Three-dimensional printing physiology laboratory technology

Matthew S. Sulkin, Emily Widder, Connie Shao, Katherine M. Holzem, Christopher Gloschat,Sarah R. Gutbrod, and Igor R. EfimovDepartment of Biomedical Engineering, Washington University in Saint Louis, Saint Louis, Missouri

Submitted 9 August 2013; accepted in final form 13 September 2013

Sulkin MS, Widder E, Shao C, Holzem KM, Gloschat C, GutbrodSR, Efimov IR. Three-dimensional printing physiology laboratory tech-nology. Am J Physiol Heart Circ Physiol 305: H1569–H1573, 2013.First published September 16, 2013; doi:10.1152/ajpheart.00599.2013.—Since its inception in 19th-century Germany, the physiology labora-tory has been a complex and expensive research enterprise involvingexperts in various fields of science and engineering. Physiologyresearch has been critically dependent on cutting-edge technologicalsupport of mechanical, electrical, optical, and more recently computerengineers. Evolution of modern experimental equipment is con-strained by lack of direct communication between the physiologicalcommunity and industry producing this equipment. Fortunately, re-cent advances in open source technologies, including three-dimensional printing, open source hardware and software, present anexciting opportunity to bring the design and development of researchinstrumentation to the end user, i.e., life scientists. Here we provide anoverview on how to develop customized, cost-effective experimentalequipment for physiology laboratories.

heart physiology; 3-D printing; open source manufacturing; opticalmapping

IN THE 19TH CENTURY, Carl Ludwig’s Institute of Physiology atthe University of Leipzig made more fundamental and lastingdiscoveries than any previous or current physiological labora-tory (10). Throughout his career, Ludwig tried to cast physi-ology as an experimental hypothesis-driven science based onthe principles of chemistry, physics, and engineering, rather thanan empirical descriptive field (3, 10). Ludwig’s exceptional im-pact is evident from his numerous disciples who are the foundersof national physiology traditions in many countries on severalcontinents.

One of his valuable contributions to life sciences was hisrecognition of the importance of engineering and technology insupporting a physiological inquiry. When Ludwig joined theLeipzig faculty in 1865, he was given the responsibility todesign an institute devoted to teaching and researching physi-ology. The institute he developed had specialized facilitiesdedicated to science (chemistry, histology, and experimentalphysiology) and to engineering (machine, electric, and glass-blowing workshops), which promoted different disciplines tounite in the elucidation of organ system functions (3). This wasin stark contrast to other world-renowned physiologists, whoperformed their studies isolated in one- or two-room laborato-ries (3).

Today’s most productive physiology and biomedical insti-tutions require not only expertise in the basic sciences but alsoadroit computer scientists to automate experiments, electricalengineers to build analog and digital hardware, and machinists

to manufacture experimental setups. Individual laboratories areoften ill equipped to overcome such diverse skill sets andconsequently purchase equipment from industry at a high cost.Open source software and hardware have had an importantimpact on diminishing expenses, and here we highlight opensource manufacturing as a new way to decrease cost andimprove a laboratory’s productivity.

Open source additive manufacturing uses three-dimensional(3-D) printing to create an object with high precision at micronresolution based on an open source digital design. Using digital3-D designs gives researchers the flexibility to make quickcustom modifications to previously designed parts, which canbe shared in the public domain (11). Currently, our laboratoryis using 3-D printing to improve the quality of experiments ona wide range of ex vivo perfused heart preparations fromdifferent species, including mouse, rabbit, canine, and human.We have designed tissue chambers, electrode holders, tissuerestrainers, and optical components to optimize all aspects ofan experiment: physiology, pharmacology, and measurements(electrical, mechanical, optical, and chemical). Thus we are nolonger dependent on industry when requiring highly complex,low-volume parts. Here, we give an overview on how to gofrom a conceptual design to a 3-D-printed object and presentseveral printed components that have benefited our researchlaboratory.

