Inhibitors of dihydroorotate dehydrogenase cooperate with molnupiravir and N4-hydroxycytidine to suppress SARS-CoV-2 replication

Summary The nucleoside analog N4-hydroxycytidine (NHC) is the active metabolite of the prodrug molnupiravir, which has been approved for the treatment of COVID-19. SARS-CoV-2 incorporates NHC into its RNA, resulting in defective virus genomes. Likewise, inhibitors of dihydroorotate dehydrogenase (DHODH) reduce virus yield upon infection, by suppressing the cellular synthesis of pyrimidines. Here, we show that NHC and DHODH inhibitors strongly synergize in the inhibition of SARS-CoV-2 replication in vitro. We propose that the lack of available pyrimidine nucleotides upon DHODH inhibition increases the incorporation of NHC into nascent viral RNA. This concept is supported by the rescue of virus replication upon addition of pyrimidine nucleosides to the media. DHODH inhibitors increased the antiviral efficiency of molnupiravir not only in organoids of human lung, but also in Syrian Gold hamsters and in K18-hACE2 mice. Combining molnupiravir with DHODH inhibitors may thus improve available therapy options for COVID-19.


Molnupiravir and DHODH inhibitors are approved drugs, facilitating clinical testing
The combination may allow lower drug doses to decrease possible toxic effects Inhibitors of nucleotide biosynthesis may boost antiviral nucleoside analogs

INTRODUCTION
During the combat of the COVID-19 pandemic, a number of vaccine approaches have been established successfully, while efficient therapeutics are still urgently needed (Doherty, 2021). Clinically evaluated therapies include the use of steroids (Horby et al., 2021;Tomazini et al., 2020), the protease inhibitor nirmatrelvir that is contained in Paxlovid (Wang and Yang, 2021), and the nucleoside analog remdesivir (Beigel et al., 2020;Goldman et al., 2020). Upon triphosphorylation at their 5 0 positions, antiviral nucleoside analogues antagonize virus propagation, by interfering with the activity of the viral RNA-dependent RNA polymerase and/or by compromising the function of the newly synthesized viral genomes through mutagenesis (Pruijssers and Denison, 2019).
Besides remdesivir, other nucleoside analogs showed promising antiviral effects against SARS-CoV-2. Most notably, molnupiravir, also known as EIDD-2801 or MK-4482, is the prodrug of N4-hydroxycytidine (NHC), or EIDD-1931(Cox et al., 2021Painter et al., 2021;Sheahan et al., 2020;Wahl et al., 2020Wahl et al., , 2021. In comparison to cytidine, NHC has the same structure but carries a hydroxylated amino group (nitrogen 4) at the pyrimidine base. This does not impair the incorporation of triphosphorylated NHC into nascent RNA by the viral RNA-dependent RNA polymerase (Kabinger et al., 2021). However, owing to a tautomeric interconversion within the NHC base, the incorporation of NHC into virus RNA leads to erroneous RNA replication (Jena, 2020). NHC can base pair with guanosine, but also with adenosine, thus leading to multiple errors in the subsequently synthesized viral RNA genomes and resulting in replication-deficient virus particles. Molnupiravir is active against SARS-CoV-2 replication in vitro and in vivo (Rosenke et al., 2021;Sheahan et al., 2020;Wahl et al., 2021), and this includes the recently emerged Omicron variant (Prince et al., 2021;Vangeel et al., 2021). It also prevents SARS-CoV-2 transmission in the ferret model (Cox et al., 2021), and it was found clinically effective in large-scale clinical trials (Jayk Bernal et al., 2021;Painter et al., 2021), exemplified by the trials registered at clinicaltrials.gov with the numbers NCT04575584, NCT04575597 and NCT04405739, leading to approval in the UK. However, although the Besides immunosuppression and direct interference with virus replication, an alternative approach of treatment against SARS-CoV-2 aims at reducing the cellular synthesis of nucleotides, thereby indirectly impairing the synthesis of viral RNA. We (Stegmann et al., 2021a) and others (Caruso et al., 2021;Zhang et al., 2021) have previously reported the high demand on cellular nucleotide biosynthesis during SARS-CoV-2 infection, resulting in an antiviral effect of folate antagonists, which impair purine synthesis. Moreover, in the context of nucleotide biosynthesis, the inhibition of dihydroorotate dehydrogenase (DHODH) represents an attractive strategy to antagonize SARS-CoV-2 replication. DHODH catalyzes a key step during pyrimidine synthesis. Unlike all cytosolic enzymes involved in this metabolic pathway, DHODH localizes to the inner mitochondrial membrane, where it transfers reduction equivalents from dihydroorotate to ubiquinone moieties of the respiration chain. As a result, orotate becomes available for the subsequent synthesis steps to obtain uridine monophosphate and later cytidine triphosphate. A number of DHODH inhibitors have become available for clinical testing or were even approved for therapy (Munier-Lehmann et al., 2013), mostly to treat autoimmune diseases, due to their selective inhibition of hyperactive immune cells. Recently, however, some DHODH inhibitors were successfully tested with regard to their efficacy in preventing the replication of viruses (Hoffmann et al., 2011;Zhang et al., 2012), including SARS-CoV-2 (Calistri et al., 2021;Hahn et al., 2020;Luban et al., 2021;Xiong et al., 2020). One DHODH inhibitor, IMU-838 (Vidofludimus calcium), was further clinically evaluated for COVID-19 therapy in hospitalized patients and was found effective according to secondary criteria, e.g. time to clinical improvement or viral burden (CALVID-1, trial identifier NCT04379271; NCT04516915).
We hypothesized that the suppression of pyrimidine synthesis should increase the ratio of NHC triphosphate versus cytidine triphosphate in infected cells, thus enhancing the incorporation of NHC into the viral RNA and resulting in the production of replication-deficient viral particles. Preliminary results that we (Stegmann et al., 2021b) and others (Schultz et al., 2022) presented on a pre-publication server revealed first in vitro-evidence of drug synergism between NHC and DHODH inhibitors, supporting our hypothesis and encouraging its further evaluation. We now found that the combination of NHC and DHODH inhibitors resulted in profoundly synergistic suppression of SARS-CoV-2 replication in vitro; DHODH inhibition also improved the performance of molnupiravir in two animal model systems, thus presenting a potential treatment strategy.

