Subscriber access provided by UNIV OF CAMBRIDGE
Article
Integrated acoustic separation, enrichment and microchip PCR detection of bacteria from blood for rapid sepsis diagnostics Pelle Daniel Ohlsson, Mikael Evander, Klara Petersson, Lisa Mellhammar, Ari Lehmusvuori, Ulla Karhunen, Minna Soikkeli, Titta Seppä, Emilia Tuunainen, Anni Spangar, Piia von Lode, Kaisu Rantakokko-Jalava, Gisela Otto, Stefan Scheding, Tero Soukka, Saara Wittfooth, and Thomas Laurell Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b00323 • Publication Date (Web): 04 Jun 2016 Downloaded from http://pubs.acs.org on June 12, 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 8
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
Integrated acoustic separation, enrichment and microchip PCR detection of bacteria from blood for rapid sepsis diagnostics Pelle Ohlsson‡a*, Mikael Evander‡a, Klara Petersson‡a, Lisa Mellhammarb,c, Ari Lehmusvuorid, Ulla Karhunend, Minna Soikkelid, Titta Seppäd, Emilia Tuunainend, Anni Spangard, Piia von Lodee, Kaisu Rantakokko-Jalavaf, Gisela Ottob, Stefan Schedingc,g, Tero Soukkad, Saara Wittfoothd and Thomas Laurella* a
Department of Biomedical Engineering, Lund University, Box 118, SE-221 00 Lund, Sweden
b
Department of Infectious Diseases, Skåne University Hospital, Lund, Sweden
c
Stem Cell Center, Lund University, BMC B10, Klinikgatan 24, Lund, Sweden
d
Department of Biochemistry / Biotechnology, University of Turku, Tykistökatu 6 A, FI-20520 Turku, Finland
e
Abacus Diagnostica Oy, Tykistökatu 4 D, FI-20520 Turku, Finland
f
Clinical microbiology, Turku University Hospital, P.O. Box 52, FI-20521 Turku, Finland
g
Department of Haematology, Skåne University Hospital, Lund, Sweden
ABSTRACT: This paper describes an integrated microsystem for rapid separation, enrichment and detection of bacteria from blood, addressing the unmet clinical need for rapid sepsis diagnostics. The blood sample is first processed in an acoustophoresis chip, where red blood cells are focused to the centre of the channel by an acoustic standing wave and sequentially removed. The bacteria-containing plasma proceeds to a glass capillary with a localized acoustic standing wave field where the bacteria are trapped onto suspended polystyrene particles. The trapped bacteria are subsequently washed while held in the acoustic trap and released into a polymer microchip containing dried PCR reagents, followed by thermocycling for target sequence amplification. The entire process is completed in less than 2 hours. Testing with Pseudomonas putida spiked into whole blood revealed a detection limit of 1 000 bacteria/ml for this first generation analysis system. In samples from septic patients, the system was able to detect Escherichia coli in half of the cases identified by blood culture. This indicates that the current system detects bacteria in patient samples in the upper part of the of clinically relevant bacteria concentration range and that a further developed acoustic sample preparation system may open the door for a new and faster automated method to diagnose sepsis. Sepsis is a systemic inflammatory response to infection, most often caused by bacteria. It is a major health problem with an annual incidence of roughly 300 cases per 100 000 people in developed countries and the number of cases is increasing1-3. Sepsis is associated with high morbidity, being one of the most common reasons for admission to intensive care units in Europe and the United States. Furthermore, it is a common cause of death with a mortality rate that reaches up to 30% 3-6. It is well established that adequate antibiotic therapy of sepsis patients administered as early as possible reduces mortality. Every hour of delay of adequate therapy is associated with a 7.6% decrease in survival during the first 6 hours in patients with septic shock7. Still, detection and identification of the causing pathogen from blood samples takes typically 29-130 h using the current standard methods based on blood culture (figure 1). In these methods, culturing of blood samples in enriched broth is followed by Gram staining and sub culturing, polymerase chain reaction (PCR)8 or matrix-assisted laser desorption ionisation time-of-flight mass spectrometry (MALDI TOF MS) for identification and antibiotic susceptibility testing. The sensitivity of these methods is limited by the sensitivity of blood culture, which can be poor if the responsible organism is slow-growing, fastidious or if the samples are taken after the start of antimicrobial therapy9. While waiting for the identification of the causing organism broadspectrum antibiotics are used. This treatment does not cover the responsible organism in 20-34% of the sepsis cases10,11. Additionally, indiscriminate use of broad-spectrum antibiotics promotes antibiotics resistance. There is an urgent need for early detection of
