I. INTRODUCTION

Section:

ChooseTop of pageABSTRACTI. INTRODUCTION <<II. EXPERIMENTAL METHODS ...III. RESULTS AND DISCUSSI...IV. CONCLUSIONSSUPPLEMENTARY MATERIALREFERENCESThe nematode Caenorhabditis elegans is a preeminent research organism in biology and medicine. Its small size, ease of culture, rapid development, and large mutant libraries make this organism a powerful model for assigning functions to genes relevant to human disease.11. D. D. Shaye and I. Greenwald, “Ortholist: A compendium of C. elegans genes with human orthologs,” PLoS One 6(5), e20085 (2011). https://doi.org/10.1371/journal.pone.0020085 These strengths are paired with unsurpassed genetic tractability, including a large molecular biological toolkit and the most comprehensively annotated genome to date.Transgenesis—the transfer of exogenous DNA into the germline of an organism—is an essential step in many C. elegans research projects. This technology has changed very little since its introduction 30 years ago.22. C. C. Mello, J. M. Kramer, D. Stinchcomb, and V. Ambros, “Efficient gene transfer in C. elegans: Extrachromosomal maintenance and integration of transforming sequences,” EMBO J. 10(12), 3959–3970 (1991). https://doi.org/10.1002/j.1460-2075.1991.tb04966.x Conventional C. elegans transgenesis involves four main steps: (i) manually mounting up to 20 worms on a dry agarose pad formed on coverslip; (ii) transferring the coverslip to a compound microscope fitted with a micromanipulator and pipet holder to position the injection pipet; (iii) inserting the pipet tip into the gonad of each worm and injecting a solution of DNA; and (iv) manually recovering each injected animal from the coverslip to a standard culture plate. In this procedure, the physical resistance of the worm to penetration by the pipet critically depends on the fact that the agarose substrate is dry, causing the worm to stick. The drawback of this method of immobilization is that it rapidly desiccates the animal. Therefore, to obtain reasonable survival rates, this procedure must be performed quickly, in approximately 1–2 min per worm. The need for both speed and dexterity makes the technique difficult to master and tiring to perform; even experienced investigators rarely inject more than four to six different genetic constructs per day. To address this problem, a computer-assisted microinjection platform system has been demonstrated.33. C. L. Gilleland, A. T. Falls, J. Noraky, M. G. Heiman, and M. F. Yanik, “Computer-assisted transgenesis of Caenorhabditis elegans for deep phenotyping,” Genetics 201(1), 39–46 (2015). https://doi.org/10.1534/genetics.115.179648 Animals are immobilized in a temperature-sensitive hydrogel, which keeps them hydrated. Microinjections are executed using computer vision and robotics. However, the cost of this system ($100 000)33. C. L. Gilleland, A. T. Falls, J. Noraky, M. G. Heiman, and M. F. Yanik, “Computer-assisted transgenesis of Caenorhabditis elegans for deep phenotyping,” Genetics 201(1), 39–46 (2015). https://doi.org/10.1534/genetics.115.179648 is prohibitive for most individual laboratories. Thus, there is a need for transgenesis methods that are not only easier to learn and less tiring to perform, but also lower in cost.At least five low-cost nematode microinjection devices have been reported.4–84. X. Zhao, F. Xu, L. Tang, W. Du, X. Feng, and B. F. Liu, “Microfluidic chip-based C. elegans microinjection system for investigating cell-cell communication in vivo,” Biosens. Bioelectron. 50, 28–34 (2013). https://doi.org/10.1016/j.bios.2013.06.0245. P. Song, X. Dong, and X. Liu, “A microfluidic device for automated, high-speed microinjection of Caenorhabditis elegans,” Biomicrofluidics 10(1), 011912 (2016). https://doi.org/10.1063/1.49419846. M. Nakajima, Y. Ayamura, M. Takeuchi, N. Hisamoto, S. Pastuhov, Y. Hasegawa et al., “High-precision microinjection of microbeads into C. elegans trapped in a suction microchannel,” in Proceedings of the IEEE International Conference on Robotics and Automation (IEEE, 2017), pp. 3678–3683.7. R. Ghaemi, J. Tong, B. P. Gupta, and P. R. Selvaganapathy, “Microfluidic device for microinjection of Caenorhabditis elegans,” Micromachines 11(3), 295 (2020). https://doi.org/10.3390/mi110302958. X. Dong, P. Song, and X. Liu, “Automated robotic microinjection of the nematode worm Caenorhabditis elegans,” IEEE Trans. Autom. Sci. Eng. 18(2), 850–859 (2021). https://doi.org/10.1109/TASE.2020.2990995 These devices solve the problem of immobilization without desiccation by trapping the worm in a fluid-filled microfluidic compartment. Validation of each of these devices has so far been limited to injection of generic solutions into the body cavity of the worm; none were used to inject DNA solutions into the worm's gonad. Nevertheless, taken together, these publications highlight the main design challenge for microfluidic microinjection devices: the need to bring the injection pipet into contact with the worm without clogging or breaking the pipet tip. Existing designs fall into two main categories: (i) open systems, in which the compartment holding the worm allows unobstructed pipet access from above44. X. Zhao, F. Xu, L. Tang, W. Du, X. Feng, and B. F. Liu, “Microfluidic chip-based C. elegans microinjection system for investigating cell-cell communication in vivo,” Biosens. Bioelectron. 