Test-strips for monitoring cancer biomarkers in point-of-care
Abstract
O cancro da mama é a forma de cancro mais prevalente, em mulheres, em todo o mundo, com crescente incidência e mortalidade. Estudos indicam que a sua deteção precoce pode potenciar o efeito da terapia e reduzir a mortalidade. No entanto, as atuais técnicas de diagnóstico apresentam desvantagens, uma vez que requerem equipamentos específicos e profissionais formados para execução da técnica e análise de resultados, sendo bastante dispendiosas e demoradas. Além disso, nos países de baixo rendimento os cuidados de saúde são de difícil acesso e, mesmo em países mais desenvolvidos, as técnicas de diagnóstico não são acessíveis para grupos socioeconómicos desfavorecidos, levando a atrasos na deteção e menores taxas de sobrevivência. Para ultrapassar este problema mundial, é urgente definir um método de deteção de baixo custo, rápido e com elevada sensibilidade e seletividade. Investigadores de todo o mundo têm desenvolvido sensores para deteção de biomarcadores do cancro, devido às suas vantagens, como o custo-benefício, portabilidade, e capacidade multiplex. O papel é uma opção de baixo custo como substrato para a construção de sensores, permitindo o desenvolvimento de um dispositivo de diagnóstico ideal, oferecendo uma resposta qualitativa e quantitativa para o utilizador final, seja o próprio paciente ou um profissional de saúde. Um resultado positivo pode encaminhar o paciente a realizar testes mais específicos para confirmação e identificação do estado da doença. Em combinação com substratos de papel, a deteção ótica é sensível, precisa e rápida. Esta tese focou-se no desenvolvimento de sensores colorimétricos em papel para biomarcadores do cancro. Foi desenvolvido um sensor colorimétrico enzimático para a glicose, como prova de conceito da combinação da deteção colorimétrica num suporte de papel. De seguida, foi produzido um imunossensor para o antigénio cancerígeno 15-3, em papel, e finalmente foi utilizado um polímero de impressão molecular para substituir o anticorpo previamente utilizado. Estes sensores permitiram detetar os alvos dentro dos valores de interesse clínico e demonstraram bom desempenho analítico em soluções tampão e amostras de soro, com uma seletividade aceitável contra interferentes. O conceito destes sensores é passível de ser aplicado a outros biomarcadores do cancro e usado em dispositivos portáteis e de baixo custo especialmente benéficos em países mais desfavorecidos.
Full text
Mariana Conde Carvalho Gonçalves Carneiro Test-strips for monitoring cancer biomarkers in point-of-care April 2024 Test-strips for monitoring cancer biomarkers in point-of-care Mariana Conde Carvalho Gonçalves Carneiro UMinho|2024 Universidade do Minho Escola de Engenharia Universidade do Minho Escola de Engenharia
Mariana Conde Carvalho Gonçalves Carneiro Test-strips for monitoring cancer biomarkers in point-of-care April 2024 Doctoral Thesis Doctoral Program in Chemical and Biological Engineering Work developed under supervision of Professor Maria Goreti Ferreira Sales and Professor Lígia Raquel Marona Rodrigues Universidade do Minho Escola de Engenharia
ii DIREITOS DE AUTOR E CONDIÇÕES DE UTILIZAÇÃO DO TRABALHO POR TERCEIROS Este é um trabalho académico que pode ser utilizado por terceiros desde que respeitadas as regras e boas práticas internacionalmente aceites, no que concerne aos direitos de autor e direitos conexos. Assim, o presente trabalho pode ser utilizado nos termos previstos na licença abaixo indicada. Caso o utilizador necessite de permissão para poder fazer um uso do trabalho em condições não previstas no licenciamento indicado, deverá contactar o autor, através do RepositóriUM da Universidade do Minho. Atribuição-NãoComercialCompartilhaIgual CC BYNC-SA https://creativecommons.org/licenses/by-nc-sa/4.0/
iii Acknowledgments I want to acknowledge to my supervisors Prof. Goreti Sales and Prof. Lígia Rodrigues for their support, encouragement, scientific knowledge, and guidance throughout this work. For their insightful comments and suggestions that made me grow as a scientist. I would like to give my warmest thanks to my co-supervisor, Felismina Moreira for her guidance and advice that carried me through all stage of laboratory work and writing my thesis. For her continuous support and belief in me. I would also like to give special thanks to all my colleagues at Biomark Sensor Research group, for helping me, for sharing with me their scientific knowledge and for the good times together. For those who become friends for life, I would like to thank for continuous encouragement and friendship. I am deeply grateful to my parents, Nazaré and Alberto, for their continuous support and understanding along these years and for letting me through all the difficulties. I would not could not have undertaken this journey without the endlessly enthusiasm of my boyfriend, Patrick Pais, that supported me and give me strength to not give up, comforting me at the end of each hard day. To my baby daughter, Beatriz, for being so lovely and providing me decent night’s sleep, keeping me sane in this final process of writing the thesis. Finally, I want to acknowledge Portuguese Foundation for Science and Technology (FCT) for funding this work through the PhD grant reference SFRH/BD/131959/2017.
iv STATEMENT OF INTEGRITY I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Code of Ethical Conduct of the University of Minho.
v Resumo Tiras de teste para monitorizar biomarcadores do cancro em point-of-care O cancro da mama é a forma de cancro mais prevalente, em mulheres, em todo o mundo, com crescente incidência e mortalidade. Estudos indicam que a sua deteção precoce pode potenciar o efeito da terapia e reduzir a mortalidade. No entanto, as atuais técnicas de diagnóstico apresentam desvantagens, uma vez que requerem equipamentos específicos e profissionais formados para execução da técnica e análise de resultados, sendo bastante dispendiosas e demoradas. Além disso, nos países de baixo rendimento os cuidados de saúde são de difícil acesso e, mesmo em países mais desenvolvidos, as técnicas de diagnóstico não são acessíveis para grupos socioeconómicos desfavorecidos, levando a atrasos na deteção e menores taxas de sobrevivência. Para ultrapassar este problema mundial, é urgente definir um método de deteção de baixo custo, rápido e com elevada sensibilidade e seletividade. Investigadores de todo o mundo têm desenvolvido sensores para deteção de biomarcadores do cancro, devido às suas vantagens, como o custo-benefício, portabilidade, e capacidade multiplex . O papel é uma opção de baixo custo como substrato para a construção de sensores, permitindo o desenvolvimento de um dispositivo de diagnóstico ideal, oferecendo uma resposta qualitativa e quantitativa para o utilizador final, seja o próprio paciente ou um profissional de saúde. Um resultado positivo pode encaminhar o paciente a realizar testes mais específicos para confirmação e identificação do estado da doença. Em combinação com substratos de papel, a deteção ótica é sensível, precisa e rápida. Esta tese focou-se no desenvolvimento de sensores colorimétricos em papel para biomarcadores do cancro. Foi desenvolvido um sensor colorimétrico enzimático para a glicose, como prova de conceito da combinação da deteção colorimétrica num suporte de papel. De seguida, foi produzido um imunossensor para o antigénio cancerígeno 15-3, em papel, e finalmente foi utilizado um polímero de impressão molecular para substituir o anticorpo previamente utilizado. Estes sensores permitiram detetar os alvos dentro dos valores de interesse clínico e demonstraram bom desempenho analítico em soluções tampão e amostras de soro, com uma seletividade aceitável contra interferentes. O conceito destes sensores é passível de ser aplicado a outros biomarcadores do cancro e usado em dispositivos portáteis e de baixo custo especialmente benéficos em países mais desfavorecidos. Palavras-chave: cancro da mama, colorimetria, immunosensor, polímeros de impressão molecular, sensores em papel.
vi Abstract Test-strips for monitoring cancer biomarkers in point-of-care Breast cancer is the most common cancer in women worldwide, with increasing incidence and mortality rates. Studies suggest that early cancer diagnosis can improve the effectiveness of therapy and thus reduce mortality rates. However, current diagnostic techniques for detecting breast cancer, have shortcomings. They require sophisticated equipment and trained professionals to perform the technique and analyse the results, which is expensive and time-consuming. Despite this, lowand middle-income countries do not have easy access to healthcare facilities, and even in developed countries, some diagnostic techniques are not affordable for vulnerable populations, resulting in delayed diagnosis and lower survival rates. To solve this problem globally, low-cost, rapid, and highly sensitive and selective methods for early diagnosis are urgently needed. In recent years, researchers worldwide have made efforts to develop sensors for the detection of cancer biomarkers, as they offer several advantages over existing diagnostic methods, such as cost-effectiveness, portability, and multiplexing capability. Paper is a cost-effective option for use as a substrate for sensor construction. Paper-based sensors fulfil the requirements for an ideal screening device as they can provide a qualitative or quantitative response and the end user can be either the patient themselves or a trained or untrained healthcare professional. If the result of the test is positive, the patient can see a doctor who will perform a full screening protocol to determine the stage of the disease. In conjunction with paperbased substrates, the optical detection techniques are sensitive, accurate, rapid, and non-invasive. The aim of this work was to develop colorimetric paper-based sensors for cancer biomarkers. For this purpose, an enzymatic colorimetric sensor for glucose was developed, which served as proof of concept for the conjugation of colorimetric detection onto a paper substrate. Subsequently, an immunosensor was assembled onto the paper substrate for cancer antigen 15-3 detection and finally a synthetic recognition element, a molecularly imprinted polymer, was used to replace the antibody in the previous device. The developed sensors work in the clinical range of the selected targets and showed good performance in both buffer solutions and serum samples, as well as acceptable selectivity towards interfering species that might be present in more complex samples. In addition, the concept of the developed sensors can be extended to other cancer biomarkers and used as point-of-care devices in resource-poor settings. Keywords: breast cancer, colorimetry, immunosensor, molecular imprinting polymers, paper-based sensors.
vii Index RESUMO ............................................................................................................................................. V ABSTRACT ......................................................................................................................................... VI INDEX ............................................................................................................................................... VII ABBREVIATIONS AND SYMBOLS ......................................................................................................... XI LIST OF FIGURES .............................................................................................................................. XV LIST OF TABLES ............................................................................................................................. XVIII CHAPTER 1 ........................................................................................................................................ 1 1. INTRODUCTION .......................................................................................................................... 2 1.1. Motivation ............................................................................................................................... 2 1.2. Structure of this thesis ............................................................................................................ 3 1.3. List of publications .................................................................................................................. 4 1.3.1. Papers published in international scientific journals (included in the thesis) ...................... 4 1.3.2. Papers published in international scientific journals (not included in the thesis) ................. 4 1.3.3. Presentations at scientific conferences ............................................................................. 5 CHAPTER 2 ........................................................................................................................................ 6 2. STATE OF THE ART .................................................................................................................... 7 2.1. Cancer overview ...................................................................................................................... 7 2.2. Cancer diagnosis .................................................................................................................. 10 2.2.1. Cancer biomarkers detection ......................................................................................... 11 2.3. Breast cancer overview ......................................................................................................... 12 2.3.1. Breast cancer diagnosis ................................................................................................. 15 2.4. Biosensors ............................................................................................................................ 20 2.4.1. Recognition elements .................................................................................................... 22 2.4.1.1. Enzymes ................................................................................................................ 23 2.4.1.2. Antibodies ............................................................................................................. 24 2.4.1.3. Molecular imprinting polymers ............................................................................... 29 2.4.2. Transducers .................................................................................................................. 34 2.4.2.1. Electrochemical biosensors .................................................................................... 34 2.4.2.2. Optical biosensors ................................................................................................. 34 2.4.2.2.1. Fluorescence ...................................................................................................... 35 2.4.2.2.2. Surface plasmon resonance ................................................................................ 36 2.4.2.2.3. Surface-enhanced Raman scattering ................................................................... 36
xiv TP53 tumor protein 53 TRIM trimethylopropane trimethacrylate U uPA urokinase plasminogen activator UV ultraviolet UV/Vis ultraviolet-visible V VEGF vascular endothelial growth factor W WHO world health organization
xv List of Figures CHAPTER 2 Figure 2-1. The hallmarks of cancer. Taken from (32). 7 Figure 2-2. Estimated number of new cases of cancer in 2020. Taken from (36). 8 Figure 2-3. Estimated number of new cases of cancer, in Portugal, in 2020. Taken from (36). 9 Figure 2-4. Structure of MUC1 in normal (A) and cancer cells (B). Adapted from (57). 18 Figure 2-5. Schematic representation of a biosensor. Analytes present on the sample bind to RE of the biosensor, producing a change which is converted into a quantifiable signal by the transducer. Signal is shown by the display system. 21 Figure 2-6. Example of types of samples and biomarkers that can be detected in biosensors. REs and transduction systems used in biosensors assembly. 21 Figure 2-7. Schematic representation of random (side-one, tail-on, head-on and flat-on) and oriented immobilisation of antibodies. Adapted from (82). 27 Figure 2-8. Steps of MIP production. Taken from (105). 30 Figure 2-9. Catalytic oxidation of TMB by HRP. 39 Figure 2-10. Molecular structure of cellulose. Adapted from (157). 41 Figure 2-11. Examples of different formats of PADs. Photograph of dipsticks (A) for urine. Adapted from (163); Spot test (B): photographs (a) and bar charts (b) of grey values for alpha fetoprotein detection. Taken from (164); LFA (C): schematic representation (a) and photograph of a LFA for microRNA-125 detection. Adapted from (165); photograph of a µPAD (D) for microRNA-21 detection. Adapted from (166). 44 Figure 2-12. Schematic representation of MIP-PADs fabrication: (A) in situ polymerisation and (B) postintroduction. Taken from (104). 48 Figure 2-13. Examples of: (A) types of samples and biomarkers, (B) types of paper, (C) REs and labels, and (D) signal readouts used in colorimetric paper sensors. Taken from (181). 50 Figure 2-14. ELISA formats. Direct ELISA (A), indirect ELISA (B), sandwich ELISA (C) and competitive ELISA (D). Taken from (77). 56 Figure 2-15. Number of papers describing colorimetric PADs (purple bar) and colorimetric PADs for cancer biomarkers detection (orange bar), published on last decade (181). 59 CHAPTER 3 Figure 3-1. Synthesis of the carboxyl-NC by TEMPO oxidation of MCC. 65 Figure 3-2. Test-strip based colorimetric assay produced by casting on the cellulose substrates the indicated solutions, and binding GOx either by adsorption (A) or by covalent bonding (B). 68 Figure 3-3. FTIR Spectra of pristine MCC (top) and NC-COOH (bottom), signalling the wavenumbers areas with the major differences between these (in pink). 70 Figure 3-4. Conductometric tirtation curve of NC-COOH. 72 Figure 3-5. TEM images of carboxyl-NC material obtained after TEMPO oxidation of MCC. 72 Figure 3-6. Digital images of the calibration curves of glucose using GOx adsorbed to the test-strip. 74 Figure 3-7. Dependence of concentration of GOx and an amount of glucose with the time. 75 Figure 3-8. Digital images of the test-strips in the presence of different glucose concentrations using ABTS as colorimetric indicator, prepared with different concentrations of carboxyl-NC (left) and the analytical calibration curves plotting the color coordinates collected against the glucose concentration (right, the carboxyl-NC assay corresponds to 5 mg mL -1 of NC-COOH). 76 Figure 3-9. Digital images of the test-strips in the presence of different glucose concentrations using ABTS as colorimetric indicator, prepared with different concentrations of GOx covalently bound (left) and the
xvi analytical calibration curves plotting the color coordinates collected against the glucose concentration (right, the carboxyl-NC assay corresponds to 1 mg mL -1 GOx). 77 Figure 3-10. Evaluation of interfering species variation in comparison with glucose. 78 CHAPTER 4 Figure 4-1. Outside (A) and inside (B) view of the dark box for image acquisition. Steps of the sensor construction (C). Covalent immobilisation of capture antibody (a), blocking step with BSA (b), incubation with CA15-3 antigen (c), incubation of detection antibody labelled with HRP (d) and color development with TMB solution (e). 82 Figure 4-2. Characterisation of paper substrate before and after washing treatment and functionalisation. FTIR spectra (A) and zoom at 1720 cm-1 (B); TGA (C) and DTG (D) spectra; pictures of paper test-strip after Bradford assay in APTES-modified or KIO4-modified paper incubated with buffer vs BSA (E). HSB coordinates of the several modifications (F). In (C) and (D). 88 Figure 4-3. Photographs of paper circles with different volumes of food coloring for reagent volume optimisation 89 Figure 4-4. Photographs (A) and respective bar charts (B) of saturation values from optimisation of the composition of washing buffer (PB vs PB with Tween20). 90 Figure 4-5. Photographs (A) and respective bar charts (B) of saturation values from capture antibody concentration optimisation with 1.61, 16.1 and 161 μg mL -1. 91 Figure 4-6. Photographs (A) and respective bar charts (B) of saturation values from detection antibody concentration optimisation with 0.5, 5 and 50-μg mL -1. 91 Figure 4-7. Photographs (A) and respective bar charts (B) of saturation values from blocking agent concentration optimisation with 0, 0.1, 1 and 10% of BSA. 92 Figure 4-8. Photographs (A) and respective bar charts (B) of saturation values from number of washing steps optimisation, including no washing, one washing step with 1000-μL of PB and three washing steps with 1000-μL of PB. 93 Figure 4-9. Photographs (A) and respective bar charts (B) of saturation values from incubation and dry temperature optimisation with 21 ºC and 37 ºC. 94 Figure 4-10. Photographs (A) and respective bar charts (B) of saturation values from papers without functionalisation and with KIO4 functionalisation. 95 Figure 4-11. Pictures of the colorimetric sensor showing the color change with increasing concentrations of CA15-3 (from 2 to 1100 U mL -1) prepared in PB, after HRP reaction with TMB (A) and respective calibration curve of hue values extracted from the photographs vs logarithmic concentration (B). Calibration curve of the sensor incubated with a 100-fold dilution of Cormay® serum spiked with CA15-3 (from 2 to 200 U mL -1) (C). Equations from calibration curve are shown. 96 Figure 4-12. Selectivity study based on comparison of the response of CA15-3 incubated alone or mixed with interfering species. Photographs (A) and respective bar chart (B) with mean ± SD. 97 Figure 4-13. Photographs (A) and calibration curve (B) of the effect of sensor functionalisation with carboxy-NC 98 CHAPTER 5 Figure 5-1. Schematic representation of the sensor construction. Production of a first layer of polymerisation with 3-APBA (A), production of MIP (B), removal step (C), rebinding step (D), HRP incubation (E), and color development with TMB solution (F). 103 Figure 5-2. Thermogravimetric analysis of cellulose paper with or without silane modification (A) and the cellulose paper with silane modified with MIP/NIP polymerisation, including stages before and after
xvii template removal (B). 107 Figure 5-3. SEM images of bare paper (A), NIP (B) and MIP (C) with 500× magnification and bare paper (D), NIP (E) and MIP (F) with 15000× magnification. 108 Figure 5-4. Photographs of NIPs and MIPs with different removal solutions (A). Respective bar charts with Q values for removal with water (B), sodium chloride (C), acetic acid + sodium dodecyl sulfate (D) and acetic acid + tween-20 (E). 110 Figure 5-5. Photographs (A) of NIPs and MIPs with PB, antibody-HRP or HRP. Respective bar charts with Q values for PB (B), antibody-HRP 5 µg mL -1 (C), HRP 5 µg mL -1 (D) and HRP 50 µg mL -1 (E). 112 Figure 5-6. Photographs (A) of NIPs and MIPs with 3, 4 or 5-µL of standard solution and HRP solution. Calibration curves with Q values of MIPs with 3 (B), 4 (C) or 5 (D) µL of standard and HRP solutions. 113 Figure 5-7. Photographs of buffer calibration in MIP (A). Calibration curve based on grey intensity for NIP (B) and MIP (C). Calibration curve based on quadrature values for NIP (D) and MIP (E). 114 Figure 5-8. Images and the respective bar chart of selectivity study based on comparison of the response of CA15-3 (250 U mL-1) incubated alone or mixed with interfering species as CEA (0.25 ng mL-1), CA125 (0.35 U mL-1), glucose (1 mg mL-1) and IgG (0.1 mg mL-1). 115 Figure 5-9. Photographs of FBS calibration in MIP (A). Calibration curve based on grey intensity for NIP (B) and MIP (C). Calibration curve based on quadrature values for NIP (D) and MIP (E). 116
xviii List of Tables CHAPTER 2 Table 2-1. Receptors expression, incidence, and prognosis of different types of BC. 13 CHAPTER 3 Table 3-1. The effects of amount and adding of sodium hypochlorite on weight yield of functionalised CNC. 70 CHAPTER 4 Table 4-1. Reagent’s volume and reaction time for the detection of CA15-3 using the reported paper-based sensor. 99 CHAPTER 5 Table 5-1. Published works for molecular imprinting polymers combined with colorimetric detection on a paper substrate. 118
xix “Nothing in life is to be feared, it is only to be understood. Now is the time to understand more, so that we may fear less.” Marie Curie
Introduction 1 Chapter 1 Introduction In this chapter, the context and motivation of the developed work is given. The research aims are listed, and the structure and framework of the thesis are described, Finally, the outcomes of the thesis are enounced, including published papers and related presentations.
Introduction 2 1. Introduction 1.1. Motivation Cancer is a major cause of mortality and morbidity worldwide (1). Breast, lung, colorectum, prostate, skin and stomach were the most common cancer types all over the world in 2020 and 60.467 new cases of cancer occurred in Portugal (2). Conventional methodologies for cancer diagnosis require significant time, human and economic resources, as they involve complex procedures and need to be performed at hospitals, by trained technicians, not being available in a point-of-care (PoC) format (3). These drawbacks often result in late diagnosis, occurring at an advanced stage of the disease. Thus, it is urgent to find ways for an early and proper cancer diagnosis, as well as for a suitable treatment. Actually, between 30% and 50% of cancer deaths could be prevented by an early diagnosis, accordingly to World Health Organization (WHO) (4). Searching for specific biomarkers that can be related with a certain cancer type and the development of methodologies for its fast and low-cost detection are two topics that are currently under the attention of researchers. PoC devices are suitable for screening protocols as they enable testing patients in a convenient way, providing useful and timely information to further carry out more accurate diagnosis (5). Early cancer diagnosis led to an increase in survival rates, an enhancement of quality of patient’s life and thus an improvement of disease overall outcome (6). These devices should be low-cost, small, and portable, and easy to use. Biosensors fulfil the requirements of a PoC device and can be a powerful tool for cancer monitoring. Up to date some biosensors for cancer biomarkers have been developed (7-9). Several types of recognition elements (REs) are available for cancer biosensor’s design, including antigens or antibodies (10, 11), enzymes (12), nucleic acids (13, 14) or aptamers (15, 16) . Most of the published works for cancer monitoring use antibodies as REs (11, 17-21). Despite the well-established role of antibodies as REs and their undoubtedly selectivity for their targets (22), antibodies present some disadvantages, as its production is expensive and present low stability (23). For that reason, researchers are searching for new REs as molecular imprinting polymers (MIPs) or aptamers (1, 24). MIPs are artificial receptors, capable of mimicking the biological recognition. They are highly specific for their templates due to the formation of specific cavities and interaction of their binding sites with the desired template and can be produced over a countless number of targets. They can be applied as REs in biosensors coupled with several transduction systems thus providing sensitive devices.
