The technology aims to speed up a process that is crucial to drug discovery and development.
Separating proteins from biological fluids such as blood is becoming increasingly important for understanding diseases and developing new treatments.
Indeed, examples include the preparation of commercial products such as enzymes (e.g. lactase), nutritional proteins (e.g. soy protein isolate), and certain biopharmaceuticals (e.g. insulin).
The engineers said that the molecular sieve would make it possible to analyse protein property, improving on the current method of choice, gel electrophoresis, which is time-consuming and less predictable.
Pore sizes in the gels vary, and the process itself is not well understood by scientists.
The sieve, developed by Massachusetts Institute of Technology (MIT) engineers is made using microfabrication technology and is the uniform size of the nanopores through which proteins are separated from biological fluids.
Millions of pores can be spread across a microchip the size of a thumbnail.
"No one has been able to measure the gel pore sizes accurately," said Jongyoon Han, the Karl Van Tassel Associate Professor of Electrical Engineering and Biological Engineering at MIT.
"With our nanopore system, we control the pore size precisely, so we can control the sieving process of the protein molecules."
Han and his team have gone on to devise a sieve that is embedded into a silicon chip.
A biological sample containing proteins is put through the sieve for separation.
"This is the first time anyone was able to experimentally confirm this theoretical idea behind molecular sieving, which has been used for more than 50 years," Han said.
"We can control the pore size, so we can do better engineering.
We can change the pore shape and engineer a better separation system."
No single technology platform dominates the proteomics market.
However, one of the leading approaches for proteomics research is liquid chromatography/mass spectrometry (LC/MS).
Liquid chromatography is a widely used technology for separating proteins from complex mixtures so they can be further analysed.
While the overall LC market is growing at single-digit levels, the LC market for proteomics is expected to grow by 10 to 15 per cent annually.
Proteomics researchers often combine an LC system with a mass spectrometer.
The sieving process is based on the Ogston sieving theoretical model.
In the model, proteins move through deep and shallow regions that act together to form energy barriers.
These barriers separate proteins by size.
The smaller proteins go through more quickly, followed by increasingly larger proteins, with the largest passing through last.
Once the proteins are separated, scientists can isolate and capture the proteins of interest.
These include the "biomarker" proteins that are present when the body has a disease.
By studying changes in these biomarkers, researchers can identify disease early on, even before symptoms show up, and potentially develop new treatments.
To date, the Ogston sieving model has been used to explain gel electrophoresis, even though no one has been able to unequivocally confirm this model in gel-based experiments.
The MIT researchers were, however, able to confirm Ogston sieving in the nanopore sieves.
The team's results appear in recent issues of Physical Review Letters, the Virtual Journal of Biological Physical Research and the Virtual Journal of Nanoscale Science and Technology.