Orally available nucleoside analog UMM-766 provides protection in a murine model of orthopox disease

ABSTRACT Although smallpox has been eradicated, other orthopoxviruses continue to be a public health concern as exemplified by the ongoing Mpox (formerly monkeypox) global outbreak. While medical countermeasures (MCMs) previously approved by the Food and Drug Administration for the treatment of smallpox have been adopted for Mpox, previously described vulnerabilities coupled with the questionable benefit of at least one of the therapeutics during the 2022 Mpox outbreak reinforce the need for identifying and developing other MCMs against orthopoxviruses. Here, we screened a panel of Merck proprietary small molecules and identified a novel nucleoside inhibitor with potent broad-spectrum antiviral activity against multiple orthopoxviruses. Efficacy testing of a 7-day dosing regimen of the orally administered nucleoside in a murine model of severe orthopoxvirus infection yielded a dose-dependent increase in survival. Treated animals had greatly reduced lesions in the lung and nasal cavity, particularly in the 10 µg/mL dosing group. Viral levels were also markedly lower in the UMM-766-treated animals. This work demonstrates that this nucleoside analog has anti-orthopoxvirus efficacy and can protect against severe disease in a murine orthopox model. IMPORTANCE The recent monkeypox virus pandemic demonstrates that members of the orthopoxvirus, which also includes variola virus, which causes smallpox, remain a public health issue. While currently FDA-approved treatment options exist, risks that resistant strains of orthopoxviruses may arise are a great concern. Thus, continued exploration of anti-poxvirus treatments is warranted. Here, we developed a template for a high-throughput screening assay to identify anti-poxvirus small-molecule drugs. By screening available drug libraries, we identified a compound that inhibited orthopoxvirus replication in cell culture. We then showed that this drug can protect animals against severe disease. Our findings here support the use of existing drug libraries to identify orthopoxvirus-targeting drugs that may serve as human-safe products to thwart future outbreaks.

. MPXV is arguably the most impactful with the greatest potential for global impact, due in large part to the cessation of smallpox vaccinations leading to immunolog ically naïve populations (7).There have been increased incidences of exported cases to various countries over the past several decades, including a 2003 outbreak in the midwestern United States as well as other sporadic (non-USA) outbreaks in subsequent years (8,9).In 2022, MPXV became a global outbreak and there were 78,000 confirmed cases involving >100 countries as of April 2023.The 2022 outbreak, in addition to the 2003 outbreak, demonstrates that orthopoxvirus zoonoses will continue to emerge outside traditional endemic areas and impact human health on a global scale.
Orthopoxviruses are large double-stranded DNA viruses that infect and replicate in the cytoplasm of a variety of mammalian cells.There are two major infectious forms of poxvirus virions (10).The first form is known as intracellular mature virions (MV).MV are thought to be important for inter-host transmission and are primarily released after cell rupture or lysis.A subset of MV is further wrapped (WV) and either released from the cell [extracellular enveloped virions (EEV)] or associated with the outer membrane of the cell [cell-associated enveloped virions (CEV)] and potentially released (rCEV) (11).The EEV and CEV/rCEV are collectively known as extracellular virions (EV), the second major form of poxvirus virions; EV are thought to be responsible for dissemination within a host (12).Because viral proteins on the outer membrane of the EV are distinct relatives to the MV form, targeting both forms is required for optimal protection by vaccines and antibodies (13)(14)(15).Given the numerous viral proteins required for the complex viral cycle of poxviruses, there are a variety of novel targets for therapeutic development that have been identified (16).Small-molecule tool compounds target specific antibodies, and some licensed antiviral drugs that block different stages of the viral life cycle have been identified (16).
To date, there are two FDA-licensed drugs for the treatment of human smallpox disease, ST-246 (TPOXX, ST-246) and brincidofovir (Tembexa, CMX001).Both orally administrated drugs were approved by the FDA under the Animal Rule (21 CFR 314).ST-246 targets the product of the F13L gene and is thought to disrupt morphogenesis and the formation of EV.Drug-resistant orthopoxvirus variants with publicly described mutations at known positions in the F13 protein have been identified in vitro (17,18).Changes in the sensitivity of virus samples to ST-246 sequentially obtained from the lesion of a patient with progressive vaccinia undergoing an extended treatment of ST-246 had been reported, but the reason for the reported resistance was unknown (19).Although seemingly efficacious against Mpox during the 2022 outbreak, resistant MPXV has been reported and is of concern for immunocompromised individuals requiring prolonged treatment (20)(21)(22).Brincidofovir is a nucleoside analog that inhibits the poxvirus DNA polymerase and also decreases the efficiency of viral DNA synthesis by incorporation into the growing viral DNA chain, thereby limiting viral replication within target cells.Specific mutations in the E9 protein can confer resistance to the drug; however, mutant viruses maybe less fit in animal systems.Brincidofovir has significant side effects, including deleterious effects on male fertility, and causes an increased risk for mortality when used for longer durations (brincidofovir, product insert), and initial assessments of efficacy and tolerability in the current Mpox outbreak have not been encouraging (20).For both approved drugs, the potential for pre-existing or drug-selec ted resistant mutants is a concern.Thus, the continued development of alternative therapeutics is warranted to ensure a countermeasure is available in the event of an outbreak involving a strain of orthopoxvirus resistant to the two approved drugs.With the globalization of Mpox and subsequent response to treat the afflicted, there is a much greater likelihood that resistant strains will emerge.
In an effort to identify inhibitors of orthopoxviruses, a panel of Merck proprietary small molecules was screened using a high-throughput high-content image-based phenotypic assay.Among the active hits from the screen, a novel nucleoside inhibitor UMM-766 was identified that demonstrated potent, broad-spectrum antiviral activity against multiple members of the pox virus family.The protective efficacy of orally delivered UMM-766 against VACV infection was evaluated in a BALB/c mouse intranasal challenge model.

