Monitoring Gene Therapy Levels

Precise localization and quantitative assessment of the magnitude and temporal variation of transgene expression is a necessary component of any gene therapy trial. Direct imaging with a transgene-specific imaging probe is ideal but neither feasible nor practical in most cases. To develop a specific probe for each individual transgene is not always technically possible; furthermore, it is necessarily labor and cost-intensive. Using

X-Gal Histochemistry

Co-Registration

Figure 13 Tracking HSV Infection with [131I]FIAU Using Autoradiography Tracking wild-type HSV-1 infection with radionuclide-based techniques can be accomplished using the virus's native thymidine kinase gene and areporter probe such as radiolabeled [131I]FIAU. To help corroborate imaging findings with histochemistry findings, a replication conditional, oncolytic recombinant HSV-1 virus vector, hrR3, containing a lacZ insertional mutation within the RR gene locus, has been prepared. Following injection of the vector into rat gliosar-comaxenografts, tumors were processed for tissue-sectioning, autoradiography, and p-galactosidase-stained histology. Image coregistration of tumor histology, HSV-1-f£-related radioactivity (assessed by [131I]FIAU autoradiography), and lacZ gene expression (assessed by p-galactosidase staining) demonstrated a characteristic pattern of gene expression around the injection sites. A narrow band of lacZ gene expression immediately adjacent to necrotic tumor areas is observed, and this zone is surrounded by a rim of HSV-1-f£-related radioactivity, primarily in viable-appearing tumor tissue. PET images (not shown) of injected tumors in the intact animal have also been performed using [124I]FIAU as a reporter probe; the areas of PET-labeled probe uptake correlate well with the p-galactosidase-stained photomicrographs. See color insert for color version of this figure. (Image reproduced with permission from Ref. 97.)

Figure 14 Tracking Transferrin Targeted Polyethylenimine (PEI)-Mediated Gene Delivery Using Optical Bioluminescence Imaging Delivery of the bioluminescence reporter gene, firefly luciferase (Fluc), by CMV-Fluc DNA/PEI polyplexes and subsequent Fluc expression can be imaged in living mice using a cooled CCD camera. Additionally, the biodistribution of modified PEI polycation complexes, altered with molecules such as transferrin and/or polyethylene glycol (PEG), can be studied in this manner. Transferrin targeting has been shown to improve the transfection efficiency in certain tumor cell lines, and PEG modification has been shown to improve circulation times of DNA/PEI complexes and prevent their nonspecific uptake by the reticuloendothelial system. All CCD images are of living mice carrying N2A xenograft 24 or 48 h after intravenous injection of various DNA/PEI polyplexes. Site of tumor is indicated (T). (a) PEI (positive control) treated animals show relatively high Fluc expression (using 1 x D-Luciferin) in the lungs as compared with the tumor. The activity on the left hind limb is from the N2A cell tumor (T). Nonspecific tail activity occurs at the DNA/PEI polyplex injection site. All Fluc expression decreases at 48 h. (b) Tf-PEI-PEG-A-, Tf-PEI-PEG-B-, and Tf-PEI-treated mice show Fluc expression (using 2 x D-Luciferin) in the tumor (T) and tail regions, but no detectable signal in the lungs. For each formulation, expression in the tumor varied over 24 to 48 h. All images are quantitated as indicated by the 2 scales (RLU/min). See color insert for color version of this figure. (Image reproduced with permission from Ref. 119.)

Figure 14 Tracking Transferrin Targeted Polyethylenimine (PEI)-Mediated Gene Delivery Using Optical Bioluminescence Imaging Delivery of the bioluminescence reporter gene, firefly luciferase (Fluc), by CMV-Fluc DNA/PEI polyplexes and subsequent Fluc expression can be imaged in living mice using a cooled CCD camera. Additionally, the biodistribution of modified PEI polycation complexes, altered with molecules such as transferrin and/or polyethylene glycol (PEG), can be studied in this manner. Transferrin targeting has been shown to improve the transfection efficiency in certain tumor cell lines, and PEG modification has been shown to improve circulation times of DNA/PEI complexes and prevent their nonspecific uptake by the reticuloendothelial system. All CCD images are of living mice carrying N2A xenograft 24 or 48 h after intravenous injection of various DNA/PEI polyplexes. Site of tumor is indicated (T). (a) PEI (positive control) treated animals show relatively high Fluc expression (using 1 x D-Luciferin) in the lungs as compared with the tumor. The activity on the left hind limb is from the N2A cell tumor (T). Nonspecific tail activity occurs at the DNA/PEI polyplex injection site. All Fluc expression decreases at 48 h. (b) Tf-PEI-PEG-A-, Tf-PEI-PEG-B-, and Tf-PEI-treated mice show Fluc expression (using 2 x D-Luciferin) in the tumor (T) and tail regions, but no detectable signal in the lungs. For each formulation, expression in the tumor varied over 24 to 48 h. All images are quantitated as indicated by the 2 scales (RLU/min). See color insert for color version of this figure. (Image reproduced with permission from Ref. 119.)