Introduction 3 Besides that, they offer stability, robustness and resistance to external conditions and its production process is reproducible and cost-effective, which offer advantages over antibodies (25). Despite the interest of new REs, a proper transduction system is also a concern. Electrochemistry is the most often transduction system applied to cancer biosensors, especially immunosensors, which require an energy source thus compromising the portability of the device. Optical transduction provides a visible response and eradicate the need of specific equipment to read the results, while allowing the sensor to be portable (22). Among optical transducers, colorimetric ones enable biomarker’s detection and quantification through color changes that could be observed by naked eye, both in solution or solid supports as paper (26). Paper is a suitable and versatile substrate due to its chemical and physical properties and the interest in paper analytical devices (PADs) has grown exponentially in the last decade, especially when conjugated with portable readers as smartphones, thus meeting the criteria for PoC devices (27-29). Hence, the motivation of this plan is to develop novel biosensors for properly but simple detection of cancer biomarkers that can be used as PoC devices for cancer screening. Cancer antigen 15-3 (CA153) was selected as breast cancer (BC) biomarker and the goal is to use alternative REs to the wellestablished antibodies, and a transduction system compatible with a portable device, as optical transduction. Paper was used as support material for the development of the biosensors. Firstly, as a proof-of-concept, an enzymatic colorimetric sensor was developed for glucose, then an immunosensor was assembled onto the paper substrate for CA15-3 detection and finally a RE from synthetic nature, a MIP, was used to replace the antibody in the previous device. The specific objectives defined for this purpose include: (1) selection of suitable substrate for sensor assembly; (2) production of synthetic REs (plastic antibodies or aptamers); (3) development of a biosensor using the selected REs; (4) and signal transduction to generate analytical data. 1.2. Structure of this thesis This thesis reports the development of colorimetric paper-based sensors for biomolecules, and it is organized in 5 chapters: Chapter I: Gives the context and motivation of the work developed. It lists the research aims and describes its structure and framework, as well as the publication outcomes and related presentations.
Introduction 4 Chapter II: Provides a state of the art about cancer, focusing on cancer diagnosis based on biomarkers detection. BC is highlighted, once BC biomarkers were selected as targets for the sensors reported in this thesis. Finally, biosensors as an innovative approach for cancer biomarkers detection are described. The following chapters from III to V report the development and characterisation of paper-based colorimetric biosensors for cancer biomarkers with different REs. Chapter III: Presents the development of a sensor in which a well-known RE, an enzyme, was applied to detect a general biomarker, glucose, in a colorimetric paper-based enzymatic sensor. Chapter IV: Describes the use of another well-known RE, an antibody, to detect a specific cancer biomarker, CA15-3, in a colorimetric paper-based immunosensor. Chapter V: Focuses on the development of an artificial RE, a MIP, to assembly a biomimetic paperbased colorimetric biosensor for CA15-3. Finally, chapter VI gathers the general conclusions of this thesis and its future perspectives. 1.3. List of publications 1.3.1. Papers published in international scientific journals (included in the thesis) ▪ Neubauerova, K., Carneiro, M. C. C. G., Rodrigues, L. R., Moreira, F. T. C., & Sales, M. G. F. (2020). Nanocellulosebased biosensor for colorimetric detection of glucose. Sensing and Bio-Sensing Research, 29, 100368. doi:10.1016/j.sbsr.2020.100368; ▪ C.C.G. Carneiro, M., Rodrigues, L. R., Moreira, F. T. C., & Goreti F. Sales, M. (2022). Paper-based ELISA for fast CA15–3 detection in point-of-care. Microchemical Journal, 181, 107756. doi:10.1016/j.microc.2022.107756; ▪ C.C.G. Carneiro, M., Rodrigues, L. R., Moreira, F., & Goreti F. Sales, M.. Paper-based biommimetic test-strip for CA15-3 with colored readout. Microchemical Journal. 2024;196:109640. 1.3.2. Papers published in international scientific journals (not included in the thesis) ▪ Carneiro, M., Rodrigues, LR., Moreira, F. T. C., & Sales, M. G. F. (2022). Colorimetric Paper-Based Sensors against Cancer Biomarkers. Sensors, 22(3221).
State of the art 11 When a person is suspected to have cancer, either by a screening test result or due to experienced symptoms, diagnostic methodologies should be carried out. Cancer diagnosis can be performed either through conventional techniques (as image methodologies like magnetic resonance imaging (MRI), endoscopy, computing tomography, X-rays, ultrasound imaging, mass spectroscopy and biopsy) or by cancer biomarkers detection (usually performed by immunoassays as enzyme-linked immunosorbent assay (ELISA), radioimmunoassay (RIA) or fluorescence immunoassay (FIA), or by electrophoresis, polymerase chain reaction (PCR) and mass spectrometry (MS)). Among these, immunoassays and PCR are the gold standard in clinical diagnosis for biomarker detection as they are highly sensitive and specific (3, 6, 8, 9, 17, 40). However, these clinical tests can be invasive, expensive, and timeconsuming for patients and healthcare systems as they imply complex and high-cost protocols with multiple analytical steps as wash steps and incubations, large consumption of reagents and samples, sophisticated equipment and are not available in some regions with lack of economic resources or trained personnel, compromising the accurate diagnosis and further effective treatment. For these reasons, these procedures are impractical in developing countries and places with low resources (1, 3, 5-7, 9, 24, 25, 40-43). Therefore, researchers keep looking for alternative and sensitive methodologies for cancer screening and diagnosis, focusing on functional images and biomarkers detection (32). 2.2.1. Cancer biomarkers detection Biomarkers are molecules responsible for several functions in our organism, including storage and transmission of genetic information, catalytic activity, regulation of biological activities or transport, and should be capable to differentiate normal and disease stages. The ideal biomarker should be specific for the disease of interest; do not exist in healthy people; and lead to the identification of the disease before the clinical diagnostic (44-46). Cancer biomarkers represent a very large group of molecules as genes, nucleic acid sequences (deoxyribonucleic acid (DNA), ribonucleic acid (RNA) and microRNA (miRNA)), proteins/enzymes/hormones, lipids, small molecules as secondary metabolites, extracellular vesicles or circulating tumor cells and changes can occur in terms of structure, expression, function or abundance (26). Several molecules have been identified and characterised as cancer biomarkers, being currently in clinical use (44-46). These molecules could be present in tissues as tumor tissues or in body fluids such as blood, urine or oral fluids (1, 3, 6, 17, 41). Biomarkers that could be detected and measured in patients’
State of the art 12 (bio)fluids must have a special attention by researchers once they can be easily collected by minimally invasive procedures, being an advantage over conventional methodologies (24, 44-46) as they enable cancer detection through less invasive processes (17, 32). They have high clinical significance once its study enables to understand more about the risk and progression of the disease and to establish a diagnosis and prognosis (and subsequently, timely and adequate treatments), as well as detect disease recurrence. They can also be useful on the assessment of patients’ response to therapy, for monitoring and prediction about drug resistance, as well as development of new treatments (3, 5-7, 9, 17, 37, 40-42, 47, 48). Despite the undoubted advantages of measuring biomarkers for cancer detection, most of the discovered biomarkers suffer from lack of sensitivity or selectivity and the use of a new biomarker in clinical diagnosis is a complex process, involving time and money consuming, once it requires analytical and clinical validation and subsequent evidence of the biomarker’s clinical value (22). The detection of biomarkers could also be confusing, as some of these molecules are released in low concentrations, which is very common in cancer, especially in the early stages of the disease (32, 41, 43). Moreover, some of them are associated with only one type of cancer, whereas other ones are related to several types of cancer. For these reasons, the detection of a biomarker in complex samples could be difficult and the detection of a single cancer biomarker may not be indicative of the disease, so a panel of biomarkers are often required for an accurate diagnosis (3, 6, 41, 43). Also, considering the organ involved, cancer could be classified in many sub-groups and a specific biomarker can show different levels in different type of cancer (6). For these reasons, actual detection of cancer biomarkers is only performed as a complement to conventional methodology (22). 2.3. Breast cancer overview BC is the most common type of cancer, accounting for 12.5% of new cancer cases worldwide (49). It is also the major cause of death in women, registering 684.996 deaths in 2020 (38). The majority of BC types are detected in lowand middle-income countries, especially at later stages due to poor health accesses (32). At present, about 80% of BC cases have more than 50 years old and their survival depends on the stage of the cancer at the moment of detection, as well as its molecular subtype (38). BC displays high heterogenicity regarding tumor morphology, molecular features and clinical response (32). BC subtypes can be described accordingly to different models of classification. Concerning the
State of the art 13 origin of the tumor and its cellular behaviour, BC can be classified into invasive or non-invasive. A BC case classified as non-invasive, means that it does not spread to the adjacent breast tissue while invasive BC, as the name suggest, spread into the surrounding breast tissue. The most common type of non-invasive BC is ductal carcinoma in situ (DCIS) and corresponds to 16% of all BCs. Lobular carcinoma in situ (LCIS) is other type of non-invasive BC and it is considered a benign BC condition and does not spread outside the lobules. Invasive ductal carcinoma (IDC) represents 80% of the cases, being the most common type of invasive BC while invasive lobular carcinoma (ILC) is the second most common type of BC, with 10% of the cases. Inflammatory BC (IBC) is a less common type of invasive BC, accounting only 1-3% of the cases. Paget’s disease of the nipple starts around the nipple and can spread to the areola and other areas of the breast. It counts with less than 5% of BC cases (22, 49). The most common and widely accepted classification is based on a five subtype classification model that takes into consideration the expression of hormone receptors and result in five subtypes (luminal A, luminal B, luminal B-like, human epidermal growth factor receptor 2 (HER2) positive, non-luminal, and triple negative BC (TNBC) whose receptor expression, incidence and prognosis is summarized in Table 2-1 (22). Table 2-1. Receptors expression, incidence, and prognosis of different types of BC. Types of BC Incidence Receptors expression Prognosis PR Estrogen alpha receptor (ERα) HER2 Luminal Luminal A 70% + + - Good Luminal B + + - Medium Luminal B-like + + + Poor Nonluminal HER2 positive 10-15% - - + Poor Basal TNBC 20% - - - Poor BC ethology is a result of a complex interaction between several modifiable and non-modifiable factors that contribute to the development of the disease. Non-modifiable factors include sex, age, genetic mutations, family history, race/ethnicity and reproductive factors, while modifiable factors are related with physical activity, body mass index, dietary and some behaviours as alcohol consumption and tobacco use, as well as environmental factors (22, 32, 38).
State of the art 14 Female sex is the major factor related with increased risk of BC, due to several reproductive and hormonal factors (38) as lower parity, earlier ages at menarche, later ages of menopause and later ages at first birth (22, 32, 33). In men, BC is a rare disease that counts with less than 1% of cases and is usually detected at advanced stages of the disease (38). Cancer has a higher occurrence in older individuals, due to accumulation of several cellular changes as well as exposition to potential carcinogenic agents (38). Personal history of BC and other non-cancerous alterations in breasts is a significant risk factor for BC. It is known that about 13 to 19% of BC patients have a first-degree relative with cancer (38). Several genetic mutations are highly related with an augmented risk of BC, especially the ones responsible for high penetration, occurring in breast cancer gene 1 (BRCA1) and breast cancer gene 2 (BRCA2), but also in tumor protein p53 (TP53), cadherin-1 (CDH1), phosphatase and tensin homolog gene (PTEN), and serin/threonine kinase 11 gene (STK11). These mutations are principally transmitted by autosomal dominant inheritance, but some sporadic mutations have been reported (38). Inequalities regarding race and ethnicity are observed and BC incidence is higher among white non-hispanic women, while higher rates of mortality and lower survival rates are observed among black women (38). Some reports indicate that some drugs (e.g. hormonal therapy, antibiotics and antidepressants) intake are related with higher risk of BC, as well as sedentary behaviour, high alcohol consumption (38), tobacco use, consumption of processed food , saturated fats, food rich in sodium or sugar. Some chemicals (e.g. polycyclic aromatic hydrocarbons, organic solvents, and insecticides) can induce epigenetic modifications and thus induce pro-carcinogen events (38). Tumor size, tumor histologic grade and hormone receptors are traditional prognostic factors in BC, being axillary lymph node statues the most important one (50, 51). However, these can only be obtained from tissue samples, which is an invasive method. Some circulating tumor markers found in serum have been investigated in several studies as potential prognostic parameters for patient outcome and response to therapy. Examples include CA15-3 and carcinoembryonic antigen (CEA), which can be easily quantified in serum by a rapid and non-invasive test. High levels of these molecules have been associated in several reports with poorer prognosis and disease-free survival in patients with BC. Patients with normal CA15-3 and CEA levels had better disease-free survival than patients with elevated CA15-3 and/or CEA levels. The association between elevated CA15-3 and CEA levels and such major parameters as tumor size, as well as node metastases and an advanced stage of the disease was also verified (50, 52). This topic will be fully discussed in section 2.3.1. Breast cancer diagnosis.
State of the art 15 Breast cancerous tissues can be surgical removed by partial mastectomy or complete mastectomy (38). Systemic treatment of BC involves preoperative or postoperative chemotherapy in which the selection of the proper drug is of main importance as different molecular BC subtypes have different response to therapy. Local treatment for BC comprises radiotherapy, usually performed after surgery and/or chemotherapy, being especially useful in the case of metastatic BC (38). 2.3.1. Breast cancer diagnosis Alongside cervical and colorectal cancer, BC is one of the cancers for which the European Council recommends screening. Screening tests for BC are essentially based on mammography for women aged 50-69 years, with a recommended screening interval of 2 years and breast examination (both clinical and self-examination) (32). Mammography is currently the golden standard methodology for BC screening, but it has some limitations as high rates of false positive results (22). Moreover, mammography is less effective in younger women and dense breasts and less sensitive to small tumors. Contrast-enhanced mammography can be performed, which is more accurate than mammography and ultrasound in dense breasts. However, this technique is not widely available due to its high cost and the inconvenience of high radiation exposure. MRI could detect small lesions that are not noticed by mammography but has lack of selectivity and is too expensive. In addition to the screening techniques mentioned, biopsies can also be performed to distinguish cancerous tissue from normal tissue. However, this is an invasive technique that is also expensive and requiring expert analysis (53). Despite the above mentioned techniques, BC can be detected by measuring BC biomarkers levels. BC biomarkers currently used in clinical diagnosis include PR and ER, HER2, BRCA1 and BRCA2, Ki-67, CA15-3 and cancer antigen 27.29 (CA27.29), CEA, urokinase plasminogen activator (uPA) and plasminogen activator inhibitor 1 (PAI-1) (22). BC biomarkers can be classified, based on omics, into genomic (DNA molecules), transcriptomic (RNA molecules), proteomic (protein molecules) and metabolomic (small metabolites) biomarkers (26). A person’s genome profile can indicate the risk of developing cancer. Alterations in genes (e.g. deletions, amplifications or mutations in certain genes such as tumor suppressor genes, protooncogenes or cell cycle regulators) increase a person's predisposition to developing cancer. A single nucleotide polymorphism (SNP) is a variation in a single nucleotide at a specific position in a particular gene. When present in genes such as cytochrome P450 family 1 subfamily A member 1
State of the art 16 (CYP1A1), RAD1, BRCA1, BRCA2, PTEN, TP53, checkpoint kinase 2 (CHEK2), ataxia telangiectasia mutation (ATM), PALB2, BRIP1, fibroblast growth factor receptor 2 (FGFR2), mitogen-activated protein kinase kinase (MAP3K1) and transforming growth factor ß 1 (TGFB1), it can lead to BC. BRCA1 and BRCA2 are the most commonly used genomic BC biomarkers and genomic screening for these genes is often performed to predict BC. Promising results have been obtained with the detection of these genes by biosensor-based techniques, especially electrochemical and optical. Circulatory tumor DNA (ctDNA) and cell-free DNA also gained a significant attention as BC biomarkers in recently years (26). Transcriptomic biomarkers, as messenger RNA (mRNA) and miRNA also provide important information about cancer development. Different specific BCassociated miRNAs, as miRNA-21 (13) and miRNA-155 (14) were successfully detected by electrochemical and optical biosensors, respectively. Although nucleic acid biomarkers can provide important information related to the tumor growth, they do not enable early diagnosis due to low concentrations (53). Since proteomic biomarkers are relatively abundant compared to RNAs and DNAs and play an important role in clinical diagnosis, they are the most frequently investigated molecules in BC. Numerous protein biomarkers are involved in the development of BC, such as CA15-3, cancer antigen 15-9 (CA15-9), cancer antigen 125 (CA125), CA27.29, cancer antigen 19-9 (CA19-9), CEA, epidermal growth factor (EGF), vascular endothelial growth factor (VEGF), TP53, cathepsin D, cyclin E, epidermal growth factor receptors (EGFR) and HER (26). Metabolomic biomarkers are low molecular weight molecules present in several biofluids whose concentrations can change during cancer development. Hence, metabolites can be potential cancer biomarkers used both in diagnosis and prognosis. In BC, lower levels of histidine and higher levels of lipids and glucose are common (26). As mentioned above, proteins are important cancer biomarkers as their concentration significantly varies in the consequence of a cancer development (9, 54). The human mucin (MUC) family comprises proteins with a high molecular weight, from MUC1 to MUC21. They can be secreted MUCs equipped with highly glycosylated tandem repeats that act as a physical barrier and limits exposure to external factors, thus protecting epithelial cells from stress and inflammation. MUCs can also be present in the form of transmembranes, which also contribute to this physical barrier and still have the task of transmitting survival and growth signals to the inside of the cell (55, 56). Deregulation of MUC production contributes to chronic inflammation response (55). Overexpression of transmembrane MUCs is related with several
State of the art 17 carcinomas once it influences some processes that leads to oncogenesis (55). The surface of several types of carcinoma cells overexpress transmembrane MUCs and this expression is induced by inflammatory cytokines (57). MUC-1, is also known as polymorphic epithelial mucin (PEM), epithelial membrane antigen (EMA) and episialin (57). MUC1-N terminal subunit (MUC1-N) is linked through stable hydrogen bounds with the MUC1-C terminal subunit (MUC1-C), forming a stable non-covalent complex that is expressed at the typical border of normal epithelial mammary cells (55, 58). MUC1-N terminus is extracellular and highly modified by O‑linked glycans, acting as a cell barrier and restricting cell-cell and cell-extracellular matrix connections (55, 58). Under normal conditions, MUC1 expression is mostly restricted to the apical side of glandular epithelial cells, acting as a barrier to protect cells from injurious environments and pathogens invasion and provide resistance to stimuli (51, 57, 58). It also participate in adhesion during metastasis and contribute to cell surface lubrification, hydration and protection from degradative enzymes (58). However, MUC1 is overexpressed in several types of human carcinomas, including BC, as well as colon, lung, liver, pancreatic and ovarian cancer (58) as a consequence of genetic alterations and transcription dysregulation (56). MUC1 glycosylation suffers an alteration in human carcinomas as consequence of changes in glycotransferase expression of cancer cells (55, 58) and thus, MUC1 is usually less glycosylated in malignant than in normal tissue (Figure 2-4) (51, 58). In cancer cells, MUC1 can play either a pro or antiinflammatory role, can limit the effectiveness of some drugs, can promotes invasion and migration of cancers and can inhibit cancer cell growth and apoptosis (58). In the presence of an epithelial stress response, MUC1-N is released from the cell surface and MUC1-C acts as a second line of defence, protecting the cell against loss of integrity. With loss of polarity in consequence of epithelial cell surface damage, MUC1-C domain is repositioned and expressed over the entire cell membrane (56, 57). MUC1 can play an anti-inflammatory role by inhibiting the response of dendritic cells and thus the inflammatory process, or it can influence Toll-like receptors through immunomodulation. On the other hand, its pro-inflammatory role is related to its interaction with macrophages and dendritic cells by regulating the recruitment of inflammatory cells, which allows the tumor to escape the immune system and thus promote metastasis (58). Overexpression of MUC1 can also limit the effectiveness of some therapies by reducing drugs intracellular uptake, promotes chemoresistance or radiotherapy resistance (58). MUC1 is also involved in the regulation of some factors that promote cell invasion and metastasis,
State of the art 18 as example, through the transforming growth factor (TGF) signalling pathway, by an increase in the production of exogenous platelet-derived growth factor (PDGF)-A or by inducing the expression of neutropilin-1 and its ligand VEGF thus promoting angiogenesis (58). Figure 2-4. Structure of MUC1 in normal (A) and cancer cells (B). Adapted from (58). Circulating MUC1 levels are determined by CA15-3 assay (55, 56), approved by US Food and Drug Administration (FDA) for monitoring BC patients during therapy and to detect disease recurrence at early stages (55). CA15-3 is the soluble form of MUC-1 (59) with ≈400 kDa, secreted by BC cells (60). It is overexpressed in 90% of BC (37, 61) and is the most widely used biomarker in serum of patients with BC (59). Despite CA15-3 has not enough sensitivity and selectivity for an early-stage BC diagnosis, it can be used for predicting prognosis and therapy monitoring of patients with metastatic disease (23, 52, 53, 59). The cut-off value for CA15-3 in clinical practice is 30 U mL -1 and superior concentrations are associated with a poor prognosis in patients with invasive BC and can be an indicative factor to search for metastases (62). For this reason, the main clinical application of CA15-3 is BC patients monitoring following surgery, the detection of recurrences in a preclinical stage (63). CA15-3 prognostic value is independent of factors as patients age, tumor size and axillary node status. It is capable to predict outcome of the disease in both node-negative and node-positive patients as well as ER-negative and ER-positive. Finally, it was demonstrated that CA15-3 maintains it prognostic value independent of the type of therapy used (e.g. hormone
State of the art 19 therapy, chemotherapy or radiotherapy) (64). Several studies have demonstrated the relation between CA15-3 levels and metastasis, poor prognosis and shorter overall survival (62-64). Shering et al. (63) performed a study (n=368), with a median follow-up time (3.28 years), during which the preoperative serum concentrations of CA15-3 were measured in BC patients. A cut-off value of 30.38 U mL -1 was used, and it was demonstrated that patients with high CA15-3 levels had a worse prognosis than those with concentrations below the cut-off value. The probability of disease-free survival at 5 years was calculated and was 44% in patients with higher CA15-3 levels compared to 65% in patients with lower CA15-3 levels. The probability of overall survival in these two groups was 67% and 83%, respectively. This study suggests that preoperative CA15-3 serum levels provide important information about patient outcome and may be a useful tool for selecting appropriate adjuvant therapy. In addition, this study found that CA15-3 levels were not dependent on age or ER status, but they were higher in patients with negative lymph node and larger tumors (63). Later, the above reports were confirmed by Duffy et al. with a larger number of patients (n=600) and a longer follow-up (6.27 years). In patients with histologically confirmed BC, the CA15-3 concentration was measured at a preoperative stage. Follow-up of these patients suggests that those with high CA15-3 levels had a significantly shorter overall survival than those with lower levels of this protein. The results confirm previous findings on the prognostic value of CA15-3, as patients with high CA15-3 levels had poorer overall survival. However, in this study, CA15-3 showed prognostic value in the node-negative subset of patients, while this was not observed in the previous study. In addition, CA15-3 predicted outcome in ER-negative patients in this study, which was not observed in the previous study (64). In a study including 2036 patients, about 200 had CA15-3 concentrations above 30 U mL -1. From these, metastases were found in 75 patients (62). CA15-3 determination in clinical context is normally performed by ELISA or electrophoresis. However, the use of antibodies in ELISA restricts the stability and the storage conditions and increases the cost of the assay, and electrophoresis is not reliable for routine practice and PoC (23). Similar to what occurs with several serum biomarkers, the main limitation of CA15-3 is lack of sensitivity once its serum levels are slightly increased in patients with localized tumors or at early stages of the disease (59). Also, CA15-3 lacks from specificity, as most of the cancer biomarkers that are not organ specific, once its levels can be increased in ovarian, pancreatic, gastric and lung cancer (57, 59). Besides that, some benign conditions could also be
State of the art 20 responsible for higher levels of CA15-3 (59). For that reason, CA15-3 should not be used alone for BC monitoring but in combination with other methodologies as physical examination, clinical history and imaging (59). Also, apparently, there are inconsistent results about the prognostic value of CA15-3 and CEA and a controversial opinion between researchers regarding which one has superior prognostic value in BC (37, 52). These inconsistency seems to be related with short follow-ups, small sample size, and variable cut-off values used in different studies (52). However, the simultaneous measurement of CA15-3 and CEA seems to be the most reasonable choice once it allows the early detection of metastasis in BC in up to 60-80% of patients (37). Besides the prognosis value of MUC1, it is also a promising target for vaccines development, as well as antibodies and drug inhibitors (55). Actually, inhibitors of MUC1-C terminal subunit have already been developed and it was proved that they block the oncogenic function of the protein, thus inducing death of BC cells, in in vitro assays (56). Also, several monoclonal antibodies have been produced against MUC1-N subunit (56). Considering that CA15-3 is one of the biomarkers approved by FDA to monitor BC patients, due to its clinical significance in monitoring the BC patients therapy, as well as in the detection of the disease recurrence in early stages, we selected this biomarker as the target case study to develop our sensors. 2.4. Biosensors Biosensors are analytical devices that combine an RE (e.g. enzymes, antigens or antibodies, metabolites, cells, nucleic acids, etc.), which is immobilised on a suitable substrate and has the function of detecting a specific target, and a transducer (e.g. optical, electrochemical, piezoelectric, photoelectrochemical, calorimetric, acoustic, etc.) responsible for converting the detection event into a measurable signal proportional to the analyte concentration and shown on a display, usually a computer or other digital device - Figure 2-5 and Figure 2-6 (65).
