Short Communication | Published:

Analysis of genetically engineered oncolytic herpes simplex viruses in human prostate cancer organotypic cultures

Gene Therapy volume 16, pages 14771482 (2009) | Download Citation

Abstract

Oncolytic herpes simplex viruses type 1 (oHSVs) such as G47Δ and G207 are genetically engineered for selective replication competence in cancer cells. Several factors can influence the overall effectiveness of oHSV tropism, including HSV-1 receptor expression, extracellular matrix milieu and cellular permissiveness. We have taken advantage of human prostate organ cultures derived from radical prostatectomies to investigate oHSV tropism. In this study, we show that both G47Δ and G207 specifically replicate in epithelial cells of the prostatic glands but not in the surrounding stroma. In contrast, both the epithelial and stromal cell compartments were readily infected by wild-type HSV-1. Analysis of oHSV replication in prostate surgical specimens 3 days post infection showed that G47Δ generated 30-fold more viral progeny than did G207. This correlated with the enhanced expression of G47Δ-derived glycoprotein gB protein levels as compared with G207. In benign prostate tissues, G207 and G47Δ titers were notably reduced, whereas strain F titers were maintained at similar levels compared with prostate cancer specimens. Overall, our results show that these oncolytic herpes vectors show both target specificity and replication competence in human prostate cancer specimens and point to the utility of using human prostate organ cultures in assessing oHSV tropism and cellular specificity.

Main

Oncolytic virotherapy is based on the selective killing of tumor cells while sparing the surrounding normal cells. Selectively replication-competent viral vectors, such as oncolytic herpes simplex viruses type 1 (oHSVs) which are genetically engineered from herpes-simplex virus (HSV) type I, represent an attractive strategy for tumor-based therapies because these viruses can specifically replicate and spread in cancer cells in situ, exhibiting oncolytic activity through direct cytopathic effects while sparing normal cells. G207 is a previously described oHSV that has been used in clinical trials.1 It contains deletions of both copies of gamma34.5 and a LacZ insertion within the ICP6 gene causing its inactivation.2 G47Δ is a replication-competent HSV-1 vector derived from G207 that also contains an additional deletion within the nonessential α47 gene.3 A deletion of α47 places the late US11 gene under the control of the immediate early α47 promoter. As a consequence, US11 expression immediately after infection results in the enhanced growth of G47Δ by precluding the shutoff of protein synthesis.4, 5 oHSV vectors have undergone extensive preclinical and clinical testing. Our laboratory has previously shown that G47Δ is more effective than G207 in several tumor models.3, 6

Importantly for prostate cancer, oHSV vectors have been shown to be nontoxic to normal prostate and the surrounding nerves after intraprostatic inoculation in mice and in non-human primates.7, 8 In addition, oHSVs are effective against human prostate cancer irrespective of hormone status or radiosensitivity,6, 9, 10 a major advantage in its application for advanced forms of the disease, in which hormone and radiation refractory tumor is a common cause of progression. Moreover, preexisting immunity does not preclude the efficacy of oHSV vectors in mice.11, 12, 13 This is critical for translational purposes, as many humans are seropositive for HSV.

Although the use of oHSV for tumor therapy has shown promise, several issues that have hampered their further development for use in clinics still remain. Some of the factors that may limit the overall effectiveness of oHSV as a cancer therapeutic modality include expression of HSV-1 entry receptors, extracellular matrix barrier and intracellular resistance mechanisms.14 Strategies developed to assess oHSV tropism in monolayer tissue culture do not recapitulate the three-dimensional tumor physiology of multiple cell types as observed in patient and in-patient specimens.

Both human and murine prostate organ cultures have been used extensively to study a wide variety of biological processes ex vivo.15, 16, 17, 18 We have taken advantage of a prostate organ culture system derived from radical prostatectomies for cancer and from surgeries removing benign prostatic tissue (BPT) to assess oHSV tropism. The advantages of prostate organ cultures are the following: (1) the use of primary human prostate cancer material (as opposed to passaged prostate cancer cell cultures or animal models); (2) the three-dimensional structure remains intact (in contrast to typical monolayer cultures; (3) tumor foci exist at different stages of tumor progression; and (4) the genetic heterogeneity of cancer cells can be studied within and between patient samples. Therefore, the factors that may affect viral entry and replication including cell–cell interactions and cell–matrix interactions remain preserved in this three-dimensional milieu and are amenable to exploration.

