1. Trang chủ
  2. » Luận Văn - Báo Cáo

Nghiên Cứu Ảnh Hưởng Của Đột Biến Lên Cấu Trúc Và Hoạt Động Của Chuỗi Peptit Amyloid Beta Hướng Đến Ức Chế Bệnh Alzheimer.pdf

40 0 0

Đang tải... (xem toàn văn)

Tài liệu hạn chế xem trước, để xem đầy đủ mời bạn chọn Tải xuống

THÔNG TIN TÀI LIỆU

Tên đề tài (Tahoma font, size 12) SỞ KHOA HỌC VÀ CÔNG NGHỆ TP HỒ CHÍ MINH VIỆN KHOA HỌC VÀ CÔNG NGHỆ TÍNH TOÁN BÁO CÁO TỔNG KẾT NGHIÊN CỨU ẢNH HƯỞNG CỦA ĐỘT BIẾN LÊN CẤU TRÚC VÀ HOẠT ĐỘNG CỦA CHUỖI PE[.]

SỞ KHOA HỌC VÀ CƠNG NGHỆ TP HỒ CHÍ MINH VIỆN KHOA HỌC VÀ CƠNG NGHỆ TÍNH TỐN BÁO CÁO TỔNG KẾT NGHIÊN CỨU ẢNH HƯỞNG CỦA ĐỘT BIẾN LÊN CẤU TRÚC VÀ HOẠT ĐỘNG CỦA CHUỖI PEPTIDES AMYLOY BETA: HƯỚNG ĐẾN ỨC CHẾ BỆNH ALZHEIMER Viện trưởng: ………………… Đơn vị thực hiện: PTN Khoa học sống Chủ nhiệm đề tài: ThS Trần Thị Minh Thư ……………… TP HỒ CHÍ MINH, THÁNG 6/2019 SỞ KHOA HỌC VÀ CƠNG NGHỆ TP HỒ CHÍ MINH VIỆN KHOA HỌC VÀ CƠNG NGHỆ TÍNH TỐN BÁO CÁO TỔNG KẾT NGHIÊN CỨU ẢNH HƯỞNG CỦA ĐỘT BIẾN LÊN CẤU TRÚC VÀ HOẠT ĐỘNG CỦA CHUỖI PEPTIDES AMYLOY BETA: HƯỚNG ĐẾN ỨC CHẾ BỆNH ALZHEIMER Viện trưởng: Đơn vị thực hiện: PTN Khoa học sống Chủ nhiệm đề tài: ThS Trần Thị Minh Thư TP HỒ CHÍ MINH, THÁNG 6/2019 Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer MỤC LỤC Trang MỞ ĐẦU ĐƠN VỊ THỰC HIỆN KẾT QUẢ NGHIÊN CỨU I Báo cáo khoa học II Tài liệu khoa học xuất III Chương trình giáo dục đào tạo IV Hội nghị, hội thảo V File liệu 8 30 30 30 TÀI LIỆU THAM KHẢO CÁC PHỤ LỤC 31 PHỤ LỤC Bài báo “Aggregation Rate of Amyloid Beta Peptide is Controlled by BetaContent in Monomeric State” 34 34 Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer MỞ ĐẦU Bệnh Alzheimer (AD) bệnh phổ biến gây hậu nghiêm trọng vật chất tinh thần người[1] Vì thế, nước phát triển Mỹ Châu Âu đẩy mạnh đầu tư việc nghiên cứu nguyên nhân gây bệnh, chế phát bệnh việc điều trị AD thông qua thực nghiệm mô phỏng[1] Một hướng nghiên cứu làm đột biến protein gây bệnh: chuỗi amyloid beta (Aβ) Cụ thể, đột biến biến đến liệu pháp ngăn ngừa độc tố đồng thời ngăn ngừa liên kết chuỗi peptide riêng lẻ để tạo thành đám rối có hại cho não giảm tốc độ tạo sợi chuỗi peptides Aβ Những residue có thứ tự từ 21-23 (vùng turn, xem hình 1) chuỗi Aβ đóng vai trị quan trọng trình tạo sợi, nhiều kết thực nghiệm[2-7] nghiên cứu lý thuyết[8-15] cho thấy với nhiều loại đột biến khác vùng turn Flemish (A21G), Dutch (E22Q), Italian (E22K), Arctic (E22G), Iowa (D23N), Osaka (ΔE22, lược bỏ residue E22) thể điều Những đột biến G25L, G29L, G33L, G33A, G33I G37L chuỗi Aβ42 làm tăng trình tạo sợi hình thành mảng β nhiều so với dạng tự nhiên Aβ42[16-19] Mặt khác, vùng residue có thứ tự từ 1-8 chuỗi Aβ40 1−16 chuỗi Aβ42 cho làm trật tự trạng thái sợi chuỗi Aβ[20-22] Những kết thực nghiệm gần gây đột biến vùng đầu N-terminal chuỗi Aβ cho thấy residue vùng N xếp có tầm quan trọng cấu trúc chuỗi Aβ Đột biến English (H6R)[23-24], Taiwanese (D7H)[25] Tottori (D7N)[24, 26-27] làm thay đổi tốc độ tạo sợi tồn tế bào mà không ảnh hưởng đến số lượng Aβ[24] Đột biến A2V cho làm tăng trình tạo mảng Aβ40, pha trộn Aβ40 dạng tự nhiên đột biến A2V chống lại bệnh AD[28] Sử dụng kính hiển vi quang học đơn phân tử (AFM) để quan sát chuỗi Aβ, thấy đầu N-terminal đóng vai trò quan trọng việc tương tác chuỗi peptides mảng Aβ[29] Các thực nghiệm đây[30] chứng tỏ đột biến Arctic (E22G) làm tăng khả tích tụ Aβ micelle thông qua việc giảm hàm lượng helix amino axit 15-25 Kết thú vị chưa giải thích đầy đủ lý thuyết Mặc dù đóng vai trị quan trọng q trình tích tụ tương tác với protein khác[31-32], đột biến đuôi C chuỗi Aβ chưa tập trung nghiên cứu nhiều Các nghiên cứu tập trung thực đột biến vào vị trí có amino acid Glycine (29, 33, 37, 38) để khảo sát vai trò cấu trúc Glycine vị trí 29, 33, 37, 38 đuôi C Các đột biến G29L, G33L, G37L làm giảm độc tính Aβ[33] Tuy nhiên, lượng oligomer sinh tăng lên ảnh hưởng G33L[17] hiệu ứng bị giảm bớt có thêm đột biến G37L[34] Gần nhóm nghiên cứu Giáo sư Christophe D Link cho rằng, việc thay Glycine vị trí residue số 37 thành Leucine C.elegan làm suy giảm đáng kể độc tố oligomers Aβ hầu hết mẫu thử [33] Trong công bố khác, Roychaudhuri cộng phát cấu trúc β-hairpin với β-turn residue Val36-Gly37 đoạn Aβ31-42 không xuất Aβ31-40[35] Ngồi ra, đột biến G33V-V36P-G38V kích thích tạo beta turn beta hairpin C kéo theo gia tăng độc tính thay đổi dạng oligomer hình thành Khơng thế, đột biến G33V-V36P-G38V có khả chuyển đổi tính chất Aβ40 thành đặc tính vốn có Aβ42 làm Aβ42 mang đột biến VPV trở thành dạng siêu-Aβ42 Cơng trình Roychaudhuri đột biến V36P-G37P làm suy giảm khả hình thành beta turn hai residue 36-37 đồng thời làm giảm Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer khả hình thành cấu trúc β Aβ Do đó, họ cho việc kích thích hình thành turn Val36-Gly37 đuôi C Aβ31-40 làm Aβ40 mang tính chất giống Aβ42 ngược lại, làm ổn định cấu trúc làm cho Aβ42 có đặc tính giống Aβ40 Khơng hồn tồn đồng với giả thiết này, công bố gần chúng tơi mơ đột biến VPV tồn chuỗi Aβ40 Aβ42 cho thấy đột biến VPV làm tăng β-turn đầu C Aβ40 tỉ lệ thấp Aβ42-WT[36] Ngoài ra, cấu trúc β-hairpin vị trí 36-37 xuất Aβ42-WT khơng xuất Aβ40-VPV Từ kết này, cho cách hành xử tương đồng Aβ40-VPV Aβ42-WT[35] không xuất β-turn β-hairpin đầu C mà gia tăng thành phần β-content tồn chuỗi [36] Có nhiều yếu tố ảnh hưởng đến tốc độ tích tụ Aβ42 bao gồm ảnh hưởng môi trường biến đổi học cấu trúc nội protein Hiểu yếu tố định đến tốc độ tụ tập protein điều thú vị đám rối protein liên quan đến nhiều bệnh suy thoái thần kinh Các nghiên cứu uy tín trước cho rằng, tính chất kị nước, điện tích thành phần xu hướng tạo sợi trạng thái monomer kiểm soát tốc độ tụ tập protein Cấu trúc tìm từ thực nghiệm cho thấy trạng thái sợi protein có cấu trúc phiến beta chéo Do đó, monomer có thành phần nhiều beta cho tạo sợi nhanh monomer có thành phần beta Tuy nhiên yếu tố quan trọng chưa minh chứng cách rõ ràng Đó lí chúng tơi thực nghiên cứu Chúng sử dụng phương pháp động lực học phân tử trao đổi nhiệt độ khác để mô 20 hệ protein bao gồm Aβ42 thể tự nhiên 19 đột biến Do nguồn lực máy tính hạn chế để tiết kiệm thời gian tính tốn, chúng tơi sử dụng dung mơi nước khơng tường minh Chúng tơi tính tốn thành phần (β) 20 hệ protein mà tốc độ tích tụ κ hệ đo lường thực nghiệm Kết đạt hệ số κ phụ thuộc vào hàm mũ β, κ ~ exp(cβ), c số Kết đạt minh chứng rõ ràng cho giả thuyết mà từ lâu