Screening collection
Targeted compounds from the Merck proprietary sample collection were provided as 10-mM DMSO solutions to USAMRIID for screening.Compounds were selected based primarily on the mechanism of action corresponding to known essential orthopoxvirus classes of enzymes.Examples include ribose and deoxyribose nucleoside polymerase inhibitors, cysteine protease inhibitors, autophagy modulators, and kinase inhibitors.Approximately 20% of the collection screened was composed of compounds that reflect the diversity of the Merck compound collection.Compounds with supporting informa tion available that enabled a more rapid path to animal studies were prioritized.

Viruses
The Western Reserve strain of VACV and the Brighton Red Strain of CPXV were obtained from Biodefense and Emerging Infections Resources Repository (BEI Resources) having catalog numbers NR-55 and NR-88, respectively.The Utrecht strain of rabbitpox virus (RPXV), VR-1591, was purchased from the American Type Culture Collection (ATCC).The USAMRIID sequence verified stocks MPXV West African clade US2003 which is now referred to clade IIa was the source of MPXV in our assay.The viruses were passaged on Vero 76 (BEI) at an multiplicity of infection (MOI) of 0.01, titrated, sequenced, and determined to be contaminant free by culture on chocolate agar plates, endotoxin test (Charles River Endosafe nexgen-PTS), MycoAlert test kit, and PCR exclusivity testing by the Unified Culture Collection (Diagnostic System Division)], USAMRIID.For the VACV stock, a confirmation of virulence study was performed in BALB/c mice (Fig. S1).Virus propagation and characterization were performed by the Unified Culture Collection group at USAMRIID.

Virus infections, immunofluorescence, and high-content quantitative image-based analysis
Infections with VACV, CPXV, and RPXV were conducted under Biosafety Laboratory 2 (BSL-2) conditions.To quantify infections, an image-based assay was developed in a 384-well format as described previously (24,25).Briefly, the optimal conditions such as the cell number/well, duration of infection, and multiplicity of infection (MOI) were determined so that greater than 60% of the cells were infected, with a Z-factor > 0.5.Z-factor was evaluated from the average and standard deviation values derived from eight wells of each mock-infected and virus-infected wells.The duration of infection was such that it encompassed more than two virus life cycles to ensure that all types of inhibitors, irrespective of the stage of the virus life cycle being targeted, were captured by the assay.Table 1 summarizes the optimal parameters identified for each of the three viruses in the two different cell types.
The treatment and infections were performed as follows.Cells were seeded into imaging 384-well assay plates (IQ-EB, Aurora) using the automated MultidropTM Combi dispenser (cat # 5840300, Thermo Fisher Scientific).After an overnight incubation, cells were pretreated for 2 hours with the compounds using an automated HP-D300 digital dispenser with each dose dispensed directly from the concentrated stocks.The compounds were tested in a 10-point dose response curve assay at threefold serial dilution starting from 30 µM, and each dose was tested in four technical replicates (n = 4) on the same plate.Cells were then infected with the optimal MOI of the virus, and after optimal infection time, the plates were fixed in 10% formalin for 24 hours.The cells were then subjected to immunofluorescence staining.
To quantify infections, the formalin-fixed cells were subjected to an immunofluorescence assay, followed by image analysis to enumerate the antigen-expressing cells.Briefly, the formalin-fixed cells were washed with phosphate-buffered saline (PBS) and then permeabilized in 0.1% Triton X-100 (Sigma) containing blocking buffer.The latter was composed of 3% Bovine Serum Albumin (Lampire, cat # 7500806) in PBS.Following blocking, cells were fluorescently stained using murine monoclonal antibody against VACV (5F8), followed by DyLight488-conjugated goat anti-mouse IgG (Thermo Fisher Scientific, catalog # 35502) to visualize viral antigen-expressing cells.To capture cell nuclei and cell cytoplasm by fluorescence imaging, cells were stained with Hoechst 33342 and HCS CellMask Red (Thermo Fisher Scientific), respectively.Images were acquired on the Opera High Content Imaging instrument (model 3842 and 5025; PerkinElmer) using the 10× air objective.Images were analyzed using Acapella 2.0, 2.6, a Dash (−) indicates that values could not be generated.
2.7 (PerkinElmer) scripts in EvoShell or the building-blocks interface in the Columbus image analysis server (PerkinElmer).Nuclei staining and cytoplasm staining were used to determine the total cell number and cell borders, respectively.Mock-infected cells were used to set the baseline intensity threshold to exclude the background noise from antibody staining.Quantification of virus-positive cells was subsequently performed based on mean fluorescent intensities in the virus-specific staining channel.Infection rates were then determined by dividing the number of virus-positive cells by the total number of cells.At least 5,000 cells were analyzed per replicate in drug-treated cells.

