research methods in critical security studies an introduction

2nd Edition

Research Methods in Critical Security Studies An Introduction

• Taylor & Francis eBooks (Institutional Purchase) Opens in new tab or window

Description

This textbook surveys new and emergent methods for doing research in critical security studies, filling a gap in the literature. The second edition has been revised and updated. This textbook is a practical guide to research design in this increasingly established field. Arguing for serious attention to questions of research design and method, the book develops accessible scholarly overviews of key methods used across critical security studies, such as ethnography, discourse analysis, materiality, and corporeal methods. It draws on prominent examples of each method’s objects of analysis, relevant data, and forms of data collection. The book’s defining feature is the collection of diverse accounts of research design from scholars working within each method, each of which is a clear and honest recounting of a specific project’s design and development. This second edition is extensively revised and expanded. Its 33 contributors reflect the sheer diversity of critical security studies today, representing various career stages, scholarly interests, and identities. This book is systematic in its approach to research design but keeps a reflexive and pluralist approach to the question of methods and how they can be used. The second edition has a new forward-looking conclusion examining future research trends and challenges for the field. This book will be essential reading for upper-level students and researchers in the field of critical security studies, and of much interest to students in International Relations and across the social sciences.

Table of Contents

Mark B. Salter is Professor in the School of Political Studies, University of Ottawa, Canada. He is the author/editor of eight books, including Making Things International 1 and 2 (2015 and 2016). He is also Editor-in-Chief of the journal Security Dialogue. Can E. Mutlu is Associate Professor of Global Politics at Acadia University in Wolfville, NS, Canada. His research interests include borders, migration, technology, and security. He is the co-editor of Architectures of Security: Design, Control, Mobility (with Benjamin J. Muller). Philippe M. Frowd is Associate Professor in the School of Political Studies at the University of Ottawa, Canada. His research focuses on the governance of irregular migration and border control in the Sahel region of West Africa. He is the author of Security at the Borders (2018).

Critics' Reviews

‘Questions of method have become increasing pertinent to the pedagogies and research practices of critical security studies. Research Methods in Critical Security Studies (2nd edition) makes a timely contribution by providing a range of answers to these questions. In doing so, RMCSS strikes a judicious balance that will appeal to seasoned researchers looking to adopt new approaches as well as students who may be embarking upon their first substantive research project in the field. While richly informed by cutting edge conceptual, methodological, and theoretical literature, the discussions are practical, precise, and plain-spoken—they cut straight to the chase in order to equip the reader with capabilities to do reflexive research in critical security studies and navigate common challenges found within and across methods. With new chapters and updated materials, the 2nd edition captures recent developments within the field while maintaining the accessibility and pragmatism that were hallmarks of the first edition. As such, the 2nd edition is an excellent teaching and research resource for everyone in the field.’ Kyle Grayson, Newcastle University, UK 'This volume shows how doing critical and reflexive research can go hand-in-hand with rigorous methodology. The book is indispensable to researchers in critical security studies broadly defined, from graduate student to project leader. It is filled with useful practical examples and fascinating case studies. I have used it in my thesis seminar for years, and it is great that we now have an updated and expanded second edition, that combines attention to state-of-the-art theory with clear advice on practical means and modes of doing research.' Marieke de Goede, University of Amsterdam, Netherlands 'If ''methods'' are off-the-shelf tools that can be casually picked up and deployed, then this is not a methods book. It is instead an invitation to critical inquiry, and a rich tapestry of examples showing how attitudes of reflexivity and a healthy skepticism about received concepts and categories are in no way incompatible with clear and sustained attention to questions of research design. This is a rich feast for critical researchers to devour.' Patrick Thaddeus Jackson, American University, USA

About VitalSource eBooks

VitalSource is a leading provider of eBooks.

• Access your materials anywhere, at anytime.
• Customer preferences like text size, font type, page color and more.
• Take annotations in line as you read.

Multiple eBook Copies

This eBook is already in your shopping cart. If you would like to replace it with a different purchasing option please remove the current eBook option from your cart.

Book Preview

The country you have selected will result in the following:

• Product pricing will be adjusted to match the corresponding currency.
• The title Perception will be removed from your cart because it is not available in this region.

We’re fighting to restore access to 500,000+ books in court this week. Join us!

Internet Archive Audio

• This Just In
• Grateful Dead
• Old Time Radio
• 78 RPMs and Cylinder Recordings
• Audio Books & Poetry
• Computers, Technology and Science
• Music, Arts & Culture
• News & Public Affairs
• Spirituality & Religion
• Radio News Archive

• Flickr Commons
• Occupy Wall Street Flickr
• NASA Images
• Solar System Collection
• Ames Research Center

• All Software
• Old School Emulation
• MS-DOS Games
• Historical Software
• Classic PC Games
• Software Library
• Kodi Archive and Support File
• Vintage Software
• CD-ROM Software
• CD-ROM Software Library
• Software Sites
• Tucows Software Library
• Shareware CD-ROMs
• Software Capsules Compilation
• CD-ROM Images
• ZX Spectrum
• DOOM Level CD

• Smithsonian Libraries
• FEDLINK (US)
• Lincoln Collection
• American Libraries
• Canadian Libraries
• Universal Library
• Project Gutenberg
• Children's Library
• Biodiversity Heritage Library
• Books by Language
• Additional Collections

• Prelinger Archives
• Democracy Now!
• Occupy Wall Street
• TV NSA Clip Library
• Animation & Cartoons
• Arts & Music
• Computers & Technology
• Cultural & Academic Films
• Ephemeral Films
• Sports Videos
• Videogame Videos
• Youth Media

Search the history of over 866 billion web pages on the Internet.

Mobile Apps

• Wayback Machine (iOS)
• Wayback Machine (Android)

Browser Extensions

Archive-it subscription.

• Explore the Collections
• Build Collections

Save Page Now

Capture a web page as it appears now for use as a trusted citation in the future.

Please enter a valid web address

• Donate Donate icon An illustration of a heart shape

Research methods in critical security studies : an introduction

Bookreader item preview, share or embed this item, flag this item for.

• Graphic Violence
• Explicit Sexual Content
• Hate Speech
• Misinformation/Disinformation
• Marketing/Phishing/Advertising
• Misleading/Inaccurate/Missing Metadata

plus-circle Add Review comment Reviews

Better World Books

DOWNLOAD OPTIONS

No suitable files to display here.

IN COLLECTIONS

Uploaded by station27.cebu on June 15, 2023

SIMILAR ITEMS (based on metadata)

Breadcrumbs Section. Click here to navigate to respective pages.

Research Methods in Critical Security Studies

DOI link for Research Methods in Critical Security Studies

Get Citation

This new textbook surveys new and emergent methods for doing research in critical security studies, thereby filling a large gap in the literature of this emerging field.

New or critical security studies is growing as a field, but still lacks a clear methodology; the diverse range of the main foci of study (culture, practices, language, or bodies) means that there is little coherence or conversation between these four schools or approaches.

In this ground-breaking collection of fresh and emergent voices, new methods in critical security studies are explored from multiple perspectives, providing practical examples of successful research design and methodologies. Drawing upon their own experiences and projects, thirty-three authors address the following turns over the course of six comprehensive sections:

• Part I: Research Design
• Part II: The Ethnographic Turn
• Part III: The Practice Turn
• Part IV: The Discursive Turn
• Part V: The Corporeal Turn
• Part VI: The Material Turn

This book will be essential reading for upper-level students and researchers in the field of critical security studies, and of much interest to students of sociology, ethnography and IR.

TABLE OF CONTENTS

Chapter | 14  pages, introduction, part i | 35  pages, research design, chapter | 9  pages, chapter | 4  pages, wondering as research attitude, criticality1, do you have what it takes, chapter | 5  pages, attuning to mess1, empiricism without positivism1, engaging collaborative writing critically, part ii | 33  pages, the ethnographic turn, chapter | 7  pages, travelling with ethnography, reflexive inquiry, listening to migrant stories, learning by feeling, how participant observation contributes to the study of (in)security practices in conflict zones1, dissident sexualities and the state, part iii | 28  pages, the practice turn, the practice of writing, researching anti-deportation, act different, think dispositif1, expertise in the aviation security field, testifying while critical, part iv | 25  pages, the discursive turn, legislative practices, medicine and the psy disciplines, speech act theory, part v | 34  pages, the corporeal turn, affect at the airport, emotional optics, affective terrain, chapter | 3  pages, theorizing the body in ir, reading the maternal body as political event1, corporeal migration, part | 34  pages, the material turn, infrastructure1, the internet as evocative infrastructure, the study of drones as objects of security, objects of security/objects of research, pictoral texts, tracing human security assemblages.

• Privacy Policy
• Terms & Conditions
• Cookie Policy
• Taylor & Francis Online
• Taylor & Francis Group
• Students/Researchers
• Librarians/Institutions

Connect with us

Registered in England & Wales No. 3099067 5 Howick Place | London | SW1P 1WG © 2024 Informa UK Limited

Sorry, there was a problem.

Download the free Kindle app and start reading Kindle books instantly on your smartphone, tablet, or computer - no Kindle device required .

Read instantly on your browser with Kindle for Web.

Using your mobile phone camera - scan the code below and download the Kindle app.

Image Unavailable

• To view this video download Flash Player

Research Methods in Critical Security Studies: An Introduction 1st Edition

This new textbook surveys new and emergent methods for doing research in critical security studies, thereby filling a large gap in the literature of this emerging field.

New or critical security studies is growing as a field, but still lacks a clear methodology; the diverse range of the main foci of study (culture, practices, language, or bodies) means that there is little coherence or conversation between these four schools or approaches.

In this ground-breaking collection of fresh and emergent voices, new methods in critical security studies are explored from multiple perspectives, providing practical examples of successful research design and methodologies. Drawing upon their own experiences and projects, thirty-three authors address the following turns over the course of six comprehensive sections:

• Part I: Research Design
• Part II: The Ethnographic Turn
• Part III: The Practice Turn
• Part IV: The Discursive Turn
• Part V: The Corporeal Turn
• Part VI: The Material Turn

This book will be essential reading for upper-level students and researchers in the field of critical security studies, and of much interest to students of sociology, ethnography and IR.

• ISBN-10 0415535395
• ISBN-13 978-0415535397
• Edition 1st
• Publisher Routledge
• Publication date October 22, 2012
• Language English
• Dimensions 7 x 0.75 x 9.75 inches
• Print length 256 pages
• See all details

Editorial Reviews

"Finally, critical security studies has its own methodological handbook. It is not only extremely broad in scope, applying methods ranging from participant observation to interviews to discourse analysis, and discussing research design, ethnography, empiricism and writing. But it is also refreshingly reflexive in its approach. Its exploration of method is intimately bound to an advancement of theory, and a critical reflection on the role of the researcher in this sensitive – and often secretive –domain. It is indispensable reading for researchers and students alike, and promises to take this important field of research to a new level."– Marieke de Goede, University of Amsterdam, The Netherlands

'This textbook moves critical security studies forward in important ways by restoring "methodology" to the full sense of concept and rescuing it from the narrowness imposed by mainstream social science. ’ -- Roxanne Lynn Doty, Arizona State University, USA

'Wide-ranging and yet systematic, rigorous and yet pluralistic, this volume makes a crucial contribution toward developing innovative methodologies able to terms with the rapidly changing politics of contemporary security. Combining sophisticated conceptual overviews with illustrations of specific research designs in practice, it is a remarkably valuable resource for students and researchers, as well as an inspiring tour d’ horizon of cutting-edge research. ' -- Michael C Williams, University of Ottawa, Canada

About the Author

Mark B. Salter is Professor at the School of Political Studies, University of Ottawa, Canada. He is editor of Mapping Transatlantic Security Relations (Routledge 2010), and author of Rights of Passage: The Passport in International Relations (2003) and Barbarians and Civilization in International Relations (2002) .

Can E. Mutlu is a PhD candidate (ABD) at the School of Political Studies at the University of Ottawa, Canada. He is the Communications Director of the International Political Sociology Section of the International Studies Association (IPS-ISA).

Product details

• Publisher ‏ : ‎ Routledge; 1st edition (October 22, 2012)
• Language ‏ : ‎ English
• Hardcover ‏ : ‎ 256 pages
• ISBN-10 ‏ : ‎ 0415535395
• ISBN-13 ‏ : ‎ 978-0415535397
• Item Weight ‏ : ‎ 1.23 pounds
• Dimensions ‏ : ‎ 7 x 0.75 x 9.75 inches

Customer reviews

• 5 star 4 star 3 star 2 star 1 star 5 star 57% 43% 0% 0% 0% 57%
• 5 star 4 star 3 star 2 star 1 star 4 star 57% 43% 0% 0% 0% 43%
• 5 star 4 star 3 star 2 star 1 star 3 star 57% 43% 0% 0% 0% 0%
• 5 star 4 star 3 star 2 star 1 star 2 star 57% 43% 0% 0% 0% 0%
• 5 star 4 star 3 star 2 star 1 star 1 star 57% 43% 0% 0% 0% 0%

Customer Reviews, including Product Star Ratings help customers to learn more about the product and decide whether it is the right product for them.

To calculate the overall star rating and percentage breakdown by star, we don’t use a simple average. Instead, our system considers things like how recent a review is and if the reviewer bought the item on Amazon. It also analyzed reviews to verify trustworthiness.

• Sort reviews by Top reviews Most recent Top reviews

Top reviews from the United States

There was a problem filtering reviews right now. please try again later..

Top reviews from other countries

• About Amazon
• Investor Relations
• Amazon Devices
• Amazon Science
• Sell products on Amazon
• Sell on Amazon Business
• Sell apps on Amazon
• Become an Affiliate
• Advertise Your Products
• Self-Publish with Us
• Host an Amazon Hub
• › See More Make Money with Us
• Amazon Business Card
• Shop with Points
• Reload Your Balance
• Amazon Currency Converter
• Amazon and COVID-19
• Your Account
• Your Orders
• Shipping Rates & Policies
• Returns & Replacements
• Manage Your Content and Devices

• Conditions of Use
• Privacy Notice
• Consumer Health Data Privacy Disclosure
• Your Ads Privacy Choices

(Stanford users can avoid this Captcha by logging in.)

• Send to text email RefWorks EndNote printer

Research methods in critical security studies : an introduction

Available online.

• Taylor & Francis eBooks

More options

• Contributors

Description

Creators/contributors, contents/summary, bibliographic information, browse related items.

• Stanford Home
• Maps & Directions
• Search Stanford
• Emergency Info
• Terms of Use
• Non-Discrimination
• Accessibility

© Stanford University , Stanford , California 94305 .

• About the School
• Research Courses
• Other Activities

PhD Course: Research Methods in Critical Security Studies

Coordinators and primary lecturers: J. Peter Burgess (PRIO) and Mark B. Salter (University of Ottawa).  

Guest lecturers: Claudia Aradau (King's College, London); Thierry Balzacq (University of Namur, Belgium); Xymena Kurowska (Central European University); Maria Stern (University of Gothenburgh).

• Dates: 14-18 October 2013
• Venue: PRIO, Oslo, Norway
• Credits: 5 ECTS
• Contact and application: [email protected]
• Deadline of application: 15 September 2013.

This course provides an introduction and overview to a range of research methods in critical security studies. Its aim is to provide tools and methods to students of critical security studies in support of clear research design and rigorous scholarly methods.  

Critical security studies can be understood as a scholarly approach that is attentive to the workings of power and exclusion inherent in social phenomena. Though objects of research can vary considerably, a four basic principles shape the field of critical analysis:

• Social and political life are interwoven without any one unifying principle or logic;
• Agency--the capacity to act--is not reserved to individual human beings, but rather is everywhere;
• Causality is emergent. In other words, critical analysis does not identify what necessarily happens, but rather what the conditions of possibility of something happening are;
• Research, writing and public engagement are inherently political.

Lectures and discussions will emphasize reapplication of classical scientific research questions for the field of critical security studies: sufficient proof, critical position, and coherency of argument, reshaped and reapplied to these four principles.

The course will build primarily on Research Methods in Critical Security Studies: An Introduction. (Mark B. Salter and Can E. Mutlu, London: Routledge), supplemented by other readings. Lectures will be given by external scholars known for their work in the field of critical security studies. The 5-day course will combine morning lectures with afternoon workshopping of relevant themes, articles, and student papers.

Requirements

One week prior to the course lectures, each student must submit a brief research proposal, related to a current or future project (approximately 500 words). The research proposal should relate to the readings and contain 3 different research paths for achieving its goals. 

In order to achieve 5 ECTS, an essay of 3-5000 words must be handed in by 20 November 2013. The essay question should be proposed by 22 October for approval by the organizers. The participants are invited to relate the essay to the methods chapter of their dissertation. 

The deadline of application is 15 September 2013. Applicants should include details about their university affiliation and a paragraph on their doctoral research (except for members of the research school, who just register). Please send your application by e-mail to the Research School Coordinator, Kristoffer Lidén at [email protected] .

There is no participation fee, but the cost of travel and accommodation, if needed, must be covered by the participants. No financial assistance is available. Applicants will be notified about the outcome of their application as quickly as possible after the deadline.

Day 1 / 14 October : Introduction to critical security research methods and design

10.15-12.00 Lecture (Burgess & Salter). Readings:

• Salter, ‘Research design: Introduction’
• Guillaume, ‘Criticality’
• Squire, ‘Attending to mess’
• c.a.s.e. collective, ‘Critical approaches to security in Europe’

13.15-15.00 Seminar

Day 2 / 15 October : Discourse and discursive approaches

10.15-12.00 Lecture (Thierry Balzacq). Readings:

• Balzacq, ‘Enquiries into methods’
• Anaïs, ‘Objects of security’
• Lobo-Guerrero, ‘Archives’
• Neal, ‘Legislative practices’

13.15-15.00 Seminar 

Day 3 / 16 October : The practical and the role of the observer

10.15-12.00 (Xymena Kurowska). Readings:

• Hughs, ‘The practice of writing’
• Law, ‘After method: An introduction’
• Friedrichs and Kratochwil, ‘On acting and knowing’
• Kurowska and Tallis, ‘Chaismatic crossings’

Day 4 / 17 October : The new materiality

10.15-12.00 (Claudia Aradau). Readings:

• Aradau, ‘Infrastructure’
• Voelkner, ‘Tracing human security assemblages’
• Barad, ‘Posthumanist performativity: toward an understanding of how matter  comes to matter’
• Adey and Anderson, ‘Anticipating emergencies’

Day 5 / 18 October : Gender and the corporeal turn

10.15-12.00 (Maria Stern). Readings:

• Wiebe, ‘Affective terrain: approaching the field in Aamjiwnaang’
• Shinko, ‘Theorizing the body in IR’
• Crane-Seeber, ‘Learning by feeling’
• Shinko, ‘Ethics after liberalism: Why (autonomous) bodies matter’.

Constructing a Novel Network Structure Weighting Technique into the ANP Decision Support System for Optimal Alternative Evaluation: A Case Study on Crowdfunding Tokenization for Startup Financing

• Research Article
• Open access
• Published: 26 August 2024
• Volume 17 , article number  222 , ( 2024 )

Cite this article

You have full access to this open access article

• Chun-Yueh Lin 1  

This study constructed a novel decision-making framework for startup companies to evaluate token financing options. A Network structure weighting (NSW) technique was developed and integrated with the analytic network process (ANP) to create a comprehensive assessment model. This innovative approach addressed the limitations of traditional multi-criteria decision-making methods by effectively capturing the complex interdependencies between factors influencing token financing decisions. The proposed model comprises three main steps: (1) utilizing a modified Delphi method to identify key factors affecting token financing, (2) developing the NSW technique to determine the network structure of these factors, and (3) integrating the NSW results into the ANP model to evaluate and rank the critical factors and alternatives. This study applied this framework to assess three token financing alternatives: Initial Coin Offerings (ICO), Initial Exchange Offerings (IEO), and Security Token Offerings (STO). The results indicate that STO is the optimal financing alternative for the analyzed startup scenario in token financing, followed by Initial Exchange Offerings and Initial Coin Offerings. The model identified platform fees, issuance costs, and financing success rate as the three most critical factors influencing the decision. This study contributes to both methodology and practice in FinTech decision-making. The NSW-ANP framework offers a more robust approach to modeling complex financial decisions, while the application to token financing provides valuable insights for startup companies navigating this emerging funding landscape. The proposed framework lays the groundwork for more informed and structured decision-making in the rapidly evolving field of cryptocurrency-based financing.

Explore related subjects

• Artificial Intelligence

Avoid common mistakes on your manuscript.

1 Introduction

Due to the rise and development of Financial Technology (FinTech), as well as the enactment of the Jumpstart Our Business Startups (JOBS) in the U.S. [ 1 ], crowdfunding has become the newest financing means for enterprises in need of external funds [ 2 , 3 ]. In 2014, the total amount of funds raised through crowdfunding reached USD 16.2 billion, which was 167% higher than that of 2013 [ 4 ]. In addition, according to the statistical results of Statista Inc. (2020) [ 5 ], the total amount of alternative financing in 2020 was USD 6.1 billion, among which crowdfunding accounted for the largest market share. For this reason, it could be said that the development scale of crowdfunding in the global financial market has been rocketing.

