Researchers face a difficult task as they try to realize the therapeutic potential of stem cells and neurons. To better understand how to manipulate these cells, they need to monitor the gene-expression patterns, as well as working out how these genes are controlled. But both stem cells and neurons are not easy to maintain in culture, and it is hard to introduce DNA or RNA molecules into them to target specific genes or pathways.

Nonetheless, researchers have been successful in determining the expression of thousands of genes and comparing expression patterns between different cells or cells grown under different conditions. Such work has allowed them to identify, for example, master regulators of stem-cell differentiation or neuronal survival. In addition, a wide variety of tools has been designed specifically for use in stem cells or neurons to control the expression of a gene of interest and study its function.

Solexa's 8-channel device for tag sequencing. Credit: SOLEXA

One useful technique allows scientists to identify all the genes involved in a particular process, such as the migration or differentiation of stem cells. Serial analysis of gene expression (SAGE) is an open platform for monitoring the expression patterns of thousands of transcripts in one sample and can lead to the discovery of novel genes. The technique relies on the generation of a library of short cDNA ‘tags’ each corresponding to a sequence near the 3′ end of every transcript in a cell or tissue sample. The tags are sequenced to reveal the identity and quantity of the corresponding transcripts. The National Institutes of Health's Cancer Genome Anatomy Project has generated several SAGE human embryonic stem-cell libraries and offers web-based tools to analyse the expression of genes in these libraries (see ‘Tools for expression analysis’). To aid researchers with the construction of SAGE libraries, Invitrogen of Carlsbad, California, sells the I-SAGE and I-SAGE Long kits.

A single SAGE experiment generates about 50,000 tags. But in mid-June Solexa in Hayward, California, will launch a sequence-based expression-analysis platform that can analyse more than one million cDNA tags. The Genome Analysis System uses a tag amplification step on the surface of a glass flow cell and features Solexa's Sequencing-by-Synthesis chemistry for automated sequencing. In principle, the number of tags sequenced in one experiment should be enough to provide very deep coverage of the transcripts expressed in a human cell and so should capture those expressed at very low levels, according to Solexa.

Methods such as SAGE and the ever-popular microarrays look at the complement of transcripts isolated from a population of cells. Using a different method, based on gene-trapping technology, scientists at the Salk Institute for Biological Studies in La Jolla, California, have achieved real-time monitoring of gene expression in individual, living mouse neuronal stem cells. “When you work with a therapeutically relevant cell system you want to keep it as close to its natural state as possible,” says Carrolee Barlow, who helped devise the system. Barlow, who is now at Brain Cells in San Diego, California, and her colleagues created a library of stem cells, each with a single retrovirus randomly integrated in its genome. The virus was often integrated within or near a ‘trapped’ gene, and carried with it a reporter gene that gave off a fluorescent signal when expressed. By detecting fluorescence, the researchers could correlate the expression of a trapped gene with a specific phenotype.

“By PCR or microarray you are looking at genes associated with a specific phenotype,” says Barlow. “But by analysing expression in real time you can identify the genes actually driving those phenotypic changes.”

Narrowing the search

Liquid handling for Gene Express's Standardized Expression Measurement Center technology. Credit: GENE EXPRESS

Large-scale gene-expression screens usually result in a subset of genes that warrant further analysis, with the polymerase chain reaction (PCR) being the usual choice for the first round of follow-up studies. In particular, real-time PCR allows products to be detected as they are being made, which provides a quantitative measurement of expression levels. With some platforms, it is also possible to look at four or five gene targets in a single reaction tube, increasing throughput and lowering costs. For example, QIAGEN in Venlo, the Netherlands, sells a proprietary PCR buffer solution that increases the specificity of each primer for the respective target sequence, even when several different primer sets are present in a single reaction mix. “It does not matter what primer–probe combinations you use; our product allows for specificity,” says Kenneth Dwyer, marketing manager at QIAGEN.

