Pharmacological descriptors related to the binding of Gp120 to CD4 corresponding to 60 representative HIV-1 strains

Abstract

The standardization of HIV-1 strains for vaccine tests include the use of viral reference panels. We determined a series of pharmacological descriptors (molecular surfaces, volumes, electrostatic energies, solvation energies, number of atoms, number of hydrogen donors or acceptors and number of rigid bonds) for the gp120 CD4-binding sites structures in the unliganded state from a reference panel of 60 diverse strains of HIV-1. We identified the descriptors that varied significantly between the strains, the outliner strains for each descriptor set and the possible correlations between the descriptors. Our results improve the knowledge about gp120, its molecular and possible neutralization properties.

Keywords::

• HIV-1 strains
• gp120
• CD4-binding site
• homology modelling
• pharmacological descriptors

Introduction

Entry of human immunodeficiency virus type 1 (HIV-1) into the T-helper lymphocytes is mediated by its envelope glycoproteins: gp120 and gp41, named according to their molecular mass and coded by the same viral gene (Env). The structure of HIV-1 gp120 glycoprotein is comprised of five variable regions (V1–V5) interposed among more conserved regions (C1-C5) that form the protein core region. The variable regions form exposed loops anchored at their bases by disulphide bonds¹. Gp120 determines the HIV tropism through a very strong and specific binding to the CD4 receptor present mainly on T-helper lymphocytes surface. Gp120 binding to CD4 is the first step for virus internalization. Binding to CD4 induces ample conformational changes in gp120 which allows it to bind one of the chemokine receptors with co-receptor role (CCR5 or CXCR4)².

The development of a vaccine is considered the only viable long-term solution to the AIDS pandemic. Given the huge variability of the virus, it is widely believed that the induction of a broadly neutralizing antibody response will be crucial for the success of a vaccine against the HIV. The development of an immunogen that elicits it still remains elusive³. Any rational attempt to design anti-HIV inhibitors or vaccines must therefore involve the characterization of the two significantly different conformational states of gp120: prior to CD4 binding and CD4-bound.

The variable regions of HIV-1 elicit a high abundance of strain-specific antibodies, whereas antibodies that target conserved regions are rare and only develop in a subset of individuals⁴. Therefore, only a few antibody species have been identified, so far, as effective in neutralizing primary HIV-1 isolates. The CD4-binding site is included in the C2-C5 conserved regions of gp120 and elicits a few known broadly neutralizing antibodies: B12⁴,⁵, HJ16 and VRC01³,⁶.

In the prefusion state, as revealed by a fully glycosylated simian immunodeficiency virus (SIV) gp120 core structure, gp120 presents a distinct antigenic surface than that from the bound structures². Comparison of the new structure of unliganded gp120 with that of the CD4-bound protein shows that part of the molecule (the “inner domain,” which includes most of the CD4 binding site) undergoes extensive conformational rearrangement upon receptor binding. Compared to the CD4-bound state⁷, its inner domain does not shift as a rigid body relative to the almost unchanged “outer domain,” but rather a few of its distinct regions move independently relative to each other².

The 60 strain panel whose pharmacological descriptors we modelled in the present study contains a combination of sensitive and resistant isolates and of CCR5 or CXCR4. The early transmitted CCR5 viruses are arguably the key for HIV-1 vaccine development and in the assessment of vaccine immunogenicity. Most of the isolates obtained for the panel were non-syncytium inducing (NSI) and utilize CCR5 as a coreceptor for viral entry. However, syncytium-inducing, CXCR4-utilizing viruses were also included in the panel⁸.

Each group of 10 sequences is included in its respective clade or circulating recombinant form (CRF) with a bootstrap value of 100%. The strains within each clade or CRF were typical of within-clade diversity and were without significant associations. The diversity of strains within each clade was similar to that observed with the complete genome sequences. The genetic analysis of the virus panel established that 40 strains from clades A through D were nonrecombinant and could be assigned to their respective clades, representing diversity within that clade with high confidence. The CRF01_AE and CRF02_AG groups strains likewise faithfully represented their respective CRF and mirrored the diversity recognized for these CRF, currently low for CRF01_AE and higher for CRF02_AG. The 60 gp120 sequences provide an envelope sequence proximal to the virus stock for neutralization analysis⁸.

A previous study proved 3D-QSAR molecular shape analysis (MSA) technique to be effective for the study of the inhibitory activity of styrylquinoline derivatives on HIV-1 integrase. The best model obtained from stepwise regression and GFA techniques shows 51.6% predicted variance (leave-one-out) and 57.3% explained variance. These models show the importance of several sterical descriptors, like total polar surface area, relative polar surface area, relative hydrophobic surface area and relative positive charge, for the design and characterization of potent HIV-1 integrase inhibitors⁹.

The volume of a selected group in a protein describes the space occupied by the group and therefore is a sterical descriptor of that group¹⁰. Protein accessible surface area is a sterical descriptor of proteins that has possible relationships to solvent excluded volume and thus to hydrophobic interaction free energy and transfer free energy of solute molecules. The surface is affected by the molecule’s conformation, but is independent of atomic charges¹¹.

The aim our study is to model through homology modelling the structures for gp120 CD4-binding sites in the unliganded form corresponding to the 60 representative strains of HIV-1, to calculate for them a broad array of pharmacological descriptors (molecular volume and solvent accessible area, hydrogen donor/acceptor bond character, number of hydrophobic centres, hydrophobic/polar surface area, number of flexible/rigid bonds, internal dipole moment) and then to identify the descriptors that present significant variation between the strains.

