HIV-1 escape from a peptidic anchor inhibitor by envelope ...

1

HIV-1 escape from a peptidic anchor inhibitor by 1 envelope glycoprotein spike stabilization 2

3 Dirk Eggink1, Steven W. de Taeye1,#, Ilja Bontjer1,#, Per Johan Klasse2, Johannes P.M. 4

Langedijk3,#, Ben Berkhout1, and Rogier W. Sanders1, 2* 5 6 7

1 Laboratory of Experimental Virology, Department of Medical Microbiology, 8 Center for Infection and Immunity Amsterdam (CINIMA), Academic Medical Center 9 of the University of Amsterdam, the Netherlands 10 2 Department of Microbiology and Immunology, Weill Medical College of Cornell 11 University, New York, United States of America 12 3 Pepscan Therapeutics BV, Lelystad, the Netherlands 13 # Current address: Crucell Holland BV, Leiden, the Netherlands 14 15 #These authors contributed equally 16 17 18 Running title: VIR165-dependent HIV-1 19 20 Keywords: HIV-1, resistance, entry inhibitors, envelope glycoprotein 21 22 23 Address correspondence to: *Rogier Sanders, Laboratory of Experimental 24 Virology, Department of Medical Microbiology, Center for Infection and Immunity 25 Amsterdam (CINIMA), Academic Medical Center of the University of Amsterdam, 26 Meibergdreef 15, 1105 AZ Amsterdam, the Netherlands. Phone: +31-20-5667859 27 Fax: +31-20-6916531 E-mail: [email protected] 28

29

JVI Accepted Manuscript Posted Online 21 September 2016J. Virol. doi:10.1128/JVI.01616-16Copyright © 2016, American Society for Microbiology. All Rights Reserved.

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

2

ABSTRACT (max. 250 words) 30

The trimeric HIV-1 envelope glycoprotein spike (Env) mediates viral entry 31

into cells using a spring-loaded mechanism that allows for the controlled insertion 32

of the Env fusion peptide into the target membrane, followed by membrane fusion. 33

Env is the focus of vaccine research aimed at inducing protective immunity by 34

antibodies as well as efforts to develop drugs that inhibit the viral entry process. 35

The molecular factors contributing to Env stability and decay need to be better 36

understood in order to optimally design vaccines and therapeutics. We generated 37

escape viruses against VIR165, a peptidic inhibitor that binds the fusion peptide of 38

the gp41 subunit and prevents its insertion into the target membrane. 39

Interestingly, a number of escape viruses acquired substitutions in the C1 domain 40

of the gp120 subunit (A60E, E64K and H66R) that rendered these viruses 41

dependent on the inhibitor. These viruses could only infect target cells when 42

VIR165 was present after CD4 binding. Furthermore, the VIR165-dependent 43

viruses were resistant to soluble CD4-induced Env destabilization and decay. 44

These data suggest that VIR165-dependent Env proteins are kinetically trapped in 45

the unliganded state and require the drug to negotiate CD4-induced 46

conformational changes. These studies provide mechanistic insight into the action 47

of the gp41 fusion peptide and its inhibitors, and provide new ways to stabilize 48

Env trimer vaccines. 49

50

IMPORTANCE (max. 150 words) 51

Because of the rapid development of HIV-1 drug resistance new drug 52

targets need to be continuously explored. The fusion peptide of the envelope 53

glycoprotein can be targeted by anchor inhibitors. Here we describe virus escape 54

from the anchor inhibitor VIR165. Interestingly, some escape viruses became 55

dependent on the inhibitor for cell entry. We show that the identified escape 56

mutations stabilize the ground state of the envelope glycoprotein and should thus 57

be useful in the design of stabilized envelope-based HIV vaccines. 58

59

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

3

INTRODUCTION 60

With over 35 million people currently being infected, HIV-1 remains a major health 61

problem. Although progress has been made in the development of antiviral drugs, 62

the potential of the virus to acquire resistance remains an issue. Neither a cure nor 63

an effective vaccine is available. HIV-1 enters target cells by using its envelope 64

glycoprotein (Env) spikes on the virus surface. Env is the target for drugs that 65

inhibit the viral entry process and for broadly neutralizing antibodies (bNAbs) that 66

researchers aim to induce with Env-based vaccines. 67

Env is a trimeric complex consisting of three gp41 transmembrane subunits 68

and three non-covalently attached gp120 subunits. The entry process is initiated 69

by binding of gp120 to the CD4 receptor. This interaction induces conformational 70

changes in Env, revealing the binding site for a chemokine co-receptor, generally 71

CXCR4 or CCR5. Additional conformational changes in gp41 involve two heptad 72

repeat regions (HR1 and HR2) in gp41 and the fusion peptide (FP) (1). The FP 73

might be partially exposed in the pre-fusion, native state of Env. In a complex of 74

the Env trimer with the FP-targeting bNAb VRC34 residues 512-519 are in contact 75

with the bNAb, suggesting that they might be solvent exposed. In contrast, residues 76

520-527 are buried within the protein (2). During the conformational changes that 77

lead to membrane fusion a trimeric coiled coil consisting of the three HR1 domains 78

is extended and the FP is inserted into the target cell membrane. Next, the three 79

HR2 domains associate with the HR1 coiled coil, resulting in the formation of a 80

stable six-helix bundle which juxtaposes the viral and cellular membranes and 81

provides the free energy for membrane fusion (3-5). The structure of an Env 82

trimer was recently solved, providing the first detailed images of this intricate 83

molecular machine (6-10). 84

Despite the weak nature of the gp120-gp41 interactions, the presence of 85

gp120 is crucial to maintain gp41 in its metastable, spring-loaded state before 86

encountering a susceptible cell. gp120 is likely to orchestrate the sequential 87

conformational changes in gp41 that culminate in membrane fusion, and 88

premature release or shedding of gp120 renders gp41 inactive (11-14). A number 89

of residues in the C1 and C5 domains of gp120 were shown to be important for the 90

interaction with gp41 (12, 13, 15-18). These domains are part of a 7-stranded β-91

sandwich with three loops directed towards the target cell, thereby organizing the 92

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

4

inner domain of gp120 in three structurally mobile layers (19-21). Layer 1 and 93

layer 2 contribute to the non-covalent interactions between gp120 and gp41 and 94

stabilize the binding of CD4, thereby linking receptor binding to the fusion 95

machinery in gp41 (22). Layer 3 helps to expose the initial site of contact with CD4 96

and provides the interaction between the gp120 inner and outer domains (23). 97

Env is the target of entry inhibitors and two such inhibitors, maraviroc, 98

which prevents co-receptor binding, and enfuvirtide/fuzeon, which prevents six-99

helix bundle formation, have been approved for clinical use. The VIRIP peptide and 100

its derivatives (e.g. VIR165, VIR353, VIR576, Fig 1A, (24)) form another class of 101

entry inhibitors termed anchor inhibitors that bind to the FP (the “anchor”) and 102

prevent its insertion into the target membrane (24). Although mutations in FP can 103

provide resistance to VIRIP-derivative VIR165 (unpublished) and limit the clinical 104

benefit of the related inhibitor VIR576, escape studies with another anchor 105

inhibitor, VIR353, selected for resistance mutations outside the binding site in the 106

C4 and C5 domain of gp120 and in HR1 and the loop domain of gp41 (25, 26). 107

Here we describe virus escape studies with the VIRIP-derivative VIR165. 108

We selected resistance mutations and describe the underlying molecular 109

mechanisms. Interestingly, we identified a number of escape mutations in the C1 110

domain of the gp120 subunit of Env (A60E, E64K, and H66R) that rendered the 111

virus dependent on the drug for CD4-induced conformational changes and viral 112

entry. The VIR165-dependent viruses were more thermostable and remarkably 113

resistant to CD4-induced virus inactivation. These results have relevance for 114

understanding HIV entry and the roles of CD4-binding and FP action, as well as for 115

entry-targeting therapeutics and Env-based vaccine design. 116

117

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

5

MATERIALS AND METHODS 118

119

Reagents 120

Reagents were obtained as gifts, or purchased, from the following sources: William 121

Olson (Progenics Pharmaceuticals) provided soluble CD4 (sCD4) and CD4-IgG2. 122

AMD3100 was a generous gift from Dr D. Schols (Rega Institute, Leuven 123

University). Peptides VIR165 (LEAIPCSIPPCFAFNKPFVF), and T20 124

(YTSLIHSLIEESQNQQEKNEQELLELDKWASLWNWF) were synthesized as 125

described previously (27, 28). 126

127

Selection of VIRIP-resistant HIV-1 variants 128

For the selection of VIR165-resistant viruses, 1 x 106 SupT1 cells were infected 129

with 1.5 ng CA-p24 wild-type (WT) HIV-1LAI molecular clone. Six independent 130

cultures were started to initiate evolution with a concentration of 3.8 μg/ml 131

VIR165, which corresponds to the IC50 of WT HIV-1LAI. Cultures were split twice 132

weekly and when HIV-induced cytopathic effects and/or increased CA-p24 133

production were apparent, virus replication was maintained by passage of cell-free 134

culture supernatant onto uninfected SupT1 cells with a 2-fold increased VIR165 135

concentration. Initially we started with passaging 100 μl cell free supernatant onto 136

fresh cells. We gradually decreased the volume of supernatant in subsequent 137

passages, from 100 μl in the second passage to a minimum of 10 μl. Cells and 138

supernatant samples were taken at regular time points and stored at -80°C. Cell 139

culturing, transfections and CA-p24 determination was performed as previously 140

reported (29). After 2½ months of culture, DNA was extracted from infected cells 141

using the QIAamp DNA mini kit (Qiagen, Valencia, CA), and the complete proviral 142

env gene was PCR amplified using primers 1 (5'-143

ATAAGCTTAGCAGAAGACAGTGGCAATG-3') and 2 (5'-GCAAAATCCTTTCCAAGCCC-144

3') and sequenced. Amino acid numbering is based on the HXB2 isolate. 145

146

Construction of HIV-1LAI molecular clones 147

The full-length molecular clone of HIV-1LAI (pLAI) (30) was used to produce WT 148

and mutant viruses. The plasmid pRS1 was used to introduce mutations as 149

described previously (27, 31) and the entire env gene was verified by DNA 150

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

6

sequencing. Mutant env genes in pRS1 were cloned back into pLAI as SalI-BamHI 151

fragments. Each virus variant was transiently transfected in C33A cells by calcium 152

phosphate precipitation as previously described (32) or by transfection in 153

HEK293T cells using lipofectamine 2000 (Invitrogen) according to the 154

manufacturers instructions. The virus containing supernatant was harvested 3 155

days post-transfection and stored at -80°C and the virus concentration was 156

quantitated by capsid CA-p24 ELISA as described previously (29). 157

158

Infectivity and IC50 determination 159

The TZM-bl reporter cell line (33, 34) stably expresses high levels of CD4 and HIV-160