indirect imaging methods by linking a portable reporter gene to a therapeutic gene allows for more flexibility, as a variety of transgenes can be individually monitored by cloning a reporter gene into appropriate sites of the vector. Coexpression of the therapeutic gene product and reporter gene product in a coordinated and regulated manner enables a correlative and quantitative relationship between the 2 genes. Thus, levels of therapeutic gene expression can be inferred by the amount measured from reporter genes, provided that the expression of both genes remains coupled. Several such approaches, ranging from the more straightforward, like the dual vector approach, to the more sophisticated, such as the bidirectional transcrip-

tional approach, are currently being developed and are briefly discussed below (78).

A. Covector Administration

One relatively simple method to monitor gene therapy in vivo is to coadminister 2 different vectors that are identical in every regard with the exception of the transgene they are carrying: one vector would encode the therapeutic gene, the other would encode the reporter gene, and both genes would be driven by the same promoter. This approach has been validated using the 2 PET reporter genes, HSV-sr39ifc and D2R, each cloned

Control fl 5 hours R LU/minute tk 24 hours %ID/g

Control fl 5 hours R LU/minute tk 24 hours %ID/g

Control fl S hours BLU/minute tk 24 hours %tD/g

Figure 15 Tracking Cationic Lipid-Mediated Reporter Gene Delivery Using Optical (Bioluminescence) and PET Imaging

Cationic lipids associate with negatively charged DNA to form complexes that bind to cell surfaces by way of electrostatic interaction, thereby allowing a nonviral means of gene transfer. Distribution of systemic administration of DNA-lipid complexes in mice is demonstrated by delivering prepared DNA-lipid complexes that carry optical and PET reporter genes. CMV-f plasmid DNA (cytomegalovirus (CMV) promoter driving expression of firefly luciferase (fl) gene) was mixed with cationic lipid, 1,2-dioleoyl-3-trimethylammonium-propoane (DOTAP) and cholesterol, to form fl DNA-lipid complexes. A similar procedure was used to prepare HSV1-sr39tk DNA-lipid complex (tk DNA-lipid complex). (A, B) Figures A and B show images following administration of 50 ^g and 75 ^g each of fl and tk DNA-lipid complexes via tail vein injection into CD-1 mice, respectively. Bioluminescent images (left images) were obtained 5 h after injection of the vector and 5 min after intraperitoneal injection of D-Luciferin. MicroPET images (right images) were obtained 24 h after vector delivery and 1 h after [18F]FHBG injection. Control mice (left) optical images were obtained prior to administration of D-Luciferin. Optical and PET images demonstrate that lungs are primary organs for transgene expression. Increased dose of DNA-lipid complex results in greater pulmonary transgene expression. Activity seen in the kidneys in the microPET images is the result of excreted, unsequestered, reporter probe, [18F]FHBG. See color insert for color version of this figure. (Image reproduced with permission from Ref. 120.)

Figure 15 Tracking Cationic Lipid-Mediated Reporter Gene Delivery Using Optical (Bioluminescence) and PET Imaging

Cationic lipids associate with negatively charged DNA to form complexes that bind to cell surfaces by way of electrostatic interaction, thereby allowing a nonviral means of gene transfer. Distribution of systemic administration of DNA-lipid complexes in mice is demonstrated by delivering prepared DNA-lipid complexes that carry optical and PET reporter genes. CMV-f plasmid DNA (cytomegalovirus (CMV) promoter driving expression of firefly luciferase (fl) gene) was mixed with cationic lipid, 1,2-dioleoyl-3-trimethylammonium-propoane (DOTAP) and cholesterol, to form fl DNA-lipid complexes. A similar procedure was used to prepare HSV1-sr39tk DNA-lipid complex (tk DNA-lipid complex). (A, B) Figures A and B show images following administration of 50 ^g and 75 ^g each of fl and tk DNA-lipid complexes via tail vein injection into CD-1 mice, respectively. Bioluminescent images (left images) were obtained 5 h after injection of the vector and 5 min after intraperitoneal injection of D-Luciferin. MicroPET images (right images) were obtained 24 h after vector delivery and 1 h after [18F]FHBG injection. Control mice (left) optical images were obtained prior to administration of D-Luciferin. Optical and PET images demonstrate that lungs are primary organs for transgene expression. Increased dose of DNA-lipid complex results in greater pulmonary transgene expression. Activity seen in the kidneys in the microPET images is the result of excreted, unsequestered, reporter probe, [18F]FHBG. See color insert for color version of this figure. (Image reproduced with permission from Ref. 120.)

into distinct adenoviral vectors and both driven by the same CMV promoter (68). While individual cell differences in expression levels may be seen, macroscopic measurements made at the tissue culture or organ level (adnenoviral-me-diated hepatic transfer) correlate quite well (r2 > 0.93). The technique may prove useful in specific experimental situations.