Crowdfunding involves a number of different forms. The first form is donation-based crowdfunding, which mainly means to raise charity funds for the implementation of programs and projects. The second form is rewards-based crowdfunding, in which the investor can receive non-monetary rewards because of capital contributions. The third form is debt-based crowdfunding, in which the relevant interest arrangements between the investor and the fundraiser are determined in line with credit contracts. The fourth form is equity-based crowdfunding, in which the fundraiser uses the equities of the target company to exchange funds from the investor, while the investor receives such equities and therefore is entitled to that company’s revenues or dividends [ 6 , 7 , 8 ]. Estrin et al. [ 9 ] pointed out that equity-based crowdfunding depends mainly on the Internet or social network platforms. This fund-raising method not only reduces the transaction cost but also stands for a new business pattern under which startup companies can establish their own goodwill and provide investors with opportunities for investment. Although crowdfunding has many advantages for startup companies, risks do exist, including uncertainty of equity ownership, lack of liquidity, and damage to stockholder equity [ 10 , 11 , 12 ]. For this reason, past studies suggested that startup companies might obtain funds by offering tokens on the basis of distributed ledger technology and the immutability of blockchains. This not only could reduce the potential risks of traditional fundraising platforms but also could promote the transparency level of the relevant transactions [ 12 , 13 , 14 ]. Howell et al. [ 15 ] indicated that token financing has become one of the important sources for enterprises to raise funds through digital platforms. Presently, the development of crowdfunding tokenization mainly involves three patterns: (1) initial coin offerings (ICO), (2) initial exchange offerings (IEO), and (3) security token offerings (STO). ICO has the advantages of low cost and high speed. However, the risks of theft and fraud exist [ 15 , 16 , 17 ]. The advantages of IEO include having the business reputation of a third-party platform as a guarantee and handling the relevant transactions directly on the transaction platform. However, the possibility of the token price being manipulated cannot be ruled out [ 17 , 18 ]. The last pattern, STO, has the advantages of the highest level of safety and of being protected by the rules and regulations of regional governments. However, the high complexity of examination and verification as well as excessively low liquidity are problems that cannot be avoided [ 17 , 19 ]. The research results of past literature also show that for startup companies, the efficiency of token financing is higher than that of equity financing [ 20 ]. Furthermore, Chod et al. [ 14 ] pointed out that enterprises may take advantage of the decentralization features of token financing to make it more convenient for token investors in their project investments and reduce the cost of encouraging token investors to join the investment platforms. In this way, it is easier for entrepreneurs in raising funds.

For this reason, the utilization of token financing for the purpose of raising operation efficiency has become an important business strategy. The aforesaid three patterns of crowdfunding tokenization have their respective advantages and disadvantages, as well as potential risks. If startup companies intend to raise funds through virtual currencies, the alternatives of financing in cryptocurrency will affect the financing efficiency and lead to the capital turnover issue. Previous studies on token financing focused more on risk-return analysis [ 21 , 22 , 23 , 24 ], token rules and regulations [ 25 , 26 , 27 ], hedging of tokens [ 28 , 29 , 30 , 31 ], and prediction of price in tokens [ 32 , 33 , 34 , 35 ]. However, there is scarce evidence and a lack of applicable measurement tools in regard to assessing the optimal solution for the token financing of startup companies. Hence, algorithms for multiple criteria decision-making can be utilized for the construction of assessment models, so that the optimal solution for assessment can be reached [ 36 , 37 , 38 ]. Past studies also suggested that the optimal solution can be solved using the analytic hierarchy process (AHP) [ 38 , 39 , 40 , 41 , 42 ]. Although AHP can be used to assess the optimal solutions in different fields, it is unsuitable to use traditional AHP methods for decision-making problems in real situations. AHP is characterized by a hierarchical structure and based upon the presumption that the variables or criteria are independent from each other. Numerous problems relating to the assessment of optimal solutions and the relevant variables are correlated to or dependent on each other; as a result, complicated internal relationships cannot be solved through hierarchical or independent methods [ 43 , 44 ]. To solve this problem, Saaty [ 45 ] proposed the analytic network process (ANP), which added a feedback mechanism and interdependency to the AHP method to solve the problems of a lack of correlation and interdependency. ANP does not require the linear relationship of traditional AHP methods, which is top-down, and can establish an assessment pattern of networked relationships. Past literature has applied ANP models in the assessment of different industries, such as traffic problems [ 46 , 47 ], environment and energy assessment [ 48 , 49 , 50 ], filtration and selection of suppliers [ 51 , 52 , 53 ], and assessment of risk factors [ 54 , 55 , 56 ]. Thus, it can be seen that the problem of correlation or interdependency between criteria or variables cannot be solved effectively through AHP during decision-making, while ANP can effectively solve this problem. Although ANP can overcome the difficulties related to the presumption of independence in AHP, the ANP algorithm cannot ascertain the strength of the dependence and relationships between variables needed to generate a network structure. Previous studies addressing the network structure issue have applied deep machine learning concepts, as demonstrated by Moghaddasi et al., Gharehchopogh et al., and subsequent works by Moghaddasi et al. [ 57 , 58 , 59 , 60 , 61 ]. However, these studies primarily focused on the relationship in the Internet of Things, implicitly highlighting the challenges in applying such approaches to multi-criteria decision-making (MCDM) problems. Additionally, several studies employed the Decision-Making Trial and Evaluation Laboratory (DEMATEL) method to resolve network structures among criteria [ 62 , 63 , 64 , 65 , 66 ]. This approach offers an alternative perspective on capturing complex interrelationships within decision-making frameworks. However, the DEMATEL method has several limitations. First, the relationships derived through DEMATEL may be biased or misleading [ 67 , 68 ]. Additionally, the method faces convergence issues, as it cannot determine relationships between criteria when the data fail to converge [ 69 ]. As evident from Table  1 , there are two primary gaps in the existing literature. First, in terms of network structure methodology, while ANP, DEMATEL, and other decision-making frameworks have been proposed, they each have limitations. Second, regarding the research problem, while many studies have examined different aspects of token financing, there is a notable absence of comprehensive, quantitative decision-making frameworks specifically designed for startup companies evaluating token financing alternatives. In view of the above, this study developed a new network structure weighting (NSW) model, and then integrated NSW into ANP to remedy ANP’s shortcoming of being unable to determine the network structure. Finally, case studies were carried out to assess the optimal solution for startup companies engaging in token financing.

For the proposed NSW-ANP model, the modified Delphi method was utilized to determine the clusters and factors influencing startup companies engaging in token financing. Then, the network structure of these clusters and factors was determined based on the NSW method. Finally, the ANP model was utilized to calculate the weights of various factors and financing schemes for startup companies engaging in token financing and then sequence them to determine the optimal token financing schemes and their key factors. While ANP has been applied in various fields, this study proposed the first application of an enhanced ANP approach (integrated with NSW) to evaluate the token financing options for startups. This novel application demonstrates the versatility and effectiveness of our integrated approach in addressing complex FinTech decision-making scenarios.

This study makes significant contributions to the existing literature in both methodological innovations and novel applications. In terms of methodological advancements, we introduce a novel NSW technique that quantifies the strength of relationships between decision factors in a network structure. Furthermore, we develop an integrated NSW-ANP framework that enhances the capabilities of the traditional ANP by incorporating a more robust method for determining network relationships. With regard to novel applications, this study breaks new ground in two key areas. Firstly, we apply this integrated NSW-ANP framework to evaluate token financing options for startup companies, an area that has not been addressed using such a comprehensive decision-making approach. Secondly, this study provides the first systematic evaluation of ICO, IEO, and STO using a multi-criteria decision-making framework. This framework resolves the complex interdependencies between various factors, offering a more nuanced understanding of these emerging financing mechanisms. By combining methodological innovation with practical application in an emerging field, this study not only advances the theoretical understanding of multi-criteria decision-making processes but also provides valuable insights for practitioners in cryptocurrency-based startup financing. Academically, the new NSW-ANP model put forward in this study could be used for determining the network relationship of a research structure, and be integrated into the ANP to remedy the ANP’s shortcomings. The new integrated decision-making pattern put forward in this study also could provide valuable references for the measurement of the interdependency and correlation among variables in the assessment of the optimal solution of token financing for startup companies. Practically, the proposed framework could provide startup companies with a measurement tool containing a network structure and is valuable, so as to determine the optimal solution of token financing for startup companies introducing token financing to their businesses.

The remainder of this paper is organized as follows: Sect.  1 is the introduction, Sect.  2 describes the research method, Sect.  3 presents the case study, and Sect.  4 offers the conclusions.

2 Methodology

In this study, the clusters and factors were acquired through collecting experts’ opinions and literature reviews via modified Delphi method (MDM) as a first step. Next, the network structure of the clusters and factors was determined on the basis of the network structure weighting (NSW) method. Finally, the analytic network process (ANP) model was utilized to calculate and sequence the weightings of the various factors and financing schemes of startup companies engaging in token financing so that the most suitable token financing scheme and the key factors could be determined. The research method is presented in the following sections.

The Delphi method is an anonymous technique of decision-making by a group of experts. To solve a certain problem or find a solution for a particular future event, these experts are treated as the appraisal targets. For the final goal of reaching a stable group consensus among the experts, the group members are anonymous to each other, and particular procedures and repetitive steps are employed. The Delphi method attempts to combine the knowledge, opinions, and speculative abilities of experts in the field in an interruption-free environment. The Delphi method can be used to deduce what will happen in the future, effectively predict future trends, or reach a consensus over a certain issue [ 70 , 71 ]. This method is based upon the judgment of experts, and multiple rounds of opinion feedback are utilized to solve complicated decision-making problems. The traditional Delphi method emphasizes the following five basic principles [ 72 , 73 ]:

The principle of anonymity: All experts voice their opinions as individuals, and they remain anonymous when doing so.

Iteration: The questionnaire issuer gathers up the experts’ opinions and sends them to other experts. This step is carried out repeatedly.

Controlled feedback: In each round, the experts are required to answer pre-designed questionnaires, and the results are served as references for the next appraisal.

Statistical group responses: Comprehensive judgments are made only after the statistics of all the experts’ opinions are conducted.

Expert consensus: The ultimate goal is to reach a consensus after the experts’ opinions are consolidated.

The procedures of the Delphi method are as follows [ 74 ]:

Select the anonymous experts.

Carry out the first round of the questionnaire survey.

Carry out the second round of the questionnaire survey.

Carry out the third round of the questionnaire survey.

Consolidate the experts’ opinions and reach a consensus.

According to the modified Delphi method, Steps C and D are carried out repeatedly until a consensus is reached among the experts, and the number of experts should be between five and nine [ 75 , 76 ].

In this study, the experts’ opinions were gathered through the Delphi method and the relevant literature was discussed, so that the clusters and factors influencing startup companies engaging in token financing could be obtained.

2.2 NSW Model

This study utilized the Delphi method to collect the clusters and factors that could influence startup companies engaging in token financing schemes. In order to effectively carry out the calculation and assessment of ANP, the network structure of these clusters and factors need to be determined as a prerequisite for the subsequent filtration and selection of the optimal token financing scheme. Therefore, this study put forward the NSW method in order to acquire the relationships and the structure chart between clusters and factors. The NSW procedure is as follows:

Step 1: Collect and confirm the decision factors

The collection and confirmation of the decision factors can be realized through common tools such as literature reviews, the Delphi method, focus group interviews, and brainstorming. When decision-makers or experts need to determine n assessment factors that are consistent with the decision-making issues, the n assessment factors may be defined as \(\{ C_{1} ,C_{2} , \ldots ,C_{n} \}\) .

Step 2: Design the questionnaire

As far as the n factors determined by the decision makers or experts in Step 1 are concerned, a nine-point Likert scale can be utilized to ascertain the correlation and correlation strength between the factors. In the event of n factors, n ( n  − 1) comparisons in line with the scale need to be carried out.

Step 3: Calculate the weight of the network structure

Each expert compares and scores the decision factors. After that, all the comparison scores of the experts are used in the matrix construction and weighted calculation. The procedure is as follows:

2.2.1 Establish the Matrix of the Network Correlation and the Correlation Diagram

The correlation matrix is established as M , while \(\{ C_{1} ,C_{2} , \ldots ,C_{n} \}\) are the decision factors. If C i is influenced by C j , \(m_{ij}\) will be the scores of a quantitative judgment given by experts. On the contrary, if \(m_{ij} = 0\) , C i is not influenced by C j . The results can be shown in matrix M ( n  ×  n ) as follows:

The column aggregation and row aggregation of matrix M are:

\({\text{Column}}_{j}\) and \({\text{row}}_{i}\) , respectively, give the scores of factor j , which affects other factors, or factor j , which is influenced by other factors.

2.2.2 Define the Transition Probability Matrix

If transition matrix A is defined by the features of the Markov chain, A  = ( a ij ), as shown in Eq. ( 2 ). A is a regular Markov matrix, and the existence of stationary distribution \(x = \left( {x_{1} ,x_{2} , \ldots ,x_{N} } \right)^{T}\) satisfies Ax  =  x and \(\sum\nolimits_{i} {x_{i} = 1}\) . The characteristic value of 1 can be acquired through the characteristic vector corresponding to the characteristic value of Matrix A , or through the iteration method \(x^{0}\) , where \(x^{k + 1} = Ax^{k}\) , to obtain the characteristic value. x stands for the distribution of probabilities of the various factors being influenced when the transition number approaches infinity, and \(x_{i}\) stands for the network node score of the i th factor.

2.2.3 Calculate the Weightings of the Network Structure

According to the results described in II above, the network node score of each factor is distributed to the correlation diagram of each expert ( n experts have n correlation diagrams). Afterwards, based on the node score of factor i , the strength score of each expert’s factor i influencing other factor j goes through a standardized distribution using the correlation diagram to obtain each expert’s weighted value of the network structure, R, as shown in Eq. ( 3 ). In the end, the \(R(C_{i} ,C_{j} )\) of n experts is averaged and standardized, as shown in Eq. ( 4 ) and Eq. ( 5 ). The standardized results can then be integrated into the ANP model to assess the optimal token financing scheme for startup companies.

Saaty put forward ANP in 1996. This method is rendered through a network structure and derived from an ANP. Practically, there are many questions about decision-making assessment that are not limited to expressing their complex interrelated properties in a hierarchical and independent manner, and they are not of purely linear relationships either. Rather, these questions have a network-like structure [ 45 , 77 , 78 , 79 ]. Based on the original presumption and prerequisite of the analytic hierarchy process (AHP), Saaty [ 45 ] integrated relationship and feedback mechanisms into the AHP model to solve the problem of correlation between different principles.

Saaty pointed out that the relationships of interactive influence between clusters and elements can be analyzed in a graphic manner. Such relationships and interactive influence can be demonstrated through arrow lines [ 45 , 80 ], as shown in Fig.  1 . This network structure is crucial for understanding the fundamental difference between hierarchical and network-based decision-making models. Unlike traditional hierarchical structures, this network allows for complex interdependencies between different elements of the decision-making process. In Fig.  1 , the bidirectional arrows indicate that influence can flow both ways between clusters, reflecting real-world complexities where factors can mutually affect each other.

Source : Ref. [ 45 ]

The network structure.

According to the relationships and strengths of different factors in the aforesaid models and structure charts of ANP, a supermatrix is utilized for demonstration, as shown in Fig.  2 . This matrix is a critical component of the ANP, allowing for the quantification of relationships between all elements in the network. It is formed when the various clusters and respective factors contained in such clusters are listed on the left side and upper part of the matrix in an orderly manner. The supermatrix consists of a number of sub-matrices, which are formulated based on the eigenvectors after the comparison of different factors. In Fig.  2 , \(W_{11} ,W_{kk} , \ldots ,W_{nn}\) are the values of the eigenvectors after the comparisons and calculations.

Source : Refs. [ 45 ] [ 80 ]

The supermatrix of a network.

ANP is an algorithm based on AHP and can be divided into four steps. In Step 1, the structures are formed step by step. In Step 2, the questions are raised. In Step 3, comparisons of interdependent clusters are made in pairs and a supermatrix is formed. In Step 4, the ultimate choice and optimal scheme are selected [ 45 , 79 ].

This study apples the ANP as the foundation of our approach due to several key advantages it offers in the context of complex decision-making scenarios. First, it is well-suited for this application because it allows for the consideration of interdependencies and feedback relationships between decision factors, which is crucial in the dynamic and interconnected world of FinTech and token financing. Furthermore, it provides a structured approach to incorporating both qualitative and quantitative factors into the decision-making process. This is particularly beneficial when evaluating token financing options, as it allows us to consider both qualitative and quantitative data. Finally, it is able to prioritize alternatives based on a comprehensive set of criteria and sub-criteria. This is especially valuable when comparing different alternatives, each of which has its own unique set of characteristics and implications. ANP allows for a more comprehensive comparison than simpler decision-making tools. Among various MCDM techniques, the ANP has a superior capacity to model complex systems with intricate interdependencies. While other MCDM techniques, such as the Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) and the VIseKriterijumska Optimizacija I Kompromisno Resenje (VIKOR) method, offer effective means for ranking alternatives, they exhibit limitations in accounting for the multifaceted interrelationships among criteria.

Consequently, this study employs the ANP method as the foundation for constructing an integrated decision-making model. A brief introduction of the construction program of the network process pattern is as follows:

Step 1: Confirm the research problems and network structure

Determine the targets according to features of the problems and search for decision-making clusters, as well as the factors contained in the various clusters by employing the proposed NSW method to acquire the influencing strength of the various factors; finally, draw the network structure models of the decision-making problems according to the results of NSW.

Step 2: Create pair-wise comparison matrices and priority vectors

Compare the factors in pairs. This step has two parts: the comparison of clusters (in pairs) and the comparison of factors within clusters (in pairs). The comparison of factors within clusters (in pairs) can be divided into the comparison within a particular group and comparisons among different clusters. The assessment scale of the comparison is similar to that of AHP. In addition, the eigenvectors, which are reached through the various comparison matrices, serve as the values of the supermatrix, which can be used to illustrate the interdependency and relative significance among the clusters. Equation ( 6 ) can be utilized to calculate the scores of relative significance in regard to the various clusters and factors. As for the strength of the interdependency among the clusters and among the factors, NSW can be utilized to determine the network structure (as described in Sect.  2.2 .)

Step 3: Construct the supermatrix

The supermatrix can effectively solve problems related to the interdependency among the various clusters and factors within the system (as shown in Fig.  2 ). The values of the supermatrix consist of small matrices, which include the comparison of different factors (in pairs) and the comparison of interdependent factors (in pairs). The numerical values of clusters or factors without the influence of feedback are 0, as shown in Eq. ( 7 ). In this study, it was suggested that the overall network structure could be confirmed by NSW. For this reason, the NSW results were integrated into the supermatrix for subsequent assessment and to determine the strength of the interdependency in the supermatrix, as shown in Eq. ( 8 ).

The ANP calculation process includes three matrices: the unweighted supermatrix, the weighted supermatrix, and the limit supermatrix. The unweighted supermatrix stands for the weightings of the original results of the comparison in pairs. In the weighted supermatrix, the weighted values of a particular element within an unweighted matrix are multiplied by the weighted values of the relevant clusters. In the limit supermatrix, the weighted matrix multiplies itself repeatedly until a stable state is attained. According to ANP, if supermatrix W is in an irreducible state of stability, all columns in the supermatrix will have similar vectors, indicating convergence can be attained. The ultimate weighted values of each cluster, factor, and scheme can be calculated through Eq. ( 9 ) during the convergence process.

Step 4: Evaluate the optimal alternative

Through the ANP framework and the calculations of the unweighted supermatrix, weighted supermatrix, and limit supermatrix, all the alternative schemes, as well as the ultimate values of the groups and factors, can be attained in the limit supermatrix. The ultimate results of the weighted values are then ranked to determine the optimal scheme.

3 Case Study

This study aimed to establish the network structure weighting (NSW) model by integrating NSW into the analytic network process (ANP) and establishing an assessment pattern to analyze the optimal scheme of token financing for startup companies, as well as the weighted values of clusters and factors. The consolidation-type diagram of the analytical process is shown in Fig.  3 . This integrated framework is a key innovation, that employs the Modified Delphi Method to identify relevant factors, and applies the NSW technique to determine the network structure. The results are then integrated into the ANP model for final calculations and analysis. This integrated approach addresses the limitations of traditional ANP by providing a more robust and objective method for determining network relationships. It combines the strengths of expert knowledge (through the Delphi method), systematic relationship quantification (via NSW), and comprehensive decision analysis (through ANP), resulting in a more reliable and nuanced decision-making tool for token financing. First, the modified Delphi method was utilized to calculate the clusters and factors influencing startup companies engaging in token financing. Second, the network structure of the clusters and factors was determined on the basis of the NSW method put forward in this study. Finally, the weighted values of the network structure of NSW were integrated into the ANP model to calculate the weighted values for the various factors and various financing schemes of startup companies engaging in token financing. These weighted values were then sequenced to obtain the optimal scheme and key factors of token financing. Figure  4 presents the integrated framework for evaluating token financing options. This model incorporates five main clusters: Finance, Laws and Regulations, Risk, Investor, and Online Community, each containing several specific factors. The model also includes three token financing alternatives: ICO, IEO, and STO. This structure allows for a comprehensive evaluation of token financing alternatives, considering a wide range of relevant factors. By inclusion of diverse clusters including financial considerations, as well as legal, risk-related, investor-focused, and community aspects, the proposed framework allows startup companies to make well-informed decisions based on a thorough analysis of all relevant factors.

