ARCAs for Prostate Cancer CV706 and CV787

A. Adenovirus: Gene Expression and Regulation

Members of the human Adenoviridae family were first cultured from the tonsils and adenoids of children in 1953 [29]. They represent 51 different serotypes which are distinguishable by antibody reactivity to epitopes on the virion surface. Each serotype is assigned to one of five subgroups (A-E). Adenovirus type 5 (Ad5), a member of Subgroup C, is associated with a self-limiting, febrile respiratory illness and ocular disease in humans; infectious virus can be recovered from the throat, sputum, urine, and rectum. Ad5 is also associated with renal impairment, hepatic necrosis, and gastric erosions in immunosuppressed individuals [30, 31]. Ad5 and the other Subgroup C viruses have little or no oncogenic potential in mammals [32].

The adenovirus type 5 genome is a double-stranded DNA molecule of 35,935 base pairs [33] containing short inverted terminal repeats [34]. Expression of the genome is a regulated cascade which is arbitrarily divided into early (E) and late (L) phases, with viral DNA replication required for maximal L gene expression. Related RNA transcripts are grouped according to the region of the genome from which they are transcribed as well as by the timing (E or L) of their expression. Viral gene expression is regulated at the levels of transcription, posttranscriptional modification (splicing), translation, and posttranslational modification. Products of the El region are essential for efficient expression of the other regions of the adenovirus genome. The El A transcription unit is the first Ad sequence to be expressed during viral infection and its products play a crucial role in a number of important biological functions in adenovirus-infected cells. The E2 region encodes several proteins that are required for viral DNA replication. These include a DNA-binding protein [35], the viral DNA-dependent DNA polymerase, and the DNA terminal protein that are required for DNA replication [36-38], The E3 region is not essential for replication in tissue culture and this region is deleted from most first-generation therapeutic adenoviruses [39, 40], Proteins encoded by the E3 region modulate host immune responses to infection by inhibiting transport of the MHC class I protein to the cell surface, thereby impairing the cytotoxic T lymphocyte (CTL) response [41-43], and by blocking TNFA-induced cytolysis of infected cells [44-46], Significantly, all natural isolates of adenovirus contain the E3 region. Seven transcripts of the E4 region have been identified. Some of the encoded proteins interact with and/or modulate the activity of El region proteins.

The onset of viral DNA replication signals the switch from E to L gene expression. Although the precise mechanisms are not fully understood, this transition requires both cis- and trans-acting factors [47-49]. Late genes primarily encode the structural components of the virion and the nonstructural scaffolding proteins that are essential for the assembly of infectious virus. It is estimated that up to 10,000 adenovirus virions accumulate per cell and most remain cell-associated [50]. The entire adenovirus replication cycle is complete in approximately 32-36 h [51]. Host range mutants of adenovirus have played a significant role in elucidation of virus functions. ARCAs using transcriptional response elements create host range mutants where replication is restricted to a particular cell type. The cytotoxicity associated with virus replication (lysis) and the vaccine nature of expressing highly visible foreign capsid antigens should be limited to a certain type of cell.

B. Tissue Specificity of ARCA

We hypothesized that tropism of a virus could be redirected if expression of an essential viral gene could be controlled. Viruses generated from this approach would have the same capsid as its parental virus and they should be able to penetrate all cell types that express the CAR receptor. Presumably, in all cells containing the CAR receptor, these viruses would follow the normal cell entry process: they would penetrate the endosome, fuse with the endosome membrane, reach the cytoplasm, find transport to the nucleus, and uncoat the viral DNA. In a normal adenovirus replication cycle the E1A gene is the only gene expressed during the first 2.5 h of infection [52-55]. In turn, the E1A proteins as transcription factors upregulate expression of the impending cascade of viral genes. However, we have genetically engineered prostate tissue-specific promoters and enhancers so as to drive the E1A genes. Viral replication should preferentially take place in cells that express the necessary transcription factors, thus enabling activation of the tissue or tumor-specific transcription regulatory elements. Thus, the E1A proteins should be preferentially expressed and the virus preferentially replicate in prostate cells.

There are several criteria important in regard to the transcriptional response element (TRE) necessary for the successful engineering of a therapeutic adenovirus: (1) the tissue-specific regulatory specificity must be tightly regulated, and transcription should be limited to tumor cells, or accessory cells with as few other sites of expression as medically tolerable, (2) the TRE must regulate the initiation of transcription of the adjacent gene, (3) the promoter must be strong enough to drive sufficient expression of essential viral genes, and (4) the TRE must be small enough to fit within the packaging limits of adenovirus. We chose prostate cancer and the TREs of PSA as our initial target.