50, 28–34 (2013). https://doi.org/10.1016/j.bios.2013.06.024 and (ii) closed systems, in which the compartment holding the worms is a channel that includes a ceiling.5–85. P. Song, X. Dong, and X. Liu, “A microfluidic device for automated, high-speed microinjection of Caenorhabditis elegans,” Biomicrofluidics 10(1), 011912 (2016). https://doi.org/10.1063/1.49419846. M. Nakajima, Y. Ayamura, M. Takeuchi, N. Hisamoto, S. Pastuhov, Y. Hasegawa et al., “High-precision microinjection of microbeads into C. elegans trapped in a suction microchannel,” in Proceedings of the IEEE International Conference on Robotics and Automation (IEEE, 2017), pp. 3678–3683.7. R. Ghaemi, J. Tong, B. P. Gupta, and P. R. Selvaganapathy, “Microfluidic device for microinjection of Caenorhabditis elegans,” Micromachines 11(3), 295 (2020). https://doi.org/10.3390/mi110302958. X. Dong, P. Song, and X. Liu, “Automated robotic microinjection of the nematode worm Caenorhabditis elegans,” IEEE Trans. Autom. Sci. Eng. 18(2), 850–859 (2021). https://doi.org/10.1109/TASE.2020.2990995 In closed systems, unobstructed access to the worm is achieved by means of a dedicated pipet channel that joins the worm channel at a T-junction in the plane of the device.Open and closed systems have reciprocal strengths and weaknesses. A key strength of open systems is that movement of the pipet is essentially unconstrained, making them well-suited to the use of conventional injection setups. Open systems also facilitate changing the injection pipet if clogged or broken (a frequent occurrence in DNA injections). In open systems, however, it is more difficult to fix the worm in position. These systems, therefore, rely on suction channels,44. X. Zhao, F. Xu, L. Tang, W. Du, X. Feng, and B. F. Liu, “Microfluidic chip-based C. elegans microinjection system for investigating cell-cell communication in vivo,” Biosens. Bioelectron. 50, 28–34 (2013). https://doi.org/10.1016/j.bios.2013.06.024 which, in turn, require additional microfluidic channels and off-chip apparatus to regulate and switch the suction. A key strength of closed systems is that they eliminate the complications of suction.5–85. P. Song, X. Dong, and X. Liu, “A microfluidic device for automated, high-speed microinjection of Caenorhabditis elegans,” Biomicrofluidics 10(1), 011912 (2016). https://doi.org/10.1063/1.49419847. R. Ghaemi, J. Tong, B. P. Gupta, and P. R. Selvaganapathy, “Microfluidic device for microinjection of Caenorhabditis elegans,” Micromachines 11(3), 295 (2020). https://doi.org/10.3390/mi110302958. X. Dong, P. Song, and X. Liu, “Automated robotic microinjection of the nematode worm Caenorhabditis elegans,” IEEE Trans. Autom. Sci. Eng. 18(2), 850–859 (2021). https://doi.org/10.1109/TASE.2020.2990995 Another strength is that the injection channel facilitates worm handling. When the injection channel is connected to a worm reservoir at one end and a recovery chamber at the other end, the process of moving an injected worm out of the injection channel automatically brings the next worm into position for injection. On the other hand, existing closed devices have the weakness that the injection pipet must be inserted without breakage through a long, narrow channel in the plane of the device, the height of which is on the order of the diameter of the worm, approximately 50 μm.5–95. P. Song, X. Dong, and X. Liu, “A microfluidic device for automated, high-speed microinjection of Caenorhabditis elegans,” Biomicrofluidics 10(1), 011912 (2016). https://doi.org/10.1063/1.49419846. M. Nakajima, Y. Ayamura, M. Takeuchi, N. Hisamoto, S. Pastuhov, Y. Hasegawa et al., “High-precision microinjection of microbeads into C. elegans trapped in a suction microchannel,” in Proceedings of the IEEE International Conference on Robotics and Automation (IEEE, 2017), pp. 3678–3683.7. R. Ghaemi, J. Tong, B. P. Gupta, and P. R. Selvaganapathy, “Microfluidic device for microinjection of Caenorhabditis elegans,” Micromachines 11(3), 295 (2020). https://doi.org/10.3390/mi110302958. X. Dong, P. Song, and X. Liu, “Automated robotic microinjection of the nematode worm Caenorhabditis elegans,” IEEE Trans. Autom. Sci. Eng. 18(2), 850–859 (2021). https://doi.org/10.1109/TASE.2020.29909959. R. Ardeshiri, B. Mulcahy, M. Zhen, and