In Vivo Two-photon Imaging of Cortical Neurons in Neonatal Mice


Your institution must subscribe to JoVE's Neuroscience section to access this content.

Fill out the form below to receive a free trial or learn more about access:



We present an in vivo two-photon imaging protocol for imaging the cerebral cortex of neonatal mice. This method is suitable for analyzing the developmental dynamics of cortical neurons, the molecular mechanisms that control the neuronal dynamics, and the changes in neuronal dynamics in disease models.

Cite this Article

Copy Citation | Download Citations

Mizuno, H., Nakazawa, S., Iwasato, T. In Vivo Two-photon Imaging of Cortical Neurons in Neonatal Mice. J. Vis. Exp. (140), e58340, doi:10.3791/58340 (2018).


Two-photon imaging is a powerful tool for the in vivo analysis of neuronal circuits in the mammalian brain. However, a limited number of in vivo imaging methods exist for examining the brain tissue of live newborn mammals. Herein we summarize a protocol for imaging individual cortical neurons in living neonatal mice. This protocol includes the following two methodologies: (1) the Supernova system for sparse and bright labeling of cortical neurons in the developing brain, and (2) a surgical procedure for the fragile neonatal skull. This protocol allows the observation of temporal changes of individual cortical neurites during neonatal stages with a high signal-to-noise ratio. Labeled cell-specific gene silencing and knockout can also be achieved by combining the Supernova with RNA interference and CRISPR/Cas9 gene editing systems. This protocol can, thus, be used for analyzing the developmental dynamics of cortical neurons, molecular mechanisms that control the neuronal dynamics, and changes in neuronal dynamics in disease models.


The precise wiring of neuronal circuits in the cerebral cortex is essential for higher brain functions including perception, cognition, and learning and memory. Cortical circuits are dynamically refined during postnatal development. Studies have investigated the process of cortical circuit formation using histological and in vitro culture analyses. However, the dynamics of circuit formation in living mammals has remained mostly unexplored.

Two-photon microscopy has been widely used for the in vivo analyses of neuronal circuits in the adult mouse brain1,2. However, owing to technical challenges, only a limited number of studies have addressed neuronal circuit formation in newborn mice. For example, Carrillo et al. performed the time-lapse imaging of climbing fibers in the cerebellum in the second postnatal week3. Portera-Cailliau et al. reported the imaging of axons in cortical layer 1 in the first postnatal week4. In the present study, we summarize a protocol for the observation of layer 4 cortical neurons and their dendrites in newborn mice. Results obtained by applying this protocol, which includes two methodologies, are reported in our recent publication5. First, we use the Supernova vector system5,6 for labeling individual neurons in the neonatal brain. In the Supernova system, the fluorescent proteins used for neuronal labeling are exchangeable and labeled cell-specific gene knockdown and editing/knockout analyses are also possible. Second, we describe a surgical procedure for cranial window preparation in fragile neonatal mice. Together, these methodologies allow the in vivo observation of individual neurons in neonatal brains.

Subscription Required. Please recommend JoVE to your librarian.


Experiments should be performed in accordance with the animal welfare guidelines prescribed by the experimenter's institution.

1. Preparation of Pups for Imaging

NOTE: Pups with sparsely labeled cortical neurons can be obtained by in utero electroporation (IUE) of Supernova vectors5,6. The Supernova system consists of the following two vectors: TRE-Cre and CAG-loxP-STOP-loxP-Gene X-ires-tTA-WPRE. In this system, sparse labeling relies on TRE leakage. In a sparse population of transfected neurons, TRE drives the weak expression of Cre and tTA. Subsequently, only in these cells, the expression of gene X is facilitated by a positive feedback of the tTA-TRE cycles. The achieved sparse and bright labeling allows the visualization of morphological details of individual neurons in vivo. Details of the IUE procedure are not described in this protocol since they have been described elsewhere7,8,9,10,11.