The integration processes

The research model

Step 1: Research the problem and confirm the decision factors

Past literature has pointed out that a research framework can be established only after experts reach a consensus on the factors [ 81 , 82 ]. Regarding the assessment of multiple principals, the number of selected experts should be between five and nine [ 76 ]. Therefore, this study included three scholars and four business starters, totaling seven experts. The goal of this study was to construct a consolidation-type pattern for the optimal scheme of token financing. Taking startup companies as examples, through a literature review and utilization of the Delphi method, a total of 17 factors, five clusters, and three token financing schemes were obtained, as shown in Fig.  4 . Relevant materials of each cluster and factors are shown as follows:

The definitions and illustrations of the clusters, factors, and token financing schemes in this study are as follows:

Finance: This includes issuance costs, platform fees, and transaction costs.

Issuance costs (C1) [ 83 , 84 ]: The costs of issuing tokens in different token financing schemes (for instance, Mint), which can vary.

Platform fees (C2) [ 83 ]: The costs for different token financing schemes to be launched on platforms (for instance, the costs for the schemes to be launched in Finance).

Transaction costs (C3) [ 83 ]: The transaction costs of different token financing schemes, which can vary (for instance, service charges).

Laws and regulations: This includes the place of issuance, government policy, token security regulations, and information disclosure transparency.

Place of issuance (C4): The laws, regulations, and rules of different countries and regions, as far as the issuance of tokens is concerned.

Policies (C5): The degree of support from government authorities on token financing.

Token security regulations (C6) [ 84 ]: The relevant policies on token security.

Information disclosure transparency (C7) [ 85 ]: Policies regarding the information disclosure of enterprises that issue tokens.

Risk: This includes financing schedules, token price fluctuations, reputation, shareholding proportion, and financing success rates.

Financing schedule (C8): The length of the financing scheme. For instance, Initial Coin Offerings (ICO) take a relatively long time, while Security Token Offerings (STO) take a relatively short time.

Token price fluctuations (C9) [ 83 ]: The price fluctuations of token transactions are obvious and influence relevant financing efficiency.

Reputation (C10) [ 86 ]: The degree of the token financing scheme’s requirements for the business reputation of the enterprises. For instance, ICO requires relatively less on the business reputation of the enterprises.

Shareholding proportion (C11): The proportion of shares corresponding to the tokens, which are held by the investors.

Financing success rates (C12) [ 87 ]: The success rates of different token financing schemes for enterprises.

The investor aspect: This includes the financing objects and financing thresholds.

Financing objects (C13): The investors being sought out by enterprises engaging in token financing. For instance, ICO and Initial Exchange Offerings (IEO) focus more on private investors, while STO focuses more on professional investors.

Financing thresholds (C14): The thresholds for enterprises to engage in token financing. For instance, the threshold of STO is relatively high.

The online community aspect: This includes the online sharing of voice, online public sentiment, and online trends.

Online sharing of voice (C15) [ 88 ]: The degree of influence of investors’ preferences of network volume in different financing platforms.

Online public sentiment (C16): The degree of influence of investor sentiment in the social network platforms of different financing platforms.

Online trends (C17): The degree of influence of the tendencies on the investors in the overall environment of token financing.

Token financing schemes: These include ICO, IEO, and STO.

ICO: The development, maintenance, and exchange for the purpose of financing, using blockchain technologies and virtual tokens.

IEO: The issuance and sales of tokens through the endorsement of exchanges. It also refers to the rules under which the exchanges are responsible for knowing your customer (KYC) compliance and anti-money laundering (AML).

STO: ICO is supervised by the government. It refers to the practice of linking the assets of enterprises to tokens through securitization, as well as the sales of such assets.

Step 2: Develop the network structure models through NSW

The results acquired in Step 1 were integrated into the NSW models suggested by this study, so as to determine the network structure. The relevant procedures are as follows:

Step 2.1: Design the questionnaire

In regards to the five clusters and 17 factors obtained by the experts in Step 1, a nine-point Likert scale was utilized to determine the strength of correlation between different factors. In the event of n factors, n ( n  − 1) comparisons of the scale were carried out. Because this study referred to seven experts for the development of the network structure model, the data involved were quite complicated. The NSW procedures were illustrated in accordance with the finance clusters, as well as the three factors of issuance costs, platform fees, and transaction costs. The questionnaire design for the finance clusters is shown in Table  2 , in which 0 indicates no influence was observed, while 9 indicates the influence was of the highest level. The strength of correlation among the three factors of finance obtained through the questionnaires of the seven experts is shown in Fig.  5 . Each expert’s assessment is represented in a separate diagram, allowing for a comparison of individual perspectives. The differences in experts’ opinions highlight the subjective nature of these assessments and underscore the importance of aggregating their opinions. The generally strong correlations between factors, particularly between issuance costs and platform fees, suggest that these financial aspects are closely interrelated in token financing decisions. This visualization is crucial for understanding the foundation of our network structure, as it forms the basis for our NSW calculations.

The strength of correlation among the three factors of finance obtained through the questionnaires of the seven experts

Step 2.2: Calculate the weight of the network structure

Each expert compared the factors and scored them in terms of strength. After that, the comparison scores provided by the experts were used in the construction of the matrices and weighted calculations. First, the correlation matrices of the finance clusters, M 1 to M 7 , were established on the basis of Eq. ( 1 ) and the scores of the strength given by the seven experts, as shown below. Second, correlation matrix M was transformed into probability matrices A 1 to A 7 through Eq. ( 2 ), as shown below, and the iteration method was used n times to obtain the characteristic values (eigenvalues) of each questionnaire and factor. Third, this study calculated the weighted values of the correlation among C 1 , C 2 , and C 3 , as well as R ( C i , C j ) 1 to R ( C i , C j ) 7 , through Eq. ( 3 ), as shown in Fig.  6 . This visualization is crucial for understanding how individual expert opinions contribute to the overall network structure. The variation in weights across experts highlights the subjective nature of these assessments and the necessity to aggregate multiple expert opinions. Notably, most experts consistently assign higher weights to the relationships between issuance costs ( C 1 ) and platform fees ( C 2 ), indicating a strong perceived connection between these two factors. In the end, the ultimate weighted values of the network structure (the scores of the correlation degree) were calculated using Eq. ( 4 ) and Eq. ( 5 ). The weighted values of the network structure of the various clusters and factors are shown in Fig.  7 . Figure  7 illustrates the final network structure weights for all five clusters and their respective factors, which is the foundation for our subsequent ANP analysis. These network structure weights provide a comprehensive understanding of the relative importance and interconnectedness of various factors in token financing decisions. They serve as a crucial input for our ANP model, ensuring that the final decision-making process accurately reflects the complex realities of token financing.

The network structure weights of finance cluster’s factors by 7 experts

The network structure weights of five cluster’s factors

Upon completing the calculations, the results of the weighted values for the network structure were integrated into the ANP models to establish the comparison matrices and calculate the eigenvectors.

Step 3: Perform pair-wise comparisons of the matrices and priority vectors

The eigenvectors of the clusters and factors were calculated through the AHP processes and pairwise comparison of features of matrices. The eigenvectors of the degree of correlation between different clusters and factors were calculated through NSW. The cases in this study involved five clusters (finance, laws and regulations, risk, investor, and online community), 17 factors (issuance costs, platform fees, transaction costs, place of issuance, government policy, token security regulations, information disclosure transparency, financing schedules, token price fluctuations, reputation, shareholding proportion, financing success rates, financing objects, financing thresholds, online share of voice, online public sentiment, and online trends), as well as three schemes.

The comparison matrices (in pairs) and the geometric method were utilized to calculate the eigenvectors, while the eigenvectors for the network structure of the correlation strength scores were obtained on the basis of NSW. The eigenvectors obtained for the various comparison matrices, as well as the eigenvectors related to the correlation strength of the factors, served as the values of the supermatrix, which was used to illustrate the correlation strength and the relative importance of different clusters. The clusters might confirm the eigenvectors of the network structure through NSW, and the scores of the relative importance were calculated using Eq. ( 6 ). The results of the eigenvectors for the network structure of the various factors are shown in Step 2.2, and the comparison matrices (in pairs) and the weighted values of the five clusters are shown in Table  3 . Table 4 contains the scores for the relative importance of the various factors against the alternative schemes. In this study, Super Decision V2.0 (software) was utilized for the subsequent assessment of the ANP models. The eigenvectors of the network structure obtained through the NSW were inputted into Super Decision V2.0 to integrate NSW and ANP and assess the optimal scheme and the key factors.

Step 4: Construct the supermatrix

The eigenvectors of the relationships among the factors, as well as the eigenvectors regarding the weights of the factors to the schemes, were determined according to the results of Step 3. In Step 4, a supermatrix is established on the basis of the eigenvectors obtained in Step 3, so that the optimal scheme for startup companies engaging in token financing could be measured. During the ANP process, the ultimate weighted values of the various factors and schemes were calculated through the unweighted supermatrix, the weighted supermatrix, and the limit supermatrix. First, the calculated eigenvectors of the NSW model for the factors and pair-wise comparison matrices were utilized to establish the unweighted supermatrix. Second, the unweighted supermatrix was multiplied by the reciprocals of the weighted values of the relevant clusters to generate the weighted supermatrix. Finally, the results of the weighted supermatrix were multiplied by themselves repeatedly until a stable probability distribution was realized. This probability distribution reflected the ultimate weighted values to be reached. The various supermatrices are shown in Tables 5 , 6 , and 7 .

Step 5: Evaluate the optimal alternative

Through the supermatrix mentioned in Step 4, as well as the operation of Super Decision, the ultimate weighted values of the various factors and schemes under the consolidated NSW network structure could be obtained, as shown in Table  8 .

This study suggested the establishment of a set of network assessment procedures integrating the new NSW technique with the ANP model, in order to analyze the optimal scheme for startup companies engaging in token financing. The findings indicated a number of results. The sequence of the weighted values for the five clusters was as follows: finance (0.307) > risk (0.294) > laws and regulations (0.211) > investors (0.106) > online community (0.082). In addition, the sequence of the weighted values for the factors was as follows: platform fees (0.083) > issuance costs (0.078) > financing success rate (0.053) > government policy (0.0049) = financing schedule (0.049) > transaction costs (0.044) > financing threshold (0.040) > information disclosure transparency (0.039) > token price fluctuations (0.032) = shareholding proportion (0.032) > financing object (0.031) > reputation (0.030) > place of issuance (0.027) > token security regulations (0.026) > online share of voice (0.022) > online public sentiment (0.019) > online trend (0.014). Finally, the sequence of the optimal scheme for startup companies engaging in token financing is as follows: ICO (0.057) > IEO (0.101) > STO (0.175). STO is the optimal scheme for startup companies to engage in token financing.

4 Conclusion and Future Work

4.1 conclusion.

The rapid development of FinTech has become one of the goals of inclusive financing. Fintech, which depends on information technology to find solutions in the financial field, is becoming the mainstream future trend in the financial industry, especially in the development of new business patterns. Startup companies might find it difficult to borrow money from traditional financial institutions due to their business operation features and financial structures. For this reason, alternative financing has gradually become an important channel for startup companies to acquire financing. Token financing is a relatively new business pattern in the field of alternative financing, and it can avoid the shortcomings and problems of crowdfunding.

However, the development history of token financing is diversified and complicated. Previous studies in this field focused more on the analysis of the values of virtual currencies. Generally speaking, when startup companies are faced with the option of token financing, which is a new business pattern, they have relatively little information available for business assessments and decision making. When startup companies assess the optimal scheme for token financing, they often use multi-principle decision-making models, which can solve the problems of filtration and selection in token financing. However, multi-principle decision-making models depend heavily on the presumption that the variables (or criteria) are independent from each other. Therefore, such models might not be suitable for the assessment of decision-making problems in the real world.

ANP can be used to solve the problem of independence assumption in traditional multi-principle decision-making models. Although ANP can overcome the problem of independence assumption, it is still unable to ascertain the strength of the dependence and relationships between variables before producing a network structure. In this study, a new model, NSW, was put forward. This new model could be used to calculate the correlation between variables and generate the network structure. In addition, NSW could be integrated into ANP to generate the network structure. In the end, the assessment of the optimal scheme for startup companies engaging in token financing served as the case study. The results of this study show that finance is the most critical cluster in the assessment aspect. In other words, when startup companies intend to engage in token financing, financial issue is the first aspect to be considered. Token financing is the most up-to-date financing method in the era of FinTech, and capital turnover and financial structure are key issues during the development of startup companies. The sequence of key factors are platform fees, issuance costs, and financing success rate. Moreover, this sequence suggests that when startup companies intend to engage in token financing, the key factors are the aspect of costs and the success rate of financing. Finally, the optimal scheme for startup companies engaging in token financing is STO. After considering financial issues, costs, and relevant risks, startup companies should, based on the cost assessment and the success rate of financing, adopt STO for token financing to promote the financial efficiency of such companies.

This study proposed the NSW technique as a novel tool for validating network structures in decision-making processes and integrated NSW into the ANP model to develop a comprehensive framework for evaluating optimal token financing strategies. The contributions of this study in token-based financing include both methodological advancement and practical application. In terms of methodology, this study integrated the NSW technique with the ANP to enhance the robustness of existing frameworks in capturing complex interrelationships within decision-making processes. This innovative approach addresses limitations in traditional methods by providing a more comprehensive quantification of the strength and directionality of relationships between decision factors. As for practical application, this study presents the first comprehensive evaluation of token financing options for startup companies utilizing this advanced decision-making approach. The integrated NSW-ANP framework can be applied to ICO, IEO, and STO, thus offering valuable options for cryptocurrency-based startup financing. This systematic evaluation considers the intricate interdependencies among various factors influencing the selection of optimal financing strategies. By bridging the gap between theoretical innovation and practical implementation, this study not only advances the field of multi-criteria decision-making but also provides startup entrepreneurs and investors with a sophisticated tool for token-based financing options. Academically, this study provided a new NSW technique, as well as the application procedures to integrate NSW into ANP. This study also presented a case study of the assessment of the optimal scheme for startup companies engaging in token financing. Practically, this new framework could provide entrepreneurs of startup companies with valuable measurement tools for promoting their company’s capital turnover rate through token financing under the rapid development of FinTech.

4.2 Limitation and Future Research

While acknowledging the substantial advantages offered by our integrated framework, it is imperative to recognize its inherent limitations. The following constraints warrant further investigation and potential mitigation in future research:

The potential complexity and mathematical technique of the proposed model, which might make it challenging to implement for organizations.

The static nature of the model, which may not fully capture the decision risks of uncertainty in the cryptocurrency and token financing landscape.

At the current stage of development, the model may not comprehensively capture the effects of factor weight variations on the rankings of alternatives.

After discussing these limitations, we will outline potential directions for future research. This section will propose several avenues for extending and refining our work:

Expanding the application of the NSW-ANP method to other areas of FinTech decision-making beyond token financing.

Integration of fuzzy set theory into the NSW-ANP model to address decision uncertainty risks.

A sensitivity analysis was conducted to ascertain the effects of factor weight variations on the rankings of alternatives.

Data Availability

Not applicable.

Ivanov, V., Knyazeva, A.: US securities-based crowdfunding under Title III of the JOBS Act. In: DERA White Paper (2017). Accessed 15 Jan 2022

Rossi, M.: The new ways to raise capital: an exploratory study of crowdfunding. Int. J. Financ. Res. 5 (2), 8–18 (2014)

Article   Google Scholar  

Walthoff-Borm, X., Schwienbacher, A., Vanacker, T.: Equity crowdfunding: first resort or last resort? J. Bus. Ventur. 33 (4), 513–533 (2018)

Massolution. (2015). 2015CF: The Crowdfunding Industry Report. Massolution. https://www.smv.gob.pe/Biblioteca/temp/catalogacion/C8789.pdf . Accessed 20 Jan 2022

Statista Inc.: Alternative Financing Report 2021 (2021). https://www.statista.com/study/47352/fintech-report-alternative-financing/ . Accessed 10 Jan 2022

Lu, Y., Chang, R., Lim, S.: Crowdfunding for solar photovoltaics development: a review and forecast. Renew. Sustain. Energy Rev. 93 , 439–450 (2018)

Bagheri, A., Chitsazan, H., Ebrahimi, A.: Crowdfunding motivations: a focus on donors’ perspectives. Technol. Forecast. Soc. Chang. 146 , 218–232 (2019)

Petruzzelli, A.M., Natalicchio, A., Panniello, U., Roma, P.: Understanding the crowdfunding phenomenon and its implications for sustainability. Technol. Forecast. Soc. Chang. 141 , 138–148 (2019)

Estrin, S., Gozman, D., Khavul, S.: Case study of the equity crowdfunding landscape in London: an entrepreneurial and regulatory perspective. FIRES Case Study 5 (2), 1–62 (2016)

Google Scholar  

Kuti, M., Madarász, G.: Crowdfunding. Public Financ. Q. 59 (3), 355 (2014)

Agrawal, A., Catalini, C., Goldfarb, A.: Some simple economics of crowdfunding. Innov. Policy Econ. 14 (1), 63–97 (2014)

Zhu, H., Zhou, Z.Z.: Analysis and outlook of applications of blockchain technology to equity crowdfunding in China. Financ. Innov. 2 (1), 29 (2016)

Baber, H.: Blockchain-based crowdfunding. In: Blockchain Technology for Industry 4.0, pp. 117–130. Springer, Singapore (2020)

Chapter   Google Scholar  

Chod, J., Trichakis, N., Yang, S.A.: Platform tokenization: financing, governance, and moral hazard. Manage. Sci. 68 (9), 6411–6433 (2021)

Howell, S.T., Niessner, M., Yermack, D.: Initial coin offerings: financing growth with cryptocurrency token sales. Rev. Financ. Stud. 33 (9), 3925–3974 (2020)

Amsden, R., Schweizer, D.: Are blockchain crowdsales the new ‘gold rush’? Success determinants of initial coin offerings (April 16, 2018) (2018)

Ante, L., Fiedler, I.: Cheap signals in security token offerings. Quant. Financ. Econ. 4 (4), 608–639 (2020)

Miglo, A.: Choice between IEO and ICO: speed vs. liquidity vs. risk. FinTech 1 (3), 276–293 (2022)

Kondova, G., Simonella, G.: Blockchain in startup financing: ICOs and STOs in Switzerland. J. Strateg. Innov. Sustain. 14 (6), 43–48 (2019)

Gryglewicz, S., Mayer, S., Morellec, E.: Optimal financing with tokens. J. Financ. Econ. 142 (3), 1038–1067 (2021)

Canh, N.P., Wongchoti, U., Thanh, S.D., Thong, N.T.: Systematic risk in cryptocurrency market: evidence from DCC-MGARCH model. Financ. Res. Lett. 29 , 90–100 (2019)

Borri, N.: Conditional tail-risk in cryptocurrency markets. J. Empir. Financ. 50 , 1–19 (2019)

Liu, Y., Tsyvinski, A.: Risks and returns of cryptocurrency. Rev. Financ. Stud. 34 (6), 2689–2727 (2021)

Zhang, W., Li, Y., Xiong, X., Wang, P.: Downside risk and the cross-section of cryptocurrency returns. J. Bank. Finance 133 , 106246 (2021)

Borri, N., Shakhnov, K.: Regulation spillovers across cryptocurrency markets. Financ. Res. Lett. 36 , 101333 (2020)

Feinstein, B.D., Werbach, K.: The impact of cryptocurrency regulation on trading markets. J. Financ. Regul. 7 (1), 48–99 (2021)

Chokor, A., Alfieri, E.: Long and short-term impacts of regulation in the cryptocurrency market. Q. Rev. Econ. Finance 81 , 157–173 (2021)

Beneki, C., Koulis, A., Kyriazis, N.A., Papadamou, S.: Investigating volatility transmission and hedging properties between Bitcoin and Ethereum. Res. Int. Bus. Financ. 48 , 219–227 (2019)

Okorie, D.I., Lin, B.: Crude oil price and cryptocurrencies: evidence of volatility connectedness and hedging strategy. Energy Econ. 87 , 104703 (2020)

Thampanya, N., Nasir, M.A., Huynh, T.L.D.: Asymmetric correlation and hedging effectiveness of gold & cryptocurrencies: from pre-industrial to the 4th industrial revolution✰. Technol. Forecast. Soc. Chang. 159 , 120195 (2020)

Sebastião, H., Godinho, P.: Bitcoin futures: an effective tool for hedging cryptocurrencies. Financ. Res. Lett. 33 , 101230 (2020)

Walther, T., Klein, T., Bouri, E.: Exogenous drivers of bitcoin and cryptocurrency volatility—a mixed data sampling approach to forecasting. J. Int. Finan. Markets. Inst. Money 63 , 101133 (2019)

Ma, F., Liang, C., Ma, Y., Wahab, M.I.M.: Cryptocurrency volatility forecasting: a Markov regime-switching MIDAS approach. J. Forecast. 39 (8), 1277–1290 (2020)

Article   MathSciNet   Google Scholar  

Köchling, G., Schmidtke, P., Posch, P.N.: Volatility forecasting accuracy for Bitcoin. Econ. Lett. 191 , 108836 (2020)

