The furin cleavage site is required for pathogenesis, but not transmission of SARS-CoV-2

SARS-CoV-2, furin cleavage site, spike, entry, QTQTN ALM and MNV contributed equally to this work. Author order was determined alphabetically.

Word count: 209 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint

The SARS-CoV-2 spike, key to viral entry, has two features that differentiate it from other sarbecoviruses: the presence of a furin cleavage site (FCS; PRRAR sequence) and an extended S1/S2 loop characterized by an upstream QTQTN amino acid motif. Our prior works show that shortening the S1/S2 loop by deleting either the FCS (ΔPRRA) or deleting an upstream sequence (ΔQTQTN), ablates spike processing, alters host protease usage, and attenuates infection in vitro and in vivo. With the importance of the loop length established, here we evaluated the impact of disrupting the FCS, but preserving the S1/S2 loop length. Using reverse genetics, we generated a SARS-CoV-2 mutant that disrupts the FCS (PQQAR) but maintains its extended S1/S2 loop. The SARS-CoV-2 PQQAR mutant has reduced replication, decreased spike processing, and attenuated disease in vivo compared to wild-type SARS-CoV- 2. These data, similar to the FCS deletion mutant, indicate that loss of the furin cleavage site attenuates SARS-CoV-2 pathogenesis. Importantly, we subsequently found that the PQQAR mutant is transmitted in the direct contact hamster model despite lacking an intact FCS. However, competition transmission showed that the mutant was attenuated compared to WT SARS-CoV-2. Together, the data argue that the FCS is required for SARS-CoV-2 pathogenesis but is not strictly required for viral transmission. .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint

1 SARS-CoV-2 emerged in late 2019 and initiated a global pandemic with a massive impact on 2 the economy and global public health1. Over the past five years, SARS-CoV-2 has continued to 3 cause infection and disease despite development of effective vaccine and therapeutics2. The 4 continued infection and spread of SARS-CoV-2 have been attributed to its ability to evolve and 5 produce variants of concern that evade key aspects of immunity and improve viral fitness3,4. 6 Much of the evolution of the virus has occurred in the spike protein with new variants able to 7 evade host immunity5. In addition, several spike mutations have improved fitness and 8 transmission of the virus in the human populations6–9. Despite all the changes that have 9 occurred in SARS-CoV-2, two key elements of the spike protein have remained intact 10 throughout the pandemic: the presence of a furin cleavage site (FCS) and an extended S1/S2 11 loop10–12. Each are unique features of SARS-CoV-2 compared to other sarbecoviruses and play 12 key roles in infection. 13 The CoV spike protein consists of a globular head (S1 subunit) and a stalk (C-terminal of 14 S1 and S2 subunit)13. The head is responsible for attachment and binding to host receptor 15 ACE2. In contrast, the stalk contains the highly conserved fusion machinery. CoV spike must 16 undergo two cleavages to enable host cell fusion, first at an S1/S2 junction site and a second at 17 an S2’ site in the stalk to activate fusion. Unlike other sarbecoviruses, SARS-CoV-2 spike 18 contains a unique furin cleavage site (FCS; RXXR) at the S1/S2 junction that aids in the 19 efficiency of its cleavage. The FCS, not shared among other group 2B coronaviruses, is 20 maintained in all SARS-CoV-2 variants. The multi-basic cleavage motif, PRRAR, exists on the 21 S1/S2 junction loop and is involved in spike cleavage as a virion exits a producer cell. Our lab 22 has previously investigated the importance of the FCS by generating an FCS deletion SARS-23 CoV-2 mutant (ΔPRRA; Fig.1A) and performing in vitro and in vivo characterization11. Our 24 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint

indicated that deleting the FCS attenuates SARS-CoV-2 infection in respiratory cells and 25 pathogenesis in vivo compared to wild-type (WA-1 strain)11. 26 Soon after, we investigated the function of the QTQTN motif immediately upstream the 27 FCS, which is commonly deleted in cell culture stocks. Our group generated a QTQTN deletion 28 mutant (ΔQTQTN; Fig. 1A) and performed in vitro and in vivo testing12. The ΔQTQTN mutant, 29 like ΔPRRA, has attenuated infection in respiratory cells and pathogenesis in vivo12. Both 30 deletion mutants shorten the spike S1/S2 loop and demonstrate that loop length is important for 31 SARS-CoV-2 pathogenesis. However, these studies do not evaluate importance of the FCS in 32 the context of the extended S1/S2 loop found in SARS-CoV-2. To address this question, we 33 generated an infectious clone of SARS-CoV-2 that disrupts the FCS without shortening the loop 34 (PQQAR mutant). We demonstrate that disruption of the FCS attenuates SARS-CoV-2 35 replication in human respiratory cells and pathogenesis in hamsters. The PQQAR mutant has 36 inhibited spike processing and altered protease usage for host cell entry. We also found that an 37 intact FCS is not required for SARS-CoV-2 contact transmission in hamsters, but plays a role in 38 transmission efficiency. Together, the data indicate that the SARS-CoV-2 FCS is important for 39 SARS-CoV-2 infection, pathogenesis, and transmission. 40 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint

41 Generation of the PQQAR mutant 42 The furin cleavage site (FCS; PRRA motif) of SARS-CoV-2 exists at the S1/S2 junction site on 43 an external disordered loop in the spike protein (Fig. 1B). Our prior work evaluated the 44 importance of the furin cleavage site by deleting the PRRA motif upstream of the S1/S2 45 cleavage site11. While the ΔPRRA mutant was attenuated, subsequent work shortening the 46 S1/S2 loop (ΔQTQTN) of SARS-CoV-2 had a similar phenotype, indicating the importance of 47 the loop length in SARS-CoV-2 pathogenesis12. However, the role of the actual FCS in the 48 context of the extended S1/S2 spike loop was still unclear. To investigate how the FCS, 49 independent of loop length, impacts infection and pathogenesis, we generated an FCS mutant 50 in the SARS-CoV WA-1 backbone that substituted arginine at position 682 and 683 to 51 glutamines (PQQAR) (Fig. 1C). The PQQAR mutant SARS-CoV-2 maintains the S1/S2 loop 52 length but disrupts the furin cleavage site allowing evaluation of the role of the FCS. Using our 53 reverse genetic system, we were able to recover the PQQAR mutant which grew to robust stock 54 titers in Vero cells. 55 PQQAR mutant attenuates viral replication in respiratory cell. 56 Deletion of the FCS and shortening of the S1/S2 loops of the spike had previously been shown 57 to impact viral replication11,12. Therefore, we first examined the PQQAR mutant in Vero E6 cells 58 (African green monkey kidney cells) which lack type I interferon responses14. Following low 59 MOI (0.01) infection, the PQQAR mutant shows no significant attenuation compared to WT 60 SARS-CoV-2 in Vero cells (Fig. 1D). In fact, the PQQAR mutant grew to slightly higher viral 61 titer, although not significantly different that WT. These results are consistent with prior studies 62 shortening the SARS-CoV-2 S1/S2 loop length and correspond with a fitness advantage in Vero 63 cells11,12. In contrast, the PQQAR mutant has attenuated replication in Calu-3 2B4 cells, a 64 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint human respiratory cell line (Fig. 1E). At 24 and 48 hours post infection (hpi), the PQQAR mutant 65 has ∼3 log reduction in titers compared to WT. Again, these results are consistent with prior 66 studies(ΔQTQTN, ΔPRRA)11,12 and demonstrate that the both the S1/S2 loop and the FCS play 67 a role in effective SARS-CoV-2 replication in human respiratory cells. 