RESULTS
NHC and DHODH inhibitors cooperate to interfere with SARS-CoV-2 replication in cultured cells, without signs of cytotoxicity The biosynthesis of pyrimidines is crucial for RNA replication ( Figure 1A). The enzyme dihydroorotate dehydrogenase (DHODH) catalyzes the oxidation of dihydroorotate to orotate, which is a precursor of cytidine triphosphate (CTP). In the presence of N4-hydroxycytidine (NHC), its active metabolite NHCTP competes with CTP for incorporation into nascent RNA (Gordon et al., 2021). We hypothesized that the suppression of cellular CTP synthesis by DHODH inhibitors will favor the incorporation of NHCTP into newly synthesized SARS-CoV-2 RNA, and will thus potentiate the antiviral efficacy of NHC. To test this, we combined both drugs for treatment of Vero E6 cells prior to infection with SARS-CoV-2. We applied NHC and the DHODH inhibitors BAY2402234, teriflunomide and IMU-838, at concentrations that only moderately suppressed virus replication as single treatments. Accordingly, neither NHC nor DHODH inhibitors alone grossly affected the development of a cytopathic effect (CPE) caused by SARS-CoV-2. Strikingly, however, the combination of NHC and DHODH inhibitors was far more efficient in preventing CPE ( Figure 1B), and it reduced virus yield more than 1000-fold, as determined by the median tissue culture infectious dose (TCID 50 /mL) of the supernatant ( Figure 1C). Combining the drugs did not produce morphologic signs of cytotoxicity in non-infected cells ( Figure 1B) and did not grossly augment the release of lactate dehydrogenase (LDH) into the culture supernatant (Figures 1D, S1A, and S1B) or reduce cell viability ( Figure S1C). Hence, the drug combination interferes with CPE and virus yield, without displaying any detectable cytotoxic effects.
Combinations of NHC and DHODH inhibitors synergistically reduce viral RNA yield upon infection with SARS-CoV-2 Figure 1. The combination of N4-hydroxycytidine (NHC) and inhibitors of dihydroorotate dehydrogenase (DHODH) strongly impairs SARS-CoV-2 replication without detectable cytotoxicity (A) Mechanistic concept for the synergistic inhibitory effect of DHODH inhibitors and N4-hydroxycytidine (NHC) on SARS-CoV-2 RNA replication. Biosynthesis of pyrimidines starts with carbamoyl phosphate and aspartate to form dihydroorotate. Dihydroorotate is further oxidized to orotate by dihydroorotate dehydrogenase (DHODH) and later converted to uridine triphosphate (UTP) and cytidine triphosphate (CTP). Molnupiravir is the prodrug of NHC, which is ll OPEN ACCESS iScience 25, 104293, May 20, 2022 3 iScience Article we determined their impact on the release of viral RNA and calculated a synergy score using the Bliss independence model ( Figure 2A). IMU-838 and NHC were added 24 h before infection (Figure 2A), at the time of infection ( Figures S2A and S2B), or 4 h after infecting the cells ( Figures 2B and S2C). The Bliss score revealed strong synergy for a subset of the drug combinations ( Figure 2A). Moreover, we combined each of the DHODH inhibitors IMU-838, BAY2402234, teriflunomide, ASLAN003, and brequinar with NHC and quantified the amount of viral RNA released into the cell culture supernatant ( Figure 2C). Strikingly, the combination treatment diminished SARS-CoV-2 RNA progeny up to 400-fold as compared to single drug treatment, and up to 1000-fold as compared to untreated controls, and the indices reflected profound synergy of the drugs as determined by the Bliss independence model. This effect was not only seen in Vero E6 cells but also in Calu-3 cells ( Figure 2D), a human lung cancer cell line used to model bronchial epithelia (Kreft et al., 2015) and susceptible to infection with SARS-CoV-2 (Saccon et al., 2021). Hence, NHC and DHODH inhibitors synergistically antagonize SARS-CoV-2 replication in two different cell lines.