  1. Prepare timed-pregnant mice for IUE.
  2. Prepare a DNA solution for IUE. For sparse labeling with RFP, use a solution containing pK031:TRE-Cre (5 ng/µL) and pK029:CAG-loxP-STOP-loxP-RFP-ires-tTA-WPRE (1 µg/µL) or a solution containing pK031:TRE-Cre (5 ng/µL) and pK273:CAG-loxP-STOP-loxP-CyRFP-ires-tTA-WPRE (1 µg/µL).
    NOTE: Various proteins can be expressed in the labeled neurons using different combinations of vectors. Also, various genes can be knocked-down or knocked-out specifically in labeled cells5,6 (e.g., a series of vectors for the Supernova system are available from RIKEN BioResource Research Center and from Addgene).
  3. Perform regular IUE7,8,9,10,11 to label cortical neurons. For the labeling of layer 4 neurons, use embryonic day-14 embryos.
  4. Wait for pup delivery and growth.

2. Surgery

  1. Anesthetize the postnatal day-5 (P5) pup using isoflurane gas (1.0%). Perform a tail-pinch test to check the level of anesthesia. If the pup responds to the pinch, increase the isoflurane concentration (up to 2.0%) or wait until the response disappears. Maintain the pup's body temperature during surgery using a heating pad.
  2. Subcutaneously inject an analgesic (carprofen, 5 mg/kg).
  3. Sterilize the pup's skin covering the skull by wiping it with 70% ethanol three times.
  4. Remove approximately 20 mm2 of the skin covering the skull using scissors sterilized with 70% ethanol (Figure 1A).
  5. Remove the fascia of the skull using sterilized forceps and a clean, sterile cotton swab (Figure 1A).
  6. Apply tissue adhesive using loading tips to the incised skin surface to stop the bleeding. Do not apply tissue adhesive to the imaging area (Figure 1B) because this makes the opening of the skull more difficult.
  7. Place the pup on a heating pad (37 °C) and allow it to recover from anesthesia. Wait until the tissue adhesive has dried and solidified (approximately 30 min).
  8. If necessary, apply more tissue adhesive and wait for it to dry and solidify.

3. Cranial Window Preparation

  1. Anesthetize the pup using isoflurane (1.0% - 2.0%) and check the anesthesia level by a tail-pinch test.
  2. Carefully open the skull with a sterilized razor blade leaving the dura intact (1 mm in diameter) (Figure 1C). Use a gelatin sponge (cut into small pieces, approximately < 2 mm3, using sterilized scissors, and apply them using tweezers) soaked in cortex buffer12 (125 mmol/L NaCl, 5 mmol/L KCl, 10 mmol/L glucose, 10 mmol/L HEPES, 2 mmol/L CaCl2, and 2 mmol/L MgSO4; pH 7.4; 300 mOsm/L; room temperature) to stop the bleeding. When opening the skull, apply a cortex buffer to keep the brain surface moist.
  3. Remove any buffer and blood from the dural surface using a gelatin sponge. Apply a thin layer of 1.0% low-melting-point agarose (dissolved in cortex buffer) using yellow tips. Using a heat block machine, maintain the temperature of the agarose solution at 42°C until application.
    NOTE: The draining of buffer and blood must be performed from the side of the craniotomy while taking care that the dry gelatin sponge does not come in contact with the dura. Failure to do so may damage the dura.
  4. Apply a round glass coverslip (No. 1, 3 mm in diameter) onto the agarose gel layer. Remove all bubbles between the coverslip and the agarose gel layer by pouring an excess of agarose gel between them. Remove the excess gel protruding from under the coverslip using tweezers (Figure 1D and 1E).
  5. Secure the coverslip using dental cement (Figure 1F).
  6. Mix the cement powder and the cement liquid. Apply the mixture using yellow tips before it becomes solidified. Do not apply dental cement onto the dura, because this may damage the brain.
  7. Attach a sterile titanium bar (custom made, approximately 30 mg, see Figure 1G) on the cranial bone using dental cement. Align the titanium bar and the coverslip (on the surface of the dura) in parallel to easily capture images.
  8. Cover the exposed skull with dental cement (Figure 1H).
  9. Recover the pup from anesthesia. Keep it on a heater (37 °C) until the dental cement has solidified (1 h).

