Accepted Manuscript

A decade of microchip electrophoresis for clinical diagnostics – A review of 2008-2017

Alain Wuethrich, Joselito P. Quirino

PII: S0003-2670(18)30946-2 DOI: 10.1016/j.aca.2018.08.009 Reference: ACA 236183

To appear in: Analytica Chimica Acta

Received Date: 27 March 2018 Revised Date: 30 July 2018 Accepted Date: 3 August 2018

Please cite this article as: A. Wuethrich, J.P. Quirino, A decade of microchip electrophoresis for clinical diagnostics – A review of 2008-2017, Analytica Chimica Acta (2018), doi: 10.1016/j.aca.2018.08.009.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT A decade of microchip electrophoresis for clinical diagnostics – A review of

2008-2017

Alain Wuethrich 1 and Joselito P. Quirino 2*

1Centre for Personalised Nanomedicine, Australian Institute for Bioengineering and

Nanotechnology (AIBN), University of Queensland, Building 75, Brisbane, QLD 4072,

Australia

2Australian Centre for Research on Separation Science (ACROSS), School of Physical

Sciences-Chemistry, University of Tasmania, Private Bag 75, Hobart, TAS 7001, Australia

MANUSCRIPT *Correspondence: Assoc./Prof. Joselito P. Quirino, Australian Centre for Research on

Separation Science (ACROSS), School of Physical Sciences, University of Tasmania, Private

Bag 75, Hobart, TAS 7001, Australia, e-mail: [email protected]; telephone: +61 3

6226 2529; fax: +61 3 6226 2858

Abbreviations: a list with abbreviations is on p. 73

ACCEPTED

ACCEPTED MANUSCRIPT

Abstract

A core element in clinical diagnostics is the data interpretation obtained through the

analysis of patient samples. To obtain relevant and reliable information, a methodological

approach of sample preparation, separation, and detection is required. Traditionally, these

steps are performed independently and stepwise. Microchip capillary electrophoresis (MCE)

can provide rapid and high-resolution separation with the capability to integrate a streamlined

and complete diagnostic workflow suitable for the point-of-care setting. Whilst standard

clinical diagnostics methods normally require hours to days to retrieve specific patient data,

MCE can reduce the time to minutes, hastening the delivery of treatment options for the

patients. This review covers the advances in MCE for disease detection from 2008 to 2017.

Miniaturised diagnostic approaches that required an electrophoretic separation step prior to

the detection of the biological samples are reviewed. In the two main sections, the discussion is focused on the technical set-up used to suit MCEMANUSCRIPT for disease detection and on the strategies that have been applied to study various diseases. Throughout these discussions MCE is

compared to other techniques to create context of the potential and challenges of MCE. A

comprehensive table categorised based on the studied disease using MCE is provided. We

also comment on future challenges that remain to be addressed.

ACCEPTED

ACCEPTED MANUSCRIPT

Introduction

A fast, reliable and accurate “sample-in–answer–out” analysis is requested in the

diagnostic environment. This is difficult to achieve since most biological samples are

complex and adequate sample preparation and separation should be performed prior to the

actual diagnostic interpretation. The separation step is crucial for obtaining qualitative and

quantitative clinical information. In general, the steps in the diagnostic process are

performed sequentially which can be tedious and resource-intensive. Thus, the idea to

integrate a whole analytical laboratory on a small glass chip seemed quite radical when it was

first proposed by Manz and colleagues in 1990 [1]. This initial “lab on a chip” concept or

micro total analysis systems (µTAS), also included the birth of microchip electrophoresis

(MCE). In MCE, electrophoretic analysis is performed in the short microchannels of the chip

and the channels can be fabricated for the analytical purpose. The small channel dimensions provide the advantage of applying high electricMANUSCRIPT field strengths that enable fast and high- resolution separation. The transition from benchto p analysis to the chip format has found

wide application in genomics, forensics peptidomics, proteomic, metabolomics, and

biomarker analysis [2–10] where microchip gel electrophoresis (MGE) and MCE can result

in performance improvements compared to classical electrophoresis. The possibility to

integrate a whole analytical workflow of sample preparation, separation, and detection to the

chip format further supports point-of-care diagnostics. This has also significantly influenced

the emerging field of single-cell analysis [11,12]. An exemplary demonstration is single cell

western blottingACCEPTED that enables the study of proteins from a single cell with a detection

sensitivity < 30,000 protein molecules [13].

Here, we highlight the advancements that amplified the potential of MCE for clinical

analysis including applications from single cell analysis. The last review on this topic

ACCEPTED MANUSCRIPT appeared more than a decade ago [14]. The focus is on targeted separations that were applied for the detection of diseases during the period 2008-2017. The keywords “microchip electrophoresis” and “clinical” were searched using the Scopus database. The review is

divided into two sections; (1) Background and integration of MCE for disease detection, and

(2) Clinical applications of MCE. The first section is for novice readers where we provide a

brief introduction of MCE and its core units followed by the separation and operation

principles. In the second section, we focus on the specific use of MCE for disease

interrogation. A comprehensive table based on the diseases studied is presented and this can

serve as an initial source of information for readers interested in MCE applications.

1. Background and integration of MCE for disease detection

A MCE setup can be described by the core elements of microfluidic structures, electrophoretic separation, and detection unit that build up the miniaturised analytical MANUSCRIPT laboratory. The application of an electric field can be used for sample manipulation (e.g., purification, sample concentration) in addition to the separation. Figure 1 shows a schematic of a typical MCE system with a T-shaped microfluidic chip and the required operational elements. The fluidic channels of the chip, sample reservoirs, and buffer reservoirs are primed for analysis using appropriate solutions (e.g., background solution, sieving gels, etc.) and fluidic controls (e.g., syringe pumps, pipetting). Once the chip has been prepared and the sample injected, a high voltage supply (HV supply) provides the electric field for separation. The detector, typicallyACCEPTED placed at the end of the separation channel, registers the separated zones and transmits the data to signal processing unit for conversion into an output format

(e.g., electropherogram). This section provides a brief overview on the MCE core elements including device fabrication, on-chip sample preparation, electrophoretic separation, and target detection.

ACCEPTED MANUSCRIPT

1.1 General approaches to device fabrication and channel design

1.1.1. Device fabrication

The microfluidic chips can be fabricated with exquisite control over the fluidic structures and these structures can be suited for disease detection. In general, a microfluidic electrophoresis chip consists of a separation channel that is connected to sample and buffer reservoirs. A wide range of materials can be used for chip fabrication and examples of common materials are glass, ceramics, and various polymers including poly(methyl methacrylate), cyclic olefin copolymer (e.g., TOPAS), polystyrene, and polycarbonate. The biocompatible and low background fluorescence poly(p-xylylene) polymers (e.g., Parylene-

C) was also explored for chip fabrication [15]. Disposable chips made from paper and fabrics have further attracted increasing attention which is mainly due to the relatively simple and low-cost device fabrication [16]. MANUSCRIPT The fabrication of microfluidic structures is generally performed by microfabrication techniques (e.g., photo-lithography, soft-lithography). Additive manufacturing is increasingly explored for microfluidic fabrication, although this fabrication process has not yet achieved the precision and resolution of lithography [17]. Similarly, direct printing on transparent polyester film with commercial office-type laser printer can be used as fabrication alternative. Although this fabrication method is rapid and cost-efficient, the resolution and quality of the printedACCEPTED microchannels are lower compared to photolithographic fabricated chips which can lead to shortcomings in separation efficiency and sensitivity. When filled with electrolyte or background solution (BGS), most of the fabrication materials provide an electroosmotic flow (EOF). The EOF can impede on separation and is thus often suppressed.

EOF suppression is obtained by dynamic and permanent channel coatings and use of

ACCEPTED MANUSCRIPT hydrogels [18–20]. In addition to EOF suppression, chemical modification of the microfluidic channels can also be applied to improve the method performance [21–24].

1.1.2. Channel design

Widely used chips have a crossed-channel or T-shaped design where the separation

channel is perpendicularly connected to sample and buffer channels (Figure 1 ). Other designs for MCE and MGE are also applied. For instance, the commercial Agilent

Bioanalyzer TM chip contains 16 reservoirs; 12 for samples and 4 for reagents and references.

The option to fabricate multiple channels and reservoirs supports high-throughput diagnostic

analysis and structures fitted for purpose.

The main modes of sample injection in MCE are injection by hydrodynamic pressure and voltage application. Hydrodynamic injection was implemented by using an on-chip peristaltic pump [28]. Novel injection strategies MANUSCRIPTinvolved controlled droplet injection using inkjet and array technology and these strategies provided advantages of high-throughput and precise nano- to picolitre volume injection [29,30]. The channel designs influenced the injection mode; electrokinetic injection was the preferred mode due to its simplicity and suitability for most channel designs. In this mode, the amount of injected sample is depending on time and applied voltage and biased towards high mobility samples (due to the electrokinetic bias [31]). Hydrodynamic injection was less common because this mode is more difficult to implement and often requires an auxiliary (e.g., pump or pipette) for injection. HydrodynamicACCEPTED injection is performed by difference in fluid levels of sample and sample waste reservoir, applying pressure, or vacuum.

The electrokinetic injection modes are pinched, floating, gated, and dynamic injections. In pinched injection, voltage is applied at sample, buffer, and buffer waste

ACCEPTED MANUSCRIPT reservoir. The sample waste reservoir is grounded. This results in injection of the sample into the channel junction. The channel junction is where the sample channel enters the separation channel. The pinched mode enables the injection of a relatively well-defined and small analyte plugs. However, the small plug of injected sample decreases the detection sensitivity. Floating injection is similar to the pinched mode; however, no potential is applied between buffer and buffer waste reservoir. This results in diffusion of the sample into the separation channel during the injection and thus increases the sample load. In gated injection, voltage is applied at sample and buffer reservoir and the two waste reservoirs are grounded.

Next, the voltage at the buffer reservoir is switched off which results in sample injection into the separation channel. This way different amounts of samples can be loaded, and the sensitivity is improved. After injection the voltage at the buffer reservoir is switched on again. The dynamic injection is a variation of the pinched mode, where an additional step is introduced. Electroosmosis is used in this additional step to pump sample into the separation channel and thus also increase sample load. MANUSCRIPT

1.2 On-chip sample preparation

The flexible fabrication of microfluidic chips offers the integration of sample

preparation including extraction, filtration, isolation/fractionation, and analyte enrichment

[7]. For instance, a solid-phase extraction unit was incorporated on a MCE device [32]. A

polymer solution was filled in a channel of the microfluidic chip and in-situ photopolymerisationACCEPTED was applied to create the monolithic absorbance material for on-chip solid-phase extraction. The extraction unit provided a model analyte enrichment of a factor

up to 80. Suitable analyte enrichment/stacking approaches for the chip setup are similar to

the approaches established in capillary electrophoresis (CE); field-

ACCEPTED MANUSCRIPT amplification/enhancement, transient isotachophoresis, sweeping, and micelle to solvent stacking [33–38].

Nucleic acids were enriched by an on-chip amplification strategy using polymerase

chain reaction (PCR) and digital loop-mediated isothermal amplification analysis [39]. A

single chip quantitative multiplexed PCR assay for phiX174 bacteriophage sequences and E.

Coli was demonstrated and enabled target amplification from as little as 50 DNA copies in the sample [40]. The assay of PCR and electrophoretic separation was completed in 45 min.

A two-chip system was demonstrated for optimised real time PCR followed by MGE separation of the amplicons [41]. The first chip was used for microfluidic real time PCR while the second chip was for separation of the amplicons. A heat sink and optimised PCR protocol was applied to shorten the DNA amplification process by factor 3 or approx. 30 min compared to typical 90 min. The heat sink enabled fast heating rates and isolated cooling within the small space of the microchip. Further, MANUSCRIPT the time for thermal equilibration was removed which further helped to speed-up the amplification process without compromise in the amplification of the model gene β2 microglobulin.

1.3 Electrophoretic separation

The electrophoretic separation of the sample is achieved by the application of an

electric field along the separation channel (Figure 1 ). The electric field results in electrophoretic ACCEPTED migration and flow of liquid called as EOF. The capillary electrophoretic separation modes are zone electrophoresis [42], electrokinetic chromatography [43,44], and

gel electrophoresis [45]. In MCE, the modes of zone electrophoresis, gel electrophoresis and

dielectrophoresis (DEP) are most commonly applied. DEP is used for isolation of particles

and cells under the influence of an non-uniform electric fields [46]. For instance, polystyrene

ACCEPTED MANUSCRIPT microbeads and PC-3 human prostate cancer cells could be separated with high confidence using self-assembled ionic liquid electrodes [47]. The capability of DEP to separate particles can also be used for on-chip extraction of plasma from whole blood [48]. In zone electrophoresis, charged analytes migrate through a liquid BGS and are separated according to their apparent electrophoretic mobility where the analytes are detected with descending mobility. Zone electrophoresis has found wide application including the separation of small molecules such as drugs and metabolites.

In gel electrophoresis, the electric field is applied along a sieving matrix. The charged

analytes migrate through the matrix where larger molecules are more retained than smaller

molecules. The sieving matrix is made from porous gels and the gel mesh size affects the

analyte resolution. For instance, an increase in the concentration of highly cross-linked gels

generally improves resolution. However, high sieving matrix concentrations also increase the viscosity and gels with high viscosity can be diffi MANUSCRIPTcult to load into the microfluidic structures. Common gels are made from polyacrylamide, cellulose, agarose, and starch. Gel

electrophoresis is used for size-based separation of large biomolecules including proteins,

DNA, and RNA. Amplification of DNA and RNA by PCR is often applied to concentrate the

analyte prior to the separation. The transition from slab gel electrophoresis to the microchip

format offers improvements in analysis time and base pair (bp) resolution [49]. For instance,

MGE typically provides analysis times of 1-10 min and can achieve single base pair resolution. ACCEPTED

1.4 Target detection

After the separation in the microfluidic channel, sample detection is performed. The

detection was performed by the modes of laser-induced fluorescence (LIF),

ACCEPTED MANUSCRIPT chemiluminescence, electrochemical, and absorbance measurement. LIF was the most frequent detection mode accounting for ~3/4 of the reviewed articles. Reasons for the popularity of LIF might be the high detection sensitivity and wide availability of this detection mode. Further, LIF is the predominant detection mode for nucleic acid analysis.

In LIF, the fluorescent analyte is excited with a laser and the emitted light when the analyte returns to the ground state provides the detection signal. Autofluorescence is relatively uncommon and thus fluorescent derivatisation is generally applied to suit the analyte for detection. The derivatisation can be performed before injection and in-channel

[50]. Fluorescence detection can also be extended to simultaneously detect multiple analytes.

For instance, a two laser setup was used for parallel excitation in the visible-light and near- infrared wavelength range of three labelled metabolites [51]. Light and laser emitting diodes have attracted interest as low-cost and portable alternatives to more expensive lasers and were often integrated into commercial chip electrop MANUSCRIPThoresis instruments. Although LIF is a highly sensitive detection mode, the small injected sample volumes and narrow structures of the microfluidic chip diminish the detection sensitivity. A fluorescence signal amplification strategy for highly sensitive detection of pM-levels of interferon-gamma by chip electrophoresis-LIF was developed [52]. Hereby, the signal amplification was achieved by a repeated cycling of cleavage and accumulation of the fluorescence reporter. This cycling was dependent on target binding to the hairpin aptamer probe. ACCEPTED In chemiluminescence detection, the analyte is excited by the adsorption of energy released from a chemical reaction. The emitted light when the excited analyte returns to its ground state is measured. This detection mode is highly sensitive and can further be enhanced by using reaction modifiers to achieve fM to aM detection limits. An example of a

ACCEPTED MANUSCRIPT common chemiluminescence reaction is the oxidation of luminol. In this reaction, luminol is

oxidised in alkaline conditions using hydrogen peroxide as an oxidiser and horseradish

peroxidase (HRP) as the catalyst. The signal can further be enhanced when other modifiers

such as p-iodophenol are present in the luminol reaction.

In electrochemical detection, the modes of amperometric and capacitively coupled

contactless conductivity are generally applied. In amperometric detection, the current change

is measured when the analyte band passes the detector. Capacitively coupled contactless

conductivity detection is similar and changes in conductivity are registered, however, the

detection unit is not in contact with the separation solution. The detection probe is placed

outside the microfluidic channel. The relatively low cost and compact design of

electrochemical detectors were early recognised for its potential in portable and point-of-care

applications [53]. MANUSCRIPT 2. Clinical application of MCE

Tables 1 summarises the reviewed papers based on the type of disease. The table

includes 33 contributions on cancer, 9 on immune disorders, 9 on neurological diseases, 8 on

genetic disorders, 8 on cardiovascular diseases, 7 on infectious diseases and pathogens, 7 on

organ diseases and dysfunctions, 5 on diabetes and pancreatic dysfunction, and 4 on

reproductive disorders. ACCEPTED 2.1 Cancer

MCE was applied to investigate lung, gastric, blood, colorectal, prostate, oral, liver,

gynecologic (i.e., ovarian and cervical), and bladder cancers. General cancer biomarkers that

ACCEPTED MANUSCRIPT are typically found in the progression of this disease were also analysed. Anticancer drug screening was further investigated, for instance by the use of DEP cellular microarrays [54].

The MCE strategies to investigate cancer were by the quantification of proteins, peptides, and small molecules, and mutation analysis of DNA and RNA from tissue samples and biological fluids. An exemplary workflow for quantification of proteins, peptides, and small molecules consisted of sample preparation (e.g., target isolation, concentration, derivatisation) followed by electrophoretic separation and detection. Chip-based immunoassays were important for antigen detection. For DNA and RNA, an exemplary workflow included extraction of the nucleic acid, target amplification by PCR, and size-based gel electrophoretic separation followed by fluorescence detection. The extraction of DNA from cells was by cell lysis including the addition of proteases and RNases to break down proteins and RNAs, and centrifugation to separate the solid debris from the supernatant. The supernatant contained the DNA for further purificat MANUSCRIPTion and isolation by ethanol precipitation, phenol-chloroform extraction, or solid-phase extraction.

Widely used RNA isolation techniques included liquid-liquid extraction (e.g., acid guanidinium thiocyanate-phenol-chloroform extraction [55]) and solid-phase extraction procedures. However, the isolation of RNA is more difficult than for DNA since RNA is prone to degradation. For instance, a study on the effects of RNA storage time and conditions showedACCEPTED that RNA could be stored up to 30 months at -80 ºC and short time storage of up to 4 h on ice [56]. However, degradation occurred at room temperature and in saline solution (i.e., in 0.9 M NaCl). The degradation was determined using the RNA integrity value. This value represents a semi-quantitative evaluation of RNA quality. The value ranges from 1 to 10 where 1 is completely degraded and 10 is unmodified RNA. The RNA

ACCEPTED MANUSCRIPT integrity value is calculated based on peak shape and height obtained by MGE [57]. The value was further applied to the evaluation of RNA quality in renal carcinoma tissue samples

[58]. The tissue samples were stored for <6 to 30 months at -80°C. In addition, tissue samples were also stored in similar conditions as typically found during tissue transportation

(e.g., at room temperature, on ice, or saline solution). Storage at -80°C for up to 30 months and 4 h on ice provided integrity values of >7 or intact RNA, while degradation was detected when the RNA was stored in saline solution and at room temperature.

