Activity-Dependent Codevelopment of the Corticospinal System and Target Interneurons in the Cervical Spinal Cord

Cervical enlargement interneurons marked with ChAT and calcium-binding proteins

Both ChAT and the CBPs marked both topographically distinct and overlapping interneuron populations. For ChAT and CB, the distributions changed during the CST developmental refinement period. Figure 1 shows representative patterns of ChAT and CBP labeling. These confocal micrographs form the basis of the quantitative localization analyses shown below. The left panel of each part of the figure shows montages obtained by deconvolution confocal microscopy. The right panels contain higher-magnification epifluorescence images. Boxes in the left panels show the locations of the higher-magnification images. At PW3 and PW4, ChAT interneurons were consistently present medially in lamina 7, close to the central canal, as previously reported in rat pups (Phelps et al., 1984). There were a few scattered ChAT interneurons in other locations. Importantly, there was abundant ChAT labeling within the lateral and medial motor pools at these young ages. At PW8 (Fig. 1A1) and older, ChAT interneurons were present both medially in lamina 7 (Fig. 1A3, arrow) and more diffusely throughout laminae 5–7 (Fig. 1A2).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Typical expression pattern of interneuron markers (ChAT, CB, CR, PV) in the cervical enlargement (C6–C8) of the spinal cord. Left, Montages obtained by deconvolution confocal microscopy. Right, Single- or double-labeled epifluorescence micrographs, at higher magnification. Boxes in left panel figures show the approximate locations of micrographs in the right panels. A, ChAT. A1 is a montage from a PW8 animal. A2 is spinal section stained for ChAT (secondary antibody conjugated to Cy3; red) showing interneurons throughout laminae 5–7. A3 shows ChAT interneurons that are consistently present medially in lamina 7. The arrow points to ChAT interneurons close to the central canal. B, CB. B1 is a montage from a PW3 animal. B2 shows CB-positive interneurons (secondary antibody conjugated to FITC, green) in the dorsal horn. B3 shows CB staining in the intermediate zone and ventral horn. B3 inset shows a double-labeled interneuron, with CB staining (FITC) and PV staining (Cy3), in the ventral lamina 7 (lower small box, B1). C, CR. C1 shows a montage from a PW3 animal. C2 shows CR-positive interneurons (FITC) clustered medially in lamina 7. The arrows point to CR interneurons that also stained for ChAT (Cy3; double labeled, yellow). C3. CR interneurons (Cy 3; red) located laterally in the deep dorsal horn and intermediate zone. Note that the dense CR motor pool staining in the deconvolution image reflects dense dendritic (or axonal) processes. D, PV. D1 is a montage from a PW3 animal. D2 and D3 are of the same regions. D2 shows PV (Cy3) interneurons among the dense plexus of afferent fiber labeling. D3 shows the PV labeling (FITC) in relation to ChAT staining (Cy3). Arrows point to PV-positive interneurons that were consistently lateral to the medial ChAT cell group. Scale bars: All 500 μm, except C3, 250 μm.

At PW3 and PW4, CB interneurons were located throughout all laminae outside the motor pools. However, by PW8 there were few labeled cells ventral to laminae 3 and 4. Figure 1B1 shows CB labeling in a PW3 animal. While the dense superficial band of labeling dominates, scattered cells are present ventrally (Fig. 1B2,B3). A small number of CB-positive neurons in the ventral lamina 7 colocalized PV (Fig. 1B3, inset).

The distributions of CR- (Fig. 1C) (PW3) and PV- (Fig. 1D) (PW3) labeled interneurons were not remarkably different across the ages examined. Laterally located CR interneurons were often large, with polarized dendrites coursing dorsally and ventrally in the gray mater. A small number (∼1 per section) of CR-positive neurons in the medial lamina 7, but lateral to the medial ChAT group, colocalized ChAT (Fig. 1C2, arrows). PV cell body staining tended to be located medially in the gray matter, often in close proximity to the dense axonal PV labeling of large-diameter afferent fibers (Fig. 1D1). Figure 1D2 shows PV-positive interneurons among the dense plexus of afferent fiber labeling (Fig. 1D2, arrow). In medial lamina 7, PV-positive interneurons were consistently present lateral to the medial ChAT cell group (Fig. 1D3; arrows point to PV interneurons, green; ChAT interneurons are red).