1 co-receptors CCR5 and CXCR4 and contains the luciferase and β-galactosidase 161

genes under the control of the HIV-1 long-terminal-repeat promoter. The TZM-bl 162

cell line was obtained through the NIH AIDS Research and Reference Reagent 163

Program, Division of AIDS, NIAID, NIH (John C. Kappes, Xiaoyun Wu, and Tranzyme 164

Inc. [Durham, NC]). One day prior to infection, 17 x 103 TZM-bl cells per well were 165

plated in a 96-well plate in Dulbecco's modified Eagle's medium containing 10% 166

fetal bovine serum and penicillin-streptomycin (both at 100 units/ml) and 167

incubated at 37°C with 5% CO2. Virus (5.0 ng CA-p24) was pre-incubated for 30 168

min at room temperature with serial dilutions of antisera or entry inhibitor (HIVIg, 169

sCD4, CD4-IgG2, AMD3100, T20 or VIR165). This mixture was added to the cells in 170

the presence of 400 nM saquinavir (Roche, Mannheim, Germany) to block 171

secondary rounds of infection and 40 µg/ml DEAE in a total volume of 200 µl. 172

DEAE was added to enhance HIV-1 infection of TZM-bl reporter cells. This did not 173

affect the sensitivity to the inhibitors tested (data not shown) (35-37). Two days 174

post-infection, the medium was removed and cells were washed once with 175

phosphate-buffered saline (50 mM sodium phosphate, pH 7.0, 150 mM NaCl; PBS) 176

and lysed in reporter lysis buffer (Promega, Madison, WI). Luciferase activity was 177

measured using a luciferase assay kit (Promega, Madison, WI) and a Glomax 178

luminometer according to the manufacturer's instructions (Turner BioSystems, 179

Sunnyvale, CA). All infections were performed in duplicate in at least two 180

independent experiments. Uninfected cells were used to correct for background 181

luciferase activity. The infectivity of each mutant without inhibitor was set at 182

100%. Nonlinear regression curves were determined using sigmoidal dose-183

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

7

response curves (variable slope, bottom constraint = 0) and 50% inhibitory 184

concentrations (IC50) were calculated using Prism software version 5.0. When full 185

inhibition was not yet reached at the highest concentration, sigmoidal curves were 186

modeled using extrapolation. Fold differences compared to WT were calculated 187

and were considered resistant when the calculated IC50 was over three fold higher 188

than that of WT (35, 36). 189

The relative infectivity of mutant viruses was calculated compared to HIV-1LAI. 190

Infections were performed in quadruple and luciferase activity without inhibitor 191

was measured corrected for background luciferase activity. Infectivity of HIV-1LAI 192

WT was set at 100% and the relative infectivity of the mutants was calculated. 193

194

Virus thermostability experiments 195

WT and mutant viruses were incubated for 1 h in the presence or absence of 0.5 196

μg/ml sCD4 at a temperature range of 37°C to 50°C using a thermocycler before 197

testing their residual infectivity on TZM-bl reporter cells as above. Because the 198

A60E, E64K and H66R mutants were not infectious in the absence of drug, VIR165 199

(10 or 30 μg/ml) was added after incubation in the presence or absence of sCD4 200

and present during infection of TZM-bl reporter cells. The midpoint of thermal 201

denaturation (Tm) was defined as the temperature at which 50% residual 202

infectivity was observed. 203

204

Virus decay experiments 205

WT and mutant viruses were incubated at 37°C in the presence or absence of 1.0 206

ug/ml sCD4 (final concentration during infection 0.5 μg/ml) for 0, 2, 4, 8, 16, 24, 32 207

and 96 hours before testing their infectivity on TZM-bl reporter cells as above. As 208

mutants A60E, E64K and H66R are not infectious in the absence of drug, VIR165 209

(10 or 30 μg/ml) was added during infection of TZM-bl reporter cells. Nonlinear 210

regression curves were determined and half-lifes (t½) were calculated using Prism 211

software version 5.0. 212

213

Statistical analysis. 214

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

8

All statistical comparisons were performed by the unpaired t test (two tailed) 215

using Prism software version 5.0 and indicated in figures with asterisks: * P < 0.05; 216

** P < 0.005. 217

218

RESULTS 219

220

VIR165 escape occurs in C1 of gp120 and HR1 of gp41 221

To investigate whether HIV-1LAI is able to escape from VIR165 (Fig. 1A), we 222

maintained six independent cultures of wild-type (WT) HIV-1LAI that were 223

passaged on SupT1 cells in the presence of increasing concentrations of VIR165. 224

The initial VIR165 concentration was 3.8 μg/ml (1.7 μM; corresponding to the IC50 225

against WT HIV-1LAI). Over the course of 2½ months (20 passages, Fig 1B), the 226

VIR165 concentration was gradually increased up to 250 μg/ml (110 μM), 64-fold 227

the IC50 for WT HIV-1LAI. The complete viral env gene of the proviral DNA was 228

subsequently sequenced. Each culture contained one or two amino acid 229

substitutions in Env (Table 1). We mapped these mutations on the Env trimer 230

structure (Fig. 1C, D) (6-8, 38). Surprisingly, most acquired amino acid 231

substitutions mapped to the C1 domain of gp120: V42I, A58V, A60E, E64K and 232

H66R. Except for V42I, these residues are located in layer 1 of the inner domain of 233

gp120 (19, 20), within the disulfide bonded loop between C54 and C74 (39, 40). 234

It’s worth noting that residue 42 is in close contact with residue 523 via extensive 235

side chain interactions (9). Residue 523 has been shown to be one of the contact 236

residues of VIR165 (24). We also observed substitutions in HR1 of gp41 (A558T 237

and Q577R). We note that Q577R was previously linked to resistance against the 238

3rd generation HR2-based fusion inhibitor T2635 (41). All positions identified in 239

these escape cultures are highly conserved: V42: 99.1%; I42 0.17%, A58: 99.6%; 240

V58: 0.02%, A60: 87.5%; E60: not found, E64: 99.8%; K64: 0.05% (found in two 241

sequences of which one contains multiple frameshifts and stop-codons), H66: 242

99.9%; R66: 0.048% (in two sequences from different subtypes), A558: 99.9%; 243

T558: 0.048%, Q577: 99.3%; R577: 0.36% (Los Alamos database (42)). We 244

observed different substitutions compared to those found in escape studies with 245

VIR353, in which resistance mutations in the C4 and C5 domain of gp120 and in 246

HR1 and the loop domain of gp41 were selected (25, 26). No escape mutations 247

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

9

were observed in FP, which is consistent with the lack of such mutations in escape 248

studies using the VIRIP-derivative VIR353 (24, 26), probably because FP-249

mutations pose a large penalty to fitness (43). 250

251

Substitutions in gp41 HR1 confer VIR165-resistance 252

To confirm the importance of the selected substitutions in providing 253

resistance against VIR165, we constructed molecular clones of HIVLAI with the HR1 254

substitution A558T or Q577R. The T20 resistant mutant V549A was included as a 255

control (28). Single cycle infection experiments were performed using the TZM-bl 256

reporter cell-line in the presence of serial dilutions of VIR165 (Fig. 2A, Table 2). 257

HR1 mutants A558T and Q577R showed 10-fold and >10-fold VIR165-resistance, 258

respectively, compared to WT. The control virus V549A did not show a significant 259

change in sensitivity to VIR165. In the absence of drug, the A558T and Q577R 260

substitutions caused a drop in viral infectivity in the single cycle infection 261

experiments in the TZM-bl reporter cell-line (35% and 20% of WT, respectively), 262

indicating that resistance comes with a significant loss of Env protein function and 263

viral fitness (Fig. 2D), consistent with what was observed upon viral escape from 264

the fusion inhibitor T2635 (41). 265

266

Substitutions in gp120 C1 confer VIR165-resistance or -dependence 267

We next evaluated the contribution of the selected C1 mutations V42I, 268

A58V, A60E, E64K, and H66R to VIR165-resistance. The D62N mutant in C1 was 269

generated to serve as a control. The H66N substitution, which was previously 270

described in a cold-resistant HIV-1 variant and shown to affect the interaction with 271

CD4, was also included (44, 45). Substitutions V42I and A58V caused 4-fold and 8-272

fold VIR165-resistance (Fig. 2B, Table 2) with little impact on viral fitness (Fig. 273

2D). More intriguing was the effect of the A60E, E64K and H66R substitutions. 274

Although these substitutions are located very close to the resistance mutation 275

A58V, they yielded a strikingly different phenotype. In the absence of VIR165, 276

these variants were barely infectious or not infectious at all (Fig. 2D). However, 277

they became infectious with increasing VIR165 concentrations (Fig. 2C). Thus, 278

these mutants display a VIR165-dependent phenotype. Drug dependence is a 279

relatively rare phenomenon in HIV-1 drug resistance. T20- or retrocyclin (RC101)-280

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

10

dependent viruses have been described that involve a combination of substitutions 281

in the HR1 binding site of the inhibitors and in HR2 (46, 47). The T20-dependent 282

virus V549A/N637K did not show cross-resistance or -dependence to VIR165 (Fig. 283

2A, Table 2). The control D62N and H66N substitutions did not confer VIR165-284

resistance or -dependence, and did not affect infectivity (Fig. 2B, C, Table 2). 285

To test whether VIR165-dependent variants could be inhibited by a high 286

dose of VIR165, we repeated the dose-inhibition experiments with VIR165 287

concentrations up to 300 μg/ml (Fig. 2E). Clear drug-induction was observed 288

reaching maximum infectivity at approximately 10 (A60E, E64K) and 30 (H66R) 289

μg/ml VIR165. However, all three VIR165-dependent variants were inhibited at 290

higher concentrations. We determined IC50 values of 60 μg/ml for A60E and E64K 291

and 100 μg/ml for H66R, translating in approximately 37-fold and 58-fold 292

resistance compared to WT. 293

Using the optimal VIR165-concentration for stimulation of each VIR165-294

dependent virus variant we reassessed the maximal infectivity of these viruses. We 295

already showed a nearly complete loss of infectivity of A60E, E64K and H66R in 296

the absence of drug (Fig. 2C, D) and we now compared WT virus infectivity without 297

VIR165 with the maximum infectivity of the dependent viruses in the presence of 298

VIR165 (10 μg/ml for A60E and E64K, and 30 μg/ml for H66R) (Fig. 2F). Infectivity 299

of WT virus was fully inhibited in the presence of 10 μg/ml VIR165. Mutants E64K 300

and H66R exhibited only ~10% infectivity in the presence of VIR165, while the 301