2.1.1 General cancer biomarkers

The general cancer biomarkers that are found in various types of cancer included telomerase activity, carcinoembryonic antigen, and DNA mutations in p53 oncogenes.

Telomerase activity has been associated with carcinogenesis where the telomerase activity inhibits the telomere shortening during MANUSCRIPT replication [59]. The telomeric repeat amplification protocol based on genetic analysis is widely used to assess the telomerase activity. This method was transferred to the chip format [60,61]. Although the chip format provided fast amplicon separation (e.g, 2 min) and sensitive detection (e.g., single cells analysis after 20 PCR cycles), multiple steps of sample preparation have to be performed manually (i.e., cell lysis, DNA extraction, and PCR amplification).

Carcinoembryonic antigen [62,63] and thymidine kinase 1 [64] were determined using a non-competitiveACCEPTED immunoassay and aptamer-based microchip electrophoresis assay with LIF detection. Carcinoembryonic antigen is a protein biomarker used in the diagnosis of colorectal, pancreatic, gastric, ovarian, and cervical carcinomas. In the non-competitive immunoassay, the sample was off-line incubated with a primary antibody for 45 min before a second step of incubation with a secondary fluorescein isothiocyanate (FITC)-labelled

ACCEPTED MANUSCRIPT antibody. Electrophoretic separation was required to separate the free FITC-labelled

secondary antibody, non-target complex of primary-secondary antibody, and target complex

of biomarker sandwiched between primary and secondary antibody. The addition of 30 mM

sodium dodecyl sulfate (SDS) to the BGS helped to suppress biomolecule adsorption on the

channel wall. This immunoassay afforded a target detection in 4 min with limits of detection

of 45.7 pg mL -1. The analysed healthy and patient serums showed carcinoembryonic antigen

concentration of 2.9-5.1 ng mL -1 and 158.3-270.1 ng mL -1, respectively. For comparison, in common fluorescence-based sandwich immunoassays, the antigen is sandwiched between capture and detection antibody. The capture antibody (e.g., anti-carcinoembryonic antigen antibody) is immobilised on a substrate. Instead of using an electrophoretic separation of free/unbound label from the antibody-antigen sandwich complex, multiple washing steps are required. Typically, this assay format requires 60 min of the result and the detection limits are < 5 ng mL -1 [65]. MANUSCRIPT Single-base mutations in the p53 gene have a higher prevalence for the development

of cancer and are the most commonly mutated gene in human cancers [66]. The p53 gene

composes more than 20000 bp, 11 exons, and 10 introns. Gene mutations in p53 were

analysed by single strand conformational polymorphism (SSCP) and heteroduplex analysis

(HA) in combination with MGE-LIF [67]. In the processes of SSCP and HA, different DNA

conformations of mutated and unmutated DNA are obtained which results in electrophoretic mobility differencesACCEPTED that can be separated. Five single-base substitutions in p53 exons 5–9 in 106 tissue samples were analysed in the blinded study and the samples were correctly

identified with 98% sensitivity and 100% specificity. The method could analyse as low as

15% mutant tissue from a mixed tissue biopsy sample that comprised of cancerous (mutant)

and healthy tissue material.

ACCEPTED MANUSCRIPT

2.1.2 Lung cancer

Mutations in the EGFR gene [68,69] and determination of neuron-specific enolase

[70] were used to study lung cancer.

Multiplexed ligation-dependent probe amplification [68] and fragment length assay

[69] followed MGE-LIF were applied to detect EGFR mutations in exons 19 and 21 of patients with non-small cell lung cancer. For the multiplexed ligation-dependent probe amplification, the tissue extracted DNA was subjected to amplification and the products were loaded into a separation gel containing 7 M urea for DNA denaturation. The modified heating element of the commercial instrument enabled to perform separations at 65 °C instead of 50 °C. The sieving matrix and elevated temperature provided fragment resolution of as low as 2 bp which was sufficient to resolve the small difference between the wild-type (i.e., 320 bp) and mutant fragment (i.e., 305 bp) MANUSCRIPTin less than 90 s of separation time with as little as 40 fluorescently labelled DNA fragments for injection.

An off-line competitive immunoassay for neuron-specific enolase based on chemiluminescence resonance energy transfer was demonstrated [70]. The competition was between HRP-labelled antigen and the unlabelled antigen in the sample. An increase in sample antigen concentration resulted in a decrease in signal intensity due to the competition with the HRP-labelledACCEPTED antigen. The chemiluminescence was triggered by HRP and luminol as the donors and fluorescein isothiocyanate as an acceptor in the presence of oxidiser solution containing hydrogen peroxide and signal enhancer p-iodophenol. The oxidiser solution was delivered from an additional channel in front of the detection point. This methodology afforded 4.5 pM detection limits and was evaluated on serum samples from

ACCEPTED MANUSCRIPT healthy subjects and lung cancer patients where >10-times higher antigen concentrations

were found in the patient samples (>96 pM). The successful integration of the immunoassay

with chemiluminescence detection required a careful optimisation of the separation

conditions. Relatively small changes in solution pH, ionic strength, and temperature affected

the chemiluminescence signal. This can result in a challenging situation for MCE devices

that aim for portability.

2.1.3 Blood cancer

The types of blood cancers that were studied included canine T-cell lymphoma, leukaemia and lymphoblastic leukaemia, follicular lymphoma, and myeloproliferative neoplasms. Canine T-cell lymphoma was studied by detection of the V and J variable regions in the T-cell receptor γ gene in an approach of target amplification followed by fast amplicon

separation in 10 s using an electric field gradient [71]. MANUSCRIPT Leukaemia and lymphoblastic leukaemia were investigated using variable number of tandem repeats (VNTRs) [72]. Hematopoietic stem cell transplantation is used in the treatment of leukaemia. To evaluate the outcome of the transplantation, the sequencing of the donor DNA in the receiver’s blood and cord blood was analysed to assess the degree of chimerism. Genomic DNA of the donor and receivers blood sample was amplified and a standard curve from mixed patient-donor VNTRs by serial dilution was created. As low as 6.3% donor DNAACCEPTED could be determined in the receiver’s blood.

Follicular lymphomas were investigated by identification of Bcl-2/IgH fusion [73] and B-cell clonality [74]. A specific translocation between the Bcl-2 and IgH gene is found in more than 85% of patients with nodal follicular lymphoma [75]. The size of the fusion

ACCEPTED MANUSCRIPT gene formed by the translocation is different from clone to clone. The fusion was detected by

MGE-LIF of the real-time PCR amplified products [73]. The method could differentiate

fusion genes with differences less than 9 bp or 5%. The B-cell clonality was analysed by

amplification of three subregions of the Ig heavy chain gene and rearrangements in the Ig κ

chain gene prior to the size-based separation and LIF detection using a simplified BIOMED-2

protocol [74]. The amplicons were subjected to heteroduplex analysis. A sensitivity of 10%

target DNA in non-target DNA was achieved. The method was compared to the same

protocol using CE, where CE provided better sensitivity and could detect as low as 1% target

DNA in non-target DNA.

Myeloproliferative neoplasms have different subtypes where polycythemia vera,

essential thrombocythemia, and primary myelofibrosis are the most common. The

identification of the subtype is important for treatment; however, the correct identification is challenging. High resolution melting analysis and MANUSCRIPT PCR-MGE-LIF were used for genotyping of 99 myeloproliferative neoplasms samples [76]. The approach analysed for mutations in

Janus kinase 2 and calreticulin genes. From the 99 samples, 60 showed a mutation in Janus

kinase 2. From the 39 samples without the Janus kinase 2 mutation, 14 had a mutation in the

calreticulin gene. The combined analysis of both genes improved the diagnostic accuracy.

2.1.4 Colorectal cancer ColorectalACCEPTED cancer was investigated by low-abundance KRAS mutations detection [77], microsatellite instability [78], and glycomic serum analysis [79].

The treatment with monoclonal antibodies cetuximab and panitumumab can help to improve the median overall survival and 5-year survival rates, however, only if the patients

ACCEPTED MANUSCRIPT have a non-mutated form of the KRAS gene [80]. A sensitive method for the detection of

mutations in the KRAS gene of colorectal cancer patients involved restriction fragment

length polymorphism and stacking MGE [77]. Figure 2 shows the methodological approach.

To detect the point mutations in codon 12, a 107-bp fragment of KRAS was amplified from

DNA templates extracted from cancer cells or tissue samples and amplified using mismatched primer PCR. The mismatched forward sense primer introduced a base to the unmutated or wild-type fragment for specific digestion with Mva I restriction endonuclease.

The digestion resulted in a 77 and 30 bp fragments that gave rise to two peaks in the electropherogram. The mutant remained undigested and a single peak was detected. The heterozygous contained wild-type and mutant DNA fragments which resulted in a combined electropherogram of the two. This combination of restriction fragment length polymorphism, stacking and fluorescence detection could detect as low as 0.01% mutant DNA mixed with wild-type DNA. For comparison, standard SDS-PAGE was three orders of magnitude less sensitive and could detect as low 10% mutant DNA. MANUSCRIPT

The characterisation of N-glycans in human serum by MALDI-MS and MGE-LIF was used to distinguish colorectal cancer patients from healthy subjects [79]. After digestion, isolation, and derivatisation of the N-glycans from 42 patients and 20 control serum samples, the analytes were electrophoretically separated and the electropherograms were correlated with the glycan mass spectrum information. The electropherogram revealed the presence of isomeric glycanACCEPTED species as two peaks which were detected as a single peak with an identical m/z ratio in the mass spectrum. An example of a patient’s N-glycan profile is shown in the electropherogram in Figure 3. Changes in the position of fucose and fucosyl isomeric species were found in the cancerous samples. The patient samples showed elevated serum

ACCEPTED MANUSCRIPT levels of fucosylated tri- and tetra-antennary glycans and this difference in concentration and

profile could be used for diagnosis.

2.1.5 Prostate cancer

Prostate cancer was studied by analysing the circulating cell-free DNA in patients

with hormone-refractory prostate cancer [81]. After extraction of the DNA from plasma,

quantitative PCR and MGE-LIF were used to determine the concentrations of cell-free DNA

during chemotherapy. The concentration of cell-free DNA increased on average from 13.3

ng mL -1 to 46.8 ng mL -1 from before treatment to after the first cycle of chemotherapy,

respectively. Cell-free DNA prior treatment was mainly composed of small DNA fragments

in the range of 160–200 bp while larger fragments of 200-10.4 kbp were found after

treatment. This was attributed to necrotic cell lysis and thereby the release of large fragments

into the bloodstream. MANUSCRIPT 2.1.6 Oral cancer

Skin cancer was investigated from patients with oral squamous cell carcinoma using

DNA methylation analysis and MGE-LIF [82]. The non-invasive strategy applied methylation-specific PCR to the extracted DNA (13 targeted genes) from oral rinse of 34 patients and 24 healthy subjects. Significantly elevated levels of DNA methylation was found in eight genes (i.e., e-cadherin (ECAD), transmembrane protein with epidermal growth factor-like and 2ACCEPTED follistatin-like domains 2 (TMEFF2), retinoic acid receptor beta (RAR β), O- 6 methyl-guanine DNA methyltransferase (MGMT), fragile histidine triad gene, WNT inhibitory factor 1, cyclin-dependent kinase inhibitor 2A, and death-associated protein kinase

1). Using a selected panel of four hyper-methylated genes (i.e., ECAD, TMEFF2, RAR β,

ACCEPTED MANUSCRIPT and MGMT), oral squamous cell carcinoma could be accurately diagnosed with 100% sensitivity and 87.5% specificity.

2.1.7 Liver cancer

Liver cancer was studied using immunoassay MCE for the determination of α- fetoprotein, carcinoembryonic antigen, prothrombin, and des-γ-carboxy prothrombin.

Amperometric detection was integrated into an enzymatic immunoassay-based MCE device for sensitive detection of α-fetoprotein and carcinoembryonic antigen in patients with liver and colonic cancers and healthy subjects [83]. Figure 4 shows the immunoassay approach integrated on the T-shaped microchip. The sample was placed in reservoir 2 and electrokinetic injection introduced the off-line labelled analytes into the separation channel

(i.e., a channel connecting 1 and 4). An additional channel and reservoir (5) were fabricated to the T-shaped chip in order to deliver the substrate for enzymatic reaction (i.e., 2- aminophenol). A voltage of 1.0 kV was applied toMANUSCRIPT this channel to introduce the substrate into the separation channel. The separation buffer contained hydrogen peroxide to complete the enzymatic assay together with HRP and the substrate to produce the electroactive product

(i.e., 3-aminophenoxazine) for amperometric detection. Amperometric detection can be label-free and has a comparable sensitivity to fluorescence detection. However, to avoid detection interferences occurring from the applied electric field, the amperometric detector had to be shielded or decoupled from the electrophoretic separation. ACCEPTED Sensitive determination of α-fetoprotein is important because it could be used to predict the recurrence of cancer after curative treatment and for detecting small tumours at an early stage of the cancer progression. Highly sensitive immunoassays for α-fetoprotein were achieved by transient isotachophoresis and fluorescence detection [84] and lectin-affinity

ACCEPTED MANUSCRIPT chip electrophoresis [85]. The lectin-affinity electrophoresis was a liquid-phase binding assay that achieved a detection sensitivity of 0.3 ng mL -1 analyte in serum. The transient isotachophoresis approach achieved 5 pM sensitivities and the stacking was further beneficial to significantly sharpen the analyte band. Both detection sensitivities were much lower than the clinical cut-off concentration of 280 pM. The stacking immunoassay was then extended for the determination of lens culinaris agglutinin-reactive α-fetoprotein, total α-fetoprotein, and des-γ-carboxy prothrombin [86]. In a similar study, the diagnostic precision of liver cancer (i.e., hepatocellular carcinoma) could be improved by considering a panel of three biomarkers (i.e., α-fetoprotein, fucosylated fraction of α-fetoprotein, and prothrombin induced by vitamin K absence-II ) that enabled the detection of 81.8% of early-stage hepatocellular carcinoma, 86.7% of small-sized tumour, and 91.7% of single tumour [87].

2.1.8 Gynecologic cancers Ovarian and cervical cancers were investigated MANUSCRIPT by the profiling of N-glycans of patient serum samples [88], identification of human papillomavirus [89,90], and determination of β-subunit of human chorionic gonadotropin [91]. The serum N-glycan from patients diagnosed with late-stage recurrent ovarian cancer were obtained after denaturation, disulfide bond reduction and enzymatic release of the glycans with peptide-N-glycosidase F.

The glycans were then isolated by solid-phase extraction on activated carbon micro spin columns followed fluorescence labelling with APTS (8-aminopyrene-1,3,6-trisulfonic acid). The labelled glycansACCEPTED were injected and the peak area of the 50-60 most intense peaks and their migration time was used for profiling and statistical evaluation (i.e., ANOVA and principal component analysis). The electropherogram profile of control, patients before and after drug treatment could clearly be differentiated. The assay was deemed as a highly accurate test according to the receiver operating characteristics (ROC) test >0.90.

ACCEPTED MANUSCRIPT

Some types of human papillomavirus (HPV) are linked to the development of cervical

cancer. A method for HPV identification applied restriction fragment length polymorphism

followed by gel electrophoretic separation with LED-induced fluorescence detection [89].

The extracted and purified DNA from tissue samples were subjected to PCR followed by the

digestion with four different endonucleases. The endonucleases cleaved the DNA amplicons

on virus type-specific sites and the separation of these fragments resulted in characteristic

electrophoretic profiles. The method provided detection limits of as low as 2 x 10 2 DNA

copies. A similar assay was applied for virus types HPV16 E6/E7 mRNA detection using

reverse transcription-polymerase chain reaction for target amplification [90].

The determination of β-subunit of the hormone human chorionic gonadotropin can be

used as a tumour marker for choriocarcinoma. A non-competitive immunoassay with chemiluminescence detection for the β-subunit MANUSCRIPT was demonstrated [91]. This assay was evaluated on serum samples from healthy subjects and ovarian cancer patients and analyte concentrations of 9.5-15.7 mIU mL -1 and 160.9-210.4 mIU mL -1 were found in healthy and cancer samples, respectively.

2.1.9 Bladder cancer

Bladder cancer was studied by multiplexed microfluidic immunoassay using DEP [92,93] and MGE-LIFACCEPTED [94]. The biomarker panel included galectin-1, lactate dehydrogenase B, and bladder cancer-specific genes FGFR3, TERT, HRAS, ALX4, RALL3, MT1A, and

RUNX3. Cell lysate from bladder cancer cell lines RT4 and T24 were analysed for protein

expression of galectin-1 and lactate dehydrogenase B by DEP immunoassay with antibody-

functionalised nanoprobes [92]. The cell lysate was incubated with the nanoprobes. DEP

ACCEPTED MANUSCRIPT was applied to focus, guide, and trap the nanoprobes on the microfluidic device using an

indium tin oxide electrode. The use of indium tin oxide was beneficial for fluorescence

imaging because it reduced the background noise level and improved the fluorescence

response.

Size-based filtration was applied as a fast approach for the isolation of bladder cancer

cells from human urine [94]. The cells were isolated using a polycarbonate membrane with 8

µm pores size. Three bladder cancer cell lines (i.e., T24, 5637, and J82) and human

umbilical vein endothelial cells were spiked into the artificial urine sample. The filter

membrane with cells was washed and the cells eluted with dichloromethane, washed,

centrifuged, and the cell pellet was dried before the genomic DNA extraction. Methylation-

specific PCR amplified the mutated and bladder cancer-specific genes FGFR3, TERT and

HRAS and methylated genes ALX4, RALL3, MT1A, and RUNX3. Identification of the amplicons was by MGE-LIF in less than 2 min. MANUSCRIPT

2.2 Immune disorders

Immune disorders including inflammatory diseases [95,96], allergic reactions and hypersensitivities [97–101], arthritis and gout [102,103] were investigated.

2.2.1 Inflammation For inflammatoryACCEPTED diseases, the general inflammation markers such as cytokines and C-reactive protein were determined by competitive and non-competitive immunoassays.

