SLAS Technology

YOU ARE NOT A ROBOT... SO DON‘T ACT LIKE ONE FREE YOURSELF FROM ROUTINE PIPETTING Making hands-free serial dilutions, reagent additions, sample reformatting, normalizations and hit picking very affordable for every lab. ASSIST PLUS Automating Handheld Pipettes SINGLE CHANNEL PIPETTING MODULE NEW D-ONE NEW VOYAGER - Adjustable Tip Spacing Pipettes VIAFLO - Electronic Pipettes D-ONE - Single Channel Pipetting Module

SLAS TECHNOLOGY Vol. 28 ISSN: 2472-6311 Editor-in-Chief: Edward Kai-Hua Chow, Ph.D. Translating Life Sciences Innovation DZ ® ?태v¨ 2³ËÀ®v¨ ³– O4 O Reprinted from SLAS Technology Ten Editor’s Selections from 2022 Reprint Collection May 20 3


SLAS Technology Volume 27 (2022) Amsterdam ∙ Boston ∙ London ∙ New York ∙ Oxford ∙ Paris ∙ Philadelphia ∙ San Diego ∙ St. Louis 0D\ 6/$6 7HFKQRORJ\ 7HQ (GLWRU·V 6HOHFWLRQV IURP 5HSULQW &ROOHFWLRQ 7KLV SULQW UHSULQW FROOHFWLRQ LV VSRQVRUHG E\ SDLG DGYHUWLVLQJ May 2023 SLAS Technology Ten Editor’s Selections from 2022 Reprint Collection This print reprint collection is sponsored by paid advertising.

We encourage your submissions to SLAS Technology, an official open access journal of SLAS. SLAS Technology publishes research that reveals how scientists adapt technological advancements for life sciences exploration and experimentation in biomedical research and development. The Editor-in-Chief is Edward Kai-Hua Chow, Ph.D. of the National University of Singapore. Submit your paper and learn more at Here are a few reasons why you should consider publishing with SLAS Technology. • The journal’s Impact Factor is 2.813. • SLAS Technology’s CiteScore is 5.6. • The journal is fully indexed in PubMed as well as 12 other international databases. • Publishing open access ensures your article can be read and cited by researchers worldwide. • Your open access article will remain discoverable and freely accessible, no matter how much time passes – that’s a guarantee. Submit your next paper to SLAS Technology!

Editor-in-Chief Edward Kai-Hua Chow, Ph.D. National University of Singapore, Singapore Associate Editors Joe Olechno, Ph.D. Beckman Coulter Life Sciences, San Jose, California, United States of America Pak Kin Wong, Ph.D. The Pennsylvania State University Department of Biomedical Engineering, University Park, Pennsylvania, United States of America Scientifi c Advisors Dino Di Carlo, Ph.D. University of California Los Angeles Department of Bioengineering, Los Angeles, California, United States of America Dean Ho, Ph.D. UCLA School of Dentistry, Los Angeles, California, United States of America Section Editors Tal Murthy, Ph.D. TTP Labtech, Cambridge, Massachusetts, United States of America David Yeo, Ph.D. Nanyang Technological University School of Chemical and Biomedical Engineering, Singapore Protocols Editor Adam Zweifach, Ph.D. University of Connecticut, Storrs, Connecticut, United States of America Editorial Board Members Giancarlo Basile, M.Sc. Lexogen GmbH, Wien, Austria Amar Basu, Ph.D. Wayne State University, Detroit, Michigan, United States of America Matthew Boeckeler, M.B.A. AstraZeneca R&D Boston, Waltham, Massachusetts, United States of America John Burczak, Ph.D. GE Global Research, Niskayuna, New York, United States of America Eun Ji Chung, Ph.D. University of Southern California Department of Biomedical Engineering, Los Angeles, California, United States of America Nathan Coussens, Ph.D. Frederick National Laboratory for Cancer Research, Frederick, Maryland, United States of America Heidi Fleischer, Ph.D. University of Rostock Center for Life Science Automation, Rostock, Germany Stephen Hughes, Ph.D. GenScript Biotech Corp, Piscataway, New Jersey, United States of America Claire Hur, Ph.D. Johns Hopkins University, Baltimore, Maryland, United States of America Sudhakar Jha, Ph.D. Oklahoma State University College of Veterinary Medicine, Stillwater, Oklahoma, United States of America Tejas Khire, Ph.D. Schenectady, New York, United States of America Felix Lenk, Ph.D. TU Dresden Institute of Natural Materials Technology, Germany Hui Li, Ph.D. Southern Illinois University Carbondale School of Electrical, Computer, and Biomedical Engineering, Carbondale, Illinois, United States of America Yiyan Li, Ph.D. Fort Lewis College, Durango, Colorado, United States of America Kevin Moore Tecan Group Ltd, Männedorf, Switzerland Michael Santillo, Ph.D. US Food and Drug Administration, Silver Spring, Maryland, United States of America Alex Shen, Ph.D. Kite Pharma Inc, Santa Monica, California, United States of America Aleidy Silva, Ph.D. University of California Los Angeles Department of Mechanical and Aerospace Engineering, Los Angeles, California, United States of America Andreas Traube, Ph.D. Fraunhofer Institute for Manufacturing Engineering and Automation IPA, Stuttgart, Germany Hideaki Tsutsui, Ph.D. University of California Riverside Department of Mechanical Engineering, Riverside, California, United States of America Katherine Wang, Ph.D. Ascendis Pharma, Palo Alto, California, United States of America Chenjie Xu, Ph.D. Nanyang Technological University School of Chemical and Biomedical Engineering, Singapore, Singapore Joe Yeong Poh Sheng, Ph.D. Singapore General Hospital, Singapore, Singapore Edmond W. K. Young, Ph.D. University of Toronto, Toronto, Ontario, Canada Ali Zarrinpar, Ph.D. University of Florida College of Medicine, Gainesville, Florida, United States of America Piero Zucchelli, Ph.D. Andrew Alliance, Genève, Switzerland Podcast Editor David Pechter, M.S.M.E. PerkinElmer, Cambridge, Massachusetts, United States of America Publishing Manager Ally Jump, M.Sc. SLAS, Pittsburgh, PA, United States of America

Learn more about these resources, services and more at Author Resources Whether you are planning a research project or working on your next manuscript, Elsevier’s author resources are there to help. OResearch & Preparation OResearcher Academy: Prepare for the next step in your research career with our free e-learning modules. OWriting OMendeley: Simplify your workflow, accelerate your research. OLanguage editing services: Ensure that your work is written in correct English before you submit. OTranslation services: Have your work translated into correct English. OSubmission & Publication OPublishing ethics resources: Access resources to help you ensure your work is carried out and published ethically. OSubmission and revision: Take the pain out of submission with streamlined systems and processes.

