing and Medical Diagnostics
- 744 Downloads
Plasmofluidics, an extension of optofluidics into the nanoscale regime, merges plasmonics and micro-/nanofluidics for highly integrated and multifunctional lab on a chip. In this chapter, we focus on the applications of plasmofluidics in the versatile manipulation and sensing of biological cell, organelles, molecules, and nanoparticles, which underpin advanced biomedical diagnostics.
1 Definition of the Topic
Plasmofluidics, an extension of optofluidics into the nanoscale regime, merges plasmonics and micro-/nanofluidics for highly integrated and multifunctional lab on a chip. In this chapter, we focus on the applications of plasmofluidics in the versatile manipulation and sensing of biological cell, organelles, molecules, and nanoparticles, which underpin advanced biomedical diagnostics.
An improved capability of manipulating and sensing living cells, viruses, bacteria, and molecules (i.e., DNAs, proteins, and drug molecules) is significant for biomedical research and disease diagnostics in global health, primary care, and point-of-care (POC ) settings. Conventional diagnostic technologies such as polymerase chain reaction (PCR) and enzyme-linked immunosorbent assay (ELISA) have been established as global gold standards. However, they have mainly been used in developed countries and major cities in developing countries due to their sophisticated instrumentation and complex operation procedures that require specially trained staffs [1, 2]. Therefore, there is still a great need for innovative portable POC devices that can provide effective high-value healthcare to underdeveloped and resource-constrained regions. Herein, we argue how plasmofluidics, a synergistic integration of plasmonics and micro-/nanofluidics , can help address current challenges in portable POC devices as well as fundamental biomedical studies. This book chapter focuses on two correlated areas: plasmofluidic tweezers and biosensors. The first part of this book chapter reviews state-of-the-art plasmofluidic tweezers for versatile manipulations of biological cells, DNAs, and proteins, including the latest strategies for overcoming the inherent drawbacks of plasmonic tweezers. The second part discusses various plasmofluidic biosensors for the detection of nucleic acids, proteins, pathogens, and drug molecules. We highlight the progress made at improving the limit of detection (LOD ), response time, and accuracy for analyzing clinical samples.
Infectious diseases, which are caused by bacterial (e.g., pneumonia), mycobacterial (e.g., tuberculosis), viral (e.g., HIV), fungal (e.g., candidiasis), and parasitic pathogens (e.g., malaria), have considerable impact on global economy, health, and security [1, 2]. For instance, recent years have witnessed large outbreaks of Ebola in Africa and Zika in Central and South America, and there have been serious concerns and heated debates on public health at the global scope. Cancers, which are characterized by the uncontrolled growth and spread of abnormal cells and induced by various factors such as tobacco, infectious organisms, genetic mutations, and immune conditions, are the leading cause of death for much of the US population . The threats from infectious diseases and cancers become increasingly prevalent due to factors such as changing trends in human and animal migration, increasing urbanization, environmental deterioration, and climate change. Early disease diagnosis and effective treatment have become critical for the clinical regulation and management of infectious diseases and cancers. Convectional disease diagnostics has relied on techniques such as optical microscopy, culture, immunoassays, and nucleic acid amplification . A standard process for disease diagnostics includes the following steps: (a) collection and transport of biological samples such as blood, urine, and tissue swabs from the point of care; (b) analysis of the samples by experienced staffs in a centralized laboratory; and (c) notification of the results to the clinicians and patients. Due to highly frequent access to energy resources like electricity, time-consuming procedures, and need for well-trained personnel, the conventional clinical methods prelude the rapid disease detection and response at the primary care. Their drawbacks are prominently revealed in resource-limited and underdeveloped areas.
Point-of-care (POC ) devices, which enable on-site test and follow-up action, are promising to improve the diagnosis and management of infectious diseases and cancers in various clinical settings. These include areas where healthcare infrastructure is limited and high-quality timely medical care is inaccessible [5, 6]. With the advancements of nanotechnologies and micro-/nanofluidic technologies, many innovative POC biosensors with optical, electrical, and mechanical interrogations have been developed. However, two major challenges, matrix effect and system integration, have prevented the further developments and uses of POC devices . Most biosensors can have excellent performance with pristine samples. However, they still need to be strictly evaluated with clinical samples due to the matrix effect. Generally, the matrix of human fluids becomes more complex in infected states, which could lead to the clogging of microfluidic channels and the decrease of the transduction signals. Transformation of POC devices from proof-of-concept and benchtop to bedside care and point-of-care settings requires the system integration. However, to integrate and package various modules for sample preparation and signal detection into a fully automated and user-friendly platform remains challenging.