These immunoassays used the specific interaction of antibodies to capture the antigens. The

incubation of the sample with antibody was performed off-line or on-chip prior to the

electrophoretic separation of the antigen-antibody complex from the free label. In the

ACCEPTED MANUSCRIPT competitive format, the target competes for antibody binding with a labelled non-target. The decrease in labelled non-target is measured and inversely proportional to the target concentration. In the non-competitive format, the target is incubated with antibody and label, and the target concentration was proportional to the signal intensity of the target-antibody- label complex.

The chronic inflammation marker C-reactive protein was analysed using an off-line immunoassay followed by MCE-LIF [95] and immunoprecipitation-MGE-LIF [96]. The immunoassay was performed in a standard microtiter plate and with functionalised polymer beads. In this format, termed as cleavable tag immunoassay, a three-step incubation procedure was performed. First, microbeads functionalised with capture antibody were incubated with the sample. Second, the beads were incubated with the fluorescently-tagged detection antibody. Third, a cleavage solution (i.e., tris(2-carboxyethyl)-phosphine) was incubated to cleave the antigen from the capture anMANUSCRIPTtibody. The cleavage solution reduced the disulfide bond of the fluorescence tag and released the tag for subsequent detection. The fluorescence tags were then electrophoretically separated and each tag was specific to a single analyte. An advantage of this assay was that the cleaved label could be detected as a free molecule rather than antibody-antigen-label. Further, baseline separation of the different free labels was not required since each label provided a characteristic signal (barcode strategy). Although the separation was fast (2 min), the incubation time for the immunoassay and the cleavageACCEPTED step required multiple hours.

The C-reactive protein from serum of sepsis patients was isolated by immunoprecipitation following quantification by MGE-LIF [96]. To suit the samples for immunoprecipitation, interfering non-target proteins (e.g., immunoglobulin G and human

ACCEPTED MANUSCRIPT serum albumin) were removed using spin column cartridges followed by fluorescence

labelling of the analytes in the eluate. The immunoprecipitation was performed from 10 µL

eluate using antibody-functionalised magnetic beads. The target protein was eluted from the

beads and subjected to gel electrophoretic separation. This approach was well suited for

small sample volumes. As little as 50 µL of patient serum was sufficient as starting material

and the assay provided an analyte sensitivity of 126 pg µL-1.

2.2.2 Allergic reactions and hypersensitivities

To investigate allergic reactions, a non-competitive on-chip immunoassay was applied for the determination of inflammatory cytokines from skin biopsies in patients with atopic dermatitis to nickel [97,98] and in the cerebral spinal fluid of preterm babies with traumatic head injury [99]. The microfluidic chips combined affinity-based isolation of cytokines, fluorescence-labelling, and separation of labelled cytokines. The determined cytokines were interferon gamma, interleukin-6, and tumour MANUSCRIPT necrosis factor alpha. The on-chip immunoassays were by functionalisation of glass fibre disks with the capture antibodies (i.e., chimeric receptors for the targets). Microdissected skin samples were extracted prior to the preconcentration and purification using the immunoaffinity disks. Figure 5 (i) shows the methodological approach for the immunoaffinity disk assay. The sample was brought in contact with antibodies immobilised on the fibre disk (“1. Inject”) for 5 min followed by a wash step to remove the sample matrix (“2. Wash”). A syringe pump delivered a flow of labelling solutionACCEPTED to fluorescently label the captured cytokines (“3. Label”) before their elution with an acidic elution buffer (“4. Elute”). A voltage was applied to guide the labelled analytes into the separation channel following separation of the three targets in less than 3 min. The whole workflow was completed in less than 15 min and the assay provided sub-pg per microliter detection sensitivity. Figure 5 (ii) shows the electropherograms from the

ACCEPTED MANUSCRIPT analysis of (A) normal subject and patients with (B) mild inflammatory lesions and (C) severe recurrent inflammatory lesions. This example of the on-chip integration of an immunoassay and electrophoretic separation was suitable for very small sample volumes of as low as 0.5 µL and provided a faster and more sensitive diagnostic alternative to commonly used enzyme-linked immunosorbent assay (ELISA).

Warfarin hypersensitivity was analysed by genotyping using single nucleotide polymorphisms for genes cytochrome P450 variant 2C9 and VKORC1 [100,101]. In these approaches, PCR and electrophoretic separation were both integrated on-chip. In [100], a tetraprimer amplification refractory mutation system was selected, in which the Taq polymerase only amplified when the primer exactly matched the DNA. This selective amplification process was applied to identify the presence or absence of the target allele. 35 patients with and without warfarin sensitivity were analysed using as little as 270 nL sample for PCR. The total analysis time of this combined MANUSCRIPT microfluidic approach was 90 min for 12 samples.

In [101], the integrated and automated microchip system performed sample preparation, PCR amplification, and amplicon separation. A photograph of the integrated microchip system is shown in Figure 6A. Sample preparation and PCR amplification were performed with the disposable elongated chips (“Plastic DNA extraction and amplification chip”). These chipsACCEPTED were connected to the gel electrophoresis chip using a PDMS block. The entire analytical workflow was completed in 2 h and required as little as 0.5 µL sample. The method was evaluated on whole blood, dried blood spots and buccal swabs. An exemplary genotyping analysis is shown in Figure 6B , where the target DNA fragments corresponded

ACCEPTED MANUSCRIPT with the typical wild type fragment size. This strategy was suggested for personalised

warfarin dosing in clinical practice to mitigate side effects from drug overdosing.

2.2.3 Arthritis and gout

Arthritis and gout were investigated by on-chip electrophoretic quantification with

electrochemical detection of lactate in synovial fluid of arthritis patients [102] and uric acid

in urine of gout patients [103]. Synovial fluid is found in synovial joints (e.g., hip joint)

where the fluid reduces the friction in the joint during movement. Increased concentrations

of lactate are found in patients suffering from arthritis. The concentration of lactate in

synovial fluid was determined by MCE with contactless conductivity detection. Synovial

fluid samples from 10 volunteers were analysed and lactate concentration in the range of 0.54

to 5.37 mM was detected. The method was sensitive to detect as low as 6.5 µM and was

rapid with 40 s separation time. MANUSCRIPT 2.3 Neurological diseases

Neurological diseases including multiple sclerosis [104], epilepsy [105,106], bipolar disorder [107,108], and Alzheimer’s disease [109–112] were investigated

2.3.1 Multiple sclerosis

A protein profiling strategy to differentiate patients with multiple sclerosis and healthy subjectsACCEPTED using MGE-LIF was established [104]. Cerebrospinal fluid and serum samples were analysed to create an electrophoretic protein profile that included common

proteins such as albumins and α-, β-, and γ-globulins. Characteristic protein profiles were found that enabled the differentiation of patient and healthy samples.

ACCEPTED MANUSCRIPT 2.3.2 Epilepsy

Phenobarbital is administered in the treatment of epilepsy patients. Competitive

immunoassays for phenobarbital in serum samples of epileptic patients were performed

[105,106]. The serum samples were subjected to liquid-liquid extraction, evaporation and

reconstitution in a buffer prior to the immunoassay procedure. The extracted drug competed

with a fluorescently-labelled phenobarbital for binding to the antibody. The signal of the

labelled drug was inversely proportional to the analyte concentration in serum. The analysed

patient samples showed concentrations in the range of 66.4–155.6 µM which were all within

the therapeutic range (i.e., 43-172 µM).

2.3.3 Bipolar disorders

Lithium is used as a mood-stabiliser in the treatment of bipolar disorders. Due to the narrow therapeutic range of this drug in the blood (i.e., 0.4-1.2 mmol L -1), a close monitoring of the blood lithium concentration is recommended. MANUSCRIPT Lithium was determined using MCE with conductivity detection from whole blood [107], and urine samples [108].

In [107], the point-of-care device for whole blood analysis consisted of a disposable

microfluidic chip that was sealed and prefilled with buffer solution. To introduce the

positively-charged ions from a droplet of blood, a voltage with a positive sign at sample inlet

reservoir was applied to introduce the ions into the orthogonal separation channel. The negatively-chargedACCEPTED blood cells and anions remained in the sample reservoir. Destacking or band broadening of the injected analytes occurred due to the high sample conductivity. To

enable quantification and to account for the effect of destacking, the sample conductivity was

on-chip measured before sample injection, and an internal standard was further added to the

sample. After injection, lithium was separated from sodium and other small inorganic ions

ACCEPTED MANUSCRIPT and electrochemically detected. The development of a disposable chip required to address

some technical challenges. For instance, buffer evaporation and temperature fluctuations had

to be considered for achieving stable device performance. The time for sample-in-answer-out

was 3.5 min including analyte separation in 30 s.

2.3.4 Alzheimer’s disease

Alzheimer’s disease is a neurodegenerative disease that is associated with elevated

levels of phosphorylated tau protein and β-amyloids, and the toxic protein aggregation in the

brain. An on-chip hyphenation of isoelectric focusing and zone electrophoresis strategy was

applied to study protein aggregation [109]. The β-amyloid peptides were investigated by

immunocapture of the peptides from cerebrospinal fluids and subsequent electrophoretic

enrichment and separation prior to the fluorescence [110] and mass spectrometric detection

[111]. The effect of β-amyloid Aβ25–35 treatment on the concentration of essential inorganic ions (i.e., Na +, K +, Ca 2+ , and Mg 2+ ) in neuron-like MANUSCRIPT PC-12 cells was studied with single-cell resolution [112]. In order to achieve single-cell injection, the cell solution was introduced to the microfluidic channel and a short voltage pulse (i.e., 1s at 800 V) was applied to introduce a single cell from sample to the perpendicular separation channel. Immediately after this pulse, a high electric field was applied for single cell lysis following separation and fluorescence detection of the inorganic ions.

A deviceACCEPTED for the isolation of β-amyloid peptides and their mass spectrometric peptide identification was demonstrated [111]. Figure 7(A) shows the methodological workflow.

Before the sample introduction, the bulk β-amyloid peptides from patient’s cerebrospinal

fluids were isolated by immunopurification. The samples were mixed with ampholytic buffer

to suit for the isoelectric focusing (“1. IEF”). Isoelectric focusing separated the β-amyloid

ACCEPTED MANUSCRIPT peptides according to their isoelectric point in a pH-gradient matrix. The peptide zones were

sampled separately with a short piece of a fused-silica capillary (“2. Sampling”), transferred,

and deposited on one micropillar of the micropillar array (“3. Deposition”). A photograph of

the micropillar with a nanodroplet of the sample is shown in the inset of “3.”. Each sample

fraction on the micropillar array was analysed by matrix-assisted laser desorption mass

spectrometry. Figure 7(B) shows the isoelectric focusing enriched β-amyloid peptides (“ii”), which enabled the mass spectrometric identification of more β-amyloid peptide species

compared to sample treatment using only immunoprecipitation (“i”).

2.4 Genetic disorders

The genetic disorders related to glucose-6-phosphate dehydrogenase deficiency [113],

spinal and bulbar muscular atrophy [114], β-thalassemia [115], and von Gierke's disease

[116] were studied using genotyping and mutation analysis combined with MGE-LIF. Further, the relationship of androgen receptor CAGMANUSCRIPT repeat genotype to muscle strength, testosterone level and finger length was studied in male subjects [117]. The D-enantiomer of

the drug penicillamine, which is administered in the therapy to treat Wilson’s disease, was

determined using a compact and inexpensive microfluidic platform with LED-induced

fluorescence detection [118]. Genetic disorders in dogs were also investigated and included

collie eye anomaly [119] and neuronal ceroid lipofuscinosis [120].

2.4.1 Glucose-6-phosphateACCEPTED dehydrogenase deficiency Mutations in the X-linked gene Xq2.8 are linked to the development of glucose-6-

phosphate dehydrogenase deficiency [113]. This deficiency is a predisposition for the

spontaneous destruction of red blood cells. Samples from 70 patients were analysed using

restriction fragment length polymorphism and fragment size separation by MGE-LIF. The

ACCEPTED MANUSCRIPT mutant fragments were of different length compared to the wild-type restriction patterns and could be separated and identified by MGE-LIF.

2.4.2 Spinal and bulbar muscular atrophy

Spinal and bulbar muscular atrophy and amyotrophic lateral sclerosis have similar symptoms and can be misdiagnosed [114]. Using CAG base repeat expansion analysis with a forward primer (i.e., 5’-ACCTACCGGCACCCAGAG-3’) and reversed primer (i.e., 5’-

CTCATCCAGGACCAGGTAGC-3’), 302 patient samples diagnosed with amyotrophic lateral sclerosis were subjected to screening analysis by MGE. The genetic diagnosis was performed by detecting the expansion of a CAG repeats in the androgen receptor gene. The expansion in repeats leads to differences in size that were electrophoretically separated. In the 302 samples, 4 could further be diagnosed with spinal and bulbar muscular atrophy that was initially diagnosed with amyotrophic lateral sclerosis. MANUSCRIPT 2.4.3 β-thalassemia

Specific mutations in the β-globin gene are associated with the development of β- thalassemia, a genetic blood disorder with little to severe symptoms. Prenatal diagnostic of the mutations in the β-globin is recommended for people carrying the gene since the disease might require long-term care. However, invasive prenatal diagnostic has risks for mother and child and is recommended for high-risk groups only [121]. A standard methodology is blood sampling of theACCEPTED chorionic villi followed by DNA extraction, amplification and sequencing. Less invasive venous blood sampling is challenging, because of the low abundance of the child’s DNA (<5%) in the mother’s blood. A peptide nucleic acid-mediated enriched PCR

(PNA-PCR) amplification followed by MGE sequencing of the mutant and wild-type DNA sequences was applied to detect foetal β-globin mutations in maternal plasma [115]. The

ACCEPTED MANUSCRIPT PNA-PCR enabled selective inhibition of the maternal wild-type allele and to preferentially amplify foetal mutated sequences. This sensitivity improving strategy was key in enabling prenatal diagnostic from venous blood of the mother. 41 prenatal diagnoses on maternal plasma were performed and compared to invasive extraction of foetal DNA from chorionic villi that were analysed by the standard procedure. The results of the chip approach were in agreement with the standard method.

2.4.4 von Gierke disease

A disposable chip for real-time PCR and high-resolution melt analysis was developed and validated for the diagnostic of von Gierke disease, an inherited glycogen storage disease

[116]. The chip made from low autofluorescence cyclic olefin thermoplastic was patterned with thin film metal electrodes. The small dimension of the chip and chip material properties enabled very fast heating and cooling rates that were crucial for PCR. A single PCR cycle was completed in 14 s and the full amplification MANUSCRIPTprocess required 8.5 min. Human genomic DNA was amplified using the G6PC primers and the obtained product was identified by on- chip high-resolution melt analysis where the device provided accurate melting temperature with a variation of 0.6 % to the expected melt temperature. The amplicons were further analysed by MGE-LIF.

2.5 Cardiovascular diseases

Acute myocardial infarction [122,123], coronary artery disease [124–128], and hypertension [129]ACCEPTED were investigated.

2.5.1 Acute myocardial infarction

ACCEPTED MANUSCRIPT Increased levels of serum deoxyribonuclease I have been reported in patients with myocardial infarction [130]. The first determination of deoxyribonuclease I using PCR followed by MGE-LIF was demonstrated [122]. Deoxyribonuclease I digests double- stranded DNA to oligonucleotides. DNA size standards of 100 and 800 bp were incubated with the enzyme for up to 80 min prior to the injection and size-based electrophoretic separation. The MGE gel contained ethidium bromide to stain the undigested DNA.

Digestion of the DNA resulted in a decrease in signal intensity that was correlated to the enzyme activity. The detection was completed in approx. 10 min which was substantially faster than widely used ELISA and single radial enzyme diffusion assay that require up to 20 hrs.

2.5.2 Coronary artery disease

Coronary artery disease as a result of atherosclerosis was investigated by determining the composition of lipoproteins in the blood. The MANUSCRIPT class of lipoproteins is divided into high- density lipoprotein (HDL) and low-density lipoproteins (LDL). The LDL are further subdivided into small, dense low-density lipoprotein (sdLDL), large, buoyant LDL (lLDL), and very low-density lipoproteins (vLDL). The isolation of these lipoproteins is difficult and is commonly performed by ultracentrifugation [131]. The serum lipoprotein contents of patients with atherosclerosis were analysed by a workflow of ultracentrifugation, fluorescent derivatisation of the isolated fractions, and electrophoretic separation with fluorescence detection [124–128].ACCEPTED Off-line derivatisation was performed with 7-nitrobenz-2-oxa-1,3- diazole-C6-ceramide. The separation of the lipoprotein fractions was difficult because the analytes were prone to channel wall adsorption and the small difference in electrophoretic mobility further complicated the separation. A dynamic channel coating with n-dodecyl-β-D- maltoside and hydroxypropyl cellulose as sieving matrix was applied and resulted in partial

ACCEPTED MANUSCRIPT to baseline separation of the lipoproteins. A further separation improvement was obtained

when 5 nm diameter gold nanoparticles were added to the sample and BGS solutions. The

nanoparticles caused a peak focusing of the various lipoproteins types and detection was

completed in 2.4 min. This method was then applied to analyse 65 serum samples; 35 with

previously diagnosed coronary heart disease and 30 healthy samples. The concentration of

sdLDL and vLDL in patients was higher than in healthy subjects. Treatment of patients with

atorvastatin, a lipoprotein-reducing drug, was shown to lower the concentrations of sdLDL

and vLDL.

2.5.3 Hypertension

The insertion and deletion polymorphism in the angiotensin-converting enzyme gene

is associated with hypertension. A strategy of MGE-LIF using an electric-field gradient was

explored for mutation detection in the Alu element of the angiotensin-converting enzyme gene [129]. The target DNA had a length of 190 MANUSCRIPT bp and 490 bp after deletion and insertion, respectively. These fragments were amplified by 10 cycles of PCR followed by the gel

electrophoretic separation using a stepwise electric field gradient of 129 to 615 V cm -1. Fast polymorphism detection was achieved in less than 25 s. The mutation analysis was compared to slab-gel electrophoresis where the MGE strategy was 300-times faster.

2.6 Infectious diseases InfectionsACCEPTED caused by viruses (i.e., hepatitis B and influenza) and bacteria (i.e., Escherichia coli and Staphylococcus aureus ) were investigated using strategies of

genotyping, immunoassays, and zone electrophoresis.

2.6.1 Viral infections

ACCEPTED MANUSCRIPT Viral infections by influenza A/B and hepatitis B were studied by MGE-LIF, DEP-

LIF, and immunoassay MCE [132–134]. Influenza A subtype H1N1 was analysed by PCR amplification of hemagglutinin (116 bp) and nucleocapsid protein (195 bp) genes prior to the subsequent amplicon separation in a multichannel MGE-LIF device [132]. An electric field gradient and channel coating with 1% poly(vinylpyrrolidone) improved amplicon separation.

