| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Experimental Therapeutics |
Division of Hematology/Oncology [T. P. C., N. H. V., Y. Y. M., M. A. C.] and Department of Pathology [M. H. C., J. D. S.], Children’s Hospital Medical Center, Cincinnati, Ohio 45229; Division of Pediatric Hematology/Oncology, University of Wisconsin Children’s Hospital and Comprehensive Cancer Center, Madison, Wisconsin 53792 [E. J. D., A. D. H., A. S.]; University of Alabama, Birmingham, Alabama 35294 [V. K., D. T. C.]; Research and Development, GenVec Inc., Gaithersburg, Maryland 20879 [T. J. W.]; University of Colorado Health Sciences Center, Denver, Colorado 80262 [J. D.]; and Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania 19104 [J. M. B.]
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
v integrins, but had low or undetectable expression of the major attachment receptor, coxsackievirus-Ad receptor (CAR). Mutant Ads with ablated CAR binding exhibited only 5–20% of transgene expression in RMS cells seen with a wild-type vector, suggesting that residual or heterogeneous CAR expression mediated the little transduction that was detectable. Immunohistochemical analysis of archived clinical specimens showed little detectable CAR expression in five embryonal and eight alveolar RMS tumors. Stable transduction of the cDNA for CAR enabled both efficient Ad gene transfer and oncolysis for otherwise resistant RMS cells, suggesting that poor CAR expression is the limiting feature. Gene transfer to RMS cells was increased >2 logs using Ads engineered with modified fiber knobs containing either an integrin-binding RGD peptide or a polylysine peptide in the exposed HI loop. The RGD modification enabled increased oncolysis for RMS cells by a conditionally replicative Ad, Ad
24RGD, harboring a retinoblastoma-binding mutation in the E1A gene. Thus, the development of replication-competent vectors targeted to cell surface receptors other than CAR is critical to advance the use of Ad for treating RMS. |
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
30% (1)
. Therefore, there is a need for more effective, tumor-specific therapies with fewer adverse sequelae. The major promise of gene therapy for cancer lies in the potential to specifically target gene delivery and expression to defined cell populations, thus potentially avoiding unwanted side effects (3) . Although a variety of methods are available for gene transfer, Ad vectors offer a number of advantages. Ads can be easily produced to high titer, are capable of transducing a wide variety of cell types, and are quite stable (4) . In general, Ad-mediated gene transfer is capable of delivering transgenes to a higher percentage of cells than most other methods. In addition, methods such as the inclusion of cationic molecules have been developed that improve Ad-mediated gene transfer to a variety of cell types (5, 6, 7, 8) including RMS cells (9) .
Ad attachment to cells is mediated by the fiber knob, which projects outward from the virion and binds to the high-affinity CAR (10 , 11) . Recently, through crystal structure and sequence mutational analyses, fiber knob amino acid residues critical for binding CAR have been identified (12 , 13) . Mutagenesis of critical CAR-binding residues and insertion of a HA epitope from influenza virus (to permit virus propagation in cells expressing a HA-binding single-chain antibody) essentially ablated transduction for cells that were otherwise highly susceptible to wild-type Ad gene transfer (12 , 14) . The availability of such mutants should allow both the determination of the role of CAR for Ad gene transfer in different cell types and the opportunity to divert Ad transduction or infection away from being mediated by CAR.
The observation that CAR is expressed in many tissues (11 , 15) may in part explain the broad tissue tropism of Ad. Nevertheless, there are many cell types that are poorly transduced by Ad, including some tumor cells. In certain cases, negatively charged cell surface molecules like sialoglycoconjugates interfere with adenoviral binding and gene transfer (16) . In other cases, Ad gene transfer is limited solely because of the lack of CAR expression (17) . However, Ads can be modified to attach to other cell surface molecules to bypass CAR dependence. For example, Ad was successfully targeted to glioma cells lacking CAR expression by the addition of a bispecific antibody, which linked the fiber protein to the epidermal growth factor receptor (18) . Genetic modification of the Ad fiber protein has also facilitated Ad gene transduction to cells that lack CAR expression by providing a "viral ligand" that binds cell surface proteins other than CAR. The insertion of a peptide containing an arginine-glycine-aspartate (RGD) integrin recognition site to the COOH terminus of the fiber protein markedly increased Ad transduction of fibroblasts, endothelial cells, and smooth muscle cells (19 , 20) . Insertion into the exposed fiber protein HI loop of a polypeptide containing an RGD motif enhanced viral transduction to ovarian carcinoma cells in culture and in vivo (21 , 22) ; gene transfer was also increased to murine liver, kidney, spleen and lung after i.v. administration (23) . A similar RGD sequence placed in the virus hexon coat protein also increased gene transfer to cells that were normally refractory to Ad transduction (24) . Similarly, the addition of a cationic polylysine sequence (pK7) to the fiber knob COOH terminus, which facilitates viral binding to negatively charged cell surface molecules such as heparan sulfates, increased transduction to macrophages, endothelial cells, smooth muscle cells, fibroblasts, and T cells (19 , 20) . Therefore, Ads have the potential to be genetically redirected in their delivery of therapeutic genes.