SLAS Technology Ten Editors Selections from 2022 Reprint Collection CONTENTS May 2023 The 2023 SLAS Technology ten: Translating life sciences innovation Edward Kai-Hua Chow 1 Comparison Between Franz Diff usion Cell and a novel Micro-physiological System for In Vitro Penetration Assay Using Diff erent Skin Models Ilaria Pulsoni, Markus Lubda, Maurizio Aiello, Arianna Fedi, Monica Marzagalli, Joerg von Hagen, Silvia Scaglione 2 Sorting single-cell microcarriers using commercial fl ow cytometers Joseph de Rutte, Robert Dimatteo, Sheldon Zhu, Maani M Archang, Dino Di Carlo 3 COVID-19 detection using chest X-ray images based on a developed deep neural network Zohreh Mousavi, Nahal Shahini, Sobhan Sheykhivand, Sina Mojtahedi, Afrooz Arshadi 4 Simple assessment of viability in 2D and 3D cell microarrays using single step digital imaging Anna A. Popova, Markus Reischl, Daniel Kazenmaier, Haijun Cui, Timo Amberger, Pavel A. Levkin 5 Enabling high throughput drug discovery in 3D cell cultures through a novel bioprinting workfl ow Martin Engel, Lisa Belfi ore, Behnaz Aghaei, Margareta Sutija 6 Quantitative determination of uric acid using paper-based biosensor modifi ed with graphene oxide and 5-amino-1,3,4-thiadiazole-2-thiol Yaw-Jen Chang, Ming-Che Lee, You-Chiuan Chien 7 DeepImageTranslator: A free, user-friendly graphical interface for image translation using deep-learning and its applications in 3D CT image analysis Run Zhou Ye, Christophe Noll, Gabriel Richard, Martin Lepage, Éric E. Turcotte, André C. Carpentier 8 AI-driven laboratory workfl ows enable operation in the age of social distancing Diego Marescotti, Chandrasekaran Narayanamoorthy, Filipe Bonjour, Ken Kuwae, Luc Graber, Florian Calvino-Martin, Samik Ghosh, Julia Hoeng 9 A Biomimetic High Throughput Model of Cancer Cell Spheroid Dissemination onto Aligned Fibrillar Collagen Hossam Ibrahim, Stephen D. Thorpe, Michael Paukshto, Tatiana S. Zaitseva, Wolfgang Moritz, Brian J. Rodriguez 10 Automation enables high-throughput and reproducible single-cell transcriptomics library preparation David Kind, Praveen Baskaran, Fidel Ramirez, Martin Giner, Michael Hayes, Diana Santacruz, Carolin K. Koss, Karim C. el Kasmi, Bhagya Wijayawardena, Coralie Viollet Note: The articles are reprinted in their original format, with the pagination of their publication in the journal. Articles should be cited by their original publication, as noted on each title page. Contents lists available at ScienceDirect SLAS Technology j o u r n a l h ome p a g e : h t t p s : / / s l a s - t e c h n o l o g y. o r g / SLAS TECHNOLOGY Vol. 28 ISSN: 2472-6311 Editor-in-Chief: Edward Kai-Hua Chow, Ph.D. Translating Life Sciences Innovation DZ ® ?태v¨ 2³ËÀ®v¨ ³– O4 O Reprinted from SLAS Technology Ten Editor’s Selections from 2022 Reprint Collection May 20 3

Aims and Scope SLAS Technology reveals how scientists adapt technological advancements for life sciences exploration and experimentation in biomedical research and development. The journal emphasizes scientifi c and technical advances that enable and improve: • Life sciences research and development • Drug delivery • Diagnostics • Biomedical and molecular imaging • Personalized and precision medicine SLAS Technology’s Editorial Scope includes: the latest news in high-throughput and other laboratory automation technologies; micro/nanotechnologies, analytical, separation and quantitative techniques; synthetic chemistry and biology; informatics (data analysis, statistics, bio genomic and chemoinformatics). SLAS Technology is of particular interest to those involved in automation, bioengineering, micro- and nanotechnology, nanomedicine, microfl uidics, 3D printing, biotechnology, bioinformatics, analytical chemistry, biomedical optics, information technology, artifi cial intelligence, spectroscopy, clinical diagnostics, pharmacogenomics, molecular biology and biomedical sciences. Publication Information: SLAS Technology (ISSN 2472-6311). For 2023, volumes 28 (1-6 issues) are scheduled for publication. Further information is available on this journal and other Elsevier products through Elsevier’s website ( Back issues: Back issues of all previously published volumes are available direct from Elsevier Offi ces (Oxford and New York). Complete volumes and single issues can be purchased for 1997–2003. Earlier issues are available in high quality photo-duplicated copies as complete volumes only. Back volumes on microfi lm are available from UMI, 300 North Zeeb Road, Ann Arbor, MI 48106, USA. Orders, claims, and journal inquiries: Please contact the Elsevier Customer Service Department nearest you: St. Louis: Elsevier Customer Service Department, 3251 Riverport Lane, Maryland Heights, MO 63043, USA; phone: (877) 8397126 [toll free within the USA]; (+1) (314) 4478878 [outside the USA]; fax: (+1) (314) 4478077; e-mail: Oxford: Elsevier Customer Service Department, The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK; phone: (+44) (1865) 843434; fax: (+44) (1865) 843970; e-mail: Tokyo: Elsevier Customer Service Department, 4F Higashi-Azabu, 1-Chome Bldg, 1-9-15 Higashi-Azabu, Minato-ku, Tokyo 106-0044, Japan; phone: (+81) (3) 5561 5037; fax: (+81) (3) 5561 5047; e-mail: The Philippines: Elsevier Customer Service Department, 2nd Floor, Building H, UP-Ayalaland Technohub, Commonwealth Avenue, Diliman, Quezon City, Philippines 1101; phone: (+65) 63490222; fax: (+63) 23521394; e-mail: Language (usage and editing services): Please write your text in good English (American or British usage is accepted, but not a mixture of these). Authors who feel their English language manuscript may require editing to eliminate possible grammatical or spelling errors and to conform to correct scientifi c English may wish to use the English Language Editing service available from Elsevier’s Author Services at language-editing/ or visit our customer support site https://service.elsevier. com for more information. Advertising information: If you are interested in advertising or other commercial opportunities please e-mail and your inquiry will be passed to the correct person who will respond to you within 48 hours. Note to contributors: Online submission is now available for this journal, to submit your article online, go to © 2023 Published by Elsevier Inc. on behalf of Society for Laboratory Automation and Screening. This journal and the individual contributions contained in it are protected under copyright, and the following terms and conditions apply to their use in addition to the terms of any Creative Commons or other user license that has been applied by the publisher to an individual article: Photocopying: Single photocopies of single articles may be made for personal use as allowed by national copyright laws. Permission is not required for photocopying of articles published under the CC BY license nor for photocopying for non-commercial purposes in accordance with any other user license applied by the publisher. Permission of the publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for non-profi t educational classroom use. Derivative Works: Users may reproduce tables of contents or prepare lists of articles including abstracts for internal circulation within their institutions or companies. Other than for articles published under the CC BY license, permission of the publisher is required for resale or distribution outside the subscribing institution or company. For any subscribed articles or articles published under a CC BY-NC-ND license, permission of the publisher is required for all other derivative works, including compilations and translations. Storage or Usage: Except as outlined above or as set out in the relevant user license, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the publisher. Permissions: For information on how to seek permission visit or call: (+44) 1865 843830 (UK) / (+1) 215 239 3804 (USA). Notice: No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, independent verifi cation of diagnoses and drug dosages should be made. Although all advertising material is expected to conform to ethical (medical) standards, inclusion in this publication does not constitute a guarantee or endorsement of the quality or value of such product or of the claims made of it by its manufacturer. Funding Body Agreements and Policies: Elsevier has established agreements and developed policies to allow authors whose articles appear in journals published by Elsevier, to comply with potential manuscript archiving requirements as specifi ed as conditions of their grant awards. To learn more about existing agreements and policies please visit Author Inquiries: You can track your submitted article at https://www. You can track your accepted article at https:// You are also welcome to contact Customer Support via The paper used in this publication meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper) Printed by CB Printforce Nederland B.V., Textielweg 12, Culemborg, NL For a full and complete Guide for Authors, please go to:

