Director, Protein Purification Center
Name: Ghil Jona
Title: Director, Protein Purification Center, Weizmann Institute of Science
Professional Background: 2005 — present, director, Protein Purification Center, Weizmann Institute of Science, Israel; 2002 — 2005, post-doctoral associate, biochemistry and molecular genetics, Yale University.
Education: 2002 — PhD, biochemistry and molecular genetics, the Weizmann Institute of Science, Israel; 1997 — MSc, biochemistry, the Weizmann Institute of Science, Israel; 1994 — BSc, biology,Tel Aviv University, Israel.
Before there was avian flu, there was severe acute respiratory syndrome, or SARS, the alarming epidemic that between November 2002 and June 2003 infected roughly 8,400 people, and killed 813. In response to SARS, three diagnostic tests were developed; an RT-PCR assay, an enzyme-linked immunosorbent (ELISA) assay, and an immunofluorescence assay.
Now, a team of researchers from Yale University, the Beijing Genomics Institute, the Agricultural University of Hebei, China, and Toronto's Mount Sinai Hospital, have added a fourth test to the list — one that utilizes microarray technology. Published in the March 7 edition of Proceedings of the National Academy of Sciences, a paper entitled "Severe acute respiratory syndrome diagnostics using a coronavirus protein microarray" describes the development of an array that can monitor SARS as well as five additional coronaviruses and demonstrates that viral infection can be monitored for many months after infection.
Moreover, the authors contend in the paper that "protein microarrays can serve as a rapid, sensitive, and simple tool for large-scale identification of viral-specific antibodies in sera." To expand on the utility of the new SARS assay as well as its potential for use in future viral outbreaks, paper co-author Ghil Jona, an investigator from Michael Snyder's lab at Yale who recently moved back to his native country Israel, spoke with BioArray News last week.
When did you move back to Israel?
At the end of December. I'm back here and I got a nice job at the Weizmann Institute of Science in Rehovot. It's roughly 15 miles south of Tel Aviv. I am setting up a new lab, kind of a proteomics center. Essentially it's a lab for protein expression and purification, but I hope in the future it will become the proteomics center at the Weizmann Institute.
I just started a couple months ago building this platform from scratch. I basically came to an empty lab and I am equipping it and building new collections and gathering a lot of cell material from all over the world. It should come out well I hope.
But you used to work at Yale.
Yes, that's correct. I was a postdoc for almost four years in Michael Snyder's lab. And that was fascinating period of my career. Not that many people get to be in Mike's lab, but those that get there are really among the high tier of scientists.
And how was the project that resulted in the SARS paper organized? I see that you have Chinese, American, and Canadian co-authors.
Actually, the whole project was inspired by Heng Zhu. He was originally from China, and he has been in Mike's lab for almost five years and now he has a job at Johns Hopkins University.
He was in China for sometime during the SARS outbreak and thought that it would be interesting to use the microarray technology that was developed in Mike's lab to build something new; a diagnostic tool. Fortunately, since he knows a lot of people in China, he was able to talk to people from the Genomic Institute in Beijing and other colleges and gather some genetic material from the SARS virus. When he came back [to Yale] in 2004, the first thing Heng did was to clone all of the genes of the SARS virus plus other coronaviruses as well, and to express, purify, and spot them on the microarray.
At that stage we decided to probe the microarray and see if we were able to detect anything. Heng went back to China and started doing some probing there, and at the same time I took some mouse sera at Yale and we were able to test in parallel. And we were able to see that SARS patients have one characteristic type of infraction on the array and one from the mouse sera had another characteristic. And that meant that we had an assay that was specific enough to go on and try something else.
At that point we figured out that the best place to go from China was to Canada, which was where the second biggest outbreak of the SARS epidemic occurred. At the end of the day we were able to talk to Tony Mazzuli and Barbara Willey from the Mt. Sinai Hospital in Toronto. I prepared some microarrays and flew over to Toronto and spent a couple of weeks probing the arrays with different sera while Heng was doing similar work back in China. It took us probably half a year of really solid work to finish the project. We finished roughly at the end of 2004.
What are the current options for clinicians dealing with an outbreak of SARS?
In 2002, they had to identify what the source [of the outbreak] was and it took them a long period of time in different labs all over the world to identify that, first of all, they were dealing with a virus. Then it took longer for them to identify that it was a coronavirus.
Right now the biggest advantage we have is we know what an outbreak of SARS looks like, how patients react and what all the symptoms are. So, essentially, the moment a physician identifies some of the characteristics of SARS, in less than one day he will be able to identify whether a patient has SARS or not. Right now the assays that are being used are the RT-PCR assays and the ELISA tests, as well as the indirect immunofluorescence test.