thừa nhận: thành phần (β) trạng thái monomer ảnh hưởng đến xu hướng kết tụ Aβ42 Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer Hình Cấu trúc dạng sợi chuỗi peptides Aβ42 phân tích thực nghiệm[37] Lời cảm ơn đến ICST Tôi xin chân thành cảm ơn Ban lãnh đạo Viện Khoa học Cơng nghệ Tính tốn Thành phố Hồ Chí Minh tài trợ kinh phí tạo điều kiện để tơi thực đề tài nghiên cứu Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer ĐƠN VỊ THỰC HIỆN Phịng thí nghiệm: Chủ nhiệm đề tài: Thành viên đề tài: Khoa học sống ThS Trần Thị Minh Thư GS TSKH Mai Xuân Lý ThS Phạm Đăng Lân ThS Phan Minh Trường ThS Nguyễn Hồng Linh CN Phạm Minh Trí Cơ quan phối hợp: Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer KẾT QUẢ NGHIÊN CỨU I BÁO CÁO KHOA HỌC Các đột biến sử dụng tính tốn Cấu trúc bậc chuỗi polypeptide định việc cuộn tự tập hợp protein [38-39] Đột biến sử dụng để làm thay đổi thứ tự amino chuỗi dẫn đến thay đổi hành vi cuộn tốc độ tạo sợi protein [40] Vùng turn Aβ42, (vùng có thứ tự residue từ 21-23) cho đóng vai trị quan trọng việc tạo sợi, đột biến vùng Flemish (A21G), Osaka (E22D), Italian (E22K), Arctic (E22G) Iowa (D23N) nghiên cứu [41-43] Trong đề tài này, nghiên cứu đột biến vùng turn vùng đầu C Aβ42, tập hợp đột biến có tốc độ cuộn đo lường thực nghiệm (bảng1)[41, 43-44] Bảng Chuỗi peptide Aβ42-WT đột biến TT 10 11 12 13 14 15 16 17 18 Aβ42 đột biến Aβ42-WT I41D, A42Q I41D, A42S I41H, A42D I41E, A42L I41H, A42N A21G I41T, A42N I41T, A42Q I41L, A42N I41Q, A42L I41T, A42M I41T, A42I I41K I41K, A42L I41R, A42R A42R E22G Chuỗi protein DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIA DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVDQ DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVDS DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVHD DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVEL DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVHN DAEFRHDSGYEVHHQKLVFFGEDVGSNKGAIIGLMVGGVVIA DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVTN DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVTQ DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVLN DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVQL DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVTM DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVTI DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVKA DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVKL DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVRR DAEFRHDSGYEVHHQKLVFFAEDVGSNKGAIIGLMVGGVVIR DAEFRHDSGYEVHHQKLVFFAGDVGSNKGAIIGLMVGGVVIA Viện Khoa học Công nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer 19 20 D23N E22K DAEFRHDSGYEVHHQKLVFFAENVGSNKGAIIGLMVGGVVIA DAEFRHDSGYEVHHQKLVFFAKDVGSNKGAIIGLMVGGVVIA Phương pháp mô động lực học phân tử Chúng tiến hành mô trạng thái cân 20 protein bảng sử dụng trường lực OPLS[45] dung dịch nước khơng tường minh mơ hình Born tổng qt[46] Chúng tơi sử dụng trường lực OPLS-AA kết tính tốn cho cấu hình Aβ tương ứng với cấu hình thực nghiệm NMR nó[47] Ngồi ra, sử dụng trường lực với mơ hình Born tổng quát mô dung dịch nước không tường minh, thu kết tin cậy cấu trúc bậc hai Aβ thể tự nhiên số đột biến [48-49] Để tăng số cấu hình đạt được, chúng tơi sử dụng phương pháp trao đổi với 12 nhiệt độ tương ứng Nhiệt độ khoảng từ 290.16 đến 490.16 K, cụ thể 290.16, 300, 311.80, 326.18, 343.14, 361.92, 380.83, 400.69, 421.86, 444.02, 466.14, 490.16 K Chúng tơi phân tích kết nhiệt độ 300K nhiệt độ trùng khớp với thực nghiệm trình xác định tốc độ tạo sợi Thời gian mô cho replica 500ns cấu hình xuất sau 10ps Cấu trúc ban đầu WT chọn từ cấu hình phổ biến nghiên cứu trước chúng tôi[48], cấu trúc đột biến khởi tạo dựa cấu trúc WT Raptor X web server[50] (Hình 2) Vì chúng tơi sử dụng phương pháp trao đổi nhiệt độ nên kết không phụ thuộc vào cấu trúc ban đầu Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer Hình Cấu trúc ban đầu cho mơ MD Aβ42-WT chọn từ cấu hình phổ biến nghiên cứu trước [48] Cấu trúc đột biến I41H-A42N khởi tạo dựa cấu trúc ban đầu Aβ42-WT web server RaptorX Các residue đột biến đánh dấu màu đỏ Khảo sát ổn dịnh hệ 3.1 Sự ổn định q trình mơ trao đổi (Replicas Exchang Molecular Dynamic - REMD) Để kiểm chứng ổn định hệ trình mơ REMD, chúng tơi theo dõi q trình trao đổi cấu hình hệ khơng gian 12 Để minh họa, vẽ trình trao đổi thứ hai cịn lại theo thời gian (hình 3, phần trên) Bản trao đổi với lại, bao gồm xa ( có nhiệt độ cao nhất) số 12, điều cho thấy phương pháp trao đổi hoạt động tốt hệ mô Nhận định củng cố trùng lấp kế cận (hình 3, phần dưới) Viện Khoa học Công nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer 4.76 I41H-A42N A21G 0.39 I41T-A42N 5.12 I41T-A42Q 4.51 I41L-A42N 10 I41Q- A42L 1.69 11 I41T-A42M 2.25 12 I41T- A42I 1.39 13 I41K 14 I41K-A42L 3.16 15 I41R-A42R 10.11 16 A42R 5.77 17 E22G 18 D23N 19 E22K -2.728 2.528 -1.773 1.211 1.5 0.535 2.319 -0.156 2.3 4.59 -2.91 -1.9 -0.14 -2.046 0.344 -1.228 0.535 3.41 -0.701 5.592 -0.443 -3.2736 -0.935 -1.228 -0.443 -1.773 -1.150 1.637 -3.41 1.410 4.365 -2.463 0.136 -0.633 -0.397 -0.067 -0.671 -0.286 -0.605 -0.228 -0.590 -0.293 -0.180 -0.295 -0.011 -0.292 0.212 -0.075 -1 -0.331 -1 -0.056 -0.379 -2 -0.392 -0.324 -1 -0.059 -0.034 -1 0.414 -1 0.578 -2 0.589 -0.837 [43, 52] [44, 52] -0.561 -0.518 0.209 [43, 52] [41] 0.238 0.545 [43, 52] Kết luận Sử dụng phương pháp mô trao đổi nhiệt độ cho toàn nguyên tử Aβ42 thể tự nhiên 19 đột biến Chúng tìm tính tương quan cấu trúc bậc hai trạng thái monomer tốc độ tạo sợi protein thực nghiệm Theo đó, chúng tơi rằng, khơng có tương quan tốc độ tạo sợi với thành phần coil α, thành phần turn, tương quan xảy thấp Sử dụng tập sữ liệu 27 chuỗi protein Chiti Dobson[40], nghiên cứu cho thấy, xét đến thay đổi điện tích residue hệ số κ không tương quan với lượng chuyển đổi từ trạng thái α sang β, ΔΔG (R< 0.5) Một lí khiến tổng lượng tương quan với vận tốc tạo sợi giá trị Pβ and Pα tính điểm đột biến[40] Nếu sử dụng tính tốn trực tiếp ΔΔG từ mơ động lực học cho tồn ngun tử tăng tương quan hai yếu tố, nhiên tính tốn cần mơ thời gian dài nằm ngồi nghiên cứu chúng tơi Một lí khác tập liệu chưa đủ lớn Trong báo cáo này, khẳng định thành phần cấu trúc β trạng thái monomer Aβ42 kiểm soát tốc độ tạo sợi loại protein Ở công thức (2), hàm phụ thuộc κ lên β thể dạng e mũ, tỉ lệ phần trăm β lớn hơn, tốc độ tạo sợi nhanh