Tissue selective transgene expression

Abstract

A method for expression of transcribable unit(s) in a target cell is provided. The method comprises the steps of: a) providing a target cell expressing a site-specific recombinase, b) providing a DNA vector characterized by a 5 to 3 vector sequence orientation. The DNA vector comprises a plurality of recombination units, wherein a single recombination unit comprises at least one transcribable unit and a first type and a second type of target site that are recognizable by the site-specific recombinase. Recombination can only occur between two target sites of the same type and the first type of target site is located at the 5 start of the recombination unit and the second type of target site is located at the 3 end of the recombination unit. For all recombination units comprised within the DNA vector, the orientation of all of the first type of target sites are the same, and the orientation of all of the second type of target sites are the same. Step c) comprises introducing the DNA vector into the target cell.

Claims

1. An in vivo method for determining phenotype alteration by a plurality of transcribable units in a non-human tissue of interest comprising the steps of: (A) administering to a non-human tissue of interest in vivo, a DNA expression vector comprising, in 5 to 3 orientation, a plurality of recombination units, wherein a single recombination unit comprises at least one transcribable unit and a first and second target site both recognizable by a site-specific recombinase, wherein said non-human tissue of interest expresses a site-specific recombinase, (B) harvesting the tissue of interest, (C) separating said tissue of interest into individual cells and sorting said individual cells according to phenotype of interest, (D) identifying said transcribable unit(s) expressed in each of said sorted individual cells; and (E) assaying for the altered phenotype that results from expression of the transcriptional unit(s) in said individual cells, wherein (i) recombination can only occur between two target sites of the same type, said first target site is located at the 5 start site of said recombination unit and said second target site is located at the 3 end of said recombination unit, and (ii) for all recombination unites comprised within said DNA vector, the orientation of all of said first target sites are the same, and the orientation of all of said second target sites are the same.

2. The method according to claim 1, wherein said phenotype of interest is characterized by a cell size, a cell morphology, a cell staining or a cell marker.

3. The method according to claim 1, wherein said plurality of recombination units comprises 2 to 80 recombination units.

4. The method according to claim 1, wherein each of said recombination units in said DNA vector additionally comprises a transcriptional terminator located between the transcribable unit closest to the 3 end of said recombination unit and said second type of target site.

5. The method according to claim 1, wherein said site-specific recombinase is selected from Cre-recombinase and FLP.

6. The method according to claim 1, wherein said DNA vector comprises a multitude of selection units, wherein each of said selection units comprises: a promoter to enable expression of the transcribable units, and a multitude of recombination units.

7. The method according to claim 1, wherein said DNA expression vector is a viral vector derived from a virus selected from the group consisting of adeno-associated virus, adenovirus, lentivirus, retrovirus, and baculovirus.

8. The method of claim 3, wherein said plurality of recombination units comprises 5 to 60 recombination units.

9. The method of claim 3, wherein said plurality of recombination units comprises 10 to 50 recombination units.

10. The method of claim 6, wherein the promoter to enable expression of the transcribable units is a U6 promoter.

Description

SHORT DESCRIPTION OF THE FIGURES

(1) FIG. 1 shows neonatal mouse cardiomyocytes (NMC) that were infected with AAV9-MLC2v-lox-STOP-lox-GFP-transgene in triplicate, in the absence (top panels) or presence (lower panels) of ectopic Cre recombinase (Ht-AAV9) (A). Samples from (A) were further assessed for GFP by immunoblotting (B) and qPCR (C) (p<0.005). Sarcomeric -actinin serves as a loading control for immunoblotting in (B) and GFP expression, as assayed by qPCR, was normalised to hypoxanthine guanine phosphoribosyl transferase (Hprt) (p<0.005).

(2) FIG. 2 shows tail-vein injection of AAV9-MLC2v-lox-STOP-lox-GFP into MHC-driven tamoxifen inducible Cre-recombinase mice, in the absence (Cre) or presence (Cre+) of tamoxifen-mediated Cre recombinase activation. GFP expression in vivo was validated by immunoblotting for GFP, with sarcomeric -actininin as a loading control. Protein lysates derived from two independent experiments are shown for the respective sets.