pathogens that leads to faster and more targeted antimicrobial therapy, resulting in improved clinical outcome for patients with sepsis. The most promising path towards faster sepsis diagnostics is PCRbased methods, where bacterial DNA is extracted directly from whole blood without culture12. A sepsis patient can have bacteria levels as low as 1-100 colony forming units (cfu) per ml whole blood or even lower13-15. To handle these extremely low bacteria concentrations, the bacteria first need to be enriched from the sample. Since almost half of the blood volume consists of red blood cells, this entails removal of the vast majority of them. Other blood components, such as plasma proteins, white blood cells and added anticoagulants also have to be removed not to inhibit the PCR16-18. In most commercially available PCR sepsis diagnosis methods this is done by lysis and DNA extraction. These assays, however, suffer from complex sample preparation, lack of integration of the different processing steps and are vulnerable to contamination. An alternative to this approach would be to separate the bacteria from the red and white blood cells. There have been a number of attempts to achieve this using various microfluidic techniques, including dielectrophoresis, inertial effects, cell margination, surface acoustic waves, affinity-based extraction and lysis19-27. Generally these methods are limited by either sample volume, throughput, bacteria recovery, efficiency in removal of cells or lysis debris or their need for specific affinity-capture of the pathogens.
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
Page 2 of 8
SEPTIFAST System prep. Sample lysis
Sample extraction
30 min
90 min
Result 150 min
30 min
Total time: 5-6 h Pipetting steps: 24 Manual handling steps: 100
BLOOD CULTURE Aerobic culture in bottle
Gram staining & microscopy 20 min
~ 5 - 130 h (avg. 15 h)
Sample prep. and MALDI-TOF
Anaerobic culture in bottle
60 min Plate culture
Result
Total time: ~29 - 130 h Pipettings steps: 14 Manual handling steps: ~ 8 Mostly automated system ~130 h until negative
~ 5 - 130 h (avg. 15 h) ~24 h
ACUSEP Acoustic sample preparation 30 min
75 min
Total time: < 2 h Pipetting steps: 1 Manual handling steps: 5 Mostly automated system
Figure 1: Comparative process timelines showing the considerable difference in time to result between Septifast (a commercial PCR-kit from Roche), blood culture (current gold standard) and the ACUSEP system presented in this article. Process times, pipetting steps and manual handling steps are from the manufacturer’s manuals, current laboratory praxis at Skåne University Hospital and from Biondi et al.28.
separated from red blood cells by acoustophoresis29-31 and enriched by acoustic trapping32-35. We have now adapted and integrated these two methods to form the ACUSEP system, a microfluidic sample preparation system delivering separated and enriched bacteria to be detected in a dry-reagent PCR-chip36-39. This approach has the advantage of being faster than existing PCR-based methods and blood culture, and it also reduces the number of manual sample handling and pipetting steps that are needed compared to commercial PCR-kits (figure 1). The ACUSEP system was evaluated using spiked whole blood from healthy volunteers and samples from septic patients. The results indicate that the current system can detect bacteria in patient samples in the upper range of clinically relevant bacteria concentrations. An acoustic sample preparation system as described herein may thus open the door for the development of a new faster and more automatable method to diagnose sepsis 7.
