AT A GLANCE
Susanne Schroeder, senior scientist, microarray development, MWG Biotech, Ebersberg, Germany.
Presentation: “Optimizing hybridization conditions of oligo-based microarrays with regard to automation” at EuroBiochips meeting, Berlin, June 27, 2002.
MWG provides oligo synthesis and sequencing services, robots and instruments for complete lab automation, as well as microarray products and services. Has subsidiaries in the UK, France, Italy, Scandinavia, and the US, as well as 30 distributors around the world. Oligo and other nucleic acid synthesis comprise the majority of the company’s business, while microarrays are gaining in importance. The company’s array lineup includes rat, mouse, human, E. coli, other microbial catalog arrays, as well as custom microarrays. Custom arrays, which can be designed online using the company website’s CompactGeneIndex, comprise at least half of the company’s microarray sales.
At the EuroBiochips meeting in Berlin last week, Susanne Schroeder, a senior scientist at MWG Biotech, presented a précis of research on an improved way to hybridize microarrays in large volumes. The group tried a number of different hybridization buffers on MWG’s Pan-Rat 5K arrays in order to find the optimal mix of reagents.
In your presentation, you describe the optimization procedure you conducted for hybridization buffers. What led your group to do these optimization procedures?
Our main goal is the automation of the total process. In order to achieve this, you have to achieve some preconditions. One of these was the optimization of the hybridization method so that you can do hybridization in large volume. If you do the purely coverslip approach, you need up to 20 to 30 microliters per reaction. Here you need, depending on the machines, up to 200 or even more microliters. And that’s not possible in the same buffer [used before]. That was the starting point. The other is that hybridization is supposed to be more reproducible. These [factors] led us to this particular project.
You said that you tested a number of reagents as the hybridization buffer: urea and the 30 percent and 10 percent formamide, and the pure salt buffer. From what I understand now, the urea and the 10 percent formamide performed well, right?
Yes but the problem was that 10 percent formamide is less specific. And this is why we got to urea.
And you also found that agitation helped?
Yes, this is what you see in all hyb chambers that are commercially available. All do agitating in different ways. There was one talk [at the conference — given by Advalytix] about sonic waves used to pump the hybridization solution around. So agitation is really an important feature.
Also you added dextran sulfate as well..
Yes, but we don’t recommend it. We tested it because we had heard of it, and it really increases the signal intensity. But it causes a lot of problems [with increased background], so we rejected this approach.
So your answer, then, to hybridization is urea, agitation, no dextran sulfate.
Right. It’s not automated yet, but it’s on the way to automation.
Do you plan to offer this system of reagents?
We do. In the middle of May, we launched this system. We call it GeneFrames. It is a kind of chamber, a frame which you slide on the slide quickly, then you [cover with] a plastic cover slip, so you generate a kind of hybridization chamber on the slide. It [can hold] about 240 microliters.
So you basically stick the frame on the slide and inject the reagents into the frame, avoiding the air bubble effect in a cover slip or any other problems...
And you allow more diffusion than with a cover slip.
Does GeneFrames come with all microarrays you offer?
Yes. Customers obtain the old buffer and the new buffer, and the GeneFrames together with all microarrays. And if anybody wants to buy just the frames, we also sell the frames and the buffer together.
Is your plan to develop an automated system?
We are strongly moving forward to automation. A couple of suppliers offer a hybridization chamber, and we are evaluating internal and external options.
Is this product offering mainly focused toward academic customers, researchers, or pharmaceutical customers?
At the beginnig of 2001, our customers were mostly academic. In the meantime, due to the strongly expanded product portfolio, we are an accepted supplier for academia, biotech, and pharma.
Do you find your customers want new different products, or higher-density probes on an array?
The higher density surely is one point, to allow these [genomic] screening projects. That is why we started offering the human 30K array a couple of weeks ago. In this product, 30,000 human genes are spotted on three different arrays. This is surely one direction, but the following step would be smaller content arrays. If people have done their screening approach, they are more interested in specific genes in a particular type of cancer, or some other diseases.
As far as cDNAs vs. oligonucleotides, there are a lot of people who are going towards oligonucleotides. What’s your view on which one offers better results?
We sell oligo arrays, and we are really convinced of our oligo arrays. They are specific, they are reproducible. And we can do the design more specifically than with the cDNA arrays. Oligos are standardized in terms of length, in terms of the GC content and the melting point, and to me they offer much better control of these [issues] and much better flexibility.
Are your oligos long or short?
They are 50-mers. In the first place, it’s more specific [than 25-mers]. If you do a huge approach with 30,000 genes on the human genome, you have a much bigger range with 60-mers than with the 25-mers in terms of statistics. In addition, with 50-mers we can guarantee a MALDI-QC process during oligo synthesis, which is not possible with oligos longer than 50-mers.
In what direction do you see the demand for microarrays among researchers going?
It is really going in the diagnostics direction. But that’s a mid-to long term goal [for MWG].
Affymetrix is pretty dominant in the US, but in Germany, do you find that more companies want to work with a supplier that is close to home?
Obviously, we feel the influence of Affymetrix here. But we apply a different approach. We have application specialists, and we are strong in our support of