P. Rezai, “A hybrid microfluidic device for on-demand orientation and multidirectional imaging of C. elegans organs and neurons,” Biomicrofluidics 10(6), 064111 (2016). https://doi.org/10.1063/1.4971157 This arrangement makes it inconvenient to exchange injection pipets during a series of injections. In some systems, the injection pipet is integrated into the device during fabrication such that when the pipet becomes clogged, the entire device must be replaced.77. R. Ghaemi, J. Tong, B. P. Gupta, and P. R. Selvaganapathy, “Microfluidic device for microinjection of Caenorhabditis elegans,” Micromachines 11(3), 295 (2020). https://doi.org/10.3390/mi11030295 Furthermore, as the micropipet channel forms a junction with the worm channel, it introduces a fluid leak to the outside. Some devices reduce leakage by adding a pressure activated control layer to compress the ceiling of the pipet channel after the pipet positioned in the channel.5–85. P. Song, X. Dong, and X. Liu, “A microfluidic device for automated, high-speed microinjection of Caenorhabditis elegans,” Biomicrofluidics 10(1), 011912 (2016). https://doi.org/10.1063/1.49419848. X. Dong, P. Song, and X. Liu, “Automated robotic microinjection of the nematode worm Caenorhabditis elegans,” IEEE Trans. Autom. Sci. Eng. 18(2), 850–859 (2021). https://doi.org/10.1109/TASE.2020.2990995 Adding this control layer complicates the fabrication process and requires the addition of microfluidic channels, plus off-chip apparatus to regulate and switch the pressure. An alternative approach is to maintain the worm channel at ambient pressure.66. M. Nakajima, Y. Ayamura, M. Takeuchi, N. Hisamoto, S. Pastuhov, Y. Hasegawa et al., “High-precision microinjection of microbeads into C. elegans trapped in a suction microchannel,” in Proceedings of the IEEE International Conference on Robotics and Automation (IEEE, 2017), pp. 3678–3683. In this approach, instead of using fluid pressure, worms are moved to the injection site by harnessing their tendency to swim in the direction of an electric field; however, suction is still required to prevent the worm from swimming during the injection.Despite the considerable inconvenience of conventional transgenesis methods in C. elegans, there appear to be no published reports citing the use of any of the above-mentioned microfluidic devices to inject DNA solutions into the worm's gonad. We suspect there are three main reasons for this apparent failure of adoption. First, relatively few C. elegans laboratories have the ability to fabricate complex, multi-layered devices in-house, and using commercial foundry services for such devices would be prohibitively expensive. Second, the need to setup and maintain peripheral apparatus to control air pressure4–84. X. Zhao, F. Xu, L. Tang, W. Du, X. Feng, and B. F. Liu, “Microfluidic chip-based C. elegans microinjection system for investigating cell-cell communication in vivo,” Biosens. Bioelectron. 50, 28–34 (2013). https://doi.org/10.1016/j.bios.2013.06.0245. P. Song, X. Dong, and X. Liu, “A microfluidic device for automated, high-speed microinjection of Caenorhabditis elegans,” Biomicrofluidics 10(1), 011912 (2016). https://doi.org/10.1063/1.49419848. X. Dong, P. Song, and X. Liu, “Automated robotic microinjection of the nematode worm Caenorhabditis elegans,” IEEE Trans. Autom. Sci. Eng. 18(2), 850–859 (2021). https://doi.org/10.1109/TASE.2020.2990995 can be a barrier for laboratories lacking this type of experience. Third, it is not clear that any of the alternative systems would be significantly easier to learn and utilize than the conventional method.In response to the challenges of injecting worms immobilized in microfluidic channels, we have developed a closed PDMS device that eliminates the need for a dedicated pipet channel. Instead, one sidewall of the injection channel is a thin PDMS membrane which the injection pipet penetrates to reach the worm. This device, called the Poker Chip, is monolithic, making it comparatively easy to fabricate. Having no valves, the device requires no peripheral apparatus except a hand-held syringe to move worms into and out of the injection channel. Additionally, the Poker Chip is reusable and can be fabricated using polyurethane masters, which last much longer than conventional silicon-SU-8 masters.1010. S. P. Desai, D. M. Freeman, and J. Voldman, “Plastic masters—Rigid templates for soft lithography,” Lab Chip 9(11), 1631–1637 (2009). https://doi.org/10.1039/b822081f This fabrication process is likely to facilitate transfer of the technology to other laboratories. We anticipate that these advantages could lower barriers to adoption and accelerate basic, translational, and industrial research in this widely used model organism.