A60E mutant could be activated by VIR165 to achieve ~50% infectivity compared 302

to WT virus without VIR165. Thus, VIR165-dependence comes at a considerable 303

loss of viral fitness, in particular for the E64K and H66R variants. 304

305

Modeling of VIR165 occupancy explains the bell-shaped dose-response curves 306

The enhancement and inhibition of VIR165-dependent HIV-1 infection was 307

modeled mathematically in order to assess whether enhancing or inhibitory effects 308

could be explained by differential FP occupancy levels by VIR165 (1, 2 or 3 VIR165 309

peptides per Env trimer). Different mathematical VIR165 occupancy models were 310

fitted to the experimentally obtained bell-shaped stimulation-inhibition data (Fig. 311

3). The model functions were derived from the binomial theorem, as described 312

(48). They postulate incremental effects of trimer ligation by VIR165. No 313

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

11

assumptions were made about total numbers of trimers or minimal thresholds of 314

functional trimers required for infectivity. Discrepancies between the modeled 315

curves and the experimental data can be partly attributed to this 316

oversimplification (48, 49). 317

The occupancy of VIR165 on a protomer was expressed as 318

p=(C/Kd)/(1+C/Kd), where C is the VIR165 concentration and Kd its dissociation 319

constant, on the simplifying assumption that neither positive nor negative 320

cooperativity between the three protomers of a trimer upon VIR165 binding 321

occurs. For the WT virus an inhibition model was applied that describes the 322

infectivity, I, as being proportional to the fraction of completely unoccupied 323

trimers: I=A(1-3p+3p2-p3), where A is a constant reflecting the maximum 324

amplitude, i.e., the maximum infectivity. An alternative model, which assumes that 325

inhibition occurs only when three VIR165 molecules are bound per trimer, was 326

also fitted to the data for comparison: I=Ap3. 327

For the VIR165-dependent viruses the experimental data were fitted to 328

three models. The first model posits that VIR165 occupancy of one or two 329

protomers per trimer allows infection, whereas occupancy of all three or none 330

does not: I = A(3p-3p2). The second model assumes the induction of infection by 331

exactly one VIR165 molecule per trimer: I = A(3p-6p2+3p3), and the third 332

postulates that infectivity requires exactly two VIR165 molecules per trimer: I = 333

A(3p2-3p3). The models were fitted to the inhibition data by non-linear regression. 334

The models postulating complete inhibition of WT Env function by a single 335

VIR165 peptide or by three peptides per trimer both fitted well to the 336

experimental data (R2=0.964 and R2=0.961, respectively). The model for induction 337

of infectivity of the VIR165-dependent mutants by one VIR165 peptide per trimer 338

fitted better to the experimental data than the other two models (R2=0.937-0.993 339

for enhancement by one VIR165 per trimer; R2=0.934-0.953 for enhancement by 340

one or two VIR165 per trimer; and R2=0.887-0.984 for enhancement by two 341

VIRIP165 per trimer). The models that fitted best to the experimental data are 342

shown in Fig. 3. 343

In conclusion, the mathematical modeling strongly supports a mechanistic 344

scenario in which partial occupancy of VIR165-dependent Env trimers by VIR165 345

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

12

enables infection, whereas occupancy of all three binding sites, which is reached at 346

high VIR165 concentrations, blocks infection. 347

348

VIR165-resistant and -dependent viruses are more sensitive to T20 349

Since all selected escape mutations are located outside the putative VIR165 350

binding site, it is possible that the kinetics of one or more steps of the entry 351

process are affected, thereby limiting the window of opportunity for the drug to 352

act (50-54). To probe whether and how the entry process was affected by the 353

VIR165-resistance and -dependence substitutions, we analyzed the sensitivity to 354

reagents that interfere with well-defined entry steps. As mutants A60E, E64K and 355

H66R are not infectious in the absence of drug, VIR165 (10 μg/ml) was always 356

present when these mutants were evaluated. First, we tested sensitivity to 357

polyclonal Ig from HIV-positive individuals (HIVIg). Neither VIR165-resistant nor 358

VIR165-dependent viruses showed altered sensitivity to HIVIg compared to WT 359

virus, indicating that these viruses did not have a generally altered sensitivity to 360

antibodies. Furthermore, the presence of VIR165 during infection of the dependent 361

variants did not alter the overall neutralization sensitivity (Table 3). 362

The windows of opportunities for VIR165 and HR2-based inhibitors 363

overlap and some mutations that confer resistance to HR2-based inhibitors confer 364

cross-resistance to VIR165 e.g. Q577R. To investigate whether the reverse was also 365

true, we tested all VIR165-resistant viruses (V42I, A58V, A558T and Q577R) and 366

two VIR165-dependent viruses (E64K and H66R) for T20-sensitivity. We did not 367

observe cross-resistance to T20. On the contrary, all VIR165-resistant and 368

dependent variants were more susceptible to the fusion inhibitor T20 (3-fold to 8-369

fold), except for Q577R (Table 3) which was not unexpected as this mutant was 370

also selected in fusion inhibitor resistance studies (41). The control T20-resistance 371

mutation V549A in HR1 (55), displayed 54-fold resistance to T20 and the D62N 372

change did not affect T20-sensitivity. Interestingly, similar observations were 373

made in a previous study, in which a combination of VIR353 resistance mutations 374

also resulted in enhanced sensitivity to T20 (26). These data suggest that the T20-375

sensitive step in entry (transition from pre-hairpin intermediate to six-helix 376

bundle), is probably slower in VIR165-resistant and -dependent virus variants, 377

resulting in an increased window of opportunity for HR1-targeting fusion 378

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

13

inhibitors. These data confirm that VIR165 and T20 inhibit a non-identical viral 379

entry step. 380

381

VIR165-dependent virus is resistant to inhibition by soluble CD4 but more 382

sensitive to inhibition by AMD3100 383

Soluble CD4 (sCD4) induces an activated Env state that is able to mediate 384

virus entry for a limited time after which Env rapidly decays into a functionally 385

inactive form (12, 56, 57). Alterations in Layer 1 in the C1 domain (e.g., H66A and 386

W69L) have been shown to affect this decay process and are thought to stabilize 387

the sCD4-activated Env intermediate, slowing its transition into a functionally 388

inactive Env form (44, 45, 57, 58). We wondered whether VIR165 resistance and 389

dependence mutations affected the interaction of Env with sCD4. We therefore 390

first performed neutralization studies with sCD4 as well as the CD4 mimetic CD4-391

IgG2. Both sCD4 and CD4-IgG2 can neutralize WT HIV-1 infection effectively (Table 392

3). Most VIR165-resistant viruses showed similar sensitivities to sCD4 and CD4-393

IgG2 compared to WT and the D62N variant. V42I was slightly less sensitive to 394

CD4-IgG2 (3-fold), but not to sCD4, while A558T was slightly more sensitive to 395

sCD4 but not to CD4-IgG2. Interestingly, the VIR165-dependent mutants E64K and 396

H66R were both less sensitive to neutralization by sCD4 (4-fold and 13-fold) and 397

CD4-IgG2 (3-fold and 6-fold). 398

The effect of mutations in either C1 or HR1 on the conformational changes 399

associated with CD4 binding could also have an effect on the affinity for the co-400

receptor. We therefore tested the sensitivity of WT virus and variants for 401

inhibition of the CXCR4 antagonist AMD3100 (Table 3). Whereas VIR165-402

resistance mutations in the C1 had no effect on sensitivity to AMD3100 inhibition, 403

resistance mutations within gp41 caused a small increase in AMD3100 sensitivity. 404

VIRIP dependency mutations E64K and H66R rendered the virus 3- to 5-fold more 405

sensitive to AMD3100, suggesting an enhanced dependence on CXCR4 for 406

proceeding with conformational changes towards fusion. 407

408

VIR165-dependent viruses are less prone to CD4-induced decay 409

We hypothesized that VIR165-dependent viruses might have an altered Env 410

stability and require VIR165 to trigger conformational changes following CD4 411

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

14

binding. Soluble CD4 induces an activated Env state that is short-lived in the 412

absence of a co-receptor and target membrane (57, 58). Substitutions at position 413

66 have been reported to decrease this CD4-induced decay (44, 45, 58). We 414

therefore tested the decay rates of WT, VIR165-resistant and VIR165-dependent 415

viruses by incubating them at physiological temperature (37˚C) for different time 416

intervals, followed by assessment of their remaining infectivity on TZM-bl cells. 417

For the VIR165 dependent viruses, VIR165 was added before the transfer to TZM-418

bl cells (Fig. 4A,B, Table 4). WT HIV-1LAI decayed with a half-life (t½) of 11.3 h, 419

which is in the range of values measured for JR-CSF (t½ = 18.9 h) and ADA (t½ = 9.0 420

h) (59). VIR165-resistant virus was less stable (A58V: t½ = 8.2 h), and so where 421

VIR165-dependent viruses (A60E: t½ = 9.4 h; E64K: t½ = 9.5 h; H66R: t½ = 8.2 h). 422

We next tested viral decay in the presence of sCD4. To avoid interference 423

with the subsequent infection, a sub-neutralizing concentration of sCD4 was used. 424

Again, VIR165 was added to the VIR165 dependent viruses before transferring the 425

viruses to TZM-bl cells. In the presence of sCD4, WT virus decayed twice as quickly 426

(1.9-fold, Δt½ = -5.2 h), indicative of CD4-induced deactivation (Fig. 4A,C, Table 4). 427

The VIR165-resistant virus (A58V), which was among the least stable viruses in 428

the absence of sCD4, also decayed more rapidly in the presence of sCD4: 1.5-fold 429

Δt½ = -2.6 h). In contrast, the VIR165-dependent variants, in particular the H66R 430

variant, remained relatively stable in the presence of sCD4 (E64K: 1.2-fold, Δt½ = -431

1.5 h; H66R: 1.1-fold. Δt½ = -0.3 h; and to a lesser extent A60E: 1.3-fold, Δt½ = -2.4 432

h). Because an H66N mutant was reported to be resistant to virus inactivation by 433

cold (44, 45), we repeated these experiments at 0˚C. None of the VIR165-434

dependent viruses were resistant to cold inactivation, but the VIR165-dependent 435

viruses were more resistant to sCD4-induced decay at 0˚C (data not shown), 436

similar to what we observed at 37˚C. We therefore conclude that VIR165-437

dependent Env variants are less prone to CD4-induced decay. 438

439

VIR165-dependent viruses have increased thermostability in the presence of 440

sCD4 441

We also tested the thermostability of WT, VIR165-resistant and -dependent 442

viruses by incubating them for 1 h at escalating temperatures, followed by testing 443

the residual infectivity on TZM-bl reporter cells (in the presence of VIR165 for the 444

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

15

dependent variants). WT virus had a midpoint of thermal denaturation (Tm) of 445

44.6˚C (Fig. 5A, B, Table 4), which is in the same range to what has been published 446

for HIV-1ADA (Tm = 40.5˚C), HIV-1JR-CSF (Tm = 44.6˚C) and HIV-1LAI (Tm = 42.0˚C) 447

