Molecular Biology and Vector Core

Director: Noriyuki Kasahara, M.D., Ph.D.

Assistant Director: Christopher Logg, Ph.D.

Research Associates: Emmanuelle Faure, Ph.D.

Vector Core Summary [PDF format]

OBJECTIVES

The objective of the proposed CURE: DDRCC Molecular Biology and Vector Core is to promote and facilitate basic and translational research in digestive diseases by providing CURE: DDRCC investigators with access to vector technologies that enable efficient gene transfer to mammalian cells in culture and in vivo.

To this end, the Molecular Biology and Vector Core will: 1) serve as an educational and advisory resource for CURE: DDRCC researchers who may have had limited experience with virus-derived gene transfer vector technologies, but who wish to utilize such technologies for efficient functional expression of genetic sequences of interest in mammalian cell culture and in animal models in vivo; 2) at minimal cost, provide various pre-made retroviral, lentiviral, and adenoviral vector stocks expressing standard marker genes to utilize in preliminary experiments, as well as a library of available vectors expressing a variety of mammalian genes and corresponding inhibitory sequences; and 3) at minimal cost, design and produce custom viral vectors that contain a specific transgene of interest (including wild type and mutant cDNAs with or without epitope tags, dominant-negative expression constructs, antisense mRNAs, siRNAs, etc.) for individual CURE: DDRCC researchers.

In this way, the Molecular Biology and Vector Core will seek to enhance the activities of the CURE: DDRCC program by providing investigators with advice, training, stock vector reagents, custom vector production services, and technical assistance. Through these services, we intend to facilitate the existing research efforts of funded CURE: DDRCC investigators, and to provide new investigators with access to vector reagents and technologies that may lead to preliminary data for future grant applications. Furthermore, by also serving as an educational and advisory resource, we seek to widely disseminate state-of-the art gene transfer and expression techniques to new investigators, post-doctoral fellows, and research staff, to teach them how to safely produce and use viral vectors, and to assist them in preparing gene transfer protocols required by the Institutional Biosafety Committee.

CORE SERVICES

Viral Vector Systems Offered by the Core
Although viruses have evolved mechanisms to efficiently enter cells and introduce their genetic material, each of the individual virus vector systems that have been developed thus far displays significant strengths as well as specific drawbacks. Therefore, we will initially offer three different virus vector technologies:

A) Retroviral Vectors:

Retroviral vectors are currently one of the most commonly used methods for gene delivery. As vehicles for the delivery of genes into eukaryotic cells, retroviruses have several advantages [1, 2] : 1) gene transfer is relatively efficient, particularly in a cell culture or ex vivo setting, as most retroviral vectors are produced from packaging cells at titers on the order of 106-7 colony-forming units (cfu) per ml; 2) stable integration into the host cell DNA is a natural part of the retroviral life cycle, and therefore the integrated provirus is passed on to all daughter cells and continues to direct the nonlytic production of its encoded products; and 3) replication-defective vectors can easily be created by deletion of all essential viral genes, which renders the vectors incapable of secondary infection [3-5] .
Most retroviral vectors in current use are traditionally based on Moloney murine leukemia virus (MLV), an oncoretrovirus which was one of the first viral genomes to be cloned and sequenced in its entirety. One characteristic of MLV is that it requires cell division during infection so that the nucleocapsid complex can gain access to the host cell genome, and hence cannot infect non-dividing cells [2, 6] . As many cell types are considered to be largely quiescent in vivo, the traditional application which has been adopted for MLV-based retroviral vectors has been to transduce cell lines in culture; when animal studies have been performed using retroviral gene delivery, this has usually been accomplished by viral infection of primary cells in culture by the ex vivo method, followed by re-implantation of the transduced cells. This approach requires surgical acquisition, isolation, and culture of autologous cells, and thus is labor-intensive and invasive, and limits the scope of ex vivo retroviral gene transfer to those cell types that can be readily accessed, maintained and manipulated in culture, and reimplanted, e.g., hematopoietic cells, skin fibroblasts, and hepatocytes. On the other hand, this absolute selectivity for actively dividing cells results in preferential infection of malignant cells, which can be advantageous for cancer-related research and therapeutics.
The generation of high titer retroviral stocks for the efficient transduction of target cells is an important technical goal for a range of gene transfer applications. Most standard methods currently use packaging cell lines expressing the gag-pol and env genes of MLV. These will package a retroviral vector genome introduced by transduction or by transfection with an appropriate plasmid. To obtain retroviral stocks of the highest titers, it is necessary to establish additional virus producer cell lines that not only contain the gag-pol and env cassettes, but also have the proviral vector genome stably integrated. To identify the highest producing lines, many subclones may then need to be screened, as greatly varying titers are observed between different subclones [4] . This screening process can take several weeks and the cell lines so established may lose their packaging ability as they are passaged [7] . As a simpler alternative system for the production of retroviral stocks without the use of packaging lines, we have utilized a MLV-based packaging system for the production of high titer helper-free virus stocks by transient transfection. Similar systems have previously been described by others [8-10] .