EXPERIMENTAL Integrated acoustic separation and enrichment. The acoustic sample preparation system was designed to separate and enrich bacteria from blood and deliver them to a dry-reagent PCR chip for detection and identification (figure 2). The blood sample was first processed in an acoustophoresis chip, removing the red blood cells. The remaining bacteria-containing plasma continued into a glass capillary where the bacteria were enriched by acoustic seed trapping34. The trapped bacteria were subsequently washed, thereafter the acoustic trapping was deactivated and the trapped bacteria released into a polymer chip containing dried PCR reagents. The disposable PCR chip was finally sealed and processed in a real-time PCR machine (GenomEra prototype from Abacus Diagnostica, Finland).
and the trapping capillary. Sample and buffer flows were controlled by seven syringe pumps (neMESYS, Cetoni GmbH, Korbussen, Germany). The four waste syringes connected to the separation chip outlets were glass syringes (5 ml 1005 TLL, Hamilton Bonaduz AG, Bonaduz, GR, Switzerland). Blood samples, sterilization liquids and KCl solution were dispensed from sterile plastic syringes to prevent cross contamination between samples. The whole system was controlled through in-house developed LabVIEW-software, minimizing manual intervention and variability. The program controlled all the seven syringe pumps, four threeway valves, two four-port switch valves, two waveform generators and a frequency tracking unit for the acoustic trapping40. Each part of the integrated system and the sample processing protocol are further described in the following sections. Acoustic separation of bacteria from red blood cells. Acoustophoresis is a label-free separation technique where particles exposed to an acoustic standing wave field are affected by an acoustic radiation force. A microchannel having a width matched to a half wavelength of the actuation frequency displays a single pressure node along the centreline of the channel. The radiation force on a particle in the standing wave field is directed either towards the pressure node or to the pressure antinode, depending on the density and compressibility of the particles and the surrounding medium. The movement of particles induced by the acoustic radiation force is dependent on the size of the particle, the acoustic pressure amplitude and the frequency of the sound wave41,42. The velocity of particles in the acoustic standing wave field scales with the radius squared, which allows for size-based separation of particles even if they are moving towards the same pressure node43-45 .
The various parts of the system were connected using Teflon tubing. External, electrically actuated valves (Cheminert C2, Valco Instruments Company Inc., Houston, TX, USA) were used to enable switching of different liquids during the sample preparation, one before the separation chip and one between the separation chip
ACS Paragon Plus Environment
Page 3 of 8
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 2: Bacteria were separated from red blood cells by acoustic separation, enriched from the remaining plasma by acoustic trapping and finally released to dry-reagent PCR chips for detection and identification. The flows were controlled by seven syringe pumps, four three-way valves and two four-port switch valves.
As a first step in the sample preparation, blood cells were separated from the bacteria-containing plasma using acoustophoresis in a microfluidic chip. The acoustophoresis channel was designed in a meandering fashion and consisted of one inlet, four blood cell waste outlets and one outlet for the bacteria-containing plasma. Red blood cells entering the channel were acoustically focused to the pressure node along the centreline of the channel and sequentially removed through three bottom outlets in the centre of the channel and one flow splitter at the end of the channel, while the bacteria, which are smaller than red blood cells and thus were less affected by the radiation force, remained in the plasma fraction along the side walls. The microchannel structure was fabricated in silicon by double sided photolithography and anisotropic potassium hydroxide wet-etching and sealed with a glass lid using anodic bonding. The design and fabrication of the acoustophoresis chip has previously been described in more detail46. A 2 MHz piezoceramic transducer (PZT26, Ferroperm piezoceramics, Denmark) was glued onto the silicon underneath the channel and was actuated using a waveform