PROTOCOL

Piezo-actuated mouse intracytoplasmic sperm

injection (ICSI)
Naoko Yoshida & Anthony CF Perry
Laboratory of Mammalian Molecular Embryology, RIKEN Center for Developmental Biology, 2-2-3 Minatojima Minamimachi, Chuo-ku, Kobe 650-0047, Japan.
Correspondence should be addressed to A.C.F.P. (tony@cdb.riken.jp).

2007 Nature Publishing Group http://www.nature.com/natureprotocols

Published online 1 March 2007; doi:10.1038/nprot.2007.7

The mouse is a genetically tractable model organism widely used to study mammalian development and disease. However, mouse
metaphase II (mII) oocytes are exquisitely sensitive and intracytoplasmic sperm injection (ICSI) with conventional pipettes generally
kills them. This problem can be solved with piezo-actuated micromanipulation, in which the piezo-electric effect (crystal deformation
in response to an externally applied voltage) propels a microinjection needle tip forward in a precise and rapid movement. Piezoactuated micromanipulation enhances the penetration of membranes and matrices, and mouse ICSI is a major application. Here we
describe a comprehensive, step-by-step mouse piezo ICSI protocol for non-specialists that can be completed in 24 h. The protocol
is a basic prelude to multiple applications, including nuclear transfer cloning, spermatid injection, blastocyst injection, mII
transgenesis, and streamlining micromanipulation in primates and livestock. Moreover, piezo ICSI can be used to obtain offspring
from dead (non-motile) sperm, enabling trivial sperm freezing protocols for mouse strain storage and shipment.

INTRODUCTION
Piezo-actuated micromanipulation harnesses the piezo-electric
effect to transmit a small crystal lattice distortion to the tip of a
pipette, driving it forward in a precise and controlled manner.
Piezo-actuated micromanipulation has multiple applications in the
study and engineering of gametes and embryos. It enabled the first
intracytoplasmic sperm injection (ICSI) to produce mice1, the first
nuclear transfer cloning of mice2 and pigs3, the first productive
frozen4 and freeze-dried sperm injections5, and the first production
of nuclear transfer embryonic stem (ES) cells6. Piezo was utilized to
generate the first transgenic offspring by injecting unfertilized
oocytes in metaphase II (mII) transgenesis7 and has been extended
to the delivery of artificial chromosome transgenes8,9. Piezo has
been employed for RNA interference (RNAi) in mII oocytes10; it
enhances blastocyst injection with ES cells11 and the manipulation
of gamete precursors12 and facilitates the renaissance of stem cell
biology13.
Piezo-actuated micromanipulation was developed by Atsushi
Mimatsu and colleagues, and its application as a biological research
tool was demonstrated in the mouse by Dr. Yasuyuki Kimura, who
used it to generate the first mouse offspring by ICSI1. Piezo was
necessary because in the mouse, oocyte plasma membranes are
exquisitely sensitive and survival rates following ICSI with conventional (i.e., manual, non-piezo) microinjection rarely exceed 50%
(ref. 14). Contrastingly, piezo ICSI can achieve mouse oocyte
survival rates of 100%, with 90% development to morula/blastocyst
stages in vitro. The efficacy of piezo is highly desirable, not only in
ICSI but also where development following microinjection is less
efficient, such as in nuclear transfer. Most mouse nuclear transfer

MATERIALS
REAGENTS

. Mice of the strain B6D2F1, female C57BL/6  male DBA/2 F1 hybrid

(suppliers include Charles River Laboratories and Shimizu Laboratory
Supplies). ! CAUTION Experimental procedures involving
animals must be carried out according to national and institutional
regulations.

296 | VOL.2 NO.2 | 2007 | NATURE PROTOCOLS

clones have been generated via piezo-actuated delivery of donor

nuclei into recipient oocytes.
Piezo displaces the microinjection needle Z0.1 mm at r40 mm
S
1 and can introduce precise, non-lethal plasma membrane breaks
even with tips of relatively large outer diameter (410 mm); indeed,
piezo can successfully drive still wider pipes (up to B30 mm outer
diameter) for other applications. The capacity to drive wide tips
for non-lethal injection augments the range of applications
possible with piezo; conventional microinjection is typically
restricted to tip diameters of 12 mmfar too small for functional
injection of eukaryotic cells. In addition, and again in contrast to
conventional micromanipulation, piezo permits multiple procedures to be performed with a single needle. Thus, although
many oocytes (such as those of humans, cattle, pigs and rabbits)
can be injected conventionally, the relative speed and ease of
piezo means that it is advantageous even in these species3,1520.
Ease of pipette fabrication (beveled tips are not required) and
speed also make piezo suitable for the injection of blastocysts with
ES cells11, and the method additionally allows partial ablation of the
inner cell mass (i.e., the removal of potentially competing embryonic cells).
The present protocol describes piezo for mouse ICSI, which is a
preferred place to learn piezo before attempting other applications,
such as nuclear transfer. Piezo ICSI shares basic steps with other
piezo applications and controls for user technique (both micromanipulation and embryological); injected oocytes should develop
efficiently in vitro, and in vivo following embryo transfer, enabling
beginners to ascertain the quality of their piezo technique.

. Kalium simplex optimized medium (KSOM)

. Nuclear isolation medium (NIM)
. Embryo-tested bovine testis hyaluronidase (Sigma, cat. no. H-4272)
. 12% (w/v) polyvinylpyrrolidone (PVP360, average MrE360,000), dissolved in
sterile, high quality distilled water

. Mineral oil (Shire or Nakalai Tesque)

