The brain is far away from the muscles that it controls. In humans, for example, the brain must be able to trigger the contraction of muscles that are more than a meter away. This task falls to specialized motor neurons that stretch from the brain to the spinal cord, and from the spinal cord to the muscles.

Neurons transmit information (in the form of electrical nerve impulses) along their length through cable-like structures called axons. The axons of the motor neurons are so long that, if they were ‘naked’, it would take at least a second for nerve impulses to travel their entire length. Such a long delay between thoughts and actions would make rapid movement impossible.

Nerve impulses are able to travel from the brain to the muscles much more quickly, because they are wrapped with a substance called myelin that acts like electrical insulation. Myelin helps the nerve impulses travel up to 100 times faster down the axon. Not surprisingly, diseases that damage myelin, such as multiple sclerosis, severely disrupt movement and sensation.

Not all of the brain’s myelin is found around the long axons of motor neurons. The outer layer of the brain, known as the cerebral cortex, also contains myelin. However, most neurons within the cerebral cortex are only a few millimeters long. So what exactly is myelin doing there?

Micheva et al. have now used electron microscopy and light microscopy to study the neurons in the cortex of the mouse brain. This revealed that up to half of the myelin in some layers of the cortex surrounds the axons of inhibitory neurons (which reduce the activity of the neurons they signal to). Moreover, one particular subtype of inhibitory neuron accounts for most of the myelinated inhibitory axons, namely inhibitory neurons that contain a protein called parvalbumin.

Exactly why parvalbumin-containing neurons are myelinated remains a mystery. Myelin covers only short segments of the axons of these neurons, so it would speed up the transmission of signals by less than a millisecond – probably not enough to make a meaningful difference. Parvalbumin-containing neurons often signal frequently, and thus require large amounts of energy. One possibility therefore is that myelin helps to meet these energy requirements or to reduce energy consumption. Further research will be needed to test this and other ideas.

Myelination is well known to be important in increasing the speed of action potential propagation in long-axon connections (Waxman and Bennett, 1972; Moore et al., 1978). The axons of the central nervous system (CNS) connecting distant brain areas are typically wrapped in myelin, as are the axons of the peripheral nervous system (PNS) that connect the CNS to all skeletal muscles and many sensory neurons. Pathologies of myelinated axons are associated with numerous neurological disorders (Fields, 2008) and are fundamental to several, such as multiple sclerosis (MS) (Popescu and Lucchinetti, 2012; Calabrese et al., 2015), neuromyelitis optica (Wingerchuk et al., 2015), and leukodystrophies (Gordon et al., 2014).

Because myelin increases the conduction velocity of action potentials and reduces their metabolic cost (Nave, 2010; but see Harris and Attwell, 2012), it has traditionally been viewed as ‘wire insulation’ for long-range neural projections. Recent findings have cast myelination as a dynamic process playing an active role in normal neural circuit function and plasticity. For example, adult-born oligodendrocytes (the myelin-producing cells of the CNS) contribute new myelin to axons in an ongoing fashion (Young et al., 2013), increased neuronal activity results in more myelination (Gibson et al., 2014), and myelin remodeling in the CNS is required for motor skill acquisition (McKenzie et al., 2014). Myelin also supplies ensheathed axons with lactate via the monocarboxylate transporter MCT1, possibly playing an important energetic support role in the CNS (Rinholm et al., 2011; Lee et al., 2012; Fünfschilling et al., 2012; Morrison et al., 2013).

The emerging diversity of roles for myelin calls for a more complete understanding of its distribution among neuronal cell types and along individual axonal arbors, particularly in the gray matter of the CNS. Excitatory pyramidal neurons in cortex are known to be a main source of long-range myelinated axons connecting the two cortical hemispheres or targeting subcortical structures. Foundational EM-based studies of individual Golgi- or dye-filled neurons (Kisvárday et al., 1986; Martin and Whitteridge, 1984; Peters and Proskauer, 1980; Tamás et al., 1997) have revealed that the local axon collaterals of both excitatory and inhibitory neurons may also be myelinated. However, population surveys of the distribution of myelin on the axonal arbor of individual neocortical neurons have been difficult to undertake with traditional methods. Golgi labeling of neurons halts where the axon first enters a myelin sheath (Peters and Proskauer, 1980; DeFelipe et al., 1986; Fairén et al., 1977), and bulk labeling of myelin reveals a laminar patterned thicket (Nieuwenhuys, 2013), rather than individual neuronal arbors.

Recent advances in large volume, high resolution neuroanatomy methods (Briggman and Bock, 2012; Micheva and Smith, 2007; Economo et al., 2016) now permit cortical myelin to be directly visualized on the axonal arbors of multiple individual identified neurons for all axons within the imaged volume. Myelin tracing in large EM datasets has revealed that myelination of the axons of pyramidal neurons in cortex is distinct from the known myelination patterns in subcortical white matter, and also differs between cortical layers (Tomassy et al., 2014). About 10–20% of neocortical neurons are inhibitory and some of these neurons may also have myelinated axons. For example, labeling of individual neocortical neurons during in vivo electrophysiological recording, followed by light-level reconstruction and electron microscopy (EM) of selected sub-arbors, has revealed that inhibitory neurons may be myelinated (DeFelipe et al., 1986; Somogyi et al., 1983; Kawaguchi and Kubota, 1998; Kisvarday et al., 1987; Thomson et al., 1996; Tamás et al., 1997). Complementary immunohistochemical studies have shown that parvalbumin-positive (i.e., putative inhibitory [Rudy et al., 2011]) axons may be myelinated (McGee et al., 2005; Chung and Deisseroth, 2013). Given that cortical interneuron types are implicated in distinct circuit functions (Kepecs and Fishell, 2014) and have distinct behavioral correlates (Kvitsiani et al., 2013), these results raise several important questions regarding the organization of neocortical neuronal networks: how widespread is inhibitory axon myelination? Is it specific to certain interneuron subtypes? Does inhibitory axon myelin differ from excitatory axon myelin at the molecular level?

Here we combine analyses of light-microscopy-based array tomography (AT) data (Micheva and Smith, 2007; Collman et al., 2015) with a publicly hosted EM dataset (Burns et al., 2013; Martone et al., 2002) generated using a custom transmission EM camera array (TEMCA) (Bock et al., 2011), to obtain an integrated view of the myelination of inhibitory neurons in cortex. We find that an unexpectedly large proportion of myelinated axons in cortex arise from a specific subtype of inhibitory neuron: the parvalbumin-positive (PV+) basket cell. These PV+ axons are especially prominent in the middle cortical layers. They constitute about half of all myelinated axons in layers 2/3, and a quarter of myelinated axons in layer 4. We also find that the inhibitory myelinated axons in cortical gray matter differ from the excitatory myelinated axons in many respects, such as their structural organization, cytoskeletal content, and myelin protein composition. Taken together, the abundance and cell-type selectivity of neocortical myelinated inhibitory axons described here raise fundamental questions about the role of myelination both in normal function and in disease.

We applied two microscopic approaches to determine the neurotransmitter content of cortical myelinated axons. Array tomography (AT) is based on digital reconstruction of images acquired from arrays of serial ultrathin sections (70 nm) attached to coverslips and imaged with different modalities, such as immunofluorescence and scanning electron microscopy (SEM). The use of ultrathin sections allows the unambiguous light level identification of individual myelinated axons even in regions with high myelin density (e.g., subcortical white matter), while the possibility of applying multiple immunofluorescent markers (10 or more) enables the molecular characterization of these axons. Conjugate SEM further reveals the underlying ultrastructure. TEMCA-based volume electron microscopy uses an array of high-speed cameras to efficiently image large numbers of serial ultrathin (40–50 nm) sections. The resulting image volumes are of sufficient size and resolution that the densely packed and intertwined axons and dendrites of the brain neuropil can be traced over relatively long distances, and the synaptic connections between them can be discerned. With this form of EM, all membranes and synapses are labeled, enabling neuronal structure and connectivity to be sampled in an unbiased fashion. Additional evidence was also obtained using immunohistochemical labeling of brain slices from transgenic mice expressing fluorophores in salient inhibitory interneuron subtypes. The EM and AT studies were initiated and conducted independently in two different laboratories and data were pooled once the complementary nature of the findings were ascertained (Materials and methods).

Array tomography shows that a large fraction of myelinated axons in upper layers of cortex contain the inhibitory neurotransmitter GABA

Immunofluorescence AT is particularly well suited for the study of myelinated axons in cortex. The abundance of myelin proteins, such as myelin basic protein (MBP), and the size of myelinated axons (>200 nm in diameter) make them an ideal target for immunofluorescence detection (Figure 1A, Figure 1—figure supplement 1, Video 1). The ultrathin AT sections provide easy access of the antibodies against MBP and do not require lipid extraction. Conjugate light-electron AT confirms that immunofluorescence for MBP precisely outlines the myelinated sheaths of axons (Figure 1B, Figure 1—figure supplement 1B, Figure 1—figure supplement 2A). Overall, the MBP immunolabel is of high quality as evidenced by the consistency of the signal from section to section (Figure 1—figure supplement 2B), and the low background (mean gray value of 248 ± 6 a.u. on resin and 279 ± 2 a.u. on nuclei, vs 5754 ± 226 a.u. on myelin, mean and standard error). In addition, the use of ultrathin sections ensures that antigens can be detected equally well within myelinated and unmyelinated portions of the axon. For example, in our AT experiments GABA and PV immunostaining within axons is not affected by myelination (Figures 1, 3), in contrast to preembedding immunocytochemistry where myelinated portions of axons are much more weakly stained (Pawelzik et al., 2002).

Figure 1 with 2 supplements see all

Download asset Open asset

Figure 1—source data 1

Data values underlying Figure 1.

https://doi.org/10.7554/eLife.15784.004

Download elife-15784-fig1-data1-v2.xlsx

Video 1

Download asset

As expected, AT detects the presence of myelinated axons in all neocortical layers. Quantification of the number of MBP profiles in single sections shows that their numbers dramatically increase with cortical depth (Figure 1D), consistent with previous quantitative myeloarchitectonic studies in non-rodent species (Nieuwenhuys, 2013; Braitenberg, 1962; Hopf, 1966). Immunostaining for the inhibitory neurotransmitter GABA reveals that a surprisingly large number of myelinated axons throughout cortex contain GABA, including more than 45% of myelinated axons in layers 2/3 (48.1 ± 3.2%, mean ± standard error, Figure 1C). Layers 4 and 5 exhibit similar numbers of GABA myelinated axons as layers 2/3 (Figure 1D), but the deeper layers contain larger numbers of non-GABA myelinated axons, presumably of projecting pyramidal cells. GABA myelinated axons are relatively sparse in layers 1, 6, and only occasionally seen in subcortical white matter.