68 PQQAR mutant has attenuated in vivo pathogenesis 69 Having established attenuation in human respiratory cells, we next evaluated how disruption of 70 the FCS impacts SARS-CoV-2 pathogenesis in vivo. Utilizing the golden Syrian hamster model, 71 three- to four-week-old male hamsters were intranasally infected with 105 focus forming units 72 (FFU) of WT or PQQAR mutant SARS-CoV-2 and followed over a 7-day time course (Fig. 2A). 73 WT-infected hamsters exhibited weight loss starting at 2 dpi with peak weight loss of ∼10-12% 74 before beginning to recover at 6 dpi (Fig. 2B). In contrast, the PQQAR mutant produced 75 minimal weight loss following infection. Similarly, disease score corresponded with weight loss 76 as WT had observable disease between days 3 and 6 characterized by ruffled fur, hunched 77 posture, and reduced activity. In contrast, no disease was noted in PQQAR mutant infected 78 hamsters. Together, the data demonstrated that disruption of the FCS attenuated disease in the 79 hamster model of SARS-CoV-2 infection. 80 PQQAR mutant has altered replication in upper and lower respiratory tract. 81 Having established attenuated disease, we next evaluated viral replication in the lung, nasal 82 wash, and trachea following infection with WT and PQQAR mutant. Examining the lung, we 83 found that both WT and the PQQAR mutant replicated in the lungs following infection (Fig. 2C). 84 However, the PQQAR mutant had a 1 log reduction in viral replication compared to WT at day 2 85 but was equal to WT at day 4. Probing the upper airway, the PQQAR mutant had a significant 86 increase (~10-fold) in viral replication from nasal washes at both days 2 and 4 as compared to 87 WT (Fig. 2D). These results are similar to prior studies that shortened the spike S1/S2 loop 88 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint length or deleted the FCS11,12. In addition, the PQQAR mutant had similar and increased 89 replication in the trachea at days 2 and 4 as compared to WT-infected hamsters (Fig. 2E). 90 Overall, the replication data demonstrate viral replication attenuation of the PQQAR mutant in 91 the lung but also augmented replication of the mutant in the upper airway tissues. 92 Viral antigen staining confirms attenuation of PQQAR mutant in vivo. 93 To further evaluate viral replication and distribution, we examined antigen staining following 94 infection with WT or PQQAR mutant at days 2 and 4 post infection. Briefly, lung sections from 95 infected hamsters were stained for SARS-CoV-2 nucleocapsid and scored in a blinded manner 96 for antigen in the airways, parenchyma, and overall as previously described8. We observed 97 wide-spread antigen staining in both the parenchyma and airways following WT SARS-CoV-2 98 infection at both day 2 and 4 (Fig. 3A-B). In contrast, the PQQAR mutant had more limited 99 antigen staining compared to WT (Fig. 3C-D). While showing modest attenuation in the airways 100 (Fig. 3E), the scores from the parenchyma (Fig. 3F) and overall (Fig. 3G) demonstrated that the 101 PQQAR mutant had a significant deficit in antigen staining as compared to WT SARS-CoV-2 102 infection. The antigen staining in the large airways may be consistent with potentially greater 103 replication in the airways (Fig. 2D-E). Similarly, the attenuation in the parenchyma and overall 104 staining corresponds to reduced viral titer in the lung of PQQAR mutant-infected animals (Fig. 105 2C). Together, these results demonstrate attenuation in PQQAR mutant as compared to WT in 106 terms of viral infection and distribution in the lung. 107 PQQAR mutant infection shows reduced immune infiltration and damage. 108 To further evaluate disease and damage, WT- and PQQAR- infected hamsters were evaluated 109 for changes in histopathology. Lung sections were H&E stained and subsequently evaluated by 110 a board-certified pathologist in a blinded manner. After infection, WT-infected hamsters 111 demonstrated significant disease and damage characterized by interstitial pneumonia, 112 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint peribrochiolitis, epithelial cytopathology, arterial mononuclear cell margination, and presence of 113 polymorphonuclear cells in the bronchioles in some animals at day 2 (Fig. 4A). In contrast, 114 significantly less disease was observed in the PQQAR mutant infected hamsters and was 115 characterized by mild bronchiolitis, and interstitial pneumonia (Fig. 4B). At day 4, WT infected 116 hamsters showed increased lung involvement by area involved with the addition of hemorrhage, 117 pulmonary edema, perivasculitis, and infiltration of mononuclear cells (Fig. 4C). Notably, while 118 increased compared to day 2, PQQAR mutant-infected animals had much less lung involvement 119 and less severe disease and damage as compared to WT (Fig. 4D). While no significant 120 disease and damage was observed in mock infected animals (Fig. 4E), the SARS-CoV-2 121 associated lesion were much more extensive in WT as compared to PQQAR mutant infected 122 hamsters. WT-infected hamsters had significant increases in SARS-CoV-2 lesion percentages 123 in lung section at days 2, 4, and 7 days post infection as compared to PQQAR mutant infection 124 (Fig. 4F). These results are consistent with the weight loss and disease data (Fig. 2B) and 125 indicate that the PQQAR mutant is attenuated in terms of pathogenesis in vivo. 126 FCS disruption attenuates spike processing 127 Spike mediated entry requires sequential cleavage at the S1/S2 junction and the S2’ site to 128 activate fusion15. Prior work has established that the SARS-CoV-2 spike is partially cleaved 129 prior to release from cells16; shortening the S1/S2 loop disrupts this cleavage on released 130 virions. Here, we evaluate if FCS disruption affected spike cleavage on purified virions from the 131 PQQAR mutant. Briefly, WT, PQQAR, and the ΔFCS mutant SARS-CoV-2 were collected from 132 Vero E6 cells, purified using ultracentrifugation sucrose cushion, inactivated, and blotted for 133 spike and nucleocapsid (Fig. 5A). Normalizing to nucleocapsid, we observe that both PQQAR 134 and ΔFCS mutants have a reduction in S1/S2 cleavage product as compared to WT (Fig. 5B). 135 While WT SARS-CoV-2 has ~40% of spike cleaved following infection of Vero cells, both 136 disruption and deletion of the FCS results in mutants with very minimal S1/S2 cleavage product 137 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint (Fig. 5C). Together, the results indicate that an intact FCS is required for spike processing prior 138 to release of SARS-CoV-2 virions. 139 TMPRSS2 plays role in attenuation of PQQAR mutant 140 SARS-CoV-2 enters the host cell through either the endosomal route mediated by host 141 cathepsins or the cell surface route mediated by host serine proteases like TMPRSS2 (Fig. 142 5D)15. The PQQAR mutant shows significant attenuation in Calu-3 2B4, but has no deficit in 143 Vero E6 cells (Fig. 1C, 1D). A major distinction between the two cell lines is expression of 144 TMPRSS2, which is highly expressed in Calu3 cells, but absent in Vero E6 cells. To determine 145 if TMPRSS2 expression contributes to attenuation, we utilize Calu3 cells with TMPRSS2 146 knocked out. We subsequently infected the TMPRSS2 KO Calu3 cells with both WT SARS-147 CoV-2 and PQQAR mutant as previously described. Following infection, we note that both 148 PQQAR and WT virus replication is not significantly different over the course of infection (Fig. 149 5E). Compared to standard Calu3 cells (Fig. 5F), WT SARS-CoV-2 infection was reduced to 150 the same level of PQQAR mutant indicating that the primary difference between replication of 151 the WT and FCS mutant virus in Calu3 cells is due to the activity of TMPRSS2. These results 152 indicate that spike processing through the FCS plays an important role in utilization of 153 TMPRSS2 mediated entry of SARS-CoV-2. 