Distinctly reduced accumulation of viral proteins in SARS-CoV-2-infected cells upon treatment with NHC and DHODH inhibitors
Besides infectious units and virus RNA, the synthesis of virus proteins in infected cells is an important readout for virus propagation. Therefore, we detected viral proteins upon treatment of Vero E6 cells with NHC and/or BAY2402234, teriflunomide or IMU-838, and infection with SARS-CoV-2. The frequency at which we detected the viral spike protein and the nucleoprotein in the cells was severely reduced upon the combined treatment, but far less by the single treatments, as determined by immunofluorescence microscopy ( Figures 3A-3C). Correspondingly, immunoblot analyses revealed that the levels of both viral proteins were strongly reduced by the drug combination ( Figures 3D-3F), whereas the single drugs at the same concentrations were much less efficient. The blots also revealed that DHODH levels were not detectably affected by the drugs. Taken together, the detection of virus proteins further confirmed that the drug combination suppresses virus propagation to a far higher degree than the single drugs in cultured cells.
The combination of NHC and DHODH inhibitors also reduces the replication of the SARS-CoV-2 variants B.1.1.7/Alpha, B.1.351/Beta, and B.1.617.2/Delta As the pandemic proceeded, new variants emerged, with potentially higher infectivity and immune-escape properties (Tegally et al., 2021;Wang et al., 2021a). To ensure the suitability of the proposed drug combination against these variants, we assessed their replication in the presence of NHC and/or DHODH inhibitors. The variants of concern Alpha (B.

Uridine and cytidine rescue virus replication in the presence of NHC and DHODH inhibitors
According to our initial considerations ( Figure 1A), we suspected that DHODH inhibitors potentiate the efficacy of NHC by suppressing the levels of endogenous pyrimidine nucleotides. We now performed rescue experiments to support this model. To elucidate the mechanism of interference with SARS-CoV-2 replication by the drug combinations, we added pyrimidine nucleosides to the culture media. We supplied uridine Figure 1. Continued further converted to the corresponding triphosphate (NHCTP), which competes with CTP for incorporation into nascent virus RNA. The suppression of CTP synthesis by inhibitors of DHODH is expected to enhance the incorporation of NHCTP into the viral RNA, causing false incorporation of nucleotides in subsequent rounds of replication. (B) Reduced cytopathic effect (CPE) by NHC and DHODH inhibitors. Vero E6 cells were treated with drugs or the DMSO control for 24 h, inoculated with SARS-CoV-2 (strain GOE_001), and further incubated in the presence of the same drugs for 48 h. Cell morphology was assessed by phase contrast microscopy. Note that the CPE was readily visible in virus-infected cells, in DMSO-treated cells and also when cells had been treated with either drug alone. However, the CPE was observed only to a far lesser extent when the cells had been incubated with both NHC and DHODH inhibitors. Bar, 100 mm.
(C) Reduction of the median tissue culture infectious dose (TCID 50 ) by the combination of NHC and the DHODH inhibitor IMU-838. Vero E6 cells were treated with NHC, IMU-838, or the combination of both compounds for 24 h before infection, and then throughout the time of infection. Cells were infected with SARS-CoV-2, strain hCoV-19/Germany/BY-Bochum-1/2020 (MOI 0.1), and further incubated for 48 h. The supernatant was titrated to determine the TCID 50 /mL (mean with SD, n = 3; logarithmic scale).
(D) Lack of measurable cytotoxicity by NHC and DHODH inhibitors. Vero E6 cells were treated with NHC and/or IMU-838, BAY2402234, and teriflunomide at the indicated concentrations for 72 h. The release of lactate dehydrogenase (LDH) to the supernatant was quantified by bioluminescence as a readout for cytotoxicity. The percentages reflect the proportion of LDH released to the media, compared to the overall amount of LDH in the cells (LDH control) (mean with SD, n = 3).

ll
OPEN ACCESS (Figures 5A and S4A) or cytidine (Figures 5B and S4B) to Vero E6 cells along with the DHODH inhibitors IMU-838, BAY2402234, or teriflunomide, combined with NHC. The addition of 5 or 10 mM uridine, or the same concentration of cytidine, prevented the inhibition of SARS-CoV-2 replication by the combination treatment. The results are in accordance with the reduced levels of UTP and CTP upon DHODH inhibition that were reported previously . This strongly suggests that the synergism of NHC with DHODH inhibitors can be explained by competition of NHC with endogenous pyrimidine nucleosides for incorporation into nascent viral RNA, as we had hypothesized initially ( Figure 1A).

