Primary sensory areas in the neocortex exhibit unique spontaneous activities during development. This article describes how to visualize individual neuron activities and primary sensory areas to analyze area-specific synchronous activities in neonatal mice in vivo.
The mammalian brain undergoes dynamic developmental changes at both the cellular and circuit levels throughout prenatal and postnatal periods. Following the discovery of numerous genes contributing to these developmental changes, it is now known that neuronal activity also substantially modulates these processes. In the developing cerebral cortex, neurons exhibit synchronized activity patterns that are specialized to each primary sensory area. These patterns markedly differ from those observed in the mature cortex, emphasizing their role in regulating area-specific developmental processes. Deficiencies in neuronal activity during development can lead to various brain diseases. These findings highlight the need to examine the regulatory mechanisms underlying activity patterns in neuronal development. This paper summarizes a series of protocols to visualize primary sensory areas and neuronal activity in neonatal mice, to image the activity of individual neurons within the cortical subfields using two-photon microscopy in vivo, and to analyze subfield-related activity correlations. We show representative results of patchwork-like synchronous activity within individual barrels in the somatosensory cortex. We also discuss various potential applications and some limitations of this protocol.
The cerebral cortex contains several sensory areas with distinct functions. The areas receive inputs originating from their corresponding sensory organs, mostly conveyed through the spinal cord or brainstem and relayed via the thalamus1,2. Notably, neurons in each primary sensory area exhibit uniquely synchronized activity during early developmental stages, which also originate from sensory organs or the lower nervous centers, but essentially differ from the activities observed in the mature cortex3.
In neonatal rodents, for example, the primary visual area (V1) displays wave-like activity, which originates in the retina (retinal wave) and propagates through the entire visual pathway while conserving retinotopy4. The primary auditory area (A1) exhibits synchronous activity organized in band-shaped subregions that correspond to the isofrequency bands in the mature brain. The activity emanates from the cochlea's inner hair cells5,6. The barrel cortex in the primary somatosensory area (S1) shows a patchwork-like activity pattern in which layer 4 neurons within individual barrels, namely, neurons responsive to individual whiskers, are synchronously activated7. Although proposed to originate from the trigeminal ganglion, the source of the activity remains unknown7. Consequently, neonatal activity patterns are specialized both within each primary sensory area and within intra-areal subfields. The simultaneous visualization of neuronal activity and structure of primary sensory areas may facilitate an inquiry into the contribution of these activity patterns to the development of sensory systems.
In this article, we summarized a series of protocols: (1) to visualize individual neuronal activities using sparse labeling of GCaMP and primary sensory areas using TCA-RFP mice that express red fluorescent protein in thalamocortical axons7, (2) to image single cell-level activity in neonatal mice using two-photon microscopy in vivo, and (3) to analyze the activity correlations within S1 barrel cortex. The representative results show patchwork-like synchronized activity within individual barrels of a postnatal day (P)6 mouse. Despite some limitations, this technique can be used for chronic imaging, wide-field imaging across multiple sensory areas, and various manipulation experiments. The multifaceted analysis of neuronal activity during development will enrich our comprehension of brain circuit formation mechanisms.
All the experiments were conducted in accordance with the guidelines for animal experimentation of Kumamoto University and the National Institute of Genetics and approved by the animal experimentation committees.
1. In utero electroporation (IUE)
2. Cranial window surgery
3. Two-photon calcium imaging
4. Analysis
Figure 1 shows the representative results of layer 4 neuron activities in the barrel cortex of a P6 pup visualized using the present protocol. Two-photon images of the green channel (GCaMP) and red channel (TCA-RFP) were temporally averaged and shown in Figure 1A. Because TCA-RFP fluorescence was much weaker than GCaMP fluorescence, the GCaMP signal leaked into the red channel (Figure 1A1,A2). Fourteen ROIs were drawn on GCaMP-expressing neurons (Figure 1A4). Though the RFP image was clear enough (Figure 1A2), to define barrel edges more precisely, a z-projection image from the 100 µm-thick tangential slice taken by one-photon confocal microscopy was used (Figure 1B,C). The dF/F of 411 second-imaging window, starting ten minutes after stopping isoflurane, is shown in Figure 1D. Movies of GCaMP signals in the same imaging window before and after the movement compensation are also provided (Supplemental Video S1 and Supplemental Video S2).