Materials and methods

We modelled the structures of gp120 and calculated the pharmacological descriptors of their CD4bs for the collection of 60 HIV-1 isolates that includes strains with very different infectivity and neutralization properties (see Table 1)⁸. The sequences for the gp120 corresponding to the 60 strains were extracted from Uniprot¹² or GenBank¹³. If both the full-length and the envelope (Env) sequences of a strain were known, we modelled the Env sequence. The 60 all-atom models of free HIV-1 strains gp120 were constructed based on the free structure of the highly related SIV gp120 structure (pdb code: 2BF1)². Regions missing from SIV structure were modelled using as template corresponding regions from available CD4-bound gp120 structures¹⁴,¹⁵. Thus, the V3 loop and the 205–216 region were built using as template a V3-loop containing CD4-bound HIV gp120 structure (pdb code: 2B4C)¹⁴, while the C5 segment of gp120 was modelled using as template the NMR solution structure of HIV gp120 C5 in 40% trifluoroethanol (pdb code: 1MEQ)¹⁵. The V1/V2 loop and the C1 region are distant to the CD4 binding site and were not included in the models.

Table 1.  The strain codes, clades, sequences, TCID50 values (median tissue culture infective dose) and percentages of neutralization in soluble CD4 (sCD4) (10 µg/mL) of the CD4-binding sites for the 60 gp120 strain panel considered.

