Nanoporous Membranes for Microfluidic Concentration Prior to

Subscriber access provided by University of Sussex Library

Article

Nanoporous membranes for microfluidic concentration prior to electrophoretic separation of proteins in urine Feng Li, Rosanne M Guijt, and Michael C. Breadmore Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b02096 • Publication Date (Web): 08 Jul 2016 Downloaded from http://pubs.acs.org on July 10, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Nanoporous membranes for microfluidic concentration prior to electrophoretic separation of proteins in urine Feng Li1,2, Rosanne M Guijt2, Michael C Breadmore1, * 1

2

Australian Centre for Research on Separation Science, School of Chemistry, University of Tasmania, Private Bag 75, Hobart, Tasmania 7001, Australia.

School of Medicine and Australian Centre for Research on Separation Science, , University of Tasmania, Private Bag 26, Hobart, Tasmania 7001, Australia

Abstract A microfluidic device with two nanoporous membranes was developed

to

seamlessly integrate

sample

preparation

and

electrophoretic separation of proteins. The device was fabricated by sandwiching two nanoporous polycarbonate track etched (PCTE) membranes with differently sized nanopores between PDMS slabs containing embedded microchannels. The first membrane contained larger (100 nm) pores and served as an initial filter to screen out particles, cells and larger proteins. The second membrane contained smaller pores (10 nm) which facilitated transport of inorganic ions and small organic molecules, but not proteins.

The sequential

combination of these two membranes allows proteins to be concentrated and purified simultaneously. The device was used for the sample-in/answer-out quantification of albumin in human urine

*

Email: [email protected]

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 28

within 2.5 min with an improvement in sensitivity of 500 fold compared to a normal pinched injection using fluorescence detection. The linear range of was 0-100 μg mL-1, with a LOD of 1.5 μg mL-1 covering the diagnostic level of microalbuminuria of 30 μg mL-1. The presented device, which is simple to make and use, provides a quantitative alternative for point-of-care detection of proteins, as demonstrated through its application to albumin in urine for the diagnoisis of (micro)albuminuria.

Introduction Sample preparation is usually defined as a series of cellular and molecular separation/fractionation steps required or recommended in order to obtain higher sensitivity and selectivity of downstream biosensing or bioanalysis [1]. Sample preparation is often the most complex and most critical step in analyzing complex biological samples like blood, urine, and saliva, aiming to eliminate the large number of components that may interfere with the target and to enrich the trace amounts to detectable levels.

Filtration and

extraction

used

are

amongst

the

most

frequently

sample

pretreatment techniques for biological sample analysis. Sample pretreatment is often conducted off-line, using timeconsuming and labor-intensive protocols with a significant risk of analyte loss, degradation of the sample and/or the introduction of contaminants. Many of these risks can be mitigated by conducting sample pretreatment in an automated manner. Microfluidic approaches are attractive for sample pretreatment, particularly due to the potential to be integrated seamlessly into a small, portable analytical device to facilitate point-of-care testing. Taking advantage


Page 3 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


of laminar flow for fluidic control and the short diffusion path lengths for speed, complex laboratory tasks can be integrated into a small microfluidic device using minimal amounts of sample and reagents [2]. Removal of particulates is the first step in many analytical processes, and this is required for most assays with ‘direct injection’ to ensure the reliability and robustness of the result where simple centrifugation of the sample is often sufficient [3-5]. This however is difficult to implement in a microchip, and other simple approaches have been developed such as exploiting laminar flow. This separates analytes from particulates and does not require a physical membrane, as demonstrated using H-filter structures for diffusion-based filtration processes [6]. Here, the small molecule targets diffuse faster than macromolecules and particulates, and hence these can be collected in an acceptor stream while leaving larger molecules behind in the original sample stream. While very simple, this approach dilutes the targets in order to collect a protein and particulate free fraction containing the small molecule targets and must be performed with both the sample and acceptor flowing to maintain the boundary between the two fluids. For collection of particles or to clean-up the sample, membranes can be integrated into a microfluidic device. These can be structured within the device as part of the microfabrication process, formed post-fabrication inside the microchannel or commercially available membranes can be incorporated during assembly [7]. Vrouwe and co-workers combined microchip capillary electrophoresis with conductivity detection and an integrated filter to measure lithium directly from whole blood. Because the target analyte is present in 3 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 28

