A self-reconfiguring metamorphic nanoinjector for injection into mouse zygotes Quentin T. Aten, Brian D. Jensen, Sandra H. Burnett, and Larry L. Howell Citation: Review of Scientific Instruments 85, 055005 (2014); doi: 10.1063/1.4872077 View online: http://dx.doi.org/10.1063/1.4872077 View Table of Contents: http://scitation.aip.org/content/aip/journal/rsi/85/5?ver=pdfcov Published by the AIP Publishing Articles you may be interested in Wavelength specific excitation of gold nanoparticle thin-films Appl. Phys. Lett. 104, 011909 (2014); 10.1063/1.4861603 Monte Carlo investigation of the increased radiation deposition due to gold nanoparticles using kilovoltage and megavoltage photons in a 3D randomized cell model Med. Phys. 40, 071710 (2013); 10.1118/1.4808150 Quantification of the specific membrane capacitance of single cells using a microfluidic device and impedance spectroscopy measurement Biomicrofluidics 6, 034112 (2012); 10.1063/1.4746249 Handling of artificial membranes using electrowetting-actuated droplets on a microfluidic device combined with integrated pA-measurements Biomicrofluidics 6, 012813 (2012); 10.1063/1.3665719 Automating fruit fly Drosophila embryo injection for high throughput transgenic studies Rev. Sci. Instrum. 79, 013705 (2008); 10.1063/1.2827516
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
REVIEW OF SCIENTIFIC INSTRUMENTS 85, 055005 (2014)
A self-reconfiguring metamorphic nanoinjector for injection into mouse zygotes Quentin T. Aten,1 Brian D. Jensen,2 Sandra H. Burnett,3 and Larry L. Howell2 1
Nexus Spine, LLC, Salt Lake City, Utah 84124, USA Department of Mechanical Engineering, Brigham Young University, Provo, Utah 84602, USA 3 Department of Microbiology and Molecular Biology, Brigham Young University, Provo, Utah 84602, USA 2
(Received 14 February 2014; accepted 9 April 2014; published online 13 May 2014) This paper presents a surface-micromachined microelectromechanical system nanoinjector designed to inject DNA into mouse zygotes which are ≈90 μm in diameter. The proposed injection method requires that an electrically charged, DNA coated lance be inserted into the mouse zygote. The nanoinjector’s principal design requirements are (1) it must penetrate the lance into the mouse zygote without tearing the cell membranes and (2) maintain electrical connectivity between the lance and a stationary bond pad. These requirements are satisfied through a two-phase, self-reconfiguring metamorphic mechanism. In the first motion subphase a change-point six-bar mechanism elevates the lance to ≈45 μm above the substrate. In the second motion subphase, a compliant folded-beam suspension allows the lance to translate in-plane at a constant height as it penetrates the cell membranes. The viability of embryos following nanoinjection is presented as a metric for quantifying how well the nanoinjector mechanism fulfills its design requirements of penetrating the zygote without causing membrane damage. Viability studies of nearly 3000 nanoinjections resulted in 71.9% of nanoinjected zygotes progressing to the two-cell stage compared to 79.6% of untreated embryos. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4872077] I. INTRODUCTION
The purpose of this work is to describe the development of a system capable of electro-physical injection of DNA into mouse zygotes (single cell embryos consisting of a fertilized egg). The objectives include achieving high cell survival, producing specified motion at the single-cell-size scale, and designing within the constraints of micro-fabrication processes. This paper describes the design, numerical modeling, in vitro biological testing, and refinement of a metamorphic nanoinjector instrument. Its origami-like folding motion begins in the fabrication plane and has motion out of that plane. Design parameters and fabrication and test results are provided for a nanoinjector specifically designed to inject mouse zygotes, and general design equations are provided to facilitate the design of nanoinjectors for other cell types. Insertion of foreign genetic material into the genome of mice and other mammals is a critical step in understanding and utilizing genes’ functions. Genetically, modified (transgenic) animals are currently used in research into such varied topics as cancer,1, 2 Alzheimer’s disease,3, 4 and diabetes.5 To genetically modify every cell within an animal, the genetic construct of interest (transgene) must be present in at least one pronucleus (the membrane surrounding each set of parental chromosomes) before the zygote’s first division.6, 7 In a mechanical sense, direct transgene delivery to a zygote requires precise motion across tens of micrometers for pronuclear penetration. Microelectromechanical system (MEMS) devices are potentially well suited to introduce genetic material into developing embryos because of their relatively small scale, potential for complex motion, and their ability to mechanically and electrically interact with individual cells.8–10 Addition0034-6748/2014/85(5)/055005/10/$30.00
ally, MEMS fabrication techniques allow for features within a single device to be as small as tens of nanometers, or as large as hundreds of micrometers. The nanoinjector’s metamorphic behavior11–16 enables it to first elevate to the cell mid-height and then pierce the cell membrane with high rates of cell survival. Because micro-fabrication constraints limit the as-fabricated position to be parallel to the substrate and the motion comes out of the substrate, the mechanism is related to lamina emergent mechanisms17–19 and origami-inspired mechanisms.20–22 Viability testing of nearly 3000 zygotes following penetration by the nanoinjector lance is presented to quantify how well the nanoinjector meets its design requirement of penetrating the zygote without permanently damaging its membranes. II. BACKGROUND A. Nanoinjection
The nanoinjector applies nanometer-scale features, precise mechanical motion over tens of microns, and electrical manipulation of DNA in an electro-physical method of gene transfer called nanoinjection. Figure 1 graphically outlines the nanoinjection process. The nanoinjector mechanism is operated while submerged in a pH buffered solution (such as phosphate buffered saline, PBS). A positive electrical charge is applied to the lance, which accumulates negatively charged DNA23 on its surface.24 The nanoinjector mechanism then penetrates the zygotic membranes, and a negative charge is applied to the lance, releasing the accumulated DNA within the cell. In vitro experiments on mouse embryos have demonstrated that nanoinjection can electro-physically transport
85, 055005-1
© 2014 AIP Publishing LLC
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-2
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
(a) SEM of a nanoinjector
FIG. 1. Schematic representation of nanoinjection of DNA into a mouse zygote. All nanoinjections occur with the nanoinjector submerged in a pH buffered saline solution. Step 1, the nanoinjector is in its as-fabricated position. Step 2, the lance is elevated and a positive charge is applied, accumulating DNA on the lance’s tip. Step 3, the lance moves at a constant height, penetrating into the target zygote. Step 4, the charge on the lance is reversed, releasing DNA into the zygote. Step 5, the lance moves at a constant elevation out of the zygote.
genetic material into embryos, and that the nanoinjected genes can be expressed by developing embryos.24
(b) Labeled SEM of a nanoinjector
B. Current methods of direct gene delivery
Currently, direct microinjection of genetic material into mammalian zygotes uses a hollow needle (a micron-scale tapered glass micropipette) driven by a micromanipulator to penetrate the zygotic membranes and a pump to expel minute volumes of a nucleic acid solution into the zygote.25, 26 Though significant improvements to microinjection equipment have been made,27, 28 the core elements of the needleand-pump design paradigm have remained essentially unchanged since microinjection hardware first appeared in the literature over 50 years ago.29, 30 Breaking from the needle-and-pump concept, MEMS DNA injection systems have been developed for adherent culture cells. These methods deliver DNA via fixed vertical nano-needles on a chip,31, 32 or modified atomic-force microscopy (AFM) probes.33, 34 The fixed nano-needles require that the target cells grow with the needle’s penetrating their surface, making fixed vertical needles impractical for transgenesis.31, 32 The AFM probe based needles require several minutes of incubation in each cell, and do not have sufficient penetration into the target cell to reach the zygote’s pronucleus.33, 34 III. INSTRUMENT DESCRIPTION AND MODELING
As a means of delivering transgenes into the zygote’s pronucleus, the nanoinjector mechanism shown in Figure 2
(c) SEM of the lance tip FIG. 2. Labeled scanning electron microscope (SEM) images of (a) the metamorphic nanoinjector posed as if it were injecting a 100 μm latex sphere, (b) the metamorphic nanoinjector in its as-fabricated position, and (c) the tip of the nanoinjector lance. The two six-bar mechanisms restraining the latex sphere in (a) are not described in this paper.
has three main functions: first, elevate the lance from its as-fabricated position on the fabrication substrate; second, move the lance horizontally at a constant height, penetrating the lance into the cell; and third, maintain electrical conductivity between the lance and an external voltage source. Constructing the nanoinjector as a self-reconfiguring metamorphic mechanism11–14 enables the lance elevation to be sequentially decoupled from the purely horizontal lance motion. Decoupling the elevation and purely horizontal motion phases ensures that the lance penetrates the zygote’s membranes and pronucleus along a linear, horizontal axis, and prevents tearing of the zygotic membranes.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-3
Aten et al.
The nanoinjector mechanism consists of one rigid-body six-bar mechanism, one compliant parallel-guiding mechanism, and two compliant electrical connections, as shown in Figure 2(b). The rigid-body six-bar mechanism provides the out-of-plane displacement in the nanoinjector’s first metamorphic subphase. The compliant parallel-guiding mechanism provides the in-plane translation toward the target cell during the second subphase. The method of self-reconfiguration and models of each of the nanoinjector’s components are described below. Prototype nanoinjectors were fabricated using MEMSCAP Inc.’s polycrystalline silicon Multiuser MEMS Processes (polyMUMPs).35 The process provides one stationary polycrystalline silicon layer (POLY0), two structural layers of polycrystalline silicon (POLY1 and POLY2), and a gold layer for increasing electrical conductivity which may be added to the POLY2 layer. The POLY1 and POLY2 layers are 2.0 μm and 1.5 μm thick, respectively. This well established process was selected for developing the nanoinjector mechanism because of the highly controlled and repeatable nature of the process from fabrication run to fabrication run. This allowed the development work to focus solely on the design and biological testing of the nanoinjector and helped mitigate deviceto-device variability caused by fabrication variability. The fabrication process is capable of producing parts with a large range in feature size. For example, the tip of the nanoinjector lance in Figure 2(c) has a minimum in-plane width of only 17 nm, while the mechanism’s overall dimensions are four orders of magnitude larger, as seen in Figure 2(b). Polycrystalline silicon has an ultimate strength of approximately between 1 GPa and 3 GPa, depending on the processing conditions and gage area.36 The ultimate strength of POLY1 and POLY2 were assumed to lie near the bottom of this distribution at 1.5 GPa.