PROTOCOL

2007 Nature Publishing Group http://www.nature.com/natureprotocols

. Elemental mercury (Hg0), optional ! CAUTION Hg0 is a cumulative

neurotoxin absorbed through the skin. Handle with gloves and dispense
waste according to local institutional guidelines.
. Fluorinert (3M Specialty Materials), optional. On a scale of 0 to 10 on which
Hg0 scores 10 in piezo, Fluorinert FC-77 scores 8 (acceptable), while FC-43
(score 4) and FC-70 (score 3) have higher viscosity and are more
sticky
. Hormones for superovulation. Consult your institutional animal facility for
reliable sources of human chorionic gonadotropin (hCG) and pregnant mare
serum gonadotropin (PMSG), also known as equine chorionic gonadotropin
(eCG)
EQUIPMENT
. Dumostar #5 fine biological forceps (Electron Microscopy Sciences, cat. no.
72705-01)
. Transfer pipette
. Holding pipettes (Eppendorf)
. Borosilicate glass capillaries (Sutter, cat. no. B100-75-100), 10 cm  1 mm
(outer diameter), 0.75 mm (inner diameter)
. Precision pipette, such as the P200 Pipetman (Gilson, Inc.)
. Pipette puller, such as the Flaming/Brown P97/IVF micropipette puller
(Sutter Instrument Co.)
. Microforge (manufacturers include Narishige and De Fonbrune)
. Stereomicroscope, such as the SZX12 (Olympus)
. Stereomicroscope stage heated platform, such as the MATS-55SZX Thermo
Plate (Tokai Hit Co.) or equivalent
. Workstation comprising inverted microscope, such as IX71 (Olympus)
equipped with MO-202U micromanipulators, IM-5 and/or -6 injectors
(Narishige), heated stage (MATS-55R30 Thermo Plate; Tokai Hit Co.) and
Hoffman modulation contrast optics (4 and 20 objectives) or equivalent
. Air-cushioned table or platform (suppliers include Technical Manufacturing
Corporation and Meiritsu, respectively). The need for insulation against
vibration depends on the location of the workstation
. Piezo-actuated micromanipulator, such as the PMAS-CT150 (Prime Tech)
. Humidified CO2 (5% (v/v) in air) incubator (suppliers include Sanyo and
Heraeus)
. Falcon 10-cm (100  20 mm) Optilux petri dishes (Becton Dickinson, cat.
no. 353003) or equivalent, bottoms are suitable for oocyte/embryo collection
and lids for micromanipulation (compatible with Hoffman modulation
optics)
. Falcon 3.5-cm (35  10 mm) dishes (Becton Dickinson, cat. no. 351008),
used for short term embryo culture (B24 h)
. Falcon 6-cm (60  15 mm) dishes (Becton Dickinson, cat. no. 351007),
used for longer culture
. Kimwipe delicate task wipers (Kimberly-Clark, cat. nos. 34155 (11.4 
21.3 cm; small) and 34133 (30  30 cm)) or equivalent low-lint tissues
. 1-ml syringe
. 26 G needle
. Microinjection pipette (also referred to as microinjection needle)
REAGENT SETUP
Chatot, Ziomek, Bavister medium (CZB) Contains 81.6 mM NaCl, 4.8 mM
KCl, 1.2 mM MgSO4, 1.7 mM CaCl2, 1.2 mM KH2PO4, 0.1 mM EDTA.Na2,
31 mM Na.lactate, 5.6 mM D-glucose, 25 mM NaHCO3, 0.3 mM Na.pyruvate,
1 mM D-glutamine, 10 mg ml
1 phenol red (0.5% (w/v) in Dulbeccos
phosphate-buffered saline (DPBS); Sigma, cat. no. P-0290), 1 mg ml
1 BSA;
pH 7.4; ref. 21. Prepare with double-distilled or higher purity water. For embryo
culture, CZB should be 0.22 mm sterile-filtered. Store and use as for KSOM. The
source of BSA can effect development, but embryo-tested Fraction V powder
(Sigma, cat. no. A-3311) is commonly used.
CZB.HEPES (CZB.H) This is essentially HEPES-buffered CZB, and contains
20 mM HEPES, 81.6 mM NaCl, 4.8 mM KCl, 1.2 mM MgSO4, 1.7 mM CaCl2,
1.2 mM KH2PO4, 0.1 mM EDTA.Na2, 31 mM Na.lactate, 5 mM NaHCO3,
0.3 mM Na.pyruvate, 0.1 mg ml
1 polyvinyl alcohol, 10 mg ml-1 phenol red
(0.5% (w/v) in DPBS); pH 7.4; ref. 22. It is used for micromanipulation, 0.22 mm
sterile-filtered. Store in 5-ml polypropylene bottles at 4 1C for up to a month.
m CRITICAL Use a fresh bottle of medium per experimental day. m CRITICAL
Any marked departure from the osmolarity of CZB H can induce oocyte lysis.
KSOM This is a frequently-used alternative to CZB for embryo culture.
KSOM+AA (Specialty, cat. no. MR-107-D) contains amino acids plus 5.6 mM
D-glucose, and should be further supplemented with 20 ml of phenol red
solution (0.5% (w/v) in DPBS) and 1 mg ml
1 BSA (Sigma, cat. no. A-3311).
m CRITICAL Store in 5-ml polypropylene bottles (Perfector Scientific, cat. no.