Yen, K.C., Cheng, H.P.: Economic policy uncertainty and cryptocurrency volatility. Financ. Res. Lett. 38 , 101428 (2021)

Hamdan, S., Cheaitou, A.: Supplier selection and order allocation with green criteria: an MCDM and multi-objective optimization approach. Comput. Oper. Res. 81 , 282–304 (2017)

Lin, S.W.: Identifying the critical success factors and an optimal solution for mobile technology adoption in travel agencies. Int. J. Tour. Res. 19 (2), 127–144 (2017)

Lin, C.Y.: Optimal core operation in supply chain finance ecosystem by integrating the fuzzy algorithm and hierarchical framework. Int. J. Comput. Intell. Syst. 13 (1), 259–274 (2020)

Kilic, B., Ucler, C.: Stress among ab-initio pilots: a model of contributing factors by AHP. J. Air Transp. Manag. 80 , 101706 (2019)

Achu, A.L., Thomas, J., Reghunath, R.: Multi-criteria decision analysis for delineation of groundwater potential zones in a tropical river basin using remote sensing, GIS and analytical hierarchy process (AHP). Groundw. Sustain. Dev. 10 , 100365 (2020)

Gündoğdu, F.K., Duleba, S., Moslem, S., Aydın, S.: Evaluating public transport service quality using picture fuzzy analytic hierarchy process and linear assignment model. Appl. Soft Comput. 100 , 106920 (2021)

Awad, J., Jung, C.: Extracting the planning elements for sustainable urban regeneration in Dubai with AHP (analytic hierarchy process). Sustain. Cities Soc. 76 , 103496 (2022)

Jorge-García, D., Estruch-Guitart, V.: Comparative analysis between AHP and ANP in prioritization of ecosystem services—a case study in a rice field area raised in the Guadalquivir marshes (Spain). Eco. Inform. 70 , 101739 (2022)

Tsai, W.H., Chou, W.C.: Selecting management systems for sustainable development in SMEs: a novel hybrid model based on DEMATEL, ANP, and ZOGP. Expert Syst. Appl. 36 (2), 1444–1458 (2009)

Saaty, T.L.: Decision Making with Dependence and Feedback: The Analytic Network Process, vol. 4922. RWS publications, Pittsburgh (1996)

Munim, Z.H., Duru, O., Ng, A.K.: Transhipment port’s competitiveness forecasting using analytic network process modelling. Transp. Policy 124 , 70–82 (2022)

Eshtiaghi, K., Aliyannezhadi, M., Najafian, A.: Identification and prioritization of factors affecting the adoption of electric vehicles using analytic network process. Int. J. Hum. Cap. Urban Manag 6 , 323–336 (2021)

Pang, N., Nan, M., Meng, Q., Zhao, S.: Selection of wind turbine based on fuzzy analytic network process: a case study in China. Sustainability 13 (4), 1792 (2021)

Asadi, A., Moghaddam Nia, A., Bakhtiari Enayat, B., Alilou, H., Ahmadisharaf, E., Kimutai Kanda, E., Chessum Kipkorir, E.: An integrated approach for prioritization of river water quality sampling points using modified Sanders, analytic network process, and hydrodynamic modeling. Environ. Monit. Assess. 193 (8), 1–15 (2021)

Toth, W., Vacik, H., Pülzl, H., Carlsen, H.: Deepening our understanding of which policy advice to expect from prioritizing SDG targets: introducing the analytic network process in a multi-method setting. Sustain. Sci. 17 (4), 1473–1488 (2022)

Fallahpour, A., Nayeri, S., Sheikhalishahi, M., Wong, K.Y., Tian, G., Fathollahi-Fard, A.M.: A hyper-hybrid fuzzy decision-making framework for the sustainable-resilient supplier selection problem: a case study of Malaysian palm oil industry. Environ. Sci. Pollut. Res. (2021). https://doi.org/10.1007/s11356-021-12491-y

Mabrouk, N.: Green supplier selection using fuzzy Delphi method for developing sustainable supply chain. Decis. Sci. Lett. 10 (1), 63–70 (2021)

Thanh, N.V., Lan, N.T.K.: A new hybrid triple bottom line metrics and fuzzy MCDM model: sustainable supplier selection in the food-processing industry. Axioms 11 (2), 57 (2022)

Erol, H., Dikmen, I., Atasoy, G., Birgonul, M.T.: An analytic network process model for risk quantification of mega construction projects. Expert Syst. Appl. 191 , 116215 (2022)

Tabatabaee, S., Mahdiyar, A., Mohandes, S.R., Ismail, S.: Towards the development of a comprehensive lifecycle risk assessment model for green roof implementation. Sustain. Cities Soc. 76 , 103404 (2022)

Rehman, O., Ali, Y., Sabir, M.: Risk assessment and mitigation for electric power sectors: a developing country’s perspective. Int. J. Crit. Infrastruct. Prot. 36 , 100507 (2022)

Moghaddasi, K., Rajabi, S., Soleimanian Gharehchopogh, F., Hosseinzadeh, M.: An energy-efficient data offloading strategy for 5G-enabled vehicular edge computing networks using double deep Q-network. Wireless Pers. Commun. 133 (3), 2019–2064 (2023)

Gharehchopogh, F.S., Abdollahzadeh, B., Barshandeh, S., Arasteh, B.: A multi-objective mutation-based dynamic Harris Hawks optimization for botnet detection in IoT. Internet of Things 24 , 100952 (2023)

Moghaddasi, K., Rajabi, S., Gharehchopogh, F.S.: Multi-objective secure task offloading strategy for blockchain-enabled IoV-MEC systems: a double deep Q-network approach. IEEE Access 12 , 3437–3463 (2024)

Moghaddasi, K., Rajabi, S., Gharehchopogh, F.S.: An enhanced asynchronous advantage actor-critic-based algorithm for performance optimization in mobile edge computing-enabled internet of vehicles networks. Peer-to-Peer Netw. Appl. 17 , 1169–1189 (2024)

Moghaddasi, K., Rajabi, S., Gharehchopogh, F.S., Ghaffari, A.: An advanced deep reinforcement learning algorithm for three-layer D2D-edge-cloud computing architecture for efficient task offloading in the internet of things. Sustain. Comput.: Inf. Syst. 43 , 100992 (2024)

Kashyap, A., Kumar, C., Shukla, O.J.: A DEMATEL model for identifying the impediments to the implementation of circularity in the aluminum industry. Decis. Anal. J. 5 , 100134 (2022)

Kamranfar, S., Azimi, Y., Gheibi, M., Fathollahi-Fard, A.M., Hajiaghaei-Keshteli, M.: Analyzing green construction development barriers by a hybrid decision-making method based on DEMATEL and the ANP. Buildings 12 (10), 1641 (2022)

Karasan, A., Ilbahar, E., Cebi, S., Kahraman, C.: Customer-oriented product design using an integrated neutrosophic AHP & DEMATEL & QFD methodology. Appl. Soft Comput. 118 , 108445 (2022)

Nezhad, M.Z., Nazarian-Jashnabadi, J., Rezazadeh, J., Mehraeen, M., Bagheri, R.: Assessing dimensions influencing IoT implementation readiness in industries: a fuzzy DEMATEL and fuzzy AHP analysis. J. Soft Comput. Decis. Anal. 1 (1), 102–123 (2023)

Du, Y.W., Shen, X.L.: Group hierarchical DEMATEL method for reaching consensus. Comput. Ind. Eng. 175 , 108842 (2023)

Si, S.L., You, X.Y., Liu, H.C., Zhang, P.: DEMATEL technique: a systematic review of the state-of-the-art literature on methodologies and applications. Math. Probl. Eng. 2018 , 1–33 (2018)

Sun, Y., Zhang, S., Huang, Z., Miao, B.: Probabilistic linguistic-based group DEMATEL method with both positive and negative influences. Complexity 2021 , 1–20 (2021)

Šmidovnik, T., Grošelj, P.: Solution for convergence problem in DEMATEL method: DEMATEL of finite sum of influences. Symmetry 15 (7), 1357 (2023)

Linstone, H.A., Turoff, M. (eds.): The Delphi method, pp. 3–12. Addison-Wesley, Reading, MA (1975)

Murry, J.W., Jr., Hammons, J.O.: Delphi: a versatile methodology for conducting qualitative research. Rev. High. Educ. 18 (4), 423–436 (1995)

Skulmoski, G.J., Hartman, F.T., Krahn, J.: The Delphi method for graduate research. J. Inf. Technol. Educ.: Res. 6 (1), 1–21 (2007)

Okoli, C., Pawlowski, S.D.: The Delphi method as a research tool: an example, design considerations and applications. Information and Management 42 (1), 15–29 (2004)

Wu, C.R., Lin, C.T., Chen, H.C.: Evaluating competitive advantage of the location for Taiwanese hospitals. J. Inf. Optim. Sci. 28 (5), 841–868 (2007)

Sung, W.C.: Application of Delphi method, a qualitative and quantitative analysis, to the healthcare management. J. Healthc. Manag. 2 (2), 11–19 (2001)

Hasson, F., Keeney, S.: Enhancing rigour in the Delphi technique research. Technol. Forecast. Soc. Chang. 78 (9), 1695–1704 (2011)

Atmaca, E., Basar, H.B.: Evaluation of power plants in Turkey using analytic network process (ANP). Energy 44 (1), 555–563 (2012)

Keramati, A., Salehi, M.: Website success comparison in the context of e-recruitment: an analytic network process (ANP) approach. Appl. Soft Comput. 13 (1), 173–180 (2013)

Meade, L., Sarkis, J.: Strategic analysis of logistics and supply chain management systems using the analytical network process. Transp. Res. E: Logist. Transp. Rev. 34 (3), 201–215 (1998)

Saaty, T.L., Vargas, L.G.: Decision making with the analytic network process, vol. 282. Springer, Berlin (2006)

Ali-Yrkko, J., Rouvinen, P., Seppala, T., Yla-Anttila, P.: Who captures value in global supply chains? Case Nokia N95 smartphone. J. Ind. Compet. Trade 11 (3), 63–278 (2011)

Linden, G., Kraemer, K.L., Dedrick, J.: Who captures value in a global innovation network? The case of Apple’s iPod. Commun. ACM 52 (3), 140–144 (2009)

Cong, L.W., Li, Y., Wang, N.: Tokenomics: dynamic adoption and valuation. Rev. Financ. Stud. 34 (3), 1105–1155 (2021)

Myalo, A.S.: Comparative analysis of ICO, DAOICO, IEO and STO. Case study. Financ. Theory Pract. 23 (6), 6–25 (2019)

Chod, J., Lyandres, E.: A theory of ICOS: diversification, agency, and information asymmetry. Manage. Sci. 67 (10), 5969–5989 (2021)

Momtaz, P.P.: Entrepreneurial finance and moral hazard: evidence from token offerings. J. Bus. Ventur. 36 (5), 106001 (2021)

Giudici, G., Adhami, S.: The impact of governance signals on ICO fundraising success. J. Ind. Bus. Econ. 46 (2), 283–312 (2019)

Fisch, C., Masiak, C., Vismara, S., Block, J.: Motives and profiles of ICO investors. J. Bus. Res. 125 , 564–576 (2021)

Download references

This research received no external funding.

Author information

Authors and affiliations.

Department of Public Finance and Tax Administration, National Taipei University of Business, 321, Sec. 1, Jinan Rd., Zhongzheng District, Taipei, 100, Taiwan

Chun-Yueh Lin

You can also search for this author in PubMed   Google Scholar

Contributions

Research design, literature review, data collection, data analysis, and manuscript writing are all conducted by Chun Yueh Lin.

Corresponding author

Correspondence to Chun-Yueh Lin .

Ethics declarations

Conflict of interest.

The authors declare no conflict of interest.

Additional information

Publisher's note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ .

Reprints and permissions

About this article

Lin, CY. Constructing a Novel Network Structure Weighting Technique into the ANP Decision Support System for Optimal Alternative Evaluation: A Case Study on Crowdfunding Tokenization for Startup Financing. Int J Comput Intell Syst 17 , 222 (2024). https://doi.org/10.1007/s44196-024-00643-0

Download citation

Received : 19 March 2024

Accepted : 19 August 2024

Published : 26 August 2024

DOI : https://doi.org/10.1007/s44196-024-00643-0

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Crowdfunding
• Token financing
• Network structure weighting (NSW)
• Analytic network process (ANP)
• Find a journal
• Publish with us
• Track your research

Select your cookie preferences

We use cookies and similar tools that are necessary to enable you to make purchases, to enhance your shopping experiences and to provide our services, as detailed in our Cookie notice . We also use these cookies to understand how customers use our services (for example, by measuring site visits) so we can make improvements.

If you agree, we'll also use cookies to complement your shopping experience across the Amazon stores as described in our Cookie notice . Your choice applies to using first-party and third-party advertising cookies on this service. Cookies store or access standard device information such as a unique identifier. The 96 third parties who use cookies on this service do so for their purposes of displaying and measuring personalized ads, generating audience insights, and developing and improving products. Click "Decline" to reject, or "Customise" to make more detailed advertising choices, or learn more. You can change your choices at any time by visiting Cookie preferences , as described in the Cookie notice. To learn more about how and for what purposes Amazon uses personal information (such as Amazon Store order history), please visit our Privacy notice .

• Politics, Philosophy & Social Sciences
• Government & Politics
• Civil Liberties & Political Activism

Sorry, there was a problem.

Download the free Kindle app and start reading Kindle books instantly on your smartphone, tablet or computer – no Kindle device required .

Read instantly on your browser with Kindle for Web.

Using your mobile phone camera - scan the code below and download the Kindle app.

Image Unavailable

• To view this video download Flash Player

Research Methods in Critical Security Studies: An Introduction Hardcover – 22 Oct. 2012

This new textbook surveys new and emergent methods for doing research in critical security studies, thereby filling a large gap in the literature of this emerging field.

New or critical security studies is growing as a field, but still lacks a clear methodology; the diverse range of the main foci of study (culture, practices, language, or bodies) means that there is little coherence or conversation between these four schools or approaches.

In this ground-breaking collection of fresh and emergent voices, new methods in critical security studies are explored from multiple perspectives, providing practical examples of successful research design and methodologies. Drawing upon their own experiences and projects, thirty-three authors address the following turns over the course of six comprehensive sections:

• Part I: Research Design
• Part II: The Ethnographic Turn
• Part III: The Practice Turn
• Part IV: The Discursive Turn
• Part V: The Corporeal Turn
• Part VI: The Material Turn

This book will be essential reading for upper-level students and researchers in the field of critical security studies, and of much interest to students of sociology, ethnography and IR.

• ISBN-10 0415535395
• ISBN-13 978-0415535397
• Edition 1st
• Publisher Routledge
• Publication date 22 Oct. 2012
• Language English
• Dimensions 17.78 x 1.91 x 24.77 cm
• Print length 256 pages
• See all details

Product description

"Finally, critical security studies has its own methodological handbook. It is not only extremely broad in scope, applying methods ranging from participant observation to interviews to discourse analysis, and discussing research design, ethnography, empiricism and writing. But it is also refreshingly reflexive in its approach. Its exploration of method is intimately bound to an advancement of theory, and a critical reflection on the role of the researcher in this sensitive – and often secretive –domain. It is indispensable reading for researchers and students alike, and promises to take this important field of research to a new level."– Marieke de Goede, University of Amsterdam, The Netherlands

'This textbook moves critical security studies forward in important ways by restoring "methodology" to the full sense of concept and rescuing it from the narrowness imposed by mainstream social science. ’ -- Roxanne Lynn Doty, Arizona State University, USA

'Wide-ranging and yet systematic, rigorous and yet pluralistic, this volume makes a crucial contribution toward developing innovative methodologies able to terms with the rapidly changing politics of contemporary security. Combining sophisticated conceptual overviews with illustrations of specific research designs in practice, it is a remarkably valuable resource for students and researchers, as well as an inspiring tour d’ horizon of cutting-edge research. ' -- Michael C Williams, University of Ottawa, Canada

About the Author

Mark B. Salter is Professor at the School of Political Studies, University of Ottawa, Canada. He is editor of Mapping Transatlantic Security Relations (Routledge 2010), and author of Rights of Passage: The Passport in International Relations (2003) and Barbarians and Civilization in International Relations (2002) .

Can E. Mutlu is a PhD candidate (ABD) at the School of Political Studies at the University of Ottawa, Canada. He is the Communications Director of the International Political Sociology Section of the International Studies Association (IPS-ISA).

Product details

• Publisher ‏ : ‎ Routledge; 1st edition (22 Oct. 2012)
• Language ‏ : ‎ English
• Hardcover ‏ : ‎ 256 pages
• ISBN-10 ‏ : ‎ 0415535395
• ISBN-13 ‏ : ‎ 978-0415535397
• Dimensions ‏ : ‎ 17.78 x 1.91 x 24.77 cm

Customer reviews

• 5 star 4 star 3 star 2 star 1 star 5 star 57% 43% 0% 0% 0% 57%
• 5 star 4 star 3 star 2 star 1 star 4 star 57% 43% 0% 0% 0% 43%
• 5 star 4 star 3 star 2 star 1 star 3 star 57% 43% 0% 0% 0% 0%
• 5 star 4 star 3 star 2 star 1 star 2 star 57% 43% 0% 0% 0% 0%
• 5 star 4 star 3 star 2 star 1 star 1 star 57% 43% 0% 0% 0% 0%

Customer Reviews, including Product Star Ratings, help customers to learn more about the product and decide whether it is the right product for them.

To calculate the overall star rating and percentage breakdown by star, we don’t use a simple average. Instead, our system considers things like how recent a review is and if the reviewer bought the item on Amazon. It also analyses reviews to verify trustworthiness.

• Sort reviews by Top reviews Most recent Top reviews

Top reviews from United Kingdom

There was a problem filtering reviews right now. please try again later..

Top reviews from other countries

• UK Modern Slavery Statement
• Amazon Science
• Sell on Amazon
• Sell on Amazon Business
• Sell on Amazon Handmade
• Associates Programme
• Fulfilment by Amazon
• Seller Fulfilled Prime
• Advertise Your Products
• Independently Publish with Us
• Host an Amazon Hub
• › See More Make Money with Us
• The Amazon Barclaycard
• Credit Card
• Amazon Money Store
• Amazon Currency Converter
• Payment Methods Help
• Shop with Points
• Top Up Your Account
• Top Up Your Account in Store
• COVID-19 and Amazon
• Track Packages or View Orders
• Delivery Rates & Policies
• Returns & Replacements
• Manage Your Content and Devices
• Amazon Mobile App
• Customer Service
• Accessibility

• Conditions of Use & Sale
• Privacy Notice
• Cookies Notice
• Interest-Based Ads Notice

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 21 August 2024

FOXO-regulated OSER1 reduces oxidative stress and extends lifespan in multiple species

• Jiangbo Song   ORCID: orcid.org/0000-0002-1349-632X 1   na1 ,
• Zhiquan Li   ORCID: orcid.org/0000-0003-3253-7606 2   na1 ,
• Lei Zhou   ORCID: orcid.org/0000-0002-6101-6669 1 ,
• Xin Chen   ORCID: orcid.org/0000-0001-8968-2711 1 ,
• Wei Qi Guinevere Sew 3 ,
• Héctor Herranz 3 ,
• Zilu Ye   ORCID: orcid.org/0000-0001-8829-6579 4 , 5 ,
• Jesper Velgaard Olsen   ORCID: orcid.org/0000-0002-4747-4938 4 ,
• Yuan Li   ORCID: orcid.org/0000-0001-8275-2916 2 ,
• Marianne Nygaard   ORCID: orcid.org/0000-0003-0703-2665 6 , 7 ,
• Kaare Christensen   ORCID: orcid.org/0000-0002-5429-5292 6 , 7 , 8 ,
• Xiaoling Tong   ORCID: orcid.org/0000-0002-2649-899X 1 ,
• Vilhelm A. Bohr   ORCID: orcid.org/0000-0003-4823-6429 2 , 9 ,
• Lene Juel Rasmussen   ORCID: orcid.org/0000-0001-6864-963X 2 &
• Fangyin Dai   ORCID: orcid.org/0000-0002-0215-2177 1  

Nature Communications volume  15 , Article number:  7144 ( 2024 ) Cite this article

1625 Accesses

14 Altmetric

Metrics details

• Genetics research
• Mechanisms of disease

FOXO transcription factors modulate aging-related pathways and influence longevity in multiple species, but the transcriptional targets that mediate these effects remain largely unknown. Here, we identify an evolutionarily conserved FOXO target gene, Oxidative stress-responsive serine-rich protein 1 ( OSER1 ), whose overexpression extends lifespan in silkworms, nematodes, and flies, while its depletion correspondingly shortens lifespan. In flies, overexpression of OSER1 increases resistance to oxidative stress, starvation, and heat shock, while OSER1-depleted flies are more vulnerable to these stressors. In silkworms, hydrogen peroxide both induces and is scavenged by OSER1 in vitro and in vivo. Knockdown of OSER1 in Caenorhabditis elegans leads to increased ROS production and shorter lifespan, mitochondrial fragmentation, decreased ATP production, and altered transcription of mitochondrial genes. Human proteomic analysis suggests that OSER1 plays roles in oxidative stress response, cellular senescence, and reproduction, which is consistent with the data and suggests that OSER1 could play a role in fertility in silkworms and nematodes. Human studies demonstrate that polymorphic variants in OSER1 are associated with human longevity. In summary, OSER1 is an evolutionarily conserved FOXO-regulated protein that improves resistance to oxidative stress, maintains mitochondrial functional integrity, and increases lifespan in multiple species. Additional studies will clarify the role of OSER1 as a critical effector of healthy aging.