Assessment of prostate organ cultures for viability and HSV-1 entry receptor expression

Human prostate specimens derived from radical prostatectomies were assessed for both tissue integrity and viability either immediately after resection or 72 h after incubation ex vivo on a semi-submersed collagen sponge (Ultrafoam, Davol, Cranston, RI, USA). A histological analysis showed that the morphological appearance of prostatic glandular regions as well as the surrounding stroma remained intact over the 72-h incubation period. It is noted that, epithelial cell–cell contacts within these glandular regions appeared to be preserved as well as epithelial cell contacts contiguous with the basement membrane (Figure 1a). The cell viability of prostate tissues was also assessed using MTS assays (Promega, Madison, WI, USA). As shown in Figure 1b, MTS enzymatic activity was similar for freshly isolated tissues as compared with those cultured over a 3-day period. Nectin-1 is a major cell-surface receptor for HSV-1 entry, the expression of which correlates with oHSV susceptibility and has been shown to be ubiquitously expressed in multiple tissues, including in the prostate.19, 20 Therefore, we assessed the tissue distribution of nectin-1 in prostate surgical specimens by immunohistochemistry. Anti-nectin-1 (R166) antibody21, 22 strongly stained epithelial cells within glandular regions of the prostate while a faint staining pattern appears in the surrounding stroma. The staining of the epithelia appeared enhanced along their edges and at epithelial cell–cell contacts (Figure 1c). Overall, organ cultures derived from prostate surgical specimens represent a viable target for assessing oHSV tropism.

Figure 1
Figure 1

Analysis of human prostate organ culture viability and nectin-1 tissue distribution. (a) Hematoxylin staining (H&E) staining of prostate tissue specimens. Tissues were obtained from the Department of Pathology at Massachusetts General Hospital (IRB approved). Tissues were either directly frozen in O.C.T. (D0) or cut into 2–4 mm3 pieces and placed on semi-submersed collagen sponge (Ultrafoam, Davol, Cranston, RI, USA) in prostate organ culture (POC) medium (RPMI-1640 medium containing 10% fetal calf serum, 2 mM non-essential amino acids and antibiotics) for 3 days and subsequently frozen. H&E-stained sections were assessed for tissue integrity. Scale bar=50 μm. (b) Assessment of prostate tissue viability. Freshly obtained (D0) or D3-derived organ cultures were cut into small cubes, incubated with MTS for 4 h at 37 °C, the absorbance was read at a wavelength of 490 nm, and protein concentrations were subsequently determined as previously described.27 (c) Anti-nectin-1 immunostaining in prostate cancer tissue. Prostate tissue specimens from pathology were immediately fixed in 10% formalin overnight and embedded in paraffin. Tissue retrieval was performed by incubation in 10% sodium citrate at 95 °C for 20 min. Serial tissue sections were either stained with normal rabbit immunoglobulin (left panels) (Dakocytomation, Carpinteria, CA, USA) or the anti-Nectin-1 R166 antibody21 for 1 h at room temperature, followed by anti-rabbit-HRP conjugated secondary antibody for 45 min (middle and right panels). Nectin-1 tissue distribution was subsequently revealed using the DAB-Chromogen system (Dakocytomation). Scale bar=50 μm. The strong staining of nectin-1 on epithelial cells and the faint staining throughout the stroma (black arrows) is noted. Photographs depict two separate images from different microscopic area of the same prostate cancer specimen.

oHSV preferentially targets epithelial cells of the prostatic ducts but not of stromal cells