2000-S) at 4 1C for up to a month; not all storage vessels are appropriate, as

toxins may leach out over prolonged periods. Use a fresh bottle of medium each
experimental day.
NIM Contains 125 mM KCl, 2.6 mM NaCl, 7.8 mM Na2HPO4, 1.4 mM
KH2PO4 and 3.0 mM EDTA (pH 7.0) and is sterilized by autoclaving for use in
sperm preparation23.
PVP solution Dissolve 1.2 g PVP360 in B8 ml high-quality, sterile water in a
50-ml Falcon tube on a roller (such as the Low Profile Roller of Stovall Life
Science, Inc.) at room temperature (2228 1C). The PVP360 takes B1 h to
dissolve. Make the final volume to 10 ml, 0.45 mm filter and store in 0.21.0 ml
aliquots at
20 1C (storage may be for up to months or probably years).
Different sources of PVP360 can affect oocyte activation and produce different
developmental outcomes following piezo; ICN (recently rebranded Valeant) and
Kanto (cat. no. 32285-20) have given good results. The in-use aliquot of PVP360
solution is useable for several weeks stored at room temperature.
Embryo-tested bovine testis hyaluronidase stock Dissolve 30 mg of bovine
testis hyaluronidase (2.254.5  104 U) in 3 ml CZB H (to give 10 mg ml
1).
Store in 50 ml aliquots at
20 1C (storage may be for at least several months). It
can be repeatedly thawed and refrozen. Use at a final concentration of 300
800 mg ml-1.
Mineral oil The mineral oil overlay of culture droplets is dispensed from a
wash bottle. To draw out plasticizers and other potential toxins from the wash
bottle, fill the bottle with oil and allow it to stand for two weeks or more (discard
this oil) prior to its first use. Sterilization of the oil is not necessary.
PMSG and hCG for superovulation Dissolve PMSG (sometimes called eCG)
and hCG separately at 50 IU ml
1 in 0.9% (w/v) NaCl solution (not PBS).
0.22 mm-filter and store in 1 ml aliquots at
20 1C for 12 months (possibly
longer depending on the batch, although the hormones gradually lose activity)
or for 13 d at 4 1C. Do not refreeze.
EQUIPMENT SETUP
Microinjection needle preparation Borosilicate glass capillaries are pulled on
a micropipette puller. Consult a local representative for instrument guidelines
because several factors can affect performance, including the model, unit-tounit variation and humidity. Pipettes may be pulled on a Flaming/Brown P97/
IVF micropipette puller equipped with a 4.5 mm box filament, and should give a
gradual taper. Suggested initial settings in this case are: heat 830; pull 75;
velocity 130; time 100. Pulls which produce tapers beginning B2.6 cm
from the shoulder of the pull to the resultant whisp-like fiber tip are acceptable,
but successful parameters vary. Flush-ended needles are generated from each
pulled capillary on a microforge. Working with the 10 objective, introduce a
ball of glass (300500 mm diameter) onto the V of the platinum-iridium
filament of the microforge and bring it into sharp focus. Move the horizontal
capillary so that it is just above the sharply defined edge of the glass ball in the
same focal plane. Lower it to touch the ball where the capillary external diameter
is 6.58.5 mm (Fig. 1a); the point of contact is where the break will occur. Apply
a brief heat (current) pulse via the microforge foot pedal, ensuring that sufficient
current flows through the filament to cause it to glow brightly. The filament
expands and the capillary locally fuses with the glass of the ball, causing it to
break when the filament contracts, producing a flush (flat) end (Fig. 1b,c). The
required level of heat is determined empirically. m CRITICAL Good needles are
essential for success in piezo ICSI. The needle tip should appear regular and
flush-ended. Next, move the needle upwards so that it is 510 mm above the glass
ball and forwards B3 mm (Fig. 1d). Starting at zero, gradually increase the heat
to introduce a bend in the needle B28 (2)1 from the horizontal (Fig. 1e).
Remove the needle and, if appropriate, introduce B23 mm Hg0 into its wide,
unworked end from a disposable 1-ml syringe fitted with a 26 G needle.
Microinjection pipettes can be stored on rolled parafilm in a 10-cm
Petri dish in ambient conditions for weeks. TIMING 10 pipettes
take B20 min to prepare.
Preparation of the microscope for micromanipulation A piezo workstation
is shown in Figure 2. Locating the workstation on an air-cushioned table or
platform helps to insulate against vibration (e.g., caused by freezers, elevators,
air-conditioning or traffic), which otherwise can make micromanipulation
unfeasible. A single holding pipette may be used for many sessions (over months
to years) without changing it. At least one change of microinjection needle is
recommended per session. Mount the microinjection needle firmly and use the
injector to apply positive pressure and advance the front (e.g., Fluorinert or Hg0)
so that it approaches the needle tip. Lower the assembly into a droplet
containing PVP360 solution and examine it at 200 magnification. The oil
(holding pipette) and Fluorinert or Hg0 (microinjection pipette) fronts should
be static unless caused to move by the injector. If either holding or

NATURE PROTOCOLS | VOL.2 NO.2 | 2007 | 297

PROTOCOL
microinjection pipette fronts recede autonomously, there is a leak in the
corresponding line, which must be repaired. Good microinjection needles have a
smooth-looking exterior; there is no evidence of external bubble nucleation,
which is caused by grease or colloidal, residual aggregated particles. The needle
tip ideally appears straight with a flush tip although slight curvature is tolerated.
Draw some of the PVP360 solution into the pipette and expel several drops of
Hg0 or Fluorinert; repeat this process to wash and lubricate the interior. To wash
the needle exterior, rub it against a globule of Hg0 in PVP360 solution or (later)
on the internal aperture of the holding pipette in the microinjection droplet. If
Hg0 is used, fill the microinjection pipette line with water. Alternatively, fill the
line with Fluorinert FC-77 and omit Hg0. ! CAUTION Elemental mercury (Hg0)
is a cumulative neurotoxin. Handle with gloves and dispose of it with care
according to local institutional guidelines. It is emphasized that Hg0 should be
avoided for human ICSI. TIMING B2 min.

2007 Nature Publishing Group http://www.nature.com/natureprotocols

PROCEDURE
Preparation and incubation of oocytes TIMING 3045 min
1| To obtain relatively large numbers of oocytes (2535/
B6D2F1), superovulate by serial intraperitoneal injection (26 G
needle) of 5 IU PMSG followed 4554 h later by 5 IU hCG (i.e.,
inject B0.1 ml of each stock per female). Typically (with
standard mouse room light/dark cycles), administer hormone
injections at 67 pm and avoid administering them after 10 pm.
Females of 810 weeks give good yields, but the optimum varies.
If the females have been shipped, allow 17 d for recovery prior
to superovulation (i.e., PMSG injection). We recommend the
strain B6D2F1 because they breed and superovulate efficiently,
their oocytes are relatively easy to micromanipulate, and
resultant embryos develop robustly in vitro and in vivo.

2| 1215 h post-hCG injection, sacrifice the mice (35 per

experiment usually suffices), collect their oviducts and place
them in CZB H. If micromanipulation and animal facilities are
not adjacent, transport the oviducts at room temperature
(i.e., avoid exposure to cold) in a 1.5-ml Eppendorf tube
containing 100200 ml CZB H. Transfer the first oviduct in a
200 ml drop of CZB H containing 615 ml bovine testis
hyaluronidase stock solution.

e
28

Figure 1 | Piezo microinjection pipette fabrication. Crafting a piezo needle

from a pulled capillary on a microforge. The capillary is lowered so that it kisses
the glass ball (a). Following a brief heat pulse, the capillary fractures (b) and is
raised above the glass ball (c) before being advanced ~3 mm (d). A gradual
increment of heat (starting at 0) is then applied to the ball, causing the newlyfashioned microinjection needle steadily to deviate upwards. Heating is stopped
when the needle has deflected by 281 (e). See Microinjection needle
preparation for detailed description. Scale bar = 100 mm.