high concentration, a detection limit of 0.15 mmol/L is more than sufficient for detection in the clinical range without a concentration step [8-9]. Nanoporous membranes have been used in microfluidics for sample filtration [10-11], sample preconcentration [12-15], sensing [16], and selective delivery [17-18]. For example, a PVAbound glass-microfiber membrane was used to remove blood cells from

plasma

before

simultaneous

analyses

of

high-density

lipoprotein cholesterol and total cholesterol [11]. A PCTE membrane was used to selectively purify and concentrate macromolecules, which was illustrated by removing primers and preconcentrating the product DNA from a PCR product mixture [17]. A microfluidic device integrating two membranes was developed for fraction collection of detected zones by Tulock et al. The first membrane was used for injection, while a second membrane positioned just after the detector was used to fractionate fluorescein isothiocyanate (FITC) labeled amino acids. This shows great potential for multidimensional separations using different electrolyte conditions for separation and detection [18]. Whilst effective in removing particulates and large interferences, filtration cannot concentrate analytes in the filtrate. For analytes present at trace concentrations, a preconcentration step prior to analysis is often necessary to enhance sensitivity. Extraction methods including both liquid-liquid extraction (LLE) [19] and solid-phase extraction (SPE) [20] can simultaneously remove particulates and matrix components and concentrate the target analytes by collecting them in a smaller volume [21], but complicates the fabrication and use of the devices because a stationary phase is needed for SPE, and fluidic handling of the eluent or acceptor phase is required. 4 ACS Paragon Plus Environment

Page 5 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Electrophoretic preconcentration methods [22-26] like fieldamplified sample stacking (FASS) [27] and sweeping [28] are conceptually simpler than extraction and can be implemented in much simpler microfluidic chips. Their use is limited by a lack of selectivity, i.e interferences are concentrated together with the analytes of interest and are not always applicable to high ionic strength

matrices

[29-30].

Isotachophoresis

overcomes

this

limitation, but the high concentrations of inorganic ions present in biological samples can destack the targets [31-34]. The use of an electric field in combination with nanoporous media gives rise to ion concentration polarization, with a ~105–106 concentration factor for proteins demonstrated using a polycarbonate track etched (PCTE) nanopore membrane sandwiched in a PDMS [12]. While this is impressive, ICP is not selective, concentrating all ions in the sample, and it generally performs best under low ionic strength conditions making it incompatible with the direct analysis of biological samples. To overcome the issues of selectivity with Ion concentration polarization (ICP), Shallan et al. presented an electrokinetic size and mobility trap for simultaneous extraction, concentration and desalting, which was coupled on the same chip with a zone electrophoresis separation to determine ampicillin from whole blood [35]. This unique approach demonstrated the first combination of two nanoporous channels of different size thus introducing some selectivity into the concentration process. Despite the ease and elegance of nanochannel fabrication [36], the limited number of nanochannels formed by dielectric breakdown limits transfer rates and detection sensitivity with the targets concentrated by a factor of less than 100. 5 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 28

In this work, we address this transport limitation of the size mobility trap by substituting the nanochannels made through controlled dielectric breakdown with those contained in two PCTE membranes of different pore size. It should be mentioned that there is only one other publication of a microfluidic device with 2 integrated membranes as described above, although these had exactly the same pore size. Our devices have two different pore sized membranes, the first larger pore sized membrane (100 nm) acts as a filter to screen out particles, cells and larger proteins, ensuring the reliability and robustness of the device when analyzing untreated samples. The second smaller pore sized membrane (10 nm) facilitated transport of inorganic ions and small organic molecules, but not proteins, allowing them to be concentrated. The analytical potential of this device for the analysis of unprocessed biological samples is demonstrated with a sample-in/answer-out analysis of human serum albumin (HSA) in human urine sample, providing a quantitative test for albuminuria.