State of the art 27 uniform distribution of the RE over the sensor surface, which translates in higher recognition rates and reproducibility (89). Fc and Fab regions are the most commonly used in the development of immunosensors for the immobilisation of antibodies exposed to antigen binding regions (83). Three major antibody immobilisation approaches are considered, including the total exposure of the Fab region, its non-exposure or its partial exposure (76). Regarding this, immobilised antibodies can assume four different orientations, namely side-on, in which one Fc and one Fab is attached to the surface; tail-on, in which only Fc bounds to the surface; headon, in which the two Fabs are linked to the surface and, flat-on, in which all the fragments are linked to the surface Figure 2-7 (83). Figure 2-7. Schematic representation of random (side-one, tail-on, head-on and flat-on) and oriented immobilisation of antibodies. Adapted from (83). Antibody-binding proteins as Protein A and Protein G can be used for oriented antibody immobilisation due to their ability to specifically bind to the Fc region of the antibody, while not interfere with the analyte binding. As Fc region do not bind to the antigen, using this Fc-binding proteins, assure that antibodies can be immobilised on the sensor surface while maintaining the Fab region available for antigen binding (90). Several techniques can be used to analyse antibodies immobilised on a sensor surface, confirming its presence, and elucidating about its orientation, such as Fourier-transform infrared spectroscopy (FTIR) which enables the identification of specific chemical groups, fluorescence microscopy that allows to see the efficient binding of analytes in surfaces functionalised with antibodies, atomic force microscopy (AFM) that provides information about the degree of coverage of sensor surface, as well as thickness of the layer and shape of immobilised antibodies (83). Immunoassays can be classified in two different categories. In homogeneous immunoassays, the biochemical reaction takes place in the solution phase, while in heterogeneous immunoassays the RE (an antibody or an antigen) is immobilised on the transducer surface,
State of the art 28 where the binding interaction occurs (3). The detection can be direct where no labels are used and the changes resulting from the immunochemical reaction and formation of the complex are directly measured or it can be indirect, where a label is coupled to the antibody or antigen (3). An approach that widely uses antibodies as REs is ELISA. It was firstly described in 1971, by Peter Perlmann and Eva Engvall, for protein detection (91) and it is considered the gold standard method for detection of several molecules as proteins, antibodies, hormones, toxins and drugs (70) being applied not only in diagnosis but also in to food industry, toxicology or drug monitoring (76, 78). Several commercial ELISA kits are available for human immunodeficiency virus (HIV) detection, Influenza, Ebola, dengue, among others. ELISA procedures start with the adsorption of an antigen or antibody to a solid substrate (e.g. plastic microplates, cellulose paper). The test sample is then added followed by the reaction mixture. A wash step enables the removal of unbounded reactants. An enzyme-labelled antibody is added and finally the substrate enables color development whose intensity is usually proportional to the analyte concentration and can be visualized and quantified (78, 92). Furthermore, the ELISA readout can be performed by spectroscopy, colorimetry (CM), fluorescence, luminescence, or chemiluminescence (CL). Conventional ELISA is typically performed in 96 well plates of polystyrene, allowing several samples to be measured at each experiment. However, ELISA protocols take several hours due to several steps, such as long incubation times and washing or blocking steps. Also, high volumes of sample and reagents (20-200 µL) are consumed and specialized instruments are needed. These reasons limit its application in PoC devices and its use in low-resources areas (93). Apart from plastic materials (well plates) used in conventional ELISA, several immunoassays on paper-based substrates have been used, such as lateral flow immunochromatography on nitrocellulose membranes and dot-immunobinding assays (DIBA) in filter paper (94). As the work developed on this thesis was based on a paper-based ELISA (PELISA) approach, this topic will be discussed in more details at the section 2.4.3.1.4. Paperbased enzyme-linked immunosorbent assay. Different signal transducers can be coupled to immunosensors, and thus they can be classified into optical, electrochemical, and piezoelectric immunosensors (3, 74, 81). Electrochemical immunoassays have high sensitivity and enable the development of low-cost devices with operational simplicity and potential automation and miniaturization. They are reliable for small and large molecules and enable the detection of trace amounts of biological molecules (3).
State of the art 29 Different electrochemical methodologies have been employed on the development of electrochemical immunosensors, as amperometry, voltammetry, potentiometry, impedimetric, electrochemiluminescence (ECL), piezoelectricity and field-effect transistor. In these sensors, amplification strategies have been used, through the use of nanomaterials or enzymes for signal amplification (3). Colorimetric immunoassays may be based on changes on optical properties of nanomaterials, in consequence of three different phenomena: aggregation-based method, based on NPs aggregation induced by the presence of the analyte; morphology-based method, based on changes on its morphology; enzyme-mediated method, based on color changes of NPs produced by enzymatic reactions (3). Colorimetric immunoassays for cancer biomarkers detection have gain significant attention (17). In these sensors, detection probes are commonly used and labelled with enzymes, in ELISA. However, enzyme-based detection has some drawbacks, related with high costs and denaturation (3). Despite the advantages of current immunoassays and their wide use in clinical diagnosis, they still face some important challenges that need further consideration. Continuous research in this area and its evolution is essential to broaden the applications of immunoassays, especially in the clinical field. Effective and stable integration of the RE into the transducer surface, adequate functionalisation of nanomaterials and biocompatibility of the nanoprobes with in vivo assays are some of the major challenges (3, 81). 2.4.1.3. Molecular imprinting polymers MIPs are synthetic polymers that act as artificial receptors with high-affinity binding sites to a specific target analyte, due to the formation of complementary cavities on the polymer network. Molecular imprinting technology (MIT) is an emergent technique that enables the production of MIPs, capable of mimicking the biological recognition event. They are considered artificial receptors and can be used in alternative to natural receptors, allowing to overcome some of its limitations (95). In 1955, Dickey (96) modified silica adsorbents with organic dies, based on the concept of molecular imprinting. However, the preparation of these materials reveals low reproducibility and only moderate selectivity for the target. Later in 1970 Wullf et al. (97) and Klotz (98) showed the impression of templates in organic polymers with formation of specific cavities. In 1993, Mosbach et al. (99) shows the utility of MIPs in replacing antibodies in biosensors through a radiolabelled ligand-binding assay (100).
State of the art 30 MIPs have some unique properties that make their application suitable in different as separation, biosensing, catalysis and drug delivery (25, 101-103). 2.8.1.3.1. Production MIPs production is a multi-step process (Figure 2-8) involving the combination of functional monomers and at some cases, cross-linking monomers and an initiator, in the presence of a target template (103-106). Figure 2-8. Steps of MIP production. Taken from (106). The above mentioned components are dissolved in a porogenic solvent and covalent, noncovalent or semi-covalent chemistry leads to the formation of a complex between the functional monomers and the template (25, 101). When monomer and template are put together, functional groups of the monomer orient through their respective groups in the template molecule and cross-linker immobilises the orientation of functional groups (95). Then, free radical polymerization or electro polymerization enables the connection of molecules of functional monomer forming a 3D-network that could be induced by different external stimuli as heat, light, charge or even chemicals (25). For MIPs production, different imprinting approaches can be considered as bulk or surface imprinting (25, 101). Lastly, the template is removed from the imprinted material allowing the generation of template-specific binding cavities at the surface of the sensor. This removal step can be achieved via physical (heat) or chemical (acids/bases, detergents, enzymes) processes. Another type of protocols, such as washing surface with excess of ions or ultrasonic treatments are also reported in the literature (95). As orientation of functional groups remains after template removal, the formed cavities will have specific binding sites complementary in size, shape, and functionality to the target (104-107). A non-imprinted polymer (NIP) is prepared in parallel, in the same way as MIP but in absence of template (100). The performance of a MIP can be affected by the combination of components in polymerization mixture and their amounts, such as type and concentration of monomer, cross-linker, initiator
State of the art 31 and solvent, as well as experimental conditions as temperature or time of polymerization (108110). Templates used in molecular imprinting could be ions, atoms, molecules or cells (110). The selected template should have great chemical stability during polymerization reaction and must contain functional groups that establish complexes with functional monomers and do not inhibit polymerization reaction (110). In some cases, for example, when the template is a rare molecule or has a high toxicity or is not stable under imprinting conditions, structural analogues could be used instead of the template (95). The number and variability of functional monomers that could be used in molecular imprinting is restricted, thus limiting selectivity and potential applications of MIPs (110). A functional monomer should be wisely chosen with respect to their functional groups to interact with template by covalent or non-covalent bindings, forming a stable complex and creating highly specific cavities (95, 102, 109). A strong interaction between functional monomer and template leads to the formation of a stable complex, thus improving the binding capacity of the MIP (109). Carboxylic acids (acrylic acid, methacrylic acid, vinylbenzoic acid), sulphonic acids (2acrylamido-2-methylpropane sulphonic acid) and heteroaromatic bases (vinylpyridine, vinylimidazole) are commonly used as functional monomers (102, 109). Molar ratio between template and functional monomer also affects the binding affinity of the MIP and should be optimised, once lower molar ratios leads to less binding sites but higher molar ratios could induce non-specific binding (109). Combinatorial and computational methods have been used to predict the conformational and chemical complementarity between functional monomer and template and thus select the most suitable combination and their ratios and so improving MIP performance (111). The cross-linker as the role to connect and fix the functional groups of the monomers around the template molecule forming a rigid 3D network polymer, thus contributing to the polymer morphology and the stability of imprinted binding sites after template removal (95, 102, 109). The type and amount of cross-linker used in the polymerization affects the binding capacity of MIPs once low amounts of cross-linker doesn’t allow to maintain the stability of the cavity configuration and could lead to non-specific binding and high amount of cross-linker could reduce number of recognition sites (109, 110). N,N’ -methylene bis-acrylamide (MBAA), ethylene glycol dimethacrylate (EGDMA) (100), divinylbenzene (DVB), trimethylopropane trimethacrylate (TRIM) (102, 109) are some used cross-linkers.
State of the art 32 When MIPs are prepared by free radical polymerization, initiators are used and reaction can be started in a thermal or photochemical way, being peroxy and azo compounds commonly used for that (110). The porogenic solvent is responsible to bring together all the components of the MIP synthesis and produce pores in the polymer that enable access to imprinted binding sites (102, 110). The solvent in which the MIP is prepared has a significant effect on the interaction between template and monomer. They should allow the dissolution of the other components but not interfere during the polymerization process. Organic solvents are usually selected for MIP production as they increase the hydrogen bonding and electrostatic interactions between monomer and template (112). The polarity of porogen influences interaction and strength of binding between template and functional monomers and in polymer morphology (109, 110). While less-polar solvents will stimulate the formation of monomer-template complex, enabling better imprinting factor, more polar solvents will hinder the formation of this complex, thus lowering the imprinting factor. However, the choice of a less polar solvent can compromise the solubility of the components thus leading to the precipitation of the produced MIP. The selection of a solvent with medium polarity will allow the dissolution of the components while not affecting so much the formation of the monomer-template complex (112). Organic solvents as toluene, chloroform, methanol, dichloromethane, tetrahydrofuran, N,N -dimethylformamide and acetonitrile are generally used (102, 110). When the template molecule or other component of the MIP is not compatible with organic solvents, for example when MIPs are designed to detect proteins, water or aqueous solutions as buffers can be used (113-117). The imprinting of biological macromolecules as proteins, cells and viruses that remains a challenge in MIT once the polymer network can obstruct the mass transfer, thus causing slow leakage of template during removal step and slowing the rebinding kinetics (118). Macromolecules can even become entrapped in the network after polymerization, and cannot be extracted during removal step, thus hindering the rebinding (110, 119). Some strategies for imprinting of macromolecules have been suggested, such as surface imprinting, epitopemediated or imprinting micro-contact imprinting (95, 109). Also, the presence of heterogeneous binding sites constitutes a problem in molecular imprinting, due to the formation of non-covalent bonds during polymerization step. Semi-covalent approach can be an alternative to surpass this question (109). In addition, when MIPs are synthetized in organic solvents, they show low binding capacity to the target in polar solvents such as aqueous media. To overcome this
State of the art 33 situation, many methodologies have been developed, such as applying a two-step extraction method or using hydrophilic monomers for the development of water-compatible MIPs (109). Another challenging aspect related to the imprinting process is that changes in the protein conformation or protein denaturation can occur, which limit protein rebinding. Also, template removal could be successful, but the molecules used in removal procedure could be adsorbed in polymer, blocking binding sites and thus compromising template rebinding (25, 95, 101, 119). 2.8.1.3.2. Advantages and disadvantages of MIPs MIPs have several advantages, such as high selectivity associated with the formation of cavities with affinity to the desired target and high sensitivity due to the transduction systems that can be applied. Stability, robustness and resistance to temperature and pressure are other advantages related to the synthetic nature of the materials. In addition, their manufacturing process is reproducible and cost-effective, with the possibility of mass production and compatibility with miniaturized devices, and the shelf life of the polymers is high. Nevertheless, MIP-based sensors can be designed to detect a variety of targets, including amino-acids, proteins, peptides, viruses, cells and chemicals as drugs or pollutants (25, 100, 102, 111). Despite the several advantages of using MIPs and the simplicity of its production, the synthesis of a MIP for a specific template is time consuming and requires several tests until obtaining optimum conditions for several variables that affect the binding capacity of the final product (102). Also, as already mentioned, despite of the imprinting of small molecules (< 1500 Da) has been widely reported, macromolecular imprinting is still a challenge (103, 107, 118). In addition to that, in some cases, polymerization conditions are usually non-physiological, with harsh conditions that can trigger conformational changes in molecules such as proteins. Also, the existence of multiple heterogeneous binding sites in macromolecules could be the reason for nonspecific binding in the imprinting of this type of molecules. In addition, the choice of monomers for imprinting macromolecules may be limited if some monomers are only soluble in organic solvents which could compromise protein structure. At least, not all the polymerization strategies are appropriate for all types of templates. For example, bulk imprinting is a standard imprinting method being successful for small molecules, but it is not suitable for macromolecules. Regarding that, many imprinting strategies have been developed to overcome these issues, as surface imprinting or epitope imprinting (107, 118).
State of the art 34 2.4.2. Transducers The transducer is the component of the biosensor responsible for converting the biochemical event that results from the interaction of the RE with the target molecule, into a measurable signal (Figure 2-5). Regarding the transducer technology, biosensors can be optical, electrochemical, piezoelectric (9). Electrochemical sensors are based on the conversion of the biological event into an electrochemical signal, while optical sensors detect the target based on modification of light absorption, fluorescence, luminescence, SPR, among other optical phenomena. On the other hand, piezoelectric sensors are mass-based transducers that identify mass changes to recognise and quantify the target (9, 120). 2.4.2.1. Electrochemical biosensors Electrochemical biosensors enable the quantification of the target analyte through the detection of an electrochemical reaction (oxidation, reduction, or transfer of charges) on the electrode surface (61). These type of sensors are usually composed by a reference electrode, a counter electrode and a working electrode, the latter containing the RE (22). The intensity of the signal depends on the target concentration. Cyclic voltammetry (CV), square wave voltammetry (SWV), differential pulse voltammetry (DPV), linear sweep voltammetry (LSV), electrochemical impedance spectroscopy (EIS) and field-effect biosensor (FET) are the main electrochemical methods used in biosensors (61). Electrochemical biosensors provide fast responses, having high selectivity and sensitivity, and displaying a high potential for miniaturization and portability. As they provide fast response at ultra-low level of biomarker’s concentration, electrochemical biosensors became a useful tool for BC biomarkers (26, 120). Several electrochemical biosensors have been reported for BC biomarkers. The fabrication of these devices is technically simple, providing fast and cost-effective detection of cancer biomarkers, using low volumes of samples and with high sensitivity provided by electrochemical transduction systems. Also, these sensors can be integrated in small and portable PoC devices. However, electrochemistry requires the use of potentiostatic equipment for sensor construction and signal acquisition (10, 60, 121-123). 2.4.2.2. Optical biosensors In optical biosensors, a RE is used in combination with an optical transducer system which provides a visible response, eliminating the need of an equipment to read the results and also
State of the art 35 enabling the portability of the sensor (22). Optical sensors can provide direct label-free detection through quantification of luminescence or fluorescence or by color change/appearance/disappearance by the measurement of absorbance, transmittance, reflectance, phosphorescence or fluorescence emissions in different regions (UV, visible or nearinfrared spectrum) (22, 41). Among these, CM is the most appropriate technique when simple and low-cost detection is intended. Also, it has potential for PoC devices and the results can be observed by naked eye (19, 124-126). Optical-based sensors for cancer detection have been developed benefiting from the advantages of optical transducers and nanomaterials (3). For BC biomarkers detection, several biosensors have been reported, using different optical transduction methodologies, including fluorescence (15, 20), SPR (121), surface-enhanced Raman scattering (SERS) (127), ECL (123) and CM (128, 129). Optical biosensors usually offer good sensitivity and specificity for cancer biomarkers detection, enabling real-time responses. Also, the use of different tag molecules as fluorophores enables the detection of different molecules on the same device. However, some optical transduction methods require expensive equipment for signal acquisition. CM, that is the detection method used on for the sensors developed in this thesis overcome the need of this equipment, due to the possibility of naked eye detection. 2.4.2.2.1. Fluorescence Among the optical properties used for monitoring a wide range of analytes, fluorescence is the most common, due to the selectivity of organic fluorophores and possibility of multiplexing (22). In fluorescence sensors, emission of fluorescence occurs due to the excitation of electrons at a certain wavelength and de-excitation at a superior wavelength. Fluorescent signal transduction highlights by their non-destructiveness, fast response, reproducibility and high signal intensity (130). Compared to conventional molecular probes (e.g. fluorescent proteins and organic dyes), luminescent nanomaterials exhibit great physicochemical properties, such as long lifetime, high luminescence efficiency, narrow emission bands and high resistance to photobleaching. These properties depend on their size, shape and composition. Advances in nanotechnology enable the production of a wide range of fluorescent materials that can be used as diagnostic probes in various medical applications, such as noble metal nanoclusters, up-conversion nanomaterials, quantum dots (QDs), graphene oxides (GOs) or CNPs.
State of the art 36 Fluorescent biosensors for cancer biomarkers as CA125 (131), CA15-3 (15), and BC-derived exosomes (132) have been reported. 2.4.2.2.2. Surface plasmon resonance SPR is a sensitive optical technique that allows the real-time study of biomolecular interactions that occur close to the transducer surface, which is usually a thin-gold-film on a glass slide (3). On SPR-based sensors the binding interaction occurs between the RE immobilised on the transducer surface and its corresponding analyte, which causes changes in the refractive index, thus a shift in the resonance angle that can be detected and provide information about the concentration of the analyte that binds to the sensor, their affinity and association/dissociation kinetics (3, 26). Resonance is affected by the immobilisation of biomolecules onto the metallic layer, as well as the conformational changes of these molecules and interaction with other molecules (3). SPR has been used as transduction method in biosensors to detect biomarkers, including the cancer ones, such as HER2 (133), CA125 (134) or miRNA-21 and miRNA-155 (135). 2.4.2.2.3. Surface-enhanced Raman scattering Raman spectroscopy measures the inelastically scattered photons that result from vibrational frequencies when a molecule is excited with monochromatic light. It has been found that Raman scattering is strongly enhanced when a molecule is close enough to noble metal surfaces (e.g. silver or gold). This phenomenon of Raman scattering enhancement is known as SERS, increases the intensity of Raman scattering by up to 10-14 orders of magnitude and can be explained by chemical and electromagnetic components. It is considered a non-invasive and non-destructive technique (3). Several cancer biomarkers, such as CEA (117), CA19-9 (136) and prostate specific antigen (PSA) (137), have been detected by SERS-based sensors. 2.4.2.2.4. Colorimetry Colorimetric biosensors provide a highly sensitive response that can be applied to PoC devices as soon as the detection of biomarkers occurs through color changes that can be observed with the naked eye, both in solution and on solid supports (e.g. paper or plastic plates) (26). Several colorimetric formats have been developed to allow an easy interpretation of the results, as
State of the art 43 simplicity (164) and consist in using melted wax to print wax patterns on paper surface that create hydrophobic barriers that allow to control the flow of liquid samples and reagents (125). The interest in PADs has grown exponentially in the last decade and their conjugation with portable readers as smartphones make them meet the criteria for PoC devices (27, 153), represent an alternative to conventional methods for the detection of biomarkers, especially in places with limited resources, if they fulfil the WHO's ASSURED requirements (i.e. affordable, sensitive, specific, user-friendly, rapid and robust and available to the end user). Affordability is met by using inexpensive materials such as cellulose and plastic carriers. High sensitivity could be achieved by NPs of different materials as colorimetric probes, and they can be easily functionalised with REs to achieve good selectivity; sensitivity avoids false-negative results, while selectivity prevents false-positive results; PADs are also user-friendly and accessible to the end user as they are easy to use, do not require additional steps between sample collection and sample application and do not use invasive samples, so results are easy to understand and do not require trained personnel or sophisticated equipment; the porosity of the cellulose allows capillary forces to act, so a rapid result is achieved; .Also, PADs are robust devices due to their constituents; patients could self-test at home without the need to go to the hospital, due to the deliverability of these devices or they can be performed in a clinic or in a hospital in a rural setting (1, 24, 29, 41, 66, 155, 157, 159, 161). Different types of samples can be collected and used in PADs, as blood, urine, saliva, water, soil, food and drug samples (29). Poor sensitivity and the ineffectiveness to provide quantitative measurements are the main challenges of these types of devices but could be overcome by the development of new designs and application of new materials. 2.4.3.1.1. Types of paper analytical devices Different designs of PADs (Figure 2-11), including spot tests, dip-sticks, LFAs, and microfluidic paper analytical devices (μPADs) (24, 75, 81, 129, 163), have been developed and used for detection of analytes in different areas as clinical diagnosis, environment monitoring and food quality (24, 125). 2.4.3.1.1.1 Spot tests Spot tests are rapid and inexpensive devices (155) and were applied, for the first time, in the 1930s and 1940s, for metal ion detection using colorimetric ligands (165).