The major hurdle to date of gene therapy strategies for cancer is the relatively poor gene transfer efficiency using currently available methods, particularly in vivo. Although progress has been made, no reliable method exists for achieving gene transfer to all cancer cells in a tumor mass or all metastatic sites in vivo. One approach to cancer therapy that bypasses the need to deliver genes to all cells up front is to exploit the fact that cells are killed by permissive viral replication.
Replicating viruses have long been proposed as effective anticancer (oncolytic) agents (25)
. Increased understanding of viral genes and virus-cell interactions has sparked an interest in the development of attenuated viruses whose replication can be targeted to specific tumor cells. Such CRAds have been developed either by the mutation of a critical virus gene that is transcomplemented by a tumor cell (type I CRAd) or by the use of tissue-specific promoters to restrict expression of critical viral genes to the targeted cells (type II CRAd; Ref. 26
). Early clinical trial results suggest that such CRAds can be relatively safe and efficacious for treating human cancers (27)
. Importantly, CRAds have been shown in animal models to be capable of causing inhibition of s.c. tumor growth at distant sites after i.v. administration (28
, 29)
. Ad24 is an example of a type I CRAd that contains a deletion in the Rb-binding domain of the E1A gene; as such, the mutant virus is attenuated in nondividing cells due to the virus’s inability to release E2F from Rb-E2F, a critical step for Ad replication (30)
. Viral replication occurs in tumor cells that harbor inactivating Rb mutations or potentially in other dividing cells due to the physiological release of E2F during the cell cycle.
One RMS cell line, RD, has been previously shown to lack CAR expression and to resist infection by group B coxsackieviruses (31)
. In an effort to determine whether Ad-mediated oncolysis would be a feasible strategy by which to target destruction of RMS cells, we tested the effect of wild-type Ad-type 5 on RD and other RMS cells in culture for efficiency of both gene transfer and lytic infection. We found that RMS cells were variably susceptible to oncolysis. Although they expressed the adenoviral internalization receptors, v integrins, they had low or absent expression of the Ad attachment receptor, CAR. Primary tumors showed no detectable CAR expression by immunohistochemistry. Stable transduction of the CAR cDNA into RMS cells rendered them highly susceptible to Ad gene transduction and oncolysis. Mutant Ads with ablated CAR binding showed decreased transduction to RMS cells, suggesting that residual CAR expression mediated the little transduction that was detectable. Ads expressing modified fiber knob proteins that bypass CAR were capable of much higher levels of gene transfer for RMS cells. Finally, a genetically modified fiber knob markedly increased the ability of the Ad24 CRAd to cause oncolysis for RMS cells.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
AdGFP and Ad type 5 strain Adsub360, a replication-competent wild-type virus, were purchased from the University of Iowa Gene Transfer Vector Core. AdLacZ (34)
, AdLux (35)
, AdLucRGD (21)
, and Ad24 (30)
have been described previously. Ad24RGD (36)
contains the same Rb-binding mutation in E1A as Ad24 but also contains the same RGD peptide in the HI loop as AdLucRGD. AdL.PB(HA) contains a deletion of eight amino acids (HAIRGDTF; amino acids 337–344) in the RGD loop of penton base that ablates binding to integrins (37)
. The deleted amino acids were replaced with the sequence SRGYPYDVPDYAGTS (the italicized amino acids represent the HA peptide motif). AdL.F(RAEK-HA) contains four point mutations in the AB loop of the fiber knob (R412S, A415G, E416G, and K417G) that ablate binding to CAR (12)
. In addition, the HI loop contains an insertion of the amino acids SRGFKSYPYDVPDYAG between amino acids Gly443 and Asp444. Again, the italicized amino acids represent the HA peptide motif that enables growth of CAR-ablated vectors on 293-HA cells (14)
. AdL.PB(HA)F(RAEK-HA) is ablated for binding to both CAR and integrins through incorporation of the fiber and penton base mutations described above. AdL.F2K(pK7) contains a replacement of the Ad5 fiber knob with the Ad2 fiber knob, both of which bind to CAR. In addition, the Ad2 fiber knob contains an insertion of the sequence KKKKKKKSR between amino acids Gly538 and Thr539 in the HI loop that confers binding to heparan sulfate-containing receptors (38)
.
