DNA branches into nanotech

Often viewed as a dream rather than a commercial prospect, DNA as a construction material in nanotechnology may be applied in practice sooner than many expect. Andrew Scott looks at the possibilities.

How much use can one chemical be? We might think that deoxyribonucleic acid (DNA) demonstrates sufficient utility and versatility by acting as the genetic material of life on earth. This central role in biochemistry has made the DNA double helix one of the most recognisable scientific icons of our time. A growing band of scientists, however, are trying to put the chemistry of DNA to new uses by exploiting it as a construction material in nanotechnology.

The dream is bold and is dismissed as over ambitious by some, but many of the most active researchers in the field believe that practical applications will arrive sooner than many expect. 'I'd say in three to five years you might get some decent biosensor applications based on this nucleic acid work, then in five to 10 years we should be able to do something useful with it in nanoelectronics,' says biochemist Thomas LaBean, a member of one of the most active teams in the field, led by John Reif and based in the department of computer science at Duke University, North Carolina, US.

LaBean's enthusiasm is matched by Nadrian Seeman of the chemistry department at New York University, US, and the most prominent pioneer exploring the potential of DNA in nanotechnology. 'People in business and industry should be paying attention to this field,' comments Seeman, 'because what we are offering ultimately is the ability to organise matter on the nanometre scale in a way that has previously not been available.'

Seeman first dreamed up the idea of using DNA as a construction material in the 1980s, when he was frustrated by his attempts to crystallise biomolecules for analysis by x-ray diffraction. He speculated that DNA double helices might be used to construct cages that would hold all the biomolecules in the same orientation, making them as ordered as they would be in a real crystal. More than 20 years after the idea first occurred to him, Seeman has still not put this concept into practice, although it remains one of his key goals. But that seminal thought has stimulated an exploration into the potential of DNA in nanotechnology that has now blossomed into a busy global research field.

The roles being proposed for DNA now also include holding in place components of proposed molecular electronic devices with nanometre-scale precision, acting as parts of molecular robots that will detect and cure disease, creating electrically conductive nanowires, forming crucial moving parts of 'nanomachines' and acting as the template for new types of polymers.

'The properties that make DNA so successful in acting as a genetic material also make it a convenient and logical molecule to use for constructing new materials on the nanometre scale,' says Seeman. As in nature, the specificity of base-pairing allows DNA strands to be bound together in a sequence controlled manner, just like scaffolding poles can be held together by clamps.

A key requirement for an effective scaffolding and construction material, however, is that it must be able to branch into other dimensions. Seeman and others have achieved this by copying from what DNA does in nature. During the process of genetic recombination temporary branched structures are created when two double helices come together. Several research teams have been able to synthesise DNA molecules that will form stable branched structures in a similar way.

To construct an extended grid, many such branches must be linked together. The base-pairing properties of DNA can be used to achieve this, by ensuring that protruding single-stranded regions of one branch have the precise base-sequence to form base pairs with complementary single-stranded regions on other branches. Using these so-called 'sticky ends' allows extended two-dimensional DNA grids to be created. The same building blocks have also spontaneously assembled into mesh-like nanoribbons about 60nm wide and 5µm long.

By varying the base sequence in the sticky ends, LaBean reports that the Duke University team is able to control the precise type of lattice that forms. Looking ahead to possible applications of this chemistry, the Duke researchers have pointed out that 'a two-dimensional lattice displaying a square aspect ratio would be useful for forming regular pixel grids for information readout from nanoarrays - for example, by encoding information in a pattern of topographic markers'.

If these achievements are also to be used to realise Seeman's original dream of constructing cages for crystallography and other applications, then the chemistry already demonstrated in two dimensions must be taken up into the third dimension. Seeman reports that much of the work now under way in his laboratory is focused on attempts to build three-dimensional DNA arrays. 'Right now we haven't got really good crystals,' he admits. 'We know we are making 3D crystals that fit with our designs but the quality of the crystals is too low for us to work out their structures in detail.'

Seeman emphasises that after so many years of concentrating on making novel DNA structures he regards the fact that they are now actually being produced as significant progress, even if the control over the process is not what would be desired. He also adds that, 'we haven't even opened the whole can of worms of how to get guest molecules in there'. But he expresses some confidence that useful applications will eventually be found, even if they don't necessarily match the hopes that are currently driving the research forward.

