Electronic DNA microarrays

DNA contains and issues the language of life. It gives cells instructions for living, and tells living organisms about their hereditary traits. This language is coded into the DNA’s famous double helix structure: Fig 1(a). Each helical strand exhibits a sequence of four chemical bases, adenine (A), guanine (G), cytosine (C), & thymine (T), e.g., CAAGTG. The two twisted strands are bound together by pairing base A always with base T, and G always with C. Due to this pairing rule, within any section of DNA, once the sequence of one strand is identified, that of the other strand is easily inferred: e.g., the complementary sequence of CAAGTG is GTTCAC. DNA sequences are the language of life. Reading them, therefore, is of prime importance.
    Electronic DNA microarrays are CMOS integrated circuits (ICs) that can rapidly decipher unknown DNA sequences [1-6]. A double-stranded DNA molecule can unzip into two complementary strands. A single-stranded DNA molecule thus obtained can bind back to its complementary sequence (either the old mate or a new one), forming again a double-stranded DNA molecule. This binding of two complementary strands, or hybridization, underlies the genetic sequencing operation of the electronic DNA chip.

The electronic DNA microarray is constructed by immobilizing single-stranded DNA molecules of different identified sequences onto different grid points on a CMOS IC [Fig. 1(b)]. The grid points are often defined by post-fabricated gold electrodes that are electrically connected to the underlying CMOS IC. Different grid points represent distinct DNA sequences. These single-stranded DNA molecules of known sequences making up the array are called DNA probes. Now consider single-stranded DNA molecules of an unknown sequence, or DNA targets. When a solution of DNA targets is introduced onto the DNA array, the target strands wander around to eventually hybridize to their complementary probe strands at a specific grid point [Fig. 1(b)]. Locating the hybridization position reveals the target sequence, for we already know the probe sequence at that position, which must be complementary to the target sequence. The CMOS IC underneath is used for electronic detection of the hybridization point.
One well-established electronic detection technique is to label DNA targets with reporter molecules of a distinctive electronic signature and to search for them. Redox enzymes are an example of such electronic labels. If a voltage is suddenly applied between an electrode where hybridization occurred and the electrolyte (DNA target solution), redox labels attached to target molecules will give up electrons to the electrode, thereby increasing the current through the electrode. The same voltage step in an electrode with no hybridization would cause no such current increase as redox labels are absent at that electrode. By applying the voltage step and monitoring current change fast across the whole array using the underlying IC, hybridization positions are rapidly detected, leading to target sequence identification. Redox-label-based CMOS DNA chip examples are found in [1-3].
While the label-based detection offers excellent sensitivity, significant efforts are being placed to develop label-free electronic DNA microarrays [4,5] because elimination of labeling steps would simplify the sample preparation. In [4], for example, the capacitance of an electrode immersed in the electrolyte is monitored to sense hybridization. No labeling is needed, as target strands added to an electrode during hybridization naturally lead to a dielectric constant change, or, capacitance change.
Field effect transistors (FETs), the basic commodity of CMOS ICs, can be also useddirectly for label-free electronic detection of DNA hybridization [7-9]. Underlying this sensing modality is the exploitation of the impact of DNA’s intrinsic negative charges upon the FET behavior. Imagine that underneath the DNA array there is a corresponding FET array integrated in the CMOS IC. The gate dielectric of each FET is linked to DNA probe strands of the same sequence in a corresponding site in the DNA array. When a target strand hybridizes to its complementary probe strand anchored to a specific FET gate, the target’s intrinsic charge alters the channel conductance and capacitance of the FET. Therefore, by monitoring the channel of each FET in the array, one can attain label-free electronic readout of target sequences. While field effect sensorsare widely used [9], there is a lot of room for development in their use in DNA microarrays. Post-processing to expose gate dielectrics to electrolyte may pose a challenge. The smallest possible FET width must be used to maximize the impact of the DNA’s charge on channel properties.
Hybridization is at the heart of many other DNA sequencing techniques. What makes electronic DNA microarrays unique is their massively parallel operation. Distinct probe sequences numbering as many as hundreds of thousands can be simultaneously available across an array. The CMOS IC monitors each site of the array fast across at a gigahertz speed, and hence, its operation may be regarded as parallel to human eyes. This parallelism allows for rapid collection of vast amounts of genetic information (far faster than non-microarray techniques), accelerating the speed at which we probe the secrets of living organisms. The parallelism is a direct outcome of using CMOS microfabrication techniques to build large microarrays, and is enhanced by the use of integrated electronics.
In the original invention, which is still the commercially dominant form of the DNA microarray, hybridization is sensed by optical means [10]: florescent dyes labeling target strands light up upon illumination, reporting hybridization points. This optical machine boasts sensitivity superior to, and parallelism similar to, its much smaller electronic cousin. Although considerable work is needed to develop a high-performance electronic DNA chip as an alternative to the optical type, decisive advantages of using ICs (small size, low cost, programmability, real-time, label-free options) are the cogent reason for the ever-growing efforts in the development of electronic DNA chips.


Other electronic biosensor microarrays

Generalizing the concepts of the electronic DNA chip, one can readily consider CMOS biosensors that can detect other biological objects such as viruses and disease marker proteins [11]. Just like a DNA strand sticks specifically to its complementary strand, a virus or a protein binds specifically to its unique biochemical mate, an antibody. This highly specific binding is analogous to the way different keys fit into different locks. Fig. 2 illustrates a scenario where nine different types of viruses can be detected by using an array of nine specifically corresponding types of antibodies. When the array is exposed to a target solution containing type-4 viruses, for instance, the viruses will specifically bind to antibodies in array point 4. This binding can be detected using the CMOS IC below in much the same ways DNA hybridization is detected. By reading the location of the binding, the presence and type of viruses are determined. The multiplexed array platform can be especially useful for the diagnostics of complex diseases like cancer.