generator (Agilent 33220A, Agilent Technologies Inc., Santa Clara, CA, USA) coupled to an amplifier (75A250A Amplifier Research, Souderton, PA, USA). The voltage over the transducer was monitored using a digital oscilloscope (TDS 1002, Tektronix UK Ltd., Bracknell, UK). A custom-built aluminium and poly(methyl methacrylate) (PMMA) chip holder was used to attach fluidic connections and a Peltier element together with a heat sink and a fan to the chip. The temperature of the chip was measured using a Pt100 sensor glued to the piezoceramic transducer. To maintain the acoustic resonance, the temperature of the chip was regulated to 28°C by a PID Peltier-controller (TC0806-RS232 CoolTronic GmbH, Beinwil am See, Switzerland). Acoustic trapping of bacteria from plasma. Acoustic trapping can be used for capturing and positioning of particles and cells in lab-on-a-chip systems33,34,47. By attaching a small piezoceramic transducer to a glass capillary, a local ultrasonic standing wave field can be generated. This means that the sound intensity will be higher close to the transducer than elsewhere along the capillary, creating a large pressure and velocity amplitude gradient in the length direction of the capillary where cells or particles can be trapped47. Particles that are too small to be trapped directly by the primary radiation force can instead be trapped onto larger seed particles due to secondary acoustic forces48 that emerge as sound waves are
scattered between particles in close proximity34, see figure 3. Acoustic capture and enrichment of bacteria has been previously demonstrated and shown to have recovery rates of 95±3% using the seed particle assisted trapping technique developed by our group34. In this paper acoustic seed particle assisted trapping was used to capture bacteria from the acoustophoresis generated blood plasma and hold the captured bacteria while PCR inhibiting plasma components were washed away. Trapping was performed in a borosilicate glass capillary with a cross section of 0.2 x 2 mm2 (Vitrotube 3520, Vitrocom, Mountain Lakes, NJ, USA). The capillary was clamped by a plastic holder to a miniaturized piezoceramic transducer that was soldered onto a printed circuit board and driven by a function generator47. To enable acoustic trapping of the small bacteria, the capillary was initially loaded with polystyrene seed particles (12 μm diameter polystyrene calibration beads, Fluka Chemie AG, Switzerland). Although the exact amount of seed particles in the trap may vary due to sedimentation during particle loading, no effect on the trapping recovery from the variation has been observed. Previous experiments with differently sized seed particles have also shown that the size is not critical for the seed trapping performance. The capillary half wavelength resonance frequency was tracked and controlled by a custom built tracking software and hardware to maintain an optimal resonance frequency in the trap, regardless of size of the trapped cluster, temperature changes or changes of fluids with different acoustic properties in the capillary40. PCR-based bacteria detection. The detection and identification of the trapped bacteria was performed in dry-reagent PCR chips. Real-time PCR assays were developed for detection of Pseudomonas spp., E. coli, Staphylococcus aureus and Streptococcus pneumonia in separate chips. Amplification of DNA was monitored with switchable lanthanide luminescence label technology49,50 by measuring fluorescence in time-resolved mode every second amplification cycle. The PCR assay reagents were preloaded and dried in plastic GenomEra chips (Abacus Diagnostica, Finland)51,52. Only the sample had to be added to the chip after which the chip was closed and placed in a GenomEra instrument for thermal cycling and signal measurement. Details of the Pseudomonas spp. assay are given in the Supplementary Information while the design of the E. coli, Staphylococcus aureus and Streptococcus pneumonia assays is described elsewhere53.
ACS Paragon Plus Environment
Analytical Chemistry
Page 4 of 8
-
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
Figure 3: Seed particle-enabled trapping of bacteria. In step 1, large seed particles (12 μm polystyrene) are trapped in the standing wave. In step 2, secondary acoustic forces attract and trap bacteria onto the large polystyrene particles. After a finished sample run the ultrasound is turned off and the cluster of seed particles and enriched bacteria is released to a dry-reagent PCR chip as step 3.