II. EXPERIMENTAL METHODS AND MATERIALS

Section:

ChooseTop of pageABSTRACTI. INTRODUCTIONII. EXPERIMENTAL METHODS ... <<III. RESULTS AND DISCUSSI...IV. CONCLUSIONSSUPPLEMENTARY MATERIALREFERENCES

A. Nematodes

The wild type (N2) strain of C. elegans was obtained from the Caenorhabditis Genetics Center at the University of Minnesota (St. Paul, USA). Worms were grown at 20 °C on NGM agar that had been previously seeded with the OP50 strain of E. coli. Worms were transferred to fresh plates of NGM with E. coli every 7–10 days to maintain well-fed stocks of worms. Synchronized populations of adult hermaphrodites were used throughout. Worms were synchronized by bleach synchronization according to established procedures1111. M. Porta-de-la-Riva, L. Fontrodona, A. Villanueva, and J. Cerón, “Basic Caenorhabditis elegans methods: Synchronization and observation,” J. Vis. Exp. 64, e4019 (2012). https://doi.org/10.3791/4019 and allowed to grow for 72 h at 20 °C, yielding a population enriched for the developmental stage required for injections (young adults, YA). At YA, worms are approximately 920 μm in length and 48 μm in diameter.1212. H. B. Atakan, F. Ayhan, and M. A. M. Gijs, “PDMS filter structures for size-dependent larval sorting and on-chip egg extraction of C. elegans,” Lab Chip 20(1), 155–167 (2020). https://doi.org/10.1039/C9LC00949C,1313. See https://www.wormatlas.org/ for “WormAtlas” (accessed January 1, 2022).

B. Solutions

Standard M9 buffer was prepared by combining 3 g KH2PO4, 6 g Na2HPO4, 5 g NaCl, and 1 ml of 1M MgSO4 then adding H2O to 1 l. Modified M9 buffer was prepared by adjusting the osmolarity of standard M9 to 315–320 mOsm by the addition of glycerol, to match the approximate osmolarity of the worm's internal fluid.

The injection mix used in the experiment of Table I and Fig. 5 contained 40 ng/μl Super-rol plasmid DNA (InVivo Biosystems, Eugene, OR), 358 ng/μl Salmon Testes DNA (Sigma-Aldrich Inc., Saint Louis, MO, USA), plus water to a final concentration of 100 ng/μl total DNA. The Super-rol plasmid is a double-stranded DNA that is similar to the widely used pRF4 plasmid containing the su100060 allele, but the Super-rol plasmid has been optimized for increased expression. Like pRF4, the Super-rol plasmid contains the R71C mutation in rol-6 causing the rolling phenotype.

