Ohio State University
Who: Haifeng Wu
Position: Assistant professor, department of pathology, director of clinical coagulation laboratory, College of Medicine and Public Health, Ohio State University, 2001 to present.
Background: Postdoc at the University of North Carolina at Chapel Hill, 1988-1993 in biochemistry; residency in pathology, Duke University, 1997-1998; residency in pathology, College of Physicians and Surgeons of Columbia University, 1998-2000; clinical fellow, transfusion medicine, New York Blood Center, 2000-2001
Haifeng Wu is Ohio State University’s lead researcher in its collaboration with Ciphergen to develop a diagnostic for thrombotic thrombocytopenic purpura, a rare blood condition in which small clots are formed in the circulatory system, resulting in the consumption of platelets and a low platelet count.
In the current issue of Spectroscopy, Wu and his colleagues describe their work using Ciphergen’s SELDI and ProteinChip array technology to clinically diagnose and evaluate patients with TTP. According to Wu, Ciphergen is “very interested” in developing a clinical diagnostic for TTP.
Below is an edited version of a conversation ProteoMonitor had last week with Wu.
Describe your research.
I’m a pathologist and I’m a blood banker, so for my practice, I treat special patients. One group of my patients has thrombotic thrombocytopenic purpura. What I do is diagnose the patient based on a clinical presentation. And after diagnosis, we treat the patient with plasma exchange. We put a patient on a machine, take out the patient’s plasma, and replace the patient’s plasma with plasma from a blood center. It’s a very expensive procedure, and because you put patients on a machine and you expose them to a donor’s plasma, it has a lot of side effects, like allergic reaction or sometimes acute lung injury.
The disease itself is a very serious disease. In the old days, we didn’t know if plasma phoresis worked, [if] plasma exchange worked. In the last 20 years, we’ve learned plasma exchange works, so if we treat patients in time, about 90 percent of patients survive.
On the other hand, aside from mortality, the patients also have a lot of morbidity. Typically, a patient presents with clotting events. A patient’s platelets form thrombosis everywhere in the blood vessels, so patients can develop thrombotic complications like renal abnormality [and] neurological deficits.
So we are dealing with very serious clinical disorders. We need to diagnose patients in time and then treat patients correctly. Another important thing is … about 50 percent of [patients] develop relapse disease. Each time a disease relapses, it has mortality and morbidity.
Clinically, there are some other disorders that overlap with TTP, so in the old days before we had a diagnostic test, about half of patients who didn’t have TTP [were treated] as if they had TTP.
I’m a clinician, I’m treating patients, so I have been thinking about treating those patients for a long time. For the patients with true TTP, they are deficient in one enzyme, a metalloproteinase ADAMTS13. So, now the question is, ‘How do you detect ADAMTS13 activity?’
The physiological substrate of ADAMTS13 is von Willebrand factor, which is a very large protein. For a couple of years, people have been doing Western blot work to detect metalloproteinase activity. Basically, you incubate the patient’s plasma with purified von Willebrand factor which has a molecular weight greater than 1,000 kilodaltons.
It’s very tedious work to do accurate quantification work, that’s number one. Number two, the turnaround time is very long, usually it’s overnight. Here, you have a patient with full-blown disease and we need to diagnose and treat the patient immediately.
So we took advantage of the recombinant von Willebrand factor which has a histidine tag. When you incubate this recombinant von Willebrand fragment with a patient’s plasma — if a patient has metalloproteinase, if we cleave this recombinant von Willebrand factor — we generate a very small peptide with a histidine tag.
The strength of SELDI is it has a protein array. Different arrays have different surface chemistries. [With immobilized metal affinity chips] you can charge IMAC chips with specific metals. In my case, I charged IMAC chips with nickel.
And nickel will bind with histidine residue specifically. With this kind of approach and technology, we are able to detect how much histidine tag is generated after you incubate recombinant von Willebrand factor with a patient’s plasma. So the amount of histidine tag [that] binds with the IMAC chip on SELDI is exactly proportional to how much enzyme activity a patient has in patient plasma. So that’s the principle.
