A modular-based integrated microsystem with multiple sample

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC

Article

A modular-based integrated microsystem with multiple sample preparation modules for automated forensic DNA typing from reference to challenging samples Yin Gu, Bin Zhuang, Junping Han, Yi Li, Xiaoyu Song, Xinying Zhou, Lei Wang, and Peng Liu Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b01560 • Publication Date (Web): 03 May 2019 Downloaded from http://pubs.acs.org on May 4, 2019

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 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 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.

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 9 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

A modular-based integrated microsystem with multiple sample preparation modules for automated forensic DNA typing from reference to challenging samples Yin Gu,1,2,# Bin Zhuang,3,# Junping Han,4 Yi Li,3 Xiaoyu Song,3 Xinying Zhou,5 Lei Wang5 and Peng Liu1,* Department of Biomedical Engineering, School of Medicine, Collaborative Innovation Center for Diagnosis and Treatment of Infectious Diseases, Tsinghua University, Beijing, 100084, China 2 State Key Laboratory of Space Medicine Fundamentals and Application, China Astronaut Research and Training Center, Beijing, 100094, China 3 Beijing CaptialBio Technology Ltd. Co., Beijing, 101111, China 4 Technology Department of Chaoyang Sub-bureau, Beijing Public Security Bureau, Beijing, 100102, China 5 CapitalBio Corporation, Beijing, 102206, China 1

ABSTRACT: The realization of an automated short tandem repeat (STR) analysis for forensic investigations is facing a unique challenge, that is DNA evidence with wide disparities in sample types, quality, and quantity. We developed a fully integrated microsystem in a modular-based architecture to accept and process various forensic samples in a “sample-in-answer-out” manner for forensic STR analysis. Two sample preparation modules (SPMs), the direct and the extraction SPM, were designed to be easily assembled with a capillary array electrophoresis (CAE) chip using a chip cartridge to efficiently achieve an adequate performance to different samples at a low cost. The direct SPM processed buccal swabs to produce STR profiles without DNA extraction in about two hours. The extraction SPM analyzed more challenging blood samples based on chitosan-modified quartz filter paper for DNA extraction. This newly developed quartz filter provided a 90% DNA extraction efficiency and the “in situ” PCR capability, which enabled DNA extraction and PCR performed within a single chamber with all the DNA concentrated in the filter. We demonstrated that minute amounts of blood (0.25 μL), highly diluted blood (0.5 μL blood in 1 mL buffer), and latent bloodstains (5-μL bloodstain on cloth washed with detergent) can be automatically analyzed using our microsystem, reliably producing full STR profiles with a 100% calling of all the alleles. This modular-based microsystem with the capability of analyzing a wide range of samples should be able to play an increasing role in both urgent situations and routine forensic investigations, dramatically extending the applications and utility of automated DNA typing.

S

hort tandem repeat (STR) analysis, the gold standard for human identification, is playing a pivotal role in criminal and civil investigations, anti-terrorism actions, military defense, and mass disaster responses.1,2 The typical procedure of a forensic STR analysis requires multiple instruments and manual operations, which can only be conducted in dedicated forensic laboratories by well-trained personnel. Although the implementation of robotic systems has enabled a streamlined STR workflow for screening casework samples and for building DNA databases automatically, these high-throughput systems are usually bulky, expensive, hard to maintain, and can only be set up in central laboratories.3,4 However, many urgent situations, such as high-profile crimes and terror attacks, demand a person’s STR profile to be obtained as soon as possible at the scenes for more prompt responses and interventions.2,5 For example, the on-site analysis of biological

evidence that was left by the perpetrator could significantly advance the investigation before the suspect has fled.6 In these situations, high throughput may not be required, but the automation and the portability are highly desired. To fulfill this need, several groups including us have reported the successful development of integrated microfluidic devices, which could generate STR profiles in a “sample-in-answer-out” manner.6-10 For examples, Le Roux et al. reported a microfluidic system that integrated DNA extraction, PCR, and capillary electrophoresis together on a single multi-layer plastic device.8 Our group also developed a fully integrated microsystem for STR typing within 2 hours.10 Commercialized systems, such as RapidHITTM (IntegenX, Pleasanton, CA) and DNAScanTM (GE Healthcare, Pittsburgh, PA),11,12 have been deployed to forensic laboratories for extensive validation. This progress announces