Plasmofluidics, which seeks to synergize plasmonics and micro-/nanofluidics for lab on a chip, is promising for advancing the next-generation POC devices that feature high compactness, high integration, and multiple functions as well as the frontiers of biomedical research . With their capability of controlling light at the nanoscale beyond the diffraction limit, surface plasmons such as surface plasmon polaritons (SPP s) and localized surface plasmon resonances (LSPR s)  are effective at optically manipulating, sensing, and analyzing biological cells and molecules [9, 10, 11]. Micro-/nanofluidics exploit rich fluidic behaviors at the micro-/nanoscale to enable low-load, high-throughput, cost-effective, and precise delivery of analyte samples. Therefore, plasmofluidic platforms can serve the purpose of processing, sensing, and analyzing biological objects in clinical samples, paving the way toward affordable and portable healthcare devices for their uses in primary care and resource-constrained settings.
There have been several successful review articles on plasmonic biosensing and medical diagnostics [1, 12, 13]. However, a comprehensive review article focused on the emerging field of plasmofluidic sensing is not available yet. To fill the blank, our chapter covers classification, working principles, design strategies, and applications of the state-of-the-art plasmofluidic systems for sensing and medical diagnostics.
4 Experimental Methodology
4.1 Manipulations with Plasmon-Enhanced Optical Near Fields
Biosensing and medical diagnostics often rely on the capability of manipulating biological particles and molecules in fluidic environments. Three prominent techniques that exploit plasmon-enhanced optical near fields to manipulate particles and molecules in fluids have been developed: (i) SPP -based plasmonic tweezers, (ii) LSPR -based plasmonic tweezers, and (iii) plasmonic tweezers based on self-induced back-action (SIBA ). Both top-down fabrication techniques such as electron beam lithography and focused ion beam lithography and bottom-up fabrication techniques such as directed assembly and self-assembly have been applied to engineer plasmonic structures for targeted tweezing platforms [14, 15].
Due to the localized electromagnetic fields associated with the excitation of LSPR s, LSPR -based tweezers are often applied to trap single or multiple particles near the plasmonic structures. Plasmonic nanoantennas such as nanodot pairs , diabolo structures , and bowtie structures  were commonly used to induce strong near-field confinement and enhancement in the nanoscale dielectric gaps known as “Hot spots,” which provide stable trapping of nanoparticles at low optical power. Trapping stiffness was enhanced by two orders of magnitude with a metal nanodot pair in a conventional optical tweezers setup, as shown in Fig. 5.1c–d . Despite the localized nature of LSPR s, recent research efforts have achieved long-range transportation of particles by using arrays of plasmonic C-shaped engravings  and gold nanoislands with a network of “hot spots” .
LSPR -based tweezers exhibit limits in manipulating particles that are smaller than 100 nm due to the high-power requirement and significant thermal effects. SIBA -based plasmonic tweezers were developed to reduce the required power intensity by more than an order of magnitude . As illustrated in Fig. 5.1e, a nanoaperture redshifts in its resonance wavelength when the trapped particle has a larger refractive index than solvent, enhancing the laser beam transmission through the nanoaperture. When the particle tends to escape from the aperture, a drop in the light transmission generates a restoring force and pulls the particle back to the equilibrium position. In other words, when the particle is moving away from the optical potential well, the SIBA force will deepen the potential well and maintain the trapping state, as illustrated in Fig. 5.1f–h.
4.2 Manipulations with Plasmon-Enhanced Photothermal Effects
The plasmon-assisted heating was exploited to control fluid motion at the microscale and nanoscale . In typical plasmofluidic systems where the characteristic size of the plasmonic structure is very small (~100 nm), thermal diffusion of the fluid dominates in heat dissipation and thermal convection plays a minor role. As shown in Fig. 5.2c–d, the thermal convection velocity at small plasmonic structures is limited to 1–100 nm/s. In this case, the thermal convection can barely contribute to the motion of the particles in the fluidic environment. The thermal convection field can be further damped by more than one order of magnitude when the height of the fluid chamber is reduced to ∼10 μm, as illustrated in Fig. 5.2e.
