Thomas et al. Journal for ImmunoTherapy of Cancer (2019) 7:214
`https://doi.org/10.1186/s40425-019-0682-1
`
`R E S E A R C H A R T I C L E
`Development of a new fusion-enhanced
`oncolytic immunotherapy platform based
`on herpes simplex virus type 1
`Suzanne Thomas1, Linta Kuncheria1, Victoria Roulstone2, Joan N. Kyula2, David Mansfield2,
`Praveen K. Bommareddy1, Henry Smith2, Howard L. Kaufman2,3, Kevin J. Harrington2 and Robert S. Coffin1*
`
`Open Access
`
`Abstract
`
`Background: Oncolytic viruses preferentially replicate in tumors as compared to normal tissue and promote
`immunogenic cell death and induction of host systemic anti-tumor immunity. HSV-1 was chosen for further
`development as an oncolytic immunotherapy in this study as it is highly lytic, infects human tumor cells broadly,
`kills mainly by necrosis and is a potent activator of both innate and adaptive immunity. HSV-1 also has a large
`capacity for the insertion of additional, potentially therapeutic, exogenous genes. Finally, HSV-1 has a proven safety
`and efficacy profile in patients with cancer, talimogene laherparepvec (T-VEC), an oncolytic HSV-1 which expresses
`GM-CSF, being the only oncolytic immunotherapy approach that has received FDA approval. As the clinical efficacy
`of oncolytic immunotherapy has been shown to be further enhanced by combination with immune checkpoint
`inhibitors, developing improved oncolytic platforms which can synergize with other existing immunotherapies is a
`high priority. In this study we sought to further optimize HSV-1 based oncolytic immunotherapy through multiple
`approaches to maximize: (i) the extent of tumor cell killing, augmenting the release of tumor antigens and danger
`-associated molecular pattern (DAMP) factors; (ii) the immunogenicity of tumor cell death; and (iii) the resulting
`systemic anti-tumor immune response.
`
`(Continued on next page)
`
`* Correspondence: heather.metcalfe@replimune.com
`1Replimune Inc, 18 Commerce Way, Woburn, MA 01801, USA
`Full list of author information is available at the end of the article
`
`© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
`International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
`reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
`the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
`(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
`
`Replimune Limited Ex. 2013 - Page 1
`Transgene and Bioinvent International AB v. Replimune Limited
`PGR2022-00014 - U.S. Patent No. 10,947,513
`
`

`

`Thomas et al. Journal for ImmunoTherapy of Cancer (2019) 7:214
`
`Page 2 of 17
`
`(Continued from previous page)
`Methods: To sample the wide diversity amongst clinical strains of HSV-1, twenty nine new clinical strains isolated
`from cold sores from otherwise healthy volunteers were screened across a panel of human tumor cell lines to
`identify the strain with the most potent tumor cell killing ability, which was then used for further development.
`Following deletion of the genes encoding ICP34.5 and ICP47 to provide tumor selectivity, the extent of cell killing
`and the immunogenicity of cell death was enhanced through insertion of a gene encoding a truncated, constitutively
`highly fusogenic form of the envelope glycoprotein of gibbon ape leukemia virus (GALV-GP-R−). A number of further
`armed derivatives of this virus were then constructed intended to further enhance the anti-tumor immune response
`which was generated following fusion-enhanced, oncolytic virus replication-mediated cell death. These viruses
`expressed GMCSF, an anti-CTLA-4 antibody-like molecule, CD40L, OX40L and/or 4-1BB, each of which is expected to
`act predominantly at the site and time of immune response initiation. Expression of these proteins was confirmed by
`ELISA and/or western blotting. Immunogenic cell death was assessed by measuring the levels of HMGB1 and ATP from
`cell free supernatants from treated cells, and by measuring the surface expression of calreticulin. GALV-GP-R− mediated
`cell to cell fusion and killing was tested in a range of tumor cell lines in vitro. Finally, the in vivo therapeutic
`potential of these viruses was tested using human A549 (lung cancer) and MDA-MB-231(breast cancer) tumor
`nude mouse xenograft models and systemic anti-tumor effects tested using dual flank syngeneic 4434
`(melanoma), A20 (lymphoma) mouse tumor models alone and in combination with a murine anti-PD1
`antibody, and 9 L (gliosarcoma) tumors in rats.