Rev. Sci. Instrum. 85, 055005 (2014)
Link A
Link B
(c) schematic of a scissor joint
(a) SEM of a scissor joint
Link A
Link B
(d) schematic of a slider joint (b) SEM of a slider joint
FIG. 3. Scanning electron microscope (SEM) images of the two types of outof-plane joints used in the nanoinjector: (a) scissor joints and (b) slider joints. Schematics of each joint type are shown in (c) and (d).
layer, the minimum sizes of features, and between-layer vias. Figure 4 shows two sources of parasitic motion in the scissor joint: one governed by minimum feature size, and the other governed by minimum gaps between features in the same
A. Out-of-plane surface micromachined mechanisms
Surface micromachined mechanisms can achieve outof-plane displacements many times greater than their asfabricated thickness through the use of specially designed joints.37–39 The out-of-plane revolute joints used in the nanoinjector can be categorized as “scissor joints” and “slider joints.” Examples of each of these joint types are shown in Figure 3. The slider joints can undergo rotations of approximately 180◦ , and the scissor joints can achieve rotations in excess of 90◦ . These joints can be combined to create complex mechanisms, such as those in Ref. 40. These joints can be fabricated from planar layers of material but enable motion out of the fabrication plane. The motion is similar to origami or pop-up mechanisms, which begin with a flat sheet of paper and out-of-plane motion occurs at folds or creases in the paper.20, 41 The MEMS scissor and slider joints can be viewed as replacements for origami creases in paper, and they allow origami-like motions to occur. The limitations of multi-layer surface micromachining processes have a significant impact on the precision of scissor joint’s motion. Parasitic motion is inherent in the joints because of the minimum gaps between features in the same
( ) (a)
(b) FIG. 4. Scanning electron micrographs of the scissor joints in the nanoinjector mechanism showing (a) parasitic motion due to minimum feature size and (b) minimum gaps between features in the same layer.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-4
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
FIG. 6. Kinematic diagram of the change-point six-bar mechanism. The location a where the electrical connection attaches to link R3 and the loads exerted by the electrical connection Fy , Fz , and Mx are also given. θ 3 is the angle of link 3 (in the case shown, θ 3 = 0 and is not shown).
FIG. 5. Schematic illustration of the nanoinjector’s motion. The folded-beam suspension’s suspension is approximated as the linear spring at left. At position 0, the nanoinjector is in its as-fabricated configuration. Between positions 0 and 1, the nanoinjector is in its first subphase, and between positions 1 and 2, the nanoinjector is in its second subphase.
layer. The kinematic models of the nanoinjector presented below account for these two types of parasitic motion.
B. Self-reconfiguration through unequal subphase mechanism stiffnesses and link contact
The nanoinjector’s sequential “up-then-forward” motion is a key element of the mechanism’s functionality. Selfreconfiguration in the nanoinjector is a consequence of the unequal force-displacement characteristics of the six-bar mechanism and the folded-beam suspension, and contact between links in the nanoinjector mechanism. The nanoinjector’s motion is shown schematically in Figure 5, and a kinematic diagram of the six-bar mechanism is shown in Figure 6. At position 0, the nanoinjector is in its as-fabricated configuration, with the six-bar mechanism mounted on to the folded-beam suspension. In other words, it is as if the six-bar mechanism is kinematically grounded to the folded-beam suspension rather than the substrate. Between positions 0 and 1, the nanoinjector is in its first subphase, with the lance moving along a kinematic path with both out-of-plane and in-plane components. This motion raises the lance to the desired level, but occurs away from the zygote to prevent damage to the cell membrane. Assuming negligible friction in the six-bar’s slider and scissor joints, actuation between positions 0 and 1 applies approximately zero force to the folded-beam suspension. With negligibly small force applied to the folded-beam suspension by the six-bar mechanism, there is no additional in-plane translation from
the deflection of the folded-beam suspension’s compliant flexures. At position 1, contact is made between the input slider and the folded-beam suspension. This effectively locks the prismatic (slider) joint shown in Figure 6 and fixes the length R6 . With this degree of freedom removed, the six-bar mechanism becomes a structure with respect to the fold-beam suspension. Between positions 1 and 2, the six-bar mechanism remains at the same elevation as position 1 while the foldedbeam suspension deflects, resulting in an in-plane translation of the lance. This in-plane translation allows a linear motion of the raised lance such that minimal damage to the cell membrane occurs during lance penetration. Pulling back on the input slider, the mechanism will proceed from position 2, to 1, back to 0.
C. Modeling the six-bar mechanism
If the scissor and slider joints in the nanoinjector six-bar mechanism are treated as idealized revolute joints, the mechanism can be modeled as shown in Figure 6. Due to the planar nature of surface micromachining, the mechanism is fabricated in a change-point configuration, with all of its links co-planar. However, the mechanism can achieve only the configuration pictured because the other kinematic configurations are only possible if one or more links move in the negative zdirection (through the substrate). In the nanoinjector, links R1 and R3 are the same length, as are links R2 and R4 . Thus, the mechanism can be modeled as a parallel-guiding (parallelogram) four-bar mechanism (links R1 through R4 ) with a driver dyad (R5 and the input slider). The parallel-guiding motion of the six-bar mechanism ensures that the lance will be horizontal through out its motion. The position of the mechanism can be calculated by θ4 = cos−1
(R4 + R5 − Yin )2 + R42 − R52 , 2R4 (R4 + R5 − Yin )
θ5 = 2π + sin−1 (R4 sin θ4 /R5 ),
(1)
(2)
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-5
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
θin = θ5 − π,
(3)
Zout = R4 sin θ4 ,
(4)
Yout = R4 (1 − cos θ4 ),
(5)
(a)
where θ 2 = θ 4 because of the parallelogram configuration. Idealizing the scissor joints as pure revolute joints can be a poor assumption given the amount off parasitic motion that is possible (see Figure 4). The parasitic motion in the scissor joints on links R2 , R4 , and R5 can approximately be modeled through adjusting the lengths of links, and the input displacement Yin by R2 = R2 + Pl ,
(6)
R4 = R4 + Pl ,
(7)
R5 = R5 + Pl ,
(8)
Yin = Yin + 3Pg ,
(9)
(b)
where Pl is the change in link length possible in the scissor joint (see Figure 4(a)), and Pg is the gap between the links connected by a scissor joint when the mechanism is in its asfabricated position (see Figure 4(b)). The factor Pg is multiplied by 3 to account for each of the three unique scissor joints in the mechanism: R2 to R3 , R3 to R5 , and R4 to R5 . The dimensions given in Table I were chosen to reduce the effects of parasitic motion on the lance’s final out-of-plane displacement (Zout ). Figure 7(a) shows kinematic models of the nanoinjector both with and without accounting for the parasitic motion in the scissor joints. The input force required to actuate the nanoinjector mechanism, Fin , is a function of the loads generated by the electrical connection during the first and second motion subphases and of the folded-beam suspension in the second moTABLE I. Dimensions of the lance six-bar mechanism shown in Figure 2(b). Link lengths and positions given correspond to the kinematic diagram in Figure 6. Link lengths and positions given in parenthesis take into account the parasitic motion given by Eq. (6)–(9). All dimensions are in μm unless otherwise noted. Link or parameter R1 = R3 R2 = R4 R5 Yin Zout Yout Lance length Lance taper Pl Pg
Value 90.0 50.0 (66.5) 120.0 37.0 (28.0) 45.0 (47.1) 28.2 (19.6) 200.0 3◦ 16.5 3.0
FIG. 7. The effect of parasitic motion on the final out-of-plane displacement (Zout ) is shown through (a) the kinematic diagram of the change-point six-bar mechanism, and through (b) Zout as a function between the input displacement (Yin ) for both the ideal mechanism with no parasitic motion (solid line), and with maximum parasitic motion (dotted line).
tion subphase, Fin = −
Fy (cos(θ3 − θ5 ) + cos(θ3 + θ5 )) 2 sin(θ3 + θ5 )
−
Fz (sin(θ3 − θ5 ) + sin(θ3 + θ5 )) 2 sin(θ3 + θ5 )
−
(Mx + aFz )(sin(θ2 − θ3 + θ5 ) + sin(θ2 − θ3 − θ5 )) R4 (cos(θ2 + θ3 + θ5 ) − cos(θ2 − θ3 − θ5 ))
−Ff b ,
(10)
where Ffb is the load applied to the input slide by the foldedbeam suspension, and Fy , Fz , and Mx are the loads exerted by the electrical connection (see Figure 6). Each angle θ i represents the angle of link i, as illustrated in Figure 6 (except θ 3 is not shown because the angle of link 3 is zero for the position shown). The term Ffb is zero until contact is made between the input slider and the folded-beam suspension. In other words, the folded-beam suspension does not contribute to the loading on the input slider until the second motion subphase. D. Modeling the folded-beam suspension
The fully compliant folded-beam suspension was analyzed using a psuedo-rigid-body model (PRBM) and nonlinear finite element analysis. The eight fixed-guided compliant flexures are 300 μm long, 5 μm wide, and 2 μm thick (POLY1). As shown in Figure 8, the PRBM of a fixedguided beam consists of one pseudo-rigid link and two
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-6
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
FIG. 8. A schematic of a fixed-guided beam and its pseudo-rigid-body model.
pseudo-springs. The input force, P, and maximum stress in the flexure, σ max , can be found by = arcsin
b , γl
α 2 = 2K ,
(11)
(12)
P = 4α 2 EI / l 2 ,
(13)
a = l(1 − γ (1 − cos )),
(14)
P a( h2 ) , (15) 2I where b and a are the y and x coordinates of the end of the beam, E is Young’s modulus for polysilicon (≈164 GPa), I is the second moment of area for a rectangular beam, γ = 0.8517, and K = 2.65 (see Ref. 42). The folded-beam suspension’s eight compliant flexures are in a parallel and series arrangement. The right and left halves of the suspension are in parallel. Within each half, the pair of outer flexures is in series with the pair of inner flexures. Additionally, the two outer flexures are in parallel with each other, and the two inner flexures are in parallel with each other. Summing the parallel and series stiffnesses and displacements, the full suspension has the net-stiffness of four parallel-guiding flexures which displace one-half the total displacement of the suspension. The calculated input force and maximum stress predicted by the pseudo-rigid-body model for a total stage displacement of 70 μm (each flexure displacing 35 μm) are given in Table II. In addition to the pseudo-rigid body model, a nonlinear (large deflection, small strain) finite element analysis (FEA) was performed on the folded-beam suspension using ANSYS 11.0. The analysis modeled the geometrically nonlinear (small strain, large rotation) deformation of one-half of σmax =
TABLE II. Comparison of pseudo-rigid-body model and finite element analyses of the folded-beam suspension.