METHODS

Design, review, and print. The first step in creating custom phys-iology laboratory equipment is turning an idea from a paper sketch intoa digital design, which is illustrated here by a screw thread in Fig. 1.Digital designs can be made in any number of free computer-aideddesign programs, including SketchUp, Blender, and OpenSCAD. Forthose who are inexperienced with computer-aided design and areinterested in creating their own designs, we refer you to the followingWeb sites: sites.google.com/site/3dprintingphysiolabefimov/ or http://efimov.wustl.edu/, where we provide links to tutorials. We recom-mend SketchUp for beginners as it takes less than 10 min to learn thebasic functions.

After the digital design has been finalized, it needs to be exportedas an stereolithography (STL) file format and visualized with an STLviewer. An STL file approximates a 3-D geometry as a collection oftriangles by storing a list of both the xyz vertices and unit normal ofevery triangle. Errors can occur converting to STL, and thus it isprudent to examine objects with an STL viewer (Netfabb StudioBasic) that specializes in detecting and repairing damages in thedigital mesh. Once the mesh is confirmed to have no damage, the STLfile can be sent to a commercial 3-D printing company, a purchased3-D printer, or a personally constructed 3-D printer such as opensource Fab@Home (7) or RepRap (4). For further information on thisprocess, including all our digital designs, software, and troubleshoot-ing tips, please refer to the above Web sites.

Our physiology equipment. We created our digital designs inAutoCAD ($99.00), Google SketchUp 7, 8, 2013 (free), or GoogleSketchUp Pro (various academic licenses available for free or at

Address for reprint requests and other correspondence: I. R. Efimov, Dept.of Biomedical Engineering, Washington Univ. in St. Louis, One BrookingsDr., St. Louis, MO 63130-4899 (e-mail: igor@wustl.edu).

Am J Physiol Heart Circ Physiol 305: H1569–H1573, 2013.First published September 16, 2013; doi:10.1152/ajpheart.00599.2013. Perspectives

0363-6135/13 Copyright © 2013 the American Physiological Societyhttp://www.ajpheart.org H1569

reduced pricing). Digital designs were viewed in Netfabb Studio Basic(free; Netfabb, Parsberg, Germany) to ensure printer-ready meshes.To date, we have printed numerous customized tissue chambers(Fig. 2A), electrode holders (Fig. 2B), tissue restrainers (Fig. 2C), andmany other experimental laboratory components (i.e., luer connectors,bubble traps, electronic holders).

The mouse chamber, electrode holders, and tissue restrainer wereprinted with Objet 24 printer using VeroWhitePlus material (Strata-sys, Eden Praire, MN). The rabbit torso chamber (Fig. 4) was printedwith Projet 3000 printer and EX2000 material (Engineering andManufacturing Services, Tampa, FL). The CLARITY electrophoretictissue clearing (ETC) chambers (Fig. 5) were printed on a StratasysDimension printer using acrylonitrile butadiene styrene (ABS) plastic(Print to 3-D, Tunkhannock, PA), with selective laser sintering usingnylon 12 material (ZoomRP, Valencia, CA), or on a MakerGear M2printer using ABS plastic (Beachwood, OH). Additional pieces, suchas optical glass, conductive material, O-rings, and Sylgard were addedpostprinting. Sylgard is a silicone elastomer that is deposited indesigned slits of tissue restrainers for tissue pinning.

Tissue preparation. To demonstrate the feasibility of using 3-Dprinted objects in the physiology laboratory, we show experimentalresults from a mouse heart preparation, which was approved by theInstitutional Animal Care and Use Committee of Washington Univer-sity in St. Louis. A single explanted mouse heart was obtained froma C57BL/6 mouse. The mouse was given an intraperitoneal injectionof 100 units of heparin followed by an anesthetic mixture of ketamine-xylazine (ketamine, 80 mg/kg; and xylazine, 10 mg/kg). The heart wasrapidly excised through a midsternal incision and placed on a Lan-gendorff apparatus where it was retrogradely perfused with oxygen-ated (95% O2-5% CO2) Tyrode solution, consisting of (in mM) 128.2NaCl, 1.3 CaCl, 4.7 KCl, 1.05 MgCl2, 1.19 NaH2PO4, 20.0 NaHCO3,and 11.1 glucose. Tissue was perfused under constant pressure(60–80 mmHg) with perfusate pH continuously monitored (OaktonInstruments, Vernon Hills, IL). Perfusate temperature was maintainedat 37 � 1°C using a heated recirculating bath (NESLAB EX7, ThermoScientific). Preplaced Ag/AgCl pellet electrodes (WPI, Sarasota, FL)monitored (PowerLab 26T; ADInstruments, Colorado Springs, CO)far-field ECG throughout experimentation.