Experimental section Materials and chemicals Polycarbonate track etch (PCTE) membranes were purchased from Sterlitech Co. (Washington, USA). Polydimethyl siloxane (PDMS) prepolymer and curing agent were purchased from Dow Corning (Sylgard 184) （Michigan, USA). R-phycoerythrin, bovine serum albumin (BSA), human serum albumin (HSA), fluorescamine for BSA and

HSA

labeling,

sodium

tris(hydroxymethyl)aminomethane

tetraborate (Tris),


boric

(borax), acid,

Page 7 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ethylenediaminetetraacetic

acid

(EDTA)

and

hydroxypropyl

methylcellulose (HPMC) for buffer preparation were purchased from Sigma-Aldrich Co (Missouri, USA). All solutions were prepared in Milli-Q water obtained from a Millipore (North Ryde, Australia) purification system. Device fabrication The microfluidic devices consist of two PCTE membranes with different sized nanopores sandwiched between two PDMS layers with embedded channels (Figure 1). The PDMS microchannels were fabricated according to established methods [37]. A 10:1 (w/w) mixture of Sylgard 184 PDMS elastomer base and curing agent was degassed, placed on positive relief SU-8 templates made in house on poly methyl methacrylate (PMMA) slides, degassed again, and heated in an oven at 70 °C for 1 h to form upper and lower PDMS layers individually. The reservoirs were punched on the upper PDMS using stainless-steel tube, and then 100 nm and 10 nm membranes (6 µm thick, pore density is 2×107-6×108/cm2) were placed on the lower PDMS. The two PDMS slabs were plasma treated using a BD-20 handheld corona discharge device (Electro Technic Products, Chicago, USA) and irreversibly sealed.

The devices were single use to

minimize adsorption of proteins onto the PDMS and fouling of the membrane. Electrokinetic process TBE buffer (100 mM Tris, 100 mM boric acid and 2 mM EDTA (the pH was adjusted using 1M NaOH) and borate buffer (25 mM sodium tetra borate and 100 mM boric acid, the pH was also adjusted using 1M NaOH) were used in this experiment. The BSA and HSA sample



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

were prepared from 2 mg/mL stock solution and labeled with fluorescamine in borate buffer. Briefly, various concentrations of stock BSA or HSA were added with 3 mg/mL fluorescamine in a ratio 3:1, and then borate buffer was added to make the final concentration, the mixture was incubated for 15 min at room temperature. The R-phycoerythrin was prepared in borate buffer and diluted to different concentrations. The sample channel and sample reservoir were filled with sample, and the waste channel and separation channel, as well as other reservoirs were filled with background electrolyte (BGE, 100 mM TBE buffer, pH 8.5 with 0.5% hydroxypropyl methyl cellulose (HPMC)). Then voltages were applied, and movement of the fluorescent molecules was observed using an inverted fluorescence microscope (Ti-U, Nikon, Tokyo, Japan) integrated with Nikon highdefinition color charge-coupled device (CCD) camera head (Digital Sight DS-Fi1c, Nikon, Japan) and operated with NIS-Elements BR 3.10 software (Melville, NY, USA). For enhancement factor experiments, the performance of the membrane device was compared with a pinched injection using 5 ppm R-phycoerithrin. In the double membrane device, both the separation channel and waste channel were filled with 100 mM TBE buffer, pH 8.5 with 0.5% HPMC. The applied voltages for injection and preconcentration were -200, +500, -200, -500 V for 90s and for separation were -300, +100, +1200, +100 V at reservoir 1-4 respectively as Figure 2 shown. During injection, the sample ions will migrate from the sample reservoir (reservoir 4) through the first membrane (100 nm) leaving particulates and larger proteins behind. Together with the smaller ions passing the first membrane, the target 8 ACS Paragon Plus Environment

Page 8 of 28

Page 9 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


analytes migrate into into the main microchanneltowards the sample waste (reservoir 2). Due to their size, the target proteins concentrate at the second membrane (10 nm) while the smaller inorganic and organic ions migrate to the waste reservoir. At the end of injection, the voltages are switched such that the electrif field moves the concentrated protein zone migrates down the main channel towards the separation waste (reservoir 3), allowing for electrophoretic separation before passing the detector. During this phase, a slight pull-back voltage is applied to the sample (4) and sample-waste (2) reservoirs to prevent bleeding.

For a comparison experiment using

normal pinched injection, all channels and reservoirs except sample reservoir were filled with 100 mM TBE buffer, pH 8.5 with 0.5% HPMC. Applied voltages and times for injection and separation were the

same

as

that

of

the

membrane

device.

