Targeting polyamines inhibits coronavirus infection by reducing cellular attachment and entry

Abstract

Coronaviruses first garnered widespread attention in 2002 when the severe acute respiratory syndrome coronavirus (SARS-CoV) emerged from bats in China and rapidly spread in human populations. Since then, Middle East respiratory syndrome virus (MERS-CoV) emerged and still actively infects humans. The recent SARS-CoV-2 outbreak and the resulting disease (coronavirus disease 2019 – COVID19) have rapidly and catastrophically spread and highlighted significant limitations to our ability to control and treat infection. Thus, a basic understanding of entry and replication mechanisms of coronaviruses is necessary to rationally evaluate potential antivirals. Here, we show that polyamines, small metabolites synthesized in human cells, facilitate coronavirus replication, and depletion of polyamines with FDA-approved molecules significantly reduces coronavirus replication. We find that diverse coronaviruses, including endemic and epidemic coronaviruses, exhibit reduced attachment and entry into polyamine-depleted cells. We further demonstrate that several molecules targeting the polyamine biosynthetic pathway are antiviral in vitro. In sum, our data suggest that polyamines are critical to coronavirus replication and represent a highly promising drug target in the current and any future coronavirus outbreaks.

Graphical Abstract

Coronaviruses are significant human pathogens with significant epidemic potential. Recurring coronavirus outbreaks wreak havoc on human health and activity, as demonstrated by the original severe acute respiratory syndrome coronavirus (SARS-CoV) outbreak of 2002, the Middle East respiratory syndrome coronavirus (MERS-CoV) of 2011 to present day¹, and the newly-emergent SARS-CoV-2 of 2019 to present². Each of these viruses emerged from zoonotic sources and rapidly transmitted through human populations, requiring urgent attention due to the number of infections, combined with mortality rates of up to 35% in the case of MERS-CoV³. In addition to their epidemic potential, coronaviruses are common human pathogens, with several endemic coronaviruses causing recurrent infections with generally mild symptoms^4,5. Treatments for SARS-CoV-2, as well as other coronaviruses, include remdesivir, a base analog that exhibited promising clinical efficacy^6,7. A dearth of additional antivirals is available to treat or prevent infections. The omnipresence, unpredictable epidemic potential, and significant impact on human health make investigation into potential antiviral compounds essential to battle current and future coronavirus outbreaks.

Polyamines are small aliphatic metabolites synthesized by mammalian cells to support cellular processes such as cell cycling, transcription, and translation^8–10. The polyamines putrescine, spermidine, and spermine (Figure 1A) are synthesized from an ornithine precursor through the enzyme ODC1. Several specific and potent inhibitors of the polyamine biosynthetic pathway have been developed and are FDA-approved. The ODC1 inhibitor difluoromethylornithine (DFMO) is approved for the treatment of trypanosomiasis¹¹. In fact, DFMO exhibits low toxicity and mild side effects¹². Topical DFMO has also been developed and is sold as Vaniqa to reduce hair growth¹³. Viruses rely on polyamines for varied stages in replication, including polymerase activity¹⁴, cell binding¹⁵, and viral protein translation^15,16. Polyamine depletion via DFMO effectively reduces replication of diverse viruses in vitro and in vivo, with minimal cytotoxicity^17,18. Prior work demonstrated that MERS-CoV was sensitive to the polyamine inhibitor DFMO in vitro at a dose of 500 μM¹⁹, suggesting that polyamines function in coronavirus replication. Whether and how polyamines function in the replication of other coronaviruses remains to be investigated.

Figure 1. Polyamine synthesis inhibitor DFMO limits coronavirus replication.

(A) Schematic of the polyamine pathway. The polyamines putrescine, spermidine, and spermine are synthesized from ornithine via ornithine decarboxylase (ODC1), which is inhibited by difluoromethylornithine (DFMO). (B) Vero-E6 cells treated with increasing doses of DFMO were infected at MOI 0.1 SARS-CoV-2 for 24h and viral genomes measured and converted to titer equivalents via a standard curve. (C) Vero-E6 cells were treated with escalating DFMO doses and viability and (D) polyamine levels were measured. (E) BHK-R cells were treated with DFMO as indicated and infected with MHV at MOI 0.01 for 48h. (F) Cellular viability levels were measured and (G) polyamine depletion was confirmed by chromatography. (H) BHK-R cells were treated with 1 mM DFMO and infected with MHV at MOI 0.01. Viral titers were determined after several rounds of replication. (I) MHV genomes from samples in (E) were quantified by qRT-PCR and compared to viral titers to quantify the relative number of genomes to PFU (J). (K) HeLa-R cells were treated with escalating doses of DFMO and infected with MHV at MOI 0.01 for 48h. (L) HeLa-R cells exposed to DFMO for four days were analyzed for viability. (M) Depletion of polyamines was confirmed by chromatography. *p<0.05, **p<0.01, ***p<0.001 by Student’s T-test. Data from at least three independent experiments.