Seegene in Seoul, South Korea, has developed a primer based on dual specific oligonucleotides that allows the length of the primer sequence to be longer than traditional primers in real-time PCR, thereby increasing specificity. “You can use more than five sets of primers in one reaction tube and never have any problems,” says Seegene's founder and chief executive, Jong-Yoon Chun. “There is no need for optimization.” The product, called GeneXP, is sold in kits for studying the expression of specific gene families.

For research that may lead to drug development, Gene Express in Toledo, Ohio, markets a technology known as StaRT-PCR. Although similar in principle to many other quantitative PCR assays, “the key differentiation is the ability to have an internal standard”, according to the company's chief executive Gerald Vardzel. The technology relies on a standardized mix of competitive cDNAs included in all the reaction mixtures, which allows numerical values to be assigned to gene-expression levels and for comparisons to be made across the drug-development pipeline. Gene Express provides the technology as a service through its Standardized Expression Measurement Center.

SuperArray Bioscience in Frederick, Maryland, sells 96-well plates that include real-time PCR primer sets for different panels of pathway- or disease-focused genes. “Even people in the field may not know all the genes related to a particular process. We have done the work for them in identifying the genes of interest,” says Sean Yu, vice-president of operations.

The company also sells cDNA and oligonucleotide microarrays for specific sets of genes, including one containing cell-type specific markers for human embryonic stem cells and another that carries representative markers for some of the neural phenotypes. The latter can distinguish between dopaminergic neurons, glial cells and pluripotent stem cells by their gene-expression profiles in a concentration-dependent manner. “The goal was to provide a practical test for stem-cell differentiation,” says Yu.

GenHunter differential display showing changes in gene regulation. Credit: GENHUNTER

Along the same lines, OriGene in Rockville, Maryland, offers TissueScan Real-Time PCR panels to study gene expression in human and mouse tissues. Each array contains PCR-ready cDNAs normalized with beta-actin. For example, the Human Brain TissueScan array contains first-strand DNA from 24 tissues in the human brain. The company markets TissueScan Real-Time PCR disease panels for gene-expression analyses across many stages of disease. “It is prohibitively expensive to get hold of high-quality disease tissues,” says Rich Hamer, vice-president for business development at OriGene.

Comparing expression

For comparing genes expressed by two different cells, for example an embryonic stem cell and one that has begun to differentiate, the usual method is differential display. Developed in 1992, the technology works by systematically amplifying portions of mRNAs from two or more samples and resolving the amplified products by denaturing polyacrylamide gel electrophoresis (PAGE). This allows sequence information to be recovered and corresponding cDNAs to be isolated for further molecular and functional characterizations. “It is especially good for samples with limited RNA or species that have not been sequenced,” says Jonathan Meade, product manager for GenHunter in Nashville, Tennessee. GenHunter sells two kits for differential display: the RNAimage Kit uses radioactive detection, and the RNAspectra Kit uses fluorescence.

Seegene's GeneFishing DEG (differentially expressed gene) discovery kit also relies on amplification with randomly designed primers. But this method uses annealing control primers, which are longer than primers used in differential display and so provide greater annealing specificity during PCR. The resulting PCR products are sufficiently long to be detected on an agarose gel. “The big advantage is that you can use agarose gels, which are easier to prepare than PAGE,” says Chun.

In most cases, microarrays and quantitative PCR technologies cannot assign a gene-expression status to a specific cell type. A better choice for this purpose is fluorescent in situ hybridization (FISH). This uses fluorescently labelled DNA oligonucleotide (or RNA) probes to visualize the expression of genes in a single cell with a microscope. Gene expression measured by FISH can be combined with other visual properties of the cell in a high-throughput, automated fashion (see ‘High-content screening’).