ACS Paragon Plus Environment

Analytical Chemistry 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

the era of automatically conducting STR analysis whenever and wherever needed is approaching. While the sensitivity and the reproducibility of these automated instruments have been successfully reached for analyzing reference and routine casework samples, a unique challenge faced by forensic practitioners, i.e., the wide disparities in sample type, quality, and quantity, has not been fully addressed.13-15 Apparently, different types of forensic evidence with various DNA contents need different sample processing strategies in order to generate the best typing profiles timely and to keep the costs affordable. For examples, reference samples, such as buccal swabs, can be analyzed using direct PCR kits without DNA extractions,16 while latent bloodstains or highly diluted evidence may require more careful DNA extractions.17,18 These common practices adopted by forensic scientists explicitly indicate that it is infeasible to handle all kinds of samples using a single device, on which all the structures are permanently integrated together during manufacture. Therefore, the modular design strategy developed in microfluidics has the potential to fill the gap between the diverse samples and the limited processing capability of an integrated microsystem. Instead of fabricating a microdevice as a whole, a modular-based design allows users to assemble a microfluidic system using basic building blocks with different functions.19,20 For instances, Shaikh et al. developed a microfluidic breadboard system, in which multiple chip modules were interconnected to form large lab-on-a-chip systems for different biochemical applications.21 Even Lego® bricks were employed to build microsystems quickly.22,23 Unfortunately, these methodologies were initially designed to produce prototype systems for testing new ideas within a short turnaround time. Thus far, no modular-based microsystems have been developed for “sample-in-answer-out” genetic analysis from various sample sources. Although several fully integrated microfluidic systems including ours were constructed by assembling several separated microchips together, only a single assembly was presented in each study without fully exploring the advantages of the modular design strategy.7,10 We should adopt this design concept to develop a more flexible microsystem, in which the sample preparation module can be changed according to the sample types. Since minute amount, latent, or highly diluted evidence is often encountered in crime scene investigations, the sensitivity of the sample preparation module needs more careful optimizations in the modular systems.17 The conventional nucleic acid extraction methods, such as silica- or magnetic bead-based solid phase extractions, were not initially designed for the microchip format.24,25 As a result, the direct translation of these methods onto chip formats often requires complicated microstructures and tedious operations. In recent years, several groups have developed new extraction methods coupled with “in situ” PCR, in which PCR is performed on the capture phases without elution.26-29 Our group demonstrated that chitosanmodified Fusion 5 filter paper can also provide a 95% DNA extraction efficiency and resist the elution for “in situ” PCR.30 These methods have been proved highly suitable for the integration into a “sample-in-answer-out” system, as a single reactor can work for both the extraction and the amplification. Additionally, since the entire nucleic acid extract can be used for amplification without elution, DNA can be concentrated from a large volume of a sample. However, one major drawback associated with the “in situ” PCR is that the solid phase usually has an inhibitory effect to amplification due to the increased

surface-to-volume ratio, compromising the advantages of “in situ” amplification.29,30 Therefore, this non-elution extraction method needs further optimization to achieve a higher sensitivity for forensic DNA analysis. In the current study, we successfully developed a fully integrated microsystem designed in a modular architecture for handling diverse forensic DNA samples in a “sample-inanswer-out” manner. According to sample types, two different sample preparation modules (SPMs) can be easily assembled with a capillary array electrophoresis (CAE) chip using a chip cartridge. For reference samples, such as buccal swabs, a direct SPM was employed in order to simplify the operation and to lower the cost. While for highly diluted blood or latent bloodstains, an extraction SPM based on chitosan-modified quartz filter paper for DNA extraction was designed. This modular-based, automated microsystem with the capability of analyzing a wide range of forensic samples should be able to significantly promote the wide applications of the automated STR analysis in forensic investigations.

Figure 1. The modular-based microsystem for forensic short tandem repeat analysis. A) Exploded view of the modular-based microdevice, consisting of a sample preparation module (SPM), a capillary array electrophoresis chip, a chip cartridge, and an electrode holder with three injection electrodes. Structures of B) the direct SPM and C) the extraction SPM. D) Schematic of the glass capillary array electrophoresis chip. Three 19-cm-long separation channels, two for samples and one for the allelic ladder, were designed on the device.



EXPERIMENTAL SECTION Fully integrated microdevice in a modular architecture. The fully integrated microdevice designed in a modular architecture consists of a sample preparation module (SPM), a glass capillary array electrophoresis chip (CAE chip), and a plastic cartridge with injection electrodes for device assembling


