Article Figures & Data
Figures
- Figure 1.
Three subdivisions of the caudal NM revealed by MAP2 immunoreactivity. A–D, Low-magnification images taken from the caudomost (A), caudal (B), middle (C), and rostral (D) regions of NM at the coronal plane. To visualize MAP2 staining in NM, the images were saturated in the surrounding tissues that are stained more strongly for MAP2 immunoreactivity than NM and NL. Dashed lines outline the border of NM. E–H, High-magnification images of the caudolateral NM. Dashed white and yellow lines outline the border of NMc1 and NMc2, respectively. Images in E and F were taken from the level between A and B. Image in G was taken from the same section in B, whereas the image in H is at a level slightly rostral to B and G. Note distinct staining pattern of MAP2 between NMcm, NMc1, and NMc2. For each image, right is lateral and up is dorsal. I, The relative location of NMc1 and NMc2 along the caudal–rostral axis in series coronal sections through NM. Abbreviations: l, lateral; d, dorsal; NM, nucleus magnocellularis; NMcm, caudomedial NM; NMc1, caudolateral NM subregion 1; NMc2, caudolateral NM subregion 2. Scale bars = 200 μm in D (applies to A–D) and 100 µm in H (applies to E–H).
- Figure 2.
Comparison of neuronal cell body size in NMcm, NMc1, and NMc2. A–C, NeuroTrace stain in NMcm (A), NMc1 (B), and NMc2 (C) on coronal sections. White dashed circles illustrate examples of measured neurons. D, Low-power image of MAP2 immunostaining on the section containing the three subregions. The NMc1 is outlined with dashed line. E, Projection of 3D color map surface plot representing the somatic area of NM neurons in relation to their location on the section shown in D. Warm colors represent larger cells. The NMc1 in D is indicated accordingly by dashed line. Arrow indicates a group of large cells along the lateral edge of NMc2. F, Bar chart of the cross-sectional relative somatic areas in NMcm, NMc1, and NMc2. ***, significant difference (P < 0.001). Data are presented as mean ± SD. Abbreviations: see Fig. 1. Scale bar = 20 μm in C (applies to A–C) and 100 μm in D.
- Figure 3.
Single-cell dye-filling shows different dendritic morphology in NMcm, NMc1, and NMc2. A, An example slice containing several filled neurons in different subregions along the lateral-to-medial axis. The black dashed circle outlines NM. Cell bodies of the filled neurons (red) are evident in this low-magnification image. B–D, Higher magnification of the boxes in A showing maximum z-projection of filled neurons in NMc2 (B), NMc1 (C), and NMcm (D). E, F, 3D reconstruction of the filled neurons in B. G, 3D reconstruction of the filled neurons in C. H, 3D reconstruction of the filled neuron in D. I, Quantitative analysis of the total dendritic branch length. J, Quantitative analysis of the number of primary trees. ***, P < 0.001; **, P < 0.01; *, P < 0.05. Data are presented as mean ± SD. Abbreviations: see Fig. 1. Scale bars = 200 μm in A and 20 µm in H (applies to B–H).
- Figure 4.
CCK is a biomarker for NMc2. The left (A1, B1, C1, D1) and middle (A2, B2, C2, D2) columns are MAP2 and CCK immunostaining, respectively. The right column (A3, B3, C3, D3) shows the merged images. A–C, Low-magnification images were taken from sections located caudal to rostral from the same animal. Dashed lines outline NMc1. Arrows in B2 indicate CCK-positive neurons in NMc1. D, High-magnification images of the box in A3. Arrows and arrowheads indicate darkly and lightly CCK-labeled NMc2 neurons, respectively. E, Bar chart of the mean grayscale of CCK-expressing neurons in NMcm, NMc1, and NMc2. F, Bar chart of the percentage of CCK-immunoreactive neurons in NMcm, NMc1, and NMc2. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant. Data are presented as mean ± SD. Abbreviations: see Fig. 1. Scale bars = 100 μm in C3 (applies to A1–C3) and 20 μm in D3 (applies to D1–D3).
- Figure 5.
