10.1128@JVI.00486-20

JVI Accepted Manuscript Posted Online 1 July 2020
J. Virol. doi:10.1128/JVI.00486-20
Copyright © 2020 American Society for Microbiology. All Rights Reserved.
1 A conformational escape reaction of HIV-1 against an allosteric integrase inhibitor
Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

3 Running Title: HIV integrase multimerization against anti-HIV agents
6 Tomofumi Nakamura,a Teruya Nakamura,b,c Masayuki Amano,a Toshikazu Miyakawa,a
7 Yuriko Yamagata,c Masao Matsuoka,a,d Hirotomo Nakataa*
a
9 Department of Hematology, Rheumatology, and Infectious Diseases, Graduate School of Medical
10 Sciences, Faculty of Life Sciences, Kumamoto University, Kumamoto 860-8556, Japan.
b
11 Priority Organization for Innovation and Excellence, Kumamoto University, Kumamoto 862-0973,
12 Japan.
c
13 Graduate School of Pharmaceutical Sciences, Kumamoto University, Kumamoto 862-0973, Japan
d
14 Laboratory of Virus Control, Institute for Frontier Life and Medical Sciences, Kyoto University,
15 Sakyo-ku, Kyoto 606-8507, Japan.
16
17 *Corresponding author: Hirotomo Nakata, nakatahi@gpo.kumamoto-u.ac.jp; Tel.: +81-096-373-5156;
18 Fax: +81-096-373-5158.
19
1
20 Abstract
21 HIV-1 often acquires drug-resistant mutations in spite of the benefits of antiretroviral therapy

22 (ART). HIV-1 integrase (IN) is essential for concerted integration of HIV-1 DNA into the host genome.
23 IN further contributes to HIV-1 RNA binding, which is required for HIV-1 maturation. Non-catalytic
24 site integrase inhibitors (NCINIs) have been developed as allosteric IN inhibitors, which performs anti-
25 HIV-1 activity by a multimodal mode of action such as inhibition of IN-LEDGF/p75 interaction in the
26 early stage and disruption of functional IN multimerization in the late stage of HIV-1 replication. Here,
27 we show that IN undergoes an adaptable conformational change to escape from NCINIs. We observed
28 that NCINI-resistant HIV-1 variants have accumulated four amino acid (AA) mutations by passage 26
29 (P26) in the IN-encoding region. We employed HPLC, thermal stability assay, and X-ray
30 crystallographic analysis to show that some AA mutations affect the stability and/or dimerization
31 interface of the IN catalytic core domains (CCD), potentially resulting in severely decreased
32 multimerization of full-length IN proteins (IN under-multimerization). This under-multimerized IN via
33 the NCINI-related mutations was stabilized by HIV-1 RNA and restored to the same level as HIV-1
34 wild type in the viral particles. Recombinant HIV-1 clones with IN under-multimerization propagated
35 similarly as HIV-1 wild type. Our study revealed that HIV-1 can eventually countervail NCINI-induced
36 IN over-multimerization by IN under-multimerization as one of the escape mechanisms. Our findings
37 provide information on the understanding of IN multimerization with or without HIV-1 RNA and may
38 influence the development of anti-HIV-1 strategies.
2
39
40 Importance

41 Understanding the mechanism of HIV-1 resistance to anti-HIV-1 drugs would lead to the
42 development of novel drugs with increased efficiency, resulting in more effective ART. ART composed
43 of more potent and long-acting anti-HIV-1 drugs can greatly improve drug adherence and also provide
44 HIV-1 prevention such as pre-exposure prophylaxis. NCINIs with the multimodal mode of action exert
45 potent anti-HIV-1 effects through IN over-multimerization during HIV-1 maturation. However, HIV-1
46 can acquire some mutations which cause IN under-multimerization to alleviate NCINI-induced IN
47 over-multimerization. This under-multimerized IN was efficiently stabilized by HIV-1 RNA and
48 restored to the same level as HIV-1 wild type. Our findings revealed that HIV-1 eventually acquires
49 such conformational escape reaction to overcome the unique NCINI actions. The investigation into the
50 drug-resistant mutations associated with HIV-1 protein multimerization may facilitate the elucidation
51 of its molecular mechanism and functional multimerization, allowing us to develop more potent anti-
52 HIV-1 drugs and unique treatment strategies.
53
54 Introduction
55 Human immunodeficiency virus type 1 (HIV-1) infection has been a chronic infectious disease.
56 Patients diagnosed with HIV are treated with antiretroviral therapy (ART), which basically consists of
57 more than two anti-HIV-1 drugs. The life expectancy of HIV-1 patients with ART has been prolonged
3
58 to the almost same extent as uninfected individuals (1, 2). In spite of the benefits of ART, HIV-1 often
59 acquires drug-resistant mutations (3) resulting in treatment failure. High adherence to ART is required

60 to sustain viral suppression in HIV-1 clinical treatment (4). Recently, the FDA has approved certain
61 anti-HIV-1 drugs which can be easily taken once-daily in the form of a single tablet, greatly improving
62 ART adherence (5). Long-acting anti-HIV-1 drugs could be also useful in ART adherence and HIV-1
63 treatment. Moreover, HIV-1 prevention, pre-exposure prophylaxis (PrEP) by such long-acting anti-
64 HIV-1 drugs, is currently being investigated in clinical trials (6). On the other hand, studying the
65 mechanism of HIV-1 resistance to anti-HIV-1 drugs would lead to the development of novel drugs with
66 increased efficiency and a greater genetic barrier to resistance, resulting in more effective ART for
67 HIV-1 treatment.
68 Four FDA-approved integrase strand transfer inhibitors (INSTIs), raltegravir (RAL), elvitegravir
69 (ELG), dolutegravir (DTG), and bictegravir (BTG) which target the active site of IN are more potent
70 and well-tolerated than other classes of anti-HIV-1 drugs due to the lack of homologous human proteins,
71 allowing these INSTIs to have been widely used for clinical HIV-1 treatment. The World Health
72 Organization (WHO) recommends INSTIs, especially DTG, for clinical HIV-1 treatment according to
73 the December 2018 Interim guidelines. A single tablet, Juluca® containing only two anti-HIV-1 drugs
74 (DTG and rilpivirine (RPV)) has been approved by the FDA in 2018 as the first maintenance therapy
75 (7), and Dovato® (DTG and 3TC) has been approved in 2019 for daily first-line HIV-1 treatment in
76 naïve patients. Therefore, the development of more potent anti-HIV-1 inhibitor targets for IN would
4
77 be useful for HIV-1 treatment.
78 In 2010, Christ et al. identified another new class of integration inhibitors (8) which interfere with

79 interaction between IN and lens epithelium-derived growth factor p75 (LEDGF/p75), thereby
80 inhibiting HIV-1 replication. LEDGF/p75 plays an important role in tethering IN to the host genome
81 and stabilizing functional IN multimerization during the integration step (9, 10). They termed such
82 inhibitors LEDGINs. In addition, it has been reported that LEDGINs exert anti-viral effects through
83 modulating IN multimerization (IN over-multimerization) in the late stage of HIV-1 replication (11-
84 14). LEDGINs are also called NCINIs (16, 19), allosteric IN inhibitors (ALLINIs) (13, 15), Integrase-
85 LEDGF allosteric inhibitors (INLAI) (41), or Multimerization Integrase Inhibitors (MINI) (22). We
86 used NCINIs for convenience due to these inhibitors exerting the multimodal mechanism of action, as
87 well as BI224436 that was termed NCINIs (19) advanced into phase I clinical trial. The IN over-
88 multimerization by potent NCINIs lead to the production of non-infectious viruses in which the
89 ribonucleoprotein (RNP) is translocated from the capsid lattice during HIV-1 maturation (15-17). In
90 2016, Kessl et al. reported that such non-infectious HIV-1 resulted from the loss of the interaction
91 between over-multimerized IN and HIV-1 RNA during HIV-1 maturation (18).
92 Many resistance mutations of general HIV-1 strains including HIV-1NL43 against NCINIs have been
93 already reported (8, 12, 27, 31-32, 41-42). In the present study, we attempted to examine the
94 mechanism underlying the NCINI-related mutations which could provide details of IN multimerization.
95 The study of drug-resistant mutations concerning such allosteric inhibitors may help to elucidate how
5
96 HIV-1 proteins undergo functional multimerization, which may aid in the development of more potent
97 anti-HIV-1 drugs and unique treatment strategies.

98
99 Results
100 Emergence of NCINI-related resistance mutations
101 NCINIs with the multimodal mode of action suppress HIV-1 replication by inducing IN over-
102 multimerization (12-15, 17). We selected HIV-1NL4-3 variants resistant to three NCINIs compounds
103 known as NCINI-1 (LEDGIN6, CX04328, and CX05168), NCINI-2 (LEDGIN7 and CX05045), and
104 NCINI-3(CX14442) (Fig. 1A) reported previously, (8, 12) by selection experiments (Fig. 1B). As the
105 concentration of the NCINIs up to 10 µM gradually increased, HIV-1 acquired several AA mutations,
106 such as A128T and K173Q at passage 15 (P15) in the catalytic core domain (CCD), and N254K at
107 passage 20 (P20) in the C-terminal domain (CTD) of IN. Four mutations, A128T, H171Q, K173Q, and
108 N254K, accumulated during passage 26 (P26) in the full-length IN region upon treatment with NCINI-
109 3 at 10 µM. A128T and G272G mutations were acquired in the IN region upon treatment with NCINI-
110 1 at passage 18, and V77S and A128T mutations were acquired upon treatment with NCINI-2 at
111 passage 12 (Fig. 1C). EC50s of the drugs (NCINI-1, -2, -3, and -4 (BI224436) (19), RAL, and darunavir
112 (DRV)) against wild type HIV-1NL4-3 (Table 1) and recombinant HIV-1NL4-3 clones (HIV-1NL4-3-INA128T,
113 -INH171Q, -INK173Q, -INN254K, -INP15, -INP20, and -INP26, respectively), and the fold changes are shown
114 in Table 2. Accumulation of these mutations gradually gave rise to NCINI-resistant HIV-1. It appeared
6
115 that each K173Q and N254K mutation did not confer strong resistance to potent NCINIs (NCINI-3
116 and -4), while A128T alone and an addition of H171Q at P26 conferred substantial resistance (over

117 10-fold EC50s) to the potent NCINIs; however, none of the clones carrying NCINI-related resistance
118 mutations exhibited cross-resistance to clinically used drugs, such as RAL and DRV.
119
120 Characteristics of recombinant IN proteins carrying NCINI-related resistance mutations
121 To examine how NCINI-3 resistance mutations acquire resistance against NCINI-3, we purified
122 recombinant His-tagged IN proteins carrying mutations, such as A128T, H171Q, K173Q, N254K, P15,
123 P20, P26, and E11K, respectively (INA128T, INH171Q, INK173Q, INN254K, INP15, INP20, INP26, and INE11K)
124 expressed in E. coli and analyzed multimerization of these IN proteins. E11K substitution was known
125 to impair IN tetramerization (10). INWT multimerization appeared as two peaks of tetramers and a
126 mixture of dimers and monomers using Size exclusion chromatography (SEC) (Fig. 2A). INH171Q and
127 INN254K multimerization were very similar to that of INWT, suggesting that H171Q and N254K
128 mutations did not affect IN multimerization, whereas INA128T multimerization was seen as a broader
129 shoulder of the two peaks. INK173Q exhibited a moderately decreased tetramer peak. Further, INP15,
130 INP20, and INP26 proteins carrying the accumulated NCINI-3 resistance mutations completely shifted
131 to one peak consisting of dimers and monomers (Fig. 2A). Next, we examined proportions of
132 monomers, dimers, and tetramers of INWT and INP26 cross-linking with BS3 (Multimers ratio) using
133 an immunoblotting assay. Multimers ratio of INP26 (19.6 % dimers and 3.3 % tetramers) slightly
7
134 decreased compared with that of INWT (22.9 % dimers and 6.5 % tetramers) (Fig. 2B). In addition, to
135 quantitatively examine IN multimerization, we performed BiFC-IN system previously reported (20).