Nghiên cứu có ý nghĩa quan trọng nguyên lý tạo sợi protein, cụ thể yếu tố nội protein ảnh hưởng đến xu hướng tự tập hợp 25 Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer Công thức (1) (2) cho phép xác định tốc độ tạo sợi protein dựa thành phần β cấu trúc monomer tính tốn mơ máy tính Nghiên cứu bước khởi đầu việc ước tính vận tốc tạo sợi chuỗi protein thực tế dựa tính tốn máy tính Hệ số c công thức (2) cần kiểm tra tập liệu lớn với nhiều loại protein khác 26 Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer II CÁC TÀI LIỆU KHOA HỌC ĐÃ XUẤT BẢN Tran Thi Minh Thu, Nguyen Truong Co, Ly Anh Tu, and Mai Suan Li, Aggregation Rate of Amyloid Beta Peptide is Controlled by Beta-Content in Monomeric State, J Chem Phys, Vol 150, Issue 22, 14th June 2019 III CHƯƠNG TRÌNH GIÁO DỤC VÀ ĐÀO TẠO III HỘI NGHỊ, HỘI THẢO IV FILE DỮ LIỆU 27 Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer TÀI LIỆU THAM KHẢO [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] A s Association, Alzheimer's & Dementia 2016, 12, 459 L Hendriks, C M Van Duijn, P Cras, M Cruts, W Van Hul, F Van Harskamp, A Warren, M G McInnis, S E Antonarakis, J.-J Martin, Nature genetics 1992, 1, 218 E Levy, M D Carman, I J Fernandez-Madrid, M D Power, I Lieberburg, S G van Duinen, G Bots, W Luyendijk, B Frangione, Science 1990, 248, 1124 O Bugiani, Neurobiol Aging 1998, 19, S238 K Kamino, H T Orr, H Payami, E M Wijsman, M E Alonso, S M Pulst, L Anderson, S O'dahl, E Nemens, J A White, American journal of human genetics 1992, 51, 998 T J Grabowski, H S Cho, J P G Vonsattel, G W Rebeck, S M Greenberg, Annals of Neurology: Official Journal of the American Neurological Association and the Child Neurology Society 2001, 49, 697 T Tomiyama, T Nagata, H Shimada, R Teraoka, A Fukushima, H Kanemitsu, H Takuma, R Kuwano, M Imagawa, S Ataka, Annals of neurology 2008, 63, 377 F Massi, J E Straub, Biophysical journal 2001, 81, 697 S Côté, P Derreumaux, N Mousseau, Journal of chemical theory and computation 2011, 7, 2584 Y.-S Lin, V S Pande, Biophysical journal 2012, 103, L47 A Huet, P Derreumaux, Biophys J 2006, 91, 3829 O Coskuner, O Wise-Scira, G Perry, T Kitahara, ACS chemical neuroscience 2012, 4, 310 S Mitternacht, I Staneva, T Härd, A Irbäck, Proteins: Structure, Function, and Bioinformatics 2010, 78, 2600 M G Krone, A Baumketner, S L Bernstein, T Wyttenbach, N D Lazo, D B Teplow, M T Bowers, J.-E Shea, Journal of molecular biology 2008, 381, 221 A Baumketner, M G Krone, J.-E Shea, Proceedings of the National Academy of Sciences 2008, 105, 6027 L W Hung, G D Ciccotosto, E Giannakis, D J Tew, K Perez, C L Masters, R Cappai, J D Wade, K J Barnham, Journal of Neuroscience 2008, 28, 11950 A Harmeier, C Wozny, B R Rost, L.-M Munter, H Hua, O Georgiev, M Beyermann, P W Hildebrand, C Weise, W Schaffner, Journal of Neuroscience 2009, 29, 7582 L M Munter, P Voigt, A Harmeier, D Kaden, K E Gottschalk, C Weise, R Pipkorn, M Schaefer, D Langosch, G Multhaup, The EMBO journal 2007, 26, 1702 Y Lu, G Wei, P Derreumaux, The Journal of Physical Chemistry B 2010, 115, 1282 A T Petkova, W.-M Yau, R Tycko, Biochemistry 2006, 45, 498 A K Paravastu, R D Leapman, W.-M Yau, R Tycko, Proceedings of the National Academy of Sciences 2008, 105, 18349 28 Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] T Lührs, C Ritter, M Adrian, D Riek-Loher, B Bohrmann, H Döbeli, D Schubert, R Riek, Proceedings of the National Academy of Sciences 2005, 102, 17342 J Janssen, J Beck, T Campbell, A Dickinson, N Fox, R Harvey, H Houlden, M Rossor, J Collinge, Neurology 2003, 60, 235 Y Hori, T Hashimoto, Y Wakutani, K Urakami, K Nakashima, M M Condron, S Tsubuki, T C Saido, D B Teplow, T Iwatsubo, Journal of biological chemistry 2007, 282, 4916 W.-T Chen, C.-J Hong, Y.-T Lin, W.-H Chang, H.-T Huang, J.-Y Liao, Y.-J Chang, Y.-F Hsieh, C.-Y Cheng, H.-C Liu, PloS one 2012, 7, e35807 Y Wakutani, K Watanabe, Y Adachi, K Wada-Isoe, K Urakami, H Ninomiya, T Saido, T Hashimoto, T Iwatsubo, K Nakashima, Journal of Neurology, Neurosurgery & Psychiatry 2004, 75, 1039 K Ono, M M Condron, D B Teplow, Journal of Biological Chemistry 2010, jbc M109 086496 G Di Fede, M Catania, M Morbin, G Rossi, S Suardi, G Mazzoleni, M Merlin, A R Giovagnoli, S Prioni, A Erbetta, Science 2009, 323, 1473 Z Lv, R Roychaudhuri, M M Condron, D B Teplow, Y L Lyubchenko, Scientific reports 2013, 3, 2880 C.-J Lo, C.-C Wang, H.-B Huang, C.-F Chang, M.-S Shiao, Y.-C Chen, T.-H Lin, Amyloid 2015, 22, J T Jarrett, E P Berger, P T Lansbury Jr, Biochemistry 1993, 32, 4693 Y.-J Chang, N H Linh, Y H Shih, H.-M Yu, M S Li, Y.-R Chen, ACS chemical neuroscience 2016, 7, 1097 V Fonte, V Dostal, C M Roberts, P Gonzales, P Lacor, J Magrane, N Dingwell, E Y Fan, M A Silverman, G H Stein, Molecular neurodegeneration 2011, 6, 61 M Decock, S Stanga, J.-N Octave, I Dewachter, S O Smith, S N Constantinescu, P Kienlen-Campard, Frontiers in aging neuroscience 2016, 8, 107 R Roychaudhuri, M Yang, A Deshpande, G M Cole, S Frautschy, A Lomakin, G B Benedek, D B Teplow, Journal of molecular biology 2013, 425, 292 N H Linh, T T M Thu, L Tu, C.-K Hu, M S Li, The Journal of Physical Chemistry B 2017, 121, 4341 Y Xiao, B Ma, D McElheny, S Parthasarathy, F Long, M Hoshi, R Nussinov, Y Ishii, Nature structural & molecular biology 2015, 22, 499 F Chiti, C M Dobson, Annu Rev Biochem 2006, 75, 333 J Nasica-Labouze, P H Nguyen, F Sterpone, O Berthoumieu, N.-V Buchete, S Coté, A De Simone, A J Doig, P Faller, A Garcia, A Laio, M S Li, S Melchionna, N Mousseau, Y Mu, A Paravastu, S Pasquali, D J Rosenman, B Strodel, B Tarus, J H Viles, T Zhang, C Wang, P Derreumaux, Chem Rev 2015, 115, 3518 F Chiti, M Stefani, N Taddei, G Ramponi, C M Dobson, Nature 2003, 424, 805 K Murakami, K Irie, A Morimoto, H Ohigashi, M Shindo, M Nagao, T Shimizu, T Shirasawa, J Biol Chem 2003, 278, 46179 29 Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] F Chiti, M Calamai, N Taddei, M Stefani, G Ramponi, C M Dobson, Proc Natl Acad Sci U.S.A 2002, 99, 16419 C Nilsberth, A Westlind-Danielsson, C B Eckman, M M Condron, K Axelman, C Forsell, C Stenh, J Luthman, D B Teplow, S G Younkin, Nat Neurosci 2001, 4, 887 W Kim, M H Hecht, J Biol Chem 2005, 280, 35069 G A Kaminski, R A Friesner, J Tirado-Rives, W L Jorgensen, J Phys Chem B 2001, 105, 6474 V Tsui, D A Case, Biopolymers 2000, 56, 275 N G Sgourakis, Y Yan, S A McCallum, C Wang, A E Garcia, J Mol Biol 2007, 368, 1448 H L Nguyen, T Thi Minh Thu, P M Truong, P D Lan, V H Man, P H Nguyen, L A Tu, Y.