|Home | About | Journals | Submit | Contact Us | Français|
Nanopatterned arrays of biomolecules are a powerful tool to address fundamental issues in many areas of biology. DNA nanoarrays, in particular, are of interest in the study of both DNA-protein interactions as well as for biodiagnostic investigations. In this context, achieving a highly specific nanoscale assembly of oligonulceotides at surfaces is critical. In this chapter we describe a method to control the immobilization of DNA on nanopatterned surfaces: the nanofabrication and the bio-functionalization involved in the process will be discussed.
The ability to control biomolecules on surfaces with nanometer resolution is of great interest to study biological events at the single biomolecule level, and provides a platform for the development of biosensing devices with unparalleled sensitivity(1–9).
In particular, the confinement of DNA at surfaces has gathered a great deal of interest, as it can be employed for bioanalytical (genomic) studies, as well as to drive the self-organization of biological (e.g. proteins) and inorganic moioties (e.g. nanoparticles) with nanometer resolution(10–17).
Various chemical strategies have been employed to immobilize DNA at surfaces, ranging from electrostatic interactions, formation of (thiolated) self-assembled monolayers (SAMs), direct covalent attachment, and through biotin-streptavidin interaction(18–24). Different fabrication methods have been used to control such immobilization in order to generate micro- and nano-scale DNA features: contact printing(25–27), AFM based methods(28), and nanopipette deposition(29) are among the most notable examples of the different approaches pursued.
In this chapter we present a strategy merging top-down nanofabrication techniques with bottom-up self-assembly, to control the confinement of DNA molecules on substrates with nanometer resolution.
We will first describe the fabrication procedure employed to nanopattern glass substrates with Au/Pd nanodots. This consists of a direct electron-beam (e-beam) writing step that creates nanometer-scale voids in a resist polymer spun on the glass substrate; a subsequent metal evaporation fills the voids with Au/Pd and produces an array of sub-50nm metal dots in the desired geometry (defined by the e-beam writing).
The Au/Pd nanodots so produced can be used as functional regions on the substrates for the controlled confinement of DNA. In this context, we will here describe a biotin-streptavidn based functionalization methodology to immobilize oligonucleotide chains on the so prepared nanopatterend surface(21, 22, 24, 30–32). The metal nanodots allow for the formation of SAMs of thiolated alkanes presenting biotin end-groups, which can be used for the subsequent immobilization of streptavidin. Employing such a strategy, we will discuss and demonstrate the ability to immobilize double stranded DNA (dsDNA) on each and every fabricated (and properly functionalized) nanodot. This functionalization procedure allows for the formation of non-sterically hindered DNA nanodomains at surfaces; an homogenous surface packing density can be envisioned, an advantage over thiolated DNA SAMs(20). We will verify the validity of our approach by EPI-fluorescence microscopy imaging.
Throughout the steps of the method described here, a great deal of attention should be focused on the cleanliness of both the laboratory working environment as well as the materials and tools employed.
Because of the nanometer scale of the features fabricated (and functionalized) we recommend carrying out the fabrication steps (section 3.1) in an ultra clean environment, such as a cleanroom. Furthermore, in order to obtain a successful bio-fucntionalization (section 3.2) of the fabricated nanopatterns, it is important to rapidly carry out the procedure with as small a time lapse as possible between steps. Moreover, all glassware and tweezers used must be dry, preferably stored in an oven at ca 70 °C and cooled in air prior use.
It should be noted that in order to verify the validity of our approach (i.e. the controlled DNA functionalization of nanopatterned surfaces) we have used a biotinylated, and Cy3-labelled, dsDNA. We have hybridized in solution, prior to the attachment to the surface, a biotinylated oligonucleotide (20-mer chain) and its complementary oligomer labeled with a Cy3 fluorophore. (In the notes section we will discuss how, and why, the same DNA immobilization procedure can be alternatively employed to bind a biotinylated single stranded DNA).
Cleaning is absolutely critical to the success of the procedure. Any contamination, even nanoscopic, not removed prior to patterning will result in defects in the surface passivation(see Note 1).
It is now possible to proceed to resist deposition (see next section).
Again, cleanliness is paramount here. All work should be performed in a cleanroom, class 10,000 or better. A bi-layer of higher MW resist is spun on top of a lower molecular weight resist to aid with the subsequent metal deposition and liftoff (discussed later). The high MW top layer has a slightly different dose curve, and will develop with a narrower opening, creating an overhang, which ensures proper liftoff (see Note 4).
It is now possible to proceed to the e-beam writing step, as described in the next section.
An electron beam writing system is a SEM, which additionally controls beam shuttering and position to generate patterns from CAD files. Process testing will be necessary to determine optimal doses for generating the desired features (see Note 7). We use a pattern which is 50 μm by 50μm, and consists of 1 um register squares spaced every 10μm, with sub-50 nm dots filling between them every 2μm. This ensures each individual dot is optically resolvable and discrete once functionalized and imaged with a fluorescence microscopy (see Note 8).