DHODH inhibitors and NHC cooperate to reduce SARS-CoV-2 replication in a human lung organoid model
So far, we had performed all experiments in cultured cell lines. The limitation of such experiments consists in their dissimilarity to bronchial and lung epithelia, i.e. the primary sites for SARS-CoV-2 in humans. For a model closer to these primary infection sites, we infected lung organoids derived from human-induced pluripotent stem cells (iPSCs) upon treatment with single and combined drugs. Here again, virus replication was strongly reduced by the combination of both drugs 24, 48, and 72 h.p.i ( Figure 6A), with tolerable cytotoxicity ( Figure 6B). In parallel, the rate of cells that contained double-stranded RNA derived from the virus was drastically reduced by the drug combination ( Figure 6C), further arguing that NHC along with DHODH inhibitors diminishes virus replication in an in vitro model of human lung tissue. Notably, however, the effects did not reach the levels of statistically significant synergy, perhaps due to the higher variations in virus yield when using primary organoids rather than cell lines for infection assays.

The drug combination ameliorates COVID-19 in a hamster model
Syrian Gold hamsters represent a well-established and acknowledged animal model of human SARS-CoV-2 infections. To evaluate the therapeutic effect of the drug combination in this system, we infected Syrian Gold hamsters with SARS-CoV-2 while treating them orally with molnupiravir (the prodrug of NHC) and/or the DHODH inhibitor teriflunomide ( Figure 7A). Teriflunomide was used in the in vivo models because it is more effective on rodent DHODH compared to e.g. IMU-838, which is more potent on human DHODH (Muehler et al., 2020b). Similarly, the pharmacokinetics (PK) of teriflunomide was favorable in hamsters and mice according to our analyses ( Figure S5A), whereas the PK of IMU-838 was previously characterized in humans (Muehler et al., 2020a). Upon drug treatment, the hamsters had reduced virus titers in nasal washes, particularly with the combined therapy ( Figure 7B). Infectious virus was no longer detectable in any of the animals at the end of the experiment 7 days p.i ( Figure S5B). The weight of the animals typically drops with the progression of the disease (Francis et al., 2021;Imai et al., 2020); using this readout, we observed that the course of COVID-19 was milder when treating the animals with the drugs under study. Of note, the drug combination was superior to the single drugs in the course of infection, as revealed by the weight of the animals at days 3 through 5 post infection ( Figures 7C and S5C). Determination of lung pathology (Figures S6A-S6C) as well as daily energy expenditure (Figures S6D) supported these findings. In conclusion, combining molnupiravir and a DHODH inhibitor proved effective in the hamster model of COVID-19. However, the degree of drug synergy was not comparable to the in vitro studies using cultured cells. We speculate that the activity of the DHODH inhibitor might be compromised somewhat by the uridine present in the serum of animals, in agreement with our results shown in Figure 5. . Strong synergism of NHC and DHODH inhibitors to diminish the release of SARS-CoV-2 RNA from cultured cells (A) Reduced release of viral RNA upon combined treatment with NHC and IMU-838. Vero E6 cells were treated with NHC and/or IMU-838, and infected as in Figure 1. A sample of the inoculum was preserved for RNA preparation. At 48 h post infection (p.i.), RNA was isolated from the cell supernatants, followed by quantitative RT-PCR to detect viral RNA and determine the amount of SARS-CoV-2 RNA copies per mL (mean, n = 3). The synergy score was calculated using the Bliss independence model. Data are presented as mean G SEM. A Bliss score >10 is generally considered to reveal strong drug synergism. (B) Diminished virus RNA progeny by NHC and DHODH inhibitors even when added 4 h after SARS-CoV-2 infection. Vero E6 cells were infected as described in Figure 1 and treated with NHC and/or DHODH inhibitors at 4 h post infection (p.i.). RNA was isolated from the cell supernatants, and SARS-CoV-2 RNA was quantified by qRT-PCR. The amount of RNA found upon infection without drug treatment was defined as 100%, and the other RNA quantities were normalized accordingly. RNA was also isolated from the virus inoculum used to infect the cells. Note that the combination treatment reduced SARS-CoV-2 replication to a greater extent compared to single drug treatments even when applied 4 h p.i. (mean with SD, n = 3). For p values, see Figure S2C.
(C) Reduced virus RNA progeny in the presence of NHC and various DHODH inhibitors. Vero E6 cells were treated with drugs and/or infected as in Figure 1, followed by quantitative detection of SARS-CoV-2 RNA. The drug combinations were found capable of reducing virus RNA yield by more than 100-fold as compared to single drug treatments (mean with SD, n = 3).
(D) In Calu-3 cells, too, the combination of NHC and the DHODH inhibitors IMU-838, BAY2402234, or teriflunomide strongly reduced the amount of viral RNA released to the supernatant (mean with SD, n = 3).  , 2007;Oladunni et al., 2020), as we described previously (Peter et al., 2021). Before and during infection, mice were treated orally with the DHODH inhibitor teriflunomide and/or molnupiravir, with a similar schedule as in the hamster experiments ( Figure 8A). Disease progression in this model was fast, leaving little change in animal weight ( Figure S7A) and requiring termination of the experiment at day 4 p.i.. Notably, the combination of both drugs reduced the virus titer in the lungs of the animals, by a factor of up to 96-fold ( Figures 8B and S7B). Moreover, lymphocyte infiltrations in the  iScience Article lungs of these animals were reduced accordingly ( Figures 8C, 8D, and S7C). This transgenic mouse model involves virus-induced encephalitis (McCray et al., 2007;Sun et al., 2020). This was reflected by substantial virus load in the brain of infected mice. Interestingly, each of the drugs suppressed detectable virus in the  Figure 1. On top of the drugs, where indicated, uridine was added to the cell culture media, at concentrations of 2, 5, or 10 mM. SARS-CoV-2 propagation was still diminished by NHC and DHODH inhibitors despite 2 mM uridine levels, but rescued in the presence of 5 or 10 mM uridine (mean with SD, n = 3), in agreement with the mechanism outlined in Figure 1A. For p values, see Figure S4A.
(B) Restored SARS-CoV-2 replication by cytidine, in the presence of NHC and DHODH inhibitors. The experiment was carried out as in (A), with the addition of cytidine instead of uridine. 5 or 10 mM cytidine restored virus replication in the presence of the drugs (mean with SD, n = 3), further confirming the mechanism outlined in Figure 1A. For p values, see Figure S4B.