3| Working under the stereomicroscope at 1520 magnification, hold the oviduct wall with one pair of forceps and tear
it with another pair of forceps where the oocyte-cumulus
complex makes a bulge. Tease the oocyte-cumulus complex
from the oviduct; it should emerge as a single mass. Discard
the oviductal remnant and repeat the process for the remaining
oviducts in the same CZB H/hyaluronidase droplet.

4| Leave the oocyte-cumulus masses in the same dish for

530 min to allow the hyaluronidase to digest the intercellular
matrix. Cells fall away from the complexes, leaving the oocytes
on a cumulus cell carpet. This step is optionally accelerated by
placing the dish on a stage heated at 37 1C for up to B20 min.
5| Remove oocytes with a transfer pipette and place them in
a drop (B200 ml) of fresh CZB H (no hyaluronidase). Wash the
oocytes by pipetting them 46 times. Repeat this washing procedure 34 times, each time in a fresh droplet (B200 ml) of
CZB H.

Figure 2 | Piezo workstation. Abbreviations: lem, left electric micromanipulator;

rem, right electric micromanipulator; lmm, left manual micromanipulator;
rmm, right manual micromanipulator; sb, supporting block; hpi, holding
pipette injector; mni, microinjection needle injector; mb, metal base; dc,
digital camera (optional); air, air-cushioned platform; im, inverted microscope;
st, microscope stage; hp, holding pipette; mn, microinjection needle; p, piezo.
298 | VOL.2 NO.2 | 2007 | NATURE PROTOCOLS

6| Transfer the oocytes to a drop of equilibrated culture medium (e.g., KSOM) under mineral oil in a 3.5-cm dish, washing

PROTOCOL

once to remove CZB H. Overlay embryo culture and micromanipulation (but not collection) droplets (520 ml) with mineral oil
to prevent evaporation. Place the 3.5-cm dish containing the oocytes in an incubator containing humidified 5% (v/v) CO2/air
mix at 37 1C for at least 15 min until required. From oviduct to incubator, oocyte isolation should take less than 60 min.
! CAUTION HEPES is mildly toxic, more so at 37 1C, so minimize exposure of the oocytes to CZB H.
m CRITICAL STEP Equilibrate culture medium in humidified 5% (v/v) CO2/air mix at 37 1C for at least 15 min prior to use.
PAUSE POINT Oocytes may be incubated in humidified 5% (v/v) CO2/air mix at 37 1C for several hours (until they are 420 h
post-hCG) prior to micromanipulation, but they are more typically injected prior to 18 h post-hCG; oocytes efficiently
retain their fertilization capacity at least 22 h post-hCG in vivo24. Following oocyte collection, spermatozoa may be
prepared as follows.

2007 Nature Publishing Group http://www.nature.com/natureprotocols

Preparation of spermatozoa TIMING 510 min

7| Collect sperm from the cauda epididymides of males that are at least 10 weeks old, yielding B5  106107 sperm per
epididymis. The cauda epididymidis is a distinctive yellow pouch of convoluted tubules surrounded by fat and muscle; it is
smaller, smoother and whiter in younger animals, and generally contains fewer sperm. Sperm may also be removed from caput
(testis proximal) epididymides, although such sperm are non-motile and need to be isolated by dicing (Step 9B). Sperm
isolation is otherwise similar for caput and cauda epididymides, although it is now described as it pertains to the cauda
epididymidis.

8| Place epididymidal tissue on a clean piece of Kimwipe (or equivalent) paper soaked in PBS or CZB H in a 3.5-cm dish to
keep it moist until ready for processing. Each epididymis is surrounded by a layer of musclethe tunica. Place the dissected
epididymis on a clean, dry Kimwipe and holding the tunica with one pair of forceps, use another to pull the epididymidal tissue
out. Cut epididymis-tunica adhesions and trim away fat pads from the cauda before further blotting it with the Kimwipe to
remove residual blood and fat.
m CRITICAL STEP Do not allow the epididymidal tissue to dry out, as this will markedly reduce sperm yield. Work swiftly,
processing a maximum of two epididymides at a time for option A and six for option B (Step 9).
9| Isolate sperm by simple epididymal puncture (option A) or by dicing epididymides with fine scissors (option B), the method
best suited to larger scale isolation.
(A) Epididymal puncture
(i) Pinch the cauda epididymidis between thumb and forefinger so that the cauda bulges, and gently squeeze. Carefully prick
the bloated cauda with a pair of fine, sharp forceps and pinch gently so that sperm ooze out onto the surface of the epididymis.
(ii) Remove the sperm by pinching the suspension between forceps that have been dunked in oil (to reduce sperm sticking to
them).
(iii) Wash the forceps by agitating them in a dish containing CZB or KSOM (equilibrated in the CO2 incubator at 37 1C for
Z15 min) under oil and return the dish to the incubator.
(iv) The drop becomes milky as the sperm actively disperse. This takes about 5 min.
(B) Dicing epididymides
(i) Mince epididymides with fine scissors in a 3.5-cm dish containing B500 ml of buffer (e.g., NIM23) or medium
(e.g., CZB H) at room temperature (2328 1C) for 25 min. Displace sperm from the tissue by brief, gentle pipetting.
(ii) Filter the suspension through a double-folded Kimwipe into a clean (sterile) 10-ml beaker, to eliminate tissue
debris. Expel residual suspension by gently squeezing the tissue (without breaking it) between thumb and forefinger over
the funnel.
PAUSE POINT Sperm may be stored at
20 1C or
70 1C for at least several months without elaborate cryopreservation;
place them directly into the freezer in medium or buffer. Storage of non-motile, dead sperm or sperm heads for
mouse ICSI and subsequent production of live offspring has been described for freeze-thaw4,25 and freeze-dry5 sperm
microinjection.

