Retroviruses are RNA-based viruses that insert a portion of their genome into the host’s genome to replicate and survive in host’s body. This property of insertion of a portion of their genome into host’s genome is exploited by scientist to develop tool and therapeutic methods to treat hereditary and defective gene conditions such as Primary Immuno Deficiencies (PIDs) including Adenosine Deaminase (ADA) Deficiency- Severe Combined Immunodeficiency (ADA-SCID) or Wiskott – Aldrich syndrome (WAS). This therapeutic procedure is called Gene Therapy. Most commonly used viral vector system is of Murine Leukaemia Virus (MLV). Integration of viral genome, in this case, is site-directed and is called Gammaretroviral integration. Host’s cellular cofactors also help in this integration. This includes Bromodomain and Extraterminal (BET) Domain family of proteins whose ET domain act as anchors by interacting with the unstructured C-Terminal tail of MLV Integrase (IN) and tethers the retroviral Pre-Integration Complex to chromatin regions enriched in BET proteins and thereby defines integration profile.
Lately, these therapies, when treated with MLV based vector, haven’t been as successful as expected due to the development of Leukaemia and Myelodysplasia as side effects. These effects are attributed to the inherent ability of the virus to activate and transform the genes involved in these disorders. This integration preference and its activation by the strong viral promoter and enhancer elements (present in the Long Terminal Repeat (LTR) of the virus) can be prevented by two of the following approaches:
- Development of Self-Inactivating (SIN) vector system. In this system, strong enhancer/promoter is replaced by a weak heterologous promoter to drive the transgene expression. Insulator sequences might also be used that blocks the interaction between enhancers and promoters.
- Another approach is to direct the integration away from these potentially unsafe regions by uncoupling the BET interaction of MLV based vectors either by deletion of the C-terminal domain or by a single specific substitution of tryptophan to alanine (A) at 390 positions in Integrase. This is represented by INW390A. This results in BET Independent (Bin) MLV vectors with Wild Type efficiency but less harmful integration profile.
Now the next generation Bin MLV vector systems are being developed by interfering with the chromatin tethering process to produce vector systems with potentially safer integration profile and lower propensity to activate nearby genes. This is achieved by the scientist from the Laboratory of Molecular Virology and Drug Discovery, Belgium.
They developed viral vector system by linking cellular peptides that bind specific epigenetic histone markers to INW390A. These peptides include chromodomain of heterochromatin-binding protein 1b (CBX1) and the chromodomain of Y-like protein (CDYL), giving rise to MLVINW390A-CBX and MLVINW390A-CDYL, respectively. Viral peptides were also fused to create two more combinations including peptide from human papillomavirus E2 protein and Kaposi sarcoma’s latency-associated nuclear antigen (LANA) developing MLVINW390A-E2 and MLVINW390A-LANA, respectively. All these are tethering peptides in their respective host and are fused with the C-terminal end of INW390A in MLV packaging plasmid. When these plasmids were tested to produce vesicular stomatitis virus glycoprotein G (VSV-G) with LTR-driven green fluorescent protein, results were positive. The next generation Bin MLV vectors transduced as efficiently as wild-type MLV vectors. Experiments were repeated with HeLa cells and results were the same showing that the addition of peptide at the C-terminal end of MLVIN_W390A does not affect the transduction efficiency compared to wild-type MLV vector and MLVIN_W390A. Next step was to investigate whether the fusion of peptide to Bin MLV vectors redistributes the integration preferences of the vector or not. For this, integration frequencies relative to four specific types of genome regions were considered including Transcription start sites (TSS), CpG islands, DNase Hypersensitive sites (DHSs). Results were as follows:
|Next generation Vector developed:||Site of integration||Relative to:||Change in integration frequency (Approx.)|
|MLVIN_W390A||TSS and CpG||MLVIN_WT||Decreased 2 folds|
|MLVINW390A-E2||TSS and CpG||MLVIN_WT||Decreased 2 folds|
|MLVINW390A-CDYL||TSS and CpG||MLVIN_WT||Decreased 2 folds|
|MLVINW390A-CBX||TSS and CpG||MLVIN_WT||Decreased 4 folds|
|MLVINW390A-LANA||TSS and CpG||MLVIN_WT||Decreased 4 folds|
|MLVINW390A-CBX||DHS||MLVIN_W390A||Decreased 3/4 folds|
|MLVINW390A-LANA||DHS||MLVIN_W390A||Decreased 3/4 folds|
|MLVINW390A-CBX||TSS||FV||Decreased 2 folds|
|MLVINW390A-LANA||TSS||FV||Decreased 2 folds|
FV= Foamy Virus
Other then the comparison within the different MLV based vectors developed; Foamy Viral vectors and Lentiviral Vectors were also selected because of their lower tendency to integrate near promoter regions as compared to MLV vectors. These results confirm the shift in the integration profile in CBX and LANA peptide fused vectors. Similar results were obtained when compared to a wider set of genomic features. An additional feature of the addition of CBX peptide to the C-terminal end of MLV Integrase is that the shift in the integration profile is from transcriptionally active chromatin region whereas other peptides had no such specific shift. Therefore CBX peptide fusion generates safer integration profile in MLV vectors.
These all experiments were based on one approach that is to substitute a single amino acid in the c-terminal domain. Another approach was also tried by deleting the C-terminal domain resulting in MLVIN_1-380. Peptides were fused with this virus type also and same four combinations were created and tested for transduction efficiency. Efficiency of MLVIN_1-380 was 3 folds lower than wild type whereas the efficiency of four fused type vectors was comparable to wild type and all 5 type of MLV vector, with original C-Terminal end deleted, were detargeted from their original integration profile and yet resulted in similar expression level of transgene, therefore, showing the significance of not so significant C-terminal domain. Integration profile of MLVIN_1-380 vectors fused with peptide was similar to MLVIN_W390A vector fused with same peptides highlighting that the integration profile is based on the specific peptide fused to MLV vector.
All of these successful experiments would have been in vain if the therapeutic potential of these modified next generation vector in reduced compared to previously used vectors. Therefore therapeutic potential to transduce clinically relevant cells was also tested by transducing CD4+ T cells (Helper T cells) and CD34+ HSCs. In this experiment MLVIN_WT, MLVIN_W390A, MLVINW390A-CBX and MLVINW390A-LANA vectors with self-inactivation (SIN) gammaretroviral vector genome with a reporter gene expression from a spleen focus-forming virus (SF) enhancer/promoter, referred to as MLV.SIN.SFIN_WT, MLV.SIN.SFIN_W390A, MLV.SIN.SFIN_W390A-CBX, and MLV.SIN.SFIN_W390A-LANA respectively were used. Transduction efficiency of CBX based and LANA based vectors were comparable to wild type and Bin MLV vectors and were sustained over time. This overtime expression is an indication of suppressed silencing of a transgene which was a problem with the previous generation of vectors due to their integration preference.
Conclusion: These Next-generation gene therapy vectors are effective with no side effects as shown in patients treated with previous generations of MLV based vectors.
Reference: Title of the publication is “Engineering Next-generation BET-Independent MLV Vectors for Safer Gene Therapy”