-C Chen, M S Li, J Phys Chem B 2016, 120, 7371 N H Linh, T T Minh Thu, L Tu, C.-K Hu, M S Li, J Phys Chem B 2017, 121, 4341 S Wang, W Li, S Liu, J Xu, Nucleic Acids Res 2016, 44, W430 D E Hicks, Thermophysical Properties of the Amyloid Beta Protein from Differential Scanning Calorimetry, University of Tennessee, 2005 S H Chong, S Ham, Angew Chem 2014, 126, 4042 W Kim, M H Hecht, Proc Natl Acad Sci U.S.A 2006, 103, 15824 S W Snyder, U S Ladror, W S Wade, G T Wang, L W Barrett, E D Matayoshi, H J Huffaker, G A Krafft, T F Holzman, Biophys J 1994, 67, 1216 M A Wälti, F Ravotti, H Arai, C G Glabe, J S Wall, A Böckmann, P Güntert, B H Meier, R Riek, Proc Natl Acad Sci U.S.A 2016, 113, E4976 CÁC PHỤ LỤC PHỤ LỤC 1: 30 Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Nghiên cứu ảnh hưởng đột biến lên cấu trúc hoạt động chuỗi peptides Amyloy Beta: hướng đến ức chế bệnh Alzheimer Bài báo : Aggregation Rate of Amyloid Beta Peptide is Controlled by BetaContent in Monomeric State Tạp chí : Journal of Chemical Physics, Vol 150 Tác giả : Tran Thi Minh Thu, Nguyen Truong Co, Ly Anh Tu, and Mai Suan Li 31 Viện Khoa học Cơng nghệ Tính tốn TP Hồ Chí Minh Aggregation rate of amyloid beta peptide is controlled by beta-content in monomeric state Cite as: J Chem Phys 150, 225101 (2019); https://doi.org/10.1063/1.5096379 Submitted: 17 March 2019 Accepted: 20 May 2019 Published Online: 10 June 2019 Tran Thi Minh Thu , Nguyen Truong Co , Ly Anh Tu , and Mai Suan Li J Chem Phys 150, 225101 (2019); https://doi.org/10.1063/1.5096379 © 2019 Author(s) 150, 225101 The Journal of Chemical Physics ARTICLE scitation.org/journal/jcp Aggregation rate of amyloid beta peptide is controlled by beta-content in monomeric state Cite as: J Chem Phys 150, 225101 (2019); doi: 10.1063/1.5096379 Submitted: 17 March 2019 • Accepted: 20 May 2019 • Published Online: 10 June 2019 Tran Thi Minh Thu,1,2,3 Nguyen Truong Co,4 Ly Anh Tu,3 and Mai Suan Li4,a) AFFILIATIONS Institute for Computational Science and Technology, SBI Building, Quang Trung Software City, Tan Chanh Hiep Ward, District 12, Ho Chi Minh City, Vietnam Department of Metarials Science and Technology, University of Science-VNUHCM, 227 Nguyen Van Cu, District 5, Ho Chi Minh City, Vietnam Department of Applied Physics, Faculty of Applied Science, Ho Chi Minh City University of Technology-VNU HCM, 268 Ly Thuong Kiet Street, District 10, Ho Chi Minh City, Vietnam Institute of Physics, Polish Academy of Sciences, Al Lotnikow 32/46, 02-668 Warsaw, Poland a) Email: masli@ifpan.edu.pl ABSTRACT Understanding the key factors that govern the rate of protein aggregation is of immense interest since protein aggregation is associated with a number of neurodegenerative diseases Previous experimental and theoretical studies have revealed that the hydrophobicity, charge, and population of the fibril-prone monomeric state control the fibril formation rate Because the fibril structures consist of cross beta sheets, it is widely believed that those sequences that have a high beta content (β) in the monomeric state should have high aggregation rates as the monomer can serve as a template for fibril growth However, this important fact has never been explicitly proven, motivating us to carry out this study Using replica exchange molecular dynamics simulation with implicit water, we have computed β of 19 mutations of amyloid beta peptide of 42 residues (Aβ42) for which the aggregation rate κ has been measured experimentally We have found that κ depends on β in such a way that the higher the propensity to aggregation, the higher the beta content in the monomeric state Thus, we have solved a long-standing problem of the dependence of fibril formation time of the β-structure on a quantitative level Published under license by AIP Publishing https://doi.org/10.1063/1.5096379 I INTRODUCTION A number of neurodegenerative diseases are believed to be associated with the aggregation of amyloid proteins For example, according to the amyloid cascade hypothesis, Alzheimer’s disease (AD) is caused by the fibrillation of amyloid beta (Aβ) peptides forming amyloid plaques.1,2 Aβ peptides, which are cleaved from amyloid precursor protein (APP) by β- and γ-secretases, have 36–43 residues, but Aβ40 (40 amino acids) and Aβ42 (42 amino acids) are most abundant As the main component of amyloid plaques in the human brain, Aβ42 is more neurotoxic than Aβ40 due to faster selfassembly.3 Because understanding key factors driving protein aggregation is important not only for basic research but also for designing J Chem Phys 150, 225101 (2019); doi: 10.1063/1.5096379 Published under license by AIP Publishing efficient drugs for many diseases, this problem has attracted a lot of attention of researchers in recent decades Accumulated experimental and theoretical data indicate that in addition to external factors (pH, T, ionic strength, and protein concentration),4 there are intrinsic factors such as the hydrophobicity, net charge, secondary structure,4–7 and chain-length polydispersity8 and the so-called population of fibril-prone state in monomeric state N∗ 7,9 which play a key role in the aggregation propensity Locking the monomer in a strandlike state by constraining the Asp23-Lys28 salt bridge accelerates the aggregation of amyloid beta peptides considerably.10,11 As shown by NMR dispersion experiments12 and simulation,13 N∗ of Fyn SH3 is a nativelike folding intermediate that is prone to aggregation A high population of fibril-prone state N∗ 150, 225101-1 The Journal of Chemical Physics promotes the formation of template for the protein accumulation leading to an exponential dependence of fibril formation time on the population of this conformation.7 The regions enriched by hydrophobic residues promote aggregation14–17 as evident from the high correlation between the hydration free energy and the aggregation rate of proteins Aβ42, HypF-N, and AcP.6 Charged residues strongly influence protein aggregation not only because they are involved in specific salt bridges but also because a nonzero net charge prevents self-assembly through repulsion between chains.18,19 Chiti and Dobson