Optical fibre-based detection of DNA hybridization - Semantic Scholar

2 downloads 0 Views 266KB Size Report
... Chen†, Marcus D. Hughes*, Kaiming Zhou†, Edward Davies†, Kate Sugden†, ..... 22 Yang Li, Y., Chun, E., Liu, Y., Pickett, N., Skabara, P.J., Cummins, S.S.,.
Advances in Nucleic Acid Detection and Quantification

Optical fibre-based detection of DNA hybridization Anna V. Hine*1 , Xianfeng Chen†, Marcus D. Hughes*, Kaiming Zhou†, Edward Davies†, Kate Sugden†, Ian Bennion† and Lin Zhang† *School of Life and Health Sciences, Aston University, Aston Triangle, Birmingham B4 7ET, U.K., and †School of Engineering and Applied Science, Aston University, Aston Triangle, Birmingham B4 7ET, U.K.

Abstract A dual-peak LPFG (long-period fibre grating), inscribed in an optical fibre, has been employed to sense DNA hybridization in real time, over a 1 h period. One strand of the DNA was immobilized on the fibre, while the other was free in solution. After hybridization, the fibre was stripped and repeated detection of hybridization was achieved, so demonstrating reusability of the device. Neither strand of DNA was fluorescently or otherwise labelled. The present paper will provide an overview of our early-stage experimental data and methodology, examine the potential of fibre gratings for use as biosensors to monitor both nucleic acid and other biomolecular interactions and then give a summary of the theory and fabrication of fibre gratings from a biological standpoint. Finally, the potential of improving signal strength and possible future directions of fibre grating biosensors will be addressed.

Introduction A brief overview of optical fibres for non-physicists The history of optical fibres as we understand them today dates from the mid-20th Century [1]. Optical fibres are often called waveguides, because the light travels down them and emerges only at the far end, rather than spilling out through the sides of the fibre. The central part, down which the light travels, is called the core. Instead of travelling straight down the middle of the core (as might be imagined), the light enters at an angle (it is refracted from the air) and when it reaches the edge (boundary) of the core, it bounces off, and continues down the fibre. For light to propagate down the waveguide or optical fibre, it must hit the boundary above a certain angle (with respect to a normal from the boundary). The smallest angle at which light can be propagated is called the critical angle. Below this angle, light simply leaves the core as soon as it encounters the core boundary (it is lost by refraction), but at or above the critical angle, the light is all reflected at the boundary. This is called total internal reflection and is the basis on which optical fibres function. In order for total internal reflection to work, the refractive index of the core must be greater than that of the surrounding medium (e.g. glass core and air surround). Although light can travel down a core that is surrounded by air, the cores of practical optical fibres are surrounded by another medium (the cladding) whose refractive index is less than but very close to that of the core. This similarity in refractive indices effectively limits Key words: biosensor, DNA hybridization, evanescent wave, long-period fibre grating (LPFG), optical fibre, unlabelled biomolecule. Abbreviations used: FBG, fibre Bragg grating; GFP, green fluorescent protein; LPFG, long-period fibre grating; ssDNA, single-stranded DNA. 1 To whom correspondence should be addressed (email [email protected]).

Biochem. Soc. Trans. (2009) 37, 445–449; doi:10.1042/BST0370445

the angles of light that can be propagated down the fibre, since the smaller the difference in refractive index between the cladding and the core, the smaller the angle of light that can be input into the fibre (the numerical aperture). Since light is an electromagnetic wave, it also has electrical properties and these cause an ‘electrical disturbance’ to run alongside the light that is totally internally reflected within the core of an optical fibre. That ‘electrical disturbance’ is called an evanescent wave and it propagates within the cladding, very close to the boundary with the core of the optical fibre. The magnitude of the evanescent wave diminishes reciprocally with distance from the core/cladding boundary. The evanescent wave is often conceptualized as a small proportion of the light penetrating into the cladding, during reflection. It is an evanescent wave that has been exploited with the current application, to sense the interaction between unlabelled biomolecules.

Optical fibres as light conduits within sensors and biosensors Optical fibres are best known for their applications in the telecommunications industry, where multiple signals (in the form of light) can be sent down a single fibre over very long distances. They are also used extensively in devices for viewing hard-to-reach locations, such as the endoscope. It is this property of using the optical fibre to conduct light to a remote, convenient location that is most frequently exploited in optical fibre biosensors. Thus most optical fibre biosensors require fluorescent or colorimetric labelling of one or more biomolecules (e.g. [2–7]). Once the labelled biomolecule has interacted with its target, the fluor is excited by the evanescent wave and the resulting fluorescence [either directly from the fluor or from a FRET (fluorescence resonance energy transfer)-based signal] is then conducted via  C The

C 2009 Biochemical Society Authors Journal compilation 

445

446

Biochemical Society Transactions (2009) Volume 37, part 2

the fibre to an analytical device that has been equipped with appropriate wavelength filters. This area is the subject of an extensive, recent review [8].