Samples are now ready for development.
A cold ultrasonic development process is used to achieve nm scale features. The cold development sharpens resist contrast, resulting in the smallest possible features from the exposed regions.
It is now possible to proceed to the metal deposition step.
Metal deposition is performed in a Semicore electron beam evaporator. A focused beam of electrons is used to heat a target material held in a crucible. As opposed to a thermal evaporator the heating is highly localized to the surface and allows for precise control of the thickness (at the angstrom level). The sample is held above the target some distance away, allowing for a highly directional and uniform material flux onto the surface.
The metalized sample now consists of a layer of metal sitting atop the unexposed PMMA, with openings in the resist where the sample was exposed to electrons during the e-beam writing session (in these holes, the Ti adhesion layer and Au/Pd are deposited on the glass surface): see figure 1.
Liftoff is the process to remove the remaining, unexposed resist now covered with metal. Recall a slight overhang was created due to the use of a bi-layer. Therefore metal deposited inside the features on the glass substrate are not connected to the bulk of the metal sitting on top of the resist. Solvent is used to dissolve the resist, removing the metal, while leaving behind only the metal deposited in the holes defined by beam writing. The metal will appear to float off of the surface, hence the term liftoff.
The Au/Pd nanopatterns on glass are now ready for surface functionalization. Figure 2 displays an Atomic Force Microscopy (AFM) image and profile of the sub-50 nm dots fabricated on a glass coverslip.
Figure 3 schematically displays the main steps of the biofunctionalization procedure.
Starting with a nano-patterned surface, i.e. Au/Pd nanodots on a glass slide:
The sample is now ready: i.e. Au/Pd nanodot properly and specifically functionalized with the oligonucleotide of interest.
In the next section we will describe how we characterized the functionalized surface, verifying by fluorescence microscopy the presence of Cy3 labeled dsDNA immobilized on sub-50nm Au/Pd nanodots on glass.
Inverted microscopes are designed to accept coverslips mounted to standard glass slides. We have devised a simple mounting scheme with our coverslips for fluorescence microscopy (see Figure 4). It is an open design, which allows for additional processes to be carried out in situ on the microscope. Microfluidics can and have been used but would warrant a whole other methods chapter unto themselves. Here we present a setup offering most of the functionality without the complexity.
An inverted fluorescence microscope capable of EPI Fluorescence microscopy is used to image the samples. Oil immersion lenses with 60x and 100x magnification are best used for imaging of the nanopatterns. The camera is a photometrics Cascade II, it is cooled to −70°C and has on chip amplification for low noise, high sensitivity imaging.
1Make sure all glassware used is clean, i.e. washed with detergent, rinsed in DI water, rinsed with ethanol and blown dry with a stream of inert gas (Ar or N2).
2When handling samples, especially washing and drying with tweezers, always make sure to rinse and blow dry towards the tweezers, this prevents the transfer of any contaminants from gloves and tweezers onto the samples
3Clean the coverslips immediately prior to the resist deposition; if they are allowed to sit in air they will accumulate dust.
4Time is of the essence; the quicker one works while spinning resist layers the less likely the surface is to be contaminated.
5Make sure the coverslip is cool before putting resist on, and never let resist sit for more than a few seconds before spinning. The solvent will evaporate slightly, especially at the edge resulting in an uneven coating, which may or may not be obvious, but will affect the obtained features.
6Observe the sample after each layer of resist is deposited. If there is streaking, this means the surface was not properly cleaned. Discard the sample and try again. Properly spun resist should uniformly coat the surface to the edges, looking smooth and glassy.
7Larger features (micron sized) are written at spot size 5, while the nano-features are written with spot size 1.
8The nanodots themselves are not visible in an optical microscope, The 1 μm registers are visible, and are intended to provide points of reference for the location of the nanodots, and to assist with focusing on the optical microscope. If you wish to characterize the nanodots, AFM is the best method.
9It is important to correct for beam tilt and stigmation. If the beam is tilted the projection of the feature will be an oval rather than a circle. This will affect feature shape and minimum achievable size, not to mention the fact that the focal plane of the beam will not be parallel with the plane of the sample. Stigmation results from the lens not being perfectly round, in fact no lens is. If you imagine the lens is an ellipse, and break the focus up into orthogonal x–y coordinates aligned along the short and long axis of the lens, it becomes apparent that the lens will have different depth of focus for each axis. This exhibits itself when imaging as the ability to get one edge in perfect, crisp focus, while edges not parallel with the sharp edge will look out of focus, worsening as the angle of the two edges approaches 90 degrees with respect to the edge. Adjusting the stigmation will correct for the aberration, resulting in better focus and thus smaller features.