Expression of the PSA gene is modulated by the prostate-specific enhancer (PSE) element that is located several thousand nucleotides upstream of the PSA

promoter [56]. When fused to a fragment (position —230 to +7, relative to the start of transcription) containing the PSA promoter, the PSE (position —5322 to —3875, relative to the start of transcription) confers tissue-specific expression on the reporter gene chloramphenicol acetyl-transferase (CAT) [56]. Sequence analysis of the PSE reveals the presence of regions with homology to steroid-response elements (SREs) and to binding sites for several cellular transcription factors including c-Fos and AP-1 [23, 56, 57]. A functional androgen-response element (ARE) within the PSE increases expression up to 100-fold in the presence of testosterone or the nonmetabolized testosterone analog R1881.

1. ARC As Containing One Prostate-Specific Transcriptional Response Element

To test the feasibility of the ARCA technology, we engineered the PSE fragment into the adenovirus genome and generated a first generation virus, CV706. CV706 contains the PSE fragment (PSA promoter and enhancer) inserted immediately upstream of the El A region and transcription of the E1A region is regulated by the PSE (Table I). Virus characterization showed that CV706 was able to efficiently replicate in PSA+ prostate carcinoma cell lines but not in the other PSA" human cell lines HBL-100, MCF-7, PANC-1, OVCAR-3. CV706 also does not replicate efficiently in DU-145, a prostate cancer cell line which does not express PSA and does not contain the androgen receptor [21]. Further study indicated that the transcription of the El A mRNA was regulated by the PSE. El A mRNA was detectable in PSA+ LNCaP cells,

Table I

ARCAs for Prostate Cancer

Table I

ARCAs for Prostate Cancer

Virus

E1A

E1B

E3

E4

Targeting cell

driven by

driven by

region

driven by

CV702

wt

wt

Deleted

wt

N/A

CV706

PSE

wt

Deleted

wt

Prostate cancer

CV711

wt

PSA

Deleted

wt

Prostate cancer

CV716

PSE

PSE

Deleted

wt

Prostate cancer

CV730

No E1A

wt

Deleted

wt

N/A

CV737

PB

wt

Deleted

wt

Prostate cancer

CV738

wt

PB

Deleted

wt

Prostate cancer

CV739

PB

PSE

Deleted

wt

Prostate cancer

CV740

PB

PB

Deleted

wt

Prostate cancer

CV757

wt

wt

Deleted

PSE

Prostate cancer

CV763

HK2

wt

Deleted

wt

Prostate cancer

CV764

PSE

HK2

Deleted

wt

Prostate cancer

CV787

PB

PSE

Full-length

wt

Prostate cancer

CV802

wt

wt

Full-length

wt

N/A

but was not detectable in PSA" cells. E1A protein was also reduced by 99% in PSA" cells, compared to that in the PSA+LNCaP cells [21]. This indicates that the inserted PSE has successfully controlled expression of the E1A gene and the host range of this adenovirus mutant has been confined to a particular cell type.

We also showed that the tropism of adenovirus could also be changed when the E1B gene or the E4 gene was placed under the control of the PSE TRE. CV711, whose E1B gene is placed under the control of PSE, and CV757, whose E4 genes are driven by PSE, both replicate similarly to wildtype adenovirus in PSA+ cells but are highly attenuated in PSA" cells. Cell specificity of CV711 viruses is similar to CV706 and replicates similarly to wild-type virus in PSA+cells. In contrast, CV757, shows significantly greater specificity for PSA+cells. While CV757 grows similarly to wild-type in PSA+ cells, it suffers a very large reduction in the ability to replicate in PSA^ cells (data not shown). Thus, adenovirus mutants can be generated to target PSA+ cells when any one of the E1A, E1B, of E4 genes are driven by the PSE.

These observations have been confirmed with other prostate-specific TREs including the TREs for probasin and hK2. The rat probasin (PB) gene is developmentally regulated in the prostate by androgens. Induction of the rat probasin gene by androgens was shown to involve the participation of two different cis-acting DNA elements that bind the androgen receptor. An expression cassette carrying 426 bp of the PB gene promoter and 28 bp of the 5'-untranslated region was found to be sufficient to target expression of a bacterial CAT reporter gene specifically to the prostate epithelium [58]. It was also shown that the same 5'-flanking region of PB gene promoter fragment fused to the SV40 TAg gene could lead to the development of progressive forms of prostate disease that histologically resemble human prostate cancer in transgenic animals [58]. The promoter of the rat probasin gene was engineered into adenovirus to drive the expression of either the E1A gene or the E1B gene to generate CV737 and CV738, respectively. Both CV737 and CV738 showed significant specificity to PSA+prostate carcinoma cells.