The study of human epidemic coronaviruses is limited due to requisite biosafety protocols. Thus, murine model systems have been developed to study coronavirus replication and pathogenesis with significantly reduced risk. Murine hepatitis virus (MHV) is a common model system for coronavirus infection. MHV, a β-coronavirus, is related to the human epidemic coronaviruses though safe to work with in the laboratory. The MHV model is well characterized^20,21, and combined with its ease of use, it provides a significant opportunity to study coronavirus infection.

Using the MHV model system, we investigated a role for polyamines in coronavirus infection. We observe that polyamine depletion via DFMO limits virus replication and that polyamines support virus infection at the stage of attachment: polyamine depletion reduces virus binding to susceptible cells. We extend this phenotype to the human endemic coronavirus NL63 as well as the pandemic SARS-CoV-2. Additionally, we demonstrate that several molecules targeting the polyamine pathway, including FDA-approved pharmaceuticals, exhibit significant antiviral activity. Together, our data highlight a role for polyamines in coronavirus replication and suggest that targeting polyamines may be a viable preventative or treatment option for coronavirus infection.

Results

Polyamine depletion limits coronavirus replication.

To determine if coronaviruses are susceptible to polyamine depletion, we treated Vero-E6 cells with doses of DFMO ranging from 750 μM to 5 mM for four days prior to infection with SARS-CoV-2 at multiplicity of infection (MOI) of 0.1 plaque forming units (pfu) per cell. Released viral genomes were then measured at 24 hours post infection (hpi) and converted to virus equivalents using a standard curve comparing known infectious titers and genomes. We observed that DFMO significantly reduced SARS-CoV-2 genomes, greater than 50-fold at 5 mM, with an IC50 (50% inhibitory concentration) value of 230 μM (Fig. 1B). To confirm that DFMO was not cytotoxic, we measured viability of Vero-E6 cells treated with escalating doses of DFMO and calculated a CC50 (50% cytotoxic concentration) value of 11 mM (Fig. 1C; summarized in Table 1). We further confirmed that DFMO treatment at these concentrations successfully depleted polyamines via thin layer chromatography (TLC; Fig. 1D).

Table 1.

Properties of compounds used in these studies.

Inhibitor	Virus	Cells	IC50 (μM)	CC50 (mM)	SI
DFMO	MHV-A59	BHK-R	380	3.8	10
		HeLa-R	7.0	16	2300
DENSpm		HeLa-R	1.6	35	7100
CPX		BHK-R	2.2	76	35000
DEF		BHK-R	54	60	1100
GC7		BHK-R	7.4	42	5600
DFMO	SARS-CoV-2	Vero-E6	230	11	48

To expand on this work, we turned to the MHV model system. As with our experiments with SARS-CoV-2, BHK-21 cells expressing the MHV receptor (BHK-R)²² were pretreated with various concentrations of DFMO 4 days before infection with an MOI of 0.01. Supernatant was collected 48 h post infection (hpi) and titers were measured with a plaque assay on BHK-Rs. Viral titers were significantly reduced in a dose dependent manner, with an IC50 value of 380 μM (Fig. 1E). We confirmed that DFMO depleted polyamines in these cells by TLC (Fig. 1G), especially levels of putrescine and spermidine, as previously described¹⁴. To test if DFMO remained antiviral over multiple rounds of infection, BHK-Rs were treated with DFMO and infected at MOI of 0.01. Supernatant was collected and titered at the indicated time points (Fig. 1H). DFMO reduced viral titers compared to no treatment group up to 72 h, and viral titers in DFMO-treated cells plateaued at a lower level than untreated controls, suggesting that MHV was unable to overcome polyamine depletion over several rounds of replication. We observed that DFMO had modest toxicity in BHK-R cells, with a CC50 of 3.8 mM (Figure 1F), giving a selectivity index (SI) of 10. Finally, we measured the relative number of genomes present in supernatant from infected cells. Following treatment and infection, we purified, reverse-transcribed and measured viral genomes by qPCR using specific primers and observed reduced genome levels with DFMO treatment (Figure 1I). Using the data from infectious titers (Fig. 1E) and viral genomes (Fig. 1I), we calculated the specific infectivity (genomes per PFU) and observed that DFMO treatment increased the ratio of genomes to titer (Fig. 1J), similar to previous findings with bunyaviruses and polyamine depletion²³. Thus, polyamines enhance coronavirus infection and maintain specific infectivity of MHV.