OPEN ACCESS
iScience 25, 104293, May 20, 2022 9 iScience Article Figure 6. Reduced SARS-CoV-2 propagation and dsRNA formation by NHC and BAY2402234 in human lung organoids (A) Reduced TCID 50 by NHC and the DHODH inhibitor BAY2402234. Human stem cell-derived lung organoids were sliced and treated with 1 mM NHC and/or 1 mM BAY2402234 for 24 h before and then throughout the time of infection. Organoid slices were infected with 35,000 PFU per well and further incubated for 24, 48, or 72 h. The supernatant was titrated to determine the TCID 50 /mL (mean with SD, n = 6). Statistical evaluation was performed using the Mann-Whitney U test.
(B) Cell viability of lung organoids was not detectably affected by NHC and/or the DHODH inhibitor BAY2402234. The release of lactate dehydrogenase (LDH) to the supernatant was quantified by bioluminescence as a readout for cytotoxicity and cell viability as in Figure 1D (mean with SD, n = 6). iScience Article brains of the animals ( Figure S7D). Of note, the synergy of the drugs found in vitro was not recapitulated in a statistically significant fashion in the mice, neither in the lungs nor in the brains. We did find a trend toward higher efficiency of the drug combination regarding the virus load and lymphocyte infiltration within the lungs of the animals. Taken together, combining molnupiravir with a DHODH inhibitor was not overtly synergistic but appears as a promising strategy even in this highly susceptible animal model of SARS-CoV-2 infection.

