Hi, everyone,
I am forwarding the compiled message from the responses I have received
regarding the question of protein-chip technology posted at ABRF boards. I
also have stripped the identities of the authors at their requests.
Kate Zhang
To: Recipients of ABRF List <abrf@aecom.yu.edu>
cc: "Vanpatten, Scott" <Scott.Vanpatten@Genzyme.com>
Subject: Protein-chip technology
Hello, everyone,
Has anybody ever look into the new proteinchip technology, for example,
Ciphergen? How does it compare with the conventional 2D gel plus mass spec.
approach in real sample for protein discovery? As far as I can tell, I am
not at all convinced by what they claim for protein identification after
protein capture? First they will identify the protein from accurate
molecular weight determination-----I am not sure. Then , they will do all
kinds of enzymatic digestion or other treatment such as deglycosylation on
the spot, and run MALDI for protein identification---I am not sure either.
They even claim for quantitative analysis????
Can anyone give me some education on this? Thanks in advance.
Kate Zhang
Xiaokui Kate Zhang, Ph.D
Genzyme Corp.
1 Mountain Rd.
Framingham, MA 01701
Tel: 508-270-2294
Fax: 508-872-9080
The following is my compilation of the responses I have received about
protein-chip technology question I posted on ABRF newgroup as above.
------------------------------------------------------------------------
Hello Kate,
I think your uncertainty about Ciphergen is valid.
Some research groups here evaluated the Ciphergen system in 1998; it's my
understanding that nothing valuable for the research projects here was
gained from several months of using the system. When Ciphergen came here,
many of the researchers were very excited about the system. However, I work
for a mass spectrometrist and he felt that this system could not live up to
its claims and he was proved correct. I went to a talk they gave here and
was unimpressed and also a bit disgusted--it didn't sound like what I call
"good science". I recall sitting in the back of the room muttering
"bullshit" under my breath. There are some problems with what Ciphergen
claims.
1) The mass spec they are using is not a high-end mass spec like something
from PerSeptive Biosystems or Bruker. The one in use here in 1998 did not
use delayed-extraction and did not have good mass accuracy.
2) At the talk I attended, very little data was presented. The speaker made
lots of claims but did not show supporting data (thus my "bullshit"
comments).
3) Ciphergen claims that their system will make 2D gels obsolete. This is
really ludicrous when you think about what people who are considered experts
in proteomics (Large Scale Biology is just one example) are doing--2D gel
electrophoresis is a core technology in those labs, it is not being done
away with. If Ciphergen was so great, would not all these experts being
using their system?
While other labs evaluated the Ciphergen system, our lab chose to use
tried-and-true technologies when we set out to do proteomics in 1998. We
have identified potential disease markers (an application touted for the
Ciphergen system), among other things, using traditional proteomics
techniques (2D gels, in-gel spot digests, peptide mass fingerprinting by
MALDI-TOF, and tandem MS by ESI-ion trap). It wasn't trivial setting all of
this up but we did accomplish what we set out to do.
The hype that Ciphergen uses to push their technology really turned my boss
and I off. What they call a "protein chip" is not at all analogous to a
gene chip--it's a metal (steel?) slide with 10 spots on it. Calling it a
chip is a stretch. They may have expanded these since the system was
installed here (i.e., more spots), but it's not really fair to call these
things "chips". Have a look at their website and you can see what they are
calling a "chip". Speaking of their website--it's been considerably updated
since I visited it last. If you really want to get a handle on what they
are doing, read some of the case studies. I looked at one on markers for
schizophrenia. It's a very slick presentation but notice that while the
slides seem to show the presence of schizophrenia biomarkers, none of the
peaks are identified. It all looks lovely, but what has the Ciphergen
system done that could not also be done on 2D gels?
For more information on what's hot in proteomics, look at the news section
of Science, 24 March issue (I'm not sure of the date--it's the Drosophila
gene issue). Evidently now that the human genome has been almost sequenced,
Celera is getting into proteomics in a big way--to the tune of close to a
billion dollars. Notice that of the technologies which Celera is interested
in pursuing as they dive into proteomics, there is no mention of Ciphergen.
Having said all that, I think the Ciphergen technology has some potential
for specific experiments where a lot is known about a given protein that
someone may be trying to look at in a moderately complex mixture. As a
general proteomics technique, however, it's my opinion that the Ciphergen
system is not a good option.
Regards,
XXXXXX
-----Original Message-----
From: Anthony Yeung
Sent: Thursday, April 20, 2000 7:52 PM
To: Zhang, Kate
Subject: Re: Protein-chip technology
I have seen one such unit in xxxx University. The MS does not look
like it has much resolution, no reflectron either. I do not believe they
have gotten much useful data from it. It does not even have a camera for
you to aim for good crystals.
Tony
>
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
I would also be somewhat hesitant with
Ciphergen's claims. While the principals are sound, ie, accurate mass,
digestion, etc., I find it hard to believe that you can do it all in one
on-line box. If that's the case, there alot of bright people wasting money
on Q-Tofs and protein purification. I think other people at NIH looked into
Ciphergen
awhile back, although I could be mistaken. I think the best current
approach to de novo protein id is chromatography (LC, gel or other) and
accurate mass and MS/MS or MSn data. On the chip front, Randy Nelson also
has some relatively mature technology if you've already got a MALDI.
Best Regards,
Jack
~~~~~~~~~~~~~~~~
Jack Simpson, Ph.D.
Laboratory Director
Center for Medical and Molecular Genetics
Armed Forces Institute of Pathology
1413 Research Boulevard
Rockville, MD 20850
Email: simpsonjt@afip.osd.mil
Hello Kate,
we have a Ciphergen machine. If you give me a call I'll be happy to
discuss its advantages and disadvantages and present other options as well.
xxxxxxx
Summary of our phone conversation:
We have purchased the system for about one yr and half. So far, we can not
say that it has generated any useful information. The technology is designed
to pull out all receptor-protein binding pair, but it is like all the
affinity selection technique, the protein may not bind to it simply because
it does not have the right confirmation. If you use sandwich technique,
then it may work but it will more complicated.
I have more problems with mass spec. technology. They don't even have
state-of-art mass spectrometer, the ms they use is pretty much like early
90's technology. No reflector, no delay extraction. Consequently, it gives
you poor resolution and poor mass accuracy. They have coined a new term
"SELDI", but it is nothing but a cover-up.
To be honest, for the last three months or longer I don't think anybody here
has even used it.
The alternative technique is to modify your MALDI target yourself. The
company such as, "Intrinsic (?) Bioprobe Inc., Arizona,( Randy Nelson) will
apply binding materials you want to any MALDI target you have. Then you can
use your existent MALDI-TOF mass spectrometer.
If you look at the data presented by Ciphergen closely, they have already
known their protein ahead of time. It is maybe useful for confirmation, but
to identify unknown protein, it does not work.
This archive was generated by hypermail 2b29 : Thu May 18 2000 - 09:47:34 EDT