Figure 1: Constructs used for transient production of MLV-based retrovirus vectors in 293 cells (see text for details).

In order to achieve maximum titers from retrovirus stocks by transient transfection, it is necessary to increase either viral gene expression or transfection efficiency of the producer cells, or both. We first boosted gene expression by the combined use of SV40 ori sequences in all the retroviral components and host cells expressing the SV40 large T antigen. 293T cells were used as they are highly transfectable. We further improved gene expression of the packaging components (gag-pol and env) in the system by directing their expression from the CMV promoter using constructs listed in Figure 1. The retroviral vectors were manipulated to replace the entire 5' U3 region with the CMV promoter while preserving the authentic RNA start site at the beginning of the R region of MLV; it has previously been demonstrated that driving transcription of the vector genome from the CMV promoter further increases infectious titers [9, 10] . This method enables us to obtain titers of up to 106/ml helper-free viral stocks without the need to establish and maintain packaging cell lines or stable producer lines, thereby allowing rapid characterization of gag-pol and env mutants and producing high titer retroviral vectors for transduction by a convenient, rapid and reproducible method. We have developed a CMV promoter-driven amphotropic envelope system as well, and this system can readily be adapted for pseudotyping (i.e., encoating with a heterologous envelope protein) using different envelope constructs.
In cases where long term production of large quantities of a specific retroviral vector is desired, however, generation of a stable producer cell line allows constitutive production and easy harvesting from the culture supernatant. Therefore, we have also obtained 293T cell lines that have been stably transfected with MLV gag/pol genes and ecotropic (i.e., mouse-specific) envelope, amphotropic (i.e., broad mammalian host range) envelope, or without envelope (NXgp-E, NXgp-A, and NXgp, respectively, generously provided by Dr. Garry Nolan, Stanford [11, 12] ), and use of these cell lines will allow us to generate stable producer cells simply by transfection of the vector construct containing the gene of interest, followed by stable selection. These third generation packaging cell lines are much easier and safer to use, because 1) the parental cell line that they are generated from, 293T, are highly transfectible, as mentioned above, 2) separation of the gag/pol and env genes greatly reduces the risk of recombination leading to the generation of replication-competent helper virus, 3) vector constructs can be transfected into the NX cells using a plasmid backbone (pLZRS) containing sequences encoding the Epstein-Barr virus nuclear antigen 1 (EBNA-1) protein, as well as the EB viral origin of replication (oriP) [11] . In latent Epstein-Barr virus infection, EBNA-1 binds to oriP and recruits host replication machinery while simultaneously binding chromosomal structures, allowing nuclear retention of the virus as an episome. The use of the LZRS-backbone for the retroviral vectors in this context thus has an added advantage in that episomal replication and nuclear retention of the plasmid allows stable maintenance of the vectors at a copy number of 5-20 per cell [13] . In practical terms, this will eliminate integration site effects and thus should allow consistent high level production of retroviral vectors from a stable pool, without the need to subclone individual producer cell sublines in order to find high level producers.