DISCUSSION
Our results demonstrate that the simultaneous application of NHC and DHODH inhibitors suppresses the replication of SARS-CoV-2 in cultured cells far more profoundly than treatment with single drugs. The combination is also effective in animal models. Because both classes of compounds are singularly undergoing advanced clinical evaluation for the treatment of COVID-19, our data raise the perspective of using both drugs together as an antiviral combination therapy.
When using cell lines as an infection model, the two drugs displayed strong degrees of statistically significant synergies (Figure 2). The results of in vivo models, while still revealing the strongest effects when putting both drugs together, did not fulfill the strict criteria of synergism. Perhaps, the uridine in the serum of the animals partially compensates the effects of the DHODH inhibitors. Higher doses of DHODH inhibitors might possibly overcome this in future studies.
Besides their cooperation, another advantage of both NHC and DHODH inhibitors is their robustness toward virus variants, as exemplified ( Figure 4). This was expected since neither of the drugs works by inhibiting a viral enzyme. DHODH inhibitors target a cellular, not a viral function, i.e. pyrimidine synthesis; NHC, in turn, works through incorrect base pairing, not by inhibition of viral enzymes. Indeed, even prolonged NHC treatment did not lead to resistance formation in other coronaviruses (Agostini et al., 2019). This raises the expectation that even novel virus variants, e.g. Omicron, remain susceptible to the drug combination.
On top of enhancing the incorporation of NHC into viral RNA, DHODH inhibitors also induce metabolic stress signaling, leading to the induction of an innate immune response independent of type I interferons.
In addition, owing to their selective inhibition of hyperactive immune cells and excessive cytokine production, they may reduce the hyperinflammation, termed ''cytokine storm'' by some authors (Fajgenbaum and June, 2020), in late stage COVID-19. This can be beneficial, as exemplified by the successful treatment of COVID-19 patients with the steroid dexamethasone (Horby et al., 2021;Tomazini et al., 2020). In fact, DHODH inhibitors are currently in clinical use to treat autoimmune diseases like multiple sclerosis and rheumatoid arthritis (Munier-Lehmann et al., 2013), further supporting their usefulness to dampen excessive immune responses. In summary, we propose that the combination of DHODH inhibitors with NHC targets and abolishes virus replication, whereas DHODH inhibition may also ameliorate COVID-19-associated immunopathology.
It remains to be determined how DHODH inhibitors will affect the T cell response in the context of COVID-19. In general, DHODH inhibition interferes with T cell activity and thus suppresses the immune response (Fragoso and Brooks, 2015). However, the T cell response in the context of COVID-19 is unusual, with CD16positive T cells mediating excessive cytotoxicity (Georg et al., 2022). It is therefore difficult to predict whether interfering with T cell proliferation will be beneficial when simultaneously counteracting virus replication.
Mechanistically, it is conceivable that reduced intracellular levels of cytidine triphosphate (resulting from DHODH inhibition) attenuate the competition and thus enhance the use of triphosphorylated NHC for incorporation into nascent virus RNA, i.e. both genomic and mRNA ( Figure 1A). This was further corroborated by the rescue of virus replication by uridine and cytidine ( Figure 5), each metabolic precursors of CTP. NHC triphosphate is generated by the salvage pathway for pyrimidines but not by de novo pyrimidine Figure 7. Continued represent the body weight of each animal (% of weight at day 0, mean with SD, n = 4). Statistical analysis was performed by one-way ANOVA followed by post hoc Tukey tests (p < 0.05) to reveal that the drug combination was statistically more effective than single drugs on days 3 through 5 post infection. For p values, see Figure S5C. iScience Article synthesis, suggesting that the levels of NHC triphosphate are not impaired by DHODH inhibition. Thus, upon combined treatment, we propose that virus RNA will contain a larger proportion of NHC versus cytidine. In subsequent rounds of virus RNA replication, this will lead to misincorporations of adenine bases instead of guanine (Gordon et al., 2021;Janion, 1978;Janion and Glickman, 1980;Kabinger et al., 2021;Salganik et al., 1973) with higher frequency, and thereby render the virus genome nonfunctional due to missense and nonsense mutations.
Previously (Stegmann et al., 2021a), we have established Methotrexate, a suppressor of purine biosynthesis, as an antagonist to SARS-CoV-2 replication, and this has been confirmed and expanded (Caruso et al., 2021;Zhang et al., 2021). The principle behind this approach is similar to that of DHODH inhibitors, which also interfere with nucleotide biosynthesis.
NHC is a mutagen to bacteria (Janion, 1978;Janion and Glickman, 1980;Jena, 2020;Negishi et al., 1983;Popowska and Janion, 1974;Salganik et al., 1973). Presumably, NHC is converted to its 2 0 -deoxy form and then incorporated into bacterial DNA, followed by false incorporation of adenine bases in the following rounds of DNA replication. It remains to be explored how this will affect the pulmonary or intestinal microbiome in patients with COVID-19. Moreover, it is currently unclear to what extent NHC induces mutagenesis in mammalian cells and possibly in patients when treated with its prodrug molnupiravir, and this raised concerns that it might be carcinogenic and/or teratogenic. A moderate level of mutations in the gene HPRT1 was found after prolonged incubation of cultured cells with NHC (Zhou et al., 2021), but it is unclear whether this might translate into relevant mutagenesis in patients. When used for a few days, as recommended, molnupiravir has not been reported to cause any unacceptable levels of toxicity so far in the target population (adult, contraception, non-pregnant) (Khoo et al., 2021;Painter et al., 2021), although long-term follow-up is pending. Remarkably, ribavirin was not found carcinogenic after being used for decades in hepatitis C treatment, although at least one of its mechanisms of action is also based on mutagenesis of virus RNA (Crotty et al., 2000;Testoni et al., 2014). Thus, there is reason for justifying the cautious use of NHC-based drugs, while prohibiting their use in individuals who are trying to become pregnant or during pregnancy.
Another concern consists in the possible occurrence of transmissible virus mutants/variants from patients treated with molnupiravir. This possibility argues in favor of a ''hit hard and early'' combination treatment strategy and sufficiently long treatment to eradicate the virus-similar to most antibiotic regimens for treating bacterial infections.
Remarkably, NHC treatment of coronavirus-infected cells did not give rise to NHC-resistant viruses, even after prolonged and repeated incubation (Agostini et al., 2019). Likewise, DHODH-inhibitors have a cellular iScience Article target that does not directly interact with a viral factor, thus providing little if any opportunity for resistant virus mutants to arise. Along with our finding that current virus variants of concern (VOC) respond similarly to the original SARS-CoV-2 strain (Figure 4), this raises the hope that the drug combination will be universally applicable to treat most if not all SARS-CoV-2 variants.