State of the art 44 2.4.3.1.1.2 Dipsticks Dipsticks are a very simple format of PAD, easy to produce and providing results easy to analyse. It is based on the impregnation of reagents in a paper strip, leading to a color change, in the presence of the analyte, directly observed by naked eye. However, it suffer from some drawbacks as low accuracy and long analysis time, only providing a qualitative response (124, 125) The first dipstick appeared at 1956 for glucose quantification in urine (166). Other examples of dipsticks are pH strips or urine test-strips that are used nowadays to simultaneously screen multiple disorders as diabetes and kidney disease (124, 157, 167). Figure 2-11. Examples of different formats of PADs. Photograph of dipsticks (A) for urine. Adapted from (168); Spot test (B): photographs (a) and bar charts (b) of grey values for alpha fetoprotein detection. Taken from (169); LFA (C): schematic representation (a) and photograph of a LFA for microRNA-125 detection. Adapted from (170); photograph of a µPAD (D) for microRNA-21 detection. Adapted from (171). 2.4.3.1.1.3 Lateral flow assays First LFA was reported and patented in the 1956, by Plotz and Singer (172) and since then LFAs have been grown over the last decades with efforts to enhancing their performance (24). Pregnancy test is the most known example of LFA, being a very simple test that does not require
State of the art 45 additional steps between sample (urine) collection and application in the test (29). LFAs are a powerful tool for several biomarkers’ detection in clinical context as well as in food safety environmental monitoring (5, 167). The principle of an LFA is that a liquid sample, containing or not the target, flows horizontally through the several pads of the device without the need of external forces and react with preimmobilised reagents (146, 161). LFAs are generally formed by NC strips assembled on a plastic backing card, generally made of polyvinyl chloride and containing the different parts: sample pad, conjugate pad, flowing membrane or test pad and adsorbent pad (1, 16, 24, 75, 81, 161, 167). The sample and adsorbent pads are usually cellulose paper or glass fibers, whereas flowing membrane is commonly a NC membrane and conjugate pad is usually made by glass fibers (1, 75). The pads are assembled adjacently to allow a continuous lateral flow of the solutions and reagents when they are added on the sample pad (16). The sample is added to the sample pad that guarantees that the analyte reaches and bind to the capture reagents (146, 167). The sample further migrates along the conjugate pad which contains colored particles (e.g. metal NPs) responsible to capture the target once they are functionalised with REs (e.g. antibodies or aptamers) (16, 146). Then, the conjugate formed between the target and the RE, flows through the strip until it reach the test pad (146). The test pad is the platform for bio-analytical and recognition reactions once it is formed by NC strips that could be easily modified with different reagents to allow capture molecules immobilisation, thus forming the test and the control line (5, 16, 24). If the analyte is present, it will lead to a response at the test line, whereas the adequate flow through the strip could be assumed by the presence of color at the control line (146). Lastly, at the end of the strip, an adsorbent pad provides the continuous flow of sample based on capillary forces and, at the same time, it wicks the excess of reagents, preventing liquids reflux (16, 146, 167). The appearance or absence of colored lines at test line could be observed by naked eye or using an adequate reader (146). Regarding the RE used, there are two principal groups of LFAs. Lateral flow immunoassays (LFIA) are the most common ones, in which the REs for the analyte of interest are labelled antibodies (146). On the other hand, aptamers could be used for target recognition, resulting in aptamer-based LFAs (167). LFAs are usually associated to colorimetric or fluorescent transduction systems (146) and colloidal AuNPs are commonly used as colorimetric probes in LFAs once they have unique optical properties and are easy to functionalise for target recognition (24).
State of the art 46 Two basic assay formats for LFA are considered: direct or sandwich LFA (sLFA), and competitive LFA (cLFA). sFLA is more suitable for high molecular weight compounds detection and includes two separated recognition steps that enhance the selectivity and sensitivity. In this assay, a primary antibody immobilised on the conjugate pad captures the analyte forming a complex that will be further captured by a transducer element immobilised at the test line, generally a colorimetric probe conjugated with a secondary antibody. The presence of the target at the test line led to the development of a positive signal that is proportional to the target concentrations. On the other hand, cLFA is more appropriate for small molecules detection, with single antigenic determinants. In this approach, the analyte competes and block the binding sites of the RE at the test line, preventing their interaction with the conjugates. Consequently, the concentration of the target is inversely proportional to the signal intensity (1, 3, 24). LFAs have several advantages as it low-cost, userfriendly format, high shelf life, high sensitivity and selectivity, the requirement of small sample volumes, the possibility of multiplex detection and a wide range of applications (24, 43). Conversely, some drawbacks are implied and need to be overcome. Regarding membrane matrix and REs, novel materials have been explored in order to enhance their performance (24). 2.4.3.1.1.4 Microfluidic paper analytical devices The first μPAD was created by a photolithography technique for colorimetric detection of glucose in urine from Whitesides laboratory in 2007 (173). μPADs are created by applying a pattern onto paper thus creating hydrophilic channels and defining different reaction zones, which enables multiplex detection. Also, these channels facilitate the direct flow of the sample, in opposite of which occurs in LFAs and spot tests (5, 143, 161). Two types of μPADs are considered, namely two-dimensional (2D) μPADs, formed by channels created by chemical or physical hydrophobic barriers, through several techniques as photolithography, wax printing, screen-printing, inkjet printing and plasma oxidation; three-dimensional (3D) μPADs constructed by folding several layers of patterned paper (5, 27, 28, 81, 161, 167). The fabrication technique should be chosen considering aspects as cost, substrate, time of fabrication and available equipment. Whatman® filter paper nº 1 is generally used in this type of devices due to uniform thickness and wicking properties. μPADs have the advantages of only require a very small amount of fluids (5 to 10 μL), could be miniaturized, could provide multiplex analysis by the creation of several channels, and simultaneously semi-quantitative and quantitative responses
State of the art 47 and could provide low response times (28, 161). This type of PADs has been used in several applications as medical ones, environmental monitoring, food safety and forensic analysis (28, 81, 157). 2.4.3.1.2. Recognition elements and transduction systems used in paper analytical devices The most common REs used in the development of PADs for cancer biomarker detection are antibodies and aptamers. Despite their broad application, researchers are continuously focused on develop new REs to improve the performance of these sensors, as MIPs. When the target is a nucleic acid, oligonucleotides or aptamers are usually applied; for proteins, enzymes and cells detection, antibodies and aptamers are used (1). These receptors are immobilised on the paper surface and on nanomaterials, used as probes, and the immobilisation methodology should be wisely selected to enable the retention of molecules without comprising its bioactivity. In paper matrix, REs could be immobilised by either physical or chemical ways whereas the immobilisation of receptors as antibodies on tags, should occur by adsorption techniques, cross-linking (e.g. glutaraldehyde (GA), carbodiimide method) or covalent binding through cellulose binding domains (periodate oxidation, tosylation). Entrapment within a gel and microencapsulation can also be applied (159). When using aptamers as REs, the attachment to NPs could occur using 5’-labelled sequences with thiol or amine groups, while for their immobilisation on the NC membrane, streptavidinbiotin interaction is commonly used (1, 159). Ge et al. where the first group that demonstrated the potential and advantages of using MIPs as REs in PADs. They applied MIT into a µPAD, by electropolymerisation of a MIP in an AuNPmodified paper. The porous morphology of the paper substrate that provides high surface area in combination with the conductivity of the AuNP layer significantly increased the sensitivity of the sensor (174). The methodology to prepare MIPs in PADs includes in situ polymerisation and post-introduction (Figure 2-12). In situ polymerisation strategy involves MIP synthesis directly on the surface of pre-modified paper fibers. Paper, which is rich in -OH groups, is usually treated with silane coupling agents to introduce other functional groups. This method is widely used due to its inherent simplicity. However, it requires the immersion of paper in solutions for a long time, which can affect the paper properties and introduce damages. In post-introduction method, the
State of the art 48 MIP is directly synthetized in a solution phase through radical polymerisation or sol-gel polymerisation and then incorporated into PADs (175). Figure 2-12. Schematic representation of MIP-PADs fabrication: (A) in situ polymerisation and (B) post-introduction. Taken from (105). Optical transduction is the universal transduction system for PADs once it is the cheapest and simplest method (159). Despite optical transduction, electrochemistry is also frequently used in PADs (72). Electrochemical is a suitable transduction system to be coupled with PADs due to the higher sensitivity, compared to optical detection (41). A three-electrode system is used, and the electrodes are deposited onto the paper surface by technologies as screen printing, physical deposition of metals, spraying or pencil-drawing of conductive inks (41, 168). As in other electrochemical devices, target analytes are analysed at the working electrode, while the redox reaction occurs at the reference electrode and the counter electrode is used to reduce the electrical current that flows to the reference electrode, thus maintaining a constant potential during the measurement (41). Optical transduction as CM, fluorescence, CL, SPR, SERS and transmittance and electrochemical transduction such as electrochemistry, photoelectrochemistry, or ECL are usually coupled to MIP-based PADs to detect the analyte of interest. MIPs provide to optical sensors high selectivity to the target analyte and versatility as MIPs can be produced against a wide range of molecules (7, 41, 125, 153, 155, 159, 161, 167). Colorimetric PADs detect the analyte of interest when it interacts with a sensing element on the paper substrate, resulting in a color change. When colorimetric tests are performed on a solid substrate such as paper, detection is usually by reflectance, measuring the light reflected from the surface of the test (139). Colorimetric PADs have been applied for the diagnosis of different diseases as cancer
State of the art 49 (128, 171), neurodegenerative (176, 177), infectious (178, 179), and chronic diseases (180, 181). As already mentioned, colorimetric sensing based on naked eye visual color changes is a very suitable approach for rapid tests and have several advantages over conventional methods once they allow a cost-effective, real-time, on-site and highly sensitive and specific detection of several molecules (162). However, these types of sensors usually suffer from lack of sensitivity, with interference of other molecules present on the sample. The combination of colorimetric detection with MIPs enables to overpass this drawback and allows these sensors to be used with complex samples (182). MIPs as RE coupled to colorimetric transduction enables a qualitative or semi-quantitative response by naked eye that can be performed in at home or in the field, without using any equipment (182). In addition to selecting a suitable RE, a suitable transduction method should also be chosen to use paper-based sensors in analysers without compromising the simplicity, portability, and costeffectiveness of the sensor. 2.4.3.1.3. Signal readout in paper analytical devices Colorimetric detection is one of the most detection technique used in PADs once it offers the advantage of simple visual detection by naked eye, without the need of trained personal or sophisticated equipment as it provides an immediate qualitative (“yes” or “no”) response or a semi-quantitative response by capturing the image with a camera and further transferred to a computer where image analysis will be performed by imaging software (153, 155, 162). However, sometimes CM suffers from lack of selectivity and sensitivity originating heterogeneous signals that could be misunderstood by the users (41). Results can also be registered by readout devices that can be simple instruments as scanners, cameras, smartphones or more complex ones as spectrophotometers or fluorimeters (1, 24, 41, 124, 125, 139, 159, 161, 162). Smartphones are widely used for this purpose due to their advantageous properties as easy-of-use, high-resolution cameras and operative systems and wireless connectivity that enables real-time analysis (11, 138, 139). However, colorimetric detection with smartphones can have some drawbacks as the influence of light conditions, focal distance, and device orientation as well as camera characteristics as lens quality and aperture dimension (29, 73). For that reason, images should be carefully captured under controlled light conditions as in black or white boxes developed for this purpose, so it could not affect assay sensitivity and repeatability (24, 41, 139). In alternative to smartphones, office scanner could
State of the art 50 also be used, being a device with high resolution and with the possibility of being portable and be used by unskilled personnel. Also, in this case, image is not affected by external light conditions (159). Still, spectrophotometers or fluorometers or more sophisticated equipment as microplate readers, photomultiplier tubes or even gel documentation systems could be used (159). Oppositely of colorimetric probes, fluorescent labels or paramagnetic particles cannot be read by naked eye thus requiring specific readers for quantitative measurements (146). Once captured, the images are transferred to a device such as a computer to be analysed using software such as Adobe Photoshop, Corel Photo Paint, ImageJ or DigitalColor Metre. This software is able to calculate parameters such as the red, green and blue (RGB) space or the hue, saturation and brightness (HSB) space. These values are used to calculate the target concentration (159). Figure 2-13 illustrate several examples of samples that can be analysed in colorimetric PADs, as well as biomarkers that can be detected. Types of paper, REs and labels used on the assembly of PADs are also represented. Different types of signal readout coupled to PADs are mentioned. Figure 2-13. Examples of: (A) types of samples and biomarkers, (B) types of paper, (C) REs and labels, and (D) signal readouts used in colorimetric paper sensors. Taken from (183).
State of the art 51 2.4.3.1.4. Paper-based enzyme-linked immunosorbent assay As above mentioned, conventional ELISA is performed in plastic well plates, but it can be done in other type of supports, as cellulose paper. The first P-ELISA was performed by Cheng et al. (184) in 2010. They used filter paper as substrate for antibody-antigen recognition, through a simple and low-cost assay. High selectivity and sensitivity were achieved, about 10 times lower than the reached by conventional ELISA procedure in 96-well plates, for the same antigenantibody pair (94). The P-ELISA utilises the advantages of the high selectivity of the ELISA and the low cost of the paper substrate and thus offers a suitable diagnostic platform for the detection of biomarkers, which is particularly important in resource-poor areas. It is usually performed on 96-microzone paper plates, which are generally produced by structuring hydrophobic polymers in hydrophilic paper sheets. P-ELISAs have the advantage of high selectivity due to the antibody-antigen interaction and the ability to analyse complex samples without pretreatment procedures. This shortens the duration of the test as it can be performed in less than an hour, whereas conventional ELISAs can take several hours if they require an hour or more per step. The fast reaction is also due to the high surface-to-volume ratio of the paper. The reagents flow over a short distance and therefore require short incubation times. In addition, due to the small reaction area and high volume-to-surface ratios, only small amounts of samples and reagents are consumed (1-10 µL), which reduces the cost of the assay. Another advantage of P-ELISA is that results can be either qualitative or quantitative once observed with the naked eye or quantified using a desktop scanner instead of an expensive microplate reader. These advantages extend the application of ELISA to non-specialised laboratories and to developing countries. However, the sensitivity could be lower than conventional ELISA, which can be due to short antibody-antigen incubation times or non-specific interaction between antibodies and the cellulose paper. Other Another disadvantage of P-ELISA is that the test zones are subjected to environmental conditions as relative humidity and temperature, which could influence the rate of evaporation of water (84, 94, 159, 184-186). 2.4.3.1.4.1. Antibodies immobilisation It was already mentioned the importance of antibodies immobilisation to the sensor surface and the related approaches for immobilisation (physical (non-covalent) or chemical (covalent). A study has shown that 40% of antibodies immobilised by physical adsorption onto cellulose paper
State of the art 52 can desorb from cellulose fibres, thus affecting the sensor performance and not providing reproducible results (187). On the other hand, the chemical approach leads to stronger binding, thus enabling the development of effective PADs (78, 85, 86). Since cellulose paper surface have few functional groups for covalent immobilisation, it should be functionalised before biomolecules immobilisation (84). Biomolecules could be covalently attached to the substrate through the modification of the surface with agents such as GA, EDC, and NHS (78, 85). Cellulose chemical modification by oxidizing agents is also common leading to the formation of aldehyde (-CHO), ketone or carboxyl groups (188). Oxidation with periodate (NaIO4) is used to graft biomolecules onto cellulose paper by cleaving the C2-C3 bond of the glucopyranoside ring, thereby converting the 1,2-dihydroxyl groups (glycol) of cellulose into aldehyde groups. Dialdehyde cellulose (DAC) is the resulting compound that is bound by antibodies through the formation of a reversible Schiff base between the aldehyde groups on the cellulose paper and the primary amine groups of the antibody (84, 85, 88, 188). Although it is a simple procedure, sodium periodate cannot be completely removed from cellulose fibers, thus causing oxidation of the immobilised antibodies (88). Chitosan could be used to modify the cellulose surface followed by GA that provide aldehyde groups for further biomolecules attachment (86). Another widely used technique is the silane coupling which enables grafting of amine functional groups that react with –OH groups on cellulose paper. 3-aminopropyltrimethoxysilane (APTES) is used to silane grafting (85). 2.4.3.1.4.2. Blocking step Blocking agents should be added to the protocol, as without appropriate blocking, non-specific reactions may occur, resulting in a high background signal and low sensitivity. The use of a blocking buffer saturates the free binding sites on the paper surface and thus improves the signal-to-noise ratio. There are a variety of blocking buffers that should be optimised for each assay, taking into account the substrate, assay format and detection system. However, an ideal blocking agent must fulfil several requirements to reduce the background signal without compromising the selectivity of the assay. Firstly, it should inhibit the non-specific binding of other components to the surface of the assay. In addition, it should act as a stabiliser, not promote non-specific interactions between proteins and have no cross-reactivity with the other assay components. It must also have low enzyme activity so as not to interfere with the detection method and it should be reproducible between batches. The most common blockers are
State of the art 59 Figure 2-15. Number of papers describing colorimetric PADs (purple bar) and colorimetric PADs for cancer biomarkers detection (orange bar), published on last decade (183). Figure 2-15 represents the number of scientific papers reporting colorimetric PADs (purple bar), highlighting the ones applied to cancer biomarkers detection (orange bar), published since 2013. Data included papers from the last decade, since 1st January 2013 to 1st December of 2023, from ISI WEB OF KNOWLEDGE, using the keywords “colorimetric AND paper-based” and “colorimetric AND paperbased AND cancer”. It can be noticed that colorimetric PADs have been increasing, being the ones applied to cancer a minority but still growing over the years. It seems that since 2020 this growing tendency has been stabilized. Although it appears that year 2023 had a decrease in the publications, no conclusions can the taken as the year is not over and search includes papers published until 1st of December. Focusing on cancer screening and diagnosis and on the purpose of this thesis, several biosensors were already reported for CA15-3 detection, in which different REs as antibodies (10, 21, 192), MIPs (60, 122, 193), aptamers (15) were coupled with different signal transduction approaches, such as CM (192), SPR (23, 121), fluorescence (15, 20), ECL (123), SERS (127) and electrochemistry (10, 21, 60, 193). In the next chapters, the sensors developed for cancer biomarkers detection over the course of this PhD will be presented. For these sensors, cellulose paper was used as substrate, in which different molecules were assembled as REs, and colorimetric detection was selected for signal transduction.
Nanocellulose-based biosensor for colorimetric detection of glucose 60 Chapter 3 Nanocellulose-based biosensor for colorimetric detection of glucose Chapter 3 describes the production of carboxyl-NC by TEMPO-mediated oxidation of microcrystalline cellulose (MCC) and its use for surface modification of cellulose paper. Glucose was selected as target based on epidemiological data that suggests that high levels of glucose have been related to an increased risk to develop BC. GOx and ABTS were used as RE and colorimetric system, respectively, to develop and enzymatic sensor for glucose colorimetric detection on a paper substrate modified with carboxyl-NC. Part of the content of this chapter was published in Sensing and Bio-sensing Research: Neubauerova, K., Carneiro, M. C. C. G., Rodrigues, L. R., Moreira, F. T. C., & Sales, M. G. F. (2020). Nanocellulosebased biosensor for colorimetric detection of glucose. Sensing and Bio-Sensing Research, 29, 100368. doi:10.1016/j.sbsr.2020.100368.
Nanocellulose-based biosensor for colorimetric detection of glucose 61 3. Nanocellulose-based biosensor for colorimetric detection of glucose 3.1. Introduction Nanocellulose (NC) materials are among the different products that may be obtained from cellulose by combining proper mechanical, chemical, and enzymatic treatments (194). They display outstanding properties, such as high surface area, mechanical strength, high thermal and chemical durability, or film-forming capacity (195-197). NC have a highly crystalline structure, with a rod-like shape ranging 2–50 nm width and 100–2000 nm length (198, 199). The most popular process for producing NC is acidic hydrolysis (200), employing typically concentrated sulphuric acid (201, 202) and/or hydrochloric acid (203, 204). In sulphuric acid-based hydrolysis, sulphate ester groups are introduced on the surface of the NC. This leads to the formation of highly stable NC suspensions, because negatively charged sulphate groups are attached to the surface of the NC. However, sulphuric acid is strongly oxidizing, and sometimes causes the degradation of cellulose. Unlike sulphuric acid hydrolysis, hydrochloric acid hydrolysis yields. The TEMPO-mediated oxidation is an alternative promising procedure, which can produce nanocrystals with modified surface in one step. TEMPO radicals catalyse oxidation of hydroxyl groups and after system supported by NaClO-KBr components converts oxidized aldehydes into final charged carboxyl groups (205-207). The use of this technique was attracting by many investigations since 1994, showing that only the primary alcohol groups of polysaccharides were oxidized, whereas the secondary hydroxyls remained unaffected (208). For cellulose, the TEMPO-mediated oxidation was applied to different types of cellulose, ranging from cotton to wood pulp, cell cellulose, rayon, and cellulose III (204, 209-213). These studies led to the preparation of a series of products, ranging from water-soluble polyuronic acid to partially derivatized cellulose products. According to previous results (205, 206), TEMPO-mediated oxidation was applied to microcrystalline cellulose (MCC) and a pre-treatment method of first oxidation was devised to prepare water-soluble polyuronic acid in high yield, which was dependent on the reaction temperature during first and second oxidation. In literature (214), the relationships between the amount of NaClO was optimised and either carboxylate or aldehyde content in oxidated NC. The carboxyl group content increased remarkably to 0.68 mmol solid -1 with the addition of at least 5 mmol of NaClO g MCC -1. The oxidation of primary alcohol groups in cellulose, catalysed by TEMPO, has been recently proposed as a more selective, faster, and better-controlled method (215, 216).