Piezo-actuated intracytoplasmic sperm injection (ICSI) TIMING 12 h

10| Mix sperm suspensions with PVP360 solution so that the final concentration of PVP360 is 515% (w/v). Mix using a pipette,
such as a Gilson Pipetman P200 or equivalent, equipped with a sterile P200 (yellow) tip truncated (cut off) B5 mm from the
end. PVP360 acts as a lubricant and retards live sperm, making them easier to collect. With a transfer pipette, remove mII
oocytes from the CO2 incubator and place them into a microinjection droplet (CZB H) on the microscope stage. Figure 3 shows
dispensation of mineral oil and a working arrangement of drops in a micromanipulation dish.
m CRITICAL STEP Sperm in PVP360 suspensions at room temperature often lose their ability to support full development after B1 h.
If good development is desired, make fresh sperm-PVP360 suspensions every 1 h. However, spermatozoa retain their ability to induce
meiotic resumption for many hours, so a single sperm-PVP360 is sufficient for beginners.

NATURE PROTOCOLS | VOL.2 NO.2 | 2007 | 299

 2007 Nature Publishing Group http://www.nature.com/natureprotocols

PROTOCOL
11| If the sperm is intact (i.e., head, midpiece and tail are
joined) decapitate it (Supplementary Video 1 online) to allow
injection of the head alone. Bring a sperm into sharp focus
(200 magnification) and draw its head into the
microinjection pipette (inner tip diameter, 5.57.5 mm; outer
tip diameter, 6.58.5 mm) so that the pipette lip touches the
head-midpiece boundary. Apply several pulses with the piezo
unit foot pedal. For initial attempts with the PMAS-CT150
(or equivalent) piezo unit, try intensity 3, frequency 6;
as a general guide in piezo, use the lowest settings that work.
The head and tail should separate. Many preparation protocols
(e.g., freeze-thawing) decapitate sperm before the sperm are
placed on the stage. Injection of a head alone minimizes the
volume introduced into the oocyte.
? TROUBLESHOOTING

Figure 3 | Microinjection dish setup. As soon as the medium and sample

drops (515 ml each) are placed on the dish lid they are overlaid with mineral
oil from a wash bottle (a) to prevent evaporation. A suggested arrangement of
the PVP, medium and sample drops is shown in (b). Abbreviations: P, PVP360
solution; S, sample; C, CZB H. Scale bars 2 cm.

12| Draw one or more sperm heads into the microinjection pipette. With experience, multiple (510 or more) heads can be
accumulated within a pipette at intervals B100 mm; although only a single head is typically injected per oocyte, collecting
multiple samples removes the need to return to the sperm-PVP360 droplet each time between injections. Working at
40 magnification, move the microscope stage so that the pipette is in the microinjection droplet (CZB H) containing 10
15 oocytes (transferred from the batch of Step 6 in the CO2 incubator), optionally first washing it in a CZB H droplet to remove
PVP360 solution from the needle exterior.
? TROUBLESHOOTING

13| Lower the holding pipette into the same droplet and position the holding and microinjection pipettes so that their tips are
on either side (usually with the holding pipette on the left) of a selected oocyte at the center of the field (Fig. 4a,b). Use
the microinjection needle to orientate the oocyte so that its mII plate is located straight up or down along the y-axis (i.e.,
12 oclock or 6 oclock). Either orientation minimizes the risk of damaging the mII plate during microinjection; even slight
mechanical damage to the spindle (i.e., mII plate) prevents normal development. Select an orientation (6 oclock in Fig. 4b)
that gives the largest space between the plasma membrane and the zona pellucida (perivitelline space) on the side of the
microinjection pipette.
? TROUBLESHOOTING
14| At 200 magnification, bring the oocyte plasma membrane into sharp focus. Use the fine z-axis control to move the
pipettes up or down so that their ends are in focus. Steps 1424 are demonstrated in Supplementary Video 2 online.
m CRITICAL STEP Accurate focus is crucial for several steps in oocyte micromanipulation.
15| Applying gentle suction within the holding pipette, push the oocyte against the holding pipette aperture with the
microinjection pipette (Fig. 4c,d). Pushing causes the oocyte zona and plasma membrane to engage the holding pipette
aperture and cover it, anchoring the oocyte and making it easier to inject (Fig. 4e).
16| Move the most advanced sperm head so that it is 50100 mm from the pipette tip. Check the focus and touch the zona
periphery with the tip (Fig. 4e).
17| Introduce a small negative pressure within the microinjection pipette and apply piezo pulses (start with intensity 3,
frequency 6) while gently pushing the microinjection pipette toward the holding pipette. The tip should rapidly pass through
the zona, emerging into the perivitelline space (Fig. 4f,g). Stop pushing as soon as you see this. Failure to control the tip risks
subjecting the plasma membrane to high intensity piezo, resulting in rapid cell lysis.
? TROUBLESHOOTING
18| The zona core is just visible (it is difficult to see without experience) entering the pipette as the tip penetrates
into the perivitteline space. Although the oocyte cytoplasm can tolerate the presence of a zona plug, the plug can
interfere with injection. Eject the zona plug prior to injection by placing the pipette tip against the plasma membrane
(without piezo pulses) and gently increasing positive pressure (Fig. 4g). When the core emerges from the pipette, it produces
an impression (dent) in the plasma membrane. Dislodge the protruding core by wiping the tip against the oocyte
(gentle movement mainly along the y-axis). Sustain positive pressure until the end-most sperm is 1050 mm from
the tip (Fig. 4h).