Optical fibre sensors that employ gratings Ideally, biosensors would not require labelled molecules. Sensing the interaction of unlabelled molecules (akin to detection of unlabelled biomolecular interactions by surface plasmon resonance) can also be achieved using optical fibres and is an emerging subdiscipline of photonics (for a recent review, see [9]). Such sensors require gratings, which are tiny repetitive structures (akin to diffraction gratings), inscribed through the side of an optical fibre. There are two types of grating that relate to optical fibres: FBGs (fibre Bragg gratings), which are used to sense strain and temperature, and LPFGs (long-period fibre gratings), which exploit an evanescent wave and can be employed as chemical sensors or biosensors. The two types of grating differ in the distance (period) between the component ‘lines’ of the grating and are manufactured differently. Both types of grating are inscribed by shining continuous UV laser (244 nm) light through the side of an optical fibre. The cladding layer is transparent to the UV light which passes straight through, but when the light enters the core of the fibre, the refractive index of the glass it hits is changed. Thus, if a fibre grating is viewed under a microscope, it appears as a series of dark lines. These dark lines are the regions of glass that have a different refractive index. The two types of grating (FBGs and LPFGs) are considerably different in size. An FBG has a period of the order of 1 μm and is inscribed using interference patterns of light, whereas an LPFG is literally printed on to the core of the fibre, by shining the laser through a template mask. LPFGs typically have periods in the region of 100– 500 μm. As light passes down the core of a fibre and encounters a grating (FBG or LPFG), certain wavelengths of the light are either reflected back up the core (FBGs; single wavelength) or pass through the grating and into the cladding (LPFGs; multiple wavelengths, Figures 1A and 1B). This process (the effective loss of certain wavelengths) is called coupling and is measured by looking at the transmission spectrum of light that exits from the core of the optical fibre. Anything that changes the periodicity of either type of grating changes the wavelength(s) of the coupled light. This phenomenon can be exploited in, for example, a remote temperature sensor such as that in an aeroplane engine [10], where a localized temperature change at the site of the grating makes the grating either expand or contract. This changes the periodicity of the grating, which in turn alters the wavelength of the coupled light that is detected in transmission at the end of the optical fibre. Similarly, mechanical strain can either compress or expand the grating, again altering the transmission spectrum of the grating. Thus fibre optic grating-based sensors have also been used to measure bending forces in yacht masts [11], bridges [12] etc. These types of sensors (temperature and strain) employ FBGs. In principle, LPFGs operate similarly. However, the factors that change the wavelength of light coupled in an  C The

C 2009 Biochemical Society Authors Journal compilation 

Figure 1 Simple representation of the principles of LPFGs and their effect on transmitted light (A) Diagram of an optical fibre showing light propagating down the core of the fibre by total internal reflection. On encountering the LPFG, certain wavelengths of that light are coupled into the cladding and are thus lost from the fibre core. Note that although light travelling down the core will set up an evanescent wave in the cladding (not shown), light coupled into the cladding will also set up an evanescent wave just outside of the cladding. It is this wave that is exploited to sense biomolecular interactions. (B) Typical transmission spectrum (wavelengths that emerge from the fibre core), after light has passed through an LPFG. Peaks represent wavelengths that are missing; the light of these wavelengths has been coupled into the cladding.

LPFG are more complex. Measurements employing LPFGs additionally exploit an external evanescent wave and thus the factors that affect the transmission spectrum of an LPFG are 4-fold: (i) the grating period, (ii) the core refractive index, (iii) the cladding structure (including its refractive index and shape), and (iv) the properties of the surrounding medium. It is this last property that is exploited to make concentration sensors, chemical sensors and biosensors from optical fibres. When light is coupled with the cladding, it itself sets up an evanescent wave just outside of the cladding and it is this evanescent wave (Figure 1A) that is affected by changes in concentration/composition of the surrounding medium or by molecular interactions occurring on the surface of the cladding.