Cervical enlargement interneurons develop distinctive regional distribution patterns by PW8

Representative examples of regional distribution maps of marked interneurons are shown in Figure 2 for PW4, PW8, and mature cats. The distributions and relative numbers of marked cells appeared to be the same at PW3 as at PW4 (see below). To generate these maps, the color scale was adjusted for each animal. The scale was set so that there was minimal color saturation (i.e., dark red) for each image. Figure 2 presents three key findings. First, ChAT interneurons increase substantially between PW4 and PW8, which is the CST refinement period. Second, CB interneurons decrease substantially between PW4 and PW8. Changes in ChAT and CB density are addressed further below when we describe the results of an unbiased stereological analysis. Third, at all ages, ChAT-, CR-, and PV-positive neurons had distinctive, often highly localized, distributions. These findings show that there is a maturation of the distribution of CB and ChAT interneurons that parallels the timing of refinement of CST axon terminations and emergence of the M1 motor map (Martin et al., 2009).

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Regional distribution of spinal interneurons at PW4, PW8, and ≥PW12. The locations of labeled interneurons were marked at 100× magnification in at least two animals per age group. Local density of interneurons (see Materials and Methods) is represented as a color scale (i.e., heat maps). The maps are averages, generated from the deconvolution images (as in Fig. 1; left column) from individual animals. Representative examples are shown. A, ChAT. Note the increase in ChAT interneurons between PW4 and PW8. B, CB. Note that CB interneuron numbers decreased remarkably between PW4 and PW8. C, CR. D, PV. No age-dependent changes were noted for either CR or PV. Scale bar: 500 μm; range for the color scale (in cells per millimeter): A, PW4: 0–33, PW8: 0–43, adult: 0–47; B, PW4: 0–60, PW8: 0–74, adult: 0–79; C, D, all 0–33.

Unbiased stereological counts recapitulate changes in regional distribution maps

To quantify our results of age-dependent changes in interneuron markers, we used the stereological method, optical fractionator, to assess cell numbers in key gray matter regions quantitatively. We combined PW3 and PW4, as a single immature group, and PW8 and >PW12, as a single mature group. This was justified because the results of a hierarchical cluster analysis (supplemental Fig. 1A, available at www.jneurosci.org as supplemental material) using a four-age-group model (PW3, PW4, PW8, and ≥PW12) as our null hypothesis revealed two principal clusters, immature (blue lines) and mature (green lines). Each cluster had a single cat that was an outlier for its own age group. A second statistical analysis, K-means clustering, also produced two principal components, immature and mature, with the same two outliers (supplemental Fig. 1B, available at www.jneurosci.org as supplemental material). This provides justification for combining ages and additional support for our hypothesis of interneuronal changes occurring during the CST refinement period.

Figure 3 plots results from the stereological cell counts. For ChAT, we sampled separately lamina 7 dorsal (Fig. 3B), which included the medial ChAT cell group, and laminae 4–6 (Fig. 3A). For both territories, we found a significant increase between the immature (open bar) and mature groups (gray bar; lamina 7, unpaired t test, p < 0.001, df = 6; laminae 4–6, unpaired t test, p = 0.0002, df = 6). Note that the increase in ChAT labeling in lamina 7 reflected substantial increases in numbers of small cells located lateral to the most medial ChAT group, which were present from PW3 (Fig. 2A). We also observed a reduction in CB labeling between the immature and mature groups, less in laminae 1–3 (Fig. 2C) (unpaired t test, p = 0.0139, df = 6) than in laminae 4–6 (Fig. 2D) (p = 0.004, df = 6). CR showed a small reduction that was not significant p = 0.34, df = 6), as did PV (p = 0.70, df = 6). These findings indicate, using an unbiased stereological approach, that there are consistent increases in ChAT-positive interneurons and decreases in CB interneurons before and after the CST refinement period. Thus, as the spinal terminations of the CS system achieve a mature topography between PW5 and PW7, so too do two spinal interneuron classes.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