TABLE I. Comparison of survival and transformation rates in the conventional method and the injection chip method. NLoaded is the number of worms placed on agarose pads in conventional injections or the number of worms loaded into the vestibule in the poker chip injections. NInjected is the number of worms in which injection mix was observed to fill the distal gonad arms. NSurvived is the number of worms that remained alive throughout the 3-day incubation period following injection. NTrans. is the number of transformed worms (those whose F1 progeny included worms having the roller phenotype). Survival and transformation rates are expressed as probabilities. Survival rate was computed as NSurvived/NInjected. Transformation rate was computed as NTrans./NInjected. Statistical comparisons (t-test), conventional vs poker chip. Survival rate: d.f. combined = 8.82, t = 1.37, p = 0.21; Transformation rate: d.f. combined = 7.22, t = 1.20, p = 0.27. CI, 95% confidence interval.

Success rateNLoadedNInjectedNRecov.NSurvivedNTrans.SurvivalTrans.Conventional121212830.670.25121212860.670.5018181817150.940.8318181813130.720.7218181813110.720.61Mean0.740.58CI0.61–0.880.31–0.86Poker Chip2512252380.830.6730203028120.900.603017302740.820.242211222140.910.362510252440.900.402514252050.640.36Mean0.830.44CI0.73–0.940.26–0.62Injection mix used in the experiment of Table II contained CRISPR dpy-10 co-injection marker, Cas9 (PNA-BIO, Newbury Park, CA, USA), dpy-10 gRNA (Synthego, Redwood City, CA, USA), and dpy-10 oligo deoxynucleotide as donor-homologous single stranded DNA (Integrated DNA Technologies, Coralville, IA, USA). Fluorescent dyes used (Table II) were tetramethylrhodamine dextran neutral and fluorescein dextran anionic (Thermo Fisher, Waltham, MA, USA). Stock solutions of dyes contained 25 mg/ml 0.2M KCl. Injection mix and dye solutions were combined to obtain the final dye concentrations indicated in Table II.

TABLE II. Absence of effect of fluorescein (Flour.) and rhodamine (Rhod.) on survival and transformation rate. Each row is an independent session of injections by the conventional method. The concentration column (Conc.) shows the final dye concentration in the injected fluid. N = number of worms as indicated by the subscript. Parenthetical values are survival and transformation rate expressed as probability.

Injection setDyeConc. (mg/ml)NinjectedNsurvived (p)Ntrans. (p)1Fluor.2.5001212 (1.0)12 (1.0)2Fluor.1.2501212 (1.0)12 (1.0)3Rhod.1.2501010 (1.0)10 (1.0)4Rhod.0.625109 (0.9)8 (0.8)

C. Device fabrication

Devices were cast in PDMS (Dow Corning Sylgard 184, Corning, NY, USA). Holes for inlet ports were cut using a catheter punch (1.52 mm ID, 1.82 mm OD, CR0720605N15R4, Syneo, Cutting Tool Division, Angleton, TX, USA). Reservoirs were formed using a 6 mm biopsy punch. PDMS castings were bonded to glass coverslips after 60 s exposure to an oxidizing air plasma (PDC-32G, Harrick Plasma, Ithaca, NY, USA).

D. DNA injection

Conventional injections followed the procedure of Mello et al.22. C. C. Mello, J. M. Kramer, D. Stinchcomb, and V. Ambros, “Efficient gene transfer in C. elegans: Extrachromosomal maintenance and integration of transforming sequences,” EMBO J. 10(12), 3959–3970 (1991). https://doi.org/10.1002/j.1460-2075.1991.tb04966.x For both types of injections, pipet pressure (40–100 psi) was regulated and switched using a digital microinjection pressure controller (MINJ-D, Tritech Research, Inc., Los Angeles, CA, USA).

III. RESULTS AND DISCUSSION

Section:

ChooseTop of pageABSTRACTI. INTRODUCTIONII. EXPERIMENTAL METHODS ...III. RESULTS AND DISCUSSI... <<IV. CONCLUSIONSSUPPLEMENTARY MATERIALREFERENCES