4. Two-photon Imaging

NOTE: The in vivo images in Figure 2 were acquired using a two-photon microscope with a titanium-sapphire laser (beam diameter [1/e2]2: 1.2 mm).

  1. Set the two-photon laser wavelength. For RFP excitation, use 1,000 nm (450 mW/mm2 at 400 µm of depth).
    NOTE: The laser power should be reduced as the z-position moves up.
  2. Wipe the surface of the coverslip with 70% ethanol.
  3. Anesthetize the pup using isoflurane (1.5% - 2.0%) and check the anesthesia level using a tail-pinch test.
  4. Attach the pup to the titanium plate on the imaging stage using a titanium bar on the pup's head (Figure 2A and 2B). Adjust the head such that the coverslip is parallel to the objective lens using the goniometer stage (Figure 2B). Maintain body temperature of the pup using a heating pad (37 °C).
  5. Set the isoflurane concentration to 0.7% - 1.0%.
    NOTE: A very high isoflurane concentration may cause accidental death of the pup during imaging.
  6. Place the imaging stage under the objective lens (20X, NA 1.0) of the two-photon microscope (Figure 2B and 2C).
  7. Apply one drop of water onto the coverslip. Use epi-fluorescence to locate the fluorescent protein-labeled neurons in the area where the dura has been exposed.
  8. Acquire z-stack images at 1.4-µm intervals. For layer 4 neuron imaging, set the z-width to 150 - 300 µm to image the entire dendritic morphology (Figure 2D and 2E). Use slow scanning and averaging to get clear images showing the neuronal morphology (it usually takes > 20 min to acquire the entire dendritic morphology).
    NOTE: The following parameters are recommended for imaging. Excitation wavelength: 1,000 nm, scanner: galvanometer type, dichroic mirror: 690 nm, emission filter: 575 - 620 nm bandpass, detector: GaAsP type, gain setting > 100, image size > 512 x 512 µm, field of view > 600 x 600 µm, pixel resolution < 1.2 µm.

5. Recovery and Nursing

  1. Detach the pup from the imaging stage.
  2. Place the pup on a heater (37 °C) and allow it to recover (15 min).
  3. Feed the pup warm milk using a micropipette at 2-h intervals and gently stimulate the stomach to allow excretion. Confirm that the pup is drinking milk by measuring its body weight.

6. Re-imaging

  1. Anesthetize the pup with isoflurane (1.5% - 2.0%), and check the anesthesia level using a tail-pinch test. Attach it to the imaging stage.
  2. Locate the previously imaged neurons and acquire a z-stack image. The identification of neurons is easy owing to their sparse labeling with Supernova.
  3. Repeat steps 5.1 - 6.2 until imaging is completed. NOTE: The pup can be imaged up to 18 h without losing weight.

Subscription Required. Please recommend JoVE to your librarian.

Representative Results

Figures 2D - 2F show representative results of two-photon time-lapse imaging of layer 4 cortical neurons using the present protocol. For the purpose of analysis, select neurons with clear dendritic morphology throughout the imaging periods. We analyzed the dendritic morphology of imaged neurons using morphological analysis software. Representative dendritic morphology reconstruction is shown in Figure 2F. Neurons showing disconnected dendrites (Figure 2G) should be excluded from analyses, because disconnected dendrites indicate cell death induced by damage during surgery or imaging. In addition, neurons with blurred dendritic tips should be excluded (e.g., the neuron with the arrowhead in Figure 2D).

Figure 1
Figure 1: Surgery, cranial window preparation, and attachment of the titanium bar. (A) This panel shows the removal of the skin covering the skull. (B) This panel shows the fixation of the gap between the skin and the skull. Be careful not to apply the bond to the imaging area. (C) This is an image of the exposed dura. A razor blade was inserted between the skull and the dura, and the skull was flapped open to the left of the exposed area (arrowhead). The bone can then be easily removed using forceps. (D) This is a schematic design showing a vertical view of the cranial window. The gap between the cover glass and the dura is filled with a thin layer of agarose gel. The coverslip is fixed to the skull using dental cement. (E) A round-shaped coverslip is placed on the agarose gel layer. (F) This is an image of the secured coverslip. (G) This panel shows the design of the titanium bar. The titanium bar contains two screw holes for attachment to the imaging stage (see Figure 2) and one flat rectangular part that is attached to the pup's head. (H) The rectangular part of the titanium bar is attached to the pup's skull using dental cement. Please click here to view a larger version of this figure.