Quantitative

measurements were performed with a photomultiplier tube (Hamamatsu Photonics KK, Hamamatsu, Japan) connected to the microscope. Data acquisition was made using an Agilent interface (35900E) connected to a laptop and operated by Agilent ChemStation for LC software (Agilent Technologies, Waldbronn, Germany). Multiband emission and excitation filters were used (λEx 390-482563-640 nm, λEm 446-523-600-677 nm, Semrock, Rochester, NY, USA). An in-house 4-channel (0-5kV) dc power supply was used to apply an electrical potential to each reservoir through a customdesigned interface connected to 6 platinum electrodes.

Results and discussions Size selective transport 9 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Before starting the extraction experiments, fluidic communication between the upper and lower channels through inefficient sealing was examined with a 10 nm PCTE membrane. The upper sample channel was filled with a 50 mM solution of fluorescein in 100 mM borate buffer whilst the lower channel was filled with the buffer only. After one hour, no fluorescein could be detected in the lower channel indicating that there was no diffusion around the membrane and confirming that the PCTE membrane was sealed into the device in a leakage free manner. The next stage was to examine the size cut-off of the membranes with electrokinetic flow. Because the membrane is hydrophilic, the nanopores are filled with BGE by capillary forces and provide an electrical contact, enabling them to function as an electrokinetic valve between the two fluidic layers. Target analytes in the upper channel could be electrokinetically transported across the membrane into the lower channel, as previously demonstrated by others [10-11, 17-18]. Electrokinetic transport through thin porous membranes is influenced by many factors such as applied field, pore diameter, pore surface chemistry, channel surface charge, channel magnitudes, solution ionic strength, etc [14]. In our experiments, a 100 mM TBE buffer pH 8.5 was used with 0.5% HPMC added to suppresses the EOF in the microchannel to ensure analyte transport was driven predominantly by electrophoretic migration. The high ionic strength buffer induces a very thin electrical double layer, and the channel is filled with bulk solution, resulting in a small EOF in the nanopores, making electromigration the dominant factor in molecular transport [16].


Page 10 of 28

Page 11 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Membranes with different pore size (10, 30, 50, 100 nm) were selected to investigate the selective permeability for fluorescein, BSA (66 kDa) and R-phycoerythrin (250 kDa). All are anionic at pH 8.5. Fluorescein passed all investigated membranes, BSA passed the 50 nm and 100 nm membrane and R-phycoerythrin could only pass the 100 nm membrane. These results confirmed that the membrane pore size determines the size of molecules that can pass through the membrane from one channel to the other. Full characterization of the transport properties of the membranes was not undertaken, but is anticipated to be similar to previous reports [12,15,17]. Electrokinetic concentration, purification and separation of proteins The use of two sequential nanoporous media in the electric field allows the creation of a size and mobility trap (SMT) [35]. This was previously

created

using

nanochannels

made

by

dielectric

breakdown, and the limited number of nanochannels produced by this process resulted in low transport of the ions, and a concentration factor of less than 100 was obtained. The use of PCTE membranes with a higher density of nanopores (2×107-6×108/cm2 giving 50015,000 pores connecting the microchannels) was anticipated to overcome this issue.

A SMT was created for the electrokinetic

concentration of R-phycoerythrin using a 100 nm/10 nm membrane combination. As illustrated in Figure 3, the R-phycoerythrin moved from the sample into the separation channel and was concentrated just above of the 10 nm membrane. This indicates R-phycoerythrin (250 kDa) passed the 100 nm membrane but not the 10 nm membrane where it concentrated via the size effect [38]. Molecules smaller than the pores will pass the 10 nm membrane, enabling their 11 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 28

removal from the concentrating protein zone, allowing desalting of the sample. For protein analysis, the combination of concentration and purification is ideal, for example to remove unreacted labels and other organic and inorganic ions. To quantify the concentration factor, the fluorescence intensity of Rphycoerythrin was measured over a measurement window specified with the rectangle in the inset of Figure 3 (b) and compared with the fluorescence intensity of the standard sample solutions (1, 2, and 4 μM) using image J software (NIH, USA) at different concentrations and plotted as shown in Figure 3(b). For all investigated concentrations, the concentration factor increased gradually. While the concentrated sample could disperse and defocuse after 90 s. For 0.2 ppm R-phycoerythrin, a concentration factor of almost 1000 was achieved within 120 s. Significant for the analysis of physiological samples, this is achieved in an electrolyte with a much higher ionic strength than many other reports for the concentration of proteins on membranes [39-41]. This enhancement is also much greater than the

