Skip to main content
Premium Trial:

Request an Annual Quote

Genisphere Execs Discuss Labeling of Arrays in Three Dimensions



Genisphere: Founded by Thor Nilsen in late 1980s as Polyprobe. Now a subsidiary of Datascope of Montvale, NJ. Has products for array detection, FISH probes, and blot probes.


Robert Getz, manager of R&D at Genisphere. Joined Genisphere after a postdoc at Lankenau Medical Research Center in Wynnewood, Penn. PhD in biochemistry from Lehigh University.


Jim Kadushin, manager of operations at Genisphere. Joined the company’s predecessor Polyprobe in February 1997. Previously at diagnostics firm Gen Trak for eight years. BS in zoology from Michigan State University.

Q: Genisphere is known in the microarray world for its dendrimer labeling technology. Could you explain exactly what the dendrimers are?

Kadushin: We think of a dendrimer as a big ball of DNA. The center is double-stranded and covalently bound together crosslinked strands. We start with 116 single-stranded bases, and with seven individual strands combined in special ways so that you have a 50-base region in the center and on either end what we call arms that are non-complementary. Imagine looking at an X with the middle fused together. The seven different strands are made from a proprietary complex and have no secondary structure. They are abiotic; they can’t bind to themselves.

We take these monomeric hybrid structures and combine them to get a latticework structure built up in layers. We are selling two types of dendrimers, four-layer and two-layer. Inside there is double-stranded DNA, all crosslinked. It never comes apart. You can boil it for a half hour under denaturing conditions. On the outside of the structure are 30-base sequences capable of binding other molecules. These single-stranded sequences are used to hybridize the fluorophores to. For four layers you can put on 375 fluors, and for two layers 44 fluors. The key to our technology in general is we have this delivery system that delivers a specific number of labels, whether fluors or biotins or enzymes. We have a small, almost nonexistent variability from lot to lot.

Getz: How I describe dendrimers to my wife is that they are like a tumbleweed. If you bump into it, it’s prickly. It causes damage. The inside is a core, inert, a lattice structure to hold everything together. The outer surface is what does all the work. It collects hundreds of fluorescent dyes and gets specified for targeting directed sequences.

There are competing methodologies that incorporate a modified oligonucleotide into a cDNA probe or target. You generate a population of probes that has anything from two fluors per probe to 50 fluors per probe. With the dendrimer system you don’t have to incorporate these modified oligonucleotides. The [target DNA] gets to work with the system that it’s used to, a normal oligonucleotide. And you deliver 375 fluors per cDNA.


Q: So it’s a lot stronger signal.

Getz: Traditionally, with microarrays, you see a tenfold improvement on average. Some messages [amplify] 50-fold, others fourfold. With direct labeling, you often use 20 and 50 µg of total RNA. With Genisphere you use two and four. It is an enabling technology. People who have less RNA to work with get the same results. And if you are looking for those low-level bottom dwellers, you can still use something like 20 µg of total RNA, and can get those.


Q: Users of microarrays have problems with expression levels on the high end of the dynamic range because the signal gets saturated. With a technology like dendrimers, I would imagine that saturation point would be reached at even lower levels. How do you recommend handling this issue?

Getz: There are two ways of dealing with that. No one knows their samples better than the scientists that use them. They know the saturation signals. And they can put those measurements that they want to measure in range. They can opt for the amount of RNA they want to use to get the low-level expressors and high-level expressors within the range. The readers and scanners of microarrays are designed to be adjustable and they tell you that there are five decades of dynamic range that you can use with the instrument. If you use a microarray you get three to three and a half decades, but with a background around 100 you lose two decades because of the of background. How you can fix this with the scanner is to do a low-level scan that leaves the high [expressors] in, and then do a high-level scan that catches the low-level expressors.

Q: I understand that dendrimer labels were originally not meant for microarrays. How did they become an array labeling technology?

Getz: Initially they were for mostly academic research labs for blot assays. A customer interested in using a dendrimer would identify a number of oligos and wanted to use it as a probe. Robin Stears [now chief microarray scientist at Telechem] was working in Steve Gullan’s [Harvard Medical School] lab on microarrays. She purchased one of our products for blots and wanted to use them on microarrays. She called up the technical support line and went through the conversions to calculate how much would be needed per microarray. She and I ended up talking and started to redesign the reagents to better meet the needs of microarrays. We switched them around so that we could make [dendrimer technology] more reproducible for her.