STAR+METHODS
Detailed methods are provided in the online version of this paper and include the following:

ACKNOWLEDGMENTS
We thank Thorsten Wolff, Daniel Bourquain, Jessica Schulz, and Christian Mache from the Robert-Koch Institute and Martin Beer from the Friedrich Loeffler Institute (FLI) for providing isolates of SARS-CoV-2 variants. We thank Anna Kraft and Gabriele Czerwinski (both FLI) for support in the preparation of samples for pathology, and Catherine Hambly (University of Aberdeen) for help with daily energy expenditure measurements. We would like to thank Cathrin Bierwirth (University Medical Center Gö ttingen), Isabell Schulz, Anne-Kathrin Donner, and Frank-Thorben Peters for excellent technician assistance and Jasmin Fertey and Alexandra Rockstroh for providing the virus stocks for the mice experiment (Fraunhofer Institute IZI Leipzig). We acknowledge support by the Open Access Publication Funds of the Gö ttingen University. KMS was a member of the Gö ttingen Graduate School GGNB during this work. This work was funded by the COVID-19 Forschungsnetzwerk Niedersachsen (COFONI) to MD, by the Federal Ministry of Education and Research Germany (Bundesministerium fü r Bildung und Forschung; BMBF; OrganSARS, 01KI2058) to SP and TM, and by a grant of the Max Planck Foundation to DG.

DECLARATION OF INTERESTS
AS, HK, EP, and DV are employees of Immunic AG and own shares and/or stock-options of the parent company of Immunic AG, Immunic Inc. Some of the Immunic AG employees also hold patents for the Immunic compounds described in this manuscript (WO2012/001,148, WO03006425). KMS, AD, and MD are employees of University Medical Center Gö ttingen, which has signed a License Agreement with Immunic AG covering the combination of DHODH inhibitors and nucleoside analogs to treat viral infections, including COVID-19 (inventors: MD, KMS, and AD

Lead contact
Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Matthias Dobbelstein (mdobbel@uni-goettingen.de).

Materials availability
This study did not generate new unique reagents.

Data and code availability
All data reported in this paper will be shared by the Lead contact upon request.
This paper does not report original code.
Any additional information required to reanalyze the data reported in this paper is available from the Lead contact upon request.

EXPERIMENTAL MODEL AND SUBJECT DETAILS
Vero E6 cells (Vero C1008) and Calu-3 cells were obtained from the German Primate Research Center Gö ttingen.

METHOD DETAILS
These sections are combined for better readability, since the methods and analyses were different for each experimental model system.

Quantitative RT-PCR for virus quantification
For RNA isolation, the SARS-CoV-2-containing cell culture supernatant was mixed (1:1 ratio) with the Lysis Binding Buffer from the MagNA Pure LC Total Nucleic Acid Isolation Kit (Roche) to inactivate the virus. The viral RNA was isolated using Trizol LS, chloroform, and isopropanol. After washing the RNA pellet with ethanol, the isolated RNA was re-suspended in nuclease-free water. Quantitative RT-PCR was performed according to a previously established RT-PCR assay involving a TaqMan probe , to quantify virus RNA yield. The following oligonucleotides were used for qRT-PCR, which amplify a genomic region corresponding to the envelope protein gene (26,141-26,253), as described .
The amount of SARS-CoV-2 RNA determined upon infection without any treatment was defined as 100%, and the other RNA quantities were normalized accordingly.

Determination of synergy between drug combinations
Vero E6 cells were treated/infected as indicated and RNA within the cell culture supernatant was isolated and quantified by qRT-PCR. The amount of SARS-CoV-2 RNA determined upon infection without any treatment was defined as 100% virus yield (0% inhibition). The other samples were normalized accordingly and stated as percentage of control. The Bliss independence model (Bliss, 1939), calculated using the synergyfinder (Ianevski et al., 2020) at https://synergyfinder.org/was used to quantify synergy between drug combinations. We did not assume mutual exclusivity of the drug effects.