Force total Maximum stress
PRBM
FEA
222 μN 993 MPa
199 μN 1334 MPa
FIG. 9. Finite element model Von Mises stress contours for the folded-beam suspension undergoing 70 μm of displacement in the y-direction. The two areas indicated in red cross-hatching were fixed in all degrees of freedom. The y-direction input displacement was applied to the surfaces indicated. Stresses are in MPa.
the folded-beam suspension with 7488 second-order shell elements. The compliant flexures were meshed with uniformly distributed quadratic elements, and the rigid portions were meshed with larger tetrahedral elements. The model was fixed at two anchor areas, had symmetry boundary conditions applied along the y-z plane, and displaced 70 μm in-plane at the point of contact between the folded-beam suspension and the input slider, as shown in Figure 9. The model was evaluated in 20 uniformly distributed load steps. In post processing, the reaction force on the displaced face was calculated from the nodal results at each step. The input force and the maximum stresses predicted by the FEA are compared with the PRBM results in Table II. The FEA results and the PRBM results agree to within 11% for the input force, and 25% for the maximum stress. The higher stress predicted by the finite element model is a result of the stress concentration at the right angle between the compliant flexures and the rigid portion of the suspension (see the inset in Figure 9). The stresses predicted by both models are below the ultimate strength of polycrystalline silicon.36
E. Modeling the compliant electrical connections
The serpentine flexure43 electrical connections between the nanoinjector and the stationary bond-pads are designed to provide electrical conductivity without interfering with the nanoinjector’s motion. The electrical connections are 1.5 μm thick, 5 μm wide, and consist of ten 250 μm long flexures, connected by nine 50 μm long segments in a serpentine pattern (see Figure 2(b)). Only a finite element analysis of the electrical connections was performed because there is no closed-form or pseudo-rigid body model for the serpentine flexures’ nonlinear three-dimensional loading. One end of the electrical
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-7
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014) TABLE III. Reaction forces applied to the lance mechanism by one or both electrical connections for a displacement of 45 μm in the z-direction (out-ofplane) and 107 μm in the y-direction (in-plane). Note that the x-direction of force, y-direction moment, and z-direction moment sum to zero because of symmetry. All forces are in μN, and all moments are in μN μm.
Fx Fy Fz Mx My Mz
FIG. 10. Finite element model Von Mises stress contours for the electrical connections undergoing 45 μm in the z-direction (out-of-plane) and 107 μm in the y-direction (in-plane). Stresses are in MPa.
connection was fixed in all degrees of freedom, while the other was displaced along the kinematic path following the point a on the six-bar mechanism (Figure 6) for a total displacement of 45 μm in the z-direction (out-of-plane) and 107 μm in the y-direction (in-plane), as indicated in Figure 10. Initially, the finite element model only included the electrical connection. However, this initial model predicted significant displacements into the substrate (negative z-direction) by part of the electrical connection. To more accurately predict the electrical connection’s final configuration and stress state, the interaction between the electrical connection and the substrate was modeled as an initially open contact pair, with the substrate being modeled as perfectly rigid. The simulation modeled the geometrically nonlinear (small strain, large rotation) deformation of the electrical connector with 5964 second-order, quadrilateral shell elements. The contact pair was modeled as a node-to-rigid surface contact pair with 5964 8-node node-to-surface contact elements on the electrical connector, and 9180 3-node tetrahedral target elements on the substrate. The model was evaluated in 30 load steps along the displaced end’s path. In post processing, the reaction forces and moments on the displaced face were calculated from the nodal results at each step. Figure 10 shows the deformed shape and Von Mises stress contours of the fully displaced electrical connection. Table III gives the reaction forces and moments for one and both electrical connections. The reaction forces are small in comparison to the total forces required to displace the compliant stage, and in testing have not interfered with the nanoinjector’s motion.
One
Both
±2.21 − 3.97 0.22 100.51 ±16.18 ±790.87
0 − 7.94 0.44 201.02 0 0
beam suspension were substituted into Eq. (10) to predict the total force displacement behavior of the mechanism. The force displacement characteristics of the full mechanism are illustrated in Figure 11. As shown in Figure 11, the electrical connections are responsible for the input force in the first motion subphase (up to 37 μm of input displacement). In the second motion subphase, at input displacements greater than 37 μm, the input slide contacts the folded-beam suspension, and the deformation of the folded beam suspension becomes the dominant load on the nanoinjector mechanism. IV. TESTING OF PROTOTYPE NANOINJECTORS
This section provides verification of the nanoinjector’s metamorphic motion through mechanical testing of the nanoinjector. The mechanism’s metamorphic motion is demonstrated, and embryo survival results are presented for nanoinjections into mouse zygotes (single-cell fertilized egg cells). The embryo survival data are especially important because these demonstrate that the nanoinjector’s metamorphic motion allows the lance to consistently penetrate the zygote’s membranes without tearing or damaging the membranes. Full nanoinjection protocols and data on successful nanoinjection of DNA into mouse embryos are presented in other literature where the biological experiments can be discussed in detail.44, 45
F. Modeling the total input force
The reaction forces and moments predicted by the finite element models of the electrical connections and the folded-
FIG. 11. Total input force (Fin ) as a function of total input displacement (Yin ).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-8
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
FIG. 12. Top view optical microscopy images of the nanoinjector’s motion.
A. Verification of metamorphic motion
The nanoinjector mechanism is actuated by applying a linear input with a tungsten micro-probe attached to a manually operated micromanipulator (Cascade Microtech model number MDL-RMS). The nanoinjector is operated under displacement control, with the micromanipualtor providing sufficient input force to actuate the nanoinjector through its full range of motion. Figure 12 shows optical microscopy images of the nanoinjector mechanism in its un-actuated position, its first metamorphic subphase (out-of-plane elevation of the lance), and its second subphase (in-plane translation). As described previously, the six-bar mechanism reaches its full out-ofplane elevation with negligible translation in the folded-beam suspension. Applying further actuation, the lance remains at a constant height above the substrate as the folded-beam suspension deforms. Figure 12 also subtly illustrates the potential for parasitic out-of-plane motion inherent in thin film compliant mechanisms. At best, the polyMUMPs process can offer out-ofplane aspect ratios ≤1.167, leading to compliant flexures with out-of-plane stiffnesses which may be less than or equal to their in-plane stiffnesses. Any eccentricity in the applied inplane loads may excite these out-of-plane displacements. A careful examination of the folded-beam suspension in the top view images in Figure 12 reveals changes in the suspension’s brightness. The darker areas are coming out-of-plane, and are reflecting light away from the microscope. Parasitic out-ofplane motions in the folded-beam suspension can be minimized through the addition of POLY2 “staples” over POLY1 features or by maximizing the out-of-plane aspect ratio.
B. Mouse zygote survival following lance penetration
The most meaningful measure of the quality and repeatability of the nanoinjector mechanism’s motion is the survival of mouse zygotes following piercing of the zygotic membranes. Ideally, penetration into a zygote occurs along the membrane’s surface-normal vector at the point of penetration. Deviation from this vector, such as in-plane or out-of-plane translation and/or rotation of the lance, will cause tearing of the zygote’s membranes and death of the cell. Zygotes for nanoinjection were harvested from superovulated CD-1 female mice. Details of the superovulation and harvesting protocols employed can be found in Ref. 46 and are not presented here for brevity. Zygotes were either untreated (placed directly into culture), or were nanoinjected fol-
FIG. 13. Optical microscopy images of before (top) and during (bottom) nanoinjection of a mouse zygote.
lowing the protocol outlined in Figure 1. For all of the nanoinjections reported here the lance was charged to 1.5 V prior to penetration into the zygote, and charged to −1.5 V for 10 s during penetration. Nanoinjected and untreated zygotes were then cultured for 24 h and the rate of progression from onecell zygotes to two-cell embryos was recorded. Figure 13 shows the nanoinjector and a zygote before and during lance penetration. Figure 14 presents the proportion of zygotes progressing to the two cell stage from untreated (1917/2407 = 79.6%) and nanoinjected (2134/2968 = 71.9%) groups. The error bars on the proportions were calculated using the logit interval presented in Ref. 47 with a 95% confidence level. The rates of survival for the untreated and nanoinjected embryos are quite similar, differing by 7.7%, and with the confidence intervals separated by 3.5%.
V. DESIGN REFINEMENTS
Observing the nanoinjector’s motion and impact on embryo viability has led to multiple improvements in the nanoinjector design. Figure 14 includes data from both experiments with higher embryo viability and experiments with lower embryo viability. The experiments with lower embryo viability provided several qualitative observations leading to improvements of the nanoinjector. Design refinements were made over the course of seven polyMUMPs production runs
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-9
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
indicates that the nanoinjector consistently fulfilled its design requirement of not tearing the target zygote’s membranes. Refinements to the nanoinjector have improved the mechanism’s performance. The equations and discussion provided here could be used for nanoinjectors for other cell types. Possible areas for improvement may include reducing parasitic motion in the scissor joints and reducing out-of-plane motion by the folded-beam suspension. A fully compliant nanoinjector, with no scissor joints or slider joints, could improve the precision of the lance’s motion by eliminating the parasitic motion caused by the rigid joints.
FIG. 14. Rate of development from single cell embryos (zygotes) to two-cell embryos for untreated (1917/2407 = 79.6%) and nanoinjected (2134/2968 = 71.9%) embryos. Confidence intervals are 95% logit confidence intervals for binomial proportions.
spanning four years. Some of the design improvements have included:
r Changing the lengths of the six-bar mechanism’s links r r
r
r
r r
to be more robust to parasitic motion in the scissor joints. Increasing length of the lance from 75 μm to 200 μm. Changing the lance from POLY1 (2 μm thick) to POLY2 (1.5 μm thick). Decreasing the thickness of the lance decreases the deformation of the cell during lance penetration. Reducing the in-plane taper on the lance from 10◦ , to 6◦ , to 5◦ , and finally to 3◦ . Decreasing the in-plane taper of the lance decreases the deformation of the cell during lance penetration. The lance tip shown in Figure 2(c) demonstrates the fine tip geometry achieved using the 3◦ taper for the mask layout. Rerouting the electrical connections in front of, rather than over, the input slider. This prevents the electrical connections from binding on the fixed portions of the slider. Reducing the width of the input slider to reduce binding. Adding markings to identify the limits of the first subphase’s out-of-plane motion and the second subphase’s in-plane motion.