Reprinted from: Publication of this reprint collection is supported by paid advertising SLAS Technology 28 (2023) 1 Contents lists available at ScienceDirect SLAS Technology journal homepage: Editorial The 2023 SLAS Technology ten: Translating life sciences innovation Looking back on the 27 th year of SLAS Technology, we are excited by the continued innovations and evolving technology that are facilitating life science and biomedical research. SLAS Technology plays a unique role in being a focused presenter of these advances and we are proud of the novel technology innovations published in SLAS Technology this year. With the rise of digital technologies, we continue to see a trend towards increased implementation of artificial intelligence (AI) and digital tools across all aspects of life science and biomedical research. Advances in laboratory automation, micro- and nanotechnologies and biomaterials continue to improve the clinical relevance of life science and biomedical research. As evidenced by the work presented here, SLAS Technology continues to stay at the forefront of research trends highlighted here. Every year, SLAS Technology Ten presents ten individual articles that stand out as the most innovative scientific achievements published in SLAS Technology in the past 12 months. The 2023 SLAS Technology Ten highlights a broad range of technologies that address a broad range of unmet needs in the laboratory and the clinic. These include advances in microtechnology to improve molecular detection and chemical characterization [1–4]. Advances in AI continue to be a major highlight of SLAS Technology. AI-driven technologies enable all aspects of life science and biomedical research and diagnostics, including directly impacting the clinic [5–7]. Automation remains a critical area of interest for SLAS Technology as automation allows for increasingly complex high-throughput drug discovery and single-cell transcriptomic library preparation workflows [8–10]. This is particularly important as life science and biomedical research moves away from 2D cell line models and towards more physiologically relevant 3D cell models. Technological advances presented here can ensure more uniform 3D cell models that are suitable for high-throughput screening with improved reproducibility. We would like to thank the authors that contributed to the work presented in these ten most highly cited SLAS Technology papers of 2022 as well as the hundreds of other researchers who also chose to share their technology advancements and discoveries in SLAS Technology in 2022. It is solely because of their contributions that SLAS Technology continues to present transformative technology that improves all aspects of life science and biomedical research. Edward Kai-Hua Chow Editor-in-Chief, SLAS Technology E-mail address: References [1] Pulsoni I, Lubda M, Aiello M, et al. Comparison between franz diffusion cell and a novel micro-physiological system for in vitro penetration assay using different skin models. SLAS Technol 2022;27:161–71. [2] de Rutte J, Dimatteo R, Zhu S, et al. Sorting single-cell microcarriers using commercial flow cytometers. SLAS Technol 2022;27:150–9. [3] Chang YJ, Lee MC, Chien YC. Quantitative determination of uric acid using paper-based biosensor modified with graphene oxide and 5-amino-1,3,4-thiadiazole-2-thiol. SLAS Technol 2022;27:54–62. [4] Popova AA, Dietrich S, Huber W, et al. Miniaturized drug sensitivity and resistance test on patient-derived cells using droplet-microarray. SLAS Technol 2021;26:274–86. [5] Mousavi Z, Shahini N, Sheykhivand S, et al. COVID-19 detection using chest X-ray images based on a developed deep neural network. SLAS Technol 2022;27:63–75. [6] Ye RZ, Noll C, Richard G, et al. DeepImageTranslator: A free, user-friendly graphical interface for image translation using deep-learning and its applications in 3D CT image analysis. SLAS Technol 2022;27:76–84. [7] Marescotti D, Narayanamoorthy C, Bonjour F, et al. AI-driven laboratory workflows enable operation in the age of social distancing. SLAS Technol 2022;27:195–203. [8] Engel M, Belfiore L, Aghaei B, et al. Enabling high throughput drug discovery in 3D cell cultures through a novel bioprinting workflow. SLAS Technol 2022;27:32–8. [9] Ibrahim H, Thorpe SD, Paukshto M, et al. A Biomimetic high throughput model of cancer cell spheroid dissemination onto aligned fibrillar collagen. SLAS Technol 2022;27:267–75. [10] Kind D, Baskaran P, Ramirez F, et al. Automation enables high-throughput and reproducible single-cell transcriptomics library preparation. SLAS Technol 2022;27:135–42. 2472-6303/© 2022 Published by Elsevier Inc. on behalf of Society for Laboratory Automation and Screening. This is an open access article under the CC BY-NC-ND license (