100-fold

enhancement

previously

reported

with

the

electrokinetic trap created by dielectric breakdown, especially considering the fact the previous report was limited to samples prepared in water to provide a field-enhanced introduction of analytes to the nanojunctions for concentration, while the current approach realizes it is prepared in 100 mM borate buffer. Having established the preconcentration capacity of the integrated device, this was combined with an electrophoretic separation. A mixture of R-phycoerythrin and fluorescamine labeled BSA was selected as the different emission colours enable visualization of the concentration and separation of both proteins. As shown in Figure 4, 12 ACS Paragon Plus Environment

Page 13 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


BSA concentrated closer to the 10 nm membrane than the Rphyscoerythrin, presumably due to the higher electrophoretic mobility of the BSA, and the establishment of an ITP-like effect as also observed by Quist et al. [42]. After preconcentration, the voltages were changed to allow the proteins to be separated by zone electrophoresis. Because of its higher mobility, BSA migrated faster that R-phycoerythrin and both proteins could be easily resolved. The performance of the dual membrane device was compared with a conventional microchannel cross with pinching and pullback injection with the result shown in Figure 5.

From the

electropherograms, the separation efficiency in the membrane device was similar to that obtained using pinched injection, and it provided a 900-fold enhancement with a 90 s injection. Direct determination of albumin in urine samples Monitoring of albuminuria is highly recommended for patients with high blood pressure and diabetes mellitus, as it is an indicator of kidney disease [43]. It is the most abundant protein present in urine, which is elevated with albuminuria, and can be detected by the current diagnostic approach of a dipstick test, with laboratory testing following a positive result. Whilst inexpensive, the dipstick test is estimated to have a 50-90% false positive rate, requiring follow-up testing, increasing costs, and causing patient anxiety [44]. Dipstick testing also fails to reliably identify diabetes patients with microalbuminuria (30 to 300 mg/L albumin in urine), and may therefore miss the early onset of kidney dysfunction [45]. For urine spot tests, HSA levels above 30 µg mL-1 need to be reliably detected, and distinguished from the lower, physiologically normal levels. A more sensitive and more reliable alternative HSA test for PoC use is 13 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 28

desirable [46], with our developed device having the potential to increase sensitivity via concentration of the HSA, and improve the specificity by including an electrophoretic separation. Urine samples (both spiked and non-spiked) from a healthy volunteer were labeled with off-chip with fluorescamine to facilitate fluorescence detection and analysed using the dual membrane device. Following

the

method

described

above,

labeled

HSA

was

concentrated near the second membrane with the electropherogram recorded during the subsequent electrophoretic separation shown in Figure 6. While the benefit of the initial 100 nm PCTE membrane is less clear in this instance, it provides an important function of removing particulate matter that would otherwise block the channel that is typically removed by centrifugation in other analytical approaches.

The peak detected at approximately 0.5 min was

identified as HSA by spiking. With a linear range of 0-100 μg mL-1 (figure 6b), a LOD of 1.5 μg mL-1, and the recovery for the HSA from urine of 81.2-116.8%, this indicates the possibility for the identification of microalbuminuria (HSA > 30 μg mL-1).

This

demonstrates the potential of our approach for the quantitative determination of proteins in urine with only off-chip labeling for sample pretreatment.

While off-chip labelling requires some

sample handling, the derivatisation is rapid (< 900 s) and requires only a single step prior to filling the microchip. It is envisaged that in the future, fluorescamine would be deposited and dried in the sample chamber for immediate reaction with proteins once filled with sample. To confirm improved performance by the integrated device, the results obtained from the dual membrane chip were compared with those obtained using a pinched injection of the same sample on 14 ACS Paragon Plus Environment

Page 15 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a cross-T chip. The electropherogram in Figure 7 shows that without the membrane the HSA peak is about 100 times smaller and there are more interferences, demonstrating that the dual membrane device concentrates and purifies the proteins in the urine sample.