Rapid separation (i.e., 20 s) and parallel sample analysis were achieved rendering the device suitable for high-throughput analysis. Using a droplet DEP device that integrated multiplexed and quantitative reverse transcription PCR, influenza A and B from nasopharyngeal and throat swabs were analysed [133]. The sample droplets were guided by

DEP to eight chambers for PCR amplification and subsequent fluorescence read-out. Eight samples were simultaneously analysed in less than 40 min and the RNA detection limit was less than 10 copies per chamber.

Hepatitis B was diagnosed using an immunoassay-base MANUSCRIPTd strategy where the enzymatic activity of alanine aminotransferase in patient serum samples was determined [134]. The determination of the enzyme activity is used as an indicator of liver cell injury such as can occur in hepatitis B patients. Alanine aminotransferase reversibly converts L-alanine and α- ketoglutarate into pyruvate and L-glutamate, respectively. Known concentrations of pyruvate and L-glutamate were added to serum samples and based on the concentration change of the substrates, the enzyme kinetic constants could be calculated using Lineweaver–Burk plots. Patient’s samplesACCEPTED showed a decrease in L-glutamate concentration compared to healthy subjects which were a consequence of higher levels of the enzyme.

2.6.2 Bacterial infections

ACCEPTED MANUSCRIPT Bacterial infections were identified on a paper-based MCE device [135] and bacterial resistance was investigated using DEP and Raman spectroscopy [136]. Furthermore, MCE with electrochemical detection was applied for the determination of antibiotic drugs in biological samples [137,138]. For instance, the antibiotic formulation sulperazone was determined using a novel gold nanoparticle-modified indium tin oxide microelectrode for amperometric detection [137]. The deposition of particles on the flat electrode surface increased the surface area and improved the detection sensitivity.

A paper-based MGE device was explored for the separation of characteristic amplicons of Escherichia coli O157:H7 (i.e., rrsH gene, 121 bp) and Staphylococcus aureus

(glnA gene, 225 bp) [135]. The T-shaped microfluidic structures were written by a plotter

and sandwiched between layers of overhead projector covers and adhesive films. The sample

reservoirs were made from PDMS. The pathogen DNA was extracted from cell culture and amplified using a benchtop PCR thermal cycler MANUSCRIPT followed by fluorescence labelling. The device that was fabricated at material costs of less than US$ 2 separated the amplicons in less

than 3 min and was sensitive to detect as little as 9.3×10 1 copies (or 0.5 pg) of target DNA.

An antibiotic resistance test for Escherichia coli was demonstrated on a microfluidic platform that applied cell sorting by DEP and single cell read-out by Raman spectroscopy

[136]. Blood samples from sepsis patients were collected and cultivated. The cultured bacteria were treatedACCEPTED with different concentration of the antibiotic ciprofloxacin. DEP was used to trap single cells in the Raman laser focus for spectrum acquisition. The changes in

Raman peaks and peak profile correlated with antibiotic resistance of the bacteria and were used to determine the minimal inhibitory concentration. The results were obtained in 2 h and were in agreement with reference methods (i.e., broth microdilution, Vitek-2, E-test).

ACCEPTED MANUSCRIPT

2.7 Organ diseases and dysfunctions

The approaches to study the function of liver [139], kidney [140–143], lacrimal gland

[144,145], and thyroid gland [146] included miniaturised immunoassays, gel electrophoresis, and zone electrophoresis.

2.7.1 Liver dysfunction

Liver dysfunction in infants with hyperbilirubinemia was investigated by immunoassay and frontal analysis chip electrophoresis to determine the free bilirubin in serum [139]. The immunoassay was applied to determine the binding constant of bilirubin with human serum albumin which then could be used to calculate the free bilirubin based on the total bilirubin and the binding constant. The measurement of the free bilirubin in the serum of neonates has been recognised as a reliable indicator to predict the toxicity of hyperbilirubinemia [147,148]. MANUSCRIPT

2.7.2 Kidney function

Kidney function was evaluated based on measuring the urea concentration in human serum [140], creatinine and protein concentration in urine [142,143] and artificial human serum [141].

An indirectACCEPTED detection of urea was applied by determining the ammonium ion, the enzymatic reaction product of urea in the presence of urease [140]. The serum samples were incubated with urease which resulted in the production of the ammonium ion. After this enzymatic conversion, the sample was injected and the ammonium ion was quantitated using contactless conductivity detection. Creatinine in human urine was determined by MCE with

ACCEPTED MANUSCRIPT LED-induced fluorescence detection [141]. To remove particulate matter, the urine was

centrifuged and filtrated before the analyte derivatisation with FITC. The detection of the

baseline separated analytes was achieved in 30 s using a LED induced excitation at 490 nm

and emission detection at 523 nm. This method was evaluated as an alternative for standard

Jaffe’s colourimetric assay and provided results concordant with the colourimetric assay.

Creatinine and albumin were also determined using pencil-drawn graphite electrodes used in

electrochemical detection [143]. This method provided baseline separation of the two

analytes in less than 2 min with migration time reproducibility of >2% RSD. The obtained

detection sensitivities were 20 and 35 µM which were relevant for clinical diagnostics.

A urine protein profiling method was based on MCE-UV [33]. To avoid the proteins

of adhering to the channel wall, a BGS of 75 mmol L-1 borate buffer containing 0.8 mmol L -1

calcium lactate and 1% ethylamine at pH 10.55 was applied. Model proteins human albumin and human transferrin were detected at detection MANUSCRIPT limits of 1.5 g L-1. The method was then evaluated on 248 pre-treated urine samples and the results were consistent with the standard

American Helena agarose gel electrophoresis methodology. However, the agarose gel electrophoresis required 45 min while the chip-based strategy was completed in 4 min.

2.7.3 Lacrimal gland

Lacrimal gland function was studied by analysing the inorganic ion composition of human tear samplesACCEPTED [145] and the effect of intravenously administered pituitary adenylate cyclase activating polypeptide (PACAP) on the tear and endolymph protein composition in

rats and chicken [144]. The inorganic ions sodium and potassium are the main electrolyte in

tear fluid and these ions are relevant in osmotic regulation and corneal epithelium

maintenance. A cost-efficient microfluidic platform based on poly(methyl methacrylate)

ACCEPTED MANUSCRIPT cross-channel chip with pencil drawn graphite electrodes for contactless conductivity

detection was applied for separation and detection of sodium and potassium in 2 min with

detection sensitivities <6.8 µM. PACAP was administered to the rats and chicken prior to the

sampling of the tear and endolymph using sterile absorbent paper strips. The sample was

eluted from the filter strips into a denaturing buffer for the followed MGE. While the protein

composition in the endolymph of chicken did not observably change due to peptide injection,

the tear composition in rats changed. The peptide injection affected the protein profile of

proteins in the molecular mass range of 50 to 70 kDa.

2.7.4 Thyroid gland

Thyroid gland function was investigated by chemiluminescence immunoassay of thyroxine in human plasma [146]. In the competitive immunoassay, the sample was incubated for 15 min with a known amount of HRP-labelled analyte and low concentration of anti-thyroxine antibody. The electrophoretic sep MANUSCRIPTaration resulted in two signal peaks; one for the HRP-labelled analyte and HRP-labelled analyte-antibody complex. The unknown analyte in the sample competed with the HRP-labelled counterpart for binding with the antibody.

This competition caused a change in the peak ratio that could be used to indirectly quantify thyroxine. Detection of the two peaks was in less than 1 min and afforded a low-nM sensitivity.

2.8 Diabetes and pancreatic dysfunction

DiabetesACCEPTED mellitus and diabetic complications [149–152], and acute pancreatitis [153] were studied by chip immunoassay and MGE-LIF.

2.8.1 Diabetes

ACCEPTED MANUSCRIPT Diabetes mellitus was studied by the analysis of haemoglobin-A1c, urinary proteins,

and human insulin-like growth factor-I. The determination of hemoglobin-A1c in the blood

can be used to assess the average three-month glucose concentration. Using an on-chip

sandwich immunoassay, the fraction of hemoglobin-A1c in total haemoglobin of human

blood was determined [149]. The assay used antibody-conjugated magnetic nanoparticles

that were prepared from layers of Au/chitosan/Fe 3O4. The anti-human hemoglobin-A1c antibody-functionalised magnetic beads were injected from one chip reservoir and immobilised in the device channel by a magnetic field. From another reservoir, the sample was injected and incubated for 8 min with the immobilised particles. After a washing step to remove any unbound sample, a solution of secondary anti-human haemoglobin antibody was incubated for 6 min. This secondary antibody bound to the captured analytes. The sandwich assay was completed by an incubation step with alkaline phosphatase conjugated goat anti- rabbit antibody that bound to the secondary antibody. The substrate solution (i.e., 1-naphthyl phosphate) was then added to the sample reservoir MANUSCRIPT and a substrate plug was injected to separation channel. Upon application of separation voltage, the substrate plug passed the immobilised alkaline phosphatase where the conversion from 1-naphthyl phosphate to the electrochemically-active 1-naphtol occurred. The analysed blood samples (after blood haemolysis) had a fraction of hemoglobin-A1c in the range of 4.84 to 10.31%. Although the assay platform provided the analytical result within 30 min, several assay parameters (e.g., particle synthesis, particle incubation time, and magnetic field) have to be carefully optimised and controlled. ACCEPTED

The urinary proteins transferrin, β2-microglobulin, human serum albumin, and immunoglobulin G are biomarkers used in the evaluation of the kidney and urinary tract function of diabetes mellitus and diabetic nephropathy patients. An integrated microfluidic

ACCEPTED MANUSCRIPT platform for sample clean-up, enrichment and detection of the four urinary proteins was demonstrated [150]. Sample clean-up and enrichment was performed by transient isotachophoresis. Transient isotachophoresis could tolerate the high ionic strength urine samples without requiring additional sample clean-up and also provided a sample enrichment of factor 40 that supported sensitive analysis. The microchip was conditioned with leading electrolyte (i.e., Tris-HCl) followed by the sample injection. The terminating electrolyte (i.e., acetic acid) was added to the inlet buffer reservoir after sample injection and the electrophoretic process was started. The device provided concordant results with common radio-immunoassay but was completed in 10 min compared to the 7 h required in the radio immunoassay.

Human serum albumin in urine samples of 48 diabetes patients was analysed by

MGE-LIF and compared to standard immunoturbidimetry measurement [151]. For MGE- LIF, the albumin in the urine sample was first heatMANUSCRIPT denatured, diluted and spiked with an internal standard (i.e., chicken albumin) prior to the sample loading into the chip reservoir.

The sieving gel matrix contained a fluorescence dye for on-chip labelling of the analyte.

Interestingly, MCE detected higher levels of albumin (mean = 167.5 mg L -1) compared to immunoturbidimetry (mean = 98.2 mg L -1) which was attributed to the ability of MCE to measure immuno-reactive albumin which remains undetected by turbidimetry.

SensitiveACCEPTED detection of human insulin-like growth factor-I (IGF-I ) in serum was demonstrated in a competitive immunoassay followed by stacking MCE [152]. Stacking was by field-amplified sample stacking which was implemented by the preparation of the pre- treated serum samples in a low conductivity matrix. The reconstitute was incubated with a known concentration of mouse anti-human IGF-I monoclonal antibody for 30 min at 37 ºC

ACCEPTED MANUSCRIPT before the addition of a known concentration of fluorescently labelled green fluorescence

protein–IGF-I. The detected concentration of the unbound labelled green fluorescence

protein–IGF-I was inversely proportional to the analyte concentration. Stacking afforded an

enrichment of the free fluorescent IGF-I of three orders of magnitude and the assay provided

a detection sensitivity of 0.68 ng mL -1.

2.8.2 Acute pancreatitis

Acute pancreatitis was analysed by the determination of pancreatic α-amylase activity using immunoassay MGE-LIF [153]. A fluorescence labelled oligosaccharide was applied as a substrate for the amylase-mediated enzymatic conversion. The sample and label were mixed offline and allowed to react for 0-160 min. In order to exclusively determine the pancreatic α-amylase unbiased from the presence of salivary amylase, a salivary amylase inhibitor was added. The enzymatic activity was measured as the increase in the peak of the formed enzymatic product (i.e., oligosaccharide hydMANUSCRIPTrolysate) after electrophoretic separation of the substrate from the enzymatic product. The use of borate in the sieving matrix (i.e., sieving BGS was 0.5% methylcellulose in 0.2 M borate buffer at pH 8.5) resulted in complexation and conformational changes of the substrate that improved separation. This method enabled the determination of pancreatic amylase in the range of 5-500 U L -1 with a limit of detection of 4.38 U L-1. The analysis of 13 plasma samples showed pancreatic amylase concentrations in the range of 13 to 74 U L-1. ACCEPTED 2.9 Reproductive disorders

Reproductive disorders were investigated by the study of androgen disorders in male subjects [154], monitoring the female reproductive system [155], and detection of preterm biomarkers [156,157]

ACCEPTED MANUSCRIPT

2.9.1 Androgen and progesterone related disorders

Serum concentrations of testosterone in men and progesterone in women were determined using a competitive immunoassay format with highly sensitive chemiluminescence detection [154,155]. Both hormones are biologically important and serve as a biomarker for reproductive disorders. The off-line competitive immunoassay for the two hormones was performed by mixing the sample with a known amount of HRP-labelled hormone and anti-hormone antibody. The unlabelled hormone in the sample and the added label competed for binding with the antibody. A gradual decrease in peak intensity of the label was obtained as a consequence of the hormone in the sample. The subsequent electrophoretic separation was required to separately detect the free label and the hormone- antibody complex. The chemiluminescence signal was triggered by luminol, HRP, and hydrogen peroxide. The limit of detections for progesterone and testosterone were 3.8 nM and 1 nM, respectively. Progesterone concentration MANUSCRIPTs of <46 nM in controls and >413 nM pregnant women were determined. Testosterone concentration in men was found in the range of 14 to 24 nM.

2.9.2 Preterm birth

The birth of babies before the 37 th week of pregnancy is considered preterm birth and is a leading cause of complication during pregnancy. The risk of preterm birth has been predicted by a panelACCEPTED of peptide biomarkers in serum [158]. One of these preterm peptide biomarkers is a 19 amino acid long and proline-rich peptide that was determined in fortified serum samples using a microfluidic device [156,157]. The device integrated sample extraction (i.e., solid-phase extraction), fluorescence derivatisation, and separation of the biomarker. The biomarker was spiked in diluted serum samples and subjected to solid-phase

ACCEPTED MANUSCRIPT extraction that was embedded in the microfluidic channel. The extraction material could

tolerate the presence of 5% serum in the sample solution, however, the extraction yield

significantly decreased at this serum concentration. A 15-fold enrichment of the biomarker

was achieved which afforded a detection limit of 50 nM.

Outlook and challenges

In the last decade, MCE has become an established analytical platform that has

influenced today’s clinical environment. While in the early 2000s MCE was in a stage where

the main driver was on fabrication and proof-of-concept studies, MCE underwent a transition

that enabled the integration of a whole analytical workflow or truly “lab-on-a-chip”.

Although fast (sub)-min separation times were already achieved in the 1990’s, the rapidly

advancing fields of lab-on-a-chip and MCE provided ingenious solutions and methods that were pivotal to the study of biomarkers. For instaMANUSCRIPTnce, MCE has become an enabler in the study of single entities (e.g., single cell analysi s) where delicate manipulation and

interrogation of the single entity could be performed using purpose-built fluidic structures,

efficient separation, and sensitive detection. However, there are still challenges that hinder

the full promise of lab-on-a-chip.

The microfluidic chips were mainly used as the separation platform – an inherent

advantage of electrophoresis where a weightless electric field and micrometre-sized fluid filled channels areACCEPTED all that is required for separation. Sample preparation, fluidic control, and detection within the MCE devices are however relevant areas that need to be improved. Most

of the reviewed approaches relied on traditional benchtop equipment such as syringe pumps,

solenoid valves, spectrometers, and microscopes. The operation of this specialised

equipment also required skilled staff. A plausible avenue to address these issues is the

ACCEPTED MANUSCRIPT miniaturisation of the auxiliaries and a simplification of the analytical workflow. There were

demonstrations of “sample-in-answer-out” platforms that made use of compact optical and

electrochemical detection and reduced the number of auxiliaries. Another bottleneck is the

fabrication of MCE devices, which generally involves a multidisciplinary team of highly

trained personnel and a long sequence of fabrication steps. A simplification of the fabrication

process would support the spread and wider acceptance of MCE for diagnostics. Promising

solutions to fabricate (disposable) microfluidic chips at relatively low cost include paper-

based and polymer-based platforms that can be mass-produced.

Today’s diagnostic is generally performed in dedicated centres, which requires an efficient sample transport system from place of sampling (e.g., doctor’s office) to the centre.

This is not only intensive in logistics but also reduces the frequency of possible sampling and diagnostic testing. For preventive care and early disease detection, however, frequent testing is crucial. Making the diagnostic test easily acce MANUSCRIPTssible for both the health professionals and person to be tested would support the efforts for disease prevention and cure. MCE has the potential for on-site diagnostics and could serve as a “generic” platform for the development of novel assays for treatment monitoring at home. However, to truly realise this promise, more efforts should be geared towards the translation from the research laboratory into the clinical setting and ultimately to the end users’ home. The translation should result in validated “sample-in-answer-out” procedures that can be performed by traditionally trained clinical chemistsACCEPTED who will be able to analyse the results rapidly for the medical doctors to use. Although this is a difficult step, the route to the successful device translation will provide new opportunities and advance the field of MCE. In the end, the development of disease-specific and validated MCE methodologies and in an easy to use small benchtop

ACCEPTED MANUSCRIPT device will be attractive to end users or clinical analytical chemists whose main objective is

only to obtain patient data for medical doctors to suggest appropriate treatment.

Conclusion

MCE in clinical diagnostic was commonly used for genotyping, mutation analysis,

immunoassays, and small molecule detection. The MCE approaches provided advantages in

assay speed and flexibility compared to the traditional non-miniaturised techniques.

Sensitive on-chip biomarker detection was demonstrated by using sensitive detectors and

dedicated on- and off-line sample preconcentration strategies. However, due to the small

sample injection volumes, the detection limits were similar or inferior to established

methodologies. Genotyping and mutation analysis was applied to diagnose inherited diseases

and to investigate the effect of environmental factors on the genetic make-up. Mutation analysis in cancer samples was most common. GenotyMANUSCRIPTping and mutation analysis of nucleic acid samples was performed by MGE-LIF after target amplification by PCR. The microchip

format provided significant advances in speed and base pair resolution (e.g., single base pair

resolution) compared to the traditional slab gel electrophoresis. PCR was generally

performed using commercial off-line benchtop instrument while on-chip PCR was infrequent.