Microscale Innovation in Life Sciences Symposium SEPTEMBER 14-15 LA JOLLA | CA, USA HILTON LA JOLLA TORREY PINES WASHINGTON DC | USA OCTOBER 12-13THE WESTIN GEORGETOWN Data Sciences and AI Symposium BASEL | SWITZERLAND NOVEMBER 14-15 2023 | SYMPOSIA SERIES Mark your calendar and make plans to join your fellow professionals at the highly anticipated SLAS Fall Symposia Series, happening in the U.S.A. and Switzerland. Don't miss out on these cutting-edge symposia—find out more and sign up today to secure your spot!

Reprinted from: Publication of this reprint collection is supported by paid advertising SLAS Technology 27 (2022) 161–171 Contents lists available at ScienceDirect SLAS Technology journal homepage: Comparison Between Franz Diffusion Cell and a novel Micro-physiological System for In Vitro Penetration Assay Using Different Skin Models Ilaria Pulsoni a,# , Markus Lubda b,# , Maurizio Aiello a,c , Arianna Fedi c , Monica Marzagalli a , Joerg von Hagen b , Silvia Scaglione a,c,∗ a React4life Srl, Genoa, Italy bMerck KGaA, Darmstadt, Germany c CNR –National Research Council of Italy, Genova, Italy a r t i c l e i n f o Keywords: Drug delivery caffeine penetration in vitro diffusion system skin model a b s t r a c t In vitro diffusive models are an important tool to screen the penetration ability of active ingredients in various formulations. A reliable assessment of skin penetration enhancing properties, mechanism of action of carrier systems, and an estimation of a bioavailability are essential for transdermal delivery. Given the importance of testing the penetration kinetics of different compounds across the skin barrier, several in vitro models have been developedThe aim of this study was to compare the Franz Diffusion Cell (FDC) with a novel fluid-dynamic platform (MIVO) by evaluating penetration ability of caffeine, a widely used reference substance, and LIP1, a testing molecule having the same molecular weight but a different lipophilicity in the two diffusion chamber systems. A 0.7% caffeine or LIP1 formulation in either water or propylene glycol (PG) containing oleic acid (OA) was topically applied on the Strat-M® membrane or pig ear skin, according to the infinite-dose experimental condition (780 ul/cm2 ). The profile of the penetration kinetics was determined by quantify the amount of molecule absorbed at different time-points (1, 2, 4, 6, 8 hours), by means of HPLC analysis. Both diffusive systems show a similar trend for caffeine and LIP1 penetration kinetics. The Strat-M® skin model shows a lower barrier function than the pig skin biopsies, whereby the PGOA vehicle exhibits a higher penetration, enhancing the effect for both diffusive chambers and skin surrogates. Most interestingly, MIVO diffusive system better predicts the lipophilic molecules (i.e. LIP1) permeation through highly physiological fluid flows resembled below the skin models. Introduction The skin tissue is an effective barrier, representing a fundamental interface between the human body and the external environment. Based on its chemical-physical features, it forms a protective layer against harmful environmental influences such as ultra-violet light, microorganisms, pollutants and environmental toxins, pesticides, or other chemical drugs. Moreover, skin regulates temperature and homeostasis of the body, particularly by limiting the loss of water [1]. Dermal absorption assays are routinely adopted to predict risks from skin exposure to chemicals, but also to demonstrate benefits after topic application of cosmetics, medical device or therapeutic active ingredient. In this context, the Organization for Economic Co-operation and Development (OECD) and the United States Environmental Protection Agency have produced guidelines for the in vitro and in vivo assessment of percutaneous absorption [2], that establishes the passage of ∗ Corresponding author: Silvia Scaglione, React4life Srl, Via Greto di Cornigliano 6/R, 16152 Genoa (Italy) E-mail address: (S. Scaglione). # These authors contributed equally to this work. compounds across the skin. This process is basically divided in three steps [3]: i) penetration, which consists of the entry of the chemical compound into the stratum corneum, build up by an intracellular lipid matrix of mainly ceramides, cholesterol and free fatty acids [4,5]; ii), permeation , namely the gradual passage of the substance through the subsequent layers, which are both functionally and anatomically distinct from the stratum corneum; iii) the uptake t into lymphatic and blood vessels [5-8]. Interestingly, according to the skin absorption outcome, the classification of these testing compounds may range from cosmetics to “medical devices made of substances”, that need to be absorbed in order to achieve their intended action [9-11]. Therefore, there is increasing demand for reliable and reproducible in vitro and ex vivo skin absorption methods that accelerate the chemicals testing and the measurement of their absorption percentage. In the last two decades, the European Union and national legislations have stipulated that animal experiments should be avoided whenever scien 2472-6303/© 2021 The Author(s). Published by Elsevier Inc. on behalf of Society for Laboratory Automation and Screening. This is an open access article under the CC BY license (