154 Intact FCS is not required for transmission of SARS-CoV-2 in vivo 155 The disruption of the FCS by the PQQAR mutant demonstrates that the importance of the motif 156 to SARS-CoV-2 infection of respiratory cells and pathogenesis in vivo. Yet, it remains unclear if 157 the furin cleavage site is required for transmission of SARS-CoV-2. While attenuated in the 158 lung, the PQQAR mutant displayed robust and augmented replication in the upper airways as 159 compared to WT SARS-CoV-2 (Fig. 2C). With the upper airway infection thought to seed 160 transmission, it left the possibility that the PQQAR mutant could still be transmited without a 161 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint functional FCS. To test this question, we performed studies with WT and PQQAR mutant SARS-162 CoV-2 to evaluate the capacity of the mutant virus to be transmitted in hamsters. Briefly, “donor” 163 hamster were intranasally infected with 105 (FFU) of WT SARS-CoV-2 or PQQAR mutant (Fig. 164 6A). After 24 hours, each donor was individually co-housed 1:1 with a naïve “recipient” with no 165 additional barriers to evaluate contact transmission for 8 hours. After cohousing, donors and 166 recipients were separated into individual cages for the duration of the study. Before being 167 returned to new cages, donor hamsters were nasal washed to measure their shedding titer on 168 the day of exposure. Animals were subsequently examined for changes in weight and disease, 169 euthanized two days post infection/exposure, and tissues/washes evaluated for viral load. 170 Following infection of donors, we observed that the shedding titers in the nasal wash 171 were similar between WT and PQAAR mutant after the exposure period (Fig. 6B). We also 172 observed that donor titers at 2 days post infection showed the PQAAR mutant had higher viral 173 load in the nasal wash, but reduced titer in the lung (Fig. 6C-D). These results are consistent 174 with observations from acute infection showing greater replication of PQQAR mutant in the 175 airways versus the lung (Fig. 2C-D). Examining transmission from donors to recipient, we 176 observed that transmission of both WT and PQQAR mutants was 100% successful. All 5 donor 177 pairs infected with either WT or PQQAR mutant resulted in detectable viral loads in the nasal 178 wash and lungs of recipient animals 2 days post exposure (Fig. 6E-F). Yet, while transmission 179 occurred, PQQAR mutant recipient hamsters had significant lower viral titers compared to WT 180 for both the lung and nasal washes (Fig. 6E-F). The results suggest that while the FCS is not 181 strictly required for transmission, its presence impacts the viral load observed in recipient 182 animals. Regardless of reduced viral load, the results indicate that an intact FCS is not required 183 for SARS-CoV-2 transmission in the direct contact hamster model. 184 FCS necessary for transmission fitness in vivo 185 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint To investigate how loss of the FCS impacts transmission fitness, we performed a competition 186 transmission study between WT and PQQAR mutant virus in hamsters. Briefly, donor hamsters 187 were intranasally infected with an equal mixture (1:1) of WT and PQQAR mutant virus with a 188 total of 105 focus forming units (Fig. 7A). At 24 hours post infection, donor hamsters were 189 cohoused 1:1 with naïve recipients for 12 hours before separation into individual cages. Upon 190 separation, the donor hamsters were nasal washed after exposure. Animals were subsequently 191 examined for weight loss and disease over a four-day time course. Animals were euthanized 192 and RNA from lung tissue/nasal washes were collected from the hamsters at 2- and 4-days 193 post-infection or post-contact. The extended time point (4 dpi) evaluated if the competition 194

changed after an prolonged period of time post infection. 195 We used Next-Generation sequencing (NGS) to evaluate transmission dynamics of WT 196 and PQQAR from in vivo competition. NGS libraries were made from extracted RNA samples 197 with “Tiled-ClickSeq”17, an approach that utilizes ”click chemistry” and >300 SARS-CoV-2 198 specific primers to fully examine the SARS-CoV-2 genome. The abundance of viral RNA 199 comprised of WT or PQQAR genotypes suggests their relative ratio during transmission 200 competition. Examining donor hamsters, we find that PQQAR mutant had substantially reduced 201 transmission efficiency than WT for 1, 2, or 4 dpi. (Fig. 7B-D). From nasal wash, PQQAR 202 mutant comprised ~2-20.1% of total viral reads (mean: 8.63% +/- 0.05%) or a ~12.3-fold 203 transmission reduction compared to WT. From lung tissue, PQQAR mutant comprised ~0.2-204 5.1% of total viral reads (mean: 1.95% +/- 0.02%) or 162.7-fold transmission reduction relative 205 to WT. Our competition results differ from the findings in the first transmission study that found 206 robust replication of the PQQAR mutant in donor animals (Fig. 6B-D). The single viral inoculum 207 experiment (Fig. 6) demonstrates that PQQAR is replication-competent in vivo, while the mixed 208 inoculum experiment (Fig. 7) shows that although PQQAR can replicate, it is outcompeted by 209 WT SARS-CoV-2. 210 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint From recipient hamsters, the PQQAR mutant was found to have even lower 211 transmission efficiency as compared to WT (Fig. 7E-F). Examination of both the lung and nasal 212 washes revealed that the PQQAR mutant was found to be <5% of the reads. Notably, despite 213 the low levels of viral reads, the PQQAR mutant RNA was detected in all recipient animals at 214 levels well above the number of background mutations (Fig. 7G) indicating that the detected 215 PQQAR genotype is a result of actual transmission, rather than random mutations. However, 216 the PQQAR mutant is unable to compete with WT virus that maintains an intact FCS. Together, 217 the results demonstrate that while not strictly required, the SARS-CoV-2 spike FCS site aids in 218 transmission efficiency. 219 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint

