Nature Biotechnology pp 143 - 148 and pp 125 - 126

Researchers have used genetic engineering to create a new type of cell line that removes a significant barrier to the practical use of immunotherapy—the difficulty of producing sufficient numbers of immune cells capable of killing diseased tissues. Arming and training a patient's own white blood cells to more vigorously attack tumors or virus-infected cells shows promise for fighting disease. If only these white cells—or cytotoxic T cells (CTLs)—could be produced in high enough numbers, the body could be trained to seek out and destroy diseased tissues by itself. One way of doing this is to remove CTLs from a patient's blood, thereby removing the blood factors and immune-regulatory mechanisms that normally prevent them from attacking target cells. Once outside the body, CTLs can be transformed into killers, and then reintroduced into the patient's blood, where they circulate and fight disease.

Unfortunately, inducing the proliferation of CTLs in sufficient numbers outside the body has proven very tricky. This is because researchers are still learning how a second type of cells—antigen-presenting cells (APCs)—"present" tumor antigens to CTLs and trigger their activation and multiplication. Ideally, APCs are also taken from a patient, but this has proved problematic because they are difficult to obtain and very variable in their activity. Now, Carl June and colleagues think that they might have come up with an alternative. They have engineered a cell line to express a set of stimulatory signals that mimics APCs and appears effective in producing large numbers of CTLs capable of killing diseased cells.

As a starting strategy, the researchers engineered a human cell line to express two types of antibody receptor (the human Fc? receptor and CD32), which were then combined with antibodies previously shown to promote the growth of another type of white blood cell (helper T cells). They also engineered another molecule (4-1BB ligand) found on natural APCs to see if it could promote the long-term growth of CTLs outside of the body. What they then found was that the resulting cell line could be used to stimulate long-term expansion of CTLs, while maintaining their ability to kill cells infected with virus. Moreover, CTLs proliferated in sufficient numbers to make transfer back to patients feasible.

The authors believe the cell line represents a more reliable and defined alternative to APCs obtained from the patient. It promises to make more practical clinical applications of CTLs against tumors and viruses.

Nature Biotechnology pp 149 - 154

In healthy individuals, the body recognizes its own proteins as "self" and so prevents the immune system from attacking its own cells. This self-protective aspect of the immune response becomes a problem, however, in cancer immunotherapy because the immune system develops self-tolerance toward the cancer cell proteins (called cancer antigens) and is unable to mount a robust immune response against them. In this issue, Rong-Fu Wang and Helen Wang propose a way to get around this self-tolerance by using dendritic cells to boost an otherwise poor immune response against implanted tumors in mice.

Dendritic cells internalize various proteins, chop them up into tiny bits, and "present" them to the cells that actually carry out the cell-mediated immune response—that is, the T cells. Efforts have been made in recent years to use dendritic cells pretreated with cancer antigens as agents (called "adjuvants") to boost a typically weak immune response. The results from these studies, though promising, have not produced complete tumor regression or therapeutic benefit. Wang and Wang suggest that this limitation possibly arises from the fact that when dendritic cells are treated with peptides from cancer antigens, they are unable to "present" these peptides efficiently for long periods of time, possibly because the antigens are not taken fully into the dendritic cells, thereby dampening the immune response.

To get around this problem, they attached a "cell-penetrating peptide" to the cancer antigen, such that the whole fusion peptide can now efficiently enter the dendritic cell and be processed and presented for prolonged periods. They show that immunization of mice with this fusion peptide results in complete protection against tumor challenge as well as inhibition of pre-existing tumors.

Their approach may be generally applicable for enhancing the efficacy of dendritic cell-based peptide vaccines against cancer and potentially many other diseases.

Nature Biotechnology pp 177 - 182

By adding unnatural amino acids into proteins, scientists hope to create new versions of proteins with enhanced properties with application in medicine, industrial production, or environmental remediation. In a first step toward this goal, Ichiro Hirao and his colleagues have developed a new method for introducing unnatural amino acids into proteins. The method is based on a novel pair of molecules, synthesized in the laboratory, that can be incorporated into DNA and RNA alongside natural base pairs.

The instructions for making proteins are carried by DNA, a long chain-like molecule composed of the four chemical "bases," A, T, C, and G. DNA is transcribed into mRNA, and every three-base sequence (e.g., AUC, CAG, AGG), or "codon," corresponds to a particular amino acid, so that the order of bases in the DNA determines the order of amino acids in the resulting protein. Protein synthesis takes place in protein-making factories called ribosomes, where molecules called tRNAs attach to codons and transfer an amino acid to the growing protein.

The problem for protein engineers who want to introduce unnatural amino acids into proteins is that all 64 possible codons are already spoken for: each one will tell the cellular system to incorporate one of the 20 common, natural amino acids. Although researchers have tried various strategies to get around this problem—for example, trying four-base codons or adding hybrid tRNAs that bind codons and attach unnatural amino acids to the protein being synthesized—these methods are very inefficient and thus limited.

Hirao and his colleagues tried a different approach. Their idea was to introduce new, artificial bases that can be added to DNA's four-letter alphabet, creating new codons that could be recognized by new types of tRNA molecules. Overcoming technical challenges that undermined an earlier attempt, they used their new base pair, which they call s·y, to introduce into a protein a close chemical relative, or analog, of the amino acid tyrosine. With appropriate modifications of the system, the approach should work for a broad range of unnatural amino acids.