4.3 Plasmonic Sensing of Particles and Molecules
Of particular interest are metal nanohole arrays integrated with complex fluidic structures for versatile and parallel sensing [48, 49]. On-chip nanohole array-based sensors feature simple optical instrument and high synergy with microfluidic schemes. In the flow-over mode (Fig. 5.3b), only the upper metal surfaces of the nanohole arrays are exploited like those in the convectional SPP -based sensors. Through scaling analysis and numerical simulation, Escobedo’s group has found that the flow-through mode features ~10-fold enhancement of time response over the flow-over mode in typical biosensing applications . Further experiments showed that the flow-through mode enabled active delivery, concentration, and sensing of analytes, leading to one order of magnitude increase in sensing efficiency and two orders of magnitude improvement in LOD .
In general, a LSPR biosensor consists of the following components: (i) a recognition element in conjunction with the plasmonic substrate, (ii) a transducer to convert the interaction between target and recognition element to optical or electrochemical output signals, and (iii) a system to interrogate the signals. Spectroscopic measurements are the most common approach for the signal analysis, which include (i) transmission spectroscopy (Fig. 5.3c), (ii) reflection spectroscopy (Fig. 5.3d), and (iii) dark-field scattering spectroscopy (Fig. 5.3e) . Surface plasmon resonance imaging (SPRI ) has been developed to perform large-area measurements at high resolution . It is a label-free method of visualizing binding activities across an arrayed biochip via a video CCD camera.
Surface-enhanced Raman spectroscopy (SERS ), which often arises from the strong electromagnetic field enhancement proximate to the metal surface, has proved as a powerful tool for label-free analysis of molecules. SERS interrogates Raman shifts originating from molecular vibrational energy levels [53, 54, 55]. Metal colloids, colloid aggregations in solutions, and metal nanoparticle arrays on substrates are commonly used for SERS -active sensors integrated with microfluidics [55, 56]. There are two major microfluidic approaches toward SERS -active plasmofluidic sensors: (i) mixing the sample with SERS -active colloids in a microfluidic channel (Fig. 5.3c) and (ii) driving the sample through a microfluidic channel with a SERS -active substrate at the bottom (Fig. 5.3d) . Such plasmofluidic sensors have the advantages of measuring biological particles and molecules in their native watery environments. However, since the molecular adsorption or binding to the SERS -active surfaces relies on the diffusion of the analytes in solutions, the Raman signal intensities will be weaker than those of SERS based on dried samples on solid-state Raman substrates.
5 Key Research Findings
5.1 Versatile Manipulations of Particles and Molecules in Plasmofluidic Systems
Manipulations of biological particles and molecules such as trapping, immobilization, and pre-concentration are becoming a critical component of biosensing and analysis in fluidic environments. Plasmonic tweezers have been employed in manipulating various bio-specimens such as cells [58, 59], DNAs [60, 61], and proteins [62, 63, 64] at high spatial resolution and low optical power. Using simple optics to create the trapping force, plasmonic tweezers can be readily incorporated into microfluidic systems to design novel plasmofluidic chips with functionalities such as single-particle trapping [57, 62], parallel trapping , co-trapping , and kinetic detection of biological objects [61, 64]. One has also recognized multiple issues of current plasmofluidic systems and proposed new solutions. First, due to the near-field nature of surface plasmons and weak thermoplasmonic convection, only particles that diffuse to the close proximity to the plasmonic structures can be captured . Second, isolated plasmonic nanostructures are commonly used to avoid collective heating and particle agglomeration, preventing long-distance transportation and dynamic manipulations. Third, in contrast to conventional optical tweezers that are capable of three-dimensional (3D) manipulations, plasmonic tweezers are often limited to two-dimensional (2D) trapping of objects at the plasmonic structures. New solutions to these issues in the plasmofluidic systems include the adoption of optical components with a higher level of spatial control  and the exploitation of loss-induced heating at the plasmonic structures .
5.1.1 Plasmon-Enhanced Trapping of Single Biological Cells and Molecules
The capability of trapping fragile biological objects such as cells, DNAs, and proteins in fluids is of significant importance in cellular and molecular biotechnology. E. coli cells were trapped in parallel with near-infrared LSPR s on the arrays of plasmonic nanoantennas in a Kretschmann optical setup . The cells were stably aligned along the antennas’ long axis, and their continuous growth and division were kept over 2 h. There was no difference in the average division time between the optically trapped cells and those outside the illuminated region. In contrast, optical tweezers have challenges in trapping the E. coli cells for their low refraction index (1.38 for visible light) and small dimensions (2 × 0.8 μm2 at infant stage).