`Results: The twenty nine clinical strains of HSV-1 isolated and tested demonstrated a broad range of tumor
`cell killing abilities allowing the most potent strain to be identified which was then used for further development.
`Oncolytic ability was demonstrated to be further augmented by the expression of GALV-GP-R− in a range of tumor cell
`lines in vitro and in mouse xenograft models in nude mice. The expression of GALV-GP-R− was also demonstrated to
`lead to enhanced immunogenic cell death in vitro as confirmed by the increased release of HMGB1 and ATP and
`increased levels of calreticulin on the cell surface. Experiments using the rat 9 L syngeneic tumor model demonstrated
`that GALV-GP-R− expression increased abscopal uninjected (anenestic) tumor responses and data using mouse 4434
`tumors demonstrated that virus treatment increased CD8+ T cell levels both in the injected and uninjected tumor, and
`also led to increased expression of PD-L1. A combination study using varying doses of a virus expressing GALV-GP-R− and
`mGM-CSF and an anti-murine PD1 antibody showed enhanced anti-tumor effects with the combination which was most
`evident at low virus doses, and also lead to immunological memory. Finally, treatment of mice with derivatives of this
`virus which additionally expressed anti-mCTLA-4, mCD40L, m4-1BBL, or mOX40L demonstrated enhanced activity,
`particularly in uninjected tumors.
`Conclusion: The new HSV-1 based platform described provides a potent and versatile approach to developing new
`oncolytic immunotherapies for clinical use. Each of the modifications employed was demonstrated to aid in optimizing
`the potential of the virus to both directly kill tumors and to lead to systemic therapeutic benefit. For clinical use, these
`viruses are expected to be most effective in combination with other anti-cancer agents, in particular PD1/L1-targeted
`immune checkpoint blockade. The first virus from this program (expressing GALV-GP-R− and hGM-CSF) has entered
`clinical development alone and in combination with anti-PD1 therapy in a number of tumor types (NCT03767348).
`Keywords: Oncolytic viruses, Herpes simplex-1, Immunotherapy, Anenestic effect
`
`Replimune Limited Ex. 2013 - Page 2
`Transgene and Bioinvent International AB v. Replimune Limited
`PGR2022-00014 - U.S. Patent No. 10,947,513
`
`

`

`Thomas et al. Journal for ImmunoTherapy of Cancer (2019) 7:214
`
`Page 3 of 17
`
`Introduction
`shown single agent
`Oncolytic immunotherapy has
`clinical activity and synergy with immune checkpoint
`blockade. However, not all patients respond, and most of
`the clinical experience has been in melanoma. With the
`objective of maximally activating a patient’s immune sys-
`tem against their own cancer to enhance synergy with
`anti-PD1/L1 blockade, we have developed a new oncoly-
`tic immunotherapy platform based on herpes simplex
`virus type 1 (HSV-1). This has the dual objectives of ro-
`bustly killing tumor to provide abundant release of
`tumor antigens, and potently activating the immune sys-
`tem against these tumor antigens once released. To aug-
`ment the natural ability of HSV-1 to kill tumors and
`activate anti-tumor immunity, the viruses developed are
`armed with therapeutic genes with the expectation that
`‘arming’ will be essential to maximizing clinical activity.
