Scientists report on the development of a platform for designing and constructing dual-function protein "switches" that can both recognize a cancer-specific marker and selectively trigger only those cells carrying the marker to produce an enzyme that renders the cell susceptible to cytotoxic drugs. The Johns Hopkins University School of Medicine team used directed evolution to generate a protein comprising two separate domains, which rendered human colon and breast cancer cells carrying the cancer marker hypoxia-inducible factor 1α (HIF-1α) susceptible to the prodrug 5-fluorocytosine (5FC) through the production of cytosine deaminase.
Chapman M. Wright, Ph.D., Marc Ostermeier, Ph.D., and colleagues report on their synthetic biology approach in PNAS in a paper titled “A protein therapeutic modality founded on molecular regulation.” They hope to start animal testing within a year.
Current approaches to the development of protein therapeutics against cancer focus largely on targeting and modulating cancer markers to trigger a downstream effect, the researchers explain. Unfortunately, though, this type of approach is limited to triggering mechanisms that occur naturally as a result of modulating the cancer marker, and also has to focus on those markers for which modulators that produce a therapeutic effect can be found.
Conversely, the ability of proteins to be regulated by molecular signals at the cellular level hasn’t yet been exploited in a therapeutic setting, the researchers continue, despite the fact that “the ability to link recognition of any cancer marker with activation of any desired therapeutic function would enormously expand the number of possible protein therapeutics.”
One approach would be to build proteins that function as switches: that is, proteins whose cellular level of activity is dependent on interactions with an input signal such as a protein or small molecule. “The design challenge for this approach is how to fuse the input (i.e., signal recognizing) and output domains (i.e., the function to be modulated) such that the input domain regulates the output domain’s function.”
To try and achieve this, the Johns Hopkins team used a directed evolution approach, involving the generation and screening of libraries of genes in which one domain is randomly inserted into the second domain. Their aim was to build a single protein that could both recognize the tumor-marker HIF-1α, and then trigger HIF-1α-producing cells to generate cytosine deaminase, which catalyzes exogenously administered prodrug 5FC into the cytotoxic drug 5-fluorouracil (5FU). Cancer cells that contain high levels of the intracellular cancer-specific marker HIF-1α are able to survive under extreme hypoxic conditions, are resistant to therapy, and have a greater potential for metastasis, the authors add.
The switch design comprised the CH1 domain from the human p300 protein as the HIF-1α recognition input domain, and yeast cytosine deaminase (yCD) as the prodrug activation output domain. To generate libraries of construct candidates, the researchers constructed a library in which DNA encoding the CH1 domain was randomly inserted into a plasmid encoding yCD. Using a two-tier genetic selection approach they then isolated yCD-CH1 hybrids that were capable of converting 5FC to 5FU in cytosine deaminase-deficient E. coli cells (strain GIA39), but only in the presence of HIF-1α. For the positive selection step library plasmids were transformed into GIA39 cells that harbored a plasmid encoding a fusion protein of GST and the C-TAD domain of HIF-1α (termed gstHIF-1α) under the control of the arabinose promoter. A recombinant gene, designed Haps59, was finally selected on the basis that it conferred to E. coli the greatest HIF-1α dependence on 5FC sensitivity.
Plasmids encoding Haps59 were then transferred to GIA39 cells that also carried either plasmids harboring the HIF-1α gene or control plasmids. The cells were plated on agar containing 5FC and treated with arabinose (to induce expression of HIF-1α in appropriate cells). As expected, only those cells that carried the gstHIF-1α gene showed increased 5FC sensitivity on the addition of arabinose. Similar results were obtained when the cells were grown in a liquid medium.
Copurification experiments indicated that Haps59 interacts with gstHIF-1α in E. coli, and this interaction, the authors suggest, must allosterically activate the cytosine deaminase activity of the switch and/or cause the increased accumulation of the switch in vivo; in fact there was a substantial increase in the accumulation of Haps59 when gstHIF-1α was co-expressed in vivo. “We conclude that increased accumulation of Haps59 in the presence of gstHIF-1α is likely to be the major mechanism by which haps59 confers HIF-1α-dependent sensitivity to 5FC,” they write.
The team then moved on to test whether Haps59 would function in human cancer cells and confer sensitivity to 5FC selectively under conditions in which HIF-1α accumulates. They first chose the colorectal cancer cell line RKO, which is known to accumulate high levels of HIF-1a in hypoxic conditions, and is sensitive to 5FU in culture. Initial tests confirmed that 5FC was highly toxic to RKO cells that expressed yCD in either hypoxic culture conditions or cultured in the presence of Co2+ (which disrupts the HIF-1α degradation pathway and allows the protein to accumulate in the cytoplasm of the cell). However, neither hypoxic growth conditions nor the addition of Co2+ led to 5FC sensitivity in cells infected with an empty vector, or changed the 5FC sensitivity of cells expressing yCD. Similar results were also obtained with an MCF7 breast cancer cell line.
Conversely, and supporting the previous results with E. coli cells, Haps59 accumulated at a higher level in the presence of HIF-1α in both RKO and MCF7 cells, whereas yCD did not. “This finding further supports the hypothesis that the increase in 5FC sensitivity results from a Haps59-HIF-1α interaction that causes increased accumulation of Haps59,” the researchers state. They in addition confirmed both that Haps59-expressing RKO cells did produce 5FU, and that production increased 2.5-fold in cells exposed to Co2+. “These results suggest that Haps59 functions as designed—increasing 5FU production in cancer cells in response to HIF-1α,” they conclude.
Haps59’s therapeutic potential is derived from its unique regulatory property of establishing a direct relationship between HIF-1α levels and intracellular production of a chemotherapeutic drug, the team states. In contrast, most existing strategies for protein cancer therapeutics target more readily accessible extracellular cancer markers, even though this markedly limits the available targets.
Developing therapeutic approaches such as the protein switches against intracellular molecules are accompanied by challenges in terms of delivering the therapeutic protein inside a cell, but, as the authors point out, this can be achieved by delivering its gene. And while a major limitation to existing CD/5FC gene-directed enzyme prodrug therapies is the need to transfer the CD gene specifically to cancer cells using viral vectors, the Johns Hopkins team says “our strategy overcomes this limitation by moving the specificity from the transductional level to the protein level, allowing efficient means of gene delivery to be used regardless of cell-type specificity...In addition, our approach is complementary to both transcriptional and transductional targeting and might be combined with these approaches to afford a double or triple layer of specificity: at the gene delivery level, at the transcription level, and at the protein level.”