Structural insights into Cullin4-RING ubiquitin ligase remodelling by Vpr from simian immunodeficiency viruses 1 Structural insights into Cullin4-RING ubiquitin ligase remodelling by Vpr from 1 simian immunodeficiency viruses 2 3 Sofia Banchenko1¶, Ferdinand Krupp1¶, Christine Gotthold1, Jörg Bürger1,2, Andrea Graziadei3, Francis 4 O’Reilly3, Ludwig Sinn3, Olga Ruda1, Juri Rappsilber3,4, Christian M. T. Spahn1, Thorsten Mielke2, Ian 5 A. Taylor5, David Schwefel1* 6 7 1 Institute of Medical Physics and Biophysics, Charité – Universitätsmedizin Berlin, corporate member 8 of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Berlin, 9 Germany 10 2 Microscopy and Cryo-Electron Microscopy Service Group, Max-Planck-Institute for Molecular 11 Genetics, Berlin, Germany 12 3 Bioanalytics Unit, Institute of Biotechnology, Technische Universität Berlin, Berlin, Germany 13 4 Wellcome Centre for Cell Biology, University of Edinburgh, Edinburgh, United Kingdom 14 5 Macromolecular Structure Laboratory, The Francis Crick Institute, London, United Kingdom 15 16 *Corresponding author 17 E-mail: david.schwefel@charite.de (DS) 18 19 ¶These authors contributed equally to this work 20 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 2 Abstract 21 Viruses have evolved means to manipulate the host’s ubiquitin-proteasome system, in order to down-22 regulate antiviral host factors. The Vpx/Vpr family of lentiviral accessory proteins usurp the substrate 23 receptor DCAF1 of host Cullin4-RING ligases (CRL4), a family of modular ubiquitin ligases involved 24 in DNA replication, DNA repair and cell cycle regulation. CRL4DCAF1 specificity modulation by Vpx 25 and Vpr from certain simian immunodeficiency viruses (SIV) leads to recruitment, poly-ubiquitylation 26 and subsequent proteasomal degradation of the host restriction factor SAMHD1, resulting in enhanced 27 virus replication in differentiated cells. To unravel the mechanism of SIV Vpr-induced SAMHD1 28 ubiquitylation, we conducted integrative biochemical and structural analyses of the Vpr protein from 29 SIVs infecting Cercopithecus cephus (SIVmus). X-ray crystallography reveals commonalities between 30 SIVmus Vpr and other members of the Vpx/Vpr family with regard to DCAF1 interaction, while cryo-31 electron microscopy and cross-linking mass spectrometry highlight a divergent molecular mechanism 32 of SAMHD1 recruitment. In addition, these studies demonstrate how SIVmus Vpr exploits the dynamic 33 architecture of the multi-subunit CRL4DCAF1 assembly to optimise SAMHD1 ubiquitylation. Together, 34 the present work provides detailed molecular insight into variability and species-specificity of the 35 evolutionary arms race between host SAMHD1 restriction and lentiviral counteraction through Vpx/Vpr 36 proteins. 37 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 3 Author summary 38 Due to the limited size of virus genomes, virus replication critically relies on host cell components. In 39 addition to the host cell’s energy metabolism and its DNA replication and protein synthesis apparatus, 40 the protein degradation machinery is an attractive target for viral re-appropriation. Certain viral factors 41 divert the specificity of host ubiquitin ligases to antiviral host factors, in order to mark them for 42 destruction by the proteasome, to lift intracellular barriers to virus replication. Here, we present 43 molecular details of how the simian immunodeficiency virus accessory protein Vpr interacts with a 44 substrate receptor of host Cullin4-RING ubiquitin ligases, and how this interaction redirects the 45 specificity of Cullin4-RING to the antiviral host factor SAMHD1. The studies uncover the mechanism 46 of Vpr-induced SAMHD1 recruitment and subsequent ubiquitylation. Moreover, by comparison to 47 related accessory proteins from other immunodeficiency virus species, we illustrate the surprising 48 variability in the molecular strategies of SAMHD1 counteraction, which these viruses adopted during 49 evolutionary adaptation to their hosts. Lastly, our work also provides deeper insight into the inner 50 workings of the host’s Cullin4-RING ubiquitylation machinery. 51 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 4 Introduction 52 A large proportion of viruses have evolved means to co-opt their host’s ubiquitylation machinery, in 53 order to improve replication conditions, either by introducing viral ubiquitin ligases and deubiquitinases, 54 or by modification of host proteins involved in ubiquitylation [1-3]. In particular, host ubiquitin ligases 55 are a prominent target for viral usurpation, to redirect specificity towards antiviral host restriction 56 factors. This results in recruitment of restriction factors as non-endogenous neo-substrates, inducing 57 their poly-ubiquitylation and subsequent proteasomal degradation [4-8]. This counteraction of the host’s 58 antiviral repertoire is essential for virus infectivity and spread [9-12], and mechanistic insights into these 59 specificity changes extend our understanding of viral pathogenesis and might pave the way for novel 60 treatments. 61 Frequently, virally encoded modifying proteins associate with, and adapt the Cullin4-RING ubiquitin 62 ligases (CRL4) [5]. CRL4 consists of a Cullin4 (CUL4) scaffold that bridges the catalytic RING-domain 63 subunit ROC1 to the adaptor protein DDB1, which in turn binds to exchangeable substrate receptors 64 (DCAFs, DDB1- and CUL4-associated factors) [13-17]. In some instances, the DDB1 adaptor serves as 65 an anchor for virus proteins, which then act as “viral DCAFs” to recruit the antiviral substrate. Examples 66 are the simian virus 5 V protein and mouse cytomegalovirus M27, which bind to DDB1 and recruit 67 STAT1/2 proteins for ubiquitylation, in order to interfere with the host’s interferon response [18-20]. 68 Similarly, CUL4-dependent downregulation of STAT signalling is important for West Nile Virus 69 replication [21]. In addition, the hepatitis B virus X protein hijacks DDB1 to induce proteasomal 70 destruction of the structural maintenance of chromosome (SMC) complex to promote virus replication 71 [22, 23]. 72 Viral factors also bind to and modify DCAF receptors in order to redirect them to antiviral substrates. 73 Prime examples are the lentiviral accessory proteins Vpr and Vpx. All contemporary human and simian 74 immunodeficiency viruses (HIV/SIV) encode Vpr, while only two lineages, represented by HIV-2 and 75 SIV infecting mandrills, carry Vpx [24]. Vpr and Vpx proteins are packaged into progeny virions and 76 released into the host cell upon infection, where they bind to DCAF1 in the nucleus [25]. In this work, 77 corresponding simian immunodeficiency virus Vpx/Vpr proteins will be indicated with their host 78 species as subscript, with the following abbreviations used: mus – moustached monkey (Cercopithecus 79 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 5 cephus), mnd – mandrill (Mandrillus sphinx), rcm – red-capped mangabey (Cercocebus torquatus), sm 80 – sooty mangabey (Cercocebus atys), deb – De Brazza’s monkey (Cercopithecus neglectus), syk – 81 Syke’s monkey (Cercopithecus albogularis), agm – african green monkey (Chlorocebus spec). 82 VprHIV-1 is important for virus replication in vivo and in macrophage infection models [26]. Recent 83 proteomic analyses revealed that DCAF1 specificity modulation by VprHIV-1 proteins results in down-84 regulation of hundreds of host proteins in a DCAF1- and proteasome-dependent manner [27], including 85 the previously reported VprHIV-1 degradation targets UNG2 [28], HLTF [29], MUS81 [30, 31], MCM10 86 [32] and TET2 [33]. This surprising promiscuity in degradation targets is also partially conserved in 87 more distant clades exemplified by Vpragm and Vprmus [27]. However, Vpr pleiotropy, and the lack of 88 easily accessible experimental models, have prevented a characterisation of how these degradation 89 events precisely promote replication [26]. 90 By contrast, Vpx, exhibits a much narrower substrate range. It has recently been reported to target 91 stimulator of interferon genes (STING) and components of the human silencing hub (HUSH) complex 92 for degradation, leading to inhibition of antiviral cGAS-STING-mediated signalling and reactivation of 93 latent proviruses, respectively [34-36]. Importantly, Vpx also recruits the SAMHD1 restriction factor to 94 DCAF1, in order to mark it for proteasomal destruction [37, 38]. SAMHD1 is a deoxynucleotide 95 triphosphate (dNTP) triphosphohydrolase that restricts retroviral replication in non-dividing cells by 96 lowering the dNTP pool to levels that cannot sustain viral reverse transcription [39-46]. Retroviruses 97 that express Vpx are able to alleviate SAMHD1 restriction and allow replication in differentiated 98 myeloid lineage cells, resting T cells and memory T cells [38, 47, 48]. As a result of the constant 99 evolutionary arms race between the host’s SAMHD1 restriction and its viral antagonist Vpx, the 100 mechanism of Vpx-mediated SAMHD1 recruitment is highly virus species- and strain-specific: The 101 Vpx clade represented by VpxHIV-2 recognises the SAMHD1 C-terminal domain (CtD), while Vpxmnd2/rcm 102 binds the SAMHD1 N-terminal domain (NtD) in a fundamentally different way [24, 49-52]. 103 In the course of evolutionary adaptation to their primate hosts, and due to selective pressure to evade 104 SAMHD1 restriction, two groups of SIVs that do not have Vpx, SIVagm, and SIVdeb/mus/syk, neo-105 functionalised their Vpr to bind SAMHD1 and induce its degradation [24, 49, 53]. Consequently, these 106 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 6 species evolved “hybrid” Vpr proteins that retain targeting of some host factors depleted by HIV-1-type 107 Vpr [27], and additionally induce SAMHD1 degradation. 108 To uncover the molecular mechanisms of DCAF1- and SAMHD1-interaction of such a “hybrid” Vpr, 109 we initiated integrative biochemical and structural analyses of the Vpr protein from an SIV infecting 110 Cercopithecus cephus, Vprmus. These studies reveal similarities and differences to Vpx and Vpr proteins 111 from other lentivirus species and pinpoint the divergent molecular mechanism of Vprmus-dependent 112 SAMHD1 recruitment to CRL4DCAF1. Furthermore, cryo-electron microscopic (cryo-EM) 113 reconstructions of a Vprmus-modified CRL4DCAF1 protein complex allow for insights into the structural 114 plasticity of the entire CRL4 ubiquitin ligase assembly, with implications for the ubiquitin transfer 115 mechanism. 116 117 Results 118 SAMHD1-CtD is necessary and sufficient for Vprmus-binding and ubiquitylation in vitro 119 To investigate the molecular interactions between Vprmus, the neo-substrate SAMHD1 from rhesus 120 macaque and CRL4 subunits DDB1/DCAF1 C-terminal domain (DCAF1-CtD), protein complexes were 121 reconstituted in vitro from purified components and analysed by gel filtration (GF) chromatography. 122 The different protein constructs that were employed are shown schematically in S1A Fig. Vprmus is 123 insoluble after removal of the GST affinity purification tag (S1B Fig) and accordingly could not be 124 applied to the GF column. No interaction of SAMHD1 with DDB1/DCAF1-CtD could be detected in 125 the absence of Vprmus (S1C Fig). Analysis of binary protein combinations (Vprmus and DDB1/DCAF1-126 CtD; Vprmus and SAMHD1) shows that Vprmus elutes in a single peak together with DDB1/DCAF1-CtD 127 (S1D Fig) or with SAMHD1 (S1E Fig). Incubation of Vprmus with DDB1/DCAF1B and SAMHD1 128 followed by GF resulted in elution of all three components in a single peak (Fig 1A, B, red trace). 129 Together, these results show that Vprmus forms stable binary and ternary protein complexes with 130 DDB1/DCAF1-CtD and/or SAMHD1 in vitro. Furthermore, incubation with any of these interaction 131 partners apparently stabilises Vprmus by alleviating its tendency for aggregation/insolubility. 