(3) FIG. 3 shows the schematic of an AAV construct for simultaneous organ-specific multi-transgenesis. Arrow indicates upstream PolIII promoter, with red and green triangles indicating loxP elements in a convergent or divergent orientation, respectively. Blue squares indicate the position of transgenes within the construct (A). The minimal repetitive unit of the construct is shown in (B).

(4) FIG. 4 shows a schematic representation of excision events leading to the presence of a single transcriptionally active transgene unit downstream of the promoter for a construct harbouring three individual transgenes (A) and four individual transgenes (B). A indicates loxP element in convergent orientation, while b indicates loxP element in a divergent orientation. Numbers in the respective boxes refer to the individual transgenes.

(5) FIG. 5 shows a schematic representation of the series of excision events leading to the presence of a single transcriptionally active transgene unit downstream of the promoter for a construct harbouring three (A), five (B), ten (C) or twenty individual transgenes (D). A indicates loxP element in convergent orientation, while b indicates loxP element in a divergent orientation. Numbers refer to individual transgenes.

(6) FIG. 6 shows the distribution of transgene expression from a 10 (A), 20 (B) and 50 (C) transgene AAV construct. Expression percentage (y-axis) indicates percentage of cardiomyocytes expressing the respective transgene. The x-axis shows position of the transgene with the respective AAV constructs.

EXAMPLES

(7) In order to overcome the above mentioned drawbacks of known methodology, the inventors have developed a novel AAV-based strategy for tissue-specific, inducible and long-term animal transgenesis. By virtue of the broad tropism of viral vectors such as AAV, this system allows for the specific targeting of multiple organs, including the brain, kidney, liver, spleen, adipose tissue, skeletal muscle and the heart. Importantly, this approach potentially allows for the high-throughput analysis of in vivo organ-specific gene function directly in animals, without the need to generate a new line of transgenic animals for each gene analyzed.

(8) This approach entails the generation of an AAV harbouring the transgene combination of interest, downstream of an RNA polymerase II (RNA PolII)- or RNA polymerase III (RNA PolIII)-driven promoter. When delivered systemically into adult rodents harbouring the cardiac-specific myosin heavy chain (MHC) -driven tamoxifen inducible Cre-recombinase (Cre), the AAV9 virus is transduced to a number of organs including the liver, kidney and heart. However, the transgene is not active despite transduction of heart cells (cardiomyocytes) by AAV9. Viral transduction can be verified by monitoring green fluorescent protein (GFP) expression, which is absent in all transduced cells. Transgene expression is induced specifically upon intraperitoneal tamoxifen delivery, upon which GFP expression is induced, in specific cell types.

Example 1: In Vitro Validation

(9) In developing this platform, the inventors first assessed the sensitivity of the AAV9 construct to Cre recombinase mediated GFP transgene induction. As shown in FIG. 1A (top panels), in the absence of Cre recombinase, GFP expression is not detected. In contrast, ectopic Cre recombinase expression driven by the myosin light chain 2v (MLC2v) promoter (Ht-AAV9) leads to expression of GFP in cardiomyocytes (FIG. 1A lower panels). The specificity of Cre recombinase-dependent GFP expression was further confirmed by immunoblotting with GFP specific antibodies and quantitative polymerase chain reaction (qPCR) (FIGS. 1B and 1C).

Example 2: In Vivo Validation

(10) To demonstrate the utility of the system in vivo, the cardiomyocyte targeting AAV9-MLC2v-lox-STOP-GFP-transgene virus was injected intravenously into mice expressing tamoxifen-inducible Cre recombinase under the control of the cardiac-specific MHC promoter. As shown in FIGS. 2A and 2B, control MHC-Cre mice did not exhibit GFP expression (FIG. 2A, upper panel and 2B), while mice harbouring the Cre recombinase transgene (MHC-Cre+) expressed high levels of GFP upon tamoxifen delivery (FIG. 2A, lower panel and 2B) in cardiac ventricles. Transgene expression was observed in hearts of MHC-Cre+ mice infected with AAV9-GFP at 10-20 weeks post-injection, but was absent in non-cardiac tissue throughout this duration.

