Page 52 - AN-2-4

Advanced Neurology Genomic insights into Alzheimer

The II serves as an indicator of a protein’s stability, with density in this study ranged from 0.0862 to 0.1832, with
a value ≤40 suggesting stability. In this study, the analysis the V717G mutation displaying the lowest density and
indicated that both the wild-type APP (40.69) and all the Flemish mutation having the highest (Table S1).
mutated APPs exhibited instability, with values ranging Consequently, these findings reveal that all the scores fell
from 40.31 to 41.37. Notably, the Swedish mutation within an acceptable range, affirming the high quality of
demonstrated the lowest II, while the Australian mutation the predicted models for further analysis. The predicted
had the highest II (Table 1). 3D protein structures of both wild-type and mutated APPs
Positive and negative GRAVY values offer insights into are illustrated in Figure 2.
the protein’s hydrophobic versus hydrophilic status. In this I-TASSER identifies the templates using 10 different
study, we observed that both wild-type and mutated APPs template prediction servers and constructs the structure
were hydrophilic. The GRAVY value for wild-type APP was based on the best-scoring template (Table S1). In this
−0.584, followed by the Taiwanese, Tottori, Iowa, London, study, the template 3KTMA was identified as common
and V717L mutations, all sharing the same threshold value among wild-type and all 21 mutations of APPs (Table S1).
of −0.584 (Table 1). In addition, templates 7WB4N (wild-type, 18 out of 20
3.2. Analysis of 3D structures in predicted wild-type mutations [excluding Indiana and M722K]), 6N7PX (wild-
and mutated APPs type, German), 4YNOB (wild-type, 19 out of 20 mutations
[excluding Indiana]), 7WOOD (wild-type, aA673V,
The full-length 3D structures of both wild-type and Arctic, Flemish, French, German, London, Taiwanese,
mutated APPs were not available in PDB. Consequently, M722K, and V717L), 6AAYA (wild-type, Australian,
we employed the I-TASSER server to predict the protein Iberian, A673V, and V717G) were utilized for the 3D
3D structure. I-TASSER predicts the protein structure structure prediction of wild-type and mutated APPs.
and chooses the optimal model based on various criteria, However, 10 templates were found to differ when used for
including the c-score, cluster density, number of decoys, the 3D structure prediction of different mutated APPs,
estimated template modeling (TM)-score, and estimated not including the wild-type APP. These templates were
RMSD score. The c-score typically ranges from −5 to 2, 5IJOJ (Indiana), 51QWQ (Indiana, Austrian, Osaka,
with larger values indicating better quality of the predicted and Taiwanese), 3UMHA (Indiana), 1QBKB (Indiana),
protein model, and a c-score exceeding −1.5 suggests a 5OWVA (Indiana), 3Q7GA (Arctic, Austrian, Flemish,
protein model with accurate global topology . In this Taiwanese), 51GWQ (French and A673V), 6Z91A (Arctic,
[23]
study, the c-score ranged from −0.68 to −1.43, differing for Australian, Iberian, Iranian, London, Swedish, Tottori,
wild-type and each mutation of APP, with the exception of M722K, T719N, V717G, and V717L), 7DRRD (Flemish,
the Tottori and Flemish mutations (−0.68) (Table S1).
Iranian, Lowa, Swedish, Tottori, M722K, T719N, V717L),
The TM-score serves as an indicator of the correct and 7RHQA (Australia, Austrian, Iberian, Iranian, Osaka,
topology of the predicted protein model, with a TM-score Swedish, T719N, and V717G) (Table S1).
>0.5 considered indicative of a good model. In this study,
the lowest TM-score, recorded at 0.54 ± 0.15, was associated 3.3. Evaluation of 3D structures in predicted wild-
with the V717G and V717L mutations, whereas the highest type and mutated APPs
TM-score, measured at 0.63 ± 0.14, was observed for the The evaluation of predicted protein structures was
Flemish and Tottori mutations. The TM-score for wild- performed based on the Ramachandran plot statistics,
type human APP was 0.60 ± 0.14 (Table S1). ERRAT score, ProSA local and overall model quality,
Comparatively, the RMSD score represents the average and Verify3D score. The Ramachandran plot assesses the
distance between residues that are structurally aligned stereochemical quality of the protein model by categorizing
according to the TM-score . In this study, the estimated residues into most favored, allowed, generously allowed,
[23]
lowest RMSD score, measuring 9.9 ± 4.6 Å, was identified and disallowed regions. A concentration of over 90% of
for the Flemish and Tottori mutations, while the highest protein residues in the core area of the Ramachandran
RMSD score, reaching 11.8 ± 4.5 Å, was associated with plot implies higher stereochemical quality in the predicted
the V717G mutation. The RMSD score for the wild-type protein model. Ramachandran plot statistics were obtained
human APP was determined to be 10.5 ± 4.6 Å (Table S1). both before and after refining the predicted protein
structures.
The number of structure decoys in a unit of space
is termed cluster density, and a higher cluster density The results of the Ramachandran plot statistics for non-
with an increased structural decoys value indicates a refined protein models predicted by I-TASSER revealed
greater level of predictability in the model . The cluster that ≥66.6% of residues are located in the most favored
[23]

Volume 2 Issue 4 (2023) 7 https://doi.org/10.36922/an.1734