for Cell Control
University of California,
Name: Chih-Ming Ho
Position: Director, Center for Cell Control, University of California, Los Angeles, 2006 to present; director, Institute for Cell Mimetic Space Exploration, UCLA, 2002 to present; Ben Rich-Lockheed Martin Professor, UCLA, 1996 to present; professor, mechanical and aerospace engineering, UCLA, 1991 to present.
Background: Associate vice chancellor for research, UCLA 2001 to 2005; director, Center for Micro Systems, UCLA, 1993-1999
Many diagnostic tests use blood as the matrix because of the high concentration of proteins found in the fluid, which in turn increases the chance of finding biomarkers.
However, saliva is more readily available, which makes it attractive as a potential matrix for diagnostic tests. Yet standard ELISAs have had trouble detecting biomarkers in saliva.
A team of researchers at the University of California, Los Angeles, has devised an optical protein sensor it says is significantly more sensitive than ELISAs and so, able to detect biomarkers in saliva that ELISAs cannot.
In an article in the August issue of Biosensors and Bioelectronics, the research team describes work they did using the sensor to detect Interleukin-8 protein, an oral cancer marker. In 20 samples, evenly split between healthy and cancerous, the sensor correctly distinguished cancerous cases from healthy ones.
ProteoMonitor spoke with Chih-Ming Ho, the senior author on the article, this week, about the sensor. Below is an edited version of the conversation.
How is detecting biomarkers in saliva different from detecting them in plasma? What kind of requirement is needed for any device which wouldn’t be for blood-based work?
This project was supported by the dental institute inside [the National Institutes of Health]. I collaborated with a faculty member in the UCLA Dental School, David Wong.
When you go see a doctor, they always ask for blood or urine, [but] nobody ever asks for saliva. But saliva is the most non-intrusive body fluid and it’s easy to get. … And saliva is really blood going through three filters around the mouth … so the concentration [of biomarkers] is very low.
We have developed a very, very sensitive biomarker sensor. That’s how we took the project that used our ultra-sensitive nucleic acid sensor to try to detect biomarkers inside saliva. [We ended up getting] very, very low-concentration detection.
At this stage, we are able to use our sensors to sort out three important markers for oral cancer.
What was the main bottleneck you were trying to get tackle? Were you trying to amplify these low-abundance proteins?
No, that’s an interesting question. For nucleic acids, there’s a very famous and very useful instrument which can amplify the RNA or DNA — that is PCR. … And that has been a powerful instrument [for amplifying] low-abundance markers.
But in our case, our sensor is so sensitive that we can directly detect biomarkers in body fluids without going through amplification. That saves a lot of effort and time and cost.
Was the main bottleneck you were addressing the high background noise?
The noise is the key issue because basically when we say [something has been detected] the signal is above the noise. What we can do is … modify the sensor surface properties or use a confocal microscope as we did in this paper, such that we can isolate most of the noise out of the detection, so that the signal can stand above the noise.
Why are ELISAs not optimal for detecting biomarkers in saliva?
ELISA is a very commonly used technology for detecting proteins, and it’s very good. But what we do is use the confocal microscope. … If you look from the top at the signal generated from the sensor surface or above the surface and you collect all the signals and include the noise … [with] the confocal microscope, we would slice … and we can move from the sample surface all the way above the surface.
In this way, we looked only at the light generated within a thin slice. And most of the noise is generally at the surface, so that’s why we found optimal location slightly above the surface, which has very high signal, very low noise, such that we can reduce the noise, and increase the signal-to-noise ratio. And that is how we can detect at such a high sensitivity.
With ELISAs you detect all the signals from the surface reaching to slightly above the surface, and we looked at only a very thin slice slightly above the surface.
What about protein arrays? How does your sensor compare to them?
Actually we also do protein arrays. There are two different types of protein arrays — one with 10 or 20 [proteins], another with 10,000 [proteins]. The reason we do small numbers is most [of the] time when we do disease diagnosis, we really don’t need a large number of biomarkers. Say 10 or 15 is the highest number that we need to operate [with].
By doing that, we can do a small array, we can use this confocal microscope and then look at each of these sensors inside this array, and then [do] very high-sensitivity detection.
This is more consuming than just collecting the light, because when you collect the light, you can just take a picture, but now we have to go to each spot and then find the optimal location.
You say in your article that you did not use enzymatic amplification. Why is that important?
When you use this amplification that means you need to spend extra time, and you look at the operating costs and you have to go through an extra step and reagents. The best method is the simplest method, so if we can eliminate one step and can get the same or better signal, everybody is looking for and working hard for that.
In our case, without the amplification, we can have the same sensitivity.
Would this device work with plasma?
Yes, it does.
Have you done any work with biomarkers in plasma?
Working with plasma, we didn’t use an optical method. We used [an] electrochemical method. In my lab we do both. It depends on the project.
With this sensor you developed, though, would it offer any advantage over electrochemical methods?
Based on the experience with saliva, we feel it should also increase the LOD [limit of detection] or reduce the noise.
Leading to or resulting in what?
The biomarker concentration in blood really depends on the type of biomarkers. Some are a lot, some are very little. Now, for very low-concentration biomarkers, we need … to use chemical amplification or use PCR amplification [to detect them].
With this [method we developed] maybe we can do this without the amplification process.
Would your sensor be too sensitive and result in high false-positives rates in plasma?
No, high false-positives or high false-negatives depend not on general noise. It depends specifically on the antibodies for the biomarkers, so that way, it’s independent of the technique.
This was designed specifically for detecting biomarkers for oral cancer. What other diseases would this sensor be applicable for?
This technique is almost a universal technique. We can use it to detect any protein or nucleic acid. And we have applied it to detect urinary tract infection, and we used it to detect genetic diseases such as Factor-5 Leiden. Now we are using it to detect biomarkers for transplant rejections.
Optimally, how would you incorporate this into a protein-based diagnostic? Are you looking at lab-on-a-chip technology?
The sensor itself is an optical sensor, so we do need optics like lens and color meters, et cetera. But when we get a sample, we need to go through sample preparation, and that sample preparation most of the time, we can do in a chip. Basically any lab-on-a-chip is a chip for sample preparation and a sensor.
Sometimes the sensor can be integrated on a chip, for example, electrochemical samples. For optical sensors, because we need optical components, we can do it, but that involves very sophisticated micro-optics operations.
How adaptable is this for a clinical setting? Are we talking about this being used by doctors or in a laboratory environment?
All the lab-on-a-chip processes, all the point-of-care processes, most of them are still in the development stage …The next stage [of this oral cancer project] is we’re going to test [this probe] in several universities. During testing, of course, we’re still going through iterations and improvement, and if the results are good, and we think they will be, the next step is we need industry to pick it up and go through the FDA approval and become a point-of-care device.
In this clinical testing step, what are you going to be testing for?
We’re going to build this device and distribute them to centers, and then they will get patients, and compare output from the patients from control subjects … and if the statistics go well, then it’s a good instrument to go further.
So no one is developing this into a diagnostic yet?
This is still in the early stage.
Has there been any commercial interest?
On this project, we do have companies working together, and …all this development will take place in different stages, and we are in the early stages.