VI. CONCLUSION
The metamorphic nanoinjector mechanism successfully meets its goal of penetrating the lance into mouse zygotes without causing significant damage to the cellular membranes. Metamorphic self-reconfiguration between the first subphase (the six-bar mechanism’s out-of-plane motion) and the second subphase (the folded-beam suspension’s in plane translation) occurs simply by advancing the mechanism’s input slider. Nanoinjection of nearly 3000 zygotes has resulted in 71.9% progression to the two-cell stage compared to 79.6% for untreated zygotes. This high embryo viability across thousands of actuations and lance penetrations into mouse zygotes
ACKNOWLEDGMENTS
Thanks go to the students of the Brigham Young University Compliant Mechanisms Research Group, and the students of the Brigham Young University Department of Molecular and Microbiology MAFIA Research Laboratory for their continued and consistent assistance throughout this project. The authors recognize the funding support provided by Crocker Ventures LLC and Nanoinjection Technologies LLC. This material is based upon work supported in part by the National Science Foundation under Grant Nos. CMS-0428532, CMMI-0800606, and 1240417. 1 A.
S. Goldstein, J. Huang, C. Guo, I. P. Garraway, and O. N. Witte, “Identification of a cell of origin for human prostate cancer,” Science 329, 568– 571 (2010). 2 S. O. Ryan, A. M. Vlad, K. Islam, J. Gariépy, and O. J. Finn, “Tumorassociated MUC1 glycopeptide epitopes are not subject to self-tolerance and improve responses to MUC1 peptide epitopes in MUC1 transgenic mice,” Biol. Chem. 390, 611–618 (2009). 3 D. A. Gimbel, H. B. Nygaard, E. E. Coffey, E. C. Gunther, J. Laurén, Z. A. Gimbel, and S. M. Strittmatter, “Memory impairment in transgenic Alzheimer mice requires cellular prion protein,” J. Neurosci. 30, 6374– 6367 (2010). 4 L. M. Billings, S. Oddo, K. N. Green, J. L. McGaugh, and F. M. LaFerla, “Intraneuronal Aβ causes the onset of early Alzheimer’s disease-related cognitive deficits in transgenic mice,” Neuron 45, 675–688 (2005). 5 J. F. Aitken, K. M. Loomes, D. W. Scott, S. Reddy, A. R. Phillips, G. Prijic, C. Fernando, S. Zhang, R. Broadhurst, P. L’Huillier, and G. J. Cooper, “Tetracycline treatment retards the onset and slows the progression of diabetes in human amylin/islet amyloid polypeptide transgenic mice,” Diabetes 59, 161–171 (2010). 6 A. W. Chan, G. Kukolj, A. M. Skalka, and R. D. Bremel, “Timing of DNA integration, transgenic mosaicism, and pronuclear microinjection,” Mol. Reprod. Dev. 52, 406–413 (1999). 7 K. Chida, R. Sueyoshi, and T. Kuroki, “Efficient and stable gene transfer following microinjection into nuclei of synchronized animal cells progressing from G1/S boundary to early S phase,” Biochem. Biophys. Res. Commun. 249, 849–852 (1998). 8 X. Zheng, H. K. Surks, and X. Zhang, “A versatile cell contractility mapping transducer utilizing moiré-based technique,” J. Microelectromech. Syst. 19, 764–773 (2010). 9 F. Caselli, P. Bisegna, and F. Maceri, “EIT-inspired microfluidic cytometer for single-cell dielectric spectroscopy,” J. Microelectromech. Syst. 19, 1029–1040 (2010). 10 N. Inomata, T. Mizunuma, Y. Yamanishi, and F. Arai, “Omnidirectional actuation of magnetically driven microtool for cutting of oocyte in a chip,” J. Microelectromech. Syst. 20, 383–388 (2011). 11 J. S. Dai and J. Jones, “Mobility in metamorphic mechanisms of foldable/erectable kinds,” J. Mech. Des. 121, 375 (1999). 12 L. Zhang and J. S. Dai, “Reconfiguration of spatial metamorphic mechanisms,” J. Mech. Rob. 1, 011012-1–011012-8 (2009). 13 L. Zhang, J. S. Dai, and T.-L. Yang, “Reconfiguration techniques and geometric constraints of metamorphic mechanisms,” in 2009 Proceedings of the ASME International Design Engineering Technical Conferences
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-10
Aten et al.
and Computers and Information in Engineering Conference, DETC2009 (ASME, 2009), Vol. 7, pp. 559–575. 14 D. Gan, J. Dai, and Q. Liao, “Constraint analysis on mobility change of a novel metamorphic parallel mechanism,” Mech. Mach. Theory 45, 1864– 1876 (2010). 15 S. Li and J. S. Dai, “Structure synthesis of single-driven metamorphic mechanisms based on the augmented Assur groups,” J. Mech. Rob. 4, 031004 (2012). 16 L. Cui and J. S. Dai, “Reciprocity-based singular value decomposition for inverse kinematic analysis of the metamorphic multifingered hand,” J. Mech. Rob. 4, 034502 (2012). 17 S. E. Wilding, L. Howell, and S. P. Magelby, “Spherical lamina emergent mechanisms,” Mech. Mach. Theory 49, 187–197 (2012). 18 P. S. Gollnick, S. P. Magelby, and L. L. Howell, “An introduction to multilayer lamina emergent mechanisms,” J. Mech. Des. 133, 081006 (2011). 19 J. O. Jacobsen, B. G. Winder, L. L. Howell, and S. P. Magleby, “Lamina emergent mechanisms and their basic elements,” J. Mech. Rob. 2, 011003 (2010). 20 H. C. Greenberg, M. L. Gong, S. P. Magelby, and L. L. Howell, “Identifying links between origami and compliant mechanisms,” Mech. Sci. 2, 217–225 (2011). 21 L. A. Bowen, C. L. Grames, S. P. Magleby, R. J. Lang, and L. L. Howell, “A classification of action origami as systems of spherical mechanisms,” J. Mech. Des. 135, 111008 (2013). 22 J. Ryu, M. D’Amato, X. Cui, K. N. Long, H. J. Qi, and M. L. Dunn, “Photoorigami — bending and folding polymers with light,” Appl. Phys. Lett. 100, 161908 (2012). 23 J. Subirana and M. Soler-Lopez, “Cations as hydrogen bond donors: A view of electrostatic interactions in DNA,” Annu. Rev. Biophys. Biomol. Struct. 32, 27–45 (2003). 24 Q. T. Aten, B. D. Jensen, S. H. Burnett, and L. L. Howell, “Electrostatic accumulation and release of DNA using a micromachined lance,” J. Microelectromech. Syst. 20, 1449–1461 (2011). 25 R. L. Brinster, H. Y. Chen, M. E. Trumbauer, M. K. Yagle, and R. D. Palmiter, “Factors affecting the efficiency of introducing foreign DNA into mice by microinjecting eggs,” Proc. Natl. Acad. Sci. U.S.A. 82, 4438–4442 (1985). 26 R. J. Wall, “Pronuclear microinjection,” Cloning Stem Cells 3, 209–220 (2001). 27 A. Pillarisetti, M. Pekarev, A. D. Brooks, and J. P. Desai, “Evaluating the role of force feedback for biomanipulation tasks,” Proc. Symp. Haptic Interfaces Virtual Environ. Teleoperator Syst. 11–18 (2006). 28 S. Zappe, M. Fish, M. Scott, and O. Solgaard, “Automated MEMS-based drosophila embryo injection system for high-throughput RNAi screens,” Lab Chip 6, 1012–1019 (2006). 29 J. Brachet, T. Kuusi, and S. Gothie, “Comparative study of the inductive power of nucleic acids and of cellular nucleoproteins when they are implanted or given by microinjection,” Arch. Biol. 63, 429–440 (1952).
Rev. Sci. Instrum. 85, 055005 (2014) 30 J.
B. Buck, “A versatile microinjection and micropipetting syringe,” Rev. Sci. Instrum. 20, 676–677 (1949). 31 S. Park, Y.-S. Kim, W. B. Kim, and S. Jon, “Carbon nanosyringe array as a platform for intracellular delivery,” Nano Lett. 9, 1325–1329 (2009). 32 A. K. Shaleka, J. T. Robinsona, E. S. Karpa, J. S. Leea, D.-R. Ahnb, M.-H. Yoona, A. Suttona, M. J. amd Rona S. Gertnera, T. S. Gujrala, G. MacBeatha, E. G. Yanga, and H. Parka, “Vertical silicon nanowires as a universal platform for delivering biomolecules into living cells,” Proc. Natl. Acad. Sci. U.S.A. 107, 1870–1875 (2010). 33 C. M. Cuerrier, R. Lebel, and M. Grandbois, “Single cell transfection using plasmid decorated AFM probes,” Biochem. Biophys. Res. Commun. 355, 632–636 (2007). 34 X. Chen, A. Kis, A. Zettl, and C. R. Bertozzi, “A cell nanoinjector based on carbon nanotubes,” Proc. Natl. Acad. Sci. 104, 8218–8222 (2007). 35 J. Carter, A. Cowen, B. Hardy, R. Mahadevan, M. Stonefield, and S. Wilcenski, PolyMUMPs Design Handbook, Revision 11 (MEMSCAP Inc., 2005). 36 B. L. Boyce, J. M. Grazier, T. E. Buchheit, and M. J. Shaw, “Strength distributions in polycrystalline silicon MEMS,” J. Microelectromech. Syst. 16, 179–190 (2007). 37 R. Yeh, E. J. Kruglick, and K. S. Pister, “Surface-micromachined components for articulated microrobots,” J. Microelectromech. Syst. 5, 10–17 (1996). 38 A. Friedberger and R. S. Muller, “Improved surface-micromachined hinges for fold-out structures,” J. Microelectromech. Syst. 7, 315–319 (1998). 39 G. H. Teichert, Q. T. Aten, S. H. Burnett, L. L. Howell, and B. D. Jensen, “Cylindrical single-degree-of-freedom spatial mechanisms for cell restraint,” J. Mech. Rob. 4, 021011 (2012). 40 K. A. Jensen, C. P. Lusk, and L. L. Howell, “An XYZ micromanipulator with three translational degrees of freedom,” Robotica 24, 305–314 (2006). 41 B. G. Winder, S. P. Magelby, and L. L. Howell, “Kinematic representations of pop-up paper mechanisms,” J. Mech. Rob. 1, 021009 (2009). 42 L. Howell, Compliant Mechanisms (Wiley, New York, 2001). 43 S. Oak, G. F. Edmiston, G. Sivakumar, and T. Dallas, “Rotating out-ofplane micromirror,” J. Microelectromech. Syst. 19, 632–639 (2010). 44 Q. T. Aten, B. D. Jensen, S. Tamowski, A. M. Wilson, L. L. Howell, and S. H. Burnett, “Nanoinjection: Pronuclear DNA delivery using a charged lance,” Transgenic Res. 21, 1279–1290 (2012). 45 A. M. Wilson, Q. T. Aten, N. C. Toone, J. L. Black, B. D. Jensen, S. Tamowski, L. L. Howell, and S. H. Burnett, “Transgene delivery via intracellular electroporetic nanoinjection,” Transgenic Res. 22, 993–1002 (2013). 46 A. Nagy, M. Gertsenstein, K. Vintersten, and R. Behringer, Manipulating the Mouse Embryo: A Laboratory Manual, 3rd ed. (Cold Spring Harbor Laboratory Press, 2003). 47 L. D. Brown, T. T. Cai, and A. DasGupta, “Interval estimation for a binomial proportion,” Statistical Science 16, 101–133 (2001).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
REVIEW OF SCIENTIFIC INSTRUMENTS 85, 055005 (2014)
A self-reconfiguring metamorphic nanoinjector for injection into mouse zygotes Quentin T. Aten,1 Brian D. Jensen,2 Sandra H. Burnett,3 and Larry L. Howell2 1
Nexus Spine, LLC, Salt Lake City, Utah 84124, USA Department of Mechanical Engineering, Brigham Young University, Provo, Utah 84602, USA 3 Department of Microbiology and Molecular Biology, Brigham Young University, Provo, Utah 84602, USA 2
(Received 14 February 2014; accepted 9 April 2014; published online 13 May 2014) This paper presents a surface-micromachined microelectromechanical system nanoinjector designed to inject DNA into mouse zygotes which are ≈90 μm in diameter. The proposed injection method requires that an electrically charged, DNA coated lance be inserted into the mouse zygote. The nanoinjector’s principal design requirements are (1) it must penetrate the lance into the mouse zygote without tearing the cell membranes and (2) maintain electrical connectivity between the lance and a stationary bond pad. These requirements are satisfied through a two-phase, self-reconfiguring metamorphic mechanism. In the first motion subphase a change-point six-bar mechanism elevates the lance to ≈45 μm above the substrate. In the second motion subphase, a compliant folded-beam suspension allows the lance to translate in-plane at a constant height as it penetrates the cell membranes. The viability of embryos following nanoinjection is presented as a metric for quantifying how well the nanoinjector mechanism fulfills its design requirements of penetrating the zygote without causing membrane damage. Viability studies of nearly 3000 nanoinjections resulted in 71.9% of nanoinjected zygotes progressing to the two-cell stage compared to 79.6% of untreated embryos. © 2014 AIP Publishing LLC. [http://dx.doi.org/10.1063/1.4872077] I. INTRODUCTION