S. no	Isolate	Strain code	Clade	Gp120 CD4-binding site sequence	TCID50	sCD4
1	92UG_029	AY713407	A	PLCVCTDNAKTSGGDEISRMWQRVGQVRDGGVGRGGGNMRD (123–126, 194, 196, 277–281, 362–365, 367, 368, 372, 419–426, 449–455, 465, 467–473)	33792	100
2	93RW_024	AY713406	A	PLCVCTNNAKTSGGDEISRMWQRTGQTRDGGNDRGGGDMRD (123–126, 198, 200, 281–285, 365–368, 370, 371, 375, 424–431, 454–460, 468, 470–476)	192	61
3	00KE_KER2008	AY736809	A	PLCVCTDNTKNSGGDEIMNMWQRTGQTRDGGNSRGGGDMRD (123–126, 203, 205, 286–290, 368–371, 373, 374, 378, 427–434, 457–463, 470, 472–478)	6947	2
4	00KE_KER2018	AY736810	A	PLCVCTDNAKTSGGDEISNMWQRVGQTRDGGNTRGGGDMRD (123–126, 192, 194, 275–279, 358–361, 363, 364, 368, 416–423, 446–452, 463, 465–471)	106	0
5	99KE_KNH1088	AY736811	A	PLCVCTDNAKTSGGDEISNMWQRTGQTRDGGNGRGGGNMKD (123–126, 200, 202, 283–287, 366–369, 371, 372, 376, 421–428, 451–457, 467, 469–475)	2172	0
6	99KE_KNH1135	AY736814	A	PLCVCTDNAKSSGGDEITRMWQRTGQTRDGGNTRGGGDMRD (123–126, 199, 201, 282–286, 365–368, 370, 371, 375, 425–432, 455–461, 469, 471–477)	874	14
7	00KE_KNH1144	AY736812	A	PLCVCTNNAKTSGGDEISNMWQRTGQTRDGGEGRGGGDMRD (123–126, 198, 200, 281–285, 365–368, 370, 371, 375, 425–432, 455–461, 469, 471–477)	874	18
8	00KE_KNH1207	AY736815	A	PLCVCTNNAKTSGGDEISNMWQRTGQTRDGGVGRGGGDMRD (123–126, 195, 197, 278–282, 362–365, 367, 368, 372, 418–425, 448–454, 462, 464–470)	905	1
9	00KE_KNH1209	AY736813	A	PLCVCTNNAKNSGGDEISNMWQRTGQTRDGGNGRGGGDMRD (123–126, 197, 199, 280–284, 364–367, 369, 370, 374, 422–429, 452–458, 466, 468–474)	192	12
10	00KE_KSM4030	AY736816	A	PLCVCTNNAKNSGGDEISNMWQRVGQTRDGGNTRGGGDMRD (123–126, 188, 190, 271–275, 354–357, 359, 360, 364, 411–418, 441–447, 455, 457–463)	309	0
11	94US_33931N	AY713410	B	PLCVCNNNAKTSGGDEISNRWQEVGKTRDGGKNRGGGDMKD (123–126, 192, 194, 275–279, 359–362, 364, 365, 369, 416–423, 446–452, 462, 464–470)	1748	55
12	90US_873	AY713412	B	PLCVCTNNAKTLGGDEISNMWQGVGKTRDGGNKRGGGDMRD (123–126, 195, 197, 278–282, 362–365, 367, 368, 372, 421–428, 451–457, 467, 469–475)	57501	17
13	85US_Ba-L	AY713409	B	PLCVCTNNAKISGGDEISNMWQEVGRTRDGGPERGGGDMRD (123–126, 199, 201, 282–286, 366–369, 371, 372, 376, 422–429, 452–458, 466, 468–474)	55938	95
14	90TH_BK132	AY736821	B	PLCVCTDNAKTSGGDEISNMWQEVGKTRDGGNQRGGGNMRD (123–126, 194, 196, 277–281, 361–364, 366, 367, 371, 416–423, 446–452, 462, 464–470)	1386	85
15	92FR_BX08	AY713411	B	PLCVCTNNAKTSGGDEISNMWQEVGKTRDGGSDRGGGDMRD (123–126, 188, 190, 271–275, 355–358, 360, 361, 365, 410–417, 440–446, 456, 458–464)	102828	5
16	89BR_BZ167	AY736820	B	PLCVCNDNAKTSGGDEISNMWQEVGKTRDGGNNRGGGDMRD (123–126, 200, 202, 283–287, 368–371, 373, 374, 378, 425–432, 455–461, 469, 471–477)	19302	89
17	84US_MNp	AY736819	B	PLCVCTDNAKTSGGDEISNMWQKVGKTRDGGKDRGGGDMRD (123–126, 201, 203, 284–288, 368–371, 373, 374, 378, 425–432, 455–461, 471, 473–479)	155	89
18	96TH_NP1538	AY713408	B	PLCVCTNNAKVSGGDEVSNMWQEVGKTRDGGNGRGGGDMRD (123–126, 187, 189, 270–274, 354–357, 359, 360, 364, 410–417, 440–446, 459, 461–467)	3920	14
19	90US_1	AY736817	B	PLCVCTNNARTSGGDEISNMWQEVGKTRDGGNKRGGGDMRD (123–126, 197, 199, 280–284, 364–367, 369, 370, 374, 420–427, 450–456, 466, 468–474)	7075	93
20	90US_4	AY736818	B	PLCVCTDNARTSGGDEISNMWQEVGKTRDGGTNRGGGNMKD (123–126, 212, 214, 295–299, 380–383, 385, 386, 390, 436–443, 466–472, 483, 485–491)	768	88
21	93IN_20635-4	AY713414	C	PLCVCTDNVKTSGGDEISNMWQEVGRVRDGGTERGGGDMRN (123–126, 195, 197, 278–282, 362–365, 367, 368, 372, 416–423, 446–452, 460, 462–468)	11841	96
22	98US_MSC5016	AY736822	C	PLCVCTNNAKTSGGDEISNMWQEVGRTRDGGTNRGGGDMRD (123–126, 200, 202, 283–287, 367–370, 372, 373, 377, 426–433, 456–462, 471, 473–479)	27877	75
23	93MW_965	AY713413	C	PLCVCTDNVKTSGGDEISNMWQGVGRTRDGGENRGGGDMKD (122–125, 196, 198, 279–283, 363–366, 368, 369, 373, 415–423, 446–452, 458, 460–466)	13033	44
24	90SE_364	AY833059	C	PLCVCTNNAKTSGGDEISNMWQGVGRTRDGGIDRGGGDMRD (123–126, 196, 198, 279–283, 362–365, 367, 368, 372, 410–417, 440–446, 454, 456–462)	71329	79
25	89SM_145	AY833060	C	PLCVCTNNAKTSGGDEISNMWQEVGQLRDGGNDRGGGDMRD (123–126, 197, 199, 280–284, 363–366, 368, 369, 373, 417–424, 447–453, 467, 469–475)	490	15
26	99ET_14	AY736823	C	PLCVCTNNAKISGGDEISNMWQEVGRTRDGGIIRGGGDMRD (122–125, 192, 194, 275–279, 359–362, 364, 365, 369, 407–414, 437–443, 450, 452–458)	362	14
27	02ET_288	AY713417	C	PLCVCTNNAKISGGDEISNMWQEVGRTRDGGMNRGGGDMRD (123–126, 204, 206, 287–291, 371–374, 376, 377, 381, 419–426, 449–455, 464, 466–472)	1349	37
28	01TZ_BD911	AY736826	C	PLCVCTNNVNTSGGDEISNMWQEVGRVRDGGGERGGGDMRD (126–129, 200, 202, 283–287, 366–369, 371, 372, 376, 421–428, 451–457, 465, 467–473)	18380	26
29	01TZ_A125	AY736824	C	PLCVCANNAKTVGGDEISNMWQRVGRTWDGGNNRGGGDMRG (123–126, 202, 204, 285–289, 370–373, 375, 376, 380, 426–433, 456–462, 470, 472–478)	426	6
30	01TZ_A246	AY736825	C	PLCVCTDNTKTSGGDEISNMWQGVGRTWDGGSGRGGGNMRD (123–126, 201, 203, 284–288, 369–372, 374, 375, 379, 427–434, 457–463, 471, 473–479)	206	9