132 Previous cell-based assays indicated that residues 583-626 of rhesus macaque SAMHD1 (SAMHD1-133 CtD) are necessary for Vprmus-induced proteasomal degradation [49]. To test this finding in our in vitro 134 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 7 system, constructs containing SAMHD1-CtD fused to T4 lysozyme (T4L-SAMHD1-CtD), or lacking 135 SAMHD1-CtD (SAMHD1-ΔCtD, Fig 1A), were incubated with Vprmus and DDB1/DCAF1-CtD, and 136 complex formation was assessed by GF chromatography. Analysis of the resulting chromatograms by 137 SDS-PAGE shows that SAMHD1-ΔCtD did not co-elute with DDB1/DCAF1-CtD/Vprmus (Fig 1A, B, 138 green trace). By contrast, T4L-SAMHD1-CtD accumulated in a single peak, which also contained 139 DDB1/DCAF1-CtD and Vprmus (Fig 1A, B, cyan trace). These results confirm that SAMHD1-CtD is 140 necessary for stable association with DDB1/DCAF1-CtD/Vprmus in vitro, and demonstrate that 141 SAMHD1-CtD is sufficient for Vprmus-mediated recruitment of the T4L-SAMHD1-CtD fusion construct 142 to DDB1/DCAF1-CtD. 143 To correlate these data with enzymatic activity, in vitro ubiquitylation assays were conducted by 144 incubating SAMHD1, SAMHD1-ΔCtD or T4L-SAMHD1-CtD with purified CRL4DCAF1-CtD, E1 145 (UBA1), E2 (UBCH5C), ubiquitin and ATP. Input proteins are shown in S2A Fig, and control reactions 146 in S2B, C Fig. In the absence of Vprmus, no SAMHD1 ubiquitylation was observed (Figs 1C and S2D), 147 while addition of Vprmus resulted in robust SAMHD1 ubiquitylation (Figs 1D and S2E). In agreement 148 with the analytical GF data, SAMHD1-ΔCtD was not ubiquitylated in the presence of Vprmus (Figs 1E 149 and S2F), while T4L-SAMHD1-CtD, was ubiquitylated with similar kinetics as the full-length protein 150 (Figs 1F and S2F). Again, these data substantiate the functional importance of SAMHD1-CtD for 151 Vprmus-mediated recruitment to the CRL4DCAF1 ubiquitin ligase. 152 153 Crystal Structure analysis of apo- and Vprmus-bound DDB1/DCAF1-CtD protein 154 complexes 155 To obtain structural information regarding Vprmus and its mode of binding to the CRL4 substrate receptor 156 DCAF1, the X-ray crystal structures of a DDB1/DCAF1-CtD complex, and DDB1/DCAF1-CtD/T4L-157 Vprmus (residues 1-92) fusion protein ternary complex were determined. The structures were solved 158 using molecular replacement and refined to resolutions of 3.1 Å and 2.5 Å respectively (S1 Table). 159 Vprmus adopts a three-helix bundle fold, stabilised by coordination of a zinc ion by His and Cys residues 160 on Helix-1 and at the C-terminus (Fig 2A). Superposition of Vprmus with previously determined Vpxsm 161 [50], Vpxmnd2 [51, 52], and VprHIV-1 [54] structures reveals a conserved three-helix bundle fold, and 162 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 8 similar position of the helix bundles on DCAF1-CtD (S3A Fig). In addition, the majority of side chains 163 involved in DCAF1-interaction are type-conserved in all Vpx and Vpr proteins (Figs S3B-G and S6A), 164 strongly suggesting a common molecular mechanism of host CRL4-DCAF1 hijacking by the Vpx/Vpr 165 family of accessory proteins. However, there are also significant differences in helix length and register 166 as well as conformational variation in the loop region N-terminal of Helix-1, at the start of Helix-1 and 167 in the loop between Helices-2 and -3 (S3A Fig). 168 Vprmus binds to the side and on top of the disk-shaped 7-bladed β-propeller (BP) DCAF1-CtD domain 169 with a total contact surface area of ~1600 Å2 comprising three major regions of interaction. The extended 170 Vprmus N-terminus attaches to the cleft between DCAF1 BP blades 1 and 2 through several hydrogen 171 bonds, electrostatic and hydrophobic interactions (S3B-D Fig). A second, smaller contact area is formed 172 by hydrophobic interaction between Vprmus residues L31 and E34 from Helix-1, and DCAF1 W1156, 173 located in a loop on top of BP blade 2 (S3E Fig). The third interaction surface comprises the C-terminal 174 half of Vprmus Helix-3, which inserts into a ridge on top of DCAF1 (S3F, G Fig). 175 Superposition of the apo-DDB1/DCAF1-CtD and Vprmus-bound crystal structures reveals 176 conformational changes in DCAF1 upon Vprmus association. Binding of the N-terminal arm of Vprmus 177 induces only a minor rearrangement of a loop in BP blade 2 (S3C Fig). By contrast, significant structural 178 changes occur on the upper surface of the BP domain: polar and hydrophobic interactions of DCAF1 179 residues P1329, F1330, F1355, N1371, L1378, M1380 and T1382 with Vprmus side chains of T79, R83, 180 R86 and E87 in Helix-3 result in the stabilisation of the sequence stretch that connect BP blades 6 and 181 7 (“C-terminal loop”, Figs 2B and S3F). Moreover, side chain electrostatic interactions of Vprmus 182 residues R15, R75 and R76 with DCAF1 E1088, E1091 and E1093 lock the conformation of an “acidic 183 loop” upstream of BP blade 1, which is also unstructured and flexible in the absence of Vprmus (Figs 2B, 184 C and S3D, F). 185 Notably, in previously determined structures of Vpx/DCAF1/SAMHD1 complexes the “acidic loop” is 186 a central point of ternary contact, providing a binding platform for positively charged amino acid side 187 chains in either the SAMHD1 N- or C-terminus [50-52]. For example, Vpxsm positions SAMHD1-CtD 188 in such a way, that SAMHD1 K622 engages in electrostatic interaction with the DCAF1 “acidic loop” 189 residue D1092 (Fig 2C, left panel). However, in the Vprmus crystal structure the bound Vprmus now blocks 190 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 9 access to the corresponding SAMHD1-CtD binding pocket, in particular by the positioning of an 191 extended N-terminal loop that precedes Helix-1. Additionally, Vprmus side chains R15, R75 and R76 192 neutralise the DCAF1 “acidic loop”, precluding the formation of further salt bridges to basic residues in 193 SAMHD1-CtD (Fig 2C, right panel). 194 To validate the importance of Vprmus residues R15 and R75 for DCAF1-CtD- and SAMHD1-binding, 195 charge reversal mutations to glutamates were generated by site-directed mutagenesis. The effect of the 196 Vprmus R15E R75E double mutant on complex assembly was then analysed by GF chromatography. 197 SDS-PAGE analysis of the resulting chromatographic profile shows an almost complete loss of the 198 DDB1/DCAF1-CtD/Vprmus/SAMHD1 complex peak (Fig 2D, fraction 6), when compared to the wild 199 type, concomitant with enrichment of (i) Vprmus R15E R75E-bound DDB1/DCAF1-CtD (Fig 2D, 200 fractions 7-8), and of (ii) Vprmus R15E R75E/SAMHD1 binary complex (Fig 2D, fraction 8-9). This 201 suggests that charge reversal of Vprmus side chains R15 and R75 weakens the strong association with 202 DCAF1 observed in wild type Vprmus, due to loss of electrostatic interaction with the “acidic loop”, in 203 accordance with the crystal structure. Consequently, some proportion of Vpr-bound SAMHD1 204 dissociates, further indicating that Vprmus side chains R15 and R75 are not central to SAMHD1 205 interaction. 206 207 Molecular mechanism of SAMHD1-targeting 208 To obtain mechanistic insight into Vprmus-recruitment of SAMHD1-CtD, we initiated cryo-EM analyses 209 of the CRL4DCAF1-CtD/Vprmus/SAMHD1 assembly. In these studies, the small ubiquitin-like protein 210 NEDD8 was enzymatically attached to the CUL4 subunit, in order to obtain its active form (S4A Fig) 211 [55]. A CRL4-NEDD8DCAF1-CtD/Vprmus/SAMHD1 complex was reconstituted in vitro and purified by GF 212 chromatography (S4B Fig). Extensive 2D and 3D classification of the resulting particle images revealed 213 considerable conformational heterogeneity, especially regarding the position of the CUL4-214 NEDD8/ROC1 subcomplex (stalk) relative to DDB1/DCAF1/Vprmus (core), (S4 Fig). 215 Nevertheless, a homogeneous particle population could be separated, which yielded a 3D reconstruction 216 at a nominal resolution of 7.3 Å that contained electron density corresponding to the core (S4C-F Fig). 217 Molecular models of DDB1 BP domains A and C (BPA, BPC), DCAF1-CtD and Vprmus, derived from 218 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 10 our crystal structure (Fig 2), could be fitted as rigid bodies into this cryo-EM volume (Fig 3A). No 219 obvious electron density was visible for the bulk of SAMHD1. However, close inspection revealed an 220 additional tubular, slightly arcing density feature, approx. 35 Å in length, located on the upper surface 221 of the Vprmus helix bundle, approximately 17 Å away from and opposite of the Vprmus/DCAF1-CtD 222 binding interface (Fig 3A, red arrows). One end of the tubular volume contacts the middle of Vprmus 223 Helix-1, and the other end forms additional contacts to the C-terminus of Helix-2 and the N-terminus of 224 Helix-3. A local resolution of 7.5-8 Å precluded the fitting of an atomic model. Considering the 225 biochemical data, showing that SAMHD1-CtD is sufficient for recruitment to DDB1/DCAF1/Vprmus, 226 we hypothesise that this observed electron density feature corresponds to the region of SAMHD1-CtD 227 which physically interacts with Vprmus. Given its dimensions, the putative SAMHD1-CtD density could 228 accommodate approx. 10 amino acid residues in a fully extended conformation or up to 23 residues in 229 a kinked helical arrangement. All previous crystal structure analyses [46], as well as secondary structure 230 predictions indicate that SAMHD1 residues C-terminal to the catalytic HD domain and C-terminal lobe 231 (amino acids 599-626) are disordered in the absence of additional binding partners. Accordingly, the 232 globular domains of the SAMHD1 molecule might be flexibly linked to the C-terminal tether identified 233 here. In that case, the bulk of SAMHD1 samples a multitude of positions relative to the DDB1/DCAF1-234 CtD/Vprmus core, and consequently is averaged out in the process of cryo-EM reconstruction. 235 The topology of CRL4DCAF1-CtD/Vprmus/SAMHD1 and the binding region of SAMHD1-CtD were further 236 assessed by cross-linking mass spectrometry (CLMS) using the photo-reactive cross-linker sulfo-SDA 237 [56]. A large number of cross-links between SAMHD1 and the C-terminal half of CUL4, the side and 238 top of DCAF1-CtD, and BP blades 6-7 of DDB1 were found, consistent with highly variable positioning 239 of the SAM and HD domains of SAMHD1 relative to the CRL4 core (Fig 3B). Moreover, multiple 240 cross-links between SAMHD1-CtD and Vprmus were observed, more specifically locating to a sequence 241 stretch comprising the C-terminal half of Vprmus Helix-1 (residues A27-E36), and to a portion of the 242 disordered Vprmus C-terminus (residues Y90, Y100). These data are in accordance with the presence of 243 SAMHD1-CtD in the unassigned cryo-EM density and its role as Vprmus tether. The remaining 244 SAMHD1-CtD cross-links were with the C-terminus of CUL4 and the “acidic loop” of DCAF1 (Fig 245 3B). Distance restraints from these SAMHD1-CtD cross-links, together with our structural models of 246 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 11 CRL4DCAF1-CtD/Vprmus (see below), were employed to visualise the interaction space accessible to the 247 centre of mass of SAMHD1-CtD. This analysis is compatible with recruitment of SAMHD1-CtD on top 248 of the Vprmus helix bundle as indicated by cryo-EM (Fig 3C). Interestingly, cross-links to Vprmus were 249 restricted to the C-terminal end of SAMHD1-CtD (residues K622, K626), while cross-links to CUL4 250 and DCAF1 were found in the N-terminal portion (residues K595, K596, T602-S606). These 251 observations are consistent with a model where the very C-terminus of SAMHD1 is immobilised on 252 Vprmus, and SAMHD1-CtD residues further upstream are exposed to the catalytic machinery 253 surrounding the CUL4 C-terminal domain. 