Q: In the past year, you have extended your microarray offerings from just labeling to a full-service microarraying facility where you do the arrays for customers. How did you get into this area of the business?

Kadushin: What led us into the service business is that a year and a half ago or so we started getting a lot of contacts from customers that were very interested in our technology or just interested in arrays. And we started becoming somewhat experts in our ability to run arrays, and how they work and do not work. Customers were calling us up and saying “we’re not interested in running the arrays ourselves.” Then we would run the array for a customer and the customer would give it back. It dawned on us that there was a nice significant market for the service business for arrays, as well as other types of assays. We decided just about a year ago that this is something that we wanted to get into, and we put together a business model that seemed to make sense. We brought in a couple of people and began really putting together the systems required for performing the service. We didn’t start offering the service until late January, and since then it has been growing. Frankly, now we are close to the saturation point. We are adding more staff and more equipment.


Q: Why do you think your microarray service business has been so successful?

Kadushin: Interestingly, the startup of our business coincided with Incyte’s shutting down their business. A lot of Incyte customers went over to us. Incyte, from what I could tell, was doing a pretty good job, but it didn’t fall that much within their business model. In terms of additional technology, we can offer customers [hybridization of] very small quantities of RNA that they cannot do with their home-grown kits. Secondly, we’ve become very good at handling problematic samples, including samples that are partially degraded and need to be purified. Three, we have flexibility in the arrays we can use. Typically, other service labs only offer one [brand of] arrays, whereas we have available to use a number of different kinds of arrays as long as they meet the criteria [for our capabilities] and are an approved vendor.


Q: You offer services for arrays from MWG Biotech and Agilent at a discount over their list price. Genomic Solutions, Takara/Pan Vera, and NEN Perkin Elmer are also listed on your web site as approved vendors. This is quite a list. Why not offer Affymetrix arrays as well?

Getz: At this time we don’t offer Affymetrix. We are primarily considering bringing in an Affymetrix system. We could use the Affymetrix system with the dendrimer labels, but at this point in time there are no protocols available to the public to merge the two technologies. That [question] has been asked of us many times, and we are considering ways of doing that.


Q: Who are your customers in the service business?

Kadushin: Academics, biotechs, and small-to-midsize pharmas. Large pharmas have their own microarray services facilities. Small-to-mid-size pharmas don’t want to spend the money on this. Our service business offers a great deal of flexibility, not only in the arrays. We also offer [sample QC] testing in a modulated format. The customer can send us a raw sample, and we will go through a number of QC tests. If we find that a cDNA sample failed, we don’t waste an array that costs hundreds to thousands of dollars.


Q: Do you test the RNA on an Agilent Bioanalyzer?

Getz: At this time we run RNA gels. We clearly understand what is good vs. bad RNA preparation. We also do ODS and OD ratios. We have tested the Agilent Bioanalyzer. It is really a wonderful machine.


Q: A final question. With the market heading toward protein chips, are you planning to get into this field?

Kadushin: That is definitely an active project that we are currently undertaking. We are working to do a couple of things, establish a collaboration with a company in the proteomics field, and trying to determine because of the newness of that field what is the best [path] for Genisphere to take.

The Scan

Study Links Genetic Risk for ADHD With Alzheimer's Disease

A higher polygenic risk score for attention-deficit/hyperactivity disorder is also linked to cognitive decline and Alzheimer's disease, a new study in Molecular Psychiatry finds.

Study Offers Insights Into Role of Structural Variants in Cancer

A new study in Nature using cell lines shows that structural variants can enable oncogene activation.

Computer Model Uses Genetics, Health Data to Predict Mental Disorders

A new model in JAMA Psychiatry finds combining genetic and health record data can predict a mental disorder diagnosis before one is made clinically.

Study Tracks Off-Target Gene Edits Linked to Epigenetic Features

Using machine learning, researchers characterize in BMC Genomics the potential off-target effects of 19 computed or experimentally determined epigenetic features during CRISPR-Cas9 editing.