Ten Technologies in Five Years

When scientists make long-term research plans, they must try to anticipate how emerging technologies will influence their work in the coming years.

Sam Jaffe(sjaffe@the-scientist.com)
Dec 5, 2004

Samples are obtained (e.g. blood sample from a patient) and target molecules (e.g. proteins like prostate specific antigen (PSA) or DNA/RNA sequences) are captured using magnetic particles that bind specifically to the target. Gold nanoparticles modified with target selective molecules and hundreds of barcode strands are used to label the captured target. The bio-barcodes on the surface of the gold nanoparticle are released and subsequently detected, heralding the presence of the target molecule.

When scientists make long-term research plans, they must try to anticipate how emerging technologies will influence their work in the coming years. To make that task a little easier, The Scientist scoured the grant wire and talked to dozens of leading scientists, research directors, and government officials. We asked: Which new technologies do you expect will be commonplace in laboratories five years from now?

From the candidates, we've chosen 10 that we think will...


Much has been made of the potential of aptamers – nucleic acids that can specifically bind to proteins or small molecules – to replace antibodies or drive biosensors. But little progress has been evident to date. Now Milan Stojanovic and Dmitry Kolpashchikov at Columbia University have engineered the aptameric equivalent of green fluorescent protein.1 "GFP revolutionized the way science studied functional genomics," says Stojanovic, "but nothing similar has worked in the RNA-to-protein portion of the equation."

Stojanovic's creation is actually a collection of three different aptamer modules. The first binds a selected analyte, the second encapsulates a fluorophore, and the third serves as a communications link between the other two. When the sensor module binds to its analyte (ATP, for instance), the cage on the reporter module closes, binding the dye (malachite green), which then fluoresces to signal the binding event.

The result is the first transfectable, self-reporting aptameric sensor. Next up: system optimization. Although malachite green works reasonably well, it is not a consistent performer. So, Stojanovic and his colleagues are testing some of the hundreds of other commonly used dyes to find one that works with his modules.


Biotech giant Genentech, whose blockbuster drug Herceptin heralded the power of antibody therapeutics, is looking to ease the process of creating new drugs (and research tools) in the future. The trick? Starting with a well-characterized antibody as a scaffold, Sachdev Sidhu and colleagues systematically scrambled solvent-exposed portions of the comple-mentarity-determining regions to create phage-display libraries capable of generating antibodies against almost any antigen.2

Constructed on the back of an antibody known as 4D5, which targets Erb2, the library contained antibodies with nanomolar affinities to 13 different test antigens. And thanks to the phage-display technology, any desired antibody can be quickly produced on a massive scale. "We can do in one week what it used to take six months to do with traditional hybridoma technology," says Sidhu. "And because most of the molecule is already human, the synthetic antibody has few if any immunogenicity problems."


Northwestern University nanotechnology wunderkind Chad Mirkin seeks to reinvent nucleic acid and protein detection. The magnetic BioBarCodes system he developed,3 soon to be available commercially from Nanosphere, uses an antibody (or oligonucleotide) library whose volumes have been tethered to nanometer-sized iron spheres.

Superficially, the system is like any bead-based sandwich assay: Antibodies capture their target analytes from the solution and then a second, gold nanosphere is added, forming a bead-analyte-bead sandwich. But the gold sphere also carries a BioBarCode, which identifies the reaction. In the last step these tags are released and quantified on a special DNA microarray.

The system promises to detect proteins (or nucleic acids) at attomolar levels. And, it can all be done cheaply and quickly, unlike traditional ELISA or PCR. "Nothing has shown this level of accuracy at detecting such miniscule portions of analyte," says Leigh Anderson of the Plasma Proteome Institute, a public/private partnership that hopes to determine the complete proteome of human blood plasma. Mirkin was recently awarded a $2.5 million Director's Pioneer Award from the National Institutes of Health to refine his system.


Neurons have long bedeviled biologists; the cells grow poorly in culture and produce faint in vivo signals. Techniques such as electroencephalography and functional magnetic resonance imaging are difficult to reproduce and extremely bulky for in vivo imaging.

Now neuroscientists have the ability to detect messages delivered by individual neurons on a free-roaming subject. Led by the groundbreaking methods developed by Miguel Nicolelis at Duke University,4 the technology is based more on software that reduces background noise than any leaps in sensor technology.

The hardware consists of nanometer-scale tungsten isonel-coated electrodes implanted into the brains of lab animals (the technology hasn't been tried on humans yet) plus a wearable battery pack and computer chip. The electrical signals are then filtered and amplified using the lab's proprietary computer program. "We can hear what an individual neuron is saying, which makes neuroscience that much more precise," says Sidharta Ribiero, a postdoc in Nicolelis' group.



The nanoscope forms images with subwavelength resolution by sampling the light scattered by an object flowing through a microfluidic channel in the near field. An array of quantum dots, randomly placed at the bottom of the microfluidic channel, serve as an optical sensor. A final image is formed in the computer by combining the emission spectra from the quantum dots and the spatial and spectral distribution of the quantum dots

What better to see on a nanoscale than with a nano-sized device? That's Demetri Psaltis' thinking. Psaltis oversees the California Institute of Technology's Optofluidics Institute, founded this summer with an $8 million grant from the Defense Advanced Research Projects Agency. Psaltis has already created a nanoscale sensor embedded in a fluid channel that literally engraves an image of its target on its face. Now his team is building a nanoscope that can do this on a consistent basis. "It's like creating a photographic plate on the nanoscale, but without a lens," Psaltis explains. "The resolution will be extremely high, but the magnification will be in the software, not in a lens," he says.