Similar content being viewed by others

SUMO promotes longevity and maintains mitochondrial homeostasis during ageing in Caenorhabditis elegans

Ageing-associated changes in transcriptional elongation influence longevity

The autophagy receptor p62/SQST-1 promotes proteostasis and longevity in C. elegans by inducing autophagy

Introduction.

Recent investigations into the molecular basis of aging suggest that the FOXO transcription factor family is a critical regulator of genes that modulate aging, lifespan, and the response to oxidative stress 1 , 2 . FOXO proteins modulate signaling pathways downstream of insulin/insulin-like growth factor-1 (IGF-1) 3 , AMPK 4 , TOR 5 , 6 , JNK, and dietary restriction (DR) 7 , and integrate signals from these pathways 8 . FOXO-regulated genes and pathways are widely implicated in age-related diseases such as cancer 9 , Alzheimer’s disease 10 , 11 , and type 2 diabetes mellitus 12 .

Elevated oxidative stress is associated with cancer 13 , chronic kidney disease 14 , and neurodegeneration 15 . FOXO proteins regulate and are regulated by redox level and oxidative stress 16 . For example, FOXOs upregulate catalase (CAT) 17 , manganese superoxide dismutase (MnSOD) 18 , and sestrins 19 , 20 , 21 while suppressing the expression of mitochondrial genes 22 and regulating the oxidative DNA damage response 23 .

A multitude of FOXO target genes have been identified and characterized in Caenorhabditis elegans ( C. elegans ) 24 , Drosophila melanogaster ( D. melanogaster ) 25 , mouse 26 , and human cells 27 , or by meta-analysis and comparative multi-species “-omics” 28 . Nevertheless, many FOXO-regulated genes and their biological roles remain poorly characterized. Meanwhile, an emerging experimental system such as silkworm ( Bombyx mori, B. mori ) could potentially reveal the identity of conserved pro-longevity genes that cannot be readily identified in the more commonly used experimental model systems such as yeast, worms, flies, and mice. In this work, we identify and characterize a FOXO target, Oxidative stress-responsive serine-rich protein 1 (OSER1), whose up- or downregulation increases or decreases lifespan, respectively, in B. mori , C. elegans , and D. melanogaster . A potential influence of OSER1 on human lifespan is supported by human subject studies.

Transcriptomic analysis of FOXO targets

Comparative analysis of FOXOs between B. mori 29 and C. elegans , D. melanogaster , and Homo sapiens revealed highly conserved structural features like a forkhead DNA-binding domain (FHD), nuclear localization signals (NLS), and a transactivation domain (TAD) 16 (Fig.  1a ) as well as their tertiary structures (Supplementary Fig.  1a, b ), which confers the possibility of employing silkworm for FOXO targets screening. To screen FOXO transcriptional targets in silkworms, plasmids harboring wild-type BmFoxO or siRNA targeting BmFoxO were constructed (Supplementary Fig.  2 ) and transfected into silkworm BmN-SWU1 cells, in which BmFoxO was up- or down-regulated, respectively, as expected (Supplementary Fig.  3a–d ). However, wild-type BmFoxO was predominantly overexpressed in BmN-SWU1 cells as an inactive cytosolic phosphorylated form (Supplementary Fig.  3e ). To solve this problem, a constitutively active BmFoxO mutant (BmFoxO CA ) was transfected into and overexpressed in BmN-SWU1 cells. The BmFoxO CA mutant lacks AKT phosphorylation sites and is primarily expressed in the nucleus of BmN-SWU1 cells (Supplementary Fig.  3a, b, d, f ). Subsequently, we identified 3185 upregulated and 467 downregulated transcripts in cells overexpressing BmFoxO CA , while 3876 transcripts were upregulated and 271 were downregulated in BmFoxO-depleted cells (Fig.  1b and Supplementary Data  1 ) out of the whole transcriptome covering 14,623 genes 30 . We identified 69 genes that were upregulated by overexpression of BmFoxO and downregulated by knockdown of BmFoxO (i.e., positive targets), and 111 genes that were downregulated by overexpression of BmFoxO and upregulated in knockdown cells (i.e., negative targets) (Fig.  1c ; Supplementary Fig.  4 ). Gene Ontology (GO) analysis revealed that these genes are involved in diverse cellular functions (Supplementary Fig.  5a ), and KEGG analysis suggested roles in metabolism, genetic and environmental information processing, and multiple organismal and cellular processes (Fig.  1d ). Several genes were linked to human diseases (Fig.  1d ; Supplementary Data  2 ). Further investigation indicated roles in the TCA cycle, protein processing, and the PI3K-Akt signaling pathway (Supplementary Fig.  5b ). Bioinformatic analyses showed that 30.94% of human genes have orthologous genes in silkworms, while nematode and Drosophila genes have known human orthologs (accounts for 26.94% and 32.20% of human genome, respectively) (Fig.  1e ). Interestingly, of the 180 BmFoxO core targets identified in the transcriptomic analysis shown in Fig.  1c , 18 targets are conserved and 162 targets were either silkworm-specific or not yet reported in other species (Fig.  1f and Supplementary Data  3 ).

a The secondary structure of FoxO proteins among B. mori , C. elegans , D. melanogaster , and Homo sapiens . Members in the FoxO family share a highly conserved forkhead DNA-binding domain (FHD), one or two nuclear localization signal(s) (NLS), a nuclear export signal (NES), and a transactivation domain (TAD). The KIX-binding domain (KBD) is only found in human FOXOs in this comparison. The presence of the FHD, KBD, and TAD was predicted using the NCBI Conserved Domains Database (NCBI CDD). The identification of NLS and NES was carried out with NLStradamus and LocNES, respectively. C. elegans and D. melanogaster FoxO do not have a functional TAD. The positions of Akt phosphorylation sites indicated were analyzed using NetPhos 3.1 Server. b BmN-SWU1 cells were selected for stable overexpression of BmFoxO CA (constitutively active) or siRNA targeting BmFoxO . Total RNA was isolated for transcriptomic analysis, and differentially-expressed genes (DEGs) were identified. c Genes with up-regulated mRNA levels in BmFoxO overexpression and down-regulated in the BmFoxO knockdown conditions were denoted as “positive targets”; the opposite pattern was designated as “negative targets”. d KEGG analyses of DEGs showing the DEGs involvement in metabolism, genetic/environmental information processing, human diseases, organismal systems, and cellular processes. e Bioinformatic analysis of orthologous genes among silkworm, nematode, Drosophila , and humans. f Venn diagram of FOXO transcriptional direct targets in the indicated species. All data are available in Supplementary Data  1 , 2 , and 3.

OSER1 extends lifespan in multiple species

To confirm the direct regulation of putative gene targets by BmFoxO, a 3 kb genomic DNA sequence in the upstream promoter region of candidate BmFoxO target genes was searched for the consensus FOXO binding motif, revealing possible matches in 42 of 180 putative target genes (Supplementary Table  1 ). Twenty-four of these genes have orthologs in C. elegans that were not reported as daf-16 targets previously (Supplementary Data  4 ), and nine out of the other 18 genes have orthologs in humans and mice (Supplementary Table  2 ). To investigate the biological functions of these putative FOXO-regulated genes, single-gene siRNA knockdown was performed in C. elegans , and the lifespan of the knockdown and control strains was measured. The results show that knockdown of 7 genes shortened lifespan ( jmjd-5 , F02E9.5 , B0303.3 , eat-3 , aco-2 , fkh-10 , rabx-5 ) while knockdown of 3 genes ( znf-782 , pygl-1 , C24A1.3a ) increased lifespan (Fig.  2 ; Supplementary Fig.  6 ; Supplementary Data  4 ). Knockdown of F02E9.5 ( C. elegans ortholog of silkworm BGIBMGA002356 gene) had the largest impact of all putative FOXO targets, causing lifespan to decrease by 17% (Fig.  2b and Supplementary Data  4 ). Having identified the positively-regulated target of BmFoxO, BGIBMGA002356 and its C. elegans ortholog, as a candidate for a conserved longevity gene, we searched for orthologous genes in other species. Phylogenetic analysis revealed that this gene is conserved from C. elegans to humans with the homologous genes forming four clusters ( i.e ., nematode, insect, amphibian and fish, mammals), all of which share one protein domain of unknown function (DUF776) (Fig.  3a ). As the human homolog was previously identified as Oxidatively-induced Serine-rich Protein 1 (OSER1), we adopted the following naming convention: B. mori BGIBMGA002356, BmOSER1 ; C. elegans F02E9.5 , Ceoser1 ; D. melanogaster CG5056 , DmOser1 (Fig.  3a ).

a – g C. elegans knockdowns that show shortened lifespan ( jmjd-5 , n  = 50; F02E9.5 , n  = 76; B0303.3 , n  = 59; eat-3 , n  = 119; aco-2 , n  = 79; fkh-10 , n  = 97; rabx-5 , n  = 54). h – j C. elegans knockdowns that show extended lifespan ( znf-782 , n  = 81; pygl-1 , n  = 53; C24A1.3a , n  = 107). Controls for i , n  = 61, for j , n  = 107. The knockdowns of a-e were performed simultaneously to compare better the lifespan changes, so the same control ( n  = 120) was used and presented in the figure. Panels f – h also share the same controls ( n  = 130). Survival curves were plotted and analyzed by Log-rank (Mantel-Cox) test using GraphPad Prism 9. * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001. Lifespan experiments were performed at least three times independently with similar observations. Source data are provided as a Source Data file.

a Phylogeny of OSER1 in C. elegans , B. mori , D. melanogaster , Xenopus tropicalis , Danio rerio , Mus musculus , Macaca mulatta , and Homo sapiens . The phylogenetic tree was generated with MEGA version 11.0.11. The scale bar 0.1 represents the number of substitutions per site. The lifespan of nematodes with Ceoser1 knockdown ( b : siCtrl, n  = 109; siCeoser1, n  = 101) or overexpression ( c : oeCtrl, n  = 143; siCeoser1, n  = 101). The lifespan of B. mori with BmOSER1 knockout ( d : koCtrl ♀, n  = 36; koCtrl ♂, n  = 54; koBmOSER1 ♀, n  = 28) or overexpression ( e : Ctrl-OE ♀, n  = 94; Ctrl-OE ♂, n  = 97; BmOSER1-OE ♀, n  = 77; BmOSER1-OE ♂, n  = 70). The lifespan of heterozygotes control (yw/DmOser1) and DmOser1 mutant flies ( f : yw/DmOser1 ♀, n  = 96; yw/DmOser1mt ♂, n  = 100; DmOser1mt/DmOser1mt ♀, n  = 79; DmOser1mt/DmOser1mt ♂, n  = 100) or flies with mild (25 °C, g ) (Act-GAL4/+ ♀, n  = 95; Act-GAL4/+ ♂, n  = 98; Act>DmOser1 ♀, n  = 100; Act>DmOser1 ♂, n  = 99) or strong overexpression of DmOser1 (29 °C, h ) (Act-GAL4/+ ♀, n  = 100; Act-GAL4/+ ♂, n  = 100; Act>DmOser1 ♀, n  = 97; Act>DmOser1 ♂, n  = 99). Survival curves were plotted and analyzed by Log-rank (Mantel-Cox) test using GraphPad Prism 9. n.s not significant. ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001. Lifespan experiments were performed at least three times independently with similar observations. Source data are provided as a Source Data file.

Additional analyses showed that depletion or overexpression of Ceoser1 in wildtype N2 nematodes decreases or increases lifespan, respectively (Fig.  3b, c ). Similarly, knockout of BmOSER1 by CRISPR-Cas9 (Supplementary Fig.  7 ) shortened silkworm lifespan (Fig.  3d ), which is consistent with the effect of BmOSER1 knockdown (Supplementary Fig.  8a–d ), while overexpression of BmOSER1 increased lifespan (Fig.  3e and Supplementary Fig.  8e ). Moreover, under conditions that promote low or high expression of DmOser1 , the average lifespan of flies was correspondingly shorter or longer, respectively (Fig.  3f–h and Supplementary Fig.  9 ). We noted that overall lifespan is extended when DmOser1 is mildly overexpressed in females, but the maximal lifespan were extended in both male and female flies (Fig.  3g and Supplementary Fig.  9d–f ). In contrast, lifespan was extended in both sexes when DmOser1 was highly overexpressed (Fig.  3h and Supplementary Fig.  9g–i ). These data demonstrate that OSER1 is a conserved regulator of longevity in nematodes, silkworms, and flies.

OSER1 is a target of FOXO

As mentioned above, the sequence 3 kb upstream of the BmOSER1 start codon includes potential BmFoxO binding sites (Supplementary Table  1 ). To explore whether BmFoxO protein binds to these sequences in vitro, electrophoretic mobility shift assays (EMSA) were performed using purified BmFoxO protein and a biotin-labeled BmOSER1 DNA fragment in the absence or presence of excess unlabeled competitor DNA (Fig.  4a ). The results demonstrate specific binding of BmFoxO protein to the promoter of BmOSER1 . To evaluate the functional consequences of this interaction, the putative BmOSER1 promoter region (from 1 kb upstream of the BmOSER1 ORF to the start codon) was cloned into the upstream of the luciferase reporter gene in vector pGL3 and used in a dual-luciferase reporter assay (Supplementary Fig.  2 ) without or with co-expression of BmFoxO CA . The results show that expression of BmOSER1 is dramatically induced in B. mori cells that overexpress BmFoxO CA (Fig.  4b ). Furthermore, similar results were observed when the same experiment was performed in Spodoptera frugiperda Sf9 cells (Fig.  4b ). Confocal microscopy showed that BmOSER1 primarily localizes to the cell nucleus (Fig.  4c ), which was also observed in human U2OS cells (Fig.  4h ). Spatiotemporal studies showed that BmOSER1 is expressed at multiple developmental stages (peaked at L5D1 and wander stage) and tissues (highly expressed in reproductive organs like ovaries and testes) (Supplementary Fig.  10a, b ) and that its expression increases with age (Supplementary Fig.  10c ).

a EMSA using biotin-labeled BmOSER1 promoter fragment and BmFoxO protein, as described in supplementary methods. b Dual-luciferase reporter gene assay quantifies transactivation of the luciferase reporter gene with co-expression of BmFoxO CA , as described in methods. Data are presented as mean with SD and statistically tested by Two-way ANOVA. c Immunofluorescence analysis of BmOSER1 subcellular localization in BmN-SWU1 cells. Red, pIZ/V5-BmOSER1-DsRed2. Blue, DAPI for nucleus stain. Magenta, the merge of BmOSER1 and DAPI. Objective lens magnification, 100X. Scale bar, 10 μm. d – g The mRNA expression levels of Catalase ( CAT ), SOD2 , and OSER1 were detected by RT-qPCR with transfection of human FOXO1 ( d ), FOXO3 ( e ), FOXO4 ( f ), and FOXO6 ( g ) overexpression plasmids in U2OS cells. The relative mRNA expression was normalized to empty vector (pCS2) transfection control. Data are presented as mean with SD of three biological replicates and statistically tested by Two-way ANOVA. h Immunofluorescence staining showed that endogenous OSER1 localizes in the cell nucleus of U2OS cells. Red: OSER1, Blue, Hoechst. Scale bar, 10 μm. The images were taken by confocal laser scanning microscope LSM710 (Zesis) under 63X objective. i Relative SOD1 and OSER1 mRNA expression in U2OS cells treated with 0, 200, or 500 μM hydrogen peroxide (H 2 O 2 ) for 12 h, respectively. Data are presented as mean with SD and statistically tested by Two-way ANOVA. n.s., not significant; ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001. All data points derive from independent cell line lysates. Source data are provided as a Source Data file.

To understand FOXO regulation of OSER1 in greater detail, we overexpressed FOXO1, FOXO3, FOXO4, and FOXO6 in human U2OS cells following bioinformatic analysis, which predicted the presence of FOXO binding motifs in the promoter region of OSER1 (Supplementary Table  3 ). Under basal conditions, direct FOXO target genes, including Catalase ( CAT ) and Superoxide dismutase 2 ( SOD2 ), are expressed at a very low level (Fig.  4 d–g). However, exposure to hydrogen peroxide increased the expression of SOD1 and OSER1 mRNA (Fig.  4i ). These data demonstrate that OSER1 is a target of FOXO and is expressed in the nucleus and induced by oxidative stress.

OSER1 regulates oxidative stress response

To investigate whether OSER1 plays a role in response to oxidative stress, percent survival was measured in DmOser1 mutant, DmOser1 overexpressing (OE), and control flies exposed to 20 mM paraquat. The results show lower percent survival in the presence of oxidative stress in DmOser1 mutant and higher percent survival in DmOser1- OE flies than in control flies (Fig.  5a, b ). Consistent with this, dihydroethidium (DHE) staining to quantify ROS abundance in the imaginal wing discs suggested lower levels of ROS/oxidative stress in DmOser1- OE flies than in control flies (Fig.  5c ) mostly due to lower ROS abundance in hinge and wing pouch regions of the wing disc (Fig.  5d ). Furthermore, DmOser1- OE flies expressed lower mRNA levels of antioxidant gene Sod2 than control flies (Fig.  5e ). This aligns with the typical roles of SOD2 that its expression is triggered by oxidative stress 18 ; SOD2 overexpressors are more resistant, and its depletions are more vulnerable to oxidative stress in Drosophila 31 , 32 . Female DmOser1 -OE flies were more resistant to starvation, while male and female DmOser1 -OE flies were more resistant to heat shock than control flies (Fig.  5f, g ). BmOSER1 knockout silkworm larvae were more sensitive to heat shock than controls (Fig.  5h ), and adult male silkworms upregulated BmOSER1 and BmSOD1 after heat shock (Supplementary Fig.  8f ). When adult silkworms were dosed with 50 mM hydrogen peroxide by injection, transcription of BmCAT and BmOSER1 increased significantly (Fig.  5i ) and transcription of BmOSER1 , BmSOD1 , BmCAT , and glutathione peroxidase 1 ( BmGpx ) increased in BmE cell line treated with hydrogen peroxide (Fig.  5j ). When BmOSER1 was overexpressed in hydrogen peroxide-treated BmE cells, the abundance of ROS declined (Fig.  5k ) and expression of BmSOD1 and BmCAT decreased (Fig.  5l ). These results indicate that expression of OSER1 regulates and is regulated by environmental conditions including oxidative stress.

Survival of DmOser1 mutant ( a ) or overexpression (OE) flies ( b ) fed with the standard diet with 20 mM paraquat. The number of flies in the lifespan experiment: ( a ) yw/DmOser1 ♀ (heterozygous control), n  = 102; yw/DmOser1mt ♂, n  = 98; DmOser1mt/DmOser1mt ♀ (homozygous mutant), n  = 100; DmOser1mt/DmOser1mt ♂, n  = 99; ( b ) Act-GAL4/+ ♀ (heterozygous control), n  = 74; Act-GAL4/+ ♂, n  = 65; Act>DmOser1 ♀ (homozygous overexpression), n  = 74; Act>DmOser1 ♂, n  = 62. c Control and DmOser1 overexpression fly larvae imaginal wing discs were stained with DHE (ROS indicator). d Quantification of wing disc size and DHE fluorescence intensity in the whole and different regions of wing imaginal discs. Unpaired t test, Two-tailed. Whiskers: minima and maxima, Center: median, Bounds of box: 25% and 75% percentile. e Transcripts of DmOser , Sod2 , Cat , and GPx were quantified by qPCR in Drosophila larval whole wing discs. n  = 6 for each group. Unpaired t test, Two-tailed. f Survival of flies overexpressing DmOser1 under conditions of starvation. Act-GAL4/+ ♀, n  = 75; Act-GAL4/+ ♂, n  = 68; Act>DmOser1 ♀, n  = 75; Act>DmOser1 ♂, n  = 72. g Survival of DmOser1 overexpressing flies under heat shock (37 °C). Act-GAL4/+ ♀, n  = 75; Act-GAL4/+ ♂, n  = 35; Act>DmOser1 ♀, n  = 75; Act>DmOser1 ♂, n  = 62. h Survival of BmOSER1 knockout silkworm larvae under heat shock (37 °C) (mixed sexes). koCtrl, n  = 23; koBmOSER1, n  = 18. i Expression of BmOSER1 and BmCAT mRNA in adult silkworms treated with 50 mM H 2 O 2 . Unpaired t test, Two-tailed. j Expression of BmOSER1 , BmSOD1 , BmCAT , and BmGpx mRNA in BmE cells after treatment with 1 mM H 2 O 2 . n  = 6 for each group. Unpaired t test, Two-tailed. k Staining of silkworm embryonic BmE cells with DCF (dichlorofluorescein) with or without BmOSER1 overexpression. Unpaired t test, Two-tailed. l Expression of BmSOD1 , BmCAT , and BmGpx mRNA after 1 mM H 2 O 2 treatment in the presence or absence of BmOSER1 overexpression. Scatter plots are presented as mean with SD, and differences between control and treatment groups were analyzed by Unpaired t -test, Two-tailed. Survival curves were analyzed using Log-rank (Mantel-Cox) test. n.s., not significant; * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001. All data points derive from independent cell line lysates, Drosophila imaginal wing discs, Drosophila , and silkworms. Lifespan experiments were performed at least three times independently with similar observations. Source data are provided as a Source Data file.