220 The presence of the furin cleavage site in SARS-CoV-2 spike protein plays a key role in 221 infection and pathogenesis. As novel variants continue to emerge, SARS-CoV-2 has maintained 222 its furin cleavage site and extended S1/S2 loop length in the spike protein indicating its 223 necessity. While we previously established the importance of the SARS-CoV-2 S1/S2 loop 224 length11,12, this study addresses how the furin cleavage site itself impacts SARS-CoV-2 infection 225 and pathogenesis. We found that disruption of the FCS significantly attenuates replication in 226 human respiratory cells in vitro and pathogenesis in vivo. The FCS mutant (PQQAR) has 227 inhibited spike processing and modified host protease usage during cell entry. Notably, despite 228 attenuated pathogenesis, the PQQAR mutant is successfully transmitted to all contact recipients 229 but has decreased transmission fitness. Together, our data indicate that the furin cleavage site 230 is critical for SARS-CoV-2 infection, pathogenesis, and transmission fitness. 231 Overall, our studies confirm that both the loop length and an intact furin cleavage site are 232 required for pathogenesis of SARS-CoV-2. Previously, deletion mutants, ΔPRRA and ΔQTQTN, 233 both had attenuated pathogeneicity in hamsters; the results showed that the SARS-CoV-2’s 234 extended S1/S2 loop played a critical role in pathogenesis11,12. However, the previous findings 235 failed to evaluate the FCS requirement in the context of an extended S1/S2 loop. In this study, 236 we demonstrate that disruption of the FCS also causes attenuation of in vivo pathogenesis and 237 has viral replication kinetics similar to the ΔPRRA and ΔQTQTN deletion mutants. The PQQAR 238 mutant maintains the extended loop length, but still results in reduced body weight loss and 239 significantly less lung pathology in hamsters. While there is slightly augmented replication in the 240 upper airways, the PQQAR mutant infected animals had lower viral loads in the lung and 241 reduced antigen staining. Together, these data highlight the importance of both an intact FCS 242 and extended S1/S2 loop in SARS-CoV-2 pathogenesis. 243 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint Combined with prior studies, our results indicate that spike processing and protease 244 usage play a key role in SARS-CoV-2 pathogenesis. Disruption of the furin cleavage site 245 (PQQAR) or shortening of the S1/S2 loop (ΔQTQTN) both independently ablate spike 246 processing of SARS-CoV-2 virions11,12. Spike processing changes on the SARS-CoV-2 virion 247 also correspond with changes in protease usage; the FCS and shortened loop mutants were 248 unable to utilize TMPRSS2-mediated pathways as effectively. The changes in spike processing 249 and protease usage also correspond with attenuation of in vivo pathogenesis. The results 250 argue that the S1/S2 loop length and the FCS both play critical roles in SARS-CoV-2 infection 251 and disease. Importantly, a number of other SARS-CoV-2 spike mutations found in variants of 252 concern have been shown to modulate both spike processing and protease usage8,18,19 . In 253 each case, these results impact disease pathogenesis and highlight the importance of both the 254 FCS and S1/S2 loop in SARS-CoV-2 disease and damage. 255 In contrast to its role in pathogenesis, our results indicate that an intact FCS is not 256 strictly required for SARS-CoV-2 transmission but likely plays a role in transmission and 257 infection efficiency. Prior studies in ferrets found that deletion of the furin cleavage site ablated 258 transmission of the virus via direct contact20. In contrast, we found that disrupting the FCS did 259 not prevent transmission of SARS-CoV-2 in the direct contact hamster model with all 5 260 recipients infected. A major distinction is the nature of the FCS mutations as the prior study 261 deleted the entire cleavage loop motif (PRRAR) ablating the S1/S2 site as well as shortening 262 the loop20. Our study maintains the S1/S2 loop length by disrupting only the furin targeted motif 263 and also maintaining the Sarbecovirus cleavage site (PQQAR)7. Our findings indicate that the 264 intact FCS is not a requirement but leaves the possibility that the S1/S2 extended loop may be 265 necessary for efficient transmission. Further studies are necessary to explore the role of the 266 extended S1/S2 loop for SARS-CoV-2 transmission. 267 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint While the PQQAR mutant was able to be passed to 100% of recipient animals, the 268 efficiency of transmission was reduced. Examining viral loads, the PQQAR mutant recipient had 269 lower viral loads as compared to WT controls following single virus transmission study. 270 Similarly, the PQQAR mutant was outcompeted by WT SARS-CoV-2 in direct competition 271 studies (Fig. 7). While the results corresponded with attenuation PQQAR mutant loads in the 272 lung upon direct challenge, surprisingly, the prior observed advantage in the upper airway 273 disappeared in recipient animals. The results suggest that furin site plays a role in efficiency of 274 transmission and infection of recipient animals. Notably, despite clear attenuation relative to 275 WT, the PQQAR mutant virus RNA was observed in every hamster pair, indicating transmission 276 occurred even in the presence of WT SARS-CoV-2. Together, the results indicate that while the 277 FCS is not strictly required, it does aid in transmission efficiency of SARS-CoV-2. 278 Overall, the manuscript data demonstrate that an intact furin cleavage site plays a critical 279 role in SARS-CoV-2 infection, pathogenesis, and transmission efficiency. Disruption of the FCS 280 in the S1/S2 loop results in ablated spike processing, altered host protease usage, and reduced 281 disease in vivo. Importantly, while not strictly required for transmission, the FCS plays a critical 282 role in efficiency of infection in recipient hamsters. Together, these results demonstrate the 283 critical importance of an intact FCS in SARS-CoV-2 infection, pathogenesis, and transmission. 284 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint Figure Legends 285 Figure 1. Generation and in vitro characterization of SARS-CoV-2 PQQAR mutant. (A) 286 Alignment of the S1/S2 cleavage site of SARS-CoV-2 WA1 and series of mutant viruses 287 generated for evaluation including deletion of the furin cleavage site (ΔFCS), truncation of the 288 extended loop (ΔQTQTN), and disruption of the furin cleavage site motif (PQQAR). (B) SARS-289 CoV-2 spike trimer structure (gray) highlighting the S1/S2 cleavage loop. WT (left) and PQQAR 290 mutant (right) are zoomed with mutated residues (Q682, Q683) in orange to disrupt the furin 291 cleavage site. (C) Schematic of SARS-CoV-2 spike with PQQAR substitutions identified. (D) 292 Viral titer from Vero E6 infected with WT (black) or PQQAR (orange) SARS-CoV-2 at an MOI of 293 0.01 (n = 3). (E) Viral titer from Calu-3 2B4 infected with WT or PQQAR SARS-CoV-2 at an MOI 294 of 0.01 (n = 3). Data are mean ± SD. Statistical analysis measured by two-tailed Student’s t test. 295 *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001 296 Figure 2. SARS-CoV-2 PQQAR mutant attenuated in golden Syrian hamsters. (A-B) 297 Schematic of golden Syrian hamster infection with WT (black) or PQQAR mutant (orange) 298 SARS-CoV-2. Three- to four-week-old male hamsters were infected with 105 pfu and monitored 299 (B) weight loss and disease for 7 days post infection. (C-E) Viral titers were measured at days 2 300 and 4 from (C) infected lung, (D) nasal wash, and (E). Data are representative of mean ± SEM. 301 Statistical analysis measured by two-tailed Student’s t test. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001. 302 Experimental schematic made in Biorender. 303 Figure 3. Attenuated antigen staining in PQQAR-infected hamsters. (A-D) Nucleocapsid 304 antigen staining of left lung section from hamsters infected with 105 ffu of either (A-B) WT or (C-305 D) PQQAR mutant at 2 or 4 dpi. Antigen staining was scored in a blinded manner by location in 306 the (E) airway, (F) parenchyma, and (G) total for WT (black) or PQQAR (orange) infected lungs. 307 Statistical analysis measured by two-tailed Student’s t test. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001 308 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint Figure 4. Reduced inflammation and damage in PQQAR-infected lungs. (A-G) 309 Representative H&E staining of left lung of hamsters infected with 105 pfu of either WT or 310 PQQAR SARS-CoV-2 at (A-B) 2 days , (C-D) 4 days, or (E,F) days post infection or (G) mock. 311 (H) WT (black), PQQAR R(orange), or PBS (grey) lung sections from each day were scored for 312 histopathological analysis with sections from individual animal averaged and representing a 313 single point. Statistical analysis measured by two-tailed Student’s t test. *P ≤ 0.05; **P ≤ 0.01; 314 ***P ≤ 0.001 315 Figure 5. Disruption of FCS alters spike processing and protease usage. (A) Schematic of 316 SARS-Cov-2 virion sucrose cushion purification approach. (B) Lysates from sucrose cushion 317 purified WT, PQQAR, and ΔFCS virions grown in Vero E6 were probed with α-Spike and α-318 Nucleocapsid (N) antibodies by Western blot. Full-length spike (FL) and S1/S2 cleavage product 319 are indicated. (C) Quantification of densitometry of the proportion between FL (black) and S1/S2 320 (red) of the total spike shown (lower). (D) Schematic of SARS-CoV-2 entry and protease usage 321 including knockout of TMPRSS2 mediated entry. (E) Viral titer from Calu-3 TMPRSS2 knock-out 322 cells infected with WT (black) or PQQAR (orange) SARS-CoV-2 at an MOI of 0.01 (n = 3). (F) 323 Viral titer at 48hpi from Calu3 WT (Fig. 1D) and Calu3 TMPRSS2-/- cells. Statistical analysis 324 measured by two-tailed Student’s t test. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001. Entry schematic 325 made in Biorender. 326 Figure 6. The FCS is not required for SARS-CoV-2 transmission. (A) Schematic of 327 transmission experiment in golden Syrian hamsters. Three- to four-week-old male donor 328 hamsters were intranasally infected with 105 pfu of WT or PQQAR SARS-CoV-2 and individually 329 housed. Donors were subsequently paired 1:1 with recipients 24 hpi and cohoused for 8 hours 330 before separating and nasal washing donors. (B-F)Nasal washes and lungs were collected at 2 331 days post infection for donors (dpi) (B-D) and post contact for recipients (E-F). Viral titers were 332 measured using focus forming assays for donor and recipient samples . Statistical analysis 333 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint measured by two-tailed Student’s t test. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001. Experimental 334 schematic made in Biorender. 335 Figure 7. The furin cleavage site impacts SARS-CoV-2 transmission efficiency. 336 (A) Schematic of transmission competition experiment in golden Syrian hamsters. Three- to 337 four-week-old male donor hamsters were intranasally infected with 105 pfu of WT:PQQAR 338 SARS-CoV-2 in a 1:1 ratio and were individually housed. (B-F) After 24 hpi, donors were paired 339 with recipients and cohoused for 12 hours before separating and nasal washing donors. Nasal 340 washes and lungs were collected at 2 and 4 days post infection for donors (dpi) and post 341 contact for recipients (dpc). Next generation sequencing was performed on extracted RNA to 342 measure the percentage of WT (grey) and PQQAR (orange) present in nasal wash and lung of 343 donors (B-D) and recipients (E-F). The expected distribution (B-F, top bar) based on NGS 344 percentage mutant/WT observed in the inoculating dose (two inoculum preparations with RNA 345 sequenced twice from each). Statistical analysis measured by two-tailed Student’s t test. *P ≤ 346 0.05; **P ≤ 0.01; ***P ≤ 0.001. Experimental schematic made in Biorender. 347 348 349 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint

350 Cells 351 Vero E6 cells and Vero E6 cells expressing TMPRSS2 (Sekisui XenoTech) were grown in 352 Dulbecco modified Eagle medium (DMEM; Gibco #11965–092) supplemented with 10% fetal 353 bovine serum (FBS) (HyClone #SH30071.03) and 1% antibiotic-antimycotic (Gibco #5240062). 354 Calu-3 2B4 cells were grown in DMEM supplemented with 10% FBS, 1% antibiotic-antimycotic, 355 and 1 mg/mL sodium pyruvate. Human TMPRSS2 knockout Calu-3 cells (Abcam #273734) 356 were grown in DMEM supplemented with 20% defined FBS (HyClone #SH30070.03), 1% 357 antibiotic-antimycotic, 1% non-essential amino acid solution (NEAA; Gibco #11140050), and 1 358 mg/mL sodium pyruvate. 359 Viruses 360 The recombinant WT and mutant SARS-CoV-2 virus sequences are based on the USA-361 WA1/2020 isolate sequence provided by the World Reference Center for Emerging Viruses and 362 Arboviruses (WRCEVA), which was originally obtained from the US Centers for Disease Control 363 and Prevention (CDC)21. Wild-type and mutant SARS-CoV-2 were generated using standard 364 cloning techniques and reverse genetics system as previously described22,23 and propagated on 365 Vero E6 cells. The mutations in PQQAR mutant have been verified by Sanger sequencing. The 366 recovered mutant virus was further sequenced with NGS to confirm the maintenance of 367 nucleotide mutations up to P2. 368 Infectious titers were measured by focus-forming assay. All experiments involving infectious 369 virus were conducted at the University of Texas Medical Branch (UTMB) in an approved 370 biosafety level (BSL) 3 laboratory with routine medical monitoring of staff. 371 In Vitro Infection 372 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint Viral infections in Vero E6, TMRPSS2-expressing Vero E6, Calu-3 2B4, and Calu-3 TMPRSS2 373 KO cells were performed as previously described22,24. Briefly, cells were washed with phosphate 374 buffered saline (PBS) and infected with WT or mutant SARS-CoV-2 at an MOI of 0.01 for 45 min 375 at 37 °C. Following absorption, cells were washed three times with PBS and fresh growth media 376 was added to represent time 0. Three or more biological replicates were collected at each time 377 point and each experiment was repeated at least twice. 378 Focus Forming Assay 379 Focus forming assays (FFAs) were performed as previously described25. Briefly, Vero E6 cells 380 were seeded in 96-well plates to be 100% confluent. Samples were 10-fold serially diluted in 381 serum-free media and 20 µl was to infect cells. Cells were incubated for 45 min at 37°C with 5% 382 CO2 before 100 µl of 0.85% methylcellulose overlay was added. Cells were incubated for 24 hrs 383 at 37°C with 5% CO2. After incubation, overlay was removed, and cells were washed three 384 times with PBS before fixed and virus inactivated by 10% formalin for 30 min at room 385 temperature. Cells were then permeabilized and blocked with 0.1% saponin/0.1% BSA in PBS 386 before incubated with α-SARS-CoV-2 Nucleocapsid primary antibody (Cell Signaling 387 Technology) at 1:1000 in permeabilization/blocking buffer overnight at 4°C. Cells are then 388 washed three times with PBS before incubated with Alexa FluorTM 555-conjugated α-mouse 389 secondary antibody (Invitrogen #A28180) at 1:2000 in permeabilization/blocking buffer for 1 h at 390 room temperature. Cells were washed three times with PBS. Fluorescent foci images were 391 captured using a Cytation 7 cell imaging multi-mode reader (BioTek), and foci were counted 392 manually. 393 Virion Purification and Western Blotting. 394 Vero E6 cells were infected with WT or mutant SARS-CoV-2 at an MOI of 0.01. Supernatant 395 was harvested 24 hpi and clarified by low-speed centrifugation. Virus particles were then 396 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint pelleted by ultracentrifugation through a 20% sucrose cushion at 26,000 rpm for 3 h using a 397 Beckman SW28 rotor. Pellets were resuspended in 2× Laemmli buffer to obtain protein lysates. 