Volume electron microscopy reveals that a large fraction of myelinated axons in layer 2/3 form symmetrical (presumed inhibitory) synapses

In parallel with the array tomography work, and blind to its results, myelinated axons were traced within a previously acquired (Bock et al., 2011), publicly available (Burns et al., 2013; Martone et al., 2002) volume electron microscopy dataset from the upper layers of mouse visual cortex V1. This dataset was of sufficient size and resolution that both myelinated and unmyelinated axons could be traced over hundreds of microns, allowing them to be categorized on the basis of their synaptic ultrastructure as inhibitory or excitatory. A section from near the middle of the EM volume was selected, and all the myelin profiles from the center portion of this section were annotated (Figure 2A). Each of these myelinated 'seeds' (e.g. Figure 2B,E) was used as a starting point for further tracing of the myelinated axon through the EM volume. Along their length, a majority of the traced axons unmyelinated and made a number of synapses (Figure 2C,F). Axons were classified according to the appearance of the pre- and postsynaptic densities of the synapses they made, using the classical definitions for asymmetrical (excitatory) and symmetrical (inhibitory) synapses (Colonnier, 1968; Peters et al., 1991) (Figure 2D,G). Of the 231 myelinated 'seeds', 73 were classified as excitatory (31.6%; Figure 2—figure supplement 1), 106 as inhibitory (45.9%; Figure 2—figure supplement 2), and 52 (22.5%; Figure 2—figure supplement 3) could not be categorized due to an absence of synapses within the volume. This analysis shows that inhibitory myelinated axons are abundant in cortical layer 2/3, comprising at least 45% of myelinated axons in the mouse visual cortex.

Figure 2 with 3 supplements see all

Download asset Open asset

Nearly all myelinated GABA axons are parvalbumin-positive

Cortical GABA neurons are a highly diverse population (Ascoli et al., 2008; Markram et al., 2004; Xu et al., 2010). About half of them contain parvalbumin (PV), a calcium buffering protein. Array tomography (Figure 3A,B and Figure 3—figure supplement 1) revealed that nearly all GABA myelinated axons (97.9 ± 0.8%, mean ± st.error from 3 animals) are immunopositive for parvalbumin and therefore originate from PV interneurons. Correspondingly, in the examined cortical AT volumes (Figure 3C–F), examples of PV and GABA-positive cell bodies forming myelinated axons are seen in layers 2/3, 4 and 5. Blind to the AT results, we conducted independent experiments in three mice expressing the fluorescent protein tdTomato driven by Cre expression in PV neurons (Pvalb-ires-Cre;Ai9) (Taniguchi et al., 2011). Confocal imaging of 40 μm thick Vibratome slices from these mice immunostained with MBP corroborated the AT findings by showing that about half (53 ± 12%) of the myelinated axons in layer 2/3 of somatosensory cortex contain tdTomato and, therefore, PV (Figure 4A–D). Similar experiments were also performed on mice expressing tdTomato from SOM-Cre or VIP-Cre driver lines (Sst-ires-Cre;Ai9, and Vip-ires-Cre;Ai9) (Taniguchi et al., 2011; Madisen et al., 2010). After careful inspection of selected subareas, no myelinated axons containing reporter for the non-PV cortical interneurons vasoactive intestinal polypeptide (VIP; Figure 4E–G) could be observed. A small fraction (4.4%) of somatostatin (SOM)-positive myelinated axons were observed (Figure 4H–J); however, the transgenic reporter line used was recently found to have a small but consistent false-positive expression pattern, wherein up to 10% of Cre-expressing neurons are SOM-negative and PV-positive when examined by immunohistochemistry (Hu et al., 2013). At least some of the SOM-positive myelinated axon profiles we observe therefore may be false positives which actually express PV, not SOM; and some may be true positives, consistent with the AT data, which show that ~2% of GABA myelinated profiles are PV negative. Regardless, the data obtained from the three Cre-lines support the conclusion that the great preponderance of myelinated inhibitory axon profiles in layer 2/3 arise from parvalbumin-positive axons.

Figure 3 with 1 supplement see all

Download asset Open asset

Figure 4

Download asset Open asset

Myelinated inhibitory axons most likely belong to cortical basket cells

Among the inhibitory neurons of cortex, parvalbumin is found in basket cells, chandelier cells (Markram et al., 2004; Kawaguchi and Kubota, 1997; DeFelipe et al., 1989), and in a sparse sub-population of long-range inhibitory axons arising from basal forebrain (Freund and Gulyás, 1991; Caputi et al., 2013). Because these different cell types are known to have distinct axonal morphology and targets, we traced the postsynaptic targets of inhibitory myelinated axons in layer 2/3 from the volume EM data set in an effort to understand their originating cell type (Figure 5). The postsynaptic targets were identified as inhibitory or excitatory according to previously described ultrastructural criteria (Bock et al., 2011) (Materials and methods). Briefly, postsynaptic dendrites were classified as inhibitory if they were largely aspinous and densely coated with asymmetric (excitatory) synapses (e.g. Figure 6E–F), while they were classified as excitatory if they were spiny and received fewer shaft excitatory synapses. Excitatory somata were distinguished from inhibitory somata based on cell shape, the presence of a large apical dendrite, and the absence of asymmetric contacts on the soma. For each synapse, the postsynaptic compartmental location and class (excitatory or inhibitory) was tabulated (Table 1), and postsynaptic class was used to color-code the synapses along the reconstructed axonal fragments (Figures 5–6).

Figure 5

Download asset Open asset

Figure 6

Download asset Open asset

Table 1

Postsynaptic target	Completely traced axons		Partially traced axons		All
	count	%	count	%	count	%
Excitatory somata	60*	15.0	46†	10.5	106	12.6
Excitatory dendritic shafts	184	45.9	225	51.4	409	48.7
Excitatory dendritic spines	119	29.7	134	30.6	253	30.2
Excitatory axons	7	1.7	8	1.8	15	1.8
All excitatory	370	92.3	413	94.3	783	93.3
Inhibitory dendritic shafts	17	4.2	20	4.6	37	4.4
Inhibitory dendritic spines	0	0	0	0	0	0
Inhibitory axons			1	0.2	1	0.1
All inhibitory	17	4.2	21	4.8	38	4.5
Uncategorized dendritic spines	14	3.5	4	0.9	18	2.1
Total	401		438		839

1. * Including 10 synapses onto somatic spines;

2. ^† Including 8 synapses on somatic spines

The observed distribution of postsynaptic targets of myelinated inhibitory axons was consistent with their originating from local axonal arbors of cortical basket cells, and inconsistent with their arising from either chandelier cells or long-range inhibitory afferents to cortex. Basket cells and chandelier cells can readily be distinguished by their morphology and synaptic targets. When the axons from multiple basket cells are labeled and examined at the light level, they form ‘baskets’ of synapses around neuronal somata, hence the name (Kisvárday, 1992). Chandelier cells have an axonal arbor which resembles a chandelier, with vertically oriented axonal cartridges that almost exclusively form synapses onto the axon initial segments of pyramidal neurons (Somogyi et al., 1982). Less than 2% of the inhibitory myelinated axons in our volume electron microscopy dataset were found to make contacts on axon initial segments (Figure 5; Table 1). In addition, the few axonal arbors that were seen contacting axon initial segments did not have the characteristic appearance of chandelier cell axons and made single contacts with axonal initial segments, instead of several (typically 3–5 for chandelier cells) synapses in a row (Figure 5A,B). Therefore we can exclude the possibility that the myelinated inhibitory axons arise from chandelier cells. The great preponderance (93%) of postsynaptic targets of myelinated inhibitory axons in layer 2/3 of cortex were excitatory, likewise eliminating the possibility that these axons originated from the arbors of the known parvalbumin-positive long-range inhibitory afferents to cortex, all of which selectively target inhibitory neurons (Freund and Gulyás, 1991; Gritti et al., 2003; Henny and Jones, 2008).

The remaining possibility, that myelinated inhibitory axons arise from the arbors of local parvalbumin-positive basket cells, was well supported by the distribution of their postsynaptic targets. Although neocortical basket cells are best known for their synapses with cell bodies, a large (but variable) fraction of their synapses are with dendritic shafts and spines (Somogyi et al., 1983; Kawaguchi and Kubota, 1998; Kisvarday et al., 1987; Tamás et al., 1997). Consistent with past surveys of basket cell synaptic targets, we found that 13% of the postsynaptic targets of the myelinated inhibitory axons were with somata, 49% were onto excitatory dendritic shafts, and 30% were onto dendritic spines (Figure 6; Table 1). Thus, we conclude that the majority of myelinated inhibitory axons belong to the intracortical axons of parvalbumin containing basket cells.

Inhibitory neurons exhibit a distinct pattern of myelination

Both our volume electron microscopy and array tomography data show that the axon of a parvalbumin interneuron becomes myelinated soon after exiting the cell body (usually within 20–50 μm), consistent with myelination commencing shortly after the axon initial segment. All of the axonal arbors that were traced back to the cell body in our volume EM dataset (n=8) had the same general appearance: a central core of partially myelinated axons (whose unmyelinated stretches rarely made synapses), from which emerged unmyelinated branches forming numerous synapses (Figures 2C, 5A,C). This is similar to the pattern reported for filled basket cells in both macaque somatosensory cortex (DeFelipe et al., 1986) and cat visual cortex (Somogyi et al., 1983).

Further analysis revealed marked differences in the myelination of GABA and nonGABA axons. We used array tomography and MBP immunostaining to quantify the lengths of the nodes of Ranvier, where the ion channels needed for nerve impulse transmission are concentrated (Figure 7A–B). Only gaps in the axonal myelin sheath of less than 4 μm length were considered to be nodes. Throughout the cortical layers, the nodes of GABA axons were shorter than for non-GABA axons (0.98 ± 0.07 μm vs. 1.51 ± 0.05 μm, mean ± standard error, p<0.0001, n=71 GABA and 209 non-GABA nodes of Ranvier, Mann-Whitney U Test; Figure 7C). Their internodes, the myelinated portion between two nodes of Ranvier, were also found to be shorter (22.54 ± 2.11 μm vs 29.29 ± 2.02 μm for non-GABA axons, mean ± standard error, p=0.032, n=23 GABA axons and 32 non-GABA axons, Mann-Whitney U Test).