(59, 60)(D. Leaman & M. Zwick, personal communication). The VIR165-resistant 448

virus had a similar Tm as WT virus (A58V: Tm = 44.8˚C), and so had the VIR165-449

dependent viruses (A60E: Tm = 44.7˚C; E64K: Tm = 44.6˚C H66R: Tm = 44.7˚C) (Fig. 450

5A, B, Table 4). Thus, we did not major differences in the themostability of the test 451

viruses. 452

To study whether the addition of sCD4 would affect the thermostability, we 453

also performed these experiments in the presence of (sub-neutralizing 454

concentrations of) sCD4. Addition of sCD4 destabilized WT HIV-1LAI and VIR165-455

resistant virus considerably (WT: ΔTm = -2.7˚C; A58V: ΔTm = -2.7˚C) (Fig 5A, C, 456

Table 4). In contrast, the VIR165-dependent viruses were less affected by the 457

addition of sCD4 (A60E: ΔTm = -1.4˚C; E64K: ΔTm = -0.9˚C; H66R: ΔTm = -0.5˚C) (Fig 458

5C, Table 4). We conclude that VIR165-dependent viruses, in particular the H66R 459

variant, are more thermostable and less prone to undergo CD4-induced 460

conformational changes that destabilize the Env spike. 461

462

DISCUSSION 463

VIRIP and VIRIP-derivatives are members of a class of peptide fusion inhibitors 464

termed anchor inhibitors and are thought to inhibit the HIV-1 entry process by 465

binding to the hydrophobic fusion peptide of gp41. A previous study used the 466

VIRIP analogue VIR353 to select resistance mutations in the HR1 and loop domain 467

of gp41 subunit and substitutions in the C4 and C5 domains of the gp120 subunit 468

(26). Here, we performed in vitro HIV-1 evolution experiments to select resistance 469

against the VIRIP analogue VIR165, one of the most potent VIRIP-derivatives. 470

Again, escape mutations were not selected in the putative FP binding site of the 471

peptide, but rather in the HR1 domain of gp41 and the C1 domain of gp120. 472

Moreover, we found that some C1 mutations conferred drug-dependence. 473

How can we mechanistically explain these escape mutations that occur 474

outside the putative binding site of VIR165? An alternative escape mechanism to 475

“decreased drug affinity” is “decreased drug opportunity” (46, 61, 62). In this 476

scenario, the drug-target affinity per se is unaltered, but the opportunity for the 477

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

16

drug to access the target site is restricted, for example because the target site is 478

exposed for a shorter period of time. This mechanism has been described for HIV-1 479

fusion inhibitors (61, 63). We think that such a window mechanism can also 480

account for the VIR165-resistance mutations. Part of the binding site for VIR165 481

(residues 513-518) might might be partially exposed in the pre-fusion state. It is 482

likely that CD4 and possibly co-receptor induced conformational changes allow 483

additional contact residues 520-523 and 525-531 to become available for VIR165 484

binding (2, 9, 24). We propose that the C1 and HR1 substitutions change the 485

kinetics of Env conformational changes, thereby decreasing the time that the FP is 486

accessible. The observation that a resistance mutation against the third-generation 487

HR2-based fusion inhibitor T2635 that did not occur in the drug binding site 488

(Q577R), provided cross-resistance to VIR165, is consistent with this hypothesis 489

(41). In fact, the Q577R substitution was selected in both T2635 (41) and VIR165-490

escape studies (Table 2). Since VIR165 and HR2-based inhibitors target 491

overlapping, although not identical, intermediate Env states, it makes sense that 492

some mutations can provide cross-resistance by changing the kinetics of this entry 493

step. We have not studied the C1 resistance mutations V42I and A58V in much 494

detail, but the A58V mutant was less stable and more prone to CD4-induced decay, 495

suggesting that it is more metastable than the WT Env spike. This might be 496

consistent with enhanced fusion kinetics and shortened FP exposure. Residue 42 is 497

in close contact with residue 523, which is one of the contact residues of Env with 498

VIR165. Alterations of the side-chain interactions between residues 42 and 523 499

might affect the exposure of the critical VIR165 contact residue 523. The selection 500

of VIR353 resistance mutations in the C4 and C5 domain of gp120 and in HR1 and 501

the loop domain of gp41, i.e. outside the inhibitor binding, might be explained by a 502

similar mechanism, especially as these substitutions also resulted in enhanced 503

sensitivity to fusion inhibitor T20 (25, 26). 504

The substitutions that confer VIR165-dependence might be explained by a 505

“decrease drug opportunity” mechanism. Drug-dependence can be classified as a 506

special form of resistance, since the virus is able to replicate in the presence of the 507

drug, but the resistance mutations cause a severe replication defect in the absence 508

of drug, which can only be overcome by addition of the drug. This phenomenon has 509

been described only for a limited number of cases, although drug stimulatory 510

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

17

effects and drug-dependence were reported for several HIV-1 inhibitors targeting 511

the Gag, Protease and Env proteins (46, 47, 64-66). For Env, enhanced fusion 512

kinetics can render HIV-1 resistant to fusion inhibitors (decreased opportunity), 513

but ‘hyperfusogenic’ Env variants can become dependent on the drug to avoid a 514

premature HR1-HR2 collapse in the absence of a target membrane (46, 47, 62). In 515

such cases the drug acts as a safety pin to prevent premature conformational 516

changes, but drug dissociation is eventually needed to allow such conformational 517

changes to occur in the presence of a target cell, leading to membrane fusion and 518

virus entry (46). VIR165-dependence is mechanistically distinct from T20-519

dependence. The VIR165 presence is not required during virus production, but 520

during an intermediate step in viral entry. We speculate that VIRIP-dependent 521

viruses have a hyperstable gp41-C1 interaction that cannot be sufficiently 522

weakened by CD4-binding and as a consequence CD4-induced conformational 523

changes in the gp120-gp41 complex are blocked. Since the FP is located at the 524

interface of gp120 and gp41, it is conceivable that binding of a peptide to the FP 525

affects the gp120-gp41 interactions. We propose that these dependent viruses 526

require VIR165 to act as a “wedge” between gp120 and gp41 to facilitate the CD4-527

induced conformational changes that allow the formation of the complete HR1(1), 528

followed by co-receptor binding and ultimately gp41-C1 dissociation and 529

membrane fusion. In this scenario, VIR165 might bind to the exposed contact 530

residues in the N-terminal part of the FP and ‘pull-out’ the rest of the fusion 531

peptide and thereby force the required conformational changes that allow 532

continuation of the entry process. In agreement with this, the VIR165-dependent 533

viruses are more stable and resistant to sCD4-induced decay. Recently, we used 534

the E64K and H66R substitutions to stabilize recombinant soluble (SOSIP) trimers 535

in the closed pre-fusion state. These substitutions facilitated the stabilization of 536

SOSIP trimers from 4 different isolates: BG505 (clade A), B41 and AMC008 (both 537

clade B), ZM197M (clade C). Hydrogen-Deuterium Exchange (HD-X) and other 538

studies demonstrated that these substitutions stabilize SOSIP trimers by blocking 539

CD4-induced conformation changes consistent with the virological observations 540

described here (67). Whereas T20-dependence involves enhanced fusion kinetics 541

that need to be tempered by T20 from the moment Env appears at the cell/virus 542

surface (61, 62, 68), VIR165-dependence involves reduced fusion kinetics 543

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

18

necessitating the presence of VIR165 as a catalyst for Env conformational changes 544

following CD4 binding. 545

If one assumes that this resistance phenotype involves a decreased window 546

of opportunity, then this might represent a paradox with delayed fusion kinetics. 547

How can reduced fusion kinetics decrease the window of opportunity for VIR165 548

to inhibit? The fact that VIR165-dependent virus is more sensitive to HR2-based 549

fusion inhibitors that target a similar stage during entry also appears paradoxical. 550

This can only be explained if the steps in which VIR165 and HR2-based inhibitors 551

act are separated in time. CD4 binding induces a short-lived activated state that 552

probably also exposes the FP. The FP is probably only fully exposed very shortly, 553

after CD4 binding and until FP is inserted in the target membrane. VIR165 554

exclusively targets this step, whereas HR2-based fusion inhibitors that may also 555

act during this stage will remain active when the FP is inserted into the target cell 556

membrane, up to the time that the six-helix bundle is formed. The windows of 557

opportunity for VIR165 and HR2-based inhibitors are therefore overlapping but 558

not identical. 559

Inhibition of VIR165-dependent viruses required a high dose of inhibitor. 560

The bell-shaped activation-inhibition curves could be mathematically explained by 561

a model in which VIR165-occupancy of one or two FPs per Env trimer could allow 562

for entry, while occupation of all three FPs would result in inhibition. It has indeed 563

been proposed that not all three FPs are required for fusion (69) (MJ. Root, 564

personal communication). This is consistent with a model in which VIR165-565

dependent Env is activated when one or two VIR165 molecules separate gp120 566

from gp41, while the remaining unbound FP(s) can insert into the target 567

membrane and mediate membrane fusion. 568

There is an interest in generating stable soluble mimics of the native Env 569

trimer for vaccine and structural studies. Efforts to generate such mimics have 570

been hampered by the metastable nature of the native Env spike. We have 571

reported a number of stabilization tricks, resulting in the stable trimer mimic 572

BG505 SOSIP.664 gp140 (18, 70-75). Additional modifications that kinetically trap 573

BG505 SOSIP.664 trimers in the unliganded state have recently been described (9, 574

38, 76, 77). We have shown that the VIR165-dependence mutations are useful for 575

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

19

further stabilization of recombinant different Env trimers to lock them in the 576

unliganded ground-state and improve their properties as immunogens (67). 577

578

ACKNOWLEDGEMENTS 579

We are grateful to Stef Heynen for technical assistance. We thank Michael Root for 580

sharing unpublished observations. 581

582

FUNDING 583

This work was supported by the Aids Fonds Netherlands, grants #2005021, 584

#2008013, and by National Institutes of Health Grants P01 AI82362, P01 585

AI110657, and UM1 AI100663 (Scripps CHAVI-ID). R.W.S. is a recipient of a Vidi 586

grant from the Netherlands Organization for Scientific Research (NWO) and a 587

Starting Investigator Grant from the European Research Council (ERC-StG-2011–588

280829-SHEV). P.J.K. is supported by NIH grant R37 AI36082.The funders had no 589

role in study design, data collection and analysis, decision to publish, or 590

preparation of the manuscript. 591 592

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

20

REFERENCES 593 594 1. Sanders RW, Moore JP. 2014. HIV: A stamp on the envelope. Nature 514:437-438. 595 2. Kong R, Xu K, Zhou T, Acharya P, Lemmin T, Liu K, Ozorowski G, Soto C, Taft JD, Bailer RT, Cale EM, Chen 596

