August 13, 2008 -- A team of researchers supported by the U.S.'s National Institute of Dental and Craniofacial Research (NIDCR), part of the National Institutes of Health, says it has developed an ultra-sensitive optical protein sensor. The sensor is the basis for much-anticipated saliva-based testing for early detection of cancer and other diseases.
The researchers' sensor can be integrated into a specially designed lab-on-a-chip programmed to bind a specific protein of interest, and generate a sustained fluorescent signal as the molecules attach. A microscope can then read the intensity of the fluorescent light -- a measure of the protein's cumulative concentration in the saliva sample – allowing scientists to gauge whether it corresponds with levels linked to developing disease.
In their initial experiments, the scientists primed the optical protein sensor to detect the IL-8 protein, which at higher than normal concentration in saliva is linked to oral cancer. Using saliva samples from 20 people -- half healthy, the others diagnosed with oral cancer -- the sensor correctly distinguished in all cases between health and disease.
The sensor achieved a limit of detection for IL-8 that is roughly 100 times more sensitive than today's blood-based Enzyme-Linked ImmunoSorbent Assay (ELISA) tests, the standard technique to measure protein in bodily fluid. The limit of detection, or LOD, refers to a sensor's ability to distinguish the lowest concentration of a protein or other target molecule apart from competing background signals.
Chih-Ming Ho, Ph.D., a scientist at the University of California at Los Angeles, is senior author of a paper describing the research that appears in the August issue of the journal Biosensors and Bioelectronics. His group's first step in widening the LOD window was to restructure the initiation of the fluorescent signal. They directly labeled the sensor-bound IL-8 with fluorescent probes, thereby cutting out the common intermediate step of using enzymes to amplify the signal. This streamlining improved the LOD of their saliva test to a level comparable with a standard ELISA blood test.
But Ho and colleagues decided to push the limit of detection further. Saliva naturally contains much lower concentrations of protein than blood, and they wanted their sensor to attain the extremely high sensitivity that some future salivary diagnostic tests will likely require. Sensitivity refers to the smallest amount of a substance, such as a protein, that a diagnostic test can detect, which Ho said he hoped to extend down to the femtomolar range, or six orders of magnitude less than one atom per cell.
To increase the sensitivity, and thus extend the lower limit of the LOD, Ho and colleagues sought novel ways to turn down the noise. Noise refers to the various ambient molecules in the saliva sample that typically stray to the sensor and bind. This creates the visual equivalent of static that heightens the intensity of the fluorescence and can lead to false positive diagnoses.
"When we talk about pushing a test's limit of detection, or LOD, we're referring to the signal to noise ratio," said Ho. "It's really a matter of figuring out how to reduce the background noise and make the signal stand above the noise."
Leyla Sabet, a member of Dr. Ho's lab and a lead author on the paper, explained that the group already filtered out other wavelengths of light that might pollute the signal. That left them to parse the fundamental and often overlooked subject of where to collect the light. Does the signal-to-noise ratio vary within or above the fluorescent pathway of light? If so, is there a precise spot that offers the highest signal and the least noise?
But to answer the where question, the group first needed a better optical tool to collect the light and see what's there. They utilized a confocal microscope, which gathers all of the fluorescence and has the added advantage of locking onto a single slice, or horizontal plane, of light and then viewing it from multiple points.
Sabet and colleagues began by locking the focus of their confocal microscope on the surface of the sample, where signal and noise typically coalesce. From there, they moved up from the surface one micron at a time, collecting the light and calculating the signal to noise ratio at each point.
"We identified a location that has the maximal signal-to-noise ratio," said Sabet. "By focusing on this signal-rich point of light, we extended the LOD by two orders of magnitude."
Winny Tan, Ph.D., also a lead author and a lab member, said the proof-of-principle tests of the sensor currently take between 30 minutes to an hour to complete. But she noted that this figure is a bit misleading. "About 90 percent of our time was spent in sample preparation, not actually performing the assay," she said. "With further integration and automation of the test, the time could be reduced significantly."
The laboratory already has developed a saliva-based electrochemical sensor, which binds the protein of interest using an electrical sensor system. Dr. Ho said the optical and electrochemical sensors, like all technologies, have their pros and cons.
"The optical sensor requires a more expensive set up because of the confocal microscope," said Ho. "So, in a small dental or doctor's office, the electrochemical sensor generally would be easier and cheaper to use. But to really push down the signal to noise ratio, the optical sensor has the advantage."
Ho said the optical sensor might be better suited for use in a more specialized central laboratory. "But the technology is advancing so rapidly, it's difficult to predict how the optical sensor might be used in the years to come," said Ho. "At this point, it has certainly pushed the envelope for the limit of detection and this will be an important capability in advancing salivary diagnostics."
The National Institute of Dental and Craniofacial Research (NIDCR) is the U.S.'s leading funder of research on oral, dental, and craniofacial health.