B) Lentiviral Vectors:
Although the life cycle of lentiviruses is similar to that of oncoretroviruses, there are several major differences. As mentioned above, MLV-based vectors can only transduce cells that divide shortly after infection, because the MLV pre-integration complex cannot migrate to the nucleus in the absence of mitosis. However, lentiviruses can infect non-proliferating cells, owing to the karyophilic properties of the lentiviral pre-integration complex which allows recognition by the cell nuclear import machinery. Correspondingly, HIV-derived vectors can transduce cell lines that are growth-arrested in culture, as well as terminally differentiated primary cells. In fact, HIV-based vector systems have been described since 1991 [14, 15] , but their efficacy in vivo has only recently been demonstrated [16-19] .
Pseudotyped lentiviral vectors have also been shown to mediate efficient delivery, integration, and sustained long-term expression of transgenes into post-mitotic cells such as adult neurons in vivo. In this case, the vector was pseudotyped (i.e., encoated with a heterologous envelope protein) using vesicular stomatitis virus glycoprotein (VSV-G) to achieve wider host range and stability of virions. VSV-G pseudotyped vectors can be concentrated up to 109 infectious particles per ml; however, the possible toxicity of vector preparations containing the highly fusogenic VSV-G protein remains a concern, especially at higher m.o.i. Furthermore, the possible toxicity of HIV accessory genes retained in lentiviral vector constructs, as well as the possibility of recombination leading to generation of wild type virus, has also been raised as a safety concern. Recently, HIV-derived multiply attenuated vector systems deleted of vif, vpr, vpu, nef and tat have been reported [20, 21]. The only remaining auxiliary gene in this system is therefore rev, which, with RRE, is required for efficient RNA export [22] . Thus both toxicity and the likelihood of recombination are reduced.
Another advantage of lentiviral vector systems is that the promoter inherent in the HIV long terminal repeat (LTR) is critically dependent on the HIV-encoded tat transactivator protein for transcriptional function. As the sequences encoding Tat are completely removed from the lentiviral vector construct, there is no possibility of promoter interference from the LTR, and transgene expression is completely dependent on the addition of an internal promoter. Although our lentiviral constructs all currently contain internal CMV promoters to drive transgene expression (see below), this dependence on internal promoters would be particularly advantageous if tissue-specific or conditional (e.g., tetracycline-responsive) promoters were to be used. This may be particularly important in vivo, as silencing of CMV promoter-driven transgene expression has been described.
Recently we have developed an HIV-based packaging system for the production of lentiviral vectors, using constructs obtained from Naldini and co-workers [16, 20] . The lentivirus packaging construct pCMVR8.91 contains the HIV gag-pol genes driven by a CMV promoter, with the packaging signal and most of the env gene deleted, along with the accessory genes vif, vpr, vpu, and nef. To replace the HIV gp160 envelope, which would result in a vector that binds only CD4+ cells, we use the vesicular stomatitis virus G protein (VSV-G) envelope; as mentioned above, this allows efficient transduction of a wide variety of cell types, as the receptor for VSV-G is thought to be a phospholipid (Fig. 2).
Figure 2: Lentiviral packaging, VSV-G envelope, and lentiviral vector constructs. (See text for details)

In the transfer vector itself, we and others have previously found that, despite the lack of significant promoter activity in the absence of Tat, promoter interference between the HIV LTR and the internal CMV promoter can still occur, thus significantly attenuating the levels of transgene expression achieved. This has been largely overcome by the use of third-generation self-inactivating (SIN) vectors, in which a portion of the U3 region of the 3' LTR has been deleted; thus, after reverse transcription, this deletion will be copied to the 5' LTR and hence result in loss of LTR promoter sequences in the integrated provirus, which therefore prevents interference with the function of the internal promoter. Our SIN vector construct pRRLsinCMViresGFP also contains a 5' LTR in which the HIV promoter sequence has been replaced with that of Rous sarcoma virus (RSV), a self-inactivating 3' LTR containing a deletion in the U3 promoter region, the HIV packaging signal, and RRE sequences linked to a marker gene cassette consisting of the Aequora jellyfish green fluorescent protein (GFP) driven by the CMV promoter (Fig. 2). As the GFP marker gene allows easy quantitation of the transduction efficiency simply by UV fluorescence microscopy or flow cytometry, we have constructed this vector containing an internal ribosome entry site (IRES) just upstream of the GFP gene with unique BamH1 and EcoRI sites available to insert genes of interest. The IRES will result in re-initiation of translation from the same mRNA, thus linking expression of the gene of interest with GFP expression. The pRRLsin-CMViresGFP vector construct is transiently co-transfected along with the pCMVR8.91 gag-pol packaging construct and pMD.G (VSV-G) env construct into 293T cells to produce virus. As described above for MLV vectors, this transient transfection system enables high level expression of viral proteins and efficient packaging of vector genomes without the need for long-term maintenance of stable packaging cell lines and thus without the attendant risk of recombination leading to generation of helper virus over time. The virus supernatants are harvested, purified by filtration, and used to transduce target cells.
Using this lentiviral vector system, we have successfully achieved efficient gene transfer to a large variety of primary cell types that have traditionally been extremely resistant to transduction, such as skin, lung, blood vessel endothelium, and cardiac muscle.