`Initially, we sampled the genetic variation between
`strains of HSV-1 by screening twenty nine new clinical
`strains isolated from volunteers who suffer from cold
`sores across a panel of human tumor cell lines to iden-
`tify the strain to be developed. This strain (RH018A)
`was then engineered for oncolytic use by deletion of the
`genes encoding ICP34.5 to reduce pathogenicity, delet-
`ing the ICP47 encoding gene to enhance viral and tumor
`antigen presentation by major histocompatibility com-
`plex-I (MHC-I), and inserting a gene encoding a potent
`fusogenic
`glycoprotein derived
`from gibbon ape
`leukemia virus (GALV-GP-R−). Expression of GALV-GP-
`R− caused increased immunogenic cell death, assessed
`by the release of danger-associated molecular pattern fac-
`tors, activated anti-tumor immunity, and enhanced sys-
`temic therapeutic activity against rat and murine tumors
`in vivo. Additionally, the virus induced expression of PD-
`L1, and demonstrated enhanced activity in combination
`with PD-1 blockade. A virus expressing GALV-GP-R− and
`hGM-CSF is currently in a Phase 1/2 clinical
`trial
`(NCT03767348). Further viruses were constructed based
`on this virus which additionally express an anti-CTLA-4
`antibody or immune co-stimulatory pathway activating li-
`gands, each of which is expected to act at the site and time
`of immune response initiation in the injected tumor and
`draining lymph nodes. These viruses demonstrated further
`increased activity in mice, particularly an enhanced ane-
`nestic effect. This data supports the potential for im-
`proved therapeutic
`activity of
`this new oncolytic
`immunotherapy platform and demonstrate its use to ex-
`press immune modulatory proteins which may provide a
`generalized strategy to improve therapy for patients with
`cancer. There have been significant advances in the im-
`munotherapy of cancer, most notably through the clinical
`development of immune checkpoint inhibitors targeting
`cytotoxic T lymphocyte antigen 4 (CTLA-4) and the pro-
`grammed cell death 1 (PD-1)/PD-1 ligand (PD-L1)
`
`pathway [1, 2]. While durable clinical responses have been
`observed across numerous solid and hematologic malig-
`nancies, many tumors do not respond or develop resist-
`ance over time [3]. The absence of tumor-specific T cells
`within the tumor microenvironment appears to be an im-
`portant feature associated with innate and acquired resist-
`ance to checkpoint blockade. New strategies that can
`induce anti-tumor immune responses with which anti-
`PD-1/L1 therapy can synergize, reverse the immune-defi-
`cient tumor microenvironment, and which can re-estab-
`lish tumor sensitivity to systemic anti-PD-1/L1 therapy
`are therefore needed. One promising approach is virus-
`based oncolytic immunotherapy [4]. Oncolytic viruses
`preferentially replicate in tumors as compared to normal
`tissue, and promote immunogenic cell death and induc-
`tion of host systemic anti-tumor immunity. The oncolytic
`immunotherapy approach has been clinically validated as
`demonstrated by the U.S. Food and Drug Administration
`(FDA) and European Medicines Agency (EMA) approval
`of talimogene laherparepvec (T-VEC), an oncolytic herpes
`simplex virus type 1 (HSV-1) encoding GM-CSF, for the
`treatment of advanced melanoma in 2015 [5]. The phase 3
`clinical trial which led to the approval of T-VEC demon-
`strated a 26.4% objective response rate, and a 10.8%
`complete response rate (rising to 17% at the time of the
`final analysis [Amgen ODAC presentation May 2015] [6]),
`in a 436-patient phase 3 study in patients with both previ-
`ously treated and previously untreated Stage IIIb-IVM1c
`disease [5].
`The therapeutic potential of T-VEC can be further
`enhanced by combination with immune checkpoint in-
`hibitors. In a small phase 1 trial in patients with melan-
`oma, T-VEC in combination with pembrolizumab
`resulted in a 62% response rate and 33% complete
`response rate [7]. Similarly promising response rates (> 50%)
`have also been seen in other small studies with either
`ipilimumab or pembrolizumab in combination with other
`oncolytic viruses, such as Cavatak (an oncolytic Coxsackie-
`virus) or HF10 (another oncolytic HSV-1) [4]. Data have also
`been reported from a 200-patient randomized controlled
`phase 2 clinical trial with T-VEC combined with ipilimumab
`compared to ipilimumab alone, where more than a doubling
`of the response rate was seen in the combination arm [8].
`While these studies were all in melanoma, it is important to
`note that none reported significant additional
`toxicity
`compared to that expected with either agent alone. Based
`on the favorable therapeutic window for T-VEC and other
`oncolytic viruses, there has been considerable interest in
`optimizing the oncolytic immunotherapy strategy and using
`such agents as part of a rational combination regimen in
`patients with solid cancers.