L, Choi CW, Chuang GY, Doria-Rose NA, Druz A, Georgiev IS, Gorman J, Huang J, Joyce MG, Louder MK, 597 Ma X, McKee K, O'Dell S, Pancera M, Yang Y, Blanchard SC, Mothes W, Burton DR, Koff WC, Connors M, 598 Ward AB, Kwong PD, Mascola JR. 2016. Fusion peptide of HIV-1 as a site of vulnerability to neutralizing 599 antibody. Science 352:828-833. 600

3. Harrison SC. 2008. Viral membrane fusion. Nat Struct Mol Biol 15:690-698. 601 4. Colman PM, Lawrence MC. 2003. The structural biology of type I viral membrane fusion. Nat Rev Mol Cell 602

Biol 4:309-319. 603 5. Melikyan GB, Markosyan RM, Hemmati H, Delmedico MK, Lambert DM, Cohen FS. 2000. Evidence that 604

the transition of HIV-1 gp41 into a six-helix bundle, not the bundle configuration, induces membrane 605 fusion. J Cell Biol 151:413-423. 606

6. Lyumkis D, Julien JP, de Val N, Cupo A, Potter CS, Klasse PJ, Burton DR, Sanders RW, Moore JP, 607 Carragher B, Wilson IA, Ward AB. 2013. Cryo-EM structure of a fully glycosylated soluble cleaved HIV-1 608 envelope trimer. Science 342:1484-1490. 609

7. Julien JP, Cupo A, Sok D, Stanfield RL, Lyumkis D, Deller MC, Klasse PJ, Burton DR, Sanders RW, Moore 610 JP, Ward AB, Wilson IA. 2013. Crystal structure of a soluble cleaved HIV-1 envelope trimer. Science 611 342:1477-1483. 612

8. Pancera M, Zhou T, Druz A, Georgiev IS, Soto C, Gorman J, Huang J, Acharya P, Chuang GY, Ofek G, 613 Stewart-Jones GB, Stuckey J, Bailer RT, Joyce MG, Louder MK, Tumba N, Yang Y, Zhang B, Cohen MS, 614 Haynes BF, Mascola JR, Morris L, Munro JB, Blanchard SC, Mothes W, Connors M, Kwong PD. 2014. 615 Structure and immune recognition of trimeric pre-fusion HIV-1 Env. Nature 514:455-461. 616

9. Garces F, Lee JH, de Val N, de la Pena AT, Kong L, Puchades C, Hua Y, Stanfield RL, Burton DR, Moore JP, 617 Sanders RW, Ward AB, Wilson IA. 2015. Affinity Maturation of a Potent Family of HIV Antibodies Is 618 Primarily Focused on Accommodating or Avoiding Glycans. Immunity 43:1053-1063. 619

10. Scharf L, Wang H, Gao H, Chen S, McDowall AW, Bjorkman PJ. 2015. Broadly Neutralizing Antibody 620 8ANC195 Recognizes Closed and Open States of HIV-1 Env. Cell 162:1379-1390. 621

11. Moore JP, McKeating JA, Weiss RA, Sattentau QJ. 1990. Dissociation of gp120 from HIV-1 virions induced 622 by soluble CD4. Science 250:1139-1142. 623

12. Thali M, Furman C, Helseth E, Repke H, Sodroski J. 1992. Lack of correlation between soluble CD4-624 induced shedding of the human immunodeficiency virus type 1 exterior envelope glycoprotein and 625 subsequent membrane fusion events. J Virol 66:5516-5524. 626

13. Helseth E, Olshevsky U, Furman C, Sodroski J. 1991. Human immunodeficiency virus type 1 gp120 627 envelope glycoprotein regions important for association with the gp41 transmembrane glycoprotein. J 628 Virol 65:2119-2123. 629

14. Helseth E, Olshevsky U, Gabuzda D, Ardman B, Haseltine W, Sodroski J. 1990. Changes in the 630 transmembrane region of the human immunodeficiency virus type 1 gp41 envelope glycoprotein affect 631 membrane fusion. J Virol 64:6314-6318. 632

15. Yang X, Mahony E, Holm GH, Kassa A, Sodroski J. 2003. Role of the gp120 inner domain beta-sandwich in 633 the interaction between the human immunodeficiency virus envelope glycoprotein subunits. Virology 634 313:117-125. 635

16. Sen J, Jacobs A, Caffrey M. 2008. Role of the HIV gp120 conserved domain 5 in processing and viral entry. 636 Biochemistry 47:7788-7795. 637

17. Wang J, Sen J, Rong L, Caffrey M. 2008. Role of the HIV gp120 conserved domain 1 in processing and viral 638 entry. J Biol Chem 283:32644-32649. 639

18. Binley JM, Sanders RW, Clas B, Schuelke N, Master A, Guo Y, Kajumo F, Anselma DJ, Maddon PJ, Olson 640 WC, Moore JP. 2000. A recombinant human immunodeficiency virus type 1 envelope glycoprotein 641 complex stabilized by an intermolecular disulfide bond between the gp120 and gp41 subunits is an 642 antigenic mimic of the trimeric virion-associated structure. J Virol 74:627-643. 643

19. Finzi A, Xiang SH, Pacheco B, Wang L, Haight J, Kassa A, Danek B, Pancera M, Kwong PD, Sodroski J. 644 2010. Topological layers in the HIV-1 gp120 inner domain regulate gp41 interaction and CD4-triggered 645 conformational transitions. Mol Cell 37:656-667. 646

20. Pancera M, Majeed S, Ban YE, Chen L, Huang CC, Kong L, Kwon YD, Stuckey J, Zhou T, Robinson JE, Schief 647 WR, Sodroski J, Wyatt R, Kwong PD. Structure of HIV-1 gp120 with gp41-interactive region reveals 648 layered envelope architecture and basis of conformational mobility. Proc Natl Acad Sci U S A 107:1166-649 1171. 650

21. Zhou T, Xu L, Dey B, Hessell AJ, Van Ryk D, Xiang SH, Yang X, Zhang MY, Zwick MB, Arthos J, Burton DR, 651 Dimitrov DS, Sodroski J, Wyatt R, Nabel GJ, Kwong PD. 2007. Structural definition of a conserved 652 neutralization epitope on HIV-1 gp120. Nature 445:732-737. 653

22. Guttman M, Garcia NK, Cupo A, Matsui T, Julien JP, Sanders RW, Wilson IA, Moore JP, Lee KK. 2014. 654 CD4-induced activation in a soluble HIV-1 Env trimer. Structure 22:974-984. 655

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

21

23. Desormeaux A, Coutu M, Medjahed H, Pacheco B, Herschhorn A, Gu C, Xiang SH, Mao Y, Sodroski J, Finzi 656 A. 2013. The highly conserved layer-3 component of the HIV-1 gp120 inner domain is critical for CD4-657 required conformational transitions. Journal of virology 87:2549-2562. 658

24. Munch J, Standker L, Adermann K, Schulz A, Schindler M, Chinnadurai R, Pohlmann S, Chaipan C, Biet T, 659 Peters T, Meyer B, Wilhelm D, Lu H, Jing W, Jiang S, Forssmann WG, Kirchhoff F. 2007. Discovery and 660 optimization of a natural HIV-1 entry inhibitor targeting the gp41 fusion peptide. Cell 129:263-275. 661

25. Forssmann WG, The YH, Stoll M, Adermann K, Albrecht U, Barlos K, Busmann A, Canales-Mayordomo A, 662 Gimenez-Gallego G, Hirsch J, Jimenez-Barbero J, Meyer-Olson D, Munch J, Perez-Castells J, Standker L, 663 Kirchhoff F, Schmidt RE. 2010. Short-term monotherapy in HIV-infected patients with a virus entry 664 inhibitor against the gp41 fusion peptide. Science translational medicine 2:63re63. 665

26. Gonzalez E, Ballana E, Clotet B, Este JA. 2011. Development of resistance to VIR-353 with cross-resistance 666 to the natural HIV-1 entry virus inhibitory peptide (VIRIP). AIDS 25:1557-1583. 667

27. Eggink D, Baldwin CE, Deng Y, Langedijk JP, Lu M, Sanders RW, Berkhout B. 2008. Selection of T1249-668 resistant human immunodeficiency virus type 1 variants. J Virol 82:6678-6688. 669

28. Eggink D, Langedijk JP, Bonvin AM, Deng Y, Lu M, Berkhout B, Sanders RW. 2009. Detailed mechanistic 670 insights into HIV-1 sensitivity to three generations of fusion inhibitors. J Biol Chem 284:26941-26950. 671

29. Jeeninga RE, Keulen W, Boucher C, Sanders RW, Berkhout B. 2001. Evolution of AZT resistance in HIV-1: 672 the 41-70 intermediate that is not observed in vivo has a replication defect. Virology 283:294-305. 673

30. Peden K, Emerman M, Montagnier L. 1991. Changes in growth properties on passage in tissue culture of 674 viruses derived from infectious molecular clones of HIV-1LAI, HIV-1MAL, and HIV-1ELI. Virology 185:661-675 672. 676

31. Sanders RW, Busser E, Moore JP, Lu M, Berkhout B. 2004. Evolutionary repair of HIV type 1 gp41 with a 677 kink in the N-terminal helix leads to restoration of the six-helix bundle structure. AIDS Res Hum 678 Retroviruses 20:742-749. 679

32. Das AT, Klaver B, Berkhout B. 1999. A hairpin structure in the R region of the human immunodeficiency 680 virus type 1 RNA genome is instrumental in polyadenylation site selection. J Virol 73:81-91. 681

33. Wei X, Decker JM, Liu H, Zhang Z, Arani RB, Kilby JM, Saag MS, Wu X, Shaw GM, Kappes JC. 2002. 682 Emergence of resistant human immunodeficiency virus type 1 in patients receiving fusion inhibitor (T-20) 683 monotherapy. Antimicrob Agents Chemother 46:1896-1905. 684

34. Derdeyn CA, Decker JM, Sfakianos JN, Wu X, O'Brien WA, Ratner L, Kappes JC, Shaw GM, Hunter E. 685 2000. Sensitivity of human immunodeficiency virus type 1 to the fusion inhibitor T-20 is modulated by 686 coreceptor specificity defined by the V3 loop of gp120. J Virol 74:8358-8367. 687

35. Montefiori DC. 2009. Measuring HIV neutralization in a luciferase reporter gene assay. Methods Mol Biol 688 485:395-405. 689

36. Sarzotti-Kelsoe M, Bailer RT, Turk E, Lin CL, Bilska M, Greene KM, Gao H, Todd CA, Ozaki DA, Seaman 690 MS, Mascola JR, Montefiori DC. 2014. Optimization and validation of the TZM-bl assay for standardized 691 assessments of neutralizing antibodies against HIV-1. J Immunol Methods 409:131-146. 692