C) Adenoviral Vectors:
Adenoviruses have also been widely used as vehicles for gene transfer, as they can efficiently infect cells from a wide variety of tissues in cell culture, and have been shown to efficiently infect many cell types in vivo by direct injection. Adenoviral replication requires multiple early and late gene functions encoded in overlapping reading frames spanning a 36 kb genome, and because adenoviruses normally exhibit a lytic life cycle due to their extremely high level of replication, stable cell lines that completely provide all packaging functions have been impossible to generate. Thus standard adenoviral vectors generally are constructed by deletion of only one critical early gene, E1, that initiates the replication cascade, and one other early gene, E3, that is involved in evading the immune response and hence is not essential for replication per se. The vectors are made by constructing a vector shuttle plasmid containing the transgene of interest cloned between the left inverted terminal repeat (ITR) and a flanking region of homology, followed by homologous recombination with a full length adenoviral genome in 293 cells, which provide a trans-complementing E1 gene function that makes this cell line permissive for E1/E3-deleted vector replication. As they replicate to extremely high titers and accumulate in the cell, eventually lysing the host 293 cell, adenoviral vectors can be produced in much higher titers than those of retroviral vectors; with adenoviral vectors, titers of 109-10 plaque-forming units (pfu) can typically be achieved. Furthermore, adenoviruses are non-enveloped DNA viruses encoated by a proteinaceous capsid, and are thus more stable than retroviruses and can be easily purified and concentrated. However, as the adenoviral vector remains episomal and does not integrate into the host cell genome, transgene expression is transient in duration, and this is further limited by cellular and humoral immune responses against wild type adenovirus gene products, which appear to be expressed at low levels in the transduced cells due to "leaky" expression, even though deletion of the E1 regulatory region should prevent initiation of the replication cascade [23, 24] . Once sensitized, a neutralizing antibody response usually precludes repeat administration by the same vector, and adenovirus-infected cells are soon eliminated by cytotoxic T lymphocytes after transduction.
The Ad5CMVlacZ vector contains a CMV promoter-driven ß-galactosidase reporter cassette inserted into the E1 region of an E1-, E3-deleted adenovirus serotype 5 construct (Quantum Biotechnologies Inc., Quebec). The adenovirus vector is serially amplified in progressively larger cultures of 293 cells by inoculation into the culture media, incubation until cytopathic effect (CPE) is observed in the culture, followed by harvesting of the cells, which are then subjected to 3 freeze-thaw cycles to obtain a crude cell lysate preparation containing the released intracellular virus. Crude virus preparations from T175 flask inoculations are pooled and the virus is isolated and purified by standard cesium chloride gradient ultracentrifugation and dialysis. Serial dilutions of both crude and purified virus preparations are titered on NIH3T3 cells by staining infected cells in situ for ß-galactosidase activity using the chromogenic substrate 5-bromo-4-chloro-3-indoyl-D-galactopyranoside (X-gal; Sigma). Briefly, 48 hours after exposure to virus, the target cells are fixed in a phosphate-buffered saline (PBS) solution containing 4% paraformaldehyde and 0.1% glutaraldehyde (pH 7.4) and subsequently stained with a PBS solution containing 1 mg/ml X-gal, 5 mM potassium ferricyanide, 5 mM potassium ferrocyanide, and 2 mM MgCl2. Positive cells are visualized by blue staining, and the highest serial dilution showing positive cells is determined as the titer of the virus preparation in infectious units (IU) per ml. We have obtained Ad5CMVlacZ virus preparations with titers of 109-10 IU/ml to be used in transduction studies as a control, and for isolation of genomic adenoviral DNA for use in transfections to generate vectors containing genes of interest.The shuttle plasmid pAdCMV-GFP contains the left ITR and a 3-kb region flanking the deleted E1 region. The E1 region has been replaced with an expression cassette containing the CMV promoter driving the GFP marker gene. A second CMV promoter with a multiple cloning site (MCS) and polyadenylation signal sequence is present just upstream of the GFP cassette. Genes of interest will be cloned into the MCS for co-expression with the GFP marker. Expression of the GFP marker can be detected by UV fluorescence microscopy or by fluorescence-activated cell sorter (FACS) analysis, and this can be used to determine the transduction efficiency for each experiment, and allow us to normalize our results.
The shuttle plasmid must then undergo homologous recombination with the genomic adenoviral DNA in order to generate a vector containing the gene of interest. Purified Ad5CMVlacZ adenoviral DNA is cut with Cla I, and 5 µg co-transfected into 293 cells along with 5 µg of pAdCMV-GFP containing the gene of interest (GOI). The 293 cells are then overlayed with Seaplaque agarose and incubated for 5-7 days until plaque formation is observed. An additional agarose overlay containing Bluo-gal (GibcoBRL) is then applied and incubated overnight. Successful homologous recombination results in replacement of the CMVlacZ gene reporter cassette with the CMV-GOI-CMV-GFP cassettes, and so plaques are selected for lack of blue staining, and individual plaque lysates inoculated into fresh 96 well plates of 293 cells. The cells are incubated until complete lysis is observed, and subsequently re-inoculated into fresh 24 well plates of 293 cells. Thereafter, the adenovirus isolates are serially amplified and harvested as described above, and individual isolates screened to confirm the presence of the transgene by PCR.
Figure 3: Homologous recombination between adenoviral genomic DNA and shuttle construct containing the gene of interest (GOI) and GFP marker. (See text for details)