VP titers were determined by absorbance at 260 nm (39)
, and plaque-forming units were determined by standard plaque assay using 293 cells. Cell survival assays were performed in 96-well dishes using the CellTiter96 assay (Promega, Madison, WI). Values plotted in Figs. 2
, 7
, and FIG 9
represent the mean of three to five samples for each point. For clarity, SDs were omitted from the figures but were all within 15% of each mean value.
|
|
|
v integrin expression was identical, except that we used the FITC-conjugated anti-CD51 monoclonal antibody (Immunotech, Marseille, France) and no secondary antibody. In the experiment depicted in Fig. 4e–h
, cells were infected with 1000 VPs of AdLacZ (control) or AdGFP per cell and analyzed 48 h later by flow cytometry as described previously (9)
.
|
Reporter Gene Assay.
Cells were plated at 200,000 cells/well in 6-well dishes, and the next day, virus was added directly to the media. Media were changed after 4 h (except in the experiments in Figs. 6
and FIG 8B
, in which media were changed after 1 h). Cells were harvested in cell lysis buffer (Tropix, Bedford, MA) after 48 h (except for the experiments in Figs. 6
and FIG 8B
, in which cells were harvested after 24 h). Assays for luciferase activity were performed using the Luciferase Assay Kit (Tropix) according to the manufacturer’s instructions using a Monolight 2010 luminometer (Analytical Luminescence Laboratory, San Diego, CA). Data in all experiments represent the mean of three to five samples for each point. Protein concentrations were measured using the detergent-compatible Lowry protein assay (Bio-Rad Laboratories, Hercules, CA).
|
|
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, RMS cell lines showed variable susceptibility to Ad-mediated gene transfer. Specifically, the cell lines RD and RhRKM-T4 were poorly transduced, showing approximately 6–12-fold less expression than HeLa cells at the highest viral load tested. In contrast, the other RMS cell lines were equally as susceptible as or more susceptible than HeLa cells to Ad-mediated gene transduction. It should be noted that RhRKM-T4 was derived from RhRKM-P4 after being explanted from a mouse xenograft and therefore may have undergone extensive mutations accounting for the differences in transduction. These results demonstrate there is a wide range of Ad transduction efficiency for RMS cells in culture.
|
demonstrate that the control cells, A549 cells, were readily susceptible to Ad infection at all three viral loads tested. These cultures showed a complete CPE induced by wild-type Ad at 7 days using as few as 10 VPs/cell. The control virus, AdLacZ, had very little effect on cell survival. In contrast, the RMS cell lines tested were less susceptible or even resistant to oncolysis by Ad. RD and RhRKM-T4 cells showed little if any CPE after exposure to wild-type Ad, even in cells exposed to 1000 VPs/cell and followed for 13 days postinfection. The FMD24#3 and Rh18 cell lines were somewhat more susceptible, with a complete CPE seen using 100 and 1000 VPs/cell at 11 and 9 days, respectively. For Rh18 cells, the higher viral loads also showed a significant CPE using AdLacZ, suggesting that these cells may have a higher capacity to support E1-defective Ad replication, which varies among cancer cell lines (40)
.
RMS Cells Express the Ad Internalization Receptor but only Weakly Express the Attachment Receptor.
The most likely cause of poor adenoviral gene transfer and oncolysis for RMS cells is low expression of the virus receptors. Ad binds cells primarily through the interaction of its protruding fiber knob protein with cell surface CAR protein. Internalization of the virus is facilitated by the interaction of the virion penton base containing the consensus integrin-binding peptide sequence RGD with two members of the v integrin family, vß3 and vß5 (41)
. We therefore tested the expression of these proteins on the surface of the RMS cell lines in an attempt to correlate their expression with the variable Ad transduction and poor oncolysis of RMS cells.
Fig. 3
depicts flow cytometry profiles of the different cell lines using an anti-CAR antibody (RmcB) and an anti-v integrin antibody (anti-CD51). HeLa and A549 cells served as positive controls. The Ramos cell line has been shown to lack v integrins (42)
and therefore served as a negative control for anti-CD51. As can be seen from the images, all of the RMS cell lines expressed extremely low levels of CAR relative to the control cells. Differences between means of experimental (RmcB) and control curves suggested that RD and RhRKM-T4 cells exhibited the lowest expression of CAR, which was essentially undetectable in these cells by this method. The detectable (albeit scarce) CAR expression in the other RMS cells correlated with their susceptibility for gene transfer by Ad, which was higher than that for RD and RhRKM-T4 cells (Fig. 1)
. In contrast to CAR, v integrins were readily expressed by all of the RMS cells.
|
Expression of CAR Enhances Ad Transduction Efficiency.