Optical trapping of single DNA molecules and DNA translocations in LiCl buffers were demonstrated using a solid-state plasmonic nanopore, as shown in Fig. 5.5b . The nanopore is positioned between the gap of a gold bowtie antenna. The enhancement of the rate of DNA translocation events was observed and attributed to the plasmon-assisted local heating and the thermophoresis in the thermal gradient. In another example, single bovine serum albumin (BSA) molecules were trapped onto a double-hole structure with a 3.5-mW laser beam focused by a 100× oil immersion objective , as shown in Fig. 5.5c. The folding and unfolding states of a trapped BSA molecule were revealed by measuring the intensity of the transmitted light. The same research group demonstrated the protein-antibody co-trapping at the double-hole structure and measured the binding kinetics of the protein-ligand interaction, opening up new avenues for studying intermolecular interactions at the single-molecule level [63, 64].
5.1.2 Three-Dimensional Manipulations and Long-Range Transportation of Objects
To solve the long-standing challenge of rapid and on-demand loading of objects at the plasmonic trapping sites, Boltasseva’s group has combined the plasmonic heating and AC electrical fields for fast delivery and trapping of nanoobjects within a few seconds . The coupling of the plasmon-enhanced temperature gradient and an applied AC electrical field induces an electrothermoplasmonic (ETP) flow, which leads to fluidic motion two orders of magnitude faster than the thermoplasmonic convection and greatly increases the particle capture efficiency, as shown in Fig. 5.6c–d. With their capability of rapid manipulation and concentration of DNA and protein samples, the hybrid plasmofluidic tweezers can be applied for nanoscale biosensors to improve the sensing throughput and efficiency.
5.2 High-Performance Sensing, Analysis, and Diagnostics in Plasmofluidic Systems
Benefiting from a synergy between plasmonic nanotechnology and micro-/nanofluidics , biosensors based on plasmofluidic platforms are attractive for POC devices [1, 2, 7]. They have been applied for portable sensing, analysis, and diagnostics in biomedicine and healthcare. The analytes include nucleic acids, protein, viruses, bacteria, and drugs. Tremendous efforts have been made to improve LOD , response speed, and accuracy of multiplexed identification. More innovative approaches are being developed to push proof-of-concept or benchtop prototypes into practical uses in directly analyzing bio-objects in body fluids for disease diagnosis. Herein, we focus on plasmofluidic POC sensing for DNAs/RNAs, proteins, viruses, bacteria, and drugs.
5.2.1 Detection of DNAs and RNAs
SERS -based plasmofluidic sensors for high-throughput detection of multiple DNAs in single assays were also developed. One example is based on self-assembled gold nanoparticles (Au NPs) on a gold nanowire (Au NW) at the presence of target DNAs . Raman signal was significantly enhanced by placing the DNA molecules in the interstices of the gold-particle-on-wire structure. The structure was constructed by sequential incubation of Au NWs premodified by thiolated probe DNAs in solutions of target DNAs and Au NPs functionalized by reporter DNAs with Raman dyes (Fig. 5.7c). A high selectivity is assured because only those target DNAs with sequences complementary to the probe DNAs and the reporter DNAs can form the particle-on-wire structure for the enhanced signal of the Raman dyes. The sensor was applied for quantitative detection of DNA concentrations, multiplexed detection of bacterial DNAs, and identification of pathogenic bacteria in real clinical samples. The results on bacteria agreed well with those obtained by conventional culture-based assays, as illustrated in Fig. 5.7d.
Vo-Dinh’s group devised a label-free DNA biosensor based on molecular sentinel (MS) immobilized on a metal film over nanosphere (MFON) [71, 74, 75, 76]. Upon DNA hybridization, the Raman label at the end of the MS is separated from the MFON’s surface to exhibit decreased Raman signal, which works as the readout and renders high selectivity in multiplexed analysis. With the initial detection of human RSAD2 gene  and Ki-67 DNA  and the multiplexed detection of IFI27 A and IFI44L , the group developed the biosensor into a bioassay-on-chip platform and detected ssDNA of dengue virus (LOD ~6 attomoles), which is the culprit of dengue fever that plagues 230–390 million people each year. The platform can be combined with microfluidics for on-chip sample preparation and detection, providing a new tool for POC diagnostics and global healthcare.
Detection of microRNAs (miRNAs) has been a key topic in cancer research, diagnosis, and prognosis. miRNAs can repress gene expression in a sequence-dependent manner  and are associated with various human diseases such as diabetes, Alzheimer’s, and cancers [78, 79, 80]. At an early stage of cancer, extremely low concentrated miRNAs circulate in human body fluid, making detection of miRNAs with improved LOD and sensitivity desirable for early disease diagnosis and implementation of new treatment options .
