A gene therapy strategy to target hepatocellular carcinoma based on conditional RNA interference

1.Description

As described in Background , spatial and/or temporal regulation of RNAi is of significant importance for basic research as well as practical applications. Since disease-specific promoters only have high activity in pathogenic cells, our RNAi becomes conditional and specific for pathogenic cells as we put genes of the RdRp and non-coding RNAs behind them (Figure 1). When the two devices become transcriptionally activated together, RNA interference occurs. This actually forms a logical “AND” gate - it behaves according to the truth table on the right.

Figure 1. The “AND” gate based on disease-specific promoters

1.1 Key procedures

DROSHA can’t cleave pri-miRNA (green) when the latter binds with the inhibitory strand (yellow), which means no RNAi (Figure 2A).

SLD3 is a short template competent for de novo RNA synthesis. It is located at the 3’ end of the inhibitory strand. NS5B, the RNA dependent RNA polymerase, interacts with the 2’-OH and 3’-OH of the two cytosines at the 3’ end of the SLD3 sequence, then starts RNA polymerization using the RNA promoter and the inhibitory strand as the template. It separates the inhibitory strand from the pri-miRNA, exposing the single-stranded area. The Microprocessor then recognizes this substrate and cleaves it into precursor miRNAs, which eventually would become mature miRNAs (Figure 2B).

Figure 2. The key procedures of the design

In the Tet-off system (Figure 2C), when Tetracycline (Tc) or Doxycycline (Dox) is absent, tetracycline transactivator (tTA) binds with the tetracycline response element (TRE), which then activates the downstream gene expression of a miRNA sponge. MiRNA sponge absorbs the miRNA targeting MAP4K4 to again block the RNAi pathway. When Tc or Dox is present, tTA changes its conformation to bind with them, thus preventing the downstream genes from expression.

1.2 Features

1.2.1 Cancer-specific promoters

Our gene therapy strategy utilizes two cancer-specific promoters (one HCC specific) to open an AND-gated system to target HCC.

hTERT promoter is the core promoter of human telomerase reverse transcriptase gene. It is one of the most well-known cancer-specific promoters. A number of factors like cellular transcriptional activator c-Myc, HIF-1 as well as the repressors p53, WT1, and Menin, most of which comprise tumor suppressor gene products, have been identified to directly or indirectly regulate the hTERT promoter, contributing its exclusive up-regulation in cancer.

HULC encodes a long non-coding RNA (lncRNA) and plays an important role in tumorigenesis. It is one of the most up-regulated genes in hepatocellular carcinoma and such activation, like the one for hTERT, is at transcriptional level. Here we only use a 132bp HULC core promoter because literature suggests that the most proximal 84nt of the Hulc gene contains several transcription factor binding sites and is capable of initiating transcription just fine.

When these two promoters work together, which only happens in tumor cells (especially malignant ones), the AND gate is “opened”, releasing miRNA and damaging themselves.

1.2.2 Controlled by small molecules

Our gene therapy strategy can be controlled by a small-molecule drug.

As we mentioned above, the Tet-off system can also adjust the amount of functional miRNA, which should be instructive for individual drug administration in vivo.

1.2.3 Wide application on other diseases

The gene therapy strategy shown here is actually a proof of concept. As a RNA interference strategy, it has its inherent flexibility to be adapted to target any messenger RNAs and to other disease settings as long as there are corresponding disease-specific promoters. This is discussed in detail in Application(link) page.

1.3 Plasmid construction

PlasmidⅠcontains hTERT promoter, transactivator tTA and RNA dependent RNA polymerase NS5B. We loaded a nuclear location sequence (NLS) to NS5B (NS5B^NLS) for this protein to be transported into the nucleus. PlasmidⅡcontains the HULC promoter, genes encoding the pri-miRNA and the inhibitory strand. Notably, tTA and TRE are located separately in two plasmids (Figure 3). More information can be found in Part.

Figure 3. Plasmid construction of our system

2.Results

To apply our parts, we chose pIRES and pcDNA3.1 as backbones for the plasmidⅠandⅡ vectors, respectively. But before connecting all the parts, we constructed plasmids containing only some of the parts to verify the proper function of each part. Here the plasmids we used involve p1 (pIRES-hTERT-tTA-NS5B^NLS), p-H(pcDNA3.1-HULC-pri-miRNA), p-H-U6(pcDNA3.1-HULC-pri-miRNA-U6-inhSi-pri-miRNA), and p2 (pcDNA3.1-HULC-pri-miRNA(MAP4K4)-U6-inhSi-pri-miRNA-tre-sponge).

2.1 Determine the promoters’ activity and specificity

Our system includes two specific promoters: hTERT and HULC. We used pGL3-Basic vector, a promoter-less vector for the luciferase assay to determine the transcriptional activity of these promoters. We added the promoters on pGL3-Basic vector and measure the OD value of the luciferase activity. We chose SV40, a highly activated promoter in both cancer cells (Figure 4A) and normal cells (Figure 4B) as positive control (PGL3-CON).

Figure 4. The efficacy and specificity of the cancer-specific promoters

2.2 Function verification of NS5B

Successful expression of NS5B^NLS of p1 is first verified by western blot (data now shown). In Figure 5, the results of immuno-fluorescence shows the ability of nuclear translocation of NS5B^NLS has improved. (See detailed information in Improve)

Figure 5. Location of NS5B

2.3 Conditional RNA interference

To test the efficiency of our system, we performed quantitative PCR (qPCR) on the effector miRNA and the targeted MAP4K4 mRNA(Figure 6A). Pri-miRNA analogue was successfully encoded and processed into miRNA, however, with the presence of the inhibitory strand, the amount of miRNA sharply decreased since DROSHA cannot cleave the pri-miRNA. This can be further confirmed in Figure 6B where a significant increase of mRNA was observed after expression of the inhibitory strand, which also indicates that our miRNA can successfully target MAP4K4.

However, from Figure 6A and 7B we found that NS5B did not function as expected, the inhibitory strand hardly removed. Taken from the results discussed above this might be due to insufficient presence of NS5B in the nucleus. However, the efficiency of nucleus translocation can be improved with e.g. adding two NLSs, thus our system might still work. Efforts will be paid regarding this issue in the future.

Figure 6. The expression of miRNA(MAP4K4) (A) and MAP4K4 mRNA (B)

For the Tet-off system, when the two plasmids worked together, we saw an increase of MAP4K4 mRNA (Figure 6B), which indicates successful down-regulation of the miRNA by the miRNA sponge.