NMc1 and NMc2 receive excitatory inputs from auditory nerve fibers. A, Schematic drawing shows injection sites of BDA in auditory and vestibular branches of the eighth nerve. See details in Materials and Methods. B, The injection site (white arrow) in the cochlear branch outlined by dashed lines. C, BDA-labeled axons and terminals in NMcm. The inset shows the end bulb morphology of a labeled terminal. D, BDA-labeled axons and terminals in NMc1 and NMc2. The images were taken from a section at the level of Fig. 1E. The inset shows the bouton-like morphology of labeled terminals. E–H, BDA-labeled axons and terminals in NMcm, NMc1, and NMc2 at the level of Fig. 1F. NMc1 is outlined by dashed line. F–H are closer views of NMcm (F), NMc1 (G), and NMc2 (H). I–L, Double-labeling of BDA (green) and the excitatory synaptic marker SNAP25 (magenta). J–L are closer views of NMcm (J), NMc1 (K), and NMc2 (L). Arrows in J indicate a number of BDA-labeled end bulbs double-stained with SNAP25. M, No labeling was observed in vestibular nuclei after injections in the cochlear branch. N, The injection site (white arrow) in the vestibular branch outlined by dashed lines. O, No labeling in NM after the injection in N. P, BDA-labeled terminals in the vestibular nucleus ventral to NM. Inset shows a labeled terminal around a vestibular neuron. Abbreviations: BDA, dextran; NeuT, NeuroTrace; Ve, vestibular nucleus; NA, nucleus angularis; NL, nucleus laminaris. Other abbreviations: see Fig. 1. Scale bars = 100 μm in B, M, N, O, and P; 50 μm in C, D; 100 µm in I (applies to E and I); 20 μm in L (applies to F–H and J–L); and 10 μm in insets.
- Figure 6.
NMc1 and NMc2 receive inhibitory inputs from SON. A–D, Distribution pattern of inhibitory synaptic marker gephyrin. NMc1 is outlined by dashed line. B–D are high-magnification observations of NMcm (B), NMc1 (C), and NMc2 (D), respectively. E–H, Anterogradely labeled axonal terminals in the caudal NM after in vivo injection of CTB into SON. Dashed lines outline NMc1 and NMc2. F–H are high-magnification observations of NMcm (F), NMc1 (G), and NMc2 (H), respectively. I, Immunostaining of CCK performed on the adjacent section of E for identifying NMc2. J, Labeled axonal terminals in NM at the level more rostral than NMc. K, L, Labeled cell bodies and neuropil in NL (K) and NA (L). M, Injection site in SON. White dashed line indicates the approximate border of the SON. The midline is indicated by black dashed line. Abbreviations: CTB, cholera toxin B; SON, superior olivary nucleus. Other abbreviations: see Fig. 1. Scale bars = 100 μm in A, E, I; 50 μm in L (applies to J–L); 20 μm in B (applies to B–D) and H (applies to F–H); and 500 μm in M.
- Figure 7.
Differential expression of calretinin in the caudal NM. The left (A1, B1, C1, D1) and middle (A2, B2, C2, D2) columns are MAP2 and calretinin immunostaining, respectively. The right column (A3, B3, C3, D3) is the merged images. A, Low-magnification images were taken from a section at the level of Fig. 1G. B–D, High-magnification images of the boxes in A1. All images were collected with the same imaging parameters and processed in the same way, except for the inset in D2 in which the brightness is enhanced to show a weakly labeled neuron in NMc2. Note calretinin-expressing neurons in NMcm and NMc1, but not NMc2. The border between NMc1 and NMc2 is indicated by dashed lines in D. E, Bar chart of the ratio of calretinin-expressing neurons in NMcm, NMc1, and NMc2. ***, P < 0.001. Data are presented as mean ± SD. Abbreviations: see Fig. 1. Scale bars = 100 μm in A3 (applies to A1–A3) and 20 μm in D3 (applies to B1–D3).
- Figure 8.
Differential expression of parvalbumin in the caudal NM. The left (A1, B1, C1, D1) and middle (A2, B2, C2, D2) columns are MAP2 and parvalbumin immunostaining, respectively. The right column (A3, B3, C3, D3) is the merged images. A, Low-magnification images were taken from a section at the level of Fig. 1G. B–D, High-magnification images of the boxes in A1. All images were collected with the same imaging parameters and processed in the same way. Arrows and arrowheads in B–D indicate labeled and unlabeled somata for parvalbumin. The border between NMc1 and NMc2 is indicated by dashed lines in D. E, Bar chart of the ratio of parvalbumin-expressing neurons in NMcm, NMc1, and NMc2. ***, P < 0.001; ns, not significant. Data are presented as mean ± SD. Abbreviations: see Fig. 1. Scale bars = 100 μm in A3 (applies to A1–A3) and 50 μm in D3 (applies to B1–D3).