A. Design strategy

A key design objective was to lower barriers to adoption by making the device compatible with conventional injection setups. In such setups, the nematodes are mounted on a coverslip resting on the stage of an inverted compound microscope. The injection target is illuminated from above and viewed from below. The injection pipet, oriented at a low angle with respect to the plane of the stage, approaches the injection target from the side. Our design replicates this arrangement and can be used in conventional injection setups (Figs. 1 and 2).To facilitate fabrication, the Poker Chip has a minimal geometry and is monolithic. It comprises (i) an inlet port (1.5 mm diam.), (ii) a tapered vestibule (h = 27 μm) connected to the inlet port, (iii) an injection channel (h = 27 μm, w = 61 μm), optimized to restrain day-1 adults having a single row of eggs, and (iv) an outlet port (6 mm diam.), which serves as a post-injection reservoir. The cross-sectional area of the injection channel (1620 μm) is approximately 11% less than the cross-sectional area of YA worms (1810 μm, see Sec. ). Thus, worms are slightly compressed by the channel. Compression contributes to restraining worms for injection. It also helps to ensure that the injection target, the worm's gonad, lies at a consistent elevation from worm to worm, obviating the need to adjust the height of the injection pipet during an injection set.The height of the tapered vestibule, being less than the diameter of an adult worm (50 μm),1414. S. M. Maguire, C. M. Clark, J. Nunnari, J. K. Pirri, and M. J. Alkema, “The C. elegans touch response facilitates escape from predacious fungi,” Curr. Biol. 21(15), 1326–1330 (2011). https://doi.org/10.1016/j.cub.2011.06.063 forces the animal to lie on its left or right lateral midline, i.e., in the correct plane for injection. The natural orientation of worms on an agar substrate is to lie on their left or right lateral midline. They adopt this orientation because worms can undulate only in the dorsoventral plane, and the vertical component of surface tension in the thin layer of fluid stretched across the worm flattens these undulations against the substrate.1515. H. R. Wallace, “Wave formation by infective larvae of the plant parasitic nematode Meloidogyne javanica,” Nematologica 15, 65–75 (1969). https://doi.org/10.1163/187529269X00100 In microfluidic chambers having a feature height less than or equal to the diameter of a worm, the vertical component of surface tension on an agar substrate is replaced by the reaction forces of the chamber's ceiling. Consequently, worms in low-height microfluidic chambers lie on their left or right lateral midline, as they do on agar surfaces.1616. C. Cáceres I de, N. Valmas, M. A. Hilliard, and H. Lu, “Laterally orienting C. elegans using geometry at microscale for high-throughput visual screens in neurodegeneration and neuronal development studies,” PLoS One 7(4), e35037 (2012). https://doi.org/10.1371/journal.pone.0035037 Thus, worms in the Poker Chip are properly oriented for injection before they enter the injection channel.Opposite the midpoint along the length of the injection channel is a nose-shaped cut-out that terminates close to the nearest sidewall of the channel, forming a narrow septum with a width of approximately 40 μm through which the injection pipet is inserted to reach the worm [Figs. 1(a), 1(b) and 2(b), 2(c)]. The walls of the cut-out are vertical such that the injection pipet can easily be lowered from above, and there is no optical interference with visualization of the injection pipet as it approaches the septum. The top surface of the chip is optically flat to ensure a clear image of the worm and the pipet tip. To close the device, the chip is plasma bonded to a glass coverslip [Fig. 1(c)]. The assembled device can be used in this form, or it can be glued into an aluminum or acrylic frame to protect the coverslip from damage during handling [Figs. 1(c) and 2(a)].The width of the septum is a critical dimension [Fig. 2(c)]. It must be sufficiently thick to survive mold release without damage and to provide a mechanically strong bond with the coverslip. On the other hand, the septum must be thin enough to enable the user to align the tip of injection pipet with the vertical center of the worm's gonad after passage through the septum. The angle of the injection pipet with respect to the microscope stage [Fig. 1(b), lower] causes the pipet tip to follow a downwardly inclined trajectory as it passes through the septum. As a result, in order to hit the vertical center of the worm's gonad, the user must choose an entry point on the septum just high enough to compensate for the vertical drop of the pipet tip. A thinner septum facilitates this alignment process. We found that a thickness of 40 μm was an optimal compromise between mechanical strength and ease of alignment.