Nanocellulose-based biosensor for colorimetric detection of glucose 62 Due to its remarkable characteristics, NC is one of the most attractive cellulose-based nanomaterial used in biosensors and biomaterials applications (114, 217-219). Cellulose and NC have been widely used as a support material for proteins/enzymes immobilisation (215, 220-227), making use of optical or electrical transduction schemes (114, 168, 194, 228-230), due its outstanding characteristics of these materials. Besides, specially NC shows a high surface area promoting an easy analyte immobilization (194). Glucose is a monosaccharide, metabolized by glycolysis to produce energy (30) in the form of adenosine triphosphate (ATP) molecules, through oxidation of carbon bonds (231). Several studies reported a positive correlation between some types of cancer (e.g. gastrointestinal, urinary, and reproductive system cancers) and diabetes. Epidemiological data suggest that people with high levels of glucose have an increased risk to develop BC (232, 233) and BC patients with higher glucose levels have poor prognosis (234). Also, it is known that diabetic patients diagnosed with cancer have lower survival rates when compared with people with a normal glucose metabolism (235). Barone et al. found that the coexistence of diabetes with cancer increases the mortality in about 40% (236). High glucose levels, insulin resistance, hyperinsulinemia and obesity are common conditions in diabetics that can increase the risk of neoplastic transformation and progression of pre-existing cancer, as they are associated with chronic inflammation and injury to the immune system. Hyperglycaemia in cancer is explained by the need of cancer cells to maintain continuous and uncontrolled proliferation, for which they require high levels of energy and substrates. Tumour cells are able to increase glucose levels by improving the expression of glucose membrane transporters (30, 235). This excess of glucose leads to the formation and accumulation of glycation end-products in the cells which promotes cell damage leading to the activation of inflammation processes in the cells, and activation of immune cells as macrophages and neutrophils, thus increasing the production of oxygen free radicals. However, glucose metabolism is differentially regulated in cancer cells and in normal cells (235, 237). Normal cells, produce energy through mitochondrial oxidative phosphorylation, in aerobic conditions. However, when oxygen is not available, cells obtain energy through glycolysis (238). In 1920, Otto Warburg discovered that cancer cells have a different metabolism to normal cells, as they produce ATP more efficiently under aerobic conditions through glycolysis, a phenomenon also known as aerobic glycolysis. The Warburg effect is thus defined by the fact that cancer cells produce a lot of lactate, despite the accessibility of oxygen (239). While in normal healthy cells glucose is fermented and converted to pyruvate to form ATP with the participation of oxygen in the Krebs cycle, cancer cells convert pyruvates to lactic acid and use glucose to synthesise DNA, RNA, proteins and lipids to
Nanocellulose-based biosensor for colorimetric detection of glucose 63 maintain their proliferation (30, 235, 240). The increased production of lactic acid and accumulation of H+ ions in cancer cells drops the pH of the extracellular matrix, thus leading to the death of normal cells and increasing collagenases activity, facilitating tumour cells migration (30, 231, 235, 237). As Warburg effect is considered a hallmark of cancer, glucose can be a promising cancer biomarker and inhibition of glycolysis could be an interesting alternative strategy for cancer treatment (30, 240). Several LFAs have been widely used for glucose analysis in urine (241). However, this procedure is rather complex, because it uses a porous membrane with specific antibodies or proteins immobilised in lines (5). Besides, LFAs show other concerns, as it is based in a “one-step” assay, in which the sensory surface is not easily washed and, consequently, may suffer from interference sample components that pre-block the strips. Beyond that, these assays demonstrate a qualitative and semiquantitative nature (241). In addition, sometimes it is necessary to label the antibody to increase the sensitivity, losing the one-step concept, becoming a more complicated and expensive assay, especially accounting the needs of a very selective antibody. A similar tool to LFA using cellulose as a support material is the well-known dipstick based sensing system (242). When the dipstick gets in touch with the sample (urine or other physiological fluid), a color change in the stick is generated (168, 242-244). This method is simple but offer a response of semi-quantitative nature, thereby limiting the accuracy of the analytical data generated. Some REs have also been employed in PADs, including enzymes (245). Enzymatic biosensors use enzymes as biological RE (246-250) and offer highly selective responses. In the case of preparing of enzyme-based biosensor, it is essential to ensure that enzymes are available to catalyse the intended reaction and must be stable under the normal reaction conditions of the biosensor (247, 248). Several enzymatic assays, with different enzymes immobilised on the cellulose paper surface and aiming to detect different analytes have been reported in the literature. These include glucose (245), lactate (251), stearate (252), catechol (253), phenol (254), among others. The main issue concerning the preparation of enzymatic based-cellulose sensors is related to the enzyme immobilisation on paper substrates. Several techniques have been reported in the literature for this purpose (255), mainly based on adsorption (256) and covalent attachment (257, 258). Adsorption is a simple methodology but hinders the typical binding capacity of enzymes for being immobilised randomly on a solid cellulose support. In addition, the enzymes are sensitive to harsh conditions as pH and temperature, leading to some concerns in terms of sensor reliability. Other methodologies include chemical cross-linking (259), thin film entrapment (255), and microencapsulation (260). These increase the effectiveness of enzyme immobilisation but may limit
Nanocellulose-based biosensor for colorimetric detection of glucose 64 the accessibility of the substrate and increase the complexity of the assay development and associated costs. This work combines the TEMPO-oxidation of MCC to produce carboxyl-NC and their application in the development of colorimetric based test-strips for glucose. The MCC oxidation is optimised, as well as the integration of the GOx within the cellulose/NC-based substrate, and the ability to generate color in the presence of H2O2. Overall, this works reports for the first time the integration of NC with enzymatic biosensor, for glucose detection in urine samples in diabetes. 3.2. Experimental section 3.2.1. Reagents Microcrystalline cellulose was obtained from Biochem Chemopharma. TEMPO (2,2,6,6-tetramethyl piperidine-1-oxyl radical), ABTS (98%), EDC (99%), MES monohydrate (2-( N -morpholino) ethanesulfonic acid monohydrate, C6H13NO4SH2O, 99%) and GOx HPS 300 (activity 260.3 U mg -1) from SCKISUI and uric acid were supplied by Sigma Aldrich. Sodium hydrogen carbonate (NaHCO3, > 99%), sodium chloride (NaCl), hydrochloric acid (0.5 M) and hydrochloric acid (37%) were purchased from Panreac. Sodium hypochlorite solution (NaClO, ca. 10% active chlorine, 15% solution) was obtained from Carlo Erba, sodium carbonate (Na2CO3) and L-ascorbic acid were supplied by Riedel-de-Haen. Glucose (dextrose monohydrate, C6H12O6H2O) Alfa Aesar. Sodium hydroxide (NaOH, solid pellets) was purchased from Eka. NHS (> 97%), HRP (150 U g -1) and creatinine were obtained from Fluka. Urea was supplied by Fragon. Ethanol (96%) was purchased from José Manuel Gomes dos Santos. PBS tablets were obtained from Amresco, dissolved in MiliQ water and pH was changed to 7.2. Purified Milli-Q water was used for all the experiments and analysis. All chemicals were used without any prior purification. 3.2.2. Pre-treatment of microcrystalline cellulose MCC was pre-treated according to literature (206). Two different protocols were used for this purpose: 1 g MCC was added to 13 mL HCl (37%), stirred and hydrolysed at 45 °C for 30 min or 1 g MCC was added to 10 mL HCl (37%) and hydrolysed at 100 °C for 15 min. After acid hydrolysis, the dispersion was washed and centrifuged three times at 12000 rpm. The last wash was carried out in a dialysis membrane with distilled water until a neutral pH was reached. The supernatant solution was sonicated for 15 min to obtain NC. The resulting MCC residue was filtered. NC was precipitated in ethanol, filtered, and dried in an oven at 60 °C.
Nanocellulose-based biosensor for colorimetric detection of glucose 65 3.2.3. TEMPO-oxidation of nanocrystals TEMPO-mediated experiments were carried out as previously published with minor modifications (205) and the pictures of the resulting materials at intermediate stages are shown in Figure 3-1. The first oxidation was carried out by dispersing 1 g of MCC in carbonate buffer solution (75 mL, pH = 10.83) in a sonicator (15 min) and adding TEMPO (30 mg) and KBr (0.32 g) to this suspension, which was kept at 30 °C. The sodium hypochlorite solution (15%, 6 mL) was then added to the resulting suspension, which was mechanically stirred. The pH was kept at 10 to prevent strong degradation of the water-soluble polymer. After 5 h of stirring, the reaction was terminated by adding 20 mL of ethanol, which reacted with the remaining TEMPO. The reaction mixture was acidified with 0.5 M HCl to pH 3 (to remove K+ cations) and then centrifuged to remove the remaining insoluble microcrystalline material. The water-soluble NC in the supernatant was precipitated by adding an excess of ethanol (up to 400 mL), followed by centrifugation. The precipitate was washed with ethanol, centrifuged several times and finally dried in an oven at 45 °C. The remaining MCC was washed with ethanol and dried in the oven at 45 °C and served as starting material for the next oxidations. The second and the third oxidation procedures consisted in repeating the previous experiments. 3.2.4. Nanomaterial characterisation 3.2.4.1. Conductometry The carboxyl content of oxidized NC samples was determined by conductometric titration (205). Dried NC samples (30-40 mg) were resuspended in 15 mL of 0.01 M HCl solution. After 10 min of sonication and stirring, the suspension was titrated with 0.01 M NaOH. According to MCC TEMPO oxidation carboxyl-NC Figure 3-1. Synthesis of the carboxyl-NC by TEMPO oxidation of MCC.
Nanocellulose-based biosensor for colorimetric detection of glucose 66 (205), the content of carboxyl groups in the material was expressed in the form of degree of oxidation (DO), as given by the following equation: 𝐷𝑂 =162×𝐶×(𝑉2−𝑉1) 𝑤−36×𝐶×(𝑉2−𝑉1) (Equation 1) where V1 and V2 are the amount of NaOH (in L), c is the NaOH concentration (mol L -1), and w is the weight of dried sample (g). The value of 36 corresponds to the difference between the molecular weight of an anydroglucose unit (AGU) and that of the sodium salt of a glucuronic acid moiety. 3.2.4.2. Fourier-transform infrared spectroscopy The NC and the MCC were analysed via FTIR spectroscopy to evaluate the chemical modifications. The resulting spectra were used to determine the structural characteristics of the NC. Infrared spectra were recorded on a FTIR Thermo Scientific spectrometer (Nicolet iS10) with an ATR (attenuated total reflectance) accessory, having a diamond crystal. Spectra were analysed from 400 to 4000 cm-1 wavenumber, with a 2 cm-1 resolution, and an accumulation of 150 scans, after background collection. 3.2.4.3. Transmission electron microscopy Dried oxidized NC was dissolved in water (0.01% w/v) and sonicated for 15 min. A drop of this suspension was deposited on the electron microscope grid and negatively stained with phosphotungstic acid for 10s. The excess of liquid was removed by a filter paper. The grids were observed in a TEM Zeiss, Model EM902 A. 3.2.5. Colorimetric assay 3.2.5.1. Binding of glucose oxidase to the nanocellulose 3.2.5.1.1. Glucose oxidase adsorption An oxidized carboxyl-NC (20 mg) was dispersed in 10 mL of 50 mM MES buffer with 500 mM NaCl buffer (pH 5.0) and kept stirring for 24 h at room temperature, before use. Then, 2.5 μL of carboxyl-cellulose was casted on the cellulose surface. After, a solution of GOx (2.0 mg mL -
Nanocellulose-based biosensor for colorimetric detection of glucose 67 1) previously dissolved in PBS, was drop-casted on the modified cellulose paper with the NC. The sensor was let dry at room temperature and stored in the fridge at 4°C before use. 3.2.5.1.2. Covalent attachment of glucose oxidase The carboxyl groups of the carboxyl-NC were activated using EDC/NHS. To this end, 12 mg of NHS (10 mM) and 8 mg of EDC (4 mM) were added into the well-dispersed mixture of carboxylNC (20 mg) in 10 mL of 50 mM MES and 500 mM NaCl buffer (pH 5). The solution was stirred for 24 h at room temperature. The excess of reactants was removed by precipitating of the activated carboxyl-NC in ethanol. The precipitate was centrifuged, cleaned with ethanol and oven-dried to eliminate ethanol residues. The activated material was dispersed in 12 mL of PBS (pH = 7.2) containing 20 mg of GOx. This dispersion was stirred 24 h at room temperature. Then, 2.5 μL of the previous solution was drop-casted in the cellulose paper. 3.2.5.2. Colorimetric assay 3.2.5.2.1. Carboxyl-nanocellulose and glucose oxidase concentration optimisation For glucose detection, oxidized-NC (0.01-20 mg mL -1, 2.5 μL) was adsorbed on a cellulose paper substrate. The next stage consisted of the immobilisation of GOx. The concentration of the enzyme was studied within 0.001 and 0.5 mg mL -1, by using 2.5 μL of the enzyme solution. For this purpose, the enzyme was casted on the cellulose/carboxyl-NC surface and let dry at room temperature. Then, 22.5 μL of a solution consisting of different concentration of glucose (0.001 to 100 mM, 1.2 μL), PBS (pH = 7.2, 18.9 μL), ABTS (5 mM, 1.2 μL) and HRP (150U g -1 solid, 1.2 μL), was dropped on the filter paper substrate with, containing the carboxyl-NC (261) (Figure 3-2). The color of each test-strip was monitored for periods of 3, 5 and 10 min of incubation. Resulting color was compared with color of the blank sample containing only glucose without enzyme. Pictures of the results were obtained using a smartphone camera. However, light conditions highly influence quality and reproducibility of the acquired images and thus the feasibility of results. Thus, the pictures were taken in controlled light conditions and a fixed focal distance. Image J software (version 1.4.3.67) have been used to analyse the results. A square tool has been used to select a constant area for the measurements. Several parameters were analysed
Nanocellulose-based biosensor for colorimetric detection of glucose 68 as RGB. The average of RGB channel image provided the best results in terms of linearity when plotted against glucose concentration and was used for analysis. 3.2.6. Selectivity assay The selectivity of the system is a crucial parameter for the analysing system when it is expected to be applied under real conditions. For this reason, the selectivity of the sensing system was further evaluated by incubating different interfering molecules, present in urine samples, onto the teststrips. Glucose (19.2 mM) plus ascorbic acid (0.01 M), glucose (19.2 mM) plus acid uric (0.06 mg mL -1), glucose (19.2 mM) plus creatinine (12 mM) and glucose (19.2 mM) plus urea (0.2 mg mL -1). Assays were prepared in PBS and dropped on the filter paper previously described in the section 3.2.5. Colorimetric assay. Images of the results were captured, and ImageJ was used to obtain the parameter values and calculate the associated errors. 3.3. Results and discussion 3.3.1. The oxidation of microcrystalline cellulose The oxidation of raw cellulose by means of the TEMPO reaction is often incomplete. On the other hand, pre-treated cellulose can give larger amount of totally oxidized water-soluble polyglucuronans (206). Thus, the oxidation of MCC by TEMPO reaction may be more effective by establishing sequential oxidative reactions. In general, at least two oxidation stages were applied herein, in Figure 3-2. Test-strip based colorimetric assay produced by casting on the cellulose substrates the indicated solutions, and binding GOx either by adsorption (A) or by covalent bonding (B).
Nanocellulose-based biosensor for colorimetric detection of glucose 75 3.3.3.3. Sensing system bound by adsorption to the oxidized nanocellulose While the chemical system for glucose detection is well-known, information about the effect of the carboxyl-NC on the color gradients or concentration ranges of detection is missing. In general, it was expected that the presence of carboxylate groups on a cellulose substrate would enhance its binding efficiency and create a suitable environment for GOx activity. In terms of binding efficiency, a regular cellulose substrate would establish interactions with GOx via hydrogen bridges only. When cellulose is doped with carboxyl-NC these interactions could be enhanced by the presence of multiple –COOH groups, which would intensify the hydrogen bridges and establish ionic interactions with multiple positive points existing in the external surface of GOx. Thus, different tests-trips were prepared by casting different amounts of NCCOOH on the cellulose substrate, ranging from 0.010 to 20 mg mL -1 and let to dry after casting. These studies were made by keeping the previously defined concentrations, having 2.0 mg mL -1 GOx and 3 min for the reaction to take place. Figure 3-8 summarizes the results obtained from different concentrations of glucose, ranging from 1.5 to 25.6 mM. When compared to the results obtained without carboxyl-NC the observed colors were more homogenously distributed along the whole surface of the test-strip and the gradient color change was improved by the presence of the carboxyl-NC. Overall, this confirmed that the presence of carboxyl-NC affected positively the results obtained by improving the color features of the glucose detection. However, there was not a great difference among the different test-strips prepared with increasing amounts of carboxyl-NC. This was probably linked to the saturation of carboxyl-NC Reaction time Glucose concentration → GOx concentration Figure 3-7. Dependence of concentration of GOx and an amount of glucose with the time.
Nanocellulose-based biosensor for colorimetric detection of glucose 76 adsorbed on the cellulose support for the selected concentration range. Thus, an intermedium concentration of 5 mg mL -1 was selected for subsequent studies, considering reproducibility and cost purposes. This would ensure highly reproducible strips at a lower cost than those employing the highest concentration tested. Comparing the analytical performance of test-strips prepared without carboxyl-NC and with 5 mg mL -1 carboxyl-NC a linear trend was observed in both by plotting (R+G+B)/3 against glucose concentration. The presence of carboxyl-NC yielded improvements in terms of lower limit of linear range, decreasing the observed value from 6.4 to 1.5 mM. Moreover, the slope increased in about 50% by the presence of carboxyl-NC, increasing the slope value from 1.0 to 1.5 a.u/mM. Additionally, the operational features in terms of lower limit of linear demonstrated improvements starting 1.5 mM with NC and 6.4 mM for the biosensor without NC. Overall, these results confirmed that the loading of GOx on the test-strips and its catalytic activity were more controlled by the presence of carboxyl-NC. 3.3.4. Sensing system covalently bound to the oxidized nanocellulose The possibility of attaching covalently GOx to the test-strip was also explored herein, aiming to increase the stability of the final device. This was made by employing EDC/NHS chemistry. In this, the carboxylate groups in the carboxyl-NC matrix were activated and underwent subsequent Figure 3-8. Digital images of the test-strips in the presence of different glucose concentrations using ABTS as colorimetric indicator, prepared with different concentrations of carboxyl-NC (left) and the analytical calibration curves plotting the color coordinates collected against the glucose concentration (right, the carboxyl-NC assay corresponds to 5 mg mL -1 of NC-COOH). Glucose, mM 1.5 3.2 6.4 12.8 19.2 25.6 0.01 [NC-COOH], mg/mL 0.1 1.0 5.0 10.0 20.0 Without y = -1,51x + 237,82 R² = 0,9853 y = -1,0312x + 224,5 R² = 0,993 190 200 210 220 230 240 0 5 10 15 20 25 30 (R+G+B)/3 [Glucose], mM with NC-COOH Without NC-COOH 240 230 220 210 200 190 (R+G+B)/3
Nanocellulose-based biosensor for colorimetric detection of glucose 77 covalent binding to the amine groups exposed in the outer surface of GOx. The concentration of GOx bound to the activated carboxylic groups ranged from 0.01 to 20 mg mL -1. The results obtained are shown in Figure 3-9. Comparing with the adsorption assay, covalent immobilisation showed much more intense colors that lead to much more sensitive readings. The slope of the 3.5-times higher, increasing from 1.5 to 5.3 a.u/mM. As expected, this increase in sensitivity was also linked to a narrower concentration range of linear response, ranging from 1.5 and 12.8 mM. Overall, the method described herein is an expeditious and low-cost approach that may be further explored in the analysis of glucose in urine. Normal levels may reach up to 150 mg L -1, meaning that normal urine samples may require 10× dilution prior to analysis with the test-strip. Moreover, urine samples shall not have intrinsic color that may interfere with this determination. 3.3.5. Selectivity assay Ascorbic acid, uric acid, creatinine, and urea were incubated onto the paper substrate simultaneously with glucose and compared with glucose incubated alone. It can be seen in Figure 3-10 that the response of the paper test-strip was not affected by other interferents as the variation of the tested interfering species is low (less than 10%) comparing to glucose value. Ascorbic acid and uric acid showed a slightly positive variation (4.6% and 5.3%, respectively) whereas creatinine and urea revealed a low negative variation (8.3% and 9.9%, respectively), over the glucose value. Glucose, mM 1.5 3.2 6.4 12.8 19.2 25.6 0.01 [GOx], mg/mL 1.0 10 50 y = -1.51x + 237.82 R² = 0.9853 y = -5.26x + 238.04 R² = 0.9898 120 140 160 180 200 220 240 260 0 5 10 15 20 25 30 (R+G+B)/3 [Glucose], mM Adsorption Covalent Binding Figure 3-9. Digital images of the test-strips in the presence of different glucose concentrations using ABTS as colorimetric indicator, prepared with different concentrations of GOx covalently bound (left) and the analytical calibration curves plotting the color coordinates collected against the glucose concentration (right, the carboxyl-NC assay corresponds to 1 mg mL -1 GOx).
Nanocellulose-based biosensor for colorimetric detection of glucose 78 3.4. Conclusions This work reports on the oxidative modification of MCC to produce water-soluble cellulose derivatives that can improve the use of test-strips using cellulose substrates as carriers. Different conditions were applied. Overall, it was possible to achieve better properties of the cellulose substrate material by using an appropriate ratio of TEMPO/hypochlorite and an appropriate number of successive oxidative procedures. The use of carboxyl-NC to improve cellulose substrates requiring enzyme binding was further tested. It was found that modified NC exhibited better color distribution and improved the process of glucose detection in terms of analytical performance. Overall, the proposed enzymatic test-strip showed good characteristics in terms of simplicity, reaction time, price and applicability and is a promising tool for PoC analysis. In addition, this device opens up possibilities for multiplex analysis by using different chromogenic reagents. By utilizing the inherent capability of the H2O2 producing oxidase, it can be extended to the detection of a wide range of analytes of interest in the food, health, and environmental fields. Figure 3-10. Evaluation of interfering species variation in comparison with glucose.
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 79 Chapter 4 Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care This chapter focused on the development of a sensor by adapting the standard ELISA protocol to a paper substrate. CA15-3 was selected as target as it is overexpressed in 90% of BCs, being one of the biomarkers approved BC diagnosis, as well as follow-up of cancer patients during treatment. Detection signal results from a colorimetric reaction based on the oxidation of the TMB substrate by a peroxidase enzyme. Part of the content of this chapter was published in Microchemical Journal: C.C.G. Carneiro, M., Rodrigues, L. R., Moreira, F. T. C., & Goreti F. Sales, M. (2022). Paper-based ELISA for fast CA15–3 detection in point-of-care. Microchemical Journal, 181, 107756. doi:10.1016/j.microc.2022.107756. Part of this chapter was presented as a poster in 31st anniversary world congress on biosensors, Live and On-demand: Carneiro, M. C. C. G., Rodrigues, L. R., Moreira, F., & Sales, M. G. F. (26-29 july, 2021). Poster P5.054: Paper-based ELISA for rapid protein detection.