To address the role of cell type, we tested the ability of DFMO to restrict virus replication in HeLa cells expressing the MHV receptor (HeLa-R). Cells were pretreated with DFMO and infected at MOI 0.01 for 48h. We observed a decline in virus replication, more pronounced than on BHK-R cells, with an IC50 value of approximately 7.0 μM (Fig. 1K). We confirmed that DFMO treatment of HeLa-R cells reduced polyamines (Fig. 1M) and verified that DFMO was not toxic to these cells (Fig. 1L). The CC50 of DFMO in HeLa cells was 16 mM, and, thus, the selectivity index was 2300.

Polyamines facilitate coronavirus replication.

DFMO restricts virus replication by reducing cellular polyamine pools¹⁹. However, excess polyamines do not enhance virus replication, according to work based on chikungunya virus¹⁴. Recent work has suggested that polyamines may serve coronavirus replication²⁴. To determine if exogenously-supplied polyamines serve an antiviral role, we treated BHK-R cells with increasing doses of the individual polyamines putrescine, spermidine, and spermine, from 1 to 100 μM. We measured cellular viability (Fig. 2A) and infected these cells to measure viral titers at 48 hpi (Fig. 2B). We observed that the polyamines significantly reduced cellular viability at approximately 5 μM, reflecting prior descriptions of high polyamine concentrations leading to apoptosis^25,26. Despite this reduction in viability, we observed modestly reduced viral titers (Fig. 2B). We measured polyamines in supplemented cells and observed that supplementation with any of the three polyamines resulted in only modest changes in cell-associated polyamines (Fig. 2B, lower), as we previously described²⁷. Reduced signal with high concentrations of spermidine and spermine likely was an artefact of reduced cellular viability at these doses.

Figure 2. Polyamines facilitate coronavirus replication.

(A) BHK-R cells were treated with escalating doses of putrescine, spermidine, and spermine and cellular viability was measure 24h later. Calculated CC50 values are inset. (B) BHK-R cells treated as in (A) were infected with MHV at MOI 0.01 and titered at 48 hpi (upper). Cell-associated polyamines were measured by chromatography (lower). (C) BHK-R cells were treated with 1 mM DFMO for four days prior to infection with MHV at MOI 0.01. At the time of infection, 1 μM polyamines were added to the cells. Viral titers were determined at 48 hpi (upper). Cell-associated polyamines were measured by chromatography (lower). (D) MHV was directly incubated with DFMO or 1 μM of the polyamines putrescine (put), spermidine (spd), and spermine (spm) for 24h and then directly titered. **p<0.01 by Student’s T-test. Data from at least two independent experiments.

To determine if that DFMO-mediated virus restriction was a result of polyamine depletion, we treated BHK-R cells with DFMO and at the time of infection, supplemented with individual polyamines putrescine, spermidine, and spermine at 1 μM, a level that did not affect cellular viability (Fig. 2A). We observed that DFMO reduced viral titers and that supplementation with any of the individual polyamines rescued virus replication to untreated levels (Fig. 2C). We confirmed the presence of these polyamines (Fig. 2C, below) and observed that supplementation with any of the individual polyamines results in the accumulation of all three polyamines in the cells. To exclude the possibility that polyamines may directly inactivate viral particles, we directly incubated MHV with DFMO and each of the individual polyamines at 10 μM for 24h and then titered the incubated virus. We observed that neither DFMO nor polyamines affected infectivity when incubated with virus, as titers were unchanged compared to untreated controls (Fig. 2D). In sum, our data suggest that polyamines support virus yield and that supplementation of polyamines supports but does not enhance this yield and can instead be cytotoxic.

Polyamine depletion is prophylactic.

DFMO depletes polyamines by inhibiting ODC1, though time is required to deplete cellular polyamine pools and subsequently reduce viral titers¹⁴. Ideally, treatment with antivirals commences after the initiation of infection; though identification of soon-to-be-infected individuals is impractical. Thus, to determine if we could treat cells with DFMO at the time of infection, which would be more clinically relevant, we determined whether DFMO suppresses plaque development when present in overlay media, as previously described¹⁹. We observed no changes in plaque size (Fig. 3A) or morphology (Fig. 3B) with DFMO supplementation to the media, suggesting that DFMO was not sufficiently antiviral when applied to cells after infection and highlighting that DFMO must be applied prophylactically to reduce coronavirus replication. We observed a small, though statistically insignificant change in HeLa-R cells (Fig. 3C).

Figure 3. Polyamine depletion is antiviral prophylactically.