136 Results of the BiFC-IN system carrying the NCINI-3 resistance mutations almost corroborated the
137 SEC data (Fig 2C).
138 Next, to understand whether under-multimerized IN affects its interaction with LEDGF/p75, we
139 performed a pull-down assay between His-tagged LEDGF/p75 and Flag-tagged IN proteins carrying
140 P26 or V165A mutation reported as a Class II mutant which fails to interact with LEDGF/p75 (21). As
141 shown in Fig. 2D, IN-LEDGF/p75 binding ratio of INV165A which fails to interact with LEDGF/p75
142 was 9.7 %, whereas those of INWT and INP26 with IN under-multimerization were 39.7 and 50.6 %,
143 respectively, suggesting that INP26 with IN under-multimerization can interact with LEDGF/p75.
144 P26 mutation with the accumulated NCINI-3 resistance mutations severely reduced IN
145 multimerization and also did not affect IN-LEDGF interaction, suggesting that IN under-
146 multimerization may be associated with NCINI-3 related resistance of HIV-1.
147
148 Characteristics of recombinant HIV-1NL4-3 clones carrying NCINI-related resistance mutations
149 NCINIs induce IN over-multimerization, leading to the production of non-infectious viruses, but
150 are not related to Gag and Gag-Pol proteolytic processing (15, 22, 42). In order to investigate
151 characteristics of HIV-1 clones carrying the NCINI-3 resistance mutations with IN under-
152 multimerization, we examined the process of viral production including Gag and Gag-Pol proteolytic
8
153 processing, and viral maturation of the recombinant HIV-1 INA128T, -INH171Q, -INK173Q, -INN254K, -INP15,
154 -INP20, -INP26, and -INE11K clones. As shown in Fig. 2A, K173Q, P15, P20, and P26 mutations

155 significantly decreased IN multimerization; however, Gag and Gag-Pol proteolytic processing of the
156 recombinant HIV-1 IN clones with or without IN under-multimerization, analyzed by immunoblotting
157 in the presence of saquinavir (SQV, an HIV-1 protease inhibitor inhibits Gag-pol proteolytic
158 processing), was similar to that of HIV-1 INWT (Fig 3A). Viral production levels (p24 values) of wild
159 type HIV-1NL4-3 (HIV-1 INWT) and such recombinant HIV-1 IN clones with or without IN under-
160 multimerization, except HIV-1 INE11K, were also similar (Fig. 3B), indicating that the recombinant
161 HIV-1 clones with IN under-multimerization had normal Gag and Gag-Pol proteolytic processing, and
162 viral production.
163 Besides, morphologies of HIV-1 INWT and HIV-1 INP26 clone in the presence or absence of NCINI-
164 3 were observed using a transmission electron microscope (TEM). HIV-1 particles were classified by
165 viral structures under three phenotypes, normal mature conical core, abnormal (or empty) core, and
166 others core including immature and ambiguous (Fig. 3C) as previously reported (15, 17, 23). The
167 morphology of HIV-1 INWT included 79.4 % normal, 14.0 % abnormal, and 6.6 % others structures.
168 By the addition of NCINI-3, abnormal morphology of HIV-1 INWT was identified in 74.4 % which
169 indicated RNP translocation from the capsid lattice. On the other hand, the morphology of HIV-1 INP26
170 clone with IN under-multimerization was similar to that of HIV-1 INWT and exhibited resistance against
171 NCINI-3 (56.6 % normal, 32.8 % abnormal, and 10.6 % others structures). Finally, we examined
9
172 replication fitness of HIV-1 INWT and the recombinant HIV-1 IN clones. The HIV-1 IN clones with IN
173 under-multimerization infected in MT-4 cells showed robust replication fitness similar to that of HIV-

174 1 INWT up to day 7 except HIV-1 INE11K clone (p24, from 1031 to 1672 ng/ml) (Fig. 3D). Thus, IN
175 under-multimerization induced by the NCINI-3 resistance mutations in this study did not significantly
176 affect viral production including Gag and Gag-Pol processing, and viral maturation in the HIV-1 life
177 cycle.
178
179 Effect of certain mutations on conformation and stability of IN catalytic core domain
180 It has been reported that NCINIs bind to two pockets formed in CCD proteins (8), carrying a
181 solubilizing AA substitution, F185K (CCDF185K) (24). To examine whether the NCINI-3 resistance
182 mutations affect the structure of CCD, we produced and purified CCDF185K proteins carrying the
183 NCINI-3 resistance mutations (CCDA128T, CCDH171Q, CCDK173Q, CCDP15, and CCDP26 del) expressed in
184 E. coli. P26 del indicates P26 without N254K mutation in the CTD. It appeared that a CCDWT peak
185 eluted at roughly the same monomer size as CCD—17 KDa by SEC in our method condition (Fig. 4A).
186 CCDH171Q which did not reduce the full-length IN multimerization eluted at the same size fraction as
187 CCDWT. Notably, CCDA128T and CCDK173Q proteins, which changed the full-length IN multimerization,
188 eluted at a decreased size compared to CCDWT. CCDP15 and CCDP26 del eluted at further decreased sizes.
189 These results suggesting that A128T, K173Q, P15, and P26 del mutations affect the form of CCD
190 monomer (Fig. 4A).
10
191 In addition, to examine the thermal stability of CCDA128T, CCDH171Q, CCDK173Q, CCDP15, and CCDP26
del
192 , we performed differential scanning fluorimetry (DSF) (25). The melting temperature value for

193 proteins generally represents the temperature at which the protein is 50% folded (Tm 50). As shown
194 in Fig. 4B and C, Tm 50 of all CCD proteins did not indicate significant differences. On the other hand,
195 we especially noticed Tm 25 value, indicating the temperature at which CCD proteins are 25% folded
196 upon heating. Tm 25 of CCDA128T, CCDK173Q, CCDP15, and CCDP26 del proteins, except CCDH171Q,
197 decreased compared with that of CCDWT (Fig. 4C), suggesting that CCD proteins carrying these
198 mutations were unstable at lower temperature from 35 to 45°C compared with CCDWT. Additionally,
199 the surface hydrophobic sites of CCDA128T, CCDK173Q, CCDP15, and CCDP26 del at 37°C increased as
200 shown in Fig. 4D and E. It seems that A128T, K173Q, P15, and P26 del mutations affect the CCD
201 conformation in the region of body temperature. These results suggest that A128T, K173Q, P15, and
202 P26 del mutations except H171Q probably change the CCD conformation, may affecting the full-
203 length IN multimerization.
204 Furthermore, to investigate binding affinity of NCINI-3 to each CCDA128T, CCDH171Q, CCDK173Q,
205 and CCDP15 protein, we used Surface Plasmon Resonance (SPR) (Fig.4F). Binding affinity (KD value)
206 of NCINI-3 to CCDWT was 0.95 µM, while that of CCDA128T, CCDH171Q, and CCDK173Q decreased by
207 2.10 µM (2.2-folds), 1.35 µM (1.4-folds) and 1.41 µM (1.5-folds), respectively. Further, CCDP15 did
208 not interact under 10 µM NCINI-3 (Fig. 4F). The interactions of NCINI-3 to CCD proteins carrying
209 A128T, H171Q, or K173Q mutation decreased, while K173Q did not confer the significant resistance
11
210 against NCINI-3 (Table2). These results suggest that K173Q causing the mild IN under-
211 multimerization may have a different resistant profile from A128T and H171Q.

212
213 The P15 mutation changes NCINIs binding pockets formed in the CCD
214 It has been reported that NCINI-related compounds bind to CCD proteins containing F185K
215 substitution by crystallographic analysis (26, 8, 14). To confirm the CCD conformation carrying P15
216 (A128T/K173Q) which mostly lost the tetramer peak of full-length IN (Fig. 2A), we analyzed a crystal
217 structure of CCDP15 carrying F185K using X-ray crystallography.
218 As shown in Fig. 5A, the NCINI-related compound binds to the binding pocket in CCDWT (PDB:
219 4GW6) (27), while a phenolic side chain (Y99 position) in CCDP15 (PDB: 6L0C) shifts closer to the
220 NCINI-binding pocket compared with that in CCDWT (Fig. 5B and C), potentially interfering with
221 NCINI to CCDP15 binding. Additionally, A128T mutation also affects the positioning of the NCINI as
222 previously reported (27) (Fig. 5C). Interestingly, the phenolic side chain of Y99 residue in a CCDWT
223 monomer (CCD1WT) participates in intermolecular interactions with two hydrogen bonds (H-bonds)
224 to side chains of E87 and Q177, respectively, in the other CCDWT monomer (CCD2WT) (Fig. 5A).
225 However, the interaction with Y99 in CCD1P15 reduces one H-bond with the side chain of H171 in
226 CCD2P15; specifically, Q173 forms new intramolecular interactions with V88 and Q177 residues in
227 CCD2P15 (Fig. 5B). Such H-bonds alteration at the dimerization interface in CCDP15 might cause an
228 unstable CCDP15 dimer compared with CCDWT, affecting NCINI binding pockets in CCDP15.
12
229 P15 mutation reduces NCINIs binding space formed in CCDP15 dimers (Fig. 5C), and also might
230 disrupt the binding pocket by producing unstable CCDP15 dimers, shifting to CCDP15 monomers,

231 potentially resulting in the reduced full-length IN multimerization (IN under-multimerization).
232
233 INP26 under-multimerization is stabilized by HIV-1 RNA and restored in the viral particles
234 Over-multimerization of IN, directed by NCINIs, fails to interact with HIV-1 RNA, leading to
235 production of the non-infectious HIV-1 particles with the abnormal core (Fig. 3C) (18). To examine
236 whether IN under-multimerization affects direct interaction with HIV-1 RNA, we analyzed the
237 interaction between purified Hisx6-tagged INP26 proteins and biotinylated transactivation response
238 (TAR) element, 54 RNA nucleotides, using Alpha assay which is a bead-based non-radioactive
239 amplified luminescent proximity homogeneous assay (Alpha assay 1) (18, 28, 29). In this assay, we
240 employed recombinant INRAKA proteins carrying R269A/K273A mutation, with reported normal IN
241 multimerization but the loss of interaction with HIV-1 RNA (18). The Alpha assay signals of each INWT
242 and INP26 at 10 nM were sufficiently increased by the addition of TAR-RNA concentration-dependent
243 up to 10 nM, fitting to saturation curves, and then decreased because of hooking effects (Fig. 6A).
244 Interestingly, binding affinity (KD value) of INRAKA in Alpha assay 1 was ambiguous, whereas that of
245 INP26 was 1.76 nM (95% confidence interval [CI], 1.22-2.30) (0.84-folds) slightly lower than that of
246 INWT at 2.06 nM (95% CI, 1.42-2.70) (Fig. 6A), suggesting that the binding ability of INP26 to TAR-
247 RNA (HIV-1 RNA) may be slightly reinforced.
13
248 Next, Alpha assay 2 was performed to confirm the effect of HIV-1 RNA on IN multimerization
249 (Fig. 6B). The Alpha assay 2 signal ratios of INWT and INP26 were determined until 1000 nM of TAR-

250 RNA. The Alpha 2 signal ratios of INWT and INP26, increased by TAR-RNA concentration-dependent
251 up to 100 nM except for INRAKA, and then showed plateau without hooking effect, fitting to sigmoid
252 curves (Fig. 6B). Hillslopes of INWT and INP26 except INRAKA were almost similar, while Log EC50 of
253 INP26 was 7.79 nM (95% CI, 2.94-12.64) (0.78-folds) slightly lower than that of INWT at 10.00 nM
254 (95% CI, 7.23-12.80) (Fig. 6B). These results suggest that INP26 multimerization may be supported
255 efficiently by HIV-1 RNA compared to INWT.
256 Finally, in order to investigate the effect of NCINIs on INWT and INP26 multimerization during HIV-
257 1 maturation, we analyzed IN multimerization of HIV-1-INWT and -INP26 clones produced in the
258 absence or presence of 20 µM NCINI-3 using cross-linking with BS3 immunoblotting assay. IN dimers
259 and tetramers obviously increased, and high-order multimers clearly appeared in HIV-1-INWT which
260 produced in the presence of NCINI-3, while IN dimers and tetramers of HIV-1-INP26 produced in the
261 absence and presence of NCINI-3 were shown to be similar (Fig. 6C), indicating that P26 mutation
262 suppressed NCINI-3-induced IN over-multimerization during HIV-1 maturation and also that INP26
263 under-multimerization was restored to the same level as INWT multimerization in the viral particle.
264 Taken together, IN under-multimerized proteins via the accumulated NCINI-related resistance
265 mutations were stabilized by HIV-1 RNA and eventually restored in HIV-1 viral particles during HIV-
266 1 maturation. These results reveal that IN under-multimerization via NCINI-resistant mutations may
14
267 be one of the HIV-1 escape mechanisms, allowing NCINI-resistant HIV-1 to propagate within the cells
268 (Fig. 7).