Immunoblot analysis
Cells were washed once in PBS and harvested in radioimmunoprecipitation assay (RIPA) lysis buffer (20 mM TRIS-HCl pH 7.5, 150 mM NaCl, 10 mM EDTA, 1% Triton X-100, 1% deoxycholate salt, 0.1% SDS, 2 M urea), supplemented with protease inhibitors. Samples were briefly sonicated and protein extracts quantified using the Pierce BCA Protein assay kit (Thermo Fisher Scientific). After equalizing the amounts of protein, samples were incubated at 95 C in Laemmli buffer for 5 min and separated by SDS iScience Article the bronchial and alveolar epithelium, diffuse alveolar damage, vasculitis or endothelialitis, pneumocyte type 2 hyperplasia/hypertrophy with atypical cells and hypertrophy/hyperplasia of the bronchial epithelium. Evaluation and interpretation was performed by a board-certified pathologist (DiplECVP) following a post examination masking approach (Meyerholz and Beck, 2018).

Determination of daily energy expenditure
The daily energy expenditure (DEE) was determined individually for nine hamsters (three not infected, three SARS-CoV-2 infected receiving no medication, three SARS-CoV-2 infected receiving Teriflunomid + Molnupiravir) for a total of four days using the doubly labeled water (DLW) method (Lifson and McClintock, 1966;Speakman, 1997), as explained in detail elsewhere (Riek et al., 2021). Briefly, hamsters were injected intraperitoneally with 1.99 G 0.03 g DLW per kg body mass, (65% 18 O and 35% 2 H; 99.90% purity). The individual dose for each hamster was determined prior to the injection according to its body mass. Subsequently, after a 1-h equilibration period, blood samples of 70-100 mL were drawn by puncturing the gingival venous plexus of each hamster at 1, 48 and 96 h after dosing to estimate the isotope elimination rates. Serum from the blood samples were stored at À20 C until determination of 18 O and 2 H enrichment. The DEE was calculated from carbon dioxide production using a single pool model as is appropriate for this size of animal (Speakman, 1993), with Equation 7.17 in (Speakman, 1997) and converted to energy expenditure assuming a respiration quotient of 0.85 and the Weir equation (Weir, 1949). Isotope analyses and calculations were done in a blinded fashion regarding the status of the animals.

Ethics statement
Mouse experiments were carried out according to the German Regulations for Animal Welfare after obtaining the necessary approval from the authorized ethics committee of the State Saxony under the permission number 25-5121/515/7 (TVV 06/21).

Mouse experiment
Female K18-hACE-2 mice (n = 8) received either 50 mg/kg twice per day (bid) Molnupiravir, 10 mg/kg/day Teriflunomide, or a mixture of both, each via oral gavage in a volume of 100 mL, 2 h prior to infection, followed by a 12h-application cycle starting 6 h after virus inoculation. The mock control group received 2.5 mL/kg of the vehicle solution PEG400. Mice were infected intranasally, under isoflurane anesthesia, with 300 FFU of SARS-CoV-2 (strain BavPat1/2020) in 50 mL total volume. At day 4 after virus inoculation, the mice were euthanized and organs were homogenized in 2 mL PBS. Viral RNA was isolated from lung homogenates and quantified by qRT-PCR. A low inoculum was used for greater sensitivity of the assay, since this puts a higher demand on virus replication. Moreover, a low inoculum is more representative of what most COVID-19 patients receive when they get accidentally infected. The treatment scheme was depicted using BioRender.com.

Detection of virus RNA
Viral RNA was isolated from 140 mL of homogenates using QIAamp Viral RNA Mini Kit (Qiagen). RT-qPCRs were performed using TaqMan Fast Virus 1-Step Master Mix (Thermo Fisher) and 5 mL of isolated RNA as a template, as described (Groß et al., 2020). Synthetic SARS-CoV-2 RNA was used as a quantitative standard to obtain viral copy numbers. Statistical evaluation of the data was performed by Mann-Whitney U test in comparison to the mock control and single treatments.

Determination of lung affection by histopathology
Evaluation of mouse pathology was performed in analogy to the hamster experiments, as described above.
A historical control of a non-infected mouse (same genotype) was added at a later time. Lymphocyte infiltration was determined in the perivascular and the interstitial regions of the murine lungs and scored each on a scale between 0 and 3. The sum of these scores was calculated for each animal.

Quantification and statistical analysis of cell-based experiments
Statistical testing was performed using Graph Pad Prism 9. Unless otherwise specified, a two-sided unpaired Student's t test was calculated, and significance was assumed where p % 0.05. Asterisks represent significance in the following way: ****, p % 0.0001; ***, p % 0.005; **, p % 0.01; *, p % 0.05. For statistical analyses and graphical illustrations, GraphPad Prism version 9.0.0 (GraphPad Software, La Jolla, CA) and SPSS version 20.0 (SPSS Inc., Chicago, IL, United States) were used. Hamster body weight was statistically analyzed by one-way ANOVA followed by post hoc Tukey tests (p < 0.05). All other data were non-parametric and tested by Kruskal-Wallis test with Dunn's correction.

ll
OPEN ACCESS