Figure 2
Figure 2: In vivo two-photon imaging of cortical neurons in neonatal mice. (A) This panel shows the design of the titanium plate. The titanium plate has two screw holes for attaching the titanium bar and four screw holes for attaching the goniometer stage, which is placed on the imaging stage. 2.5 mm (in diameter) x 2 mm (in length) screws are used for fixation. (B) This is a representative image of the pup attached to the imaging stage. The pup's body temperature is maintained using a heater. (C) The pup is anesthetized using isoflurane during the in vivo imaging procedure. (D) This panel shows a representative Z-stack time-lapse image of layer 4 cortical neurons of a P5 pup. The arrowhead indicates the neuron with blurred dendrites, which should be removed from the analysis. (E) This panel shows higher magnification time-lapse images of the neuron with arrows in panel D. Blue arrowheads: dendritic tips that are retracted in 4.5 h, yellow arrowheads: dendritic tips that are elongated in 4.5 h, small white arrowheads: axon of a neighboring cell. (F) This is a 3-D model of the dendritic trace of neuron in the left figure of panel E. Blue circles indicate the position of the cell body. (G) This panel shows representative neurons with disconnected dendrites, which are not included in the analysis. The sample data in panels D - G contain NR1 (an essential subunit of NMDA-type glutamate receptor)-knocked-out neurons (because of limited data available from the authors). Please click here to view a larger version of this figure.

Subscription Required. Please recommend JoVE to your librarian.


Critical Steps in the Protocol and Troubleshooting:

The most critical step of the protocol is the removal of the skull (Protocol step 3.2). Upon insertion, the razor blade often adheres to the dura, causing dural bleeding and damage to the brain. This can be avoided by adding a drop of cortex buffer on the skull and removing the skull in cortex buffer.

Bleeding from the dura and the skin after cranial window preparation leads to occlusion of the window. To avoid this, the tissue adhesive and dental cement used should be allowed to completely dry before proceeding to the next step. Multiple pups with a cranial window should be prepared since it is difficult to completely avoid bleeding. In general, if the cranial window remains clear and stable for 2 - 3 h post-surgery, the pup can be used for time-lapse imaging.

Significance of the Method with Respect to Existing/Alternative Methods:

Here, we have described a method for in vivo two-photon imaging of the neonatal cortex. This protocol has several merits compared with previously reported methods. These are listed as follows.

1) Cortical neurons can be sparsely and brightly labeled by transfecting Supernova vectors using IUE. IUE has been widely used for the labeling of cortical neurons during developing stages. However, a simple IUE is unsuitable for the imaging of individual neurons since neurons may be labeled too densely7,8,9,10. Moreover, using the Supernova system, various types of fluorescent proteins can be used for labeling sparse populations of neurons. For example, using Supernova-mediated sparse labeling of a genetically encoded calcium indicator GCaMP13,14, we have recently performed a functional analysis of individual neurons in the developing cortex layer 4 (P3 to P13)15.

2) The method described in this study is suitable for elucidating the molecular mechanisms underlying neuronal circuit development. The Supernova system allows sparsely labeled cell-specific knockout of any gene. Thus, the dynamics of a neuron containing a specific gene disruption can be observed. For this purpose, genetics-based systems such as MADM16 and SLICK17 have been previously reported. However, because the breeding of mouse lines is essential for these systems, they require much time and cost than the protocol described here.

3) This study utilizes a razor blade for skull removal. This minimizes bleeding from the dura and allows the opening of a wide area of the skull (< 2 mm in diameter). Using this procedure, it was possible to observe the spontaneous activity of layer 4 neurons in the entire large barrel area within the somatosensory cortex15.

