Label-free Photonics Biosensor Transducing Nano-Biological Events

HISTORY Received: 12th of April 2014 Received in revised form: 14th of June 2014 Accepted: 21st of June 2014 Available online: 3rd of July 2014


Introduction
Nearly two decades has passed since the first draft of the human genome had been published in June 2000 that leaves a remarkable milestone in the history of science [1,2].Since then, the genome of many other living creatures has been successfully read, and numerous information about the genes can be extracted and extrapolated in order to understand biological processes.One of the most powerful analytical tools to be addressed in these fields is the DNA microarray technology, which can analyze mRNA transcript levels expressed under various conditions [3][4][5][6].However, it is known that the mRNA expression level and the corresponding protein abundances does not always correlate with each other.This is due to the changes in translation rates and protein lifetimes [7,8].Besides, researchers have no clue in predicting protein activities from the genomic information collected.This is because, the analysis of mRNA expression, as well as DNA sequence, does not provide any information regarding to protein functions, their interactions, activities, three dimensional structures and post-translational modifications such as proteolysis, phosphorylation, glycosylation, methylation, and alkylation.Therefore, posttranslational produced proteins need to be analyzed directly in order to obtain such information.Such challenges are now being classified as -omics studies such as transcriptomics, proteomics, and metabolomics.
Studying protein characteristics and their molecular interactions can provide an important route to investigate protein networks in living cell.It can also postulate the functions of newly discovered genes or proteins, thus, holds great value for understanding disease mechanisms [9][10][11] and provides suitable diagnostic for global-threatening diseases [12,13] (Figure 1).To date, works have vastly being done in these areas (proteomics studies) in order to develop simple and noninvasive test that can indicate disease risk at early stage [14].One of the keys that plays important role in these research areas is nanobio-sensor technology.This technology provides analytical tools upon knowing the abundance, and also the qualitative characterization of a particular biomolecule.Therefore, the development of biosensor technology is crucial, and is expected to meet the needs of analyzing protein behaviors (interactions, activities, etc.) accurately and effectively.

Nanobio-sensor technology
A biosensor is a device that designed to detect or quantify a nano-bio-chemical molecule such as metabolites, nucleic acids, protein, viruses and nanoparticles.It is a self-contained integrated device [15] that can provide quantitative or qualitative analytical information, in which the biorecognition agent interacts with a target at interfaces thus conveying the information to some physical transduction elements.In biological research involving protein interactions, many biosensors used are solid-supported and affinity-based, in which surface-immobilized capture agents are used to capture targeted analyte selectively and induce some physical changes at localized surface.This change can be made in a variety of ways because the presence of target protein has a variety of physical significance.For example, nano-electrical biosensors detect electrical charge changes at the localized surface, thus relying on the measurement of current or voltage to detect the bindings or interactions.Others would be subjected to light absorbance / transmittance changes, surface tension changes, and so on.
There are two main purposes of using a nanobio-sensor [16,17], i) To detect or quantify biomolecules abundance.Jeewon Lee et al. reported the use of quantum dots along with virus particles for detecting ultralow levels (unit) of the protein marker relevant to heartattack patients [18].This allows the detection of the abundance of a specific biomarker that would lead to further disease diagnosis.Rica and Stevens used gold nano particles that grew on ELISA technique to make cancer marker at the level of 10 - 18 M visible to the naked eye [19].These are the examples of quantitative analytical information. ii) To study the characteristics and properties of a particular biomolecule (protein) such as its kinetic behaviour, physical and chemical characteristic, reactions capability, and interactions networking.
Igor Sokolov et al. [20] has successfully studied the differences of cilia (cell surface brushes) behaviour in normal and cancerous cell using atomic force microscopy (AFM) technique.They found that the cilia of cancerous cell is somewhat differs in length and concentrations in compared to the normal one.This would leads to the characterization of some specific behaviour of a protein on cell surface to differentiate normal cell from the cancerous one.The later one is qualitative analytical information.Thus, depending on these specific purposes, various types of biosensors are being utilized in these two main avenues of nanobiological research.
It is important to realize that the advances in biosensor technologies for in vitro diagnostics does have the potential to transform the practice of biological science, medicine and bioterrorism defense.To date, viruses [21][22][23][24], disease biomarkers [25][26][27], cancer related gene [28], bacteria [29,30], dangerous substances/chemicals [31,32], and protein networks [33] have been studied or confirmed thanks to the development or various biosensor technologies.Figure 2 depicts the numbers of scientific literatures and articles published during a period from 1982 to 2013, suggesting that a lot of progress in biosensor field have been made.Despite rapid progress in the biosensor field, there is still no general sensing platform that can be ubiquitously applied to clinically detect the constellation of biomolecules in variety type of samples (saliva, urine, serum, or cell lysates) with high sensitivity, large linear dynamic range, rapid, low cost and using less sophisticated devices.Major limitations such as; i) unsatisfied sensitivity, ii) non-native signals interference iii) using too sophisticated devices (unpractical for certain conditions), iv) sensor delicateness (causing errors, high signal-to-noise ratio and doubtful signal), v) small or non-linear dynamic range, and vi) time consuming, forced researchers to keep improving fundamental concepts or exploratory rooms of biosensors by manipulating basic principles or combining the field of studies (e.g.merging of biosensor and electronics [34] or material science [35]).