`It is now generally accepted that patients responding
`to immunotherapy need to have tumors that are im-
`munologically ‘hot’, i.e. have a T cell-inflamed phenotype,
`
`Replimune Limited Ex. 2013 - Page 3
`Transgene and Bioinvent International AB v. Replimune Limited
`PGR2022-00014 - U.S. Patent No. 10,947,513
`
`

`

`Thomas et al. Journal for ImmunoTherapy of Cancer (2019) 7:214
`
`Page 4 of 17
`
`although the specific mechanisms that regulate T cell re-
`cruitment
`into established tumors are incompletely
`understood [9]. Additional factors that favor immune-
`mediated rejection include high mutation burden,
`presence of pre-existing immune responses to tumor an-
`tigens, particularly tumor neoantigens, and expression of
`a pro-inflammatory gene signature [10]. While a number
`of approaches are in development aimed at correcting
`these deficiencies in non-responsive patients, oncolytic
`immunotherapies may have particular promise for this
`purpose as they kill tumors in a highly inflammatory
`context. This effect is highly immunogenic,
`including
`activation of both innate and adaptive immunity, with
`the potential to create a vaccine “in situ” within the pa-
`tient against their own cancer. The local production of
`type 1 interferons induced by oncolytic viruses also
`results in increased expression of several immune regu-
`latory proteins, including MHC class I and PD-L1 [4].
`Thus, oncolytic immunotherapy appears to be particu-
`larly well suited for combination strategies with immune
`checkpoint blockade. We sought to further optimize the
`approach by maximizing (i) the extent of tumor cell
`killing, augmenting the release of tumor antigens and
`danger-associated molecular pattern (DAMP) factors; (ii)
`the immunogenicity of tumor cell death; and (iii) the
`resulting systemic anti-tumor immune response. While a
`range of viral species were considered for development,
`HSV-1 was selected for several reasons. First, HSV-1 is a
`very lytic DNA virus;
`it
`infects human tumor cells
`broadly, and when ICP34.5 is deleted exhibits preferen-
`tial replication in neoplastic tissue. Second, HSV-1 kills
`mainly by necrosis and activates
`innate immunity,
`including through the cGAS/STING pathway. Third,
`HSV-1 has a large capacity for
`the insertion of
`additional, potentially therapeutic, exogenous genes.
`Finally, HSV-1 has a proven safety and efficacy profile in
`patients with cancer. While intravenous administration
`was also considered, an intratumoral approach, i.e. local
`administration providing systemic immune-based bene-
`fit, was selected based on prior clinical validation and
`the
`considerable,
`and
`potentially
`insurmountable,
`biological hurdles to effective intravenous dosing [4, 11].
`HSV-1 causes cold sores in humans and is widely preva-
`lent in the population, with up to 90% of individuals
`testing seropositive by the age of 65 [12]. However,
`substantial natural variation might be expected amongst
`clinical strains of HSV-1 (i.e. as sampled from individ-
`uals suffering from cold sores) with respect to evolved
`biological properties such as virulence. This natural
`variation might also translate into differences in non-
`evolved properties, such as the ability to infect and kill
`human tumor cells. Based on the hypothesis that proto-
`typical ‘laboratory’ strains of HSV-1, such as Strain 17+,
`KOS or Strain F may have become attenuated through
`
`extended serial passage or may otherwise not be optimal
`strains for cancer therapy, T-VEC was initially derived
`from a clinical strain of HSV-1 after comparing two
`clinical isolates to Strain 17+. Both of the clinical strains
`were superior for human tumor cell killing compared to
`Strain 17+, and the most promising of the two, Strain
`JS1, was chosen and engineered into T-VEC [13].
`In this
`report, we describe the generation and
`characterization of a new HSV-1-based oncolytic im-
`munotherapy platform which utilizes a strain of HSV-1
`selected from twenty nine newly isolated clinical strains
`on the basis of increased oncolytic activity in vitro. This
`was then engineered for tumor selectivity and to express
`a potent fusogenic membrane glycoprotein (GALV-GP-
`R−) to increase the extent and immunogenicity of tumor
`cell death. Various fusogenic proteins including from
`measles virus and various retroviruses have previously
`been tested in replicative and non-replicative virus-me-
`diated gene therapy approaches to the treatment of can-
`cer in pre-clinical models [14], including when delivered
`by oncolytic versions of HSV [15]. Fusogenic cell death
`has also previously been demonstrated to be highly im-
`munogenic [14]. Genes encoding GM-CSF, an anti-
`CTLA-4 antibody-like molecule and a number of im-
`mune co-stimulatory pathway-activating ligands were
`then inserted, intending to further enhance the systemic,
`immune-mediated, anti-tumor effects achieved.