31	98UG_57128	AY736829	D	PLCVCTNNAKISGGDEISNMWQGVGKTRDGGGARGGGDMRD (124–127, 207, 209, 290–294, 374–377, 379, 380, 384, 435–442, 465–471, 481, 483–489)	7769	25
32	93UG_065	AY713418	D	PLCVCTNNAKNSGGDEISNMWQGVGKTRDGGGNRGGGDMRD (123–126, 201, 203, 284–288, 369–372, 374, 375, 379, 426–433, 456–462, 471, 473–479)	309	12
33	99UG_A03349	AY736834	D	PLCVCTNNAKISGGDEISNMWQGVGKTRDGGGNRGGGDMRD (123–126, 191, 193, 274–278, 357–360, 362, 363, 367, 409–416, 439–445, 454, 456–462)	4345	61
34	99UG_A07412	AY736828	D	PLCVCTNNAKTSGGDEISNMWQGVGKTRDGGNTRGGGDMRD (123–126, 200, 202, 283–287, 367–370, 372, 373, 377, 421–428, 451–457, 465, 467–473)	11088	68
35	99UG_A08483	AY736827	D	PLCVCTNNAKISGGDEISNTWQGVGRTRDGGTNRGGGDMRD (123–126, 202, 204, 285–289, 370–373, 375, 376, 380, 427–434, 457–463, 472, 474–480)	33492	30
36	99UG_J32228	AY736831	D	PLCVCTNNAKISGGDEISNMWQGVGKTRDGGADRGGGDMRD (123–126, 187, 189, 270–274, 353–356, 358, 359, 363, 407–414, 437–443, 452, 454–460)	11089	30
37	99UG_D26830	AY736830	D	PLCVCTDNAKISGGDEITNLWQKVGRTRDGGHNRGGGDMRN (124–127, 194, 196, 277–281, 361–364, 366, 367, 371, 416–423, 446–452, 461, 463–469)	309	76
38	99UG_E08364	AY736832	D	PLCVCTNNVKNSGGDETSNMWQGVGRTRDGGNNRIGGDMRD (123–126, 187, 189, 270–274, 354–357, 359, 360, 364, 404–411, 434–440, 448, 450–456)	48	38
39	99UG_E13613	AY736833	D	PLCVCTNNAKTSGGDEISNMWQGVGTTRDGGADRGGGDMRD (122–125, 185, 187, 268–272, 351–354, 356, 357, 361, 408–415, 438–444, 454, 456–462)	693	74
40	00KE_NKU3006	AY736835	D	PLCVCTNNAKTSGGDEISNMWQEVGKTRDGGNTRGGGDMRD (123–126, 209, 211, 292–296, 375–378, 380, 381, 385, 435–442, 465–471, 479, 481–487)	23412	18
41	96TH_M02138	AY713424	A/E	PLCVCTNNAKTSGGDEVHNMWQETGQTRDGGINRGGGNMKD (123–126, 198, 200, 281–285, 364–367, 369, 370, 374, 418–425, 448–454, 462, 464–470)	464	10
42	90TH_CM235	AY736837	A/E	PLCVCTNNAKTSGGDEIHNMWQGAGQTRDGGAIRGGGNIKD (123–126, 200, 202, 283–287, 367–370, 372, 373, 377, 421–428, 451–457, 465, 467–473)	2896	1
43	90TH_CM240	AY736838	A/E	PLCVCTNNAKTSGGDEIHNMWQGAGQTRDGGVNRGGGNIKD (123–126, 199, 201, 282–286, 365–368, 370, 371, 375, 419–426, 449–455, 463, 465–471)	288	14
44	90TH_CM244	AY713425	A/E	PLCVCTNNAKTSGGDEIHNMWQGAGQTRDGGATRGGGNIKD (123–126, 204, 206, 287–291, 371–374, 376, 377, 381, 425–432, 455–461, 470, 472–478)	66048	2
45	96TH_NI1046	AY713421	A/E	PLCVCTNNAKTSGGDEIHNMWQEAGQTRDGGNNRGGGNMKD (123–126, 198, 200, 281–285, 365–368, 370, 371, 375, 425–432, 455–461, 469, 471–477)	724	11
46	99TH_NI1052	AY713423	A/E	PLCVCTDNAKTSGGDEIHNMWQGVGQTRDGGNTREGGNIKD (123–126, 215, 217, 298–302, 382–385, 387, 388, 392, 435–442, 465–471, 480, 482–488)	490	0
47	96TH_NI1049	AF070707	A/E	PLCVCTNNAKTSGGDEIHNMWQGAGQTRDGGANRGGGDIRD (123–126, 194, 196, 277–281, 361–364, 366, 367, 371, 415–422, 445–451, 459, 461–467)	327	19
48	98TH_N91251	AY713422	A/E	PLCVCTNNAKTSGGDEIHNMWQRAGQTRDGGGAREGGNIKD (123–126, 187, 189, 270–274, 354–357, 359, 360, 364, 408–415, 438–444, 454, 456–462)	25219	25
49	97TH_NP1525	AY713420	A/E	PLCVCTNNAKTSGGDEIHNMWQGVGQTRDGGDNRGGGNTKD (123–126, 201, 203, 284–288, 368–371, 373, 374, 378, 428–435, 458–464, 474, 476–482)	1629	22
50	97TH_NP1695	AY713419	A/E	PLCVCTNNAKTSGGDEIHNMWQGVGQTRDGGNNRGGGNIKD (123–126, 202, 204, 285–289, 368–371, 373, 374, 378, 423–430, 453–459, 466, 468–474)	1448	4
51	98US_MSC5007	AY736840	A/G	PLCVCTNNAKTSGGDEISNMWQKVGQTRDGGNNRGGGDMRD (121–124, 189, 191, 272–276, 357–360, 362, 363, 367, 415–422, 445–451, 459, 461–467)	1024	20
52	01CM_0002BBY	AY736843	A/G	PLCVCTNNAKTSGGDEISNMWQRVGQTRDGGNNREGGDMRN (122–125, 197, 199, 280–284, 364–367, 369, 370, 374, 420–427, 450–456, 465, 467–473)	38	7
53	01CM_0005BBY	AY736844	A/G	PLCVCTNNAKTSGGDEISNLWQKVGQTRDGGVNRGGGDMRD (122–125, 193, 195, 276–280, 360–363, 365, 366, 370, 413–420, 443–449, 458, 460–466)	53032	4
54	01CM_0008BBY	AY736845	A/G	PLCVCTNNAKNSGGDEISNMWQRVGQTRDGGNNRGGGDMRD (122–125, 191, 193, 274–278, 358–361, 363, 364, 368, 412–419, 442–448, 457, 459–465)	46092	8
55	02CM_0013BBY	AY736846	A/G	PLCVCTNNAKTSGGDEIMNMWQRVGQTRDGGNNRGGGDMRD (122–125, 200, 202, 283–287, 367–370, 372, 373, 377, 426–433, 456–462, 469, 471–477)	35502	77
56	02CM_0014BBY	AY736847	A/G	PLCVCTNNAKTSGGDEISNMWQKVGQTRDGGNGRGGGNMKD (120–123, 190, 192, 273–277, 358–361, 363, 364, 368, 416–423, 446–452, 465, 467–473)	29489	30
57	02CM_0015BBY	AY736848	A/G	PLCVCTNNAKTSGGDEISNMWQKVGQTRDGGNNRGGGDMRD (122–125, 199, 201, 282–286, 365–368, 370, 371, 375, 424–431, 454–460, 468, 470–476)	4370	0
58	01CM_1475MV	AY736842	A/G	PLCVCTNNAKTSGGDEISRMWQRVGQTRDGGNNRGGGDMRD (122–125, 194, 196, 277–281, 361–364, 366, 367, 371, 418–425, 448–454, 462, 464–470)	437	10
59	02CM_1970LE	AY736841	A/G	PLCVCTNNAKISGGDEISNMWQRVGQTRDGGNNRGGGDMRD (122–125, 184, 186, 267–271, 351–354, 356, 357, 361, 409–416, 439–445, 453, 455–461)	1024	30
60	91FR_DJ263	AY736839	A/G	PLCVCTNNAKNSGGDEISNMWQKVGLTRDGGSNRGGGDMRD (122–125, 204, 206, 287–291, 370–373, 375, 376, 380, 435–442, 465–471, 479, 481–487)	5803	33