300 | VOL.2 NO.2 | 2007 | NATURE PROTOCOLS

PROTOCOL

2007 Nature Publishing Group http://www.nature.com/natureprotocols

19| Ensure that the plasma membrane

is in sharp focus and fine-adjust the tip
(z-axis) so that it is also in focus. Avoid
the first polar body (Pb1), which can be
displaced; it is not usually attached to
the oocyte. Maintain a marginal positive
pressure within the pipette.
20| Steadily advance the tip towards
the opposite side of the oocyte, where
it is being anchored by the holding
pipette (Fig. 4ik). Stop when the tip
has advanced B95% of the oocyte
diameter. This produces a deep
invagination in the (remarkably elastic)
oocyte plasma membrane, which is now
stretched around the microinjection
needle. Advance of the needle tip
takes 25 sec.
? TROUBLESHOOTING
21| Change the piezo pre-set channel
(foot switch) to a gentler setting (e.g.,
intensity 1, frequency 1). While
maintaining zero net pressure (or a very
small negative one) within the
microinjection pipette, apply a single
piezo pulse.
m CRITICAL STEP Ensure that the
oocyte plasma membrane visibly relaxes
along the shaft of the needle (Fig. 4l).
This is an indispensable indication that
the membrane has been punctured.
? TROUBLESHOOTING
22| Deposit the sperm head in the
cytoplasm (Fig. 4m) with the
application of small positive pressure in
the microinjection pipette. Introduce
the minimum extraneous medium and
do not suck cytoplasm into the pipette.
? TROUBLESHOOTING
23| As soon as the sperm has been
deposited, withdraw the needle
smoothly from the oocyte (Fig. 4np).
The membrane should return to its
original position and seal. If the plasma
membrane clings to the pipette, pause
briefly during withdrawal. From zona
penetration to needle withdrawal takes
2535 sec.
? TROUBLESHOOTING

k
12
9

3
6

mll

RIP

Zy

Figure 4 | Mouse piezo ICSI. Successive micromanipulation steps (aq) are described in the text. Panel
(a) was obtained at 40 magnification by the microscope (4 objective), while all subsequent panels
were obtained at 200 magnification (20 objective). The Hg0 front in (a) is marked with a white
arrowhead. In (b), the oocyte orientation is visualized by the position of its mII bump (red arrowhead)
on an imagined clock face (in this case, it is at 6 oclock). Circles mark the position of the sperm head
prior to injection in (f), (g) and (h) and immediately after it (m). Insets schematically represent
membrane invagination (j, k) and piezo-induced puncture (l), and show evisceration of a different oocyte
(o) and the resulting exit wound (p) caused when the membrane adheres to the withdrawing pipette (this
oocyte died). In (r), panels show a metaphase II oocyte (mII), a dead oocyte (requiescat in pace, RIP),
and a pronuclear zygote (Zy) 6 h after piezo ICSI with pronuclei and a Pb2 (white arrowhead). Scale
bars 20 mm throughout.

24| Release the injected oocyte by gently applying positive pressure within the holding pipette (Fig. 4q). Any minor distortion
caused by holding rapidly disappears. Allow oocytes at least 5 min (some workers prefer 1030 min) to recover before being
returned to the CO2 incubator. Fatal mechanical trauma to the oocyte during micromanipulation usually results in the onset of
lysis within 30 sec. Remove dead eggs (Fig. 4r) from the survivors during transfer to the CO2 incubator. For embryo culture,
NATURE PROTOCOLS | VOL.2 NO.2 | 2007 | 301

PROTOCOL

2007 Nature Publishing Group http://www.nature.com/natureprotocols

wash the oocytes B4 times in equilibrated culture medium to remove CZB H (HEPES is toxic at 37 1C) before incubation. For
optimal development, limit the time the oocytes are out of the incubator; r40 min is a guideline. Experienced workers can
successfully inject 15 oocytes in o10 min. When learning, one oocyte can take 30 min or much longer to inject. A breakdown
of typical times taken to complete Steps 124 is shown in Table 1.
? TROUBLESHOOTING
25| Examine the outcomes of injections at 200 magnification, 58 h later. Oocyte activation produces a cytoplasm that is
darker and more heterogeneous, and the plasma membrane is no longer smooth but irregular (Fig. 4r). The mII bump will have
disappeared and there will be a second polar body (Pb2), the Pb1 having degenerated, although a remnant often remains. Male
and female pronuclei will be visible, sometimes containing multiple pronucleoli whose numbers decrease with time. Visible light
damages these structures, so for further culture, examine zygotes rapidly, wash them 46 times in CO2-equilibrated culture
medium (e.g., KSOM) and return them to the CO2 incubator.
? TROUBLESHOOTING
26| Next day (corresponding to embryonic day 1.5, E1.5), wash 2-cell embryos once in CO2-equilibrated culture medium
(removing dead cells) and return them to the CO2 incubator to monitor in vitro development further or transfer them to a
surrogate mother. Embryo development in vitro following piezo ICSI to BE3.5 is shown in Supplementary Video 3 online.

TIMING
The first (PMSG) injection for superovulation is administered three days prior to piezo. For a piezo micromanipulation
experiment on Monday (morning), mice should be injected with PMSG on Friday (afternoon or evening) and hCG on Sunday
(afternoon or evening), for example. After collection (taking 3060 min), oocytes are placed in CO2-equilibrated medium in a
CO2 incubator, after which spermatozoa are collected (510 min) and then the micromanipulation dish setup (5 min).
Experienced workers place 1520 oocytes in an injection droplet and inject them sequentially, which takes 1020 min. At least
5 min after injection of the last oocyte, the entire batch is washed in drops of equilibrated culture medium and returned to the
CO2 incubator. From oocyte collection to injection of the final sample, piezo ICSI typically takes 24 h for 40120 oocytes. The
timing of the entire procedure, including superovulation, is summarized in Table 1.
? TROUBLESHOOTING
Step 11: Sperm are difficult to collect
Selection of motile sperm from a freshly prepared suspension ensures that they were viable just prior to injection, maximizing
developmental potential. However, if they are difficult to pick up, they can be immobilized by touching them and applying one
or two high or low intensity piezo pulses.
Steps 11, 17 and 21: Piezo pulses have no effect
This common problem has multiple possible causes. (i) The needle is inadequate. If the tip is not flush-ended, only part of it
contacts the sample. Critical flaws in pipette fabrication or damage during use may not be visible microscopically but they
require that the needle is replaced. (ii) Hg0 or Fluorinert is too far from the tip. Piezo power dissipates the further the Hg0 or
Fluorinert front is from the pipette tip. To remedy the problem, advance the front. As a guide, there should be sufficient power
even when the Hg0 front is outside the field (as a guide, this corresponds to B500 mM from the tip at 200 magnification
when the front is just outside the field). (iii) Air bubbles interrupt the Hg0 or Fluorinert column. (iv) The pipette is
inadequately secured within its housing and/or the housing itself is not tightly mounted onto the workstation (everything
needs to be firmly secured). (v) Failure to engage the sample, often because the pipette tip is either above or below it; the
plane of focus should pass through (or near) the oocyte center and the pipette tip should be in the same plane. To achieve this,
bring the sample periphery (e.g., outer face of the zona pellucida or plasma membrane) into sharp focus and then adjust the tip
TABLE 1 | A timeline for mouse piezo ICSI.
Step 1
Day 3
PMSG