The role of OSER1 in responding to and mitigating oxidative stress was also explored in Ceoser1 knockdown nematodes. Importantly, Ceoser1 knockdown nematodes had a shorter lifespan than control nematodes (Fig.  6a and Supplementary Fig.  8g ), but lifespan was specifically normalized (lengthened) by N-acetylcysteine (NAC), while control experiments showed that NAC does not extend the lifespan of control nematodes (Fig.  6a ). Interestingly, Ceoser1 overexpression increased the lifespan of nematodes with high levels of ROS after treatment with 4 mM TBHP (Fig.  6b ; Supplementary Fig.  2 ; Supplementary Fig.  8h ). Since the Ceoser1 regulation of lifespan is related to oxidative stress (Fig.  6a ), we detected the cellular ROS levels using DCF staining 33 following Ceoser1 knockdown and found that depletion of Ceoser1 increased DCF fluorescence (Fig.  6c ). This is also confirmed by Ceoser1 depletion in HyPer expression worms that showed elevated hydrogen peroxide levels (Fig.  6d ). Furthermore, when Ceoser1 was knocked down in nematodes carrying an integrated oxidative stress-inducible GFP tag, GFP fluorescence increased to a level that was further elevated by heat shock (Fig.  6e, f ). These results suggest that OSER1 improves resilience and resistance to oxidative stress, which could directly contribute to lifespan extension.

a Survival of control and Ceoser1 knockdown C. elegans in the presence or absence of 5 mM NAC (N-acetylcysteine). siCtrl, n  = 107; siCeoser1, n  = 113; siCtrl+NAC, n  = 106; siCeoser1+NAC, n  = 113. b Survival of control and Ceoser1 overexpressing C. elegans stimulated by 4 mM TBHP (tert-Butyl hydroperoxide). oeCtrl, n  = 74; oeCeoser1, n  = 77. c Quantification of 2’, 7’-dichlorofluorescin diacetate (DCF-DA) fluorescence intensity. Data are presented as mean with SD and analyzed by Unpaired t -test, Two-tailed. d Quantification of the HyPer fluorescence intensity. Data are presented as mean with SD and analyzed by Unpaired t -test, Two-tailed. Ceoser1 control or knockdown nematodes were imaged for ROS indicator gst-4p::gfp at 20 °C or 25 °C on day 1 or day 2 of adulthood ( e ) and quantified in ( f ). Data are presented as mean with SD and analyzed by two-way ANOVA. g Representative confocal image of mitochondrial morphology in control and Ceoser1 knockdown nematode strains with GFP and RFP tagged mitochondria. h Quantification of mitochondria length from ( g ). Data are presented as mean with SD and analyzed by Unpaired t -test, Two-tailed. i ATP levels in control and Ceoser1 knockdown nematodes. Data are presented as mean with SD and analyzed by Unpaired t -test, Two-tailed. Relative expression of Ceoser1 mRNA in various longevity mutant backgrounds, including daf-16 , daf-2 , eat-2 , pmk-1 , and an skn-1 gain‐of‐function mutant (#) cultured at 20 °C ( j ), and Ceoser1 mRNA expressions of glp-1 mutant cultured at 25 °C ( k ). Data in scatter plots are presented as mean with SD and statistically tested by One-way ANOVA ( j ) and Unpaired t -test, Two-tailed ( k ). l Survival of daf-2 mutants with normal or low levels of Ceoser1 expression. All survival curves were plotted and analyzed by Log-rank (Mantel-Cox) test using GraphPad Prism 9. daf-2 siCtrl, n  = 315; daf-2 siCeoser1, n  = 350. n.s., not significant; * p  < 0.05, ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001. All data points derive from independent nematodes. Lifespan experiments were performed at least three times independently with similar observations. Source data are provided as a Source Data file.

Transcriptomic analyses in BmOSER1 -overexpressing silkworms and Ceoser1 -depleted nematodes revealed the potential involvement of OSER1 in mitochondrial biogenesis and other mitochondrial functions (Supplementary Data  5 ). This was explored by knocking down Ceoser1 in [ myo-3p::TOM20::RFP(mit) ] and SJ4103 [ myo-3::GFP(mit) ] nematodes with fluorescently labeled mitochondria. Interestingly, fragmented mitochondria were more abundant in Ceoser1 knockdown nematodes than in control nematodes, and the concentration of intracellular ATP decreased significantly (Fig.  6g–i ).

Previous studies showed that in the DAF-2/insulin signaling pathway, daf-2 / InR (insulin receptor) mutants are more than 100% longer-lived 34 ; age-1/PI3K mutants are 40% longer-lived 35 ; and daf-16/FOXO mutants are 12% shorter-lived 36 in C. elegans . Here, we show that the expression of Ceoser1 is lower in daf-16/FOXO mutants and higher in daf-2/InR mutants (Fig.  6j ). In the dietary restriction mimicking mutant eat-2 37 , Ceoser1 expression increases (Fig.  6j ). Furthermore, in skn-1/NRF2 gain-of-function nematodes with enhanced detoxification, mitochondrial biogenesis, and longer lifespan 38 , Ceoser1 expression is upregulated (Fig.  6j ). However, the expression of Ceoser1 was not changed in control and pmk-1/p38 and glp-1/NOTCH mutant animals (Fig.  6j, k ). In contrast to other known major DAF-16 targets, Ceoser1 knockdown in daf-2 mutant nematodes only shortened lifespan by approximately 10% (Fig.  6l ). Thus, we conclude that Ceoser1 is upregulated in pro-longevity daf-2 and eat-2 mutants and skn-1 gain-of-function mutant and downregulated in anti-longevity daf-16 mutants. This indicates that OSER1 acts downstream of the convergent FOXO node.

OSER1 promotes reproduction

The biological roles of human OSER1 were examined in U2OS cells overexpressing or depleted for OSER1 using a lentiviral expression system. GO analysis of DEGs in OSER1-depleted cells revealed enrichment for genes involved in cellular senescence, cell cycle, senescence-associated secretory phenotype (SASP), oxidative stress-induced senescence, and reproduction and substantial enrichment for genes involved in cell cycle-related pathways in OSER1-OE cells (Fig.  7a–d ; Supplementary Data  6 ; Supplementary Data  7 ), which is consistent with the nuclear localization of OSER1 in silkworms (Fig.  4c ) and U2OS cells (Fig.  4h ).

a, b KEGG enrichment in U2OS cells with OSER1 knockdown (kd) ( a ) or overexpression (oe) ( b ). The cnetplot of OSER1 knockdown ( c ) and overexpression cells ( d ), based on data summarized in ( a ) and ( b ), respectively. kdCtrl, the lentivirus control U2OS cell line. kdOSER1, the lentivirus-targeted knockdown of OSER1 U2OS cell line. oeCtrl, the U2OS cells transiently transfected with the control plasmid. oeOSER1, the U2OS cells transiently overexpressed with OSER1. All data are available in Supplementary Data  6 and 7 .

A possible role of OSER1 in reproduction was confirmed by showing that BmOSER1 overexpression increased the fertility of eggs and enhanced male silkworm mating activity, leading to more successful mating events (Fig.  8a–c ). Interestingly, depletion of BmOSER1 decreased egg laying (Fig.  8d ). In C. elegans , the number of progeny was not affected by Ceoser1 knockdown, but it increased in Ceoser1 -OE nematodes (Fig.  8e, f ).

a Representative images of eggs laid by wildtype (oeCtrl) and BmOSER1 overexpression (oeBmOSER1) silkworms. As shown, one male silkworm mates with 6 females in the wildtype Dazao and produces hatchable eggs, while in BmOSER1 overexpression silkworms, one male mates with 23 females that laid hatchable eggs. b Duration of the entire fertile mating period between male and female adult silkworms. c Number of successful matings per male adult silkworms. d The number of eggs laid per female adult silkworms in wildtype and BmOSER1 knockout. e – f  The total brood sizes of Ceoser1 RNAi ( Ceoser1 RNAi) ( e ) and overexpression (oeCeoser) ( f ) throughout life. Data in all scatter plots are presented as mean with SD and analyzed by Unpaired t -test, Two-tailed. n.s., not significant; ** p  < 0.01, *** p  < 0.001, **** p  < 0.0001. All data points derive from independent silkworms or nematodes. Source data are provided as a Source Data file.

OSER1 is associated with longevity in humans

Finally, we investigated OSER1 genetic variants in humans. A total of 49 common single nucleotide polymorphisms (SNPs) in OSER1 , representing seven independent ( r 2  < 0.1) LD-based groups of SNPs, fulfilled the applied quality criteria and were detected in 90 + -year-old individuals as well as younger controls (Supplementary Table  4 and 5 ). The minor allele dosage for two of 49 SNPs from one independent LD-based group was significantly ( P  < 0.0024) associated with longevity. In addition, the minor allele dosage of five SNPs from another LD-based group was associated with longevity at a nominal significance level ( P  < 0.05) (Supplementary Table  6 ). No SNPs showed a significant association with age at menopause in long-lived or younger women. However, several nominally significant associations were found (Supplementary Table  7 ) between higher minor allele dosage and higher age at menopause, especially in long-lived women. In summary, human subject studies support the idea that OSER1 also influences lifespan in humans.

Oxidative stress is widely considered a primary driver of aging and age-related diseases 39 . In response to oxidative stress, FOXO proteins upregulate anti-oxidative genes and suppress mitochondrial genes 22 , 40 . In this study, emerging and well-established animal models of aging were used to identify and characterize the longevity protein OSER1, a highly conserved FOXO-regulated determinant of longevity in silkworms, flies, and nematodes (Fig.  9 ). The results presented here show that up- or downregulation of OSER1 affects lifespan somewhat more significantly in silkworms and flies than in nematodes. While the molecular basis of this observation is not yet known, we postulate that the physiological function(s) of OSER1 homologs could differ in these diverse species. Transcriptomic data followed by in vivo studies demonstrate that OSER1 provides strong protection against oxidative stress at the organismal and cellular levels, suggesting that it might mitigate oxidative stress downstream of FOXO, similar to SOD2, CAT, and GPX1 in multiple species. In unstressed silkworm and human cells, OSER1 localized to the nucleus, but the network of cytosolic and nuclear proteins that mediates the biological effects of OSER1 remains poorly understood.

Transcriptomic analyses of FOXO transcriptional targets in silkworms identified an evolutionarily conserved FOXO direct target OSER1. Mechanistically, FOXO protein binds to the promoter of OSER1 to activate its transcription. OSER1 expression is essential for normal mitochondrial morphology and functions, as well as low levels of reactive oxygen species (ROS), possibly via interacting with other as yet unidentified proteins. The in vivo studies showed the critical roles of OSER1 in longevity and stress resistance (especially oxidative stress response) in silkworms, flies, and nematodes. Interestingly, human subject studies support the idea that OSER1 also influences the human lifespan. This figure was created by BioMedVisual.com.

Heat stress can increase the lifespan of yeast 41 and C. elegans 42 , which could potentially reflect and be dependent on increased expression of heat shock factor (HSF-1) 43 . HSF1 and FOXO cooperate to activate the expression of small heat-shock proteins to extend lifespan 44 . Downregulation of DAF-16/FOXO and HSF-1 activity has been linked to a shorter lifespan in C. elegans with glucose supplementation 45 . Heat stress stimulates the production of ROS and the expression of both MnSOD 46 and BmOSER1 (Supplementary Fig.  8f ). A recent study in zebrafish also demonstrated a correlation of heat shock with increased expression of OSER1 and ROMO1 47 . Nevertheless, the mechanism by which OSER1 confers thermotolerance remains unclear.

In summary, the current study presents strong evidence that OSER1 increases resilience and resistance to oxidative stress, heat shock, and other types of cellular stress, while at the same time causing an apparent increase in lifespan in multiple species. Finally, we emphasize the need for further research on the biological roles and mechanism of action of the longevity protein OSER1.

Cell culture and transcriptomics

Cell culture.

Bombyx mori BmN-SWU1 cells were cultured in TC-100 insect medium 48 , and BmE cells were cultured in Grace insect medium 49 . Spodoptera frugiperda Sf9 cells were cultured in Sf-900 TM III SFM insect medium (Thermo Fisher Scientific, USA) 50 . These cell cultures were grown at 27 °C under sterile conditions, as previously reported 48 , 50 .

Transcriptomics

Cell lines pIZ/V5-FoxO CA and pIZ/V5-siFoxO were grown under selection by zeocin for 48 h post-transfection. For each sample, cells from three independent cell culture flasks were pooled. Total RNA was extracted using Trizol Reagent (Invitrogen, USA) according to the manufacturer’s instructions. A differential gene expression library was constructed and analyzed by GENE DENOVO Corporation (Guangzhou, China). Bombyx mori BmFoxO Raw sequence reads data can be found at PRJNA943622 [ https://www.ncbi.nlm.nih.gov/bioproject/PRJNA943622/# . RNA-Seq data for Control (Accession No. SRX20534814 ), BmFoxO-RNAi (Accession No. SRX20534813 ), and FoxO CA -OE (Accession No. SRX20534812 ).

Analysis of differentially-expressed genes (DEGs)

The Omicshare website tool ( https://www.omicshare.com/ ) was used to identify differentially-expressed genes across samples or groups. We identified genes with a fold-change ≥2 and a false discovery rate (FDR) threshold ≤0.05. DEGs were then analyzed for enrichment of GO functions and KEGG pathways using the Omicshare website tool.

Phylogenetic analysis

Each OSER1 protein sequence was downloaded on the NCBI website, and the Muscle software was employed to perform multiple sequence alignment 51 . Bayesian inferences were performed using MrBayes v3.2.7 to calculate the tree statistics, including the mean and variance of split or clade frequencies and branch rates 52 , 53 . The Java-jarprotest-3.3.jar was employed to detect the most suitable model to reconstruct the phylogenetic tree 54 . The “MCMC” command was used for running the process 54 , 55 .

Dual-luciferase reporter assay

The upstream regulatory regions (−1500 bp to 0 bp) of the oser1 gene from the silkworm wildtype Dazao strain were amplified using primers PGL-F and PGL-R. The PCR products were cloned into the PGL3-Basic vector (Promega). The PGL3- oser1 -luciferase recombinant plasmid and the FoxO CA overexpression plasmid were cotransferred with the reference plasmid (containing Renilla luciferase gene driven by Ie1 promoter) separately into silkworm BmN-SWU1 cells and Spodoptera frugiperda Sf9 cells using X-tremeGENE HP DNA Transfection Reagent (Roche, USA) was used according to the manufacturer’s instructions. At 72 h post-transfection, cells were collected for luciferase assays.

Quantitative real-time PCR (RT-qPCR)

Silkworm tissues were collected to quantify BmOSER1 mRNA at different developmental stages, and total RNA was isolated using TRIzol (Invitrogen) according to the manufacturer’s instructions. Reverse transcription and quantitative real-time PCR (RT-qPCR) were performed as previously described 56 . The primers used for RT-qPCR are listed in Supplementary Table  8 . sw22934 / BmeIF4A was used as a previously-validated RT-qPCR reference gene for the silkworm 57 . As shown in the Supplementary data, our RT-qPCR experiments follow almost all the criteria of the MIQE guidelines 58 .

Vector construction

For gene overexpression, selected promoter, ORF, and poly-A regions were inserted into recipient vectors using standard cloning procedures (Supplementary Fig.  2 ). For knockdown constructs, the interference fragment for selected target genes was inserted into commercially-available vectors according to standard procedures. To express the product with the interference effect (Supplementary Fig.  2 ).

Prokaryotic protein expression

The pET-32a (+) vector with His tag was used to express the BmFoxO protein in E. coli , and a histidine affinity column (HisTrap HP) was used for protein purification. Recombinant protein-containing culture supernatant was filtered using a 0.22 μm membrane prior to affinity chromatography. Bound protein was washed using 10, 20, and 30 column volumes of binding buffer and was eluted with 20 column volumes of elution buffer containing 80 mM Imidazole. Eluted protein was collected in separate tubes, and the column was washed with 5 column volumes of Elution Buffer (100 mM Imidazole) to elute residual protein. Finally, the column was washed with 10 column volumes of Binding Buffer (40 mM Imidazole), 5 column volumes of 20% ethanol, and the column was stored at 4 °C. In general, a flow rate of 1 mL/min was maintained.

Electrophoretic mobility shift assay (EMSA)

The 3000 bp genomic sequence upstream of the BmOSER1 start codon was input to JASPAR for the prediction of putative transcription factor (TF) binding sites. For the EMSA assay, the labeled probe and competitor probe were designed according to the binding motif with or without biotin, respectively. Probes were diluted to 10 µM and annealed to respective complementary probes. 0.5×TBE was used as the electrophoresis and transfer buffer, and protein transfer was performed for 8 min. Then, the 254 nm UV light was used for cross-linking for 15 min. Biotin-labeled probes were detected by chemiluminescence.

Confocal microscopy

BmN-SWU1 cells were plated in a 24-well plate and transfected with 500 ng of the expression vector (pIZ/V5-His, pIZ/V5-His-BmFoxO, pIZ/V5-His-BmFoxO CA ). At 96 h post-transfection, cells were fixed with 4% paraformaldehyde in PBS for 15 min at room temperature. Fixed cells were permeabilized in 0.1% Triton X-100 in PBS for 10 min. Cells were blocked with 10% normal goat serum in PBS for 1 h at 37 °C, followed by incubation with anti-His antibody (1:200, Beyotime) for 1 h at 37 °C. Cells were stained with anti-mouse Alexa 555 (1:500) antibody for 1 h and mounted with 0.1 g/ml DAPI (Sigma, USA) for observation under a confocal microscope (Olympus, Japan). BmN-SWU1 cells were plated in a 24-well plate and transfected with 500 ng of the expression vector (pIZ/V5-His-BmOSER1-DsRed2). At 48 h post-transfection, cells were fixed with 4% paraformaldehyde in PBS for 10 min at room temperature. Fixed cells were permeabilized in 0.1% Triton X-100 in PBS for 10 min and stained with DAPI (Beyotime) for imaging under a confocal microscope (Olympus, Japan).

Lifespan assay

Silkworm wildtype Dazao strain was reared under standard conditions (25 °C, in approximately 75% relative humidity with a 12 L/12D photoperiod) with clean and fresh mulberry leaves during the whole larval stage in the Silkworm Gene Bank at Southwest University. Survival was monitored every 3 h. Silkworms that did not move when gently prodded were marked as dead and recorded. Maximum lifespan refers to the upper 10% of the distribution of lifespan.

Transgene overexpression and gene interference in individual BmOSER1 silkworms

BmOSER1 was overexpressed or subject to RNAi knockdown in wild-type silkworms (Dazao strain). The BmOSER1 overexpression plasmid was constructed using a 3p3×piggyBac-EGFP basic vector by inserting the IE2- oser1 (ORF)-SV40 expression box. The oser1 overexpression plasmid was mixed with the A3 helper vector and injected into silkworm eggs within 2 h after laying. EGFP fluorescence in the silkworm compound eye was used as a marker for transgene-expressing individual worms. Samples were collected separately for RT-qPCR analysis at larval and adult stages.

The BmOSER1 -dsRNA fragment was obtained by PCR amplification from the full-length cDNA with a pair of primers containing the T7 RNA polymerase binding site (14). The BmOSER1 -dsRNA (ds BmOSER1 ) and the red fluorescent protein gene dsRNA (dsRed) were synthesized using the T7 RiboMAX Express RNAi kit (Promega, USA) according to manufacturer’s instructions. The dsRNA (ds BmOSER1 and dsRed, 120 μg) was injected into hemolymph through the penultimate spiracle on the 10th day after pupation. Samples were collected for RT-qPCR analysis 12 and 24 h after molting. Lifespan analysis was conducted as described.

Stress assay

Cell and individual silkworm exposure to oxidative stress.

Cells were distributed in a 12-well cell culture plate and grown to approximately 80% maximum cell density, prior to transfection with vector control, pIZ/V5, or pIZ/V5- BmOSER1 overexpression plasmid. After 48 h, cells were washed gently with PBS, and exposed to fresh culture medium containing 600 μl 1 mM H 2 O 2 (Sigma, USA) per well; after 1 h, cells were collected, treated with lysis solution and then stored at −20 °C. For the treatment of individual silkworms, injection devices, syringes, and capillaries were prepared. Each worm was injected with 10 μl H 2 O 2 by the needle through the intersegment membrane on day 1 after the moth eclosion in male and female individuals. The control group was injected with the same volume of 1 × PBS solution. These materials were used for checking the ROS content by the Reactive Oxygen Species (ROS) Assay Kit (Comin, China) or for detection of gene expression levels by the microRNA extraction kit (Omega, USA).

ROS detection

DCFH-DA was diluted into serum-free medium at a ratio of 1:1000 and added to cells. Cells were incubated for 20 min at 27 °C and, and then washed 3x with serum-free medium. Cells were imaged and fluorescence was detected on a microplate reader using 488 nm excitation and 525 nm emission.