Label-free detection
Methods in detecting molecular interactions have been studied since 1960's when radioisotope label has been rapidly used to quantitatively determine molecular reactions [36,37].The system is sensitive, simple, yet bringing about some safety concern due to the usage of radioisotope material.Since then, many label system (other than radioisotope label) that detects molecular interactions have been developed such as Fluorescence/Fourier resonance energy transfer (FRET) [38,39], chemiluminescence [40,41], active enzymes with an easily detectable product, allowing facile target conjugation and convenient detection.These technologies are much safer [42], and being vastly used, producing a lot of sciences in molecular biology [43][44][45].
Thus, labeled type of detection e.g.with dye molecule or GFP [46] that chemically attached to target molecule can be found in most of the classical examples of biosensors.It can be sitespecifically targeted and shows clear signal output that can be easily recognized [47].During signal readout, the amount of label detected is expected to correspond to the amount of bound targets.However, several steps are required before the molecule can be put in the biosensing system.Target molecule and label molecule coupling reaction, and purification (with broad range of yield) are simply required.
In addition, labeling a biomolecule may change its intrinsic properties [48,49].
Changing structure of any proteins caused by label probe or labeling process was not commonly reported in the literatures of protein sciences.However, by labeling a target protein or ligand, its chemical properties (if not physical) would also change.These could lead to different protein behavior during interactions.These are relatively minor problems upon studying DNA interactions, but they are a bigger concern for protein interactions.
Until recent years, labeled detection systems are being mainly developed in searching for more simple and efficient chemistries (in terms of probe size and labeling process), and more sensitive detection.But now, focus has been shifted to the development of so called 'label-free detection system', which requires none of these labels [17,[50][51][52][53][54][55].Although there have been only a few studies comparing molecular interactions read-out between labeled and label-free detection system, it is generally accepted that molecular interactions with label probes are involving non-native interactions, and could possibly leads to false information [46,57,58].
Label-free detection systems utilize only native physical properties of a target molecule such as weight, refractive index, and molecular charges.Therefore, the detection classes is divided into three; i) nano-mechanical system that are sensitive to weigh of the molecule, ii) optical system that are sensitive to the refractive indices, and iii) electrical system that are sensitive to molecular charges [53].In a typical biosensing process, ligand is immobilized onto the sensor surface using chemical reaction or physical adsorption to the surface.Then, a protein solution will be allowed to interact with the surface-bound ligand.When interacts to the surface-immobilized ligand, the target protein will transfer its information in a form of mechanical, optical or electrical signals.These will be produced from the physical presence of the target proteins that changed the local environment of the sensor surface, thus, made the protein interactions detectable without any label probe.The main advantage for label-free detection over label detection is that more direct information can be acquired as the method only uses native proteins and ligands.Table 1 summarized the main advantages of the label-free detection system.
Table 1 Main advantages of label-free detection system compare to the detection with label probes.a This is a disadvantage of a labelfree detection system.

Types of detection Label-free Label
Signal

Three key technologies
In order to develop nanobio-sensor technology, the following three key technologies must be tied in with each other: i) capturing agents, ii) surface chemistry, and iii) detection systems (Figure 3) [59,60].Although many of capturing agent like peptides and proteins are commercially available, the other two key technologies (detection system and surface chemistry) are now in a rapid development.To date, many interesting approaches on both technologies to overcome problems and current limitations of the label-free methodology have been reported.

