Recombinant Myxoma Virus Expressing Walleye Dermal Sarcoma Virus orfC Is Attenuated in Rabbits

The poxvirus, myxoma virus (MYXV) has shown efficacy as an oncolytic virus (OV) in some cancer models. However, MYXV replication within murine cancer models and spontaneous canine sarcomas is short-lived. In mice, successful treatment of tumors requires frequent injections with MYXV. We hypothesize that treatment of cancer with a recombinant MYXV that promotes apoptosis could improve the efficacy of MYXV. The orfC gene of walleye dermal sarcoma virus (WDSV), which induces apoptosis, was recombined into the MYXV genome (MYXVorfC). A marked increase in apoptosis was observed in cells infected with MYXVorfC. To ensure that expression of WDSV orfC by MYXV does not potentiate the pathogenesis of MYXV, we evaluated the effects of MYXVorfC inoculation in the only known host of MYXV, New Zealand white rabbits. Virus dissemination in rabbit tissues was similar for MYXVorfC and MYXV. Virus titers recovered from tissues were lower in MYXVorfC-infected rabbits as compared to MYXV-infected rabbits. Importantly, rabbits infected with MYXVorfC had a delayed onset of clinical signs and a longer median survival time than rabbits infected with MYXV. This study indicates that MYXVorfC is attenuated and suggests that MYXVorfC will be safe to use as an OV therapy in future studies.


Introduction
There is a growing interest in using viruses to eliminate cancers. The first oncolytic virus (OV) approved for use in the United States is marketed for treatment of melanoma in humans [1]. Although there are ongoing clinical trials testing the efficacy of oncolytic virotherapy in many types of human cancers, new OVs that are safer and more globally effective are actively being researched.
The poxvirus, MYXV is an excellent candidate oncolytic virotherapeutic because, unlike other OVs, it does not cause disease in humans or other vertebrates, with the exception of rabbits [2][3][4][5][6][7][8][9]. In spite of its species specificity, MYXV productively infects cultured cancer cells from several animal species [10][11][12][13]. In culture, data suggest that MYXV can replicate in neoplastic cells which have activated Akt [11,14] and lack appropriate Type I interferon responses to virus infection [12,15]. In rodent cancer models, MYXV treatment has eliminated some glioma xenografts [16] and reduced tumor burden of different allografts [17][18][19][20], but to date allografts have not been eliminated by MYXV treatment alone. We hypothesize that a recombinant MYXV with an enhanced oncolytic effect in cell culture will be a more potent anti-cancer therapy. However, before a new MYXV OV is tested in cancer models, its safety profile must be determined in its natural host, rabbits.