The lack of CAR expression on RMS cell lines was thought to likely be responsible for poor Ad transduction in the RD and RhRKM-T4 RMS cell lines. However, sialoglycoproteins are expressed on RMS cells (43)
and have been shown to interfere with Ad binding (16)
. It was therefore possible that these or other molecules might play an inhibitory role in adenoviral gene transfer to RMS cells. To test whether expression of CAR would be sufficient to achieve efficient gene transfer to poorly transduced RMS cells, RD and RhRKM-T4 were stably transduced with a retrovirus encoding hCAR or empty vector (LSXN). G418-resistant cells were analyzed for CAR expression by flow cytometry. The isolates that displayed the highest levels of CAR expression were used for subsequent experiments (Fig. 4, a–d)
.
Analysis by flow cytometry of gene transfer to hCAR- and LXSN-transduced cells after exposure to AdGFP demonstrated that robust transgene expression was closely related to CAR expression (Fig. 4, e–h)
. Interestingly, in both RD-LXSN and T4-LXSN control cells, a small subpopulation appeared to be transduced by AdGFP as indicated by the "shoulder" of fluorescence (see arrow, Fig. 4, e and g
). To more closely quantitate the effect of CAR on transduction efficiency, these cell lines were transduced with different viral loads of AdLux (Fig. 5)
. The results for both RD- and RhRKM-T4-derived cell lines demonstrated that hCAR expression enhanced Ad-mediated gene transfer more than 10-fold. At virus loads of 102, 103, and 104 VPs/cell, transgene expression in RD-hCAR cells exceeded that in control-transduced cells (RD-LXSN) by 18-, 30-, and 25-fold, respectively. These virus loads in RhRKM-T4-hCAR cells resulted in transgene expression that was 61-, 82-, and 67-fold above that seen in T4-LXSN cells, respectively. Thus, the lack of CAR expression in RMS cells was a significant barrier to Ad-mediated transgene expression that was overcome by hCAR gene transfer.
|
was in fact due to their residual CAR expression, the v integrins, or other cell surface molecules. We examined the effects on transduction of four point mutations in the AB loop of the fiber knob that ablate fiber attachment to CAR in vector AdL.F(RAEK-HA). This mutant also contains a HA epitope in the fiber knob HI loop that allows virus propagation in 293-HA cells (14)
. We also tested the effect of a mutation that deleted the RGD motif from the penton base in vector AdL.PB(HA). Gene transfer was measured in Rh18 cells, which were most susceptible to transfer by wild-type Ad, and RD cells, which were least susceptible to transfer by wild-type Ad. To ascertain the effect of exogenous CAR expression on gene transfer, viruses were tested in both stable control-transduced (LXSN) and hCAR-transduced cells. Cells were exposed to virus for only 1 h and harvested at 24 h to maximize differences due to cell binding and entry.
As shown in Fig. 6
, transduction by vector AdL.F(RAEK-HA), which is defective in its attachment to CAR, was significantly less efficient than transduction by the wild-type Ad vector in both Rh18 and RD cells. The effect was 2.4–5.2-fold more dramatic in Rh18 cells (the mutant exhibited 4–7% of AdL expression) compared with RD cells (the mutant exhibited 17–21% of AdL expression), consistent with the data in Fig. 3
that indicated slightly higher CAR expression in Rh18 cells. A less dramatic effect was observed with deletion of the RGD sequence involved in integrin interactions using vector AdL.PB(HA). This vector exhibited a 
40% reduction of gene transfer in Rh18 cells but a minimal reduction in RD cells (there was a trend of reduced activity in RD-hCAR cells that was significant for the double mutant). These studies support the notion that residual CAR plays a role in mediating Ad gene transfer to RMS cells, despite being expressed at very low (Rh18) or undetectable (RD) levels by flow cytometry. In addition, the results suggest that the penton base integrin-binding sequence is primarily important only under conditions in which the fiber knob-CAR interaction is above a critical threshold, consistent with its previously described role in virus internalization subsequent to attachment (41)
.
Low CAR Expression Is the Limitation in Ad-mediated Oncolysis for RMS Cells.
Our data clearly show that lack of CAR expression is a barrier for Ad gene transfer to some of the RMS cells. However, several of the cell lines exhibited relatively robust Ad gene transfer similar to that seen in HeLa cells (Fig. 1)
but were not as readily killed by wild-type Ad as control cells (Fig. 2)
. Thus, resistance to Ad oncolysis might not be solely a function of viral entry into these cells. For oncolysis to occur, all of the steps in the viral life cycle, from cell entry, viral uncoating, nuclear transport, viral genome replication, and encapsidation to viral release, must occur. It was therefore possible that one or more of these steps might be limiting in the cells that showed relative resistance to oncolysis by Ad.