Joshi et al. designed a regenerative LSPR -based miRNA sensor for early diagnosis of pancreatic ductal adenocarcinoma (PDAC), a deadly cancer with an overall 5-year survival rate of 7%. The cancer is hard to be detected when the tumor is small and at nonmetastatic stage [82, 83]. The sensor is based on gold nanoprisms functionalized with –S-C6-ssDNAs . Direct hybridization between the –S-C6-ssDNAs and the target miRNAs forms DNA duplex (Fig. 5.7e), which increases the refractive index in local environment of the nanoprism and redshifts the LSPR peak wavelength. The high sensitivity down to 10−18 M for specific miRNA-10b benefits from multiple factors: (a) atomically flat surface of the nanoprisms allowing efficient duplex formation, (b) high charge density of the duplex greatly altering the local refractive index, and (c) strong field enhancement at the tips of nanoprisms. Detection of miRNA-10b at a low concentration was achieved in pancreatic cancer cell lines, derived tissue culture media, human plasma, and plasma exosomes. Quantification of miRNA-10b in highly purified exosomes that were isolated from patients with PDAC and chronic pancreatitis (CP) was shown in Fig. 5.7f. PDAC patients had much higher miRNA-10b levels than CP patients. Integration of the sensor into microfluidics is expected to further enhance the miRNA measurements and the disease diagnosis with identifiable miRNA signatures.
5.2.2 Detection of Proteins
Proteins in biofluids such as serum are highly associated with biological functionalities. Detection of proteins and analysis of their unique sequences have high clinical relevance [84, 85]. For instance, a protein biomarker called SVEGFR-1 was increasingly expressed in patients with myelodysplastic syndromes (MDS), i.e., a diverse group of clonal disorders of the hematopoietic stem cell. Prostate-specific antigen (PSA) shows increased levels above the normal limits (4 ng mL−1) in serum for possible prostate malignancy, which can be used for diagnosis and prognosis of prostate cancer . Plasmofluidic immunoassays and immunosensors feature parallel, label-free, and real-time detection of proteins with high sensitivity, selectivity, multiplicity, and reproducibility [30, 87, 88, 89, 90]. They are also applied to investigate protein-protein interaction , monitor live cell secretory events , and identify diagnosis-related cells .
SPP -based plasmofluidic sensors for label-free detection of proteins have recently been implemented on a smartphone platform . A plasmonic sensor with dispersionless microfluidics was developed to detect protein biomarker of MDS disease (i.e., sVEGFR-1) via interaction with its high-affinity bio-receptor VEGF-A . A detection limit of 25 ng/mL was achieved in 2% human blood plasma by using the sequential injection approach. However, the traditional experimental design that relies on a specific bio-receptor (known as “lock-and-key” approach) is limited by interference from molecules that are structurally or chemically alike. Moreover, most diseases are associated with multiple biomarkers, which necessitate the detection of a total protein distribution rather than a particular protein. To address these challenges, Choi et al. developed a “cross-reactive”-based sensor using a substrate of multiple segments of surfaces that are pre-adsorbed by different types of proteins. A distinctive pattern of SPP angle shifts arose from the interaction of the substrate with a sample of variable proteins . The SPP angle pattern was used to clarify the concentration distributions of proteins and a specific biomarker, C-reactive protein (CRP), in a cocktail sample.
More challenges arise for SPP -based sensing in human serum samples due to (i) high nonspecific interaction between the sensor surface and serum proteins and (ii) matrix effects of serum such as high refractive index and viscosity that mask the binding events between the recognition elements on the sensor surface and the analytes in the serum . Uludag et al. used a matrix elimination buffer to eliminate nonspecific binding of 98% serum proteins and performed a sandwich assay, in which the target proteins were sandwiched between a bottom layer of capture antibody and a top layer of antibody-modified Au NPs . The Au NPs reduced the refractive-index mismatch between the buffer and the sample, leading to amplified SPP signal and enhanced sensitivity. A LOD of 0.29 ng mL−1 for tPSA in 75% human serum was attained, which was comparable to that achieved by a quartz crystal microbalance (QCM) sensor. SPP and QCM sensors were further combined to detect sub-attomolar human α-thrombin . The sandwich immunoassays with SPP -based sensors are expected to improve early-stage cancer diagnosis and prognosis.
LSPR -based plasmofluidic sensors enable high-end miniaturization down to the single-nanoparticle scale [87, 97, 98, 99, 100]. Zijlstra et al. reported real-time, label-free detection of single-protein binding events by monitoring the LSPR of a nanorod with an ultrasensitive photothermal assay . When a single protein binds to the receptors on the surface of the nanorod, a small redshift of the longitudinal LSPR wavelength changes the absorption cross section of the nanorod at the wavelength of the heating beam, which can be measured with the transduced temperature change. Unlike the resonance Rayleigh scattering, the contrast scale of photothermal microscopy permits the measurement of mode volume that is commensurate to the size of a protein.