Technical challenges including probe adsorption on fluidic walls, sample evaporation, and

temperature control which might have been some of the reasons that hampered wider

integration of on-chip PCR . Immunoassays were applied for activity and concentration measurements ofACCEPTED proteins and peptides in circulating biological fluids. This was clinically- relevant in the diagnosis and treatment of cancer, immune disorders, infectious-,

neurological-, and reproductive diseases. In these immunoassays, the immuno-specific

interaction of antibody and antigen was performed on- and off-chip prior to the

electrophoretic separation of the free label and labelled target complex. The two main

ACCEPTED MANUSCRIPT detection approaches were LIF and chemiluminescence. For chemiluminescence, an

additional channel for delivery of luminescence reagents was integrated on the chip before

the detection point.

Small molecules and inorganic ions were separated by zone electrophoresis with absorbance and electrochemical detection. These analytes were drugs quantified in blood and blood-derived samples after adequate sample preparation. The on-chip implementation of compact electrochemical detectors enabled the fabrication of point-of-care device for lithium detection in whole blood. This example reflects what it is probably the most significant contribution of MCE to clinical diagnostics: the merge and match of the advances in microfabrication with powerful and compact electrophoretic separation. The advances in microfabrication made it possible to fabricate chips with purpose-built structures for diagnostics. However, qualitative and quantitative diagnosis on a miniaturised chip generally relies on a highly efficient separation step in ordMANUSCRIPTer to achieve analyte separation in the smallest channel dimensions. This is where electrophoresis as an ideal match to complete the diagnostic device found its application. At its core, electrophoresis only requires an electric field and supporting media to achieve separation and this is relatively unaffected by the small dimensions of the chip. Retrospectively, MCE has enriched the diagnostic field in the last decade and enabled high-speed analysis requiring only smallest sample sizes that were in favour of both the patients and clinicians.

AcknowledgementACCEPTED AW thanks the Swiss National Science Foundation SNSF (P2SKP2_168309) and the

University of Queensland Development Fellowship (UQFEL1831057). JPQ thanks the

Australian Research Council for funding (Discovery Project, DP180102810).

The authors have declared no conflict of interest.

ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED

ACCEPTED MANUSCRIPT References

[1] A. Manz, N. Graber, H.M. Widmer, Miniaturized Total Chemical Analysis Systems: A Novel Concept for Chemical Sensing, Sensors Actuators B Chem. 1 (1990) 244–248.

[2] S. Št ěpánová, V. Kaši čka, Analysis of proteins and peptides by electromigration methods in microchips, J. Sep. Sci. 40 (2017) 228–250. doi:10.1002/jssc.201600962.

[3] S. Št ěpánová, V. Kaši čka, Recent developments and applications of capillary and microchip electrophoresis in proteomic and peptidomic analyses, J. Sep. Sci. 39 (2016) 198–211. doi:10.1002/jssc.201500973.

[4] V. Kaši čka, Recent developments in capillary and microchip electroseparations of peptides (2013–middle 2015), Electrophoresis. 37 (2016) 162–188. doi:10.1002/elps.201700295.

[5] J.R. Kraly, R.E. Holcomb, Q. Guan, C.S. Henry, Review: Microfluidic applications in metabolomics and metabolic profiling, Anal. Chim. Acta. 653 (2009) 23–35. doi:10.1016/j.aca.2009.08.037.

[6] T.M. Sikanen, Microchip Technology in Metabolomics, in: Chromatogr. Methods Metabolomics, Royal Society of Chemistry, 2013: pp. 138–182.

[7] M. Sonker, V. Sahore, A.T. Woolley, Recent advances in microfluidic sample preparation and separation techniques for molecular biomarker analysis: A critical review, Anal. Chim. Acta. 986 (2017) 1–11. doi:10.1016/j.aca.2017.07.043.

[8] P. Liu, X. Li, S.A. Greenspoon, J.R. MANUSCRIPT Scherer, R.A. Mathies, Integrated DNA purification, PCR, sample cleanup, and capillary electrophoresis microchip for forensic human identification, Lab Chip. 11 (2011) 1041–1048. doi:10.1039/c0lc00533a.

[9] Y.T. Kim, H.Y. Heo, S.H. Oh, S.H. Lee, D.H. Kim, T.S. Seo, Microchip-based forensic short tandem repeat genotyping, Electrophoresis. 36 (2015) 1728–1737. doi:10.1002/elps.201400477.

[10] J.P. Han, J. Sun, B. Zhuang, P. Liu, X.C. Zhao, W.S. Li, A.Q. Ji, J. Ye, Y. Liu, C.X. Li, The Electrophoresis on Microfluidic Chips and Its Application in Forensic Science, Biochem. Biophys. 42 (2015) 1128–1135.

[11] R.N. Zare, S. Kim, Microfluidic Platforms for Single-Cell Analysis, Annu. Rev. Biomed. Eng. 12 (2010) 187–201. doi:10.1146/annurev-bioeng-070909-105238.

[12] Y. Lu, L. Yang, W. Wei, Q. Shi, Microchip-based single-cell functional proteomics for biomedicalACCEPTED applications, Lab Chip. 17 (2017) 1250–1263. doi:10.1039/C7LC00037E.

[13] C.C. Kang, J.M.G. Lin, Z. Xu, S. Kumar, A.E. Herr, Single-cell western blotting after whole-cell imaging to assess cancer chemotherapeutic response, Anal. Chem. 86 (2014) 10429–10436. doi:10.1021/ac502932t.

[14] S.F.Y. Li, L.J. Kricka, Clinical analysis by microchip capillary electrophoresis, Clin. Chem. 52 (2006) 37–45. doi:10.1373/clinchem.2005.059600.

ACCEPTED MANUSCRIPT [15] S. Sukas, A.E. Erson, C. Sert, H. Kulah, A parylene-based dual channel micro- electrophoresis system for rapid mutation detection via heteroduplex analysis, Electrophoresis. 29 (2008) 3752–3758. doi:10.1002/elps.200800164.

[16] T. Narahari, D. Dendukuri, S.K. Murthy, Tunable electrophoretic separations using a scalable, fabric-based platform, Anal. Chem. 87 (2015) 2480–2487. doi:10.1021/ac5045127.

[17] N.P. Macdonald, J.M. Cabot, P. Smejkal, R.M. Guijt, B. Paull, M.C. Breadmore, Comparing Microfluidic Performance of Three-Dimensional (3D) Printing Platforms, Anal. Chem. 89 (2017) 3858–3866. doi:10.1021/acs.analchem.7b00136.

[18] V. Kaši čka, Recent developments in capillary and microchip electroseparations of peptides (2011-2013), Electrophoresis. 35 (2014) 69–95. doi:10.1002/elps.201300331.

[19] A. Wuethrich, P.R. Haddad, J.P. Quirino, Zero net-flow in capillary electrophoresis using acrylamide based hydrogel, Analyst. 139 (2014) 3722–3726.

[20] A. Wuethrich, P.R. Haddad, J.P. Quirino, Off-line sample preparation by electrophoretic concentration using a micropipette and hydrogel, J. Chromatogr. A. 1369 (2014) 186–190.

[21] F. Kitagawa, S. Nakagawara, I. Nukatsuka, Y. Hori, K. Sueyoshi, K. Otsuka, Simple and rapid immobilization of coating polymers on poly(dimethyl siloxane)-glass hybrid microchips by a vacuumdrying method, Anal. Sci. 31 (2015) 1171–1175. doi:10.2116/analsci.31.1171.

[22] K. Mesbah, R. Verpillot, M. Chiari, A. PallandMANUSCRIPTre, M. Taverna, Neutral polymers as coatings for high resolution electrophoretic separation of A 2 peptides on glass microchips, Analyst. 139 (2014) 6547–6555. doi:10.1039/c4an01296h.

[23] B.G. Abdallah, A. Ros, Surface coatings for microfluidic-based biomedical devices, Woodhead Publishing Limited, 2013. doi:10.1533/9780857097040.1.63.

[24] Z.D. Kihm, E.M. Veen, J.D. Bergen-Hartigan, Y. Zhang, Y. Liu, Modification of electroosmotic flow for a polydimethylsiloxane electrophoresis microchip via polyelectrolyte coating., Anal. Sci. 28 (2012) 183–6. doi:10.2116/analsci.28.183.

[25] A.C. Henry, T.J. Tutt, M. Galloway, Y.Y. Davidson, C.S. McWhorter, S.A. Soper, R.L. McCarley, Surface modification of poly(methyl methacrylate) used in the fabrication of microanalytical devices, Anal. Chem. 72 (2000) 5331–5337. doi:10.1021/ac000685l.

[26] Y. Liu, J.C.ACCEPTED Fanguy, J.M. Bledsoe, C.S. Henry, Dynamic coating using polyelectrolyte multilayers for chemical control of electroosmotic flow in capillary electrophoresis microchips, Anal. Chem. 72 (2000) 5939–5944.

[27] J. Horvath, V. Dolník, Polymer wall coatings for capillary electrophoresis, Electrophoresis. 22 (2001) 644–655.

[28] V. Sahore, S. Kumar, C.I. Rogers, J.K. Jensen, M. Sonker, A.T. Woolley, Pressure- actuated microfluidic devices for electrophoretic separation of pre-term birth

ACCEPTED MANUSCRIPT biomarkers, Anal. Bioanal. Chem. 408 (2016) 599–607. doi:10.1007/s00216-015- 9141-0.

[29] T. Yasui, Y. Inoue, T. Naito, Y. Okamoto, N. Kaji, M. Tokeshi, Y. Baba, Inkjet injection of DNA droplets for microchannel array electrophoresis, Anal. Chem. 84 (2012) 9282–9286. doi:10.1021/ac3020565.

[30] S.U. Hassan, H. Morgan, X. Zhang, X. Niu, Droplet Interfaced Parallel and Quantitative Microfluidic-Based Separations, Anal. Chem. 87 (2015) 3895–3901. doi:10.1021/ac504695w.

[31] X. Huang, M.J. Gordon, R.N. Zare, Bias in Quantitative Capillary Zone Electrophoresis Caused by Electrokinetic Sample Injection, Anal. Chem. 60 (1988) 375–377.

[32] S. Kumar, V. Sahore, C.I. Rogers, A.T. Woolley, Development of an integrated microfluidic solid-phase extraction and electrophoresis device, Analyst. 141 (2016) 1660–1668. doi:10.1039/C5AN02352A.

[33] M.C. Breadmore, A. Wuethrich, F. Li, S.C. Phung, U. Kalsoom, J.M. Cabot, A.I. Shallan, A.S. Abdul Keyon, H.H. See, M. Dawod, J.P. Quirino, Recent advances in enhancing the sensitivity of electrophoresis and electrochromatography in capillaries and microchips (2014-2016), Electrophoresis. 38 (2017) 33–59. doi:10.1002/elps.201400420.

[34] A. Wuethrich, J.P. Quirino, Sensitivity enhancing injection from a sample reservoir and channel interface in microchip electrophoresis, J. Sep. Sci. 40 (2017). doi:10.1002/jssc.201601064. MANUSCRIPT [35] A. Šlampová, Z. Malá, P. Gebauer, P. Bo ček, Recent progress of sample stacking in capillary electrophoresis (2014–2016), Electrophoresis. 38 (2017) 20–32. doi:10.1002/elps.201600292.

[36] A. Wuethrich, P.R. Haddad, J.P. Quirino, The electric field - An emerging driver in sample preparation, TrAC - Trends Anal. Chem. 80 (2016). doi:10.1016/j.trac.2016.04.016.

[37] A. Wuethrich, P.R. Haddad, J.P. Quirino, Green sample preparation for liquid chromatography and capillary electrophoresis of anionic and cationic analytes, Anal. Chem. 87 (2015). doi:10.1021/ac504765h.

[38] Z. Malá, A. Šlampová, L. K řivánková, P. Gebauer, P. Bo ček, Contemporary sample stacking in analytical electrophoresis, Electrophoresis. 36 (2015) 15–35. doi:10.1002/elps.201400313.ACCEPTED

[39] Y.-D. Ma, K. Luo, W.-H. Chang, G.-B. Lee, A microfluidic chip capable of generating and trapping emulsion droplets for digital loop-mediated isothermal amplification analysis, Lab Chip. 18 (2017) 296–303. doi:10.1039/C7LC01004D.

[40] Y. Liu, C. Li, Z. Li, S.D. Chan, D. Eto, W. Wu, J.P. Zhang, R.L. Chien, H.G. Wada, M. Greenstein, S. Satomura, On-chip quantitative PCR using integrated real-time detection by capillary electrophoresis, Electrophoresis. 37 (2016) 545–552.

ACCEPTED MANUSCRIPT doi:10.1002/elps.201500298.

[41] V.N. Hoang, G. V. Kaigala, A. Atrazhev, L.M. Pilarski, C.J. Backhouse, Strategies for enhancing the speed and integration of microchip genetic amplification, Electrophoresis. 29 (2008) 4684–4694. doi:10.1002/elps.200800351.

[42] J.W. Jorgenson, K.D. Lukacs, Capillary Zone Electrophoresis, Science. 222 (1983) 266–272.

[43] S. Terabe, K. Otsuka, I. Kunimichi, T. Akihiro, T. Ando, Electrokinetic Separations with Micellar Solutions and Open-Tubular Capillaries, Anal. Chem. 56 (1984) 111– 113.

[44] J.P. Quirino, S. Terabe, Electrokinetic chromatography., J. Chromatogr. A. 856 (1999) 465–482.

[45] A.S. Cohen, D.R. Najarian, A. Paulus, A. Guttman, J.A. Smith, B.L. Karger, Rapid separation and purification of oligonucleotides by high-performance capillary gel electrophoresis., Proc. Natl. Acad. Sci. U. S. A. 85 (1988) 9660–9663. doi:10.1073/pnas.85.24.9660.

[46] H.A. Pohl, The motion and precipitation of suspensoids in divergent electric fields, J. Appl. Phys. 22 (1951) 869–871. doi:10.1063/1.1700065.

[47] M. Sun, P. Agarwal, S. Zhao, Y. Zhao, X. Lu, X. He, Continuous on-chip cell separation based on conductivity-induced dielectrophoresis with 3D self-assembled ionic liquid electrodes, Anal. Chem. 88 (2016) 8264–8271. doi:10.1021/acs.analchem.6b02104. MANUSCRIPT [48] C. Szydzik, K. Khoshmanesh, A. Mitchell, C. Karnutsch, Microfluidic platform for separation and extraction of plasma from whole blood using dielectrophoresis, Biomicrofluidics. 9 (2015) 1–16. doi:10.1063/1.4938391.

[49] J.-F. Hsieh, S. Sinchaikul, B.-M. Ou, S.-T. Chen, Studies on the analysis of human serum proteins by the gel-based microchip and SDS-PAGE, J. Food Drug Anal. 18 (2010) 434–439.

[50] A. Wuethrich, J.P. Quirino, Derivatisation for separation and detection in capillary electrophoresis (2015-2017), Electrophoresis. 39 (2017) 82–96. doi:10.1002/elps.201700252.

[51] Q. Li, P. Chen, Y. Fan, X. Wang, K. Xu, L. Li, B. Tang, Multicolor Fluorescence Detection-Based Microfluidic Device for Single-Cell Metabolomics: Simultaneous QuantitationACCEPTED of Multiple Small Molecules in Primary Liver Cells, Anal. Chem. 88 (2016) 8610–8616. doi:10.1021/acs.analchem.6b01775.

[52] Y. Qin, L. Zhang, S. Li, J. Zhao, Y. Huang, S. Zhao, Y.-M. Liu, A microchip electrophoresis-based fluorescence signal amplification strategy for highly sensitive detection of biomolecules, Chem. Commun. 53 (2017) 455–458. doi:10.1039/C6CC08911A.

[53] J. Wang, Electrochemical biosensors: Towards point-of-care cancer diagnostics,

ACCEPTED MANUSCRIPT Biosens. Bioelectron. 21 (2006) 1887–1892. doi:10.1016/j.bios.2005.10.027.

[54] L.-C. Hsiung, C.-L. Chiang, C.-H. Wang, Y.-H. Huang, C.-T. Kuo, J.-Y. Cheng, C.-H. Lin, V. Wu, H.-Y. Chou, D.-S. Jong, H. Lee, A.M. Wo, Dielectrophoresis-based cellular microarray chip for anticancer drug screening in perfusion microenvironments., Lab Chip. 11 (2011) 2333–2342. doi:10.1039/c1lc20147f.

[55] P. Chomczynski, N. Sacchi, The single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction: Twenty-something years on, Nat. Protoc. 1 (2006) 581–585. doi:10.1038/nprot.2006.83.

[56] J. Botling, K. Edlund, U. Segersten, S. Tahmasebpoor, M. Engström, M. Sundström, P.-U. Malmström, P. Micke, Impact of Thawing on RNA Integrity and Gene Expression Analysis in Fresh Frozen Tissue, Diagnostic Mol. Pathol. 18 (2009) 44–52. doi:10.1097/PDM.0b013e3181857e92.

[57] J.G. Slagter-Jäger, C.A. Nicolette, I.Y. Tcherepanova, Evaluation of a microfluidics- based platform and slab electrophoresis for determination of size, integrity and quantification of in vitro transcribed RNA used as a component in therapeutic drug manufacturing, J. Pharm. Biomed. Anal. 70 (2012) 657–663. doi:10.1016/j.jpba.2012.04.031.

[58] H. Sun, R. Sun, M. Hao, Y. Wang, X. Zhang, Y. Liu, X. Cong, Effect of Duration of Ex Vivo Ischemia Time and Storage Period on RNA Quality in Biobanked Human Renal Cell Carcinoma Tissue., Ann. Surg. Oncol. (2015) 1–8. doi:10.1245/s10434- 014-4327-9. [59] N.W. Kim, M.A. Piatyszek, K.R. Prowse, MANUSCRIPT C.B. Harley, M.D. West, P.L. Ho, G.M. Coviello, W.E. Wright, S.L. Weinrich, J.W. Shay, Specific association of human telomerase activity with immortal cells and cancer, Science. 266 (1994) 2011–2015. doi:10.1126/science.7605428.

[60] S.J. Kim, S.J. Choi, R. Neelamegam, T.S. Seo, Microchip-based capillary electrophoretic analysis of telomerase activity for cancer diagnostics, Biochip J. 4 (2010) 42–48. doi:10.1007/s13206-010-4107-y.

[61] K. Karasawa, H. Arakawa, Detection of telomerase activity using microchip electrophoresis, J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 993–994 (2015) 14–19. doi:10.1016/j.jchromb.2015.04.032.