254 To further probe the interaction, Vprmus amino acid residues in close proximity to the putative SAMHD1-255 CtD density were substituted by site-directed mutagenesis. Specifically, Vprmus W29 was changed to 256 alanine to block a hydrophobic contact with SAMHD1-CtD involving the aromatic side chain, and 257 Vprmus A66 was changed to a bulky tryptophan, in order to introduce a steric clash with SAMHD1-CtD 258 (Fig 3D). This Vprmus W29A A66W double mutant was then assessed for complex formation with 259 DDB1/DCAF1-CtD and SAMHD1 by analytical GF. In comparison to wild type Vprmus, the W29A 260 A66W mutant showed a reduction of DDB1/DCAF1-CtD/Vprmus/SAMHD1 complex peak intensity (Fig 261 3E, fraction 6), concomitant with (i) enrichment of DDB1/DCAF1-CtD/Vprmus ternary complex, sub-262 stoichiometrically bound to SAMHD1 (Fig 3E, fraction 7), (ii) excess DDB1/DCAF1-CtD binary 263 complex (Fig 3E, fraction 8), and (iii) monomeric SAMHD1 species (Fig 3E, fractions 9-10). In 264 conclusion, this biochemical analysis, together with cryo-EM reconstruction at intermediate resolution 265 and CLMS analysis, locate the SAMHD1-CtD binding site on the upper surface of the Vprmus helix 266 bundle. 267 These data allow for structural comparison with neo-substrate binding modes of Vpx and Vpr proteins 268 from different retrovirus lineages (Fig 4A-D). VpxHIV-2 and Vpxsm position SAMHD1-CtD at the side of 269 the DCAF1 BP domain through interactions with the N-termini of Vpx Helices-1 and -3 (Fig 4B) [50]. 270 Vpxmnd2 and Vpxrcm bind SAMHD1-NtD using a bipartite interface comprising the side of the DCAF1 271 BP and the upper surface of the Vpx helix bundle (Fig 4C) [51, 52]. VprHIV-1 engages its ubiquitylation 272 substrate UNG2 using both the top and the upper edge of the VprHIV-1 helix bundle (Fig 4D) [54]. Of 273 note, these upper-surface interaction interfaces only partially overlap with the Vprmus/SAMHD1-CtD 274 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 12 binding interface identified here and employ fundamentally different sets of interacting amino acid 275 residues. Thus, it appears that the molecular interaction interfaces driving Vpx/Vpr-mediated neo-276 substrate recognition and degradation are not conserved between related SIV and HIV Vpx/Vpr 277 accessory proteins, even in cases where identical SAMHD1-CtD regions are targeted for recruitment. 278 279 Cryo-EM analysis of Vprmus-modified CRL4-NEDD8DCAF1-CtD conformational states and 280 dynamics 281 A reanalysis of the cryo-EM data using strict selection of high-quality 2D classes, followed by focussed 282 3D classification yielded three additional particle populations, resulting in 3D reconstructions at 8-10 Å 283 resolution, which contained both the Vprmus-bound CRL4 core and the stalk (conformational states-1, -2 284 and -3, Figs 5A and S4G-J). The quality of the 3D volumes was sufficient to fit crystallographic models 285 of core (Fig 2) and the stalk (PDB 2hye) [15] as rigid bodies (Figs 5B and S5A). For the catalytic RING-286 domain subunit ROC1, only fragmented electron density was present near the position it occupies in the 287 crystallographic model (S5A Fig). In all three states, electron density was selectively absent for the C-288 terminal CUL4 winged helix B (WHB) domain (residues 674-759), which contains the NEDD8 289 modification site (K705), and for the preceding α-helix, which connects the CUL4 N-terminal domain 290 to the WHB domain (S5A Fig). In accordance with this observation, the positions of CRL5-attached 291 NEDD8 and of the CRL4 ROC1 RING domain are sterically incompatible upon superposition of their 292 respective crystal structures (S5B Fig) [57]. 293 Alignment of 3D volumes from states-1, -2 and -3 shows that core densities representing DDB1 BPA, 294 BPC, DCAF1-CtD and Vprmus superimpose well, indicating that these components do not undergo major 295 conformational fluctuations and thus form a rigid platform for substrate binding and attachment of the 296 CRL4 stalk (Fig 5). However, rotation of DDB1 BPB around a hinge connecting it to BPC results in 297 three different orientations of state-1, -2 and -3 stalk regions relative to the core. BPB rotation angles 298 were measured as 69° between state-1 and -2, and 50° between state-2 and -3. Furthermore, the 299 crosslinks between DDB1 and CUL4 identified by CLMS are satisfied by the state-1 model, but 300 increasingly violated in states-2 and -3, validating in solution the conformational variability observed 301 by cryo-EM. (S5C Fig). Taken together, this places the CRL4 catalytic machinery, sited at the distal end 302 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 13 of the stalk, appropriately to approach the Vprmus-tethered bulk of SAMHD1 for ubiquitylation at a wide 303 range of angles (Fig 5B). 304 These data are in line with previous prediction based on extensive comparative crystal structure 305 analyses, which postulated an approx. 150° rotation of the CRL4 stalk around the core [13, 15, 16, 19, 306 58]. However, the left- and rightmost CUL4 orientations observed here, states-1 and -3 from our cryo-307 EM analysis, indicate a slightly narrower stalk rotation range (119°), when compared to the outermost 308 stalk conformations modelled from previously determined crystal structures (143°) (S5D Fig). An 309 explanation for this discrepancy comes from inspection of the cryo-EM densities and fitted models, 310 revealing that along with the main interaction interface on DDB1 BPB there are additional molecular 311 contacts between CUL4 and DDB1. Specifically, in state-1, there is a contact between the loop 312 connecting helices D and E of CUL4 cullin repeat (CR)1 (residues 161-169) and a loop protruding from 313 BP blade 3 of DDB1 BPC (residues 795-801, S5E Fig). In state-3, the loop between CUL4 CR2 helices 314 D and E (residues 275-282) abuts a region in the C-terminal helical domain of DDB1 (residues 1110-315 1127, S5F Fig). These auxiliary interactions might be required to lock the outermost stalk positions 316 observed here in order to confine the rotation range of CUL4. 317 318 Discussion 319 Our X-ray crystallographic studies of the DDB1/DCAF1-CtD/Vprmus assembly provide the first 320 structural insight into a class of “hybrid” SIV Vpr proteins. These are present in the SIVagm and 321 SIVmus/deb/syk lineages of lentiviruses and combine characteristics of related VprHIV-1 and SIV Vpx 322 accessory proteins. 323 Like SIV Vpx, “hybrid” Vpr proteins down-regulate the host restriction factor SAMHD1 by recruiting 324 it to CRL4DCAF1 for ubiquitylation and subsequent proteasomal degradation. However, using a 325 combination of X-ray, cryo-EM and CLMS analyses, we show that the molecular strategy, which Vprmus 326 evolved to target SAMHD1, is strikingly different from Vpx-containing SIV strains. In the two clades 327 of Vpx proteins, divergent amino acid sequence stretches just upstream of Helix-1 (variable region 328 (VR)1, S6A Fig), together with polymorphisms in the SAMHD1-N-terminus of the respective host 329 species, determine if HIV-2-type or SIVmnd-type Vpx recognise SAMHD1-CtD or SAMHD1-NtD, 330 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 14 respectively. These recognition mechanisms result in positioning of SAMHD1-CtD or -NtD on the side 331 of the DCAF1 BP domain in a way that allows for additional contacts between SAMHD1 and DCAF1, 332 thus forming ternary Vpx/SAMHD1/DCAF1 assemblies with very low dissociation rates [50-52, 59]. 333 In Vprmus, different principles determine the specificity for SAMHD1-CtD. Here, VR1 is not involved 334 in SAMHD1-CtD-binding at all, but forms additional interactions with DCAF1, which are not observed 335 in Vpx/DCAF1 protein complexes (S6A Fig). Molecular contacts between Vprmus and SAMHD1 are 336 dispersed on Helices-1 and -3, facing away from the DCAF1 interaction site and immobilising 337 SAMHD1-CtD on the top side of the Vprmus helix bundle (S6A Fig). Placement of SAMHD1-CtD in 338 such a position precludes stabilising ternary interaction with DCAF1-CtD, but still results in robust 339 SAMHD1 ubiquitylation in vitro and SAMHD1 degradation in cell-based assays [24]. 340 Predictions regarding the molecular mechanism of SAMHD1-binding by other “hybrid” Vpr 341 orthologues are difficult due to sequence divergence. Even in Vprdeb, the closest relative to Vprmus, only 342 approximately 50% of amino acid side chains lining the putative SAMHD1-CtD binding pocket are 343 conserved (S6A Fig). Previous in vitro ubiquitylation and cell-based degradation experiments did not 344 show a clear preference of Vprdeb for recruitment of either SAMHD1-NtD or –CtD [24, 49]. 345 Furthermore, it is disputed if Vprdeb actually binds DCAF1 [60], which might possibly be explained by 346 amino acid variations in the very N-terminus and/or in Helix-3 (S6A Fig). Vprsyk is specific for 347 SAMHD1-CtD [49], but the majority of residues forming the binding platform for SAMHD1-CtD 348 observed in the present study are not conserved. The SIVagm lineage of Vpr proteins is even more 349 divergent, with significant differences not only in possible SAMHD1-contacting residues, but also in 350 the sequence stretches preceding Helix-1, and connecting Helices-2 and -3, as well as in the N-terminal 351 half of Helix-3 (S6A Fig). Furthermore, there are indications that recruitment of SAMHD1 by the 352 Vpragm.GRI sub-type involves molecular recognition of both SAMHD1-NtD and –CtD [49, 53]. In 353 conclusion, recurring rounds of evolutionary lentiviral adaptation to the host SAMHD1 restriction 354 factor, followed by host re-adaptation, resulted in highly species-specific, diverse molecular modes of 355 Vpr-SAMHD1 interaction. In addition to the example presented here, further structural characterisation 356 of SAMHD1-Vpr complexes will be necessary to illustrate the manifold outcomes of this particular 357 virus-host molecular “arms race”. 358 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 15 Previous structural investigation of DDB1/DCAF1/VprHIV-1 in complex with the neo-substrate UNG2 359 demonstrated that VprHIV-1 engages UNG2 by mimicking the DNA phosphate backbone. More precisely, 360 UNG2 residues, which project into the major groove of its endogenous DNA substrate, insert into a 361 hydrophobic cleft formed by VprHIV-1 Helices-1, -2 and the N-terminal half of Helix-3 [54]. This 362 mechanism might rationalise VprHIV-1’s extraordinary binding promiscuity, since the list of potential 363 VprHIV-1 degradation substrates is significantly enriched in DNA- and RNA-binding proteins [27]. 364 Moreover, promiscuous VprHIV-1-induced degradation of host factors with DNA- or RNA-binding 365 activity has been proposed to induce cell cycle arrest at the G2/M phase border, which is the most 366 thoroughly described phenotype of Vpr proteins so far [26, 27, 61]. In Vprmus, the N-terminal half of 367 Helix-1 as well as the bulky amino acid residue W48, which is also conserved in Vpragm and Vpx, 368 constrict the hydrophobic cleft (S6A, B Fig). Furthermore, the extended N-terminus of Vprmus Helix-3 369 is not compatible with UNG2-binding due to steric exclusion (S6C Fig). In accordance with these 370 observations, Vprmus does not down-regulate UNG2 in a human T cell line [27]. However, Vprmus, Vprsyk 371 and Vpragm also cause G2/M cell cycle arrest in their respective host cells [60, 62, 63]. This strongly 372 hints at the existence of further structural determinants in Vprmus, Vprsyk, Vpragm and potentially VprHIV-1, 373 which regulate recruitment and ubiquitylation of DNA/RNA-binding host factors, in addition to the 374 hydrophobic, DNA-mimicking cleft on top of the three-helix bundle. Future efforts to structurally 375 characterise these determinants will further extend our understanding of how the Vpx/Vpr helical 376 scaffold binds, and in this way adapts to a multitude of neo-substrate epitopes. In addition, such efforts 377 might inform approaches to design novel CRL4DCAF1-based synthetic degraders, in the form of 378 proteolysis-targeting chimera-(PROTAC-) type compounds [64, 65]. 