`
`Methods
`Assessment of GALV-GP-R− mediated fusion
`The cell
`lines used for the fusion assays were A549
`(ECACC 91072201), HT29 (ECACC 91072201), HT1080
`(ECACC 85111505), MDA-MB-231 (ECACC 92020424),
`miaPaCa-2 (ECACC 85062806) and SK-mel-28 (ATCC®
`HTB-72™). The monolayers were infected using a range
`of multiplicity of infection (MOI) from 0.01 to 0.0001.
`The infected cell monolayers were observed for GFP
`expression at 24 h. and 48 h. post-infection and then
`fixed and stained with crystal violet.
`
`Western blots and ELISA
`For detection of anti-CTLA-4 expressed from Virus
`27, supernatant was used from BHK cells infected at
`MOI =1 in serum-free mediuma for 24 h. The proteins
`were separated on 10–20% sodium dodecyl polyacryl-
`amide gel (Thermo Fisher CAT No: XP10200BOX) and
`transferred to polyvinylidene difluoride membrane (Life
`Technologies Cat No: LC2005). The membrane was
`probed with goat anti-mouse IgG1 heavy chain (alkaline
`phosphatase)
`(Abcam Cat No: ab97237). BCIP®/NBT
`Liquid Substrate System (Sigma Aldrich Cat No: B1911)
`was used for the detection.
`For detection of CD40L, 4-1BBL and OX40L from
`Viruses 32, 33 and 35, respectively, BHK cells were
`
`Replimune Limited Ex. 2013 - Page 4
`Transgene and Bioinvent International AB v. Replimune Limited
`PGR2022-00014 - U.S. Patent No. 10,947,513
`
`

`

`Thomas et al. Journal for ImmunoTherapy of Cancer (2019) 7:214
`
`Page 5 of 17
`
`infected at MOI =1 for 24 h. To confirm expression of
`4-1BBL from Virus 33, microplates were coated with the
`capture antibody (0.5μg/ml, R&D Systems Cat No:-
`AF1246) and incubated overnight at 4 °C. Following
`blocking, standards (R&D Systems Cat No 1256-4 L, 40
`ng/ml- 0.63 ng/ml)
`and samples were
`added and
`incubated at 37 °C. The wells were then probed with
`anti-mouse 41BBL (Bioxcell Cat No: BE0110) after
`which HRP Tagged antibody (Sigma Aldrich Cat No:
`A5795) was added and incubated for 1 h. TMB was
`added and incubated for 5 mins and sulphuric acid was
`added to stop the reaction. The plates were read at 450
`nm. ELISA for CD40L (Abcam Cat No: ab119517) and
`OX40L (Thermo Fisher Cat No: EMTNFSF4) was
`the manufacturer’s
`performed using
`kits
`as per
`instructions.
`
`ATP release
`Cells were plated at 2 × 105 cells per well in 1 mL, in 12-
`well plates, and incubated overnight. Cells were then
`infected with Virus 23 or Virus 17 the following day.
`Twenty-four and 48 h after treatment, cell supernatants
`were collected and centrifuged at 2000 rpm for 4 mins.
`Cell-free supernatants were then measured for ATP by
`CellTiter-Glo Luminescent Cell Viability Assay (CTG,
`Promega, UK). Fifty microliter of CTG was added per
`200 uL sample and incubated for 10 min. Luminescence
`was measured on a Victor 2 V plate reader (Perkin
`Elmer).
`
`High mobility group box 1 protein (HMGB1) release
`Cells were plated at 2 × 105 cells per well in 1 mL, in 12-
`well plates, and incubated overnight. Cells were infected
`with Virus 23 or Virus 17 the following day. Forty-eight
`h after treatment, cell supernatants were collected and
`centrifuged at 2000 rpm for 4 mins. Cell-free superna-
`tants were then measured for HMGB1 by an ELISA
`Assay (IBL International GmbH Cat No: ST51011) as
`per the manufacturer’s instructions.