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 80 4. Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 4.1. Introduction CA15-3 is a transmembrane glycoprotein that is overexpressed in 90% of BCs (37) and is one of the biomarker approved not only for BC diagnosis, but also for follow-up of cancer patients during treatment (23, 59, 92). ELISA is the most used technique for detection of CA15-3 in serum but this methodology involves long incubation times and multiple washing steps (265). Regarding this, new screening devices with rapid response and high sensitivity and selectivity such as biosensors are needed (266). Biosensors meet the requirements for PoC analysis as they are easy to use, fast responding, have high sensitivity and specificity and offer the possibility of multiplex detection (9, 26). There are several biosensors reported for CA15-3, involving equipment-based transduction (10, 23, 121). Colorimetric assays can provide instrument-free PoC analysis (41) being particularly attractive for environmental and cost reasons when cellulose paper is used as a substrate (155). PADs have therefore experienced exponential growth in the last decade (27, 81) and are suitable alternatives for cancer diagnosis compared with actual methods (267). A quantitative result can be provided by colorimetric PADs by using simple readout devices as smartphones to capture the results (1, 24, 41, 124, 125, 159, 161, 162, 268), which are then analysed using adequate software to calculate optical parameters using as RGB or HSB coordinates that can be used for biomarker quantification (159). Scientists tried to overcome the disadvantages of the traditional ELISA by replacing the conventional 96-well plate with cellulose paper substrates and developing the P-ELISA which can be used as a PoC device, being particularly important in resource-poor settings (94, 185, 186). It takes advantage of the high specificity of conventional ELISA, but in a fast procedure performed in less than 1h and requiring small amounts of samples and reagents (1-10 µL), which lowers the cost of the test. There are several reports of colorimetric PADs for various cancer biomarkers, not only proteins (CEA (19, 27, 129, 269, 270), p16 (271), HPV 16/18 E6 oncoprotein (272), cytochrome c (273)) but also for nucleic acids (miRNA-21 (171)) and other molecules (citrate (140)). Most of them are based on immunoassays that include the enzymatic reaction of HPR with TMB as chromogenic substrate (142, 182). As far as we know, this method has never been reported for the detection CA15-3 in paper substrate. This paper reports on the development of a colorimetric P-ELISA for CA15-3. Based on a sandwich
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 81 immunoassay on a cellulose substrate, which was modified and optimised accordingly. The detection is performed by a colorimetric reaction based on the oxidation of the TMB substrate by a peroxidase enzyme, HRP, in the presence of H2O2. The color development on the paper substrate can be visualized and recorded in the ImageJ software for further analysis, allowing the concentration of the analyte of interest to be determined. The characterisation and application of the test-strip is also presented. 4.2. Experimental section 4.2.1. Reagents and solutions All chemicals were of analytical grade and water was ultrapure Milli-Q laboratory grade. Sodium phosphate dibasic dihydrate was acquired from Panreac. Sodium dihydrogen phosphate dihydrate was from Scharlau. Sodium Hydroxide was obtained from EKA. Potassium periodate was from May&Baker. APTES was obtained from Acros Organics. Glucose monohydrate was purchased from Alfa Aesar. Bradford reagent was from BioRad. TMB Liquid Substrate System for ELISA and BSA were purchased from Sigma Aldrich. Tween20 was acquired from Merck. Cormay® serum HN was from PZ Cormay®. CEA was obtained from EastCostBio. CA125 was from Hytest. CA15-3 from host human (reference MBS536585) was purchased from MyBioSource. Capture antibody (Mucin 1 monoclonal antibody Vu-2G7, reference SC-69644) was acquired from Santa Cruz Biotechnology and detection antibody (Recombinant monoclonal antibody to MUC1 labelled with HRP, reference EPR1023) was from Abcam. Stock solutions of proteins and antibodies were prepared in phosphate buffer (PB), having 0.081 M Na2HPO4 and 0.019 M NaH2PO4, with pH 7.7. Whatman® quantitative filter paper (ashless, Grade 40, 110 mm diameter, 210 μm thickness, 8 μm pore size) was used as substrate. 4.2.2. Apparatus FTIR and thermogravimetric analysis (TGA) were used to characterise the cellulose paper at consecutive stages of sensor assembly. FTIR spectra were recorded using a Nicolet iS10 spectrometer from Thermo Scientific, coupled to an ATR accessory with a diamond crystal. All spectra were acquired after background correction from 400 to 4000 cm-1 with a resolution of 8 cm-1 and 250 scans. Data analysis was performed using OMNIC 9 software. TGA measurements were made on and Hitachi TGA DTA/7200, using ≈5.4 mg paper samples in an aluminum holder at a heating rate of 5⁰C per second, from 30 to 500⁰C, under a nitrogen atmosphere at 300 mL min -1. A homemade dark box (Figure 4-1A and Figure 4-1B) was used for picture collection. A
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 82 cardboard box (19 cm width × 13.5 cm height × 13.5 cm depth) was painted inside and outside with a black spray ink. An aperture (2 cm × 1 cm) was made on the top of the box where a smartphone lens exactly fits for image acquisition. A light-emitting diode (LED) tape (high luminosity, 4000K, 50/60 Hz, DC12V) was stacked on the inside, covering all the four faces of the box. 4.2.3. Paper pre-treatment: washing and functionalisation Whatman® filter paper circles (area ≈ 95 cm2) were soaked in 8% NaOH solution in a horizontal shaker for 1 h to remove contaminants. They were then washed several times with ultrapure water to remove the NaOH content. The pH of the washed solution was measured with an indicator paper until a neutral value was reached. The paper circles were then dried in an oven at 60 °C. Figure 4-1. Outside (A) and inside (B) view of the dark box for image acquisition. Steps of the sensor construction (C). Covalent immobilisation of capture antibody (a), blocking step with BSA (b), incubation with CA15-3 antigen (c), incubation of detection antibody labelled with HRP (d) and color development with TMB solution (e). B A C a b c d e
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 83 The cellulose substrate was first modified to allow subsequent antibody binding (159). For this purpose, silanization with APTES and oxidation with potassium periodate was tested. Silanization was performed by immersing the paper in a 10% APTES (230) solution in ethanol:water (95:5, v/v) and shaking horizontally for 2 h at room temperature. The paper was then washed with fresh ethanol:water solution to remove unbound silanes and thermally treated in an oven at 80 °C for 2 h. Oxidation of cellulose was carried out with periodate functionalisation, as described in (86). Potassium periodate was employed in a solution of 3.1×10-3 M prepared in water (protected from light by aluminium foil and prepared daily). The filter paper was immersed in 50 mL of potassium periodate and allowed to react at 65 °C for 2 h to produce aldehyde functions allowing covalent binding of an antibody. After this reaction, the filter paper was washed twice with water and dried at 60 °C. The chemical changes on the cellulose surface were then monitored by FTIR. To determine which functionalisation (APTES or KIO4) provided the best binding to proteins/antibodies, the functionalised paper was incubated with 1% BSA for 30 min, washed with 1 mL PB, and subjected to Bradford reagent to detect the bound amino groups (-NH2) of BSA on the cellulose-oxidized surface. 4.2.4. Assembly of the paper-based sandwich enzyme-linked immunosorbent assay Whatman® paper was cut into small circles of 8 mm diameter with a hole puncher. The minimum volume of reagent solution required to ensure a uniform and complete coverage of the entire paper circles was investigated using red food coloring. A sandwich ELISA is the principle of the assay used in this work and each step is represented in Figure 4-1C. For covalent immobilisation of capture antibody on the aldehyde-modified paper circle, a solution containing monoclonal antibodies (3 µL, 161 µg mL -1) against CA15-3 antigen was added and let to incubate at 37 °C in the oven for 10 min. Then, any remaining aldehyde functions and nonspecific binding regions of the antibodies were blocked by addition of BSA solution (3 µL, 1%). After incubation for 10 min at 37 °C, a washing and a drying step were followed. At this stage, the paper circles were ready to detect CA15-3.
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 84 After each incubation, a washing step was performed with 1 mL of PB. Washing steps between the incubation steps are crucial to remove unbound material from the sensor surface, which could increase background noise. Despite washing, the drying steps proved to be essential in this assay. For this reason, washing and drying steps were carried, after each incubation step and before the next one, as follows. Washing buffer (PB, 1 mL) was allowed to pass through the circle paper with the aid of a Pasteur pipette. An auxiliary filter paper was positioned under the reaction paper circle to absorb the excess of washing buffer and the paper was allowed to dry at 37 °C in the oven for 15 min. At this stage, the paper circles were ready to detect CA15-3. 4.2.5. Detection of cancer antigen 15-3 The detection ability of the test-strips was tested by incubating CA15-3 antigen standard solutions with different concentrations (from 2 to 2000 U mL -1) prepared in PB. The control was made by incubating only PB solution and this was considered the background signal. After finishing the incubation time, the test-strips were washed to remove unbounded target, and dried. Antigen binding was detected by adding HRP-labelled antibody (5 µg mL -1) to the test-strips. It recognises the CA15-3 previously bound to the capture Ab and forms a sandwich structure. After this interaction, the paper was washed and allowed to dry. Color development was achieved by pouring 5 µL of a ready-to-use redox indicator solution of TMB onto each reaction paper. A multichannel micropipette was used for this step to ensure the same reaction time for all teststrips. A color change from colorless to blue immediately occurred on the paper zone and images were immediately captured in the dark box (Figure 4-1A and Figure 4-1B). 4.2.6. Quantitative data and sample analysis Quantitative data were obtained by capturing images with a smartphone and analysing them with ImageJ software. The coordinates of the RGB and HSB color systems were considered with the aim of obtaining a linear trend as a function of concentration. HSB system was chosen to optimally represent the absence or presence of the protein proportional to its concentration. The performance of this paper-based assay was evaluated in Human Cormay® serum. Thus, various concentrations of CA15-3 (from 2 to 2000 U mL -1) were added into Cormay® serum aliquots using a 100-fold dilution of serum.
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 91 difference. The color of the blank sample was very light and appeared greenish, while the color of the positive sample was a clear blue. This condition was therefore selected for further testing. A B Figure 4-5. Photographs (A) and respective bar charts (B) of saturation values from capture antibody concentration optimisation with 1.61, 16.1 and 161 μg mL -1 . B A Figure 4-6. Photographs (A) and respective bar charts (B) of saturation values from detection antibody concentration optimisation with 0.5, 5 and 50-μg mL -1.
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 92 4.3.2.4. Blocking agent concentration BSA is frequently used in ELISA as a blocking agent to prevent non-specific bindings of other molecules (e.g. proteins, antibodies) to the surface (282). Absence and presence of BSA as blocking buffer was evaluated in three different concentrations (0.1, 1 and 10%). Results are shown in Figure 4-7. Increasing concentrations of BSA provided higher background signals in negative controls. It has already been reported in a previous work that some BSA formulations (including globulin or endotoxin) used for blocking step cause high background signals (282). Despite that, it was also reported that some antibodies used in ELISA can cross-react with BSA (283). Nevertheless, due to the occurrence of a considerable background signal at controls, possibly resulting from the facts described before (interference from BSA formulation and/or antibodies cross-reaction), 1% of BSA was selected as the best concentration for the blocking step since it provided a more pronounced difference between the control and a positive result. A B Figure 4-7. Photographs (A) and respective bar charts (B) of saturation values from blocking agent concentration optimisation with 0, 0.1, 1 and 10% of BSA.
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 93 4.3.2.5. Number of washing steps As already stated, the washing step in ELISA assays is crucial to reduce background signals provided from unspecific bindings. Besides the composition of the washing buffer, already optimised, the number of washing steps (no washing versus one washing step with 1000 μL of PB versus three washing steps with 1000 μL of PB) was also tested (Figure 4-8). No washing results in highly colored papers due to background signal from unbounded or unreacted reagents and a small difference between the positive result and the negative control. One washing step led to a higher difference between the papers and was proved to be efficient in the removal of unbounded reagents. Finally, three washing steps showed a small difference between papers, when compared with only washing step, and leads to an increase in HSB values of negative control. This may be due to the fact of the excessive wash can promote the removal of more capture antibody, thus promoting nonspecific bindings and increasing the background signal. Therefore, only one washing step with 1000-μL of PB was selected for further assays. A B Figure 4-8. Photographs (A) and respective bar charts (B) of saturation values from number of washing steps optimisation, including no washing, one washing step with 1000-μ L of PB and three washing steps with 1000-μL of PB.
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 94 4.3.2.6. Incubation and dry temperature Temperature of incubation and drying steps have been mentioned as important factors in the assay performance. Incubation and drying at room temperature (21 °C) were tested against 37 °C (Figure 4-9). Room temperature provided very colorful papers but seemed to promote unspecific response in the negative control as it shows higher HSB values, especially on saturation parameter, compared to the positive result. Thus, 37 °C was selected as the best condition. 4.3.2.7. Effect of functionalisation The effectiveness of the functionalisation step with KIO4 was already showed through the characterisation by FTIR, TGA and Bradford assay. However, it is important to assess if this step is or not essential to the performance of the assay. Results (Figure 4-10) showed that paper without functionalisation exhibits no significant differences between control and positive result, hence justifying the need to functionalise to assure a proper capture antibody mobilization and further recognition of the target protein. B A Figure 4-9. Photographs (A) and respective bar charts (B) of saturation values from incubation and dry temperature optimisation with 21 ºC and 37 ºC.
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 95 4.3.3. Main analytical features The intensity of the blue color of oxidized TMB depends on the amount of peroxidase enzyme catalysing the reaction. This is determined by the amount of HRP-labelled antibody bound to the target analyte. When CA15-3 is present in the sample, the HRP coupled to the detecting antibody allows the colorless TMB to oxidize to TMB+, resulting in a blue color complex that is visible to the naked eye. Therefore, the intensity of the blue color of the oxidized TMB is proportional to the CA15-3 present in the sample. It is worth noting that the background signals for blank paper circle could come from residual HRPmarked Ab that remain on the sensor surface. This is because the blocking agent could not completely block the reactive sites, making these sites available for the HRP-labelled antibody. Moreover, the HRP-labelled antibody used as the detection antibody could also bind to the capture antibody. These results are consistent with previous paper-based ELISA assays using colorimetric detection, which also showed color development in negative samples (269). Figure 4-11A displays the sensor color change with increasing concentrations of CA15-3 (from 2 to 1100 U mL -1) after HRP reaction with TMB. The data obtained showed a linear trend in hue B A Figure 4-10. Photographs (A) and respective bar charts (B) of saturation values from papers without functionalisation and with KIO4 functionalisation.
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 96 values against logarithm concentration, corresponding to Hue = 3684.3 × log(CA15-3, U mL -1) + 125098m with R-squared > 0.997 (Figure 4-11B). Each data point represents an average of three independent experiments, and the error bars indicate the standard deviation (SD). 4.3.3.1. Spiked serum samples Figure 4-11C demonstrates the sensor response for the detection of CA15-3 in Cormay® serum. It showed a linear trend in hue values against logarithm concentration, corresponding to Hue = 4878 × log(CA15-3, U mL -1) + 121236 (R-squared > 0.994). Compared with the experiments in PB, we obtained similar linearity but a narrower linear range (from 2 to 200 U mL -1). The assay was performed in triplicate. Each data point represents the average of the assays, and the error bars indicate SD. Thus, it was confirmed that this P-ELISA can be used to detect CA153 in human serum with good performance. 4.3.3.2. Selectivity assay Glucose, CEA and CA125 were incubated simultaneously with CA15-3 on the paper substrate and compared with CA15-3 incubated alone. It can be seen from Figure 4-12 that the response of the paper test-strip was not significantly affected by other interfering factors, as the deviations of the tested interfering species compared to the CA15-3 value were 1.6%, 7%, 12.1% (all below A B C Figure 4-11. Pictures of the colorimetric sensor showing the color change with increasing concentrations of CA15-3 (from 2 to 1100 U mL -1) prepared in PB, after HRP reaction with TMB (A) and respective calibration curve of hue values extracted from the photographs vs logarithmic concentration (B). Calibration curve of the sensor incubated with a 100-fold dilution of Cormay® serum spiked with CA15-3 (from 2 to 200 U mL -1) (C). Equations from calibration curve are shown.
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 97 15%), respectively. The test was performed in triplicate and the data are presented as mean ± SD. 4.3.3.3. Carboxy-nanocellulose modification of the sensor surface Despite the good performance obtained in the reported sensor, we want to verify if the modification of sensor surface with a carboxy-NC synthetized in a previous work (284) provide any improvement. It is expected that the presence of -COOH groups on cellulose surface provided by carboxyl-NC enhance its binding efficiency to the primary antibody, thus increasing the colorimetric signal. However, results in Figure 4-13 shown that neither one nor three layers of carboxy-NC improved the performance of the sensor as it was not possible to visually distinguish between the different concentrations of protein. This highlighted that each system has its own characteristics, and each sensor needs to be carefully optimised considering each parameter that can affect the sensor performance. Figure 4-12. Selectivity study based on comparison of the response of CA15-3 incubated alone or mixed with interfering species. Photographs (A) and respective bar chart (B) with mean ± SD. CA15-3 (25 U mL -1) ( ▌); glucose (1 mg mL -1) ( ▌) CEA (0.25 ng mL -1) ( ▌); CA125 (0.35 U mL -1) ( ▌) B A
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 98 4.4. Conclusions The biosensing strips described here are a rapid and sensitive method for the detection of cancer proteins on a paper substrate, consisting of a sandwich ELISA method on a surface-modified cellulose filter paper with colorimetric detection based on TMB oxidation. The combination of an inexpensive and biodegradable substrate such as filter paper with simple colorimetric detection makes this sensor suitable for PoC detection of cancer biomarkers such as the CA15-3 protein. In addition, colorimetric detection as a transduction method allowed the qualitative result to be obtained with the naked eye, and data analysis with the ImageJ made it possible to determine the color coordinates and extrapolate the target concentration by comparison with a calibration curve. A good correlation coefficient was obtained with both buffer and human Cormay® serum (R2 >0.99), suggesting that the sensor is suitable for clinical practice. In addition, the sensor proved to be selective for the target analyte as it was not significantly affected by other interfering B A Figure 4-13. Photographs (A) and calibration curve (B) of the effect of sensor functionalisation with carboxy-NC ● one layer of carboxy-NC ◼ three layers of carboxy-NC
Paper-based enzyme-linked immunosorbent assay for fast cancer antigen 15-3 detection in point-of-care 99 molecules. A major disadvantage of colorimetric sensors is their lack of sensitivity and high LODs. In this work, however, CA15-3 could be detected in the clinical range without the need for lower LODs. The entire test (including the construction of the sensor) can be performed in approximately 2 h, which is shorter than the duration of a conventional ELISA. The target can be detected in 35 min (Table 4-1). Table 4-1. Reagent’s volume and reaction time for the detection of CA15-3 using the reported paper-based sensor. Reagents Volume (µL) Time (min) Sample 3 10 Washing buffer 1000 - Dry Step - 15 Detection antibody 3 10 Washing buffer 1000 - Dry Step - 15 Revealing step (TMB) 5 - The short incubation times used for the assay may be due to the high surface-to-volume ratio and porous structure of cellulose. These properties of cellulose also allow the use of a low volume of reagents and samples, which is an advantage of the P-ELISA over the conventional ELISA. Despite the advantages of this simple and rapid detection of CA15-3, there are still costs associated with the use of antibodies as detection elements. A future improvement could therefore be to replace natural antibodies with other synthetic biomimetic materials, such as molecularly imprinted polymers or aptamers.
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 100 Chapter 5 Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout Chapter 5 focused on the replacement of the antibody in paper-based ELISA protocol described in Chapter 4 by a MIP. The target and colorimetric system used were the same of the previous work. Part of the content of this chapter was published in Microchemical Journal: C.C.G. Carneiro, M., Rodrigues, L. R., Moreira, F., & Goreti F. Sales, M.. Paper-based biommimetic teststrip for CA15-3 with colored readout. Microchemical Journal. 2024;196:109640.
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 107 the loss of the polymeric network, that was not firmly attached to the sensor film. In general, the TGA information confirms the formation of the imprinted film and the effects of removal on the modified cellulose substrates, resulting in sensing materials that contain similar a percentage of polymer above the silane/cellulose composite. Figure 5-2. Thermogravimetric analysis of cellulose paper with or without silane modification (A) and the cellulose paper with silane modified with MIP/NIP polymerisation, including stages before and after template removal (B). 5.3.1.2. Scanning electronic microscopy analysis The morphology of the bare and modified cellulose substrates was characterised using SEM. Figure 5-3 shows magnification images of the cellulose paper showing the typical microfibers of cellulose (A), which were also modified with polydopamine in the form of NIP (B) or MIP (C). The images show that the cellulose fibers have been coated with polydopamine during the polymerisation step, resulting in an increase in fiber diameter (B and C) compared to the naked cellulose separator (A). Higher magnification images show the polydopamine fibers and the rough polymeric surface in MIP (F) compared to bare cellulose (D) and to NIP (E), which is still smooth (287). This roughness suggests that the presence of the template on MIP led to a different structural polymer growth. Nevertheless, the embossed voids on MIP cannot be observed because the dimensions of these binding sites are below the resolution limit of the technique.
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 108 Figure 5-3. SEM images of bare paper (A), NIP (B) and MIP (C) with 500× magnification and bare paper (D), NIP (E) and MIP (F) with 15000× magnification. 5.3.2. Synthesis of the biomimetic material The biomimetic material was assembled as shown in Figure 5-1, where a cellulose substrate was modified with silane chemistry, resulting in the interaction of the template with boronic acid and subsequent polymerisation with dopamine to form the MIP layer. An important aspect of any chemosensor is reproducibility, which depends first on the homogeneity of the starting material. Therefore, all paper substrates were washed with H2O2 before any further modification. H2O2 is a strong oxidising agent that is said to oxidise impurities, increasing their solubility in water. It also adds hydroxyl groups to the paper surface, which helps to increase the efficiency of subsequent chemical modifications with involving covalent hydroxyl bonds (as in silane chemistry) (289). Physical adsorption would also be possible as an immobilisation technique for any intended biological RE, but desorption of the molecules may occur during the washing steps. Therefore, cellulose was modified with silane chemistry by incubating the paper in silanes. Binding to the vicinal hydroxyl groups of the cellulose (treated with H2O2) and cross-linking through the porous structure of the cellulose provided stable chemistry of the silane/cellulose matrix (84, 88). Next, 3-APBA was incubated on the silane/cellulose composite followed by the protein (in the case of the MIP), to undergo the reversible reaction between boronic acid (in 3APBA) and the cis-diol groups (in glycoproteins) (118). This stage corresponds to the formation of a template-monomer complex in a pre-polymerisation solution, giving rise to a cyclic bromate ester. This ester is easily
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 109 reversed by a change in pH/ions, releasing the protein as needed. The stability and suitability of this complex is critical, as this stage establishes the functional groups to which CA15-3 will bind after contact with the MIP in the sample. The subsequent addition of a monomer/cross-linker and an initiator allowed the formation of the MIP film (105). In this work, dopamine was considered as the monomer. It is a small molecule containing catechol and amine functional groups that self-polymerise in a weak alkaline pH (295). Its polymerisation is based on the oxidation of catechol group to quinone and further reaction with amine groups and other catechols and quinones, allowing the formation of a biocompatible and hydrophilic polymeric film. This polymer is highly reactive towards amine and thiol groups, which has advantages for covalent binding of target biomolecules for sensing purposes (296). To our knowledge, the first polydopamine-based MIP was reported for nicotine sensing (116), but since then dopamine has been used in the preparation of MIPs as REs for sensing of macromolecules, including proteins (296). Removal of the template was the next step, a critical step in molecular imprinting. Efficient template removal should be performed to create the imprinted sites and allow subsequent rebinding of CA15-3 present in the sample. In addition, residual template molecules should be minimised as they are responsible for background signals (286). Template removal was optimised by incubating NIPs and MIPs under different conditions that affect the reaction between the diol and phenylboronic acid and the conformation of the protein, while the effects on the polymer surface should be minimal. In this work, water (113) or sodium chloride (0.1 M) (292) or acetic acid (10%) and sodium dodecyl sulphate (SDS, 1%) (297) or acetic acid (3%) and Tween-20 (0.1%) (115) were tested, under the same experimental conditions (temperature and time). This study was made by checking how the color of the MIP/NIP sensor paper changes between the control solution and the CA15-3 standard solution of 2000 U mL-1 to check the effect of removal conditions upon the sensitivity of the system. The collected images and the analytical output are shown in Figure 5-4. The results showed that the use of a detergent with acetic acid was undesirable, and the use of water or sodium chloride gave similar results, but with the naked eye, sodium chloride (ΔMIP-NIP= 2.90) was slightly better than water (ΔMIP-NIP=2.58), due to the greater difference between the results of the control and the protein detection.