379 Our cryo-EM reconstructions of CRL4DCAF1-CtD/Vprmus/SAMHD1, complemented by CLMS, also 380 provide insights into the structural dynamics of CRL4 assemblies prior to ubiquitin transfer. The data 381 confirm previously described rotational movement of the CRL4 stalk, in the absence of constraints 382 imposed by a crystal lattice, creating a ubiquitylation zone around the Vprmus-modified substrate receptor 383 (Figs 5 and 6A) [13, 15, 16, 19, 58]. Missing density for the neddylated CUL4 WHB domain and for 384 the catalytic ROC1 RING domain indicates that these distal stalk elements are highly mobile and likely 385 sample a multitude of orientations relative to the CUL4 scaffold (Fig 6B). These observations are in line 386 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 16 with structure analyses of CRL1 and CRL5, where CUL1/5 neddylation leads to re-orientation of the 387 cullin WHB domain, and to release of the ROC1 RING domain from the cullin scaffold, concomitant 388 with stimulation of ubiquitylation activity [57]. Moreover, recent cryo-EM structure analysis of 389 CRL1β-TRCP/IκBα demonstrated substantial mobility of pre-catalytic NEDD8-CUL1 WHB and ROC1 390 RING domains [66]. Such flexibility seems necessary to structurally organise multiple CRL1-dependent 391 processes, in particular the nucleation of a catalytic assembly, involving intricate protein-protein 392 interactions between NEDD8, CUL1, ubiquitin-charged E2 and substrate receptor. This synergistic 393 assembly then steers the ubiquitin C-terminus towards a substrate lysine for priming with ubiquitin [66]. 394 Accordingly, our cryo-EM studies might indicate that similar principles apply for CRL4-catalysed 395 ubiquitylation. However, to unravel the catalytic architecture of CRL4, sophisticated cross-linking 396 procedures as in reference (65) will have to be pursued. 397 Intrinsic mobility of CRL4 stalk elements might assist the accommodation of a variety of sizes and 398 shapes of substrates in the CRL4 ubiquitylation zone and might rationalise the wide substrate range 399 accessible to CRL4 ubiquitylation through multiple DCAF receptors. Owing to selective pressure to 400 counteract the host’s SAMHD1 restriction, HIV-2 and certain SIVs, amongst other viruses, have taken 401 advantage of this dynamic CRL4 architecture by modification of the DCAF1 substrate receptor with 402 Vpx/Vpr-family accessory proteins. By tethering either SAMHD1-CtD or -NtD to DCAF1, and in this 403 way flexibly recruiting the bulk of SAMHD1, the accessibility of lysine side chains both tether-proximal 404 and on the SAMHD1 globular domains to the CRL4 catalytic assembly might be further improved (Fig 405 6C, D). This ensures efficient Vpx/Vpr-mediated SAMHD1 priming, poly-ubiquitylation and 406 proteasomal degradation to stimulate virus replication. 407 408 Methods 409 Protein expression and purification 410 Constructs were PCR-amplified from cDNA templates and inserted into the indicated expression 411 plasmids using standard restriction enzyme methods (S2 Table). pAcGHLT-B-DDB1 (plasmid #48638) 412 and pET28-UBA1 (plasmid #32534) were obtained from Addgene. The pOPC-UBA3-GST-APPBP1 413 co-expression plasmid, and the pGex6P2-UBC12 plasmid were obtained from MRC-PPU Reagents and 414 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 17 Services (clones 32498, 3879). Bovine erythrocyte ubiquitin and recombinant hsNEDD8 were 415 purchased from Sigma-Aldrich (U6253) and BostonBiochem (UL-812) respectively. Point mutations 416 were introduced by site-directed mutagenesis using KOD polymerase (Novagen). All constructs and 417 variants are summarised S3 Table. 418 Proteins expressed from vectors pAcGHLT-B, pGex6P1/2, pOPC and pET49b contained an N-terminal 419 GST-His-tag; pHisSUMO – N-terminal His-SUMO-tag; pET28, pRSF-Duet-1 – N-terminal His-tag; 420 pTri-Ex-6 – C-terminal His-tag. Constructs in vectors pAcGHLT-B and pTri-Ex-6 were expressed in 421 Sf9 cells, and constructs in vectors pET28, pET49b, pGex6P1/2, pRSF-Duet-1, and pHisSUMO in E. 422 coli Rosetta 2(DE3). 423 Recombinant baculoviruses (Autographa californica nucleopolyhedrovirus clone C6) were generated 424 as described previously [67]. Sf9 cells were cultured in Insect-XPRESS medium (Lonza) at 28°C in an 425 Innova 42R incubator shaker (New Brunswick) at a shaking speed of 180 rpm. In a typical preparation, 426 1 L of Sf9 cells at 3×106 cells/mL were co-infected with 4 mL of high titre DDB1 virus and 4 mL of 427 high titre DCAF1-CtD virus for 72 h. 428 For a typical E. coli Rosetta 2 (DE3) expression, 2 L of LB medium was inoculated with 20 mL of an 429 overnight culture and grown in a Multitron HT incubator shaker (Infors) at 37°C, 150 rpm until OD600 430 reached 0.7. At that point, temperature was reduced to 18°C, protein expression was induced by addition 431 of 0.2 mM IPTG, and cultures were grown for further 20 h. During co-expression of CUL4 and ROC1 432 from pRSF-Duet, 50 µM zinc sulfate was added to the growth medium before induction. 433 Sf9 cells were pelleted by centrifugation at 1000 rpm, 4°C for 30 min using a JLA 9.1000 centrifuge 434 rotor (Beckman). E. coli cells were pelleted by centrifugation at 4000 rpm, 4°C for 15 min using the 435 same rotor. Cell pellets were resuspended in buffer containing 50 mM Tris, pH 7.8, 500 mM NaCl, 4 436 mM MgCl2, 0.5 mM tris-(2-carboxyethyl)-phosphine (TCEP), mini-complete protease inhibitors (1 437 tablet per 50 mL) and 20 mM imidazole (for His-tagged proteins only). 100 mL of lysis buffer was used 438 for resuspension of a pellet from 1 L Sf9 culture, and 35 mL lysis buffer per pellet from 1 L E. coli 439 culture. Before resuspension of CUL4/ROC1 co-expression pellets, the buffer pH was adjusted to 8.5. 440 5 µL Benzonase (Merck) was added and the cells lysed by passing the suspension at least twice through 441 a Microfluidiser (Microfluidics). Lysates were clarified by centrifugation at 48000xg for 45 min at 4°C. 442 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 18 Protein purification was performed at 4°C on an Äkta pure FPLC (GE) using XK 16/20 chromatography 443 columns (GE) containing 10 mL of the appropriate affinity resin. GST-tagged proteins were captured 444 on glutathione-Sepharose (GSH-Sepharose FF, GE), washed with 250 mL of wash buffer (50 mM Tris-445 HCl pH 7.8, 500 mM NaCl, 4 mM MgCl2, 0.5 mM TCEP), and eluted with the same buffer 446 supplemented with 20 mM reduced glutathione. His-tagged proteins were immobilised on Ni-Sepharose 447 HP (GE), washed with 250 mL of wash buffer supplemented with 20 mM imidazole, and eluted with 448 wash buffer containing 0.3 M imidazole. Eluent fractions were analysed by SDS-PAGE, and appropriate 449 fractions were pooled and reduced to 5 mL using centrifugal filter devices (Vivaspin). If applicable, 100 450 µg GST-3C protease, or 50 µg thrombin, per mg total protein, was added and the sample was incubated 451 for 12 h on ice to cleave off affinity tags. As second purification step, gel filtration chromatography 452 (GF) was performed on an Äkta prime plus FPLC (GE), with Superdex 200 16/600 columns (GE), 453 equilibrated in 10 mM Tris-HCl pH 7.8, 150 mM NaCl, 4 mM MgCl2, 0.5 mM TCEP buffer, at a flow 454 rate of 1 mL/min. For purification of the CUL4/ROC1 complex, the pH of all purification buffers was 455 adjusted to 8.5. Peak fractions were analysed by SDS-PAGE, appropriate fractions were pooled and 456 concentrated to approx. 20 mg/mL, flash-frozen in liquid nitrogen in small aliquots and stored at -80°C. 457 Protein concentrations were determined with a NanoDrop spectrophotometer (ND 1000, Peqlab), using 458 theoretical absorption coefficients calculated based upon the amino acid sequence by ProtParam on the 459 ExPASy webserver [68]. 460 461 Analytical gel filtration analysis 462 Prior to gel filtration analysis affinity tags were removed by incubation of 30 µg GST-3C protease with 463 6 µM of each protein component in a volume of 120 µL wash buffer, followed by incubation on ice for 464 12 h. In order to remove the cleaved GST-tag and GST-3C protease, 20 μL GSH-Sepharose FF beads 465 (GE) were added and the sample was rotated at 4 °C for one hour. GSH-Sepharose beads were removed 466 by centrifugation at 4°C, 3500 rpm for 5 min, and 120 µL of the supernatant was loaded on an analytical 467 GF column (Superdex 200 10/300 GL, GE), equilibrated in 10 mM Tris-HCl pH 7.8, 150 mM NaCl, 468 4 mM MgCl2, 0.5 mM TCEP, at a flow rate of 0.5 mL/min. 1 mL fractions were collected and analysed 469 by SDS-PAGE. 470 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 19 471 In vitro ubiquitylation assays 472 160 µL reactions were prepared, containing 0.5 µM substrate (indicated SAMHD1 constructs, S2 Fig), 473 0.125 µM DDB1/DCAF1-CtD, 0.125 µM CUL4/ROC1, 0.125 µM HisSUMO-T4L-Vprmus (residues 1-474 92), 0.25 µM UBCH5C, 15 µM ubiquitin in 20 mM Tris-HCl pH 7.8, 150 mM NaCl, 2.5 mM MgCl2, 475 2.5 mM ATP. In control reactions, certain components were left out as indicated in S2 Fig. A 30 µl 476 sample for SDS-PAGE analysis was taken (t=0). Reactions were initiated by addition of 0.05 µM UBA1, 477 incubated at 37°C, and 30 µl SDS-PAGE samples were taken after 1 min, 2 min, 5 min and 15 min, 478 immediately mixed with 10 µl 4x SDS sample buffer and boiled at 95°C for 5 min. Samples were 479 analysed by SDS-PAGE. 480 481 In vitro neddylation of CUL4/ROC1 482 For initial neddylation tests, a 200 µL reaction was prepared, containing 8 µM CUL4/ROC1, 1.8 µM 483 UBC12, 30 µM NEDD8 in 50 mM Tris-HCl pH 7.8, 150 mM NaCl, 2.5 mM MgCl2, 2.5 mM ATP. 2x 484 30 µL samples were taken for SDS-PAGE, one was immediately mixed with 10 µL 4x SDS sample 485 buffer, the other one incubated for 60 min at 25°C. The reaction was initiated by addition of 0.7 µM 486 APPBP1/UBA3, incubated at 25°C, and 30 µL SDS-PAGE samples were taken after 1 min, 5 min, 487 10 min, 30 min and 60 min, immediately mixed with 10 µL 4x SDS sample buffer and boiled at 95°C 488 for 5 min. Samples were analysed by SDS-PAGE. Based on this test, the reaction was scaled up to 1 mL 489 and incubated for 5 min at 25°C. Reaction was quenched by addition of 5 mM TCEP and immediately 490 loaded onto a Superdex 200 16/600 GF column (GE), equilibrated in 10 mM Tris-HCl pH 7.8, 150 mM 491 NaCl, 4 mM MgCl2, 0.5 mM TCEP at a flow rate of 1 mL/min. Peak fractions were analysed by SDS-492 PAGE, appropriate fractions were pooled and concentrated to ~20 mg/mL, flash-frozen in liquid 493 nitrogen in small aliquots and stored at -80°C. 494 495 X-ray crystallography sample preparation, crystallisation, data collection and structure 496 solution 497 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 20 DDB1/DCAF1-CtD complex. DDB1/DCAF1-CtD crystals were grown by the hanging drop vapour 498 diffusion method, by mixing equal volumes (1 µL) of DDB1/DCAF1-CtD solution at 10 mg/mL with 499 reservoir solution containing 100 mM Tri-Na citrate pH 5.5, 18% PEG 1000 and suspending over a 500 500 µl reservoir. Crystals grew over night at 18°C. Crystals were cryo-protected in reservoir solution 501 supplemented with 20% glycerol and cryo-cooled in liquid nitrogen. A data set from a single crystal was 502 collected at Diamond Light Source (Didcot, UK) at a wavelength of 0.92819 Å. Data were processed 503 using XDS [69] (S1 Table), and the structure was solved using molecular replacement with the program 504 MOLREP [70] and available structures of DDB1 (PDB 3e0c) and DCAF1-CtD (PDB 4cc9) [50] as 505 search models. Iterative cycles of model adjustment with the program Coot [71], followed by refinement 506 using the program PHENIX [72] yielded final R/Rfree factors of 22.0%/27.9% (S1 Table). In the model, 507 94.5 % of residues have backbone dihedral angles in the favoured region of the Ramachandran plot, the 508 remainder fall in the allowed regions, and none are outliers. Details of data collection and refinement 509 statistics are presented in S1 Table. Coordinates and structure factors have been deposited in the PDB, 510 accession number 6zue. 511 DDB1/DCAF1-CtD/T4L-Vprmus (1-92) complex. The DDB1/DCAF1-CtD/Vprmus complex was 512 assembled by incubation of purified DDB1/DCAF1-CtD and HisSUMO-T4L-Vprmus (residues 1-92), at 513 a 1:1 molar ratio, in a buffer containing 