To test the role of CAR in Ad-mediated oncolysis, we exposed the hCAR-transduced RMS cells to wild-type Ad and followed them for cell survival. As shown in Fig. 7
, the control cells (RD-LXSN and T4-LXSN) were relatively resistant to oncolysis. The minimal toxicity seen with these cells was not different from that seen with the negative control, AdLacZ, except at the highest viral load (1000 VPs/cell). Even at that level of virus, these cultures did not reach a complete CPE by the end of the experiment (after 13 days). In contrast, cells stably transduced by hCAR were readily susceptible to Ad oncolysis. Both RD-hCAR and T4-hCAR showed a complete CPE by the first day tested (day 4) with 1000 VPs/cell and by day 7 with 100 VPs/cell. Both cell lines reached a complete CPE by day 13 with 10 VPs/cell. These results are similar to those seen with control A549 cells (see Fig. 2
) and demonstrate that low CAR expression is the major limitation to Ad-mediated oncolysis of RMS cells.
Ad Fiber Knob Modification Markedly Improves Transduction for RMS Cells.
The preceding results demonstrate that restoration of CAR expression markedly enhanced Ad-mediated gene transfer and oncolysis for RMS cells in culture. The corollary hypothesis is that viral modifications that obviate the requirement for CAR by attaching to other cell surface molecules might also enhance Ad-mediated gene transfer and oncolysis. We thought it likely that vectors modified to attach to cell surface molecules more abundant than CAR would provide more efficient gene delivery to CAR-deficient RMS cells. Because RMS cells express v integrins, we tested an Ad whose fiber knob has been modified to express a peptide containing an integrin-binding RGD motif in the exposed HI loop. This virus has been previously reported to be superior for gene transfer to some primary cancers and other cells (21
, 22)
.
As shown in Fig. 8A
cell lines were transduced with the modified vector (AdLucRGD) or the corresponding control vector (AdCMVLuc) at 1 pfu/cell. For each cell line, the modified vector showed higher transgene expression. In the RMS cell lines, the fold increase was highest (nearly 2 orders of magnitude) for those cells that exhibited the lowest transgene expression using the control vector (RD, T4-LXSN, and RD-LXSN); only a modest increase was seen in the other cells that had higher transgene expression using the unmodified virus (10-fold). An identical pattern was seen using 0.1 or 10 pfu/cell (data not shown). Interestingly, AdLucRGD transduction in RD-LXSN and T4-LXSN was better than control virus (AdCMVLuc) transduction in the hCAR-transfected cell lines (RD-hCAR and T4-hCAR). AdLucRGD transduction was further improved by the expression of hCAR in RhRKM-T4 and RD cells, although the improvement was less dramatic than that seen in the LXSN-transduced cells due to their higher baseline expression.
We also tested an Ad whose fiber knob was modified by the addition of a polylysine peptide into the fiber knob HI loop, AdL.F2K(pK7), in comparison with the corresponding virus containing a wild-type fiber knob, AdL. A very similar pattern to the effect of the RGD modification was observed, showing increased transduction in RMS cells lacking CAR with less effect in cells stably transduced with hCAR cDNA (Fig. 8B)
.
Ad Fiber Knob Modification Markedly Improves Oncolysis for RMS Cells.
To test whether the improved transduction of RMS cells with the RGD-modified fiber knob translated to improved oncolysis, we compared the Ad24 CRAd with a derivative containing the same fiber knob modification, Ad24RGD. Both viruses contain a 24-bp deletion in the Rb-binding domain of E1A (30)
. As such, they cannot replicate in quiescent cells that contain Rb-E2F complexes, but they can replicate in tumor cells harboring Rb mutations or in other dividing cells secondary to the normal, transient E2F release. Ad24, which expresses the wild-type fiber knob, had virtually no effect on the growth of RD-LXSN or RD-hCAR cells at the three virus loads tested (Fig. 9)
. However, Ad24RGD showed a marked oncolytic effect for both cell lines. The effect was most dramatic in the RD-hCAR cells, consistent with the higher gene transfer for these cells (Fig. 8)
, likely a result of two avenues of virus entry (mediated by CAR and RGD-binding integrins). 293 cells were used as a positive control for oncolysis by Ad24 because they express wild-type E1A (and CAR) that transcomplements the 24 mutation.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our experiments demonstrated that some RMS cell lines had equal or even higher Ad transgene expression than HeLa cells. These cell lines showed a correspondingly increased sensitivity to oncolysis by wild-type Ad as compared with RMS cell lines that were less transducible. Thus, there appeared to be a correlation between efficiency of transduction and oncolysis. However, even the cell lines that were readily transducible (Rh18 and FMD24#3) showed a slower CPE compared with control A549 cells. These cells had only slightly detectable CAR expression by flow cytometry. In addition, Ad transduction was apparent (although at a lower level) in RD and RhRKM-T4 cells, which were completely resistant to oncolysis. Whereas studies using Ads with ablated CAR binding suggested that CAR is important for Ad entry even for these cells, higher gene transfer and a high sensitivity to oncolysis could be achieved by forced overexpression of CAR. However, exogenous CAR expression did not have a marked effect for Rh18 cells, suggesting that low but detectable endogenous CAR expression was sufficient to mediate high gene transfer. These data are consistent with the report that RD cells, which are normally resistant to lytic infection by coxsackievirus B3, became susceptible to coxsackievirus B3 infection after undergoing high passage and acquiring CAR expression (44) .