Viruses
Wild-type MYXV (Lausanne strain) and recombinant MYXV-red (originally designated vMyx-tdTr [24]) was originally acquired from Dr. Grant McFadden. MYXVorfC was constructed by recombining a 2385 base pair (bp) PCR product into wild-type MYXV between the genetic open reading frames M135 and M136 using a pBluescript plasmid vector ( Figure 1). Plasmid constructs were amplified by transformation into Escherichia coli DH5α chemically competent cells (Rapid5-a, Hardy Diagnostics, Santa Monica, CA, USA). Successful construction was confirmed using restriction enzyme digests and Sanger sequencing. DNA for transfection was prepared using PCR OneTaq Mastermix (New England Biolabs, Ipswitch, MA, USA) containing 0.5 µM of M13 forward primer (5 GTA AAA CGA CGG CCA GT 3 ), 0.5 µM of M13 reverse primer (5 CAG GAA ACA GCT ATG ACC 3 ), and 1.0 µg plasmid DNA. Amplification was performed in a thermocycler under the conditions: 94 • C for 1 min, followed by 30 cycles of 94 • C for 30 sec, 52 • C for 1 min, and 68 • C for 5 min, and a final 68 • C for 5 min step. The PCR product contained DNA sequences of: (1) the 5 fragment of MYXV M135, (2) tandem dimer tomato red (tdTomato) under the transcriptional control of a synthetic early/late poxvirus promoter (vvSynE/L), (3) hemagglutinin (HA)-tagged WDSV orfC under the control of a late poxvirus promoter (p11), and (4) the 3 fragment of MYXV M136. Transfection of the PCR fragment was performed using a modified method described by Rice et al. 2011 [25]. Briefly, RK-13 cells were infected at a multiplicity of infection (moi) of 0.01 with wild-type MYXV, transfected with 0.2 µg of PCR product DNA, and combined with Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA) to potentiate recombination of the PCR product and viral DNA. Cells were scraped into growth media at 72 h post-inoculation (hpi), centrifuged at 400× g for 15 min, washed in phosphate buffered saline (PBS), re-suspended in media lacking FBS, frozen and thawed 3 times, and sonicated. Viral lysates were serially diluted in media lacking FBS and incubated on RK-13 cells for 30 min. A solid overlay of 1 part 2× growth media and 1 part 1% agarose was placed on the infected cells. Viral foci that formed were screened for fluorescent red protein expression using a 560/40 nm bandpass excitation filter and a Leica DMI4000B inverted microscope. Fluorescent foci were isolated and expanded in RK-13 cells. The process of picking foci and growing viruses was repeated 9 times until only foci that expressed red fluorescent protein were observed. Viral purification was confirmed using PCR and next generation sequencing. For injection into rabbits, viruses were grown and titered in RK-13 cell cultures. Cellular debris was removed by sucrose pad purification as previously described [26].
Viruses 2020, 12, x FOR PEER REVIEW 3 of 17 overlay of 1 part 2× growth media and 1 part 1% agarose was placed on the infected cells. Viral foci that formed were screened for fluorescent red protein expression using a 560/40 nm bandpass excitation filter and a Leica DMI4000B inverted microscope. Fluorescent foci were isolated and expanded in RK-13 cells. The process of picking foci and growing viruses was repeated 9 times until only foci that expressed red fluorescent protein were observed. Viral purification was confirmed using PCR and next generation sequencing. For injection into rabbits, viruses were grown and titered in RK-13 cell cultures. Cellular debris was removed by sucrose pad purification as previously described [26].

Viral Growth Curves
Growth media was removed from wells of RK-13 cells when they were 80% confluent. Cells were inoculated with MYXV or MYXVorfC in media lacking FBS (moi = 0.1, n = 10 per group). Cells were incubated with virus for 1 h at 5% CO2 and 37 °C . Viral inoculum was removed, cells were rinsed with PBS, and growth media was added to wells. Cells were scraped into growth media at

Viral Growth Curves
Growth media was removed from wells of RK-13 cells when they were 80% confluent. Cells were inoculated with MYXV or MYXVorfC in media lacking FBS (moi = 0.1, n = 10 per group). Cells were incubated with virus for 1 h at 5% CO 2 and 37 • C. Viral inoculum was removed, cells were rinsed Viruses 2020, 12, 517 4 of 16 with PBS, and growth media was added to wells. Cells were scraped into growth media at designated time-points post-inoculation, frozen and thawed 3 times, and sonicated. Plaque assays were then performed to determine the number of infectious virions per mL of media. To perform plaque assays, viral lysates were serially diluted in media lacking FBS and incubated on RK-13 cells for 30 min. A solid overlay of 1 part 2× growth media and 1 part 1% agarose was placed on the infected cells. Viral foci were counted 4 days later. The log of plaque/focus-forming units (pfu) per mL was calculated and plotted versus time.

Detection of Exogenous Protein Production by MYXVorfC
Fluorescence from tdTomato protein expression by MYXVorfC was detected using a 560/40 nm bandpass excitation filter and a Leica DMI4000B inverted microscope. Production of the HA-tagged OrfC protein was detected using a Western immunoblot. Briefly, RK-13 cells were grown in 35 mm diameter plates to 90% confluency and inoculated with MYXVorfC (moi = 0.5). Infected cells were collected into cell lysis buffer at several time points pi. Total protein concentration was determined with a standard Bradford assay and 20 µg of protein from each cell lysate was separated using SDS-PAGE (10%). The SDS-PAGE-separated proteins were transferred to a nitrocellulose membrane. Membranes were incubated in blocking buffer for 1 h. Membranes were washed then incubated with rat anti-HA IgG (Milipore Sigma, St Louis, MO, USA) diluted 1:1000 in blocking buffer overnight at 4 • C. Membranes were washed then incubated for 1 h at room temperature with a horseradish peroxidase-conjugated goat anti-rat IgG antibody (ImmunoReagents, Raleigh, NC, USA) diluted 1:1000 in blocking buffer. A chemiluminescent Western immunoblot detection kit (GE Healthcare, Marlborough, MA, USA) was used to detect antigen-antibody complexes. These experiments were repeated a minimum of three times.