Figure 7

Download asset Open asset

Figure 7—source data 1

Data values and statistics underlying Figure 7.

https://doi.org/10.7554/eLife.15784.019

Download elife-15784-fig7-data1-v2.xlsx

GABA and non-GABA myelin differ in protein composition, but not thickness

Myelin thickness is thought to be directly correlated to axon diameter (constant 'g-ratio'), based mostly on studies of the peripheral nervous system (Fraher and Dockery, 1998). However, we found no such correlation within mouse cerebral cortex. Using a deconvolved array tomography dataset with MBP and GABA immunofluorescence, we found that the average myelin thickness was 0.13 ± 0.002 μm for both GABA and non-GABA axons (mean ± standard error, n=163 GABA and 238 non-GABA axons from layers 4 and 5, p=0.88, Mann-Whitney U Test) and it did not correlate with axon thickness (R=0.008; n=163 GABA and 238 non-GABA axons from layers 4 and 5). Greater axon diameters not accompanied by changes in myelin thickness result in the g-ratio increasing as a function of axon diameter (Figure 8B). GABA axons are thicker than non-GABA axons (0.54 ± 0.001 μm vs. 0.45 ± 0.001 μm, n=163 GABA and 238 non-GABA axons, p<0.0001, Mann-Whitney U Test; and Figure 8A), thus their average g-ratio is higher (0.66 ± 0.006 vs. 0.61 ± 0.006, n=163 GABA and 238 non-GABA axons, p<0.0001, Mann-Whitney U Test).

Figure 8

Download asset Open asset

Figure 8—source data 1

Data values and statistics underlying Figure 8.

https://doi.org/10.7554/eLife.15784.021

Download elife-15784-fig8-data1-v2.xlsx

Next we compared the protein composition of the myelin of GABA and non-GABA axons. The two major proteins in myelin are MBP and proteolipid protein (PLP), which together constitute about 80% of all proteins in myelin (Baumann and Pham-Dinh, 2001; Rosetti and Maggio, 2007). While the average PLP immunofluorescence was very similar in both types of axons (2229 ± 60 a.u. vs 2282 ± 40 a.u., for GABA and non-GABA axons respectively; mean ± standard error, n=254 GABA axons and 489 non-GABA, p=0.2713, Mann-Whitney U Test, Figure 8D), GABA myelinated axons had almost 20% higher MBP immunofluorescence (5263 ± 117 a.u. vs. 4486 ± 63 a.u., mean ± standard error, n=254 GABA axons and 489 non-GABA, p<0.0001, Mann-Whitney U Test). This difference in MBP content was not related to the thickness of the axons (Figure 8C). The higher MBP immunofluorescence of the myelin sheath of inhibitory axons was confirmed in experiments in two more animals (A1: 3024 ± 100 a.u. vs. 2634 ± 86 a.u., mean ± standard error, n=322 PV axons and 314 non-PV from layers 2 to 5, p=0.0114, Mann-Whitney U Test; A2: 3038 ± 117 a.u. vs. 2585 ± 120 a.u., mean ± standard error, n=233 PV axons and 119 non-PV from layer 4, p=0.0455, Mann-Whitney U Test). Although immunofluorescence has not been calibrated to underlying MBP concentration, the observation that GABA axons have more MBP than neighboring non-GABA axons was statistically significant and qualitatively consistent across multiple experiments.

The cytoskeletal composition of myelinated GABA axons is different from myelinated non-GABA axons and from unmyelinated GABA axons

The most striking difference between GABA and non-GABA myelinated axons is in their cytoskeleton. Immunostaining for the neurofilament heavy chain and for alpha tubulin revealed that myelinated GABA axons are rich in neurofilaments and have relatively low microtubule content, while the cytoskeleton of myelinated non-GABA axons is dominated by microtubules (Figure 9). In addition, a larger fraction of the tubulin in myelinated GABA axons appears to be acetylated. While GABA myelinated axons contain on average only 66% of the alpha tubulin found in non-GABA myelinated axons, they have 86% of the acetylated alpha tubulin content of non-GABA myelinated axons. These differences in the cytoskeletal composition of myelinated axons were observed throughout the cortical layers.

Figure 9

Download asset Open asset

Figure 9—source data 1

Data values and statistics underlying Figure 9.

https://doi.org/10.7554/eLife.15784.023

Download elife-15784-fig9-data1-v2.xlsx

The differences in cytoskeletal composition of myelinated axons could be due to a general difference between GABA and non-GABA axons, regardless of their myelination status. To explore this possibility, we identified individual axons that contained both a myelinated and a non-myelinated portion within the dataset volume. Nodes of Ranvier were excluded and only non-myelinated stretches longer than 4 μm were considered. This analysis revealed several interesting differences (Figure 9D). First, the cytoskeletal composition of axons is indeed related to the presence of GABA. Both the myelinated and the unmyelinated regions of the GABA axons have higher neurofilament content and lower microtubule content compared to non-GABA axons. There are also significant differences in cytoskeletal content between the different regions of GABA axons, where myelinated portions have higher neurofilament and lower microtubule content compared to the unmyelinated portions of the same axons. While excitatory axons show similar trends of differences in cytoskeletal composition depending on myelination, the differences are more pronounced for GABA axons, especially regarding neurofilament content.

The present work is the first to survey the targets of myelin in neocortex using recently developed high resolution, large-volume imaging methods (Briggman and Bock, 2012; Kleinfeld et al., 2011; Economo et al., 2016; Micheva and Smith, 2007). These methods reveal that a surprisingly large fraction of myelin in cortical gray matter ensheathes the axons of a single inhibitory neuron subtype, the parvalbumin-positive basket cell. Based on cellular morphology and postsynaptic targets, other PV-expressing neuron types were essentially ruled out, and we found minimal evidence for the myelination of non-PV inhibitory axons. Although this cell-type-specificity of axon myelination was unexpected, especially since myelin has generally been thought to ensheath long-range axons of excitatory neurons, it is nevertheless consistent with earlier observations of individual myelinated inhibitory axons, including basket cell axons (Peters and Proskauer, 1980; DeFelipe et al., 1986; Somogyi et al., 1983; Keller and White, 1987).

The more comprehensive survey presented in the present work was achieved through analysis of data arising from two different imaging modalities: array tomography (Micheva and Smith, 2007; Collman et al., 2015) and large-scale serial-section electron microscopy (Bock et al., 2011). Work based on each method was initiated independently and in parallel, and initial annotation and analysis was performed blind to preliminary results arising from the other method. The results from AT and volume EM were complementary. The AT data spanned all cortical laminae and revealed that many myelinated axons in mouse somatosensory cortex (almost half of layer 2/3, and a quarter of layer 4 myelinated axons) contain the inhibitory neurotransmitter GABA. Practically all of these GABA axons also contain PV. AT-based comparison of GABA and nonGABA myelinated axons revealed several important differences. The nodes of Ranvier and the internodes are significantly shorter on GABA axons. In addition, AT multiplex imaging detected molecular differences, such as an enrichment of MBP in the myelin of GABA axons, as well as a high neurofilament and low microtubule content of their cytoskeleton.

The volume EM data provided independent evidence that half of the categorized myelinated axons in layer 2/3 of mouse visual cortex are inhibitory, as indicated by the symmetric pre- and postsynaptic densities of the synapses that they formed after exiting the myelin sheath (Colonnier, 1968; Peters et al., 1991; Gray, 1959). Immunolabeling of myelin in a PV-Cre reporter line showed that about half of the myelin in layer 2/3 ensheathes PV+ axons, and similar experiments in VIP- and SOM-Cre lines (which comprise almost all non-PV interneurons [Xu et al., 2010]) failed to show any myelinated axons arising from VIP+ neurons and very few (if any) from SOM+ neurons. Ultrastructural analysis also allowed us to deduce which of the two PV-expressing subtypes in neocortex was myelinated, by revealing that the myelinated axons’ postsynaptic targets were largely non-axonal, instead targeting the dendrites and somata of pyramidal cells. This finding ruled out PV-positive chandelier cells (which synapse onto axons), and strongly suggested that the other predominant PV subtype, basket cells, is the myelinated subtype. PV basket cells comprise about half of all neocortical interneurons (Markram et al., 2004). Taken together, the quantitative and qualitative consistency between the AT, volume EM, and immunohistochemistry datasets provides strong evidence that PV+ basket cell axons are the predominant myelinated inhibitory axons in neocortex.

Our axon reconstructions showed that the distribution of myelin on basket cell axons is patchy and seemingly localized to the axon arbor near the cell body (e.g. Figures 3D–F, 5A–C). This patchiness qualitatively resembles the distribution of myelin on the descending axons of excitatory pyramidal cells measured in the same EM data set (Tomassy et al., 2014), and is consistent with an observation of patchy myelination of an inhibitory neuron axon in an earlier study in mouse somatosensory cortex (Keller and White, 1987). The concentration of myelin on the proximal arbor is also consistent with more complete reconstructions of individual filled cells in cat (Somogyi et al., 1983) and macaque (DeFelipe et al., 1986).

There are several caveats and limitations to this study. First, our experiments allow for the possibility that a small fraction of somatostatin or other non-PV axons are myelinated. We found by AT that ~2% of myelinated GABA axons are PV-negative, and examination of the transgenic mouse line SOM-Cre;Ai9-TdTom revealed a small percentage of TdTomato-positive myelinated axons. However, the lack of PV immunostaining in some myelinated axons may reflect limitations of immunodetection rather than proving that these axons are PV-negative. Furthermore, the SOM-Cre-line used in this study was found to have a small but consistent false-positive expression pattern, where Cre is expressed in neurons that immunohistochemically are somatostatin-negative but PV-positive (Hu et al., 2013). Although additional studies of somatostatin-positive interneurons, as well as other inhibitory interneuron subtypes, will be needed to resolve this issue, we believe it does not materially affect our finding that the great preponderance of myelinated inhibitory axon profiles in layer 2/3 arise from parvalbumin positive basket cells.

A second limitation stems from the incomplete characterization of the distribution of myelin on the full arbors of inhibitory neuron axons. Myelinated axon profiles were traced until they exited the myelin sheath and formed a sufficient number of synapses to categorize the axon as excitatory or inhibitory, at which point tracing was halted (except in two cases, Figure 5A and B, where the entire arbor within the 450 × 350 × 50 μm EM volume was traced). The result was a set of axon fragments (Figure 5C) that cannot be positively identified according to morphological type. Their identification as arising from basket cells is based on a combination of immunolabeling results (that myelinated inhibitory axons are nearly always PV+) with the identification of the fragments’ post-synaptic targets as being almost entirely non-axonal (so the axons do not arise from chandelier cells, the other PV+ cell type besides basket cells). Ideally, direct confirmation of this deduction would also be obtained from complete volume EM axonal arbor reconstructions and subsequent confirmation of morphological type. However, the additional tracing effort to achieve this would be substantial (Helmstaedter, 2013), and, perhaps more importantly, such an investment might better be made in a larger cortical EM volume (e.g [Lee et al., 2016]), so that a greater fraction of each neuron’s axonal arbor could be reconstructed. Such an approach would also permit assessment of whether all PV-containing basket cells have myelinated axons. Considering the functional and morphological heterogeneity of PV basket cells (Kawaguchi and Kubota, 1997; Wang et al., 2002; Li and Huntsman, 2014), it is possible that some subtypes preferentially form myelinated axons.