The purpose of this work is to describe the development of a system capable of electro-physical injection of DNA into mouse zygotes (single cell embryos consisting of a fertilized egg). The objectives include achieving high cell survival, producing specified motion at the single-cell-size scale, and designing within the constraints of micro-fabrication processes. This paper describes the design, numerical modeling, in vitro biological testing, and refinement of a metamorphic nanoinjector instrument. Its origami-like folding motion begins in the fabrication plane and has motion out of that plane. Design parameters and fabrication and test results are provided for a nanoinjector specifically designed to inject mouse zygotes, and general design equations are provided to facilitate the design of nanoinjectors for other cell types. Insertion of foreign genetic material into the genome of mice and other mammals is a critical step in understanding and utilizing genes’ functions. Genetically, modified (transgenic) animals are currently used in research into such varied topics as cancer,1, 2 Alzheimer’s disease,3, 4 and diabetes.5 To genetically modify every cell within an animal, the genetic construct of interest (transgene) must be present in at least one pronucleus (the membrane surrounding each set of parental chromosomes) before the zygote’s first division.6, 7 In a mechanical sense, direct transgene delivery to a zygote requires precise motion across tens of micrometers for pronuclear penetration. Microelectromechanical system (MEMS) devices are potentially well suited to introduce genetic material into developing embryos because of their relatively small scale, potential for complex motion, and their ability to mechanically and electrically interact with individual cells.8–10 Addition0034-6748/2014/85(5)/055005/10/$30.00
ally, MEMS fabrication techniques allow for features within a single device to be as small as tens of nanometers, or as large as hundreds of micrometers. The nanoinjector’s metamorphic behavior11–16 enables it to first elevate to the cell mid-height and then pierce the cell membrane with high rates of cell survival. Because micro-fabrication constraints limit the as-fabricated position to be parallel to the substrate and the motion comes out of the substrate, the mechanism is related to lamina emergent mechanisms17–19 and origami-inspired mechanisms.20–22 Viability testing of nearly 3000 zygotes following penetration by the nanoinjector lance is presented to quantify how well the nanoinjector meets its design requirement of penetrating the zygote without permanently damaging its membranes. II. BACKGROUND A. Nanoinjection
The nanoinjector applies nanometer-scale features, precise mechanical motion over tens of microns, and electrical manipulation of DNA in an electro-physical method of gene transfer called nanoinjection. Figure 1 graphically outlines the nanoinjection process. The nanoinjector mechanism is operated while submerged in a pH buffered solution (such as phosphate buffered saline, PBS). A positive electrical charge is applied to the lance, which accumulates negatively charged DNA23 on its surface.24 The nanoinjector mechanism then penetrates the zygotic membranes, and a negative charge is applied to the lance, releasing the accumulated DNA within the cell. In vitro experiments on mouse embryos have demonstrated that nanoinjection can electro-physically transport
85, 055005-1
© 2014 AIP Publishing LLC
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-2
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
(a) SEM of a nanoinjector
FIG. 1. Schematic representation of nanoinjection of DNA into a mouse zygote. All nanoinjections occur with the nanoinjector submerged in a pH buffered saline solution. Step 1, the nanoinjector is in its as-fabricated position. Step 2, the lance is elevated and a positive charge is applied, accumulating DNA on the lance’s tip. Step 3, the lance moves at a constant height, penetrating into the target zygote. Step 4, the charge on the lance is reversed, releasing DNA into the zygote. Step 5, the lance moves at a constant elevation out of the zygote.
genetic material into embryos, and that the nanoinjected genes can be expressed by developing embryos.24
(b) Labeled SEM of a nanoinjector
B. Current methods of direct gene delivery
Currently, direct microinjection of genetic material into mammalian zygotes uses a hollow needle (a micron-scale tapered glass micropipette) driven by a micromanipulator to penetrate the zygotic membranes and a pump to expel minute volumes of a nucleic acid solution into the zygote.25, 26 Though significant improvements to microinjection equipment have been made,27, 28 the core elements of the needleand-pump design paradigm have remained essentially unchanged since microinjection hardware first appeared in the literature over 50 years ago.29, 30 Breaking from the needle-and-pump concept, MEMS DNA injection systems have been developed for adherent culture cells. These methods deliver DNA via fixed vertical nano-needles on a chip,31, 32 or modified atomic-force microscopy (AFM) probes.33, 34 The fixed nano-needles require that the target cells grow with the needle’s penetrating their surface, making fixed vertical needles impractical for transgenesis.31, 32 The AFM probe based needles require several minutes of incubation in each cell, and do not have sufficient penetration into the target cell to reach the zygote’s pronucleus.33, 34 III. INSTRUMENT DESCRIPTION AND MODELING
As a means of delivering transgenes into the zygote’s pronucleus, the nanoinjector mechanism shown in Figure 2
(c) SEM of the lance tip FIG. 2. Labeled scanning electron microscope (SEM) images of (a) the metamorphic nanoinjector posed as if it were injecting a 100 μm latex sphere, (b) the metamorphic nanoinjector in its as-fabricated position, and (c) the tip of the nanoinjector lance. The two six-bar mechanisms restraining the latex sphere in (a) are not described in this paper.
has three main functions: first, elevate the lance from its as-fabricated position on the fabrication substrate; second, move the lance horizontally at a constant height, penetrating the lance into the cell; and third, maintain electrical conductivity between the lance and an external voltage source. Constructing the nanoinjector as a self-reconfiguring metamorphic mechanism11–14 enables the lance elevation to be sequentially decoupled from the purely horizontal lance motion. Decoupling the elevation and purely horizontal motion phases ensures that the lance penetrates the zygote’s membranes and pronucleus along a linear, horizontal axis, and prevents tearing of the zygotic membranes.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-3
Aten et al.
The nanoinjector mechanism consists of one rigid-body six-bar mechanism, one compliant parallel-guiding mechanism, and two compliant electrical connections, as shown in Figure 2(b). The rigid-body six-bar mechanism provides the out-of-plane displacement in the nanoinjector’s first metamorphic subphase. The compliant parallel-guiding mechanism provides the in-plane translation toward the target cell during the second subphase. The method of self-reconfiguration and models of each of the nanoinjector’s components are described below. Prototype nanoinjectors were fabricated using MEMSCAP Inc.’s polycrystalline silicon Multiuser MEMS Processes (polyMUMPs).35 The process provides one stationary polycrystalline silicon layer (POLY0), two structural layers of polycrystalline silicon (POLY1 and POLY2), and a gold layer for increasing electrical conductivity which may be added to the POLY2 layer. The POLY1 and POLY2 layers are 2.0 μm and 1.5 μm thick, respectively. This well established process was selected for developing the nanoinjector mechanism because of the highly controlled and repeatable nature of the process from fabrication run to fabrication run. This allowed the development work to focus solely on the design and biological testing of the nanoinjector and helped mitigate deviceto-device variability caused by fabrication variability. The fabrication process is capable of producing parts with a large range in feature size. For example, the tip of the nanoinjector lance in Figure 2(c) has a minimum in-plane width of only 17 nm, while the mechanism’s overall dimensions are four orders of magnitude larger, as seen in Figure 2(b). Polycrystalline silicon has an ultimate strength of approximately between 1 GPa and 3 GPa, depending on the processing conditions and gage area.36 The ultimate strength of POLY1 and POLY2 were assumed to lie near the bottom of this distribution at 1.5 GPa.
Rev. Sci. Instrum. 85, 055005 (2014)
Link A
Link B
(c) schematic of a scissor joint
(a) SEM of a scissor joint
Link A
Link B
(d) schematic of a slider joint (b) SEM of a slider joint
FIG. 3. Scanning electron microscope (SEM) images of the two types of outof-plane joints used in the nanoinjector: (a) scissor joints and (b) slider joints. Schematics of each joint type are shown in (c) and (d).
layer, the minimum sizes of features, and between-layer vias. Figure 4 shows two sources of parasitic motion in the scissor joint: one governed by minimum feature size, and the other governed by minimum gaps between features in the same
A. Out-of-plane surface micromachined mechanisms
Surface micromachined mechanisms can achieve outof-plane displacements many times greater than their asfabricated thickness through the use of specially designed joints.37–39 The out-of-plane revolute joints used in the nanoinjector can be categorized as “scissor joints” and “slider joints.” Examples of each of these joint types are shown in Figure 3. The slider joints can undergo rotations of approximately 180◦ , and the scissor joints can achieve rotations in excess of 90◦ . These joints can be combined to create complex mechanisms, such as those in Ref. 40. These joints can be fabricated from planar layers of material but enable motion out of the fabrication plane. The motion is similar to origami or pop-up mechanisms, which begin with a flat sheet of paper and out-of-plane motion occurs at folds or creases in the paper.20, 41 The MEMS scissor and slider joints can be viewed as replacements for origami creases in paper, and they allow origami-like motions to occur. The limitations of multi-layer surface micromachining processes have a significant impact on the precision of scissor joint’s motion. Parasitic motion is inherent in the joints because of the minimum gaps between features in the same
( ) (a)
(b) FIG. 4. Scanning electron micrographs of the scissor joints in the nanoinjector mechanism showing (a) parasitic motion due to minimum feature size and (b) minimum gaps between features in the same layer.
