key: cord-0914117-4tnhcgow
authors: Thakur, Nazia; Gallo, Giulia; Newman, Joseph; Peacock, Thomas P.; Biasetti, Luca; Hall, Catherine N.; Wright, Edward; Barclay, Wendy; Bailey, Dalan
title: SARS-CoV-2 variants of concern Alpha, Beta, Gamma and Delta have extended ACE2 receptor host-ranges
date: 2021-11-24
journal: bioRxiv
DOI: 10.1101/2021.11.23.469663
sha: 9ce5d3e3db6f51a35357fae82f249c2f9626120f
doc_id: 914117
cord_uid: 4tnhcgow

Following the emergence of SARS-CoV-2 in China in late 2019 a number of variants have emerged, with two of these – Alpha and Delta – subsequently growing to global prevalence. One characteristic of these variants are changes within the Spike protein, in particular the receptor binding domain (RBD). From a public health perspective these changes have important implications for increased transmissibility and immune escape; however, their presence could also modify the intrinsic host-range of the virus. Using viral pseudotyping we examined whether the variants of concern (VOCs) Alpha, Beta, Gamma and Delta have differing host ACE2 receptor usage patterns, focusing on a range of relevant mammalian ACE2 proteins. All four VOCs were able to overcome a previous restriction for mouse ACE2, with demonstrable differences also seen for individual VOCs with rat, ferret or civet ACE2 receptors, changes which we subsequently attribute to N501Y and E484K substitutions within the Spike RBD.

ACE2 proteins. All four VOCs were able to overcome a previous restriction for mouse ACE2, 30 with demonstrable differences also seen for individual VOCs with rat, ferret or civet ACE2 31 receptors, changes which we subsequently attribute to N501Y and E484K substitutions within 32 the Spike RBD. 33 34

SARS-CoV-2, the β-coronavirus responsible for the Covid-19 pandemic, is thought to have 36 emerged from a bat reservoir, potentially via an as yet unidentified intermediate mammalian 37

host. These conclusions are supported by knowledge of the origins of SARS-CoV-1 as well 38

as the sequence data of related viruses isolated in bats e.g., the RaTG13 isolate [1] . Recently, 39 Temmam et al., identified sarbecoviruses in bats that are nearly identical within the Spike 40 RBD, which adds further weight to this conclusion and highlights that coronaviruses with high 41 human ACE2 affinity are actively circulating in wild reservoir populations [2] . Interestingly, 42 experimental infection of ferrets, mice, bats, primates and other animals, together with natural 43

infections in a range of species including cats, dogs and mink indicate these viruses may have 44 a broader host-range than bats and humans [3] [4] [5] [6] . Understanding SARS-CoV-2 infection in 45 animals is important for three main reasons. Firstly, to assess the direct risk to livestock, 46 companion animals and wildlife. Secondly, to examine whether these animals can act as 47 secondary reservoirs for SARS-CoV-2. And lastly, to help establish and validate animal 48 models for Covid-19 which can then be used for the development of therapeutics and 49 vaccines, and to better understand the mechanisms of viral pathogenesis. To support these 50 research endeavours, we and others have shown that SARS-CoV-2 Spike has a broad tropism 51 for mammalian ACE2 proteins [7] [8] [9] . Using lentiviral pseudotyping combined with live virus 52 experiments we showed that SARS-CoV-2 can use a wide range of mammalian ACE2s 53 including dog, cat, cattle, sheep, pangolin and rabbit, but is restricted with rat, ferret and a 54 subset of bat and bird receptors [7, 10] . 55 Importantly, the continued evolution of SARS-CoV-2 in human populations has led to the 56 emergence of variants, a natural result of RNA virus replication. Informed by epidemiological, 57 virological and immunological data, public health bodies such as the World Health 58

Organisation (WHO) and Public Health England (PHE) have assigned some as variants of 59 concern (VOCs). Over the course of the pandemic two of these VOCs (B.1.1.7 [Alpha] and 60 B.1.617.2 [Delta]) have independently risen to prominence, rapidly replacing the previously 61 circulating strain across multiple regions (G614) [11] . The mechanisms underpinning this 62 replacement appear to correlate with continued evolution to the human host, through 63 increases in particle infectivity, replicative capacity, transmission potential, innate immune 64 antagonistic properties and potentially immune escape [12] [13] [14] [15] [16] . However, whether these 65 variants have expanded host-ranges and the consequences for the ongoing pandemic, in 66 terms of the factors described above, are not well characterised. 67