Conclusions We have demonstrated an integrated device for automated and fast detection of proteins in body fluids, and demonstrated this with a quantitative assay for albumin in urine for the diagnosis of (micro)albuminuria. The proteins are extracted, purified and concentrated using a size mobility trap, followed by an on-chip electrophoretic separation and fluorescence detection. The size mobility trap was created using a simple fabrication approach, integrating commercially available PCTE membranes with pore size of 10 and 100 nm into a multilayer PDMS device. The upper membrane containing larger (100 nm) nanopores acted as a sizeselective filtering passing some proteins and smaller molecules but not larger debris like cells. The lower, smaller pore size membrane (10 nm) did not pass the proteins, but allowed the passage of small molecules. Consequently, proteins concentrated at the membrane whilst small molecule interferences were removed. By using this dual membrane device, multiple proteins can be extracted, concentrated and purified, to be separated and detected with an enhancement factor as high as 900 with a 90 s injection. To achieve a similar outcome by LLE or SPE, all of the proteins from 900 μL of sample would need to be extracted into 1 μL in 90 s. The device was successfully applied for the detection of HSA in urine, enabling quantitative determination to diagnose microalbumuria in patients



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 28

with HSA levels above 30 µg/mL. The fast and sensitive detection of proteins from biological fluids without the need for external sample pretreatment is highly attractive for point of care diagnostics .

ACKNOWLEDGEMENTS FL would like to acknowledge the University of Tasmania for the provision of a scholarship. MCB wish to acknowledge an Australian Research

Council

Future

Fellowship

award

(FT130100101,

respectively). RMG would like to acknowledge the Alexander von Humboldt Foundation for the award of a fellowship for Experienced Researchers.


Page 17 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References [1] Han, J.; Fu, J.; Wang, Y. C.; Song, Y. A. Encyclopedia of Microfluidics and Nanofluidics 2008, 96-106. [2] Whitesides, G. M. Nature 2006, 442, 368–373. [3] Thomas, A.; Görgens, C.; Guddat, S.; Thieme, D.; Dellanna, F.; Schänzer, W.; Thevis, M. J. Sep. Sci. 2016, 39, 333–341. [4] Jeong, Y. D.; Kim, M. K.; Suh, S. I.; In, M. K.; Kim, J. Y.; Paeng, K. J. Forensic Sci. Int. 2015, 257, 84–92. [5] Huybrechts, B.; Martins, J. C.; Debongnie, P.; Uhlig, S.; Callebaut, A. Arch. Toxicol. 2015, 89, 1993–2005. [6] Weigl, B. H.; Yager, P. Science 1999, 283, 346-347. [7] De Jong, J.; Lammertink, R. G. H.; Wessling, M. Lab Chip, 2006, 6, 1125-1139. [8] Vrouwe, E. X.; Lüttge, R.; Vermes, I.; Van Den Berg, A. Clin. Chem. 2007, 53, 117-123.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

[9] Vrouwe, E. X.; Lüttge, R.; Olthuis, W.; Van den Berg, A. Electrophoresis 2005, 26, 3032-3042. [10] Mohamed, H.; Russo, A. P.; Szarowski, D. H.; McDonnell, E.; Lepak, L. A.; Spencer, M. G.; Martin, D. L.; Caggana, M.; Turner, J. N. J. Chromatogr. A 2006, 1111, 214–219. [11] Kim, J. E.; Cho, J. H.; Paek, S. H. Anal. Chem. 2005, 77, 7901–7907. [12] Wu, D.; Steckl, A. J. Lab Chip 2009, 9, 1890-1896. [13] Kovarik, M. L., Jacobson, S. C. Anal. Chem. 2008, 80, 657-664. [14] Cannon, D. M., Kuo, T. C., Bohn, P. W., Sweedler, J. V. Anal. Chem. 2003, 75, 2224-2230. [15] Kuo, T. C., Cannon, D. M., Shannon, M. A., Bohn, P. W., Sweedler, J. V. Sensor Actuat A-Phys 2003,102, 223-233. [16] Toda, K.; Ohira, S. I.; Ikeda, M. Anal. Chim. Acta 2004, 511, 3-10. [17] Long, Z.; Liu, D.; Ye, N.; Qin, J.; Lin, B. Electrophoresis 2006, 27, 4927-4934. [18] Tulock, J. J., Shannon, M. A., Bohn, P. W., Sweedler, J. V. Anal. Chem. 2004, 76, 6419-6425. [19] Assmann, N.; Ładosz, A.; Rudolf von Rohr, P. Chem. Eng. Technol. 2013, 36, 921-936. [20] Giordano, B. C.; Burgi, D. S.; Hart, S. J.; Terray, A. Anal. Chim. Acta 2012, 718, 11-24. [21] Al-Hetlani, E. Electrophoresis 2013, 34, 1262-1272. [22] Breadmore, M. C.; Tubaon, R. M.; Shallan, A. I.; Phung, S. C.; Abdul Keyon, A. S.; Gstoettenmayr, D.; Prapatpong, P.; Alhusban, A. A.;