Taken together, our results suggest that the most likely explanation for RD and RhRKM-T4 resistance to oncolysis is their only rare, heterogeneous expression of CAR. In support of our hypothesis, flow cytometry analysis of gene transfer by AdGFP on a cell-by-cell basis showed that a distinct, small population of cells was transduced (Fig. 4, e and g)
. If only a subset of cells is capable of uptaking Ad, then the virus would not be able to spread as a lytic infection to other cells. Such heterogeneous cell populations would explain the absence of significant oncolysis in the face of measurable gene transduction when analyzed by testing the cell culture as a whole. Our analysis of primary tumor specimens was consistent with low or absent CAR expression in RMS.
Ad-mediated oncolysis is a treatment strategy that is gaining increasing attention for a number of different human cancers. Whereas many cancer-derived cell lines in culture have been shown to be susceptible to Ad infection, low or heterogeneous expression of Ad receptors that attenuates viral oncolysis may be a common feature of primary tumors. Differential CAR expression has been shown to cause variable Ad gene transfer to head and neck cancers and was postulated to be responsible, at least in part, for the wide difference in responses seen in clinical trials (46) . Similarly, variable CAR expression limited the efficiency of Ad-mediated gene transfer for glioma (18) and bladder cancer cells (47) . Although the physiological function of CAR is unknown, forced CAR expression in prostate cancer cells slowed tumor growth (48) , suggesting that it may be involved in transmitting extracellular antigrowth signals and may therefore be teleologically down-regulated in cancer.
A variety of methods have been devised to increase Ad-mediated gene transfer to cells that are normally poorly transduced by Ad. Most notably, these involve the inclusion of cationic substances that facilitate virus-cell interactions (5, 6, 7, 8) . We have found that the addition of cationic lipids increased Ad-mediated gene transfer for RMS cells in vitro and after intratumoral injection (9) . Such methods may be useful for augmenting gene transfer; however, they should not in principle facilitate viral oncolysis (other than for the first wave of virus infection). Conversely, modifications of the Ad genome that improve viral attachment and replication and that will be carried by the Ad progeny would be ideal for improving Ad-mediated oncolysis. As such, a polylysine modification of the fiber knob improved oncolysis by the ONYX-015 E1B-deleted CRAd for glioma cells (49) . The modification of the Ad fiber knob by the addition of an RGD-containing peptide in the HI loop increased gene transfer to all RMS cells tested here. The fact that a much less dramatic increase was seen using the modified virus in cells that overexpress CAR (RD-hCAR and T4-hCAR) is consistent with low or absent CAR being the limiting factor in the other cell lines and argues against any trivial explanation (such as variability in viral quantitation assays) for the differences seen between the two modified and unmodified viruses.
The use of the RGD-modified fiber knob modification enabled a CRAd to kill RMS cells. Because an otherwise wild-type virus with the modified fiber knob might pose safety concerns due to expanded cell tropism, we used the Ad24 attenuated virus containing a Rb-binding mutation in E1A. Our data demonstrate that the RGD modification enabled oncolysis for RD RMS cells that lacked CAR expression. Oncolysis of RMS cells was further increased by exogenous CAR expression in the RD-hCAR cells, suggesting that the RGD modification might be useful for enhancing CRAd-mediated oncolysis for CAR-positive tumors as well. At lower virus concentrations, Ad24RGD appeared to be a more potent oncolytic agent than wild-type virus, inducing a complete CPE of RD-hCAR cells infected with 100 VPs/cell after 6 days as compared with 11 days for wild-type virus. The enhanced potency could be due in part to the E1A mutation because a similar Ad E1A mutant (dl922–947) with a wild-type fiber knob was recently shown to also have enhanced oncolytic potency compared with wild-type virus and other attenuated mutants both in vitro and in vivo (50)
. Our data suggest, however, that the increased potency of Ad24RGD resulted from enhanced cell entry because Ad24 (with a wild-type fiber knob) had no effect on RMS cells stably transduced with hCAR. This finding also suggests that the fiber knob modification was more effective at facilitating virus entry and spread than hCAR expression. Apparently, the enhanced virus entry gained by hCAR expression was sufficient to support wild-type Ad replication but not Ad24 mutant virus replication.