have demonstrated5 that the variation in the aggregation rate upon mutation depends on the free energy change in conversion from the α-helix to β-sheet conformation, ∆∆G = ∆∆Gcoil-α + ∆∆Gβ-coil Here, ∆∆Gcoil-α and ∆∆Gβ-coil refer to free energy changes in coil-helix and beta-coil transformations, respectively However, these quantities were estimated using α-helix and β-sheet propensities of individual wildtype (WT) and mutant residues,5 leaving the question of the dependence of the fibril formation rate on the monomer β-content open In order to clarify the impact of secondary structures on the propensity to aggregation, one has to compute them consistently taking into account the contribution of all residues To so, in this paper, we have performed the all-atom replica exchange molecular dynamics (REMD) simulation using the optimized potentials for liquid simulations (OPLS) force field and implicit solvent for the wildtype (WT) and 19 mutants of Aβ42 We showed that the experimentally determined aggregation rate is strongly correlated with the β-content of the monomer, while no correlation with the α-content was observed So far, it is believed that the higher the β-content in the monomeric state, the faster the formation of fibrils, but there is no convincing evidence Here, based on the results obtained for a large dataset, for the first time, we strongly supported this widespread opinion II CHOICE OF SEQUENCES It is well known that the sequence of a protein controls not only its folding ability but also the propensity to self-assembly.20,21 Mutation is used to change the sequence resulting in a variation of the aggregation rate.5 Because the turn region21–23 of Aβ42, for example, plays a key role in the fibril formation, some mutations, such as Flemish (A21G), Osaka (E22D), Italian (E22K), Arctic (E22G), and Iowa (D23N), have been intensively studied.18,22,23 Table S1 of the supplementary material lists 19 Aβ42 mutants, for which the aggregation rate has been measured experimentally.22–24 We will perform a REMD simulation for this set of mutants Note that by performing 100 ns all-atom conventional MD simulation in explicit water, Chong and Ham showed that the experimentally determined self-assembly rates of these mutants strongly correlate with hydrophobicity in such a way that the higher the hydrophobicity, the faster the aggregation6 implying the important role of hydration III MOLECULAR DYNAMICS SIMULATION We have calculated secondary structures of WT and 19 mutants (Table S1 of the supplementary material) in the monomeric state J Chem Phys 150, 225101 (2019); doi: 10.1063/1.5096379 Published under license by AIP Publishing ARTICLE scitation.org/journal/jcp using the OPLS force field25 and the generalized Born (GB) model for an implicit solvent.26 The OPLS-AA force field was chosen because for the Aβ monomer it provided conformations consistent with the NMR data.27 In addition, using this force field in combination with the GB model for water, we previously obtained a reliable estimation of the secondary structures not only of WT but also of mutations of Aβ.28,29 To improve sampling, REMD was employed with 12 replicas in the temperature range from 290.16 to 490.16 K (see the supplementary material for more details), and for each replica, 500 ns MD simulation was performed One of the representative structures, obtained in our previous work for the monomer,28 was chosen as the initial structure for all replicas in WT simulation, while for the mutants we used the same structure but the corresponding mutation was made using the Raptor X web server30 (Fig S1 of the supplementary material) Because we used the REMD method, the results should not depend on the starting configuration IV EQUILIBRATION A Effectiveness of REMD simulation To show the effectiveness of REMD simulation, we monitored the evolution of each replica in the replica space For illustration, we plot the time dependence of exchanges of the second replica with the remaining partners [Fig S2 (upper panel) of the supplementary material] Clearly, this replica was exchanged with everyone else, including the farthest 12th, implying that the replica exchange method worked well for our system This is also confirmed by a strong overlap between adjacent distributions of potential energy of 12 replicas [Fig S2 (lower panel) of the supplementary material] For contiguous temperatures, the probability of exchanges is about 20%– 30% and this is consistent with the overlap of the potential energy distributions B Time dependence of RMSD, total energy, and beta content We monitored the time dependence of the root mean square displacement (RMSD), which was calculated using the initial structure as a reference structure and the coordinates of Cα-atoms We assumed that the system has reached equilibrium when RMSD gets saturation As evident from Fig S3 (upper panel) of the supplementary material, the equilibration time is about 210 ns for WT and mutants This conclusion is further supported by the time dependence of the total energy and beta content of three sequences (Fig S3 of the supplementary material) C Heat capacity in two time windows To ensure that 210 ns is enough for equilibration, we have computed the specific heat for time windows [210–400] and [210–500] ns using the formula Cv = (⟨E2 ⟩ − ⟨E⟩2 )/(kB T ), where E is the potential energy and ⟨ .