269
270 Discussion
271 In the present study, we found that IN undergoes an adaptable conformational alteration to escape
272 NCINI activity. It has been reported that CCD dimers formed at the central core are material component
273 of full-length IN tetramers with HIV-1 nucleotides (10, 30). However, without HIV-1 nucleotides, the
274 P15 (A128T/K173Q) mutation changed the conformation of CCD monomers, might causing unstable
275 CCD dimers, and also affecting the form of NCINI binding sites. Such conformational changes could
276 induce severe reductions of full-length IN multimerization before IN binding to HIV-1 RNA. INP26
277 under-multimerized proteins carrying A128T/K173Q was efficiently recovered by HIV-1 TAR-RNA
278 and were restored to the same level as INWT multimerization in the viral particles. The morphology
279 and replication fitness of HIV-1-INP26 clone was similar to HIV-1 wild type. These results suggest that
280 IN under-multimerized proteins carrying A128T/K173Q still retained intrinsic IN function.
281 On the other hand, NCINIs exert the multimodal mode of action including anti-IN-LEDGF
282 interaction in early stage of HIV-1 replication. A128T mutation was firstly identified in HIV-1 mutants
283 which were selected using cell lines that stably expressed a C-terminal fragment of LEDGF/p75 (43),
284 indicating that A128T is related to IN-LEDGF interaction in the early stage of HIV-1 replication. In
285 this study, we did not investigate the roles of IN under-multimerization by A128T/K173Q mutation in
15
286 the early stage, which requires further experimental data to fully understand the NCINI-related
287 resistance mutations of HIV-1 against the multimodal mode of action. Additionally, A128T mutation

288 has been reported as a typical and critical NCINI-related resistance mutation (8, 11-12, 15, 22, 27, 41).
289 We observed that A128T emerged early at P15 passage together with K173Q in the presence of NCINI-
290 3, and at last P17 and P12 of NCINI-1 and -2, respectively. As shown in Table2, HIV-1 INA128T
291 exhibited strong resistance to NCINI-3 and -4, suggesting that A128T represents a critical mutation
292 against the NCINI action.
293 HIV-1 INK173Q clone carrying K173Q, which moderately reduces IN multimerization, showed
294 resistance (over 10 µM) against weak NCINIs (NCINI-1 and -2), but no observed resistance against
295 potent NCINIs (NCINI-3 and -4). By the addition of A128T, HIV-1 INP15 clones acquired severe
296 resistance to the potent NCINIs (Table2). Collectively, K173Q moderately decreased IN
297 multimerization (Fig.2A) and reduced direct binding of NCINI-3 (Fig.4F), but, EC50 of HIV-1 INK173Q
298 clone did not indicate critical resistance compared with that of HIV-1 INA128T clone, suggesting that
299 K173Q plays a secondary role among the NCINI-related resistance mutations. H171Q mutation has
300 been reported as NCINI-related resistance mutations (11, 41). In this study, H171Q mutation, which
301 did not affect IN multimerization, was acquired at P26 together with A128T, K173Q, and N254K, and
302 conferred sufficient resistance against the potent NCINIs (Table2). The characteristics of H171Q were
303 similar to H171T previously reported (31) at the same AA position. Moreover, we constructed and
304 examined HIV-1 INA128T/H171Q clone carrying a double and potent NCINI-resistance mutation.
16
305 Replication fitness of HIV-1 INA128T/H171Q clone was severely lower than that of HIV-1 wild type (data
306 not shown), suggesting that HIV-1 INA128T/H171Q clone may require secondary mutations such as K173Q

307 and N254K to propagate within the cells.
308 Y99H mutation was also reported (12, 41) as an important resistant mutation of NCINI-related
309 mutations. Y99H was not acquired by our selection assay in this study. Interestingly, according to the
310 crystal structure of CCDP15 (Fig.5), the side chain of Q173 in one CCDP15 pushed that of Y99 in the
311 other CCDP15 to NCINI binding pocket, suggesting that K173Q might be a substitute mutation for
312 Y99H. T124N/T174I resistance mutation against a pyridine-based inhibitor, KF116 (MINI), was also
313 reported to disrupt IN multimerization (32). T124N/T174I decreased Gag-Pol proteolytic processing,
314 viral maturation, and viral infectivity of the HIV-1 INT124N/T174I variant compared with those in HIV-1
315 wild type. It seems that T124N/T174I mutation, which induced abnormal IN multimerization, confers
316 a similar phenotype to class II IN mutant HIV-1 (33); however, A128T/K173Q showed a clearly
317 different phenotype from the class II IN mutant. Moreover, it was reported that the CTD of IN plays a
318 critical role of IN over-multimerization (45), and NCINIs can interact both CCD and CTD which are
319 contributed by two different dimers so that IN forms an open polymer mediated by inhibitor-bridged
320 contacts in the crystal lattice (46). Although N254K did not confer sufficient resistance against potent
321 NCINIs, N254K is possibly associated with inhibition of such IN over-multimerization mechanism.
322 In the viral particles, HIV-1 nucleocapsid (NC) binds to HIV-1 RNA, while IN also binds to HIV-
323 1 RNA at different positions from NC (18, 23). It has yet to be shown how IN bindings to HIV-1 RNA
17
324 maintain RNP in the capsid lattice during HIV-1 maturation. NCINIs are not yet approved by the FDA
325 and NCINI-related resistance mutations in vivo (clinical use) are unknown, which presents the major

326 limitations of our study.
327 Our findings reveal that HIV-1 can countervail NCINI-inducing IN over-multimerization by IN
328 under-multimerization as one of the escape mechanisms, may allowing the NCINI-resistant HIV-1 to
329 propagate within the cells. The investigation into the drug-resistant mutations associated with HIV-1
330 protein multimerization may facilitate the elucidation of its molecular mechanisms, allowing the
331 development of more potent anti-HIV-1 drugs and unique treatment strategies.
332
333 Materials and Methods
334 Cells
335 293T (JCRB cell bank, Japan) and HEK293T (ATCC, CRL-11268) cells were maintained in
336 Dulbecco’s Modified Eagle’s Medium (DMEM; WAKO, Japan) supplemented with 10% fetal bovine
337 serum (FBS; SIGMA-ALDRICH), penicillin (P+) (100 I.U/mL), and kanamycin (K+) (100 mg/mL)
338 (MEIJI, Japan) at 37°C and 5% CO2. MT-2 and MT-4 cells (JCRB) were in RPMI-1640 medium
339 (WAKO, Japan) supplemented with 10% FBS, P+, and K+ at 37°C and 5% CO2.
340
341 Plasmids constructs
342 Full-length IN and CCD sequences derived from pNL4-3 were introduced to pET30a vectors (Novagen)
18
343 with Hisx6 tagged at the C-terminus or N-terminus, respectively, producing pET30a IN-6xHis (INHis)
344 or pET30a 6xHis-CCD (CCDHis) by using In-Fusion® HD Cloning Kit (TAKARA Bio Inc., Japan).

345 Flag-tagged IN (F-IN) was also introduced to pET30a vector at the N-terminus by the In-Fusion
346 method, producing pET30a Flag-IN. LEDGF/p75 sequence derived from 293T cells was introduced to
347 pET30a vectors (Novagen) with Hisx6 tagged at the N-terminus, producing pET30a His-LEDGF/p75
348 (LEDGF/p75His). The site-directed mutagenesis of single mutations was performed using
349 PrimeSTAR® Max (TAKARA) to introduce A128T, H171Q, K173Q, N254K, E11K, V165A, and
350 R269A/K273A into the pET30a INHis, producing pET30a INA128T, INH171Q, INK173Q, INN254K, INE11K,
351 INV165A, and INRAKA vectors, respectively. The NCINI-3 resistance mutations in the CCCD were
352 introduced into pET30a CCDHis with F185K, producing pET30a CCDA128T, CCDH171Q, and CCDK173Q
353 vectors. The plural resistance mutations such as P15 (A128T/K173Q), P20 (A128T/K173Q/N254K),
354 and P26 (A128T/H171Q/K173Q/N254K) were introduced to pET30a INHis or CCDHis vector by In-
355 Fusion method, producing pET30a INP15, INP20, INP26, and pET30a CCDP15 with F185K, respectively.
356 These NCINI-3 resistance mutations were also introduced to pNL4-3 and pBiFC-IN vectors (20) by
357 above the method.
358
359 Proteins expression and purification
360 IN and CCD proteins were produced from the pET30a INHis and CCDHis, respectively in E. coli
361 RosettaTM (DE3) pLysS Competent Cells (Novagen) grown in LB medium supplemented with K+ and
19
362 chloramphenicol (Cam+) at 37°C and induced with 1.0 mM Isopropyl β-D-1-thiogalactopyranoside
363 (IPTG) for 4 h at 30°C. The bacterial cells were harvested and stored at -80°C. The pellets of INHis

364 were resuspended and sonicated in lysis buffer (50 mM HEPES-NaOH pH 7.4, 100 mM NaCl). The
365 lysates were cleared by centrifugation for 15 min at 3,500 rpm at 4°C. The supernatants were removed,
366 and the pellets of IN were resuspended in IN reservoir buffer (50 mM HEPES pH 7.4, 1 M NaCl, 7.5
367 mM CHAPS, 10 mM MgSO4, 10% glycerol) supplemented with 5 mM 2-mercaptoethanol (BME), 10
368 mM imidazole and 0.5 mM Phenylmethylsulfonyl fluoride (PMSF). The pellets of CCD were
369 resuspended and sonicated in CCD reservoir buffer (50 mM HEPES pH 7.4, 500 mM NaCl, 7.5 mM
370 CHAPS, 10 mM MgSO4, 5% glycerol) supplemented with 5 mM BME, 10 mM imidazole, and 0.5
371 mM PMSF. The lysates were precleared for 30 min at 15,000 rpm at 4°C, and filtered through a 0.45
372 μm filter. IN and CCD proteins were loaded onto a His-TALON column (TAKARA), and the columns
373 were washed with IN wash buffer (20 mM HEPES pH 7.4, 1 M NaCl, 3.75 mM CHAPS) or CCD
374 wash buffer (20 mM HEPES pH 7.4, 500mM NaCl, 3.75 mM CHAPS). IN or CCD proteins were
375 eluted with IN or CCD reservoir buffer, respectively supplemented with 300 mM imidazole using
376 AKTAprime plus (GE Healthcare). IN or CCD fractions were concentrated using Amicon Ultra-30K
377 or -10K device (Merck Millipore) in each reservoir buffer. LEDGF/p75 was produced in E. coli
378 RosettaTM (DE3) pLysS competent cells grown in LB medium supplemented with 2% ethanol, K+,
379 and Cam+ at 37°C and induced with 0.5 mM IPTG for 5 h at 30°C. The bacterial cells were harvested
380 and sonicated in LEDGF reservoir buffer (50 mM Tris-HCl pH 7.4, 500 mM NaCl, 5 mM BME)
20
381 supplemented with 10 mM imidazole and 0.5 mM PMSF. The lysates were precleared for 30 min at
382 15,000 rpm at 4°C, and filtered. LEDGF/p75 was loaded onto the His-TALON column and the column

383 was washed with LEDGF wash buffer (20 mM Tris-HCl pH 7.4, 500 mM NaCl). LEDGF/p75 was
384 eluted with LEDGF reservoir buffer supplemented with 300 mM imidazole. LEDGF/p75 was
385 concentrated using an Amicon Ultra-50K device in LEDGF reservoir buffer. F-IN was expressed as
386 above the method and purified by previous reports (13). The protein concentration was determined
387 using a BCA Protein Assay Reagent Kit (Thermo Fisher Scientific).
388
389 Recombinant HIV-1 clones
390 293T cells (1.5 × 105/ml) were seeded on TC 6-well plate (Greiner Bio-one) and incubated for 24 h at
391 37°C. Cells were transfected with 5 μg of pNL4-3 encoding NCINI-3 resistance mutations (A128T,
392 H171Q, K173Q, N254K, P15, P20, and P26), or E11K in the IN region using Attractene Transfection
393 Reagent (QIAGEN). Cells were washed at 24 h later, and cell supernatants containing viruses were
394 collected after 48 h incubation. The HIV-1 clones were measured by HIV-1 p24 ELISA
395 (LUMIPULSE® G1200; FUJIREBIO Inc., Japan) and normalized to determine the viral concentration,
396 and stored at -80°C.
397
398 Western blotting
399 293T cells were plated on TC 6-well plates (1.5 × 105/ml) and incubated for 24 h at 37◦C in 5% CO2.
21
400 Cells were co-transfected with BiFC-IN vectors carrying NCINI-3 resistance mutations using
401 Attractene and incubated for 48 h. Subsequently, the cells were lysed in lysis buffer (20 mM Pipes pH