Example 3: Use of a Plurality of Transcribable Units

(11) To further exploit this transgene delivery platform for in vivo, in situ functional genomics screens, a human U6 RNA promoter (RNA PolIII-driven transcription) based system for the expression of multiple coding genes was developed, including non-coding RNAs (ncRNA), micro RNAs (miRNA), small interfering RNAs (siRNA), short-hairpin RNAs (shRNA) or a combination thereof within a single AAV, such that only one transgene or shRNA is expressed per cell of the target organ. Cell specific expression of the individual transgene or shRNA is achieved through combinatorial utilization of loxP elements in a convergent or divergent orientation as shown in FIG. 3A. In FIG. 3B, the minimal repetitive unit of the construct is shown.

(12) Upon Cre recombinase expression in the target tissue, a series of excision and recombination events occur (FIGS. 4A and 4B) culminating in the presence of a single transgene downstream of the promoter amenable to transcription. Transcription of other transgenes further downstream is inhibited due to the presence of a T5 transcriptional stop motif leading to fall off of RNA Polymerase III after transcription of the transgene directly downstream of the promoter. In FIG. 5A-D, a simplified representation of events is shown for a number of larger transgene sets.

(13) This system was successfully applied for expression of up to 50 individual shRNAs to target a total of 48 gene targets. Two shRNAs, positioned at the 5 and 3 ends, carry non-silencing shRNAs serving as controls to ensure specificity of the other shRNAs and to maintain some degree of targeted organ function. In FIG. 6, the percentage of cells within the heart expressing a specific transgene is shown for an AAV9 harbouring 10 (FIG. 6A), 20 (FIG. 6B) or 50 (FIG. 6C) shRNAs. The shRNAs contained within the respective cells were determined by qPCR and/or sequencing of RNA derived from enzymatically dissociated heart cells subjected to FACS sorting. As indicated in FIG. 6A-C, transgene expression is strongly correlated to its position within the AAV9 construct. Transgenes at the flanks are most heavily represented, while those positioned at the centre are least represented. However, in the context of the adult mouse heart, within a 50 transgene construct, the lowest expressed transgene was detected in approximately 0.2% of cardiomyocytes within the heart corresponding to roughly 40 000 heart cells.

Example 4: In Vivo Assessment of a Plurality of Transcribable Units for Phenotype Variation

(14) As an example for the in vivo assessment of gene function for phenotype variation using the current invention the study of fatty liver disease is detailed. Mice expressing Cre-recombinase in the liver were maintained on a high fat diet to induce fatty livers. Next, the mice were injected with a virus containing shRNAS targeting 50 individual genes that are hypothesised to effect fat accumulation in the liver. Thereafter, the liver was dissociated using standard methods yielding individual hepatocytes. Some of those hepatocytes carry a fat load while others do less so or haven't any fat load. Using a fat dye that enters and stains lipids accumulated in the hepatocytes can be sorted by fluorescense-activated cell sorting (FACS). This yields defined populations of cells containing varying amounts of fat based on the staining intensity. Knockdown of genes hypothesised to promote lipid accumulation would be predicted to inhibit lipid accumulation. Hence, selection of cell populations containing negligible amounts of the fat-dye and performing targeted sequencing (using unique primer specific for a sequence within the expression vector) or quantitative PCR (of all 50 shRNAs) of the cell, identifies shRNAs that prevented fat accumulation in these cells. In doing so, genes involved in fat accumulation in the liver are identified.

(15) This method can be applied for other possible phenotypic readouts such as cell size, cell morphology, surface or internal protein expression, apoptosis, proliferation, etc. Basically any parameter that serves as readout for a gene function that can be distinguished by staining, size or cell morphology can be used.

Example 5: Novel Mouse System for In Vivo Orthotropic Identification of Disease-Associated Genes

(16) The invention can also be employed for the generation of a novel mouse system for in vivo orthotropic identification of disease-associated genes. This is achieved by introduction of multiple small interfering RNAs (siRNA), short hairpin RNAs (shRNA) or coding or non-coding genes flanked by a combination of two different loxP motifs (that are mutually exclusive but of similar Cre recombinase sensitivity and efficiency) into a single transgene at the mouse ROSA26 locus. This system enables the functional assessment of multiple genes in a single mouse, in a cell type and organ-specific manner.