`
`Cell surface calreticulin expression
`Cells were plated at 2 × 105 cells per well in 1 mL, in 12-
`well plates, and incubated overnight. Cells were infected
`with Virus 23 or Virus 17 the following day at various
`MOI. Forty-eight h after treatment, un-permeabilized
`samples were stained with viability dye (Thermo Fisher
`Cat No: 65–0865-14), with anti-calreticulin antibody
`(Abcam Cat No: ab92516), or isotype control antibody
`(Abcam Cat No: ab172730), and flow cytometry was
`performed. Surface calreticulin expression was shown as
`median fluorescence intensity (MFI). Data was analyzed
`using FlowJo software.
`
`In vivo efficacy testing
`Bilateral mouse A20 lymphoma tumors were grown in
`of Balb/c mice or human A549 or MDA-MB-231 tumors
`grown in the right flanks of Balb/c nude mice until aver-
`age tumor diameters were > 5 mm. Right flank tumors
`were then injected 3 times (every other day) with the
`indicated virus and dose in 50 μl or with vehicle (PBS)
`and tumor diameters were then followed. For experi-
`ments in rats, rat 9 L glioma tumors were grown in the
`left and right flanks of Fischer 344 rats until tumors were
`0.75-1 cm in diameter and right flank tumors then dosed
`5x (approximately every other day) with the indicated
`virus at a dose of 5 × 106 pfu in 50 μl or with vehicle and
`tumor diameters then followed. For experiments in
`combination with anti-murine PD1, clone RMP1–14
`(BioXCell) was given by the intraperitoneal route at 10
`mg/kg every 3 days for a total of 9 doses.
`
`Vectra staining
`Vectra staining was performed on tumors to identify
`tumor infiltrating immune cells as previoulsy described
`[16]. Bi-flank 4434 murine melanoma tumours grown in
`C57BL/6 mice were treated with Virus 16 on days 1, 3
`and 5, then collected at day 10 after the first injection,
`fixed overnight in 10% neutral buffered formalin and
`then transferred to PBS prior
`to processing and
`embedding. Tissue sections were labeled with immuno-
`fluorescent stains as follows; CD8 (Cat No: 14–0808-82),
`CD4 (Cat No:14–9766-82), and foxp3 (Cat No: 14–
`5773-82), all from eBioscience. Images were then quanti-
`fied by an automated cell segmentation and phenotyping
`algorithm, using inForm analysis
`software
`(Perkin
`Elmer). Four thousand four hundred thirty-four cells are
`a murine melanoma tumor cell line generated in house
`at The Institute of Cancer Research, London.
`
`FACS analysis of tumours
`C57BL/6 mice were subcutaneously implanted with
`4 × 106 4434 murine melanoma cells suspended in 0.1
`mL PBS per flank in a bi-flank model. Tumours were
`allowed to grow to 6–8 mm and randomized into
`study groups. The right flank was injected with 5 × 106
`plaque forming units (pfu) of Virus 16 in 50 μl or a mock
`group received formulation buffer (vehicle), given on days
`1, 3 and 5. Mice were euthanized when a tumour reached
`15 mm in any direction. Tumours were harvested and
`minced with scissors in digestion mix (0.01% trypsin,
`2.5 mg/mL collagenase, 2 mg/mL dispase and 1 mg/mL
`DNAse in RPMI), and incubated at 37 °C for 30 min.
`Thereafter, samples were kept on ice. Suspensions were
`passed through a 70 μm strainer using a 2.5 mL syringe
`plunger and washed through with RPMI + 5 mM EDTA
`until only connective tissue remained. Samples were
`centrifuged at 1500 rpm,
`for 5 mins at 4 °C) and
`
`Replimune Limited Ex. 2013 - Page 5
`Transgene and Bioinvent International AB v. Replimune Limited
`PGR2022-00014 - U.S. Patent No. 10,947,513
`
`

`

`Thomas et al. Journal for ImmunoTherapy of Cancer (2019) 7:214
`
`Page 6 of 17
`
`transferred into a V-well 96 plate. Samples were stained in
`FACS buffer (PSB + 5% FCS) with the following extracellu-
`lar antibodies for 30 mins, on ice and protected from light;
`CD3 (Cat No: 100236), CD4 (Cat No:100406), CD8 (Cat
`No: 100732) all from BioLegend, PD-L1 (BD Biociences
`Cat No: 558091), and viability dye (Thermo Fisher Cat
`No: 65–0865-14). Cells were then washed in FACS buffer
`and permeabilized and stained with intracellular antibody
`to foxp3 (Thermo Fisher Cat No: 48–5773-80). Samples
`were then washed and fixed (1–2% PFA) prior to analysis
`of tumour infiltrating lymphocytes by flow cytometry.