The numbering of the strains (S. no) shown here will be further used in plotting the descriptor values versus strain number (Figures 2 and 3). In the clades column (Cl.) the circulating recombinant form CRF01_AE and CRF01_AE are noted A/E and A/G respectively. The CD4bs sequences of each strain, comprising of 41 amino acids, are followed by the amino acids position in the whole gp120 sequence (written between brackets). The localization of these residues in the sequence of strain AY713410 is presented in Figure 1. The differences between the each of the 60 CD4bs sequences and the consensus sequence (PLCVCTNNAKTSGGDEISNMWQGVGQTRDGGNNRGGGDMRD) are highlighted by using bold, underlined characters.

The alignment and sequence analysis (including the CD4bs consensus sequence of the 60 strains, see Table 1) was done with Jalview 2.7 software¹⁶. Homologous Cys involved in disulfide bonds formation were preferentially aligned. The sequence alignment between one of the 60 HIV-1 gp120 strains considered (AY713410), the main SIV gp120 template (PDB ID: 2BF1) and the other templates used is presented in Figure 1. The model was built using a homology modelling method and optimized by short molecular dynamics simulations using MODELLER 9v3 software¹⁷.

Figure 1.  Homology modelling of free gp120 corresponding to strain code AY71341 (sequence number 11 out of the 60 strains considered). The 41 amino-acids of the CD4-binding site are bolded and underlined. The alignment follows Chen et al.².

The CD4 binding site (CD4bs) was defined in a previous study¹⁸ as the residues being in contact with CD4 in the bound gp120 structure along with residue S375 (the disposition of these residues through gp120 sequence for the AY713410 strain of the panel is presented in Figure 1). Even if S375 is more than 5 Å apart from CD4, it was included in the binding site, as the S375W mutation is known to partially stabilize the CD4-bound state¹⁹. CD4bs includes a total of 41 amino-acids that can be grouped in six segments from the C2, C3, C4, C5 conserved regions of the protein. The corresponding positions in the HIV-1 HXB2 reference strain are: 124–127, 196, 198, 279–283, 365–368, 370, 371, 375, 425–432, 455–461, 469 and 471–477. The regions don’t present insertions or deletions among the 60 strain-panel considered. For our analysis of protein surfaces, we chose all the 41 amino acids of CD4bs.

We calculated the descriptors for all the 41 amino-acids of CD4bs. The analysis of structures was performed using CHARMM c34b1²⁰ and VMD 1.8.7²¹ software. The electrostatic surface potential energy (U_E) of solvated gp120 CD4 binding site regions were calculated using the PBEQ module²² implemented in CHARMM²⁰ using a grid centred in the centre of the molecule, with a minimum distance between the solute and the grid boundary of 10 Å and a coarse grid spacing of 1.5 Å before focusing and 1 Å after focusing. For the dielectric boundary, we considered the molecular contact and reentrant surface created using the water radius (1.4 Å) and the atomic radii of protein atoms from the Charmm27 force field²³. We performed calculations using a protein dielectric constant of 1, a bulk-solvent dielectric constant of 80 and a salt concentration of 0.15 M/L, at 310 K. The solvation free energy of the binding site regions were calculated as the difference between the electrostatic surface potential energy of the structure in water and the electrostatic surface potential energy in a reference environment, with a dielectric constant of 1.