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-4
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
FIG. 6. Kinematic diagram of the change-point six-bar mechanism. The location a where the electrical connection attaches to link R3 and the loads exerted by the electrical connection Fy , Fz , and Mx are also given. θ 3 is the angle of link 3 (in the case shown, θ 3 = 0 and is not shown).
FIG. 5. Schematic illustration of the nanoinjector’s motion. The folded-beam suspension’s suspension is approximated as the linear spring at left. At position 0, the nanoinjector is in its as-fabricated configuration. Between positions 0 and 1, the nanoinjector is in its first subphase, and between positions 1 and 2, the nanoinjector is in its second subphase.
layer. The kinematic models of the nanoinjector presented below account for these two types of parasitic motion.
B. Self-reconfiguration through unequal subphase mechanism stiffnesses and link contact
The nanoinjector’s sequential “up-then-forward” motion is a key element of the mechanism’s functionality. Selfreconfiguration in the nanoinjector is a consequence of the unequal force-displacement characteristics of the six-bar mechanism and the folded-beam suspension, and contact between links in the nanoinjector mechanism. The nanoinjector’s motion is shown schematically in Figure 5, and a kinematic diagram of the six-bar mechanism is shown in Figure 6. At position 0, the nanoinjector is in its as-fabricated configuration, with the six-bar mechanism mounted on to the folded-beam suspension. In other words, it is as if the six-bar mechanism is kinematically grounded to the folded-beam suspension rather than the substrate. Between positions 0 and 1, the nanoinjector is in its first subphase, with the lance moving along a kinematic path with both out-of-plane and in-plane components. This motion raises the lance to the desired level, but occurs away from the zygote to prevent damage to the cell membrane. Assuming negligible friction in the six-bar’s slider and scissor joints, actuation between positions 0 and 1 applies approximately zero force to the folded-beam suspension. With negligibly small force applied to the folded-beam suspension by the six-bar mechanism, there is no additional in-plane translation from
the deflection of the folded-beam suspension’s compliant flexures. At position 1, contact is made between the input slider and the folded-beam suspension. This effectively locks the prismatic (slider) joint shown in Figure 6 and fixes the length R6 . With this degree of freedom removed, the six-bar mechanism becomes a structure with respect to the fold-beam suspension. Between positions 1 and 2, the six-bar mechanism remains at the same elevation as position 1 while the foldedbeam suspension deflects, resulting in an in-plane translation of the lance. This in-plane translation allows a linear motion of the raised lance such that minimal damage to the cell membrane occurs during lance penetration. Pulling back on the input slider, the mechanism will proceed from position 2, to 1, back to 0.
C. Modeling the six-bar mechanism
If the scissor and slider joints in the nanoinjector six-bar mechanism are treated as idealized revolute joints, the mechanism can be modeled as shown in Figure 6. Due to the planar nature of surface micromachining, the mechanism is fabricated in a change-point configuration, with all of its links co-planar. However, the mechanism can achieve only the configuration pictured because the other kinematic configurations are only possible if one or more links move in the negative zdirection (through the substrate). In the nanoinjector, links R1 and R3 are the same length, as are links R2 and R4 . Thus, the mechanism can be modeled as a parallel-guiding (parallelogram) four-bar mechanism (links R1 through R4 ) with a driver dyad (R5 and the input slider). The parallel-guiding motion of the six-bar mechanism ensures that the lance will be horizontal through out its motion. The position of the mechanism can be calculated by θ4 = cos−1
(R4 + R5 − Yin )2 + R42 − R52 , 2R4 (R4 + R5 − Yin )
θ5 = 2π + sin−1 (R4 sin θ4 /R5 ),
(1)
(2)
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-5
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
θin = θ5 − π,
(3)
Zout = R4 sin θ4 ,
(4)
Yout = R4 (1 − cos θ4 ),
(5)
(a)
where θ 2 = θ 4 because of the parallelogram configuration. Idealizing the scissor joints as pure revolute joints can be a poor assumption given the amount off parasitic motion that is possible (see Figure 4). The parasitic motion in the scissor joints on links R2 , R4 , and R5 can approximately be modeled through adjusting the lengths of links, and the input displacement Yin by R2 = R2 + Pl ,
(6)
R4 = R4 + Pl ,
(7)
R5 = R5 + Pl ,
(8)
Yin = Yin + 3Pg ,
(9)
(b)
where Pl is the change in link length possible in the scissor joint (see Figure 4(a)), and Pg is the gap between the links connected by a scissor joint when the mechanism is in its asfabricated position (see Figure 4(b)). The factor Pg is multiplied by 3 to account for each of the three unique scissor joints in the mechanism: R2 to R3 , R3 to R5 , and R4 to R5 . The dimensions given in Table I were chosen to reduce the effects of parasitic motion on the lance’s final out-of-plane displacement (Zout ). Figure 7(a) shows kinematic models of the nanoinjector both with and without accounting for the parasitic motion in the scissor joints. The input force required to actuate the nanoinjector mechanism, Fin , is a function of the loads generated by the electrical connection during the first and second motion subphases and of the folded-beam suspension in the second moTABLE I. Dimensions of the lance six-bar mechanism shown in Figure 2(b). Link lengths and positions given correspond to the kinematic diagram in Figure 6. Link lengths and positions given in parenthesis take into account the parasitic motion given by Eq. (6)–(9). All dimensions are in μm unless otherwise noted. Link or parameter R1 = R3 R2 = R4 R5 Yin Zout Yout Lance length Lance taper Pl Pg
Value 90.0 50.0 (66.5) 120.0 37.0 (28.0) 45.0 (47.1) 28.2 (19.6) 200.0 3◦ 16.5 3.0
FIG. 7. The effect of parasitic motion on the final out-of-plane displacement (Zout ) is shown through (a) the kinematic diagram of the change-point six-bar mechanism, and through (b) Zout as a function between the input displacement (Yin ) for both the ideal mechanism with no parasitic motion (solid line), and with maximum parasitic motion (dotted line).
tion subphase, Fin = −
Fy (cos(θ3 − θ5 ) + cos(θ3 + θ5 )) 2 sin(θ3 + θ5 )
−
Fz (sin(θ3 − θ5 ) + sin(θ3 + θ5 )) 2 sin(θ3 + θ5 )
−
(Mx + aFz )(sin(θ2 − θ3 + θ5 ) + sin(θ2 − θ3 − θ5 )) R4 (cos(θ2 + θ3 + θ5 ) − cos(θ2 − θ3 − θ5 ))
−Ff b ,
(10)
where Ffb is the load applied to the input slide by the foldedbeam suspension, and Fy , Fz , and Mx are the loads exerted by the electrical connection (see Figure 6). Each angle θ i represents the angle of link i, as illustrated in Figure 6 (except θ 3 is not shown because the angle of link 3 is zero for the position shown). The term Ffb is zero until contact is made between the input slider and the folded-beam suspension. In other words, the folded-beam suspension does not contribute to the loading on the input slider until the second motion subphase. D. Modeling the folded-beam suspension
The fully compliant folded-beam suspension was analyzed using a psuedo-rigid-body model (PRBM) and nonlinear finite element analysis. The eight fixed-guided compliant flexures are 300 μm long, 5 μm wide, and 2 μm thick (POLY1). As shown in Figure 8, the PRBM of a fixedguided beam consists of one pseudo-rigid link and two
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-6
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
FIG. 8. A schematic of a fixed-guided beam and its pseudo-rigid-body model.
pseudo-springs. The input force, P, and maximum stress in the flexure, σ max , can be found by = arcsin
b , γl
α 2 = 2K ,
(11)
(12)
P = 4α 2 EI / l 2 ,
(13)
a = l(1 − γ (1 − cos )),
(14)
P a( h2 ) , (15) 2I where b and a are the y and x coordinates of the end of the beam, E is Young’s modulus for polysilicon (≈164 GPa), I is the second moment of area for a rectangular beam, γ = 0.8517, and K = 2.65 (see Ref. 42). The folded-beam suspension’s eight compliant flexures are in a parallel and series arrangement. The right and left halves of the suspension are in parallel. Within each half, the pair of outer flexures is in series with the pair of inner flexures. Additionally, the two outer flexures are in parallel with each other, and the two inner flexures are in parallel with each other. Summing the parallel and series stiffnesses and displacements, the full suspension has the net-stiffness of four parallel-guiding flexures which displace one-half the total displacement of the suspension. The calculated input force and maximum stress predicted by the pseudo-rigid-body model for a total stage displacement of 70 μm (each flexure displacing 35 μm) are given in Table II. In addition to the pseudo-rigid body model, a nonlinear (large deflection, small strain) finite element analysis (FEA) was performed on the folded-beam suspension using ANSYS 11.0. The analysis modeled the geometrically nonlinear (small strain, large rotation) deformation of one-half of σmax =
TABLE II. Comparison of pseudo-rigid-body model and finite element analyses of the folded-beam suspension.