Page 2 of 9

Page 3 of 9 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 (Figure 1A and S1). While the CAE chip is permanently glued with the cartridge, the SPM can be selectively assembled with the cartridge by users according to the sample types that are analyzed. To assemble the device, the SPM is inserted into the cartridge and fixed by using the electrode holder with the injection electrodes. The SPM is formed by thermally bonding two poly(methyl methacrylate) (PMMA) layers together, which are fabricated using injection molding. We designed two different SPMs for processing different forensic samples: a direct SPM and an extraction SPM. The direct SPM contains an inlet, a 15-μL PCR reactor with two microvalves on each side, an injection chamber, and an outlet (Figure 1B). The extraction SPM has a similar structure except that a sample chamber with a microvalve is connected to the PCR reactor and a piece of quartz filter paper is embedded into the reactor for DNA extraction and “in situ” PCR (Figure 1C). The CAE chip contains three 19-cm-long separation channels, one for the analysis of the allelic ladder and the other two for samples (Figure 1D). The connection between the SPM and the CAE chip is realized using the injection electrodes developed previously by our group.31 The modular-designed microdevice is operated on a custom-built control and detection instrument (Figure S2), which has dimensions of 48 × 35 × 35 cm. The detail information of this instrument can be found in our previous study.10,31 Modification of quartz filter paper with chitosan. Chitosan (low molecular weight), MES (2-(N-Morpholino) ethanesulfonic acid), SDS (sodium dodecyl sulfonate), and GPTMS ((3-Glycidyloxypropyl) trimethoxysilane) were all purchased from Sigma-Aldrich (St.Louis, MO). Quartz filter paper was from Whatman (QHA, GE Healthcare, Pittsburgh, PA). All solutions were prepared in water purified to 18.2 MΩ.cm by Milli-Q Advantage A10 (Millipore, Massachusetts, MA). In the process of chitosan modification, a 47-mmdiameter piece of quartz filter was first treated with GPTMS (2.5 % in methanol) for 1 hour to add epoxy groups on the surfaces of filter fibers. Then, the treated filter was submerged into a 20 mL MES containing 10 μL 1% chitosan (prepared with 0.1% acetic acid, pH 6.0) followed by an overnight incubation on a shaker. After modification, the filter paper was washed with acetic acid (0.1%, pH 6.0) for three times and dried completely at 50 °C in a vacuum drying oven. After that, 2-mmdiameter discs of the quartz filter paper were punched off and stored in a sealed petri dish at room temperature until use. As a comparison, Fusion 5 filter paper (GE Healthcare, Pittsburgh, PA) was modified with chitosan according to the protocol published previously by our group.30 Evaluation of chitosan-modified quartz filter for DNA extraction. Standard K562 human genomic DNA (Promega, Madison, WI) was employed to verify the DNA capture efficiency of the chitosan-modified quartz filter paper. DNA samples were prepared with MES (pH 5.0) to a concentration of 12.5 ng/μL. The 2-μL DNA sample was pipetted onto the surface of the filter paper disc embedded into the microchip (Figure S3). Then, the reaction chamber was sealed with a piece of ARseal™ adhesive tape (Adhesives Research, Glenrock, PA). 200 μL of DI water was aspirated through the reaction chamber at a speed of 200 μL/min to wash the filter paper disc. Finally, the microchip was disassembled and the disc was taken out for real-time PCR analysis in an Eppendorf tube. DNA samples prepared in 1×TE buffer (pH 7.0 and 9.0) were also captured by the modified filter paper to verify the electrostatic capture mechanism. As a control, unmodified quartz filter paper

was used to capture DNA at pH 5.0 in parallel. To test the DNA retention during elution, 25-ng K562 DNA was first captured on the filter discs and then washed with 200-μL DI water. After that, the filter discs were washed under different elution conditions, including 1-mL DI water, 100-μL PCR buffer (pH 8.5), and 100-μL 1×TE (pH 9.0). After elution, both the eluants and the filter discs with remained DNA were quantified by realtime PCR. To evaluate the capture capacity of the filter paper, a series of samples containing 25, 50, 100, 200, and 300 ng K562 DNA were captured by the filters using the same protocol. To test the PCR inhibitory effect, a serially diluted K562 DNA of 1, 5, 10, and 25 ng were pipetted onto the surface of the modified Fusion 5 filter paper and quartz filter paper, respectively. Then the paper discs were quantified by real-time PCR with or without additional DNA polymerases (2.5 more Units in 25-μL reactions). All the DNA quantifications in the study were performed using real-time PCR on a Bio-Rad iQ5 system (Bio-Rad, Hercules, CA). A pair of primers (forward: 5’CCCTGGGCTCTGTAAAGAA; and reverse: 5’-ATCAGA GCTTAAACTGGGAAGCTG) which amplifies a 106-bp and a 112-bp fragment from the X and the Y chromosomes, respectively, was employed to quantitate human genomic DNA. Standard curves for quantification were generated using a series of diluted standard K562 DNA with or without filter paper according to the tested samples. A 25-μL mixture for real-time PCR was composed of 0.425 μL of each primer (Sangong, Beijing, China), 12.5 μL of PowerUp 2× SYBR real-time PCR premix (Thermo Fisher, Waltham, MA), 11.65 μL of DI water, and the filter paper disc. The thermal cycling included an initial activation of Dual-Lock™ Taq polymerases at 95 °C for 5 min, followed by 35 cycles of 95 °C for 30 s, 60 °C for 30 s, and 72 °C for 30 s, and a final extension for 10 min at 72 °C. Evaluation of DNA extraction. Human whole blood samples were obtained from volunteers after informed consents for research and were anticoagulated in evacuated blood collection tubes. The lysis of blood cells was performed in an Eppendorf tube by mixing blood with 200-μL or 1000-μL lysis buffer (0.1% CTAB (cetyltrimethylammonium bromide), 1.5M NaCl, MES, pH 5.0) and incubating at room temperature for 15 minutes. The lysate was aspirated into the DNA extraction chamber of the microchip (Figure S3) at a flow rate of 100 μL/min, followed by washing with 200 μL of 1% SDS and 200 μL of DI water at a speed of 200 μL/min. Finally, the microchip was disassembled, and the filter disc was taken out for real-time PCR analysis or STR typing. Conventional DNA extractions using QIAamp® DNA Micro kits (Qiagen, Germantown, MD) were also conducted by following the user manual. The PowerPlex® 21 System kit (Promega) was employed to test the extracted DNA for STR analysis. The QF discs with captured DNA were dropped into Eppendorf tubes with a 25μL PCR mixture consisted of 5 μL of 5× Primer Pair Mix, 5 μL of 5× Master Mix, 0.5 μL of AmpliTaq Gold™ DNA polymerase, and 14.5 μL of DI water. A thermal cycling protocol included an initial activation at 96 °C for 1 min, followed by 25 cycles of 94 °C for 10 s, 59 °C for 60 s and 72 °C for 30 s, and a final extension for 20 min at 60 °C. Automated STR analysis of buccal swabs on the microsystem. To collect buccal swabs samples, volunteers with informed consents first rinsed their mouths with water and cells were then collected from buccal mucosa with a cotton swab (Huachenyang, Shenzhen, China). This swab was gently rinsed in 100-μL TE buffer to release cells. Then, 25 μL of the TE