Dose response curve analysis
Each compound's activity against VACV including EC 50 or EC 90 , which is the effective concentration to achieve 50% or 90% infection inhibition, respectively, and CC 50 (50% cytotoxicity) was accomplished by running the dose response curve analysis (DRA) using the GeneData Explorer software.The percentage infection inhibition and percentage cell viability were evaluated using the following controls.Sixteen wells/plate served as neutral control (NC), where in the cells were infected but treated with DMSO (mock vehicle control).Sixteen wells per plate were uninfected and served as blank control (BC), to establish a fluorescence intensity threshold for virus-specific staining.In addition, every plate included dose response with ST-246 that served as a reference inhibitor.
Since the in vitro efficacy of ST-246 against rabbitpox virus has not previously been reported, RPXV was screened for informational purposes but not reported here (e.g., utilized as a reference inhibitor).The plates were considered "pass" for analysis if the statistical parameter, Z-factor, that measures the robustness of the assay was greater than 0.5.The Z-factor was evaluated by using the average and standard deviations derived from the NC and BC.In addition, ST-246 showed a dose response with EC 50 in the 1-10 nM range and CC 50 values > 30 µM in all the plates that were considered pass.The percentage of infection inhibition was determined as follows: the percentage infection from the individual wells was first normalized with the average infection from NC and then multiplied by 100.Finally, the resulting value was subtracted from 100.Similarly, the percentage cell viability was calculated by normalizing the cell number in the individual wells with the average cell number from NC and multiplying the result by 100.
The percentage inhibition and the percentage cell viability were then used to determine the EC 50 , EC 90 , and CC 50 values by applying the GeneData Condoseo software with the Levenberg Marquardt algorithm (LMA) for curve fitting strategy.Validity criteria, including the chi-square statistic (χ 2 ), the standard error of the logEC 50 , and the number of valid data points, were applied to indicate if curve fitting was successful (Table 1).The R 2 value quantifies goodness of fit.Fitting strategy was considered acceptable if R 2 > 0.8.The relative effectiveness of the compound is defined in terms of its selectivity index (SI), a value that indicates the relationship between the compound's effective and toxic concentrations and is calculated as follows: SI = CC 50 /EC 50 .It is therefore desirable for a compound to have a high SI value, indicating maximum antiviral activity and minimal cell toxicity.

Animals and experimental design
Six-to eleven-week-old, male and female BALB/c (NCI) mice were obtained from Charles River.A drug tolerability study was performed before performing efficacy studies.Four groups of four mice per group (6-8 weeks), equal male and female, were orally treated for 7 days with either saline or three different doses of UMM-766, 1 mg/kg, 3 mg/kg, or 10 mg/kg.Animals were observed for changes in natural and provoked behaviors, as well as physical condition/appearance. Two efficacy experiments were performed in either mice aged 7 weeks or in a group of mice aged 10-11 weeks.Use of older mice reduces the level of cachexia resulting in fewer animals reaching euthanasia criteria.For both efficacy experiments, mice were anesthetized using isoflurane and exposed via intranasal instillation to a target dose of 5.5 × 10 5 PFU on Day 0. Animals were administered 1, 3, or 10 mg/kg of UMM-766 or vehicle by oral gavage starting on Day 1 and continuing through Day 7. Mice were weighed as indicated and observed for signs of disease.The in-life portion of the experiments were 21 days after virus exposure.Pathology studies were conducted using 7-week-old mice (n = 4 per group) and a scheduled time point on Day 6. Mice that survived disease to the endpoint (Day 21) were also used for pathology samples.

Pathology
On Day 6 post-exposure, four mice per group were euthanized and tissues were harvested and submitted for histopathology.Tissues from a subset of animals in each treatment dose group were harvested and submitted for histopathology.Group one animals (controls/saline administration) included 2 males/2 females; group two (1 mg/kg UMM-766 therapeutic) contained 2 males/3 females; group 3 (3 mg/kg UMM-766 therapeutic) contained 4 males/3 females; group 4 (10 mg/kg UMM-766 therapeutic) contained 4 males/4 females.Histopathological [including immunohistochemistry (IHC)] evaluation was performed on a full complement of tissues in the four control animals (group 1) (see Appendix A) to determine which tissues contained significant lesions.Based upon that evaluation, only the following target tissues were evaluated for groups 2-4 (therapeutic treatment groups): nasal cavity/turbinates, olfactory bulb, eye, ear, brain, trachea, esophagus, thymus, thyroid gland, heart, lung, salivary gland, mandibular lymph node, ovary, and uterus.Animals were necropsied, and tissues were fixed by immersion into labeled containers of 10% neutral-buffered formalin.Tissues were then trimmed, processed, embedded in paraffin, cut by microtomy, hemoxylin and eosin (H&E) stained, and screened by a board-certified veterinary pathologist, and coverslips were added.