We also recently cloned the TRE of the human glandular kallikrein (hK2) gene. The hK2 gene is located 12 kb downstream from the PSA gene in a head-to-tail fashion, whereas the hKl gene is located 30 kb upstream of the PSA gene in head-to-head fashion [59]. The PSA and hK2 gene share DNA (80%) and amino acid (78%) sequence homologies that suggests they evolved by gene duplication from the same ancestral gene [60, 61]. Interestingly, the hK2 protein was recently shown to be expressed in every prostate cancer, and the expression of hK2 protein incrementally increased from benign epithelium, to high-grade prostatic intraepithelial neoplasia, to adenocarcinoma. We recently described CV763 containing the hK2 promoter and enhancer driving the Ad5 El gene. CV763 behaved identically to CV706 [23].

Thus, the replication of adenovirus can be restricted to prostate cancer cells when one of the essential adenovirus genes E1A, E1B, or E4 is placed under the control of any one of three different prostate-specific TREs.

2. ARCAs Containing Two Prostate-Specific Transcriptional Response Elements

Since both the E1A and E1B genes are essential for adenovirus replication, we reasoned that it was possible to create a virus with significantly higher specificity if both the E1A and E1B genes were under independent control of two TREs. To test this hypothesis, we generated an adenovirus mutant CV716, in which both the E1A gene and the E1B gene were under the control of PSE. In vitro study showed that CV716 replicated well in the PSA-producing prostate cancer cells. However, replication of CV716 was highly attenuated in nonprostate human cell lines. Compared to CV706, the efficiency of CV716 replication in nonprostate cancer cells has been further reduced by another 100-fold, giving specificity for PSA+ cells compared to PSA" cells of nearly 10,000:1 (data not shown). The high degree of specificity for PSA+ cells of CV716 as compared to PSA~cells was found to be universally true [22, 23]. CV740, containing duplicate copies of the rat probasin promoter, also showed this high level of specificity. Unfortunately, CV716 and CV740 are genetically unstable, resulting in self-inactivation of the virus. The El A gene and one copy of the tissue-specific TRE inserts are deleted during replication. Southern blot analysis of stocks of CV716 indicated a new band when annealed with an ElB-labeled probe. DNA sequence analysis of the cloned deletion mutant indicated that self-inactivation is due to homologous recombination between two identical inserted TREs.

In order to make a stable tissue-specific adenovirus we employed two different TREs to drive expression of early essential viral genes. In CV739 the E1A gene and the E1B genes are under the control of the TRE of the rat probasin gene and PSA gene, respectively. CV739 replicates well in PSA+ prostate cancer cells, but poorly in nonprostate human cancer cell lines. The cell specificity of CV739 was similar to that of CV716, again showing the roughly 10,000:1 selectivity for PSA+ cells as compared to PSA- cells. However, CV739 is stable. No replication-defective mutants with deleted genomes were found after extensive passages. The same is true for other CV739-like viruses including CV764. CV764 is a stable ARCA variant containing the PSE driving the El A genes and the hK2 promoter and enhancer driving the E1B genes. The sequences of the PSE and hK2 promoter and enhancer are 80% identical, yet the virus is genetically stable. Again, CV764 has the high therapeutic index of the other viruses containing two prostate-specific TREs with a cell specificity of 10,000:1 for PSA+ cells compared to PSA-cells [22, 23],

Taken together, these adenovirus variants show that tropism of adenovirus can be redirected by placing essential viral genes under the control of tissue-specific regulators. The cell selectivity of a stable oncolytic virus can be over 10,000:1 when the expression of more than one viral gene is driven by two different tissue-specific TREs. The phenomenal success of creating adenovirus host range mutants with specificity for target cells compared to nontarget cells of over 10,000:1 is one of the major achievements of the ARCA technology.