402 7.0, 400 mM NaCl, 10% (w/v) sucrose, 0.5 mM dithiothreitol (DTT), 0.5% NP-40, Halt Protease
403 Inhibitor Cocktail (Thermo Fisher Scientific) (9). After centrifugation for 30 min at 15,000 rpm at 4°C,
404 the samples were titrated using BCA Protein Assay Kit and stored at -80°C. Wild type and recombinant
405 HIV-1 clones were filtered and purified by ultracentrifugation (at 35,000 rpm for 30 min) in 15%
406 sucrose PBS, and normalized by the p24 levels and stored in PBS supplemented with Halt Protease
407 Inhibitor Cocktail at -80°C. The samples were prepared in NuPAGE® LDS and reducing sample buffer
408 (Thermo Fisher Scientific), and separated with SDS-PAGE (4–12% Bis-Tris Gel; Thermo Fisher
409 Scientific) and transferred a nitrocellulose membrane. The samples were detected with anti-HIV-1 IN
410 antibody (abcam; ab66645), anti-HIV-1 Gag (p55 + p24 + p17) antibody (abcam; ab63917), anti-HIV1
411 p24 [39/5.4A] (abcam; ab9071), and second mouse or rabbit antibody (MBL co., LTD.), and anti-beta
412 actin antibody (HRP conjugated) (abcam), and then visualized using SuperSignal WestPico
413 Chemiluminescent Substrates (Thermo Fisher Scientific).
414
415 Cross-linking IN proteins with BS3
416 Purified IN proteins were diluted at 300 nM final concentration in buffer (20mM HEPES pH 7.4, 100
417 mM NaCl, 2 mM DTT, 5 mM MgCl2, and 15 µM ZnCl2). Samples were mixed with BS3 (Thermo
418 Scientific Pierce) at 0.05 mM final concentration and incubated for 20 min at RT, and the reactions
22
419 were quenched by the addition of mixed LDS and reducing sample buffer. The cross-linked proteins
420 with BS3 were visualized by immunoblotting. In the case of IN in viral particles (16), the high

421 concentrated wild type and recombinant HIV-1 clones were titrated, and treated with 0.5% Triton X
422 in PBS, and cross-linked with 0.05 mM BS3. The samples employed at 100 ng (p24 level) per wells
423 were visualized by immunoblotting with anti-IN antibody and SDS PAGE (4–12% Bis-Tris Gel).
424
425 Generation of drug-resistant HIV-1 variants
426 Selection experiments of resistant HIV-1 variants were performed as previously described (34). In brief,
427 MT-4 cells (1.0 × 105/ml) were exposed to HIV-1NL4-3 (100 TCID50) and cultured in the presence of
428 NCINIs, each at an initial concentration of each EC50. Viral replication in MT-4 cells was monitored
429 by p24 levels of the culture medium at intervals of one week. The selection procedure was continued
430 until the compound concentration over 10 µM. Proviral DNA sequences of resistant HIV-1 variants
431 were extracted from the infected into MT4 cells using NucleoSpin® Tissue (TAKARA). DNA
432 sequences of the IN regions were amplified with 1st primers (5’-AAA TTT AAA TTA CCC ATA CAA
433 AAG GAA ACA TGG GAA GC-3’ and 5’-GGT CTG CTA GGT CAG GGT CTA CTT GTG TGC
434 TAT ATC-3’) and following 2nd nested primers (5’-AAA AGG AAA AAG TCT ACC TGG CAT GGG
435 TAC CAG CAC AC-3’ and 5’-AGT CCT TAG CTT TCC TTG AAA TAT ACA TAT GG-3’), and then
436 analyzed using IN sequence primers (5’-AAA AGT TAT CTT GGT AGC AGT TCA TGT AGC CAG
437 TGG-3’ and 5’-TT TAG TTT GTA TGT CTG TTG CTA TTA TGT CTA CTA TTC-3’).
23
438
439 Biomolecular fluorescence complementation (BiFC) analysis

440 BiFC-IN system was performed as previously described (20). This assay was modified as follows. In
441 brief, HEK293T cells were seeded (2 × 104/well) on Collagen I 96-well microplates (BD BioCoatTM)
442 and incubated for 24 h at 37◦C in 5% CO2. The cells were co-transfected with pBiFC-IN vectors (VN-
443 IN and VC-IN) carrying the NCINI-3 resistance mutations using Attractene and incubated for 48 h at
444 37◦C. Then Venus fluorescence in the cells was measured using a FACS (FACSVerse flow cytometer,
445 BD Bioscience). The area under the curve (AUC) of the histogram emitting from BiFC-IN was
446 measured using ImageJ. Data were compared as AUC ratios to BiFC-INWT.
447
448 IN and LEDGF/p75 binding assay
449 Pull-down assay was performed to confirm the interaction of F-IN with LEDGF/p75His (17, 32). Briefly,
450 IN (2 μM) and LEDGF/p75 (2 μM) in binding buffer (50 mM HEPES pH 7.0, 300 mM NaCl, 2 mM
451 MgCl2, 2 mM BME, 5% Glycerol, and 0.2% (v/v) Nonidet P-40) supplemented with 20 mM Imidazole
452 were incubated for 30 min at RT and loaded on His 60 Ni resin (TAKARA) after equilibrating. The
453 reaction mixture was incubated for 30 min at RT, and then were repeated washing by wash buffer (50
454 mM HEPES pH 7.0, 1M NaCl, 5 mM BME, and 5% Glycerol). The samples were eluted by the binding
455 buffer supplemented with 300 mM imidazole and visualized by Coomassie Brilliant Blue staining.
456
24
457 HPLC
458 Recombinant INHis proteins were examined with a TSKgel-SWXLG3000 column (TOSOH cor., Japan)

459 at 1.0 ml/min in running buffer (20 mM HEPES pH 7.4, 750 mM NaCl, 3.75 mM CHAPS) (10). IN
460 proteins (50 μM) in sample buffer (50 mM HEPES pH 7.4, 1 M NaCl, 7.5 mM CHAPS, 10 mM MgSO4,
461 50 µM ZnCl2, 2 mM DTT, 10% Glycerol) were subjected to Agilent Technologies 1220 Infinity LC
462 system (TOSOH) for SEC, and were detected by OD at 280 nm. Recombinant CCDHis proteins were
463 examined with a COSMOSIL 5Diol-120-II column (nakarai tesque, Japan) at 1.0 ml/min in running
464 buffer (20 mM HEPES pH 7.4, 500 mM NaCl, 3.75 mM CHAPS). CCD proteins (50 μM) in sample
465 buffer (50 mM HEPES pH 7.4, 500 mM NaCl, 7.5 mM CHAPS, 10 mM MgSO4, 5 mM BME, 5%
466 Glycerol) were subjected for SEC. The column was calibrated with standard proteins containing
467 Bovine gamma-globulin (158 KDa), Chicken ovalbumin (44 KDa), and Horse myoglobin (17 KDa).
468 All the procedures were controlled at 4°C as possible.
469
470 Crystallization, X-ray data collection and structure determination
471 The crystallization procedure was referred to a previous report (14). Briefly, crystallization was
472 performed by the hanging drop vapor diffusion method using EasyXtal 15-Well Tools (QIAGEN).
473 CCDP15 proteins carrying A128T, K173Q, and F185K mutations were expressed and purified as
474 described above. The solution buffer was changed to 50 mM MES-NaOH pH 5.5, 50 mM NaCl, and
475 5 mM DTT using an Amicon Ultra-10K device (Millipore). The protein concentration was adjusted at
25
476 2 mg/mL. The reservoir solution consists of 50 mM sodium cacodylate-HCl pH 6.5 and 1.2 M
477 ammonium sulfate. The crystals reached 0.2–0.4 mm within 1 week at 16°C. The crystals were

478 transferred to a reservoir solution supplemented with 25% glycerol and flash-frozen at 100K. The X-
479 ray diffraction experiments were carried out on beamlines of Photon Factory (Tsukuba, Japan) and of
480 SPring-8 (Hyogo, Japan). The diffraction data used for the following structure determination were
481 collected on BL17A of Photon Factory, and were processed with HKL2000 (35). The phases were
482 determined by molecular replacement with MOLREP (36) using coordinates of CCDWT in complex
483 with the NCINI-related compound (PDB: 4GW6) as a search model (27). Structural refinements were
484 carried out using PHENIX (37) and COOT (38). Molecular graphics were prepared using PyMOL
485 version 2.0 (Schrödinger, LLC). Atomic coordinates and structure factors were deposited in the Protein
486 Data Bank under the accession code 6L0C. Data collection and refinement statistics of CCDP15 (PDB:
487 6L0C) were listed in Fig. 5D.
488
489 DSF
490 Recombinant CCD proteins (12.5, 25, and 50 µM) carrying A128T, H171Q, K173Q, and P15,
491 respectively, together with F185K were prepared in sample buffer (50 mM HEPES pH 7.4, 500 mM
492 NaCl, 7.5 mM CHAPS, 10 mM MgSO4, 5 mM BME, 5% Glycerol). SYPRO Orange (Life
493 Technologies) was added to the samples (final concentration of SYPRO orange: 5×) (25). The samples
494 were successively heated from 25 to 95°C, and the increasing fluorescence intensity was detected by
26
495 the real-time PCR system 7500 Fast (Applied Biosystems). Data was indicated as a relative ratio
496 between minimum and maximum intensity of SYPRO orange from 25 to 95°C detected at each sample.

497
498 SPR
499 SPR was examined using a Biacore X100 (GE Healthcare) at RT. Recombinant CCD proteins carrying
500 A128T, H171Q, K173Q, and P15(A128T/K173Q) mutations, respectively, together with F185K at 20
501 μg/mL diluted in HBS-EP buffer (GE Healthcare) were immobilized on two flow-cells of a CM5 sensor
502 chip by direct amine coupling to 3,000 response units (RU). NCINI-3 diluted in HBS-EP buffer was
503 flowed with a 2 min injection at 30 μL/min followed by a 2 min dissociation. KD values of NCINI-3
504 to the CCD proteins were calculated from the sensorgrams using the Biacore X100 control software.
505
506 TEM
507 Viruses were produced by transfection with pNL4-3 INWT or INP26 into 293T cells. The cells were
508 washed at 12 h later, added fresh medium with or without 20 μM of NCINI-3. After 48 h incubation,
509 the samples were fixed in a solution of 4% formaldehyde and 4% glutaraldehyde in 0.1 M phosphate
510 pH 7.4 buffer for 1 h at 4°C. After centrifugation, the supernatants were filtered through a 0.45 μm
511 filter and then performed by ultracentrifugation using KUBOTA model 7780 (KUBOTA, Japan) at
512 22,000 rpm for 5 h at 4ºC. Virus pellets were resolved in minuscule quantity of 2% glutaraldehyde in
513 0.1 M phosphate pH 7.4 buffer and shipped to Tokai Electron Microscopy, Inc. Japan. Digital imaged
27
514 (3296x2472 pixels) were taken with a CCD camera (EM-14830RUBY2; JEOL Ltd., Tokyo, Japan).
515 The virus images over 100 per the samples were classified under three types by the locations of RNP

516 and the capsid lattice in the viral particles as previous reports (15, 17).
517
518 Alpha assay
519 Alpha assay 1 (18, 28) was used the recombinant INHis proteins (10 nM final concentration) which
520 were incubated with various concentration of synthetic biotinylated TAR-RNA (biotin 5’-dT/GGU
521 CUC UCU GGU UAG ACC AGA UCU GAG CCU GGG AGC UCU CUG GCU AAC UAG GGA
522 ACC/3’-biotin dT) (Integrated DNA Technologies) at 4°C for 3 h in the AlphaLISA buffer
523 (PerkinElmer). Anti-His Donor and Streptavidin Acceptor AlphaLISA beads (PerkinElmer) were
524 added at the final concentration of 10 mg/mL to each well and incubated on RT for 3 h. The Alpha
525 signal was detected by an EnSpire Plate Reader (PerkinElmer) with an AlphaScreen module setting.
526 KD values were determined by saturation binding curves using one site-specific binding analysis of
527 Graphpad Prism® (39). Alpha assay 2 was used the recombinant INHis and F-IN proteins which were
528 incubated with various concentrations of synthetic TAR-RNA (Integrated DNA Technologies) at 4°C
529 for 3 h in the AlphaLISA buffer. Anti-His Donor and anti-Flag Acceptor AlphaLISA beads were added
530 at the final concentration of 10 mg/mL to each well and incubated on RT for 3 h, and then the Alpha
531 signal was also detected by the same above method. Hillslope and EC50 were determined using
532 sigmoidal x is log analysis of Graphpad Prism®
28
533
534 HIV-1 replication fitness