Confluent monolayers of BHK-R cells were treated with DFMO at the time of infection with MHV and overlaid with agarose to form plaques. (A) Plaques were quantified two days later. (B) Representative plaques are shown from (A). (C) HeLa-R cells were infected and treated as in (A) and plaque area measured. (D) BHK-R cells were treated with 1.5 mM DFMO either in one, three, or four doses prior to infection with MHV at MOI 0.01 Viral titers were determined by plaque assay at 48 hpi. (E) Cells were treated with 1 mM DFMO and then subsequently refed with fresh medium not containing DFMO 24h before infection with MHV at MOI 0.01. Titers were determined at 24 hpi. **p<0.01, ***p<0.001 by Student’s T-test. Representative data from at least three independent experiments. Individual plaque sizes are indicated in (A) and (C).

Use of DFMO as an anti-trypanosomal agent requires repeated dosing to reach a serum concentration high enough to inhibit trypanosome growth^11,12,28. To mimic this, we treated BHK-R cells with 1.5 mM DFMO either in a single treatment, three 500-μM treatments, or four 375-μM treatments equally spaced over the four days prior to infection. We observed that cells receiving multiple doses of the drug reduced cellular capacity to support virus production, similar to when the drug was delivered in a single dose (Fig. 3D), suggesting that repeated dosing is as effective as a single drug treatment in vitro. Finally, we tested whether the antiviral effects of DFMO were reversed by removing the drug and replenishing cells with fresh media. We treated cells with 1 mM for four days and then removed media and replaced with fresh media without drug for 24h before infection. When we measured viral titers, we observed the DFMO treatment reduced infectious virus production, and this effect was completely reversed by replenishing the cells with fresh media, suggesting that DFMO’s antiviral effect can be reversed by removing the drug.

Polyamines facilitate coronavirus binding and entry.

We previously demonstrated that enteroviruses, flaviviruses, and bunyaviruses rely on polyamines for viral attachment and entry¹⁵. To test whether coronaviruses similarly rely on polyamines for attachment and entry, we performed a binding assay in cells treated with DFMO. We treated BHK-R cells with DFMO for four days prior to infection with 1000 pfu. After five minutes, unbound virus was washed away with PBS and cells were overlaid with agarose for a plaque assay, lacking DFMO in the overlay. Thus, DFMO is present only during binding of MHV to target cells and not for any other portion of the viral life cycle. Plaques were then allowed to form and were enumerated, representing bound and entered infectious virus. We observed that attachment and plaque formation was significantly inhibited with DFMO treatment (Fig. 4A, representative plaques for NT and 1 mM in Fig. 4B). We expanded these results by infecting DFMO-treated BHK-R cells with MHV for five minutes, washing away unbound virus, and quantifying bound virus by qRT-PCR. We observed that, similar to our plaque formation assays, bound viral genomes were significantly reduced with DFMO treatment (Fig. 4C). We confirmed that the cells expressed the MHV receptor (CECAM-1) by immunofluorescence (Fig. 4D), regardless of DFMO treatment, suggesting that DFMO treatment was not directly reducing receptor availability. We recapitulated these results with HeLa-R cells and observed a similar reduction in virus binding with DFMO treatment (Fig. 4E).

Figure 4. Polyamines facilitate coronavirus cellular attachment.

Confluent BHK-R cells were treated with DFMO prior to MHV binding assay. Virus was inoculated on cells for 5 min prior to removal, wash, and agarose overlay in media without DFMO. Plaques formed were quantified (A) and representative images from 1 mM DFMO treatment are shown (B). (C) BHK-R cells were treated as in (A), but after washing unattached virus, attached virus and cells were collected for RNA purification, reverse transcription, and viral genome quantitation by qPCR. Attached genomes were normalized to input virus. (D) MHV receptor (CECAM-1) was visualized on untreated and 1 mM DFMO treated cells. (E) HeLa-R cells were treated with 1 mM DFMO, infected with MHV, and processed as in (C). (F) Vero-E6 cells were treated with DFMO and analyzed for binding ability of HCoV-NL63 as in (C). (G) ACE2 expression on Vero-E6 cells was confirmed by immunofluorescence. (H) Pseudotyped viral particles with no viral glycoproteins or the SARS-CoV-2 spike were tested for their ability to bind and enter Vero-E6 cells. (I) Spike protein was confirmed to be present on pseudotyped virus by western blot. (J) Vero-E6 cells were treated with DFMO and infected with SARS-CoV-2 pseudoparticles as in (C). *p<0.05, **p<0.01, ***p<0.001 by Student’s T-test. Data from at least three independent experiments.

To expand to human coronaviruses, we tested HCoV-NL63 for binding to DFMO-treated Vero-E6 cells. Again, we observed that attachment was sensitive to DFMO treatment (Fig. 4E). Again, we noted no change in viral receptor expression (ACE2, Fig. 4G). Taking advantage of a pseudoparticle system as previously described²⁹ and adapted to SARS-CoV-2, we measured attachment of SARS-CoV-2 to DFMO-treated cells. We verified that these pseudoparticles were competent to bind susceptible cells compared to “bald” particles, lacking viral spike (Fig. 4H), and we confirmed the presence of the spike protein on purified particles by western blot (Fig. 4I). We observed a reduction in cellular attachment of SARS-CoV-2 pseudoparticles in a dose-dependent manner (Fig. 4H). Thus, human coronaviruses, including pandemic SARS-CoV-2, are sensitive to polyamine depletion and rely on polyamines for efficient cellular attachment.