Consistent with the possibility that not all basket cells (or basket cell subtypes) are myelinated, the fraction of somatic post-synaptic targets made by the inhibitory myelinated axons described here is at the low end of the previously reported range for cortical basket cells. The majority of myelinated inhibitory axons in our study contact dendritic shafts (53%) and spines (32%), and only 13% of the targets are cell bodies. These proportions are similar to those found in a study in rat frontal cortex of four individually filled fast-spiking basket cells, which made on average 53% of synapses onto dendritic shafts, 29% onto spines, and 18% onto cell bodies (Kawaguchi and Kubota, 1998). However, the four cells showed wide variability in the proportion of their somatic targets, which ranged between 4% and 35%. Studies in cat neocortex also reveal a wide variability between individual basket cells (between 20% and 69% of synapses targeting somata) (Somogyi et al., 1983; Kisvarday et al., 1987; Tamás et al., 1997). These findings in neocortex contrast with measurements in hippocampus, where more than 50% of the synaptic targets of PV basket cells are cell bodies, and dendritic spines are only occasionally targeted (Seress and Ribak, 1990; Gulyás et al., 1993; Halasy et al., 1996). The fraction of inhibitory post-synaptic targets made by the axons in this study (4.5%) is similar to, but somewhat lower than that previously observed in rat (8%) and cat (9%) neocortex (Somogyi et al., 1983; Kawaguchi and Kubota, 1998). This variability is consistent with a recent study in mouse visual cortex which found wide variation between individual animals in the fraction of synapses made onto inhibitory interneurons by layer 2/3 pyramidal cells (Bopp et al., 2014). Overall, the anatomical literature briefly described above shows that the post-synaptic targets of a given cell type can vary widely between species, brain regions, animals, and individual neurons. Determining whether discrete basket cell subtypes underlay any of this variability will require further study.

Our findings raise the question: what is the role of myelin on inhibitory interneurons in neocortex? The patchy distribution of myelin along GABA axons and the lack of correlation between myelin thickness and axon diameter imply that increased conduction velocity may not be the main consequence of this myelination. Rather, the presence of myelin may be due to the unique cellular properties and circuit functions (Hu et al., 2014) of PV basket cells. Nearly all cortical PV basket cells are fast-spiking (Kawaguchi and Kubota, 1997; Galarreta and Hestrin, 1999, 2002; Gibson et al., 1999), and fast-spiking neurons have very high tonic activity compared to other cell types in cortex (Gentet et al., 2010; Hofer et al., 2011) — they generate long trains of action potentials with very high frequency (>100 Hz) (Kawaguchi and Kubota, 1997; Galarreta and Hestrin, 1999, 2002; Gibson et al., 1999). Fast-spiking cells exhibit a number of biophysical specializations that enable high firing rates but incur an increased energy cost per action potential (Hu and Jonas, 2014; Carter and Bean, 2009; Hasenstaub et al., 2010). Axonal myelination could help with these high energy demands in two ways: first, myelin is known to improve the energy efficiency of action potential propagation (Hartline and Colman, 2007). Secondly, myelin can provide metabolic support to the axon by supplying lactate as an energy source (Rinholm et al., 2011; Lee et al., 2012; Fünfschilling et al., 2012; Morrison et al., 2013). Thus, myelin could reduce the basket cells’ metabolic needs and provide an additional external energy supply from oligodendrocytes.

At the circuit level, PV basket cells participate in regulating excitatory/inhibitory balance, in neuronal synchronization and cortical rhythm generation, and in plasticity associated with experience and learning (Hu et al., 2014). Could basket-cell myelin hasten the delivery of synaptic input to their postsynaptic targets to an extent that would be functionally relevant? No estimates of conduction velocities of cortical myelinated PV axons exist in the literature; however, in young (~P30) mice, PV axon conduction velocities of ~0.5 m/s have been recorded (Casale et al., 2015; Li et al., 2014). The majority of PV axons at this age are most likely unmyelinated, since cortical myelination is not complete until much later (after P60) (Barrera et al., 2013; Mengler et al., 2014). Using this conduction velocity estimate, an unmyelinated PV axon would conduct a spike in 0.4 ms to synapses located 200 μm away, which was the typical path length from cell body to proximal synapses in our EM-based reconstructions. Thus, any reduction in the spike arrival time to proximal synapses due to myelination will inevitably be in the sub-millisecond range, likely too small a difference to affect post-synaptic dendritic integration or spike-timing-dependent plasticity, given the much slower timescales of those functions (Dan and Poo, 2006; Feldman, 2012; Stuart and Spruston, 2015). However, spike time arrival differences on a sub-millisecond timescale might affect the magnitude of coordinated oscillations (Pajevic et al., 2014), which are a known circuit role of basket cells.

The present study also bears on the genesis and dynamics of myelination. Cortical myelin is dynamically remodeled during learning and development (McKenzie et al., 2014; Bergles and Richardson, 2016; Makinodan et al., 2012) and it redistributes during normal aging, where internode length decreases with age in primates (Peters, 2009). It is possible that these dynamics may differ between the myelin wrapping inhibitory and excitatory axons. This could be due to differences arising from local interactions between oligodendrocytic processes and individual axons, or from a possible allocation of distinct oligodendrocytes (or oligodendrocyte subtypes [Tomassy et al., 2016]) to axons of inhibitory or excitatory neurons (de Hoz and Simons, 2015).

Further studies will also be needed to address the interesting question of myelin diversity at the molecular level. Between the two major proteins of myelin, our experiments reveal that inhibitory axons have higher MBP content than excitatory axons, while the proteolipid protein (PLP) content appears similar. Because MBP is one of the potential target antigens in MS (Wucherpfennig and Strominger, 1995; Fujinami and Oldstone, 1985; Garg and Smith, 2015), this finding has important implications for pathologies of myelin. It is likely that other differences in the protein or lipid content of myelin exist, considering the accumulating evidence for CNS oligodendrocyte heterogeneity (Tomassy et al., 2016). Characterizing these differences as a function of neuronal type will be an important aspect of future studies.

The abundance of myelinated inhibitory axons in cortical gray matter also raises the possibility of their involvement in myelin perturbations in normal aging (Peters, 2009) and disease (Fields, 2008; Popescu and Lucchinetti, 2012; Calabrese et al., 2015; Wingerchuk et al., 2015; Gordon et al., 2014; Lucchinetti et al., 2011; Kang et al., 2013). For example, gray matter myelin degeneration is observed in the early stages of Alzheimer’s disease (AD) (Bartzokis, 2004; de la Monte, 1989), and, in relapsing remitting multiple sclerosis (RRMS), the presence of cortical lesions is strongly correlated with seizure occurrence (Calabrese et al., 2008). Several other diseases predicated on CNS demyelination, including neuromyelitis optica, acute disseminated encephalomyelitis, and progressive multifocal leukoencephalopathy may also display seizure activity (Jarius et al., 2014; Hynson et al., 2001; Lima et al., 2006). Demyelination of PV interneurons may perturb the balance of large-network neuronal excitation and inhibition, biasing neocortical microcircuits toward seizure.

By combining data from two independent efforts, we show here that a large fraction of cortical myelin is allocated to the axons of inhibitory neurons, specifically to PV basket cells. This work alters our understanding of the distribution of myelin in neocortical gray matter, and opens new avenues in basic neurobiology, systems neuroscience, and neurological disease research. Finally, this study leverages an openly available EM data set (Burns et al., 2013; Martone et al., 2002) that has been used in several other studies of neocortex (Tomassy et al., 2014; Bopp et al., 2014) since its publication (Bock et al., 2011), illustrating the value of image data sharing in biomedical research.

1. 1. Bock DD
   2. Lee WC
   3. Kerlin AM
   4. Andermann ML
   5. Hood G
   6. Wetzel AW
   7. Yurgenson S
   8. Soucy ER
   9. Kim HS
   10. Reid RC

   (2011) Network anatomy and in vivo physiology of visual cortical neurons

   Publicly available at the Open Connectome Project (Mouse V1; http://openconnecto.me/bock11/).

   http://openconnecto.me/catmaid/?pid=4&zp=0&yp=59903&xp=67711&tool=navigator&sid0=4&s0=10

1. 1. Álvarez-Quilón A
   2. Serrano-Benítez A
   3. Lieberman JA
   4. Quintero C
   5. Sánchez-Gutiérrez D
   6. Escudero LM
   7. Cortés-Ledesma F

   (2014) ATM specifically mediates repair of double-strand breaks with blocked DNA ends

   Nature Communications 5:3347.

   https://doi.org/10.1038/ncomms4347

   • Google Scholar
2. 1. Ascoli GA
   2. Alonso-Nanclares L
   3. Anderson SA
   4. Barrionuevo G
   5. Benavides-Piccione R
   6. Burkhalter A
   7. Buzsáki G
   8. Cauli B
   9. Defelipe J
   10. Fairén A
   11. Feldmeyer D
   12. Fishell G
   13. Fregnac Y
   14. Freund TF
   15. Gardner D
   16. Gardner EP
   17. Goldberg JH
   18. Helmstaedter M
   19. Hestrin S
   20. Karube F
   21. Kisvárday ZF
   22. Lambolez B
   23. Lewis DA
   24. Marin O
   25. Markram H
   26. Muñoz A
   27. Packer A
   28. Petersen CC
   29. Rockland KS
   30. Rossier J
   31. Rudy B
   32. Somogyi P
   33. Staiger JF
   34. Tamas G
   35. Thomson AM
   36. Toledo-Rodriguez M
   37. Wang Y
   38. West DC
   39. Yuste R
   40. Petilla Interneuron Nomenclature Group

   (2008) Petilla terminology: nomenclature of features of GABAergic interneurons of the cerebral cortex

   Nature Reviews Neuroscience 9:557–568.

   https://doi.org/10.1038/nrn2402

   • Google Scholar
3. 1. Barrera K
   2. Chu P
   3. Abramowitz J
   4. Steger R
   5. Ramos RL
   6. Brumberg JC

   (2013) Organization of myelin in the mouse somatosensory barrel cortex and the effects of sensory deprivation

   Developmental Neurobiology 73:297–314.

   https://doi.org/10.1002/dneu.22060

   • Google Scholar
4. 1. Bartzokis G

   (2004) Age-related myelin breakdown: a developmental model of cognitive decline and Alzheimer's disease

   Neurobiology of Aging 25:5–18.

   https://doi.org/10.1016/j.neurobiolaging.2003.03.001

   • Google Scholar
5. 1. Baumann N
   2. Pham-Dinh D

   (2001)

   Biology of oligodendrocyte and myelin in the mammalian central nervous system

   Physiological Reviews 81:871–927.