`Tumours were weighed on collection and counting beads
`were added when running the analysis in order to calcu-
`late cells per mg of tumour.
`
`Viral replication
`Bi-flank 4434 tumours were collected by dissection,
`homogenized with 600 μl of serum-free DMEM and cen-
`trifuged at 3600 rpm. For 5 mins. Tumor draining lymph
`nodes corresponding to the injected and contralateral
`tumors
`and
`spleens were
`collected
`separately.
`Supernatants were titred on BHK cells plated at 1 × 104
`per well in 96 well plates. Cytopathic effect (CPE) was
`scored 48–72 h later and viral titer was determined by
`TCID50 assay.
`
`Viral propagation
`All viruses used in the study were propagated using a
`standard laboratory HSV-1 propagation protocol as de-
`scribed previously [17]. In brief, monolayers of vero cells
`were infected and virus allowed to seed for 2–3 h after
`which the monolayer was washed with growth media
`which was replaced and the cells then left in culture
`until 100% CPE was observed. Virus was harvested from
`the supernatant and a standard HSV-1 plaque assay per-
`formed to quantify the virus [18].
`
`Statistical analysis
`All statistical analyses were performed using GraphPad
`Prism software version 7.0a. Tumor growth curves, flow
`cytometric data and immunohistochemistry counts were
`compared using an unpaired student’s t test (two-tailed),
`one-way ANOVA or two-way ANOVA when multiple
`comparisons were performed. P values of less than 0.05
`were considered significant. The Figures use the follow-
`ing indications of the level of significance: * = p < 0.05,
`** = p < 0.01, *** = p < 0.001, **** = p < 0.0001.
`
`Results
`Selection of the virus strain for development
`We sought to extend the hypothesis that strains of HSV-
`1 with greater oncolytic potential could be derived from
`a larger sampling of HSV-1 cold sore isolates. To accom-
`plish this, we recruited 126 volunteers who suffered
`
`from herpes cold sores between May 2015 and Aug
`2015 and, after obtaining informed consent, collected
`viral swab samples from these volunteers during a recur-
`rent episode of cold sores. Samples were cultured from
`twenty nine volunteers. These were confirmed to be
`HSV-1 by anti-HSV-1 antibody staining of infected BHK
`cell monolayers and then compared against each other
`across a panel of human tumor cell lines representative
`of different tumor histologies for their ability to infect
`and kill rapidly and at low virus dose. As expected, con-
`siderable variation in these abilities was seen, with
`roughly one-third of the isolates being relatively poor,
`roughly one-third being ‘average’, and nine clearly being
`more effective than the rest. These nine isolates were
`then compared more thoroughly across the cell
`line
`panel, allowing the generation of a rank order of the top
`five isolates. Representative data at only an individual
`time point and MOI
`in each case are shown in
`(Additional file 1: Figure S1A). Strain RH018 was chosen
`as the strain for further development on the basis that it
`scored either first or second most effective at cell killing
`on each of the cell lines tested. Compared to a represen-
`tative ‘average’ strain from the screen, i.e. a strain from
`the middle-third group (isolate RH065), RH018 yielded
`approximately a 10-fold increase in cytotoxic potency, as
`defined by isotoxic efficacy at a 10-fold lower multipli-
`city of infection (Additional file 1: Figure S1B). Isolate
`RH018 was sequenced, confirming the presence of the
`expected HSV-1 encoded genes, but with a variety of
`small changes across the genome as compared to the
`originally sequenced prototype HSV-1 genome sequence,
`Strain 17+ (Genbank NC_001806.2). No attempt was
`made to determine which of
`the observed changes,
`individually or in combination, may be responsible for
`the improved (as compared to the ‘average’ clinical strain
`of HSV)
`tumor cell cell killing properties observed.
