HDIs and oncology’s great hope

The US Food and Drug Administration has already approved the first HDI, Merck’s Vorinostat, as a treatment for cutaneous T-cell lymphoma. Several other HDIs are in late-phase clinical trials.

But Dr Ricky Johnstone, head of the Gene Regulation Laboratory at the Peter MacCallum Cancer Research Institute’s Cancer Immunology Division, says unlocking HDIs’ full potential requires a detailed understanding of how they work, alone or in combination with current chemotherapy agents like cis-platins, taxols, and monoclonal antibodies.

Histone acetylation and deacetylation are opposing epigenetic processes with central roles in regulating gene activity. Acetylation activates genes by modifying chromatin.

Histone acetyltransferase enzymes (HATs) attach acetyl groups to lysine residues on histone proteins. The acetylated histone spools relax their hold on the DNA wound around them, allowing transcription factors to access the promoters of genes at that chromosomal locus.

Acetylated regions of chromosomes have a puffy appearance. The cell shuts down activated loci by synthesising histone deacetylase enzymes, which move into chromatin and remove the acetyl molecules from the histone spools. The deacetylated histones bind strongly to the DNA coiled around them, rendering gene promoters inaccessible to transcription complexes.

Johnstone and his colleagues are using knockout-mutant mice to explore how HDIs induce apoptosis and how cancerous cells escape cell-cycle and apoptotic checks to proliferate and survive.

RNA interference has gone from a laboratory curiosity to a powerful tool for exploring gene networks in cell culture. Short interfering RNAs (siRNAs) now make it a relatively simple matter to knock down the activity of genes of interest in normal and cancerous cell lines.

HDIs originally came to attention for their ability to induce cellular differentiation in tumour cells, but more recently have caused excitement for their potential to treat neurological disorders involving gene-transcription defects, including motor neuron disease.

Researchers from California’s Scripps Institute reported in a paper in PNAS in October that the HDI 4b ameliorates the symptoms of Huntington’s disease in transgenic mice expressing a pathological form of the human huntingtin gene, as well as restoring a more normal pattern of huntingtin expression.

It is the capacity of HDIs to augment the tumour-killing power of conventional cancer-killing agents that excites oncology researchers. As monotherapies, HDIs show their greatest potential for treating blood cancers.

Johnstone says they are very effective for the treatment of cutaneous T-cell lymphoma. While they are not quite as potent as current chemotherapeutics, they are considerably less toxic. But used in combination with more familiar therapies, they appear to work synergistically to force cancerous cells to differentiate and undergo apoptosis – programmed cell death.

Ultimately, says Johnstone, their efficacy will probably hinge on genotyping patients’ cancers to determine their vulnerability to particular combinations of HDIs and other cell-killing agents.

---PB--- Molecular targets for HDIs

The molecular targets for histone deacetylase inhibitors are 11 different enzymes, representing three different classes of deacetylases that help regulate gene activity by modifying the histone “spools” around which supercoiled nuclear DNA is wound, forming chromatin – the stuff of chromosomes.

A characteristic of some malignant cells is that they over-express histone deacetylases, shutting down genes at random, and disrupting normal patterns of gene expression. Unregulated deacetylation can repress cell-cycle genes that normally maintain a tight rein on cell growth and replication, allowing cancerous cells to proliferate uncontrollably.

Acetylases and deacetylases regulate the activity of many proteins other than histones, according to Johnstone, and deacetylation can also inactivate apoptosis networks by acting directly upon proteins whose function depends on acetylation, including tumour suppressor proteins like P53.

Knockout mutations of this key tumour-suppressor gene occur in well over 50 per cent of all tumours, but deacetylation of the P53 protein itself can block apoptosis even in the absence of disruptive mutations in the p53 gene.

HDIs neuter the ability of deacetylases to strip acetyl groups from histones by binding to the active sites in the enzymes. With normal cell-cycle checks restored, the mutant cell is able to differentiate and undergo apoptosis.

“The good thing about HDIs is that they have inherently low toxicity profiles – they are not nearly as potent as common chemotherapeutic agents, so they tend not to make patients sick,” Johnstone says.

While HDIs show some promise as monotherapies against the “liquid” cancers, lymphomas and leukaemias, they are less effective against solid tumors.

Johnstone’s research group is using mouse models of leukaemias, lymphomas and solid tumours to test combinations of HDIs and other anti-cancer agents, including targeted therapeutics like monoclonal antibodies.

“We now have mouse models of leukaemias, lymphomas and solid tumours that accurately reflect human cancers,” he says.

“For example, we are using the Eu-Myc mouse model of Burkitt’s lymphoma developed by Jerry Adams and Alan Harris at the Hall Institute in 1985 – it’s an absolutely brilliant model to work with.

“We can manipulate primary Eu-Myc lymphomas by knocking out genes or using RNAi to knock down expression of target genes. We can also over-express genes of interest, to identify apoptotic proteins and the pathways necessary for the activity of the various drugs.

“Once we know a drug is inducing apoptosis by a particular pathway in a cancerous cell line, or in vivo in mice, we can selectively knock out key genes in that pathway to see if the drug retains its activity. This clearly identifies pathways that are necessary and sufficient for that drug’s activity, or which are not required at all.”