37. Mascola JR, D'Souza P, Gilbert P, Hahn BH, Haigwood NL, Morris L, Petropoulos CJ, Polonis VR, Sarzotti 693 M, Montefiori DC. 2005. Recommendations for the design and use of standard virus panels to assess 694 neutralizing antibody responses elicited by candidate human immunodeficiency virus type 1 vaccines. J 695 Virol 79:10103-10107. 696

38. Do Kwon Y, Pancera M, Acharya P, Georgiev IS, Crooks ET, Gorman J, Joyce MG, Guttman M, Ma X, 697 Narpala S, Soto C, Terry DS, Yang Y, Zhou T, Ahlsen G, Bailer RT, Chambers M, Chuang GY, Doria-Rose 698 NA, Druz A, Hallen MA, Harned A, Kirys T, Louder MK, O'Dell S, Ofek G, Osawa K, Prabhakaran M, Sastry 699 M, Stewart-Jones GB, Stuckey J, Thomas PV, Tittley T, Williams C, Zhang B, Zhao H, Zhou Z, Donald BR, 700 Lee LK, Zolla-Pazner S, Baxa U, Schon A, Freire E, Shapiro L, Lee KK, Arthos J, Munro JB, Blanchard SC, 701 Mothes W, Binley JM, et al. 2015. Crystal structure, conformational fixation and entry-related 702 interactions of mature ligand-free HIV-1 Env. Nature structural & molecular biology 22:522-531. 703

39. van Anken E, Sanders RW, Liscaljet IM, Land A, Bontjer I, Tillemans S, Nabatov AA, Paxton WA, 704 Berkhout B, Braakman I. 2008. Only five of 10 strictly conserved disulfide bonds are essential for folding 705 and eight for function of the HIV-1 envelope glycoprotein. Mol Biol Cell 19:4298-4309. 706

40. Leonard CK, Spellman MW, Riddle L, Harris RJ, Thomas JN, Gregory TJ. 1990. Assignment of intrachain 707 disulfide bonds and characterization of potential glycosylation sites of the type 1 recombinant human 708 immunodeficiency virus envelope glycoprotein (gp120) expressed in Chinese hamster ovary cells. J Biol 709 Chem 265:10373-10382. 710

41. Eggink D, Bontjer I, Langedijk JP, Berkhout B, Sanders RW. 2011. Resistance of Human Immunodeficiency 711 Virus Type 1 to a Third-Generation Fusion Inhibitor Requires Multiple Mutations in gp41 and Is 712 Accompanied by a Dramatic Loss of gp41 Function. J Virol 85:10785-10797. 713

42. database LA. HIV Sequence Database. http://www.hiv.lanl.gov/content/index. Accessed 714 43. Steffy KR, Kraus G, Looney DJ, Wong-Staal F. 1992. Role of the fusogenic peptide sequence in syncytium 715

induction and infectivity of human immunodeficiency virus type 2. Journal of virology 66:4532-4535. 716

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

22

44. Kassa A, Finzi A, Pancera M, Courter JR, Smith AB, 3rd, Sodroski J. 2009. Identification of a human 717 immunodeficiency virus type 1 envelope glycoprotein variant resistant to cold inactivation. J Virol 718 83:4476-4488. 719

45. Kassa A, Madani N, Schon A, Haim H, Finzi A, Xiang SH, Wang L, Princiotto A, Pancera M, Courter J, 720 Smith AB, 3rd, Freire E, Kwong PD, Sodroski J. 2009. Transitions to and from the CD4-bound 721 conformation are modulated by a single-residue change in the human immunodeficiency virus type 1 722 gp120 inner domain. J Virol 83:8364-8378. 723

46. Baldwin CE, Sanders RW, Deng Y, Jurriaans S, Lange JM, Lu M, Berkhout B. 2004. Emergence of a drug-724 dependent human immunodeficiency virus type 1 variant during therapy with the T20 fusion inhibitor. J 725 Virol 78:12428-12437. 726

47. Cole AL, Yang OO, Warren AD, Waring AJ, Lehrer RI, Cole AM. 2006. HIV-1 adapts to a retrocyclin with 727 cationic amino acid substitutions that reduce fusion efficiency of gp41. J Immunol 176:6900-6905. 728

48. Klasse PJ. 2007. Modeling how many envelope glycoprotein trimers per virion participate in human 729 immunodeficiency virus infectivity and its neutralization by antibody. Virology 369:245-262. 730

49. Klasse PJ. 2012. The molecular basis of HIV entry. Cellular microbiology 14:1183-1192. 731 50. Ray N, Blackburn LA, Doms RW. 2009. HR-2 mutations in human immunodeficiency virus type 1 gp41 732

restore fusion kinetics delayed by HR-1 mutations that cause clinical resistance to enfuvirtide. Journal of 733 virology 83:2989-2995. 734

51. Tolstrup M, Selzer-Plon J, Laursen AL, Bertelsen L, Gerstoft J, Duch M, Pedersen FS, Ostergaard L. 2007. 735 Full fusion competence rescue of the enfuvirtide resistant HIV-1 gp41 genotype (43D) by a prevalent 736 polymorphism (137K). AIDS 21:519-521. 737

52. Xu L, Pozniak A, Wildfire A, Stanfield-Oakley SA, Mosier SM, Ratcliffe D, Workman J, Joall A, Myers R, 738 Smit E, Cane PA, Greenberg ML, Pillay D. 2005. Emergence and evolution of enfuvirtide resistance 739 following long-term therapy involves heptad repeat 2 mutations within gp41. Antimicrobial agents and 740 chemotherapy 49:1113-1119. 741

53. Reeves JD, Gallo SA, Ahmad N, Miamidian JL, Harvey PE, Sharron M, Pohlmann S, Sfakianos JN, Derdeyn 742 CA, Blumenthal R, Hunter E, Doms RW. 2002. Sensitivity of HIV-1 to entry inhibitors correlates with 743 envelope/coreceptor affinity, receptor density, and fusion kinetics. Proceedings of the National Academy 744 of Sciences of the United States of America 99:16249-16254. 745

54. Platt EJ, Durnin JP, Kabat D. 2005. Kinetic factors control efficiencies of cell entry, efficacies of entry 746 inhibitors, and mechanisms of adaptation of human immunodeficiency virus. Journal of virology 79:4347-747 4356. 748

55. Shafer RW, Schapiro JM. 2008. HIV-1 drug resistance mutations: an updated framework for the second 749 decade of HAART. AIDS Rev 10:67-84. 750

56. Chertova E, Bess Jr JW, Jr., Crise BJ, Sowder IR, Schaden TM, Hilburn JM, Hoxie JA, Benveniste RE, Lifson 751 JD, Henderson LE, Arthur LO. 2002. Envelope glycoprotein incorporation, not shedding of surface 752 envelope glycoprotein (gp120/SU), Is the primary determinant of SU content of purified human 753 immunodeficiency virus type 1 and simian immunodeficiency virus. J Virol 76:5315-5325. 754

57. Haim H, Si Z, Madani N, Wang L, Courter JR, Princiotto A, Kassa A, DeGrace M, McGee-Estrada K, 755 Mefford M, Gabuzda D, Smith AB, 3rd, Sodroski J. 2009. Soluble CD4 and CD4-mimetic compounds 756 inhibit HIV-1 infection by induction of a short-lived activated state. PLoS Pathog 5:e1000360. 757

58. Haim H, Strack B, Kassa A, Madani N, Wang L, Courter JR, Princiotto A, McGee K, Pacheco B, Seaman 758 MS, Smith AB, 3rd, Sodroski J. 2011. Contribution of intrinsic reactivity of the HIV-1 envelope 759 glycoproteins to CD4-independent infection and global inhibitor sensitivity. PLoS pathogens 7:e1002101. 760

59. Agrawal N, Leaman DP, Rowcliffe E, Kinkead H, Nohria R, Akagi J, Bauer K, Du SX, Whalen RG, Burton 761 DR, Zwick MB. 2011. Functional stability of unliganded envelope glycoprotein spikes among isolates of 762 human immunodeficiency virus type 1 (HIV-1). PloS one 6:e21339. 763

60. Leaman DP, Zwick MB. 2013. Increased functional stability and homogeneity of viral envelope spikes 764 through directed evolution. PLoS pathogens 9:e1003184. 765

61. Ray N, Blackburn LA, Doms RW. 2009. HR-2 mutations in human immunodeficiency virus type 1 gp41 766 restore fusion kinetics delayed by HR-1 mutations that cause clinical resistance to enfuvirtide. J Virol 767 83:2989-2995. 768

62. Baldwin C, Berkhout B. 2008. Mechanistic studies of a T20-dependent human immunodeficiency virus 769 type 1 variant. J Virol 82:7735-7740. 770

63. Bai X, Wilson KL, Seedorff JE, Ahrens D, Green J, Davison DK, Jin L, Stanfield-Oakley SA, Mosier SM, 771 Melby TE, Cammack N, Wang Z, Greenberg ML, Dwyer JJ. 2008. Impact of the Enfuvirtide Resistance 772 Mutation N43D and the Associated Baseline Polymorphism E137K on Peptide Sensitivity and Six-Helix 773 Bundle Structure. Biochemistry doi:10.1021/bi702509d. 774

64. Aberham C, Weber S, Phares W. 1996. Spontaneous mutations in the human immunodeficiency virus 775 type 1 gag gene that affect viral replication in the presence of cyclosporins. J Virol 70:3536-3544. 776

65. Menzo S, Monachetti A, Balotta C, Corvasce S, Rusconi S, Paolucci S, Baldanti F, Bagnarelli P, Clementi 777 M. 2003. Processivity and drug-dependence of HIV-1 protease: determinants of viral fitness in variants 778 resistant to protease inhibitors. AIDS 17:663-671. 779

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

23

66. Baldwin C, Berkhout B. 2007. HIV-1 drug-resistance and drug-dependence. Retrovirology 4:78. 780 67. de Taeye SW, Ozorowski G, Torrents de la Pena A, Guttman M, Julien JP, van den Kerkhof TL, Burger JA, 781

Pritchard LK, Pugach P, Yasmeen A, Crampton J, Hu J, Bontjer I, Torres JL, Arendt H, DeStefano J, Koff 782 WC, Schuitemaker H, Eggink D, Berkhout B, Dean H, LaBranche C, Crotty S, Crispin M, Montefiori DC, 783 Klasse PJ, Lee KK, Moore JP, Wilson IA, Ward AB, Sanders RW. 2015. Immunogenicity of Stabilized HIV-1 784 Envelope Trimers with Reduced Exposure of Non-neutralizing Epitopes. Cell 163:1702-1715. 785

68. Reeves JD, Gallo SA, Ahmad N, Miamidian JL, Harvey PE, Sharron M, Pohlmann S, Sfakianos JN, Derdeyn 786 CA, Blumenthal R, Hunter E, Doms RW. 2002. Sensitivity of HIV-1 to entry inhibitors correlates with 787 envelope/coreceptor affinity, receptor density, and fusion kinetics. Proc Natl Acad Sci U S A 99:16249-788 16254. 789