buffer was pipetted into another tube to dissolve freeze-dried PCR mixture powders, which were prepared from 25 μL of PCR mixture consisting of 5 μL of 5× Primer Pair Mix, 5 μL of 5× Master Mix, and 15 μL of DI water. After that, 15 μL of the PCR mixture was filled into the reactor of the direct SPM and the inlets were sealed with ARseal™ tape (Figure 1B). Once the micro-device was loaded into the instrument, the rest of the STR analysis was automatically conducted under the control of a custom-built program. The on-chip PCR protocol was the same as that of the tube reaction shown above. After the amplification, PCR amplicons in the reactor were drawn into the upper compartment of the injection chamber. The amplicons together with the sizing standards and formamide in the lower compartment were transferred to the sample reservoir of the CAE chip, which were filled with 4.5% (w/v) linear polyacrylamide (LPA) with 6 M urea in 1×Tris TAPS EDTA (TTE). Following the sample injection step by applying an electric field of 100 V/cm for 90 s between the sample and the waste reservoirs, the electrophoretic separation was performed with an electric field of 180 V/cm between the cathode and anode at 60 oC. Once the analysis was finished, the SPM was disposed of and the CAE chip was cleaned with DI water for next use. Automated STR analysis of blood samples on the microsystem. Three types of blood samples, whole blood, highly diluted blood, and latent bloodstains, were tested on the microsystem. Human whole blood samples were lysed with 200 μL of lysis buffer in 15 minutes. The lysate was then loaded into the sample chamber of the extraction SPM for the rest of the STR analysis. First, the lysate was aspirated into the reactor, where DNA was captured by the modified quartz filter. Then, 200 μL of 1% SDS and 200 μL of DI water were drawn through the filter at a speed of 200 μL/min. After the PCR mixture was loaded into the reactor from the inlet, the thermal cycling was conducted, followed by the sample injection and capillary electrophoretic detection. Highly diluted blood samples were prepared by diluting 0.1 or 0.5 μL of whole blood in a 1-mL lysis buffer. After an incubation of 15 minutes, the lysate was aspirated into the microdevice (Figure 1C) at a speed of 100 μL/min, followed by the same on-chip analysis as that of whole blood. To prepare latent bloodstains, 5 μL of whole blood was pipetted onto each 5-mm-diameter substrate of either 100% polyester or 100% cotton cloth. After dried at room temperature for 20 min, the bloodstains were washed with laundry detergent in a beaker with hands until the blood became invisible. The stains were further washed with water at least 3 times to remove any residual detergent. A luminol solution (Pioneer Forensics, Loveland, CO) was sprayed onto the substrates for visualizing the washed bloodstains. Then, the dried textile samples were dropped into 600 μL of lysis buffer and lysed for 20 minutes with vortexing for 15 seconds every 10 minutes at room temperature. The lysate was transferred to the sample chamber of the extraction SPM for the analysis on the microsystem.  RESULTS AND DISCUSSION Analysis of buccal swabs with the direct SPM. Buccal swabs, the most common reference samples, usually contain a large number of epithelial cells and less inhibitory components, making the direct amplification without DNA purification feasible. Several previous studies have verified that buccal swabs can be directly amplified with a high success rate.32,33 Apparently, this direct PCR procedure can shorten the turnaround time, lower the cost, and simplify the structure of

the microdevice for handling buccal swabs automatically. To ensure the successful analysis of swabs on our microsystem, we first tested the sensitivity of the direct SPM by loading a series of PCR solutions containing 5.0, 2.5, and 1.0 ng of K562 standard genomic DNA into the PCR reactor for the STR amplification followed by the electrophoretic detection on the modular-based device. As demonstrated in Figure S4, 1 ng of input DNA can still generate a full STR profile using the Promega PowerPlex® 21 System on our microdevice automatically. The allele calls from these profiles were always concordant with the conventional results. The successful separation of the internal sizing standards along with the PCR products proved that the injection chamber coupled with the injection electrode can sufficiently mix all the reagents together.

Figure 2. Automated short tandem repeat analysis of buccal swabs on the integrated microsystem with the direct SPM. A) Protocol for processing buccal swabs. A buccal swab was first gently rinsed in 100-μL TE buffer to release cells. Then, 25 μL of the TE buffer was pipetted into another tube to dissolve freeze-dried PCR mixture powders. After that, 15 μL of the mixture was loaded into the device for STR analysis. B) A typical STR profile automatically


Page 4 of 9

Page 5 of 9 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 obtained from a buccal swab on the microsystem with the direct SPM. C) Mean peak height ratios calculated from 24 buccal swabs. Error bars represent one standard deviation.