The number of hydrophobic and polar atoms, and of flexible and rigid bonds CD4-binding site region were calculated with CHARMM²⁰. We considered as rigid bonds the bonds within aromatic (Phe, Tyr, Trp) and non-aromatic (Pro) amino-acid cycles. We considered as hydrophobic atoms the carbon and hydrogen atoms and as polar atoms the oxygen, nitrogen, and sulfur atoms.

The solvent accessible surface (SAS) of gp120 CD4-binding site is important in defining the hydrophobic term of the interaction with CD4. We used CHARMM²⁰ to calculate SAS for the 41-amino acid CD4-binding site regions by rolling a probe radius of 1.4 Å along the target surface of the binding site. We thus measured for all the 60 strains the SAS of all atoms, of the positively and the negatively-charged atoms, as well as of the hydrophobic and the polar atoms (as defined above).

Results

The studied 60 strain panel data is presented in Table 1 (the codes, clades CD4-binding site sequences). For reference, Table 1 includes two measured biological properties⁸: the viral 50% toxicity citotoxic concentration (TCID50) and the percentage of virus neutralized by a 10 µg/mL soluble CD4 (sCD4). The rank of the 60 models of gp120 strain variants is the same as in the original panel⁸. The CD4 binding site considered in our study (41 amino-acids) has a global identity conservation of ~84%, with 17 identical amino-acid positions in all the strains and 24 positions with at least one amino-acid variation. We calculated three types of pharmacological descriptors for the 60-strain panel: sterical, energetical and structural descriptors.

Sterical descriptors

We calculated solvent accessible surface (SAS) for all atoms, for positive and negative atoms and for hydrophobic and polar atoms, and the total volume. We also calculated the total dipole moment and its components on the X, Y and Z axes (see Table 2). We analyzed the variation of the calculated molecular descriptors that characterize the scale of their variations compared to the median values.

Table 2.  Significant structural, energetical and sterical descriptors of the CD4-binding sites for the 60 gp120 strains and their statistics.

Strain	SAS of positive atoms (Å^2)	SAS ofnegative atoms (Å^2)	SAS of hydrophobic atoms (Å^2)	SAS of polar atoms (Å^2)	Solvation free energy (kcal/mol)	UE in vacuum (kcal/mol)	Total dipole (debyes)	Number of H-bond donor atoms
AY713407	755.80	1435.66	1219.73	973.18	−134.63	141.28	133.14	81
AY713406	690.74	1403.23	1021.70	1074.50	−138.33	143.26	166.58	85
AY736809	712.34	1511.98	1078.16	1148.93	−221.74	225.60	108.64	82
AY736810	697.24	1409.25	1115.62	992.78	−35.09	41.15	87.33	80
AY736811	678.67	1341.25	1012.59	1009.45	−91.09	96.63	122.70	80
AY736814	656.39	1371.05	1003.58	1025.20	−59.07	67.18	94.24	84
AY736812	650.36	1290.22	958.69	985.13	−55.67	55.44	156.58	80
AY736815	706.29	1408.77	1100.87	1015.69	−79.64	86.86	151.18	80
AY736813	728.15	1535.55	1132.63	1133.56	−108.78	111.16	168.83	83
AY736816	721.36	1441.85	1201.20	964.25	−176.75	181.11	156.83	83
AY713410	813.65	1369.31	1179.26	1004.39	−124.76	131.86	228.95	84
AY713412	694.14	1383.35	1120.40	958.75	−115.33	126.42	162.85	79
AY713409	777.51	1468.44	1254.34	1007.16	−89.09	92.50	177.83	75
AY736821	712.61	1390.65	1087.74	1020.89	−114.18	127.13	180.66	79
AY713411	778.42	1337.44	1119.78	999.34	−48.04	55.17	173.17	76
AY736820	707.39	1386.79	1155.50	941.07	−71.62	75.86	145.66	78
AY736819	700.07	1291.07	1096.83	903.57	−126.46	128.86	119.49	79
AY713408	699.97	1258.53	1123.68	839.13	−101.83	105.72	168.39	76
AY736817	753.33	1347.82	1120.84	982.29	−45.88	50.11	215.95	82
AY736818	723.72	1328.97	1128.50	925.35	−90.28	96.72	158.15	78
AY713414	744.36	1426.92	1130.18	1049.83	11.97	−12.75	244.05	77
AY736822	764.04	1461.23	1170.58	1056.13	−27.57	28.78	175.68	80
AY713413	751.17	1695.97	1174.68	1275.96	−69.65	69.26	204.99	77
AY833059	769.24	1428.13	1207.33	993.95	−14.06	17.09	160.12	77
AY833060	716.78	1376.04	1149.01	950.57	−22.69	26.63	263.53	75
AY736823	805.47	1439.43	1310.25	937.87	−6.97	3.00	143.69	76
AY713417	802.06	1481.91	1261.78	1026.50	−33.30	36.81	186.45	78
AY736826	736.52	1465.10	1185.39	1016.52	23.31	−26.73	195.89	75
AY736824	697.64	1424.26	1125.21	1001.04	−42.28	44.72	125.16	80
AY736825	709.96	1419.89	1140.64	993.75	−55.07	54.51	215.59	75
AY736829	788.07	1370.94	1165.63	994.79	−109.44	114.83	130.67	74
AY713418	754.40	1394.15	1160.41	991.24	−150.30	156.10	117.42	78
AY736834	695.04	1397.88	1113.74	982.02	−99.21	111.93	95.30	76
AY736828	857.41	1380.50	1266.69	974.46	−54.34	60.25	148.49	78
AY736827	732.15	1470.85	1156.78	1053.00	−132.93	141.18	97.26	80
AY736831	727.36	1337.05	1140.51	925.28	−131.97	139.88	162.79	74
AY736830	810.56	1400.76	1248.93	965.34	−97.55	107.20	98.73	82
AY736832	690.54	1463.77	1116.53	1043.13	−77.77	83.44	118.95	83
AY736833	681.78	1473.20	1107.23	1052.78	−78.97	86.36	189.81	73
AY736835	801.33	1421.42	1227.47	997.74	−102.82	107.64	192.50	78
AY713424	711.43	1425.76	1088.93	1050.60	−120.07	125.22	259.80	77
AY736837	740.08	1291.04	1150.37	882.58	−69.50	74.95	214.20	74
AY736838	755.64	1342.59	1161.51	938.28	−75.54	83.87	238.78	76
AY713425	631.48	1247.97	962.23	919.98	−42.85	47.34	230.68	75
AY713421	648.01	1438.78	1023.01	1065.56	−93.95	102.28	193.84	78
AY713423	697.34	1341.38	1036.84	1003.69	−116.70	118.87	172.21	75
AF070707	700.28	1461.25	1095.67	1069.48	−35.36	34.52	248.91	76
AY713422	635.19	1273.35	1010.71	900.07	−103.34	102.86	191.16	79
AY713420	652.60	1322.23	995.86	979.75	−81.37	88.69	222.19	77
AY713419	722.62	1391.75	1132.82	983.56	−151.32	163.62	202.67	78
AY736840	682.05	1389.20	1076.87	1002.15	−74.02	79.20	144.13	81
AY736843	688.17	1413.28	1102.87	1006.72	−122.04	124.86	179.85	85
AY736844	769.72	1339.15	1200.17	909.38	−57.79	61.81	186.59	79
AY736845	722.93	1538.81	1202.02	1060.06	−58.40	64.57	157.25	84
AY736846	715.06	1510.28	1095.18	1132.44	−137.72	143.97	156.04	82
AY736847	702.75	1371.71	1122.78	959.55	−117.58	125.42	143.97	79
AY736848	651.95	1411.73	1052.30	1014.91	−137.01	148.10	167.09	81
AY736842	694.86	1421.14	1049.84	1067.65	−117.59	126.38	80.33	86
AY736841	738.73	1530.61	1158.96	1112.74	−83.71	81.74	146.11	82
AY736839	828.32	1413.30	1259.24	986.19	−113.43	119.07	149.90	79
Mean	724.72	1405.27	1129.48	1003.86	−86.82	91.79	167.17	78.88
Standard deviation	49.41	76.96	76.86	70.62	45.93	47.81	44.23	3.18
Minimum	631.48	1247.97	958.69	839.13	−221.74	−26.73	80.33	73
Median	715.92	1406.00	1126.86	998.54	−89.68	94.57	164.72	79
Maximum	857.41	1695.97	1310.25	1275.96	23.31	225.60	263.52	86