Using a lentiviral (HIV-1) pseudotyping approach for SARS-CoV-2 (described previously, [7, keeping with previous findings that early SARS-CoV-2 isolates were restricted with this 94 specific receptor [7] . Of note, the increase in mouse ACE2 usage was less notable for Delta, 95 the only non-N501Y-containing variant in our study. Similarly, the SARS-CoV-2 Spikerat 96 ACE2 restriction we reported previously [7] was only overcome by N501Y-containing VOC 97

Spikes (Alpha, Beta, Gamma), with no significant increase seen with Delta ( Figure 1 ). For 98 civet ACE2 significant increases in entry were observed only with the Beta VOC, while for 99 ferret ACE2 Beta and Gamma were both significantly higher than WT D614. Interestingly, no 100 significant differences were observed when hamster or pig ACE2 were used as receptors 101 although in our 2020 study neither of these receptors were restrictive to early isolates of 102 SARS-CoV-2 entry, unlike rat, mouse, civet and ferret ACE2 [7] . Analysis of the same data on 103 radar plots highlights the extended host-range of the N501Y-containing VOCs (Alpha, Beta 104

and Gamma), while simultaneously demonstrating the similar host-range profiles of WT D614 105 and Delta (Figure 1 A-D right panels). 106

To examine the role of individual amino acid changes in Spike in overcoming host-range 107

restrictions we subsequently focused on the mouse, rat and civet ACE2 interactions. Plasmids were performed in a N501Y-containing background, which could potentially obscure 116 synergistic effects between these two mutations. Interestingly for civet ACE2, the N501Y or 117 K417N changes were inhibitory, with E484K, in contrast, increasing entry efficiency of the 118 respective SARS-CoV-2 pseudotype ( Figure 2C ). The small, but repeatable, increase in civet 119 ACE2 usage with the Beta VOC Spike ( Figure 1B ) may therefore be a combinatorial effect of 120 these mutations, with E484K compensating for the inhibitory properties of the 501 and 417-121 specific changes. Accordingly, an Alpha + E484K Spike-based pseudotype was able to rescue 122 the defect in civet ACE2 usage seen with Alpha alone ( Figure 2C ; left panel). 123

The continuing evolution of SARS-CoV-2 in human populations raises significant concerns. 124

Principally, these relate to the human pandemic and human disease, e.g., antigenic escape 125 from vaccine-or natural infection-derived immunity, or the acquisition of enhanced 126 transmission potential or increased pathogenicity. Separately, however, the question of 127 whether SARS-CoV-2 will develop enhanced reverse zoonotic potential is also important. The Regarding infection, our data show that from an ACE2 receptor usage perspective the host-137 range of Alpha, Beta and Gamma (and to a lesser extent Delta) is broader than the virus that 138 initially emerged in China (WT D614). Several host ACE2 proteins that were previously 139 refractory to viral entry (mouse, rat and ferret) are used more efficiently by VOC Spikes (Figure  140 1). This change in tropism is attributable to specific amino acid changes in the RBD of Spike, 141 in particular N501Y (Figure 2 ). This is consistent with previous work showing amino acid 142 substitutions within the RBD overcome restriction with murine ACE2. These mutations were 143 identified by structure-led approaches [19] or following passage of SARS-CoV-2 directly in 144 mice [20], results which have been subsequently confirmed in vitro [21, 22] . Although these 145 changes are not identical, they fit within an overall pattern that small adaptations in Spike are 146 needed to overcome specific incompatibilities between the viral attachment protein and non-147 cognate receptors. We observed a similar trend during adaptation of SARS-CoV-2 in ferrets, 148

with Spike mutations Y453F, F486L and N501T leading to increased ferret ACE2 usage [10] . Similar results were also reported by Shuai et al in mice and rats [27] . This has contributed to 160 speculation that rodents could play an important role in the transmission of SARS-CoV-2 to 161 people [28]; however, these data also have important ramifications beyond reverse zoonosis. 162