LPFG construction LPFGs, with comparatively short periods of 161 μm, were inscribed into single mode, H2 -loaded SMF-28 optical fibres (Corning) as described in [13]. Inscription generally takes of the order of seconds to minutes (depending on the composition of the core) and it is possible to monitor light transmitted down the fibre during the inscription process. Thus it is possible to generate a grating that gives almost 100% coupling of specific wavelengths of light into the cladding (Figure 1 of [13]; note that the transmission scale, dB, is a logarithmic scale and thus 99% coupling ≈ −20 dB

Advances in Nucleic Acid Detection and Quantification

transmission). To increase sensitivity, the grating was then gently etched [14] with 15% HF for approx. 10 min to remove some of the cladding in the area of the grating. It is important to note that this etching process is relatively gentle and thus although the thickness of the cladding is reduced by etching, the optical fibre itself is not damaged by the etching process.

Figure 2 Chemical derivatization to cross-link biomolecules to the surface of the fibre cladding (A–D) Chemical derivatization of the glass surface as described in the ‘Sensor molecule immobilization’ section (full experimental details are described in Chen et al. [14]); (E) fluorescent micrograph of a derivatized optical fibre, cross-linked with GFP.

Sensor molecule immobilization To generate a biosensor, it is necessary to immobilize one of the two interacting molecules in the surface of the sensor. In our preliminary study, the whole outer surface of the fibre in the region of the grating was derivatized first by silanization. Both the core and the cladding SMF-28 fibre are made of glass and thus the outer surface of the fibre can be derivatized by conventional silanization in which the hydroxy groups exposed on the outer surface of the cladding (Figure 2A) are made to react with (3-aminopropyl)triethoxysilane (Aldrich) to yield primary amine surface groups [15] (Figure 2B). Subsequent reaction with dimethyl suberimidate (Aldrich) results in a glass surface from which chemical linkers project that are terminated with imidoester groups [16] (Figure 2C). These reactive groups can then react with primary amines within proteins (i.e. the N-terminus or side groups of exposed lysine residues) or N-terminated oligonucleotides [17], to covalently link biomolecules to the surface of the glass (Figure 2D). In the first instance, GFP (green fluorescent protein) was cross-linked to the fibre surface to check whether its chemical derivatization had been successful. The derivatized fibre was compared with a non-cross-linked fibre under a fluorescent microscope. No fluorescence was observed from the noncross-linked fibre (results not shown), but Figure 2(E) demonstrates that the cross-linking with GFP had indeed been successful. Therefore a duplicate chemically derivatized fibre, containing a dual-peak LPFG with a period of 161 nm, was similarly cross-linked with an amino-modified target DNA. This ssDNA (single-stranded DNA)-derivatized fibre was used in subsequent experiments.

Detection of DNA binding and reuse of the sensor Unsurprisingly, the act of cross-linking the sensor DNA molecule to the surface of the LPFG modified the wavelength of the coupled light. Thus attachment of the sensor molecule caused a blue shift of the coupled light of 254 pm (Figure 3A; note the leftward displacement of the transmission spectrum after immobilization). The fibre was then equilibrated in hybridization buffer [6 × SSPE; 1 × SSPE = 0.15 M NaCl/ 10 mM sodium phosphate (pH 7.4)/1 mM EDTA] and then an unlabelled, fully complementary 15-mer DNA oligonucleotide (1 μM) was hybridized to the biosensor at room temperature. The hybridization of the two DNA strands can be visualized by the red shift of the coupled light with time, from 1591.91 to 1592.63 nm: a total red shift of 715 pm over the 1 h hybridization (Figure 3B). Stripping of the hybridized

DNA again caused a blue shift of the coupled light, of 1257 pm (Figure 3C, note the leftward displacement of the bold curve), whereas after washing, drying and equilibration in hybridization buffer, a second DNA hybridization (2 μM complementary DNA) once again red-shifted the coupled light, by 1165 pm, from a starting wavelength of 1590.88 nm to 1592.05 nm (Figure 3D). It is encouraging to note that a doubling of the test DNA concentration from 1 to 2 μM led to a near doubling (78%) of the sensor’s response on reuse. From Figures 3(B) and 3(D), it can be seen that an initial period of rapid hybridization (approx. 3 min in each case) was followed by a prolonged period of more gradual hybridization over the time course of 1 h  C The