B. Fabrication of molds and chips

Fabrication of the Poker Chip required development of a mold in which a macroscopic feature, the nose-shaped cut-out, is located with micrometer-order precision next to a microscopic feature, the injection channel. To overcome this challenge, we designed an adjustable two-part brass mold (Fig. 3). The bottom plate of the mold includes a micromachined feature, which forms the injection channel. The top plate includes a cavity that creates the overall outline of the chip, including the nose feature. To assemble the mold, the top plate is placed in contact with washers on the bottom plate. The plates are then screwed together by four screws (1), one in each corner. This step positions the nose feature approximately 100 μm from the outside edge of the injection channel. The nose feature is then moved closer to the channel feature by turning screw (2), which presses against the backside of the nose feature. The nose feature is then locked in place by a screw (3), which passes through a clearance hole in the nose and threads into the bottom plate. The final position of the nose feature was found by an iterative process of casting a PDMS positive, measuring the thickness of the septum in a calibrated image, and repositioning the nose feature as needed.The brass mold is limited to casting one PDMS positive at a time. To overcome this limitation, we made multiplexed polyurethane molds1010. S. P. Desai, D. M. Freeman, and J. Voldman, “Plastic masters—Rigid templates for soft lithography,” Lab Chip 9(11), 1631–1637 (2009). https://doi.org/10.1039/b822081f by casting against a set of six PDMS positives generated from the brass mold. The final chips were made by filling the cavities of the polyurethane mold with degassed PDMS pre-polymer (10:1 by weight). To ensure an optically flat top surface, each cavity was slightly over-filled with PDMS. A silanized glass plate [1 h exposure to (tridecafluoro-1,1,2,2-tetrahydrooctyl)tricholorosilane], large enough to cover all six cavities, was then pressed down onto the mold, extruding excess PDMS. The glass slide was secured with a spring clamp and the assembled mold was cured for 3 h at 65 °C. To release chips from the mold, we removed the glass plate, then flooded the top of the mold with methanol and used a Teflon coated spatula to lift each chip from its cavity. Chips were plasma bonded to 24 × 60 mm2 No. 0 coverslips (Gold Seal Cover Glass, Fisher Scientific Co., Boston, MA, USA). The chip and coverslip assemblies were glued to the frame using spray adhesive (Elmer's Spray Adhesive, Newell Brands, Westerville, OH, USA).

C. Injection pipets

The Poker Chip is designed to be compatible with the glass micropipets used in conventional injections. To lower barriers to adoption, we used widely available pipettes, obtained from a commercial supplier of nematode DNA injection apparatus (TriTech, Los Angeles, CA, USA). The pipets were formed from borosilicate capillary tubes (1.0 mm OD, 0.60 mm ID, with filament) using a PC-100 pipet puller (Narashige International USA, Amityville, NY, USA) [Fig. 3]. The optimal pipet shape to minimize disruption of the septum is a long, gentle taper of about 2.8° starting approximately 140 μm from the tip (Fig. 4). When the pipet is in position within the gonad, the pipet diameter at the point of septum entry and exit is approximately 5.4 and 1.8 μm, respectively.To demonstrate the ability of the injection pipet to penetrate the septum without clogging, we filled the injection channel with mineral oil and placed a drop of oil at the base of the nose feature. The pipet was filled with a standard DNA injection mixture. When the pipet lumen was pressurized, ejection of fluid could be detected by the formation of a droplet at the tip (Video 1 in the supplementary material). We found that droplets could be formed inside the injection channel after the tip passed through the septum. When the pipet was then withdrawn from the septum, a droplet could be formed within the nose feature, indicating that the pipet remained open after traversing the septum. Visual inspection of the pipet (40×) indicated that the pipet tip remained intact. Pressurizing the injection channel after withdrawing the pipet did not generate a fluid bubble in the nose feature, indicating that the septum resealed after penetration (not shown). This was the case even after 100 penetrations of the septum, each at a different position along its length. This test demonstrates that a single Poker Chip could be used for making a long series of injections, including across multiple days if properly cleaned. As this test utilized the same injection pipet for all 100 penetrations, it also demonstrates that pipets are not easily broken by passage through the septum.