Solvent accessible surface of positive atoms (see Figure 2a) has as the five lowest outliners the strains AY713425 (631.48 Ų) (lowest), AY713422, AY713421, AY736812 and AY736848 and as the five highest outliers the strains AY736823, AY736830, AY713410, AY736839 and AY736828 (857.41 Ų) (highest). AY713425 also presents the lowest SAS of negative atoms (see below), the lowest total SAS (data not shown) and the second lowest SAS of hydrophobic atoms. The presence in AY713425 strains’ CD4bs of small amino-acids in several variable positions (one treonine and two alanine residues instead of a valine and two arginine residues present in most strains) is in good correlation with its small surfaces.

Figure 2.  Variation of significant structural descriptors of the CD4-binding sites for the 60 gp120 strains. (a) SAS of positive atoms; (b) SAS of and negative atoms; (c) SAS of hydrophobic atoms; (d) SAS of polar atoms. The mean values and the standard deviations are noted in the higher right corner. The Y-axes are adjusted for a clearer representation of the descriptor variation.

Solvent accessible surface of negative atoms (see Figure 2b) has as the five lowest outliners the strains AY713425 (1247.97 Ų) (lowest), AY713408, AY713422, AY736812 and AY736837, and as the five highest outliers the strains AY736809, AY736841, AY736813, AY736845 and AY713413 (1695.97 Ų) (highest).

Solvent accessible surface of hydrophobic atoms (see Figure 2c) has as the five lowest outliners the strains AY736812 (958.69 Ų) (lowest), AY713425, AY713420, AY736814 and AY713422, and as the five highest outliers the strains AY713409, AY736839, AY713417, AY736828, AY736823 (1310.25 Ų) (highest). As can be seen from the above values, the recorded changes are significant.

Solvent accessible surface of polar atoms (see Figure 2d) has as the five lowest outliners the strains AY713408 (839.13 Ų) (lowest), AY736837, AY713422, AY736819 and AY736844, and as the five highest outliers the strains AY736841, AY736846, AY736813, AY736809, and AY713413 (1275.96 Ų) (highest). The recorded changes are therefore significant.