Detection system
The detection system used is divided into three classes, i) optical-based, ii) mechanical-based, and iii) electrical-based detection.While nano-mechanical-based methods and electrical-based detection methods are being reviewed elsewhere [61,62], herein we concluded several most used photonic-based methods (specifically using plasmonic approaches) in recent years.
Conceivably, the widespread label-free method in the detection of molecular interactions is either optical or photonic-based method.Several methodologies are have been already widely implemented and commercialized.Among these, the most prominent technique is the surface plasmon resonance (SPR) spectroscopy technique.Other specific methods are the localized plasmon using nanoparticle, surface plasmon using nanohole array [63], Raman scattering (especially with surface-enhanced mode), ellipsometry method [64], fabry-perot based technique using porous silicon [65][66][67], anomalous reflection (AR) of gold method [68], second harmonic generation (SHG) method [69,70], surface-immobilized gold nanospheres (SIGN) [71], and interference-based techniques for example using multilayered silicon surface [72].The light used therein is in all range from ultra violet (UV), visible light, and into midinfrared region.Molecular interactions are being monitored either by reflection, transmission or scattering (including dark-field scattering) of the light.The advantages of using optical methods over any other methods (electrical and mechanical) are that it is simple, fast, can be easily miniaturized, and the light does not influence the interactions that are being monitored.

Surface plasmon resonance (SPR) spectroscopy
The surface plasmons (SPs) are light waves that are trapped on the surface of a metal because of their resonance interaction with the free moving electrons oscillate on metallic surface.It is also known as surface plasmon polaritons (SPP), referring to the hybrid nature between light waves and free electron oscillation.It was first predicted by Ritchie in 1957, where plasma energy losses increases with the decreasing of metallic foil thickness [73].Since then, many scientists such as Raether, Kretschmann and Otto have extensively studied the surface phenomenon.
In the interaction between light wave (irradiated) and free electrons at metal surface, the electrons are responding collectively by oscillating at resonance wavelength.The resonant interaction between the surface charge oscillation and the electromagnetic field of the light constitutes the SP and gives rise to its unique properties.The wave vector (k SP ) of SPs is always larger than the wave vector of the incoming photon (k 0 ).Thus, the irradiated light cannot directly excites the SPs wave [74].In 1968, Otto introduced the idea of total internal reflection (TIR), where the SPs wave of a metal surface (thinfilm) can be excited with the help of a prism [75].The metal thinfilm has been put close enough to the interface of the prism surface and air so that the evanescent field produced by the TIR can reach the metal.
The wave vector of the evanescent wave along the interface direction is k = k 0 nsinα, which can be larger than k 0 (when n sinα> 1).Thus, the SPs wave can be excited at some certain point where k = k(ω) as illustrated (Figure 4).However, this experiment is not easy to repeat since it requires a very precise control of the thickness of the air gap, which is in the order of a couple hundred nm.In 1971, Kretschmann demonstrated a modified Otto geometry [76].The metal film was put in between the dielectric layer and air, and was kept thin enough (approx.50 nm) so that the concept of TIR at the interface would still working, while the experimental setup is less requiring.
Figure 4 The dispersion curve for a SP mode shows the momentum mismatch problem between illuminated light (blue line) and SP modes (orange line).This problem must be overcome in order to couple light and SP modes together, with the SP mode always lying beyond the light line, that is, it has greater momentum (kSP) than a free space photon (k0) of the same frequency (dotted line).One way to overcome the problem is by using evanescent wave produced in TIR, which k = k0 n sinα (red line).
The SPR technique was first used as a sensor in 1980's, where it was used to detect gas [77,78].Then, an imaging system of SPR (also known as SPRi or SPR microscopy) was invented in the late 1980's [79][80][81][82], and forged the 2diamensional SPR chips that we see today.In 1995, S. Kawata pioneered the concept of grating-coupled SPR, where there is no longer need for a prism and TIR to excite the SPs [83].Collective momentum of illuminated light produced by the periodic grating seems to match that of the SPs of metal.This concept allowed direct reflection method of SPR, thus the idea of high-throughput study using imaging system from a CCD camera device could be easily implemented [84][85][86][87][88].The research progress of SPR is numerous and unbeatable by any other detection method [89][90][91][92].The widespread of Biacore® instrument of SPR devices made the SPR one of the most reliable tools and sources of studying protein interactions especially in thermodynamics and kinetics analyses.Remarkable achievements of SPR analytical tools includes a peny-sized 2-dimensional microfluidic chip format detection [93,94], and the detection of small molecules like arsenic and calcium ions by manipulating molecular conformational changes [95,96], reflects its preeminence.