50 mM Bis-tris propane pH 8.5, 0.5 M NaCl, 4 mM MgCl2, 0.5 514 mM TCEP, containing 1 mg of HRV-3C protease for HisSUMO-tag removal. After incubation on ice 515 for 12 h, the sample was loaded onto a Superdex 200 16/600 GF column (GE), with a 1 mL GSH-516 Sepharose FF column (GE) connected in line. The column was equilibrated with 10 mM Bis-tris propane 517 pH 8.5, 150 mM NaCl, 4 mM MgCl2, and 0.5 mM TCEP. The column flow rate was 1 mL/min. GF 518 fractions were analysed by SDS-PAGE, appropriate fractions were pooled and concentrated to 4.5 519 mg/mL. 520 Crystals were prepared by the sitting drop vapour diffusion method, by mixing equal volumes (200 nL) 521 of the protein complex at 4.5 mg/mL and reservoir solution containing 8-10% PEG 4000 (w/v), 200 mM 522 MgCl2, 100 mM HEPES-NaOH, pH 7.0-8.2. The reservoir volume was 75 µL. Crystals grew after at 523 least 4 weeks of incubation at 4°C. Crystals were cryo-protected in reservoir solution supplemented with 524 20% glycerol and cryo-cooled in liquid nitrogen. Data sets from two single crystals were collected, 525 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 21 initially at BESSY II (Helmholtz-Zentrum Berlin, HZB) at a wavelength of 0.91841 Å, and later at 526 ESRF (Grenoble) at a wavelength of 1 Å. Data sets were processed separately using XDS [69] and 527 XDSAPP [73]. The structure was solved by molecular replacement, using the initial BESSY data set, 528 with the program PHASER [74], and the following structures as search models: DDB1/DCAF1-CtD 529 (this work) and T4L variant E11H (PDB 1qt6) [75]. After optimisation of the initial model and 530 refinement against the higher-resolution ESRF data set, Vprmus was placed manually into the density, 531 using an NMR model of VprHIV-1 (PDB 1m8l) [76] as guidance. Iterative cycles of model adjustment 532 with the program Coot [71], followed by refinement using the program PHENIX [72] yielded final 533 R/Rfree factors of 21.61%/26.05%. In the model, 95.1 % of residues have backbone dihedral angles in 534 the favoured region of the Ramachandran plot, the remainder fall in the allowed regions, and none are 535 outliers. Details of data collection and refinement statistics are presented in S1 Table. Coordinates and 536 structure factors have been deposited in the PDB, accession number 6zx9. 537 538 Cryo-EM sample preparation and data collection 539 Complex assembly. Purified CUL4-NEDD8/ROC1, DDB1/DCAF1-CtD, GST-Vprmus and rhesus 540 macaque SAMHD1, 1 µM each, were incubated in a final volume of 1 mL of 10 mM Tris-HCl pH 7.8, 541 150 mM NaCl, 4 mM MgCl2, 0.5 mM TCEP, supplemented with 1 mg of GST-3C protease. After 542 incubation on ice for 12 h, the sample was loaded onto a Superdex 200 16/600 GF column (GE), 543 equilibrated with the same buffer at 1 mL/min, with a 1 mL GSH-Sepharose FF column (GE) connected 544 in line. GF fractions were analysed by SDS-PAGE, appropriate fractions were pooled and concentrated 545 to 2.8 mg/mL. 546 Grid preparation. 3.5 µl protein solution containing 0.05 µM CUL4-NEDD8/ROC1/DDB1/DCAF1-547 CtD/Vprmus/SAMHD1 complex and 0.25 µM UBCH5C-ubiquitin conjugate (S4 A, B Fig) were applied 548 to a 300 mesh Quantifoil R2/4 Cu/Rh holey carbon grid (Quantifoil Micro Tools GmbH) coated with an 549 additional thin carbon film as sample support and stained with 2% uranyl acetate for initial 550 characterisation. For cryo-EM, a fresh 400 mesh Quantifoil R1.2/1.3 Cu holey carbon grid (Quantifoil 551 Micro Tools GmbH) was glow-discharged for 30 s using a Harrick plasma cleaner with technical air at 552 0.3 mbar and 7 W. 3.5 µl protein solution containing 0.4 µM CUL4-NEDD8/ROC1/DDB1/DCAF1-553 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 22 CtD/Vprmus/SAMHD1 complex and 2 µM UBCH5C-ubiquitin conjugate were applied to the grid, 554 incubated for 45 s, blotted with a Vitrobot Mark II device (FEI, Thermo Fisher Scientific) for 1-2 s at 555 8°C and 80% humidity, and plunged in liquid ethane. Grids were stored in liquid nitrogen until imaging. 556 Cryo-EM data collection. Initial negative stain and cryo-EM datasets were collected automatically 557 for sample quality control and low-resolution reconstructions on a 120 kV Tecnai Spirit cryo-EM (FEI, 558 Thermo Fisher Scientific) equipped with a F416 CMOS camera (TVIPS) using Leginon [77, 78]. 559 Particle images were then analysed by 2D classification and initial model reconstruction using SPHIRE 560 [79], cisTEM [80] and Relion 3.07 [81]. These data revealed the presence of the complexes containing 561 both DDB1/DCAF1-CtD/Vprmus (core) and CUL4/ROC1 (stalk). High-resolution data was collected on 562 a 300 kV Tecnai Polara cryo-EM (FEI, Thermo Fisher Scientific) equipped with a K2summit direct 563 electron detector (Gatan) at a nominal magnification of 31000x, with a pixel size of 0.625 Å/px on the 564 object scale. In total, 3644 movie stacks were collected in super-resolution mode using Leginon [77, 78] 565 with the following parameters: defocus range of 0.5-3.0 µm, 40 frames per movie, 10 s exposure time, 566 electron dose of 1.25 e/Å2/s and a cumulative dose of 50 e/Å2 per movie. 567 568 Cryo-EM computational analysis 569 Movies were aligned and dose-weighted using MotionCor2 [82] and initial estimation of the contrast 570 transfer function (CTF) was performed with the CTFFind4 package [83]. Resulting micrographs were 571 manually inspected to exclude images with substantial contaminants (typically large protein aggregates 572 or ice contaminations) or grid artefacts. Power spectra were manually inspected to exclude images with 573 astigmatic, weak, or poorly defined spectra. After these quality control steps the dataset included 2322 574 micrographs (63% of total). At this stage, the data set was picked twice and processed separately, to 575 yield reconstructions of the core (analysis 1) and states-1, -2 and -3 (analysis 2). 576 For analysis 1, particle positions were determined using template matching with a filtered map 577 comprising core and stalk using the software Gautomatch (https://www2.mrc-578 lmb.cam.ac.uk/research/locally-developed-software/zhang-software/). 712,485 particle images were 579 found, extracted with Relion 3.07 and subsequently 2D-classified using cryoSPARC [84], resulting in 580 505,342 particle images after selection (S4C, D Fig). These particle images were separated into two 581 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 23 equally sized subsets and Tier 1 3D-classification was performed using Relion 3.07 on both of them to 582 reduce computational burden (S4D Fig). The following parameters were used: initial model=“core”, 583 number of classes K=4, T=10, global step search=7.5°, number of iterations=25, pixel size 3.75 Å/px. 584 From these, the ones possessing both core and stalk were selected. Classes depicting a similar stalk 585 orientation relative to the core were pooled and directed into Tier 2 as three different subpopulations 586 containing 143,172, 193,059 and 167,666 particle images, respectively (S4D Fig). 587 For Tier 2, each subpopulation was classified separately into 4 classes each. From these 12 classes, all 588 particle images exhibiting well-defined densities for core and stalk were pooled and labelled 589 “core+stalk”, resulting in 310,801 particle images in total. 193,096 particle images representing classes 590 containing only the core were pooled and labelled “core” (S4D Fig) 591 For Tier 3, the “core” particle subset was separated into 4 classes which yielded uninterpretable 592 reconstructions lacking medium- or high-resolution features. The “core+stalk” subset was separated into 593 6 classes, with 5 classes containing both stalk and core (S4D Fig) and one class consisting only of the 594 core with Vprmus bound. The 5 classes with stalk showed similar stalk orientations as the ones obtained 595 from analysis 2 (see below, S5 Fig), but refined individually to lower resolution as in analysis 2 and 596 were discarded. However, individual refinement of the core-only tier 3 class yielded a 7.3 Å 597 reconstruction (S4E, F Fig). 598 For analysis 2, particle positions were determined using cisTEMs Gaussian picking routine, yielding 599 959,155 particle images in total. After two rounds of 2D-classification, 227,529 particle images were 600 selected for further processing (S4G, H Fig). Using this data, an initial model was created using Relion 601 3.07. The resulting map yielded strong signal for the core but only fragmented stalk density, indicating 602 a large heterogeneity in the stalk-region within the data set. This large degree of compositional (+/- 603 stalk) and conformational heterogeneity (movement of the stalk relative to the core) made the 604 classification challenging. Accordingly, alignment and classification were carried out simultaneously. 605 The first objective was to separate the data set into three categories: “junk”, “core” and “core+stalk”. 606 Therefore, the stalk was deleted from the initial model using the “Eraser”-tool in Chimera [85]. This 607 core-map was used as an initial model for the Tier 1 3D-classification with Relion 3.07 at a decimated 608 pixel size of 2.5 Å/px. The following parameters were used: number of classes K=6, T=10, global step 609 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 24 search=7.5°, number of iterations= 25. The classification yielded two classes containing the stalk 610 (classes 3 and 5 containing 23% and 22% of the particle images, respectively) (S4H Fig). These particles 611 were pooled and directed into Tier 2 3D-classification using the following parameters: number of classes 612 K=6, T=10, global step search=7.5°, number of iterations=25. Three of these classes yielded medium-613 resolution maps with interpretable features (states-1, -2 and -3, S4H Fig). These three classes were 614 refined individually using 3D Relion 3.07, resulting in maps with resolution ranging from 7.8 Å – 8.9 615 Å (S4H-J Fig). 616 617 Molecular visualisation, rigid body fitting, 3D structural alignments, rotation and 618 interface analysis 619 Density maps and atomic models were visualised using Coot [71], PyMOL (Schrödinger) and UCSF 620 Chimera [85]. Rigid body fits and structural alignments were performed using the program UCSF 621 Chimera [85]. Rotation angles between extreme DDB1 BPB domain positions were measured using the 622 DynDom server [86] (http://dyndom.cmp.uea.ac.uk/dyndom/runDyndom.jsp). Molecular interfaces 623 were analysed using the EBI PDBePISA server [87] (https://www.ebi.ac.uk/msd-srv/prot_int/cgi-624 bin/piserver). 625 626 Multiple sequence alignment 627 A multiple sequence alignment was calculated using the EBI ClustalOmega server [88] 628 (https://www.ebi.ac.uk/Tools/msa/clustalo/), and adjusted manually using the program GeneDoc [89]. 629 630 Cross-linking mass spectrometry (CLMS) 631 Complex assembly. Purified CUL4/ROC1, DDB1/DCAF1-CtD, GST-Vprmus and rhesus macaque 632 SAMHD1, 1 µM each, were incubated in a volume of 3 mL buffer containing 10 mM HEPES pH 7.8, 633 150 mM NaCl, 4 mM MgCl2, 0.5 mM TCEP, supplemented with 1 mg GST-3C protease. After 634 incubation on ice for 12 h, the sample was loaded onto a Superdex 200 16/600 GF column (GE), 635 equilibrated with the same buffer, at a flow rate of 1 mL/min with a 1 mL GSH-Sepharose FF column 636 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 25 (GE) connected in line. GF fractions were analysed by SDS-PAGE, appropriate fractions were pooled 637 and concentrated to 6 mg/mL. 638 Photo-Crosslinking. The cross-linker sulfo-SDA (sulfosuccinimidyl 4,4′-azipentanoate) (Thermo 639 Scientific) was dissolved in cross-linking buffer (10 mM HEPES pH 7.8, 150 mM NaCl, 4 mM 640 MgCl2, 0.5 mM TCEP) to 100 mM before use. The labelling step was performed by incubating 641 18 μg aliquots of the complex at 1 mg/mL with 2, 1, 0.5, 0.25, 0.125 mM sulfo-SDA, added, 642 respectively, for an hour. The samples were then irradiated with UV light at 365 nm, to form cross-643 links, for 20 min and quenched with 50 mM NH4HCO3 for 20 min. All steps were performed on 644 ice. Reaction products were separated on a Novex Bis-Tris 4–12% SDS−PAGE gel (Life 645 Technologies). The gel band corresponding to the cross-linked complex was excised and digested 646 with trypsin (Thermo Scientific Pierce) [90] and the resulting tryptic peptides were extracted and 647 desalted using C18 StageTips [91]. Eluted peptides were fractionated on a Superdex Peptide 3.2/300 648 increase column (GE Healthcare) at a flow rate of 10 µL/min using 30% (v/v) acetonitrile and 0.1 649 % (v/v) trifluoroacetic acid as mobile phase. 