Force total Maximum stress
PRBM
FEA
222 μN 993 MPa
199 μN 1334 MPa
FIG. 9. Finite element model Von Mises stress contours for the folded-beam suspension undergoing 70 μm of displacement in the y-direction. The two areas indicated in red cross-hatching were fixed in all degrees of freedom. The y-direction input displacement was applied to the surfaces indicated. Stresses are in MPa.
the folded-beam suspension with 7488 second-order shell elements. The compliant flexures were meshed with uniformly distributed quadratic elements, and the rigid portions were meshed with larger tetrahedral elements. The model was fixed at two anchor areas, had symmetry boundary conditions applied along the y-z plane, and displaced 70 μm in-plane at the point of contact between the folded-beam suspension and the input slider, as shown in Figure 9. The model was evaluated in 20 uniformly distributed load steps. In post processing, the reaction force on the displaced face was calculated from the nodal results at each step. The input force and the maximum stresses predicted by the FEA are compared with the PRBM results in Table II. The FEA results and the PRBM results agree to within 11% for the input force, and 25% for the maximum stress. The higher stress predicted by the finite element model is a result of the stress concentration at the right angle between the compliant flexures and the rigid portion of the suspension (see the inset in Figure 9). The stresses predicted by both models are below the ultimate strength of polycrystalline silicon.36
E. Modeling the compliant electrical connections
The serpentine flexure43 electrical connections between the nanoinjector and the stationary bond-pads are designed to provide electrical conductivity without interfering with the nanoinjector’s motion. The electrical connections are 1.5 μm thick, 5 μm wide, and consist of ten 250 μm long flexures, connected by nine 50 μm long segments in a serpentine pattern (see Figure 2(b)). Only a finite element analysis of the electrical connections was performed because there is no closed-form or pseudo-rigid body model for the serpentine flexures’ nonlinear three-dimensional loading. One end of the electrical
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-7
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014) TABLE III. Reaction forces applied to the lance mechanism by one or both electrical connections for a displacement of 45 μm in the z-direction (out-ofplane) and 107 μm in the y-direction (in-plane). Note that the x-direction of force, y-direction moment, and z-direction moment sum to zero because of symmetry. All forces are in μN, and all moments are in μN μm.
Fx Fy Fz Mx My Mz
FIG. 10. Finite element model Von Mises stress contours for the electrical connections undergoing 45 μm in the z-direction (out-of-plane) and 107 μm in the y-direction (in-plane). Stresses are in MPa.
connection was fixed in all degrees of freedom, while the other was displaced along the kinematic path following the point a on the six-bar mechanism (Figure 6) for a total displacement of 45 μm in the z-direction (out-of-plane) and 107 μm in the y-direction (in-plane), as indicated in Figure 10. Initially, the finite element model only included the electrical connection. However, this initial model predicted significant displacements into the substrate (negative z-direction) by part of the electrical connection. To more accurately predict the electrical connection’s final configuration and stress state, the interaction between the electrical connection and the substrate was modeled as an initially open contact pair, with the substrate being modeled as perfectly rigid. The simulation modeled the geometrically nonlinear (small strain, large rotation) deformation of the electrical connector with 5964 second-order, quadrilateral shell elements. The contact pair was modeled as a node-to-rigid surface contact pair with 5964 8-node node-to-surface contact elements on the electrical connector, and 9180 3-node tetrahedral target elements on the substrate. The model was evaluated in 30 load steps along the displaced end’s path. In post processing, the reaction forces and moments on the displaced face were calculated from the nodal results at each step. Figure 10 shows the deformed shape and Von Mises stress contours of the fully displaced electrical connection. Table III gives the reaction forces and moments for one and both electrical connections. The reaction forces are small in comparison to the total forces required to displace the compliant stage, and in testing have not interfered with the nanoinjector’s motion.
One
Both
±2.21 − 3.97 0.22 100.51 ±16.18 ±790.87
0 − 7.94 0.44 201.02 0 0
beam suspension were substituted into Eq. (10) to predict the total force displacement behavior of the mechanism. The force displacement characteristics of the full mechanism are illustrated in Figure 11. As shown in Figure 11, the electrical connections are responsible for the input force in the first motion subphase (up to 37 μm of input displacement). In the second motion subphase, at input displacements greater than 37 μm, the input slide contacts the folded-beam suspension, and the deformation of the folded beam suspension becomes the dominant load on the nanoinjector mechanism. IV. TESTING OF PROTOTYPE NANOINJECTORS
This section provides verification of the nanoinjector’s metamorphic motion through mechanical testing of the nanoinjector. The mechanism’s metamorphic motion is demonstrated, and embryo survival results are presented for nanoinjections into mouse zygotes (single-cell fertilized egg cells). The embryo survival data are especially important because these demonstrate that the nanoinjector’s metamorphic motion allows the lance to consistently penetrate the zygote’s membranes without tearing or damaging the membranes. Full nanoinjection protocols and data on successful nanoinjection of DNA into mouse embryos are presented in other literature where the biological experiments can be discussed in detail.44, 45
F. Modeling the total input force
The reaction forces and moments predicted by the finite element models of the electrical connections and the folded-
FIG. 11. Total input force (Fin ) as a function of total input displacement (Yin ).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-8
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
FIG. 12. Top view optical microscopy images of the nanoinjector’s motion.
A. Verification of metamorphic motion
The nanoinjector mechanism is actuated by applying a linear input with a tungsten micro-probe attached to a manually operated micromanipulator (Cascade Microtech model number MDL-RMS). The nanoinjector is operated under displacement control, with the micromanipualtor providing sufficient input force to actuate the nanoinjector through its full range of motion. Figure 12 shows optical microscopy images of the nanoinjector mechanism in its un-actuated position, its first metamorphic subphase (out-of-plane elevation of the lance), and its second subphase (in-plane translation). As described previously, the six-bar mechanism reaches its full out-ofplane elevation with negligible translation in the folded-beam suspension. Applying further actuation, the lance remains at a constant height above the substrate as the folded-beam suspension deforms. Figure 12 also subtly illustrates the potential for parasitic out-of-plane motion inherent in thin film compliant mechanisms. At best, the polyMUMPs process can offer out-ofplane aspect ratios ≤1.167, leading to compliant flexures with out-of-plane stiffnesses which may be less than or equal to their in-plane stiffnesses. Any eccentricity in the applied inplane loads may excite these out-of-plane displacements. A careful examination of the folded-beam suspension in the top view images in Figure 12 reveals changes in the suspension’s brightness. The darker areas are coming out-of-plane, and are reflecting light away from the microscope. Parasitic out-ofplane motions in the folded-beam suspension can be minimized through the addition of POLY2 “staples” over POLY1 features or by maximizing the out-of-plane aspect ratio.
B. Mouse zygote survival following lance penetration
The most meaningful measure of the quality and repeatability of the nanoinjector mechanism’s motion is the survival of mouse zygotes following piercing of the zygotic membranes. Ideally, penetration into a zygote occurs along the membrane’s surface-normal vector at the point of penetration. Deviation from this vector, such as in-plane or out-of-plane translation and/or rotation of the lance, will cause tearing of the zygote’s membranes and death of the cell. Zygotes for nanoinjection were harvested from superovulated CD-1 female mice. Details of the superovulation and harvesting protocols employed can be found in Ref. 46 and are not presented here for brevity. Zygotes were either untreated (placed directly into culture), or were nanoinjected fol-
FIG. 13. Optical microscopy images of before (top) and during (bottom) nanoinjection of a mouse zygote.
lowing the protocol outlined in Figure 1. For all of the nanoinjections reported here the lance was charged to 1.5 V prior to penetration into the zygote, and charged to −1.5 V for 10 s during penetration. Nanoinjected and untreated zygotes were then cultured for 24 h and the rate of progression from onecell zygotes to two-cell embryos was recorded. Figure 13 shows the nanoinjector and a zygote before and during lance penetration. Figure 14 presents the proportion of zygotes progressing to the two cell stage from untreated (1917/2407 = 79.6%) and nanoinjected (2134/2968 = 71.9%) groups. The error bars on the proportions were calculated using the logit interval presented in Ref. 47 with a 95% confidence level. The rates of survival for the untreated and nanoinjected embryos are quite similar, differing by 7.7%, and with the confidence intervals separated by 3.5%.
V. DESIGN REFINEMENTS
Observing the nanoinjector’s motion and impact on embryo viability has led to multiple improvements in the nanoinjector design. Figure 14 includes data from both experiments with higher embryo viability and experiments with lower embryo viability. The experiments with lower embryo viability provided several qualitative observations leading to improvements of the nanoinjector. Design refinements were made over the course of seven polyMUMPs production runs
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-9
Aten et al.
Rev. Sci. Instrum. 85, 055005 (2014)
indicates that the nanoinjector consistently fulfilled its design requirement of not tearing the target zygote’s membranes. Refinements to the nanoinjector have improved the mechanism’s performance. The equations and discussion provided here could be used for nanoinjectors for other cell types. Possible areas for improvement may include reducing parasitic motion in the scissor joints and reducing out-of-plane motion by the folded-beam suspension. A fully compliant nanoinjector, with no scissor joints or slider joints, could improve the precision of the lance’s motion by eliminating the parasitic motion caused by the rigid joints.
FIG. 14. Rate of development from single cell embryos (zygotes) to two-cell embryos for untreated (1917/2407 = 79.6%) and nanoinjected (2134/2968 = 71.9%) embryos. Confidence intervals are 95% logit confidence intervals for binomial proportions.
spanning four years. Some of the design improvements have included:
r Changing the lengths of the six-bar mechanism’s links r r
r
r
r r
to be more robust to parasitic motion in the scissor joints. Increasing length of the lance from 75 μm to 200 μm. Changing the lance from POLY1 (2 μm thick) to POLY2 (1.5 μm thick). Decreasing the thickness of the lance decreases the deformation of the cell during lance penetration. Reducing the in-plane taper on the lance from 10◦ , to 6◦ , to 5◦ , and finally to 3◦ . Decreasing the in-plane taper of the lance decreases the deformation of the cell during lance penetration. The lance tip shown in Figure 2(c) demonstrates the fine tip geometry achieved using the 3◦ taper for the mask layout. Rerouting the electrical connections in front of, rather than over, the input slider. This prevents the electrical connections from binding on the fixed portions of the slider. Reducing the width of the input slider to reduce binding. Adding markings to identify the limits of the first subphase’s out-of-plane motion and the second subphase’s in-plane motion.
VI. CONCLUSION
The metamorphic nanoinjector mechanism successfully meets its goal of penetrating the lance into mouse zygotes without causing significant damage to the cellular membranes. Metamorphic self-reconfiguration between the first subphase (the six-bar mechanism’s out-of-plane motion) and the second subphase (the folded-beam suspension’s in plane translation) occurs simply by advancing the mechanism’s input slider. Nanoinjection of nearly 3000 zygotes has resulted in 71.9% progression to the two-cell stage compared to 79.6% for untreated zygotes. This high embryo viability across thousands of actuations and lance penetrations into mouse zygotes
ACKNOWLEDGMENTS
Thanks go to the students of the Brigham Young University Compliant Mechanisms Research Group, and the students of the Brigham Young University Department of Molecular and Microbiology MAFIA Research Laboratory for their continued and consistent assistance throughout this project. The authors recognize the funding support provided by Crocker Ventures LLC and Nanoinjection Technologies LLC. This material is based upon work supported in part by the National Science Foundation under Grant Nos. CMS-0428532, CMMI-0800606, and 1240417. 1 A.