535 Wild type or recombinant HIV-1NL4-3 clones carrying E11K or NCINI-3 resistance mutations at 30
536 ng/ml of p24 were exposed to MT-4 cells (2.4 × 105/ml) in TC 6-well plates for 3 h, and washed the
537 cells, and divided into three fractions in fresh medium, and each cultured p24 levels were measured by
538 HIV-1 p24 ELISA at 1, 3, 5, 7 days (34).
539
540 Drug Susceptibility Assay
541 Evaluation for EC50 and CC50 of drugs and test compounds was previously reported (34).
542
543 Compounds
544 NCINI-1, -2, -3, and DRV were synthesized using the published synthetic methods (8, 12, 40). RAL,
545 EVG, and DTG were purchased from Selleck Chemicals and NCINI-4 (BI224436) from Med Chem
546 Express LLC.
547
548 Data Availability
549 Atomic coordinates and structure factors of CCDP15 were deposited in the Protein Data Bank
550 (http://www.rcsb.org) under the accession code 6L0C.
551
29
552 Acknowledgments
553 We thank Haruo Aikawa and Hirokazu Tamamura for NCINIs synthesis, Sachiko Otsu for technical

554 assistance with the experiments, and Tokai Electron Microscopy Inc. staff for technical advice on the
555 TEM sample preparation and the data collection. We also thank the beamline staff at Photon Factory
556 and SPring-8 for their help with the data collection.
557 This work was supported by Japan Society for the Promotion of Science grants JP16K15521,
558 JP18K16180, JP18K08436, and Development of Novel Drugs for Treating HIV-1 Infection and AIDS.
559 TN designed and performed almost all the experiments. TN and YY performed X-ray
560 crystallography. TM and MM discussed the data and supported the preparation of the MS. HN
561 supervised, managed the project, and acquired the necessary funding. TN and HN wrote and edited the
562 MS. All authors read, commented on, and approved the final manuscript.
563 We declare that we have no conflicts of interest.
564
565 References
566 1. May MT, Gompels M, Delpech V, Porter K, Orkin C, Kegg S, Hay P, Johnson M, Palfreeman A,
567 Gilson R, Chadwick D, Martin F, Hill T, Walsh J, Post F, Fisher M, Ainsworth J, Jose S, Leen C, Nelson
568 M, Anderson J, Sabin C. 2014. Impact on life expectancy of HIV-1 positive individuals of CD4+ cell
569 count and viral load response to antiretroviral therapy. AIDS 28:1193–202.
570 https://doi.org/10.1097/QAD.0000000000000243.
30
571
572 2. Trickey A, May MT, Vehreschild JJ, Obel N, Gill MJ, Crane HM, Boesecke C, Patterson S, Grabar

573 S, Cazanave C, Cavassini M, Shepherd L, Monforte AD, van Sighem A, Saag M, Lampe F, Hernando
574 V, Montero M, Zangerle R, Justice AC, Sterling T, Ingle SM, Sterne JAC. 2017. Antiretroviral Therapy
575 Cohort Collaboration, Survival of HIV-positive patients starting antiretroviral therapy between 1996
576 and 2013: a collaborative analysis of cohort studies, Lancet HIV., 4 (8) e349-e356.
577 https://doi.org/10.1016/S2352-3018(17)30066-8.
578
579 3. Günthard HF, Calvez V, Paredes R, Pillay D, Shafer RW, Wensing AM, Jacobsen DM, Richman DD.
580 2019. Human Immunodeficiency Virus Drug Resistance: 2018 Recommendations of the International
581 Antiviral Society-USA Panel. Clin. Infect. Dis. 68:177–187. https://doi.org/10.1093/cid/ciy463.
582
583 4.Kanters S, Park JJ, Chan K, Socias ME, Ford N, Forrest JI, Thorlund K, Nachega JB, Mills EJ. 2017.
584 Interventions to improve adherence to antiretroviral therapy: a systematic review and network meta-
585 analysis. Lancet HIV 4:e31–e40. https://doi.org/10.1016/S2352-3018(16)30206-5.
586
587 5. Flexner C, Tierney C, Gross R, Andrade A, Lalama C, Eshleman S, Aberg J, Sanne I, Parsons T,
588 Kashuba A, Rosenkranz S, Kmack A, Ferguson E, Dehlinger M, Mildvan D. 2010. Comparison of
589 Once-Daily versus Twice-Daily Combination Antiretroviral Therapy in Treatment-Naive Patients:
31
590 Results of AIDS Clinical Trials Group (ACTG) A5073, a 48-Week Randomized Controlled Trial. Clin
591 Infect Dis 50:1041–1052. https://doi.org/10.1086/651118.

592
593 6. Gulick RM, Flexner C. 2019. Long-Acting HIV Drugs for Treatment and Prevention. Annu Rev
594 Med 70:137–150. https://doi.org/10.1146/annurev-med-041217-013717.
595
596 7. Dowers E, Zamora F, Barakat LA, Ogbuagu O. 2018. Dolutegravir/rilpivirine for the treatment of
597 HIV-1 infection. HIV AIDS (Auckl) 10:215–224. https://doi.org/10.2147/HIV.S157855.
598
599 8. Christ F, Voet A, Marchand A, Nicolet S, Desimmie BA, Marchand D, Bardiot D, der Veken N,
600 Remoortel B, Strelkov SV, Maeyer M, Chaltin P, Debyser Z. 2010. Rational design of small-molecule
601 inhibitors of the LEDGF/p75-integrase interaction and HIV replication. Nature Chemical Biology
602 6:442–448. https://doi.org/10.1038/nchembio.370.
603
604 9. Cherepanov P, Maertens G, Proost P, Devreese B, Van Beeumen J, Engelborghs Y, De Clercq E,
605 Debyser Z. 2003. HIV-1 integrase forms stable tetramers and associates with LEDGF/p75 protein in
606 human cells. J Biol Chem 278:372–81. https://doi.org/10.1074/jbc.M209278200.
607
608 10. Hare S, Nunzio DF, Labeja A, Wang J, Engelman A, Cherepanov P. 2009. Structural basis for
32
609 functional tetramerization of lentiviral integrase. PLoS Pathog 5:e1000515.
610 https://doi.org/10.1371/journal.ppat.1000515.

611
612 11. Tsiang M, Jones GS, Niedziela-Majka A, Kan E, Lansdon EB, Huang W, Hung M, Samuel D,
613 Novikov N, Xu Y, Mitchell M, Guo H, Babaoglu K, Liu X, Geleziunas R, Sakowicz R. 2012. New
614 class of HIV-1 integrase (IN) inhibitors with a dual mode of action. J Biol Chem 287:21189–203.
615 https://doi.org/10.1074/jbc.M112.347534.
616
617 12. Christ F, Shaw S, Demeulemeester J, Desimmie BA, Marchand A, Butler S, Smets W, Chaltin P,
618 Westby M, Debyser Z, Pickford C. 2012. Small-molecule inhibitors of the LEDGF/p75 binding site of
619 integrase block HIV replication and modulate integrase multimerization. Antimicrob Agents
620 Chemother 56:4365–74. https://doi.org/10.1128/AAC.00717-12.
621
622 13. Kessl JJ, Jena N, Koh Y, Taskent-Sezgin H, Slaughter A, Feng L, de Silva S, Wu L, Le Grice SF,
623 Engelman A, Fuchs JR, Kvaratskhelia M. 2012. Multimode, cooperative mechanism of action of
624 allosteric HIV-1 integrase inhibitors. J Biol Chem 287:16801–11.
626
627 14. Rouzic LE, Bonnard D, Chasset S, Bruneau JM, Chevreuil F, Strat LF, Nguyen J, Beauvoir R,
33
628 Amadori C, Brias J, Vomscheid S, Eiler S, Levy N, Delelis O, Deprez E, Saib A, Zamborlini A,
629 Emiliani S, Ruff M, Ledoussal B, Moreau F, Benarous R. 2013. Dual inhibition of HIV-1 replication

630 by integrase-LEDGF allosteric inhibitors is predominant at the post-integration stage. Retrovirology
631 10:144. https://doi.org/10.1186/1742-4690-10-144.
632
633 15. Jurado KA, Wang H, Slaughter A, Feng L, Kessl JJ, Koh Y, Wang W, Ballandras-Colas A, Patel
634 PA, Fuchs JR, Kvaratskhelia M, Engelman A. 2013. Allosteric integrase inhibitor potency is
635 determined through the inhibition of HIV-1 particle maturation. Proceedings of the National Academy
636 of Sciences 110:8690–8695. https://doi.org/10.1073/pnas.1300703110.
637
638 16. Balakrishnan M, Yant SR, Tsai L, O Sullivan C, Bam RA, Tsai A, Niedziela-Majka A, Stray KM,
639 Sakowicz R, Cihlar T. 2013. Non-Catalytic Site HIV-1 Integrase Inhibitors Disrupt Core Maturation
640 and Induce a Reverse Transcription Block in Target Cells. PLoS ONE 8:e74163.
641 https://doi.org/10.1371/journal.pone.0074163.
642
643 17. Desimmie BA, Schrijvers R, Demeulemeester J, Borrenberghs D, Weydert C, Thys W, Vets S,
644 Remoortel VB, Hofkens J, Rijck DJ, Hendrix J, Bannert N, Gijsbers R, Christ F, Debyser Z. 2013.
645 LEDGINs inhibit late stage HIV-1 replication by modulating integrase multimerization in the virions.
646 Retrovirology 10:57. https://doi.org/10.1186/1742-4690-10-57.
34
647
648 18. Kessl JJ, Kutluay SB, Townsend D, Rebensburg S, Slaughter A, Larue RC, Shkriabai N, Bakouche

649 N, Fuchs JR, Bieniasz PD, Kvaratskhelia M. 2016. HIV-1 Integrase Binds the Viral RNA Genome and
650 Is Essential during Virion Morphogenesis. Cell 166:1257–1268.e12.
651 https://doi.org/10.1016/j.cell.2016.07.044.
652
653 19. Fenwick C, Amad M, Bailey MD, Bethell R, Bös M, Bonneau P, Cordingley M, Coulombe R,
654 Duan J, Edwards P, Fader LD, Faucher A-MM, Garneau M, Jakalian A, Kawai S, Lamorte L, LaPlante
655 S, Luo L, Mason S, Poupart M-AA, Rioux N, Schroeder P, Simoneau B, Tremblay S, Tsantrizos Y,
656 Witvrouw M, Yoakim C. 2014. Preclinical profile of BI 224436, a novel HIV-1 non-catalytic-site
657 integrase inhibitor. Antimicrob Agents Chemother 58:3233–44. https://doi.org/10.1128/AAC.02719-
658 13.
659
660 20. Nakamura T, Campbell JR, Moore AR, Otsu S, Aikawa H, Tamamura H, Mitsuya H. 2016.
661 Development and validation of a cell-based assay system to assess human immunodeficiency virus
662 type 1 integrase multimerization. J. Virol. Methods 236:196–206.
663 https://doi.org/10.1016/j.jviromet.2016.07.023.
664
665 21. Lu R, Limón A, Devroe E, Silver PA, Cherepanov P, Engelman A. 2004. Class II integrase mutants
35
666 with changes in putative nuclear localization signals are primarily blocked at a postnuclear entry step
667 of human immunodeficiency virus type 1 replication. J Virol 78:12735–46.

668 https://doi.org/10.1128/JVI.78.23.12735-12746.2004.
669
670 22. Sharma A, Slaughter A, Jena N, Feng L, Kessl JJ, Fadel HJ, Malani N, Male F, Wu L, Poeschla E,
671 Bushman FD, Fuchs JR, Kvaratskhelia M. 2014. A new class of multimerization selective inhibitors
672 of HIV-1 integrase. PLoS Pathog 10:e1004171. https://doi.org/10.1371/journal.ppat.1004171.
673
674 23. Fontana J, Jurado KA, Cheng N, Ly NL, Fuchs JR, Gorelick RJ, Engelman AN, Steven AC. 2015.
675 Distribution and Redistribution of HIV-1 Nucleocapsid Protein in Immature, Mature, and Integrase-
676 Inhibited Virions: a Role for Integrase in Maturation. J Virol 89:9765–80.
677 https://doi.org/10.1128/JVI.01522-15.
678
679 24. Jenkins TM, Hickman AB, Dyda F, Ghirlando R, Davies DR, Craigie R. 1995. Catalytic domain
680 of human immunodeficiency virus type 1 integrase: identification of a soluble mutant by systematic
681 replacement of hydrophobic residues. Proc Natl Acad Sci U S A 92:6057–61.
682 https://doi.org/10.1073/pnas.92.13.6057.
683
684 25. Niesen FH, Berglund H, Vedadi M. 2007. The use of differential scanning fluorimetry to detect
36
685 ligand interactions that promote protein stability. Nat Protoc 2:2212–2221.
686 https://doi.org/10.1038/nprot.2007.321.