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 110 5.3.3. Colorimetric protein detection The presence of CA15-3 was detected by a color change from colorless to blue because the incubation of HRP was followed by addition of a ready-to-use peroxidase substrate with TMB. The more intense the blue coloration, the more HRP was bound to the sensor surface. This meant that HRP was bound to CA15-3 by non-specific protein-protein interactions. This also meant that this HRP-biomarker interaction was more intense than any interaction of HRP with polydopamine. B C D E Figure 5-4. Photographs of NIPs and MIPs with different removal solutions (A). Respective bar charts with Q values for removal with water (B), sodium chloride (C), acetic acid + sodium dodecyl sulfate (D) and acetic acid + tween-20 (E).
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 111 Because this interaction was non-specific, a larger amount of HRP would also result in color development as it binds to different sites on the sensor than the protein CA15-3. Therefore, the best experimental conditions should be optimised. This principle was confirmed by testing different conditions for color development, as shown in Figure 5-5. The incubation of a TMB solution in PB on the sensor papers did not lead to any color development, either on the NIP or on the MIP. The same applies to the sensor paper incubated with CA15-3 standard solution (1000 U mL-1), which remained colorless. This shows that the peroxidase substrate is stable with TMB and this substrate is not oxidised in the absence of HRP. As with conventional ELISA methods using antibodies, incubation of an antibody labelled with HRP was also attempted for CA15-3. It showed a good response with a higher value for ΔMIP than for ΔNIP, but this condition also gave high background signals. This was certainly due to the nonspecific adsorption of the antibody to the polymeric material, as the epitope region of the antibody corresponds to only a small part of the antibody and the control tests do not contain CA15-3. This was also confirmed by the more intense color obtained in the control test of NIP (compared to MIP) where CA15-3 was not present. This confirms that the MIP surface is less susceptible to nonspecific adsorption. The higher color intensity of the MIP signals compared to the signals from NIP in the presence of CA15-3 proves the selective response of the MIP surface to CA15-3. Incubation of HRP alone (5 µg ml-1), not in combination with an antibody, was subsequently tested. This condition resulted in a significant reduction in background signal compared to antibody binding, as no color was obtained in controls. MIP incubated with CA15-3 also showed a good (color intensive) response, which was also significantly higher than the response obtained with NIP. Increasing the amount of HRP (to 50 µg mL-1) showed a very weak response to the MIP control, but the signal of the paper incubated with CA15-3 also had a blue color. Since the ΔMIP/ΔNIP ratio is higher at the highest HRP concentration, the following tests were performed with 50 µg mL-1 HRP.
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 112 Figure 5-5. Photographs (A) of NIPs and MIPs with PB, antibody-HRP or HRP. Respective bar charts with Q values for PB (B), antibody-HRP 5 µg mL -1 (C), HRP 5 µg mL -1 (D) and HRP 50 µg mL -1 (E). From another point of view, the volume of the solution used was important for the homogeneity of the sensor surface and the sensitivity of the response, as shown in Figure 5-6. In the previous tests, 3 μl CA15-3 solution (or buffer) and then 3 μl HRP solution were incubated on each paper circle. The results showed that only high concentrations of CA15-3 could lead to color development under these conditions and that no color was visible to the naked eye at lower concentrations. For this
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 113 reason, higher volumes (4 and 5 μL) were tested. Incubation of 4 μL or 5 μL CA15-3 and HRP solutions resulted in an increase in colorimetric reaction intensity with increasing concentration of CA15-3. It was also found that not only MIPs can also recognise lower concentrations of this protein, but also NIPs can adsorb some targets. However, the recognition capacity of the MIP surface is obviously greater than the adsorption that occurs on NIPs. Five microlitres were selected for further investigation as this gave the best linear response, as shown in Figure 5-6 (B to C). Under this condition, the quadrature signal of the sensor was saturated for concentrations > 500 U mL-1, which by far includes the CA15-3 values > 30 U mL-1 that are clinically important in BC. Figure 5-6. Photographs (A) of NIPs and MIPs with 3, 4 or 5-µL of standard solution and HRP solution. Calibration curves with Q values of MIPs with 3 (B), 4 (C) or 5 (D) µL of standard and HRP solutions.
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 114 5.3.4. Biomimetic enzyme-linked immunosorbent assay calibration Figure 5-7 shows the images of the paper sensors incubated in different concentrations of CA153, from 3 to 500 U mL-1, in buffered solutions. The average points obtained from triplicate measurements of each data point were plotted against grey intensity or Quadrature. The error bars included correspond to the SD. The linear regression equation of the MIP was Grey intensity = - 0.0858 × CA15-3 U mL-1 + 149.44 (R-squared 0.9892) or Quadrature = -0.0102 × CA15-3 U mL-1 + 11.032 (R-squared 0.9794). The NIP showed no linear response. Figure 5-7. Photographs of buffer calibration in MIP (A). Calibration curve based on grey intensity for NIP (B) and MIP (C). Calibration curve based on quadrature values for NIP (D) and MIP (E). 5.3.5. Selectivity Selectivity is a crucial parameter to be investigated when evaluating the performance of a biosensensor to assess whether it can be used in real-world conditions. The selectivity assay was A B D E C
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 115 performed by incubating CA15-3 solutions prepared alone or in the presence of a possible interfering substance. Assays were made triplicate and data was expressed as mean ± SD. Figure 5-8 shows the mean values in terms of percent signal change compared with CA15-3 alone. CEA (1%) and CA125 (5.5%) resulted in small differences in grey scale compared with the effect of glucose (10.4%) and IgG (11%). Overall, the effect of the four interfering species tested in this assay was negligible with error values equal or below 11%. The results show that the MIP surface has highly selectivity for CA15-3 compared to the interfering molecules tested. This is mainly due to the presence of the imprinted cavities for CA15-3. These results suggest that these sensors could be used to detect CA15-3 in complex matrices. Figure 5-8. Images and the respective bar chart of selectivity study based on comparison of the response of CA15-3 (250 U mL-1) incubated alone or mixed with interfering species as CEA (0.25 ng mL-1), CA125 (0.35 U mL-1), glucose (1 mg mL-1) and IgG (0.1 mg mL-1). 5.3.6. Biomimetic enzyme-linked immunosorbent assay in serum samples Figure 5-9 shows the sensor response for the detection of CA15-3 in FBS. Calibrations were performed in a background medium of serum to achieve an analytical level of communication that
Paper-based biomimetic test-strip for cancer antigen 15-3 with colored readout 116 is well matched to the composition of the sample. For this purpose, the standard samples of CA153 contained real serum in their composition. As with the buffered solutions, calibrations were performed in triplicate and data are reported as mean ± SD. The overall behaviour showed a linear trend for both the grey-scale and the quadrature values as a function of CA15-3 concentration (U mL-1), corresponding to grey value = -0.0956 × CA15-3 U mL-1 + 158.05 (R-squared 0.997) and quadrature = -0.0098 × CA15-3 U mL-1 + 11.006 (R-squared 0.9903). Compared with the experiments with buffer, the slope values were similar, as was the range of the linear response. The results obtained indicate that this sensor can be used for the detection of CA15-3 in serum samples. Figure 5-9. Photographs of FBS calibration in MIP (A). Calibration curve based on grey intensity for NIP (B) and MIP (C). Calibration curve based on quadrature values for NIP (D) and MIP (E). A B C D E
General conclusions and future perspectives 123 least, to use this sensor in the field, proper conditions to acquire the results should be guaranteed in order to not affect assay sensitivity and repeatability. In the described works, a homemade darkroom was constructed to capture the photographs under controlled conditions, namely light intensity, and focal distance. For that reason, a portable and robust darkroom can be constructed, where a smartphone can be coupled to capture the results. In this smartphone, several software can be used to calculate the color coordinates and a specific one can be developed to automatically convert these values to the concentration of the detected target, given a real-time response. Regarding the complexity of cancerous process and the wide range of available biomarkers, the development of future biosensors for cancer detection should consider a panel of analytes in a multiplex device, enabling the detection of several analytes at the same time, using the same biologic sample. Despite the good performance obtained in Chapter 5 with a MIP being used as an alternative to antibodies as RE, other synthetic REs (e.g., and aptamer) with promising recognition ability for cancer biomarkers could be tested.
124 References rdsd Chapter 7 References
125 References rdsd 7. References 1. Mahmoudi T, de la Guardia M, Baradaran B. Lateral flow assays towards point-of-care cancer detection: A review of current progress and future trends. TrAC Trends in Analytical Chemistry. 2020;125:115842. 2. Ferlay J, Ervik M, Lam F, Colombet M, Mery L, Piñeros M, et al. Cancer Today: International Agency for Research on Cancer; 2020 [Available from: https://gco.iarc.fr/today/home. 3. Shankaran DR. Chapter 8 - Nano-Enabled Immunosensors for Point-of-Care Cancer Diagnosis. In: Mohan Bhagyaraj S, Oluwafemi OS, Kalarikkal N, Thomas S, editors. Applications of Nanomaterials: Woodhead Publishing; 2018. p. 205-50. 4. Organization WH. Cancer [Available from: https://www.who.int/news-room/factsheets/detail/cancer. 5. Carrell C, Kava A, Nguyen M, Menger R, Munshi Z, Call Z, et al. Beyond the lateral flow assay: A review of paper-based microfluidics. Microelectronic Engineering. 2019;206:45-54. 6. Tothill I, Altintas Z. Molecular biosensors: promising new tools for early detection of cancer. Nanobiosensors in Disease Diagnosis. 2015:1. 7. Tothill IE. Biosensors for cancer markers diagnosis. Seminars in Cell & Developmental Biology. 2009;20(1):55-62. 8. Ranjan R, Esimbekova EN, Kratasyuk VA. Rapid biosensing tools for cancer biomarkers. Biosensors and Bioelectronics. 2017;87:918-30. 9. Jayanthi VSPKSankara A, Das AB, Saxena U. Recent advances in biosensor development for the detection of cancer biomarkers. Biosensors and Bioelectronics. 2017;91:15-23. 10. Rebelo TSCR, Ribeiro JA, Sales MGF, Pereira CM. Electrochemical immunosensor for detection of CA 15-3 biomarker in point-of-care. Sensing and Bio-Sensing Research. 2021;33:100445. 11. Hosu O, Ravalli A, Lo Piccolo GM, Cristea C, Sandulescu R, Marrazza G. Smartphone-based immunosensor for CA125 detection. Talanta. 2017;166:234-40. 12. Cheraghi S, Taher MA, Karimi-Maleh H, Karimi F, Shabani-Nooshabadi M, Alizadeh M, et al. Novel enzymatic graphene oxide based biosensor for the detection of glutathione in biological body fluids. Chemosphere. 2022;287:132187. 13. Pothipor C, Aroonyadet N, Bamrungsap S, Jakmunee J, Ounnunkad K. A highly sensitive electrochemical microRNA-21 biosensor based on intercalating methylene blue signal amplification and a highly dispersed gold nanoparticles/graphene/polypyrrole composite. Analyst. 2021;146(8):2679-88. 14. Bodulev OL, Galkin II, Zhao S, Pletyushkina OY, Sakharov IY. Quantitation of MicroRNA-155 in Human Cells by Heterogeneous Enzyme-Linked Oligonucleotide Assay Coupled with Mismatched Catalytic Hairpin Assembly Reaction. Biosensors [Internet]. 2022; 12(8). 15. Wu Y, Chen X, Wang X, Yang M, Xu F, Hou C, et al. A fluorescent biosensor based on prismatic hollow Metal-polydopamine frameworks and 6-carboxyfluorescein (FAM)-labeled protein aptamer for CA15-3 detection. Sensors and Actuators B: Chemical. 2021;329:129249. 16. Yu Q, Zhao Q, Wang S, Zhao S, Zhang S, Yin Y, et al. Development of a lateral flow aptamer assay strip for facile identification of theranostic exosomes isolated from human lung carcinoma cells. Analytical Biochemistry. 2020;594:113591. 17. Yin Y, Cao Y, Xu Y, Li G. Colorimetric immunoassay for detection of tumor markers. International journal of molecular sciences. 2010;11(12):5077-94. 18. Liu W, Yang H, Ding Y, Ge S, Yu J, Yan M, et al. Paper-based colorimetric immunosensor for visual detection of carcinoembryonic antigen based on the high peroxidase-like catalytic performance of ZnFe2O4– multiwalled carbon nanotubes. Analyst. 2014;139(1):251-8. 19. Chen Y, Chu W, Liu W, Guo X. Distance-based carcinoembryonic antigen assay on microfluidic paper immunodevice. Sensors and Actuators B: Chemical. 2018;260:452-9. 20. Mohammadi S, Salimi A, Hamd-Ghadareh S, Fathi F, Soleimani F. A FRET immunosensor for sensitive
126 References rdsd detection of CA 15-3 tumor marker in human serum sample and breast cancer cells using antibody functionalized luminescent carbon-dots and AuNPs-dendrimer aptamer as donor-acceptor pair. Analytical Biochemistry. 2018;557:18-26. 21. Li H, He J, Li S, Turner APF. Electrochemical immunosensor with N-doped graphene-modified electrode for label-free detection of the breast cancer biomarker CA 15-3. Biosensors and Bioelectronics. 2013;43:25-9. 22. Pereira AC, Sales MGF, Rodrigues LR. Chapter 3 - Biosensors for Rapid Detection of Breast Cancer Biomarkers. In: Inamuddin, Khan R, Mohammad A, Asiri AM, editors. Advanced Biosensors for Health Care Applications: Elsevier; 2019. p. 71-103. 23. Santos ART, Moreira FTC, Helguero LA, Sales MGF. Antibody Biomimetic Material Made of Pyrrole for CA 15-3 and Its Application as Sensing Material in Ion-Selective Electrodes for Potentiometric Detection. Biosensors. 2018;8(1):1-16. 24. Kasetsirikul S, Shiddiky MJA, Nguyen N-T. Challenges and perspectives in the development of paperbased lateral flow assays. Microfluidics and Nanofluidics. 2020;24(2):1-18. 25. Frasco MF, Truta LAANA, Sales MGF, Moreira FTC. Imprinting Technology in Electrochemical Biomimetic Sensors. Sensors (Basel, Switzerland). 2017;17(3):523. 26. Ranjan P, Parihar A, Jain S, Kumar N, Dhand C, Murali S, et al. Biosensor-based diagnostic approaches for various cellular biomarkers of breast cancer: A comprehensive review. Analytical Biochemistry. 2020;610:113996. 27. Wang K, Yang J, Xu H, Cao B, Qin Q, Liao X, et al. Smartphone-imaged multilayered paper-based analytical device for colorimetric analysis of carcinoembryonic antigen. Analytical and Bioanalytical Chemistry. 2020;412(11):2517-28. 28. Ozer T, McMahon C, Henry CS. Advances in Paper-Based Analytical Devices. Annual Review of Analytical Chemistry. 2020;13(1):85-109. 29. Murray LP, Mace CR. Usability as a guiding principle for the design of paper-based, point-of-care devices – A review. Analytica Chimica Acta. 2020;1140:236-49. 30. Tekade RK, Sun X. The Warburg effect and glucose-derived cancer theranostics. Drug Discovery Today. 2017;22(11):1637-53. 31. Organization WH. Cancer [Available from: https://www.who.int/news-room/factsheets/detail/cancer. 32. CP W, E W, BW S. World Cancer Report: Cancer Research for Cancer Prevention. 2020. 33. Cancer types: National Cancer Institute; [Available from: www.cancer.gov/types. 34. Azab MY, Hameed MFO, Obayya SSA. Overview of Optical Biosensors for Early Cancer Detection: Fundamentals, Applications and Future Perspectives. Biology [Internet]. 2023; 12(2). 35. Cancer tomorrow: International Agency for Research on Cancer; 2020 [Available from: https://gco.iarc.fr/tomorrow/en. 36. Ferlay J, Ervik M, Lam F, Colombet M, Mery L, Piñeros M, et al. Global Cancer Observatory - Cancer Today: International Agency for Research on Cancer; 2020 [Available from: https://gco.iarc.fr/today/home. 37. Hirata B, Massayo Maeda Oda J, Guembarovski R, Ariza C, Oliveira C, Watanabe M. Molecular Markers for Breast Cancer: Prediction on Tumor Behavior. Disease markers. 2014;2014:1-12. 38. Łukasiewicz SA-O, Czeczelewski M, Forma AA-O, Baj JA-O, Sitarz R, Stanisławek A. Breast Cancer - Epidemiology, Risk Factors, Classification, Prognostic Markers, and Current Treatment Strategies-An Updated Review. Cancers. 2021;13(2072-6694 (Print)). 39. About cancer: National Cancer Institute; [Available from: https://www.cancer.gov/. 40. Song C, Yang Y, Yang B, Min L, Wang L. Combination assay of lung cancer associated serum markers using surface-enhanced Raman spectroscopy. Journal of Materials Chemistry B. 2016;4(10):1811-7. 41. Suntornsuk W, Suntornsuk L. Recent applications of paper-based point-of-care devices for biomarker detection. ELECTROPHORESIS. 2020;41(5-6):287-305. 42. Mazzu-Nascimento T, Morbioli GG, Milan LA, Donofrio FC, Mestriner CA, Carrilho E. Development and statistical assessment of a paper-based immunoassay for detection of tumor markers. Analytica Chimica
127 References rdsd Acta. 2017;950:156-61. 43. Ratajczak K, Stobiecka M. High-performance modified cellulose paper-based biosensors for medical diagnostics and early cancer screening: A concise review. Carbohydrate Polymers. 2020;229:115463. 44. Suh KS, Park SW, Castro A, Patel H, Blake P, Liang M, et al. Ovarian cancer biomarkers for molecular biosensors and translational medicine. Expert Review of Molecular Diagnostics. 2010;10(8):1069-83. 45. Weber JA, Baxter DH, Zhang S, Huang DY, Huang KH, Lee MJ, et al. The microRNA spectrum in 12 body fluids. Clinical chemistry. 2010;56(11):1733-41. 46. Wong DTW. Salivary extracellular noncoding RNA: emerging biomarkers for molecular diagnostics. Clinical therapeutics. 2015;37(3):540-51. 47. Fracchiolla NS, Artuso S, Cortelezzi A. Biosensors in clinical practice: focus on oncohematology. Sensors (Basel, Switzerland). 2013;13(5):6423-47. 48. Kabel AM. Tumor markers of breast cancer: New prospectives. Journal of Oncological Sciences. 2017;3(1):5-11. 49. DePolo J. 2023 [Available from: www.breastcancer.org. 50. Lee JS, Park S, Park JM, Cho JH, Kim SI, Park BW. Elevated levels of preoperative CA 15-3 and CEA serum levels have independently poor prognostic significance in breast cancer. Annals of Oncology. 2013;24(5):1225-31. 51. Duffy MJ, Shering S Fau - Sherry F, Sherry F Fau - McDermott E, McDermott E Fau - O'Higgins N, O'Higgins N. CA 15-3: a prognostic marker in breast cancer. The International journal of biological markers. 2000;15(4):330-33. 52. Park B, Oh J, Kim J, SH P, Kim K, Kim J, et al. Preoperative CA 15-3 and CEA serum levels as predictor for breast cancer outcomes. Annals of oncology. 2008;18(4):675-81. 53. Wang L. Early Diagnosis of Breast Cancer. Sensors. 2017;17(7). 54. Meirinho SG, Dias LG, Peres AM, Rodrigues LR. Voltammetric aptasensors for protein disease biomarkers detection: A review. Biotechnology Advances. 2016;34(5):941-53. 55. Kufe DW. Mucins in cancer: function, prognosis and therapy. Nature Reviews Cancer. 2009;9(14741768 (Electronic)):874-85. 56. Kufe DW. MUC1-C oncoprotein as a target in breast cancer: activation of signaling pathways and therapeutic approaches. Oncogene. 2013;32(9):1073-81. 57. Duffy MJ. CA 15-3 and related mucins as circulating markers in breast cancer. Annals of clinical biochemistry. 1999(0004-5632 (Print)):579-86. 58. Chen W, Zhang Z, Zhang S, Zhu P, Ko JK, Yung KK. MUC1: Structure, Function, and Clinic Application in Epithelial Cancers. International Journal of Molecular Sciences. 2021;22(12). 59. Duffy MJ, Evoy D, McDermott EW. CA 15-3: Uses and limitation as a biomarker for breast cancer. Clinica Chimica Acta. 2010;411(23):1869-74. 60. Ribeiro JA, Pereira CM, Silva AF, Sales MGF. Disposable electrochemical detection of breast cancer tumour marker CA 15-3 using poly(Toluidine Blue) as imprinted polymer receptor. Biosensors and Bioelectronics. 2018;109:246-54. 61. Hong R, Sun H, Li D, Yang W, Fan K, Liu C, et al. A Review of Biosensors for Detecting Tumor Markers in Breast Cancer. Life [Internet]. 2022; 12(3). 62. Chourin S Fau - Georgescu D, Georgescu D Fau - Gray C, Gray C Fau - Guillemet C, Guillemet C Fau - Loeb A, Loeb A Fau - Veyret C, Veyret C Fau - Basuyau JP, et al. Value of CA 15-3 determination in the initial management of breast cancer patients. Annals of Oncology. 2009;20(5). 63. Shering SG, Sherry F Fau - McDermott EW, McDermott Ew Fau - O'Higgins NJ, O'Higgins Nj Fau - Duffy MJ, Duffy MJ. Preoperative CA 15-3 concentrations predict outcome of patients with breast carcinoma. American Cancer Society. 1998(0008-543X (Print)). 64. Duffy M, Duggan C, Keane R, Arnold K, McDermott E, Crown J, et al. High Preoperative CA 15-3 Concentrations Predict Adverse Outcome in Node-Negative and Node-Positive Breast Cancer: Study of 600 Patients with Histologically Confirmed Breast Cancer. Clinical chemistry. 2004;50(3):559-63. 65. Metkar SK, Girigoswami K. Diagnostic biosensors in medicine – A review. Biocatalysis and Agricultural