Ament et al. utilized single gold nanorods to monitor single-protein conformational dynamics and equilibrium coverage fluctuations . The latter contain the information on the binding kinetics and nonequilibrium thermodynamics. By using a white light laser, an intensified CCD camera, and an engineered nanoparticle geometry, they achieved considerable improvement of signal-to-noise (SNR) ratio and time resolution over previous techniques to identify single-molecule binding events. Rather than utilizing passive surface binding to capture the proteins, Gordon’s group used a double nanohole to actively trap single proteins and to measure the single-protein binding events .
LSPR biosensors based on nanoparticle ensembles and microfluidics are more suitable for clinical applications because the ensembles allow statistically significant data analysis, relax constraints by nanoparticle variations, and enable simple instrumentation, fast data acquisition, and improved signal-to-noise ratio . Plasmonic nanotechnologies such as plasmonic arrays [87, 88, 101, 102, 103] and SPRI [104, 105] and innovative microfluidic techniques such as integrated concentration gradient generator  and multi-well fluidic measurement  have been intensely pursued to detect and quantify cancer biomarkers with enhanced sensitivity, robustness, integrity, high throughput, and multiplexity.
Measuring binding affinities between proteins is also of fundamental interest and clinical significance. A “nanoSPR” method was proposed to simultaneously characterize the binding affinities among many macromolecular partners . Briefly, multiple batches of gold nanorods functionalized with different proteins and one batch without functionalization (as a reference) were driven through the flow cell consecutively to generate a positon-encoded sensor. The encoding of the sensor was carried out by repetitive recording of the position of each randomly deposited gold nanorod, as shown in Fig. 5.9c. As a demonstration, the sensor was applied to study the interactions of three bacterial division proteins with a target protein FtsZ, which is an essential element of the division machinery in most bacterial systems. The results had an excellent agreement with those by conventional composition gradient static light scattering and fluorescence anisotropy. With the small device size, reduced sample volume, and built-in statistics, the sensor is envisioned to be a powerful tool in drug screening and discovery applications.
In light of the important role of cell secretion in a variety of physiological processes, Wu et al. developed a label-free, ultrasensitive, plasmofluidic platform based on gold nanoslits with strong Fano resonance to monitor dynamic live cell secretory activities . The THP1 cells were trapped and cultured in microfluidic channels with the sensing surface of gold nanoslits ~3–5 μm away from the cell membrane. Upon stimulation with continuous lipopolysaccharide, the cells were found to secrete MMP-9 proteins in 2 h. The plasmofluidic sensor needs less than ten cells for the study, while ELISA analysis demands at least a few thousand cells. The sensor can be applied for fundamental study of secretory events in different physiological environments and medical diagnosis of various diseases at the single-cell level.
Formation of stable electromagnetic “hot spots” in nanoparticle aggregates is another strategy for SERS -based protein analysis with high reproducibility and sensitivity. Saha et al. designed a paper-based microfluidic SERS system to detect streptavidin and glycoprotein with pico to femtomolar concentration, as shown in Fig. 5.10b . Two channels with Ag/Au nanoparticles and proteins merge into a small round-shaped reaction chamber to generate SERS -active nanoparticle aggregates via protein-assisted cross-linking. Notably, the porous feature of the microchannels regulates the extent of nanoparticle aggregation and prevents the screening of “hot spots” in large aggregates.
SERS protein biotags (SBTs) have been developed for identification of cells and pathogens in body fluids, which is significant for early disease detection and monitoring of patient response to therapy . In the microfluidic cell identification system, two spectroscopically distinguishable SBTs that target distinct cell epitopes were used to label cells in a single focused line by hydrodynamic flow, as illustrated in Fig. 5.10c . The identification was accomplished by measuring the relative signal from the cancer-specific SBT versus the cell-identifying universal control SBT. The discrimination efficiency can reach 1 cancer cell from a population of 100 normal cells due to spectroscopic richness of the Raman bands of the reporter molecules on the two SBTs and algorithmic effectiveness of the two deconvolution strategies for the composite spectrum. The SBT approach provides a continuous, low-cost, and nondestructive tumor cell identification and paves the way toward clinical cancer diagnosis at the single-cell level.