Figure 2 shows the analysis of activity correlation. The pairwise correlation coefficients within the same barrels (ROI#1-2, ROI#3-7, and ROI#11-12) were higher than those across different barrels (Figure 2A–D). Though the distance from ROI#3 to ROI#4-7 (the same barrel neurons) was longer than that to ROI#12 (a different barrel neuron) (Figure 1C), the correlation was much stronger in the former (Figure 2A,C), indicating that activities synchronized more within individual barrels. The average of correlation coefficients within the same barrels was significantly higher than that calculated from 10,000 sets of surrogate data (Figure 2E). The correlation within the same barrels was significantly strong among three different time windows (Figure 2F,G; compared with surrogate data, all P < 0.001), suggesting that the patchwork-like pattern is independent of anesthetic depth. The septal neuron activity was correlated with both barrel neurons and septal neurons (Figure 2A–C). The specificity of activity correlation from septal neurons to other neurons is yet to be examined.
Figure 1: Calcium activity of barrel area neurons at P6. (A) Two-photon images averaged within the imaging session. (A1) GCaMP signal in the green channel. (A2) RFP signal in the red channel. (A3) Merged image of green and red channels. (A4) Thirteen ROIs drawn on the GCaMP image. (B) One-photon image of the fixed brain tangential slice. (C) The barrel edges defined on the one-photon image. The ROIs in A4 were aligned. The barrel IDs (B1, C2, C3, and D2) are shown.(D) The dF/F of each ROI in a 411-second imaging session starting ten minutes after stopping isoflurane. The location of each ROI (barrel ID or septa) is shown on the left. Scale bars = 100 µm (A1,A2,A3,A4, B,C). Abbreviations: RFP = red fluorescent protein; ROIs = regions of interest; S = septa. Please click here to view a larger version of this figure.
Figure 2: Analysis of calcium activity correlations. (A) The pairwise Pearson's correlation coefficient between all ROIs shown in Figure 1. (B) The P-value of correlation between positively correlated pairs of ROIs. (C) Schematic diagram of correlation strength between the ROIs. The neurons with significant (P < 0.05) positive correlation were connected by lines whose width describes their correlation coefficients. (D) The pairwise correlation coefficient within the same barrels and that across different barrels are plotted. Mean ± S.E.M and P-value of a two-tailed unpaired t-test are shown. (E) The correlation coefficients within the same barrels. Red arrow: actual data; gray bars: histogram of 10,000 surrogate data. (F) The correlations in three separate time windows are shown as in C. Left; first window (1-137 s). Middle; second window (138-274 s). Right; third window (275-411 s).(G) The correlation coefficients are shown as in D for all three windows. Abbreviation: ROIs = regions of interest. Please click here to view a larger version of this figure.
Supplemental Video S1: Patchwork-like activity of L4 neurons in a 411 s imaging session at P6 (10-fold speed) before the movement compensation. Please click here to download this Video.
Supplemental Video S2: Patchwork-like activity of L4 neurons in a 411 s imaging session at P6 (10-fold speed) after the movement compensation. Please click here to download this Video.
Given that the spontaneous activities emerge from the sensory organ or lower nervous system and travel to the primary sensory area through a pathway equivalent to that of a mature nervous system3, it is crucial to define the primary sensory area and the location of imaged neurons within the area. In this protocol, we addressed this requirement by employing transgenic mice that visualize thalamocortical axons and the Supernova system that expresses GCaMP sparsely8. These techniques enabled the analysis of activity patterns within each sensory area as well as within the intra-areal subfields, offering a means of investigating the developmental mechanisms of sensory systems.
While this protocol summarizes how to observe patchwork-like activity within the S1 barrel area, it is also applicable to observe unique activity patterns in other cortical areas, as demonstrated in V17. Some activities synchronize or propagate across multiple cortices, which have mainly been studied using single-photon imaging3,17,18,19. The recently established wide-field two-photon imaging techniques20,21, combined with gene-introducing techniques using virus vectors and transgenic mice22,23, will enable the simultaneous assessment of area-specific and inter-areal activities at a single-cell resolution.
For the analysis of interneuronal correlations, signal contamination between neurons can lead to erroneous outcomes. One solution is to express GCaMP sparsely, as we performed with the Supernova system8. Another solution is to reduce signal contamination using computational methods24 or cell body-targeted GCaMP expression25. Additionally, while ROIs can be determined manually with Fiji or other software7,14, computer programs with appropriate algorithms provide more rigorous and reproducible ROI settings. Automated segmentation tools like CaImAn26, Suite2p27, and EZcalcium15 can enable such settings. Experimenters should consider employing a suitable tool based on the data type they deal with, available software, and programming skills.
TCA-RFP mice enabled us to delineate primary sensory areas and layer 4 during imaging7, ensuring precise imaging. However, transgenic mice's introduction and timed mating can be time-consuming and demanding. Alternatively, primary sensory areas can be visualized after imaging by VGluT2 immunostaining that labels thalamocortical axon terminals. A two-dimensional arrangement of these areas can be visualized with tangential slices and imaged by one-photon confocal microscopy7.