D) Library of Currently Available Vectors

As shown in the table below, we currently have accumulated a library of previously constructed vectors that will be made available to all CURE: DDRCC investigators. These include lentivirus, retrovirus, and adenovirus vector constructs that express angiogenesis inhibitors, signal transduction pathway intermediates, immunostimulatory cytokines and co-stimulator proteins, anti-inflammatory proteins, as well as other proteins related to cellular DNA damage repair and senescence.

Table 1. Currently available vector library

Vector type Insert sequences
Angiogenesis inhibitors

lentivirus Thrombospondin-1 human
Endostatin murine
Angiostatin murine
soluble VEGF-R II human, murine

adenovirus Thrombospondin-1 human
Endostatin murine
Angiostatin murine

Signal transduction pathways

lentivirus FAK wild type, Y397F, K454R, FRNK
Ras wild type (k-ras), N116Y, C-del
Src wild type, v-Src, Y416F, Y529F
PKC-d wild type, K376K
Rho family RhoA (L63 and C3/N19)
Rac1 (L61 and N-17)
Cdc42 (L61 and N-17)
Nck (a & b) wild type, SH2, SH3-I, SH3-II, SH3-III, 3SH3
Pak1 wild type, K299R, P13A, H83+86L, K299R+H83+86L, T423E
MAPKs ERK1/2 (wt and DN of Mek1)
JNK (wt and DN of JNKK2)
ERK5 (wt and DN of MEK5)
p38 (wt and DN of p38a, p38b, and MKK3/6)
TGFbR-II wild type, K-
Akt-1 wild type, K-
STATs Stat 1 (wt and DN)
Stat 3 (wt and DN)
SHP-2 wild type, Phosphatase -
adenovirus Ras wild type (k-ras), N116Y, C-del
Immunostimulatory cytokines and co-stimulators:

retrovirus GM-CSF human, murine
IL-2 human, murine
IL-4 human, murine
mM-CSF murine
CD80 human, murine

lentivirus FLT3L human, murine
GM-CSF human, murine
IL-4 human, murine
CD28 human, murine
CD40L human, murine
CD70 human
CD80 human, murine
L-selectin human, murine
Anti-inflammatory proteins:

lentivirus IL-1 receptor antagonist protein (IRAP) murine
soluble IL-1 type 2 receptor murine
Suppressor of cytokine signaling (SOCS)-3 murine

Other/DNA damage-related:

lentivirus Telomerase reverse transcriptase (hTERT) human
Helicase (Werner’s syndrome) human
p53 human

BACK