5.2.3 Detection of Pathogen s
The detection of infectious agents such as viruses and bacteria is critical for public health, homeland security, and food industry [113, 115, 117]. Viruses have been responsible for a number of epidemic outbreaks (e.g., H1N1 flu and SARS) in recent years. Timely virus detection becomes essential for recognizing and controlling future epidemics. Conventional techniques such as cell culture methods, PCR, and ELISA can detect and quantify pathogens with high sensitivity and specificity [118, 119]. However, they require lab-intensive procedures, expensive equipment, and well-trained operators. Alternative biosensing techniques that can achieve reliable, accurate, and sensitive detection and analysis of pathogens under the variable settings, including source-limited settings and primary care settings, are in great need. Among the various detection platforms that use different mechanisms such as electrical, mechanical, and optical signal transduction [2, 5, 120], plasmofluidic biosensors are promising for label-free detection of infectious agents.
Wang et al. achieved label-free imaging, detection, and mass/size measurement of single viral particles with high-resolution surface plasmon resonance spectroscopy . The viral particles were imaged as diffraction patterns from their scattering of SPP . The particle size and mass were determined from the image intensities. Two viruses, i.e., H1N1 influenza A/PR/8/34 and HCMV, were studied with a mass detection limit of 1 ag, which is four orders of magnitude lower than that by conventional surface plasmon resonance method.
Inci et al. developed a plasmofluidic platform for label-free, high-sensitive, high-specific, and reproducible HIV viral load quantification from unprocessed whole blood sample . Surface chemistry and highly specific antibody immobilization were applied to capture the viruses at the electromagnetic “hot spots” and to minimize background signal from nonspecific binding of blood cells. Briefly, poly-L-lysine-modified polystyrene with terminal amine group was used as substrates to capture gold nanoparticles. The gold nanoparticles were treated with different chemicals consecutively, including 11-mercaptoundecanoic acid, N-ethyl-N-(3-dimethylaminopropyl)carbodiimide hydrochloride, N-hydroxysulfosuccinimide, NeutrAvidin, and biotinylated anti-gp120 polyclonal antibody. The antibody was positioned in a favorable orientation to have a high capture efficiency for the viruses. A SPP platform based on DNA-RNA hybridization was also developed to detect viruses plaguing plant food and feed crop plants .
Rapid detection of bacteria makes a difference in monitoring and preventing outbreaks of infections [2, 118]. For this purpose, Walter et al. developed microfluidic SERS to discriminate bacteria based on strain levels in a fast and reliable fashion . In their plasmofluidic system, aqueous analytes and colloid solution form droplets in mineral oil, creating a segmented flow for stable optical measurements and reproducible bacterial identification, as shown in Fig. 5.11b. Besides colloidal nanoparticles, solid-state SERS substrate functionalized with recognition molecules was also employed to capture and analyze E. coli cells with high specificity and sensitivity .
Demirci’s group developed a portable microfluidic SPP platform that can rapidly detect and quantify E. coli and S. aureus . All the elements, including light source, optical components, CMOS sensor, circuitry, and microfluidic chip, were packaged in a portable box with dimensions of 13.5 cm × 10 cm × 5.2 cm, as shown in Fig. 5.11c. Effective capture and detection of E. coli at concentrations from ~105 to 3.2 × 107 CFUs/mL in PBS and peritoneal dialysis fluid were demonstrated. Multiplicity and specificity were tested with S. aureus in PBS solution, implying the potential of the platform for pathogen diagnostics at both POC and primary care settings.
5.2.4 Detection of Drug s
Quick and cost-effective pharmaceutical analysis can benefit healthcare, biomedicine, and food safety . For example, drug screening is a critical step in new drug discovery. Detection of drug residues in food and human liquids is frequently needed for food safety and healthcare. The capability of monitoring antibody-polypeptide reaction is desirable for multiplexed drug discovery. Chen et al. proposed a SPRI-based visualization method for monitoring antibody-polypeptide binding in a label-free and high-throughput format . The plasmonic platform can support the dynamic imaging of polypeptide microarray and the time tracing of 3D histograms for monitoring antibody-polypeptide reaction on the surface.
Chiral compounds have significant impacts on pharmacological and biological processes. Effective strategies for chiral discrimination and enantioseparation are highly needed to meet the increasing need for enantiomerically pure compounds. For a pair of enantiomers, it is possible that one configuration is active drug, while the other configuration remains inactive, contributes to side effects, displays toxicity, or works as an antagonist [126, 127]. As an example, Guo et al. demonstrated α-thrombin-functionalized LSPR sensor integrated with a microfluidic chip for enantioselective analysis of melagatran. Recently, we have achieved high-sensitive label-free chiral sensing for drug molecules based on moiré chiral metamaterials .