Have you been able to validate your methods and results?
It’s been used repeatedly in my laboratory. It’s not validated in other laboratories yet, so that’s what we’re working on. It’s been used very successfully in my laboratory, validated in the patient. For TTP, there’s no established test, so now you validate in the patient. For the patient with true TTP, the test is positive. For the patient without TTP, the test is negative.
Is this a test that can be done in a clinical setting?
Right now, it’s done in my research laboratory and whenever we develop a test, in the early phase, it’s more result-oriented. It’s not fully automated, so I think at this stage it can be done in the research lab, it can be done in the reference lab where you have more skillful technologies.
I’m in the middle of trying to do some automation, to streamline the test or to make the test more robotic. So, I think in the future, it can be done in the general clinical laboratory after the test becomes fully automated.
I can perceive that [the test] can be done in the tertiary center, in medical schools, in large hospitals with more than 500 patient-beds, in those medical centers where they have strong personnel, technical support for the laboratories.
What’s the difficulty in turning this into a point-of-care clinical test?
For point-of-care … you need very fast turnaround time. Then you need to make clinical decisions right away. For point-of-care tests, you really need a cost-effective test. This is a mass spectrometer-based test. The machine itself is somewhere between $100,000 and $200,000. I can’t imagine someone would spend that amount of money to do a point-of-care test.
What’s the turnaround time for your test?
Somewhere between three to four hours.
As opposed to what, with other tests?
As opposed to Western blot, you need at least two days. And then, the Western blot is not very quantitative. And this test is very quantitative.
Was the SELDI platform key to your research?
Right now in the clinical laboratory, the most popular and prominent technology is immunoassays or immunofluorescent assays. And the test depends on specific interactions between antigens and antibodies.
I think SELDI has very similar features. The protein arrays in SELDI contain surface chemistries, such as IMAC, that allow for selection of the analyte specifically. For example, in my assay, I incubated patient plasma with the recombinant substrate containing a histidine tag. After the reaction, I just added the plasma on the SELDI plate and applied high stringency washing conditions to remove all unbound proteins. That’s similar to what we achieve in any immune-based assays. Basically, you have a platform that washes off thousands of nonspecific plasma proteins. At the end, the specific analyte, a histidine tagged cleavage product which binds to the IMAC chips, is measured effectively.
So I think [SELDI] does have its own features. I was lucky to take advantage of it, so it worked out very well. If I had another mass spectrometer, I would not be able to develop the test.
The SELDI has had issues concerning its accuracy and reproducibility. How were you able to work around it, or was it even a concern?
I think that’s the limitation with all mass spectrometers. Basically, the mass spectrometer cannot provide accurate quantification because of intrinsic limitations that are not [specific only to] SELDI but to all mass spectrometers.
What I did was use internal controls. We added internal controls, so after reactions, we put the patient’s plasma on the SELDI protein array. We added fixed amounts of internal controls, histidine tag fragments. It has different molecular weight [than] the analytes we detected, but it has the same histidine tags.
After we added the internal controls, the analyte is quantified as the ratio of the analyte to the internal control. By doing that, we minimized the intrinsic variation with mass spectrometry.
This is not unique to SELDI, this should be the strategy for all mass spectrometry-based tests.
Who would benefit the most from any sort of test that could be developed from your research? New cases or people who may relapse?
I think both people can benefit. I think for the patient with TTP, we diagnose the patient, treat the patient. Ninety percent of the patients respond very well. For the patient without TTP but who clinically overlaps with TTP, I think it’s very important to make sure they don’t have TTP. If the patient has cancer, we treat the cancer. If the patient has HIV, we treat the HIV as opposed to treating the patient for TTP. That’s one end.
The second end, for the patient who has TTP already, we know they’re going to relapse. So I think it’s very important, once you establish diagnosis, you follow those patients longitudinally. So if they have early relapse … if you can catch the patient in early relapse, we can provide early intervention. The patient may need only very minor treatment.
Where does it go from here?
Ciphergen is very interested in developing a clinical diagnostic test. We are working on that right now, trying to provide full clinical validation for this test.