A limitation of this protocol is that isoflurane anesthesia affects spontaneous neuronal activity in neonates4,7,28. In the representative results, activities within the same barrels remained consistently synchronized as the anesthesia lessened, suggesting that the patchwork-like pattern is resilient to variations in anesthetic depth. Nevertheless, a new technique is required to remove restraint stress from neonatal pups and enable fully awake state imaging.
The synchronous activity patterns are primarily shaped by peripheral activity via thalamocortical inputs3,4,17, which may be altered as intracortical connections mature and come to predominate within the cortical network29,30,31. Intracortical connection types may also contribute to the pattern changes. The connections between cortical excitatory neurons are largely governed by gap junctions that facilitate synchronous activity during the first postnatal week16,32,33,34,35, which are replaced by recurrent synaptic connections in the second postnatal week32,35. These transitions may change activity correlations, which can be evaluated by neuronal assembly detection techniques36 (also see37). Ideally, chronic imaging spanning two weeks may fully assess such changes in the same animal. However, chronic neonatal imaging is currently limited to up to four days38, due to the resin encasing that affects brain development and the inflammatory response that lowers pial surface transparency. A novel technique to image neuronal activities in broader space and time will lead to a better understanding of brain circuit formation.
The authors have nothing to disclose.
This work was supported by the Japan Society for the Promotion of Science Grants-in-Aid for Transformative Research Areas (B) (22H05092, 22H05094) and for Scientific Research Grants 20K06876, AMED under Grant Number 21wm0525015, the Takeda Science Foundation, the Naito Foundation, the Kato Memorial Bioscience Foundation, the Kowa Life Science Foundation, NIG-JOINT (24A2021) (to H.M.); and Japan Society for the Promotion of Science Grants-in-Aid for Scientific Research Grants 19K06887 and 22K06446, the Kodama Memorial Fund for Medical Research, the Uehara Memorial Foundation, the Kato Memorial Bioscience Foundation, and the Takeda Science Foundation (to N.N-T.). We thank Dr. Takuji Iwasato for the TCA-RFP mice.
20× objective lens (water immersion) | |||
250 mL Vacuum Filter/Storage Bottle System | Corning | 431096 | |
4%-paraformaldehyde phosphate buffer solution (4% PFA) | Nacalai | 09154-85 | |
Acrylic resin (UNIFAST II) | GC | N/A | |
Agarose | Sigma | A9793 | |
Aspirator tube assembly | Drummond | 2-040-000 | |
CaCl2•2H2O | Nacalai | 06731-05 | |
Electroporator | BEX | GEB14 | |
Eye drop (Scopisol) | Senju Pharmaceutical | N/A | |
Fluorescence stereo microscope | Leica | M165FC | |
Glucose | Nacalai | 16806-25 | |
Heating pad | Muromachi Kikai | FHC-HPS | |
HEPES | Gibco | 15630-080 | |
Isoflurane | Pfizer | N/A | |
KCl | Nacalai | 28514-75 | |
MgSO4•7H2O | Wako | 131-00405 | |
Micropipette puller | Narishige | PC-100 | |
Multiphoton laser | Spectra-Physics | Mai Tai eHP DeepSee | |
Multiphoton microscope | Zeiss | LSM 7MP | |
NaCl | Nacalai | 31320-05 | |
Non-woven fabric (Kimwipe) | Kimberly Clark | S-200 | |
Phosphate buffered saline (PBS) | Nacalai | 27575-31 | |
Plasmid: CAG-loxP-STOP-loxP-GCaMP6s-ires-tTA-WPRE | Addgene | pK175 | |
Plasmid: TRE-nCre | Addgene | pK031 | |
Precision calibrated micropipets | Drummond | 2-000-050 | |
Razor blade | Feather | FA-10 | |
Rimadyl (50 mg/mL Carprofen) | Zoetis JP | N/A | |
Round cover glass, 3-mm-diameter | Matsunami | CS01078 | |
Saline | Otsuka | 035175315 | |
Sodium pentobarbital | Nacalai | 26427-72 | |
Stage for imaging living pup (two single-axis translation stage for XY positioning, two-axis goniometer, base plate, adjustable pillar for z positioning) | ThorLabs | LT1/M, GN2/M, BM2060/M, MLP01/M | |
TCA-RFP mouse | N/A | N/A | Mizuno et al., 2018a |
Tissue adhesive (Vetbond) | 3M | 1469SB | |
Titanium bar | Endo Scientific Instrument | N/A | Custom made (Mizuno et al., 2018b) |
Titanium bar fixing plate | N/A | Custom made (Mizuno et al., 2018b) | |
Trypan blue | Sigma | T8154 | |
Tweezers with platinum plate electrode, 5 mm diameter | BEX | CUY650P5 | |
Wild-type ICR mouse | Nihon SLC | Slc:ICR |