Psaltis points out that electron microscopy is based on electron backscatter, and atomic force microscopy produces a recreated image based on physical feedback. "The image that comes back from our nanoscope will hold all the optical properties of the target just like a photograph does. It should show us things that we've never seen before." The first operating nanoscope should come online sometime in 2005.



Courtesy of Philippe Zeitoun

Schematic of the soft x-ray laser. The soft x-ray amplifier, a hot plasma (1 billion °C) created from a gas heated up by high energy infrared laser (in green), is seeded by a "perfect" soft x-ray beam (in yellow) generated by non-linear interaction of an ultra-short laser with noble gas. After amplification, the energy of the seeded soft x-ray laser (in purple) has been increased up to 600 times. The seed spectrum (left) is made of a comb of wavelengths, only one of them being amplified by the plasma as shown on the right spectrum. The spectrum colors are false.

X-ray crystallography is a more conventional technology for visualizing the nanoworld. The technique provides stunning images of protein architecture but at the expense of the protein sample itself. Three different billion-dollar projects (two in Germany and one at Stanford University) are presently underway to build massive free-electron lasers that will create X-rays in short-enough pulses to prevent the destruction of themolecule being viewed.

Now a team of physicists at the Laboratory of Applied Optics in Palaiseau, France, might have found a much cheaper alternative.5 The team created a tabletop device that seeds a laser, which then fires in femtosecond pulses that are too quick to damage their target, and then amplifies the beam so that it can view things at the nanometer level.

"We've proved that we can create a pulse strong enough and short enough to do in vivo biological imaging," says Phillipe Zeitoun, the first author of the paper announcing the team's results. Zeitoun expects the laser to cost less than $250,000 when mass-produced, a mere fraction of the cost of building traditional free-electron lasers.


Without a doubt the power of grid networks and supercomputers will continue to grow in the coming years. Both forms of computing are useful to bioinformaticians, but the field will likely benefit from the successful meshing of the two. "Big, centralized computers are good at certain tasks and distributed networks are good at others," says Grant Heffelfinger, the deputy director of materials at Sandia National Laboratories. "If we can farm out the right calculations to the right system at the right time, we can optimize a hybrid system." That's the goal of a $78 million program overseen by Heffelfinger that aims to create a hybrid supercomputer/grid network.

The project includes teams from more than 30 institutions throughout the world. Its aim is to join Sandia's experimental hyperspectral scanner (originally designed for nuclear weapons research) with the new hybrid computing system to determine the exact pathway Synecococcus (a cyanobacterium) uses to sequester carbon dioxide. While it is hoped that the basic biology of carbon sequestration will be served, the main goal is to create a template and a working hybrid computing system that can be used for just about any molecular modeling project.


When diagrammed, genetic networks resemble a more complex version of an electrician's wiring scheme for a house. That has led many to hope that someday genetic engineering can merge human-made circuits with natural genetic circuits. That promise is already being fulfilled at Boston University, where James Collins and colleagues have devised a genetic switch that can toggle mRNA translation on and off.

The switch has two components. The first is a sequence, engineered into the 5'-untranslated region of a bacterial mRNA, which folds back to form a stem-loop structure in the ribosome-binding site (RBS). The second component is a trans-acting sequence element that can bind to and disrupt the loop, thereby exposing the RBS and allowing translation to proceed.

Collins has already shown that the technique works to regulate GFP expression in Escherichia coli.6 Now he's created its first practical application: a bacterial colony that emits toxins to digest the extracellular matrix in a biofilm and then switches to manufacturing a separate toxin when they encounter the actual bacteria inside the film. "It's the first real solution for one of the biggest hygiene problems in the healthcare field," says Collins.

Collins sees a bright future for riboswitches as a flexible alternative to the much-hyped RNAi, because while riboswitches can be turned on and off, RNAi can only silence. As such, he expects riboswitches to be standard genetics fare within three years.


The real promise for riboswitches might go even further than the bench lab. Richard Mulligan of Harvard's Children's Hospital used viral vectors to implant a similar riboswitch into a mouse genome that can be turned on and off with a small inducer molecule.7 Such a system can be an extremely efficient way to study the DNA-to-RNA portion of the gene-expression pathway.

It also holds out a far more lucrative future. If proven safe, eventually human cells can be reprogrammed with such a riboswitch to manufacture a specific protein only when the person takes a pill that is otherwise neutral to the system. "Instead of taking a pill to alter your biochemistry, this could turn the body into a drug factory," says Collins of Mulligan's work.


Few goals are more important in drug development today than finding an easy-to-breed model organism that can predict drug safety and efficacy in the human body. Markus Manz and colleagues at the Institute for Biomedicine in Bellinzona, Switzerland, took a giant step towards that goal recently when they successfully reconstituted a human adaptive immune system in immunodeficient mice from human cord blood stem cells.8

Although it will be a while before Manz can create a mouse with a completely human immune system, the technology does open the door for more accurate predictive results in preclinical pharmacology. More important to Manz, though, it opens up a new route for bench researchers to study how the human immune system fights human viruses without endangering the safety of human test subjects. "This is a great system to test infections and to tinker with stem cell research," Manz says, who has already used the altered knockouts in his HIV research.

Interested in reading more?

Become a Member of

Receive full access to digital editions of The Scientist, as well as TS Digest, feature stories, more than 35 years of archives, and much more!
Already a member?