Apoptosis in Cell Culture
Growth media was removed from wells of RK-13 cells and cancer cells when they were 80% confluent. Cells were inoculated in media lacking FBS without virus (mock), with MYXV, or with MYXVorfC (moi = 1, n ≥ 22 per group). Cells were incubated with virus for 1 h at 5% CO 2 and 37 • C then growth media was added to the wells. A RealTime-Glo Annexin V Apoptosis and Necrosis Assay kit (Promega, Madison, WI, USA) was used per the manufacturer's instructions to label cells expressing phosphatidylserine. Luminescence and fluorescence (485 nm excitation/525 nm emission) were detected using a BioTek Synergy H1 microplate reader. The percent luminescent signal from cells undergoing apoptosis and the percent fluorescent signal from partially ruptured (necrotic) cells were graphed relative to mock-infected controls. The ratio of percent apoptotic to percent necrotic cells was also calculated and graphed.

Rabbits
All animal procedures were approved by the Institutional Animal Care and Use Committee at Colorado State University (protocol #17-7708A). Twelve 8-week-old, female, New Zealand white rabbits were purchased from Western Oregon Rabbit Company (Philomath, OR, USA) and were acclimated for 5 days. Prior to starting the study, all rabbits were deemed healthy by physical examination including determination of heart rate and respiratory rate; auscultation of the heart and lungs; palpation of the abdomen and lymph nodes; measurement of weight and rectal temperature; evaluation of the injection site, haired skin, and mucous membranes for evidence of lesion formation; and assessment of food and water intake, grooming behavior, and mentation. An area of fur on the lateral aspect of the upper right thigh was shaved and disinfected with an accelerated hydrogen peroxide solution. Rabbits were inoculated with virus via an intradermal injection of 50 pfu of sucrose pad purified wild-type MYXV (n = 6) or MYXVorfC (n = 6) in 100 µL PBS. The area was cleaned with an accelerated hydrogen peroxide solution after the virus is administered. Rabbits were given physical examinations daily. The lesion length, width, and height at the site of inoculation were measured daily. Lesion volume was Viruses 2020, 12, 517 5 of 16 calculated as volume = length × width × height. The clinical scoring system outlined in Rice et al., JVI 2014 was used [27]. Animals were euthanized as soon as signs of respiratory distress were observed; if rectal temperature was less than 36 • C; if the rabbit appeared mentally dull with poor response to noise or light; if the rabbit stopped grooming, eating, or drinking; or at Day 21 post-inoculation.

Sample Collection
Prior to euthanasia, rabbits were anesthetized with a freshly-mixed combination of 50 mg/kg ketamine and 10 mg/kg xylazine by intramuscular injection into the proximal and lateral aspect of the left thigh. Fully anesthetized rabbits (unresponsive to noxious stimulation, such as pinching the toes with forceps) were administered an overdose of intracardiac pentobarbital (120 mg/kg) to ensure they were deceased. After euthanasia, a necropsy was performed to collect skin lesions and all internal organs. Two sections of each tissue were collected; one was flash frozen in liquid nitrogen for isolation of virus and detection of viral DNA and one was preserved in 10% buffered formalin for histopathology.

Detection of Virus in Tissues
Flash frozen tissue sections were weighed and homogenized. Samples were serially diluted in media without FBS, sonicated, and inoculated onto RK-13 cells for 30 min. A solid overlay of 1 part 2× growth media and 1 part 1% agarose was placed on the infected cells. Viral foci were counted 4-6 days later. Pfu per mg of tissue calculated. Additionally, DNA was isolated from sample homogenates using a tissue DNA isolation kit (DNeasy Blood and Tissue Kit, QIAGEN, Hilden, Germany). Next, 5 µL of DNA was used in PCR reactions with two primer sets. A set of primers that detect a multigenic region of the MYXV genome that includes M032R, M033R, M034L (DNA polymerase), and M035R was used to detect MYXV in tissues (Forward 5 CAC CCT CTT TAG TAA AGT ATA CAC C 3 , Reverse 5 GAA ATG TTG TCG GAC GGG 3 ). An 800 bp product is detected for both MYXV and MYXVorfC with these primers. A second set of primers was also used (Forward 5 ACA TAC GAC ATC GGA CAG CA 3 , Reverse 5 CGT CGA TCG CTG TGT AAG AA 3 ) that covered the region of the gene insert (M135-M136); a 320 bp product is expected for wild-type MYXV and a 1547 bp product is expected for MYXVorfC. Amplification was performed in a thermocycler under the conditions: 94 • C for 1 min, followed by 30 cycles of 94 • C for 30 sec, 55 • C for 1 min, 72 • C for 2 min and a final 72 • C for 10 min elongation step.