We initially assessed the target specificity of G47Δ within prostate surgical specimens by X-gal staining. G47Δ contains a LacZ insertion in the ICP6 gene, which allows for viral tracking. Organ culture conditions were optimized to detect G47Δ infectivity while minimizing incubation times to preserve tissue integrity. G47Δ (1.5 × 106 p.f.u. (plaque forming units)) was placed in a medium with prostate tissue specimens for 1.5 h, and the specimens were subsequently transferred into a semi-submersed collagen-matrix for an additional 2–3 days resulting in the detection of β-gal+ cells (Figure 2a). It is noted that, X-gal staining appeared to be confined to epithelial cells within the prostatic glands with negligible staining detected in the surrounding stroma. Senescence-induced β-gal expression in prostate tissues has been detected by X-gal staining;23 however, β-gal expression was not detected in uninfected tissues (data not shown). Immunostaining of serial sections with anti-HSV gB, which recognizes the cell-surface glycoprotein B, a late gene product of HSV-1, confirmed the replication of G47Δ within these glandular regions but not in the stroma (Figure 2a). Staining with the luminal-specific epithelial anti-cytokeratin-8/18 antibody, further showed that both β-gal+ and gB+ regions likely represented luminal epithelial cells. LacZ+ and gB+ staining were frequently observed toward the center of the tissue section, suggesting viral penetration. The nuclear protein p63 is expressed in normal basal epithelial cells of the prostate. Furthermore, anti-p63 staining is frequently used as a diagnostic aid the expression of which is either decreased or absent in prostate adenocarcinomas but is present in BPT.24 Staining with anti-p63 on serial sections derived from these organ cultures revealed minimal detection in the prostate cancer specimens, whereas the nuclear expression of p63 staining was indeed observed in an organ culture derived from a BPT specimen used as a control (Figure 2b).

Figure 2
Figure 2

Analysis of G47Δ infection of prostate surgical samples. (a) Tissues were cut into 2–4 mm3 pieces, incubated with G47Δ (1.5 × 106 p.f.u.) for 1.5 h in 250 μl of POC medium and thereafter, placed on a semi-submersed collagen sponge in prostate organ culture (POC) medium. Tissue specimens were frozen in O.C.T 2 or 3 days post infection, sectioned (5 μm) and stained for β-galactosidase activity (upper panels) as previously described.28 Anti-HSV-1 gB (Virusys Corp., Sykesville, MD, USA) (middle panels) or anti-cytokeratin-8/18 (UCD/PR-10.11, Dakocytomation, Carpinteria, CA, USA) (lower panels) staining was also performed on serial sections. Tissue sections were acetone-fixed and rehydrated. After serum and peroxidase blocking (Dakocytomation), sections were incubated with primary antibody for 1 h, followed by a secondary anti-mouse immunoglobulin coupled to HRP (GE Healthcare, Piscataway, NJ, USA). Antibody staining was detected using the DAB-Chromogen system (Dakocytomation). Sections were counterstained with hematoxylin. It is noted that anti-HSV gB and X-gal staining overlap with the cytokeratin-8/18+ staining. Similar results were observed in all 12 prostate specimens examined. Pt.1 refers to patient specimen-1; Gleason Score 3+3=6 (Age 50, TNM tumor stage T2). Scale bar=50 μm. (b) Nuclear p63 expression in prostate organ cultures. Anti-p63 (4A4; Santa Cruz, Santa Cruz, CA, USA) staining was performed on serial sections from the above-infected D2 and D3-organ cultures (Pt.1) or on a D3 organ culture derived from a benign prostatic tissue (BPT). A representative prostatic gland is depicted. The dark brown nuclear stain of anti-p63 and the hematoxylin counterstain (violet) is noted. Arrows denote p63 positive staining. An isotype-matched control mouse antibody (Neg. Co; Dakocytomation) was used to show background staining. Scale bar=10 μm. PCa denotes prostate cancer.

