Nature Biotechnology pp 47 - 52 and pp 32 - 33

In a step toward making gene therapy a reality, researchers have succeeded in partially repairing a cystic fibrosis-associated gene using an approach that targets the RNA copy of the gene. The gene in question encodes a protein (known as the cystic fibrosis transmembrane conductance regulator, or CFTR) that controls the flow of chloride ions across cell membranes; this gene can be defective in people with cystic fibrosis. Working on cystic fibrosis tissue that had been transplanted into mice, the researchers have restored chloride ion conductance to 22% of that seen in normal cells-a level that should alleviate disease symptoms if it can be achieved in individuals with cystic fibrosis.

The goal of gene therapy is to substitute "normal" genes for defective genes associated with disease. Such genes give rise to faulty proteins through a series of steps, and gene therapy researchers can try to correct the genetic defect at any of these steps. In this issue, John Engelhardt and his team focused on the point at which an RNA copy of the gene is processed by a cellular machine that splices out long stretches of the molecule to create a mature RNA, which then goes on to direct the production of a protein. The researchers used a harmless adenovirus to carry the correct genetic sequence into cells, where it was spliced into the mature RNA. This "corrected" RNA was then used by the cells to synthesize a normal version of the CFTR protein.

Unlike many more traditional forms of gene therapy, in which a "normal" gene sequence is introduced into cells to counteract the deleterious effects of a defective gene, Engelhardt's approach has the advantages of both correcting the genetic defect and ensuring that the "corrected" RNA is expressed at levels appropriate to ensure attaining airway function closer to that seen in healthy individuals. Although the results are very encouraging, modifications will be required before this RNA-targeted approach can be tested on people with cystic fibrosis. For example, an alternative virus carrier will be needed, because adenovirus does not readily enter human lung cells.

Nature Biotechnology pp 64 - 69

Dendritic cells are central to the regulation, maturation, and maintenance of the body's cellular immune response against cancer. Part of their job is to activate cancer-killing T cells by "presenting" tumor-associated antigens on their surface. However, dendritic cells have proved difficult to employ in cancer vaccine therapies because, normally, they have to be isolated from individual patients, loaded with tumor-specific antigens in the laboratory, and then administered back to the same patients. In this issue, Akira Takashima and colleagues, working in mice, circumvent the need to manipulate dendritic cells outside the body through the use of rod-like implants that, after insertion into the skin, both trap skin dendritic cells and load them with tumor antigens.

They achieved this by implanting two types of rods under the skin of each mouse. The first released a substance called a chemokine, which temporarily "traps" dendritic cells by attracting them to the rod as they migrate from the (skin) epidermis to the lymph nodes. (A chemical was applied to the skin of the mice to trigger migration of the cells.) The second implanted rod released tumor-associated antigens that could be taken up by the trapped dendritic cells.

Mice that had been treated this way and were then inoculated with tumors were protected against tumor growth. Moreover, vaccinating mice that already had tumors significantly inhibited tumor growth. This anti-tumor vaccine approach could prove cost effective in dendritic cell-based cancer immunotherapies.

Nature Biotechnology pp 58 - 63 and pp 27 - 28

Despite the increasing number of genomes that have been "fully" sequenced, researchers still struggle to determine conclusively how many genes are hidden within the jungle of genetic sequence. Now, Michael Snyder and colleagues have developed an integrated approach to address this problem that should be easy to scale up for use in organisms with larger genomes. In the process, they have detected over a hundred yeast genes that had previously been overlooked.

To achieve this, the authors applied a multistep approach. First, gene sequences that were "active," and might therefore potentially code for proteins, were detected by tagging with a foreign DNA sequence (termed a Tn3 transposon) that also encoded beta-galactosidase, an enzyme that is detectable using a colored stain. Staining allowed yeast colonies containing the tagged DNA to be identified; their DNA was then sequenced, and any sequences that did not match those of known genes were selected as potential new genes. Next, to (1) check that the "new" sequences did indeed code for genes and (2) find their location on the DNA double helix's two strands (the sense, or coding, and antisense strands), RNA from the tagged genes was screened against microarrays of short DNA oligonucleotides. Independently, the "new" sequences were also screened against databases of genes from the genomes of other organisms to look for matches that could provide clues to the sequences' function.

Using this scheme, the researchers identified 137 new genes. Around 75% were shorter than the minimal length (around 300 bases) previously thought for a valid gene meaning that they would have been excluded from consideration under criteria previously used to identify genes in DNA sequence. Unexpectedly, 15 of the new genes were located within other known genes, but on the antisense strand opposite these known genes on the DNA double helix.

The researchers believe that this approach complements other gene-detection techniques, enabling gene hunters to identify elusive genes. In the four years since the yeast genome was released in 1997, functional and comparative genomics studies have identified only 65 additional genes. In their single paper, Snyder and colleagues describe over twice that number-around 2% of those previously known to exist in yeast.