69. Salzwedel K, Berger EA. 2009. Complementation of diverse HIV-1 Env defects through cooperative 790 subunit interactions: a general property of the functional trimer. Retrovirology 6:75. 791

70. Binley JM, Sanders RW, Master A, Cayanan CS, Wiley CL, Schiffner L, Travis B, Kuhmann S, Burton DR, 792 Hu SL, Olson WC, Moore JP. 2002. Enhancing the proteolytic maturation of human immunodeficiency 793 virus type 1 envelope glycoproteins. J Virol 76:2606-2616. 794

71. Sanders RW, Vesanen M, Schuelke N, Master A, Schiffner L, Kalyanaraman R, Paluch M, Berkhout B, 795 Maddon PJ, Olson WC, Lu M, Moore JP. 2002. Stabilization of the soluble, cleaved, trimeric form of the 796 envelope glycoprotein complex of human immunodeficiency virus type 1. J Virol 76:8875-8889. 797

72. Ringe RP, Sanders RW, Yasmeen A, Kim HJ, Lee JH, Cupo A, Korzun J, Derking R, van Montfort T, Julien 798 JP, Wilson IA, Klasse PJ, Ward AB, Moore JP. 2013. Cleavage strongly influences whether soluble HIV-1 799 envelope glycoprotein trimers adopt a native-like conformation. Proceedings of the National Academy of 800 Sciences of the United States of America 110:18256-18261. 801

73. Sanders RW, Derking R, Cupo A, Julien JP, Yasmeen A, de Val N, Kim HJ, Blattner C, de la Pena AT, 802 Korzun J, Golabek M, de Los Reyes K, Ketas TJ, van Gils MJ, King CR, Wilson IA, Ward AB, Klasse PJ, 803 Moore JP. 2013. A next-generation cleaved, soluble HIV-1 Env trimer, BG505 SOSIP.664 gp140, expresses 804 multiple epitopes for broadly neutralizing but not non-neutralizing antibodies. PLoS pathogens 805 9:e1003618. 806

74. Klasse PJ, Depetris RS, Pejchal R, Julien JP, Khayat R, Lee JH, Marozsan AJ, Cupo A, Cocco N, Korzun J, 807 Yasmeen A, Ward AB, Wilson IA, Sanders RW, Moore JP. 2013. Influences on trimerization and 808 aggregation of soluble, cleaved HIV-1 SOSIP envelope glycoprotein. Journal of virology 87:9873-9885. 809

75. Sanders RW, van Gils MJ, Derking R, Sok D, Ketas TJ, Burger JA, Ozorowski G, Cupo A, Simonich C, Goo L, 810 Arendt H, Kim HJ, Lee JH, Pugach P, Williams M, Debnath G, Moldt B, van Breemen MJ, Isik G, Medina-811 Ramirez M, Back JW, Koff WC, Julien JP, Rakasz EG, Seaman MS, Guttman M, Lee KK, Klasse PJ, 812 LaBranche C, Schief WR, Wilson IA, Overbaugh J, Burton DR, Ward AB, Montefiori DC, Dean H, Moore 813 JP. 2015. HIV-1 VACCINES. HIV-1 neutralizing antibodies induced by native-like envelope trimers. Science 814 349:aac4223. 815

76. Kwon YD, Pancera M, Acharya P, Georgiev IS, Crooks ET, Gorman J, Joyce MG, Guttman M, Ma X, 816 Narpala S, Soto C, Terry DS, Yang Y, Zhou T, Ahlsen G, Bailer RT, Chambers M, Chuang GY, Doria-Rose 817 NA, Druz A, Hallen MA, Harned A, Kirys T, Louder MK, O'Dell S, Ofek G, Osawa K, Prabhakaran M, Sastry 818 M, Stewart-Jones GB, Stuckey J, Thomas PV, Tittley T, Williams C, Zhang B, Zhao H, Zhou Z, Donald BR, 819 Lee LK, Zolla-Pazner S, Baxa U, Schon A, Freire E, Shapiro L, Lee KK, Arthos J, Munro JB, Blanchard SC, 820 Mothes W, Binley JM, et al. 2015. Crystal structure, conformational fixation and entry-related 821 interactions of mature ligand-free HIV-1 Env. Nat Struct Mol Biol 22:522-531. 822

77. Guenaga J, Dubrovskaya V, de Val N, Sharma SK, Carrette B, Ward AB, Wyatt RT. 2016. Structure-Guided 823 Redesign Increases the Propensity of HIV Env To Generate Highly Stable Soluble Trimers. J Virol 90:2806-824 2817. 825

78. Garces F, Lee JH, de Val N, Torrents de la Pena A, Kong L, Puchades C, Hua Y, Stanfield RL, Burton DR, 826 Moore JP, Sanders RW, Ward AB, Wilson IA. 2015. Affinity Maturation of a Potent Family of HIV 827 Antibodies Is Primarily Focused on Accommodating or Avoiding Glycans. Immunity 43:1053-1063. 828

829 830

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

24

TABLES 831 832 Table 1. Mutations selected in VIR165 escape studies 833 834

835 1Env numbering according to HXB2 sequence, gp41 numbering indicated in brackets 836

*Mixed sequences detected 837

838

C1 HR1 Env position 421 58 60 64 66 558 (47) 577 (66)

Culture 1

A58A/V* H66R

2 Q577R

3 E64K/E

* A558T/A

* 4 A558T 5 V42I E64K

6 A60E/A

*

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

25

Table 2. VIR165-resistance in single cycle infection assays 839

VIR1651

Virus Mutant IC50 (μg/ml)

Fold resistance

WT 2.7 1.0 V42I 10.6 3.8 A58V 21.2 7.7 A60E 602 37 C1 D62N 3.2 1.2 E64K 602 37 H66N 3.2 1.1 H66R 1002 58 A522V 2.2 0.8 HR1 A558T 27.33 9.93 Q577R

A549A A549A/N637K

>30 3.5 2.4

>10 1.3 0.9

840 1Fold -resistance values indicated in bold represent resistance of 3-fold or more compared to WT 841 virus. 842 2VIR165-dependent viruses, IC50 values estimated based on Fig. 2E. 843 3Estimated IC50 and fold-resistance values as no complete sigmoidal dose-response curve was 844 obtained. 845

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

26

Table 3. Sensitivity of VIR165-resistant and -dependent mutants to a series of antibodies, entry and fusion inhibitors 846

HIV-Ig2 sCD42 CD4-IgG22 AMD31002 T203

Virus mutant IC50 n-fold4 IC50 n-fold IC50 n-fold IC50 n-fold IC50 n-fold

WT 2.40 1.0 2.30 1.0 2.29 1.0 1.85 1.0 54.9 1.0 V42I 4.90 2.0 2.57 1.1 6.98 3.1 2.02 1.1 68.9 1.3 A58V 2.58 1.1 3.01 1.3 2.14 0.9 1.96 1.1 14.7 0.3 D62N 3.08 1.3 2.10 0.9 3.76 1.6 2.08 1.1 53.3 1.0 E64K1 2.76 1.2 9.84 4.3 7.88 3.4 0.29 0.2 22.0 0.4 H66R1 1.84 0.8 30.30 13.2 14.42 6.3 0.58 0.3 16.5 0.3 A558T 2.40 1.0 0.99 0.4 2.07 0.9 1.26 0.7 6.7 0.1 A549A ND ND ND ND 2983 54.3 A549A/N637K ND ND ND ND 1426 26.0 Q577R ND ND ND ND 71.4 1.3

847 1 Because VIR165-dependent mutants are not infectious in the absence of VIR165, 10 μg/ml VIR165 was present for the A60E, E64K and H66R mutants 848 2 Values presented in μg/ml 849 3 Values presented in ng/ml 850 4 Fold-resistance (n-fold) compared to WT. Values indicated in bold represent 3-fold or more resistance compared to WT; values indicated in bold italic represent 3- fold 851 or more sensitivity compared to WT 852

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

27

Table 4. Stability of VIR165-resistant and -dependent mutants in the absence or presence of sCD4 853

854 855 856 857 858 859 860

1 Average values are shown from at least three independent experiments, each performed in duplo 861 with SEM values shown in brackets. 862

863

Decay at 37°C Thermostability t1/2 (h) t1/2 (+sCD4) Δt1/2 Tm (°C) Tm (°C) ΔTm (°C)

+sCD4 +sCD4 WT 11.3 (0.9) 1 6.1 (0.3) -5.2 (1.0) 44.6 (0.3) 41.9 (0.2) 2.7 (0.3)

A58V 8.2 (1.1) 5.6 (0.6) -2.6 (0.5) 44.8 (0.2) 42.1 (0.1) 2.7 (0.2) A60E 9.4 (0.5) 7.0 (0.4) -2.4 (0.6) 44.7 (0.2) 43.3 (0.1) 1.4 (0.1) E64K 9.5 (1.4) 8.0 (1.2) -1.5 (0.9) 44.6 (0.2) 43.6 (0.1) 0.9 (0.1) H66R 8.2 (1.3) 7.9 (1.1) -0.3 (0.5) 44.7 (0.2) 44.2 (0.4) 0.5 (0.3)

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

28

FIGURE LEGENDS 864 865 Figure 1. Escape of HIV-1LAI form anchor inhibitor VIR165. (A) Sequence of the 866

natural peptide VIRIP and the more potent and stable derivatives VIR165 and 867

VIR353, both cyclized by a disulfide bond, and the di-peptide VIR576. The p in 868

VIR353 indicates the introduction of a D-proline. (B) Step-wise increase of the 869

VIR165 concentration during passage experiments. Cultures were passaged twice a 870

week for 2½ months with increasing concentrations of VIR165. (C) Linear 871

representation of HIVLAI gp160 showing positions of VIR165 escape mutations in C1 872

and HR1, as well as previously identified escape mutations to VIRIP analogue 873

VIR353 (26). (D) The residues involved in VIR165-resistance and -dependence were 874

mapped on the structure of the BG505 SOSIP.664 trimer structure containing the 875

complete gp41 interactive domain (PDB accession code 5CEZ (78)) using Pymol 876

(DeLano Scientific; http://pymol.sourceforge.net). The residues are shown as 877

spheres on a cartoon model of one of the protomers, with gp41 in sand and gp120 in 878

green. A surface model of the two other protomers is shown in white and grey. 879

Residues at position 42, 58, 558 and 577 which are involved in VIR165-resistance, 880

are indicated in red. The control position 62 is indicated in cyan, and residues 60, 64 881

and 66, which are involved in VIR165-dependence, are depicted in yellow. For 882

comparison, residues involved in resistance to VIRIP analogue VIR353 are indicated 883

in blue (26). The right panel shows a detail of the region that includes residues 58, 884