Page 18 of 28

Page 19 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Ranjbar, L.; See, H. H.; Dawod, M.; Quirino, J. P. Electrophoresis 2015, 36, 36-61. [23] Breadmore, M. C.; Shallan, A. I.; Rabanes, H. R.; Gstoettenmayr, D.; Abdul Keyon, A. S.; Gaspar, A.; Dawod, M.; Quirino, J. P. Electrophoresis 2013, 34, 29-54. [24] Breadmore, M. C.; Dawod, M.; Quirino, J. P. Electrophoresis 2011, 32, 127-148. [25] Breadmore, M. C.; Thabano, J. R.; Dawod, M.; Kazarian, A. A.; Quirino, J. P.; Guijt, R. M. Electrophoresis 2009, 30, 230-248. [26] Breadmore, M. C. Electrophoresis 2007, 28, 254-281. [27] Chien, R. L.; Burgi, D. S. Anal. Chem. 1992, 64, 489A-496A. [28] Quirino, J. P.; Terabe, S. Science 1998, 282, 465-468. [29] Palmer, J.; Landers, J. P. Anal. Chem. 2000, 72, 1941-1943. [30] Palmer, J. F. J. Chromatogr. A 2004,1036, 95-100. [31] Gebauer, P.; Křivánková, L.; Pantůčková, P.; Boček, P.; Thormann, W. Electrophoresis 2000, 21, 2797-2808. [32] Gebauer, P.; Thormann, W.; Boc̆ ek, P. J. Chromatogr. A 1992, 608, 47-57. [33] Křivánková, L.; Vraná, A.; Gebauer, P.; Boček, P. J. Chromatogr. A 1997, 772, 283-295. [34] Gebauer, P.; Thormann, W.; Boček, P. Electrophoresis 1995, 16, 2039-2050. [35] Shallan, A. I.; Guijt, R. M.; Breadmore, M. C. Angew. Chem. Int. Edt. 2015, 127, 7467-7470.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

[36] Shallan, A. I.; Gaudry, A. J.; Guijt, R. M.; Breadmore, M. C. Chem. Commun. 2013, 49, 2816-2818. [37] Duffy, D. C.; McDonald, J. C.; Schueller, O. J. A.; Whitesides, G. M. Anal. Chem. 1998, 70, 4974–4984. [38] Wang, C.; Li, S. J.; Wu, Z. Q.; Xu, J. J.; Chen, H. Y.; Xia, X. H. Lab Chip 2010, 10, 639–646. [39] Foote, R. S.; Khandurina, J.; Jacobson, S. C.; Ramsey, J. M. Anal. Chem. 2005, 77, 57-63. [40] Lee, J. H.; Song, Y. A.; Han, J. Lab Chip 2008, 8, 596-601. [41] Hlushkou, D.; Dhopeshwarkar, R.; Crooks, R. M.; Tallarek, U. Lab Chip 2008, 8, 1153-1162. [42] Quist J.; Vulto P.; Van der Linden H.; Hankemeier, T. Anal. Chem. 2012, 84, 9065-9071. [43] Futrakul, N.; Sridama, V.; Futrakul, P. Renal failure 2009, 31, 140-143. [44] Samal, L.; Linder, J. A. Clin. J. AM. Soc. Nephro. 2013, 8, 131-135. [45] Nagrebetsky, A.; Jin, J.; Stevens, R.; James, T.; Adler, A.; Park, P.; Craven, A.; Shine, B.; Farmer, A. Fam. Pract. 2013, 30, 142-152. [46] de Jong, P. E.; Curhan, G. C. J. AM. Soc. Nephro. 2006, 17, 21202126.


Page 20 of 28

Page 21 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42


After bonding

Figure 1. Schematic drawing of PDMS/nanopore membrane fabrication process. PDMS chips were 5 mm-thick, 25 mm wide and 75 mm long cuboid, in which the channel was of 70 µm deep and 50 µm wide. Top PDMS chip with reservoirs was bonded with the bottom one after ACS Paragon Plus Environmentwere sandwiched in the channel plasma treating, and two different pore size membranes cross interjections between these two PDMS chips.