- Figure 9.
Electrophysiological protocols applied to NMc1 and NMc2 neurons. A, Neurobiotin-labeled NMc1/NMc2 neuron. Inset shows low-magnification image of the entire coronal NM region with the labeled NMc1/NMc2 neuron. Dorsal, top; lateral, right. Scale bar = 20 μm (200 μm in inset). B, Current clamp protocol to measure passive membrane properties. Upper trace shows the representative voltage response (average of 30 repetitive trials) recorded from an NMc1/NMc2 neuron in response to a hyperpolarizing current injection (lower trace, –10 pA). A single exponential was fitted to a 30-ms time window after the current injection (superimposed red line), to calculate time constant (tau), input resistance, and membrane capacitance. C, Population data showing membrane capacitance (CMEMBRANE) sampled from the first (also referred as caudomost [caud-mos]) slice, second/third slices (caud), and middle to rostral slices (mid-ros, mid- to high-frequency NM; data modified from Hong et al. [2016]). Asterisk represents significance at p < 0.05. Error bars show SE. D, Metrics used to measure AP properties. Representative first APs (30 superimposed trials) were recorded from an NMc1/NMc2 neuron in response to current injections with the strength 25% above the threshold current (duration 100 ms). Several AP properties were characterized: rise rate (D), fall rate (E), half width (F), and reliability range (G). Population data of AP properties are shown in the corresponding panels in Fig. 11.
- Figure 10.
Heterogeneous voltage responses to current injections recorded from NMc1 and NMc2 neurons. A, B, C, (top) Representative voltage responses recorded from three NMc1/NMc2 neurons to current injections from –100 to 80 pA in steps of 20 pA (bottom). (middle) Representative voltage responses (30 superimposed trials) recorded from the same three NMc1/NMc2 neurons shown in the top panel, respectively. Bottom, current injections with the strength 25% above threshold current. Arrowhead in A shows widespread AP peak occurrences for this neuron. Arrow in A shows spontaneous activity. Insets in B and C show the enlargement of 30 superimposed APs. (bottom) Representative voltage responses recorded from three NMc1/NMc2 neurons to current injections with the strength of 200 pA (A and B) or 500 pA (C). D, Population data showing the number of APs elicited as a function of current injections from 0 to 200 pA, steps of 20 pA. NMc1/NMc2 neurons are divided into three subgroups: neurons displaying voltage responses similar to the neurons shown in A, B, and C are noted as A-like, B-like, and C-like, respectively (see Results for objective classification details). Mid- to high-frequency NM neurons (M-HF) are also shown as a reference (data modified from Hong et al., 2016). E, F, Population data showing resting membrane potential (RMP, E) and input resistance (F) of A-like, B-like, and C-like NMc1/NMc2 neurons. The duration of all current injections in this figure is 100 ms. Error bars show SE.
- Figure 11.
Heterogeneity of AP properties of NMc1 and NMc2 neurons. A, Number of APs generated in response to 200-pA current injections plotted as a function of threshold current for individual NMc1/NMc2 neurons. Three filled and labeled circles represent the neurons shown in Fig. 10A, B, and C, respectively. Correlation coefficient r and p values are shown. B, C, Threshold current plotted as a function of membrane capacitance (B) and input resistance (C) for individual NMc1/NMc2 neurons. D–G, Population data of AP rise rate (D), fall rate (E), half width (F), and reliability range (G) are plotted for individual NMc1/NMc2 neurons, as a function of threshold current (left), membrane capacitance (middle), and input resistance (right). Correlation coefficient r and p values are shown. Three filled circles represent the neurons shown in Fig. 10A, B, and C, respectively. Note that in G, five outliers with extremely large range (>30 ms) were removed.
- Figure 12.