Day 1
hCG

(67 pm)

(67 pm)

Steps 26

Steps 79

Oocyte
collection
(79 am)

Day 0
Sperm
collection
(99:15 am)

Steps 1024

Step 25

ICSI

1215 h

302 | VOL.2 NO.2 | 2007 | NATURE PROTOCOLS

24 h

Day 1
Wash embryo
Embryo transfer

Day 3

2-cell

Morula (early)
Blastocyst (later)

(am)
Zygote

4554 h

Step 26

58 h

B1224 h

48 h

PROTOCOL

2007 Nature Publishing Group http://www.nature.com/natureprotocols

along the z-axis so that it falls sharply into the same focus. (vi) The pipette tip may not be in full contact because the sample
surface is irregular, in which case a small negative pressure (suction) within the microinjection pipette may help the two to
engage. However, high negative pressures and/or piezo settings should be unnecessary and are likely to damage samples.
(vii) In some models it is possible to set piezo actuation in the wrong direction so that the needle is caused to move in reverse
rather than forwards; flip the L to R setting at the rear of the transformer. (viii) The piezo unit may be brokenit is sensitive
to being knocked or dropped even a short distance. When working well with Hg0, the lowest intensity setting may be adequate
for all piezo steps.
Steps 11, 17 and 21: Piezo pulses violently displace the microinjection needle tip
If the needle tip and/or sample appear to jump when piezo pulses are applied, either the setting is too high (use a lower
setting), the assembly may not be fastened securely or the pipette has been filled with injection medium (not PVP360 solution)
near its end. Relatively dense solutions are apparently needed near the pipette tip to add momentum and stability, ensuring
that the piezo pulse is efficiently transmitted in a controlled manner.
Steps 12 and 22: Sperm heads stick to each other and the microinjection pipette interior
Samples stick to the pipette interiorparticularly as they get nearer the tipif the needle is too narrow; an internal diameter
of B6.5 mm should be comfortably within the acceptable range. If the problem persists, change the needle for one with a wider
tip. Sample stickiness is affected by the preparation method. For example, extraction with some detergents (e.g., 1% (w/v)
SDS) causes sperm stickiness whereas others (e.g., 1% (v/v) Triton X-100) tend not to. If sperm heads stick to each other, load
the pipette with fewer, possibly only one head. Increase the concentration of PVP360 (next time). Heads that stick to
the pipette can sometimes be dislodged with 13 low intensity piezo pulses. If this fails, apply slight suction and repeat the
piezo pulse, before pushing, again with a piezo pulse. Alternatively, toggle back to the high (zona drilling) piezo
intensity with the foot switch and apply a short burst of stronger pulses. Although higher intensity piezo applied to the
plasma membrane typically kills the oocyte, once the pipette tip is within the cytoplasm, the oocyte can usually tolerate
a higher intensity. If these remedies are unsuccessful the pipette interior may be dirty, so withdraw from the oocyte
(which must be discarded) and wash the needle thoroughly by expelling several drops of Hg0 or Fluorinert in PVP360 solution,
or change the needle.
Step 13: The mII plate is not visible
When learning, we recommend the use of B6D2F1 oocytes, in which the mII plate (containing the spindle and chromosomes)
is a relatively discernible (compared with that of some other strains) translucent disc lying just under the plasma membrane,
making a pronounced cortical bump. The disc is easier to see at higher temperatures because spindle assembly is more
complete. However, stage temperatures above 30 1C cause oocyte lysis after microinjection. Injection works well at room
temperatures in the range 1628 1C.
Step 20: The oocyte plasma membrane breaks as the microinjection tip is advanced
Do not apply piezo pulses until the pipette tip is fully advanced to the opposite side of the oocyte. If breakage occurs to the
membranes of more than one oocyte during or immediately after advance even without piezo, the tip may be damaged and the
pipette should be replaced. Occasionally, batches of oocytes are idiopathically hypersensitive and have to be discarded.
Step 23: The oocyte plasma membrane sticks to the microinjection pipette
If this occurs, the membrane is pulled outwards as the pipette is withdrawn to produce a surface exit wound resembling an
explosion, eviscerating the oocyte (Fig. 4o,p insets). Not surprisingly, these oocytes usually die. The problem can be
reduced by positioning the sperm sufficiently far back prior to pipette advance (Step 20), and exerting a constant (but
marginal) positive pressure during advance. This displaces injection media from the advancing tip to make a sleeve around the
needle, reducing adherence between the needle and the plasma membrane. Additionally, needles can be purged between
injections; interiors can be cleansed by displacing Hg0 or Fluorinert in the PVP solution washing drop, and exteriors by rubbing
the shafts on Hg0 drops (in the PVP drop) or within the aperture of the holding pipette (in a drop of CZB H). If stickiness
persists, change the needle.

Step 24: Many oocytes are dead

This is normal for beginners. Improvement requires stamina built up through regular practice. One or two injection sessions per
week are insufficient for most learners to achieve this; four or five times per week (without gaps between weeks) is more likely
to yield success. Injection sessions are typically limited to one per day. A high proportion of manipulated oocyte deaths can
occur even for experienced operators. This probably reflects submicroscopic needle damage or injectionist fatigue, which is why
stamina is necessaryin particular with respect to optical focusing on samples. As an illustration of this, oocyte survival may

NATURE PROTOCOLS | VOL.2 NO.2 | 2007 | 303

PROTOCOL
decline as the experiment proceeds, which should be controlled against where appropriate. The number of oocytes killed is also
generally proportional to the needle tip diameter.