Localized plasmon resonance (LPR) of nanoparticle
The use of nanoparticle goes back to the time of the Romans (4th century AD), when it was used as decorative pigments in the glass of the famous Lycurgus Cup.Analysis of the glass reveals that it contains a small amount of metal crystals (~70 nm) containing Ag and Au.However, it was not until 1857, when Michael Faraday announced a systematic study of the synthesis and colors of colloidal gold [100], that nanoparticle-based phenomenon began to draw attention.In 2000, Richard Van Duyne invented the concept of a "localized" SPR [101].Since the pioneering work, thousands of scientific studies have been published that leads to today's understanding of the so-called localized plasmon resonance (LPR) of nanoparticle.
LPR is basically the same plasmon resonance wave as SPR.
The only different is that LPR wave is localized around the metallic nanoparticle, rather than propagates along the metal surface like SPR.Advances in nano-fabrication techniques allow researchers to tune the LPR wavelength throughout the visible, near-infrared, and into the Mid infrared region of the EM spectrum, by varying the shape, size, and material of the nanoparticles that support the surface plasmons (Figure 5) [102][103][104][105][106].This offers additional flexibility when designing LPR biosensing experiments [107].LPR biosensing is usually observed by the wavelength-shift mode.For sphere nanoparticle geometries, the simplest theoretical approach available for modeling the optical properties is the Mie theory [108].For other geometries, different approaches have been devised, such as the theoretical calculation derived by Gans [109], which has been used for nanorod geometry [110], and discrete dipole approximation (DDA) method [111], which has been applied to the calculation of nanoprisms [112,113].
The LPR biosensing is a rapidly growing field.The future of this field looks exceptionally bright and promising with awaited wide range of applications from detections to rapid disease diagnostics.The lower fabrication-cost of LPR biosensor compared to SPR biosensor has made it a good candidate for commercialization.However, some limitations have bottlenecked the progress of LPR biosensor, such as i) the difficulties to fabricate a uniform structure of nanoparticle, ii) comparatively shorter linear dynamic range, and iii) immobilization of nanoparticle to a solid surface.More accurate nanomaterial fabrication with desirable design, and the preparation of an LPR-based biochip for a quantitative and rapid screening are the current challenges for this technology.

Conclusion
Plasmonic methods applied in nanobio-sensors are very sensitive to changes at molecular level.Manipulation of the plasmonic principles, combine with other techniques (such as nano-fabrication and bio-techniques) has created many versatile approaches that are able to detect the presence of biomolecule, or characterize the molecule of interest qualitatively.It has been proven that plasmonic methods are able to perform as sensing platform at ultrasensitive (<10 -12 M) level of detection.Paradigm-shifting thoughts are still needed, however, to made it practical as an on-field technique that can be used as point-of-care diagnostics world widely.These include massive cost reduction and the usage of non-sophisticated machinery [138].
Despite the world-widely usage of SPR and LPR techniques, many other optical-based nanobio-sensors have been invented and progressively being re-invented from time to time.Anomalous reflections (AR) of gold technique [139], surface-immobilized gold nano-sphere (SIGN) [71,140], simple reflection on alloy and composite thin layer surfaces [141], metal-insulator-metal platforms [142,143], gold nanoparticle growth ELISA [19] etc., are just a handful examples of the exponentially growing development.It is hoped that the nanobio-sensing technology can be benefited to relief global-threatening diseases, out breaks, and bioterrorism violence to a larger population of the world, and thus contributes to a better social structure of today's modern living.

Figure 1 .
Figure 1.Global examples of emerging and re-emerging infectious diseases.Red represents newly emerging diseases; blue, reemerging/resurging diseases.Figure was taken from literature 12 and republished with permission.

Figure 2 .
Figure 2. Number of scientific publications for label and label-free biosensor from 1982-2013 (four years period each).The survey was conducted using the Google Scholar search engine.

Figure 3 .
Figure 3. Three key technologies of developing biosensor; i) Capturing agents (CA), ii) surface chemistry, and iii) detection system.CA = capturing agent, CA-TP = capturing agents-target protein.

Figure 5 .
Figure 5. Range of plasmon resonances for a variety of nanoparticle morphologies.Republished with permission of Nature Publishing Group, from Nano-optics from sensing to waveguiding, Surbhi Lal et.al., 1, 2007; permission conveyed through Copyright Clearance Center,