I. Pulsoni, M. Lubda, M. Aiello et al. SLAS Technology 27 (2022) 161–171 tifically feasible, in line with the 3Rs (reduce, replace, refine) approach [2]. Furthermore, owing to the difference in skin structure, animal studies do not always reflect the human situation [2]. The estimation of percutaneous absorption of compounds using excised animal skin is widely accepted for the toxicological risk assessment [12]. Porcine skin has been re-cognized as an appropriate tissue for prediction of human skin permeability for some years, despite the lower barrier function of this tissue compared with human skin [13-17]. Recently, numerous skin surrogate systems and human skin equivalents (HSEs) have been developed to study skin penetration to overpass limits of animal sources. HSEs are typically constructed by culturing human keratinocytes on appropriate substrates [18], providing a good-quality control of the system and species relevance. Consequently, some reconstructed tissues (i.e. Episkin and EpiDerm) have already been validated under specific applications, such as for skin irritation [19,20]. In general, they have proved a lower barrier function than human skin [18,21,22], but for hydrophilic compounds (i.e. ethyl ascorbic acid [23] and caffeine [21]) some HSEs (i.e. LabSkin, EpiDerm, EpiSkin, SkinEthic) represent a valuable option to carry out transdermal delivery investigations, even though some regulation guidelines still have to be updated [10,24]. Artificial membranes have been also fabricated and employed as synthetic skin alternatives [25]. Non-lipid- and lipid-based membranes are cost-effective and reproducible tools to study the underlying physicochemical mechanisms of the passive drug diffusion, [24,26]. Moreover, being non-biological models, these skin models may support highthroughput screenings, without lot-to-lot variability, safety and storage limitations [4]. The parallel artificial membrane permeability assay (PAMPA) may be represent an alternative to simulate dermal absorption of some compounds [25,27,28], whose results displayed a good correlation with . full-thickness skin (R2>0.7). The Strat-M® membrane is an animal free, multiple layer polyether sulfone membrane, which is coated with skin lipids (e.g., ceramides, cholesterol, and free fatty acids), especially designed to mimicking the skin structure for transdermal diffusion testing . The hydrophobic lipidic structure, which is coated on the membrane is composed of the main stratum corneum lipids, paired with the polyether sulfone membrane cut-off this membrane mimics a skin penetration more closely than other membranes, which are limited only on their cut-off definition [29]. These chemico-physical porperties make Strat-M® membrane a good skin model alternative to evaluate penetration flux and permeation of molecules, under infinite and finite dosing conditions [30]. However, due to their lower barrier function, artificial membranes typicaly lead to a higher penetration ability and penetration rate [12,13,31]. For ex vivo skin penetration studies the OECD guideline recommends pig ear skin as suitable skin surrogate to mimic human percutaneous penetration [24,32]. Pig ear skin shows similarities in morphology as well as penetration abilities and corresponds to that of human skin [33].Typically, the passive diffusion of active ingredients is tested by culturing surrogate skin in a Franz Diffusion Cell (FDC) setup [34]. The FDC system can be either a static or flow-through setup. Although both are compliant with the OECD Test Guidelines 428, static FDC setup remains simpler, lower-cost and more widely used diffusive system [12,24]. It consists of a receptor compartment filled with a physiological buffer solution, in which the compound is released after penetrating through the skin surrogate. Onto this surrogate a finite (≤10μL/cm2 ) or infinite (≥10μL/cm2 ) formulation dose can be topically applied into the donor compartment and allows the evaluation of penetration kinetics over time [35,36]. Finite dosing represents more closely application and usage condition, whereas an infinite dosing helps to understand and elaborate permeation abilities due to a steady state penetration and a constant high formulation concentration. The penetration kinetics interpretation underlines the predicted bioavailability of the active ingredient, which is important to ensure the efficacy and the exposure to the living cell entity and a targeted drug delivery. The hydrophilic alkaloid caffeine is recommended as a model compound by the OECD guideline for in vivo, ex vivo, and in vitro percutaneous penetration testing due to its well-known penetration behaviour [32]. To optimize and ensure targeted delivery of ingredients into the skin, the formulation needs to be designed individually [37]. Most ingredients are formulated into a standard formulation (vehicle) containing water, glycols and fatty acids for penetration testing based on their physico-chemical properties and solubility characteristics [38]. Thereby, propylene glycol (PG) is the most used glycol and is often combined with other penetration enhancers like oleic acid (OA) to test ingredient penetration [39]. In this work, we compared the FDC with a novel ready-to-use, compartmental technology , named MIVO® - Multi In Vitro Organ device, compliant with the OECD 428 guideline’s definition of a diffusion cell and able to properly resemble the mono-directional physiological capillary-like flow below the tissue. MIVO has been already adopted to carry out diffusion studies in gut absorption [40], tumor cells intravasation [41] and cancer drug efficacy testing [42]. Here, molecules having different lipophilicity (expressed as log p values) although same molecular weight (i.e. caffeine and LIP1, acronym for 1,3-Benzodioxol-5-ylmethylurea lipophilic molecule) have been adopted as testing molecules for skin permeation assays through two human surrogate models, the Strat-M® membrane and the pig ear skin, carried out employing both FDC, as it is the standard in vitro method for this type of study, and MIVO®. Also, computational fluiddynamic (CFD) simulations have been performed to inspect the flow field beneath the skin in the receptor of both diffusion cells considered. Materials and method Diffusive chambers MIVO® system The MIVO® device is a disposable cell culture chamber able to host living tissues (e.g. cellular monolayers, 3D reconstructed tissues, tissue biopsies) or artificial membranes under physiological conditions, providing a multiple fluidic circulation that mimics the human circulatory system with the vascularization of the tissue of interest. The diffusion cell designed and implemented in this work is schematically represented in Figure 1, panel A, showing features in compliance with the OECD 428 rule for in vitro skin absorption method. Specifically, the human skin surrogate separates the donor compartment from the receiver one, exposing a surface area of 0.43 cm2 suitable for drug administration. The receiver compartment has been designed to be connected to a peristaltic pump inducing a monodirectional flow: then, a capillary velocity can be set up below the skin barrier, emulating the real physiological conditions. A three-way valve placed in the fluidic circuit allows the sampling of the media over time, without affecting the sterility environment and the tissue. The whole system is hosted within the cell culture incubator with controlled atmosphere (i.e., 32°C, 5% CO2). Strat-M® membranes and pig skin biopsy discs were placed within MIVO®; an add-on equipped with a biocompatible o-ring avoiding any fluid leakage blocks the skin models inside the MIVO® chamber. The donor compartment was filled with a volume of caffeine formulations according to the infinite-dose experimental condition (780uL/cm2 ), and the receiver one was filled with 2.3 mL of calcium and magnesium enriched phosphate buffered saline (PBS), at a flow rate of 2 mL/min in order to have below the skin surrogate a mean flow velocity (of 0.1 cm/s), resembling the capillary flow [42,43]. The experiment with Strat-M® was performed in triplicate, while the one with pig ear skin was performed with 6 replicates. For both skin models, 400 L (or 20%) of the circulating solution in the receiver compartment were collected at different time-points (1, 2, 4, 6, 8 hours), in order to assess the quantity of caffeine/LIP1 accumulating into the receiver chamber over time. The samples were filtered through a cellulose acetate membrane filter (0.22 m pore size) and analyzed through High-Performance Liquid Chromatography (HPLC). 162