Next, G47Δ infectivity was compared with G207 and the nonengineered parental HSV-1 (strain F). Anti-gB and anti-cytokeratin-8/18 staining indicated that G207 infectivity was confined to the epithelium of the prostatic glands similar to that of G47Δ (Figure 3a). This was in striking contrast to the pattern of infectivity observed for strain F as previously observed in mice.7 By day 3 post infection with wild-type strain F, anti-HSV gB-positive staining appeared not only within the epithelial-containing glandular regions but was also prevalent in the stroma as verified by anti-cytokeratin 8/18 immunostaining (Figure 3a). That is, HSV-1 gB expression was noted in regions that lacked anti-cytokeratin 8/18 staining. Thus, these data show that although G47Δ and G207 infection appears confined to the epithelia within the prostatic glands, wild-type HSV-1 is more promiscuous in nature, spreading throughout both the epithelia as well as the stroma. In addition, these data also suggest that although G47Δ has the capacity to infect both epithelial and stroma cells similar to strain F, its genetic alterations restrict its ability to spread in the epithelium. In BPT specimens, G47Δ infectivity resulted in limited distribution to glands at the periphery of tissue, whereas strain F infectivity was frequently observed in both glandular and non-glandular areas throughout the tissue (Figure 3c). In parallel, prostate cancer tissue lysates derived from G47Δ and G207-infected organ cultures were monitored by immunoblot analysis over a 3-day period for HSV-1 gB expression levels, an indicator of viral replication (Figure 3b). Although gB expression was undetected for either vector by D1 post infection, G47Δ-expressed gB was clearly visible by D3 and moreover, was substantially more prominent than for G207. Finally, we assessed oHSV replication by quantifying viral progeny in these prostate surgical specimens. Tissue samples (n=6) were infected as described above with G207, G47Δ or strain F, and infectious viral titers were subsequently determined 3 days post infection. Figure 3c summarizes these results. Although, as anticipated, both engineered oHSV replicated less well than did wild-type, G47Δ (1.2 × 106 p.f.u. mg−1±4.5 × 105) was consistently superior to G207 (4.3 × 104 p.f.u. mg−1±1.15 × 104) achieving a 28-fold increase in viral titer. We estimate that on average, the input for each infection was 6 × 104 p.f.u. mg−1 of protein from 25 mg of total tissue weight. The increase in strain F viral titers likely reflects, at least in part, its non-selectivity in its ability to spread and replicate in both epithelial and stromal cell compartments. Finally, oHSV infection of BPT specimens resulted in notably reduced viral titers for both G207, G47Δ, whereas viral titers for strain F were maintained at comparable levels relative to prostate cancer samples (Figure 3d).

Figure 3
Figure 3

A comparative analysis of oncolytic herpes simplex virus (oHSV) and wild-type HSV-1 infectivity and replication. (a) Tissue distribution of G47Δ, G207 and strain F 3 days post infection. Prostate cancer tissues were infected as detailed above and 3 days post infection, specimens were fixed, sectioned and stained with either the anti-HSV gB antibody (left panels) or anti-cytokeratin-8/18 antibody (right panels) on serial sections. It is noted that G207 is confined to cytokeratin-8/18+ epithelial cells, whereas strain F is present in both epithelial as well as stromal cells. Sections were counterstained with hematoxylin. Pt.5 was infected with either G47Δ or wild-type strain F. Scale bar=50 μm. Pt.2–6 refers to patient specimens 2 through 6 representing Gleason Scores of 3+2=5 (Pt.2: age 69 years, TNM tumor stage T2); 3+3=6 (Pt.5: age 58 years, TNM tumor stage T2) 3+4=7 (Pt.3: age 58 years, TNM stage T2; Pt.4: age 64 years, TNM tumor stage T3; Pt.6: age 66 years, TNM tumor stage T2). (b) Kinetic analysis of G47Δ and G207-derived gB protein expression. Prostate specimens were inoculated with either G47Δ or G207 (1 × 106 p.f.u.) for 1.5 h and placed on a collagen matrix for 1 or 3 days. Immunoblot analysis was performed on infected tissue lysates using either anti-gB (Virusys Corp., Sykesville, MD, USA) or anti-cytokeratin-18 antibody (DC-10; Santa Cruz, Santa Cruz, CA, USA). Anti-β-actin was used as a loading control. Approximately 50 μg of protein was loaded per lane (c) Tissue distribution of G47Δ and strain F in benign prostatic tissue (BPT) samples. Tissues were infected and at 3 days post infection, specimens were fixed, sectioned and stain with anti-HSV gB. It is noted that the majority of anti-gB staining in G47Δ infected tissues was identified in glandular regions near the rim of the tissue, whereas strain F infectivity resulted in a ubiquitous anti-gB staining to both glandular and non-glandular regions. (d) Quantification of oHSV viral titers in prostate organ cultures. Prostate cancer (left panel) and benign prostatic (right panel) tissue specimens infected with G207, G47Δ or strain F were maintained on a collagen support for 3 days. Virus titers were determined by titration of tissue homogenates on Vero cells using plaque assays.3 Tissues weights were determined to correct for viral titer concentrations. (Average±s.e.m; paired Student's t-test). Each color dot indicates a different prostate surgical specimen that was separately infected with one of the three indicated viruses (n=6). Pt.7–12 refers to patient specimens 7 through 12 representing Gleason Scores of 3+3=6 (Pt.8: age 69 years, TNM tumor stage T3; Pt.11: age 64 years, TNM tumor stage T2), 3+4=7 (Pt.9: age 66 years, TNM tumor stage T3; Pt.10: age 61 years, TMN tumor stage T2), 4+3=7 (Pt.7: age 54 years, TMN tumor stage T2; Pt.12: age 63 years, TMN tumor stage T2). Pt.13–17 refers to patients 13–17 representing BPT samples (n=5). PCa denotes prostate cancer.