Histopathology
Formalin-fixed tissue was paraffin embedded, sectioned, and stained with hematoxylin and eosin (H&E) for histologic evaluation. The extent of inflammation, necrosis, and edema observed in skin lesions histologically was graded as previously described [20]. Similarly, additional sections were used to detect apoptosis using a TUNEL (TdT-mediated dUTP Nick End Labeling) assay (Click-iT TUNEL, Life Technologies, Carlsbad, CA, USA) and the extent of apoptosis was graded.

Statistical Analysis
Median survival time was calculated using Kaplan-Meier survival curves. The apoptosis to necrosis ratio in RK-13 cells; clinical scores, respiratory rate, heart rate, body temperature, and lesion volume in rabbits; virus titers in tissues; and the extent of inflammation, necrosis, edema, and apoptosis in histologic sections were compared using t-tests. All data were analyzed using GraphPad Prism version 8.1.0 software (GraphPad, San Diego, CA, USA).

Validation of MYXVorfC Construction
A diagram of the PCR product introduced between genes M135 and M136 to create the recombinant MYXVorfC virus is shown in Figure 2A. Restriction enzyme digestions of MYXVorfC indicated the Viruses 2020, 12, 517 6 of 16 predicted DNA fragment lengths. MYXVorfC sequence analysis indicated that no unexpected genetic changes were caused during recombination of the PCR fragment into wild-type MYXV ( Figure 2B).
Viruses 2020, 12, x FOR PEER REVIEW 6 of 17 unexpected genetic changes were caused during recombination of the PCR fragment into wild-type MYXV ( Figure 2B). Growth curve kinetics of wild-type MYXV and MYXVorfC in RK-13 cells were statistically indistinguishable ( Figure 3), indicating that insertion of the PCR product did not negatively affect MYXVorfC replication in permissive RK-13 cells. Next, protein expression of orfC and tdTomato by MYXVorfC was assessed. Red fluorescence was detected in RK-13 cell cultures inoculated with MYXVorfC ( Figure 4). The number of cells expressing tdTomato was similar when MYXV-red and MYXVorfC infections were compared ( Figure 4A). However, subtle morphologic differences were noted. Some MYXVorfC-infected cells contained large cytoplasmic vacuoles, which were not observed in MYXV-red-infected cells ( Figure 4B). OrfC protein expression was detected by 12 hpi using Western blot analysis ( Figure 5). Together, these data indicate that tdTomato and orfC were successfully inserted into wild-type MYXV to create a recombinant MYXVorfC that expresses detectable amounts of tdTomato and orfC proteins.   Growth curve kinetics of wild-type MYXV and MYXVorfC in RK-13 cells were statistically indistinguishable (Figure 3), indicating that insertion of the PCR product did not negatively affect MYXVorfC replication in permissive RK-13 cells. Next, protein expression of orfC and tdTomato by MYXVorfC was assessed. Red fluorescence was detected in RK-13 cell cultures inoculated with MYXVorfC ( Figure 4). The number of cells expressing tdTomato was similar when MYXV-red and MYXVorfC infections were compared ( Figure 4A). However, subtle morphologic differences were noted. Some MYXVorfC-infected cells contained large cytoplasmic vacuoles, which were not observed in MYXV-red-infected cells ( Figure 4B). OrfC protein expression was detected by 12 hpi using Western blot analysis ( Figure 5). Together, these data indicate that tdTomato and orfC were successfully inserted into wild-type MYXV to create a recombinant MYXVorfC that expresses detectable amounts of tdTomato and orfC proteins.

Apoptosis Is Induced By MYXVorfC
The purpose of using MYXV to express the orfC protein was to enhance apoptosis in infected cancer cells. Apoptosis was greatly increased in RK-13 cells infected with MYXVorfC as compared to MYXV-red infected or mock-infected cells ( Figure 6).