C 2009 Biochemical Society Authors Journal compilation 

447

448

Biochemical Society Transactions (2009) Volume 37, part 2

Figure 3 Effect of DNA immobilization, hybridization, stripping and rehybridization on the wavelength of light coupled in an etched, long-period FBG, of a period of 161 μm c 2007 Optical Society of America), where full experimental procedures Reproduced, with permission, from Chen et al. [13] ( have been described. (A) Transmission spectrum of the fibre, measured before (fine line) and after (bold line) immobilization of 1 μM ssDNA (5 -GCACAGTCAGTCGCC-NH2 -3 ) in PBS buffer for 16 h. (B) Change in the wavelength of coupled light (the peak wavelength of transmission loss) with time, as 1 μM of complementary ssDNA (5 -GGCGACTGACTGTGC-3 ) hybridizes in 6 × SSPE buffer to the DNA immobilized on the FBG. (C) Transmission spectrum of the fibre, measured before (fine line) and after (bold line) stripping of the hybridized DNA in 5 mM Na2 HPO4 /0.1% (w/v) SDS for 3 × 30 s at 95◦ C. (D) Reuse of the sensor; change in the wavelength of coupled light with time, as 2 μM of ssDNA (5 -GGCGACTGACTGTGC-3 ) hybridizes in 6 × SSPE buffer to the DNA immobilized on the FBG.

(sensed as a red shift of 9 pm/min for 1 μM of complementary DNA and 16 pm/min for 2 μM of DNA; again an increase in signal of 78% for a doubling of DNA concentration). However, the initial rates of hybridization may be of most interest in a functional biosensor, and here again, it is encouraging to note that a doubling of the DNA concentration led to an increase in the red shift from 86 to 119 pm/min (a 38% increase) over this initial, rapid part of the hybridization curve.

Why are grating-based biosensors important? Aside from the obvious convenience of employing unlabelled biomolecules, we believe that LPFG-based biosensors may ultimately offer greater sensitivity than fluorescence-based sensors. Fluors necessarily give off light in all directions, and of this, only a tiny proportion will ever enter the fibre, to travel back to the detector. In contrast, LPFGs rely only on  C The

C 2009 Biochemical Society Authors Journal compilation 

the effect of the biomolecule on the evanescent wave, which in turn affects the wavelength of coupled light within the fibre. Thus the fact that biomolecular interactions take place on the fibre surface becomes unimportant.

Future improvements in sensitivity Clearly, our work with LPFGs is in its early stages and although we have made initial attempts to maximize sensitivity by employing dual-peak LPFGs and by etching away some of the fibre’s cladding, there remain many avenues to explore. Sample volumes can be reduced relatively easily, by designing tiny vessels that hold the test biomolecule only in the region of the grating, rather than bathing the whole fibre in the sample. The ultimate extension of such work would be to ablate microfluidic channels [18] within the cladding to achieve absolute minimization. Likewise, we have yet to explore any sensitivity gains offered by D-shaped fibres [19], or by refining the etching processes [14,20].

Advances in Nucleic Acid Detection and Quantification

Future directions We have yet to examine whether the sensitivity is sufficient to detect mismatches in oligonucleotide hybridization. Likewise, particularly because the biomolecules are unlabelled, we are interested in examining the potential of investigating protein–protein, protein–DNA and protein–substrate interactions. Moreover, would an LPFG-based sensor be useful in screening combinatorial protein libraries [21] or in general proteomic applications? Finally, it will be interesting to determine whether we can combine the best of both worlds by embedding fluors such as quantum dots [22] within the optical fibres.

Funding We gratefully acknowledge the Engineering and Physical Sciences Research Council for support [grant number EP/D500427/I to L.Z., I.B, K.S. and A.V.H.].

References 1 Hecht, J. (2004) City of Light: the Story of Fiber Optics, pp, xi–x, Oxford University Press, Oxford 2 Ko, S. and Grant, S.A. (2006) A novel FRET-based optical fiber biosensor for rapid detection of Salmonella typhimurium. Biosens. Bioelectron. 21, 1283–1290 3 Li, Z.H., Hayman, R.B. and Walt, D.R. (2008) Detection of single-molecule DNA hybridization using enzymatic amplification in an array of femtoliter-sized reaction vessels. J. Am. Chem. Soc. 130, 12622–12623 4 Niu, S.Y., Wang, S.J., Shi, C. and Zhang, S.S. (2008) Studies on the fluorescence fiber-optic DNA biosensor using p-hydroxyphenylimidazo[f]1,10-phenanthroline ferrum(III) as indicator. J. Fluoresc. 18, 227–235 5 Leung, A., Shankar, P.M. and Mutharasan, R. (2008) Model protein detection using antibody-immobilized tapered fiber optic biosensors (TFOBS) in a flow cell at 1310 nm and 1550 nm. Sensors Actuators B 129, 716–725 6 Wei, H., Guo, Z., Zhu, Z., Tan, Y., Du, Z. and Yang, R. (2007) Sensitive detection of antibody against antigen F1 of Yersinia pestis by an antigen sandwich method using a portable fiber optic biosensor. Sensors Actuators B 127, 525–530 7 Chou, C., Hsu, H.Y., Wu, H.T., Tseng, K.Y., Chiou, A., Yu, C.J., Lee, Z.Y. and Chan, T.S. (2007) Fiber optic biosensor for the detection of C-reactive protein and the study of protein binding kinetics. J. Biomed. Opt. 12, 024025