Thermotolerance

Developmentally consistent female and male individuals were selected on day 1 after eclosion and incubated at 37 °C or 25 °C. Survival was measured as described.

Drosophila melanogaster techniques

The DmOser1 mutant stock was obtained from the Bloomington Drosophila Stock Center (#21830), and the overexpression stock was from FlyORF (#F002710). The yw and w1118 were reared as previously described 59 . For lifespan assay, newly emerged female and male flies were separated, and healthy flies with similar body sizes were selected. For each group, 4 vials, and 25 flies/vial were used at the start of survival measurement. Food was changed every three days, and the number of dead flies was recorded. Control and DmOser1 mutant flies were incubated at 25 °C with controlled humidity. For mild DmOser1 overexpression, flies were also incubated at 25 °C, while flies with higher DmOser1 overexpression and control flies were incubated at 29 °C.

For oxidative stress assay, flies were collected in vials containing standard food containing 20 mM paraquat. For each group, 4 vials (25 flies/vial) were incubated at 25 °C with controlled humidity. Survival was monitored every three hours. For the starvation assay, experimental conditions were the same, except standard food was replaced with 1% agar and survival was monitored every hour. For the thermotolerance assay, flies were placed in vials containing standard food and incubated at 25 °C or 29 °C.

RNA extraction and RT-qPCR for adult Drosophila were performed as previously described 59 . For larval Drosophila : wing discs were dissected from 25 to 30 L3 wandering Drosophila larvae, and total RNA was extracted according to the Trizol protocol (Life Technologies, #15596026). DNase treatment was performed on RNA samples with RQ1 DNase I (Promega, #M6101), and cDNA was prepared using the SuperScript III Reverse Transcriptase kit (Life Technologies, #18080-044). RT-qPCR was run on the QuantStudio 6 Flex Real-Time PCR machine (Applied Biosystems) using the 5x HOT FIREpol EvaGreen qPCR Mix Plus (Solis BiodDyne, #08-24-00001). DNA sequences of all primers are shown in Supplementary Table  8 .

Dihydroethidium (DHE) staining

L3 wandering larvae were dissected in Schneider’s Drosophila media (Gibco, #11720). Samples were then stained with DHE (Invitrogen, #D11347) diluted in Schneider’s media for 5 to 7 min at room temperature under conditions of darkness. All subsequent steps were performed at room temperature in the dark. DHE-containing medium was removed, discs were washed 3 times for 5 min each with Schneider’s media and then incubated in 7% formaldehyde / PBS for 5 min. Samples were washed once for 5 min with PBS and mounted for imaging in 90% glycerol / PBS /0.05% N-propyl gallate. Stained samples were imaged with a Leica SP8 confocal laser-scanning microscope.

Caenorhabditis elegans techniques

C. elegans strains and maintenance.

C. elegans were cultured at 20 °C on a standard nematode growth medium (NGM) seeded with E. coli OP50- 60 unless stated otherwise. The wild-type N2 Bristol, CL2166[ gst-4p::gfp ], KU25[ pmk-1(km25 )], SPC167[ skn-1(lax120) ], DR1572[ daf-2(e1368) ], CB4037[ glp-1(e2141) ], DA465[ eat-2(ad465 )], CF1038[ daf-16(mu86) ] strains were provided by the Caenorhabditis Genome Center. The muscle mitochondria-targeted fluorescent protein strain [ myo-3p::TOM20::RFP(mit) ] was provided by Dr. Shanshan Pang and SJ4103 [ myo-3::GFP(mit) ] was provided by Dr. Shiqing Cai. FYD19[ eft-3p::F02E9.5a ] transgenic worms were constructed by cloning the promoter of eft-3 and full-length F02E9.5a ( Ceoser1 ) genomic DNA into pPD95_79 plasmid followed by standard microinjection.

RNA interference treatment

For the RNAi experiment, HT115 bacteria containing specific dsRNA-expression plasmids (Ahringer library) 61 were cultured overnight at 37 °C in LB containing 100 µg/mL carbenicillin and seeded on NGM plates containing 5 mM IPTG. RNAi Plasmid L4440 (2790 bp, Addgene, #1654) was used as a control in the RNAi experiment. RNAi was induced at 25 °C for 24 h after seeding. Then, synchronized L1 worms were added to RNAi plates to knock down the indicated genes. Primer sequences are listed in Supplementary Table  8 .

Vector construction for Ceoser1 overexpression

The pPD95_79 Plasmid was a gift from Andrew Fire (Addgene, #1496) and was employed as the basic recipient vector. The eft-3 promoter (600 bp) was inserted from Hind III to Xba I, and F02E9.5a genomic DNA (1147 bp) was inserted from Xba I to Sma I. The pPD95_79 basic plasmid was set as control.

RT-qPCR was performed as previously described 62 . Briefly, day1 worms were collected, washed in M9 buffer, and then homogenized in TRIzol reagent (Life Technologies). RNA was extracted according to the manufacturer’s protocol. DNA contamination was digested with DNase I (Thermo Fisher Scientific), and RNA was subsequently reverse-transcribed to cDNA using the RevertAid First Strand cDNA Synthesis Kit (Thermo Fisher Scientific). Quantitative PCR was performed using SYBR Green (Bio-Rad). The expression of snb-1 was used to normalize samples. Primer sequences are listed in Supplementary Table  8 .

Fluorescence microscopy

Microscopic imaging was performed as previously described 33 . Briefly, unless otherwise indicated, day 1 adult worms were collected, washed in M9 buffer, and then paralyzed with 1 mM levamisole. Fluorescence microscopic images were taken after mounting on 2% agarose pad slides.

Lifespan assays were performed as previously described 63 with modifications. Briefly, synchronized L1 worms were added to NGM plates seeded with different E . coli strains. Worms were transferred every day during the reproductive period. Worms that died of vulva burst, bagging, or crawling off the plates were censored. For NAC treatment, concentrated NAC (Sigma) was added to NGM plates at a final concentration of 5 mM NAC.

ROS measurement

ROS measurement for hydrogen peroxide assays was performed as previously described 64 . Briefly, the day 2 adult HyPer expression worms jrIs1[Prpl-17::HyPer] were mounted on slides, and excited with GFP and CFP channels to measure the oxidized and reduced HyPer, respectively. Fluorescence microscopic images were then taken, and fluorescent density was measured using ImageJ software. The ratio of oxidized to reduced HyPer intensity was calculated as the hydrogen peroxide levels. ROS measurement assays using DCF-DA (Sigma) were performed as described previously 33 . Briefly, approximately 1000 adult worms (day 2) were collected, washed three times in M9 buffer, then homogenized in PBST (PBS, 0.1% Tween 20), and subsequently centrifuged twice at 13,000 rpm at 4 °C for 10 min. The supernatant was used for protein concentration determination (bicinchoninic acid (BCA) assay) and ROS levels measurement. The supernatant containing 20 µg of protein was pre-incubated with 200 µM of DCF-DA in 100 µl of PBS at 37 °C for 1 h. Fluorescent intensity was then measured at the excitation wavelength of 485 nm and the emission wavelength of 530 nm every 10 min for 1 h. Fluorescent intensity was normalized, and background fluorescence was subtracted.

ATP measurement

ATP levels were measured as previously described 64 . In brief, about 300 adult worms (day 2) were collected and washed three times in M9 buffer. The worms were then homogenized by boiling for 20 min and centrifuged at 13,000 rpm at 4 °C. The supernatant was used for ATP concentration determination according to the manufacturer’s protocol (ENLITEN ATP Assay; Promega). The protein content in the supernatant was determined by employing a BCA assay to normalize the ATP levels.

Silkworm and C. elegans OSER1 transcriptomics

Eggs of wild-type Dazao silkworm were injected into a mixture of piggyBac-BmOSER1 overexpression plasmid and pHA3PIG A3-helper plasmid. After culture and screening, the BmOSER1 overexpression strain was successfully constructed. The male adults on the first day of the adult stage ( n  = 3) from wild-type Dazao and BmOSER1 overexpression strains were sampled. Synchronized L1 worms were added to Ceoser1 RNAi and control plates separately and cultured to day 1 of adulthood at 20 °C. Total RNA was extracted using the TRIzol reagent kit (Invitrogen, Carlsbad, CA, USA), and the extraction procedure was performed according to the manufacturer’s protocol. RNA quality was assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA, USA) and checked using RNase-free agarose gel electrophoresis. The sequencing Libraries were thereafter constructed and sequenced using an Illumina Novaseq 6000 150PE by Gene Denovo Biotechnology Co., Ltd. (Guangzhou, China). The input data for differential gene expression analysis were the read counts obtained from gene expression level analysis. The analysis was performed using the DESeq2 software, which was divided into three main parts, firstly, normalization of read counts. then calculation of p-values based on the model; Finally, multiple hypothesis testing correction is performed using Benjamini-Hochberg method to obtain the FDR values (False Discovery Rate). Based on the results of the differential analysis, genes between the control group and the treated group with the parameter of the false discovery rate (FDR) < 0.05 and absolute fold change (FC) ≥ 2 were considered differentially expressed genes (DEGs). Significance test of enrichment analysis was performed by calculating the corresponding q-value which is obtained by multiple testing of the p-value. It was considered that the gene had been significantly enriched when the q- value was <0.05. The raw sequencing data generated from this study were deposited in NCBI under the BioProject ID PRJNA978397 and PRJNA978499. The identified DEGs were subsequently subjected to enrichment analysis of GO and KEGG with the OmicShare website tool. The various significantly enriched GO terms and KEGG pathways were defined with a hypergeometric test (corrected p -value < 0.05).

Human cells

Human U2OS osteosarcoma cells were cultured in DMEM (Gibco, #31966021) supplemented with 10% FBS and 1% Penicillin-Streptomycin (Gibco, #15140122). Cells were cultured in a 37 °C incubator with 5% CO 2 .

5x HOT FIREPol® EvaGreen® qPCR Mix Plus (ROX) (Solis BioDyne, #08-36-00020) was used for determining gene expression on the StepOne™ Real-Time PCR System. Experiments were performed following the manufacturer’s instructions. RT-qPCR primers were included in Supplementary Table  8 .

Transient transfection of human pCS2, FOXO1, FOXO3, FOXO4, and FOXO6

U2OS cells were seeded to 6-well plates the day before transfection. The next day, pCS2, FOXO1, FOXO3, FOXO4, and FOXO6 were transfected into U2OS cells at 80% confluency. These plasmids are generously provided by Stefan Koch 65 . 1 ug plasmid was transfected to each well of 6-well plates by using Polyjet transfection reagent (SignaGen, #SL100688). RNA samples were collected after 72 h to measure gene expression.

Immunofluorescence and microscopy

U2OS cells were seeded on the coverslips of the 24-well plate. The next day, cells were washed with PBS once and then fixed by adding 4% paraformaldehyde (Sant Cruz, #30525-89-4) for 15 min at room temperature. Then, cells were rinsed three times in PBS for 5 min each. Afterward, cells were blocked by blocking&permeabilization buffer (5% goat serum +0.2% Triton-100 in PBS) for 60 min at room temperature first and then incubated with blocking&permeabilization buffer diluted OSER1 antibody (Merck, # HPA045125) overnight at 4 °C. The next day, coverslips were rinsed with PBS three times for 5 min each, and the coverslips were incubated with PBS-diluted Goat-anti-Rabbit secondary antibody (ThermoFisher, #A-11011) for 2 h at room temperature. Then coverslips were rinsed with PBS three times for 5 min each after secondary antibody incubation and then stained with PBS diluted Hoechst (ThermoFisher, #62249) (ThermoFisher, #P36965) for 5 min and washed with PBS twice for 5 min each. Afterward, the coverslips were mounted with a Prolong-gold mounting medium for imaging by confocal laser scanning microscope LSM710 (Zeiss).

Lentivirus packaging and infection

The lentiviruses were packaged in HEK293T cells together with packaging vectors (plasmid (9 μg), pMDLg/pRRE (4.5 μg), pRSV-Rev (4.5 μg), and pMD2.G (4.5 μg) using 45 μg 1 ug/uL PEI (polyethylenimine, transfection reagent). Lentivirus-containing media were collected 48 h later and filtered through a 0.45 μm filter to remove the cell debris and used to infect control and AD fibroblasts. After 8 h infection for lentivirus, the medium was changed to the normal culture medium.

Mass spectrometry proteomics analysis

Cells were washed twice with cold phosphate-buffered saline (1 × PBS) and then rapidly lysed, and cysteines were reduced and alkylated in a single step. Briefly, boiling 6 M guanidine hydrochloride (Gnd-HCl) containing 10 mM chloroacetamide and 5 mM tris(2-carboxyethyl)phosphine was added directly to the cells. Lysis buffer containing cells was boiled for an additional 10 min at 99 °C, followed by sonication for 2 min. Samples are diluted using 50 mM ammonium bicarbonate to approximately 0.3 M GdnCl final concentration and an aliquot containing 12 µg protein was digested with Trypsin (Sigma-Aldrich) (0.5 μg/μl) for 18 h at pH 8.5, 37 °C.

After spinning down for 5 min at 17,000 × g, 500 ng peptides were immobilized and purified on EvoTips. All samples were analyzed on an Orbitrap Exploris 480 mass spectrometer coupled with an Evosep One system using an in-house packed 15 cm, 150 μm i.d. capillary column with 1.9 μm Reprosil-Pur C18 beads (Dr. Maisch, Ammerbuch, Germany) using the pre-programmed gradients (30 samples per day) in data-independent acquisition (DIA) mode as described before 66 . Briefly, full MS resolutions were set to 120,000 at m/z 200 and full MS AGC target was 300% with an IT of 45 ms. Mass range was set to 350–1400. AGC target value for fragment spectra was set at 1000%. 49 windows of 13.7 Da were used with an overlap of 1 Da. Resolution was set to 15,000 and ion injection time to 22 ms. Normalized collision energy was set at 27%. All data were acquired in profile mode using positive polarity and peptide match was set to off, and isotope exclusion was on. All samples were acquired in four replicates ( n  = 4) and a total number of 24 samples were analyzed including four control samples.

All DIA raw files were processed with Spectronaut (v15, Biognosys, Zurich, Switzerland) in DirectDIA mode using human Uniprot Reference Proteome without isoforms (21,074 entries). We used standard settings in Spectronaut for peptide identification, protein quantification, and determination of significantly regulated proteins. Briefly, cysteine carbamylation was set as a fixed modification, whereas methionine oxidation and protein N-terminal acetylation were set as variable modifications. Precursor filtering was set as Q value, and cross-run normalization was checked. The results of the differential abundance testing were exported from Candidates table node. The table is filtered by a q -value (Benjamini-Hochberg multiple testing corrected p -value) of 0.05 and an absolute log2 ratio of 0.58. Functional enrichments were conducted with in-house written R-scripts using clusterProfiler. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD036579.

Human studies

The study is registered in SDU’s internal list (notification no. 11.648) and complies with the rules in the General Data Protection Regulation. Written informed consent was obtained from all participants. The collection and use of biological material and survey and registry information were approved by the Regional Committees on Health Research Ethics for Southern Denmark.

Long-lived individuals (LLIs), i.e ., individuals >90 years old, were drawn from six nationwide surveys collected at the University of Southern Denmark; the Study of Danish Old Sibs (DOS), the 1905 birth cohort study, the 1910 birth cohort study, the 1911-12 birth cohort study, the 1915 birth cohort study, and the Longitudinal Study of Ageing Danish Twins (LSADT). The younger controls were drawn from the Middle-Aged Danish Twins (MADT) study. Briefly, DOS was initiated in 2004 and included families in which at least two siblings were ≥90 years of age at intake. The 1905, 1910, and 1915 birth cohort studies are prospective follow-up studies initiated in 1998, 2010, and 2010, when participants were 92-93, 100, and 95 years of age, respectively 67 . The 1911-1912 birth cohort study consists of individuals reaching the age of 100 years in the period from May 2011 to July 2012 68 , and LSADT was initiated in 1995 and included Danish twins ≥70 years of age 69 . The younger controls were obtained from MADT that was initiated in 1998 and included twins randomly chosen from the birth years 1931–1952 69 . From DOS and LSADT, one individual from each sib-ship or twin pair was randomly selected among participants that had reached an age of at least 91 years for DOS and 90 years for LSADT. From the 1905 and 1915 birth cohort studies, participants were selected among individuals reaching an age of at least 96 years. From MADT, one individual from each monozygotic twin pair was randomly selected. From dizygotic twin pairs, both twins were included.

DNA was extracted from whole blood using standard methods 70 , or from filter cards using the Extract-N-Amp Blood PCR Kit (Sigma Aldrich, St. Louis, MO, USA) followed by amplification using the GenomePlex Complete Whole Genome Amplification (WGA) Kit (Sigma Aldrich, St. Louis, MO, USA).

Genotype data

The included individuals were genotyped as part of one of two data sets. In data set 1, individuals from DOS, 1905, 1910, 1911-12, and 1915 birth cohort studies, and LSADT were genotyped using the Illumina Human OmniExpress Array (Illumina San Diego, CA, USA). Pre-imputation quality control included filtering SNPs on genotype call rate <95%, HWE P  < 10 −4 , and MAF = 1%, and individuals on sample call rate <95%, relatedness, and gender mismatch. Pre-phasing and imputation to the 1000 Genomes phase I v.3 reference panel were performed using IMPUTE2 version 2.3.2 71 . In data set 2, individuals from LSADT and MADT were genotyped using the Illumina Infinium PsychArray (Illumina San Diego, CA, USA). Pre-imputation quality control included filtering SNPs on genotype call rate <98%, HWE P  < 10 −6 , and MAF = 0, and individuals on sample call rate <99%, relatedness, and gender mismatch. Pre-phasing and imputation to the 1000 Genomes phase 3 reference panel were performed using IMPUTE2 version 2.3.2 71 . After imputation, genotype probabilities were converted to hard-called genotypes in Plink 72 using a cut-off of 90%, meaning that only genotypes with a probability of more than 90% were called. Variants with no genotype probabilities above 90% were set to missing.

Data on single nucleotide polymorphisms (SNPs) in the OSER1 gene, including 5000 bp upstream and 1000 bp downstream to cover regulatory regions (total region; chr20:42,823,581-42,844,546), was extracted from the hard-called genotype files. Only bi-allelic SNPs with MAF > 0.01, HWE P  > 0.01, and INFO > 0.80 were included. Independent LD-based groups of SNPs were identified using European populations linkage disequilibrium patterns from 1000 G (genome build GRCh37) and the NIH National Cancer Institute SNP clip Tool ( https://ldlink.nci.nih.gov/?tab=snpclip ).

Phenotype data

Descriptive information on cases and controls can be found in Supplementary Table  4 .

Information on individual survival status up to May 1st, 2022, was retrieved from the Danish Central Population Register 73 .

Self-reported age at menopause was collected as part of comprehensive interviews in the 1905, 1910, and 1915 birth cohort studies and MADT, focusing on health and lifestyle issues and assessments of cognitive and physical abilities. Among the younger controls, age at menopause was available for 529 women (mean = 49.5 years, sd = 5.1 years, and range = 35–65 years), and among the long-lived individuals for 282 women (mean = 49.4 year, sd = 4.6 years, and range = 35–60 years). Women with self-reported age at menopause <35 years ( N  = 15) or >65 years ( N  = 1, 71 years) were excluded. In general, applying these exclusion criteria resulted in an attenuation of effect sizes, primarily for positively associated SNPs.

The association between longevity and SNP minor allele dosage (coded as 0, 1, or 2 depending on the number of minor alleles) was analyzed in STATA v. 17.0 using a logistic regression model including sex as a covariate and with twin pair number as a random effect to account for the relatedness between twins from the same twin pair. OSER1 SNPs associated ( P  < 0.05) with longevity. The analysis was performed using a logistic regression model comparing long-lived cases to younger controls and including sex as a covariate. P -values are uncorrected.

The association between age at menopause and SNP minor allele dosage (coded as 0, 1, or 2 depending on the number of minor alleles) was analyzed in STATA v. 17.0 using a linear regression model with twin pair number as a random effect to account for the relatedness between twins from the same twin pair in the analysis of the younger controls. OSER1 SNPs associated ( P  < 0.05) with age at menopause. The analysis was performed in long-lived and younger women using a linear regression model. P -values are uncorrected.

A Bonferroni-corrected significance level of P  < 0.0024 (corresponding to correction for 21 tests; seven independent LD-based groups of SNPs, two phenotypes, one study cohort for longevity, and two study cohorts for age at menopause) was applied. However, given the a priori hypothesis of an association between OSER1 and longevity and fertility/age at menopause, uncorrected P -values are reported.

Statistics & reproducibility

Scatter plots were analyzed using Prism 9 and presented as mean ± SD. Survival curves were generated and analyzed by Log-rank (Mantel-Cox) test using Prism 9. Significant differences between treatments/groups were analyzed using a Student’s t- test, one-way ANOVA, or two-way ANOVA: * p  < 0.05; ** p  < 0.01; *** p  < 0.001; **** p  < 0.0001. No data were excluded from the analyses. The experiments were randomized. The investigators were not blinded to allocation during experiments and outcome assessment except for the immunofluorescence analysis.