D. Method of use

Preparation of worms. To obtain the tens of worms needed for a series of injections, a culture of synchronized worms, enriched for the developmental stage required for injections (YA), is used. Worms are washed off the culture plate in 2 ml of standard M9 worm buffer (see Sec. ), transferred to a 2 ml Eppendorf tube, and rinsed several times by pelleting and aspiration to remove debris and bacteria. This method of rinsing is sufficient to eliminate debris large enough to clog the injection channel. After the final rinse, worms are concentrated into a loose pellet by allowing them to settle in the Eppendorf tube at room temperature or in a 4 °C refrigerator.Setup. The Poker Chip is prepared by filling the vestibule and injection channel with modified worm buffer (see Sec. ) contained in a 10 ml syringe fitted with 30–40 cm of polyethylene tubing (PE-9, Scientific Commodities, Inc., Lake Havasu City, AZ, USA). The post-injection reservoir is left mostly empty so it can accommodate worms and associated buffer that accumulate during an injection series. Worms are bulk loaded into the inlet port by drawing 1–3 μl of fluid from the pellet using a 10 μl micropipetter (3-000-510, Drummond Scientific Company, Broomall, PA, USA) fitted with a glass capillary tube (1.2 mm OD) and ejecting the fluid into the port. An air-filled 10 ml syringe is fitted with 30–40 cm of modified M9-filled PE-9 tubing; the free end of the tubing is inserted into the injection port. The chip is placed on the stage of a microscope fitted with Hoffman optics, with the middle of the septum centered in field of view. The chip is fixed in place using tape or plasticine modeling clay. A drop of hydrocarbon oil is placed in contact with the septum at the end of the nose feature for testing pipet function (see below).DNA injections. We use the term injection set to refer to a group of worms injected consecutively with the same DNA injection mix. We use the term injection series to refer to consecutive injection sets. An injection pipet filled with injection mix is inserted into a conventional micropipet holder attached to a micromanipulator. The pipet is moved into position above the nose feature, then lowered until its tip can be seen in the oil. At this point, injection pressure can be adjusted by raising it until the ejected fluid droplet is the desired size (see Video 1 in the supplementary material). A worm is moved into the injection channel using pressure from the syringe. The approximate vertical center of the gonad arm is found by focusing up and down through the arm. At this focal plane, the pipet tip is brought into focus just outside the septum by adjusting the vertical axis of the micromanipulator. The pipet is then raised slightly to compensate for its downward trajectory through the septum. At this point, the pipet is inserted through the septum and into the gonad, injection mix is injected, and the pipet is withdrawn from the worm. This process is illustrated in Video 2 in the supplementary material. As can be seen, the worm is firmly restrained and does not visibly respond to the injection. In any given injection series, two or three practice injections into one or two worms may be required to find the correct elevation of the pipet tip; this process takes no more than 5 min. Once this elevation is found, it is not necessary to adjust the elevation for injection of subsequent worms. After each injection, the pipet is withdrawn to a resting position in which the pipet tip is located at the horizontal center of septum, ready for the next worm to inject. This approach is illustrated in Video 2 in the supplementary material. Worms remain hydrated and apparently healthy for an hour or more in the chip. It is, therefore, possible to switch pipets and injection mixes many times without having to reload and remount the chip.For reasons described above, worms in the injection channel of the Poker Chip lie on either their left or right lateral midline such that their dorsal or ventral surface is apposed to the injection pipet. The standard injection site is the distal arm of the gonad, which is located dorsally. Thus, in the Poker Chip, only worms whose dorsal surface is apposed to the injection pipet are injected. Worms whose ventral surface is apposed to the pipet are skipped and pushed directly into the post-injection reservoir (Video 2 in the supplementary material). The dorsoventral orientation of worms in the injection channel is expected to be random. Indeed, in six replicate injection sets, the average probability of dorsal orientation was 0.53 ± 0.037 SEM. As a result of variations in rate of development, some culture plates may contain a number of worms that are too small to be firmly restrained in the injection channel. Such worms are also skipped. If most of the worms in the culture are too small or too large, the chip is reloaded with worms from a different culture plate.

E. Quantitative assessment of Poker Chip performance

To compare the success rate of DNA injections using the Poker Chip to the success rate of the conventional method, we used the same injection mix in parallel sets of injections, with 12–20 injected worms per set for each method (Table I). Both types of injections were carried out by the same injectionist (S.P.), an expert with eight years of experience with the conventional method who was also familiar with the Poker Chip method. The DNA marker in the mix was a semi-dominant allele of rol-6, which produces the so-called roller phenotype in which a cuticle defect causes worms to crawl in tight circles. This salient phenotype is widely used to positively identify transformed progeny of injected adults. Whenever possible, both gonad arms were injected in each worm. Worms were recovered from the Poker Chip by withdrawing them from the post-injection reservoir with a Pasteur pipet and placing them on a food-laden recovery plate. Twelve hours later, each worm was placed on its own recovery plate. After incubation for 3 days at 24 °C, the F1 progeny of each worm were scored for the roller phenotype. We computed survival rate as the fraction of injected worms that survived after being injected. We computed transformation rate as the fraction of injected worms whose progeny included rollers. We found that survival and transformation rates were statistically indistinguishable between the t