8 Leung, A., Shankar, P.M. and Mutharasan, R. (2007) A review of fiberoptic biosensors. Sensors Actuators B 125, 688–703 9 Fan, X., White, I.M., Shopova, S.I., Zhu, H., Suter, J.D. and Sun, Y. (2008) Sensitive optical biosensors for unlabelled targets: a review. Anal. Chim. Acta 620, 8–26 10 Hagen, C.L., Schmidt, J.R. and Sanders, S.T. (2006) Spectroscopic sensing via dual-clad optical fiber. IEEE Sens. J. 6, 1227–1231 11 Read, I.J. and Foote, P.D. (2001) Sea and flight trials of optical fiber Bragg grating strain sensing systems. Smart Mater. Struct. 10, 1085–1094 12 Kister, G., Winter, D., Gebremichael, Y., Leighton, J., Badcock, R.A., Tester, P.D., Krishnamurthy, S., Boyle, W.J.O., Grattan, K.T.V. and Fernando, G.F. (2007) Methodology and integrity monitoring of foundation concrete piles using Bragg grating optical fiber sensors. Eng. Struct. 29, 2048–2055 13 Cheng, X., Zhang, L., Zhou, K., Davies, E., Sugden, K., Bennion, I., Hughes, M. and Hine, A. (2008) Real-time detection of DNA interactions with long-period fiber-grating-based biosensor. Opt. Lett. 32, 2541–2543 14 Chen, X., Zhou, K., Zhang, L. and Bennion, I. (2007) Dual-peak long-period fiber gratings with enhanced refractive index sensitivity by finely tailored mode dispersion that uses the light cladding etching technique. Appl. Opt. 46, 451–455 15 Langer, J.J., Galinski, ˜ R. and Gibinski, ˜ T. (1998) Electrically conducting and fluorising self-assembled monomolecular layers. Adv. Mater. Opt. Electron. 8, 31–37 16 Chen, J.C., von Lintig, F.C., Jones, S.B., Huvar, I. and Boss, G.R. (2002) High-efficiency solid-phase capture using glass beads bonded to microcentrifuge tubes: immunoprecipitation of proteins from cell extracts and assessment of Ras activation. Anal. Biochem. 302, 298–304 17 Debyser, Z. and Clercq, E.D. (1996) Chemical crosslinking of the subunits of HIV-1 reverse transcriptase. Protein Sci. 5, 278–286 18 Zhou, K.M., Lai, Y.C., Chen, X.F., Sugden, K., Zhang, L. and Bennion, I. (2007) A refractometer based on a micro-slot in a fiber Bragg grating formed by chemically assisted femtosecond laser processing. Opt. Express 15, 5848–15853 19 Chen, X.F., Zhou, K.M., Zhang, L. and Bennion, I. (2005) Simultaneous measurement of temperature and external refractive index by use of a hybrid grating in D fiber with enhanced sensitivity by HF etching. Appl. Opt. 44, 178–182 20 Yang, J., Sandhu, P., Liang, W.G., Xu, C.Q. and Li, Y.F. (2007) Label-free fiber optic biosensors with enhanced sensitivity. IEEE J. Selected Top Quantum Electron. 13, 1691–1696 21 Hughes, M.D., Zhang, Z.-R., Sutherland, A.J., Santos, A.F. and Hine, A.V. (2005) Discovery of active proteins directly from combinatorial randomized protein libraries without display, purification or sequencing: identification of novel zinc finger proteins. Nucleic Acids Res. 33, e32 22 Yang Li, Y., Chun, E., Liu, Y., Pickett, N., Skabara, P.J., Cummins, S.S., Ryley, S., Sutherland, A.J. and O’Brien, P. (2005) Synthesis and characterization of CdS quantum dots in polystyrene microbeads. J. Mater. Chem. 15, 1238–1243 Received 4 November 2008 doi:10.1042/BST0370445

 C The

C 2009 Biochemical Society Authors Journal compilation 

449