The clinical utility of Ad24RGD to selectively treat RMS has yet to be determined. Although selective in its replication because of the E1A mutation, Ad24RGD theoretically will replicate in dividing cells that provide E1A-like functions in trans (40)
or that physiologically release E2F from Rb during the S phase of the cell cycle. Indeed, the similar E1A mutant dl922–947 showed some replication (albeit attenuated) in proliferating normal endothelial and epithelial cells (50)
. Whether the replication of Ad24RGD is sufficiently restricted for use in treating RMS tumors, which can occur at essentially all body sites, is not yet known. Nevertheless, our data clearly demonstrate that the development of oncolytic Ads more specifically restricted for RMS must include modifications that permit CAR-independent cell entry.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 Supported in part by University of Wisconsin Comprehensive Cancer Center Grant P30-CA-14520-25. T. P. C. was supported by NIH Grant CA75113, the Midwest Athletes Against Childhood Cancer, and an institutional faculty development award to the University of Wisconsin from the Howard Hughes Medical Institute. J. D. was supported by the V Foundation, the Howard Hughes Medical Institute, and NIH Grant CA77314. J. M. B. was supported by NIH Grants HL54734 and AI35667 and by an Established Investigator Award from the American Heart Association. D. T. C. was supported by NIH Grants CA68245, HL50255, and CA74242. 
2 To whom requests for reprints should be addressed, at Division of Hematology/Oncology, Children’s Hospital Medical Center, 3333 Burnet Avenue, Cincinnati, OH 45229. Phone: (513) 636-7241; Fax: (513) 636-3549; E-mail: t.cripe{at}chmcc.org
3 The abbreviations used are: RMS, rhabdomyosarcoma; Ad, adenovirus; CAR, coxsackievirus-Ad receptor; Rb, retinoblastoma; HA, hemagglutinin; CRAd, conditionally replicative Ad; hCAR, human CAR; AdGFP, AdCMV-humanized green fluorescent protein; VP, viral particle; CPE, cytopathic effect; pfu, plaque-forming unit; CMV, cytomegalovirus.
Received 1/31/00. Accepted 1/30/01.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
vß3 and vß5 promoter adenovirus internalization but not virus attachment. Cell, 73: 309-319, 1993.[Medline]
This article has been cited by other articles:
|
|
 |
|
  H.-J. Li, M. Everts, M. Yamamoto, D. T. Curiel, and H. R. Herschman Combined Transductional Untargeting/Retargeting and Transcriptional Restriction Enhances Adenovirus Gene Targeting and Therapy for Hepatic Colorectal Cancer Tumors Cancer Res., January 15, 2009; 69(2): 554 - 564. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Tsuruta, L. Pereboeva, M. Breidenbach, D. T. Rein, M. Wang, R. D. Alvarez, G. P. Siegal, P. Dent, P. B. Fisher, and D. T. Curiel A Fiber-Modified Mesothelin Promoter-Based Conditionally Replicating Adenovirus for Treatment of Ovarian Cancer Clin. Cancer Res., June 1, 2008; 14(11): 3582 - 3588. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M.H. Hassan, N. Khatoon, D.T. Curiel, F.M. Hamada, H.M. Arafa, and A. Al-Hendy Toward gene therapy of uterine fibroids: targeting modified adenovirus to human leiomyoma cells Hum. Reprod., March 1, 2008; 23(3): 514 - 524. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Tsuruta, L. Pereboeva, J. N. Glasgow, D. T. Rein, Y. Kawakami, R. D. Alvarez, R. P. Rocconi, G. P. Siegal, P. Dent, P. B. Fisher, et al. A Mosaic Fiber Adenovirus Serotype 5 Vector Containing Reovirus {sigma}1 and Adenovirus Serotype 3 Knob Fibers Increases Transduction in an Ovarian Cancer Ex vivo System via a Coxsackie and Adenovirus Receptor Independent Pathway Clin. Cancer Res., May 1, 2007; 13(9): 2777 - 2783. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Gomez-Manzano, M. M. Alonso, W.K. A. Yung, F. McCormick, D. T. Curiel, F. F. Lang, H. Jiang, B. N. Bekele, X. Zhou, R. Alemany, et al. Delta-24 Increases the Expression and Activity of Topoisomerase I and Enhances the Antiglioma Effect of Irinotecan Clin. Cancer Res., January 15, 2006; 12(2): 556 - 562. