Hypoxia is tumor specific and represent a target for therapy. Up to 60 % percent of advanced solid tumors are characterized by hypoxia areas (0). Cancer cells present in these hypoxic regions are resistant to both chemotherapy and radiotherapy (1) and must be targeted by hypoxia-selective therapy. These cells are exposed to a very low oxygen tension of pO2 < 10 mmHg, equivalent to < 1.3% O2 in vitro (1).

Besides conferring resistance to cancer treatments, hypoxia also promotes tumor progression and metastasis. Hypoxia up-regulates over 80 genes associated with tumor progression, glycolysis, angiogenesis and metastasis (2) through HIF activity. Patients with a high proportion of hypoxic cell have low survival rate after surgical resection of the primary tumor (3). As 90 % of cancer deaths are attributed to metastatic spreading (3), targeting hypoxic cells could have a deep impact on reducing metastasis and the overall survival.

In order to detect this tumor hypoxic environment, we implemented a hypoxia bio-sensor in the yeast S. cerevisiae.

Design of the hypoxia bio-sensor

The choice of the right promoter is crucial in a bio-sensor design. Yeasts are facultative anaerobes that can naturally detect very low oxygen concentration with the corresponding promoters already present. However, these promoters need an addition fermentable carbon for induction, like ADH2 (4, 5). On the contrary, our system is intended to be used in vivo after encapsulation in alginate beads, without fermentable carbon source. Thus, we cloned the CMV minimal promoter as an inducer of the hypoxia gene reporter. Then, we need a Hypoxia Response Element (HRE) to start transcription factor binding to our CMV minimal promoter. HRE are DNA sequence that enhance transcription activity when the Hypoxia inducing Factor (HIF) binds to it. We assembled 4 HRE sequences ahead of the CMV minimal promoter to increase transcription in presence of hypoxia. The HRE was derived from the human HRE ahead of EPO promoter, as the EPO/HRE was more strongly induced than VEGF/HRE (6) by HIF in presence of hypoxia.

HIF transcription factors are composed of two sub-units. The α-subunits of the HIF transcription factors are degraded by proteasomal pathways during normoxia but are stabilized under hypoxic conditions (6). On the contrary, the beta-subunit of HIF is always expressed and maintained stable in the cytosol. We codon optimized the human HIF-alpha and HIF-beta for yeast and cloned these proteins in yeast S. cerevisiae under control of GAL1, a galactose inducible promoter.

Figure 1 : Hypoxia inducible promoter

To measure the activity of the bio-sensor, the gene reporter RFP was expressed under control of this HRE-CMV promoter. We used cobalt, a transition metal, to induce hypoxia. Cobalt mimics hypoxia by causing the stabilization of HIF-α (7).

Cloning steps were successful as showed sequencing data and colony PCR (figure 2). To remove the promoter GAL1 from our plasmid pYGG1 and replace it with the HRE/CMV promoter, we performed site-directed mutagenesis. First, we amplified the plasmid around GAL1 to remove it. Then, we phospholyrated with a kinase the blunt ends, we added DpnI to remove the template DNA and we ligated with T4 kinase. Colony PCR followed by digestion confirmed the successful assembly by golden gate:

Figure 2: Colony PCR for biosensor 2 (HIF alpha and beta) and for biosensor 3 (HRE-CMV-RFP)

Due to yeast autofluorescence in the red channel, we were not able to perform a fluorescence measurement with the Red Fluorescent Protein We will further characterize our bio-sensor with a LacZ assay instead of RFP.

Inducing apoptosis in hypoxic environment

We have a way to detect hypoxia area to locate tumor cells, now it’s time to destroy them. In order to kill hypoxic tumor cells, we created a yeast secretion system for perforin and granzyme B under control of our hypoxia bio-sensor. Perforin is a pore-forming enzyme, allowing granzyme B to enter the cell. Granzyme B is a serine protease capable of inducing cell apoptosis by caspase activation. Co-expressed, these enzymes have proven their potential to induce cancer cell death :

A. In vitro delivery of granzyme B to colon cancer cells with perforin resulted in caspase 3 activation and features of apoptosis in those cells : chromatin condensation, nucleus fragmentation and internucleosomal DNA fragmentation (8)

B. Recombinant Granzyme B and perforin co-expressed in E. coli induced apoptosis and directly inhibited the growth of human laryngeal cancer Hep-2 cells in vitro (9)

Figure 3 : Yeast encapsulated secreting perforine and granzyme B upon hypoxia detection

For optimal conditions of translation, we placed both proteins on the same mRNA under control of the CMV minimal promoter. The Internal Ribosome Entry Site IRES URE2, effective in yeast, was placed ahead of the second protein to ensure two translations on the same mRNA.

Two different secretion signal, Matalpha leader peptide and BGL2 signal peptide, were fused to each protein to avoid internal recombination during expression.

Sequencing data showed the correct cloning of all the constructions producing Granzyme B and perforine. In vitro characterization remains to be done with these constructions.

References