For example, significant weight is given to animal model data on SARS-CoV-2 pathogenicity 163 during antiviral screening or vaccine development [6, 29] . If, for example, N501Y containing 164

VOCs cause increased pathogenicity because they are now able to use mouse ACE2 more 165 efficiently, this could complicate or potentially invalidate comparisons between WT D614 and 166 VOC-based mouse data and more broadly correlations with human data, where the difference 167 in human ACE2 usage is marginal. In addition, those animal models that endogenously 168 express restricted ACE2s (ferret, mouse and rat) are likely to drive in-animal adaptation, as 169

we witnessed with ferrets [10], which further complicates conclusions on viral phenotypes such 170 as transmission potential. Fortunately, in hamsters which have emerged as a highly tractable 171 model for SARS-CoV-2, the situation appears less convoluted as differences in hamster and 172 human ACE2 usage between WT D614 and VOC pseudotypes look comparable (Figure 1) . 173

replicate in these hosts correlates with changes in the Spike protein, it is difficult to make a 175 broader set of conclusions. Why certain hosts develop disease and others don't and how this 176 could impact the Covid-19 pandemic is still a relatively unanswered question. A relevant 177 example is pigs, which we have established encode a cognate receptor for SARS-CoV-2 entry 178 ( Figure 1) . Pigs have proven time and again to be an adequate reservoir for human tropic 179 viruses (e.g. influenza and Nipah), yet are apparently refractory to SARS-CoV-2 infection [30, 180 31] . Recently, we demonstrated tissue-specific differences in ACE2 expression between 181 various animals, e.g. ACE2 was present in the nasal mucosa epithelium of Eptesicus serotinus 182 (serotine bat) but not in pigs (Sus scrofa domestica) [32] which may provide some mechanistic 183 insight into the varying susceptibility of hosts to SARS-CoV-2. Beyond entry, various species-184 specific restrictions, for example at the level of the innate immunity response, may also be 185 playing a role. 186

Nevertheless, the fundamental question that remains is whether existing or emerging SARS-187

CoV-2 variants can establish themselves in an animal reservoir in a way that is consequential humans by SARS-CoV-2 infected animals appears much less common, albeit not totally 195 absent [23, 24] . The conceivable 'worst case scenario' for SARS-CoV-2 reverse zoonosis is 196 that the virus establishes itself in a new reservoir and at such a level that antibody selection 197 pressure takes place and/or prolonged antigenic drift leading to escape mutants that are 198 relevant to immune human populations. Our previous data indicated that SARS-CoV-2 was a 199 'generalist' with a broad host-range [7] . Significantly, the VOCs have even broader host ranges 200 and, in the majority of cases, the amino acid changes involved e.g., N501Y, which enhances 201 human ACE2 binding (providing evidence of ongoing human adaptation), do not lead to a loss 202 of activity with other host ACE2s. In other words, the 'generalist' nature of SARS-CoV-2 is 203 being maintained and extended over time. Whether this trend will continue and whether this 204

increases the probability of human-relevant reverse zoonosis events remains to be 205 determined. 206

This work was supported by the G2P-UK National Virology Consortium funded by MRC funded 208 grant G2P-UK; A National Virology Consortium to address phenotypic consequences of 209 SARS-CoV-2 genomic variation (MR/W005611/1). DB, NT, JN and GG were also funded by 210

The Pirbright Institute's BBSRC institute strategic programme grant (BBS/E/I/00007038 and 211 BBS/E/I/00007034) with NT receiving studentship support from BB/T008784/1. EW, CH and 212 LB were supported by the MRC grant MR/V036750/1. 213

The authors have no conflicts of interest to declare. 215 the BrightGlo reagent (Promega). Right panels: The same data was replotted on radar plots 228 to illustrate the broadening host-range of the Alpha (red), Beta (green) and Gamma (blue) 229

VOCs, as opposed to Delta (yellow). In each plot the signal for WT D614 is provided in grey 230 (transparent polygon). For entry assays, BHK-21 cells were plated at 5x10 5 /well in a 6-well 231 dish and transfected with the indicated ACE2 expression plasmids (human, pig, rat, hamster, 232 ferret, civet, mouse) 24 hours later. The next day, cells were harvested using 2mM EDTA-PBS 233 and diluted to 2x10 5 /ml with 100µL being seeded into white-bottomed 96-well plates and 234 incubated for 24 hrs at 37oC, 5% CO2. The media was removed from plated cells and infected 235 with 100µl of the indicated pseudotyped virus and incubated for 48 hours. 