Next, as illustrated in Figure 2A, we developed a simplified protocol for processing buccal swabs on the microsystem. The release of cells performed manually in a tube allowed the preservation of a portion of the samples for the future validation. We analyzed a total of 24 buccal swabs collected from 12 volunteers to demonstrate the performance of our system. As summarized in Table S1, full STR profiles were obtained from 22 out of 24 samples, resulting in a success rate of 92%. Figure 2B demonstrated a typically full STR profile obtained using our system, showing all the alleles together with the internal sizing standards were baseline resolved. The allele calling of these 21 loci in the PowerPlex® 21 System was 100% concordant with that provided by the conventional method. The peak height ratios (PHR), defined as the ratio of the signal strengths between the lower and the higher allele in a heterozygous locus, were also calculated from these samples for all the loci. As shown in Figure 2C, the PHRs were about 0.6-1.0, illustrating the amplification balance within each locus. While the overall performance of our system is acceptable, we found there were stochastic drop-ins and relatively wide variations in peak heights across all the loci. We believe this is because the Promega PowerPlex® 21 System was directly used in our microsystem without any optimization. Several previous studies have proved the changes in the PCR conditions may require further fine tuning of the amplification system.6,10 Since the detailed information of the Promega PowerPlex® 21 was not available to us, we decided to focus our work on the development of the microsystem in the current study. Nevertheless, since no DNA extraction was involved and a single-chamber structure was designed for sample preparation, this fully integrated microdevice with the direct SPM possesses the potential to become a more reliable and affordable genetic analyzer for rapidly identify a person in various locations, such as crime scene, security checkpoint, and police custody suite. Chitosan-modified quartz filter for nucleic acid extraction. Forensic casework investigations demand a high sensitivity to analyze minute amount, diluted, or latent evidence. While we have developed a DNA extraction method based on chitosan-modified Fusion 5 filter paper,30 we found that PCR performed with this filter paper has a lower efficiency due to the adsorption of DNA polymerases and the modification of the filter has a poor reproducibility owing to the nonspecific coating of chitosan. In addition, since the branded Fusion 5 filter paper contains glass microfibers bound with unknown organic binders, the modification was not completely under the control. Here we employed another type of filter paper - quartz filter (QF) which contains near 100% SiO2 as the capture phase to replace Fusion 5. We modified the QF with GPTMS ((3Glycidyloxypropyl) trimethoxysilane) followed by the specific linking of chitosan via the epoxy-amino reaction (Figure S5). This method ensures the addition of a uniform layer of chitosan onto the fibers, improving the reproducibility of the modification (data not shown). As demonstrated in our previous study,30 we have proved the chitosan-modified Fusion 5 filter captures DNA via a “twomode” mechanism. In a pH below 6.3, anionic DNA “actively” binds to the chitosan-modified fibers via electrostatic adsorption. In addition, the physical entanglement of the longchain DNA with the fibers can enhance the DNA capture,

resulting in an efficiency over 90%. In a pH around 8.5, the deprotonated chitosan cannot capture DNA any more. However, the physical trapping of DNA within the fiber matrix effectively prevents the elution and enable “in situ” PCR. Here we believe the chitosan-modified quartz filter should have the same DNA capture effect as that of the Fusion 5. To prove it, we conducted a similar test to that published using standard K562 human genomic DNA. As shown in Figure 3A, unmodified quartz filter can only capture 23.2% of the input K562 DNA. This result is close to that of the unmodified Fusion 5 filter reported previously, indicating that DNA molecules were only seized in the QF matrix by physical entanglement.2628,30 When the chitosan-modified quartz filter was employed, the DNA capture efficiency was dramatically increased to 97.2% at pH 5. By contrast, the capture efficiencies remained low at pH 7 and 9. This observation was consistent with the DNA capture of the chitosan-modified Fusion 5 filter.30 Next, we verified the “in situ” PCR of the chitosan-modified QF paper. As shown in Figure 3B, only 0%, 10.4% (2.59 ng), and 7.5% (1.84 ng) were detected in the eluants with 1 mL of DI water, 100 μL of TE buffer (pH=9.0), and 100 μL of PCR mastermix (pH=8.5), respectively, the majority of DNA were left on the filter. These results proved that PCR mix can be directly loaded into the chamber to enable the “in situ” amplification without the worry of DNA loss.

Figure 3. Evaluation of the DNA capture by chitosan-modified quartz filter. A) Comparisons of DNA capture efficiencies of K562 genomic DNA among unmodified quartz filter (QF) in pH of 5, chitosan-modified QF in pH of 5, 7, and 9. B) K562 DNA capture capacity of a 2-mm-diameter piece of chitosan-modified quartz filter. C) Elution of captured K562 DNA from filter paper in different conditions. About 0%, 10.4%, and 7.5% of the captured DNA were washed off with water, TE buffer, and PCR buffer, respectively. D) Comparisons of the Ct values of real-time PCR among chitosan-modified Fusion 5 filter with more enzymes, chitosan-modified QF, chitosan-modified QF with more enzymes, and control without any filter. (n=3, error bars represent one standard deviation.)