But, I think the more interesting idea here will not be for SARS, but for other viruses as well. We now know SARS. But there are dozens and dozens of viruses out there that we humans have never encountered. People have come to the conclusion that SARS came from other animals, not just some random mutation. It was a mutation that made it compatible to attack human beings. My belief is that we have to be ready for other types of viruses as well that can come and cross that barrier and become something that can attack us. Today the avian flu is something that is very disturbing. Other coronaviruses may cross over someday. So our big challenge will be to come and identify those new viruses and try to deal with them.
To give us the greatest advantage, we could make a microarray of hundreds of different viruses. We can at the moment put something like 12,000 to 14,000 proteins on a microarray. We could, if we wanted, put ten proteins for each virus. They don't have to be viruses from humans. They can be other viruses, from dogs, from mice, from any animal of interest.
There are other technologies in use, like RT-PCR and ELISA. What are the advantages and disadvantages of protein array technology over existing technologies?
Nothing is perfect in our world. RT-PCR is very sensitive and it's a very fast technology. The big disadvantage of that is that after about ten days it is difficult to trace the agent that is causing symptoms. So a test that tests for antibodies, which remain for a long time after the agent has gone, has an advantage over RT-PCR technology.
The disadvantage of ELISA tests and immunofluorescence tests is that you are using a lot of material. Whereas with the microarray we are able to use as little as one microliter of serum sample in order to identify a SARS patient. This amount is one tenth to one twentieth of what is regularly being used in the other tests.
Direct immunofluorescence tests I think are the most specific tests on the market, but they lack some things. One is the speed of the experiment. A human being has to sit at the microscope and look at the cells, to decide which of the serum samples originates from a real SARS patient.
The thing with the microarray is that we have specific proteins, and each one of them is spotted in a different place. We are able to trace more specifically and look for more specific cross-reactivity as well. The more you are able to trace cross-reactivity, then you are able to take your results in a more cautious way because you have enough controls on your microarray to exclude cross-reactivity.
That's why a lot of people are doing microarrays. Because you are taking a lot of variables into account and you are reducing the noise.
One problem with our microarray is that we had a small percentage of false-negative sera. There are no two people that are exactly the same, and even though we found that the best condition to do our assay is, for example, diluting our sera 1:200. We need to do two things. First of all, we need to repeat the experiment. If you are going to deal with sera, you have to repeat every sera at least twice, to be very confident that the results are true. In certain cases you need to go and reduce the dilution factor. Because, if one patient is very allergic or has a lot of antibodies against other pathogens, they will mask the patient sera and will reduce your ability to go and identify that that patient is really infected with the target pathogen. And that is something that we are still debating how to deal with.
Can you roughly go over what was on your array?
Well we dealt with proteins from the SARS virus, as well as different portions of different proteins from the SARS virus. We also dealt with additional immune viruses — the human coronavirus 229E and the human coronavirus OC43. These are two viruses that infect humans that usually result in flu-like symptoms. It is believed that roughly 30 percent of flu cases are not really from an influenza virus, but rather from a coronavirus infection. In addition, we had murine coronavirus, a bovine coronavirus, and a feline coronavirus. We tried to tile across their genomes as much as we could, to get a better representation of the proteins that may be targets of cellular antibodies.
This came about in an interesting way. We found out that a portion of the people that we tested, some of those were SARS positive and SARS negative, we were able to see that they had antibodies against another human coronavirus — the human 229E coronavirus.
One of the reasons we decided to put so many proteins on the array was that we wanted to be sure that we didn't get cross-reactivity, and that we could get high specificity against one species over the other. In most cases, it was indeed the case. With the SARS we didn't see any cross-reactivity with the murine, bovine, or feline coronaviruses, which is very encouraging. That means that the day that one of the viruses crosses the barrier, and infects humans, we can all of sudden see that human sera has antibodies against a feline coronavirus, then we can specifically trace the event and determine that the pathogen is a feline coronavirus. That's a really big advantage of the assay.
Proteins are sometimes seen as more sensitive than cDNA arrays. Do you think the technology would hold up under widespread use outside of a research setting?
That's one of the points we were very much trying to persuade people about. The fact is that, yes [they are capable of widespread usage], but it is not black or white. One of the things that came out of my discussions with Barbara Willey is that there is no one technology that can cover everything. But rather the way things should be done is using two or three technologies in parallel to cover all the possible angles and the possibilities that you have. There is no one assay that can give you all the right answers.
RT-PCR will give you the good answer at the beginning, but only for the first week to ten days of infection. Later on you need other assays, which are currently the ELISA test, immunfluorescence, and the microarray.
How would you design your experiment using those technologies?
The way I would do the experiment is that I would first use RT-PCR to try and see whether it's positive or negative. If it's positive, then it's positive. If it's negative, you are left debating whether it's a real negative, or whether you have passed the window where you can detect the virus in a patient.
At that point I would go and do at least two or three types of assays to detect whether this is a patient or not. Then I would use an immunofluoresence and probably an ELISA test as well as a microarray. The moment that you tackle the question in the second phase with these three assays, you are better able to assure yourself of the sera's nature.