Collectively, our studies show the suitability of using prostate surgical specimens to assess oHSV tropism. In accordance with previously published studies, prostate tissues can be maintained on a matrix support system.15, 16 The heterogeneous nature of prostate cancer specimens as well as the presence of Nectin-1, a major HSV-1 entry receptor, makes their use in organ cultures particularly attractive to evaluate oHSV tropism. Although both G47Δ and G207 were observed to preferentially target cytokeratin-8/18-positive epithelial cells, wild-type HSV-1 was also prevalent throughout the stroma. Moreover, the replication efficacy of G47Δ was consistently superior to that of G207 in these surgical samples. The molecular basis for these differences in replication is due to the immediate-early expression of Us11 after G47Δ infection that compensates for the loss of gamma34.5. Us11 protein has been shown to counteract the activities of eIF2α kinase PKR, PACT and 2′-5′-oligoadenylate synthetase, which are all critical for host defense.25 Finally, although we did not observe a correlation between tumor grade and viral infectivity more samples may need to be assessed. Thus, the unique genetic modifications built into these oHSVs confer cell-type specific replication competence in prostate surgical specimens. When these data are considered in conjunction with the equivalent lack of neurotoxicity of G47Δ and G207 after direct injection in the brains of HSV susceptible mice,3, 26 these data suggest that G47Δ should be considered for further development for prostate cancer therapy. The specificity of this model system should also allow assessment of the efficacy of oHSV virotherapy in combination with chemotherapeutics and appropriate small molecule pharmaceuticals.

Conflict of interest

The authors declare no conflict of interest.

References

  1. 1.

    , , , , , et al. Conditionally replicating herpes simplex virus mutant, G207 for the treatment of malignant glioma: Results of a phase I trial. Gene Therapy 2000; 7: 867–874.

  2. 2.

    , , , , . Attenuated multi-mutated herpes simplex virus-1 for the treatment of malignant gliomas. Nat Med 1995; 1: 939–943.

  3. 3.

    , , , . Oncolytic herpes simplex virus vector with enhanced MHC class I presentation and tumor cell killing. Proc Natl Acad Sci USA 2001; 98: 6396–6401.

  4. 4.

    , , , , , . Suppression of the phenotype of gamma(1)34.5- herpes simplex virus 1: failure of activated RNA-dependent protein kinase to shut off protein synthesis is associated with a deletion in the domain of the alpha47 gene. J Virol 1997; 71: 6049–6054.

  5. 5.

    , , . The second-site mutation in the herpes simplex virus recombinants lacking the gamma134.5 genes precludes shutoff of protein synthesis by blocking the phosphorylation of eIF-2alpha. J Virol 1998; 72: 7005–7011.

  6. 6.

    , , . Oncolytic herpes simplex virus vector therapy of breast cancer in C3(1)/SV40 T-antigen transgenic mice. Cancer Res 2005; 65: 1532–1540.

  7. 7.

    , , , , , et al. Preclinical safety evaluation of G207, a replication-competent herpes simplex virus type 1, inoculated intraprostatically in mice and nonhuman primates. Hum Gene Ther 2001; 12: 999–1010.

  8. 8.

    , , , , , et al. Attenuated multimutated herpes simplex virus-1 effectively treats prostate carcinomas with neural invasion while preserving nerve function. FASEB J 2008; 22: 1839–1848.

  9. 9.

    , , , , , et al. Local and systemic therapy of human prostate adenocarcinoma with the conditionally replicating herpes simplex virus vector G207. Hum Gene Ther 1999; 10: 2237–2243.

  10. 10.

    , , , , , et al. Ionizing radiation does not alter the antitumor activity of herpes simplex virus vector G207 in subcutaneous tumor models of human and murine prostate cancer. Neoplasia 2001; 5: 451–556.

  11. 11.

    , , , , . Effect of prior exposure to herpes simplex virus 1 on viral vector-mediated tumor therapy in immunocompetent mice. Gene Therapy 1999; 6: 1751–1758.

  12. 12.