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

a)

Page 22 of 28

b)

-200V

1

-300V

1

FLOAT SW

4 -500V

FLOAT SW

4 +100V

+500V 2

BW FLOAT

+100V 2

BW FLOAT

3

3

-200V

+1200V

Figure 2. Schematic diagram showing the voltages applied to the fluid reservoirs for (a) sample injection, preconcentration and purification, and (b) separation and quantification. ACS Paragon Plus Environment

Page 23 of 28


a) 0s

50 s

b) 1200

0.2 ppm

1000 2.0 ppm

90 s

120 s

Enhancement factor

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

800 5.0 ppm 600 400 200 0

0

20

40

60

80

100

120

140

Time (s)

Figure 3. Enhancement factor calculation. a) Fluorescence micrographs captured during preconcentration of a 5 ppm R-phycoerythrin sample in 100 mM borate buffer (pH 8.5), the BGE was 100 mM TBE buffer, pH 8.5 with 0.5% HPMC and the applied voltages were 200, +500, -200, -500 V at reservoir 1-4 . The horizontal blue lines represent the position of the second channel and the position of the 10 nm sized membrane. b) Enhancement factor of 0.2, 2, 5 ppm R-phycoerythrin, error bars represent standard error from duplicate ACS Paragon Plus Environment experiments.


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

A) a

Page 24 of 28

b B)

c

d

1

SW

4

2

BW

c

a,b

d

3

Figure 4. A) Fluorescence micrographs captured during preconcentration (top, a-30s, b70s) and separation (bottom, c-20s, e-30s) of mixtures of 10 ppm fluorescamine labeled BSA and 2 ppm R-phycoerythrin in 100 mM borate buffer (pH 8.5), the BGE was 100 mM TBE buffer, pH 8.5 with 0.5% HPMC. The applied voltages for injection and preconcentration were -200, +500, -200, -500 V for 90s and for separation were -300, ACS Paragon Plus Environment +100, +1200, +100 V at reservoir 1-4 . B) Detection point for a, b, c, d

Page 25 of 28


BSA 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

RPE

Figure 5. Electropherograms comparing membrane device (b) with pinched injection (a and magnified inserted image c) using 10 ppm labeled BSA and 2 ppm R-phycoerythrin (RPE) in 100 mM borate buffer (pH 8.5), the BGE was 100 mM TBE buffer, pH 8.5 with 0.5% HPMC. The applied voltages for injection and preconcentration were -200, +500, -200, 500 V for 90s and for separation were -300, +100, +1200, +100 V at reservoir 1-4. ACS Paragon Plus Environment Membrane device showed 900-fold enhancement

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42


Page 26 of 28

b) 300

HSA

y = 2.3892x + 5.9554 R² = 0.9975

250

Peak height

a)

HSA

200 150 100 50 0 0

20

40

60

80

100

120

Concentration (μg mL-1)

Figure 6. Analysis of HSA in urine. a) Electropherograms for blank urine (bottom) and urine spiked with 5 ppm HSA (top). the BGE was 100 mM TBE buffer, pH 8.5 with 0.5% HPMC. The applied voltages for injection and preconcentration were -200, +500, -200, 500 V for 90s and for separation were -300, +100, +1200, +100 V at reservoir 1-4. b) The Plus Environment linear calibration curve for HASACSinParagon urine, error bars represent standard error from duplicate experiments.

Page 27 of 28

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42


HSA

HSA

(c) interferences

(b) (a)

Figure 7 . Electropherograms comparing membrane device (a) with pinched injection (b and magnified inserted image c) using labeled urine, the BGE was 100 mM TBE buffer, pH 8.5 with 0.5% HPMC. The applied voltages for injection and preconcentration were 200, +500, -200, -500 V for 90s and for separation were -300, +100, +1200, +100 V at reservoir 1-4. Membrane device showed protein concentration efficacy (about 100fold), as well as interfering substances removing ability. ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42

Microfluidic channels

Page 28 of 28

Microfluidic channels

Graphic abstract

100 nm PCTE membrane 10 nm PCTE membrane 100 nm PCTE membrane

10 nm PCTE membrane

Microchip separation

1

Cells Various proteins Inorganic ions

CCD camera

W

S

2

W

Detection point ACS Paragon Plus Environment 3

Nanoporous Membranes for Microfluidic Concentration Prior to

Recommend Documents