`Based on this screen, the RH018A strain of HSV-1 was,
`therefore, selected as a foundation for further development.
`
`Engineering for use as an oncolytic virus
`To render strain RH018 non-pathogenic and replication-
`selective for tumors, the HSV-1 genes encoding infected
`cell protein (ICP) 34.5 and ICP47 were deleted. ICP34.5,
`the so-called neurovirulence factor, has functions which
`include overcoming host anti-viral (i.e. interferon-medi-
`ated) responses which would otherwise block virus repli-
`cation in normal tissue, and the expression of which is
`essential for pathogenicity [19, 20]. Deletion of ICP34.5
`inhibits replication in normal tissue but ICP34.5 is dis-
`pensable for replication in tumors [14] by virtue of their
`generally having impaired interferon-mediated responses
`through various mechanisms [21]. ICP47 is an inhibitor
`of antigen presentation in HSV-1 infected cells [22], the
`deletion of which also increases the expression of the
`
`Replimune Limited Ex. 2013 - Page 6
`Transgene and Bioinvent International AB v. Replimune Limited
`PGR2022-00014 - U.S. Patent No. 10,947,513
`
`

`

`Thomas et al. Journal for ImmunoTherapy of Cancer (2019) 7:214
`
`Page 7 of 17
`
`HSV US11 gene by placing the coding sequence for
`US11 adjacent
`to the immediate/early promoter for
`ICP47 [13]. US11 has
`functional
`redundancy with
`ICP34.5 and immediate/early
`expression of US11
`restores to HSV-1 some level of resistance to interferon
`[23]. This increases replication in tumors, without redu-
`cing the tumor selectivity achieved through the deletion
`of
`ICP34.5 [13].
`ICP34.5-
`and ICP34.5/47-deleted
`versions of HSV-1 have been extensively tested in clin-
`ical trials [24] and T-VEC (which has both the ICP34.5
`and ICP47 deletions) is U.S. FDA approved for the treat-
`ment of advanced melanoma. In all cases, these viruses
`have been shown to be well tolerated, including through
`direct intracerebral
`injections in patients with glioma
`[25]. This proven safety and efficacy profile provided the
`basis for using the same disabling approach here. All
`viruses were generated by recombination of viral and
`plasmid DNA using standard methods,
`followed by
`clone selection based on the presence or absence of GFP
`[26]. The genome structures of the viruses constructed
`and tested in this paper are shown in (Fig. 1). The details
`of the construction of each virus is described in the
`Additional file 1.
`
`Augmenting the natural ability of HSV-1 to kill tumor
`cells
`In order to augment the natural ability of HSV-1 to kill
`tumor cells, a codon-optimized version of a potent fuso-
`genic membrane glycoprotein (GP) from gibbon ape
`leukemia virus (GALV) was additionally encoded in the
`virus backbone. Here, the R sequence was deleted (R−),
`which provides constitutive fusion properties to the
`GALV-GP [14]. The initial viruses constructed to test
`this approach expressed either GFP or GFP together
`with GALV-GP R− (Virus 10 and Virus 12) (Fig. 1),
`which were first tested on a range of tumor cell lines in
`vitro. This demonstrated that potent cell-to-cell fusion
`was achieved through the expression of GALV-GP-R−,
`and that the plaques generated by these viruses were
`greatly enlarged, as visualized by expression of GFP
`(Fig. 2a). Cell killing potency was also greatly increased,
`with substantially greater killing being achieved at
`equivalent virus doses through the expression of GALV-
`GP-R− across multiple cell
`lines (Fig. 2b). Next, the
`effects of GALV-GP-R− were assessed in human tumor
`models in nude mice in which A549 and MDA-MB-231
`tumor cells were grown in the flanks of mice and various
`doses of the viruses were tested for their ability to treat
`these pre-existing tumors. Again, expression of GALV-
`GP-R− was seen to significantly enhance anti-tumor
`activity (Fig. 2c-d), even when the viruses were used at
`low dose levels (data with the viruses used at a 5 × 103
`pfu dose level are shown).
`
`As GM-C