398 Relative viral protein levels were determined by sodium dodecyl sulfate-polyacrylamide gel 399 electrophoresis (SDS-PAGE) followed by Western blot analysis as previously described11,12. In 400 brief, sucrose-purified WT and mutant SARS-CoV-2 virions were inactivated by boiling in 401 Laemmeli buffer. Samples were loaded in equal volumes into 4 to 20% Mini-PROTEAN TGX 402 Gels (Bio-Rad #4561093) and electrophoresed by SDS–PAGE. Protein was transferred to 403 polyvinylidene difluoride (PVDF) membranes. Membranes were probed with SARS-CoV S-404 specific antibodies (Novus Biologicals #NB100-56578) and followed with horseradish 405 peroxidase (HRP)-conjugated anti-rabbit antibody (Cell Signaling Technology #7074). 406 Membranes were stripped and reprobed with SARS-CoV N-specific antibodies (Novus 407 Biologicals #NB100-56576) and the HRP-conjugated anti-rabbit secondary IgG to measure 408 loading. Signal developed using Clarity Western ECL substrate (Bio-Rad #1705060) or Clarity 409 Max Western ECL substrate (Bio-Rad #1705062) and imaging on a ChemiDoc MP System (Bio-410 Rad #12003154). Densitometry was performed using ImageLab 6.0.1 (Bio-Rad #2012931). 411 Hamster Infection Study 412 Male golden Syrian hamsters (3 to 4 weeks old) were purchased from Envigo (HsdHan:AURA 413 strain). All studies were conducted under a protocol approved by the UTMB Institutional Animal 414 Care and Use Committee and complied with USDA guidelines in a laboratory accredited by the 415 Association for Assessment and Accreditation of Laboratory Animal Care. Procedures involving 416 infectious SARS-CoV-2 were performed in the Galveston National Laboratory ABSL3 facility. 417 Hamsters were intranasally inoculated with 105 pfu of WT or PQQAR SARS-CoV-2 in100 µl. 418 Infected hamsters were weighed and monitored for illness over 7 days. Hamsters were 419 anesthetized with isoflurane (Henry Schein Animal Health) and nasal washes were collected 420 with 400 µl of PBS on endpoint days (2, 4, and 7 dpi). Hamsters were euthanized by CO2 for 421 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint organ collection. Nasal wash and lung were collected to measure viral titer. Left lungs were 422 collected for histopathology. 423 Hamster Transmission 424 Male golden Syrian hamsters (3-4 weeks old) were purchased from Envigo. Each group (WT, 425 PQQAR, or PBS) had 10 donor hamsters that were intranasally infected with 100 uL of 105 pfu 426 of virus or PBS depending on the group. 24 hrs post infection, donor hamsters were cohoused 427 1:1 with a recipient hamster for 8 hrs for contact transmission. After 8 hrs, hamster pairs were 428 separated into individual housing and the donors were nasal washed. At 2 days post infection 429 for donors and post contact for recipients, hamsters were nasal washed with 400 µl of PBS and 430 euthanized for lung and nasal wash collection. Nasal washes and lungs were processed in 431 TRIzol, and RNA was extracted to perform next generation sequencing as previously described. 432 Hamster Transmission Competition 433 Male golden Syrian hamsters (3-4 weeks old) were purchased from Envigo. ten donor hamsters 434 were intranasally infected with a 1:1 ratio of WT:PQQAR SARS-CoV-2 totaling 105 pfu in 100 µl 435 and were subsequently individually housed. At 24 hrs post infection, donor hamsters were 436 cohoused 1:1 with a recipient hamster for 12 hrs for contact transmission. After 8 hrs, hamster 437 pairs were separated into individual housing and the donors were nasal washed for D1 values. 438 At 2 and 4 days post infection for donors and post contact for recipients, hamsters were nasal 439 washed with 400 µl of PBS and euthanized for lung and nasal wash collection. Nasal washes 440 and lungs were processed in TRIzol. RNA was extracted to perform next generation 441 sequencing. 442 Histology 443 Histopathology was performed as previously described26,27. Briefly, left lungs were harvested 444 from hamsters and fixed in 10% buffered formalin solution for at least 7 d. Fixed tissue was then 445 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint embedded in paraffin, cut into 5 µM sections, and stained with hematoxylin and eosin on a 446 SAKURA VIP6 processor by the University of Texas Medical Branch Surgical Pathology 447 Laboratory. H&E staining was performed by the University of Texas Medical Branch Histology 448 Laboratory and then analyzed and scored by a blinded pathologist. 449 Immunohistochemistry 450 Antigen staining was performed as previously described28, Briefly, fixed and paraffin-embedded 451 left lung lobes from hamsters were cut into 5 µM sections and mounted onto slides by the 452 University of Texas Medical Branch Surgical Pathology Laboratory. Paraffin-embedded sections 453 were warmed at 56°C for 10 min, deparaffinized with xylene (3x 5-min washes) and graded 454 ethanol (3x 100% 5-min washes, 1x 95% 5-min wash), and rehydrated in distilled water. After 455 rehydration, antigen retrieval was performed by steaming slides in antigen retrieval solution (10 456 mM sodium citrate, 0.05% Tween-20, pH 6) for 40 min (boil antigen retrieval solution in 457 microwave, add slides to boiling solution, and incubate in steamer). After cooling and rinsing in 458 distilled water, endogenous peroxidases were quenched by incubating slides in TBS with 0.3% 459 H2O2 for 15 min followed by 2x 5-min washes in 0.05% TBST. Sections were blocked with 10% 460 normal goat serum in BSA diluent (1% BSA in 0.05% TBST) for 30 min at room temperature. 461 Sections were incubated with primary anti-N antibody (Sino #40143-R001) at 1:1000 in BSA 462 diluent overnight at 4°C. Following overnight primary antibody incubation, sections were washed 463 3x for 5 min in TBST. Sections were incubated in secondary HRP-conjugated anti-rabbit 464 antibody (Cell Signaling Technology #7074) at 1:200 in BSA diluent for 1 hour at room 465 temperature. Following secondary antibody incubation, sections were washed 3x for 5 min in 466 TBST. To visualize antigen, sections were incubated in ImmPACT NovaRED (Vector 467 Laboratories #SK-4805) for 3 min at room temperature before rinsed with TBST to stop the 468 reaction followed by 1x 5-min wash in distilled water. Sections were incubated in hematoxylin for 469 5 min at room temperature to counterstain before rinsing in water to stop the reaction. Sections 470 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint were dehydrated by incubating in the previous xylene and graded ethanol baths in reverse order 471 before mounted with coverslips. 472 Structural Modeling 473 Structural models previously generated were used as a base to visualize residues mutated in 474 Omicron12. Briefly, structural models were generated using SWISS-Model to generate homology 475 models for WT and PQQAR SARS-CoV-2 spike protein based on the SARS-CoV-1 trimer 476 structure (Protein Data Bank code 6ACD). Homology models were visualized and manipulated 477 in PyMOL (version 2.5.4) to visualize the PQQAR mutation. 478 Next Generation Sequencing and data analysis 479 Next generation sequencing (NGS) method was used to determine viral RNA populations from 480 infected animals. Briefly, total cellular RNA samples were extracted from animal tissues and 481 NGS libraries were prepared with Tiled-ClickSeq method17,25. A modified pre-RT annealing 482 protocol was applied as previously described25 to reduce mis-priming. The final libraries 483 comprising of 300–600 bps fragments were pooled and sequenced on an ElementBio Aviti 484 platform with paired-end sequencing (120 bp R1 and 30 bp R2). The raw Illumina data of the 485 Tiled-ClickSeq libraries were processed with established bioinformatics pipelines 486 (https://github.com/andrewrouth/TCS). The relative ratio between WT and PQQAR was 487 calculated based on the average Pilon-reported29 G-to-A mutation rate at specific genomic loci 488 (G23607A, G23608A, G23610A, G23611A). A baseline mutation rate is also evaluated by 489 averaging the background mutation rates from nts. 23606-23612. 490