Limitations of the method:

A disadvantage of in vivo imaging of the mouse neonatal brain, compared with the imaging of transparent animals such as zebrafish larvae and Xenopus tadpoles, is the lower spatial and temporal resolution. Slow scanning and averaging should be performed for yielding clear images of the neuronal morphology because more light scattering occurs in the mouse brain. Light scattering may possibly be reduced using a longer wavelength laser for fluorescent protein excitation and proteins with longer fluorescence emission wavelengths.

Another limitation of the present protocol could be that the surgery for cranial window implantation may affect the formation of a normal cortical circuit due to brain inflammation12. However, there is a high likelihood that in vivo imaging is more physiological compared with in vitro imaging such as time-lapse imaging of the brain slice preparation, which should also give rise to severe inflammation by ischemia and slicing. It has also been reported that repetitive exposure to isoflurane may affect several neuronal processes18. Control experiments should be performed for verifying the appropriateness of results obtained by in vivo imaging. In the case of our imaging of layer 4 neurons in the somatosensory cortex, we confirmed a normal increase in total dendritic length and acquisition of orientation bias of dendritic projections of spiny stellate neurons5,20.

Recent studies reported > 1-mm-depth imaging in an adult mouse brain wherein the dura was removed during surgery19. On the other hand, we have been able to report up to 400-µm-depth imaging in neonates5. Since the dura cannot be removed in neonates and this, in turn, leads to a high light scattering, we consider that deep imaging in neonates is more challenging than in adults. Future improvement in fluorescent probes, lasers, and detectors should allow deep imaging in neonatal brains.

Thus far, we have reported 18-hour time-lapse imaging using the present protocol5. Recently, we have succeeded in 72-hour time-lapse imaging by improving the protocol20. We will continue to refine the protocol presented here (e.g., longer-term, higher time, and/or spatial resolution imaging) for revealing dynamic mechanisms of neuronal circuit development.

Subscription Required. Please recommend JoVE to your librarian.


The authors have nothing to disclose.


The authors thank T. Sato, M. Kanbayashi, and S. Kouyama for their technical assistance. This work was supported by JSPS KAKENHI Grant Numbers JP15K14322 and JP16H06143, the Takeda Science Foundation, the Uehara Memorial Foundation, and the Collaborative Research Project of Niigata University Brain Research Institute 2017-2923 (H.M.) and by KAKENHI JP16K14559, JP15H01454, and JP15H04263 and Grant-in Scientific Research on Innovation Areas "Dynamic regulation of Brain Function by Scrap & Build System" (JP16H06459) from MEXT (T.I.).


Name Company Catalog Number Comments
pK031. TRE-Cre Authors - Available from RIKEN BRC and Addgene
pK029. CAG-loxP-STOP-loxP-RFP-ires-tTA-WPRE Authors - Available from RIKEN BRC and Addgene
pK273. CAG-loxP-STOP-loxP-CyRFP-ires-tTA-WPRE Authors - Available from authors
Isoflurane Wako 099-06571
410 Anaesthesia Unit (isoflurane gas machine) Univentor 8323101
Vetbond (tissue adhesive) 3M 084-1469SB
MµltiFlex Round (loading tip) Sorenson 13810
Gelfoam (gelatin sponge) Pfizer 09-0353-01
Agarose Sigma A9793 Low melting point
Round-shaped coverslip Matsunami - Custom made
Unifast 2 (dental cement) GC -
Titanium bar Authors - Custom made (see Figure 1G)
Rimadyl (carprofen) Zoetis - Injectable
2-photon microscope Zeiss LSM7MP
Titanium-sapphire laser Spertra-Physics Mai-Tai eHPDS
Titanium plate Authors - Custom made (see Figure 2A)
IMARIS, FilamentTracer, MeasurementPro BITPLANE
Goniometer stage Thorlabs GN2/M