The DNA capture capacity of the QF was measured by capturing a series of K562 DNA samples containing a range of 25-300 ng. We found more than 90% of K562 DNA can be retained in the QF when the input DNA amount was between



25 and 200 ng (Figure 3C). When the input was increased to 300 ng, the retention rate was dropped to 85.4% due to the saturation of the capacity. Therefore, we estimated that the capacity of the QF paper is ~285 ng (assuming a capture efficiency of 90%). We then evaluated the PCR inhibitory effect of the chitosan-modified quartz filter together with the Fusion 5. As shown in Figure 3D, the PCR amplification with Fusion 5 filter paper demonstrated the highest Ct values even with more DNA polymerases (2.5 more Units in 25 μL) added into the reaction. By contrast, the PCR efficiencies of the quartz filters with additional DNA polymerases can even become similar to those of the controls in which no filter discs were added and were almost 10 times higher than those of the Fusion 5 filter. The low inhibition of the QF to PCR could enable more sensitive tests in our system. To explore the extraction capability, we compared the quartz filter with the QIAamp® DNA Micro kit in processing human whole blood samples. As illustrated in Figure 4, DNA extractions from 1, 0.5, and 0.25 μL blood could produce 70.0±15.0, 34.1±7.1, and 18.5±3.7 ng of DNA, respectively, using our modified quartz filter, while the Qiagen kit only provided 24.1±1.0, 13.7±1.1, and 1.1±0.3 ng of DNA with the elution volume of 50 μL. Since the entire filter disc with extracted DNA can be used directly in a 25-μL PCR, the template concentration extracted from 1 μL blood (2.8 ng/μL) was much higher (~72 times) than that provided by the Qiagen kit (0.039 ng/μL, assuming 2 μL of eluted DNA was used). For more challenging samples, such as highly diluted blood, our method can still extract 6.5±1.4 ng of DNA from 0.1 μL whole blood diluted 10,000-fold, while no DNA was obtained with the Qiagen kit.

a series of human whole blood samples to verify the performance of our microsystem with the extraction SPM. As demonstrated in Figure S6, the full STR profiles could be obtained even from 0.25 μL of blood, showing all the alleles were accurately resolved and called. Such a detection limit enabled by the high capture efficiency and the low PCR inhibition of the QF paper has not been reported yet. As shown in Figure 4, only 1.1 ng DNA can be extracted from 0.25 μL blood using the QIAamp® DNA Micro kit. Since the elution volume is 50 μL, it is challenging to obtain full STR profiles from such a minute amount of blood even using the conventional method.

Figure 4. Comparison of DNA extractions from whole blood and highly diluted blood samples using the chitosan-modified quartz filter paper embedded in the microchip and the QIAamp® DNA Micro kit. (n=3, error bars represent one standard deviation.)

Automated analysis of minute amount and highly diluted blood samples. Due to the “in situ” PCR capability, our chitosan-modified quartz filter should be an excellent choice for sample preparation in a “sample-in-answer-out” microsystem. As shown in Figure 1C, we embedded a piece of quartz filter paper into the sample preparation module to enable the automated analysis of forensic evidence. As illustrated in Figure 5A, samples, such as blood or bloodstains, were first mixed with the lysis buffer followed by an incubation of 10-20 min. This lysis procedure performed outside of the chip can enable that a portion of samples may be retained for further validation in caseworks. The lysates were manually loaded into the sample chamber of the extraction SPM. The rest of the operations, including DNA extraction, PCR amplification, and CE detection, were all conducted by the instrument automatically under the control of the program. We first tested


Page 6 of 9

Page 7 of 9 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 Figure 5. STR analysis of highly diluted blood samples on the fully integrated microsystem with the extraction SPM. A) Processing protocol of blood and stain samples tested on the fully integrated microsystem with the extraction SPM. Blood samples were mixed with lysis buffer and incubated for 10-20 minutes. The lysates were then loaded into the sample chamber of the extraction SPM for automated DNA analysis. B) and C) STR electropherograms of these two highly diluted blood samples obtained using the fully integrated microsystem.

Figure 6. Forensic STR analysis conducted from latent bloodstains using the fully integrated microsystem with the extraction SPM. A)

Photographs of polyester and cotton substrates, dried bloodstains, washed bloodstains, and luminol stains. DNA extractions were performed three times and error bars represent one standard deviation. B) and C) Full PowerPlex® 21 profiles successfully obtained from latent bloodstains on polyester and cotton using the microsystem.