Apoptosis Is Induced By MYXVorfC
The purpose of using MYXV to express the orfC protein was to enhance apoptosis in infected cancer cells. Apoptosis was greatly increased in RK-13 cells infected with MYXVorfC as compared to MYXV-red infected or mock-infected cells ( Figure 6).

MYXVorfC Is Attenulated in Rabbits
As expected, all New Zealand white rabbits infected with 50 pfu wild-type MYXV needed to be euthanized due to respiratory distress. One of six rabbits infected with 50 pfu of MYXVorfC survived. The median survival time of rabbits inoculated with MYXVorfC was significantly greater than that of rabbits inoculated with MYXV ( Figure 7). The median survival times for MYXV-and MYXVorfC-infected rabbits were 9 and 12 Days pi (dpi), respectively.

MYXVorfC Is Attenulated in Rabbits
As expected, all New Zealand white rabbits infected with 50 pfu wild-type MYXV needed to be euthanized due to respiratory distress. One of six rabbits infected with 50 pfu of MYXVorfC survived. The median survival time of rabbits inoculated with MYXVorfC was significantly greater than that of rabbits inoculated with MYXV ( Figure 7). The median survival times for MYXV-and MYXVorfCinfected rabbits were 9 and 12 Days pi (dpi), respectively. Development of clinical signs of disease (including respiratory distress and dissemination of virus to form edematous, ulcerative skin lesions) was slightly delayed in rabbits infected with MYXVorfC ( Figure 8). Slightly, but significantly, lower clinical scores were observed in MYXVorfC rabbits on Days 5, 6, and 8 after virus injection ( Figure 9A). These differences were attributed to the delay in development of secondary lesions on mucus membranes and ear pinnae in rabbits infected with MYXVorfC ( Figure 9B) and detection of fever (body temperature > 40 °C , Figure 10A). Small differences in mean respiratory rate ( Figure 10B) and heart rate ( Figure 10C) were found on certain dpi, but no definitive trends were seen over time. No significant differences in lesion volume were calculated, but the average volume of the primary lesion of rabbits infected with MYXVorfC decreased in animals that survived beyond Day 12 pi ( Figure 10D). These data indicate that MYXVorfC is less pathogenic than wild-type MYXV.  Development of clinical signs of disease (including respiratory distress and dissemination of virus to form edematous, ulcerative skin lesions) was slightly delayed in rabbits infected with MYXVorfC ( Figure 8). Slightly, but significantly, lower clinical scores were observed in MYXVorfC rabbits on Days 5, 6, and 8 after virus injection ( Figure 9A). These differences were attributed to the delay in development of secondary lesions on mucus membranes and ear pinnae in rabbits infected with MYXVorfC ( Figure 9B) and detection of fever (body temperature > 40 • C, Figure 10A). Small differences in mean respiratory rate ( Figure 10B) and heart rate ( Figure 10C) were found on certain dpi, but no definitive trends were seen over time. No significant differences in lesion volume were calculated, but the average volume of the primary lesion of rabbits infected with MYXVorfC decreased in animals that survived beyond Day 12 pi ( Figure 10D). These data indicate that MYXVorfC is less pathogenic than wild-type MYXV.

MYXVorfC Is Attenulated in Rabbits
As expected, all New Zealand white rabbits infected with 50 pfu wild-type MYXV needed to be euthanized due to respiratory distress. One of six rabbits infected with 50 pfu of MYXVorfC survived. The median survival time of rabbits inoculated with MYXVorfC was significantly greater than that of rabbits inoculated with MYXV ( Figure 7). The median survival times for MYXV-and MYXVorfCinfected rabbits were 9 and 12 Days pi (dpi), respectively. Development of clinical signs of disease (including respiratory distress and dissemination of virus to form edematous, ulcerative skin lesions) was slightly delayed in rabbits infected with MYXVorfC ( Figure 8). Slightly, but significantly, lower clinical scores were observed in MYXVorfC rabbits on Days 5, 6, and 8 after virus injection ( Figure 9A). These differences were attributed to the delay in development of secondary lesions on mucus membranes and ear pinnae in rabbits infected with MYXVorfC ( Figure 9B) and detection of fever (body temperature > 40 °C , Figure 10A). Small differences in mean respiratory rate ( Figure 10B) and heart rate ( Figure 10C) were found on certain dpi, but no definitive trends were seen over time. No significant differences in lesion volume were calculated, but the average volume of the primary lesion of rabbits infected with MYXVorfC decreased in animals that survived beyond Day 12 pi ( Figure 10D). These data indicate that MYXVorfC is less pathogenic than wild-type MYXV.