[62] L. Pan, J. Zhao, Y. Huang, S. Zhao, Y.M. Liu, Aptamer-based microchip electrophoresis assays for amplification detection of carcinoembryonic antigen, Clin. Chim. Acta.ACCEPTED 450 (2015) 304–309. doi:10.1016/j.cca.2015.09.002. [63] F. Ye, M. Shi, Y. Huang, S. Zhao, Noncompetitive immunoassay for carcinoembryonic antigen in human serum by microchip electrophoresis for cancer diagnosis, Clin. Chim. Acta. 411 (2010) 1058–1062. doi:10.1016/j.cca.2010.03.035.

[64] J. V. Pagaduan, M. Ramsden, K. O’Neill, A.T. Woolley, Microchip immunoaffinity electrophoresis of antibody-thymidine kinase 1 complex, Electrophoresis. 36 (2015) 813–817. doi:10.1002/elps.201400436.

ACCEPTED MANUSCRIPT [65] S. Sharma, R. Raghav, R. O’Kennedy, S. Srivastava, Advances in ovarian cancer diagnosis: A journey from immunoassays to immunosensors, Enzyme Microb. Technol. 89 (2016) 15–30. doi:10.1016/j.enzmictec.2016.03.002.

[66] A.J. Levine, J. Momand, C.A. Finlay, The p53 tumour suppressor gene, Nature. 351 (1991) 453–456. doi:10.1038/351453a0.

[67] C.N. Hestekin, J.S. Lin, L. Senderowicz, J.P. Jakupciak, C. O’Connell, A. Rademaker, A.E. Barron, Blinded study determination of high sensitivity and specificity microchip electrophoresis-SSCP/HA to detect mutations in the p53 gene, Electrophoresis. 32 (2014) 2921–2929.

[68] M. Minarik, M. Gassman, B. Belsanova, M. Pesek, J. Schouten, R. Chudoba, B. Gas, L. Benesova, A novel high-resolution chipCE assay for rapid detection of EGFR gene mutations and amplifications in lung cancer therapy by a combination of fragment analysis, denaturing CE and MLPA, Electrophoresis. 31 (2010) 3518–3524. doi:10.1002/elps.201000156.

[69] U. Malapelle, S. Russo, F. Pepe, R. Sgariglia, C. De Luca, C. Bellevicine, P. Pallante, G. Troncone, EGFR mutation detection by microfluidic technology: a validation study, J. Clin. Pathol. 66 (2013) 982–984. doi:10.1136/jclinpath-2013-201730.

[70] T. Yang, M. Vdovenko, X. Jin, I.Y. Sakharov, S. Zhao, Highly sensitive microfluidic competitive enzyme immunoassay based on chemiluminescence resonance energy transfer for the detection of neuron-specific enolase, Electrophoresis. 35 (2014) 2022– 2028. doi:10.1002/elps.201300630. [71] K.K. Suresh, M.J. Lee, J. Park, S.H. Kang,MANUSCRIPT Microchip gel electrophoresis with programmed field strength gradients for ultra-fast detection of canine T-cell lymphoma in dogs, Talanta. 75 (2008) 49–55. doi:10.1016/j.talanta.2007.10.028.

[72] S.S. Daud, K. Ibrahim, S.S. Choong, L. Vengidasan, L.A. Chong, H. Ariffin, Microfluidic chip-based assay for post-hematopoietic stem cell transplantation chimerism monitoring using polymorphic tandem repeat markers, Anal. Biochem. 397 (2010) 181–185. doi:10.1016/j.ab.2009.10.008.

[73] Y. Tabe, Y. Kawase, K. Miyake, N. Satoh, N. Aritaka, Y. Isobe, K. Oshimi, N. Komatsu, T. Miida, A. Ohsaka, Identification of Bcl-2/IgH fusion sequences using real-time PCR and chip-based microcapillary electrophoresis, Clin. Chem. Lab. Med. 49 (2011) 809–815. doi:10.1515/CCLM.2011.141.

[74] A. Zamò, A. Bertolaso, A.W.M. Van Raaij, F. Mancini, M. Scardoni, M. Montresor, F. Menestrina,ACCEPTED J.H.J.M. Van Krieken, M. Chilosi, P.J.T.A. Groenen, A. Scarpa, Application of microfluidic technology to the BIOMED-2 protocol for detection of B- cell clonality, J. Mol. Diagnostics. 14 (2012) 30–37. doi:10.1016/j.jmoldx.2011.07.007.

[75] J.C. Aster, J.A. Longtine, Detection of BCL2 Rearrangements in Follicular Lymphoma, Am. J. Pathol. 160 (2002) 759–763. doi:10.1039/c2an15866c.

[76] N. Matsumoto, S. Mori, H. Hasegawa, D. Sasaki, H. Mori, K. Tsuruda, D. Imanishi, Y. Imaizumi, T. Hata, N. Kaku, K. Kosai, N. Uno, Y. Miyazaki, K. Yanagihara, 54

ACCEPTED MANUSCRIPT Simultaneous screening for JAK2 and calreticulin gene mutations in myeloproliferative neoplasms with high resolution melting, Clin. Chim. Acta. 462 (2016) 166–173. doi:10.1016/j.cca.2016.09.023.

[77] H. Zhang, J. Song, H. Ren, Z. Xu, X. Wang, L. Shan, J. Fang, Detection of Low- Abundance KRAS Mutations in Colorectal Cancer Using Microfluidic Capillary Electrophoresis-Based Restriction Fragment Length Polymorphism Method with Optimized Assay Conditions, PLoS One. 8 (2013). doi:10.1371/journal.pone.0054510.

[78] M. Odenthal, N. Barta, D. Lohfink, U. Drebber, F. Schulze, H.P. Dienes, S.E. Baldus, Analysis of microsatellite instability in colorectal carcinoma by microfluidic-based chip electrophoresis., J. Clin. Pathol. 62 (2009) 850–852. doi:10.1136/jcp.2008.056994.

[79] C.M. Snyder, W.R. Alley, M.I. Campos, M. Svoboda, J.A. Goetz, J.A. Vasseur, S.C. Jacobson, M. V. Novotny, Complementary Glycomic Analyses of Sera Derived from Colorectal Cancer Patients by MALDI-TOF-MS and Microchip Electrophoresis, Anal. Chem. 88 (2016) 9597–9605. doi:10.1021/acs.analchem.6b02310.

[80] C. Desantis, J. Ma, L. Bryan, A. Jemal, Breast Cancer Statistics , 2013, CA Cancer J. Clin. 64 (2013) 52–62. doi:10.3322/caac.21203.

[81] S. Kwee, M.A. Song, I. Cheng, L. Loo, M. Tiirikainen, Measurement of Circulating Cell-Free DNA in Relation to 18F-Fluorocholine PET/CT Imaging in Chemotherapy- Treated Advanced Prostate Cancer, Clin. Transl. Sci. 5 (2012) 65–70. doi:10.1111/j.1752-8062.2011.00375.x. [82] S. Nagata, T. Hamada, N. Yamada, S. Yokoyama,MANUSCRIPT S. Kitamoto, Y. Kanmura, M. Nomura, Y. Kamikawa, S. Yonezawa, K. Sugihara, Aberrant DNA methylation of tumor-related genes in oral rinse, Cancer. 118 (2012) 4298–4308. doi:10.1002/cncr.27417.

[83] S. Zhang, W. Cao, J. Li, M. Su, MCE enzyme immunoassay for carcinoembryonic antigen and alpha-fetoprotein using electrochemical detection, Electrophoresis. 30 (2009) 3427–3435. doi:10.1002/elps.200800805.

[84] T. Kawabata, H.G. Wada, M. Watanabe, S. Satomura, “Electrokinetic Analyte Transport Assay” for α-fetoprotein immunoassay integrates mixing, reaction and separation on-chip, Electrophoresis. 29 (2008) 1399–1406. doi:10.1002/elps.200700898.

[85] M. Kobayashi, T. Hosaka, K. Ikeda, Y. Seko, Y. Kawamura, H. Sezaki, N. Akuta, F. Suzuki, Y.ACCEPTED Suzuki, S. Saitoh, Y. Arase, H. Kumada, Highly sensitive AFP-L3% assay is useful for predicting recurrence of hepatocellular carcinoma after curative treatment pre- and postoperatively, Hepatol. Res. 41 (2011) 1036–1045. doi:10.1111/j.1872- 034X.2011.00858.x.

[86] F. Kanke, T. Kumada, H. Toyoda, S. Satomura, Reference change values for lens culinaris agglutinin-reactive α-fetoprotein and des-γ-carboxy prothrombin in patients with chronic hepatitis C, Clin. Chem. Lab. Med. 50 (2012) 957–960. doi:10.1515/cclm-2011-0912.

ACCEPTED MANUSCRIPT [87] J.Y. Choi, S.W. Jung, H.Y. Kim, M. Kim, Y. Kim, D.G. Kim, E.J. Oh, Diagnostic value of AFP-L3 and PIVKA-II in hepatocellular carcinoma according to total-AFP, World J. Gastroenterol. 19 (2013) 339–346. doi:10.3748/wjg.v19.i3.339.

[88] I. Mitra, W.R. Alley, J.A. Goetz, J.A. Vasseur, M. V. Novotny, S.C. Jacobson, Comparative profiling of N-Glycans isolated from serum samples of ovarian cancer patients and analyzed by microchip electrophoresis, J. Proteome Res. 12 (2013) 4490– 4496. doi:10.1021/pr400549e.

[89] L. Yi, X. Xu, X. Lin, H. Li, Y. Ma, J.-M. Lin, High-throughput and automatic typing via human papillomavirus identification map for cervical cancer screening and prognosis., Analyst. 139 (2014) 3330–5. doi:10.1039/c4an00329b.

[90] Q. Liu, X. Lin, L. Lin, L. Yi, H. Li, J.-M. Lin, A comparative study of three different nucleic acid amplification techniques combined with microchip electrophoresis for HPV16 E6/E7 mRNA detection, Analyst. 140 (2015) 6736–6741. doi:10.1039/C5AN00944H.

[91] S. Li, T. Yang, J. Zhao, Y. Huang, S. Zhao, Chemiluminescence noncompetitive immunoassay based on microchip electrophoresis for the determination of β-subunit of human chorionic gonadotropin, J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 1053 (2017) 42–47. doi:10.1016/j.jchromb.2017.03.031.

[92] C.-H. Chuang, T.-F. Wu, C.-H. Chen, K.-C. Chang, J.-W. Ju, Y.-W. Huang, V. Van Nhan, Lab on a chip for multiplexed immunoassays to detect bladder cancer using multifunctional dielectrophoretic manipulations, Lab Chip. 15 (2015) 3056–3064. doi:10.1039/C5LC00352K. MANUSCRIPT [93] C.H. Chuang, Y.W. Huang, Y.T. Wu, Dielectrophoretic chip with multilayer electrodes and micro-cavity array for trapping and programmably releasing single cells, Biomed. Microdevices. 14 (2012) 271–278. doi:10.1007/s10544-011-9603-x.

[94] Y. Deng, L. Yi, X. Lin, L. Lin, H. Li, J.M. Lin, A non-invasive genomic diagnostic method for bladder cancer using size-based filtration and microchip electrophoresis, Talanta. 144 (2015) 136–144. doi:10.1016/j.talanta.2015.05.065.

[95] Y. Sameenoi, M.M. Mensack, B.M. Murphy, C.S. Henry, Competitive, non- competitive, and mixed format cleavable tag immunoassays, Methods. 56 (2012) 166– 173. doi:10.1016/j.ymeth.2011.09.018.

[96] E. Herwig, M. Marchetti-Deschmann, C. Wenz, A. Rüfer, H. Redl, S. Bahrami, G. Allmaier, Sensitive detection of C-reactive protein in serum by immunoprecipitation- microchipACCEPTED capillary gel electrophoresis, Anal. Biochem. 478 (2015) 102–106. doi:10.1016/j.ab.2015.03.009.

[97] T.M. Phillips, H. Kalish, E. Wellner, Receptor affinity CE for measuring bioactive inflammatory cytokines in human skin biopsies, Electrophoresis. 30 (2009) 3947– 3954. doi:10.1002/elps.200900311.

[98] H. Kalish, T.M. Phillips, Assessment of chemokine profiles in human skin biopsies by an immunoaffinity capillary electrophoresis chip, Methods. 56 (2012) 198–203. doi:10.1016/j.ymeth.2011.12.003. 56

ACCEPTED MANUSCRIPT [99] T.M. Phillips, E. Wellner, Detection of cerebral spinal fluid-associated chemokines in birth traumatized premature babies by chip-based immunoaffinity CE, Electrophoresis. 34 (2013) 1530–1538. doi:10.1002/elps.201200634.

[100] B.L. Poe, D.M. Haverstick, J.P. Landers, Warfarin genotyping in a single PCR reaction for microchip electrophoresis, Clin. Chem. 58 (2012) 725–731. doi:10.1373/clinchem.2011.180356.

[101] B. Zhuang, J. Han, G. Xiang, W. Gan, S. Wang, D. Wang, L. Wang, J. Sun, C.-X. Li, P. Liu, A fully integrated and automated microsystem for rapid pharmacogenetic typing of multiple warfarin-related single-nucleotide polymorphisms, Lab Chip. 16 (2016) 86–95. doi:10.1039/C5LC01094B.

[102] Z. Song, Y. Xu, Z. Chen, J. Yang, X. Li, Z. Zhang, Quantification of lactate in synovia by microchip with contactless conductivity detection, Anal. Biochem. 434 (2013) 73– 77. doi:10.1016/j.ab.2012.11.006.

[103] A. Fernández-la-Villa, V. Bertrand-Serrador, D.F. Pozo-Ayuso, M. Castaño-Álvarez, Fast and reliable urine analysis using a portable platform based on microfluidic electrophoresis chips with electrochemical detection, Anal. Methods. 5 (2013) 1494. doi:10.1039/c2ay26166a.

[104] O. Tren čevska, V. Aleksovski, K. Stojanoski, Advanced Techniques in Clinical Practice: Use of Lab-on-a-Chip Electrophoresis and Other Methods in Protein Profiling, J. Med. Biochem. 28 (2009) 274–278. doi:10.2478/v10011-009-0026-7.

[105] Y. Huang, S. Zhao, M. Shi, J. Liu, H. Liang, Competitive immunoassay of phenobarbital by microchip electrophoresis MANUSCRIPT with laser induced fluorescence detection, Anal. Chim. Acta. 694 (2011) 162–166. doi:10.1016/j.aca.2011.03.036.

[106] Y. Huang, S. Zhao, M. Shi, H. Liang, One-way multiplexed immunoassay strategy for simultaneous determination of multi-analytes by microchip electrophoresis, Analyst. 136 (2011) 2119. doi:10.1039/c0an00836b.

[107] S. Staal, M. Ungerer, A. Floris, H.W. Ten Brinke, R. Helmhout, M. Tellegen, K. Janssen, E. Karstens, C. van Arragon, S. Lenk, E. Staijen, J. Bartholomew, H. Krabbe, K. Movig, P. Dubský, A. van den Berg, J. Eijkel, A versatile electrophoresis-based self-test platform, Electrophoresis. 36 (2015) 712–721. doi:10.1002/elps.201400428.

[108] P. Kubá ň, P.C. Hauser, Evaluation of microchip capillary electrophoresis with external contactless conductivity detection for the determination of major inorganic ions and lithium in serum and urine samples, Lab Chip. 8 (2008) 1829. doi:10.1039/b802973c.

[109] S. Yu, J. ACCEPTEDXu, K. Huang, J. Chen, J. Duan, Y. Xu, H. Qing, L. Geng, Y. Deng, A novel method to predict protein aggregations using two-dimensional native protein microfluidic chip electrophoresis, Anal. Methods. 8 (2016) 8306–8313. doi:10.1039/C6AY02011A.

[110] R.M. Mohamadi, Z. Svobodova, Z. Bilkova, M. Otto, M. Taverna, S. Descroix, J.L. Viovy, An integrated microfluidic chip for immunocapture, preconcentration and separation of β-amyloid peptides, Biomicrofluidics. 9 (2015) 1–11. doi:10.1063/1.4931394. 57

ACCEPTED MANUSCRIPT [111] S. Mikkonen, J. Jacksén, J. Roeraade, W. Thormann, Å. Emmer, Microfluidic Isoelectric Focusing of Amyloid Beta Peptides Followed by Micropillar-Matrix- Assisted Laser Desorption Ionization-Mass Spectrometry, Anal. Chem. 88 (2016) 10044–10051. doi:10.1021/acs.analchem.6b02324.

[112] L. Li, Y. Fan, Q. Li, R. Sheng, H. Si, J. Fang, L. Tong, B. Tang, Simultaneous Single- Cell Analysis of Na + , K + , Ca 2+ , and Mg 2+ in Neuron-Like PC-12 Cells in a Microfluidic System, Anal. Chem. 89 (2017) 4559–4565. doi:10.1021/acs.analchem.6b05045.

[113] A. Minucci, E. Delibato, M. Castagnola, P. Concolino, F. Ameglio, C. Zuppi, B. Giardina, E. Capoluongo, Identification of RFLP G6PD mutations by using microcapillary electrophoretic chips (ExperionTM), J. Sep. Sci. 31 (2008) 2694–2700. doi:10.1002/jssc.200800216.

[114] H. Maruyama, H. Morino, Y. Izumi, K. Noda, H. Kawakami, Convenient diagnosis of spinal and bulbar muscular atrophy using a microchip electrophoresis system., Am. J. Neurodegener. Dis. 2 (2013) 35–9.

[115] S. Galbiati, B. Foglieni, M. Travi, C. Curcio, G. Restagno, L. Sbaiz, M. Smid, F. Pasi, A. Ferrari, M. Ferrari, L. Cremonesi, Peptide-nucleic acid-mediated enriched polymerase chain reaction as a key point for non-invasive prenatal diagnosis of β- thalassemia, Haematologica. 93 (2008) 610–614. doi:10.3324/haematol.11895.

[116] A. Sposito, V. Hoang, D.L. DeVoe, Rapid real-time PCR and high resolution melt analysis in a self-filling thermoplastic chip, Lab Chip. 16 (2016) 3524–3531. doi:10.1039/C6LC00711B. MANUSCRIPT [117] J.P. Folland, T.M. Mc Cauley, C. Phypers, B. Hanson, S.S. Mastana, Relationship of 2D:4D finger ratio with muscle strength, testosterone, and androgen receptor CAG repeat genotype, Am. J. Phys. Anthropol. 148 (2012) 81–87. doi:10.1002/ajpa.22044.

[118] F. Yang, X.C. Li, W. Zhang, J. Bin Pan, Z.G. Chen, A facile light-emitting-diode induced fluorescence detector coupled to an integrated microfluidic device for microchip electrophoresis, Talanta. 84 (2011) 1099–1106. doi:10.1016/j.talanta.2011.03.020.