⟩ denotes the thermodynamics average The heat capacity, obtained in two time windows at 300 K, for WT, A21G, and E22K, is shown in Fig S5 of the supplementary material Since CV is almost the same for two windows, we conclude that the data were equilibrated Note that together with WT, we select A21G and E22K as representatives for displaying data because the former is 150, 225101-2 The Journal of Chemical Physics among the sequences with low β-content, while the latter has the highest β-content (see below) Our result obtained for WT (Fig S4 of the supplementary material) is consistent with the data reported by Hicks31 that Cv slightly increases with the temperature in the range of 300–350 K D Secondary structure in two time windows We have also calculated the secondary structures for the [210– 400] and [210–500] ns time windows using the STRIDE algorithm.32 Since within error the obtained results are identical for these windows (Fig S5 of the supplementary material), the Aβ42 variants are at equilibrium after 210 ns (see also our previous works for WT28,29 ) The secondary structures obtained in the [210–500] ns window will be used for data analysis Although the MD simulation has been performed for 12 temperatures, we will consider T = 300 K where experimental data were collected Details of MD simulation are provided in the supplementary material ARTICLE scitation.org/journal/jcp analysis (dPCA) method in which only the first two important components V1 and V2 were kept (see the supplementary material for more details) Structures (S) representing major basins clearly show that the mutations alter secondary structures This is also evident from the per-residue distributions and the most important structures of three sequences shown in Fig For WT, these distributions were discussed in detail in our previous works.28,29 For E22K, the β-propensity at the C-terminal levels up compared with WT, while the reduction occurs in this region for A21G The β-content is compatible in the 18–23 fragment for WT and E22K Overall, E22K increases the β-content from 21% (WT) to 29%, but A21G reduces it to 13.7% (Table I) This is consistent with the experiments showing that E22K speeds up aggregation, while Flemish A21G slows it down.6,23 A Linear and exponential dependence Figure shows that the experimentally measured aggregation rate linearly depends on β-content κmut /κwt = − 0.29 + 0.0635 β, V AGGREGATION RATE DEPENDS ON β-CONTENT The free energy surface was constructed for Aβ42-WT and two selected mutants A21G and E22K (Fig S6 of the supplementary material) by using the dihedral angle principle component (1) where β is measured in percentage The correlation level is high as R = 0.85 Thus, for the first time, we explicitly showed that the β-content of the monomer controls the aggregation propensity for Aβ peptides and this is expected to be true for other proteins FIG (Upper) Most representative structures obtained by the clustering method at equilibrium (see structures S1 in Fig S3 of the supplementary material) for WT, A21G, and E22K The mutated residue is in the all-atom presentation (Lower) Per-residue distributions of secondary structures of Aβ42WT, A21G, and E22K at 300 K Results were obtained at equilibrium J Chem Phys 150, 225101 (2019); doi: 10.1063/1.5096379 Published under license by AIP Publishing 150, 225101-3 The Journal of Chemical Physics ARTICLE scitation.org/journal/jcp TABLE I The relative computational β-content and observed aggregation propensities of all mutations The simulation result was obtained at 300 K, the aggregation rate was taken from experiment, and reference is at the last column Simulation Mutations Aβ42 I41D-A42Q I41D-A42S I41H-A42D I41E-A42L I41H-A42N A21G I41T-A42N I41T-A42Q I41L-A42N I41Q- A42L I41T-A42M I41T- A42I I41K I41K-A42L I41R-A42R A42R E22G D23N E22K Experiment β-content (%) α-content (%) log (κmut /κwt ) 20.94 ± 1.91 12.92 ± 2.37 12.15 ± 1.93 11.99 ± 2.38 13.83 ± 2.32 15.06 ± 2.85 13.7 ± 2.02 11.82 ± 1.74 13.31 ± 1.82 14.61 ± 2.66 19.79 ± 3.17 12.78 ± 1.8 13.72 ± 1.96 19.00 ± 4.12 18.56 ± 3.46 16.54 ± 2.07 15.49 ± 1.9 20.52 ± 2.43 23.91 ± 2.39 29.06 ± 3.53 0.07 ± 1.70 1.98 ± 0.47 0.58 ± 0.15 3.32 ± 0.87 8.00 ± 2.16 2.37 ± 0.75 7.63 ± 2.43 0.73 ± 0.21 0.69 ± 0.30 0.60 ± 0.14 1.35 ± 0.42 4.72 ± 1.63 0.26 ± 0.06 3.05 ± 0.72 3.97 ± 1.39 4.92 ± 1.54 0.67 ± 0.22 0.80 ± 0.14 1.28 ± 0.36 −0.964 −0.913 −0.708 −0.445 −0.837 −0.671 −0.605 −0.590 −0.561 −0.295 −0.292 −0.075 −0.518 −0.379 −0.324 −0.034 0.209 0.238 0.545 Reference and 24 and 23 and 24 and 23 22 and 23 Although linear fit [Fig and Eq (1)] works well, we can show that the exponential fit is also possible As evident from Fig 2, there is also a high correlation (R = 0.8) between ln(κmut /κwt ) and β-content and their relationship can be described by an exponential function κ = κ0 exp (cβ), c = 0.071, (2) where κ0 is a fitting constant and β is measured in percentage A mutation that levels up self-aggregation enhances the β-propensity in the monomeric state and slows it down otherwise One can expect that the exponential dependence is valid for other systems as the α-β conversion is a barrier crossing event but constant c may be not universal Using lattice models, Li et al have observed the exponential dependence of κ on the population of fibril-prone state N∗ On the other hand, the N∗ conformation is rich in β-sheets implying that the exponential behavior [Eq (2)] is in the line with this work From this point of view, exponential fit is more favorable than linear, despite the fact that the correlation of the linear fit is higher One of the reasons why it is difficult to distinguish two dependences is that the β-content varies in a narrow interval We have to study a larger dataset to solve this problem B Aggregation rate does not correlate with helix-, turn-, and coil-propensity At 300 K, the helix-content of Aβ42-WT is practically zero, and the mutations E22K and A21G increase it to 1.28% and 7.63%, J Chem Phys 150, 225101 (2019); doi: 10.1063/1.5096379 Published under license by AIP Publishing FIG Dependence of the relative aggregation rate (upper panel) and the logarithm of the relative aggregation rate on the β-content (lower panel) The red circle refers to WT Linear fits are y = −0.29 + 0.0635 × (R = 0.85) and y = −1.534 + 0.071 × (R = 0.80) respectively (Fig 1) The enhancement occurs mainly in the 10– 18 region With the exception of I41K, all mutations increase the propensity to helix formation, varying from 0.26 (I41T-A42I) to 8% (I41E-A42L) (Table I) A small change in this quantity as a result of mutations leads to a poor correlation between the self-assembly rate and the α-content (Fig S7 of the supplementary material, R = 0.16) This is also consistent with the experiment showing that Aβ42-WT is much more aggregation prone than Aβ40-WT33 although they have almost the same helix content.28 The turn is highly populated at the C-terminal and in the 22– 29 region (Fig 1) In A21G, it increases mainly at residues 9, 19– 21 and 31–34, while a notable reduction occurs at positions 11, 12, and 15–18, which leads to a slight increase from 61.17% (WT) to 63.47% (A21G) Upon E22K replacement, the turn propensity increases at residues 1–3 and 