Perhaps the major application foreseen for DNA grids and lattices is to use them to position other molecules in the very precise way that would be required to construct molecular electronic devices, such as nanosensors and the components of bio-molecular computers. The team at Duke University has done some proof of concept work in this area. The researchers chemically modified their basic DNA branched 'tile' by incorporating a biotin group at a specific location. When the protein streptavidin was added to the self-assembled nanogrids the interaction of streptavidin with biotin produced arrays with streptavidin bound at periodic intervals.

Another proof of concept achievement by the Duke University team, and others, has been to metallise their mesh-like DNA nanoribbons with silver. This generated highly conductive nanowires, prototypes for the wires needed to construct nanocomputers. Other groups have used bacteriophage DNA as a template for depositing other metals, including gold, copper, platinum and palladium.

The possibility of using DNA lattices as parts of medical biosensors is the application that researchers seem most eager to speculate about. LaBean is prepared to imagine a simple example of what might be clinically useful within the three to five year timescale he reckons is feasible. He proposes a molecular scale agent that would have a biosensing component - perhaps just a specific molecule that could detect a disease state - combined with another component that would release a drug at the precise location where the disease state has been detected. DNA might form the structure used to hold these two components together, and DNA parts may also play a more active role in binding to particular genes or other biomolecules. Even before such basic devices have been constructed, however, LaBean is prepared to look farther ahead, saying: 'the long term goal would be a range of programmable biochemical machines that could be used for medical applications'. So instead of prescribing individual molecules, as they generally do today, GP's in the future may choose from a selection of molecular machines able to seek out and fix a wide range of biochemical maladies.

DNA is not the only material being proposed as the substrate for nanodevices, but the chemistry of DNA offers some particular advantages.

Researchers point to the high functional group density of the DNA molecule, which provides many opportunities to position components that can bind to the DNA by selective reaction with chosen functional groups. Also, the widespread use of DNA in biotechnology has led to the development of automated techniques to synthesise any chosen DNA molecule and join different DNAs together. This allows the budding DNA nanoengineers to gain complete control over the spacing between the branch points in a DNA lattice and the length and binding specificity of any protruding single strands. Single-stranded DNA can bind specifically to the DNA of selected genes, or to selected RNAs, so single-stranded DNA arms might serve as active components on nanodevices. These arms might bind to and detect genes and other biomolecules, for example, while double-stranded DNA serves as a less active scaffolding material within the same device.

DNA can also be induced to form structures that differ from the standard right-handed (B-DNA) double helix found in living cells. For example, left-handed DNA (Z-DNA) and more exotic structures such as triple helices can be formed. There are also a huge number of possibilities for using modified nucleic acids that share the base-pairing characteristics of DNA but also open up other chemical reactivities not found in DNA itself.

Several groups are exploring the possibilities of peptide nucleic acids (PNAs), which are analogues of DNA but with a peptide backbone replacing the sugar-phosphate backbone of DNA. The PNA backbone is electrically neutral, unlike DNA's negatively charged sugar phosphate backbone. This electrical neutrality may make PNA a better scaffold for holding together the components of nanoelectronic devices, since the scaffold will be less likely to interfere with the electrical processes taking place within the devices.

Work to convert the most adventurous imaginings of DNA nanotechnologists into reality has taken its first steps. Seeman's group, for example, has assembled DNA molecules into prototype switches, based on movement caused by the transition between B- and Z-DNA, tweezers and even a prototype bipedal walking device which the team hopes ultimately 'may be used to move molecules around for nanoscale construction and manufacturing'.

After many years in which people largely speculated on the possibilities, while only a few laboratories did real chemistry, active research into the nanotechnology possibilities of DNA and other nucleic acids is now expanding fast.

'The field really is exploding somewhat,' says Seeman, 'and that can actually lead to some frustration in our lab, because there was a time when we had the whole area to ourselves. But if it is going to be a genuine field and not just a personal idiosyncratic brand of science then we have to welcome that other people are going to work on it'.