[119] H.-S. Chang, K. Mizukami, A. Yabuki, M.A. Hossain, M.M. Rahman, M.M. Uddin, T. Arai, O. Yamato, A novel rapid genotyping technique for Collie eye anomaly: SYBR Green-based real-time polymerase chain reaction method applicable to blood and saliva specimens on Flinders Technology Associates filter paper., J. Vet. Diagnostic Investig. ACCEPTED22 (2010) 708–715. doi:10.1177/104063871002200506. [120] K. Mizukami, H.-S. Chang, A. Yabuki, T. Kawamichi, N. Kawahara, D. Hayashi, M.A. Hossain, M.M. Rahman, M.M. Uddin, O. Yamato, Novel rapid genotyping assays for neuronal ceroid lipofuscinosis in Border Collie dogs and high frequency of the mutant allele in Japan., J. Vet. Diagn. Invest. 23 (2011) 1131–1139. doi:10.1177/1040638711425590.

[121] D.Z. Li, Y.D. Yang, Invasive prenatal diagnosis of fetal thalassemia, Best Pract. Res. Clin. Obstet. Gynaecol. 39 (2017) 41–52. doi:10.1016/j.bpobgyn.2016.10.011.

ACCEPTED MANUSCRIPT [122] J. Fujihara, M. Tabuchi, T. Inoue, T. Yasuda, Y. Fujita, H. Takeshita, Rapid measurement of deoxyribonuclease I activity with the use of microchip electrophoresis based on DNA degradation, Anal. Biochem. 413 (2011) 78–79. doi:10.1016/j.ab.2011.02.011.

[123] A.R. Kim, J.Y. Kim, K. Choi, D.S. Chung, On-chip immunoassay of a cardiac biomarker in serum using a polyester-toner microchip, Talanta. 109 (2013) 20–25. doi:10.1016/j.talanta.2013.03.028.

[124] H. Zheng, H. Cong, H. Wang, Q. Jin, J. Zhao, A Microchip-Based Method for Rapid Separation of Subclasses of High-Density Lipoprotein, Chromatographia. 71 (2010) 1119–1124. doi:10.1365/s10337-010-1602-1.

[125] H. Wang, W. Zhang, J. Wan, W. Liu, B. Yu, Q. Jin, M. Guan, Microchip-based human serum atherogenic lipoprotein profile analysis, Anal. Biochem. 467 (2014) 75–83. doi:10.1016/j.ab.2014.08.031.

[126] H. Wang, C. Han, H. Wang, Q. Jin, D. Wang, L. Cao, G. Wang, Simultaneous determination of high-density lipoprotein, very low-density lipoprotein and low- density lipoprotein subclass in human serum by microchip CE, Chromatographia. 74 (2011) 799–805. doi:10.1007/s10337-011-2147-7.

[127] H. Wang, H.-M. Wang, Q.-H. Jin, H. Cong, G.-S. Zhuang, J.-L. Zhao, C.-L. Sun, H.- W. Song, W. Wang, Microchip-based small, dense low-density lipoproteins assay for coronary heart disease risk assessment, Electrophoresis. 29 (2008) 1932–1941. doi:10.1002/elps.200700631. [128] H. Wang, D. Wang, J. Wang, H. Wang, J. Gu,MANUSCRIPT C. Han, Q. Jin, B. Xu, C. He, L. Cao, Y. Wang, J. Zhao, Application of poly(dimethylsiloxane)/glass microchip for fast electrophoretic separation of serum small, dense low-density lipoprotein, J. Chromatogr. A. 1216 (2009) 6343–6347. doi:10.1016/j.chroma.2009.06.086.

[129] Y. Sun, S.K. Kim, P. Zhang, N. Woo, S.H. Kang, Fast high-throughput screening of angiotensin-converting enzyme insertion/deletion polymorphism by variable programmed electric field strength-based microchip electrophoresis, J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 1028 (2016) 77–85. doi:10.1016/j.jchromb.2016.06.007.

[130] Y. Kawai, M. Yoshida, K. Arakawa, T. Kumamoto, N. Morikawa, K. Masamura, H. Tada, S. Ito, H. Hoshizaki, S. Oshima, K. Taniguchi, H. Terasawa, I. Miyamori, K. Kishi, T. Yasuda, Diagnostic use of serum deoxyribonuclease I activity as a novel early-phase marker in acute myocardial infarction, Circulation. 109 (2004) 2398–2400. doi:10.1161/01.CIR.0000129232.61483.43.ACCEPTED [131] V.C. Menys, Y. Liu, M.I. Mackness, M.J. Caslake, S. Kwok, P.N. Durrington, Measurement of plasma small-dense LDL concentration by a simplified ultracentrifugation procedure and immunoassay of apolipoprotein B, Clin. Chim. Acta. 334 (2003) 95–106. doi:10.1016/S0009-8981(03)00231-6.

[132] P. Zhang, H. Nan, S. Lee, S.H. Kang, Fast High-Throughput Screening of H1N1 Virus by Parallel Detection with Multichannel Microchip Electrophoresis, in: Methods Mol.

ACCEPTED MANUSCRIPT Biol., 2015: pp. 183–195. doi:10.1007/978-1-4939-2353-3.

[133] R. Prakash, K. Pabbaraju, S. Wong, A. Wong, R. Tellier, K. Kaler, Multiplex, Quantitative, Reverse Transcription PCR Detection of Influenza Viruses Using Droplet Microfluidic Technology, Micromachines. 6 (2014) 63–79. doi:10.3390/mi6010063.

[134] X. Mu, L. Qi, J. Qiao, H. Zhang, H. Ma, Study on alanine aminotransferase kinetics by microchip electrophoresis, Anal. Biochem. 421 (2012) 499–505. doi:10.1016/j.ab.2011.11.037.

[135] J.W. Lee, D. Lee, Y.T. Kim, E.Y. Lee, D.H. Kim, T.S. Seo, Low-cost and facile fabrication of a paper-based capillary electrophoresis microdevice for pathogen detection, Biosens. Bioelectron. 91 (2017) 388–392. doi:10.1016/j.bios.2016.12.053.

[136] J. Kirchhoff, U. Glaser, J.A. Bohnert, M.W. Pletz, J. Popp, U. Neugebauer, Simple ciprofloxacin resistance test and determination of minimal inhibitory concentration (MIC) within two hours using Raman spectroscopy, Anal. Chem. (2017) acs.analchem.7b03800. doi:10.1021/acs.analchem.7b03800.

[137] G. Zhu, Q. Song, W. Liu, X. Yan, J. Xiao, C. Chen, A Gold Nanoparticles-Modified Indium Tin Oxide Microelectrode for In-channel Amperometric Detection in Dual- Channel Microchip Electrophoresis, Anal. Methods. 9 (2017) 4319–4326. doi:10.1039/C7AY01008G.

[138] K.C. Chong, L.Y. Thang, J.P. Quirino, H.H. See, Monitoring of vancomycin in human plasma via portable microchip electrophoresis with contactless conductivity detector and multi-stacking strategy, J. Chromatogr. A. 1485 (2017) 142–146. doi:10.1016/j.chroma.2017.01.012. MANUSCRIPT [139] Z. Nie, Y.S. Fung, Microchip capillary electrophoresis for frontal analysis of free bilirubin and study of its interaction with human serum albumin, Electrophoresis. 29 (2008) 1924–1931. doi:10.1002/elps.200700596.

[140] A. Schuchert-Shi, P.C. Hauser, Monitoring the enzymatic conversion of urea to ammonium by conventional or microchip capillary electrophoresis with contactless conductivity detection, Anal. Biochem. 376 (2008) 262–267. doi:10.1016/j.ab.2008.02.030.

[141] S. Wang, X. Li, J. Yang, X. Yang, F. Hou, Z. Chen, Rapid determination of creatinine in human urine by microchip electrophoresis with LED induced fluorescence detection, Chromatographia. 75 (2012) 1287–1293. doi:10.1007/s10337-012-2324-3.

[142] H. Song, H. Wang, S. Ju, Q. Jin, C. Jia, H. Cong, Simplified Microchip ElectrophoresisACCEPTED for Rapid Separation of Urine Proteins, J. Clin. Lab. Anal. 28 (2014) 104–109. doi:10.1002/jcla.21651.

[143] C.L.S. Chagas, F.R. de Souza, T.M.G. Cardoso, R.C. Moreira, J.A.F. da Silva, D.P. de Jesus, W.K.T. Coltro, A fully disposable paper-based electrophoresis microchip with integrated pencil-drawn electrodes for contactless conductivity detection, Anal. Methods. 8 (2016) 6682–6686. doi:10.1039/C6AY01963C.

[144] V. Gaal, L. Mark, P. Kiss, I. Kustos, A. Tamas, B. Kocsis, A. Lubics, V. Nemeth, A.

ACCEPTED MANUSCRIPT Nemeth, L. Lujber, J. Pytel, G. Toth, D. Reglodi, Investigation of the effects of PACAP on the composition of tear and endolymph proteins, J. Mol. Neurosci. 36 (2008) 321–329. doi:10.1007/s12031-008-9067-5.

[145] C.L.S. Chagas, L. Costa Duarte, E.O. Lobo-Júnior, E. Piccin, N. Dossi, W.K.T. Coltro, Hand drawing of pencil electrodes on paper platforms for contactless conductivity detection of inorganic cations in human tear samples using electrophoresis chips, Electrophoresis. 36 (2015) 1837–1844. doi:10.1002/elps.201500110.

[146] Y. Huang, S. Zhao, M. Shi, Y.M. Liu, Chemiluminescent immunoassay of thyroxine enhanced by microchip electrophoresis, Anal. Biochem. 399 (2010) 72–77. doi:10.1016/j.ab.2009.11.036.

[147] P.A. Gustafson, D.W. Boyle, Bilirubin index: A new standard for intervention?, Med. Hypotheses. 45 (1995) 409–416. doi:10.1016/0306-9877(95)90213-9.

[148] S.U. Sarici, M.A. Serdar, A. Korkmaz, G. Erdem, O. Oran, G. Tekinalp, M. Yurdakök, S. Yigit, Incidence, course, and prediction of hyperbilirubinemia in near-term and term newborns., Pediatrics. 113 (2004) 775–780. doi:10.1542/peds.113.4.775.

[149] S.P. Chen, X.D. Yu, J.J. Xu, H.Y. Chen, Gold nanoparticles-coated magnetic microspheres as affinity matrix for detection of hemoglobin A1c in blood by microfluidic immunoassay, Biosens. Bioelectron. 26 (2011) 4779–4784. doi:10.1016/j.bios.2011.06.001.

[150] R. Wu, W.S.B. Yeung, Y.S. Fung, 2-D t-ITP/CZE determination of clinical urinary proteins using a microfluidic-chip capillary electrophoresis device, Electrophoresis. 32 (2011) 3406–3414. doi:10.1002/elps.201100214. MANUSCRIPT [151] S. Pandey, C.M. Lu, D.A. Herold, Measurement of microalbuminuria using protein chip electrophoresis, Am. J. Clin. Pathol. 129 (2008) 432–438. doi:10.1309/4JU0XQH62D3YLTGK.

[152] Q. Pan, S. Hong, X. Zhu, M. Zhao, L.P. Lee, On-line electrophoretic sample clean-up for sensitive and reproducible µCE immunoassay, Lab Chip. 12 (2012) 932. doi:10.1039/c2lc21111d.

[153] S. Yatsushiro, M. Kataoka, Application of microchip electrophoresis for clinical tests, Electr. Eng. Japan (English Transl. Denki Gakkai Ronbunshi). 129 (2009) 282–287. doi:10.1002/eej.21158.

[154] Y. Huang, M. Shi, S. Zhao, H. Liang, A sensitive and rapid immunoassay for quantification of testosterone by microchip electrophoresis with enhanced chemiluminescenceACCEPTED detection, Electrophoresis. 32 (2011) 3196–3200. doi:10.1002/elps.201100218.

[155] F. Ye, J. Liu, Y. Huang, S. Li, S. Zhao, Competitive immunoassay of progesterone by microchip electrophoresis with chemiluminescence detection, J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 936 (2013) 74–79. doi:10.1016/j.jchromb.2013.08.002.

[156] M. Sonker, R. Yang, V. Sahore, S. Kumar, A.T. Woolley, On-Chip Fluorescent Labeling using Reversed-phase Monoliths and Microchip Electrophoretic Separations

ACCEPTED MANUSCRIPT of Selected Preterm Birth Biomarkers, Anal. Methods. 8 (2016) 7739–7749. doi:10.1021/jacs.5b07154.Total.

[157] M. Sonker, R. Knob, V. Sahore, A.T. Woolley, Integrated electrokinetically driven microfluidic devices with pH-mediated solid-phase extraction coupled to microchip electrophoresis for preterm birth biomarkers, Electrophoresis. 38 (2017) 1743–1754. doi:10.1002/elps.201700054.

[158] M.S. Esplin, K. Merrell, R. Goldenberg, Y. Lai, J.D. Iams, B. Mercer, C.Y. Spong, M. Miodovnik, H.N. Simhan, P. Van Dorsten, M. Dombrowski, Proteomic identification of serum peptides predicting subsequent spontaneous preterm birth, Am. J. Obstet. Gynecol. 204 (2011) 391.e1-391.e8. doi:10.1016/j.ajog.2010.09.021.

MANUSCRIPT

ACCEPTED

ACCEPTED MANUSCRIPT

Table 1 . MCE applied for different diseases.

Disease Biomarker Biomarker Sample Off -line sample Separation Separation Sensitivity (or Focus Ref type preparation time and detection sample requirement)

(i) Cancer Human breast Microarray for Model Breast cancer Cells cancer cell line n/a n/a DEP-LIF n/a anticancer drug [54] cell line MCF7 screening Total RNA from Human RNA extraction RNA stability Renal cancer RNA renal cell n/a MGE-LIF n/a [58] tissue and PCR analysis carcinoma General cancer Telomerase Model Target extraction Telomeric repeat DNA < 5 min MGE-LIF 5 cells [60] marker activity cell lines and PCR amplification analysis

General cancer Telomerase Model Target extraction Telomeric repeat DNA 2 min MGE-LIF 1 cell [61] marker activity cell lines and PCR amplification analysis General cancer Aptamer-based Carcinoembryo Human -1 Aptamer-based marker Protein target 1.5 min MGE-LIF 68 pg ml [62] nic antigen serum assay amplification General cancer Carcinoembryo Human -1 Non-competitive Protein n/a MANUSCRIPT < 2 min MCE-LIF 45.7 pg ml [63] marker nic antigen serum immunoassay

General cancer Thymidine Standar -1 Enzyme none 20 s MGE-LIF 2 µg mL Immunoaffinity assay [64] marker kinase 1 ds

Single strand conformational General cancer Exons 5–9 of Human Target extraction DNA 6 min MGE-LIF n/a polymorphism and [67] marker p53 gene tissue and PCR heteroduplex analysis 5% of mutated EGFR exon 19 alleles in the Human Target extraction EGFR gene Lung cancer DNA and 21 <2 min MGE-LIF presence of 95% [68] tissue and PCR mutations alterations ACCEPTED wild-type alleles fragments

ACCEPTED MANUSCRIPT

EGFR exon 19 5% mutant Human Target extraction EGFR gene Lung cancer DNA and 21 n/a MGE-LIF fragment in 95% [69] tissue and PCR mutations alterations wild-type

Neuron- Human Competitive enzyme Lung cancer Enzyme specific n/a 1 min MCE-CL 4.5 pM [70] serum immunoassay enolase Variable region in the T-cell Canine Target extraction Rapid diagnostic of Blood cancer DNA <10 s MGE-LIF n/a [71] receptor γ blood and PCR T-cell lymphoma gene Variable number of Human Assay for post- tandem repeat blood Target extraction hematopoietic stem Blood cancer DNA markers ~3 min MGE-LIF 0.02 ng DNA [72] and cord and PCR cell transplantation D1S11, YNZ, blood chimerism monitoring D1S80, and ApoB Human Identification of Bcl-2/IgH blood Target extraction Blood cancer DNA 2 min MGE-LIF n/a fusion sequences in [73] fusion gene and and PCR follicular lymphoma tissue MANUSCRIPT 10% target DNA in Ig heavy chain Human Target extraction B-cell clonality Blood cancer DNA n/a MGE-LIF 90% non-target [74] gene tissue and PCR analysis DNA Human bone Mutation analysis in Calreticulin marrow Target extraction Blood cancer DNA n/a MGE-LIF n/a myeloproliferative [76] gene and and PCR neoplasms peripher al blood Mutation detection Colorectal KRAS codon Human Target extraction 0.01% target DNA using restriction DNA 5 min MGE-LIF [77] cancer 12 mutation tissue and PCR in non-target DNA fragment length polymorphism Microsatellite ACCEPTED Analysis of loci Bat25, Colorectal Human DNA extraction microsatellite DNA Bat26, <2 min MGE-LIF n/a [78] cancer tissue and PCR instability in D2S123, colorectal carcinoma D5S346 and 64

ACCEPTED MANUSCRIPT

D17S250

Denaturation, disulfide bond reduction, Colorectal Polysacchari Human N-glycans digestion, glycan 2 min MGE-LIF n/a N-glycan profiling [79] cancer de serum isolation by solid- phase extraction, derivatisation Analysis of GSTP1 and Human Target extraction circulating free DNA Prostate cancer DNA 2.5 min MGE-LIF n/a [81] RARB2 gene blood and PCR in response to chemotherapy 13 genes (ECAD, RAR β, TMEFF2, Analysis of aberrant MGMT, DAPK, Human Target extraction DNA methylation in Oral cancer DNA p15, p16, oral n/a MGE-LIF n/a [82] and PCR squamous cell TIMP3, FHIT, rinse carcinoma HIN-1, SPARC, APC, MANUSCRIPT WIF) Liver (and Carcinoembryo Human 0.25 and 0.13 ng Non-competitive colorectal) Proteins nic antigen and None 1 min MCE-EC -1 [83] serum mL immunoassay cancer α-fetoprotein

Sensitive Standar Liver cancer Protein α-fetoprotein None <3 min ITP-MGE LIF 5 pM immunoassay [84] d strategy

Non-competitive α-fetoprotein- Human -1 immunoassay for Liver cancer Protein n/a 2.5 min MCE-LIF 0.3 ng mL [85] L3 serum hepatocellular ACCEPTED carcinoma α-fetoprotein Chip-based liquid- and des-γ- Human phase binding assay Liver cancer Protein n/a n/a MCE-LIF n/a [86] carboxy serum for hepatocellular prothrombin carcinoma

ACCEPTED MANUSCRIPT

α-fetoprotein- Evaluation of the L3 (AFP-L3) diagnostic value of and -1 0.3 ng mL AFP- combined AFP-L3 prothrombin Human Liver cancer Protein n/a n/a MCE-LIF L3 and 5 AU/L and PIVKA-II [87] induced by serum PIVKA-II detection for vitamin K hepatocellular absence-II carcinoma (PIVKA-II) Denaturation, disulfide bond reduction, Human Ovarian cancer Polymer N-glycans digestion, glycan 1.5 min MCE-LIF n/a N-glycan profiling [88] serum isolation by solid- phase extraction, derivatisation