Immunohistochemistry
Replicate tissue sections were placed on positively charged slides and stained.Immu nohistochemistry on formalin-fixed paraffin-embedded tissue sections was performed using the Dako Envision system (Dako Agilent Pathology Solutions, Carpinteria, CA, USA).After deparaffinization, peroxidase blocking, and antigen retrieval, sections were covered with a rabbit anti-vaccinia virus antibody (#1287, USAMRIID) at a dilution of 1:5,000 and incubated at room temperature for 45 minutes.They were rinsed, and the peroxidase-labeled polymer (secondary antibody) was applied for 30 minutes.Slides were rinsed, and a brown chromogenic substrate 3,3′ Diaminobenzidine (DAB) solution (Dako Agilent Pathology Solutions) was applied for 8 minutes.The substrate-chromogen solution was rinsed off the slides, and slides were counterstained with hematoxylin and rinsed.The sections were dehydrated, cleared with Xyless, and then coverslipped.

Statistics
Mice survival and weight change graphs, normalizations, calculations, and statistics were performed using GraphPad Prism 9. Statistical analysis of survival data was performed by comparing combinations of two groups using the log-rank (Mantel-Cox) test.Corrected baseline values were calculated as the percentage difference relative to Day 0 (preexposure weights) for individual animals (repeated measures).Statistical comparisons between group percentage difference in weights were conducted using a two-way ANOVA with Tukey's multiple comparison test.A full model fit was utilized for the first mouse experiment (semi-lethal), whereas a combination of a full model (weight change on Days 0 through Day 8) and a main effect model (all weight change data captured) was utilized on the second experiment.

In vitro activity
A library of proprietary small molecules provided by Merck, encompassing compounds directed at different essential enzyme classes/mechanisms (i.e., nucleosides, kinases, cysteine proteases, and autophagy modulators) as well as compounds reflecting the diversity of the compound collection, was screened for antiviral activity against VACV in both MRC-5 and RAW264.7 cells.ST-246, a small-molecule inhibitor of envelopment and viral spread, was utilized as a positive control in these assays.Initial screens with VACV identified a subset of molecules with favorable selective indices (SIs).Active compounds were further characterized in assays for additional orthopoxviruses, RPXV and CPXV, in the same cell lines.Among a set of hits identified from the screening library, compound UMM-766 was selected for further progression due to its favorable activity profile against orthopoxviruses as well as an SI in a desirable range.(Table 1; Fig. 1).UMM-766 had an EC 50 value of <1 µM in both cell types against VACV and only in RAW264.7 cells against rabbitpox virus.The selective index ranged from >3.7 to >110 .UMM-766 was relatively less effective against CPXV having EC 50 of 2.17 and 8.09 µM in RAW and MRC5 cells, respectively.Regardless of the virus, there was less antiviral activity when using MRC5 cells having EC 50 and EC 90 values roughly 3-to 15-fold higher relative to RAW cells.For nucleosides, cell-to-cell differences in inhibitory potency may be due to the amount and/or differential rates of uptake (transport) or enzymatic conversion to the active component, the nucleoside triphosphate.Another variable is that MRC-5 cells are highly susceptible to the virus infection, and hence, the assay duration is much shorter as compared with RAW264.7 cells.Overall, ST-246 had the highest SI (Table 1).

In vivo pharmacokinetic and maximum tolerated dose studies
The tolerability of UMM-766 was previously determined in rats and dogs where a dose of 10 mg/kg was the highest tolerable dose.This dose could be considered below the predicted tolerable dose using other models (26).To confirm the tolerability data in mice, BALB/c mice were dosed orally with 1, 3, and 10 mg/kg of UMM-766 once daily for 7 days.Animals were observed for physical appearance and behavioral signs of toxicity for 14 days, as well as survival.Based on these criteria, the drug was well tolerated at all three doses, as none of the animals presented with any adverse physical or behavio ral signs and all animals survived to the endpoint.Pharmacokinetic parameters were measured in C57BL/6 mice wherein clearance (Cl) was measured at 154.5 mL/min/kg, t 1/2 was 2.9 hours, and %F is 48.8.Mouse plasma protein binding was measured at 15.6% with solubility equal to 3.3 mg/mL in PBS (pH 7.4).