Additional molecules targeting the polyamine and hypusination pathway quell coronavirus infection.

Several pharmaceuticals target polyamine biosynthesis (Fig. 5A) and exhibit antiviral activity^16,19, though coronaviruses have not been thoroughly tested. Additionally, an offshoot of the polyamine pathway, the cellular hypusination pathway, in which spermidine is conjugated to eIF5A, plays roles in cellular translation³⁰ is critical for filovirus replication^16,31. To determine if additional compounds targeting the polyamine and hypusination pathways exhibited activity against coronaviruses, we treated cells with increasing doses of inhibitors of these pathways. We used the molecules diethylnorspermidine (DENSpm; enhances spermidine-spermine acetyltransferase [SAT1] activity and polyamine breakdown), GC7 (inhibits deoxyhypusine synthase [DHPS] to prevent eIF5A hypusination), deferiprone and ciclopirox (DEF and CPX; inhibit deoxyhypusine hydroxylase [DOHH]) because they target distinct steps in the polyamine pathway. DENSpm significantly reduced polyamine levels and MHV titers at 10 μM (Fig. 5B), with an IC50 of 1.6 μM. We observed minimal toxicity (Fig. 5C), with a CC50 value of 35 μM resulting in a selectivity index of 23. Because DENSpm activates SAT1 to deplete polyamines, its action is much more rapid than DFMO. Thus, we tested whether DENSpm exhibited antiviral activity after the initiation of infection. Cells were treated with DENSpm at times before and after infection and virus titers measured at 48 hpi. We observed that DENSpm was most effective when treatment initiated before infection; however, significant antiviral activity was maintained when DENSpm was added as late as 4 hpi (Fig. 5D). Finally, we tested the hypusination inhibitors GC7, DEF, and CPX. We observed in all cases that these molecules significantly reduced MHV replication in a dose-dependent manner (Fig. 5E–G) and that cytotoxicity was minimal (Fig. 5H; summarized in Table 1). Thus, molecules targeting an additional branch of the polyamine biosynthetic pathway exhibit anti-coronavirus activity.

Figure 5. Additional molecules targeting the polyamine pathway quell coronavirus infection.

(A) Schematic of the polyamine and hypusination pathway. (B) HeLa-R cells were treated with escalating doses of DENSpm for 16 h prior to infection with MHV at MOI 0.01. Viral titers were determined at 48 hpi (upper). Polyamine levels were measured by chromatography (lower). (C) Treated cells were assayed for viability after 16 h of treatment with DENSpm. (D) HeLa-R cells were treated with DENSpm at the indicated times before and after infection with MHV at MOI 10. Viral titers were determined at 48 hpi. BHK-R cells were treated with (E) CPX, (F) DEF, and (G) GC7 for 24h prior to infection with MHV at MOI 0.01. Viral titers were determined at 48 hpi. (H) BHK-R cells treated with CPX, DEF, and GC7 were analyzed for cell viability after 24h treatment. *p<0.05, **p<0.01, ***p<0.001 by Student’s T-test. Data from at least three independent experiments.

Discussion

Polyamines are crucial to virus infection; however, their role(s) in coronavirus replication was previously unknown. Here, we demonstrated that polyamines are essential to coronavirus infection and that attachment is at least one of the steps affected by polyamine depletion. We previously demonstrated that enteroviruses are similarly sensitive to polyamine depletion and rely on these polyamines for attachment¹⁵. This phenotype is conserved among several viral families, suggesting that polyamines may mediate a common attachment factor. Whether polyamines directly facilitate attachment by mediating virus-cell connections is unclear, and future work will be needed to delineate precisely how polyamines affect attachment. In addition to attachment, we observed that inhibitors of the hypusination pathway impact virus replication, suggesting that polyamines may facilitate infection through eIF5A hypusination, as observed for Ebolavirus¹⁶. Given the role of hypusinated eIF5A in translation, inhibitors of this pathway may directly inhibit viral translation or may inhibit the translation of a cellular factor essential to virus replication. Regardless, targeting polyamines may affect multiple steps in infection, and additional work will be needed to uncover additional mechanisms.