S. Goldstein, J. Huang, C. Guo, I. P. Garraway, and O. N. Witte, “Identification of a cell of origin for human prostate cancer,” Science 329, 568– 571 (2010). 2 S. O. Ryan, A. M. Vlad, K. Islam, J. Gariépy, and O. J. Finn, “Tumorassociated MUC1 glycopeptide epitopes are not subject to self-tolerance and improve responses to MUC1 peptide epitopes in MUC1 transgenic mice,” Biol. Chem. 390, 611–618 (2009). 3 D. A. Gimbel, H. B. Nygaard, E. E. Coffey, E. C. Gunther, J. Laurén, Z. A. Gimbel, and S. M. Strittmatter, “Memory impairment in transgenic Alzheimer mice requires cellular prion protein,” J. Neurosci. 30, 6374– 6367 (2010). 4 L. M. Billings, S. Oddo, K. N. Green, J. L. McGaugh, and F. M. LaFerla, “Intraneuronal Aβ causes the onset of early Alzheimer’s disease-related cognitive deficits in transgenic mice,” Neuron 45, 675–688 (2005). 5 J. F. Aitken, K. M. Loomes, D. W. Scott, S. Reddy, A. R. Phillips, G. Prijic, C. Fernando, S. Zhang, R. Broadhurst, P. L’Huillier, and G. J. Cooper, “Tetracycline treatment retards the onset and slows the progression of diabetes in human amylin/islet amyloid polypeptide transgenic mice,” Diabetes 59, 161–171 (2010). 6 A. W. Chan, G. Kukolj, A. M. Skalka, and R. D. Bremel, “Timing of DNA integration, transgenic mosaicism, and pronuclear microinjection,” Mol. Reprod. Dev. 52, 406–413 (1999). 7 K. Chida, R. Sueyoshi, and T. Kuroki, “Efficient and stable gene transfer following microinjection into nuclei of synchronized animal cells progressing from G1/S boundary to early S phase,” Biochem. Biophys. Res. Commun. 249, 849–852 (1998). 8 X. Zheng, H. K. Surks, and X. Zhang, “A versatile cell contractility mapping transducer utilizing moiré-based technique,” J. Microelectromech. Syst. 19, 764–773 (2010). 9 F. Caselli, P. Bisegna, and F. Maceri, “EIT-inspired microfluidic cytometer for single-cell dielectric spectroscopy,” J. Microelectromech. Syst. 19, 1029–1040 (2010). 10 N. Inomata, T. Mizunuma, Y. Yamanishi, and F. Arai, “Omnidirectional actuation of magnetically driven microtool for cutting of oocyte in a chip,” J. Microelectromech. Syst. 20, 383–388 (2011). 11 J. S. Dai and J. Jones, “Mobility in metamorphic mechanisms of foldable/erectable kinds,” J. Mech. Des. 121, 375 (1999). 12 L. Zhang and J. S. Dai, “Reconfiguration of spatial metamorphic mechanisms,” J. Mech. Rob. 1, 011012-1–011012-8 (2009). 13 L. Zhang, J. S. Dai, and T.-L. Yang, “Reconfiguration techniques and geometric constraints of metamorphic mechanisms,” in 2009 Proceedings of the ASME International Design Engineering Technical Conferences
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
055005-10
Aten et al.
and Computers and Information in Engineering Conference, DETC2009 (ASME, 2009), Vol. 7, pp. 559–575. 14 D. Gan, J. Dai, and Q. Liao, “Constraint analysis on mobility change of a novel metamorphic parallel mechanism,” Mech. Mach. Theory 45, 1864– 1876 (2010). 15 S. Li and J. S. Dai, “Structure synthesis of single-driven metamorphic mechanisms based on the augmented Assur groups,” J. Mech. Rob. 4, 031004 (2012). 16 L. Cui and J. S. Dai, “Reciprocity-based singular value decomposition for inverse kinematic analysis of the metamorphic multifingered hand,” J. Mech. Rob. 4, 034502 (2012). 17 S. E. Wilding, L. Howell, and S. P. Magelby, “Spherical lamina emergent mechanisms,” Mech. Mach. Theory 49, 187–197 (2012). 18 P. S. Gollnick, S. P. Magelby, and L. L. Howell, “An introduction to multilayer lamina emergent mechanisms,” J. Mech. Des. 133, 081006 (2011). 19 J. O. Jacobsen, B. G. Winder, L. L. Howell, and S. P. Magleby, “Lamina emergent mechanisms and their basic elements,” J. Mech. Rob. 2, 011003 (2010). 20 H. C. Greenberg, M. L. Gong, S. P. Magelby, and L. L. Howell, “Identifying links between origami and compliant mechanisms,” Mech. Sci. 2, 217–225 (2011). 21 L. A. Bowen, C. L. Grames, S. P. Magleby, R. J. Lang, and L. L. Howell, “A classification of action origami as systems of spherical mechanisms,” J. Mech. Des. 135, 111008 (2013). 22 J. Ryu, M. D’Amato, X. Cui, K. N. Long, H. J. Qi, and M. L. Dunn, “Photoorigami — bending and folding polymers with light,” Appl. Phys. Lett. 100, 161908 (2012). 23 J. Subirana and M. Soler-Lopez, “Cations as hydrogen bond donors: A view of electrostatic interactions in DNA,” Annu. Rev. Biophys. Biomol. Struct. 32, 27–45 (2003). 24 Q. T. Aten, B. D. Jensen, S. H. Burnett, and L. L. Howell, “Electrostatic accumulation and release of DNA using a micromachined lance,” J. Microelectromech. Syst. 20, 1449–1461 (2011). 25 R. L. Brinster, H. Y. Chen, M. E. Trumbauer, M. K. Yagle, and R. D. Palmiter, “Factors affecting the efficiency of introducing foreign DNA into mice by microinjecting eggs,” Proc. Natl. Acad. Sci. U.S.A. 82, 4438–4442 (1985). 26 R. J. Wall, “Pronuclear microinjection,” Cloning Stem Cells 3, 209–220 (2001). 27 A. Pillarisetti, M. Pekarev, A. D. Brooks, and J. P. Desai, “Evaluating the role of force feedback for biomanipulation tasks,” Proc. Symp. Haptic Interfaces Virtual Environ. Teleoperator Syst. 11–18 (2006). 28 S. Zappe, M. Fish, M. Scott, and O. Solgaard, “Automated MEMS-based drosophila embryo injection system for high-throughput RNAi screens,” Lab Chip 6, 1012–1019 (2006). 29 J. Brachet, T. Kuusi, and S. Gothie, “Comparative study of the inductive power of nucleic acids and of cellular nucleoproteins when they are implanted or given by microinjection,” Arch. Biol. 63, 429–440 (1952).
Rev. Sci. Instrum. 85, 055005 (2014) 30 J.
B. Buck, “A versatile microinjection and micropipetting syringe,” Rev. Sci. Instrum. 20, 676–677 (1949). 31 S. Park, Y.-S. Kim, W. B. Kim, and S. Jon, “Carbon nanosyringe array as a platform for intracellular delivery,” Nano Lett. 9, 1325–1329 (2009). 32 A. K. Shaleka, J. T. Robinsona, E. S. Karpa, J. S. Leea, D.-R. Ahnb, M.-H. Yoona, A. Suttona, M. J. amd Rona S. Gertnera, T. S. Gujrala, G. MacBeatha, E. G. Yanga, and H. Parka, “Vertical silicon nanowires as a universal platform for delivering biomolecules into living cells,” Proc. Natl. Acad. Sci. U.S.A. 107, 1870–1875 (2010). 33 C. M. Cuerrier, R. Lebel, and M. Grandbois, “Single cell transfection using plasmid decorated AFM probes,” Biochem. Biophys. Res. Commun. 355, 632–636 (2007). 34 X. Chen, A. Kis, A. Zettl, and C. R. Bertozzi, “A cell nanoinjector based on carbon nanotubes,” Proc. Natl. Acad. Sci. 104, 8218–8222 (2007). 35 J. Carter, A. Cowen, B. Hardy, R. Mahadevan, M. Stonefield, and S. Wilcenski, PolyMUMPs Design Handbook, Revision 11 (MEMSCAP Inc., 2005). 36 B. L. Boyce, J. M. Grazier, T. E. Buchheit, and M. J. Shaw, “Strength distributions in polycrystalline silicon MEMS,” J. Microelectromech. Syst. 16, 179–190 (2007). 37 R. Yeh, E. J. Kruglick, and K. S. Pister, “Surface-micromachined components for articulated microrobots,” J. Microelectromech. Syst. 5, 10–17 (1996). 38 A. Friedberger and R. S. Muller, “Improved surface-micromachined hinges for fold-out structures,” J. Microelectromech. Syst. 7, 315–319 (1998). 39 G. H. Teichert, Q. T. Aten, S. H. Burnett, L. L. Howell, and B. D. Jensen, “Cylindrical single-degree-of-freedom spatial mechanisms for cell restraint,” J. Mech. Rob. 4, 021011 (2012). 40 K. A. Jensen, C. P. Lusk, and L. L. Howell, “An XYZ micromanipulator with three translational degrees of freedom,” Robotica 24, 305–314 (2006). 41 B. G. Winder, S. P. Magelby, and L. L. Howell, “Kinematic representations of pop-up paper mechanisms,” J. Mech. Rob. 1, 021009 (2009). 42 L. Howell, Compliant Mechanisms (Wiley, New York, 2001). 43 S. Oak, G. F. Edmiston, G. Sivakumar, and T. Dallas, “Rotating out-ofplane micromirror,” J. Microelectromech. Syst. 19, 632–639 (2010). 44 Q. T. Aten, B. D. Jensen, S. Tamowski, A. M. Wilson, L. L. Howell, and S. H. Burnett, “Nanoinjection: Pronuclear DNA delivery using a charged lance,” Transgenic Res. 21, 1279–1290 (2012). 45 A. M. Wilson, Q. T. Aten, N. C. Toone, J. L. Black, B. D. Jensen, S. Tamowski, L. L. Howell, and S. H. Burnett, “Transgene delivery via intracellular electroporetic nanoinjection,” Transgenic Res. 22, 993–1002 (2013). 46 A. Nagy, M. Gertsenstein, K. Vintersten, and R. Behringer, Manipulating the Mouse Embryo: A Laboratory Manual, 3rd ed. (Cold Spring Harbor Laboratory Press, 2003). 47 L. D. Brown, T. T. Cai, and A. DasGupta, “Interval estimation for a binomial proportion,” Statistical Science 16, 101–133 (2001).
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitationnew.aip.org/termsconditions. Downloaded to IP: 81.246.229.144 On: Sat, 17 May 2014 08:15:37