Aureon Laboratories in Yonkers, New York, has developed a method called peT-FISH (paraffin embedded tissue FISH) for detecting signals from nascent RNA molecules localized at the transcription site of genes being expressed. “We are looking at the presence of nascent RNA that correlates with early events of gene regulation in response to stimuli,” says Paola Capodieci, one of Aureon's scientists. “The transcript is still in the nucleus and we don't know if it will become mRNA or protein. So we can see what happens at the beginning of the expression chain.” So far, the method has detected up to five genes in a single cell. One of the advantages of this technology is that tissue is not destroyed. “We preserve the tissue morphology,” says Capodieci.

Homing in on individual genes

To study the function of a gene, researchers typically alter its expression, either overexpressing it or knocking it out. A popular way to turn down the expression of specific genes is through RNA interference. This uses synthetic double-stranded RNA oligonucleotides known as small interfering RNAs (siRNAs), short-hairpin RNAs (shRNAs), which are expressed from a polIII promoter on a plasmid vector, or microRNAs (see ‘A micro perspective on stem cells’). But slow-growing stem cells and neuronal cells are difficult to transfect with these molecules.

One way to introduce plasmids or siRNAs into cells is to use lipid-based transfection products. These work by forming a complex with DNA or RNA that interacts with the cell membrane. Invitrogen's Lipofectamine 2000 works with many mammalian cells, and can introduce an siRNA corresponding to the transcription factor Oct-4 in stem cells. “We got about 80% delivery,” says Peter Welch, director for research and development.

Several companies have developed transfection reagents to introduce plasmids and siRNAs specifically into cells that are hard to transfect. Mirus Bio in Madison, Wisconsin, has the TransIT LT reagent for neuronal cell lines, which is a combination of an endogenous cellular protein, histone H1 and lipoamine. “The histone H1 is the primary cationic carrier and reduces the amount of lipid needed, reducing toxicity,” says James Hagstrom, vice-president for scientific operations at Mirus.

This year, Panomics in Fremont, California, will launch a peptide-based reagent called DeliverX, which it claims will have high efficiency and low toxicity. “It has been validated with one primary cell and this summer more primary and suspension cells will be added,” says Ian Ley, the company's vice-president for marketing.

When transfection agents fail, electroporation is the brute-force approach for delivering genetic material inside cells. The nucleofector technology marketed by Amaxa Biosystems of Cologne, Germany, is aimed at cell lines that are difficult to transfect. A series of gentle pulses allows transfected DNA directly to enter the nucleus. “Other non-viral transfection methods rely on cell division for the transfer of DNA into the nucleus. Nucleofection provides the ability to transfect even non-dividing cells such as neurons and resting blood cells,” says spokeswoman Kimberly Stevenson. Depending on the cell type, a researcher would choose one of many proprietary nucleofector solutions and program the machine with the appropriate parameters. For experiments that require stably transfected cells or in vivo work, viral vectors are the obvious choice (see ‘Hitching a ride on a virus’).

DNA delivery directly into the nucleus visualized with different fluorescent labels. Credit: AMAXA

Regardless of the method of gene expression used, some steps in the process will probably require growing cells in culture, using formulated liquid media supplemented with growth factors and other substances that promote cellular replication and govern differentiation. The conditions and the reagents used may affect the gene-expression programmes of a cell, a concern that is particularly relevant to stem-cell research. “For the stem-cell field to move forward it is necessary to have standardized reagents and assays to allow results to be compared,” says Sharon Louis, senior scientist at StemCell Technologies in Vancouver, Canada. “Where we see ourselves is providing standardized tissue-culture reagents.” The media reagents provided by the company come with detailed protocols for either maintaining cells in an undifferentiated pluripotent state or for inducing differentiation towards a specific lineage.

Each in their own way, stem cells and neuronal cells present problems for the scientists who want to manipulate them. But many scientists and companies have risen to this challenge by developing increasingly sophisticated tools and technologies. With this tool kit in hand, studying gene expression in stem and neuronal cells is a realistic goal that has been embraced by researchers with an enthusiasm matching the promises of their field.