Reporting summary

Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

Data availability

The numerical data and uncropped representative images used for quantifications are available as Source Data file accompanying this paper. Other data reported in the current study can be found in the Supplementary Information. All cellular and animal data are available from the corresponding author upon reasonable request. The transcriptomics raw sequence reads data for BmFoxO knockdown and overexpression can be found at PRJNA943622 . RNA-Seq data for Control (Accession No. SRX20534814 ), BmFoxO-RNAi (Accession No. SRX20534813 ), and FoxOCA-OE (Accession No. SRX20534812 ). The transcriptomics raw data for BmOSER1 overexpression silkworms (NCBI BioProject Accession #: PRJNA978397 ) and Ceoser1 knockdown C. elegans (NCBI BioProject Accession #: PRJNA978499 ) are available at NCBI. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD036579 . According to the Danish and EU legislations, transfer and sharing of individual-level data require prior approval from the Research & Innovation Organization at the University of Southern Denmark or the Danish Data Protection Agency and require that data sharing requests be dealt with on a case-by-case basis. For this reason, the raw data cannot be deposited in a public database. However, we welcome any inquiries regarding collaboration and individual requests for data sharing. Inquiries and requests can be sent to Marianne Nygaard ([email protected]). Further information about data access can be found here: https://www.sdu.dk/en/om-sdu/institutter-centre/ist_sundhedstjenesteforsk/centre/dtr/researcher .  Source data are provided with this paper.

van der Horst, A. & Burgering, B. M. Stressing the role of FoxO proteins in lifespan and disease. Nat. Rev. Mol. Cell Biol. 8 , 440–450 (2007).

Article   PubMed   Google Scholar  

Martins, R., Lithgow, G. J. & Link, W. Long live FOXO: unraveling the role of FOXO proteins in aging and longevity. Aging Cell 15 , 196–207 (2016).

Article   CAS   PubMed   Google Scholar  

Kenyon, C. J. The genetics of ageing. Nature 464 , 504–512 (2010).

Article   ADS   CAS   PubMed   Google Scholar  

Greer, E. L. et al. The energy sensor AMP-activated protein kinase directly regulates the mammalian FOXO3 transcription factor. J. Biol. Chem. 282 , 30107–30119 (2007).

Hay, N. Interplay between FOXO, TOR, and Akt. Bba-Mol. Cell Res 1813 , 1965–1970 (2011).

CAS   Google Scholar  

Robida-Stubbs, S. et al. TOR Signaling and Rapamycin Influence Longevity by Regulating SKN-1/Nrf and DAF-16/FoxO. Cell Metab. 15 , 713–724 (2012).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Jiang, Y., Yan, F., Feng, Z., Lazarovici, P. & Zheng, W. Signaling Network of Forkhead Family of Transcription Factors (FOXO) in Dietary Restriction. Cells-Basel 9 . https://doi.org/10.3390/cells9010100 (2019)

Mukhopadhyay, A., Oh, S. W. & Tissenbaum, H. A. Worming pathways to and from DAF-16/FOXO. Exp. Gerontol. 41 , 928–934 (2006).

Dansen, T. B. & Burgering, B. M. T. Unravelling the tumor-suppressive functions of FOXO proteins. Trends Cell Biol. 18 , 421–429 (2008).

Manolopoulos, K. N., Klotz, L. O., Korsten, P., Bornstein, S. R. & Barthel, A. Linking Alzheimer’s disease to insulin resistance: the FoxO response to oxidative stress. Mol. Psychiatr. 15 , 1046–1052 (2010).

Article   CAS   Google Scholar  

Hwang, I. et al. FOXO protects against age-progressive axonal degeneration. Aging Cell 17 . https://doi.org/10.1111/acel.12701 (2018)

Du, S. & Zheng, H. Role of FoxO transcription factors in aging and age-related metabolic and neurodegenerative diseases. Cell Biosci. 11 , 188 (2021).

Sosa, V. et al. Oxidative stress and cancer: an overview. Ageing Res Rev. 12 , 376–390 (2013).

Daenen, K. et al. Oxidative stress in chronic kidney disease. Pediatr. Nephrol. 34 , 975–991 (2019).

Niedzielska, E. et al. Oxidative Stress in Neurodegenerative Diseases. Mol. Neurobiol. 53 , 4094–4125 (2016).

Klotz, L. O. et al. Redox regulation of FoxO transcription factors. Redox Biol. 6 , 51–72 (2015).

Nemoto, S. & Finkel, T. Redox regulation of forkhead proteins through a p66shc-dependent signaling pathway. Science 295 , 2450–2452 (2002).

Kops, G. J. et al. Forkhead transcription factor FOXO3a protects quiescent cells from oxidative stress. Nature 419 , 316–321 (2002).

Chen, C. C. et al. FoxOs inhibit mTORC1 and activate Akt by inducing the expression of Sestrin3 and Rictor. Dev. Cell 18 , 592–604 (2010).

Lee, J. H. et al. Sestrin as a Feedback Inhibitor of TOR That Prevents Age-Related Pathologies. Science 327 , 1223–1228 (2010).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Hagenbuchner, J. et al. FOXO3-induced reactive oxygen species are regulated by BCL2L11 (Bim) and SESN3. J. Cell Sci. 125 , 1191–1203 (2012).

Ferber, E. C. et al. FOXO3a regulates reactive oxygen metabolism by inhibiting mitochondrial gene expression. Cell Death Differ. 19 , 968–979 (2012).

Li, Z., Anugula, S. & Rasmussen, L. J. in Aging 275-295 (Elsevier, 2023).

Lee, S. S., Kennedy, S., Tolonen, A. C. & Ruvkun, G. DAF-16 target genes that control C. elegans life-span and metabolism. Science 300 , 644–647 (2003).

Alic, N. et al. Genome-wide dFOXO targets and topology of the transcriptomic response to stress and insulin signalling. Mol. Syst. Biol. 7 , 502 (2011).

Audesse, A. J. et al. FOXO3 directly regulates an autophagy network to functionally regulate proteostasis in adult neural stem cells. Plos Genet. 15 , e1008097 (2019).

Eijkelenboom, A. et al. Genome‐wide analysis of FOXO3 mediated transcription regulation through RNA polymerase II profiling. Mol. Syst. Biol. 9 , 638 (2013).

Article   PubMed   PubMed Central   Google Scholar  

Webb, A. E., Kundaje, A. & Brunet, A. Characterization of the direct targets of FOXO transcription factors throughout evolution. Aging Cell 15 , 673–685 (2016).

Song, J. et al. Genome-wide identification and characterization of Fox genes in the silkworm, Bombyx mori. Funct. Integr. Genomics 15 , 511–522 (2015).

Li, Y. et al. Transcriptome analysis of the silkworm (Bombyx mori) by high-throughput RNA sequencing. Plos One 7 , e43713 (2012).

Duttaroy, A. et al. The manganese superoxide dismutase gene of Drosophila: structure, expression, and evidence for regulation by MAP kinase. DNA Cell Biol. 16 , 391–399 (1997).

Paul, A. & Duttaroy, A. Genomic regions responsible for manganese superoxide dismutase regulation in Drosophila melanogaster. Aging Cell 2 , 223–231 (2003).

Tang, H. & Pang, S. Proline Catabolism Modulates Innate Immunity in Caenorhabditis elegans. Cell Rep. 17 , 2837–2844 (2016).

Kenyon, C., Chang, J., Gensch, E., Rudner, A. & Tabtiang, R. A C. elegans mutant that lives twice as long as wild type. Nature 366 , 461–464 (1993).

Friedman, D. B. & Johnson, T. E. A mutation in the age-1 gene in Caenorhabditis elegans lengthens life and reduces hermaphrodite fertility. Genetics 118 , 75–86 (1988).

Uno, M. et al. Neuronal DAF-16-to-intestinal DAF-16 communication underlies organismal lifespan extension in C. elegans. iScience 24 , 102706 (2021).

Lakowski, B. & Hekimi, S. The genetics of caloric restriction in Caenorhabditis elegans. Proc. Natl. Acad. Sci. USA 95 , 13091–13096 (1998).

Itoh, K., Ye, P., Matsumiya, T., Tanji, K. & Ozaki, T. Emerging functional cross-talk between the Keap1-Nrf2 system and mitochondria. J. Clin. Biochem Nutr. 56 , 91–97 (2015).

Tan, B. L., Norhaizan, M. E., Liew, W. P. & Sulaiman Rahman, H. Antioxidant and Oxidative Stress: A Mutual Interplay in Age-Related Diseases. Front Pharm. 9 , 1162 (2018).

Article   Google Scholar  

Calnan, D. R. & Brunet, A. The FoxO code. Oncogene 27 , 2276–2288 (2008).

Shama, S., Lai, C. Y., Antoniazzi, J. M., Jiang, J. C. & Jazwinski, S. M. Heat stress-induced life span extension in yeast. Exp. Cell Res. 245 , 379–388 (1998).

Lithgow, G. J., White, T. M., Melov, S. & Johnson, T. E. Thermotolerance and Extended Life-Span Conferred by Single-Gene Mutations and Induced by Thermal-Stress. P Natl. Acad. Sci. USA 92 , 7540–7544 (1995).

Article   ADS   CAS   Google Scholar  

Morley, J. F. & Morimoto, R. I. Regulation of longevity in Caenorhabditis elegans by heat shock factor and molecular chaperones. Mol. Biol. Cell 15 , 657–664 (2004).

Hsu, A. L., Murphy, C. T. & Kenyon, C. Regulation of aging and age-related disease by DAF-16 and heat-shock factor. Science 300 , 1142–1145 (2003).

Lee, S. J., Murphy, C. T. & Kenyon, C. Glucose Shortens the Life Span of C. elegans by Downregulating DAF-16/FOXO Activity and Aquaporin Gene Expression. Cell Metab. 10 , 379–391 (2009).

Mustafi, S. B., Chakraborty, P. K., Dey, R. S. & Raha, S. Heat stress upregulates chaperone heat shock protein 70 and antioxidant manganese superoxide dismutase through reactive oxygen species (ROS), p38MAPK, and Akt. Cell Stress Chaperon-. 14 , 579–589 (2009).

Zhang, Q., Kopp, M., Babiak, I. & Fernandes, J. M. O. Low incubation temperature during early development negatively affects survival and related innate immune processes in zebrafish larvae exposed to lipopolysaccharide. Sci. Rep. 8 , 4142 (2018).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Pan, M. H. et al. Establishment and characterization of an ovarian cell line of the silkworm, Bombyx mori. Tissue Cell 42 , 42–46 (2010).

Lynn, D. E. Methods for maintaining insect cell cultures. J. Insect Sci. 2 , 9 (2002).

PubMed   PubMed Central   Google Scholar  

Rhiel, M., MitchellLogean, C. M. & Murhammer, D. W. Comparison of Trichoplusia ni BTI-Tn-5B1-4 (High Five((TM))) and Spodoptera frugiperda Sf-9 insect cell line metabolism in suspension cultures. Biotechnol. Bioeng. 55 , 909–920 (1997).

Edgar, R. C. MUSCLE: multiple sequence alignment with high accuracy and high throughput. Nucleic Acids Res. 32 , 1792–1797 (2004).

Ronquist, F. & Huelsenbeck, J. P. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19 , 1572–1574 (2003).

Ronquist, F. et al. MrBayes 3.2: efficient Bayesian phylogenetic inference and model choice across a large model space. Syst. Biol. 61 , 539–542 (2012).

Sun, W. et al. Phylogeny and evolutionary history of the silkworm. Sci. China Life Sci. 55 , 483–496 (2012).

Nylander, J. A., Wilgenbusch, J. C., Warren, D. L. & Swofford, D. L. AWTY (are we there yet?): a system for graphical exploration of MCMC convergence in Bayesian phylogenetics. Bioinformatics 24 , 581–583 (2008).

Song, J. et al. Astragalus Polysaccharide Extends Lifespan via Mitigating Endoplasmic Reticulum Stress in the Silkworm, Bombyx mori. Aging Dis. 10 , 1187–1198 (2019).

Wang, G. H. et al. Reference genes identified in the silkworm Bombyx mori during metamorphism based on oligonucleotide microarray and confirmed by qRT-PCR. Insect Sci. 15 , 405–413 (2008).

Bustin, S. A. et al. The MIQE guidelines: minimum information for publication of quantitative real-time PCR experiments. Clin. Chem. 55 , 611–622 (2009).

Eichenlaub, T. et al. Warburg Effect Metabolism Drives Neoplasia in a Drosophila Genetic Model of Epithelial Cancer. Curr. Biol. 28 , 3220–3228.e3226 (2018).

Gut, P. & Verdin, E. The nexus of chromatin regulation and intermediary metabolism. Nature 502 , 489–498 (2013).

Pietrocola, F., Galluzzi, L., Bravo-San Pedro, J. M., Madeo, F. & Kroemer, G. Acetyl coenzyme A: a central metabolite and second messenger. Cell Metab. 21 , 805–821 (2015).

Zhou, L., He, B., Deng, J., Pang, S. & Tang, H. Histone acetylation promotes long-lasting defense responses and longevity following early life heat stress. Plos Genet 15 , e1008122 (2019).

Brenner, S. The genetics of Caenorhabditis elegans. Genetics 77 , 71–94 (1974).

Pang, S. & Curran, S. P. Adaptive capacity to bacterial diet modulates aging in C. elegans. Cell Metab. 19 , 221–231 (2014).

Moparthi, L. & Koch, S. A uniform expression library for the exploration of FOX transcription factor biology. Differentiation 115 , 30–36 (2020).

Bekker-Jensen, D. B. et al. A Compact Quadrupole-Orbitrap Mass Spectrometer with FAIMS Interface Improves Proteome Coverage in Short LC Gradients. Mol. Cell Proteom. 19 , 716–729 (2020).

Rasmussen, S. H. et al. Cohort Profile: The 1895, 1905, 1910 and 1915 Danish Birth Cohort Studies - secular trends in the health and functioning of the very old. Int J. Epidemiol. 46 , 1746 (2017).

Robine, J. M. et al. Centenarians Today: New Insights on Selection from the 5-COOP Study. Curr. Gerontol. Geriatr. Res 2010 , 120354 (2010).

PubMed   Google Scholar  

Pedersen, D. A. et al. The Danish Twin Registry: An Updated Overview. Twin Res Hum. Genet 22 , 499–507 (2019).

Miller, S. A., Dykes, D. D. & Polesky, H. F. A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic acids Res. 16 , 1215 (1988).

Howie, B. N., Donnelly, P. & Marchini, J. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. Plos Genet 5 , e1000529 (2009).

Chang, C. C. et al. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience 4 , 7 (2015).

Pedersen, C. B., Gotzsche, H., Moller, J. O. & Mortensen, P. B. The Danish Civil Registration System. A cohort of eight million persons. Dan. Med Bull. 53 , 441–449 (2006).

Download references

Acknowledgements

We thank Dr. Shilin Song and Prof. Shanshan Pang (Chongqing University), Kaige Hao, Tao Sun, Li Xiao, and Yunzhu Shang for their generous technical support with D. melanogaster , C. elegans , and B. mori assays. We would also like to thank associate Professor Stefan Koch from the Department of Biomedical and Clinical Sciences (BKV), Linköping University, Sweden, for generously sharing pCS2, FOXO1, FOXO3, FOXO4, FOXO6 plasmids. We also thank BioMedVisual.com for helping with the working model illustrations. We express our sincere gratitude to the editors and anonymous reviewers for their meticulous examination and valuable feedback on our manuscript. This work is supported by grants from the National Natural Science Foundation of China (Grant Nos. 32272939, 32330102, and 31902215), Natural Science Foundation of Chongqing, China (Grant no. cstc2021jcyj-cxttX0005), and Funds of China Agriculture Research System of MOF and MARA (No. CARS-18-ZJ0102). L.J.R. receives financial support from the Nordea-Fonden and Novo Nordisk Foundation (NNF23OC0084974 and NNF17OC0027812). L.J.R. is a member of the Clinical Academic Group: Recovery Capacity After Acute Illness in An Aging Population (RECAP). The studies behind the individuals included here received funding from The National Program for Research Infrastructure 2007 (grant no. 09-063256) from the Danish Agency for Science Technology and Innovation, the Velux Foundation, the US National Institute of Health (P01 AG08761), the Danish Agency for Science, Technology and Innovation/The Danish Council for Independent Research (grant no. 11-107308), the European Union’s Seventh Framework Programme (FP7/2007-2011) under grant agreement n° 259679, the INTERREG 4 A programme Syddanmark-Schleswig-K.E.R.N. (by EU funds from the European Regional Development Fund), the CERA Foundation (Lyon), and the AXA Research Fund, Paris. Genotyping of data set 2 was conducted by the SNP&SEQ Technology Platform, Science for Life Laboratory, Uppsala, Sweden ( http://snpseq.medsci.uu.se/genotyping/snp-services/ ) and supported by NIH R01 AG037985. L.Z. is supported by a grant from NSFC (No. 32000810). V.A.B. receives financial support from the Nordea-fonden and Novo Nordisk Foundation (NNF17OC0027812). Z.Y. was supported by the Non-profit Central Research Institute Fund of the Chinese Academy of Medical Sciences (grant no. 2023-RC180-03), the Chinese Academy of Medical Sciences (CAMS) Innovation Fund for Medical Sciences (2022-I2M-2-004, 2023-I2M-2-005) and the NCTIB Fund for the R&D Platform for Cell and Gene Therapy. Work at The Novo Nordisk Foundation Center for Protein Research (CPR) was funded in part by a donation from the Novo Nordisk Foundation (NNF14CC0001).

Author information

These authors contributed equally: Jiangbo Song, Zhiquan Li.

Authors and Affiliations

State Key Laboratory of Resource Insects, Key Laboratory for Sericulture Biology and Genetic Breeding, Ministry of Agriculture and Rural Affairs, College of Sericulture, Textile and Biomass Sciences, Southwest University, Chongqing, 400715, China

Jiangbo Song, Lei Zhou, Xin Chen, Xiaoling Tong & Fangyin Dai

Center for Healthy Aging, Department of Cellular and Molecular Medicine, University of Copenhagen, 2200, Copenhagen, Denmark

Zhiquan Li, Yuan Li, Vilhelm A. Bohr & Lene Juel Rasmussen

Department of Cellular and Molecular Medicine, University of Copenhagen, 2200, Copenhagen, Denmark

Wei Qi Guinevere Sew & Héctor Herranz

Novo Nordisk Foundation Center for Protein Research, University of Copenhagen, 2200, Copenhagen, Denmark

Zilu Ye & Jesper Velgaard Olsen

Key Laboratory of Common Mechanism Research for Major Diseases, Suzhou Institute of Systems Medicine, Chinese Academy of Medical Sciences & Peking Union Medical College, Suzhou, China

Epidemiology, Biostatistics and Biodemography, Department of Public Health, University of Southern Denmark, Odense, Denmark

Marianne Nygaard & Kaare Christensen

Department of Clinical Genetics, Odense University Hospital, Odense, Denmark

Department of Clinical Biochemistry, Odense University Hospital, Odense, Denmark

Kaare Christensen

Section on DNA repair, National Institute on Aging, National Institutes of Health, Baltimore, MD, 21224, USA

Vilhelm A. Bohr

You can also search for this author in PubMed   Google Scholar

Contributions

J.S.: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing - Original Draft, Writing - Review & Editing, Visualization, Project administration, Funding acquisition. Z.L.: Conceptualization, Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing - Original Draft, Writing - Review & Editing, Visualization, Project administration, Funding acquisition. L.Z.: Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing - Review & Editing, Funding acquisition. X.C.: Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing - Review & Editing. W.Q.G.S.: Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing - Review & Editing. H.H.: Methodology, Resources, Writing - Review & Editing, Funding acquisition. Z.Y.: Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing - Review & Editing, Visualization. J.V.O.: Resources, Writing - Review & Editing, Funding acquisition. Y.L.: Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing - Review & Editing. M.N.: Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing - Review & Editing, Funding acquisition. K.C.: Methodology, Validation, Formal analysis, Investigation, Data Curation, Writing - Review & Editing, Funding acquisition. X.T.: Writing - Review & Editing. V.A.B.: Writing - Review & Editing, Funding acquisition. L.J.R.: Conceptualization, Resources, Writing - Original Draft, Writing - Review & Editing, Supervision, Project administration, Funding acquisition. F.D.: Conceptualization, Resources, Writing - Original Draft, Writing - Review & Editing, Supervision, Project administration, Funding acquisition.

Corresponding authors

Correspondence to Lene Juel Rasmussen or Fangyin Dai .

Ethics declarations

Competing interests.

The authors declare no competing interests.

Peer review

Peer review information.

Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, description of additional supplementary files, supplementary data 1, supplementary data 2, supplementary data 3, supplementary data 4, supplementary data 5, supplementary data 6, supplementary data 7, reporting summary, source data, source data, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/ .

Reprints and permissions

About this article

Cite this article.

Song, J., Li, Z., Zhou, L. et al. FOXO-regulated OSER1 reduces oxidative stress and extends lifespan in multiple species. Nat Commun 15 , 7144 (2024). https://doi.org/10.1038/s41467-024-51542-z

Download citation

Received : 22 October 2022

Accepted : 12 August 2024

Published : 21 August 2024

DOI : https://doi.org/10.1038/s41467-024-51542-z

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.