In vivo efficacy testing
Next, the post-exposure efficacy of UMM-766 at three doses was evaluated in BALB/c mice exposed to VACV via intranasal instillation.The model was chosen because the route of infection is similar to natural transmission of pertinent orthopoxviruses in humans (VACV, MPXV).Also, information regarding viral lung burden, weight data, and mortality serves as an indicator of the overall severity of disease.To distinguish the ability of the drug to ameliorate different magnitudes of disease, the drug was tested in both severe and sublethal models, one more reflective of the outcome of smallpox in humans.To accomplish this, different age groups of mice were utilized, while still exposing the animals to a similar dose of virus (27).For both experiments, BALB/c mice were exposed to a target dose of 5.5 × 10 5 PFU of VACV via the intranasal route.Starting 1 day after exposure, groups of 10 animals were orally treated with vehicle (PBS), 10 mg/kg, 3 mg/kg, or 1 mg/kg of UMM-766 for 7 days.Animals were monitored for weight and survival and scored for disease progression for 21 days.
In the first experiment, 11-week-old mice were challenged and treated as indicated above.Two animals each from the 1 mg/kg and 3 mg/kg dosing groups and one animal from the vehicle control group died or met euthanasia criteria on Day 7 (Fig. 2A).Five additional animals in the vehicle control group met euthanasia criteria or died over the course of the next 2 days (Days 8 and 9) along with one animal from the 1 mg/kg group.In contrast, no animals were euthanized or died in the 10 mg/kg group.Forty percent of the mice in the vehicle control group, 66.7% of the 1 mg/kg group, 80% of the 3 mg/kg group, and 100% of the 10 mg/kg group survived exposure.Only survival between the 10 mg/kg and the vehicle control group was statistically significant (log-rank; P = 0.0044).(Fig.2C) In terms of changes in weight, vehicle-only infected control animals began losing weight 2 days post-exposure and reached a trough on Day 7 (-25.6%), and weight slowly increased from Day 8 to the end of the study (Fig. 2B).The onset of weight loss was similar between the animals in vehicle control and the 10 mg/kg-treated group; however, the other two treated groups (1 mg/kg and 3 mg/kg) started losing weight on Day 1. Maximum weight changes in the 10 mg/kg-, 3 mg/kg-, and 1 mg/kg-treated groups were -17.3, -24.6, and −28.5 and occurred on Days 7, 7, and 8, respectively.The recovery of the 3-mg/kg and 1-mg/kg animals was slower and less complete, relative to the 10 mg/kg-treated group as well as the vehicle control group.There were statistical differences in weight change between the 10 mg/kg group and both of the other UMM-766-treated groups starting on Day 5 and continuing through (and including) Day 12 (Table S1).There was also a statistical difference between the 10 mg/kg and low-dosed group (1 mg/kg) weight changes on Day 15 (Table S1).The 10 mg/kg-dosed group weight changes were also significantly different from those of the vehicle control group starting on Day 6 through and including Day 11.There was also a significant difference in weight changes between the 1 mg/kg and the vehicle control groups on Day 11.Other than a single statistical difference on Day 11 between the 1 mg/kg and the vehicle control group, there were no additional statistical differences between the control group weight changes, 1 mg/kg group, and/or 3 mg/kg groups on any days examined (Table S1).We conclude that only the oral 10 mg/kg dose of UMM-766 protected mice against disease in the semi-lethal murine model.
To assess UMM-766-mediated protection in a more aggressive and lethal model, mice were exposed to the pathogen at approximately 7 weeks of age.We also infected an additional set of four mice per group to allow for pathological assessment at a critical disease timepoint (Day 6).This day was chosen because animals typically succumb to disease or meet euthanasia criteria on Days 5-8 in this model (Fig. S1).In terms of survival, all animals in the vehicle control group met euthanasia criteria or died (median survival of 7 days), as opposed to only two animals treated with 10 mg/kg of UMM-766 (P = 0.0001) (Fig. 2D and F).Only 10% of the mice survived in the 1 mg/kg group (median survival of 8 days) and 30% in the 3 mg/kg group (median survival of 8 days).Of the two lower dose groups, only the 3 mg/kg group had a significant difference in survival relative to the vehicle control group (P = 0.0048).Animals treated with 10 mg/kg yielded survival curves that were also significantly different from the 3 mg/kg (P = 0.0258) and the 1 mg/kg group (P = 0.0015) (Fig. 2F).From these data, we conclude that both 10 mg/kg and 3 mg/kg doses of UMM-766 provide efficacy in terms of survival but the 10 mg/kg treatment provides the superior survival advantage.
All four groups of animals had decreased weights (and corresponding percentages) up to and including Day 8 (Fig. 2E).The vehicle control group had the greatest maximum percentage weight change (−31.1%),followed by the 1 mg/kg group (−30.5%), 3 mg/kg group (−29.3%), and 10 mg/kg (−24.4%).There were statistical differences between the 10 mg/kg group and the vehicle control group (Days 6-8), the 3 mg/kg group (Days 6-8), and the 1 mg/kg group (Days 7 and 8).From Days 0 to 8, the 1 mg/kg and 3 mg/kg groups were not significantly different from the vehicle control group or each other.Since all animals in the vehicle control group were euthanized or succumbed to disease on Day 8, subsequent statistical comparisons of the UMM-766-treated groups with the vehicle were not possible.Therefore, a main effects model was used to compare the four groups which included all the captured, normalized weight data.Again, the 10 mg/kg group was significantly different than the vehicle group (P < 0.0001), the 1 mg/kg group (P = 0.0016), and the 3 mg/kg group (P < 0.0001).From these weight data, we conclude that only the 10 mg/kg oral dosing is efficaciously relative to the vehicle and the two lower dose groups.

