HAMBURG, Germany - Last week, the German chemical industry association DECHEMA e.V., of Frankfurt, held a two-day conference about the state of German DNA chip technology. The conference was attended by more than 300 participants from Germany's biotechnology industry, pharmaceutical companies and university research groups.

While in the U.S. the development of biochips has been funded by the government's Advanced Technology Program since the early 1990s, the Europeans discovered the application and market potential of this technology only lately, Hans-Peter Saluz, head of the molecular biology department at the Hans Knvll Institute, of Jena, Germany, said at the conference. "So, the U.S. is ahead, with many patents and a lot of companies," he said, but only five have been able to enter the market with biochips. "Due to new funding programs and an enormous potential in engineering skills, pharmacy and chemistry, Europe is catching up now."

Jvrg Hoheisel, a specialist in DNA chip technology at the German Cancer Research Center DKFZ, of Heidelberg, said some "really good technologies have been established with proof of principle during the last two years" in Germany. The main problem of all DNA chips, he added, is lack of reproducibility, so quality control is very expensive and routine medical application impossible.

"And as there is no really broad and deep application and almost no market yet, there is plenty of room for innovative ideas still," he said. "But once the technology is mature, the market will be really huge."

Development of hardware is the main focus of biochip developers in Germany. Some research groups are trying to improve photolithography methods to synthesize oligonucleotides on a chip, and hope to sell their inventions to Affymetrix Inc., of Santa Clara, Calif., the leading company in this field. Some try to build chips with peptide nucleic acids, which are able to hybridize large sample molecules, while others use fluorescence-marked microbeads for hybridization, which can be identified in a flow cytometer.

While most DNA chip producers use optical detection systems, the Fraunhofer Institute of Silicium Technology (ISIT), of Itzehoe, Germany, has produced prototypes of chips with electrical detection of hybridization or binding events. Hybridization changes the electrical impedance that can be measured at small electrodes; if the DNA bound to the chip is marked with enzymes producing redox molecules, even a quantitative measurement is possible. "The advantage of this detection method," said Rainer Hintsche, senior researcher at ISIT, "is that it is very simple. Technologically, it is very close to microelectronics, so it should be cheap to produce and to handle. It works even in turbid solutions and suspensions." ISIT has filed 11 patents covering various aspects of this technology.

An entirely different approach is followed by Sequenom, with sites in Hamburg and San Diego. The company already markets a technology to scan an array of samples on a chip format by mass spectroscopy. The method is suitable for sequencing and expression analysis, as well as for mutation and allele detection, and even allows the identification of compound heterozygotes. "Many analytes can be processed in a single spectrum," said Charles Cantor, chief scientific officer of Sequenom, "and with mass spectroscopy, you have an in-built quality control of the entire process."

GPC AG, of Martinsried, is interested in chips suitable for gene expression analysis. Sebastian Meier-Ewert, vice president of biology, said these tools were an important and fast instrument for the analysis of gene expression patterns. "One should not confuse such data with knowledge about status and metabolism of a cell," however, he warned.

The pharmaceutical industry's needs were outlined by Peter Spreyer, of Bayer AG in Leverkusen, Germany, and Jutta Heim, of Novartis AG in Basel, Switzerland.

Spreyer said there is growing demand for parallel and automated methods of screening libraries of compounds, cDNAs, etc. "We need cheap, high-quality chips from mass production and excellent software for the analysis of hundreds of thousands of data, and a profiling service would be a good business idea," he said.

Heim described the parallel analysis of expression data as performed by DNA microchips as "indispensable" in pharmacogenomics and a great help in oncology. "We hope for marker genes, filtered from tens of thousands of genes and represented on an initial chip, which could aid the clinician to monitor the patient's reaction to a drug at a very early stage," he said. "There is justified hope that this will accelerate the development of cheaper and more efficient drugs."