MYXVorfC Tissue Tropism Is Not Altered
Dissemination of wild-type MYXV and MYXVorfC during infection of rabbits appeared similar. Viral DNA was detected in skin, lungs, heart, spleen, liver, and kidney in all MYXV-infected rabbits

MYXVorfC Tissue Tropism Is Not Altered
Dissemination of wild-type MYXV and MYXVorfC during infection of rabbits appeared similar. Viral DNA was detected in skin, lungs, heart, spleen, liver, and kidney in all MYXV-infected rabbits Figure 10. Average daily measurements during physical examination of rabbits infected with 50 pfu of MYXV or MYXVorfC (n = 6 per group). Small, but statistically significant differences were observed on a few days when average (A) body temperature, (B) respiratory rate, and (C) heart rate were compared. No significant differences were observed when (D) average daily lesion volumes were calculated (error bars = SEM, * p-values < 0.046).

MYXV and MYXVorfC Induce Similar Histologic Lesions
No overt histologic abnormalities were appreciated in heart or kidney sections. Viral pneumonia was observed in lung tissue from all rabbits with mild to moderate edema and diffuse suppurative inflammation. Sloughing of bronchial epithelial cells into the bronchial lumen was noted. Mild hyperplasia of perivascular lymphoid tissue of the lungs was observed in both MYXV-and MYXVorfC-infected rabbits. Evidence of disease in the liver was more variable. Mild diffuse lymphocytic inflammation was observed in the livers of 67% (4/6) MYXV-infected rabbits and 33% (2/6) MYXVorfC-infected rabbits. Mild diffuse mixed inflammation (including neutrophils, histiocytes, lymphocytes, and plasma cells) was observed in the liver of 50% (3/6) MYXVorfC-infected rabbits. No inflammation was observed in the liver of the remaining 33% (2/6) MYXV-infected and 17% (1/6) MYXVorfC-infected rabbits. Tissue sections from primary and secondary skin lesions of all rabbits had the classic appearance of myxomatosis in European rabbits. A large focus of full-thickness dermal necrosis was associated with several large reactive fibroblasts and myxedema (Figure 12). Fibroblasts and epithelial cells sometimes contained a large basophilic cytoplasmic viral inclusion. Marked suppurative inflammation was present within the necrotic lesion and moderate mononuclear inflammation surrounded the damaged tissue. The average number of TUNEL positive cells in ten 40× objective microscopic fields in sections of primary skin lesions was not significantly different in MYXVand MYXVorfC-infected rabbits ( Figure 13). The secondary lesion from one MYXVorfC-infected rabbit was beginning to heal with extension of the epidermis under a layer of fibrin and cellular debris.

MYXV and MYXVorfC Induce Similar Histologic Lesions
No overt histologic abnormalities were appreciated in heart or kidney sections. Viral pneumonia was observed in lung tissue from all rabbits with mild to moderate edema and diffuse suppurative inflammation. Sloughing of bronchial epithelial cells into the bronchial lumen was noted. Mild hyperplasia of perivascular lymphoid tissue of the lungs was observed in both MYXV-and MYXVorfC-infected rabbits. Evidence of disease in the liver was more variable. Mild diffuse lymphocytic inflammation was observed in the livers of 67% (4/6) MYXV-infected rabbits and 33% (2/6) MYXVorfC-infected rabbits. Mild diffuse mixed inflammation (including neutrophils, histiocytes, lymphocytes, and plasma cells) was observed in the liver of 50% (3/6) MYXVorfC-infected rabbits. No inflammation was observed in the liver of the remaining 33% (2/6) MYXV-infected and 17% (1/6) MYXVorfC-infected rabbits. Tissue sections from primary and secondary skin lesions of all rabbits had the classic appearance of myxomatosis in European rabbits. A large focus of full-thickness dermal necrosis was associated with several large reactive fibroblasts and myxedema (Figure 12). Fibroblasts and epithelial cells sometimes contained a large basophilic cytoplasmic viral inclusion. Marked suppurative inflammation was present within the necrotic lesion and moderate mononuclear inflammation surrounded the damaged tissue. The average number of TUNEL positive cells in ten 40× objective microscopic fields in sections of primary skin lesions was not significantly different in MYXV-and MYXVorfC-infected rabbits ( Figure 13). The secondary lesion from one MYXVorfCinfected rabbit was beginning to heal with extension of the epidermis under a layer of fibrin and cellular debris.