Human Genotyping by papillomavirus Human Target extraction 2 Cervical cancer DNA 4 min MGE-LIF 2 x 10 copies restriction fragment [89] DNA tissue and PCR length polymorphism PGMY09/11

Comparison of three Human amplification Model Target extraction Cervical cancer RNA papillomavirus MANUSCRIPT1.5 min MGE-LIF 1 cell techniques in [90] cell line and PCR E6/E7 mRNA combination with MGE

β-subunit of human Human -1 Non-competitive Ovarian cancer Protein None 1 min MCE-CL 0.36 mIU mL [91] chorionic serum immunoassay gonadotropin

Galectin-1 and Multiplexed Protein/Enzy lactate Model Cell lysis, immunosensor for Bladder cancer 60 min DEP-LIF n/a [92] me dehydrogenas cell line derivatisation protein expression e B ACCEPTED analysis

ACCEPTED MANUSCRIPT

Human Controlled trapping bladder cancer Model Bladder cancer Cells n/a ~2 min DEP Single cell and release of single [93] cells cell line cells TSGH8301

FGFR3, TERT, HRAS, ALX4, Artificial DNA mutation Target extraction Bladder cancer DNA RALL3, MT1A, human 2 min MGE-LIF 5 cells analysis for bladder [94] and PCR and RUNX3 urine cancer diagnostic genes (ii) Immune disorders 3-nitrotyrosine, Protein and thyroxine, and Standar -1 Cleavable tag Inflammation small None 2 min MCE-LIF 0.3 to 5 µg mL [95] C-reactive ds immunoassay molecule protein Sensitive target C-reactive Human Immunoprecipitati -1 Inflammation Protein 30 s MGE-LIF 126 pg µL detection from 50 µL [96] protein serum on sample Interleukin 6, gamma Human interferon, -1 Receptor affinity Inflammation Proteins skin Tissue extraction 2.5 min MCE-LIF 0.5 pg mL [97] tumour disks biopsies MANUSCRIPT necrosis factor alpha 6 chemokines (CXCL1, Human CXCL5, -1 Receptor affinity Inflammation Proteins skin Tissue extraction 2.5 min MCE-LIF 0.2 pg mL [98] CXCL8, CCL1, disks biopsies CCL3, and CCL5 6 chemokines Human (CCL2, CCL19, cerebral -1 Chip-based Inflammation Proteins CCL21,CXCL8, Dilution 2 min MCE-LIF 0.5 pg mL [99] spinal immunoaffinity assay CXCL12, and fluid CXCL-13) PrepurifiACCEPTED Genes ed On-chip genotyping Hypersensitivity DNA VKORC1 and human none 2 min MGE-LIF n/a of warfarin [100] CYP2C9 genomic hypersensitivity DNA

ACCEPTED MANUSCRIPT

Human Genes “Sample-in–answer- blood Hypersensitivity DNA VKORC1 and none ~15 min MGE-LIF 0.5 µL raw sample out” genotyping [101] and CYP2C9 device saliva Protein Human precipitation, Arthritis Metabolite Lactate synovial 40 s MCE-EC 6.5 µM Method development [102] centrifugation, fluid evaporation Human Gout Metabolite Uric acid Dilution 90 s MCE-EC 4 µM Method development [103] urine (iii) Neurological diseases Human serum General Cystatin C and and neurological Protein general protein n/a <1 min MGE-LIF n/a Protein profiling [104] cerebros function profile pinal fluid

Small Human Liquid-liquid Competitive Epilepsy Phenobarbital 50 s MCE-LIF 3.4 nM [105] molecules plasma extraction immunoassay

Phenobarbital (and Multiplexed Small Human MANUSCRIPT Epilepsy carbamazepine none 45 s MCE-LIF 1.8 nM competitive [106] molecules serum , phenytoin, immunoassay theophylline Human serum, blood, Small + and Bipolar disorders Li None 30 s MCE-EC ~0.09 mM Point-of-care device [107] molecule urine. Bovine venous blood Li +, NH 4+ , K +, + 2+ Human Na , Ca , Fast determination of Small 2+ − serum Bipolar disorders Mg , Cl , ACCEPTEDDilution <50 s MCE-EC 1 to 7.5 µM inorganic ions in body [108] molecules − 2− and NO 3 , SO 4 , fluids 3- urine PO 4

ACCEPTED MANUSCRIPT

PC12 cells Study of protein Alzheimer Protein α- synuclein and rat n/a n/a IEF-MCE-LIF n/a [109] aggregation brain tissue β-amyloid Human On-chip peptides 1–37, cerebros immunocapture, Alzheimer Peptides None 2 min MCE-LIF n/a [110] 1–39, 1–40, pinal enrichment, and and 1–42 fluid separation Human β-amyloid Enrichment of nL- cerebros Immunopurificaito Alzheimer Peptides peptides 1–30 10-15 min IEF-MS n/a samples and MS [111] pinal n to 1–42 identification fluid

Small Na + ,K+, Ca2 +, PC 12 Derivatisation and Alzheimer 2+ 1 min MCE-LIF 0.003 to 0.316 fM Single cell analysis [112] molecule and Mg cells cell lysis

(iv) Genetic disorders

Glucose-6- Glucose-6- Mutation detection 1 µL of restricted phosphate phosphate Human Target extraction - using restriction DNA 2 min MGE-LIF DNA (~0.1 ng µL [113] dehydrogenase dehydrogenase blood and PCR 1 fragment length ) deficiency gene polymorphism Spinal and MANUSCRIPT Androgen Human Target extraction bulbar muscular DNA n/a MGE-LIF n/a CAG repeat analysis [114] receptor gene serum and PCR atrophy Human Target extraction Non-invasive prenatal β-thalassemia DNA β-globin gene n/a MGE-LIF n/a [115] blood and PCR diagnostic Disposable chip for von Gierke's Standar qPCR and high DNA G6PC gene None n/a MGE-LIF n/a [116] disease ds resolution melt analysis Determination of CAG repeat Skeletal muscle Androgen Human Target extraction DNA ~3 min MGE-LIF n/a polymorphisms in [117] function receptor gene blood and PCR relation to skeletal ACCEPTED muscle performance Small Standar −1 Use of light emitting Wilson’s disease Penicillamine Derivatisation 2.5 min MCE-LIF 1.1 µmol L [118] molecule d diode for LIF

ACCEPTED MANUSCRIPT

Canine Solid-phase Collie eye Intron 4 in blood DNA extraction using ~3 min MGE-LIF n/a Rapid genotyping [119] anomaly NHEJ1 gene and filter cards saliva Canine Neuronal ceroid Exon 4 in blood Target extraction DNA 3 min MGE-LIF n/a Rapid genotyping [120] lipofuscinosis CLN5 gene and and PCR saliva (v) Cardiovascular diseases

Coronary heart C-reactive Human -1 On-chip Protein Dilution 10 s MCE-LIF 0.3 mg L [123] disease protein serum immunoassay

Myocardial Deoxyribonucle Standar Determination of Enzyme None ~10 min MGE-LIF n/a [122] ischemia ase I ds enzyme activity

Coronary heart Rapid separation of High-density Human Centrifugation, disease and Protein 3 min MGE-LIF n/a subclasses of high- [124] lipoproteins serum derivatisation atherosclerosis density lipoprotein Lipoproteins Improving lipoprotein (high-density Human separation using gold Atherosclerosis Protein Derivatisation 2.5 min MCE-LIF n/a [125] and low serum MANUSCRIPT nanoparticles for density) separation Lipoproteins (high-density Human -1 Non-competitive Atherosclerosis Protein Derivatisation 3 min MCE-LIF 5-15 µg L [126] and low serum immunoassay density) Lipoproteins Coronary heart (high-density Human Centrifugation, Alternative method to disease and Protein 3 min MCE-LIF n/a [127] and low serum derivatisation enzymatic assays atherosclerosis density) Lipoproteins Coronary heart Method development (high-density Human Centrifugation, disease and Protein 3 min MGE-LIF n/a for lipoprotein [128] and low serum derivatisation atherosclerosis separation density) Angiotensin- ACCEPTED Use of an electric- Human Target extraction Hypertension DNA converting 25 s MGE-LIF n/a field gradient for fast [129] blood and PCR enzyme gene genotyping

(vi) Infectious diseases 70

ACCEPTED MANUSCRIPT

Influenza A DNA Hemagglutinin H1N1 Target extraction 20 s MGE-LIF n/a High-throughput [132] and virus and PCR screening for H1N1 nucleocapsid DNA virus protein gene sample Influenza A and RNA Matrix gene Nasal Target extraction n/a DEP-LIF 10 RNA copies On-chip reverse [133] B and non- and (per 5µL) transcription PCR structural gene throat swaps Hepatitis B Enzyme Alanine Human Derivatisation 3 min MCE-LIF 0.2-0.4 µM Study on alanine [134] aminotransfera serum aminotransferase se activity Bacterial DNA rrsH and glnA Model Target extraction 3 min MGE-LIF 9.3×10 1 to Paper-based device [135] infection genes cell line and PCR 1.6×10 2 copies for pathogen detection Bacterial Bacteria E. coli Human Cell culture n/a DEP-Raman n/a On-chip antibiotic [136] infection blood spectroscopy resistance test Bacterial Small p-aminophenol, Standar None 1.5 min MCE-ED 0.24 to 0.42 µM Amperometric [137] infection molecules o-aminophenol ds detection with gold and m- nanoparticle-modified aminophenol indium tin oxide electrodes Bacterial Small Vancomycin Human Protein 3 min MCE-EC 1.2 µg mL -1 On-chip Stacking for [138] infection molecule plasma precipitation, MANUSCRIPT sensitive vancomycin centrifugation, determination evaporation (vii) Organ diseases and dysfunctions Renal Protein and Albumin and Standar None 2 min MCE-EC 20 and 35 µM Disposable paper- [143] insufficiency metabolite creatinine ds based device Renal Small NH 4+ Human Dilution 1 min MCE-EC n/a Monitoring of [140] insufficiency molecule blood enzymatic reactions and serum Renal Metabolite Creatinine Human Centrifugation, 30 s MCE-LIF 2.9 µM Method development [141] insufficiency urine filtration, derivatisation Renal Protein Human HumanACCEPTED n/a 4 min MCE-UV 0.4 g L-1 Rapid protein urine [142] insufficiency transferrin and urine analysis human albumin Ocular Peptide Pituitary Chicken Direct sampling 45 s MGE-LIF n/a Effect of pituitary [144] dysfunction adenylate endolym using filter paper adenylate cyclase 71

ACCEPTED MANUSCRIPT

cyclase ph and strips activating polypeptide activating rat tear on protein polypeptide composition in tears and endolymph Ocular Small K+, Na +, and Human Dilution 2 min MCE-EC 4.9, 6.8, and 9.0 Disposable paper- [145] dysfunction molecules Li + tear µM based device Thyroid gland Small Thyroxine Human Protein 1 min MCE-CL 2.2 nM Sensitive [146] dysfunction molecule serum precipitation, chemiluminescence centrifugation immunoassay (viii) Diabetes and pancreatic dysfunction On-chip Diabetes Hemoglobin Human −1 immunoassay using Protein Hemolysis 3 min MCE-EC 0.025 µg mL [149] mellitus A1c blood magnetic nanoparticles Transferrin, β2- Integrated platform microglobulin, for sample Diabetes human serum Human Centrifugation, -1 preparation, mellitus, diabetic Protein 5 min ITP-MCE-UV 0.05-0.6 mg L [150] albumin and urine filtration enrichment and nephropathy immunoglobuli detection of 4 urinary n G proteins Diabetes Alternative chip Human Human Protein -1 mellitus, diabetic Protein <1 min MGE-LIF 7.5 mg L assay to standard [151] albumin urine denaturation nephropathy MANUSCRIPT enzymatic assay Centrifugation, Diabetes Insulin-like Human -1 Sensitive competitive Protein precipitation, 1 min FASI-MCE-LIF 0.68 ng mL [152] mellitus growth factor-I serum immunoassay reconstitution Alternative chip Pancreatic α-amylase Human Density gradient -1 Enzyme <2 min MGE-LIF 4.38 U L assay to standard [153] dysfunction isoenzyme plasma centrifugation enzymatic assay (ix) Reproductive disorders Androgen Small Human Anticoagulation, Competitive Testosterone 30 s MCE-CL 1.0 nM [154] disorders molecule serum denaturation immunoassay Protein precipitation, Progesterone- Small Human Competitive Progesterone centrifugation, 1 min MCE-CL 3.8 nM [155] related disorders molecule serum immunoassay ACCEPTEDevaporation, reconstitution Preterm birth Peptides/Prot Standar On-chip solid-phase Preterm birth biomarker None <50 s MCE-LIF n/a [156] eins ds extraction peptides, 72

ACCEPTED MANUSCRIPT

lactoferrin, ferritin Preterm birth Human On-chip solid-phase Preterm birth Peptide biomarker Dilution 10 s MCE-LIF 50 nM [157] serum extraction peptide 1 CL , chemiluminescence; DEP , dielectrophoresis; EC , electrochemical; FASI , field-amplified sample stacking; IEF , isoelectric focusing; ITP , isotachophoresis; LIF , laser- induced fluorescence, MCE , microchip zone electrophoresis; MGE , microchip gel electrophoresis; MS , mass spectrometry; n/a , not applicable; PCR , polymerase chain reaction.

MANUSCRIPT

ACCEPTED

ACCEPTED MANUSCRIPT Abbreviations

APTS 8-aminopyrene-1,3,6-trisulfonic acid AUC Area-under-the-curve BGS Background solution bp Base pairs CE Capillary electrophoresis DEP Dielectrophoresis EGFR Epidermal growth factor receptor ELISA Enzyme-linked immunosorbent assay EOF Electroosmotic flow FITC Fluorescein isothiocyanate HA Heteroduplex analysis HDL High-density lipoprotein HPV Human papillomavirus HRP Horseradish peroxidase IGF-I Insulin-like growth factor-I JAK2 Janus kinase 2 LDL Low-density lipoproteins LED Light-emitting diode LIF Laser-induced fluorescence lLDL Large, buoyant low-density lipoprotein IU International unit MANUSCRIPT MCE Microchip electrophoresis MGE Microchip gel electrophoresis PACAP Pituitary adenylate cyclase activating polypeptide PCR Polymerase chain reaction PIVKA-II Prothrombin induced by vitamin K absence-II PMMA Poly(methyl methacrylate) PNA-PCR Peptide nucleic acid-mediated enriched polymerase chain reaction ROC Receiver operating characteristics sdLDL Small, dense low-density lipoprotein SDS Sodium dodecyl sulfate SSCP Single strand conformational polymorphism vLDL Very low-density lipoproteins VNTR Variable number of tandem repeats ACCEPTED

ACCEPTED MANUSCRIPT Figure captions

Figure 1. Schematic of a typical microchip electrophoresis system.

Figure 2 Methodological approach for KRAS mutation detection using restriction fragment length polymorphism followed by microchip electrophoresis [77] .

Figure 3. Electropherogram showing the serum N-glycans profile from a patient with colorectal cancer. Reprinted with permission from [79]. Copyright (2016) American

Chemical Society.

Figure 4 Schematic of immunoassay for carcinoembryonic antigen and α‐fetoprotein. The

offline HRP-antibody-labelled biomarkers were injected from sample reservoir 2 into the

separation channel where the labelled biomarkers were separated. A continuous supply of MANUSCRIPT substrate during separation was provided from reservoir 5. Upon reaction of HRP with the substrate, a product was formed that was amperometrically detected at 4.

Figure 5 (i) Schematic of the affinity disk chip for the determination of bioactive cytokines.

The bioactive cytokines (open circle) are captured using specific protein receptors

immobilised on a glass substrate. The sample matrix (squares) and the non-bioactive

cytokines (black circle) are washed away. The immobilised cytokines are labelled with a fluorescence tagsACCEPTED and subsequently eluted and electropheretically separated. In (ii), the electropherograms obtained from the analysis of skin biopsy samples of (A) normal subject

and (B) patient with mild inflammatory lesions and (C) patient with severe recurrent

inflammatory lesions are shown. Figure reprinted from [97] with permission.

ACCEPTED MANUSCRIPT Figure 6. (A) Photograph of “sample-in-answer-out” microfluidic chip for genotyping of warfarin hypersensitivity. (B) shows the electropherograms from dried blood stain and oral swab obtained by a “sample-in-answer-out” microfluidic genotyping platform. Reprinted with permission from [101].

Figure 7. (A) Microfluidic device for isoelectric focusing of β-amyloid peptides. (B) Mass spectrum of 5 nL of (i) immunoprecipitated cerebrospinal fluid and (ii) immunoprecipitated and isoelectrically-focused cerebrospinal fluid. Reprinted with permission from reference

MANUSCRIPT

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT Highlights

• Microchip electrophoresis (MCE) performed immunoassays, genotyping, and small molecule determination that were relevant in clinical analysis.

• MCE enabled fast separation of biomarkers which was a crucial step for disease identification.

• There were only a few reports that fulfilled the promise of point-of-care diagnostics.

• The MCE strategies are discussed based on the targeted diseases.

• Recent advancements and challenges of MCE are highlighted.

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT

Alain Wuethrich is an early career researcher who graduated with MSc in Life Sciences from the University of Applied Sciences Northwestern Switzerland in 2012. He obtained his PhD in 2016 from the Australian Centre for Research on Separation Science (University of Tasmania) under the supervision of A/Prof Quirino. After his PhD, Alain was awarded an Early Postdoc Mobility Fellowship from the Swiss National Science Foundation to research on microfluidic approaches for early cancer detection at the Australian Institute for Bioengineering and Nanotechnology at the University of Queensland (UQ). In 2018, he was awarded a 3-year UQ Development Fellowship to continue the initiated research program on microfluidic disease detection strategies.

MANUSCRIPT

ACCEPTED ACCEPTED MANUSCRIPT

Joselito P. Quirino graduated BSc in Industrial Pharmacy from the University of the Philippines Manila in 1992. He finished his MSc/PhD in Analytical Chemistry with Prof. Shigeru Terabe at the Himeji Institute of Technology Japan in 1998/1999. He was then a postdoctoral researcher in Prof. Richard Zare’s laboratory at Stanford University USA until 2001. This was followed by an industrial stint in pharmaceutical companies in California USA until 2007. He went back to academia at the ACROSS in the University of Tasmania where he is currently an Associate Professor. He has been awarded an Australian Research Council Future Fellowship and Japanese Society for the Promotion of Science Fellowships.

MANUSCRIPT

ACCEPTED