491 Special thanks to UTMB next generation sequencing core staff (Haiping Hao, Jill K. Thompson) 492 for next-generation sequencing support. 493 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint Funding: 494 Research was supported by grants from NIAID of the NIH to (AI168232 and AI153602 to VDM; 495 R24-AI120942 (WRCEVA) to KSP). Research was also supported by STARs Award provided by 496 the University of Texas System to VDM and Data Acquisition award provided by the Institute for 497 Human Infections and Immunity at UTMB to MNV. Trainee funding provided by NIAID of the NIH 498 to MNV (T32-AI060549). ZY was supported by an Institute of Human Infection and Immunity at 499 UTMB COVID-19 Research Fund. Research was also supported by Burroughs Wellcome Fund 500 Investigators in Pathogenesis to VDM. 501 Competing Interests 502 VDM has filed a patent on the reverse genetic system for SARS-CoV-2. All other authors declare 503 no conflicts of interest. Other authors declare no competing interests. 504 Author contributions 505 Conceptualization: ALM, MNV, VDM 506 Formal analysis: MNV, ALM, YZ, VDM 507 Funding acquisition: MNV, MSS, BAJ, KSP, VDM 508 Investigation: ALM, MNV, YZ, WMM, REA, KGL, YPA, LKE, JAP, DHW, KSP, VDM 509 Methodology: ALM, MNV, YZ, KGL, VDM 510 Project Administration: MNV, ALM, VDM 511 Supervision: MSS, DHW, KSP, VDM 512 Visualization: ALM, MNV, DHW, VDM 513 Writing – original draft: ALM, VDM 514 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint Writing – review and editing: ALM, MNV, KGL, DHW, VDM 515 516 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint

517 1 Gralinski LE, Menachery VD. Return of the Coronavirus: 2019-nCoV . Viruses 2020;12:135. 518 https://doi.org/10.3390/v12020135. 519 2 Amanat F , Krammer F . SARS-CoV-2 Vaccines: Status Report. Immunity 2020;52:583–9. 520 https://doi.org/10.1016/j.immuni.2020.03.007. 521 3 Plante JA, Mitchell BM, Plante KS, Debbink K, Weaver SC, Menachery VD. The variant gambit: 522 COVID-19’s next move. Cell Host Microbe 2021;29:508–15. 523 https://doi.org/10.1016/j.chom.2021.02.020. 524 4 DeGrace MM, Ghedin E, Frieman MB, Krammer F , Grifoni A, Alisoltani A, et al. Defining the risk of 525 SARS-CoV-2 variants on immune protection. Nature 2022;605:640–52. 526 https://doi.org/10.1038/s41586-022-04690-5. 527 5 Tomalka JA , Suthar MS, Deeks SG, Sekaly RP. Fighting the SARS-CoV-2 pandemic requires a 528 global approach to understanding the heterogeneity of vaccine responses. Nat Immunol 529 2022;23:360–70. https://doi.org/10.1038/s41590-022-01130-4. 530 6 Plante JA, Liu Y, Liu J, Xia H, Johnson BA, Lokugamage KG, et al. Spike mutation D614G alters 531 SARS-CoV-2 fitness. Nature 2021;592:116–21. https://doi.org/10.1038/s41586-020-2895-3. 532 7 Liu Y , Liu J, Plante KS, Plante JA, Xie X, Zhang X, et al. The N501Y spike substitution enhances 533 SARS-CoV-2 infection and transmission. Nature 2022;602:294–9. 534 https://doi.org/10.1038/s41586-021-04245-0. 535 8 Vu MN, Alvarado RE, Morris DR, Lokugamage KG, Zhou Y, Morgan AL, et al. Loss-of-function 536 mutation in Omicron variants reduces spike protein expression and attenuates SARS-CoV-2 537 infection. BioRxiv Prepr Serv Biol 2023:2023.04.17.536926. 538 https://doi.org/10.1101/2023.04.17.536926. 539 9 Yajima H, Nomai T, Okumura K, Maenaka K, Genotype to Phenotype Japan (G2P-Japan) 540 Consortium, Ito J, et al. Molecular and structural insights into SARS-CoV-2 evolution: from BA.2 541 to XBB subvariants. mBio 2024;15:e0322023. https://doi.org/10.1128/mbio.03220-23. 542 10 Coutard B, Valle C, de Lamballerie X, Canard B, Seidah NG, Decroly E. The spike glycoprotein of 543 the new coronavirus 2019-nCoV contains a furin-like cleavage site absent in CoV of the same 544 clade. Antiviral Res 2020;176:104742. https://doi.org/10.1016/j.antiviral.2020.104742. 545 11 Johnson BA, Xie X, Bailey AL, Kalveram B, Lokugamage KG, Muruato A, et al. Loss of furin 546 cleavage site attenuates SARS-CoV-2 pathogenesis. Nature 2021;591:293–9. 547 https://doi.org/10.1038/s41586-021-03237-4. 548 12 Vu MN, Lokugamage KG, Plante JA, Scharton D, Bailey AO, Sotcheh S, et al. QTQTN motif 549 upstream of the furin-cleavage site plays a key role in SARS-CoV-2 infection and pathogenesis. 550 Proc Natl Acad Sci 2022;119:e2205690119. https://doi.org/10.1073/pnas.2205690119. 551 13 Vu MN, Menachery VD. Binding and entering: COVID finds a new home. PLOS Pathog 552 2021;17:e1009857. https://doi.org/10.1371/journal.ppat.1009857. 553 14 Desmyter J, Melnick JL, Rawls WE. Defectiveness of interferon production and of rubella virus 554 interference in a line of African green monkey kidney cells (Vero). J Virol 1968;2:955–61. 555 https://doi.org/10.1128/JVI.2.10.955-961.1968. 556 15 Jackson CB, Farzan M, Chen B, Choe H. Mechanisms of SARS-CoV-2 entry into cells. Nat Rev Mol 557 Cell Biol 2022;23:3–20. https://doi.org/10.1038/s41580-021-00418-x. 558 16 Shang J, Wan Y , Luo C, Ye G, Geng Q, Auerbach A, et al. Cell entry mechanisms of SARS-CoV-2. 559 Proc Natl Acad Sci 2020;117:11727–34. https://doi.org/10.1073/pnas.2003138117. 560 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint 17 Jaworski E, Langsjoen RM, Mitchell B, Judy B, Newman P , Plante JA, et al. Tiled-ClickSeq for 561 targeted sequencing of complete coronavirus genomes with simultaneous capture of RNA 562 recombination and minority variants. eLife 2021;10:e68479. https://doi.org/10.7554/eLife.68479. 563 18 Saito A, Irie T, Suzuki R, Maemura T, Nasser H, Uriu K, et al. Enhanced fusogenicity and 564 pathogenicity of SARS-CoV-2 Delta P681R mutation. Nature 2022;602:300–6. 565 https://doi.org/10.1038/s41586-021-04266-9. 566 19 Liu Y , Liu J, Johnson BA, Xia H, Ku Z, Schindewolf C, et al. Delta spike P681R mutation enhances 567 SARS-CoV-2 fitness over Alpha variant. Cell Rep 2022;39:110829. 568 https://doi.org/10.1016/j.celrep.2022.110829. 569 20 Peacock TP , Goldhill DH, Zhou J, Baillon L, Frise R, Swann OC, et al. The furin cleavage site in the 570 SARS-CoV-2 spike protein is required for transmission in ferrets. Nat Microbiol 2021;6:899–909. 571 https://doi.org/10.1038/s41564-021-00908-w. 572 21 Harcourt J, Tamin A, Lu X, Kamili S, Sakthivel SK, Murray J, et al. Severe Acute Respiratory 573 Syndrome Coronavirus 2 from Patient with Coronavirus Disease, United States. Emerg Infect Dis 574 2020;26:1266–73. https://doi.org/10.3201/eid2606.200516. 575 22 Xie X, Muruato A, Lokugamage KG, Narayanan K, Zhang X, Zou J, et al. An Infectious cDNA Clone 576 of SARS-CoV-2. Cell Host Microbe 2020;27:841-848.e3. 577 https://doi.org/10.1016/j.chom.2020.04.004. 578 23 Xie X, Lokugamage KG, Zhang X, Vu MN, Muruato AE, Menachery VD, et al. Engineering SARS-579 CoV-2 using a reverse genetic system. Nat Protoc 2021;16:1761–84. 580 https://doi.org/10.1038/s41596-021-00491-8. 581 24 Josset L, Menachery VD, Gralinski LE, Agnihothram S, Sova P , Carter VS, et al. Cell Host 582 Response to Infection with Novel Human Coronavirus EMC Predicts Potential Antivirals and 583 Important Diherences with SARS Coronavirus. mBio 2013;4:10.1128/mbio.00165-13. 584 https://doi.org/10.1128/mbio.00165-13. 