   • Google Scholar
6. 1. Beirowski B
   2. Babetto E
   3. Golden JP
   4. Chen YJ
   5. Yang K
   6. Gross RW
   7. Patti GJ
   8. Milbrandt J

   (2014) Metabolic regulator LKB1 is crucial for Schwann cell-mediated axon maintenance

   Nature Neuroscience 17:1351–1361.

   https://doi.org/10.1038/nn.3809

   • Google Scholar
7. 1. Bergles DE
   2. Richardson WD

   (2016) Oligodendrocyte Development and Plasticity

   Cold Spring Harbor Perspectives in Biology 8:a020453.

   https://doi.org/10.1101/cshperspect.a020453

   • Google Scholar
8. 1. Bock DD
   2. Lee WC
   3. Kerlin AM
   4. Andermann ML
   5. Hood G
   6. Wetzel AW
   7. Yurgenson S
   8. Soucy ER
   9. Kim HS
   10. Reid RC

   (2011) Network anatomy and in vivo physiology of visual cortical neurons

   Nature 471:177–182.

   https://doi.org/10.1038/nature09802

   • Google Scholar
9. 1. Bopp R
   2. Maçarico da Costa N
   3. Kampa BM
   4. Martin KA
   5. Roth MM

   (2014) Pyramidal cells make specific connections onto smooth (GABAergic) neurons in mouse visual cortex

   PLoS Biology 12:e1001932.

   https://doi.org/10.1371/journal.pbio.1001932

   • Google Scholar
10. 1. Braitenberg V

    (1962) A note on myeloarchitectonics

    Journal of Comparative Neurology 118:141–156.

    https://doi.org/10.1002/cne.901180202

    • Google Scholar
11. 1. Briggman KL
    2. Bock DD

    (2012) Volume electron microscopy for neuronal circuit reconstruction

    Current Opinion in Neurobiology 22:154–161.

    https://doi.org/10.1016/j.conb.2011.10.022

    • Google Scholar
12. 1. Burns R
    2. Roncal WG
    3. Kleissas D
    4. Lillaney K
    5. Manavalan P
    6. Perlman E
    7. Berger DR
    8. Bock DD
    9. Chung K
    10. Grosenick L
    11. Kasthuri N
    12. Weiler NC
    13. Deisseroth K
    14. Kazhdan M
    15. Lichtman J
    16. Reid RC
    17. Smith SJ
    18. Szalay AS
    19. Vogelstein JT
    20. Vogelstein RJ

    (2013) The open connectome project data cluster: scalable analysis and vision for high-throughput neuroscience

    Scientific and Statistical Database Management.

    https://doi.org/10.1145/2484838.2484870

    • Google Scholar
13. 1. Calabrese M
    2. De Stefano N
    3. Atzori M
    4. Bernardi V
    5. Mattisi I
    6. Barachino L
    7. Rinaldi L
    8. Morra A
    9. McAuliffe MM
    10. Perini P
    11. Battistin L
    12. Gallo P

    (2008) Extensive cortical inflammation is associated with epilepsy in multiple sclerosis

    Journal of Neurology 255:581–586.

    https://doi.org/10.1007/s00415-008-0752-7

    • Google Scholar
14. 1. Calabrese M
    2. Magliozzi R
    3. Ciccarelli O
    4. Geurts JJ
    5. Reynolds R
    6. Martin R

    (2015) Exploring the origins of grey matter damage in multiple sclerosis

    Nature Reviews Neuroscience 16:147–158.

    https://doi.org/10.1038/nrn3900

    • Google Scholar
15. 1. Caputi A
    2. Melzer S
    3. Michael M
    4. Monyer H

    (2013) The long and short of GABAergic neurons

    Current Opinion in Neurobiology 23:179–186.

    https://doi.org/10.1016/j.conb.2013.01.021

    • Google Scholar
16. 1. Cardona A
    2. Saalfeld S
    3. Schindelin J
    4. Arganda-Carreras I
    5. Preibisch S
    6. Longair M
    7. Tomancak P
    8. Hartenstein V
    9. Douglas RJ

    (2012) TrakEM2 software for neural circuit reconstruction

    PLoS One 7:e38011.

    https://doi.org/10.1371/journal.pone.0038011

    • Google Scholar
17. 1. Carter BC
    2. Bean BP

    (2009) Sodium entry during action potentials of mammalian neurons: incomplete inactivation and reduced metabolic efficiency in fast-spiking neurons

    Neuron 64:898–909.

    https://doi.org/10.1016/j.neuron.2009.12.011

    • Google Scholar
18. 1. Casale AE
    2. Foust AJ
    3. Bal T
    4. McCormick DA

    (2015) Cortical interneuron subtypes vary in their axonal action potential properties

    Journal of Neuroscience 35:15555–15567.

    https://doi.org/10.1523/JNEUROSCI.1467-13.2015

    • Google Scholar
19. 1. Chung K
    2. Deisseroth K

    (2013) CLARITY for mapping the nervous system

    Nature Methods 10:508–513.

    https://doi.org/10.1038/nmeth.2481

    • Google Scholar
20. 1. Collman F
    2. Buchanan J
    3. Phend KD
    4. Micheva KD
    5. Weinberg RJ
    6. Smith SJ

    (2015) Mapping synapses by conjugate light-electron array tomography

    Journal of Neuroscience 35:5792–5807.

    https://doi.org/10.1523/JNEUROSCI.4274-14.2015

    • Google Scholar
21. 1. Colonnier M

    (1968) Synaptic patterns on different cell types in the different laminae of the cat visual cortex. An electron microscope study

    Brain Research 9:268–287.

    https://doi.org/10.1016/0006-8993(68)90234-5

    • Google Scholar
22. 1. Curchoe CL
    2. Maurer J
    3. McKeown SJ
    4. Cattarossi G
    5. Cimadamore F
    6. Nilbratt M
    7. Snyder EY
    8. Bronner-Fraser M
    9. Terskikh AV

    (2010) Early acquisition of neural crest competence during hESCs neuralization

    PLoS One 5:e13890.

    https://doi.org/10.1371/journal.pone.0013890

    • Google Scholar
23. 1. Dan Y
    2. Poo MM

    (2006) Spike timing-dependent plasticity: from synapse to perception

    Physiological Reviews 86:1033–1048.

    https://doi.org/10.1152/physrev.00030.2005

    • Google Scholar
24. 1. de Hoz L
    2. Simons M

    (2015) The emerging functions of oligodendrocytes in regulating neuronal network behaviour

    BioEssays 37:60–69.

    https://doi.org/10.1002/bies.201400127

    • Google Scholar
25. 1. de la Monte SM

    (1989) Quantitation of cerebral atrophy in preclinical and end-stage Alzheimer's disease

    Annals of Neurology 25:450–459.

    https://doi.org/10.1002/ana.410250506

    • Google Scholar
26. 1. DeFelipe J
    2. Hendry SH
    3. Jones EG

    (1986) A correlative electron microscopic study of basket cells and large GABAergic neurons in the monkey sensory-motor cortex

    Neuroscience 17:991–1009.

    https://doi.org/10.1016/0306-4522(86)90075-8

    • Google Scholar
27. 1. DeFelipe J
    2. Hendry SH
    3. Jones EG

    (1989) Visualization of chandelier cell axons by parvalbumin immunoreactivity in monkey cerebral cortex

    PNAS 86:2093–2097.

    https://doi.org/10.1073/pnas.86.6.2093

    • Google Scholar
28. 1. Economo MN
    2. Clack NG
    3. Lavis LD
    4. Gerfen CR
    5. Svoboda K
    6. Myers EW
    7. Chandrashekar J

    (2016) A platform for brain-wide imaging and reconstruction of individual neurons

    eLife 5:e10566.

    https://doi.org/10.7554/eLife.10566

    • Google Scholar
29. 1. Fairén A
    2. Peters A
    3. Saldanha J

    (1977) A new procedure for examining Golgi impregnated neurons by light and electron microscopy

    Journal of Neurocytology 6:311–337.

    https://doi.org/10.1007/BF01175194

    • Google Scholar
30. 1. Feldman DE

    (2012) The spike-timing dependence of plasticity

    Neuron 75:556–571.

    https://doi.org/10.1016/j.neuron.2012.08.001

    • Google Scholar
31. 1. Feng G
    2. Mellor RH
    3. Bernstein M
    4. Keller-Peck C
    5. Nguyen QT
    6. Wallace M
    7. Nerbonne JM
    8. Lichtman JW
    9. Sanes JR

    (2000) Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP

    Neuron 28:41–45.

    https://doi.org/10.1016/S0896-6273(00)00084-2

    • Google Scholar
32. 1. Fields RD

    (2008) White matter in learning, cognition and psychiatric disorders

    Trends in Neurosciences 31:361–370.

    https://doi.org/10.1016/j.tins.2008.04.001

    • Google Scholar
33. 1. Fraher J
    2. Dockery P

    (1998) A strong myelin thickness-axon size correlation emerges in developing nerves despite independent growth of both parameters

    Journal of Anatomy 193:195–201.

    https://doi.org/10.1046/j.1469-7580.1998.19320195.x

    • Google Scholar
34. 1. Freund TF
    2. Gulyás AI

    (1991) GABAergic interneurons containing calbindin D28K or somatostatin are major targets of GABAergic basal forebrain afferents in the rat neocortex

    Journal of Comparative Neurology 314:187–199.

    https://doi.org/10.1002/cne.903140117

    • Google Scholar
35. 1. Fujinami RS
    2. Oldstone MB

    (1985) Amino acid homology between the encephalitogenic site of myelin basic protein and virus: mechanism for autoimmunity

    Science 230:1043–1045.