Optical mapping and data analysis. Optical mapping of transmem-brane potential (Vm) and calcium transients (CaT) was done aspreviously described (5, 9). Briefly, tissue was stained with voltage-sensitive dye (RH-237; 5 �l in 1 mg/ml DMSO; Life Technologies,Carlsbad, CA) and immobilized by addition of blebbistatin (10 �M;Tocris Bioscience, Ellisville, MO) to reduce motion artifact. Admin-istration of calcium indicator rhod 2-AM (Life Technologies) enabledimaging of CaT. Optical mapping data were recorded from a MiCAMULTIMA (SciMedia, Costa Mesa, CA) imaging system at high-spatial(100 � 100 pixel) and -temporal (1,000 frames/s) resolution. Opticalsignals were processed using a freely available MATLAB programdescribed in Laughner et al. (6). All optical data were filtered using a3 � 3 pixel spatial filter, a 0–100-Hz finite impulse response filter,and normalized. Activation times were defined as the maximum firstderivative of the fluorescence signal upstroke.

Results and Discussion

To improve the quality of experimentation, we designed andprinted species-specific equipment. Figure 2A shows the digitaldesign and printed chamber for a vertical Langendorff-per-fused mouse heart, where optical imaging is done through theside glass window. Unlike imaging a horizontally orientedheart from the surface (5), optical mapping of a verticallysubmerged preparation reduces artifact of the rippling fluidcreated by a superfusion pump (9). Figure 2B displays twoexamples of electrode holders that slide over the columns ofthe tissue restrainer (Fig. 2C) and are secured to the restrainerby screws. After tissue dissection and aortic cannulation, thepreplaced stimulating electrodes are easily positioned onto theright atrial appendage and apex. Previously, we had to te-diously place stimulating electrodes onto the surface of theheart using micromanipulators, delaying experimentation.

Figure 2C shows an example tissue restrainer that keeps allhearts in the same orientation and slides into the mouse tissuechamber. Consistent orientation between experiments is vitalfor calculating electrical intervals from a far-field ECG that isdependent on heart-lead orientation. In contrast to glass tissuechambers, the printed mouse chamber does not require a heatedwater jacket to maintain temperature because of the material’slow thermal conductivity. Overall, this setup reduces the timeof tissue preparation by 30 min and thus enables the comple-tion of longer physiological protocols. Furthermore, there is astriking volumetric difference between the printed mousechamber (80 ml; Fig. 2A) and our previous commerciallyavailable chamber (200 ml, No. 166060, Radnoti, Monrovia,CA). Consequently, the quantity of both dyes and pharmaco-logical agents per experiment can be reduced. A cost analysisof the 3-D printed equipment versus industry-purchased equip-ment is summarized in Fig. 2D. The overall savings of printingequipment is extraordinary: �$3,200. Two printed electrodeholders have a material cost of $0.14, and the addition ofplatinum-iridium wire (No. 778000, A-M systems, SequimWA) postprinting is $13.34. By 3-D printing our customdesigns, we can avoid the necessity of using expensive equip-ment such as micromanipulators, and the savings is enough topurchase the newest generation 3-D printer.

Figure 3 demonstrates the high-quality data collected fromthe 3-D printed mouse setup. Vm and CaT were simultaneouslyimaged on the posterior surface of the ex vivo Langendorffheart preparation. When the preparation was being paced fromthe apex, Vm spread across the epicardial surface in 13 ms (Fig.3B). Figure 3C displays representative Vm (black) and CaT(teal) signals during S1–S1 (180 ms) pacing with Vm preceding

Fig. 1. A digital design of a screw thread converted to a stereolithography (STL) file and printed. An STL file stores the three-dimensional (3-D) geometry asa collection coordinates (x,y,z) and surface normals (n).