In the UK, Andrew Turberfield's group at Oxford University is exploring the possibilities of using DNA as what he calls an 'addressable glue' to hold the components of nanosystems in precise orientations. He is also interested in the potential for using DNA molecules to perform computational procedures and has also been involved in the growing efforts to make moving parts of DNA-based machines.

In Germany, Friedrich Simmel of Munich University leads a group that, amongst other things, is exploring the potential of DNA to promote the self-assembly of nanoelectronic circuits. These would be used in molecular scale computing devices, with the DNA backbone acting as a template on which conducting and semi-conducting materials are deposited.

Most of these proposed new chemical applications for DNA would involve the manufacture of tiny quantities of very sophisticated components. There is one possibility, however, that might make an impact on the massively large-scale polymer production industry. Seeman and others have used nucleic acid molecules as templates on which to synthesise nylon-based polymers. Their hope is that polymers may be produced with very specific topologies, such as different degrees of helical winding, that would give them a range of new and readily controlled properties.

These possibilities are yet another reason why Seeman feels that industrial and commercial interests should be paying attention to this area of chemistry which, at times, can seem more like a land of dreams than one of hard commercial possibilities.

'Assuming any of this stuff winds up being effective it will be a highly disruptive technology,' warns Seeman. 'People in industry have a huge investment in the way things are done now. I expect our methods could be relatively cheap and so sooner or later those people are going to be forced to do something with it.'

Acknowledgements

Andrew Scott is a writer and lecturer based in Perth, UK, and a founding board member of the International Center for First-year Undergraduate Chemistry Education, based at Illinois University, US

References

Y Benenson et al, Nature, 2004, 429, 423-9

Further Reading

N C Seeman, Scientific American, 2004, 290, 64

H Yan et al, Science 2003, 301, 1882

First steps towards medical nanomachines

 

Scientists at the Weizmann Institute in Israel have already taken the first steps towards using nucleic acids to form medical nanomachines that can diagnose and treat disease. Ehud Shapiro?s research team has created a molecular computer that can detect key biochemical indicators of some types of cancer and release an anti-cancer molecule in response. Their system has only been tested in vitro, and in its present form could not work in the body, but it gives a first indication of the potential of DNA-based nanotechnology to transform medicine.

At the heart of the system is a synthetic DNA molecule designed to achieve both diagnosis and therapy. This DNA molecule includes a loop at the end that, if released by the action of a DNA-cutting enzyme, forms a much smaller single-stranded DNA with anti-cancer activity. This DNA drug works by binding to natural nucleic acids in the cell and interfering with the activity of the corresponding genes.

The drug is released after a multi-step diagnostic process, depending on the level of natural messenger ribonucleic acid (mRNA) molecules associated with specific types of cancer. The systems used in the trials were targeted against prostate cancer and lung cancer.

Each step in the diagnosis involves a nucleic acid binding by base pairing to a protruding 'sticky-end' on the DNA device. This binding initiates an automatic process in which an enzyme cuts off part of the end of the DNA device and exposes another diagnostic sticky-end. This second sticky-end is designed to bind to another nucleic acid, and so on. If all the diagnostic steps occur in the manner that indicates the nucleic acids associated with the disease state are present, then the single stranded DNA drug is automatically cut out and released.

The system was made more sophisticated by the presence of a distinct but similar DNA device that releases a drug suppressor if the disease state is not detected. If this method were used in vivo, the balance between drug and drug-suppressor would determine whether effective therapy was administered.

When the results of the trials of this device were released in May 2004, Shapiro commented: 'The best way to think about it is as a smart drug. that has some conditions encoded for its release. It will be released and activated only at the right time and at the right location when a disease is diagnosed.'

The key simplicity of this prototype nanomachine is that the diagnostic and therapeutic components are both based on DNA, and enzymes dissolved in the surrounding medium activate them. But simply injecting these nucleic acids and enzymes into a person would wreak havoc on their biochemistry.

More adventurous but more useful designs may have all the components held together by being attached to a DNA scaffold.

There are challenging obstacles between this test-tube prototype and the synthesis of any practical therapeutic device. Nevertheless, many researchers are thinking about more complicated and useful possibilities.