C. Antitumoral Efficacy of ARCA

In vivo studies evaluating intratumoral and intravenous administration of prostate-specific adenoviruses were conducted in the nu/nu mouse containing human tumor xenografts. Tumors were produced by subcutaneous injection of PSA+-producing prostate cancer LNCaP cells into each flank of each mouse, and after establishment of palpable tumors (mean tumor volume 300 mm3), the tumors were directly injected with purified virus or vehicle (PBS and 10% glycerol). Tumor growth was then followed for 6 weeks, at which time the mean tumor volume in each group was determined. A significant antitumoral activity was observed in the in vivo study for CV706. Tumor volume dropped by more than 80% in the animal group that was treated with CV706 by a single intratumoral injection. These residual tumor masses were shown by histology to be scar tissue devoid of PSA+ cells. After 6 weeks, 5 of 10 mice were visually free of tumor [21].

In contrast, DU145 is a prostate cancer cell line that is PSA" and does not produce the androgen receptor. Tumors of DU145 cells were induced in nude mice and challenged with buffer, wild-type Ad5 but E3 virus CV702 and CV706. The results showed that CV702 inhibited growth of DU145 tumors, whereas CV706 has no effect on tumor growth. Thus, the prostate-specific CV706 virus not only shows efficacy but also selectivity for PSA+cells in vivo [21].

The E3 region has long been considered unnecessary for replication of adenovirus in vitro. It has been universally deleted from Ad5 gene therapy constructs until recent efforts to prolong transgene expression from replication-defective Ad5 gene therapy constructs [39,40, 62-64], To test the possibility of increasing virus cytotoxicity, we created CV787 from its parent virus CV739 by engineering the full-length E3 region back into the viral genome. Thus CV787 contains the rat probasin promoter driving the E1A gene the PSE driving the E1B gene. Otherwise, CV787 is identical to the recombinant wild-type adenovirus CV802. CV787 retained the high specificity of characteristics of two TRE-containing viruses driving the E1A and E1B genes. Cell viability assay and virus yield assay demonstrated that addition of E3 aids virus replication and increases virus cytotoxicity. Thus CV787 has a stronger cytotoxicity than CV739 [22],

The increased cytotoxicity due to the Ad5 E3 region was also confirmed in vivo in the LNCaP xenograft animal model. A single intratumoral injection of CV739 and CV787 yielded identical reduction of LNCaP xenografts. However, CV739 required 100-fold more virus to achieve the same effect as CV787 [22], A single intratumoral CV787 at a dose of 1 x 108 particles/mm3 was curative for animals 6 weeks after treatment (n — 8). A single intravenous injection of CV739 at a dose of 5 x 1010 particles could stop tumor growth, whereas CV787 at this dose level caused a fourfold reduction in tumor volume [22]. Six weeks following a single intravenous injection of 1 x 1011 particles, the sizes of tumors were reduced to less than 5% of their original size, and 8 of 14 mice were visually free of tumors. The residual tumors measurably present were immunohistologically devoid of PSA [22]. The serum PSA levels in mice injected intravenously with CV787 decreased to 5% of their starting values within 4 weeks. Intravenous administration of CV787 designed to treat LNCaP xenografts showed that 1 x 10n particles could eliminate 300 mm3 preexistent LNCaP xenografts, whereas 1 x 10n particles of CV706 administered intravenously only stabilizes tumors. A dose-response curve of 1 x 109 and 1 x 1010 CV787 particles administered as a single intravenous dose can stabilize and regress tumors, respectively, but not eliminate tumors. These data indicate that CV787 has a significantly improved antitumor activity and a single dose of intratumoral or intravenous administration can eliminate pre-existent tumors in animal models.

D. Mechanism for Cell-Killing of ARCA

Infection with adenovirus causes profound changes in host-cell macro-molecular synthesis that ultimately lead to cell death. Virion fiber protein inhibits macromolecular synthesis when applied directly to cells bearing the adenovirus receptor [65]; soluble penton protein causes cytopathic effects (CPEs) in susceptible cells that are similar to those caused by infectious virus [66]. Cell-specific DNA synthesis, export of cellular mRNAs from the nucleus to the cytoplasm, and cell-specific translation are all inhibited after infection, but the precise mechanisms are not completely understood.

The 243E1A protein induces the full range of classical apoptotic events by increasing the level of the host cellular tumor suppressor protein p53. The 289E1A protein induces apoptosis by a p53-independent mechanism that requires a product of the E4 region [67, 68]. The El A-induced activation of the apoptosis pathway(s) must be modulated by E1B proteins to ensure efficient virus replication prior to cell death [69]. Activation of the interferon-inducible RNase L pathway by the adenovirus-associated type I (VAI) RNA [70] may also contribute to the stimulation of apoptotic pathways in adenovirus-infected cells [71]. The E3 11.6-kDa adenovirus death protein also has a role in cell-killing and promotes the release of progeny virions from the cell [72, 73].