I. Pulsoni, M. Lubda, M. Aiello et al. SLAS Technology 27 (2022) 161–171 Figure 1. Diffusive chambers. Schematic representation of the diffusive chambers where the tissue (Strat-M® or pig skin) is placed for performing dynamic absorption studies. Panel A: the MIVO® fluidic chamber; Panel B: Franz Diffusion Cell. Table 1 Technical features of MIVO and FDC. Diffusive chamber MIVO FDC Flow regimes Laminar, monodirectional, spatially homogeneous Rotational, slight turbulent, spatially non homogeneous Reynolds number ∼20 <2000 ∼1700 <2000 Driving force Peristaltic pump Stirring bar Media Sampling system Through a three-way valve Through the sampling port Receptor Media volume Flexible, ranging from 2.5 to 5 ml From 5 to 20 ml according to the model Skin models Cells monolayer on inserts, membranes (eg StratM, Permeapad), 3D reconstructed skin tissues (e.g. epiDerm, epiSkin SkinEthic, Labskin), biopsies Membranes, biopsies Skin clamping Auto-centering and blocking of inserts, add-on with o-ring for biopsies Metallic clamps Sterilization Already sterile and ready to be used Autoclavable, after a pre-clean/wash FDC system The FDC system consists of a receiver compartment filled with 5 mL PBS, in which the compound is released after penetrating through the skin surrogate. Onto this surrogate the infinite-dose experimental condition (780 L/cm2 ) is topical applied into the donor compartment and allows the evaluation of penetration kinetics over time [33-35,44]. Figure 1B shows a schematic illustration of the FDC with its magnetic stirrer and its thermostatically controlled water bath, to maintain a controlled temperature of 32°C. The Strat-M® membranes and pig skin biopsy discs as human skin surrogates were placed onto the FDC (Ø 9 mm, diffusion area 0.64 cm2 ) from Logan Instruments Corp. (Somerset, USA). The Strat-M® membrane was placed shiny side up and the pig skin dermal side down onto the receiver compartment. The experiment with the Strat-M® membrane was performed with 6 replicates, while the one with pig ear skin was performed with 8 replicates. After an experimental equilibration for 30 min, different exposure times (1, 2, 4, 6, 8 hours) were adopted and 1000 L (or 20%) of the receiver compartment were collected and filtered through a cellulose acetate membrane filter (0.22 m pore size). The solution was HPLC analyzed and replaced with fresh PBS. A summary of the main features of MIVO and FDC was shown in Table 1. Chemicals Caffeine, 1,3-Benzodioxol-5-ylmethylurea (LIP1), Milli-Q® (water), acetonitrile, were purchased from Merck KGaA (Darmstadt, Germany). Propylene glycol (PG), Oleic acid (OA), and calcium and magnesium enriched PBS solution were purchased from Sigma Aldrich by Merck KGaA (St. Louis, USA). Formulations containing either water or PGOA (95:5) with either 0.7% caffeine (w/w) or LIP1 were prepared. Caffeine and LIP1 were selected as reference molecules having same molecular weight and different lipophilicity (Table 2). Skin models: Strat-M® and pig skin biopsies The Strat-M® membrane purchased from Merck KGaA (Darmstadt, Germany) is a non-animal based synthetic membrane build up by multiple layers of polyether sulfones coated with skin lipids, which mimics the penetration ability of human skin (Figure 2). This lipid coating, which mimics the intercellular lipid matrix of the human stratum corneum, contains a combination of ceramides, cholesterol, free fatty acids, and other components with a similar specific lipid ratio to human skin surface [4]. Pig ear skin was used as a human skin surrogate for penetration testing. The pig ears (German domestic pigs, 6-month-old) were obtained from a local slaughterhouse (Brensbach, Germany). Freshly slaughtered, they were cleaned with water, dried using soft tissue and stored at +4°C. The skin from the back of the pig ear was dermatomed with an electrical dermatome from Humeca BV (Borne, Netherlands). 500 μm thick splitskin punches with a diameter of 26 mm were obtained from each ear and stored for the maximum of 6 month at -20°C. HPLC The quantitative concentration of caffeine and LIP1 was determined using a HPLC (VWR-Hitachi ELITE LaChrom system) system. A Chromolith® Performance RP-18e 100-4.6 mm (Merck KGaA, Darmstadt) column was used as the stationary phase at 30°C and at flow rate of 2.0 163

I. Pulsoni, M. Lubda, M. Aiello et al. SLAS Technology 27 (2022) 161–171 Table 2 Physico-chemical properties of caffeine and LIP1. Caffeine 1,3-Benzodioxol-5-ylmethylurea (LIP1) Chemical structure Molecular Formula C8 H10 N4 O2 C9 H10 N2 O3 Molecular weight [g/mol] 194.19 194.19 Log P -0.1 (exp.) 0.6 (exp.) CAS number 58-08-2 65609-28-1 Figure 2. Strat-M® skin model. Schematic illustration of the skin morphology and the multilayered structure of the Strat-M® membrane to compare the different skin layers with a scanning electron microscopic image of a Strat-M® cross-section. mL/min, with an isocratic mobile phase of 90% water and 10% acetonitrile (HPLC gradient grade, Merck KGaA, Darmstadt) for caffeine and a mobile phase of 80:20 for LIP1. The caffeine amount was determined at a detection wavelength of 272 nm, and LIP1 at 285 nm, using a DAD l-2450 detection unit. Prior the analysis, the samples were mixed in auto sampler screw vials and analyzed with an injection volume of 60 μL of each sample (n=4). The specificity of the HPLC run was controlled via a blank injection and an internal standard solution. The quantification linearity was confirmed by a six-point calibration series (0.5-250.0 g/mL) with linear regression confirmation of R2> 0.99 in all cases. Accuracy and precision of the HPLC run was determined within the internal acceptance criteria variation <2 %. Statistical analysis Calculations were made using Microsoft® Excel® Office 365 and statistical analysis were performed using GraphPad Prism 8.03. All data sets are shown as the mean ± standard deviation (SD) with statistically significant differences determined by t-test (the Bonferroni method for multiple comparisons was also applied) with probability (p) values <0.05. Results Caffeine penetration kinetics using Strat-M® The cumulative amount of caffeine penetrated through Strat-M®was derived for the two experimental conditions (PGOA and Milli-Q® as donor solutions), showing statistical difference between the two diffusive chambers, although similar trend were observed (Figure 4). The passage of caffeine was enhanced (∼10X) when using PGOA as a vehicle (Figure 4A); moreover, the caffeine absorption through the StratM®membrane shows a linear trend during time in the Milli-Q® solution (Figure 4B), while in PGOA an initial plateau was observed after 6 hours. Caffeine penetration kinetics using pig skin biopsies The cumulative amount of caffeine penetrated through pig skin tissues was measured for both formulations (PGOA and Milli-Q®), compar164