Summary drawings of neuronal features in caudal NM. Based on cytoarchitecture, the caudal NM (regions outlined by solid black lines) is divided into three subdivisions, NMcm, NMc1, and NMc2. Borders between subregions are indicated by dashed lines. Left is medial and up is dorsal. A, Morphology and molecular signatures. NMcm, NMc1, and NMc2 neurons exhibit different dendritic complexity and cell body size. Compared with NMcm neurons with few dendrites, NMc1 and NMc2 neurons preserve more dendrites. Notably, NMc2 neurons show longer total dendritic branch lengths than NMc1. On average, NMcm are larger than NMc1 and NMc2 neurons in somatic size, and the majority of cells with the smallest cell body sizes are located in NMc2. Moreover, neurons in the three subregions also show distinct expression patterns of calretinin (magenta), parvalbumin (green), and CCK (blue). Most neurons in NMcm and NMc1 express calretinin, whereas neurons in NMc2 are not immunoreactive for this protein. A substantial number of NMc1 neurons coexpress calretinin and parvalbumin (half green and half magenta), but only a few neurons in NMcm and NMc2 show parvalbumin expression. Whether CCK-positive neurons in NMc2 are immunoreactive for parvalbumin or calretinin is not determined. Blue circles only represent CCK immunoreactivity of NMc2 neurons, not indicating restricted expression in cell bodies. Extensive neuropil staining of parvalbumin (short green lines) is observed in all three subregions. In NMcm, parvalbumin neuropil staining shows perisomatic pattern (green rings), whereas in NMc1 and NMc2, parvalbumin-positive neuropils (green spots) scatter between cell bodies. B, Connectivity. Neurons in NMcm receive excitatory (blue) and inhibitory (orange) inputs via end bulbs and small boutons, respectively. The inhibitory inputs form bouton synapses on the cell bodies. In contrast, NMc1 and NMc2 neurons receive both excitatory and inhibitory inputs via bouton terminals, primarily in the neuropil (presumably dendrites). C, Biophysics. NMcm neurons generate single-onset AP in response to sustained depolarization, whereas NMc1/NMc2 neurons display the ability of generating multiple action potentials to suprathreshold sustained depolarization and are spontaneously active. Abbreviations: see Fig. 1 for anatomic terms. CR, calretinin; PV, parvalbumin; CCK, cholecystokinin.
Tables
Antibody Manufacturer RRID Host species Working concentration Calretinin Millipore; AB5054 AB_2068506 Rabbit 1:5000 CCK Sigma; C2581 AB_258806 Rabbit 1:2000 CTB List Biological Lab; 703 AB_10013220 Goat 1:12,000 Gephyrin Synaptic Systems; mAb7a AB_2314591 Mouse 1:500 MAP2 Millipore; MAB3418 AB_94856 Mouse 1:1000 Parvalbumin Sigma; P3088 AB_477329 Mouse 1:5000 SNAP25 Millipore; MAB331 AB_94805 Mouse 1:1000 CCK, Cholecystokinin; CTB, Cholera toxin B; MAP2, microtubule-associated protein 2; SNAP25, synaptosome associated protein 25.
- Table 2.
Comparison of passive membrane and AP properties between NMc1/NMc2 and mid- to high-CF NM neurons
Property NMc1/NMc2 (n) Mid- to high-CF NMa P, t test Passive membrane properties RMP (mV)b –50.55 ± 9.74 (30) –66.52 ± 8.49 (28) <0.0001 Time constant tau (ms) 20.39 ± 17.25 (29) 3.18 ± 1.33 (20) <0.0001 Input resistance (MΩ) 467.2 ± 342.5 (29) 123.90 ± 49.90 (20) <0.0001 Membrane capacitance (pF) 41.25 ± 21.74 (29) 26.15 ± 4.60 (20) <0.01 Action potential properties Threshold current (pA) 38.96 ± 25.96 (22) 321.70 ± 121.00 (28) <0.0001 Max rise rate (mV/ms) c 136.1 ± 41.01 (22) 155.60 ± 42.19 (28) 0.107 Max fall rate (mV/ms) c –69.14 ± 22.16 (22) –104.40 ± 29.79 (28) <0.0001 AP half width (ms) c 1.45 ± 0.48 (22) 0.97 ± 0.17 (28) <0.0001 AP reliability range (ms)c 6.98 ± 5.87 (17)d 0.21 ± 0.14 (28) <0.0001 ↵a Data from Hong et al., 2016. Experimental conditions (e.g., temperature) and recording parameters (e.g., membrane capacitance) for both studies are the same.
↵b Numeric values without the correction of –10 mV junction potential.
↵c Measured from APs in response to current injections 25% above threshold current.
↵d Five outliner neurons with reliability range >30 ms were removed.
In this issue