    https://doi.org/10.1126/science.2414848

    • Google Scholar
36. 1. Fünfschilling U
    2. Supplie LM
    3. Mahad D
    4. Boretius S
    5. Saab AS
    6. Edgar J
    7. Brinkmann BG
    8. Kassmann CM
    9. Tzvetanova ID
    10. Möbius W
    11. Diaz F
    12. Meijer D
    13. Suter U
    14. Hamprecht B
    15. Sereda MW
    16. Moraes CT
    17. Frahm J
    18. Goebbels S
    19. Nave KA

    (2012) Glycolytic oligodendrocytes maintain myelin and long-term axonal integrity

    Nature 485:517–521.

    https://doi.org/10.1038/nature11007

    • Google Scholar
37. 1. Gabbott PL
    2. Martin KA
    3. Whitteridge D

    (1987) Connections between pyramidal neurons in layer 5 of cat visual cortex (area 17)

    Journal of Comparative Neurology 259:364–381.

    https://doi.org/10.1002/cne.902590305

    • Google Scholar
38. 1. Galarreta M
    2. Hestrin S

    (1999) A network of fast-spiking cells in the neocortex connected by electrical synapses

    Nature 402:72–75.

    https://doi.org/10.1038/47029

    • Google Scholar
39. 1. Galarreta M
    2. Hestrin S

    (2002) Electrical and chemical synapses among parvalbumin fast-spiking GABAergic interneurons in adult mouse neocortex

    PNAS 99:12438–12443.

    https://doi.org/10.1073/pnas.192159599

    • Google Scholar
40. 1. Garg N
    2. Smith TW

    (2015) An update on immunopathogenesis, diagnosis, and treatment of multiple sclerosis

    Brain and Behavior 5:e00362.

    https://doi.org/10.1002/brb3.362

    • Google Scholar
41. 1. Gentet LJ
    2. Avermann M
    3. Matyas F
    4. Staiger JF
    5. Petersen CC

    (2010) Membrane potential dynamics of GABAergic neurons in the barrel cortex of behaving mice

    Neuron 65:422–435.

    https://doi.org/10.1016/j.neuron.2010.01.006

    • Google Scholar
42. 1. Gibson EM
    2. Purger D
    3. Mount CW
    4. Goldstein AK
    5. Lin GL
    6. Wood LS
    7. Inema I
    8. Miller SE
    9. Bieri G
    10. Zuchero JB
    11. Barres BA
    12. Woo PJ
    13. Vogel H
    14. Monje M

    (2014) Neuronal activity promotes oligodendrogenesis and adaptive myelination in the mammalian brain

    Science 344:.

    https://doi.org/10.1126/science.1252304

    • Google Scholar
43. 1. Gibson JR
    2. Beierlein M
    3. Connors BW

    (1999) Two networks of electrically coupled inhibitory neurons in neocortex

    Nature 402:75–79.

    https://doi.org/10.1038/47035

    • Google Scholar
44. 1. González-Albo MC
    2. Elston GN
    3. DeFelipe J

    (2001) The human temporal cortex: characterization of neurons expressing nitric oxide synthase, neuropeptides and calcium-binding proteins, and their glutamate receptor subunit profiles

    Cerebral Cortex 11:1170–1181.

    https://doi.org/10.1093/cercor/11.12.1170

    • Google Scholar
45. 1. Gordon HB
    2. Letsou A
    3. Bonkowsky JL

    (2014) The leukodystrophies

    Seminars in Neurology 34:312–320.

    https://doi.org/10.1055/s-0034-1386769

    • Google Scholar
46. 1. Gray EG

    (1959)

    Axo-somatic and axo-dendritic synapses of the cerebral cortex: an electron microscope study

    Journal of Anatomy 93:420–433.

    • Google Scholar
47. 1. Gritti I
    2. Manns ID
    3. Mainville L
    4. Jones BE

    (2003) Parvalbumin, calbindin, or calretinin in cortically projecting and GABAergic, cholinergic, or glutamatergic basal forebrain neurons of the rat

    Journal of Comparative Neurology 458:11–31.

    https://doi.org/10.1002/cne.10505

    • Google Scholar
48. 1. Gulyás AI
    2. Miles R
    3. Hájos N
    4. Freund TF

    (1993) Precision and variability in postsynaptic target selection of inhibitory cells in the hippocampal CA3 region

    European Journal of Neuroscience 5:1729–1751.

    https://doi.org/10.1111/j.1460-9568.1993.tb00240.x

    • Google Scholar
49. 1. Halasy K
    2. Buhl EH
    3. Lörinczi Z
    4. Tamás G
    5. Somogyi P

    (1996) Synaptic target selectivity and input of GABAergic basket and bistratified interneurons in the CA1 area of the rat hippocampus

    Hippocampus 6:306–329.

    https://doi.org/10.1002/(SICI)1098-1063(1996)6:3&amp;lt;306::AID-HIPO8&amp;gt;3.0.CO;2-K

    • Google Scholar
50. 1. Harris JJ
    2. Attwell D

    (2012) The energetics of CNS white matter

    Journal of Neuroscience 32:356–371.

    https://doi.org/10.1523/JNEUROSCI.3430-11.2012

    • Google Scholar
51. 1. Hartline DK
    2. Colman DR

    (2007) Rapid conduction and the evolution of giant axons and myelinated fibers

    Current Biology 17:R29–R35.

    https://doi.org/10.1016/j.cub.2006.11.042

    • Google Scholar
52. 1. Hasenstaub A
    2. Otte S
    3. Callaway E
    4. Sejnowski TJ

    (2010) Metabolic cost as a unifying principle governing neuronal biophysics

    PNAS 107:12329–12334.

    https://doi.org/10.1073/pnas.0914886107

    • Google Scholar
53. 1. Helmstaedter M

    (2013) Cellular-resolution connectomics: challenges of dense neural circuit reconstruction

    Nature Methods 10:501–507.

    https://doi.org/10.1038/nmeth.2476

    • Google Scholar
54. 1. Henny P
    2. Jones BE

    (2008) Projections from basal forebrain to prefrontal cortex comprise cholinergic, GABAergic and glutamatergic inputs to pyramidal cells or interneurons

    The European Journal of Neuroscience 27:654–670.

    https://doi.org/10.1111/j.1460-9568.2008.06029.x

    • Google Scholar
55. 1. Hodgson AJ
    2. Penke B
    3. Erdei A
    4. Chubb IW
    5. Somogyi P

    (1985) Antisera to gamma-aminobutyric acid. I. Production and characterization using a new model system

    Journal of Histochemistry & Cytochemistry 33:229–239.

    https://doi.org/10.1177/33.3.3973378

    • Google Scholar
56. 1. Hofer SB
    2. Ko H
    3. Pichler B
    4. Vogelstein J
    5. Ros H
    6. Zeng H
    7. Lein E
    8. Lesica NA
    9. Mrsic-Flogel TD

    (2011) Differential connectivity and response dynamics of excitatory and inhibitory neurons in visual cortex

    Nature Neuroscience 14:1045–1052.

    https://doi.org/10.1038/nn.2876

    • Google Scholar
57. 1. Hopf A

    (1966)

    [On a method for the objective registration of the myeloarchitectonics of the cerebral cortex]

    Journal Hirnforschung 8:301–313.

    • Google Scholar
58. 1. Hu H
    2. Cavendish JZ
    3. Agmon A

    (2013) Not all that glitters is gold: off-target recombination in the somatostatin-IRES-Cre mouse line labels a subset of fast-spiking interneurons

    Frontiers in Neural Circuits 7:195.

    https://doi.org/10.3389/fncir.2013.00195

    • Google Scholar
59. 1. Hu H
    2. Gan J
    3. Jonas P

    (2014) Interneurons. Fast-spiking, parvalbumin⁺ GABAergic interneurons: from cellular design to microcircuit function

    Science 345:1255263.

    https://doi.org/10.1126/science.1255263

    • Google Scholar
60. 1. Hu H
    2. Jonas P

    (2014) A supercritical density of Na(+) channels ensures fast signaling in GABAergic interneuron axons

    Nature Neuroscience 17:686–693.

    https://doi.org/10.1038/nn.3678

    • Google Scholar
61. 1. Hynson JL
    2. Kornberg AJ
    3. Coleman LT
    4. Shield L
    5. Harvey AS
    6. Kean MJ

    (2001) Clinical and neuroradiologic features of acute disseminated encephalomyelitis in children

    Neurology 56:1308–1312.

    https://doi.org/10.1212/WNL.56.10.1308

    • Google Scholar
62. 1. Jarius S
    2. Wildemann B
    3. Paul F

    (2014) Neuromyelitis optica: clinical features, immunopathogenesis and treatment

    Clinical and Experimental Immunology 176:149–164.

    https://doi.org/10.1111/cei.12271

    • Google Scholar
63. 1. Kamura K
    2. Kobayashi D
    3. Uehara Y
    4. Koshida S
    5. Iijima N
    6. Kudo A
    7. Yokoyama T
    8. Takeda H

    (2011) Pkd1l1 complexes with Pkd2 on motile cilia and functions to establish the left-right axis

    Development 138:1121–1129.

    https://doi.org/10.1242/dev.058271

    • Google Scholar
64. 1. Kang SH
    2. Li Y
    3. Fukaya M
    4. Lorenzini I
    5. Cleveland DW
    6. Ostrow LW
    7. Rothstein JD
    8. Bergles DE

    (2013) Degeneration and impaired regeneration of gray matter oligodendrocytes in amyotrophic lateral sclerosis

    Nature Neuroscience 16:571–579.

    https://doi.org/10.1038/nn.3357

    • Google Scholar
65. 1. Kawaguchi Y
    2. Kubota Y

    (1997) GABAergic cell subtypes and their synaptic connections in rat frontal cortex

    Cerebral Cortex 7:476–486.

    https://doi.org/10.1093/cercor/7.6.476

    • Google Scholar
66. 1. Kawaguchi Y
    2. Kubota Y

    (1998) Neurochemical features and synaptic connections of large physiologically-identified GABAergic cells in the rat frontal cortex

    Neuroscience 85:677–701.

    https://doi.org/10.1016/S0306-4522(97)00685-4

    • Google Scholar
67. 1. Keller A
    2. White EL

    (1987) Synaptic organization of GABAergic neurons in the mouse SmI cortex

    The Journal of Comparative Neurology 262:1–12.

    https://doi.org/10.1002/cne.902620102

    • Google Scholar
68. 1. Kepecs A
    2. Fishell G

    (2014) Interneuron cell types are fit to function

    Nature 505:318–326.

    https://doi.org/10.1038/nature12983

    • Google Scholar
69. 1. Kisvarday ZF
    2. Martin KA
    3. Friedlander MJ
    4. Somogyi P

    (1987) Evidence for interlaminar inhibitory circuits in the striate cortex of the cat