(17) The DNA construct shown in FIGS. 3A and 3B, where FIG. 3B represents the minimal modular unit of this system, while FIG. 3A depicts a schematic of the transgene, is introduced at the ROSA26 locus. Expression of the respective siRNAs in the transgene in FIG. 3A is tightly regulated by Cre recombinase activity (Cre) activity, such that in any given cell only one siRNA (or shRNA or coding or non-coding gene) is active. This occurs upon completion of Cre-mediated recombination as shown in FIGS. 4A and 4B, where the arrows represent the multiple routes taken to reach completion of Cre-activity in a three (FIG. 4A) or four (FIG. 4B) siRNA transgene.

(18) Crossing mice carrying the multi-construct transgene at the ROSA26 locus with mice expressing Cre in an organ-specific manner, allows the relevance of specific genes in the context of disease to be assessed. Through siRNA/shRNA-mediated knockdown or ectopic expression of multiple genes in a cell-type specific manner, while simultaneously ensuring that only one gene or RNA is upregulated or inhibited in one cell, the function of multiple genes or RNAs can be addressed simultaneously in one mouse. Using this approach, the in vivo function of a protein family (consisting of 50 members) can be interrogated in a specific cell type with 1-3 mice as opposed to the 50 mice that would be required using conventional methods.

(19) Upon derivation of mice carrying both the multi-transgene construct and a tissue-specific Cre, these mice can be subjected to pathologic stress stimuli to induce a disease state. The relevance of a specific siRNA (or shRNA, or coding or non-coding gene) will thereafter be assessed through a phenotypic screen.

(20) Taking heart disease as a model, cell size in response to pathologic stress can be used as readout. Heart cells can be isolated using the established Langendorf method and fluorescence activated cell sorting (FACS) performed to distinguish large cells from small cells. Once the cell populations have been separated, parallel sequencing for siRNAs, shRNAs or genes enriched in the population of large or small cells can be determined. siRNAs enriched in the small cell population would be predicted to inhibit pathologic growth in response to stress. In doing so, the in vivo functional significance of a specific siRNA or gene can be determined. Other readouts such as cell proliferation, ROS levels, induction of specific protein, etc. can also be used using the same principle. This system can also be used without a stress stimulus to assess if siRNA-mediated gene inactivation or ectopic gene expression is sufficient to induce a disease phenotype.

(21) Concept Behind Invention:

(22) This novel system for large-scale in vivo organ-targeted mammalian transgenesis is applicable for the tissue specific and inducible expression of transgenes in a broad range of embryonic, newborn and adult animals. The use of this platform enables the high-throughput and economical gain-of-function and loss-of function assessment of coding and non-coding gene function in vivo in an organ-specific manner. Importantly, by facilitating the knockdown or ectopic expression of multiple genes/ncRNAs in a cell-type specific manner in one animal, while simultaneously ensuring that only one gene or RNA is upregulated or inhibited per cell, the function of multiple genes or ncRNAs can be addressed simultaneously. Using this approach, the in vivo function of a protein family (consisting of up to 50 members) can be interrogated in a particular cell type with several animals as opposed to the around 50 animals that would be required using conventional methods. Thereafter, function of specific transgenes can be addressed using functional or phenotypic readouts including apoptosis, cell proliferation, ROS levels, induction of specific protein, etc. followed by cost-effective targeted qPCR or sequencing techniques to identify phenotype-associated gene targets. This platform not only facilitates easy in vivo gene validation, but is also advantageous by reducing the number of animals used in experiments. This technology can be further extended for in vivo screening of DNA/RNA-based therapeutic compounds in rodents and non-rodent/large animals, for optimization of plant growth in agriculture and used in combination with CRISPR/Cas technology for simultaneous multi-loci genome editing for generation of mutations, knockouts or knock-ins in a single cell-specific manner for downstream functional genomic analysis.

(23) This platform serves as the missing link for translation between animals and human clinical studies by facilitating the ability to cost-effectively and efficiently extend preliminary findings in small animal models to more physiologically relevant large animal models such as the pig, sheep and non-human primate, prior to targeted therapies in humans.