Biochemical evidence of furin specificity and potential for phospho-regulation at Spike protein S1/S2 cleavage site in SARS-CoV2 but not in SARS-CoV1 or MERS-CoV

Abstract

The Spike protein of the novel coronavirus SARS-CoV2 contains an insertion 680 SPRRAR↓SV 687 forming a cleavage motif RxxR for furin-like enzymes at the boundary of S1/S2 subunits. Cleavage at S1/S2 is important for efficient viral entry into target cells. The insertion is absent in other CoV-s of the same clade, including SARS-CoV1 that caused the 2003 outbreak. However, an analogous insertion was present in the Spike protein of the more distant Middle East Respiratory Syndrome coronavirus MERS-CoV. We show that a crucial third arginine at the left middle position, comprising a motif RRxR is required for furin recognition in vitro, while the general motif RxxR in common with MERS-CoV is not sufficient for cleavage. Further, we describe a surprising finding that the two serines at the edges of the insert SPRRAR↓SV can be efficiently phosphorylated by proline-directed and basophilic protein kinases. Both phosphorylations switch off furin's ability to cleave the site. Although phosphoregulation of secreted proteins is still poorly understood, further studies, supported by a recent report of ten in vivo phosphorylated sites in the Spike protein of SARS-CoV2, could potentially uncover important novel regulatory mechanisms for SARS-CoV2.

Ferrets not infected by SARS-CoV-2 in a high-exposure domestic setting

“…S686G changes a neutral polar residue to a non-polar one, which we estimated to decrease furin efficiency. Furthermore, S686 completes a novel glycosaminoglycan (GAG)-binding motif (XBBXBX/PRRARS) that enhances binding and the two flanking serines in the S1/S2 site (SPRRAR↓SV) have been shown to be permissive to host phosphorylation and consequent down regulation of furin activity, (26,34). We were surprised to see evidence of positive selection over time for this potentially unfavorable mutation in ferrets as described by Richard et al for these reasons (19).…”

Implications of Spike-glycoprotein processing at S1/S2 by Furin, at S2’ by Furin and/or TMPRSS2 and shedding of ACE2: cell-to-cell fusion, cell entry and infectivity of SARS-CoV-2

“…1D). This reflects the multi-basic nature of the S1/S2 recognition sequence R RA R 685 ↓S, whereby RR AA 685 and A R A R 685 are cleavable, but not ARAA 685 (μS1/S2), suggesting the importance of the P3 Arg 683 [43]. The S686A mutant was based on the prediction that Ser 686 could be O-glycosylated [44], which may hamper processing at S1/S2 [45].…”

SARS-CoV-2 spike-glycoprotein processing at S1/S2 and S2’and shedding of the ACE2 viral receptor: roles of Furin and TMPRSS2 and implications for viral infectivity and cell-to-cell fusion

“…1D). This reflects the multi-basic nature of the S1/S2 recognition sequence RRAR685↓S, whereby RRAA685 and ARAR685 are cleavable, but not ARAA685 (µS1/S2), suggesting the importance of the P3 Arg683 [43]. The S686A mutant was based on the prediction that Ser686 could be O-glycosylated [44], which may hamper processing at S1/S2 [45].…”

“…2B), these single mutants were partially cleaved by overexpressed Furin (SI-Fig. 1C), reflecting the multi-basic nature of the S1/S2 recognition sequence and suggesting the importance of the P3 Arg683 (37). The S686A mutant was based on the prediction that Ser686 could be O-glycosylated (38), which may hamper processing at S1/S2 (39).…”

“…2A). This reflects the multi-basic nature of the S1/S2 recognition sequence RRAR 685 ↓S, whereby ARAR 685 and RRAA 685 are cleavable, but not ARAA 685 (μS1/S2), suggesting the importance of the P3 site 26 . Finally, Ala-mutants of each underlined residue in S2': KPSKR 815 ↓SFIE 15 resulted in an ER-retained S-protein (not shown).…”