Perspectives

H1570 PRINTING A PHYSIOLOGY LABORATORY

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00599.2013 • www.ajpheart.org

CaT. Ventricular fibrillation was initiated with S1–S2 (140–55ms) pacing and uncouples Vm from CaT.

For a specific set of experiments, our laboratory created atissue chamber in the shape of a human torso derived from acomputed tomography scan of a patient and scaled proportion-ally to a rabbit heart (Fig. 4). Twelve holders were made in thedigital design to accommodate electrodes in the configurationof a standard 12-lead ECG. Collecting a 12-lead ECG is rare inex vivo experimentation due to lengthy instrumentation setupand low signal-to-noise recordings. Multiple leads are impor-tant because they mitigate spurious findings of a single leadand are capable of identifying anatomical structures of cardio-myopathy. Having stable, high signal-to-noise 12-lead ECGsmakes signal analysis straightforward and identification ofdisease reliable.

Fig. 2. Digital designs and 3-D printed objects for physiology experiments.A: mouse tissue chamber. B: example electrode holders encapsulating plati-num-iridium wire. C: mouse heart restrainer. D: cost analysis between printedand industry bought products. Black asterisk, cannula; pink asterisk, atrialpacing electrode; teal asterisk, ventricular pacing electrode. *Radnoti TissueCamber, No. 166060, 200 ml; **Harvard Apparatus, Coaxial StimulationElectrode, No. 730219; ***Tritech Research, Narishige Magnet-Mount ThreeAxis Coarse Micromanipulator, MB-PP2. NA, not applicable.

Fig. 3. Physiological experiment of a Langendorff-perfused mouse heart.A: posterior surface of mouse heart, where white dotted line and black/tealsquare indicate location of activation map (B) and optical signals (C), respec-tively. B: activation map of the transmembrane potential across the epicardialsurface of mouse heart. C: representative transmembrane potential (Vm) andcalcium transient (CaT) signals during pacing (top) and induction of ventric-ular fibrillation (bottom). White square pulse, location of pacing electrode.

Perspectives

H1571PRINTING A PHYSIOLOGY LABORATORY

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00599.2013 • www.ajpheart.org

In addition to improving the quality of experimental data anddecreasing cost, using digital designs has given our laboratorythe flexibility to make quick modifications to experimentalsetups without assistance from industry. In Fig. 5, we demon-strate the prototyping process for custom-designed laboratoryequipment. Figure 5, A–E, are various iterations of an ETCchamber designed for CLARITY, which is a method to rendertissue optically transparent while preserving 3-D molecularstructure (2). This is achieved by first transforming the tissueinto a hydrogel-hybridized form and then subjecting it to theETC process in a detergent buffer. Figure 5A is the originalchamber constructed based on the CLARITY protocol, withoutassistance from 3-D printing, whereas Fig. 5, B–E, are printedchambers (top) and their digital designs (bottom). Given thedetergent buffer and conditions for ETC, the flexibility tocontinually redesign, prototype, and test new chambers hasbeen critical to developing a stable, leak-resistant chamber forthis process.

Although open source manufacturing offers many advan-tages, progress in 3-D printing technology is essential for opensource manufacturing to be used to create fully functionalexperimental setups. Currently, the materials available for

printing physiological setups are somewhat limited. Two plas-tics used by our laboratory with inexpensive 3-D printers (cost,$400–2,500) include ABS (�$35.00/kg) and polylactic acid(�$35.00/kg). Both plastics have melting points well beyondthe physiological range (�100°C), making them ideal forphysiology equipment. When compared with polylactic acid,ABS has improved mechanical strength and flexibility but ismore prone to warping during the printing process. Polycar-bonate is gaining prevalence for 3-D printing in industrial anddesktop settings, because of its strength, heat resistance, andoptical transparency. In addition, professional selective laser-sintering printers are able to print using nylon plastic, whichoffers high heat and chemical resistance. Printers capable ofprinting metal parts are currently expensive ($15,000–1,000,000�), and all conductive material shown here wasadded postprinting. In the future a single printer that is capableof incorporating both structural components as well as electri-cal circuits (1, 7, 8) will enable researchers to create fullyfunction experimental setups for a wide range of biomedicalinvestigations customized for each individual experiment.