B. Single Vector Approaches

The use of an internal ribosomal entry site (IRES) is a hallmark of the bicistronic approach to coupling genes (98-100). In a bicistronic expression cassette, an IRES sequence is interposi-tioned between the therapeutic and reporter gene, usually the first and second cistron, respectively. Both genes are under the control of the same promoter, and transcription of this construct results in a single mRNA molecule. Initiation of translation of the first cistron is by way of the usual cap-dependent manner, but translation of the second cistron is facilitated by the IRES sequence in a cap-independent mechanism, which allows binding a translation by a second ribo-some. This approach has been verified in a few studies. For example, an IRES derived from an encephalomyocarditis virus has been used to construct a bicistronic vector from which both D2R and HSV1-sr39tk reporter genes are coex-pressed from a common CMV promoter (pCMV-D2R-IRES-HSV1-sr39tk) (47). The levels of D2R andHSV1-sr39tk activity demonstrate a high degree of correlation (r2 = 0.97) using [18F]FESP and [18F]FHBG as imaging probes, respectively. Another vector that encodes Renilla luciferase (Rluc) in a bic-istronic configuration, pCMV-Rluc-IRES-sr39tk or pCMV-sr39tk-IRES-Rluc, also shows excellent correlation (47). Similar relationships have also been seen with the use of an HSV1-tk gene that has been ''IRES-linked'' to the lacZ gene (101); imaging with iodinated FIAU (SPECT reporter probe for HSV1-tk) correlates well with p-galactosidase activity seen by light microscopy. These studies corroborate the use of ra-dionuclide and optical reporter genes as a means of quantitatively determining relative levels of target gene expression.

One interesting finding in this approach, however, is that expression levels from the gene upstream to IRES sequence is consistently more robust then the levels seen from the gene downstream to the IRES (47). This may have to do with cell-specific differential translation from the IRES sequence, but, regardless, emphasizes the need for a highly sensitive reporter system, as levels of the reporter gene product will be significantly diminished compared to the upstream gene. Further understanding and exploitation of regulatory ''modules'' recently found within the IRES may help circumvent this problem in the future (102). Alternatively, 2 different genes from 2 distinct, but identical, promoters within a single vector can be expressed to avoid the attenuation problem and tissue variation issues experienced with the IRES-based approach. This is a variation of the bicistronic approach and is otherwise known as the ''dual-promoter'' approach (76). Strong correlation between 2 reporter genes, hsstr2 and HSV1-tk, each driven by an independent but identical CMV promoter, has been exhibited (Fig. 10).

In some situations, gene therapists will want to externally control the levels of transgene expression and, additionally, will need to verify the extent of control with imaging techniques. One particularly novel indirect imaging method makes this entire scenario possible by the use of a single inducible bidirectional tetracycline-responsive element and 2 flanking minimal CMV promoters (Fig. 16A). The fusion transactiva-tor protein rtetR-VP16, which is constitutively expressed and can potentially be incorporated into the same bidirectional vector, binds to TRE only in the presence of tetracycline or 1 of its analogs. By varying levels of an exogenously added inducer such as doxycycline, a gene therapist can control transcription and magnitude of expression, which can be verified by the accompanying reporter gene. Proof of principle has been shown in rat xenograft models using the reporter genes D2R and HSV1-sr39tk (Fig. 16B) (103).

The fusion gene/protein approach is yet another powerful means of indirect monitoring of gene therapy. Constructs in this protocol contain 2 or more genes linked together within the same reading frame so that a single protein is translated. The resultant hybrid or fusion protein will have therapeutic and reporter properties, and the expression of the fused gene can be closely monitored since the expression of the therapeu tic component is stoichiometrically coupled to the reporter component of the protein. HSV1-tk-gfp, HSV1-tk-Fluc, gfp-Fluc, and HSV1-sr39tk-Rluc are a few successful examples (21,104-106). The fusion gene is engineered such that the proteins are linked via a short peptide spacer. However, while appealing in concept, it is quite challenging in practice to produce generalizable proteins, since fusion proteins are often inactive or less active than their individual counterparts; furthermore, fusion proteins also may not localize to appropriate compartments since appropriate signaling mechanisms are either masked or unavailable (78). Using this technique for every therapeutic gene developed would prove to be a daunting task, since each newly generated fusion protein has its own peculiarities. Future technological improvements in the physical linkage of 2 proteins may help minimize discrepant behavior and allow more flexibility for this technique.

Was this article helpful?

0 0

Post a comment