MYXVorfC Stimulated Follicular Hyperplasia in Lymphoid Tissues
Spleen and lymph nodes from uninfected rabbits and rabbits infected with MYXV had minimal lymphoid reactivity ( Figure 14A,B). In contrast, MYXVorfC-infected rabbits developed lymphoid hyperplasia ( Figure 14C) suggestive of a more activated adaptive systemic immune response to the virus.

Discussion
Development of safe, effective OVs could revolutionize cancer treatment, particularly for neoplasms that are refractory to currently available therapies. MYXV is nonpathogenic in humans [6] and shows potential as a safe OV [29,30]. However, the effectiveness of MYXV is somewhat limited. We predict that OVs that induce apoptosis of tumor cells quickly can expose the immune response to tumor antigens more rapidly and negate limitations of rapid removal of MYXV (which we hypothesize is due to the anti-viral response induced by healthy cells within the tumor microenvironment). With this in mind, we designed a pro-apoptotic, recombinant MYXV, MYXVorfC.
MYXVorfC contains an HA-tagged orfC gene isolated from WDSV. WDSV orfC may promote seasonal tumor regression in fish infected with WDSV by inducing cytochrome C release from mitochondria [21]. Expression of orfC by MYXVorfC was placed under the control of a late poxvirus promoter element (vaccinia p11) with the expectation that complete replication of MYXVorfC and production of infectious virions would occur before significant apoptosis was induced in infected

MYXVorfC Stimulated Follicular Hyperplasia in Lymphoid Tissues
Spleen and lymph nodes from uninfected rabbits and rabbits infected with MYXV had minimal lymphoid reactivity ( Figure 14A,B). In contrast, MYXVorfC-infected rabbits developed lymphoid hyperplasia ( Figure 14C) suggestive of a more activated adaptive systemic immune response to the virus.

MYXVorfC Stimulated Follicular Hyperplasia in Lymphoid Tissues
Spleen and lymph nodes from uninfected rabbits and rabbits infected with MYXV had minimal lymphoid reactivity ( Figure 14A,B). In contrast, MYXVorfC-infected rabbits developed lymphoid hyperplasia ( Figure 14C) suggestive of a more activated adaptive systemic immune response to the virus.

Discussion
Development of safe, effective OVs could revolutionize cancer treatment, particularly for neoplasms that are refractory to currently available therapies. MYXV is nonpathogenic in humans [6] and shows potential as a safe OV [29,30]. However, the effectiveness of MYXV is somewhat limited. We predict that OVs that induce apoptosis of tumor cells quickly can expose the immune response to tumor antigens more rapidly and negate limitations of rapid removal of MYXV (which we hypothesize is due to the anti-viral response induced by healthy cells within the tumor microenvironment). With this in mind, we designed a pro-apoptotic, recombinant MYXV, MYXVorfC.
MYXVorfC contains an HA-tagged orfC gene isolated from WDSV. WDSV orfC may promote seasonal tumor regression in fish infected with WDSV by inducing cytochrome C release from mitochondria [21]. Expression of orfC by MYXVorfC was placed under the control of a late poxvirus promoter element (vaccinia p11) with the expectation that complete replication of MYXVorfC and production of infectious virions would occur before significant apoptosis was induced in infected