Overall, open source manufacturing combined with opensource hardware and software offers life scientists a cost-

Fig. 4. Digital design and 3-D printed tissuechamber in the shape of a human torso andscaled proportionally for a rabbit heart.

Fig. 5. Prototyping a CLARITY chamber. A: photograph of the original CLARITY chamber that was not printed. B–E: sequential 3-D printed (top) CLARITYchambers and digital designs (bottom).

Perspectives

H1572 PRINTING A PHYSIOLOGY LABORATORY

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00599.2013 • www.ajpheart.org

effective technique that will improve both the quality andproductivity of biomedical research.

GRANTS

This research and development was supported by National Institutes ofHealth Grants R21-HL-112278, R01-HL-114395, R21-HL-108617, R01-HL-095010, and R01-EB-008999.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

M.S.S. and I.R.E. conception and design of research; M.S.S., E.W., C.C.S.,and K.M.H. performed experiments; M.S.S., E.W., C.C.S., K.M.H., C.G.,S.R.G., and I.R.E. analyzed data; M.S.S., E.W., C.C.S., K.M.H., C.G., S.R.G.,and I.R.E. interpreted results of experiments; M.S.S. prepared figures; M.S.S.and I.R.E. drafted manuscript; M.S.S., K.M.H., and I.R.E. edited and revisedmanuscript; M.S.S. and I.R.E. approved final version of manuscript.

REFERENCES

1. Calvert P. Inkjet printing for materials and devices. Chem Mater 13:3299–3305, 2001.

2. Chung K, Wallace J, Kim SY, Kalyanasundaram S, Andalman AS,Davidson TJ, Mirzabekov JJ, Zalocusky KA, Mattis J, Denisin AK,

Pak S, Bernstein H, Ramakrishnan C, Grosenick L, Gradinaru V,Deisseroth K. Structural and molecular interrogation of intact biologicalsystems. Nature 497: 332–337, 2013.

3. Fye WB. Carl Ludwig and the Leipzig Physiological Institute: ‘a factoryof new knowledge’. Circulation 74: 920–8, 1986.

4. Jones R, Haufe P, Sells E, Iravani P, Olliver V, Palmer C, Bowyer A.RepRap—the replicating rapid prototyper. Robotica 29: 177–191, 2011.

5. Lang D, Sulkin MS, Lou Q, Efimov IR. Optical mapping of action potentialsand calcium transients in the mouse heart. J Vis Exp 55: e3275, 2011.

6. Laughner JI, Ng FS, Sulkin MS, Arthur RM, Efimov IR. Proces-sing and analysis of cardiac optical mapping data obtained withpotentiometric dyes. Am J Physiol Heart Circ Physiol 303: H753–H765, 2012.

7. Lipson H. Homemade: the future of functional rapid prototyping. IEEESpectrum 42: 24–31, 2005.

8. Lopes AJ, MacDonald E, Wicker RB. Integrating stereolithography anddirect print technologies for 3D structural electronics fabrication. RapidPrototyping J 18: 129–143, 2012.

9. Pietka TA, Sulkin MS, Kuda O, Wang W, Zhou D, Yamada KA, YangK, Su X, Gross RW, Nerbonne JM, Efimov IR, Abumrad NA. CD36protein influences myocardial Ca2� homeostasis and phospholipid metab-olism: conduction anomalies in CD36-deficient mice during fasting. J BiolChem 287: 38901–38912, 2012.

10. Rothschuh KE. History of Physiology. Huntington, New York: R. E.Krieger, 1973, p. 204–212, 248–258.

11. Zhang C, Anzalone NC, Faria RP, Pearce JM. Open-source 3D-printable optics equipment. PLoS One 8: e59840, 2013.

Perspectives

H1573PRINTING A PHYSIOLOGY LABORATORY

AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00599.2013 • www.ajpheart.org