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Kuhnel, B. Schulte, T. Wirth, N. Woller, S. Schafers, L. Zender, M. Manns, and S. Kubicka Protein Transduction Domains Fused to Virus Receptors Improve Cellular Virus Uptake and Enhance Oncolysis by Tumor-Specific Replicating Vectors J. Virol., December 15, 2004; 78(24): 13743 - 13754. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Chu, D. E. Post, F. R. Khuri, and E. G. Van Meir Use of Replicating Oncolytic Adenoviruses in Combination Therapy for Cancer Clin. Cancer Res., August 15, 2004; 10(16): 5299 - 5312. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Gomez-Manzano, W.K. A. Yung, R. Alemany, and J. Fueyo Genetically modified adenoviruses against gliomas: From bench to bedside Neurology, August 10, 2004; 63(3): 418 - 426. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Gu, A. Ogose, H. Kawashima, M. Ito, T. Ito, A. Matsuba, H. Kitahara, T. Hotta, K. Tokunaga, H. Hatano, et al. High-Level Expression of the Coxsackievirus and Adenovirus Receptor Messenger RNA in Osteosarcoma, Ewing's Sarcoma, and Benign Neurogenic Tumors among Musculoskeletal Tumors Clin. Cancer Res., June 1, 2004; 10(11): 3831 - 3838. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Qin, S. Chen, T. Yu, B. Escuadro, S. Sharma, and R. K. Batra Coxsackievirus Adenovirus Receptor Expression Predicts the Efficiency of Adenoviral Gene Transfer into Non-Small Cell Lung Cancer Xenografts Clin. Cancer Res., October 15, 2003; 9(13): 4992 - 4999. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 O. Hamid, M. L. Varterasian, S. Wadler, J. R. Hecht, A. Benson III, E. Galanis, M. Uprichard, C. Omer, P. Bycott, R. C. Hackman, et al. Phase II Trial of Intravenous CI-1042 in Patients With Metastatic Colorectal Cancer J. Clin. Oncol., April 15, 2003; 21(8): 1498 - 1504. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. W. Schoggins, J. G. D. Gall, and E. Falck-Pedersen Subgroup B and F Fiber Chimeras Eliminate Normal Adenovirus Type 5 Vector Transduction In Vitro and In Vivo J. Virol., December 20, 2002; 77(2): 1039 - 1048. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. S. Haviv, J. L. Blackwell, A. Kanerva, P. Nagi, V. Krasnykh, I. Dmitriev, M. Wang, S. Naito, X. Lei, A. Hemminki, et al. Adenoviral Gene Therapy for Renal Cancer Requires Retargeting to Alternative Cellular Receptors Cancer Res., August 1, 2002; 62(15): 4273 - 4281. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Hemminki, K. R. Zinn, B. Liu, T. R. Chaudhuri, R. A. Desmond, B. E. Rogers, M. N. Barnes, R. D. Alvarez, and D. T. Curiel In Vivo Molecular Chemotherapy and Noninvasive Imaging With an Infectivity-Enhanced Adenovirus J Natl Cancer Inst, May 15, 2002; 94(10): 741 - 749. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. W. van Beusechem, J. Grill, D. C. J. Mastenbroek, T. J. Wickham, P. W. Roelvink, H. J. Haisma, M. L. M. Lamfers, C. M. F. Dirven, H. M. Pinedo, and W. R. Gerritsen Efficient and Selective Gene Transfer into Primary Human Brain Tumors by Using Single-Chain Antibody-Targeted Adenoviral Vectors with Native Tropism Abolished J. Virol., February 22, 2002; 76(6): 2753 - 2762. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Kim, T. Smith, N. Idamakanti, K. Mulgrew, M. Kaloss, H. Kylefjord, P. C. Ryan, M. Kaleko, and S. C. Stevenson Targeting Adenoviral Vectors by Using the Extracellular Domain of the Coxsackie-Adenovirus Receptor: Improved Potency via Trimerization J. Virol., February 15, 2002; 76(4): 1892 - 1903. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
) or a modified fiber knob (
). Fiber knob modifications included insertions into the HI loop of an RGD (A) or a polylysine (B) peptide. In A, 1 pfu/cell of each virus was added to the cells for 4 h before media change, and the cells were harvested at 48 h. In B, 10 VPs/cell of each virus were added to the cells for 1 h before media change, and the cells were harvested at 24 h.