Discussion
Development of safe, effective OVs could revolutionize cancer treatment, particularly for neoplasms that are refractory to currently available therapies. MYXV is nonpathogenic in humans [6] and shows potential as a safe OV [29,30]. However, the effectiveness of MYXV is somewhat limited. We predict that OVs that induce apoptosis of tumor cells quickly can expose the immune response to tumor antigens more rapidly and negate limitations of rapid removal of MYXV (which we hypothesize is due to the anti-viral response induced by healthy cells within the tumor microenvironment). With this in mind, we designed a pro-apoptotic, recombinant MYXV, MYXVorfC.
MYXVorfC contains an HA-tagged orfC gene isolated from WDSV. WDSV orfC may promote seasonal tumor regression in fish infected with WDSV by inducing cytochrome C release from mitochondria [21]. Expression of orfC by MYXVorfC was placed under the control of a late poxvirus promoter element (vaccinia p11) with the expectation that complete replication of MYXVorfC and production of infectious virions would occur before significant apoptosis was induced in infected cells. Additionally we engineered MYXVorfC to encode the reporter element tdTomato under the control of an early/late poxvirus promoter (vvSynE/L) to enable us to screen for and plaque purify recombinant virus. Our sequencing data indicated that the HA-tagged orfC gene and the reporter element tdTomato were successfully recombined into wild-type MYXV. The virus growth curve in RK-13 cells was not altered by recombination of tdTomato or HA-orfC into MYXV. Protein expression of tdTomato was observed as early as 8 hpi and HA-tagged orfC was detected by 12 hpi. Importantly, MYXVorfC significantly increased apoptosis in cell culture as compared to wild-type MYXV. These data indicate that MYXVorfC is able to replicate in cells, express exogenous proteins, induce apoptosis, spread to neighboring cells, and continue to replicate until all cells have been infected. Apoptosis did not reduce the virus growth rate because virus particles that are formed during infection remain infectious when the cells loose viability.
Next, we infected New Zealand white rabbits with MYXVorfC to ensure that the pathogenicity of the recombinant virus was not more severe than wild-type MYXV. During wild-type MYXV infection, virus replicates to high titers at the primary site of virus inoculation. Within 2 to 4 dpi, viremia occurs allowing for systemic dissemination of infectious viral particles. MYXV localizes to the skin and mucous membranes by 5 dpi, replicates to high titers, and causes secondary myxomatous lesions, fever, and viral pneumonia. This leads to severe obstruction of the respiratory tract, respiratory arrest, and death between 8 and 12 dpi [28]. In comparison, rabbits infected with MYXVorfC had a significantly longer median survival time. Five animals required euthanasia between 11 and 16 dpi, but one animal recovered from MYXVorfC infection. Differences in overall clinical scores were minimal and mostly due to a short delay in the development of secondary lesions and fever in rabbits infected with MYXVorfC. The clinical similarities in infected rabbits are not unexpected, as no MYXV genes were disrupted by insertion of orfC into the virus. Therefore, any attenuation observed was due to expression of orfC by MYXVorfC.
Although MYXVorfC infection was attenuated, the dissemination of MYXVorfC was similar to MYXV. Virus DNA and infectious virions were detected for both viruses in all organs tested. However, significantly lower MYXVorfC titers were recovered from the primary skin lesion at the site of inoculation, the lung tissue, and the spleen. This could reflect a mild reduction of MYXVorfC replication in rabbit tissues or increased clearance of the virus by the immune response. MYXVorfC rabbits had clear evidence of lymphoid hyperplasia, which was lacking in MYXV-infected rabbits. This suggests that rabbits mount a more effective immune response against MYXVorfC. Lymphocytic inflammation is also observed during regression of tumors caused by WDSV [22]. If lymphocytic inflammation is induced by orfC expression during oncolytic virotherapy, it could greatly improve outcomes for cancer patients being treated with MYXVorfC.
MYXVorfC was engineered to increase apoptosis during infection. This was confirmed using RK-13 cells in tissue culture. In rabbits, the average number of TUNEL positive cells in MYXV-infected primary lesions (26.7 ± 11.7 /hpf) was lower than that of MYXVorfC-infected lesions (37.2 ± 7.1 /hpf). Although this difference was not statistically significant, a slight increase in TUNEL positive cells was consistently observed in MYXVorfC-infected tissue. It is possible that MYXV-induced apoptosis will be more enhanced in cancerous cells.
In conclusion, MYXVorfC is pro-apoptotic in cell culture and attenuated in rabbits. The marked lymphoid hyperplasia observed in rabbits infected with MYXVorfC suggests that expression of orfC allows rabbits to mount a more effective immune response against MYXVorfC than wild-type MYXV. A very mild increase in apoptosis at the site of MYXVorfC inoculation was also observed. It is possible that the combination of increased immunogenicity and apoptosis contribute to the decreased pathogenicity of the virus. Future studies will test MYXVorfC treatment in animal cancer models to determine if MYXVorfC is more effective than MYXV at eliminating cancer.