23, but decreases at positions 11–15 and 29–31, resulting in a decline by about 5% Compared to the α-content, the turn propensity varies over a wider range from 31.81 (D23N) to 73.54 (I41T-A42N) leading to a higher correlation with experiments (R = 0.46, Fig S8 of the supplementary material) The rationale for this observation is that the formation of fibril contacts, such as the Asp23-Lys28 salt bridge, in the turn segment plays a key role in the fibril growth.10,11 Nevertheless, it cannot be concluded that the turn-propensity controls the aggregation rate as the correlation coefficient remains below 0.5 (Fig S8 of the supplementary material) 150, 225101-4 The Journal of Chemical Physics In WT, the coil structure is about 16% (Table S3 of the supplementary material) and it is predominantly populated at several first and last residues (Fig 1) After mutation, the per-residue distribution changes (Fig 1), but the total amount slightly changes remaining in the interval of 13%–21% (Table S3 of the supplementary material) Consequently, as in the case of α-content, the coil propensity poorly correlates with the aggregation rate, having R = 0.15 (Fig S9 of the supplementary material) One of the possible causes of poor correlation with the α- and coil-content is that these structures are poorly populated in the N∗ state.34 VI EFFECT OF CHARGE, HYDROPHOBICITY, AND PROPENSITY TO CONVERSION FROM AN α-HELICAL TO A β-SHEET CONFORMATION A Aggregation rate poorly correlates with the free energy for conversion from an α-helical to a β-sheet conformation ∆∆G The change in free energy from the coil to the β-state, ∆∆Gβ-coil , was calculated using the formula ∆∆Gβ-coil = 13.64(Pβ wt − Pβ mut ), where Pβ wt and Pβ mut are the β sheet propensities of the wildtype and mutant residues, respectively, and 13.64 is the conversion constant from the normalized scale to the unit kJ/mol, and ∆∆Gβ-coil is measured in kJ/mol Using Pβ for the 20 amino acids provided in the supplementary material of Chiti and Dobson5 and in Table I of Street and Mayo,35 we obtained ∆∆Gβ-coil for all sequences studied (Table S4 of the supplementary material) The change in free energy from α-helix to the coil, ∆∆Gcoil-α , was estimated using the equation ∆∆Gcoil-α = RT ln(Pα wt /Pα mut ), ARTICLE scitation.org/journal/jcp where Pα wt and Pα mut are α-helical propensities of the wildtype and mutated sequences at the mutation site, respectively, and gas constant R = 0.008 314 kJ mol−1 K−1 The helix percentage was calculated using the AGADIR algorithm (www.http://agadir.crg.es/), and the results are shown in Table S3 of the supplementary material The free energy of conversion from an α-helical to a β-sheet conformation, ∆∆G, is defined as ∆∆G = ∆∆Gβ-coil + ∆∆Gcoil-α and its values are also given in Table S3 of the supplementary material As shown in Fig 3(a), the correlation between the experimental aggregation rates and ∆∆G is quite low (R = 0.40), suggesting that the β-content of the monomeric state is a better indicator for the selfassembly propensity compared to the propensity to conversion from the α-helix to the β-sheet conformation By confining to mutations that not change the net charge, Chiti and Dobson showed5 that for a set of 15 sequences, the correlation level between κ and ∆∆G is noticeably higher (R = 0.71) than ours This result seems to contradict our result, but using the entire set of 27 sequences,5 we obtained R = 0.41 (Fig S10 of the supplementary material) which is close to our value 0.40 From this prospect, our result is consistent with that of Chiti and Dobson.5 Nevertheless, further study is needed to clarify the correlation with ∆∆G B Aggregation rate correlates with the change in the hydrophobicity (∆Hydr) and overall charge (∆Charge) due to mutation Following Dobson et al.,5 hydrophobicity ∆Hydr = Hydrwt Hydrmut are the hydrophobicity sequences, respectively Similarly, we calculated the change in − Hydrmut , where Hydrwt and of the wildtype and mutant a change in the net charge is FIG Dependence of the logarithm of the relative aggregation rate on the predicted change in propensity to convert from an α-helical to a β-sheet conformation (a), hydrophobicity (b), charge (c), and the calculated aggregation rate obtained by Eq (3) (d) The correlation level and the slope are also shown J Chem Phys 150, 225101 (2019); doi: 10.1063/1.5096379 Published under license by AIP Publishing 150, 225101-5 The Journal of Chemical Physics ARTICLE defined as ∆Charge = |Chargemut | − |Chargewt |, where Chargemut and Chargewt are the charge of the mutation and the wildtype, respectively We calculated ∆Hydr and ∆Charge for all studied sequences (Table S4 of the supplementary material) using the values of hydrophobicity and charge for 20 amino acids provided in the supplementary material of Dobson et al.5 The level of correlation between the experimental aggregation rate and ∆Hydr and ∆Charge is R = 0.661 and 0.683, respectively [Figs 3(b) and 3(c)] These values are compatible with 0.545 and 0.721 reported by Dobson et al.,5 but they are noticeably lower than the correlation level between the aggregation rate and the beta content We have calculated the change in the aggregation rate due to mutation, using the following equation:5 ln(κmut /κwt ) = A ⋅ ∆Hydr + B ⋅ ∆∆G + C ⋅ ∆Charge (3) Here, A, B, and C are the slopes obtained from the linear fit between ln(κmut /κwt ) and ∆Hydr, ∆∆G, and ∆Charge, respectively, as follows from Figs 3(a)–3(c) are −0.081, 0.063 and −0.304, respectively The predicted aggregation rates, obtained by using Eq (3), are shown in column of Table S4 of the supplementary material Despite the poor correlation with ∆∆G, its combination with hydrophobicity and charge significantly improves the correlation between the predicted and experimental aggregation rates as we have R = 0.863 [Fig 3(d)] This correlation level is as high as that for correlation with the beta content Interestingly, because the slope in Fig 3(d) is 1.064 (close to 1), the predicted and experimentally measured rates are almost the same VII CONCLUSION We have carried out the all-atom REMD simulation in implicit water for various mutations of Aβ42 The correlation between the secondary structures obtained in the monomeric state and the experimentally determined aggregation rates