2007 Nature Publishing Group http://www.nature.com/natureprotocols

Step 25: Pronucleus formation has not occurred after 6 h

This is typically because the plasma membrane was not broken in Step 21. Also, oocyte activating capacity declines with sperm
sample aging and is abolished by certain treatments, such as exposure to dimethyl sulphoxide or heating at 48 1C in NIM23,26.
ANTICIPATED RESULTS
Survival rates 30 min after injection should approach 100%, but this value declines with increasing microinjection pipette tip
diameter. Optimally, tip inner diameters are 5.57.5 mm for sperm injection. Oocyte activation following ICSI also promotes cell
lysis independently of injection per se; lysis is more likely if the oocytes were activated by, or soon after, injection. However, it
is not uncommon for 80% or more of injected oocytes to yield pronuclear zygotes and experienced operators can produce
developmentally competent zygotes in excess of 90%. These can produce offspring in up to B50% of cases following embryo
transfer (2030% is more common), depending on how skillfully the transfer is performed and to what host. ICR (CD-1) is a
favored recipient strain because it supports pre- and post-term development even when litter sizes are small (as low as one).

Note: Supplementary information is available via the HTML version of this article.
ACKNOWLEDGMENTS Grant support from the RIKEN Presidents Fund is gratefully
acknowledged.
COMPETING INTERESTS STATEMENT The authors declare competing financial
interests (see the HTML version of this article for details).
Published online at http://www.natureprotocols.com
Reprints and permissions information is available online at http://npg.nature.com/
reprintsandpermissions
1. Kimura, Y. & Yanagimachi, R. Intracytoplasmic sperm injection in the mouse. Biol.
Reprod. 52, 709720 (1995).
2. Wakayama, T., Perry, A.C.F., Zuccotti, M., Johnson, K.R. & Yanagimachi, R. Fullterm development of mice from enucleated oocytes injected with cumulus cell
nuclei. Nature 394, 369374 (1998).
3. Onishi, A. et al. Pig cloning by microinjection of fetal fibroblast nuclei. Science
289, 11881190 (2000).
4. Wakayama, T., Whittingham, D.G. & Yanagimachi, R. Production of normal
offspring from mouse oocytes injected with spermatozoa cryopreserved with or
without cryoprotection. J. Reprod. Fertil. 112, 1117 (1998).
5. Wakayama, T. & Yanagimachi, R. Development of normal mice from oocytes
injected with freeze-dried spermatozoa. Nature Biotechnol. 16, 639641 (1998).
6. Munsie, M.J. et al. Isolation of pluripotent embryonic stem cells from
reprogrammed adult mouse somatic cell nuclei. Curr. Biol. 10, 989992 (2000).
7. Perry, A.C.F. et al. Mammalian transgenesis by intracytoplasmic sperm injection.
Science 284, 11801183 (1999).
8. Perry, A.C.F. et al. Efficient metaphase II transgenesis with different transgene
archetypes. Nature Biotechnol. 19, 10711073 (2001).
9. Moreira, P.N. et al. Efficient generation of transgenic mice with intact yeast
artificial chromosomes by intracytoplasmic sperm injection. Biol. Reprod. 71,
19431947 (2004).
10. Shoji, S. et al. Mammalian Emi2 mediates cytostatic arrest and transduces the
signal for meiotic exit via Cdc20. EMBO J. 25, 834845 (2006).
11. Kawase, Y. et al. Application of the piezo-micromanipulator for injection of
embryonic stem cells into mouse blastocysts. Contemp. Top. Lab. Anim. Sci. 40,
3134 (2001).
12. Tanemura, K. et al. Birth of normal young by microinsemination with frozenthawed round spermatids collected from aged azoospermic mice. Lab. Anim. Sci.
47, 203204 (1997).

304 | VOL.2 NO.2 | 2007 | NATURE PROTOCOLS

13. Kanatsu-Shinohara, M. et al. Generation of pluripotent stem cells from neonatal

mouse testis. Cell 119, 10011012 (2004).
14. Rybouchkin, A., Dozortsev, D., de Sutter, P., Qian, C. & Dhont, M. Intracytoplasmic
injection of human spermatozoa into mouse oocytes: a useful model to
investigate the oocyte-activating capacity and the karyotype of human
spermatozoa. Hum. Reprod. 10, 11301135 (1995).
15. Huang, T., Kimura, Y. & Yanagimachi, R. The use of piezo micromanipulation for
intracytoplasmic sperm injection of human oocytes. J. Assist. Reprod. Genet. 13,
320328 (1996).
16. Ogura, A. et al. Recent advances in the microinsemination of laboratory animals.
Int. J. Androl. 23 (Suppl 2): 6062 (2000).
17. Inoue, K. et al. Improved postimplantation development of rabbit nuclear transfer
embryos by activation with inositol 1,4,5-trisphosphate. Cloning Stem Cells 4,
311317 (2002).
18. Zou, X. et al. Production of cloned goats from enucleated oocytes injected
with cumulus cell nuclei or fused with cumulus cells. Cloning 3, 3137
(2001).
19. Choi, Y.H. et al. Production of nuclear transfer horse embryos by piezo-driven
injection of somatic cell nuclei and activation with stallion sperm cytosolic
extract. Biol. Reprod. 67, 561567 (2002).
20. Hirabayashi, M. et al. Offspring derived from intracytoplasmic injection of
transgenic rat sperm. Transgenic Res. 11, 221228 (2002).
21. Chatot, C.L. et al. An improved culture medium supports development of randombred 1-cell mouse embryos. J. Reprod. Fertil. 86, 679688 (1989).
22. Lawitts, J.A. & Biggers, J.D. Culture of preimplantation embryos. Methods
Enzymol. 225, 153164 (1993).
23. Perry, A.C.F., Wakayama, T. & Yanagimachi, R. A novel trans-complementation
assay suggests full mammalian oocyte activation is coordinately initiated by
multiple, submembrane sperm compartments. Biol. Reprod. 60, 747755
(1999).
24. Marston, J.H. & Chang, M.C. The fertilizable life of ova and their morphology
following delayed insemination in mature and immature mice. J. Exp. Zool. 155,
237251 (1964).
25. Kuretake, S., Kimura, Y., Hoshi, K. & Yanagimachi, R. Fertilization and
development of mouse oocytes injected with isolated sperm heads. Biol. Reprod.
55, 789795 (1996).
26. Yoshida, N., Brahmajosyula, M., Shoji, S., Amanai, M. & Perry, A.C.F. Epigenetic
discrimination by mouse metaphase II oocytes mediates asymmetric
chromatin remodeling independently of meiotic exit. Dev. Biol. 301, 464477
(2007).