I. Pulsoni, M. Lubda, M. Aiello et al. SLAS Technology 27 (2022) 161–171 Figure 3. Pig biopsy vs. Strat-M®. Image of a 500 μm thick pig split-skin biopsy (on the left) and the Strat-M® membrane (on the right). Figure 4. Cumulative amount of Caffeine penetrated through the Strat-M® membrane. Comparison of the caffeine penetration kinetics through Strat-M® membrane in FDC and MIVO® systems with PGOA (A) and Milli-Q® water (B) as vehicle solution. Values are presented as mean ±SD. ing the FDC and the MIVO® chambers (Figure 5) and showing statistical difference between the two diffusive systems only in MilliQ® conditions. The PGOA formulation (Figure 5A) led to an increasing amount of caffeine in the receiver chamber over time, despite the resulting penetration (ug/cm2 ) was overall lower than for Strat-M® (∼12X for FDC and ∼17X for MIVO®). Likewise, when caffeine was topically applied within the Milli-Q® vehicle (Figure 5B), its penetration over time through the pig skin was lower than the one resulted with Strat-M® membrane. LIP1 penetration kinetics using Strat-M® and pig skin biopsies Similarly to the caffeine in PGOA (Fig. 4A and Fig. 5A), the LIP1 in PGOA formulation displayed a gradual increase of its passage through Strat-M and pig skin, with an initial plateau observed at the ending points for Strat-M (Figure 6A) and a linear trend for pig skin (Figure 6B) for both diffusive systems. Like for caffeine, also LIP1 was more absorbed through the Strat-M than the pig skin (∼6X for FDC and ∼3X for MIVO®). The cumulative amount of LIP1 was statistically higher in MIVO than in FDC already after 2 hours using Strat-M and 6 hours using pig skin. Interestingly, values of caffeine in PGOA are much higher (∼4X after 8hr) than those of LIP1 using Strat-M in FDC system (Figure 6C), while no significant differences are observed between caffeine and LIP1 using pig skin in FDC system (Figure 6D). On the other side, as expected for lipophilic molecules applied onto skin tissues, MIVO displays a higher LIP1 permeation (∼3X after 8hr) than caffeine using pig skin model (Figure 6D), while artificial Strat-M membrane displayed similar permeation behavior for caffeine and LIP1 (Figure 6C). A comparison of the caffeine and LIP1 penetration for all implemented experimental conditions after 8 hours, with statistical analysis, was also shown (Figure 7). Mathematical approach to determine the diffusion coefficient Permeation of an active ingredient through the skin’s stratum corneum is described as diffusion process in which active transport plays no role [45]. Mathematically, skin absorption can be described by Fick’s laws of diffusion. The first Fick’s law is specific to an infinite dose condition [45], which is described for experiments with a topical amount higher than 100 μL/cm2 (or higher than 10 mg/cm2 ): = − ∕ (1) where J is the rate of transfer per unit area (flux) (g/cm2 /h), dC is the concentration gradient (g/cm3 ), dx is the linear distance travelled (cm) and D is the diffusion coefficient (cm2 /h). 165

I. Pulsoni, M. Lubda, M. Aiello et al. SLAS Technology 27 (2022) 161–171 Figure 5. Cumulative amount of Caffeine penetrated through pig skin. Comparison of the caffeine penetration kinetics through pig skin biopsies in FDC and MIVO® systems with PGOA (A) and Milli-Q® water (B) as vehicle solution. Values are presented as mean ±SD. Figure 6. Cumulative amount of LIP1 penetrated through the Strat-M® membrane and pig skin. Comparison of the LIP1 penetration kinetics through Strat-M® membranes (A) and pig skin biopsies (B) with FDC and MIVO® and in comparison with Caffeine penetration kinetics through Strat-M® membranes (C) and pig skin biopsies (D). Values are presented as mean ±SD, the values obtained by using the two systems are compared using paired t-Test statistics (p <0.05). Thereby a steady-state flux, J ss , is commonly assessed in vitro and ex vivo in diffusion cells (e.g., FDC and MIVO®), consisting of a donor compartment separated from the receiver compartment by a human skin surrogate. The ingredient is applied to the stratum corneum side of the skin, and an accumulation of active ingredient in the receiver compartment is monitored by repeated concentration measurements in the receiver medium over time. Under infinite dose conditions, dC can be replaced by the known donor concentration, c D , and the permeated mass per time is assumed to be constant. Therefore, a plot of the permeated mass per unit area versus unit time, yields a linear function with a slope which represents the steady-state flux. The apparent permeation coefficient, P app , which represents an independent measure of the skin resistance against permeation of the examined active ingredient, is frequently calculated as: = ∕ (2) 166