687
688 26. Dyda F, Hickman AB, Jenkins TM, Engelman A, Craigie R, Davies DR. 1994. Crystal structure of
689 the catalytic domain of HIV-1 integrase: similarity to other polynucleotidyl transferases. Science
690 266:1981–6. https://doi.org/10.1126/science.7801124.
691
692 27. Feng L, Sharma A, Slaughter A, Jena N, Koh Y, Shkriabai N, Larue RC, Patel PA, Mitsuya H, Kessl
693 JJ, Engelman A, Fuchs JR, Kvaratskhelia M. 2013. The A128T resistance mutation reveals aberrant
694 protein multimerization as the primary mechanism of action of allosteric HIV-1 integrase inhibitors. J
695 Biol Chem 288:15813–20. https://doi.org/10.1074/jbc.M112.443390.
696
697 28. Demeulemeester J, Tintori C, Botta M, Debyser Z, Christ F. 2012. Development of an AlphaScreen-
698 based HIV-1 integrase dimerization assay for discovery of novel allosteric inhibitors. J Biomol Screen
699 17:618–28. https://doi.org/10.1177/1087057111436343.
700
701 29. Ullman EF, Kirakossian H, Singh S, Wu ZP, Irvin BR, Pease JS, Switchenko AC, Irvine JD, Dafforn
702 A, Skold CN. 1994. Luminescent oxygen channeling immunoassay: measurement of particle binding
703 kinetics by chemiluminescence. Proc. Natl. Acad. Sci. U.S.A. 91:5426–30.
37
704 https://doi.org/10.1073/pnas.91.12.5426.
705

706 30. Passos DO, Li M, Yang R, Rebensburg SV, Ghirlando R, Jeon Y, Shkriabai N, Kvaratskhelia M,
707 Craigie R, Lyumkis D. 2017. Cryo-EM structures and atomic model of the HIV-1 strand transfer
708 complex intasome. Science 355:89–92. https://doi.org/10.1126/science.aah5163.
709
710 31. Slaughter A, Jurado K, Deng N, Feng L, Kessl J, Shkriabai N, Larue R, Fadel H, Patel P, Jena N,
711 Fuchs J, Poeschla E, Levy R, Engelman A, Kvaratskhelia M. 2014. The mechanism of H171T
712 resistance reveals the importance of Nδ-protonated His171 for the binding of allosteric inhibitor BI-D
713 to HIV-1 integrase. Retrovirology 11:100. https://doi.org/10.1186/s12977-014-0100-1.
714
715 32. Hoyte AC, Jamin AV, Koneru PC, Kobe MJ, Larue RC, Fuchs JR, Engelman AN, Kvaratskhelia
716 M. 2017. Resistance to pyridine-based inhibitor KF116 reveals an unexpected role of integrase in HIV-
717 1 Gag-Pol polyprotein proteolytic processing. J Biol Chem 292:19814–19825.
719
720 33. Engelman A. 1999. In vivo analysis of retroviral integrase structure and function. Adv Virus Res.
721 52:411-26. https://doi.org/10.1016/S0065-3527.60309-7.
722
38
723 34. Amano M, Koh Y, Das D, Li J, Leschenko S, Wang YF, Boross PI, Weber IT, Ghosh AK, Mitsuya
724 H. 2007. A novel bis-tetrahydrofuranylurethane-containing nonpeptidic protease inhibitor (PI), GRL-

725 98065, is potent against multiple-PI-resistant human immunodeficiency virus in vitro. Antimicrob
726 Agents Chemother 51:2143–55. https://doi.org/10.1128/AAC.01413-06.
727
728 35. Otwinowski Z, Minor W. 1997. Processing of X-ray diffraction data collected in oscillation mode.
729 Methods Enzymol. 276:307–26. https://doi.org/10.1016/S0076-6879.76066-X.
730
731 36. Vagin A, Teplyakov A. 2010. Molecular replacement with MOLREP. Acta Crystallogr. Sect. D Biol.
732 Crystallogr. 66:22–5. https://doi.org/10.1107/S0907444909042589.
733
734 37. Adams PD, Afonine PV, Bunkóczi G, Chen VB, Davis IW, Echols N, Headd JJ, Hung LW, Kapral
735 GJ, Grosse-Kunstleve RW, McCoy AJ, Moriarty NW, Oeffner R, Read RJ, Richardson DC, Richardson
736 JS, Terwilliger TC, Zwart PH. 2010. PHENIX : a comprehensive Python-based system for
737 macromolecular structure solution. Acta Crystallogr. Sect. D Biol. Crystallogr. 66:213–221.
738 https://doi.org/10.1107/S0907444909042589.
739
740 38. Emsley P, Cowtan K. 2004. Coot: model-building tools for molecular graphics. Acta Crystallogr D
741 Biol Crystallogr. 60(Pt 12 Pt 1):2126–32. https://doi.org/10.1107/S0907444904019158.
39
742
743 39. Cassel JA, Blass BE, Reitz AB, Pawlyk AC. 2010. Development of a novel nonradiometric assay

744 for nucleic acid binding to TDP-43 suitable for high-throughput screening using AlphaScreen
745 technology. J Biomol Screen 15:1099–106. https://doi.org/10.1177/1087057110382778.
746
747 40. Ghosh AK, Leshchenko S, Noetzel M. 2004. Stereoselective photochemical 1,3-dioxolane addition
748 to 5-alkoxymethyl-2(5H)-furanone: synthesis of bis-tetrahydrofuranyl ligand for HIV protease
749 inhibitor UIC-94017 (TMC-114). J Org Chem 69:7822–9. https://doi.org/10.1021/jo049156y
750
751 41. Amadori C, Velden Y, Bonnard D, Orlov I, Bel N, Rouzic E, Miralles L, Brias J, Chevreuil F,
752 Spehner D, Chasset S, Ledoussal B, Mayr L, Moreau F, García F, Gatell J, Zamborlini A, Emiliani S,
753 Ruff M, Klaholz B, Moog C, Berkhout B, Plana M, Benarous R. 2017. The HIV-1 integrase-LEDGF
754 allosteric inhibitor MUT-A: resistance profile, impairment of virus maturation and infectivity but
755 without influence on RNA packaging or virus immunoreactivity. Retrovirology14:50.
756 https://doi.org/10.1186/s12977-017-0373-2.
757
758 42. Desimmie BA, Weydert C, Schrijvers R, Vets S, Demeulemeester J, Proost P, Paron I, Rijck J De,
759 Mast J, Bannert N, Gijsbers R, Christ F, Debyser Z. 2015. HIV-1 IN/Pol recruits LEDGF/p75 into viral
760 particles. Retrovirology 12:16. https://doi.org/10.1186/s12977-014-0134-4.
40
761
762 43. Hombrouck A, Rijck J De, Hendrix J, Vandekerckhove L, Voet A, Maeyer M De, Witvrouw M,

763 Engelborghs Y, Christ F, Gijsbers R, Debyser Z. 2007. Virus evolution reveals an exclusive role for
764 LEDGF/p75 in chromosomal tethering of HIV. PLoS Pathog. 3:e47. https://doi.org/
765 10.1371/journal.ppat.0030047.
766
767 44. Feng L, Dharmarajan V, Serrao E, Hoyte A, Larue RC, Slaughter A, Sharma A, Plumb MR, Kessl
768 JJ, Fuchs JR, Bushman FD, Engelman AN, Griffin PR, Kvaratskhelia M. 2016. The Competitive
769 Interplay between Allosteric HIV-1 Integrase Inhibitor BI/D and LEDGF/p75 during the Early Stage
770 of HIV-1 Replication Adversely Affects Inhibitor Potency. ACS Chem. Biol 11:1313–21.
771 https://doi.org/ 10.1021/acschembio.6b00167.
772
773 45. Shkriabai N, Dharmarajan V, Slaughter A, Kessl JJ, Larue RC, Feng L, Fuchs JR, Griffin PR,
774 Kvaratskhelia M. 2014. A critical role of the C-terminal segment for allosteric inhibitor-induced
775 aberrant multimerization of HIV-1 integrase. J Biol Chem 289:26430–40. https://doi.org/
776 10.1074/jbc.M114.589572.
777
778 46. Gupta K, Turkki V, Sherrill-Mix S, Hwang Y, Eilers G, Taylor L, McDanal C, Wang P, Temelkoff
779 D, Nolte RT, Velthuisen E, Jeffrey J, Duyne GD Van, Bushman FD. 2016. Structural Basis for Inhibitor-
41
780 Induced Aggregation of HIV Integrase. PLoS Biol 14:e1002584.
781 https://doi.org/10.1371/journal.pbio.1002584.

782
783 Figure legends
784 FIG 1 NCINIs information and the emergence of NCINI-related resistance mutations
785 in the IN region.
786 (A) Chemical structures and names of NCINIs in this study. (B) Selection experiments of NCINI-1, -
787 2, and -3. HIV-1NL4-3 propagated in MT-4 cells in the presence of increasing concentrations of NCINI-
788 1 (▲), NCINI-2 (●), or NCINI-3 (■), and the resistant viral selection to the NCINIs continues up to
789 10 µM. (C) Dominant NCINI-related resistance mutations in the IN region (%) are shown in the
790 schematic of HIV-1 integrase and at following passage ranges 1-10, 11-15, 16-20, and over 21. These
791 mutations at each passage were identified from cellular DNA in the infected MT-4 cells with the
792 NCINI-related resistant HIV-1.
793
794 FIG 2 Characteristics of recombinant IN proteins carrying the NCINI-3 resistance
795 mutations.
796 (A) SEC analysis of IN proteins carrying the NCINI-3 resistance mutations. The elution profile of
797 INWT proteins at 50 µM concentration was shown together with estimated IN multimers and two
798 molecular weight standard proteins (158 and 44 KDa). Those of INA128T, INH171Q, INK173Q, INN254K,
42
799 INP15, INP20, INP26, and INE11K proteins, respectively were shown compared with INWT. (B) Multimers
800 ratio of IN proteins (INWT and INP26) cross-linking with BS3. Each multimer’s intensity of INWT and

801 INP26 measured by ImageJ was shown as multimers ratios (%). Molecular weight (MW), buffer (Lane
802 1), INWT and INP26 (Lanes 2 and 4), and INWT and INP26 cross-linked with BS3 at 0.05 mM (Lanes 3
803 and 5) were shown by immunoblotting with an anti-IN antibody. The figure is representative data and
804 the ratios represent mean values (±standard deviation, SD) from three independent experiments. (C)
805 Results of the BiFC-IN system carrying the NCINI-3 resistance mutations and E11K. Fluorescence
806 intensity derived from multimers of BiFC-INA128T, -INH171Q, -INK173Q -INN254K, -INP15, -INP20, -INP26,
807 and -INE11K in HEK293T cells measured by FACS was shown as a ratio of area under the curve (AUC)
808 compared to BiFC-INWT. Error bars indicate ±SD from three independent experiments. The BiFC-IN
809 system is summarized in ref. (20) Statistical significance is examined using a parametric two-tailed
810 Student t-test; *, P < 0.05. (D) LEDGFHis and F-IN pull-down assay. Loading samples of F-INWT, F-
811 INP26, F-INV165A (Lanes 1, 2, and 3), LEDGFHis (Lane 4), and LEDGFHis together with F-INWT, F-INP26,
812 and F-INV165A, respectively (Lanes 5, 6, and 7) were shown at the left panel. Pull down samples of F-
813 INWT, F-INP26, F-INV165A (Lanes 8, 9, and 10), LEDGFHis (Lane 11), and LEDGFHis incubated for 30
814 min with F-INWT, F-INP26, and F-INV165A (Lanes 5, 6, and 7), respectively were shown after His-tag
815 column selection at the right panel. The binding ability of IN proteins carrying these mutations to
816 LEDGF/p75 was shown as IN-LEDGF/p75 ratios (%) by measuring the band intensity using ImageJ.
817 The ratios represent mean values (±SD) from two independent experiments.
43
818
819 FIG 3 Characteristics of recombinant HIV-1 IN clones carrying the NCINI-3 resistance