The total dipole moment (see Figure 3a) has as the five lowest outliners the strains AY736842 (80.33 debyes) (lowest), AY736810, AY736814, AY736834 and AY736827, and as the five highest outliers the strains AY736838, AY713414, AF070707, AY713424 and AY833060 (263.52 debyes) (highest), with very significant changes between strains. The highest dipole moment presented by strain AY833060 is associated to the additional presence of two acidic residues compared to other strains (one aspartic acid and one glutamic acid residues instead of glycine and asparagine in the consensus sequence) (see Table 1).

Figure 3.  Variation of significant energetical and sterical descriptors of the CD4-binding sites for the 60 gp120 strains. (a) free solvation energy; (b) electrostatic surface potential energy (UE) in vacuum; (c) total dipole moment; (d) the number of hydrogen-bond donors. The mean values and the standard deviations are noted in the higher right corner. The Y-axes are adjusted for a clearer representation of the descriptor variation.

Energetical descriptors

We measured the solvation free energy, electrostatic surface potential energy in water, electrostatic surface potential energy in vacuum (see Table 2 and Figure 3).

Solvation free energy (see Figure 3b) has as the five lowest outliners the strains AY736809 (−221.74 kcal/mol) (lowest) AY736816, AY713419, AY713418 and AY713406, and as the five highest outliers the strains AY833060, AY833059, AY736823, AY713414 and AY736826 (23.31 kcal/mol) (highest). As we can see, the largest value is positive, indicating a slightly energetically unfavourable solvation for the AY736826 strain’s gp120 CD4-binding site. Unsurprisingly, the AY736826 strain also has the lowest (most favourable) electrostatic potential energy in vacuum (see below). Its CD4bs sequence contains non-polar amino-acids (two additional valine residues and a glycine residue compared to the consensus sequence).

Electrostatic surface potential energy in vacuum (U_E in vacuum) (see Figure 3c) has as the five lowest outliners the strains AY736826 (−26.73 kcal/mol) (lowest) AY713414, AY736823, AY833059 and AY833060, and as the five highest outliers the strains AY736848, AY713418, AY713419, AY736816 and AY736809 (225.6 kcal/mol) (highest).

Structural descriptors

We determined the following structural descriptors: total charge, the number of hydrophobic atoms, number of polar atoms, number of bonds that allow free rotation, number of rigid bonds, number of H-bond donor atoms and H-bond acceptor atoms (see Table 2 and Figure 2).

The number of hydrogen bond donors (see Figure 3d) has as the four lowest outliners the strains AY736833 (73) (lowest) AY736829, AY736837 and AY736831, and as the six highest outliers the strains AY736845, AY736814, AY713410, AY713406, AY736843 and AY736842 (86) (highest). The presence of small non-polar alanine and threonine stands out in the sequence of AY736833 CD4bs, while AY736842 CD4bs sequence is characterized by the presence of two additional asparagine residues compared to the consensus sequence.

Discussion

Descriptors calculation in the CD4 binding sites of gp120 strains has shown that significant variations were obtained for: solvent accessible surface (SAS) of positive atoms (from 631.48 Ų (in AY713425 strain) to 857.41 Ų (in AY736828 strain). The descriptor SAS of negative atoms recorded quite large variations (from 1247.97 Ų, AY713425 to 1695.97 Ų, AY713413). The same strain (AY713413) also presented the largest solvent SAS of polar atoms (which varied from 839.13 Ų for AY713408, to 1275.96 Ų for AY713413), and the largest SAS on all atoms (data not shown). Its CD4bs sequence differences from other strains studied are several bulky polar residues (glutamic acid and lysine).

Significant variations were recorded when the descriptor SAS of hydrophobic atoms (from 958.69 Ų, AY736812 to 1310.25 Ų, AY736823) was considered. AY736823 strain CD4bs also presents the third lowest electrostatic energy in vacuum and solvation free energy, and the lowest electrostatic energy in water (data not shown). The presence of three hydrophobic isoleucine residues stands out compared to the CD4bs consensus sequence.

The solvation free energy and electrostatic surface potential energy in vacuum presents very large and inversely correlated variations (see above and Figure 3b and 3c). The total dipole moment presents large variations among the 60 protein sites (see Figure 3a). The electrostatic surface potential energy in water varied in a very narrow range (but with a standard deviation of more than half the mean value), owing to the shielding effect of water on an electric field²⁴ (data not shown).

Our analysis has shown that among the structural descriptors, the number of hydrogen bond donors presents significant variations (see above and Figure 3d). The variations present in the other sterical descriptors are relatively small and can be considered non-significant (data not shown). Given the constant number of amino-acids of the CD4-binding site (41), and the relatively high overall identity rate (~84%) this fact is explainable despite the numerous mutations present.

Conclusion

Molecular simulation techniques will continue to give important information about vaccine and inhibitor design). We identified eight pharmacological descriptors for the unliganded gp120 CD4-binding sites (SAS of positive and negative atoms, hydrophobic and polar SAS, free solvation energy, electrostatic surface potential energy in vacuum, total dipole moment and number of hydrogen-bond donors) that vary significantly between the strains of a 60 member HIV-1 strain collection. We also identified the outliner strains for each descriptor set and the correlations between the descriptors and sequence for a better calibration of both simulations and experiments. Our results improve the knowledge about gp120 and its weaknesses and are useful together with the data provided by other molecular simulation techniques in the design of broad-spectrum anti-HIV-1 inhibitors and vaccines.

Declaration of interest

The authors acknowledge the financial support of the Romanian Ministry of Education, Research, Youth and Sport through the “IDEAS” grant 137/2011.