Since all the DNA can be concentrated within the filter for the downstream amplification, the extraction SPM should be highly suitable for analyzing diluted samples. To prove the performance of our system, we prepared a male and a female blood sample by diluting 0.5 μL of blood in 1-mL lysis buffer (2000-fold dilution). DNA extractions could produce 25.2±0.2 and 25.6±0.7 ng of DNA from these two highly diluted samples, respectively, using the microchip with the embedded QF. In the following automated STR analyses performed on the microsystem, both STR profiles were successfully obtained with the accurate information of all the 21 loci (Figure 5B and 5C). These results indicated that our fully integrated system with the extraction SPM could extract and enrich DNA in the reactor from large-volume samples, significantly improving the sample processing capability of the fully integrated microsystem. Highly sensitive typing of latent bloodstains. A perpetrator often attempts to hide bloodstains by washing surfaces and objects with cleaning agents after committing a crime. These latent bloodstains usually have low DNA contents, contain PCR inhibitors, and are possibly degraded by washing agents.34 As a result, the capability of analyzing such a challenging sample is a perfect showcase to illustrate how the performance of our microsystem is. So far, no fully integrated devices have demonstrated such a high sensitivity yet. In the current study, two types of fabrics, 100% polyester and 100% cotton, were employed as substrates to prepare latent bloodstain samples. Figure 6A shows that the bloodstains on the fabrics cannot be seen by our eyes after wash and even the luminol reactions just showed dim lights. The washed fabrics were vortexed with the lysis buffer vigorously and incubated for 20 min. The 600-μL lysates were then aspirated into the microdevice with the extraction SPM for automated STR analysis. The latent blood samples of polyester and cotton can provide 19.6±3.6 and 22.7±2.5 ng of DNA, respectively, which were approximately equal to the amount of DNA from 0.3 μL of whole blood (Figure 6A). The fully automated STR analyses on the microsystem generated full STR profiles containing all the alleles from these two latent blood samples (Figure 6B and 6C). The allele calling was 100% concordant to that of the conventional method. Negative controls were performed regularly to ensure no contaminations throughout the study. These results proved that our new quartz filter-based DNA extraction method can be well integrated into the microfluidic device to achieve an excellent sensitivity. The automated analyses of the minute amount of blood, highly diluted blood, and latent bloodstains on our microsystem was achieved due to the use of the chitosan-modified quartz filter. This new extraction method demonstrated i) over 90% of the DNA capture efficiency by the filter, ii) the complete use of extracted DNA enabled by non-elution feature, and iii) low PCR inhibition effect of the modified quartz filter. When we translated this filter paper into the microchip platform, more prominent advantages could be attained due to the “in situ” PCR capability. First, only a single chamber was required for both DNA extraction and amplification, simplifying the microstructures of the device. Second, PCR reagents could be



easily loaded and fully mixed with the captured DNA without the mixing issue, improving the reliability of the operations. Third, the entire DNA extract could be used for amplification, leading to an improved sensitivity. Owing to the use of this quartz filter, we demonstrated that the extraction SPM coupled with the modular-designed device has one of the simplest structures among fully integrated microdevices reported in literature so far.  CONCLUSION In summary, a fully integrated microsystem in a modularbased architecture was successfully developed for processing and typing diverse forensic DNA samples in a “sample-inanswer-out” manner. This modular-based design strategy, in which all the components can be flexibly assembled together, possesses the potential to dramatically expand the processing capabilities of an integrated microfluidic system. In the current study, for the first time, we demonstrated a truly modular-based microfluidic system for forensics, in which two sample preparation modules: the direct and the extraction type, can be selectively assembled into a chip cartridge to form a selfcontained microdevice. This design allows the user to choose the most suitable way to analyze each type of sample in order to make the STR typing cheaper, faster, and simpler. In the future, more SPMs can be developed to fit into the microsystem to fully take the advantages of the modular design. In the protocols of processing crude samples, several operations, such as mixing of samples with lysis buffers, were still performed outside of the device manually. Large vertical chambers can be easily attached to the inlets of the SPM for accepting and processing raw samples with solid carriers, such as swabs and stains, to realize a truly “sample-in-answer-out” analysis. Moreover, the modular-based microsystem developed here is still a proof-of-principle demonstration. Thus, it will be of great interest to perform more thorough validation tests to evaluate the robustness and the reliability of the system. In addition, the further optimization of the Promega PowerPlex® 21 System is also required. We believe the extensibility provided by the modular design strategy together with the outstanding sensitivity achieved with the quartz filter will be highly likely to expand the applications and utility of our fully integrated microsystem in the future.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Additional information as noted in text (PDF).

AUTHOR INFORMATION Corresponding Author *E-mail: [email protected]. Phone: +86-10-62798732. Fax: +86-10-62798732.

Author Contributions 1

Y.G. and B.Z. contributed equally to this work.

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT

Financial support was provided by the National Key Research and Development Program of China (No. 2016YFC0800703) from the Ministry of Science and Technology of China.