50 μL fractions were collected and vacuum-dried. 650 CLMS acquisition. Samples for analysis were resuspended in 0.1% (v/v) formic acid, 3.2% (v/v) 651 acetonitrile. LC-MS/MS analysis was performed on an Orbitrap Fusion Lumos Tribrid mass 652 spectrometer (Thermo Fisher) coupled on-line with an Ultimate 3000 RSLCnano HPLC system 653 (Dionex, Thermo Fisher). Samples were separated on a 50 cm EASY-Spray column (Thermo Fisher). 654 Mobile phase A consisted of 0.1% (v/v) formic acid and mobile phase B of 80% (v/v) acetonitrile with 655 0.1% (v/v) formic acid. Flow rates were 0.3 μL/min using gradients optimized for each chromatographic 656 fraction from offline fractionation, ranging from 2% mobile phase B to 55% mobile phase B over 657 90 min. MS data were acquired in data-dependent mode using the top-speed setting with a 3 s cycle 658 time. For every cycle, the full scan mass spectrum was recorded using the Orbitrap at a resolution of 659 120,000 in the range of 400 to 1,500 m/z. Ions with a precursor charge state between 3+ and 7+ were 660 isolated and fragmented. Analyte fragmentation was achieved by Higher-Energy Collisional 661 Dissociation (HCD) [92] and fragmentation spectra were then recorded in the Orbitrap with a resolution 662 of 50,000. Dynamic exclusion was enabled with single repeat count and 60 s exclusion duration. 663 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 26 CLMS processing. A recalibration of the precursor m/z was conducted based on high-confidence 664 (<1% false discovery rate (FDR)) linear peptide identifications. The re-calibrated peak lists were 665 searched against the sequences and the reversed sequences (as decoys) of cross-linked peptides using 666 the Xi software suite (v.1.7.5.1) for identification [93]. Final crosslink lists were compiled using the 667 identified candidates filtered to <1% FDR on link level with xiFDR v.2.0 [94] imposing a minimum of 668 20% sequence coverage and 4 observed fragments per peptide. 669 CLMS analysis. In order to sample the accessible interaction volume of the SAMHD1-CtD consistent 670 with CLMS data, a model for SAMHD1 was generated using I-TASSER [95]. The SAMHD1-CtD, 671 which adopted a random coil configuration, was extracted from the model. In order to map all crosslinks, 672 missing loops in the complex structure were generated using MODELLER [96]. An interaction volume 673 search was then submitted to the DisVis webserver [97] with an allowed distance between 1.5 Å and 22 674 Å for each restraint using the "complete scanning" option. The rotational sampling interval was set to 675 9.72° and the grid voxel spacing to 1Å. The accessible interaction volume was visualised using UCSF 676 Chimera [85]. 677 678 Acknowledgments 679 We thank the MPI-MG for granting access to the TEM instruments of the microscopy and cryo-EM 680 service group. We thank Manfred Weiss and the scientific staff of the BESSY-MX (Macromolecular X-681 ray Crystallography)/Helmholtz Zentrum Berlin für Materialien und Energie at beamlines BL14.1, 682 BL14.2, and BL14.3 operated by the Joint Berlin MX-Laboratory at the BESSY II electron storage ring 683 (Berlin-Adlershof, Germany) as well as the scientific staff of the ESRF (Grenoble, France) at beamlines 684 ID30A-3, ID30B, ID23-1, ID23-2, and ID29 for continuous support. We acknowledge Diamond Light 685 Source (Didcot, UK) for access and support of the synchrotron beamline I04 and cryo-EM facilities at 686 the UK's national Electron Bio-imaging Centre (eBIC). Furthermore, the authors acknowledge the 687 North-German Supercomputing Alliance (HLRN) and the HPC for Research cluster of the Berlin 688 Institute of Health for providing HPC resources. The pHisSUMO plasmid was a generous gift from Dr. 689 Evangelos Christodoulou (The Francis Crick Institute, UK). The rhesus macaque SAMHD1 cDNA 690 template was a generous gift from Prof. Michael Emerman (Fred Hutchinson Cancer Research Center, 691 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 27 Seattle, USA). Recombinant BAC10:1629KO bacmid was a generous gift from Prof. Ian Jones 692 (University of Reading, UK). pAcGHLT-B-DDB1 was a gift from Ning Zheng (Addgene plasmid 693 48638). pET28-mE1 was a gift from Jorge Eduardo Azevedo (Addgene plasmid 32534). 694 695 Data availability 696 The coordinates and structure factors for the crystal structures have been deposited at the Protein Data 697 Bank (PDB) with the accession codes 6ZUE (DDB1/DCAF1-CtD) and 6ZX9 (DDB1/DCAF1-698 CtD/T4L-Vprmus 1-92). Cryo-EM reconstructions have been deposited at the Electron Microscopy Data 699 Bank (EMDB) with the accession codes EMD-10611 (core), EMD-10612 (conformational state-1), 700 EMD-10613 (state-2) and EMD-10614 (state-3). CLMS data have been deposited at the PRIDE database 701 [98] with the accession code PXD020453, reviewer password fCrQG2u8. 702 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 28 References 703 1. Randow F, Lehner PJ. Viral avoidance and exploitation of the ubiquitin system. Nat Cell Biol. 704 2009;11(5):527-34. doi: 10.1038/ncb0509-527. 705 2. Isaacson MK, Ploegh HL. Ubiquitination, ubiquitin-like modifiers, and deubiquitination in viral 706 infection. Cell host & microbe. 2009;5(6):559-70. doi: 10.1016/j.chom.2009.05.012. 707 3. Gustin JK, Moses AV, Fruh K, Douglas JL. Viral takeover of the host ubiquitin system. Front 708 Microbiol. 2011;2:161. doi: 10.3389/fmicb.2011.00161. 709 4. Barry M, Fruh K. Viral modulators of cullin RING ubiquitin ligases: culling the host defense. Science's 710 STKE : signal transduction knowledge environment. 2006;2006(335):pe21. Epub 2006/05/18. doi: 711 10.1126/stke.3352006pe21. 712 5. Mahon C, Krogan NJ, Craik CS, Pick E. Cullin E3 ligases and their rewiring by viral factors. 713 Biomolecules. 2014;4(4):897-930. Epub 2014/10/15. doi: 10.3390/biom4040897. 714 6. Becker T, Le-Trilling VTK, Trilling M. Cellular Cullin RING Ubiquitin Ligases: Druggable Host 715 Dependency Factors of Cytomegaloviruses. Int J Mol Sci. 2019;20(7). doi: 10.3390/ijms20071636. 716 7. Seissler T, Marquet R, Paillart JC. Hijacking of the Ubiquitin/Proteasome Pathway by the HIV 717 Auxiliary Proteins. Viruses. 2017;9(11). doi: 10.3390/v9110322. 718 8. Zheng N, Shabek N. Ubiquitin Ligases: Structure, Function, and Regulation. Annu Rev Biochem. 719 2017;86:14.1-29. 720 9. Sauter D, Kirchhoff F. Key Viral Adaptations Preceding the AIDS Pandemic. Cell host & microbe. 721 2019;25(1):27-38. doi: 10.1016/j.chom.2018.12.002. 722 10. Sharp PM, Hahn BH. Origins of HIV and the AIDS pandemic. Cold Spring Harbor perspectives in 723 medicine. 2011;1(1):a006841. Epub 2012/01/10. doi: 10.1101/cshperspect.a006841. 724 11. Hatziioannou T, Del Prete GQ, Keele BF, Estes JD, McNatt MW, Bitzegeio J, et al. HIV-1-induced 725 AIDS in monkeys. Science. 2014;344(6190):1401-5. Epub 2014/06/21. doi: 10.1126/science.1250761. 726 12. Malim MH, Bieniasz PD. HIV Restriction Factors and Mechanisms of Evasion. Cold Spring Harbor 727 perspectives in medicine. 2012;2(5):a006940. Epub 2012/05/04. doi: 10.1101/cshperspect.a006940. 728 13. Fischer ES, Scrima A, Bohm K, Matsumoto S, Lingaraju GM, Faty M, et al. The molecular basis of 729 CRL4DDB2/CSA ubiquitin ligase architecture, targeting, and activation. Cell. 2011;147(5):1024-39. Epub 730 2011/11/29. doi: 10.1016/j.cell.2011.10.035. 731 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 29 14. Lee J, Zhou P. DCAFs, the missing link of the CUL4-DDB1 ubiquitin ligase. Molecular cell. 732 2007;26(6):775-80. Epub 2007/06/26. doi: 10.1016/j.molcel.2007.06.001. 733 15. Angers S, Li T, Yi X, MacCoss MJ, Moon RT, Zheng N. Molecular architecture and assembly of the 734 DDB1-CUL4A ubiquitin ligase machinery. Nature. 2006;443(7111):590-3. Epub 2006/09/12. doi: 735 10.1038/nature05175. 736 16. Scrima A, Konickova R, Czyzewski BK, Kawasaki Y, Jeffrey PD, Groisman R, et al. Structural basis of 737 UV DNA-damage recognition by the DDB1-DDB2 complex. Cell. 2008;135(7):1213-23. Epub 2008/12/27. doi: 738 10.1016/j.cell.2008.10.045. 739 17. Zimmerman ES, Schulman BA, Zheng N. Structural assembly of cullin-RING ubiquitin ligase 740 complexes. Current opinion in structural biology. 2010;20(6):714-21. Epub 2010/10/01. doi: 741 10.1016/j.sbi.2010.08.010. 742 18. Andrejeva J, Young DF, Goodbourn S, Randall RE. Degradation of STAT1 and STAT2 by the V 743 proteins of simian virus 5 and human parainfluenza virus type 2, respectively: consequences for virus replication 744 in the presence of alpha/beta and gamma interferons. Journal of virology. 2002;76(5):2159-67. doi: 745 10.1128/jvi.76.5.2159-2167.2002. 746 19. Li T, Chen X, Garbutt KC, Zhou P, Zheng N. Structure of DDB1 in complex with a paramyxovirus V 747 protein: viral hijack of a propeller cluster in ubiquitin ligase. Cell. 2006;124(1):105-17. Epub 2006/01/18. doi: 748 10.1016/j.cell.2005.10.033. 749 20. Trilling M, Le VT, Fiedler M, Zimmermann A, Bleifuss E, Hengel H. Identification of DNA-damage 750 DNA-binding protein 1 as a conditional essential factor for cytomegalovirus replication in interferon-gamma-751 stimulated cells. PLoS pathogens. 2011;7(6):e1002069. doi: 10.1371/journal.ppat.1002069. 752 21. Paradkar PN, Duchemin JB, Rodriguez-Andres J, Trinidad L, Walker PJ. Cullin4 Is Pro-Viral during 753 West Nile Virus Infection of Culex Mosquitoes. PLoS pathogens. 2015;11(9):e1005143. doi: 754 10.1371/journal.ppat.1005143. 755 22. Decorsiere A, Mueller H, van Breugel PC, Abdul F, Gerossier L, Beran RK, et al. Hepatitis B virus X 756 protein identifies the Smc5/6 complex as a host restriction factor. Nature. 2016;531(7594):386-9. doi: 757 10.1038/nature17170. 758 23. Murphy CM, Xu Y, Li F, Nio K, Reszka-Blanco N, Li X, et al. Hepatitis B Virus X Protein Promotes 759 Degradation of SMC5/6 to Enhance HBV Replication. Cell reports. 2016;16(11):2846-54. doi: 760 10.1016/j.celrep.2016.08.026. 761 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 30 24. Lim ES, Fregoso OI, McCoy CO, Matsen FA, Malik HS, Emerman M. The ability of primate 762 lentiviruses to degrade the monocyte restriction factor SAMHD1 preceded the birth of the viral accessory protein 763 Vpx. Cell host & microbe. 2012;11(2):194-204. Epub 2012/01/31. doi: 10.1016/j.chom.2012.01.004. 764 25. Romani B, Cohen EA. Lentivirus Vpr and Vpx accessory proteins usurp the cullin4-DDB1 (DCAF1) 765 E3 ubiquitin ligase. Current opinion in virology. 2012;2(6):755-63. Epub 2012/10/16. doi: 766 10.1016/j.coviro.2012.09.010. 767 26. Fabryova H, Strebel K. Vpr and Its Cellular Interaction Partners: R We There Yet? Cells. 2019;8(11). 768 doi: 10.3390/cells8111310. 769 27. Greenwood EJD, Williamson JC, Sienkiewicz A, Naamati A, Matheson NJ, Lehner PJ. Promiscuous 770 Targeting of Cellular Proteins by Vpr Drives Systems-Level Proteomic Remodeling in HIV-1 Infection. Cell 771 reports. 2019;27(5):1579-96 e7. doi: 10.1016/j.celrep.2019.04.025. 772 28. Schrofelbauer B, Yu Q, Zeitlin SG, Landau NR. Human immunodeficiency virus type 1 Vpr induces the 773 degradation of the UNG and SMUG uracil-DNA glycosylases. Journal of virology. 2005;79(17):10978-87. doi: 774 10.1128/JVI.79.17.10978-10987.2005. 775 29. Lahouassa H, Blondot ML, Chauveau L, Chougui G, Morel M, Leduc M, et al. HIV-1 Vpr degrades the 776 HLTF DNA translocase in T cells and macrophages. Proceedings of the National Academy of Sciences of the 777 United States of America. 2016;113(19):5311-6. doi: 10.1073/pnas.1600485113. 