585 25 Johnson BA, Zhou Y , Lokugamage KG, Vu MN, Bopp N, Crocquet-Valdes PA, et al. Nucleocapsid 586 mutations in SARS-CoV-2 augment replication and pathogenesis. PLoS Pathog 587 2022;18:e1010627. https://doi.org/10.1371/journal.ppat.1010627. 588 26 Garvanska DH, Alvarado RE, Mundt FO, Lindqvist R, Duel JK, Coscia F , et al. The NSP3 protein of 589 SARS-CoV-2 binds fragile X mental retardation proteins to disrupt UBAP2L interactions. EMBO 590 Rep 2024;25:902–26. https://doi.org/10.1038/s44319-023-00043-z. 591 27 Yang Z, Johnson BA, Meliopoulos VA, Ju X, Zhang P, Hughes MP, et al. Interaction between host 592 G3BP and viral nucleocapsid protein regulates SARS-CoV-2 replication and pathogenicity. Cell 593 Rep 2024;43:113965. https://doi.org/10.1016/j.celrep.2024.113965. 594 28 Schindewolf C, Lokugamage K, Vu MN, Johnson BA, Scharton D, Plante JA, et al. SARS-CoV-2 595 Uses Nonstructural Protein 16 To Evade Restriction by IFIT1 and IFIT3. J Virol 2023;97:e01532-22. 596 https://doi.org/10.1128/jvi.01532-22. 597 29 Walker BJ, Abeel T, Shea T, Priest M, Abouelliel A, Sakthikumar S, et al. Pilon: an integrated tool 598 for comprehensive microbial variant detection and genome assembly improvement. PloS One 599 2014;9:e112963. https://doi.org/10.1371/journal.pone.0112963. 600 601 602 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint 603 604 Figure 1. Generation and in vitro characterization of SARS-CoV-2 PQQAR mutant. (A) Alignment of 605 the S1/S2 cleavage site of SARS-CoV-2 WA1 and series of mutant viruses generated for evaluation 606 including deletion of the furin cleavage site (ΔFCS), truncation of the extended loop (ΔQTQTN), and 607 disruption of the furin cleavage site motif (PQQAR). (B) SARS-CoV-2 spike trimer structure (gray, PDB 608 6ACD) highlighting the S1/S2 cleavage loop. WT (left) and PQQAR mutant (right) are zoomed with 609 mutated residues (Q682, Q683) in orange to disrupt the furin cleavage site. (C) Schematic of SARS-CoV-610 2 spike with PQQAR substitutions identified. (D) Viral titer from Vero E6 infected with WT (black) or 611 PQQAR (orange) SARS-CoV-2 at an MOI of 0.01 (n = 3). (E) Viral titer from Calu-3 2B4 infected with WT 612 or PQQAR SARS-CoV-2 at an MOI of 0.01 (n = 3). Data are mean ± SD. Statistical analysis measured by 613 two-tailed Student’s t test. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001. 614 615 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint 616 Figure 2. SARS-CoV-2 PQQAR mutant attenuated in golden Syrian hamsters. (A-B) Schematic of 617 golden Syrian hamster infection with WT (black) or PQQAR mutant (orange) SARS-CoV-2. Three- to 618 four-week-old male hamsters were infected with 105 pfu and monitored (B) weight loss and disease for 7 619 days post infection. (C-E) Viral titers were measured at days 2 and 4 from (C) infected lung, (D) nasal 620 wash, and (E). Data are representative of mean ± SEM. Statistical analysis measured by two-tailed 621 Student’s t test. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001. Experimental schematic made in Biorender. 622 623 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint 624 Figure 3. Reduced antigen staining in PQQAR-infected hamsters. (A-D) Nucleocapsid antigen 625 staining of left lung section from hamsters infected with 105 ffu of either (A-B) WT or (C-D) PQQAR 626 mutant at 2 or 4 dpi. Antigen staining was scored in a blinded manner by location in the (E) airway, (F) 627 parenchyma, and (G) total for WT (black) or PQQAR (orange) infected lungs. Statistical analysis 628 measured by two-tailed Student’s t test. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001 629 630 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint 631 Figure 4. Reduced inflammation and damage in PQQAR-infected lungs. (A-G) Representative H&E 632 staining of left lung of hamsters infected with 105 pfu of either WT or PQQAR SARS-CoV-2 at (A-B) 2 633 days , (C-D) 4 days, or (E,F) days post infection or (G) mock. (H) WT (black), PQQAR (orange), or PBS 634 (grey) lung sections from each day were scored for histopathological analysis with sections from 635 individual animal averaged and representing a single point. Statistical analysis measured by two-tailed 636 Student’s t test. *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001 637 638 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint 639 Figure 5. Disruption of FCS alters spike processing and protease usage. (A) Schematic of SARS-640 Cov-2 virion sucrose cushion purification approach. (B) Lysates from sucrose cushion purified WT, 641 PQQAR, and ΔFCS virions grown in Vero E6 were probed with α-Spike and α-Nucleocapsid (N) 642 antibodies by Western blot. Full-length spike (FL) and S1/S2 cleavage product are indicated. (C) 643 Quantification of densitometry of the proportion between FL (black) and S1/S2 (red) of the total spike 644 shown (lower). (D) Schematic of SARS-CoV-2 entry and protease usage including knockout of TMPRSS2 645 mediated entry. (E) Viral titer from Calu-3 TMPRSS2 knock-out cells infected with WT (black) or PQQAR 646 (orange) SARS-CoV-2 at an MOI of 0.01 (n = 3). (F) Viral titer at 48hpi from Calu3 WT (Fig. 1D) and 647 Calu3 TMPRSS2-/- cells. Statistical analysis measured by two-tailed Student’s t test. *P ≤ 0.05; **P ≤ 648 0.01; ***P ≤ 0.001. Entry schematic made in Biorender. 649 650 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint 651 Figure 6. The furin cleavage site is not required for SARS-CoV-2 transmission. (A) Schematic of 652 transmission experiment in golden Syrian hamsters. Three- to four-week-old male donor hamsters were 653 intranasally infected with 105 pfu of WT or PQQAR SARS-CoV-2 and individually housed. Donors were 654 subsequently paired 1:1 with recipients 24 hpi and cohoused for 8 hours before separating and nasal 655 washing donors. (B-F)Nasal washes and lungs were collected at 2 days post infection for donors (dpi) (B-656 D) and post contact for recipients (E-F). Viral titers were measured using focus forming assays for donor 657 and recipient samples . Statistical analysis measured by two-tailed Student’s t test. *P ≤ 0.05; **P ≤ 0.01; 658 ***P ≤ 0.001. Experimental schematic made in Biorender. 659 660 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint 661 Figure 7. The furin cleavage site impacts SARS-CoV-2 transmission efficiency. 662 (A) Schematic of transmission competition experiment in golden Syrian hamsters. Three- to 663 four-week-old male donor hamsters were intranasally infected with 105 pfu of WT:PQQAR 664 SARS-CoV-2 in a 1:1 ratio and were individually housed. (B-F) After 24 hpi, donors were paired 665 with recipients and cohoused for 12 hours before separating and nasal washing donors. Nasal 666 washes and lungs were collected at 2 and 4 days post infection for donors (dpi) and post 667 contact for recipients (dpc). Next generation sequencing was performed on extracted RNA to 668 measure the percentage of WT (grey) and PQQAR (orange) present in nasal wash and lung of 669 donors (B-D) and recipients (E-F). The expected distribution (B-F, top bar) based on NGS 670 percentage mutant/WT observed in the inoculating dose (two inoculum preparations with RNA 671 sequenced twice from each). (G) Ratios of PQQAR mutations (orange) and background random 672 mutations (Black). Statistical analysis measured by two-tailed Student’s t test. *P ≤ 0.05; **P ≤ 673 0.01; ***P ≤ 0.001. Experimental schematic made in Biorender. 674 .CC-BY-ND 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted March 11, 2025. ; https://doi.org/10.1101/2025.03.10.642264doi: bioRxiv preprint