    Journal of Comparative Neurology 260:1–19.

    https://doi.org/10.1002/cne.902600102

    • Google Scholar
70. 1. Kisvárday ZF
    2. Martin KA
    3. Freund TF
    4. Maglóczky Z
    5. Whitteridge D
    6. Somogyi P

    (1986) Synaptic targets of HRP-filled layer III pyramidal cells in the cat striate cortex

    Experimental Brain Research 64:541–552.

    https://doi.org/10.1007/BF00340492

    • Google Scholar
71. 1. Kisvárday ZF

    (1992) GABAergic networks of basket cells in the visual cortex

    Progress in Brain Research 90:385–405.

    https://doi.org/10.1016/S0079-6123(08)63623-7

    • Google Scholar
72. 1. Kleinfeld D
    2. Bharioke A
    3. Blinder P
    4. Bock DD
    5. Briggman KL
    6. Chklovskii DB
    7. Denk W
    8. Helmstaedter M
    9. Kaufhold JP
    10. Lee WC
    11. Meyer HS
    12. Micheva KD
    13. Oberlaender M
    14. Prohaska S
    15. Reid RC
    16. Smith SJ
    17. Takemura S
    18. Tsai PS
    19. Sakmann B

    (2011) Large-scale automated histology in the pursuit of connectomes

    Journal of Neuroscience 31:16125–16138.

    https://doi.org/10.1523/JNEUROSCI.4077-11.2011

    • Google Scholar
73. 1. Kvitsiani D
    2. Ranade S
    3. Hangya B
    4. Taniguchi H
    5. Huang JZ
    6. Kepecs A

    (2013) Distinct behavioural and network correlates of two interneuron types in prefrontal cortex

    Nature 498:363–366.

    https://doi.org/10.1038/nature12176

    • Google Scholar
74. 1. Lee WC
    2. Bonin V
    3. Reed M
    4. Graham BJ
    5. Hood G
    6. Glattfelder K
    7. Reid RC

    (2016) Anatomy and function of an excitatory network in the visual cortex

    Nature 532:370–374.

    https://doi.org/10.1038/nature17192

    • Google Scholar
75. 1. Lee Y
    2. Morrison BM
    3. Li Y
    4. Lengacher S
    5. Farah MH
    6. Hoffman PN
    7. Liu Y
    8. Tsingalia A
    9. Jin L
    10. Zhang PW
    11. Pellerin L
    12. Magistretti PJ
    13. Rothstein JD

    (2012) Oligodendroglia metabolically support axons and contribute to neurodegeneration

    Nature 487:443–448.

    https://doi.org/10.1038/nature11314

    • Google Scholar
76. 1. Li P
    2. Huntsman MM

    (2014) Two functional inhibitory circuits are comprised of a heterogeneous population of fast-spiking cortical interneurons

    Neuroscience 265:60–71.

    https://doi.org/10.1016/j.neuroscience.2014.01.033

    • Google Scholar
77. 1. Li T
    2. Tian C
    3. Scalmani P
    4. Frassoni C
    5. Mantegazza M
    6. Wang Y
    7. Yang M
    8. Wu S
    9. Shu Y

    (2014) Action potential initiation in neocortical inhibitory interneurons

    PLoS Biology 12:e1001944.

    https://doi.org/10.1371/journal.pbio.1001944

    • Google Scholar
78. 1. Lima MA
    2. Drislane FW
    3. Koralnik IJ

    (2006) Seizures and their outcome in progressive multifocal leukoencephalopathy

    Neurology 66:262–264.

    https://doi.org/10.1212/01.wnl.0000194227.16696.11

    • Google Scholar
79. 1. Lucchinetti CF
    2. Popescu BF
    3. Bunyan RF
    4. Moll NM
    5. Roemer SF
    6. Lassmann H
    7. Brück W
    8. Parisi JE
    9. Scheithauer BW
    10. Giannini C
    11. Weigand SD
    12. Mandrekar J
    13. Ransohoff RM

    (2011) Inflammatory cortical demyelination in early multiple sclerosis

    The New England Journal of Medicine 365:2188–2197.

    https://doi.org/10.1056/NEJMoa1100648

    • Google Scholar
80. 1. Madisen L
    2. Zwingman TA
    3. Sunkin SM
    4. Oh SW
    5. Zariwala HA
    6. Gu H
    7. Ng LL
    8. Palmiter RD
    9. Hawrylycz MJ
    10. Jones AR
    11. Lein ES
    12. Zeng H

    (2010) A robust and high-throughput Cre reporting and characterization system for the whole mouse brain

    Nature Neuroscience 13:133–140.

    https://doi.org/10.1038/nn.2467

    • Google Scholar
81. 1. Makinodan M
    2. Rosen KM
    3. Ito S
    4. Corfas G

    (2012) A critical period for social experience-dependent oligodendrocyte maturation and myelination

    Science 337:1357–1360.

    https://doi.org/10.1126/science.1220845

    • Google Scholar
82. 1. Marek KW
    2. Kurtz LM
    3. Spitzer NC

    (2010) cJun integrates calcium activity and tlx3 expression to regulate neurotransmitter specification

    Nature Neuroscience 13:944–950.

    https://doi.org/10.1038/nn.2582

    • Google Scholar
83. 1. Markram H
    2. Toledo-Rodriguez M
    3. Wang Y
    4. Gupta A
    5. Silberberg G
    6. Wu C

    (2004) Interneurons of the neocortical inhibitory system

    Nature Reviews Neuroscience 5:793–807.

    https://doi.org/10.1038/nrn1519

    • Google Scholar
84. 1. Martin KA
    2. Whitteridge D

    (1984) Form, function and intracortical projections of spiny neurones in the striate visual cortex of the cat

    Journal of Physiology 353:463–504.

    https://doi.org/10.1113/jphysiol.1984.sp015347

    • Google Scholar
85. 1. Martone ME
    2. Gupta A
    3. Wong M
    4. Qian X
    5. Sosinsky G
    6. Ludäscher B
    7. Ellisman MH

    (2002) A cell-centered database for electron tomographic data

    Journal of Structural Biology 138:145–155.

    https://doi.org/10.1016/S1047-8477(02)00006-0

    • Google Scholar
86. 1. McGee AW
    2. Yang Y
    3. Fischer QS
    4. Daw NW
    5. Strittmatter SM

    (2005) Experience-driven plasticity of visual cortex limited by myelin and Nogo receptor

    Science 309:2222–2226.

    https://doi.org/10.1126/science.1114362

    • Google Scholar
87. 1. McGuire BA
    2. Gilbert CD
    3. Rivlin PK
    4. Wiesel TN

    (1991) Targets of horizontal connections in macaque primary visual cortex

    Journal of Comparative Neurology 305:370–392.

    https://doi.org/10.1002/cne.903050303

    • Google Scholar
88. 1. McKenzie IA
    2. Ohayon D
    3. Li H
    4. de Faria JP
    5. Emery B
    6. Tohyama K
    7. Richardson WD

    (2014) Motor skill learning requires active central myelination

    Science 346:318–322.

    https://doi.org/10.1126/science.1254960

    • Google Scholar
89. 1. Mengler L
    2. Khmelinskii A
    3. Diedenhofen M
    4. Po C
    5. Staring M
    6. Lelieveldt BP
    7. Hoehn M

    (2014) Brain maturation of the adolescent rat cortex and striatum: changes in volume and myelination

    NeuroImage 84:35–44.

    https://doi.org/10.1016/j.neuroimage.2013.08.034

    • Google Scholar
90. 1. Micheva KD
    2. Busse B
    3. Weiler NC
    4. O'Rourke N
    5. Smith SJ

    (2010) Single-synapse analysis of a diverse synapse population: proteomic imaging methods and markers

    Neuron 68:639–653.

    https://doi.org/10.1016/j.neuron.2010.09.024

    • Google Scholar
91. 1. Micheva KD
    2. Smith SJ

    (2007) Array tomography: a new tool for imaging the molecular architecture and ultrastructure of neural circuits

    Neuron 55:25–36.

    https://doi.org/10.1016/j.neuron.2007.06.014

    • Google Scholar
92. 1. Moore JW
    2. Joyner RW
    3. Brill MH
    4. Waxman SD
    5. Najar-Joa M

    (1978) Simulations of conduction in uniform myelinated fibers. Relative sensitivity to changes in nodal and internodal parameters

    Biophysical Journal 21:147–160.

    https://doi.org/10.1016/S0006-3495(78)85515-5

    • Google Scholar
93. 1. Morrison BM
    2. Lee Y
    3. Rothstein JD

    (2013) Oligodendroglia: metabolic supporters of axons

    Trends in Cell Biology 23:644–651.

    https://doi.org/10.1016/j.tcb.2013.07.007

    • Google Scholar
94. 1. Nave KA

    (2010) Myelination and support of axonal integrity by glia

    Nature 468:244–252.

    https://doi.org/10.1038/nature09614

    • Google Scholar
95. 1. Nieuwenhuys R

    (2013) The myeloarchitectonic studies on the human cerebral cortex of the Vogt-Vogt school, and their significance for the interpretation of functional neuroimaging data

    Brain Structure & Function 218:303–352.

    https://doi.org/10.1007/s00429-012-0460-z

    • Google Scholar
96. 1. Pajevic S
    2. Basser PJ
    3. Fields RD

    (2014) Role of myelin plasticity in oscillations and synchrony of neuronal activity

    Neuroscience 276:135–147.

    https://doi.org/10.1016/j.neuroscience.2013.11.007

    • Google Scholar
97. 1. Pawelzik H
    2. Hughes DI
    3. Thomson AM

    (2002) Physiological and morphological diversity of immunocytochemically defined parvalbumin- and cholecystokinin-positive interneurones in CA1 of the adult rat hippocampus

    The Journal of Comparative Neurology 443:346–367.

    https://doi.org/10.1002/cne.10118

    • Google Scholar
98. Book

    1. Peters A
    2. Palay SL
    3. Webster H. d

    (1991)

    The Fine Structure of the Nervous System: Neurons and Their Supporting Cells (3rd edn)

    Oxford University Press.