778 30. Laguette N, Bregnard C, Hue P, Basbous J, Yatim A, Larroque M, et al. Premature activation of the 779 SLX4 complex by Vpr promotes G2/M arrest and escape from innate immune sensing. Cell. 2014;156(1-2):134-780 45. Epub 2014/01/15. doi: 10.1016/j.cell.2013.12.011. 781 31. Zhou X, DeLucia M, Ahn J. SLX4-SLX1 Protein-independent Down-regulation of MUS81-EME1 782 Protein by HIV-1 Viral Protein R (Vpr). The Journal of biological chemistry. 2016;291(33):16936-47. doi: 783 10.1074/jbc.M116.721183. 784 32. Romani B, Shaykh Baygloo N, Aghasadeghi MR, Allahbakhshi E. HIV-1 Vpr Protein Enhances 785 Proteasomal Degradation of MCM10 DNA Replication Factor through the Cul4-DDB1[VprBP] E3 Ubiquitin 786 Ligase to Induce G2/M Cell Cycle Arrest. The Journal of biological chemistry. 2015;290(28):17380-9. doi: 787 10.1074/jbc.M115.641522. 788 33. Lv L, Wang Q, Xu Y, Tsao LC, Nakagawa T, Guo H, et al. Vpr Targets TET2 for Degradation by 789 CRL4(VprBP) E3 Ligase to Sustain IL-6 Expression and Enhance HIV-1 Replication. Molecular cell. 790 2018;70(5):961-70 e5. doi: 10.1016/j.molcel.2018.05.007. 791 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 31 34. Su J, Rui Y, Lou M, Yin L, Xiong H, Zhou Z, et al. HIV-2/SIV Vpx targets a novel functional domain 792 of STING to selectively inhibit cGAS-STING-mediated NF-kappaB signalling. Nat Microbiol. 2019;4(12):2552-793 64. doi: 10.1038/s41564-019-0585-4. 794 35. Chougui G, Munir-Matloob S, Matkovic R, Martin MM, Morel M, Lahouassa H, et al. HIV-2/SIV viral 795 protein X counteracts HUSH repressor complex. Nat Microbiol. 2018;3(8):891-7. doi: 10.1038/s41564-018-796 0179-6. 797 36. Yurkovetskiy L, Guney MH, Kim K, Goh SL, McCauley S, Dauphin A, et al. Primate 798 immunodeficiency virus proteins Vpx and Vpr counteract transcriptional repression of proviruses by the HUSH 799 complex. Nat Microbiol. 2018;3(12):1354-61. doi: 10.1038/s41564-018-0256-x. 800 37. Hrecka K, Hao C, Gierszewska M, Swanson SK, Kesik-Brodacka M, Srivastava S, et al. Vpx relieves 801 inhibition of HIV-1 infection of macrophages mediated by the SAMHD1 protein. Nature. 2011;474(7353):658-802 61. Epub 2011/07/02. doi: 10.1038/nature10195. 803 38. Laguette N, Sobhian B, Casartelli N, Ringeard M, Chable-Bessia C, Segeral E, et al. SAMHD1 is the 804 dendritic- and myeloid-cell-specific HIV-1 restriction factor counteracted by Vpx. Nature. 2011;474(7353):654-805 7. Epub 2011/05/27. doi: 10.1038/nature10117. 806 39. Powell RD, Holland PJ, Hollis T, Perrino FW. Aicardi-Goutieres syndrome gene and HIV-1 restriction 807 factor SAMHD1 is a dGTP-regulated deoxynucleotide triphosphohydrolase. The Journal of biological chemistry. 808 2011;286(51):43596-600. Epub 2011/11/10. doi: 10.1074/jbc.C111.317628. 809 40. Goldstone DC, Ennis-Adeniran V, Hedden JJ, Groom HC, Rice GI, Christodoulou E, et al. HIV-1 810 restriction factor SAMHD1 is a deoxynucleoside triphosphate triphosphohydrolase. Nature. 811 2011;480(7377):379-82. Epub 2011/11/08. doi: 10.1038/nature10623. 812 41. Zhu C, Gao W, Zhao K, Qin X, Zhang Y, Peng X, et al. Structural insight into dGTP-dependent 813 activation of tetrameric SAMHD1 deoxynucleoside triphosphate triphosphohydrolase. Nature communications. 814 2013;4:2722. Epub 2013/11/13. doi: 10.1038/ncomms3722. 815 42. Kim B, Nguyen LA, Daddacha W, Hollenbaugh JA. Tight interplay among SAMHD1 protein level, 816 cellular dNTP levels, and HIV-1 proviral DNA synthesis kinetics in human primary monocyte-derived 817 macrophages. The Journal of biological chemistry. 2012;287(26):21570-4. Epub 2012/05/17. doi: 818 10.1074/jbc.C112.374843. 819 43. Lahouassa H, Daddacha W, Hofmann H, Ayinde D, Logue EC, Dragin L, et al. SAMHD1 restricts the 820 replication of human immunodeficiency virus type 1 by depleting the intracellular pool of deoxynucleoside 821 triphosphates. Nature immunology. 2012;13(3):223-8. Epub 2012/02/14. doi: 10.1038/ni.2236. 822 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 32 44. St Gelais C, de Silva S, Amie SM, Coleman CM, Hoy H, Hollenbaugh JA, et al. SAMHD1 restricts 823 HIV-1 infection in dendritic cells (DCs) by dNTP depletion, but its expression in DCs and primary CD4+ T-824 lymphocytes cannot be upregulated by interferons. Retrovirology. 2012;9:105. Epub 2012/12/13. doi: 825 10.1186/1742-4690-9-105. 826 45. Rehwinkel J, Maelfait J, Bridgeman A, Rigby R, Hayward B, Liberatore RA, et al. SAMHD1-827 dependent retroviral control and escape in mice. The EMBO journal. 2013;32(18):2454-62. Epub 2013/07/23. 828 doi: 10.1038/emboj.2013.163. 829 46. Morris ER, Taylor IA. The missing link: allostery and catalysis in the anti-viral protein SAMHD1. 830 Biochem Soc Trans. 2019;47(4):1013-27. doi: 10.1042/BST20180348. 831 47. Baldauf HM, Pan X, Erikson E, Schmidt S, Daddacha W, Burggraf M, et al. SAMHD1 restricts HIV-1 832 infection in resting CD4(+) T cells. Nature medicine. 2012;18(11):1682-7. Epub 2012/09/14. doi: 833 10.1038/nm.2964. 834 48. Shingai M, Welbourn S, Brenchley JM, Acharya P, Miyagi E, Plishka RJ, et al. The Expression of 835 Functional Vpx during Pathogenic SIVmac Infections of Rhesus Macaques Suppresses SAMHD1 in CD4+ 836 Memory T Cells. PLoS pathogens. 2015;11(5):e1004928. doi: 10.1371/journal.ppat.1004928. 837 49. Fregoso OI, Ahn J, Wang C, Mehrens J, Skowronski J, Emerman M. Evolutionary toggling of Vpx/Vpr 838 specificity results in divergent recognition of the restriction factor SAMHD1. PLoS pathogens. 839 2013;9(7):e1003496. Epub 2013/07/23. doi: 10.1371/journal.ppat.1003496. 840 50. Schwefel D, Groom HC, Boucherit VC, Christodoulou E, Walker PA, Stoye JP, et al. Structural basis of 841 lentiviral subversion of a cellular protein degradation pathway. Nature. 2014;505(7482):234-8. Epub 2013/12/18. 842 doi: 10.1038/nature12815. 843 51. Schwefel D, Boucherit VC, Christodoulou E, Walker PA, Stoye JP, Bishop KN, et al. Molecular 844 Determinants for Recognition of Divergent SAMHD1 Proteins by the Lentiviral Accessory Protein Vpx. Cell 845 host & microbe. 2015;17(4):489-99. Epub 2015/04/10. doi: 10.1016/j.chom.2015.03.004. 846 52. Wu Y, Koharudin LM, Mehrens J, DeLucia M, Byeon CH, Byeon IJ, et al. Structural Basis of Clade-847 specific Engagement of SAMHD1 (Sterile alpha Motif and Histidine/Aspartate-containing Protein 1) Restriction 848 Factors by Lentiviral Viral Protein X (Vpx) Virulence Factors. The Journal of biological chemistry. 849 2015;290(29):17935-45. doi: 10.1074/jbc.M115.665513. 850 53. Spragg CJ, Emerman M. Antagonism of SAMHD1 is actively maintained in natural infections of simian 851 immunodeficiency virus. Proceedings of the National Academy of Sciences of the United States of America. 852 2013;110(52):21136-41. Epub 2013/12/11. doi: 10.1073/pnas.1316839110. 853 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 33 54. Wu Y, Zhou X, Barnes CO, DeLucia M, Cohen AE, Gronenborn AM, et al. The DDB1-DCAF1-Vpr-854 UNG2 crystal structure reveals how HIV-1 Vpr steers human UNG2 toward destruction. Nature structural & 855 molecular biology. 2016;23(10):933-40. doi: 10.1038/nsmb.3284. 856 55. Enchev RI, Schulman BA, Peter M. Protein neddylation: beyond cullin-RING ligases. Nature reviews 857 Molecular cell biology. 2015;16(1):30-44. doi: 10.1038/nrm3919. 858 56. Schneider M, Belsom A, Rappsilber J. Protein Tertiary Structure by Crosslinking/Mass Spectrometry. 859 Trends in biochemical sciences. 2018;43(3):157-69. Epub 2018/02/06. doi: 10.1016/j.tibs.2017.12.006. 860 57. Duda DM, Borg LA, Scott DC, Hunt HW, Hammel M, Schulman BA. Structural insights into NEDD8 861 activation of cullin-RING ligases: conformational control of conjugation. Cell. 2008;134(6):995-1006. doi: 862 10.1016/j.cell.2008.07.022. 863 58. Fischer ES, Bohm K, Lydeard JR, Yang H, Stadler MB, Cavadini S, et al. Structure of the DDB1-864 CRBN E3 ubiquitin ligase in complex with thalidomide. Nature. 2014;512(7512):49-53. doi: 865 10.1038/nature13527. 866 59. DeLucia M, Mehrens J, Wu Y, Ahn J. HIV-2 and SIVmac accessory virulence factor Vpx down-867 regulates SAMHD1 enzyme catalysis prior to proteasome-dependent degradation. The Journal of biological 868 chemistry. 2013;288(26):19116-26. doi: 10.1074/jbc.M113.469007. 869 60. Berger G, Lawrence M, Hue S, Neil SJ. G2/M cell cycle arrest correlates with primate lentiviral Vpr 870 interaction with the SLX4 complex. Journal of virology. 2014. Epub 2014/10/17. doi: 10.1128/JVI.02307-14. 871 61. Guenzel CA, Herate C, Benichou S. HIV-1 Vpr-a still "enigmatic multitasker". Front Microbiol. 872 2014;5:127. doi: 10.3389/fmicb.2014.00127. 873 62. Stivahtis GL, Soares MA, Vodicka MA, Hahn BH, Emerman M. Conservation and host specificity of 874 Vpr-mediated cell cycle arrest suggest a fundamental role in primate lentivirus evolution and biology. Journal of 875 virology. 1997;71(6):4331-8. 876 63. Planelles V, Jowett JB, Li QX, Xie Y, Hahn B, Chen IS. Vpr-induced cell cycle arrest is conserved 877 among primate lentiviruses. Journal of virology. 1996;70(4):2516-24. 878 64. Schapira M, Calabrese MF, Bullock AN, Crews CM. Targeted protein degradation: expanding the 879 toolbox. Nat Rev Drug Discov. 2019;18(12):949-63. doi: 10.1038/s41573-019-0047-y. 880 65. Hanzl A, Winter GE. Targeted protein degradation: current and future challenges. Curr Opin Chem 881 Biol. 2020;56:35-41. doi: 10.1016/j.cbpa.2019.11.012. 882 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 34 66. Baek K, Krist DT, Prabu JR, Hill S, Klugel M, Neumaier LM, et al. NEDD8 nucleates a multivalent 883 cullin-RING-UBE2D ubiquitin ligation assembly. Nature. 2020;578(7795):461-6. doi: 10.1038/s41586-020-884 2000-y. 885 67. Zhao Y, Chapman DA, Jones IM. Improving baculovirus recombination. Nucleic acids research. 886 2003;31(2):E6-. Epub 2003/01/16. 887 68. Wilkins MR, Gasteiger E, Bairoch A, Sanchez JC, Williams KL, Appel RD, et al. Protein identification 888 and analysis tools in the ExPASy server. Methods Mol Biol. 1999;112:531-52. Epub 1999/02/23. doi: 10.1385/1-889 59259-584-7:531. 890 69. Kabsch W. Xds. Acta crystallographica Section D, Biological crystallography. 2010;66(Pt 2):125-32. 891 Epub 2010/02/04. doi: 10.1107/S0907444909047337. 892 70. Vagin A, Teplyakov A. Molecular replacement with MOLREP. Acta crystallographica Section D, 893 Biological crystallography. 2010;66(Pt 1):22-5. Epub 2010/01/09. doi: 10.1107/S0907444909042589. 894 71. Emsley P, Cowtan K. Coot: model-building tools for molecular graphics. Acta crystallographica Section 895 D, Biological crystallography. 2004;60(Pt 12 Pt 1):2126-32. Epub 2004/12/02. doi: 896 10.1107/S0907444904019158. 897 72. Liebschner D, Afonine PV, Baker ML, Bunkoczi G, Chen VB, Croll TI, et al. Macromolecular structure 898 determination using X-rays, neutrons and electrons: recent developments in Phenix. Acta Crystallogr D Struct 899 Biol. 2019;75(Pt 10):861-77. Epub 2019/10/08. doi: 10.1107/S2059798319011471. 900 73. Sparta KM, Krug M, Heinemann U, Mueller U, Weiss MS. XDSAPP2.0. Journal of Applied 901 Crystallography. 2016;49(3):1085-92. doi: doi:10.1107/S1600576716004416. 902 74. McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC, Read RJ. Phaser 903 crystallographic software. Journal of Applied Crystallography. 2007;40(4):658-74. doi: 904 doi:10.1107/S0021889807021206. 905 75. Kuroki R, Weaver LH, Matthews BW. Structural basis of the conversion of T4 lysozyme into a 906 transglycosidase by reengineering the active site. Proceedings of the National Academy of Sciences of the 907 United States of America. 1999;96(16):8949-54. Epub 1999/08/04. doi: 10.1073/pnas.96.16.8949. 908 76. Morellet N, Bouaziz S, Petitjean P, Roques BP. NMR structure of the HIV-1 regulatory protein VPR. 909 Journal of molecular biology. 2003;327(1):215-27. Epub 2003/03/05. 910 77. Carragher B, Kisseberth N, Kriegman D, Milligan RA, Potter CS, Pulokas J, et al. Leginon: an 911 automated system for acquisition of images from vitreous ice specimens. J Struct Biol. 2000;132(1):33-45. Epub 912 2000/12/21. doi: 10.1006/jsbi.2000.4314. 