has been thoroughly analyzed We have found that there is no correspondence between the experimental rate and the helix and coil propensities, while the fit with the turn is relatively poor Using the data set of 27 sequences from Chiti and Dobson,5 we demonstrated that if the change in net charge was taken into account, then the experimental κ does not correlate with the propensity to conversion from α-helical to β-sheet conformation free energy ∆∆G as R is below 0.5 One of the possible reasons is that ∆∆Gβ-coil and ∆∆Gcoil-α were determined by the empirical propensities Pβ and Pα A direct estimate of ∆∆G from the all-atom simulation may improve the correlation but this requires for a long simulation that is beyond the scope of this paper Another possible scenario is that the poor correlation between κ and ∆∆G is due to a small dataset To clarify this problem, we need more experimental as well as simulation data that are more convincing rather than the existing ones We have found a strong correlation between the experimental aggregation rate and β-propensity in the monomer The dependence of κ on β is expressed by an exponential function in such a way that the higher the β-propensity, the faster the formation of fibril [Eq (2)] But linear dependence [Eq (1)] is not excluded, probably due to the fact that the dataset is not large enough Nevertheless, our result sheds light on our understanding of major principles that J Chem Phys 150, 225101 (2019); doi: 10.1063/1.5096379 Published under license by AIP Publishing scitation.org/journal/jcp regulate the self-aggregation propensity of proteins, in particular, intrinsically disordered proteins Since an estimation of the aggregation rate of long polypeptide chains is beyond existing computational facilities, Eq (2) [or maybe Eq (1)] is very useful because it would allow us to predict the aggregation rate based on the β-content, which can be easily obtained by REMD simulation SUPPLEMENTARY MATERIAL See supplementary material for “Material and Method”; initial structure for MD simulation of WT; plots showing the performance of the RE method; time dependence of RMSD, total energy, and beta-content of the three representative sequences; temperature dependence of the heat capacity, per-residue distributions of the beta-content, and free energy surfaces of WT, A21G, and E22K; dependence of the aggregation rate on ∆∆G, helix, turn, and coil propensity; tables showing sequences of Aβ42-WT and mutations, characteristics of structures representing major basins on the free energy surface of WT, A21G, and E22K, turn and coil propensities, ∆∆Gβ-coil and ∆∆Gcoil-α , estimated for 20 sequences through the β- and α-propensities of amino acids, and ∆Hydr and ∆Charge ACKNOWLEDGMENTS This work was supported by Narodowe Centrum Nauki in Poland (Grant No 2015/19/B/ST4/02721) and Department of Science and Technology at Ho Chi Minh City, Vietnam (Grant No 19/2017/H–D-KHCNTT) Allocation of CPU time at the supercomputer center TASK in Gdansk (Poland) is highly appreciated This research was also supported in part by PLGrid Infrastructure in Poland REFERENCES D J Selkoe, Science 275, 630 (1997) J Hardy and G Higgins, Science 256, 184 (1992) D J Selkoe, Physiol Rev 81, 741 (2001) M Belli, M Ramazzotti, and F Chiti, EMBO Rep 12, 657 (2011) F Chiti, M Stefani, N Taddei, G Ramponi, and C M Dobson, Nature 424, 805 (2003) S H Chong and S Ham, Angew Chem 126, 4042 (2014) M S Li, N T Co, G Reddy, C K Hu, J E Straub, and D Thirumalai, Phys Rev Lett 105, 218101 (2010) A Hernik-Magon, W Pulawski, B Fedorczyk, D Tymecka, A Misicka, P Szymczak, and W Dzwolak, Biomacromolecules 17, 1376 (2016) B Tarus, J E Straub, and D Thirumalai, J Am Chem Soc 128, 16159 (2006) 10 G Reddy, J E Straub, and D Thirumalai, J Phys Chem B 113, 1162 (2009) 11 K L Sciarretta, D J Gordon, A T Petkova, R Tycko, and S C Meredith, Biochemistry 44, 6003 (2005) 12 P Neudecker, P Robustelli, A Cavalli, P Walsh, P Lundstrom, A ZarrineAfsar, S Sharpe, M Vendruscolo, and L E Kay, Science 336, 362 (2012) 13 P I Zhuravlev, G Reddy, J E Straub, and D Thirumalai, J Mol Biol 426, 2653 (2014) 14 C Tanford, Science 200, 1012 (1978) 15 K A Dill, Biochemistry 29, 7133 (1990) 16 A Ben-Naim, Hydrophobic Interactions (Springer Science & Business Media, 2012) 17 W Kim and M H Hecht, Proc Natl Acad Sci U S A 103, 15824 (2006) 150, 225101-6 The Journal of Chemical Physics 18 F Chiti, M Calamai, N Taddei, M Stefani, G Ramponi, and C M Dobson, Proc Natl Acad Sci U S A 99, 16419 (2002) 19 A T Petkova, Y Ishii, J J Balbach, O N Antzutkin, R D Leapman, F Delaglio, and R Tycko, Proc Natl Acad Sci U S A 99, 16742 (2002) 20 F Chiti and C M Dobson, Annu Rev Biochem 75, 333 (2006) 21 J Nasica-Labouze et al., Chem Rev 115, 3518 (2015) 22 K Murakami, K Irie, A Morimoto, H Ohigashi, M Shindo, M Nagao, T Shimizu, and T Shirasawa, J Biol Chem 278, 46179 (2003) 23 C Nilsberth et al., Nat Neurosci 4, 887 (2001) 24 W Kim and M H Hecht, J Biol Chem 280, 35069 (2005) 25 G A Kaminski, R A Friesner, J Tirado-Rives, and W L Jorgensen, J Phys Chem B 105, 6474 (2001) 26 V Tsui and D A Case, Biopolymers 56, 275 (2000) 27 N G Sgourakis, Y Yan, S A McCallum, C Wang, and A E Garcia, J Mol Biol 368, 1448 (2007) J Chem Phys 150, 225101 (2019); doi: 10.1063/1.5096379 Published under license by AIP Publishing ARTICLE scitation.org/journal/jcp 28 H L Nguyen, T T M Thu, P M Truong, P D Lan, V H Man, P H Nguyen, L A Tu, Y.-C Chen, and M S Li, J Phys Chem B 120, 7371 (2016) 29 N H Linh, T T M Thu, L Tu, C.-K Hu, and M S Li, J Phys Chem B 121, 4341 (2017) 30 S Wang, W Li, S Liu, and J Xu, Nucleic Acids Res 44, W430 (2016) 31 D E Hicks, “Thermophysical properties of the amyloid beta protein from differential scanning calorimetry,” Ph.D thesis (University of Tennessee, 2005) 32 M Heinig and D Frishman, Nucleic Acids Res 32, W500 (2004) 33 S W Snyder, U S Ladror, W S Wade, G T Wang, L W Barrett, E D Matayoshi, H J Huffaker, G A Krafft, and T F Holzman, Biophys J 67, 1216 (1994) 34 M A Wälti, F Ravotti, H Arai, C G Glabe, J S Wall, A Böckmann, P Güntert, B H Meier, and R Riek, Proc Natl Acad Sci U S A 113, E4976 (2016) 35 A G Street and S L Mayo, Proc Natl Acad Sci U S A 96, 9074 (1999) 150, 225101-7

Ngày đăng: 05/10/2023, 17:08

Xem thêm:

TÀI LIỆU CÙNG NGƯỜI DÙNG

TÀI LIỆU LIÊN QUAN

w