Polyamines play several important roles in the cell, and recent work has highlighted pathways that may modulate virus replication. Spermidine, as a substrate for eIF5A hypusination, enhances cellular autophagy³², and recent work has suggested that spermidine-enhanced autophagy limits SARS-CoV-2 replication²⁴. In contrast, our results suggest that spermidine does not inhibit or restrict virus replication, at concentrations up to 100 μM. We do observe that spermine supplementation to cells reduces virus replication; however, we also observe significant cytotoxicity with spermine treatment, likely the reason for reduced titers. Additionally, we observed that DFMO treatment reduced polyamine content within cells, even at concentrations that did not affect viral titers. These data suggest that coronaviruses require a low concentration of polyamines for robust replication. In fact, bunyaviruses and alphaviruses replicate with supplementation of 100 nM polyamines in polyamine depleted cells^14,33. Thus, these viruses may have evolved mechanisms to take advantage of polyamines even at low concentrations. This has implications for the molecular interactions between polyamines and coronaviruses, and perhaps coronaviruses evolved strong affinity for polyamines. This also has implications for infection in an organism, as polyamines are transported between cells. In cancer therapies, polyamine transporter inhibitors can be used in conjunction with polyamine depletion agents to prevent both polyamine synthesis and internalization. While this has not been tested in the context of virus infection, the combination of inhibitors of synthesis and transport may more effectively reduce virus replication. Precisely how polyamines play into cellular signaling and metabolic pathways to affect virus replication is not clear; however, the complex interplay between metabolism and virus replication affords several opportunities to target virus replication with small molecule inhibitors.

With the current SARS-CoV-2 outbreak and the resulting coronavirus disease of 2019 (COVID-19), novel antivirals are urgently needed. Remdesivir has shown significant potential clinically^34–37, and additional clinical testing will determine if this molecule can effectively treat infected individuals. Our work with DFMO and polyamine depletion highlights another promising intervention. An important limitation of our study is the relatively high concentration of drug required to reduce coronavirus infection. Given this caveat, the possibility of off-target antiviral effects cannot be ignored. However, we show that DFMO’s antiviral effects are reversed by addition of exogenous polyamines, suggesting specificity. Our prior results with chikungunya, Zika, Coxsackie, and Rift Valley fever viruses, have demonstrated that low micromolar concentrations of inhibitor are sufficient to reduce virus replication. Thus, coronaviruses may be less reliant on polyamines than other viral families, though this requires significant further investigation.

Importantly, we observe that polyamine depletion functions prophylactically to reduce virus replication and may not reduce viral titers after the initiation of infection. Given the limits to our in vitro studies, we cannot say for certain whether polyamine depletion in a human could reduce virus replication or improve clinical outcome when treatment commences after exposure. However, polyamine depletion may be most effective as a prophylactic, preventing infection of vulnerable populations. With the significant spread of SARS-CoV-2, protecting vulnerable populations, including the elderly and immunocompromised is essential to reducing mortality. Further, prophylactically treating frontline workers, including medical staff, could reduce spread of the virus in a hospital setting. Currently, DFMO is approved for human use as a topical and intravenous agent. Nebulization and direct lung delivery, if effective at depleting polyamines in the lung, would be a benefit for reducing virus burden, but this remains untested. Whether DFMO or other polyamine-targeting drugs are sufficient to do this is unclear and requires significantly more investigation, including in animal models.

Methods

Cell Culture.

Cells were maintained at 37°C in 5% CO₂, in Dulbecco’s modified Eagle’s medium (DMEM; Life Technologies) with bovine serum and penicillin-streptomycin. Vero cells were obtained through BEI Resources, and VeroE6 cells were obtained through ATCC (CRL-1586). Vero and HeLa-R cells were supplemented with 10% new-born calf serum (NBCS; Thermo-Fischer). BHK-R cells were kindly provided by Dr. Susan Baker and were supplemented with 10% fetal bovine serum (FBS; Thermo-Fischer). HEK293T cells were grown in 10% FBS.

Infection and Enumeration of Viral Titers.

MHV-A59 was derived from the first passage of virus in BHK-R cells from an infectious clone. SARS-CoV-2, isolate BetaCoV/France/IDF0372/2020 C2, was supplied by the National Reference Centre for Respiratory Viruses hosted at Institut Pasteur (Paris, France). The human sample from which this strain was isolated has been provided by Dr. X. Lescure and Pr. Y. Yazdanpanah from the Bichat Hospital, Paris, France. Viral stocks were prepared by propagation in Vero E6 cells in DMEM supplemented with 2% FBS. All experiments involving live SARS-CoV-2 were performed in compliance with Institut Pasteur guidelines for Biosafety Level 3 work. For all infections, DFMO was maintained throughout infection as designated. Viral stocks were maintained at −80°C. For infection, virus was diluted in serum-free DMEM for a multiplicity of infection (MOI) of 0.01 on BHK-R cells, unless otherwise indicated. Viral inoculum was added to the cells and supernatants were collected at specified timepoints. To quantify viral titers via plaque assay, dilutions of cell supernatant were prepared in serum-free DMEM and used to inoculate confluent monolayers of BHK-R cells for 10 to 15 min at 37°C. Cells were overlain with 0.8% agarose in DMEM containing 2% FBS. MHV-A59 samples were incubated for 2 days at 37°C. Cells were fixed with 4% formalin and revealed with crystal violet solution (10% crystal violet; Sigma-Aldrich). Plaques were enumerated and used to back-calculate the number of plaque forming units (PFU) per milliliter of collected volume.