  1. Lendvai, B., Stern, E. A., Chen, B., Svoboda, K. Experience-dependent plasticity of dendritic spines in the developing rat barrel cortex in vivo. Nature. 404, (6780), 876-881 (2000).
  2. Grutzendler, J., Kasthuri, N., Gan, W. B. Long-term dendritic spine stability in the adult cortex. Nature. 420, (6917), 812-816 (2002).
  3. Carrillo, J., Nishiyama, N., Nishiyama, H. Dendritic translocation establishes the winner in cerebellar climbing fiber synapse elimination. The Journal of Neuroscience. 33, (18), 7641-7653 (2013).
  4. Portera-Cailliau, C., Weimer, R. M., De Paola, V., Caroni, P., Svoboda, K. Diverse modes of axon elaboration in the developing neocortex. PLoS Biology. e272 (2005).
  5. Mizuno, H., et al. NMDAR-regulated dynamics of layer 4 neuronal dendrites during thalamocortical reorganization in neonates. Neuron. 82, (2), 365-379 (2014).
  6. Luo, W., et al. Supernova: A Versatile Vector System for Single-Cell Labeling and Gene Function Studies in vivo. Scientific Reports. 6, 35747 (2016).
  7. Mizuno, H., Hirano, T., Tagawa, Y. Evidence for activity-dependent cortical wiring: formation of interhemispheric connections in neonatal mouse visual cortex requires projection neuron activity. The Journal of Neuroscience. 27, (25), 6760-6770 (2007).
  8. Saito, T., Nakatsuji, N. Efficient gene transfer into the embryonic mouse brain using in vivo electroporation. Developmental Biology. 240, (1), 237-246 (2001).
  9. Tabata, H., Nakajima, K. Efficient in utero gene transfer system to the developing mouse brain using electroporation: visualization of neuronal migration in the developing cortex. Neuroscience. 103, (4), 865-872 (2001).
  10. Mizuno, H., Hirano, T., Tagawa, Y. Pre-synaptic and post-synaptic neuronal activity supports the axon development of callosal projection neurons during different post-natal periods in the mouse cerebral cortex. European Journal of Neuroscience. 31, (3), 410-424 (2010).
  11. Matsui, A., Yoshida, A. C., Kubota, M., Ogawa, M., Shimogori, T. Mouse in utero electroporation: controlled spatiotemporal gene transfection. Journal of Visualized Experiments. (54), e3024 (2011).
  12. Holtmaat, A., et al. Long-term, high-resolution imaging in the mouse neocortex through a chronic cranial window. Nature Protocols. 4, (8), 1128-1144 (2009).
  13. Nakai, J., Ohkura, M., Imoto, K. A high signal-to-noise Ca(2+) probe composed of a single green fluorescent protein. Nature Biotechnology. 19, (2), 137-141 (2001).
  14. Chen, T. W., et al. Ultrasensitive fluorescent proteins for imaging neuronal activity. Nature. 499, (7458), 295-300 (2013).
  15. Mizuno, H., et al. Patchwork-Type Spontaneous Activity in Neonatal Barrel Cortex Layer 4 Transmitted via Thalamocortical Projections. Cell Reports. 22, (1), 123-135 (2018).
  16. Zong, H., Espinosa, J. S., Su, H. H., Muzumdar, M. D., Luo, L. Mosaic analysis with double markers in mice. Cell. 121, (3), 479-492 (2005).
  17. Young, P., et al. Single-neuron labeling with inducible Cre-mediated knockout in transgenic mice. Nature Neuroscience. 11, (6), 721-728 (2008).
  18. Liu, J., et al. Neonatal Repeated Exposure to Isoflurane not Sevoflurane in Mice Reversibly Impaired Spatial Cognition at Juvenile-Age. Neurochemical Research. 42, (2), 595-605 (2017).
  19. Kondo, M., Kobayashi, K., Ohkura, M., Nakai, J., Matsuzaki, M. Two-photon calcium imaging of the medial prefrontal cortex and hippocampus without cortical invasion. eLIFE. 6, e26839 (2017).
  20. Nakazawa, S., Mizuno, H., Iwasato, T. Differential dynamics of cortical neuron dendritic trees revealed by long-term in vivo imaging in neonates. Nature Communications. 9, (1), 3106 (2018).



    Post a Question / Comment / Request

    You must be signed in to post a comment. Please or create an account.

    Usage Statistics