    , , , , , et al. Effects of preexisting immunity on the response to herpes simplex-based oncolytic viral therapy. Hum Gene Ther 2000; 18: 2465–2472.

  13. 13.

    , . Role of the immune response during neuron-attenuated herpes simplex virus-mediated tumor destruction in a murine intracranial melanoma model. Cancer Res 2000; 60: 5714–5722.

  14. 14.

    , , , . Reprogrammed viruses as cancer therapeutics: targeted, armed and shielded. Nat Rev Microbiol 2008; 6: 529–540.

  15. 15.

    , , , , , . Hormone regulation of human prostate in organ culture. Cancer Res 1993; 53: 5199–5207.

  16. 16.

    , , , , , et al. Boosting antitumor reponses of T lymphocytes infiltrating human prostate cancers. J Exp Med 2005; 201: 1257–1268.

  17. 17.

    , , , , , . Prolactin and prolactin receptors are expressed and functioning in human prostate. J Clin Invest 1997; 99: 618–627.

  18. 18.

    , , , , , et al. Roles of hedgehog signaling in androgen production and prostate ductal morphogenesis. Dev Biol 2004; 267: 387–398.

  19. 19.

    , , , , , et al. Nectin-1 expression by squamous cell carcinoma is a predictor of herpes oncolytic sensitivity. Mol Therapy 2007; 15: 103–113.

  20. 20.

    , , , . The novel receptors that mediate the entry of herpes simplex viruses and animal alphaerpesviruses into cells. Rev Med Virol 2000; 10: 305–319.

  21. 21.

    , , , , , et al. Herpes simplex virus glycoprotein D can bind to poliovirus receptor-related protein 1 or herpes virus entry mediator, two structurally unrelated mediators of virus entry. J Virol 1998; 72: 7064–7074.

  22. 22.

    , , , . Nectin-1 expression in the normal and neoplastic human uterine cervix. Arch Path Lab Med 2006; 130: 1193–1195.

  23. 23.

    , , , , , et al. Expression of senescence-associated B-galactosidase in enlarged prostates from me with benign prostatic hyperplasia. Urology 2000; 56: 160–166.

  24. 24.

    , , , . P63 protein expression is rare in prostate adenocarcinoma: implications for cancer diagnosis and carcinogenesis. Urology 2001; 58: 619–624.

  25. 25.

    , . Inhibition of cellular 2′-5′ oligoadenylate synthetase by the herpes simplex virus type 1 Us11 protein. J Virol 2007; 81: 3455–3464.

  26. 26.

    , , , . Attenuated, replication-competent herpes simplex virus type 1 mutant G207: safety evaluation in mice. J Virol 2000; 74: 3832–3841.

  27. 27.

    , , , , , et al. Herpes simplex virus type 1 preferentially targets human colon carcinoma: role of extracellular matrix. J Viol 2008; 82: 999–1010.

  28. 28.

    , , , , , . Enhanced therapeutic efficacy of IL-12, but not GM-CSF expressing oncolytic herpes simplex virus for transgenic mouse derived prostate cancers. Cancer Gene Ther 2005; 13: 253–265.

Download references

Acknowledgements

We would like to thank Drs T. Kasic and A. Viola (University of Padova; Padova, Italy) for their helpful advice on setting up conditions for prostate organ cultures. We are grateful to Dr CA Krummenacher (University of Pennsylvania School of Dental Medicine; Philadelphia, PA, USA) for supplying the R166 antibody. We thank Dr G Fulci (Massachusetts General Hospital) for expertise on immunohistochemistry and Ms. Melissa Marinelli for laboratory assistance. RLM an SDR are consultants to MediGene Ag, which has a license from Georgetown University for G207. Support for this study was in part obtained from a grant to RLM (RO1 CA102139).

Author information

Affiliations

  1. Department of Neurosurgery, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA

    • B J Passer
    • , S D Rabkin
    •  & R L Martuza
  2. Department of Pathology, Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA

    • C-l Wu
    •  & S Wu

Authors

  1. Search for B J Passer in:

  2. Search for C-l Wu in:

  3. Search for S Wu in:

  4. Search for S D Rabkin in:

  5. Search for R L Martuza in:

Corresponding author

Correspondence to B J Passer.

About this article

Publication history

Received

Revised

Accepted

Published

DOI

https://doi.org/10.1038/gt.2009.94

Further reading