Respiratory pathology
The ability of UMM-766 to mitigate tissue pathology in a subset of 7-week-old mice (n = 4 per group) on Day 6 post-infection was assessed.Additionally, the histopathology in mice surviving to Day 21 was examined.(Table 2).Histopathological (H&E staining and IHC) evaluation was performed on a full complement of tissues in the four vehicle control animals to determine which tissues contained significant lesions, and based on these findings, a down-selected set of tissues was evaluated for the groups of UMM-766 treated animals.To assess UMM-766-mediated protection against tissue injury and viral burden, nasal cavity/turbinates, olfactory bulb, eye, ear, brain, trachea, esophagus, thymus, thyroid gland, heart, lung, salivary gland, mandibular lymph node, ovary, and uterus were examined in animals harvested on Day 6 and in animals surviving to Day 21.A cursory graphical view of both H&E and IHC is provided to highlight differences between groups (Fig. 3).
The most significant VACV-induced lesions were present in the nasal cavity, which included lytic necrosis of the olfactory epithelium and, to a lesser extent, the respiratory epithelium.The necrosis extended below the epithelium and was present within the lamina propria/submucosa, surrounded olfactory nerve fibers, included central nasal glands, and, in some cases, included peripheral Steno's glands surrounding the maxillary sinus (Fig. 3).Necrosis of the olfactory epithelium ranged from moderate to severe in animals from the vehicle control group and was generally less severe in the animals from the UMM-766-treated group.On Day 6, animals treated with 10 mg/kg of UMM-766 had the least severe degree of necrosis, with two animals having minimal necrosis and two with moderate necrosis.This trend was also seen for lesions in the respiratory epithelium (degeneration and necrosis) where the severity in UMM-766-treated mice was generally less in Day 6 animals than that seen in the vehicle control group.No vehicle control animal survived to Day 21; but in the surviving UMM-766-treated animals, necrosis was reduced compared with those of Day 6 animals.
Inflammation within the tissue (as opposed to nasal cavity) was predominantly within the lamina propria and submucosa.Inflammation ranged from minimal to moderate across all groups and was absent in a subset of UMM-766-treated animals on Day 6 but was present in all vehicle-treated animals.
Bright eosinophilic, round to oval, cytoplasmic globules ranging in size from 2 to 4 µm were identified within certain areas of the nasal tissue in low numbers, most commonly in the lamina propria or nerve fiber layer, including mesenchymal support cells surround ing olfactory nerve fibers, fibroblasts, and periosteal cells.In some instances, they were also seen within the olfactory and respiratory epithelia (Fig. 4; Fig. S2 and S3).Within the nerve fiber layer, these inclusions were mostly in areas of necrosis and abundant labeling by IHC.However, not all these inclusions contained viral protein.Inclusions were more readily identified in the lamina propria/nerve fiber layer than in the epithelium and were observed in all vehicle control animals in that region, whereas they were rarely identified in drug-treated mice.
IHC labeling in the nasal cavity/turbinates was generally present near areas of lytic necrosis in all groups (Fig. 3 and 4).The severity of labeling (including the extent of tissues and cell types labeled and overall degree of labeling) was generally highest in vehicle-only-treated mice and markedly reduced in animals treated with 10 mg/kg of UMM-766 (Fig. 4).Cell types identified with viral IHC labeling in the vehicle control animals include the following: the olfactory and respiratory epithelia, generalized labeling in the lamina propria/submucosa including the nerve fiber layer and  periphery of skeletal muscle and rarely extending into skeletal muscle.In at least one animal from the vehicle control group, viral protein labeling was identified in the subepithelium underlying the oral mucosa and in periodontal tissue.Labeling was less extensive and overall severity was less in animals treated with UMM-766, corresponding to the lesser degree of necrosis, and was most commonly seen in the olfactory epithe lium and adjacent submucosa/lamina propria, surrounding olfactory nerve fiber bundles, and in nasal/Steno's glands as well as the adjacent periosteal tissue of the turbinate bones.Among the Day 6 animals, the severity of labeling in the 10 mg/kg group was slightly less severe compared with those in the other groups.Day 21 animals in the 1 mg/kg and 3 mg/kg groups had minimal or absent labeling, and labeling was absent for all Day 21 animals in the 10 mg/kg group.Within the lung, the most significant lesions included necrosis of alveolar septa, bronchiole epithelium (either segmentally or diffusely), and adjacent blood vessels (Fig. 4).A distinct difference in the frequency or severity of necrosis between the vehicle control group and the UMM-766-treated groups is not identified in the lesion scoring data.
IHC labeling in the lung generally tracked very closely with the presence of necrosis, similar to the pattern seen in the nasal cavity.This included strong positive labeling in the necrotic bronchioles (epithelium and surrounding interstitium), vessels (including endothelium/interna, tunica media smooth muscle and in some cases the outer tunica adventitia), and alveoli (septa) (Fig. 5).Labeling was also multifocally present in alveolar septa that appeared less affected (i.e., necrosis was less histologically apparent) but had edema in the adjacent alveolar lumen, suggesting some degree of damage to the septa and associated capillaries.The severity/degree of IHC labeling in the lung of the vehicle control group was greater (moderate to marked) than that seen in other groups (mostly mild).Day 21 animals either had minimal or absent labeling and when present was in areas of histiocytic infiltrate and had a very faint labeling pattern.Overall, 10 mg/kg of UMM-766 resulted in a decreased respiratory injury compared with the vehicle control.