10A faceted, crystalline particle such as Al2O3 is a good target for stigmation correction. When all edges appear clear, regardless of orientation, stigmation is minimized. Sprinkling these particles at the corners of the slide and using them to define the four point focus as well is a commonly used technique. The edge is actually a few hundred nm above the surface however, so the focus is not entirely accurate. This method is best for features down to ~35nm, below that, the “spot” method should be used (see below).
11“Spot” method: This requires using a 15nm film of aluminum as the discharge layer instead of Aquasave. Focus as best you can on the surface. Use the spot function in the scan menu to manually expose a spot on the surface, and then subsequently optimize focus and stigmation on the spot until it is sharp. This must be done iteratively, the resulting spot coming out smaller and sharper each time enabling further refinement of the stigmation and focus. When the spot size is no longer getting smaller focus is optimized. Always adjust focus first, before stigmation is adjusted. This is the best method for performing four-point focus, as it verifies that the focus is optimized for exposure of the smallest possible area. Spotting four corners in a box, immediately around the area to be written ensures the best focus for writing.
12We recommend keeping the sonicator, and water/IPA development solution in a refrigerator prior to and during development. If that isn’t possible be sure bath and H2O/IPA are chilled to 4°C before proceeding.
13Development solutions can be reused, but should be changed periodically (every 10 uses) and stored at 4°C.
14The development and quenching in room temperature IPA are time critical, so prepare everything in advance and move as quickly as possible. It is recommended to bring the IPA over to the cold sonication and after one minute, transfer the sample directly into the IPA at room temperature.
15It is important to prevent any re-deposition of the lifted off metal.
16When placing the samples in the liftoff, keep them mostly vertical, but tilted slightly face down so the metal will fall away from the surface. Use one of the slotted Teflon racks as a holder.
17Never allow the samples to go dry in the acetone or upon removal during the rinsing steps.
18Always use separate dedicated syringes and needles for the anhydrous solvents employed. Rinse the needles and syringes after use, and store them in an oven at 70°C: the syringe and the needle used for toluene should be rinsed in acetone, then Ethanol. Blow-dry the needles before storage in oven.
19Incubate the samples in the biotin solution directly after plasma cleaning: allowing the sample to sit too long in air has a detrimental effect over proper SAM formation.
20When preparing the PEG-silane solution make sure to work in a cold environment chamber (temperature below 5°C).
21A functionalization procedure incubating the sample in the PEG-silane solution first and then thiolating the gold works but gives rise to a less reliable passivation, probably due to physisorbtion of biotin-alkane chains on the PEG layer, that can then play an active role as functional spots for the attachment of streptavidin on surfaces.
22The purpose of Albumin is to block non-specific binding between streptavidin and any passivation defects on the glass surface.
23It is recommended to incubate in the streptavidin solution immediately after drying the sample.
24It is possible to check the intermediate step by using a fluorescent labeled Streptavidn (Invitrogen): suggested fluorophores include alexa-488 and Cy5. It is also possible to use Avdin or Neutravidin, instead of Streptavidin, although Avidin is not recommended due to a high degree of non-specific adsorption.
25Either single stranded (ssDNA) or dsDNA can be immobilized on the metal nanodots depending on the application: dsDNA nanodomains, for example, can be meaningful to study protein-DNA interactions, while bound biotinylated ssDNA is available for in-situ hybridization to its complementary sequence, a process of interest for bioanalytical applications. Furthermore immobilized ssDNA can serve as an anchoring point on the surface to drive the self-assembly of biological and inorganic nano-objects properly functionalized with the complementary strand: the information encoded in the double-helix can be a powerful tool for the fine control of such organization, noteworthy at the nanoscale.
For ssDNA: at step 15 of section 3.2, incubate the sample in a ssDNA solution at a concentration of 2 mM in TPBS at RT for at least 3h (up to 12/18h, i.e. overnight). In order to proceed with a RT in-situ hybridization on the nanodots (fabricated as described above, and functionalized with ssDNA), incubate then the sample in a 2mM solution of the complementary strand in TPBS adding 3mM NaCl. The purpose of the high salt concentration is to screen any repulsive interactions taking place between the phosphate groups of the complementary DNA strands, as they can prevent an effective hybridization from taking place. The incubation should be done in at least 1.5 mL of solution and left in a container sealed with parafilm and covered with Al foil over-night (minimum incubation time 12h, up to 36h): do not shake the container during the incubation. After incubation rinse the sample in PBS and incubate in PBS for 1h: the rinsing is very important to get rid of residual NaCl salt physisorbed on the substrate.
26The incubation with biotinylated dsDNA can go up to 12/18h (i.e. overnight).
27Imaging is always a balancing act between exposure time, and amplification to achieve detection threshold. Longer exposure times integrate more light and result in better signal to noise, but result in loss of time resolution and lead to faster photobleaching. Increasing the gain amplification allows for lower exposures but increases the effect of noise, which can swamp out the fluorescence signal. We typically work at exposures of 100–300ms as this provides adequate signal to noise with most fluorophores we work with.