820 mutations with IN under-multimerization.
821 (A) Gag and Gag-Pol proteolytic processing products of recombinant HIV-1NL4-3 carrying NCINI-
822 related mutations or INE11K clones, respectively. The HIV-1 IN clones were produced in the presence
823 (even lanes) or absence (odd lanes) of 2 µM SQV. Gag and Gag-Pol proteolytic processing products
824 normalized with p24 levels of each clone were visualized by immunoblotting with anti-HIV-1 Gag, IN,
825 and RT antibody. Three times the amount of INE11K (Lanes 19 and 20) was examined due to the epitope
826 of the anti-IN antibody including E11 residue. (B) Viral production, p24 levels of HIV-1 INWT and
827 HIV-1 clones carrying NCINI-3 resistance mutations or E11K mutation. Data indicated relative ratios
828 normalized to HIV-1 INWT and mean values (±SD) from three independent experiments. Statistical
829 significance was examined using a two-tailed Student t-test; *, P < 0.05. (C) Morphologies of HIV-1
830 INWT and HIV-1 INP26 clone in the presence or absence of 20 µM NCINI-3 using TEM. Representative
831 images of normal, abnormal, and others HIV-1 particles (Magnification, 168,000 x scale bar, 100 nm)
832 were shown at the upper panel. Over 100 numbers of the HIV-1 particles were examined from several
833 images of each HIV-1 sample, and the percentages of HIV-1 morphology classified as normal,
834 abnormal, and others were shown in the lower graph. (D) Replication kinetics of HIV-1 carrying the
835 NCINI-3 resistance mutation, or INE11K clones. MT-4 cells were exposed to HIV-1 preparation
836 normalized with the p24 levels, and the production of each HIV-1 clone from the MT-4 cells was
44
837 monitored at day 1, 3, 5, and 7 by p24 ELISA. The assays were performed in triplicate, and error bars
838 indicate ±SD from two independent experiments.

839
840 FIG 4 Effect of the NCINI-3 resistance mutations on CCD.
841 (A) SEC analysis of CCD proteins carrying the NCINI-3 resistance mutations together with F185K.
842 The elution profile of CCDWT was shown with two molecular weight standards (44 and 17 KDa) and
843 estimating the size position of CCD dimers and monomers, respectively. That of CCDA128T, CCDH171Q,
844 CCDK173Q, CCDP15, and CCDP26 del, respectively at 50 µM concentration was shown compared with
845 CCDWT. (B) DSF analysis of CCD proteins carrying the NCINI-3 resistance mutations. Fluorescence
846 of CCDWT (○), CCDA128T (◊), CCDH171Q (△), CCDK173Q (□), CCDP15 (*), and CCDP26 del (+) was
847 detected and plotted as a relative ratio from 25 to 95°C at the left panel. Expanded plots from 30 to
848 50°C were shown at the right panel. (C) ΔTm 25 or 50 indicates that Tm 25 or 50 differences between
849 CCDWT and CCD carrying the mutations at 12.5, 25, and 50 µM CCD concentration. The assays were
850 performed in triplicate, and error bars indicate ±SD from two independent experiments. Statistical
851 significance of ΔTm 25 or 50 was examined using a two-tailed Student t-test; *, P < 0.05. (D) Relative
852 ratios of SYPRO orange fluorescence of CCDWT and CCD carrying the mutations from 30 to 40°C at
853 25 μM concentration. (E) The proportion of surface hydrophobic sites of CCD proteins at 37°C. The
854 relative ratios were obtained from SYPRO orange binding to the surface hydrophobic sites of CCD
855 carrying the mutations to CCDWT shown at 12.5, 25, and 50 μM concentration. The assays were
45
856 performed in triplicate, and error bars indicate ±SD from two independent experiments. Statistical
857 significance was examined using Student t-test; *, P < 0.05. (F) SPR analysis of NCINI-3 interactions

858 with CCDWT, CCDA128T, CCDH171Q, CCDK173Q, and CCDP15, respectively. Data shown represents
859 derived from two independent experiments. KD values of NCINI-3 to CCDWT and CCD carrying the
860 mutations, and fold changes indicate KD ratios of CCD carrying the mutations to CCDWT. Error bars
861 indicate ±SD from two independent experiments.
862
863 FIG 5 P15 mutation changes NCINIs binding space formed in the CCD.
864 (A) The X-ray crystal structure of CCDWT dimer (PDB: 4GW6). NCINI-related compound interacts
865 with the binding pocket formed from CCD1WT and CCD2WT monomers colored in cyan and grey,
866 respectively. Intermolecular hydrogen bonds between CCD1WT and CCD2WT were shown by the red
867 dashed lines. Locations of side chains for Y99 and K173 residues in the monomer-monomer interface
868 of CCDWT were indicated. (B) The X-ray crystal structure of CCDP15 dimer (PDB: 6L0C). CCD1P15
869 and CCD2P15 monomers were colored in green and light green, respectively. Intermolecular and
870 intramolecular hydrogen bonds in CCDP15 dimers were shown by the red dashed lines. Locations of
871 side chains for Y99 and Q173 residues in the CCDP15 were indicated. (C) Overlay of the crystal
872 structures of CCDWT in complex with the NCINI-related compound (PDB: 4GW6) and CCDP15 (PDB:
873 6L0C). (D) X-ray data collection and refinements.
874
46
875 FIG 6 IN under-multimerization is stabilized by HIV-1 RNA and restored in the viral
876 particles.

877 (A) Alpha assay 1 for direct binding between Bio-TAR and INHis. Schematic illustration indicates that
878 streptavidin-coated Donor (D) and anti-His Accepter (A) beads bind to Bio-TAR and INHis complex at
879 the left panel. Alpha signal intensity between INHis (INWT, INP26, and INRAKA at 10 nM) and Bio-TAR
880 (0 from 1000 nM) was shown at left under the panel. The intensity of INWT, INP26, and INRAKA at 0
881 from 10 nM of Bio-TAR was fitting to saturated curves at right under the panel. The curves and KD
882 values (±SD) were determined from representative fitting data of three independent experiments. (B)
883 Alpha assay 2 for indirect IN multimerization between F-IN and INHis. Schematic illustration indicates
884 that anti-Flag coated D and anti-His coated A beads bind to a dimer of F-IN and INHis. Alpha signal
885 ratios between F-IN and INHis proteins (INWT, INP26, and INRAKA, respectively) by the addition of 0.1
886 from 1000 nM concentration of TAR-RNA were fitting to sigmoid curves. Hillslopes and Log EC50s
887 (±SD) were calculated by representative sigmoid curves from three independent experiments. (C) IN
888 multimerization of HIV-1NL4-3 INWT and INP26 viral particles generated in the presence or absence of
889 20 μM NCINI-3. The HIV-1 INWT and INP26 clones were purified by ultracentrifugation. The high
890 concentrated HIV-1 INWT and INP26 were cross-linked with (Lanes 2, 4, 6, and 8) or without (Lanes 1,
891 3, 5, and 7) BS3, and visualized by SDS-PAGE with anti-HIV-1 IN antibody. Monomers (Mo), Dimers
892 (Di), Tetramers (Te), and High-order multimers (Hi) of INWT and INP26 were shown as multimer ratios
893 (%) by measuring each band intensity using ImageJ. The ratios represent mean values (±SD) from two
47
894 independent experiments.
895

896 FIG 7 Graphical article of HIV-1 escape reaction only during HIV-1 maturation.
897 The schematic illustration focuses on the interaction of IN multimerization with HIV-1 RNA and the
898 location of RNP in the viral particles during HIV-1 maturation from Gag-Pol proteolytic processing in
899 this study. The upper illustration indicates that normal interaction of functional IN multimerization with
900 HIV-1 RNA results in mature HIV-1. The middle illustration shows that over-multimerized IN proteins
901 induced by NCINIs fail to interact with HIV-1 RNA, producing immature HIV-1 in which the RNP
902 complex translocated from the capsid lattice. The lower illustration indicates that under-multimerized
903 INP26 proteins can escape from NCINIs binding and are restored to the same level as INWT
904 multimerization in the viral particles, producing mature HIV-1 in the presence of NCINIs as one of the
905 HIV-1 escape mechanisms.
48
FIG. 1
A.
NCINI-1 NCINI-2 NCINI-3 NCINI-4
(LEDGIN6) (LEDFIN7) (CX14442) (BI224436)
(CX04328) (CX05045)
(CX05168)

B.
10
Drug concentration (μM)
NCINI-1
0.1 NCINI-2
NCINI-3
0.01
0 5 10 15 20 25 30
Passage number
C. HIV-1 integrase
V77S A128T H171Q K173Q N254K G272R
NTD CCD CTD

50 212 288
Dominant mutations (%) of NCINIs

Passage NCINI-1 NCINI-2 NCINI-3
1-10 - - WT
11-15 - V77S, A128T (83.3%) A128T, K173Q (44.4%)
16-20 A128T, G272R (50.0%) A128T, K173Q, N254K (50.0%)
21- A128T, H171Q, K173Q, N254K (77.7%)

FIG. 2
A.
Tetramer INWT
INA128T INH171Q INK173Q
OD280 (mAU)
Dimer & Monomer
158
44 KDa

ml
INN254K INP15 INP20 INP26 INE11K
B. INWT INP26
C.
KDa
- + - + BS3 1.5
120
BiFC-IN
Tetramer
100
AUC ratio
80
1.0
Dimer
60 * * *
50
* *
0.5
40
Monomer
30
1 2 3 4 5
0.0
Multimers ratio (%)
+BS3
Monomer Dimer Tetramer
INWT 70.6 ±1.7 22.9 ±1.3 6.5 ±0.6
INP26 77.1 ±0.8 19.6 ±0.8 3.3 ±0.6
D.
Load Pull down
KDa
62 LEDGF
49 IN-LEDGF binding ratio (%)

INWT 39.7 ± 2.3
38
INP26 50.6 ±10.6
INV165A 9.7 ± 5.0
IN
28
1 2 3 4 5 6 7 8 9 10 11 12 13 14
INWT INP26 INV165A INWT INP26 INV165A INWT INP26 INV165A INWT INP26 INV165A
LED LED LED LED LED LED LED LED
HIV-1 INWT -INH171Q -INN254K -INP20 -INE11K
FIG. 3
A. -INA128T -INK173Q -INP15 -INP26
KDa + + + + + + + + + SQV
100
80
60 Gag
p49
50
p41 Anti-Gag
40

30
CA
20
MA
-INE11K (x3)
120
100 Pol
RT-IN
80
60
50
Anti-IN
RNase-IN
40
30 IN
120 Pol
100 RT-IN
80
60 p66 (RT) Anti-RT
50 p51 (RT)
40
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
B. 2
D.
Relative ratio (/WT)
HIV-1NL43
3000
1 WT A128T H171Q
*
K173Q N254K P15
Average of P24 (ng/ml)
P20 P26 E11K

0
2000
C. Normal Abnormal Others 1000
0
100 nm Average of p24 (ng/ml)
Day 1 3 5 7
100 0 30 ± 35 830 ± 198 1313 ± 58
Normal Abnormal Others HIV-1NL4-3 INWT
79.4
74.4 75.5 -INA128T 0 49 ± 18 675 ± 102 1031 ± 118
80
-INH171Q 0 71 ± 40 946 ± 127 1221 ± 193
56.6
60 -INK173Q 0 22 ± 24 1074 ± 283 1672 ± 461
-INN254K 0 19 ± 22 980 ± 233 1601 ± 230
%
32.8 0 42 ± 12 971 ± 81 1307 ± 265

40 -INP15
19.6 -INP20 0 25 ± 19 1035 ± 63 1412 ± 216
14.0 15.0
20 6.6
10.6 10.6 -INP26 0 10 ± 14 795 ± 352 1443 ± 340
4.9
-INE11K 0 0 0 0
0
FIG. 4
Monomer
A.
Dimer
CCDWT CCDA128T CCDH171Q
OD280
44 17 KDa

ml
CCDK173Q CCDP15 (A128T/K173Q) CCDP26 del (A128T/H171Q/K173Q)
* CCD
P15
CCDWT CCDA128T CCDH171Q CCDK173Q ＋ CCDP26 del
B. 0.5
Tm 50
1
CCD 25 µM
0.75
Relative ratio
0.25
Tm 50 Tm 25
0.5
0.25 Tm 25
0 0
25 35 45 55 65 75 85 95 30 40 50
Temperature (℃)
C. ΔTm 25 ΔTm 50
(CCD concentration) (CCD concentration)
5 10
12.5 µM 25 µM 50 µM 12.5 µM 25 µM 50 µM
*
Temperature (℃)
5
0
-5
-5
*
*
-10
-10 *
FIG. 4
D. CCDWT CCDA128T CCDH171Q

