The document analyzes the D30N/L90M dual mutation in HIV-1 protease that confers drug resistance. Molecular dynamics simulations were performed on the wild type, D30N, and D30N/L90M variants. Dynamic residue interaction network (dRIN) analysis found that the dual mutation greatly reduced hydrogen bonding between residues near the dimer interface and increased a van der Waals interaction between residue 90M and 25D. This disrupts dimer formation in protease conformational changes, explaining why the dual mutation is rarely observed in vivo.

1. Dynamic Residue Interaction Network
Analysis of D30N/L90M dual mutation in
Protease conferring Drug Resistance in
HIV-1 virus
Ryouga miyawaki | Mohini Yadav | Norifumi Yamamoto
CBI 2021
P02-03(F-3)
Chiba Institute of Technology
26th – 27th October 2021

2. Introduction： HIV
• HIV stands for human immunodeficiency virus and is the virus that
causes AIDS.HIV cannot grow on its own but uses infected T cells.
• At the end of 2020, there will be approximately 37.7 million people
living with HIV.
HIV-1

3. Forget to drink
Introduction：Drug Resistance in HIV-1 virus
Resistant HIV
HIV
Regular dose
Lower limit of effective
blood concentration
Drug
blood
concentration
duration of therapy

4. Introduction： HIV mutation
L90
D30
D25
D25
HIV-1 WT PR PDBID(3oxc)
D25
D25’
L90
D30
D30’
L90’

5. Introduction： HIV-1 protease experimental results
WT
D30N/L90M
1990
770
kcat/Km
[mM
−1
s
−1
]
D30N
960
catalytic efficiency
WT
-54.4±0.2 D30N/L90M
-41.0±0.4
D30N
-48.6±0.4
ΔG
[kJmol
−1
]
Binding energy parameters of nelfinavir binding to HIV-1 protease
Kožíšek et al., J. Mol. Bio. , Vol. 374, pp. 1005-1016 (2007)

6. Introduction： HIV-1 protease experimental results
Perrin et al., J Virol. Sept. 2003, p. 10172–10175

7. This study
To elucidate why HIV-1 Proteases with D30N/L90M dual
mutation are rarely found in vivo
Aim
The effect of D30N/L90M dual mutation on residue
interactions in HIV-1 protease is still unknown
Problem
Perform Dynamic Residue Interaction Network(dRIN)
analysis to elucidate why D30N/L90M dual mutation is
rarely found in vivo
Solution

8. Method: Molecular Dynamics (MD) simulation
Initial Structure PDB ID:2Q64(WT, D30N) 2Q63(D30N/L90M)
Force Field
Protein: ff14SB
Water: TIP3P
Condition NPT ensemble
Length 100 ns
Thermostat Weak-coupling algorithm
Barostat Berendsen algorithm
Software AMBER 20

9. Result：RMSD

10. Result：RMSF
sheet helix loop
Secondary
Structure

11. Method: Residue interaction network (RIN)
• The residue interaction network helps in visualizing
these interactions using simple graph with nodes and
edges.
• Nodes represents the residues and Edges represents
the inter-residue interactions.

12. Method: Residue interaction network (RIN)
Hydrogen bond Van der Waals interactions Disulfide bridges
Salt bridges π-π stacking interactions π-cation interactions
Piovesan et al., Nucleic Acids Res, Vol. 44, pp. W367-374 (2016)

13. Method: Residue interaction network (RIN)
X-Ray or NMR
Single protein structure
Molecular Dynamics (MD) Simulation
Residue Interaction Network
(RIN) analysis
Multiple protein structure
Dynamic Residue Interaction Network
(dRIN) analysis

14. Result： Interaction change of D30N/L90M and D30N
D30N/L90M
D30N/L90M
-D30N
HB VDW
D30N

15. Result： D30N+L90M-D30N(hb)
hb [%] vdw [%]
D30N 59.5 84.5
D30N/L90M 0 0
ASN98 –ASN98’

16. Result： D30N+L90M-D30N(vdw)
hb [%] vdw [%]
D30N 59.5 84.5
D30N/L90M 0 0
ASN98 –ASN98’

17. Result： VMD snapshot of ASN98-ASN98’
D30N D30N/L90M
ASN98
ASN98’
A chain
B chain
Existence probability of interaction of
ASN98 –ASN98’
hb [%] vdw [%]
D30N 59.5 84.5
D30N/L90M 0 0
ASN98’
ASN98’
ASN98
ASN98

18. Result： D30N+L90M-D30N(vdw)
hb [%] vdw [%]
D30N 0 63
D30N/L90M 0 92.5
ASH25’ –LEU/MET90’

19. Result： VMD snapshot of ASH25’ – LEU/MET90’
D30N D30N/L90M
ASH25’
LEU/MET 90’
Existence probability of interaction of ASH25’ – LEU/MET90’
hb [%] vdw [%]
D30N 0 63
D30N/L90M 0 92.5
ASH25’
LEU90’
ASH25’
MET90’

20. Conclusion
This allows us to infer that the 30N/90M mutant has a low
mutant proliferative potential because it does not produce
dimers when a conformational change of flap occurs, such as
a half-open or large open form.
dRIN analysis of D30N/L90M mutant HIV-1 PR
Hydrogen bonding of
amino acid residues near
the dimer interface is
greatly reduced.
A van der Waals
interaction is formed
between the 90M
residue and 25D.

21. BACK

22. BACK：Dynamic Residue Interaction Network (dRIN) graphs
WT
30N
30N-WT
HB VDW ALL

23. BACK：Dynamic Residue Interaction Network (dRIN) graphs
30N
30N/90M
30N/90M-
30N
HB VDW ALL

24. BACK：Flap
Yu et al., RSC Adv., 2017, 7, 45121

25. BACK：
HIV-1 protease experimental results of catalytic
efficiency

26. BACK：Conditions for the Existence of Interactions
Hydrogen bond
Van der Waals interactions
Disulfide bridges
Hydrogen Bond, the distance between acceptor
and donor atom must be less than or equal to 3.5Å
and the angle formed by donor atom, hydrogen
atom and acceptor atom (DHA) must be less than
or equal to 63°.
Disulphide bridges are covalent bonds and the
distance between SG atoms of cysteine pairs must
be less than or equal to 2.5Å.
For Van der Waals interactions, the distance
between the surface of two atoms subtracting
their van der Waals radii must be less than or
equal to 0.5Å.

27. BACK：Conditions for the Existence of Interactions
Salt bridges occurs between residues with
opposite charges and the distance between the
mass centers of the charged groups must be less
than or equal to 4Å.
π-cation interactions occurs between positively
charged amino acids (Arg, Lys) and an aromatic
side chain. The distance between the mass center
of charged group and any atom of the π-system
must be less than 5Å and the angle between the
distance vector and the ring plane has to
guarantee that the mass center of the cation lies
above (or below) the ring area.
π-π stacking interactions occurs between aromatic
residues (His, Tyr, Trp, Phe) and the distance
between the two ring barycenters must be less
than or equal to 6.5Å.
Salt bridges
π-π stacking interactions
π-cation interactions