REFERENCES (1) Jobling, M. A.; Gill, P. Nat. Rev. Genet. 2004, 5, 739-751. (2) Bruijns, B.; van Asten, A.; Tiggelaar, R.; Gardeniers, H. Biosensors 2016, 6, 41. (3) Graham, E. A. Forensic Sci. Med. Pathol. 2005, 1, 285-288. (4) Sim, J. E.; Park, S. J.; Lee, H. C.; Kim, S. Y.; Kim; J. Y.; Lee, S. H. J. Forensic Sci. 2013, 58, 989-992. (5) Horsman, K. M.; Bienvenue, J. M.; Blasier, K. R.; Landers, J. P. J. Forensic Sci. 2007, 52, 784-799. (6) Liu, P.; Seo, T. S.; Beyor, N.; Shin, K. J.; Scherer, J. R.; Mathies, R. A. Anal. Chem. 2007, 79, 1881-1889. (7) Hopwood, A. J., Hurth, C., Yang, J., Cai, Z., Moran, N., LeeEdghill, J. G., Nordquist, A., Lenigk, R., Estes, M. D., Haley, J. P., McAlister, C. R., Chen, X., Brooks, C., Smith, S., Elliott, K., Koumi, P., Zenhausern, F., Tully, G. Anal. Chem. 2010, 82, 6991-6999. (8) Le Roux, D.; Root, B. E.; Hickey, J. A.; Scott, O. N.; Tsuei, A.; Li, J.; Saul, D. J.; Chassagne, L.; Landers, J. P.; de Mazancourt, P. Lab Chip 2014, 14, 4415-4425. (9) Kim, Y. T.; Lee, D.; Heo, H. Y.; Sim, J. E.; Woo, K. M.; Kim, D. H.; Im, S. G.; Seo, T. S. Biosens. Bioelectron. 2016, 78, 489-496. (10) Han, J.; Gan, W.; Zhuang, B.; Sun, J.; Zhao, L.; Ye, J.; Liu, Y.; Li, C. X.; Liu, P. Analyst 2017, 142, 2004-2012. (11) Date-Chong, M.; Hudlow, W. R.; Buoncristiani, M. R. Forensic Sci. Int. Genet. 2016, 23, 1-8. (12) Tan, E.; Turingan, R. S.; Hogan, C.; Vasantgadkar, S.; Palombo, L.; Schumm, J. W.; Selden, R. F. Investig. Genet. 2013, 4, 16. (13) Wickenheiser, R. A. J. Forensic Sci. 2002, 47, 442-450. (14) Turingan, R. S.; Vasantgadkar, S.; Palombo, L.; Hogan, C.; Jiang, H.; Tan, E.; Selden, R. F. Investig. Genet. 2016, 7, 2. (15) Cavanaugh, S. E.; Bathrick, A. S. Forensic Sci. Int. Genet. 2018, 32, 40-49. (16) Wang, D. Y.; Gopinath, S.; Lagace, R. E.; Norona, W.; Hennessy, L. K.; Short, M. L.; Mulero, J. J. Forensic Sci. Int. Genet. 2015, 19, 148-155. (17) van Oorschot, R. A.; Ballantyne, K. N.; Mitchell, R. J. Investig. Genet. 2010, 1, 14. (18) Liu, P.; Scherer, J. R.; Greenspoon, S. A.; Chiesl, T. N.; Mathies, R. A. Forensic Sci. Int. Genet. 2011, 5, 484-492. (19) Rhee, M.; Burns, M. A. Lab Chip 2008, 8, 1365-1373. (20) Yuen, P. K.; Bliss, J. T.; Thompson, C. C.; Peterson, R. C. Lab Chip 2009, 9, 3303-3305. (21) Shaikh, K. A.; Ryu, K. S.; Goluch, E. D.; Nam, J. M.; Liu, J.; Thaxton, C. S.; Chiesl, T. N.; Barron, A. E.; Lu, Y.; Mirkin, C. A.; Liu, C. Proc. Natl. Acad. Sci. U. S. A. 2005, 102, 9745-9750. (22) Lind, K. R.; Sizmur, T.; Benomar, S.; Miller, A.; Cademartiri, L. Plos One 2014, 9, e100867. (23) Nie, J.; Gao, Q.; Qiu, J. J.; Sun, M.; Liu, A.; Shao, L.; Fu, J. Z.; Zhao, P.; He, Y. Biofabrication 2018, 10, 035001. (24) Price, C. W.; Leslie, D. C.; Landers, J. P. Lab Chip 2009, 9, 2484-2494. (25) Reinholt, S. J.; Baeumner, A. J. Angew. Chem. Int. Ed. 2014, 53, 13988-14001. (26) Jangam, S. R.; Agarwal, A. K.; Sur, K.; Kelso, D. M. Biosens. Bioelectron. 2013, 42, 69-75. (27) Oblath, E. A.; Henley, W. H.; Alarie, J. P.; Ramsey, J. M. Lab Chip 2013, 13, 1325-1332. (28) Gan, W.; Zhuang, B.; Zhang, P.; Han, J.; Li, C. X.; Liu, P. Lab Chip 2014, 14, 3719-3728. (29) Pandit, K. R.; Nanayakkara, I. A.; Cao, W.; Raghavan, S. R.; White, I. M. Anal. Chem. 2015, 87, 11022-11029. (30) Gan, W.; Gu, Y.; Han, J.; Li, C. X.; Sun, J.; Liu, P. Anal. Chem. 2017, 89, 3568-3575. (31) Zhuang, B.; Han, J.; Xiang, G.; Gan, W.; Wang, S.; Wang, D.; Wang, L.; Sun, J.; Li, C. X.; Liu, P. Lab Chip 2016, 16, 86-95. (32) Myers, B. A.; King, J. L.; Budowle, B. Forensic Sci. Int. Genet. 2012, 6, 640-645.


Page 8 of 9

Page 9 of 9 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 (33) Flores, S.; Sun, J.; King, J.; Budowle, B. Forensic Sci. Int. Genet. 2014, 10, 33-39. (34) Vandewoestyne, M.; Lepez, T.; Van Hoofstat, D.; Deforce, D. J.

Forensic Sci. 2015, 60, 707-711.


A modular-based integrated microsystem with multiple sample

Recommend Documents