Nervous system
The most significant nervous system lesions were present in the olfactory bulb and included meningeal necrosis which was characterized by multifocal or focally extensive lytic necrosis in the meninges, often involving small vessels, in the ventral or lateral meninges (rarely extending dorsally) (Fig. 3; Fig. S2).Another common lesion in the olfactory bulb was hemorrhage in the olfactory nerve layer (outermost functional layer) of the olfactory bulb, likely resulting from damage to small vessels or hemodynamic changes (Fig. S2).Minimal to mild mononuclear inflammation was also seen in the meninges in several animals.The necrosis and hemorrhage were seen in all vehicle controls (consolidated in Fig. 3).For animals collected on Day 6, necrosis occurred in all vehicle control animals and 1 mg/kg UMM-766-treated animals but was absent in two 3-mg/kg animals and two 10-mg/kg animals.Hemorrhage was seen in all vehicle control animals but was seen less frequently and was of less severity in 1 mg/kg and 3 mg/kg UMM-766-treated animals.Hemorrhage was completely absent in animals receiving 10 mg/kg UMM-766.Meningeal necrosis and hemorrhage in the olfactory bulb were not present in Day 21-harvested animals, but inflammation was present in the 1 mg/kg-and 3 mg/kg-treated animals.
Inflammation in the meninges of the olfactory bulb was minimal to mild in all cases and occurred more commonly in UMM-766-treated animals than in the vehicle control.Viral protein labeling by IHC was minimal to mild and included labeling in the olfactory nerve layer and/or meninges of the olfactory bulb.The pattern of labeling in these regions was often extensive and appeared confluent with the labeling in the submucosa/lamina propria underlying the dorsal most olfactory epithelium of the nasal cavity (separated by only a thin layer of bone).Labeling appears to surround nerve fiber bundles of the olfactory nerve layer, extending to the periosteum surrounding the olfactory bulb.Labeling generally corresponded to areas of necrosis, inflammation, and Lesions in the brain were rare and included minimal mononuclear inflammation in the meninges of one vehicle control animal and minimal necrosis in the meninges of one 1 mg/kg-treated Day 6 animal.