We have investigated how our oncolytic viruses kill tumors in the nu/nu mouse model. Immunohistochemical analyses were performed to assay for the de novo synthesis of CV787-encoded proteins in tumor xenografts and to examine the effects of treatment with CV787 on tumor morphology in vivo. The mice bearing human LNCaP tumors were injected intravenously on day 0 with 1 x 1011 particles of CV787 per animal. Tumors were excised from two animals on days 1, 3, 7, 14, 21, and 28. The tumors were cut into six pieces and each piece fixed, embedded in paraffin, and sectioned. Sections were stained for the presence of adenovirus protein by a double-antibody protocol with rabbit anti-Ad antibodies and Fast Red stain followed by a hematoxylin counterstain.

On day 1, intracellular staining for adenovirus protein was detected in less than 1% of the tumor cells examined in 12 sections from two tumors. Occasional small clusters of stained cells, as well as dispersed single stained cells, were visible. By day 3, large clusters of cells expressing adenovirus proteins were detected in one of the two excised tumors. In some instances, areas of tumor necrosis were adjacent to clusters of adenovirus protein positive cells. On day 7, intracellular staining for adenovirus proteins was detected in greater than 10% of the tumor cells examined in 12 sections from both excised tumors. Virus-infected cells within the tumor sections were prominent on Day 21 and increased to more than 90% of the microscopic field of the section by Day 28. These results demonstrated that CV787 replicated in and expressed virus-encoded gene products in the LNCaP xenografts. The increased distribution of virus protein-positive cells indicated that infectious progeny CV787 spread to adjacent cells within the tumor which was associated with progressive necrosis in vivo [22].

Adenovirus-induced apoptosis causes cell death in vitro, specifically at the late stage of infection [67], and this process may contribute to the therapeutic effect of oncolytic virus in vivo. LNCaP xenografts in athymic mice were treated on day 0 with vehicle alone or a total dose of 3.2 x 107 particles of CV706 per mm3 of tumor. Tumor biopsy specimens were taken on day 14, and 5-|xm sections were prepared and examined for apoptosis. Extensive areas containing apoptotic nuclei were detected in sections of tumors treated with CV706. More than 25% of nuclei were apoptotic in some sections from CV706-treated tumors. In contrast, less than 2% of nuclei were apoptotic in sections from tumors treated with vehicle alone.

Additionally, a visual change of tumor color has been documented when the animals receive tumor specific adenovirus variants. For example, the color of the established human LNCaP xenografts is black. However, the tumors will become white within 1 week after receiving a dose of CV706 or CV787 through either intratumoral or intravenous administration. Histological H&E staining analysis found that the virus-treated tumors had significantly fewer numbers of blood vessels when compared to the tumors treated with vehicle. It is unclear at this time as to the precise mechanism by which this reduction in blood vessel number is achieved. This can achieved either though direct

A. Vehicle B. CV706

v ■ v,/ t!i ■ * "-■■■' r^*^ ♦

  • i i ' M. f 1 * * S _
  • i * , V" « ' S "

> -i-

  • gt; * ?, t * i vyi im
  • v - , . C ' . . ' V*

Figure 1 CD31 staining for newly developed blood vessel. Tumors were harvested 14 days after receiving (A) vehicle or (B) 1 x 107 particles/mm3 tumor of CV706 and stained with anti-CD31 monoclonal antibody by immunohistochemical analysis.

damage of endothelial cells or indirectly through the destruction of tumor vasculature by extensive necrosis. CD31 is expressed constituitively on the surface of adult and embryonic endothelial cells and has been used as a marker to detect angiogenesis [74], Immunohistochemical staining was performed to examine the effect of virus treatment on tumor angiogenesis by using monoclonal antibody against CD31. Tumors treated with CV706 showed a significantly lower level of CD31-positive cells in the vessel when compared to vehicle treated tumors (Fig. 1). This observation suggests that CV706 may be inhibiting tumor angiogenesis to a significant extent. The change of tumor color from black to white has been a reliable early indication for antitumor efficacy of a virus.

10 Ways To Fight Off Cancer

10 Ways To Fight Off Cancer

Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.

Get My Free Ebook


Post a comment