    • Google Scholar
99. 1. Peters A
    2. Proskauer CC

    (1980) Smooth or sparsely spined cells with myelinated axons in rat visual cortex

    Neuroscience 5:2079–2092.

    https://doi.org/10.1016/0306-4522(80)90126-8

    • Google Scholar
100. 1. Peters A

     (2009) The effects of normal aging on myelinated nerve fibers in monkey central nervous system

     Frontiers in Neuroanatomy 3:11.

     https://doi.org/10.3389/neuro.05.011.2009

     • Google Scholar
101. 1. Popescu BF
     2. Lucchinetti CF

     (2012) Pathology of demyelinating diseases

     Annual Review of Pathology 7:185–217.

     https://doi.org/10.1146/annurev-pathol-011811-132443

     • Google Scholar
102. 1. Preibisch S
     2. Saalfeld S
     3. Tomancak P

     (2009) Globally optimal stitching of tiled 3D microscopic image acquisitions

     Bioinformatics 25:1463–1465.

     https://doi.org/10.1093/bioinformatics/btp184

     • Google Scholar
103. 1. Puthussery T
     2. Gayet-Primo J
     3. Taylor WR
     4. Haverkamp S

     (2011) Immunohistochemical identification and synaptic inputs to the diffuse bipolar cell type DB1 in macaque retina

     Journal of Comparative Neurology 519:3640–3656.

     https://doi.org/10.1002/cne.22756

     • Google Scholar
104. 1. Rinholm JE
     2. Hamilton NB
     3. Kessaris N
     4. Richardson WD
     5. Bergersen LH
     6. Attwell D

     (2011) Regulation of oligodendrocyte development and myelination by glucose and lactate

     Journal of Neuroscience 31:538–548.

     https://doi.org/10.1523/JNEUROSCI.3516-10.2011

     • Google Scholar
105. 1. Rosetti CM
     2. Maggio B

     (2007) Protein-induced surface structuring in myelin membrane monolayers

     Biophysical Journal 93:4254–4267.

     https://doi.org/10.1529/biophysj.107.112441

     • Google Scholar
106. 1. Rudy B
     2. Fishell G
     3. Lee S
     4. Hjerling-Leffler J

     (2011) Three groups of interneurons account for nearly 100% of neocortical GABAergic neurons

     Developmental Neurobiology 71:45–61.

     https://doi.org/10.1002/dneu.20853

     • Google Scholar
107. 1. Schindelin J
     2. Arganda-Carreras I
     3. Frise E
     4. Kaynig V
     5. Longair M
     6. Pietzsch T
     7. Preibisch S
     8. Rueden C
     9. Saalfeld S
     10. Schmid B
     11. Tinevez JY
     12. White DJ
     13. Hartenstein V
     14. Eliceiri K
     15. Tomancak P
     16. Cardona A

     (2012) Fiji: an open-source platform for biological-image analysis

     Nature Methods 9:676–682.

     https://doi.org/10.1038/nmeth.2019

     • Google Scholar
108. 1. Seress L
     2. Ribak CE

     (1990) The synaptic connections of basket cell axons in the developing rat hippocampal formation

     Experimental Brain Research 81:500–508.

     https://doi.org/10.1007/BF02423498

     • Google Scholar
109. 1. Somogyi P
     2. Freund TF
     3. Cowey A

     (1982) The axo-axonic interneuron in the cerebral cortex of the rat, cat and monkey

     Neuroscience 7:2577–2607.

     https://doi.org/10.1016/0306-4522(82)90086-0

     • Google Scholar
110. 1. Somogyi P
     2. Kisvárday ZF
     3. Martin KA
     4. Whitteridge D

     (1983) Synaptic connections of morphologically identified and physiologically characterized large basket cells in the striate cortex of cat

     Neuroscience 10:261–294.

     https://doi.org/10.1016/0306-4522(83)90133-1

     • Google Scholar
111. 1. Stuart GJ
     2. Spruston N

     (2015) Dendritic integration: 60 years of progress

     Nature Neuroscience 18:1713–1721.

     https://doi.org/10.1038/nn.4157

     • Google Scholar
112. 1. Tamás G
     2. Buhl EH
     3. Somogyi P

     (1997) Fast IPSPs elicited via multiple synaptic release sites by different types of GABAergic neurone in the cat visual cortex

     Journal of Physiology 500:715–738.

     https://doi.org/10.1113/jphysiol.1997.sp022054

     • Google Scholar
113. 1. Taniguchi H
     2. He M
     3. Wu P
     4. Kim S
     5. Paik R
     6. Sugino K
     7. Kvitsiani D
     8. Kvitsani D
     9. Fu Y
     10. Lu J
     11. Lin Y
     12. Miyoshi G
     13. Shima Y
     14. Fishell G
     15. Nelson SB
     16. Huang ZJ

     (2011) A resource of Cre driver lines for genetic targeting of GABAergic neurons in cerebral cortex

     Neuron 71:995–1013.

     https://doi.org/10.1016/j.neuron.2011.07.026

     • Google Scholar
114. 1. Thomson AM
     2. West DC
     3. Hahn J
     4. Deuchars J

     (1996) Single axon IPSPs elicited in pyramidal cells by three classes of interneurones in slices of rat neocortex

     Journal of Physiology 496:81–102.

     https://doi.org/10.1113/jphysiol.1996.sp021667

     • Google Scholar
115. 1. Tomassy GS
     2. Berger DR
     3. Chen HH
     4. Kasthuri N
     5. Hayworth KJ
     6. Vercelli A
     7. Seung HS
     8. Lichtman JW
     9. Arlotta P

     (2014) Distinct profiles of myelin distribution along single axons of pyramidal neurons in the neocortex

     Science 344:319–324.

     https://doi.org/10.1126/science.1249766

     • Google Scholar
116. 1. Tomassy GS
     2. Dershowitz LB
     3. Arlotta P

     (2016) Diversity Matters: A Revised Guide to Myelination

     Trends in Cell Biology 26:135–147.

     https://doi.org/10.1016/j.tcb.2015.09.002

     • Google Scholar
117. 1. Tomer R
     2. Ye L
     3. Hsueh B
     4. Deisseroth K

     (2014) Advanced CLARITY for rapid and high-resolution imaging of intact tissues

     Nature Protocols 9:1682–1697.

     https://doi.org/10.1038/nprot.2014.123

     • Google Scholar
118. 1. Wang Y
     2. Gupta A
     3. Toledo-Rodriguez M
     4. Wu CZ
     5. Markram H

     (2002) Anatomical, physiological, molecular and circuit properties of nest basket cells in the developing somatosensory cortex

     Cerebral Cortex 12:395–410.

     https://doi.org/10.1093/cercor/12.4.395

     • Google Scholar
119. 1. Waxman SG
     2. Bennett MV

     (1972) Relative conduction velocities of small myelinated and non-myelinated fibres in the central nervous system

     Nature New Biology 238:217–219.

     https://doi.org/10.1038/newbio238217a0

     • Google Scholar
120. 1. Wingerchuk DM
     2. Banwell B
     3. Bennett JL
     4. Cabre P
     5. Carroll W
     6. Chitnis T
     7. de Seze J
     8. Fujihara K
     9. Greenberg B
     10. Jacob A
     11. Jarius S
     12. Lana-Peixoto M
     13. Levy M
     14. Simon JH
     15. Tenembaum S
     16. Traboulsee AL
     17. Waters P
     18. Wellik KE
     19. Weinshenker BG
     20. International Panel for NMO Diagnosis

     (2015) International consensus diagnostic criteria for neuromyelitis optica spectrum disorders

     Neurology 85:177–189.

     https://doi.org/10.1212/WNL.0000000000001729

     • Google Scholar
121. 1. Wucherpfennig KW
     2. Strominger JL

     (1995) Molecular mimicry in T cell-mediated autoimmunity: viral peptides activate human T cell clones specific for myelin basic protein

     Cell 80:695–705.

     https://doi.org/10.1016/0092-8674(95)90348-8

     • Google Scholar
122. 1. Xu X
     2. Roby KD
     3. Callaway EM

     (2010) Immunochemical characterization of inhibitory mouse cortical neurons: three chemically distinct classes of inhibitory cells

     Journal of Comparative Neurology 518:389–404.

     https://doi.org/10.1002/cne.22229

     • Google Scholar
123. 1. Young KM
     2. Psachoulia K
     3. Tripathi RB
     4. Dunn SJ
     5. Cossell L
     6. Attwell D
     7. Tohyama K
     8. Richardson WD

     (2013) Oligodendrocyte dynamics in the healthy adult CNS: evidence for myelin remodeling

     Neuron 77:873–885.

     https://doi.org/10.1016/j.neuron.2013.01.006

     • Google Scholar

Author details

1. Kristina D Micheva

   Department of Molecular and Cellular Physiology, Stanford University, Stanford, United States

   Contribution

   KDM, Conceived and designed the study, Performed IF in AT experiments, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   kmicheva@stanford.edu

   Competing interests

   KDM: Has founder's equity interests in Aratome, LLC (Menlo Park, CA), an enterprise that produces array tomography materials and services. Also listed as inventor on two US patents regarding array tomography methods that have been issued to Stanford University (US patents 7,767,414 and 9,008,378)

2. Dylan Wolman

   Janelia Research Campus, Howard Hughes Medical Institute, Ashburn, United States

   Contribution

   DW, Performed immunohistochemistry experiments, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   No competing interests declared.

   "This ORCID iD identifies the author of this article:" 0000-0002-5012-1690

3. Brett D Mensh

   Janelia Research Campus, Howard Hughes Medical Institute, Ashburn, United States

   Contribution

   BDM, Contributed to data analysis and interpretation, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   No competing interests declared.

4. Elizabeth Pax

   Janelia Research Campus, Howard Hughes Medical Institute, Ashburn, United States

   Contribution

   EP, Traced volume EM data, Analyzed immunohistochemistry data, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   No competing interests declared.

5. JoAnn Buchanan

   Department of Molecular and Cellular Physiology, Stanford University, Stanford, United States

   Present address

   Allen Institute for Brain Science, Seattle, United States

   Contribution

   JAB, Performed SEM in AT experiments, Acquisition of data, Drafting or revising the article

   Competing interests

   No competing interests declared.

6. Stephen J Smith

   Department of Molecular and Cellular Physiology, Stanford University, Stanford, United States

   Contribution

   SJS, Contributed to data analysis and interpretation, Drafting or revising the article

   Competing interests

   SJS: Has founder's equity interests in Aratome, LLC (Menlo Park, CA), an enterprise that produces array tomography materials and services. Also listed as inventor on two US patents regarding array tomography methods that have been issued to Stanford University (US patents 7,767,414 and 9,008,378)

7. Davi D Bock

   Janelia Research Campus, Howard Hughes Medical Institute, Ashburn, United States

   Contribution

   DDB, Conceived and designed the study, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   bockd@janelia.hhmi.org

   Competing interests

   No competing interests declared.

   "This ORCID iD identifies the author of this article:" 0000-0002-8218-7926

• 6,383

  views

• 1,422

  downloads

• 225

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.