128 References rdsd Biotechnology. 2019;17:271-83. 66. Kamel S, Khattab T. Recent Advances in Cellulose-Based Biosensors for Medical Diagnosis. Biosensors. 2020;10:1-26. 67. Bhalla N, Jolly P, Formisano N, Estrela P. Introduction to biosensors. Essays in Biochemistry. 2016(1744-1358 (Electronic)). 68. IUPAC Gold Book: International Union of Pure and Applied Chamistry; [Available from: https://goldbook.iupac.org/terms/search. 69. Samuel VR, Rao KJ. A review on label free biosensors. Biosensors and Bioelectronics: X. 2022;11:100216. 70. Sadani K, Nag P, Thian XY, Mukherji S. Enzymatic optical biosensors for healthcare applications. Biosensors and Bioelectronics: X. 2022;12:100278. 71. Syahir A, Usui K, Tomizaki K-y, Kajikawa K, Mihara H. Label and Label-Free Detection Techniques for Protein Microarrays. Microarrays [Internet]. 2015; 4(2):[228-44 pp.]. 72. Dukle AA-O, Nathanael AA-O, Panchapakesan BA-O, Oh TH. Role of Paper-Based Sensors in Fight against Cancer for the Developing World. Biosensors. 2022;12(9). 73. Nadar SA-O, Patil PA-O, Tiwari MA-O, Ahirrao DJ. Enzyme embedded microfluidic paper-based analytic device (μPAD): a comprehensive review. Critical Reviews in Biotechnology. 2021;41(1549-7801). 74. Liu L, Hao Y, Deng D, Xia N. Nanomaterials-Based Colorimetric Immunoassays. Nanomaterials (Basel, Switzerland). 2019;9(3):316. 75. Mahato K, Srivastava A, Chandra P. Paper based diagnostics for personalized health care: Emerging technologies and commercial aspects. Biosensors & bioelectronics. 2017(1873-4235):246-59. 76. Gopinath SCB, Tang T-H, Citartan M, Chen Y, Lakshmipriya T. Current aspects in immunosensors. Biosensors and Bioelectronics. 2014;57:292-302. 77. Sharma S, Byrne H, O'Kennedy RJ. Antibodies and antibody-derived analytical biosensors. Essays in biochemistry. 2016;60(1):9-18. 78. Cerda-Kipper A, Montiel B, Hosseini S. Radioimmunoassay and Enzyme-Linked Immunosorbent Assay. 2018. 79. Lindenmann J. Origin of the terms 'antibody' and 'antigen'. 1984(0300-9475 (Print)). 80. Yalow Rs Fau - Berson SA, Berson SA. Assay of plasma insulin in human subjects by immunological methods. 1959(0028-0836 (Print)). 81. Zhu G, Yin X, Jin D, Zhang B, Gu Y, An Y. Paper-based immunosensors: Current trends in the types and applied detection techniques. TrAC Trends in Analytical Chemistry. 2019;111:100-17. 82. Singh A, Chaudhary S, Agarwal A, Verma AS. Chapter 15 - Antibodies: Monoclonal and Polyclonal. In: Verma AS, Singh A, editors. Animal Biotechnology. San Diego: Academic Press; 2014. p. 265-87. 83. Trilling AK, Beekwilder J, Zuilhof H. Antibody orientation on biosensor surfaces: a minireview. Analyst. 2013;138(6):1619-27. 84. Zhu X, Xiong S, Zhang J, Zhang X, Tong X, Kong S. Improving paper-based ELISA performance through covalent immobilization of antibodies. Sensors and Actuators B: Chemical. 2018;255:598-604. 85. Peng Y, Gelder VV, Amaladoss A, Patel KH. Covalent Binding of Antibodies to Cellulose Paper Discs and Their Applications in Naked-eye Colorimetric Immunoassays. 2016(1940-087X (Electronic)). 86. Fu H, Liu X. Experimental comparison of surface chemistries for biomolecule immobilization on paperbased microfluidic devices. Journal of Micromechanics and Microengineering. 2019;29(12):124003. 87. Ahmed S, Bui M-PN, Abbas A. Paper-based chemical and biological sensors: Engineering aspects. Biosensors and Bioelectronics. 2016;77:249-63. 88. Hong W, Jeong S-G, Shim G, Kim DY, Pack SP, Lee C-S. Improvement in the Reproducibility of a Paper-based Analytical Device (PAD) Using Stable Covalent Binding between Proteins and Cellulose Paper. Biotechnology and bioprocess engineering : BBE. 2018;23(6):686-92. 89. Soler M, Lechuga LM. Biochemistry strategies for label-free optical sensor biofunctionalization: advances towards real applicability. Analytical and Bioanalytical Chemistry. 2022;414(18):5071-85. 90. Gao S, Guisán JM, Rocha-Martin J. Oriented immobilization of antibodies onto sensing platforms - A
129 References rdsd critical review. Analytica Chimica Acta. 2022;1189:338907. 91. Engvall E, Perlmann P. Enzyme-linked immunosorbent assay (ELISA) quantitative assay of immunoglobulin G. Immunochemistry. 1971;8(9):871-4. 92. Jeong S, Park MJ, Song W, Kim HS. Current immunoassay methods and their applications to clinically used biomarkers of breast cancer. Clinical biochemistry. 2020;78(1873-2933):43-57. 93. Gonzalez A, Gaines M, Gallegos LY, Guevara R, Gomez FA. Threadpaper, and fabric enzyme-linked immunosorbent assays (ELISA). Methods. 2018;146:58-65. 94. Chi J, Gao B, Sun M, Zhang F, Su E, Liu H, et al. Patterned Photonic Nitrocellulose for Pseudopaper ELISA. Analytical Chemistry. 2017;89(14):7727-33. 95. Schirhagl R. Bioapplications for Molecularly Imprinted Polymers. Analytical Chemistry. 2014;86(1):250-61. 96. Dickey FH. Specific Adsorption. The Journal of Physical Chemistry. 1955;59(8):695-707. 97. Wulff G, Sarhan A, Zabrocki K. Wulff, G., Sarhan, A. & Zabrocki, K. Enzyme-analog built polymers and their use for the resolution of racemates. Tetrahedron Lett. 44, 4329-43321973. 4329–32 p. 98. Takagishi T, Klotz IM. Macromolecule-small molecule interactions; introduction of additional binding sites in polyethyleneimine by disulfide cross–linkages. Biopolymers. 1972;11(2):483-91. 99. Vlatakis G, Andersson LI, Müller R, Mosbach K. Drug assay using antibody mimics made by molecular imprinting. Nature. 1993;361:645. 100. Cieplak M, Kutner W. Artificial Biosensors: How Can Molecular Imprinting Mimic Biorecognition? Trends in Biotechnology. 2016;34(11):922-41. 101. Ertürk G, Mattiasson B. Molecular Imprinting Techniques Used for the Preparation of Biosensors. Sensors (Basel, Switzerland). 2017;17(2):288. 102. Vasapollo G, Sole RD, Mergola L, Lazzoi MR, Scardino A, Scorrano S, et al. Molecularly Imprinted Polymers: Present and Future Prospective. International Journal of Molecular Sciences. 2011;12(9). 103. Xu J, Miao H, Wang J, Pan G. Molecularly Imprinted Synthetic Antibodies: From Chemical Design to Biomedical Applications. Small. 2020;16(27):1906644. 104. Zhang C, Cui H, Han Y, Yu F, Shi X. Development of a biomimetic enzyme-linked immunosorbent assay based on molecularly imprinted polymers on paper for the detection of carbaryl. Food Chemistry. 2018;240:893-7. 105. Li W, Zhang X, Li T, Ji Y, Li R. Molecularly imprinted polymer-enhanced biomimetic paper-based analytical devices: A review. Analytica Chimica Acta. 2021;1148:238196. 106. Saylan Y, Denizli A. Molecularly Imprinted Polymer-Based Microfluidic Systems for Point-of-Care Applications. Micromachines. 2019;10(11):766. 107. Ansari S, Masoum S. Molecularly imprinted polymers for capturing and sensing proteins: Current progress and future implications. TrAC Trends in Analytical Chemistry. 2019;114:29-47. 108. Saylan Y, Yilmaz F, Özgür E, Derazshamshir A, Yavuz H, Denizli A. Molecular Imprinting of Macromolecules for Sensor Applications. Sensors (Basel, Switzerland). 2017;17(4):898. 109. Chen L, Xu S, Li J. Recent advances in molecular imprinting technology: current status, challenges and highlighted applications. Chemical Society Reviews. 2011;40(5):2922-42. 110. Chen L, Wang X, Lu W, Wu X, Li J. Molecular imprinting: perspectives and applications. Chemical Society Reviews. 2016;45(8):2137-211. 111. Uzun L, Turner APF. Molecularly-imprinted polymer sensors: realising their potential. Biosensors and Bioelectronics. 2016;76:131-44. 112. Hasanah AN, Safitri N, Zulfa A, Neli N, Rahayu D. Factors Affecting Preparation of Molecularly Imprinted Polymer and Methods on Finding Template-Monomer Interaction as the Key of Selective Properties of the Materials. Molecules [Internet]. 2021; 26(18). 113. Uysal A, Demirel G, Turan E, Çaykara T. Hemoglobin recognition of molecularly imprinted hydrogels prepared at different pHs. Analytica Chimica Acta. 2008;625(1):110-5. 114. Tavares APM, Ferreira NS, Truta L, Sales MGF. Conductive Paper with Antibody-Like Film for Electrical Readings of Biomolecules. Scientific Reports. 2016;6.
130 References rdsd 115. Bossi A, Piletsky SA, Piletska EV, Righetti PG, Turner APF. Surface-Grafted Molecularly Imprinted Polymers for Protein Recognition. Analytical Chemistry. 2001;73(21):5281-6. 116. Liu K, Wei Wz Fau - Zeng J-X, Zeng Jx Fau - Liu X-Y, Liu Xy Fau - Gao Y-P, Gao YP. Application of a novel electrosynthesized polydopamine-imprinted film to the capacitive sensing of nicotine. Analytical and Bioanalytical Chemistry. 2006;385(1618-2642). 117. Carneiro MCCG, Sousa-Castillo A, Correa-Duarte MA, Sales MGF. Dual biorecognition by combining molecularly-imprinted polymer and antibody in SERS detection. Application to carcinoembryonic antigen. Biosensors and Bioelectronics. 2019;146:111761. 118. Liu Z, He H. Synthesis and Applications of Boronate Affinity Materials: From Class Selectivity to Biomimetic Specificity. Accounts of Chemical Research. 2017;50(9):2185-93. 119. Erdőssy J, Horváth V, Yarman A, Scheller FW, Gyurcsányi RE. Electrosynthesized molecularly imprinted polymers for protein recognition. TrAC Trends in Analytical Chemistry. 2016;79(Supplement C):17990. 120. Quazi S. Application of biosensors in cancers, an overview. Frontiers in Bioengeneering and biotechnology. 2023(2296-4185 (Print)). 121. Ribeiro JA, Sales MGF, Pereira CM. Electrochemistry-Assisted Surface Plasmon Resonance Biosensor for Detection of CA 15–3. Analytical Chemistry. 2021;93(22):7815-24. 122. Pacheco JG, Silva MSV, Freitas M, Nouws HPA, Delerue-Matos C. Molecularly imprinted electrochemical sensor for the point-of-care detection of a breast cancer biomarker (CA 15-3). Sensors and Actuators B: Chemical. 2018;256:905-12. 123. Huang J, Zhang T, Zheng Y, Liu J. Dual-Mode Sensing Platform for Cancer Antigen 15-3 Determination Based on a Silica Nanochannel Array Using Electrochemiluminescence and Electrochemistry2023; 13(3). 124. Kuswandi B, Ensafi AA. Perspective—Paper-Based Biosensors: Trending Topic in Clinical Diagnostics Developments and Commercialization. Journal of The Electrochemical Society. 2019;167(3):037509. 125. Singh AT, Lantigua D, Meka A, Taing S, Pandher M, Camci-Unal G. Paper-Based Sensors: Emerging Themes and Applications. Sensors. 2018;18(9):2838. 126. Zhang Z, Wang H, Chen Z, Wang X, Choo J, Chen L. Plasmonic colorimetric sensors based on etching and growth of noble metal nanoparticles: Strategies and applications. Biosensors and Bioelectronics. 2018;114:52-65. 127. Khosroshahi MA-O, Patel Y, Chabok R. Non-invasive optical characterization and detection of CA 153 breast cancer biomarker in blood serum using monoclonal antibody-conjugated gold nanourchin and surface-enhanced Raman scattering. Lasers in Medical Science. 2022;38(1). 128. Ranganathan V, Srinivasan S, Singh A, DeRosa MC. An aptamer-based colorimetric lateral flow assay for the detection of human epidermal growth factor receptor 2 (HER2). Analytical Biochemistry. 2020;588:113471. 129. Alizadeh N, Salimi A, Hallaj R. Mimicking peroxidase activity of Co2(OH)2CO3-CeO2 nanocomposite for smartphone based detection of tumor marker using paper-based microfluidic immunodevice. Talanta. 2018;189:100-10. 130. Sousa DA, Carneiro M, Ferreira D, Moreira FTC, Sales MGF, Rodrigues LR. Recent Advances in the Selection of Cancer-Specific Aptamers for the Development of Biosensors. Curr Med Chem 2022;29(1875-533X (Electronic)):5850-80. 131. Abou-Omar MN, Attia MS, Afify HG, Amin MA, Boukherroub R, Mohamed EH. Novel Optical Biosensor Based on a Nano-Gold Coated by Schiff Base Doped in Sol/Gel Matrix for Sensitive Screening of Oncomarker CA-125. ACS Omega. 2021;6(32):20812-21. 132. Chen W, Zhang Y, Di K, Liu C, Xia Y, Ding S, et al. A Washing-Free and Easy-to-Operate Fluorescent Biosensor for Highly Efficient Detection of Breast Cancer-Derived Exosomes. Frontiers in Bioengineering and Biotechnology. 2022(2296-4185 (Print)). 133. Loyez M, Lobry M, Hassan EM, DeRosa MC, Caucheteur C, Wattiez R. HER2 breast cancer biomarker detection using a sandwich optical fiber assay. Talanta. 2021;221:121452. 134. Szymańska B, Lukaszewski Z, Hermanowicz-Szamatowicz K, Gorodkiewicz E. A biosensor for
131 References rdsd determination of the circulating biomarker CA125/MUC16 by Surface Plasmon Resonance Imaging. Talanta. 2020;206:120187. 135. Xue T, Liang W, Li Y, Sun Y, Xiang Y, Zhang Y, et al. Ultrasensitive detection of miRNA with an antimonene-based surface plasmon resonance sensor. Nature Communications. 2019;10(1):28. 136. Castaño-Guerrero Y, Romaguera-Barcelay Y, Moreira FT, Brito WR, Fortunato E, Sales MG. Poly(Thionine)-Modified Screen-Printed Electrodes for CA 19-9 Detection and Its Properties in Raman Spectroscopy. Chemosensors [Internet]. 2022; 10(3). 137. Turan E, Zengin A, Suludere Z, Kalkan NÖ, Tamer U. Construction of a sensitive and selective plasmonic biosensor for prostate specific antigen by combining magnetic molecularly-imprinted polymer and surface-enhanced Raman spectroscopy. Talanta. 2022;237:122926. 138. Aydindogan E, Ceylan AE, Timur S. Paper-based colorimetric spot test utilizing smartphone sensing for detection of biomarkers. Talanta. 2020;208:120446. 139. Fernandes GM, Silva WR, Barreto DN, Lamarca RS, Lima Gomes PCF, Flávio da SPJ, et al. Novel approaches for colorimetric measurements in analytical chemistry - A review. 2020(1873-4324 (Electronic)). 140. Abarghoei S, Fakhri N, Borghei YS, Hosseini M, Ganjali MR. A colorimetric paper sensor for citrate as biomarker for early stage detection of prostate cancer based on peroxidase-like activity of cysteine-capped gold nanoclusters. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy. 2019;210:251-9. 141. ImmunoChemistry Technologies L. HRP Redox Reaction Driven TMB Color Development. 2020. 142. Busa LSA, Maeki M, Ishida A, Tani H, Tokeshi M. Simple and sensitive colorimetric assay system for horseradish peroxidase using microfluidic paper-based devices. Sensors and Actuators B: Chemical. 2016;236:433-41. 143. Vazquez-Alvarado M, Vanasupa S, Valdez EH, Pama AM, Crowder MJ, Vanasupa L, et al. Evaluation of chromogenic substrates for horseradish peroxidase on paper-based microfluidic devices. Sensors and Actuators B: Chemical. 2023;377:133028. 144. Harpaz D, Eltzov E, Ng TSE, Marks RS, Tok AIY. Enhanced Colorimetric Signal for Accurate Signal Detection in Paper-Based Biosensors. Diagnostics. 2020(2075-4418 (Print)). 145. Ren W, Mohammed SI, Wereley S, Irudayaraj J. Magnetic Focus Lateral Flow Sensor for Detection of Cervical Cancer Biomarkers. Analytical Chemistry. 2019;91(4):2876-84. 146. Koczula KM, Gallotta A. Lateral flow assays. Essays in biochemistry. 2016;60(1):111-20. 147. Aldewachi H, Chalati T, Woodroofe MN, Bricklebank N, Sharrack B, Gardiner P. Gold nanoparticlebased colorimetric biosensors. Nanoscale. 2018;10(1):18-33. 148. Yuan Z, Hu C-C, Chang H-T, Lu C. Gold nanoparticles as sensitive optical probes. Analyst. 2016;141(5):1611-26. 149. Zhang Y, McKelvie ID, Cattrall RW, Kolev SD. Colorimetric detection based on localised surface plasmon resonance of gold nanoparticles: Merits, inherent shortcomings and future prospects. Talanta. 2016;152:410-22. 150. Jazayeri MH, Aghaie T, Avan A, Vatankhah A, Ghaffari MRS. Colorimetric detection based on gold nano particles (GNPs): An easy, fast, inexpensive, low-cost and short time method in detection of analytes (protein, DNA, and ion). Sensing and Bio-Sensing Research. 2018;20:1-8. 151. Ma Z, Sui S-F. Naked-Eye Sensitive Detection of Immunoglubulin G by Enlargement of Au Nanoparticles In Vitro. Angewandte Chemie International Edition. 2002;41(12):2176-9. 152. Liu H, Rong P, Jia H, Yang J, Dong B, Dong Q, et al. A Wash-Free Homogeneous Colorimetric Immunoassay Method. Theranostics. 2016;6(1):54-64. 153. Liana DD, Raguse B, Gooding JJ, Chow E. Recent advances in paper-based sensors. Sensors. 2012;12(9):11505-26. 154. Watt J, Banks J. On a new method of preparing a test liquor to shew the presence of acids and alkalies in chemical mixtures. Philosophical Transactions of the Royal Society of London. 1784;74:419-22. 155. Agarwal C, Csóka L. Recent Advances in Paper-Based Analytical Devices: A Pivotal Step Forward in Building Next-Generation Sensor Technology. Sustainable Polymer Composites and Nanocomposites: Springer International Publishing; 2019. p. 479-517.
132 References rdsd 156. Pelton R. Bioactive paper provides a low-cost platform for diagnostics. TrAC Trends in Analytical Chemistry. 2009;28(8):925-42. 157. Yamada K, Shibata H, Suzuki K, Citterio D. Toward practical application of paper-based microfluidics for medical diagnostics: state-of-the-art and challenges. Lab on a Chip. 2017;17(7):1206-49. 158. Klemm D, Heublein B, Fink H-P, Bohn A. Cellulose: Fascinating Biopolymer and Sustainable Raw Material. Angewandte Chemie International Edition. 2005;44(22):3358-93. 159. Nery EW, Kubota LT. Sensing approaches on paper-based devices: a review. Analytical and Bioanalytical Chemistry. 2013;405(24):7573-95. 160. Ye J, Hutchison JA, Uji-i H, Hofkens J, Lagae L, Maes G, et al. Excitation wavelength dependent surface enhanced Raman scattering of 4-aminothiophenol on gold nanorings. Nanoscale. 2012;4(5):160611. 161. Lim WY, Goh BT, Khor SM. Microfluidic paper-based analytical devices for potential use in quantitative and direct detection of disease biomarkers in clinical analysis. Journal of Chromatography B. 2017;1060:42442. 162. Shende P, Prabhakar B, Patil A. Color changing sensors: A multimodal system for integrated screening. TrAC Trends in Analytical Chemistry. 2019;121:115687. 163. Nie J, Zhang Y, Lin L, Zhou C, Li S, Zhang L, et al. Low-Cost Fabrication of Paper-Based Microfluidic Devices by One-Step Plotting. Analytical Chemistry. 2012;84(15):6331-5. 164. Kim TH, Hahn YA-O, Kim MA-O. Recent Advances of Fluid Manipulation Technologies in Microfluidic Paper-Based Analytical Devices (μPADs) toward Multi-Step Assays. Micromachines. 2020(2072-666X). 165. West PW. Selective Spot Test for Copper. Industrial & Engineering Chemistry Analytical Edition. 1945;17(11):740-1. 166. Free AH, Adams EC, Kercher ML, Free HM, Cook MH. Simple Specific Test for Urine Glucose. Clinical Chemistry. 1957;3(3):163-8. 167. Liu L, Yang D, Liu G. Signal amplification strategies for paper-based analytical devices. Biosensors and Bioelectronics. 2019;136:60-75. 168. Hu J, Wang SQ, Wang L, Li F, Pingguan-Murphy B, Lu TJ, et al. Advances in paper-based point-ofcare diagnostics. Biosensors & Bioelectronics. 2014;54:585-97. 169. Ma F, He L, Lindner E, Wu D-Y. Highly porous poly(l-lactic) acid nanofibers as a dual-signal paperbased bioassay platform for in vitro diagnostics. Applied Surface Science. 2021;542:148732. 170. Gao X, Xu H, Baloda M, Gurung AS, Xu L-P, Wang T, et al. Visual detection of microRNA with lateral flow nucleic acid biosensor. Biosensors and Bioelectronics. 2014;54:578-84. 171. Fakhri N, Abarghoei S, Dadmehr M, Hosseini M, Sabahi H, Ganjali MR. Paper based colorimetric detection of miRNA-21 using Ag/Pt nanoclusters. Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy. 2020;227:117529. 172. Singer JM, Plotz CM. The latex fixation test: I. Application to the serologic diagnosis of rheumatoid arthritis. The American Journal of Medicine. 1956;21(6):888-92. 173. Martinez AW, Phillips ST, Butte MJ, Whitesides GM. Patterned paper as a platform for inexpensive, low-volume, portable bioassays. Angewandte Chemie (International ed in English). 2007;46(8):1318-20. 174. Ge L, Wang S, Yu J, Li N, Ge S, Yan M. Molecularly Imprinted Polymer Grafted Porous Au-Paper Electrode for an Microfluidic Electro-Analytical Origami Device. Advanced Functional Materials. 2013;23(24):3115-23. 175. Xu S, Xu Z, Liu Z. Paper-Based Molecular-Imprinting Technology and Its Application. Biosensors [Internet]. 2022; 12(8). 176. Cheng H, Wang Y, Wang Y, Ge L, Liu X, Li F. A visualized sensor based on layered double hydroxides with peroxidase-like activity for sensitive acetylcholinesterase assay. Analytical Methods. 2023;15(30):37008. 177. Brazaca LC, Moreto JR, Martín A, Tehrani F, Wang J, Zucolotto V. Colorimetric Paper-Based Immunosensor for Simultaneous Determination of Fetuin B and Clusterin toward Early Alzheimer’s Diagnosis. ACS Nano. 2019;13(11):13325-32.