A 3D-printed hand-powered centrifuge for molecular biology

The centrifuge is an essential tool for many aspects of research and medical diagnostics. However, conventional centrifuges are often inaccessible outside of standard laboratory settings, such as remote field sites, because they require a constant external power source and can be prohibitively costly in resource-limited settings and Science, technology, engineering, and mathematics (STEM)-focused programs. Here we present the 3D-Fuge, a 3D-printed hand-powered centrifuge, as a novel alternative to standard benchtop centrifuges. Based on the design principles of a paper-based centrifuge, this 3D-printed instrument increases the volume capacity to 2 mL and can reach hand-powered centrifugation speeds up to 6,000 rpm. The 3D-Fuge devices presented here are capable of centrifugation of a wide variety of different solutions such as spinning down samples for biomarker applications and performing nucleotide extractions as part of a portable molecular lab setup. We introduce the design and proof-of-principle trials that demonstrate the utility of low-cost 3D-printed centrifuges for use in remote field biology and educational settings.

The chromoprotein plasmid construct (Scrooge Orange) was purchased from the corporation ATUM Bio (ATUM, 2018) to test in the competent E. coli cells. The samples obtained from ATUM Bio were hydrated in 10 µL of ddH20, and transformed into New England Biolabs * authors contributed equally † Please address correspondence to M.S.B: (saadb@chbe.gatech.edu) DH5a E. coli using the NEB High Efficiency Transformation Protocol. The resulting transformation solutions were plated onto LB agar containing the antibiotic carbenicillin. Single colonies from the sample plate were diluted into 40 µL of water, and 1 µL of the colony dilution was inoculated into three different 5 mL LB broth liquid cultures with 5 µL of carbenicillin (100 µg/ml). These were induced with IPTG at concentrations of 0 M, 100 µM and 1 mM respectively. Liquid cultures grown for 24 hours in an 37 • Celsius incubator set to shake at 170 rpm, and then removed from the incubator and allowed to sit undisturbed for 1 hour. 75 µL of the settled particulate was transferred into a 0.2 mL PCR tube and centrifuged for 5 minutes using the 3D-Fuge. The supernatant was discarded and the bacterial pellets were placed into the sample illumination chamber for measurements. An image captured with a phone and an RGB color-analysis tool (https://imagecolorpicker.com) were utilized to determine the hue values of each bacterial pellet [1]. These were then compared with the concentration of IPTG induction for analysis. As the color analysis tool uses a single pixel to determine the hue value, the pixel selected was representative of chromoprotein expression wherein the region within the bacterial pellet containing visible color expression (in the case of induced samples) contained the selected pixel.

B. 3D-Fuge Design and Materials
In Case Study 1, one male individual (age 29) utilized a 3D-Fuge capable of holding microcentrifuge tubes as well as a combination of spin columns and flow-through tubes. In Case Study 2, four identical 3D-Fuges were utilized for the centrifugation of the chromoprotein samples. All devices were printed using a custom-built 3D Printer and CAD files modified for the inclusion of tube holders to ensure PCR tubes are securely placed in the device.
The 3D-Fuge design files can be found on Github https://github.com/bhamla-lab/ 3D-fuge-PlOS-Biology-2019 and example models with dimensions can be found in SI   The string used in this study was the Dorisea Extreme Braid 500 lb 2.0 mm Fishing Line, purchasing 300m at a cost of approximately $40 as of March 2018. String lengths were maintained at a standard 104 cm, and 3D-Fuges were coiled 50 times prior to centrifugation. We also note that fishing-line string breaks can occur during centrifugation and insist the use of safety goggles when using the 3D-Fuge, as the device and components can reach high speeds. For Design 2, tape can additionally be used to secure the PCR tubes in place along with the existing latch for added protection.

C. Nucleotide Extractions
The 3-D centrifuge was used to perform DNA extractions using the Quick-DNA Miniprep Plus kit (Zymo Research) according to the protocol. For DNA extraction comparisons in the lab, a benchtop centrifuge (Eppendorf model 5414 D) was used and each centrifugation step was performed at 13,000 rpm for 1 minute. Solid tissues collected in the field were lysed and added to a microcentrifuge tube with 95 µL of water, 95 µL of solid tissue buffer, and 10 µL of proteinase K for 1-3 hours. 400 µL of Genomic Binding Buffer was added to the supernatant and transferred to a Zymo-spin Column in a collection tube. The tubes were placed into the 3D-Fuge and spun by hand for approximately one to two minutes. The flow through was discarded, then 400 µL of DNA pre-wash buffer was added and centrifuged using the 3D-Fuge. Two rounds of washing steps were performed using gDNA Wash Buffer with 3D-Fuge centrifugation steps for one minute each followed by discarding of the flow through. 30 µL of DNA elution buffer was added and a final centrifugation step was carried out in a clean 1.5 µL tube. The final purified DNA was used for downstream molecular experiments, including polymerase chain reaction. Primers for long-range PCR were designed to amplify mitochondrial DNA (Fig 1F) [2] or ribosomal DNA [3]. Amplicons were pooled and sequenced on the Min-ION platform (Oxford Nanopore Technologies) [3]. The sequencing data can be found on GitHub.

D. High-Speed Video Analysis
The angular velocity of the 3D-Fuge was determined through high-speed video tracking. A marker was placed on the centrifuge body, and was spun at maximum force by user which was captured using high-speed video at 1,057 frames per second (fps). Tracker (software) was used to track the marker through the high speed video frame by frame [4]. The angular velocity of 5 full cycles were tracked for both designs. The raw angular velocity data was smoothed using a LOESS function with a factor 0.02, and RPM and RCF were plotted in Figures S4 and  S5. This is a second trial of the data shown in the main manuscript in Figure 1K, which is a representative of this data from a different trial. Average RPM and RCF is given as the square root mean of the peak rpm. Code for high-speed analysis can be found on https://github. com/bhamla-lab/3D-fuge-PlOS-Biology-2019. with their respective Nanodrop DNA quantifications. Long-range mitochondrial PCR products using these extracts can be found in Figure 1E. (C) Gel electrophoresis of samples that were extracted in the field using the 3D-Fuge and subsequently PCR amplified with ribosomal DNA primers (left to right: Solanaceae, Heliconius butterfly, Mosquito, and Butterfly eggs). (D) NCBI distance of tree results from a consensus sequence generated in the field from the bloodfed mosquito sample.