Johnstone says that, with an understanding of how particular HDIs are inducing apoptosis, his team selects HDIs that activate apoptosis pathways complementary to those activated by conventional chemotherapy agents or monoclonal antibodies. They then experiment with combination therapies to see if they are more effective in killing cancerous cells the same chemotherapy agents or monoclonals used alone.

They work closely with Dr Miles Prince, a clinical oncologist at the Peter Mac, who is conducting early-phase clinical trials of combination therapies in human volunteers with late-stage cancers that have become resistant to monotherapies with conventional therapeutic agents.

Johnsone says one promising candidate for a combination HDI therapy is the experimental drug, ABT-737, a targeted inhibitor of the Bcl-2 family of pro-survival proteins. Some Bcl-2 proteins prevent mitochondrial membranes becoming permeable, resulting in activation of caspases in the cytoplasm, where they induce apoptosis by cleaving a host of cellular proteins proteins.

Multiple genetic pathways lead to apoptosis, and Johnstone says that combining selected HDIs with ABT-737 boosts apoptotic activity, apparently by engaging complementary apoptosis pathways.

---PB--- Combination therapy

Johnstone’s team has been trialling Merck’s newly approved HDI Vorinostat in combination with targeted monoclonal antibody therapies. Vorinostat induces apoptosis by selectively activating TRAIL (Tumour Necrosis Factor-related apoptosis-inducing ligand) receptors.

HDIs and TRAIL-R activators have different molecular targets, and different mechanisms of action, but both agents selectively induce apoptosis of tumour cells.

In a paper in Proceedings of the National Academy of Science in August, Johnstone and his Peter Mac colleagues described the results of combination HDI/TRAIL-R therapy in a mouse model of human breast cancer. The research was done in collaboration with researchers at the University of Melbourne, the National Centre for Tumour Illnesses in Heidelberg, Germany, and the Department of Immunology at Juntendo University in Japan.

Vorinostat induces the intrinsic apoptosis pathway, and by combining it with a mouse monoclonal antibody, MD5-1, which selectively activates the DR5 death receptor, they obtained a synergistic reaction that rapidly induced apoptosis in breast cancer cells in vitro and in vivo.

The in vivo activity of the combination therapy caused established tumours in mice to regress, with few signs of toxicity. Used alone, Vorinostat and the MD5-1 receptor activator had little effect on the same tumour types.

A functional analysis of changed patterns of gene activity in the tumour cells produced a surprise: Vorinostat did not enhance apoptosis by activating the TRAIL-R mediated intrinsic apoptosis pathway, but by down-regulating C-FLIP, an intracellular inhibitor of apoptosis.

“The challenge is to target the various classes and types of HDIs, because the current compounds tend to hit all or most deacetylases,” Johnstone says.

“So far there are no HDIs that specifically target individual deacetylases, so chemists are trying to develop compounds with greater specificity. It’s proving difficult because the enzyme molecules are notoriously difficult to produce as recombinant proteins, and they’re very hard to crystallise to solve the 3-D structures of their active sites.

“HDIs are very promising anti-cancer agents, but there’s a long way to go before we fully understand how they work, and have the ability to identify the right patient cohorts to treat.”

Johnstone says HDIs exhibit pleiotropic properties – for example, they enhance the immune response against tumours, as indicated by changes in cytokine profiles, and inhibit the growth of new blood vessels that nourish and oxygenate solid tumours.

“Eventually we will employ genetic profiling to characterise individual patients’ cancers, and identify patient cohorts that are more likely to respond to certain combination therapies than others – but that’s going to be difficult to do.

“Working with Miles Prince, we are obtaining samples of patients’ cancers before and during treatment with HDI combination therapies, and using gene-expression arrays to profile changing patterns of gene expression, to try to understand which genes are being affected.

“We need large cohorts of responding and non-responding patients to provide the statistical power to would allow us to be confident that combination therapies are providing consistent response signatures. But it could be a powerful way to use these therapies further down the track.”

Prince is running phase 1 and 2 clinical trials of promising HDIs in combination with other drugs, in collaboration with Johnstone and Dr Richard Bates of the National Cancer Institute in the US.

Among the combination HDI-conventional therapies Prince is trialling are Novartis’ experimental HDI LBH589, a potent anti-myeloma agent, that potentiates the activity of the frontline multiple myeloma drug, dexamethasone.

There is also Romidepsin (depsipeptide), an HDI that shows promise as a component of combination therapies for cutaneous T-cell lymphoma, and Panabinostat, an HDI that shows promise for treating blood malignancies including B- and T-cell lymphomas, myelodysplasias, and Hodgkins lymphoma.

Prince says the sheer number of novel HDIs that have appeared this decade actually set back the search for more effective cancer therapies. “We had no real idea of what the targets were, so we couldn’t build a coherent explanation around any particular malignancy,” he says.

The most promising progress is in the area haematological malignancies, including T-cell lymphoma, leukaemias, myelodysplasias, Hodgkins lymphoma, and multiple myeloma – in some cases HDI combination therapies were delivering “incredibly prolonged responses”.

These malignancies were easier to study than solid tumours because samples could be easily obtained for serial biopsies during the course of treatment.

But Prince said while phase 1 trials were delivering large amounts of important, data, and some very promising indications of efficacy, it was still too early to draw any broad conclusions about the clinical promise of HDIs.