E.
CCDK173Q * CCDP15 ＋CCDP26 del 12.5 µM 25 µM 50 µM
3 *
0.2
Relative ratio (SYPRO orange)
Relative ratio (/CCDWT) at 37℃

CCD 25 µM *
*

*
2
0.1 *
0 0
30 35 37℃ 40
Temperature (℃)
F. 10 µM 10 µM
30 CCDWT + NCINI-3 5
30 CCDA128T + NCINI-3 5
2 2
Response (RU)
1 20 1
20 0.5 0.5
0.2
0.1
10 0.05 10
0 0
-10 -10
0 50 100 150 0 50 100 150
Time (s)
30 10 µM
30 10 µM
CCDH171Q + NCINI-3 5
CCDK173Q + NCINI-3 5
2
1 2
20 20 1
0.5
0.2 0.5
0.2
10 10
0 0
-10 -10
0 50 100 150 0 50 100 150
30 10 µM
CCDP15 + NCINI-3 5 Mutations Binding affinity, KD (µM) Fold change
2
20 1 WT 0.95 ± 0.14 1.0
A128T 2.10 ± 0.36 2.2
10
H171Q 1.35 ± 0.20 1.4
0 K173Q 1.41 ± 0.13 1.5
P15 N.D. N.D.
-10
0 50 100 150
FIG. 5
A. CCDWT B. CCDP15 (A128T/K173Q)
NCINI

His171 His171
Tyr99
Tyr99
Lys173 Gln173
Val88 Val88
Gln177 Gln177
Glu87 Glu87
CCD1WT CCD2WT CCD1P15 CCD2P15
C. Overlay of CCDWT and CCDP15
NCINI NCINI
A128T
Tyr99 K173Q
FIG. 5
D.
CCD (A128T/K173Q/F185K)
Data collection
Wavelength (Å) 0.98

Space group P3121
Unit-cell lengths (Å) a = b = 72.769, c = 64.259
50.0-2.10
Resolution range (Å)
(2.14-2.10)
No. of observed reflections 113,571
No. of unique reflections 11,817
Completeness (%) 100 (100)
Rmerge (%) a 6.4 (51.4)
<I / sigmaI> 43.0 (2.7)
Refinement statistics
Resolution range (Å) 50.0-2.10
No. of reflections used 11,791
Completeness (%) 99.9
Rwork /Rfree (%) b 19.3/24.7
R.m.s.d. in bonds (Å) 0.007
R.m.s.d. in angles (deg.) 0.929
Ramachandran plot
Favoured (%) 96.2
Allowed (%) 3.8
Highest resolution shell is shown in parenthesis.

merge = Σhkl Σi |Ii (hkl) - <I(hkl)>| / Σhkl Σi Ii (hkl), where Ii (hkl) is the observed intensity and <I(hkl)> is
a R
the mean value of Ii (hkl).

b Rwork = Σ||Fo| - |Fc|| / Σ|Fo|. Rfree was calculated from the test set (5% of the total data).
FIG. 6
A. INHis Mutations KD (nM) [fold change]
680 nm
615 nm
Bio-TAR IN INWT 2.06 ±0.19 [1]
IN
+
Bio D Bio
INP26 1.75±0.23 [0.84]
His A
His
INRAKA N.D.
INWT INP26 INRAKA

INWT INP26 INRAKA
20000 20000
Signal intensity
15000 15000
Signal intensity
10000 10000
5000 5000
0 0
0.01 0.1 1 10 100 1000 0 2 4 6 8 10
TAR-RNA (nM) TAR-RNA (nM)
B.
2 .0 INWT INP26 INRAKA
Flag Flag D
TAR
IN IN
+ IN
Signal ratio
His
His AA
1 .5
Mutations Hill slope LogEC50 (nM)

1 .0
WT
IN 0.13 ±0.11 [1] 10.00 ±1.34 [1]
INP26 0.12 ±0.10 [0.92] 7.79 ± 2.35 [0.78]

RAKA 0 .5
IN N.D. N.D.
0 .1 1 10 100 1000
TAR-RNA (nM)
C.
HIV-1 -INWT -INP26
+ + - - + + NCINI-3
- -
- + - + - + - + BS3 100
Band intensity (%) of IN multimers
KDa Higher-order
Multimer (Hi) 80
120
Tetramer (Te) 60
100
(%)
80 40
60 Dimer (Di)
20
50
0
WT+ P26+
WT P26
NCI-3 NCI-3
40
Mo 75.9 62.0 67.9 67.6
Monomer (Mo) Di 23.4 36.2 30.7 29.1

30 Te 0.8 1.5 1.4 3.3
Hi 0.0 0.3 0.0 0.0
1 2 3 4 5 6 7 8
FIG. 7
HIV-1 maturation
Pol HIV-1 viral particle
Capsid lattice
PR Gag-Pol HIV-1 RNA
proteolytic Nucleocapsid
processing

IN IN
INWT RT
Functional multimerization
IN
RNP complex
+NCINIs
NCINIs
PR
INWT RT IN IN
IN Over-multimerization
PR
INP26 RT IN
Under-multimerization Functional multimerization

IN
a b
TABLE 1 Antiviral activities of NCINIs and INSTIs against HIV-1LAI and HIV-1NL4-3, and cytotoxicities
EC50 (μM) CC50 (μM)

drugs MT-2 MT-4
MT-4 HEK
HIV-1 LAI HIV-1 NL4-3

NCINI-1 2.29 ± 1.19 7.01 ± 5.41 42.23 ± 9.44 34.26 ± 4.19
NCINI-2 0.77 ± 0.13 2.60 ± 0.34 66.59 ± 1.70 >100
NCINI-3 0.029 ± 0.003 0.19 ± 0.13 45.67 ± 6.17 >100
NCINI-4 0.032 ± 0.001 0.027 ± 0.004 36.09 ± 1.59 >100
RAL 0.0024 ± 0.0012 0.0019 ± 0.0008 >100 >100
EVG 0.0019 ± 0.0020 0.0036 ± 0.0002 2.89 ± 0.59 4.19 ± 3.89
DTG 0.0010 ± 0.0010 0.0029 ± 0.0010 14.33 ± 0.64 21.39 ±0.97
a
MT-2 cells were exposed to 100 TCID50s of HIV-1LAI and cultured in the presence of various concentration of
NCINIs and INSTIs, and EC50s were determined by using MTT assay.
b
Cytotoxicities (CC50s) of NCINIs and INSTIs against MT-4 and HEK293T cells were also determined by
MTT assay. All assays were conducted in duplicate, and data shown represent mean values (±SD) from three
independent experiments.
a
TABLE 2 Antiviral activities of NCINIs, RAL, and DRV against NCINI-3-resistance HIV-1NL4-3 clones
NCINI-1 NCINI-2 NCINI-3 NCINI-4

recombinant HIV-1 clone
EC50 (µM) fold change EC50 (µM) fold change EC50 (µM) fold change EC50 (µM) fold change
WT
HIV-1NL4-3 IN 7.01 ± 5.41 1 2.60 ± 0.34 1 0.19 ± 0.13 1 0.027 ± 0.004 1

A128T
-IN > 10 N.D. > 10 N.D. 3.12 ± 0.12 16.1 0.31 ± 0.01 11.6
P15
-IN > 10 N.D. > 10 N.D. 3.36 ± 0.11 17.3 0.32 ± 0.06 12.2
P20
-IN - - - - 2.14 ± 0.14 11.1 0.39 ± 0.04 14.7
P26
-IN - - - - 34.83 ± 2.21 179.8 2.43 ± 0.61 91.7
H171Q
-IN > 10 N.D. > 10 N.D. 0.42 ± 0.08 2.2 0.31 ± 0.04 11.7
K173Q
-IN > 10 N.D. > 10 N.D. 0.14 ± 0.03 0.7 0.021 ± 0.01 0.8
N254K
-IN - - - - 0.27 ± 0.04 1.4 0.092 ± 0.03 3.5
N.D., not determined - , not examined
RAL DRV
recombinant HIV-1 clone
EC50 (µM) fold change EC50 (µM) fold change
WT
HIV-1NL4-3 IN 0.0017 ± 0.0025 1 0.0037 ± 0.0003 1
A128T
-IN 0.0028 ± 0.0056 1.7 0.0094 ± 0.0008 2.5
P15
-IN 0.0028 ± 0.0058 1.7 0.0097 ± 0.0031 2.6
P20
-IN 0.0008 ± 0.0041 0.5 0.0074 ± 0.0018 2.0
P26
-IN 0.0017 ± 0.0025 1.0 0.0034 ± 0.0016 0.9
H171Q
-IN 0.0027 ± 0.0051 1.6 0.0085 ± 0.0024 2.3
K173Q
-IN 0.0022 ± 0.0011 1.3 0.0032 ± 0.0011 0.8
N254K
-IN 0.0019 ± 0.0010 1.1 0.0054 ± 0.0020 1.4
a 4
MT-4 cells (1 × 10 /ml) are exposed to 100 TCID50s of each infectious HIV-1 clone, and the inhibition of p24 production by the
drug was used as the endpoint on day 7 in culture. The fold changes represent EC50 ratios against each HIV-1 clone to HIV-1 wild
type. All assays are performed in duplicate, and the data shown are mean values (±SD) derived from the results of three
independent experiments.

10.1128@JVI.00486-20

Uploaded by

Copyright:

Available Formats

You might also like

10.1128@JVI.00486-20

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

10.1128@JVI.00486-20

Uploaded by

Copyright:

Available Formats

JVI Accepted Manuscript Posted Online 1 July 2020

1 A conformational escape reaction of HIV-1 against an allosteric integrase inhibitor

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

6 Tomofumi Nakamura,a Teruya Nakamura,b,c Masayuki Amano,a Toshikazu Miyakawa,a

7 Yuriko Yamagata,c Masao Matsuoka,a,d Hirotomo Nakataa*

10 Sciences, Faculty of Life Sciences, Kumamoto University, Kumamoto 860-8556, Japan.

15 Sakyo-ku, Kyoto 606-8507, Japan.

17 *Corresponding author: Hirotomo Nakata, nakatahi@gpo.kumamoto-u.ac.jp; Tel.: +81-096-373-5156;

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

32 multimerization of full-length IN proteins (IN under-multimerization). This under-multimerized IN via

36 IN over-multimerization by IN under-multimerization as one of the escape mechanisms. Our findings

38 influence the development of anti-HIV-1 strategies.

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

46 can acquire some mutations which cause IN under-multimerization to alleviate NCINI-induced IN

47 over-multimerization. This under-multimerized IN was efficiently stabilized by HIV-1 RNA and

52 HIV-1 drugs and unique treatment strategies.

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

91 between over-multimerized IN and HIV-1 RNA during HIV-1 maturation (18).

97 anti-HIV-1 drugs and unique treatment strategies.

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

100 Emergence of NCINI-related resistance mutations

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

120 Characteristics of recombinant IN proteins carrying NCINI-related resistance mutations

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

137 SEC data (Fig 2C).

146 multimerization may be associated with NCINI-3 related resistance of HIV-1.

148 Characteristics of recombinant HIV-1NL4-3 clones carrying NCINI-related resistance mutations

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

156 recombinant HIV-1 IN clones with or without IN under-multimerization, analyzed by immunoblotting

162 viral production.

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

190 monomer (Fig. 4A).

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

203 length IN multimerization.

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

217 structure of CCDP15 carrying F185K using X-ray crystallography.

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

247 RNA (HIV-1 RNA) may be slightly reinforced.

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

255 efficiently by HIV-1 RNA compared to INWT.

268 (Fig. 7).

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

280 IN under-multimerized proteins carrying A128T/K173Q still retained intrinsic IN function.

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

292 against the NCINI action.

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

333 Materials and Methods

341 Plasmids constructs

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

357 above the method.

359 Proteins expression and purification

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

367 mM CHAPS, 10 mM MgSO4, 10% glycerol) supplemented with 5 mM 2-mercaptoethanol (BME), 10

Downloaded from http://jvi.asm.org/ on July 4, 2020 at Univ of Massachusetts Amherst

389 Recombinant HIV-1 clones

396 and stored at -80°C.

398 Western blotting