Drug Treatments.

Difluoromethylornithine (DFMO; TargetMol) was diluted to 100 mM solution in sterile PBS. For DFMO treatments, cells were trypsinized (Zymo Research) and reseeded with fresh medium supplemented with 2% FBS. Cells were treated with 1 mM DFMO unless otherwise indicated. Cells were incubated with DFMO for 96 hours to allow for depletion of polyamines. Experiments involving polyamine rescues were performed using 1 μM polyamines (Sigma-Aldrich) added to either the cell supernatant or viral inoculum as indicated.

Thin-layer chromatography of polyamines.

Polyamines were separated by thin-layer chromatography as previously described ³⁸. For all samples, cells were treated as described prior to being trypsinized and centrifuged. Pellets were washed with PBS and then resuspended in 200 μL 2% perchloric acid. Samples were then incubated overnight at 4°C. Supernatant or 1 μmole of putrescine, spermidine, and spermine standards was combined with 5 mg/mL dansyl chloride (Sigma Aldrich) in acetone and saturated sodium bicarbonate. Samples were incubated in the dark overnight at room temperature. Excess dansyl chloride was cleared by incubating the reaction with proline (Sigma Aldrich). Dansylated polyamines were extracted with toluene (Sigma Aldrich) and centrifuged. Sample was added in spots to silica gel matrix TLC plates (Sigma Aldrich) and exposed to ascending chromatography with 1:1 cyclohexane: ethyl acetate. The plate was dried and visualized via exposure to UV.

Plaque formation attachment assay.

BHK-R cells were seeded in 6 well plates and grown to confluence in DMEM with 2% FBS. The cells were treated for 96 hours with varying concentrations of DFMO. After the 96 hour DFMO treatment, the media was aspirated from the cells and replaced with 500 μL serum free media containing 1000 PFU MHV. The infected cells incubated at room temperature or on ice for 5 minutes. Cells were then washed 3x with PBS and then overlaid with 0.8% agarose containing DMEM with 2% FBS. The plates were incubated at 37°C for 2 days for plaques to develop. The cells were fixed with 4% formalin, and the plaques were visualized with crystal violet staining.

Plaque size measurement.

BHK-R or HeLa-R cells were seeded in 6 well plates and grown to confluence. Approximately 100 PFU of MHV-A59 was diluted in a 500 μL inoculum of serum free DMEM. The inoculum was incubated on the cells for approximately 30 mins at 37°C. After 30 mins, an overlay of 8mL 0.8% agarose containing DMEM with 2% FBS (BHK-R) or NBCS (HeLa-R) was added to each well. The dishes were incubated at 37°C for 2 days to allow plaque formation. The cells were fixed with 4% formalin and the agarose plugs removed. The fixed cells were stained with crystal violet. Plaque size was determined using ImageJ software ³⁹.

Immunofluorescence imaging.

Cells grown on cover slips were either treated with 1 mM DFMO or untreated. Cells were fixed with 4% formalin for 15 minutes, washed with PBS, permeabilized and blocked with 0.2% Triton X-100 and 2% BSA in PBS (blocking solution) for 30 minutes at room temperature (RT). Cells were sequentially incubated as follows: Primary rabbit anti-hACE2 (Invitrogen, SN0754) (Vero E6) or rabbit anti-mCEACAM (BHK-R) antibodies 1:500 in blocking for 2 hrs at room temperature. Cells were subsequently washed then incubated with secondary goat anti-rabbit (1:500 in blocking, 30 min, RT). Mounting media with DAPI was used to visualize nuclei. Samples were imaged with Zeiss Axio Observer 7 with Lumencor Spectra X LED light system and a Hamamatsu Flash 4 camera using appropriate filters using Zen Blue software with a 40X objective.

Western blot.

Samples were collected with Bolt LDS Buffer and Bolt Reducing Agent (Invitrogen) and run on polyacrylamide gels. Gels were transferred using the iBlot 2 Gel Transfer Device (Invitrogen). Membranes were probed with primary antibodies for mouse anti-rhodopsin (EMD Millipore, MAB5356), anti-VSV-M (23H12), and anti-mouse IgG (Goat, HRP-Labelled, NEF822001EA). Membranes were treated with SuperSignal West Pico PLUS Chemiluminescent Substrate (ThermoFisher Scientific) and visualized on ProteinSimple FluorChem E imager.