Plasmofluidic sensors were also developed for label-free and rapid detection of drug residues in food and clinical samples . Fernandez et al. used a portable six-channel SPP sensor to analyze antibiotic residues from different families, i.e., fluoroquinolones, sulfonamides, and phenicols, in whole milk samples . The sensor showed good regenerative ability and repeatability and accomplished LOD s far below the maximum residue levels established by the European Union for the antibiotics. Zhao et al. used a multichannel SPP sensor to analyze the concentrations of methotrexate, an anticancer drug, in the serum of a patient undergoing chemotherapy treatments. The results agreed with those by fluorescence polarization immunoassay .
Andreou et al. developed a microfluidic SERS to identify drugs at clinical levels without specialized chemicals and bio-reagents . As shown in Fig. 5.11d, a laminar flow consisted of a central stream of analytes (i.e., methamphetamine) and two sheath streams of silver nanoparticle and salt solutions was created. At the interrogation region, SERS -active nanoparticle dimers and small-order aggregates with methamphetamine predominantly formed. Due to the low affinity of methamphetamine to the silver nanoparticles, the salt was added to induce the aggregation. Trace concentrations of methamphetamine in saliva were detected within minutes. The microfluidic SERS will find applications in detecting many other health-related molecules such as toxins and pollutants.
6 Conclusions and Future Perspectives
Tremendous progress has been made in plasmofluidics for biological analysis and medical diagnostics. In particular, plasmonic tweezers achieved noninvasive trapping of biological cells, DNAs, and proteins at the single-entity level. 3D manipulations were demonstrated with plasmonic tweezers based on an optical fiber tip with high-resolution mechanical motion control. Long-range and rapid delivery of nanoparticles was achieved with electrothermoplasmonic s . Most of the current tweezers used microfluidic channels as passive containers for sample solutions. It is anticipated that future plasmonic tweezers will fully reap the benefits of micro-/nanofluidics to further enhance their functionalities for biosensing and diagnostic applications. In particular, the light-fluid interactions at the nanoscale can be explored to achieve versatile multifunctional plasmofluidic tweezers . Our preliminary study indicated that the plasmon-enhanced photothermal effects such as thermophoresis are one of the promising strategies for optical manipulations in variable fluidic environments.
A variety of plasmofluidic sensors has been developed to detect and analyze nucleic acids, proteins, pathogens, and drugs in life sciences, disease diagnostics, and drug discovery. Innovative microfluidic design, plasmonic engineering, and surface functionalization have led to sensitive, robust, and label-free POC biosensors. The next-generation plasmofluidic sensors are expected to target at translational and clinical applications . For the purpose, the sensors need to be a highly integrated, robust, and user-friendly system that can handle a variety of clinical samples such as urine, blood, and saliva and enables POC diagnostics at field settings.
In summary, the highly interdisciplinary field of plasmofluidics presents exciting opportunities for new discoveries and POC devices. With the stronger collaborations among researchers from physics, chemistry, biology, pharmacology, and engineering, as well as clinicians and entrepreneurs, plasmofluidics for biosensing and disease diagnostics will continue to grow and contribute to global healthcare.
The authors acknowledge the financial support of the Beckman Young Investigator Program and the Office of Naval Research Young Investigator Program.
- 6.Han KN, Li CA, Seong GH (2013) Microfluidic chips for immunoassays. In: Cooks RG, Pemberton JE (eds) Annual review of analytical chemistry, vol 6. Annual Reviews, Palo Alto, pp 119–141Google Scholar
- 14.Huang JS, Yang YT (2015) Origin and future of plasmonic optical tweezers. Nano 5(2):1048–1065Google Scholar
- 85.Masson J-F, Zhao SS (2015) Plasmonic sensors for analysis of proteins and an oncologic drug in human serum. In: Vestergaard MDC et al (eds) Nanobiosensors and nanobioanalyses. Springer Japan, Tokyo, pp 305–333Google Scholar
- 118.Wang S et al (2012) Portable microfluidic chip for detection of Escherichia coli in produce and blood. Int J Nanomedicine 7:2591–2600Google Scholar
- 119.Durmus NG et al (2013) Fructose enhanced reduction of bacterial growth on nanorough surfaces. MRS Proc 1498:73–78Google Scholar
- 128.Wu Z, Zheng YB (2017) Moiré chiral metamaterials. Adv Opt Mater 5. (in Press)Google Scholar