Acoustic separation test. The red blood cell removal efficiency and bacteria recovery of the acoustic separation was tested separately. E. coli were grown 2.5 h at 37°C without IPTG, then 16 h at 30°C with IPTG and finally centrifuged and resuspended in isotonic saline. (Further details for all bacteria culture can be found in the Supplementary Information.) Whole blood was collected in citrate tubes (BD Vacutainer) and the blood cells were washed by centrifugation. The cells were resuspended in the same volume of isotonic saline containing bovine serum albumin (0.1 kg/L), methyl α-Dmannopyranoside (25 g/L) and E. coli (final concentration 3 x 108 bacteria/ml). The sample was injected into the separation chip at 80 μl/min and focused blood cells were removed through the blood outlets at 20, 20, 15 and 15 μl/min respectively using a syringe pump (neMESYS, Cetoni GmbH, Korbussen, Germany). The remaining plasma was collected through the plasma outlet and the concentration of bacteria and red blood cells was compared to the sample entering the system using flow cytometry (BD FACSCanto II, BD Biosciences, San Jose, CA, USA). The chip, without holder and temperature control, was actuated at 1.96 MHz with a peak-topeak amplitude of 6.6 V. Bacteria enrichment test. To test the acoustic bacteria enrichment, eGFP-producing E. coli were suspended at a high concentration (22.2 x 106 bacteria/ml) in isotonic saline (9 g/l). The bacteria suspension was perfused at 10 μl/min over a cluster of 12 μm polystyrene particles that were held in an acoustic trap using a 4 MHz standing wave. Dry-reagent PCR chip test. The Pseudomonas spp. chips were tested in real-time PCR using 10, 100, 1 000 and 10 000 P. putida bacteria in 35 μl 100 mM KCl per reaction. The S. aureus, S. pneumonia and E.coli PCR assays have been tested previously, showing detection limits of 3 cfu, 2 cfu and 400 cfu per reaction, respectively53. Their specificity has also been verified using 21 different sepsis causing bacteria species53. System sterilization test. The sterilization protocol for the acoustic sample preparation system was tested by running the sterilization protocol (see Supporting Information) and then collecting sterilization samples of KCl solution (100 mM) from the trapping capillary. To check for DNA contamination, 35 μl of the sterilization sample was analysed by PCR in chips for P. putida detection and compared to sterile KCl solution. Sterilization samples were also spiked with 1 000 copies of P. putida DNA and compared to sterile KCl solution spiked the same way, to check if any sterilization compounds that would inhibit the PCR remained in the system. System testing with spiked blood samples. The sensitivity of the ACUSEP system was tested using citrated blood from healthy volunteers spiked with known amounts of P. putida bacteria. Ve-
nous blood samples from healthy volunteers were collected in citrate tubes (BD Vacutainer, 90% whole blood and 10% buffered trisodium citrate solution after filling). Samples were spiked with known concentrations of P. putida and sterile isotone sodium chloride was added to dilute samples to 70% citrated blood (corresponding to 63% whole blood). To keep the bacteria concentration constant, sodium azide was added to a final concentration of 0.05% to avoid bacterial growth. A volume of 1.43 ml of each blood sample (corresponding to 1 ml citrated whole blood) was processed in the ACUSEP system. The extracted bacteria were released into a PCR chip for Pseudomonas spp. detection and analysed in the PCR instrument. This was done with triplicate samples for 10, 100, 1 000 and 10 000 bacteria/ml citrated blood and seven unspiked controls. Further details of the sample processing protocol are given in the Supplementary Information. System testing with clinical samples. From January through March 2014, seventy-one patients were enrolled in a prospective study comparing the ACUSEP system to blood culture (BacT/Alert) at the Clinic of Haematology and the Clinic of Infectious Diseases at Skåne University Hospital in Lund, Sweden. The study protocol was approved by the local ethics committee (decision number 2013/300) and informed consent was obtained from all patients. The inclusion criteria were neutropenic fever (38°C temperature) or suspected sepsis due to urinary tract infection, pneumonia or severe sepsis. Suspected sepsis was defined according to the American College of Chest Physicians/ Society of Critical Care Medicine (ACCP/SCCM) Consensus Conference as an infection and the presence of at least two of the systemic inflammatory response syndrome (SIRS) criteria: temperature >38°C or 90 beats/minute, respiratory rate >20 breaths/min, or white blood cell count (WBC) >12 000 or