RNA purification and cDNA synthesis.

Media was cleared from cells and Trizol reagent (Zymo Research) directly added. Lysate was then collected, and RNA was purified through Zymo RNA extraction kit. Purified RNA was subsequently used for cDNA synthesis using High Capacity cDNA Reverse Transcription Kits (Thermo-Fischer), according to the manufacturer’s protocol, with 10–100 ng of RNA and random hexamer primers.

qPCR-based attachment assay.

BHK-R, HeLa-R, and Vero cells were seeded at 4×10⁴ cells per well in 24 well plates in DMEM with either 2% FBS or NBCS. The cells were treated for 96 hours with varying concentrations of DFMO. After 96 hours, the media was aspirated from the cells and replaced with 100 uL of serum free media containing virus. The infected cells incubated for 5 minutes at room temperature or on ice. The cells were then washed 1x with PBS and then 200 μL of Trizol was added to the cells. The RNA was extracted with the Zymo RNA extraction kit, converted to cDNA, and quantified by real-time PCR with SYBR Green (DotScientific) using a one-step protocol using QuantStudio 3 (ThermoFisher Scientific). Relative genomes calculated using the ΔC^T method, normalized to β-actin qRT-PCR control and calculated as fraction of the unwashed samples. Primer sequences for NL63 are 5’-TGT-CAA-CGA-GGT-TTT-GCA-TTA-AAT-3’ (F) and 5’-ACT-GGC-CTA-CCA-TTG-TGT-GTA-AGA-3’ (R); for MHV are 5’-ATC-CTC-AAG-AAG-ACC-ACT-TGG-GCT-GAC-3’ (F) and 5’-GAG-TAA-TGG-GGA-ACC-ACA-CTC-CCG-3’ (R); β-actin are 5’-CAC-TCT-TCC-AGC-CTT-CCT-TC-3’ (F) and 5’-GTA-CAG-GTC-TTT-GCG-GAT-GT-3’ (R). Primers were verified for linearity using eight-fold serial diluted cDNA and checked for specificity via melt curve analysis. SARS-CoV-2 genomes were detected using a Taqman assay (IP4 set), as described on the WHO website (https://www.who.int/docs/default-source/coronaviruse/real-time-rt-pcr-assays-for-the-detection-of-sars-cov-2-institut-pasteur-paris.pdf?sfvrsn=3662fcb6_2)

Synthesis of SARS-CoV-2 pseudovirus.

293T cells were seeded into a 150mm tissue culture dish 24 hours prior to transfection. Cells were transfected with SARS-CoV-2 spike using LipoD293 following manufacture’s protocol. After 6 hours, the media was removed and replaced with fresh, prewarmed 10% FBS media. The following day, cells were infected with a 1 to 100 dilution of Fluc-VSV-Junin Virus and allowed to infect for two hours. After the two hours, cells were washed 2x with serum free DMEM and fresh, prewarmed media was placed onto the cells. On the following day, cell supernatant was collected and spun down via differential centrifugation (300×g for 10 min at 4°C, 4500×g for 10 min at 4°C) and stored at −80°C.

Pseudovirus attachment assay.

Vero E6 cells were seeded at 4×10⁴ in 24 well plates in DMEM with 2% NBCS. Cells were then treated with varying concentrations of DFMO for 96 hours. After 96 hours, the media was aspirated from the cells and replaced with 100 uL of serum free media containing pseudovirus. The infected cells incubated for 5 minutes at room temperature. Pseudovirus was then aspirated off and washed 1x with PBS (bound), and new prewarmed 2% NBCS media was added to the cells and allowed to sit overnight(input and bound). The following day, the media was aspirated from the cells and the cells were then washed with 1x PBS. Subsequently, cells were then lysed with 75 uL 1x Pierce Lysis Buffer (Promega) for 10 minutes. The lysed samples were then transferred to a white, opaque 96 well plate. 150 uL of Luciferase was added to wells and read with a GlowMax luminometer. Percent bound was calculated by normalizing luciferase values to input.

Genome/PFU ratio.

Cells were infected with the MHV-A59 at an MOI of 0.01. Supernatant was collected 48 hpi. The infectious virus (PFU) was quantified via plaque assay as described. RNA was extracted, reverse-transcribed, and quantified with qPCR as described above.

Statistical Analysis.

Prism 6 (GraphPad) was used to generate graphs and perform statistical analysis. For all analyses, two-tailed Student’s t test was used to compare groups, unless otherwise noted, with a = 0.05. For tests of sample proportions, p values were derived from calculated Z scores with two tails and a = 0.05.