I. Pulsoni, M. Lubda, M. Aiello et al. SLAS Technology 27 (2022) 161–171 Figure 7. Caffeine and LIP1 permeation in all the experimental conditions. Quantity of caffeine penetrated after 8 hours through Strat-M® (A) and pig skin (B) with FDC and MIVO® and quantity of LIP1 through Strat-M® (C) and pig skin (D) with FDC and MIVO®. Values are presented as mean ±SD, the values obtained by using the two systems are compared using paired t-Test statistics (p <0.05). Table 3 The table summarizes the values of the steady-state flux and the apparent permeation coefficient for all the experimental conditions. Steady-state Flux J ss and Apparent Permeation Coefficient P app Skin Model Strat-M® Pig Skin FDC MIVO® FDC MIVO® LIP1 in PGOA J ss (μg/cm2 /h) 83.59 123.13 14,53 31.97 P app (cm/s X 10 −6 ) 3.32 4.88 0,63 1.27 Caffeine in PGOA J ss (μg/cm2 /h) 268.18 216.01 17.2 10.05 P app (cm/s X 10 −6 ) 10.64 8.57 0.68 0.40 Caffeine in Milli-Q® J ss (μg/cm2 /h) 20.94 16.51 2.54 4.75 P app (cm/s X 10 −6 ) 0.83 0.65 0.1 0.19 because this value only depends on the donor concentration used in the given experiment. J ss and P app were derived for the two skin models and the caffeine/LIP1 vehicle from the linear part of the respective absorption curves by linear regression, by using an automated approach [46], if at least four data points are within the linear part of a curve (R2 >0.92). The time to achieve steady-state conditions, under infinite dose conditions is referred as lag time (t) and the preceding period is the lagphase. Lag time is a function of the active ingredient loading the stratum corneum and dermis, diffusivity, and thickness of the skin. Lag time is the time required for the diffusion flow to become stable. By using the Lag time calculation [47], the diffusion coefficient can be derived: = 2 ∕6 . (3) Table 3 and Table 4 report respectively the steady state flux and the percentages of caffeine and LIP1 absorbed after 8 hours under different conditions. Table 4 Percentage values of penetrated caffeine and LIP1 through Strat-M®membranes and pig skin biopsies in all the experimental conditions. % of penetrated through the skin barrier Skin Model Strat-M® Pig Skin FDC MIVO® FDC MIVO® LIP1 in PGOA 11.6% 17.6% 2.5% 5.4% Caffeine in PGOA 36.2% 28.6% 3.1% 1.7% Caffeine in Milli-Q® 3.2% 2.6% 0.4% 0.8% Computational fluid dynamic simulation of fluid flow within MIVO and FDC Fluid dynamic simulations were performed both in MIVO and in FDC environments to simulate the fluid flow profiles, mean velocities 167

I. Pulsoni, M. Lubda, M. Aiello et al. SLAS Technology 27 (2022) 161–171 Figure 8. Computational Fluid-dynamic simulations of FDC and MIVO. Velocities distribution at the plane below the skin within FDC and MIVO (panel A). Velocity profiles at the cut line, compared to the physiological values, within FDC and MIVO (panel B). and shear stresses below the skin model (Figure 8). The 3D domain was drawn based on the dimensions of the two diffusive chambers used for skin permeation assays. The physical outputs were modeled using Comsol Multiphysics 5.6 (Laminar Fluid Flow module). The fluid was supposed to be laminar, incompressible, and not turbulent. The velocity profiles were calculated according to Navier-Stokes and the continuity equation [48,49]. In the MIVO environment, the flow rate was set to Q=2 mL/min to generate velocity resembling the capillary blood flow, while in the FDC the flow was generated by a stirring bar (length of 7 mm, rotational frequency 10Hz). An iterative geometric multigrid (GMRES) algorithm was used to solve the equations. A no-slip condition was fixed on the boundary of the geometry. Discussion In the fields of pharmaceutical industry, reliable skin penetration data of active ingredients are indispensable, since it affects the bioavailability, defined as the amount of molecules that reaches the systemic circulation [50]. Various studies have considered the impact of different physico-chemical formulations and skin models on permeation assays. Moreover, the use of in vitro diffusion cellss has evolved into a major research methodology, providing key insights towards more reliable, reproducible, and standardized in vitro and ex vivo methods. In vitro models frequently involve the use of artificial membranes to model realistic human skin penetration features. Although these membranes do not model the cellular mediated phenomena affecting the molecule passage through a living tissue, diffusion studies can be carried out. In particular, artificial membranes may be preferred to skin biopsies as they are more easily resourced, less expensive, structurally simpler and lead to a faster outcome [51]. Another challenging approach is based on in vitro human epidermis and dermis cells models, building human skin equivalent (HSE). Some commercially available HSE like Graftskin TM , SkinEthic TM , LAbskin TM , EpiDerm and Episkin, have been already adopted for penetration assays [52,53] and to understand metabolic skin response [13,31] within exploratory assays, while regulatory bodies are currently involved in guidelines updates., Also artificial membranes, although to a higher penetration ability and penetration rate [12,13,31], are currently adopted to provide useful permeation measurement for multiple formulations with higher throughput [4]. Indeed, they exhibit superior data reproducibility, as in vivo variables such as donor age, sex and anatomical site are excluded [51]. In this context, the Strat-M® represents an interesting option among synthetic skin models, since it is a membrane-based model with diffusion characteristics well-correlated to human skin [4]. However, it does not fully resemble the heterogeneous multi-layer structure of human skin. For these reasons, an effective alternative is represented by animal skin biopsies as indicated in the OECD TG 428 guidelines [32]. Among these, pig ear skin is currently the most widely used, given its histological similarities to human skin, with a comparable stratum corneum thickness [54,55]. In this work, a comparative study was performed by using both synthetic non-animal-based membrane (i.e., Strat-M®) and pig ear skin biopsies for assessing the penetration of two molecules with similar molecular weight but different lipophilicity: caffeine, a reference OECD 428 substance, and LIP1, a lipophilic test substance. By applying such testing molecules to the skin under infinite dose conditions, we expected that these penetrate into and diffuse through the stratum corneum, depending on their physico-chemical properties. [56,57] Indeed, a continuous increase of the caffeine and LIP1 penetration was observed up to 8 hours in all experimental conditions; in particular, Strat-M® turned out to be more permeable than pig biopsy, as already reported for synthetic membranes, especially for hydrophilic molecules [13]; this i can be due to a different morphological and histological structure of the skin tissue. Moreover, despite artificial skin membranes do not fully resemble the proper passive route of molecule permeation, a lower variability, and therefore a higher reproducibility, was observed for the StratM® model than pig ear biopses, as expected [31]. Although the same anatomical site (i.e., pig ear) was selected for all donors, this high variability is intrinsic to the model because of the biopsy structure, age and hydration state of the skin that may differ from one donor to another [45]. Transdermal pharmacokinetic studies allow to recognize the fate of the new formulations/drugs applied to the skin, to evaluate what fraction of the applied doses have been effectively absorbed, and also to determine the bioequivalence of the generic products [58]. To better resemble the in vivo situation, where systemic circulation rapidly clears permeants , diffusion chambers hosting skin models have been widely adopted as more reliable alternative to the static condition [59]. Among these, FDC systems are used since many years, under the two available configurations: static and flow-through [34]. In both systems the receptor fluid is stirred in a non-physiological way, but the pivotal difference between the two configurations is the continuous fluid replacement in 168