DISCUSSION
The 2022 global MPXV outbreak demonstrated that orthopoxviruses remain a threat to human health.Other orthopoxviruses, such as cowpox virus, also continue to cause sporadic human disease (28).Because locations of these outbreaks are unpredictable, the use of a widespread vaccination campaign, such as the one used to eradicate smallpox, is unlikely to occur.Rather, the targeted use of post-exposure therapeutics, accompa nied by limited vaccination of close contacts, is the most likely mitigation strategy.Several small-molecule therapeutics were developed and stockpiled in response to the biological weapons threat imposed by variola virus or other orthopoxviruses.ST-246 and brincidofovir are essential to mitigating disease caused by orthopoxviruses.However, there is a risk that resistant strains of orthopoxvirus could emerge especially in individu als undergoing prolonged treatment.Thus, the continued development of anti-poxvirus compounds is needed.
This study provides a template for high-throughput screening of drug libraries to identify compounds with efficacy against orthopoxviruses.Through antiviral screening of over 12,000 compounds, UMM-766 was rapidly discovered to have in vitro and in vivo anti-orthopoxvirus activity.More specifically, the activity of these compounds was rapidly evaluated for efficacy using at least three orthopoxviruses in multiple cell lines to down-select candidates before testing in animal models.The oral availability and tolerability of UMM-766 was previously tested in rats (23) and confirmed in our studies to be tolerable in Balb/C mice.The impact of UMM-766 on weight was not assessed in our tolerability screens nor during in vivo efficacy testing.Therefore, it is not possible to state what, if any, impact the drug alone had on weight changes observed in the infected mice.
UMM-766 provided complete protection from lethality at 10 mg/kg administered orally post-exposure and limited the extent of tissue injury.The pathology of vehicletreated animals showed that the nasal turbinate and lungs were the most affected by viral infection, consistent with findings from others (29).Six days after infection, IHC scores, and to a lesser extent H&E scores, were lower in treated animals relative to control counterparts (Fig. 3), suggesting UMM-766 can decrease viral load in tissues and the ensuing tissue pathology.Other tissues analyzed had a similar trend, but the decreased incidence of lesions in these tissues precludes an overall conclusion based on those tissues alone.
Eosinophilic, intracytoplasmic globules were observed within the cytoplasm of either bronchiole and/or tracheal epithelium (but more commonly in the lamina propria or nerve fiber) in low numbers in all the vehicle control mice and animals treated with either 1 or 3 mg/kg of UMM-766 (Fig. S2 and S3).The presence of what are interpreted as inclusions did not consistently correspond to viral labeling by IHC in the affected cells.The inclusions do closely resemble poxvirus viral inclusion bodies previously described for cowpox virus and found in the same region of the lung (30).Cowpox virus can form A-type inclusions (ATI) that embed mature virions, but the viral ATI protein responsible for the inclusion formation is truncated in the WR strain of vaccinia virus (31,32).Without additional diagnostics (i.e., inclusion-specific IHC or EM), it is not possible to say with certainty whether the inclusion bodies occlude virions, only comprised of viral proteins, or are strictly a cellular phenomena induced by viral insult (32,33).If virally induced, the bodies become more interesting as they can be found in tissues from recovered animals on Day 21 and may be used as a marker of disease in this model.
An important factor to consider is whether the mechanism of UMM-766 action is distinct from brincidofovir.In the development of novel orthopoxvirus-targeting drugs, it may be critical to identify those compounds with distinct and non-redundant mecha nisms of action to help prevent the emergence of escape mutants.Future efforts will focus on determining the mechanism of action of UMM-766 against poxvirus; however, the compound was previously determined to decrease NS5B-mediated replication of hepatitis C virus RNA by chain termination (23).Based on the ability of UMM-766 to interfere with RNA polymerization, the DNA-dependent RNA polymerase expressed by poxviruses will be explored as a potential target for UMM-766.Future studies will determine if the compound inhibits poxviruses by a mechanism independent of brincidofovir and help rationalize UMM-766 (or analogs) for advancement.Our work supports the screening of existing drug libraries to identify compounds that will enhance the toolbox of compounds capable of inhibiting current and emerging orthopoxviruses.

FIG 2
FIG 2 Efficacy of UMM-766 in a lethal and semi-lethal exposure of VACV in mice.BALB/c mice were exposed to an intranasal challenge of 5.5 × 10 5 pfu VACV and treated for 7 days starting on Day 1 (A-F).The percent survival (A and D ) and changes in weight (B and E) were plotted over time (days post exposure) using GraphPad Prism.To decrease the severity of disease, older (approximately 11-to 12-week-old) mice were utilized and are shown in panels A and B, whereas a more severe disease model using animals that were approximately 7 weeks of age is shown in panels C and D. The tables show Pvalues derived from log-rank Mantel-Cox (GraphPad Prism) tests between groups for sublethal (C) and lethal (F) challenge experiments.Bold text indicates significant differences.For weights, data were baseline corrected as a percentage difference and presented in GraphPad Prism.Standard errors of the means are shown.

FIG 3
FIG 3 Summary of histopathology for UMM-766-treated or vehicle control mice exposed to vaccinia virus.Tissues were harvested from infected 7-week-old animals (n = 4) euthanized on Day 6 (A, C, E) post-exposure and animals that survived until Day 21 (B, D).Cumulative pathology scores are based on the extent and intensity of the lesion in the affected tissue by examination by H&E staining (A and B).Lesions included in the score include inflammation, edema, hyperplasia, infiltrates, and occlusions.Scores were added for each group (A) or averaged (B) per group.The presence of poxvirus protein by IHC is also presented (C, D, and E).Again, the magnitude and extent of staining was scored for each tissue/area.More common findings (C) and less common findings (E) are given for animals euthanized on Day 6, as well as findings for animals surviving to Day 21 (D).Mean with standard deviation are shown (bars) with individual animal data (points) in C, D, and E.

FIG 4
FIG 4 Histopathologic examination of the nasal cavity in mice exposed to VACV and orally treated with vehicle control or UMM-766.Images from the nasal cavity of a representative vehicle control animal (left column), compared to an animal treated with 10 mg/kg UMM-766 (right column).Black boxes in images A, B, C, and D (4× magnification) correspond to images E, F, G, and H (20× magnification), respectively.H&E of dorsal-caudal nasal cavity and ventral olfactory bulb (A and B).IHC of identical region in A and B (C and D): H&E corresponding to boxes in A and B (E and F) and IHC corresponding to boxes in C and D (G and H).

TABLE 2
Summary of animals for pathology