60, 64 and 66, all situated in layer 1 of gp120. Gp41 is depicted in orange, gp120 is 885

in green, with layer 1 in magenta and layer 2 in blue. 886

887

Figure 2. HIV-1LAI VIR165 escape variants can be resistant to or dependent on VIR165. 888

Single cycle infection experiments were performed as described in the material and 889

methods section. Inhibition of HIV-1LAI variants containing (A) VIR165-resistance 890

mutations in gp41, (B) VIR165-resistance mutations in the C1 domain, (C) VIR165-891

dependence mutations within C1. A D62N control virus, the T20-resistant variant 892

V549A, and the T20-dependent virus V549A/N637K (46) were included as control 893

viruses. (D) Infectivity of VIR165-resistant and -dependent viruses in the absence of 894

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

29

VIR165 in a single cycle infection assay relative to WT virus. The virus mutants 895

indicated with an (*) were VIR165-dependent, i.e. they were not infectious in the 896

absence of VIR165. (E) VIR165-dependent virus variants are inhibited at high 897

VIR165 concentrations. Single cycle infection experiments were performed using 898

concentrations of VIR165 up to 300 μg/ml revealing a bell-shaped dose-response 899

curve. (F) Maximum infectivity of VIR165-dependent mutants relative to WT in the 900

presence of VIR165. The infectivity of VIR165-dependent viruses was obtained in 901

the presence of VIR165 (indicated with an *) at 10 μg/ml VIR165 (A60E and E64K), 902

or 30 μg/ml (H66R). WT infectivity was measured in the absence of VIR165 (WT) or 903

in the presence of 10 μg/ml VIR165 (WT*). 904

Figure 3. Modeling of trimer occupancy by VIR165 explains the bell-shaped dose-905

response curves of VIR165-dependent viruses. The enhancement and inhibition of WT 906

and VIR165-dependent viruses was modeled mathematically to assess whether 907

enhancing or inhibitory effects of VIR165 could be associated with occupancy levels 908

of 1, 2, or 3 VIR165 molecules per trimer. Different mathematical occupancy models 909

were fitted to the experimentally obtained bell-shaped stimulation-inhibition data 910

of Fig. 2E. The best fit models are shown. 911

912

Figure 4. VIR165-dependent viruses are less prone to CD4-induced decay. (A) 913

Representative experiment of at least three independent experiments in which 914

VIR165-resistant and VIR165-dependent viruses were incubated at physiological 915

temperature (37˚C) in the absence or presence of sCD4, for different time intervals, 916

followed by assessing the remaining infectivity on TZM-bl cells in the absence of 917

VIR165 for WT virus or in the presence of 10 or 30 μg/ml VIR165 for the dependent 918

variants. (B) Half life (t1/2) for WT and VIR165 resistant and dependent mutants, 919

measured in at least three independent experiments performed in duplo. (C) The 920

difference in t1/2 when incubated in the absence or presence of sCD4 during 921

incubation at 37˚C. Statistical significance is indicated with asterisks: * P < 0.05; ** P 922

< 0.005. 923

924

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

30

Figure 5. VIR165-dependent viruses show increased thermostability in the presence of 925

sCD4. (A) The thermostability of WT, VIR165-resistant and VIR165-dependent 926

viruses was tested by incubating them for 1 h at escalating temperatures, followed 927

by testing the remaining infectivity on TZM-bl reporter cells in the absence of 928

VIR165 (WT) or the presence of 10 or 30 μg/ml VIR165 (dependent variants). A 929

representative experiment is show of at least three independent experiments. (B) 930

The midpoints of thermal denaturation (Tm) of WT and mutant viruses measured in 931

at least three independent experiments performed in duplo. (C) Changes in Tm in the 932

presence of sCD4. Statistical significance is indicated with asterisks: ** P < 0.005; *** 933

P < 0.0005. 934

935

Figure 6. Hypothetical model of the inhibiting and enhancing modes of action of 936

VIRIP-derived peptides of WT and VIR165-dependent viruses. 937

(A) The FP becomes fully exposed upon binding of the CD4 receptor and co-receptor 938

and is inserted into the target membrane, followed by subsequent fusion of the two 939

membranes. (B) VIR165 binds to the FP in the short-lived intermediate state that is 940

induced by CD4-binding, thereby inhibiting insertion of the FP into the target 941

membrane. (C) In VIR165-dependent viruses a hyperstable gp120-gp41 interaction 942

cannot be sufficiently weakened by CD4-binding and CD4-induced conformational 943

changes in the gp120-gp41 complex and subsequent entry steps are blocked. (D) 944

VIR165 in sub-saturating amounts acts as a “wedge” between gp120 and gp41, 945

destabilizing the gp120-gp41 interaction and facilitating CD4-induced 946

conformational changes that allow subsequent entry steps. We do not know 947

whether in this scenario VIR165 remains associated with the one or two FP(s) that 948

is occupies and only the free FP(s) insert into the cell membrane, or whether it 949

dissociates to allow all three FPs to insert. (E) High concentrations of VIR165 result 950

in occupancy of all three FPs of the Env trimer, effectively inhibiting infection. 951

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

LEAIPMSIPPEVKFNKPFVF VIRIPLEAIPCSIPPCFAFNKPFVF VIR165

LEAIPCSIPpCFLFNKPFVF VIR353

LEAIPCSIPPEFLFGKPFVF(x2) VIR576

BA

S S

Figure 1

C

D

S S

90°58

64

66

60

0 5 10 15 202

4

8

16

32

64

128

256

IC50 WT HIV-1LAI

time (passages)

VIR

165

(µg/

ml)

V1 V2 C2 V3 C3 V4 C4 V5 C5 C1 HR1 HR2 TM 131# 511#470#460#418#385#331#296#196#157#

FP

42#58#60#

64#

66# 563#577#

gp120# gp41#

558433 489545 570 612

62

S612

V570

V489

A433

A558

V42

A60

E64H66E62

A558Q577

A58

V42

A60E64

H66E62

S612

V489

A433

A558

L545 L545 on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

WTWT*

A60E*

E64K*

H66R*

0

25

50

75

100

125

150

Max

imum

infe

ctiv

ity (%

WT)

A B C

D

Figure 2

E F 01030

WTV42

IA58

VA60

ED62

NE64

KH66

NH66

RA51

7VA55

8T

Q577R

0

25

50

75

100

125

150

Max

imum

infe

ctiv

ity (%

WT)

* * *

(μgVIR165/ml)

0.001 0.01 0.1 1 10 100 10000

25

50

75

100

125

150WT

E64KH66R

A60E

[VIR165 (µg/ml)]

Rel

ativ

e In

fect

ivity

(% m

ax)

WTWT*

A60E*

E64K*

H66R*

106

107

108

Luci

fera

se u

nits

01030(μgVIR165/ml)

0.1 1 10 1000

25

50

75

100

125

150

D62N

WTV42IA58V

[VIR165 (µg/ml)]

Rel

ativ

e In

fect

ivity

(%)

0.1 1 10 1000

25

50

75

100

125

150

A60ED62NE64KH66R

WT

[VIR165 (µg/ml)]

Rel

ativ

e In

fect

ivity

(%)

0.1 1 10 1000

25

50

75

100

125

150 WT V549A

Q577RA558TV549A/N637K

[VIR165 (µg/ml)]

Rel

ativ

e In

fect

ivity

(%)

0.1 1 10 100 10000

25

50

75

100

125

150 BG505-WTBG505-E64KBG505-H66R

[VIR165] µg/ml

Rel

ativ

e In

fect

ivity

(% m

ax)G

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

Figure 3

WT A60E

E64K H66R

R2 0.964 R2 0.964

R2 0.937 R2 0.993

0.001 0.010 0.100 1 10 100 10000

25

50

75

100

125

150

[VIR165 (µg/ml)]

Rel

ativ

e In

fect

ivity

(% m

ax)

0.001 0.010 0.100 1 10 100 10000

25

50

75

100

125

150

[VIR165 (µg/ml)]

Rel

ativ

e In

fect

ivity

(% m

ax)

0.001 0.010 0.100 1 10 100 10000

25

50

75

100

125

150

[VIR165 (µg/ml)]

Rel

ativ

e In

fect

ivity

(% m

ax)

0.001 0.010 0.100 1 10 100 10000

25

50

75

100

125

150

[VIR165 (µg/ml)]

Rel

ativ

e In

fect

ivity

(% m

ax)

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

0.1 1 10 1000

25

50

75

100

125

time (hrs)

Rel

ativ

e In

fect

ivity

(%)

0.1 1 10 1000

25

50

75

100

125

time (hrs)

Rel

ativ

e In

fect

ivity

(%)

0.1 1 10 1000

25

50

75

100

125

time (hrs)

Rel

ativ

e In

fect

ivity

(%)

0.1 1 10 1000

25

50

75

100

125

time (hrs)

Rel

ativ

e In

fect

ivity

(%)

0.1 1 10 1000

25

50

75

100

125

no sCD4+ sCD4

time (hrs)

Rel

ativ

e In

fect

ivity

(%)

Figure 4

A

WT A58V A60E

E64K H66R

WT A58V A60E E64K H66R0

1

2

3

4

5

6

7

-Δt 1/

2 (ho

urs,

+sC

D4)

* ***

B

C*

WT A58V A60E E64K H66R

6

8

10

12

14

t 1/2 (

hour

s)

*

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

Figure 5

B

C

A

WT A58V A60E

E64K H66R

36 41 46 510

25

50

75

100

125

temp (°C)

Rel

ativ

e In

fect

ivity

(%)

36 41 46 510

25

50

75

100

125

temp (°C)

Rel

ativ

e In

fect

ivity

(%)

36 41 46 510

25

50

75

100

125

temp (°C)

Rel

ativ

e In

fect

ivity

(%)

36 41 46 510

25

50

75

100

125

temp (°C)

Rel

ativ

e In

fect

ivity

(%)

36 41 46 510

25

50

75

100

125

no sCD4+ sCD4

temp (°C)

Rel

ativ

e In

fect

ivity

(%)

WT A58V A60E E64K H66R0.0

0.5

1.0

1.5

2.0

2.5

3.0

ΔT m

(°C

)+s

CD

4 co

mpa

red

to n

o sC

D4 ** ***

***WT A58V H66R A60E E64K

43.0

43.5

44.0

44.5

45.0

T m (°

C)

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from

Figure 6

A

B

C

D

E

CD4Cell

Virus

VIR165

CXCR4

WT

WT+VIR165

VIR165-dependent

VIR165

VIR165Dependent+highVIR165

VIR165

VIR165-dependent+lowVIR165

post-fusionstateNativepre-fusionstate CD4-boundstate Intermediate1 Intermediate2

on April 4, 2018 by guest

http://jvi.asm.org/

Dow

nloaded from