A pneumonia outbreak associated with a new 255 coronavirus of probable bat origin

CoV-2-like receptor-binding domain allowing ACE2-mediated entry into human cells isolated from bats 258 of Indochinese peninsula

Susceptibility of ferrets, cats, dogs, and other 260 domesticated animals to SARS-coronavirus 2

SARS-CoV-2 neutralizing serum antibodies in 262 cats: a serological investigation

Infection and Rapid Transmission of SARS-CoV-264 2 in Ferrets

CoV-2 transmission

The SARS-CoV-2 Spike protein has a 268 broad tropism for mammalian ACE2 proteins

SARS-CoV-2 and Three Related Coronaviruses Utilize 270

Multiple ACE2 Orthologs and Are Potently Blocked by an Improved ACE2-Ig

Broad and Differential Animal Angiotensin-272 Converting Enzyme 2 Receptor Usage by SARS-CoV-2

CoV-2 to mustelid hosts do not increase fitness in the human airway

Changing composition of 277 SARS-CoV-2 lineages and rise of Delta variant in England

The furin cleavage site in the SARS-279

CoV-2 spike protein is required for transmission in ferrets

Recurrent emergence of SARS-CoV-2 spike 281 deletion H69/V70 and its role in the Alpha variant B.1.1.7

Evolution of 283 enhanced innate immune evasion by the SARS-CoV-2 B.1.1.7 UK variant

AZD1222-induced neutralising antibody 286 activity against SARS-CoV-2 Delta VOC

CoV-2 VOCs B.1.617.2 and B.1.351 by BNT162b2 vaccination

Production of Recombinant Replication-defective 290

Lentiviruses Bearing the SARS-CoV or SARS-CoV-2 Attachment Spike Glycoprotein and Their 291

Deletion of the SARS-CoV-2 Spike Cytoplasmic 293

Tail Increases Infectivity in Pseudovirus Neutralization Assays

A mouse-adapted model 295 of SARS-CoV-2 to test COVID-19 countermeasures

Adaptation of SARS-CoV-2 in BALB/c mice for testing 297 vaccine efficacy

Effects of Spike Mutations in 299 SARS-CoV-2 Variants of Concern on Human or Animal ACE2-Mediated Virus Entry and Neutralization

Analysis of SARS-CoV-2 variant mutations 302 reveals neutralization escape mechanisms and the ability to use ACE2 receptors from additional 303 species

SARS-CoV-2 and the human-animal interface: outbreaks on mink farms

The B1.351 and P.1 variants 310 extend SARS-CoV-2 host range to mice

SARS-CoV-2 B.1.1.7 and B.1.351 variants 312 of concern induce lethal disease in K18-hACE2 transgenic mice despite convalescent plasma therapy

Emerging SARS-CoV-2 315 variants expand species tropism to rodents

SARS-CoV-2 N501Y variants of concern and their 317 potential transmission by mouse

Tree Shrew, and Non-human Primates

SARS-CoV-2 in fruit bats, ferrets, pigs, 321 and chickens: an experimental transmission study

Pigs are not susceptible to 323 SARS-CoV-2 infection but are a model for viral immunogenicity studies

Differential susceptibility of SARS-326

CoV-2 in animals: Evidence of ACE2 host receptor distribution in companion animals, livestock and 327 wildlife by immunohistochemical characterisation

Multiple spillovers and 329 onward transmission of SARS-Cov-2 in free-living and captive White-tailed deer (<em>Odocoileus 330 virginianus</em>)

Evidence of exposure to SARS-CoV-334 2 in cats and dogs from households in Italy