913 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 35 78. Suloway C, Pulokas J, Fellmann D, Cheng A, Guerra F, Quispe J, et al. Automated molecular 914 microscopy: the new Leginon system. J Struct Biol. 2005;151(1):41-60. Epub 2005/05/14. doi: 915 10.1016/j.jsb.2005.03.010. 916 79. Moriya T, Saur M, Stabrin M, Merino F, Voicu H, Huang Z, et al. High-resolution Single Particle 917 Analysis from Electron Cryo-microscopy Images Using SPHIRE. J Vis Exp. 2017;(123). Epub 2017/06/02. doi: 918 10.3791/55448. 919 80. Grant T, Rohou A, Grigorieff N. cisTEM, user-friendly software for single-particle image processing. 920 Elife. 2018;7. Epub 2018/03/08. doi: 10.7554/eLife.35383. 921 81. Zivanov J, Nakane T, Forsberg BO, Kimanius D, Hagen WJ, Lindahl E, et al. New tools for automated 922 high-resolution cryo-EM structure determination in RELION-3. Elife. 2018;7. Epub 2018/11/10. doi: 923 10.7554/eLife.42166. 924 82. Zheng SQ, Palovcak E, Armache JP, Verba KA, Cheng Y, Agard DA. MotionCor2: anisotropic 925 correction of beam-induced motion for improved cryo-electron microscopy. Nat Methods. 2017;14(4):331-2. 926 Epub 2017/03/03. doi: 10.1038/nmeth.4193. 927 83. Mindell JA, Grigorieff N. Accurate determination of local defocus and specimen tilt in electron 928 microscopy. J Struct Biol. 2003;142(3):334-47. Epub 2003/06/05. doi: 10.1016/s1047-8477(03)00069-8. 929 84. Punjani A, Rubinstein JL, Fleet DJ, Brubaker MA. cryoSPARC: algorithms for rapid unsupervised 930 cryo-EM structure determination. Nat Methods. 2017;14(3):290-6. Epub 2017/02/07. doi: 10.1038/nmeth.4169. 931 85. Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC, et al. UCSF Chimera--a 932 visualization system for exploratory research and analysis. J Comput Chem. 2004;25(13):1605-12. Epub 933 2004/07/21. doi: 10.1002/jcc.20084. 934 86. Hayward S, Lee RA. Improvements in the analysis of domain motions in proteins from conformational 935 change: DynDom version 1.50. J Mol Graph Model. 2002;21(3):181-3. Epub 2002/12/05. doi: 10.1016/s1093-936 3263(02)00140-7. 937 87. Krissinel E, Henrick K. Inference of macromolecular assemblies from crystalline state. Journal of 938 molecular biology. 2007;372(3):774-97. Epub 2007/08/08. doi: 10.1016/j.jmb.2007.05.022. 939 88. Madeira F, Park YM, Lee J, Buso N, Gur T, Madhusoodanan N, et al. The EMBL-EBI search and 940 sequence analysis tools APIs in 2019. Nucleic acids research. 2019;47(W1):W636-W41. Epub 2019/04/13. doi: 941 10.1093/nar/gkz268. 942 89. Nicholas KB, Nicholas Jr., H. B., Deerfield II., D. W. GeneDoc: Analysis and Visualization of Genetic 943 Variation. embnetnews. 1997;4(2):1-4. 944 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 36 90. Shevchenko A, Tomas H, Havlis J, Olsen JV, Mann M. In-gel digestion for mass spectrometric 945 characterization of proteins and proteomes. Nature protocols. 2006;1(6):2856-60. Epub 2007/04/05. doi: 946 10.1038/nprot.2006.468. 947 91. Rappsilber J, Ishihama Y, Mann M. Stop and go extraction tips for matrix-assisted laser 948 desorption/ionization, nanoelectrospray, and LC/MS sample pretreatment in proteomics. Anal Chem. 949 2003;75(3):663-70. Epub 2003/02/15. doi: 10.1021/ac026117i. 950 92. Kolbowski L, Mendes ML, Rappsilber J. Optimizing the Parameters Governing the Fragmentation of 951 Cross-Linked Peptides in a Tribrid Mass Spectrometer. Anal Chem. 2017;89(10):5311-8. Epub 2017/04/14. doi: 952 10.1021/acs.analchem.6b04935. 953 93. Mendes ML, Fischer L, Chen ZA, Barbon M, O'Reilly FJ, Giese SH, et al. An integrated workflow for 954 crosslinking mass spectrometry. Mol Syst Biol. 2019;15(9):e8994. Epub 2019/09/27. doi: 955 10.15252/msb.20198994. 956 94. Fischer L, Rappsilber J. Quirks of Error Estimation in Cross-Linking/Mass Spectrometry. Anal Chem. 957 2017;89(7):3829-33. Epub 2017/03/08. doi: 10.1021/acs.analchem.6b03745. 958 95. Yang J, Zhang Y. Protein Structure and Function Prediction Using I-TASSER. Curr Protoc 959 Bioinformatics. 2015;52:5 8 1-5 8 15. Epub 2015/12/19. doi: 10.1002/0471250953.bi0508s52. 960 96. Webb B, Sali A. Comparative Protein Structure Modeling Using MODELLER. Curr Protoc 961 Bioinformatics. 2016;54:5 6 1-5 6 37. Epub 2016/06/21. doi: 10.1002/cpbi.3. 962 97. van Zundert GC, Trellet M, Schaarschmidt J, Kurkcuoglu Z, David M, Verlato M, et al. The DisVis and 963 PowerFit Web Servers: Explorative and Integrative Modeling of Biomolecular Complexes. Journal of molecular 964 biology. 2017;429(3):399-407. Epub 2016/12/13. doi: 10.1016/j.jmb.2016.11.032. 965 98. Perez-Riverol Y, Csordas A, Bai J, Bernal-Llinares M, Hewapathirana S, Kundu DJ, et al. The PRIDE 966 database and related tools and resources in 2019: improving support for quantification data. Nucleic acids 967 research. 2019;47(D1):D442-D50. Epub 2018/11/06. doi: 10.1093/nar/gky1106. 968 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 37 Figures 969 970 Fig 1. Biochemical analysis of Vprmus-induced CRL4DCAF1 specificity redirection. 971 (A) GF analysis of in vitro reconstitution of protein complexes containing DDB1/DCAF1-CtD, Vprmus 972 and SAMHD1 constructs. A schematic of the SAMHD1 constructs is shown above the chromatograms. 973 SAM – sterile α-motif domain, HD – histidine-aspartate domain, T4L – T4 Lysozyme. (B) SDS-PAGE 974 analysis of fractions collected during GF runs in A, boxes are colour-coded with respect to the 975 chromatograms. Note that during preparation of the GF run containing SAMHD1-ΔCtD (green trace), 976 the GST-affinity tag, which forms dimers in solution, was not removed completely from DDB1. 977 Accordingly, the GF trace contains an additional dimeric GST-DDB1/DCAF1-CtD/Vprmus component 978 in fractions 4-5. (C-F) In vitro ubiquitylation reactions with purified protein components in the absence 979 (C) or presence (D-F) of Vprmus, with the indicated SAMHD1 constructs as substrate. Reactions were 980 stopped after the indicated times, separated on SDS-PAGE and visualised by staining. 981 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 38 982 Fig 2. Crystal structure of the DDB1/DCAF1-CtD/Vprmus complex. 983 (A) Overall structure of the complex in two views. DCAF1-CtD is shown as grey cartoon and semi-984 transparent surface. Vprmus is shown as a dark green cartoon with the co-ordinated zinc ion shown as 985 grey sphere. T4L and DDB1 have been omitted for clarity. (B) Superposition of apo-DCAF1-CtD (light 986 blue cartoon) with Vprmus-bound DCAF1-CtD (grey/green cartoon). Only DCAF1-CtD regions with 987 significant structural differences between apo- and Vprmus-bound forms are shown. Disordered loops are 988 indicated as dashed lines. (C) Comparison of the binary Vprmus/DCAF1-CtD and ternary Vpxsm/DCAF1-989 CtD/SAMHD1-CtD complexes. For DCAF1-CtD, only the N-terminal “acidic loop” region is shown. 990 Vprmus, DCAF1-CtD and bound zinc are coloured as in A; Vpxsm is represented as orange cartoon and 991 SAMHD1-CtD as pink cartoon. Selected Vpr/Vpx/DCAF1-CtD side chains are shown as sticks, and 992 electrostatic interactions between these side chains are indicated as dotted lines. (D) In vitro 993 reconstitution of protein complexes containing DDB1/DCAF1-CtD/Vprmus or the Vprmus R15E/R75E 994 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 39 mutant, and SAMHD1, analysed by analytical GF. SDS-PAGE analysis of corresponding GF fractions 995 is shown next to the chromatogram. 996 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 40 997 Fig 3. Mechanism of SAMHD1-CtD recruitment by Vprmus. 998 (A) Two views of the cryo-EM reconstruction of the CRL4-NEDD8DCAF1-CtD/Vprmus/SAMHD1 core. 999 The crystal structure of the DDB1/DCAF1-CtD/Vprmus complex was fitted as a rigid body into the cryo-1000 EM density and is shown in the same colours as in Fig 2A. The DDB1 BPB model and density was 1001 removed for clarity. The red arrows mark additional density on the upper surface of the Vprmus helix 1002 bundle. (B) Schematic representation of Sulfo-SDA cross-links (grey lines) between CRL4DCAF1/Vprmus 1003 and SAMHD1, identified by CLMS. Proteins are colour-coded as in A, CUL4 is coloured orange, 1004 SAMHD1 black/white. SAMHD1-CtD is highlighted in red, and cross-links to SAMHD1-CtD are 1005 highlighted in violet. (C) The accessible interaction space of SAMHD1-CtD, calculated by the DisVis 1006 server [97], consistent with at least 14 of 26 observed cross-links, is visualised as grey mesh. DCAF1-1007 CtD and Vprmus are oriented and coloured as in A. (D) Detailed view of the SAMHD1-CtD electron 1008 density. The model is in the same orientation as in A, left panel. Selected Vprmus residues W29 and A66, 1009 which are in close contact to the additional density, are shown as red space-fill representation. (E) In 1010 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 41 vitro reconstitution of protein complexes containing DDB1/DCAF1-CtD, Vprmus or the Vprmus 1011 W29A/A66W mutant, and SAMHD1, assessed by analytical GF. SDS-PAGE analysis of corresponding 1012 GF fractions is shown below the chromatogram. 1013 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 42 1014 Fig 4. Variability of neo-substrate recognition in Vpx/Vpr proteins. 1015 Comparison of neo-substrate recognition modes of Vprmus (A), Vpxsm (B), Vpxmnd2 (C) and VprHIV-1 (D) 1016 proteins. DCAF1-CtD is shown as grey cartoon and semi-transparent surface, Vprmus – green, Vpxsm – 1017 orange, Vpxmnd2 – blue and VprHIV-1– light brown are shown as cartoon. Models of the recruited 1018 ubiquitylation substrates are shown as strongly filtered, semi-transparent calculated electron density 1019 maps with the following colouring scheme: SAMHD1-CtD bound to Vprmus – yellow, SAMHD1-CtD 1020 (bound to Vpxsm, PDB 4cc9) [50] – mint green, SAMHD1-NtD (Vpxmnd2, PDB 5aja) [51] – magenta, 1021 UNG2 (VprHIV-1, PDB 5jk7) [54] – light violet. 1022 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 43 1023 Fig 5. Cryo-EM analysis of CRL4-NEDD8DCAF1-CtD conformational states. 1024 (A) Two views of an overlay of CRL4-NEDD8DCAF1-CtD/Vprmus/SAMHD1 cryo-EM reconstructions 1025 (conformational state-1 – light green, state-2 – salmon, state-3 – purple). The portions of the densities 1026 corresponding to DDB1 BPA/BPC, DCAF1-CtD and Vprmus have been superimposed. (B) Two views 1027 of a superposition of DDB1/DCAF1-CtD/Vprmus and CUL4/ROC1 (PDB 2hye) [15] molecular models, 1028 which have been fitted as rigid bodies to the corresponding cryo-EM densities; the models are oriented 1029 as in A. DDB1/DCAF1-CtD/Vprmus is shown as in Fig 2A, CUL4 is shown as cartoon, coloured as in A 1030 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 44 and ROC1 is shown as cyan cartoon. Cryo-EM density corresponding to SAMHD1-CtD is shown in 1031 yellow, to illustrate the SAMHD1-CtD binding site in the context of the whole CRL4 assembly. 1032 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931 45 1033 Fig 6. Schematic illustration of structural plasticity in Vprmus-modified CRL4DCAF1-CtD, and 1034 implications for ubiquitin transfer. 1035 (A) Rotation of the CRL4 stalk increases the space accessible to catalytic elements at the distal tip of 1036 the stalk, forming a ubiquitylation zone around the core. (B) Modification of CUL4-WHB with NEDD8 1037 leads to increased mobility of these distal stalk elements (CUL4-WHB, ROC1 RING domain) [57], 1038 further extending the ubiquitylation zone and activating the formation of a catalytic assembly for 1039 ubiquitin transfer (see also D) [66]. (C) Flexible tethering of SAMHD1 to the core by Vprmus places the 1040 bulk of SAMHD1 in the ubiquitylation zone and optimises surface accessibility. (D) Dynamic processes 1041 A-C together create numerous possibilities for assembly of the catalytic machinery (NEDD8-CUL4-1042 WHB, ROC1, ubiquitin-(ubi-)charged E2) on surface-exposed SAMHD1 lysine side chains. Here, three 1043 of these possibilities are exemplified schematically. In this way, ubiquitin coverage on SAMHD1 is 1044 maximised. 1045 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted January 1, 2021. ; https://doi.org/10.1101/2020.12.31.424931doi: bioRxiv preprint https://doi.org/10.1101/2020.12.31.424931