Unbiased measurements of the density of spinal interneurons. The optical fractionator method was used to determine changes in the density of labeled interneurons. The insets in each part show the spinal region measured (gray). Animals were grouped as immature (PW3 and PW4) and mature (≥PW8). As discussed in the Results, cluster analysis provided justification for pooling into these two age groups. Cell density was calculated, based on estimated cell counts for each animal in each age group (5 × 40 μm sections/animal, 4 animals in each group). The mean density for each group and SEM are presented. A, ChAT, within laminae 4–6. B, ChAT, within lamina 7 dorsal. C, CB, within laminae 1–3. D, CB, within laminae 4–6. E, CR, within laminae 4–6. F, PV, within laminae 4–6. Note that, for the two age groups, there were significant increases in ChAT cell density in both laminae 4–6 (A) and lamina 7 (B) and significant decreases in CB density in laminae 1–3 (C) and 4–6 (D). In contrast, E and F show the lack of a significant change in density for CR- and PV-positive neurons in the same animals. lam, Laminae.

Interneuron clusters are marked by ChAT and CBP staining

We next defined the locations of clusters of marked interneurons by fitting contours around the densest areas (60%; see Materials and Methods) of the interneuron regional distribution maps for each cat. This analysis (Fig. 4) combined data across all animals and provides spatial quantification of the locations of interneurons. ChAT neurons were consistently present only in medial lamina 7 at PW3 and PW4 (Fig. 4A, left two columns). This group persisted into maturity. On a given 40 μm section, there typically were two to five ChAT-positive cells in this region. By PW8, there was a substantial increase in ChAT labeling throughout the dorsal horn and intermediate zone.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

Interneuron clusters marked by ChAT and CBPs. The left half of the figure shows the locations of clusters (see Materials and Methods for how these were obtained) fitted for all animals (2 per age group). The right two columns are semischematic representations, combining and summarizing the immature (PW3 and PW4) and mature (PW8 and >PW12) groups. The size of all spinal cords was adjusted so that the dorsoventral length of the gray matter was the same. A, ChAT neurons were consistently present only in medial lamina 7 at PW3 and PW4 (left two columns). Note the emergence of clusters in the older animals. B, CB neurons show a clear shift from a diffuse pattern in the immature animals to a focused pattern in laminae 1–3 in the mature animals. C and D show a diffuse distribution of CR and PV cells, with no apparent change with age. Rexed's laminae are defined in one of the top summary column images.

CB-positive interneurons were distributed widely throughout the gray matter outside the motor pools at PW3 and PW4 (Fig. 4B, left two columns). At PW8 and mature animals, CB-positive interneurons were localized primarily to the superficial laminae. There were a few CB-positive cells located ventrally (Fig. 2B). One consistent group was located in the ventral portion of lamina 7, although because of the small number of cells, it fell below the 60% cutoff. There were no systematic changes in the distribution of clusters of CR and PV interneurons (Fig. 4C,D).

Semischematic summaries of the contour data, split according to whether the animal had an immature CST spinal termination pattern (PW3 and PW4) or a mature pattern (PW8 and mature) are shown in the right two columns of Figure 4. These summaries clearly show marked differences in ChAT and CB interneurons at immature and mature ages.

From the perspective of the spinal gray matter, superposition of the contours, both in immature and mature animals, shows that there are areas with unique interneuron populations and areas of mixed interneuron groups. The only gray matter zone selective for a single interneuron class in immature animals was the most medial portion of lamina 7, which contains the medial ChAT cell group. Older than PW8, additional predominantly selective zones emerged: laminae 1–3, for CB; lamina 4 primarily contained PV interneurons; the middle portion of laminae 5–7 dorsal contained ChAT interneurons primarily; and lateral 5–7 contained predominantly CR cells. The remaining areas had, to varying degrees, combinations of one or more of the markers studied. Supplemental Figure 2 (available at www.jneurosci.org as supplemental material) shows a semischematic summary of the gray matter territories containing the different cell groups. This is discussed further below, in the context of CST termination patterns. Our findings suggest, on the basis of the distributions of interneuron markers, that most zones contain overlapping populations of interneuron classes, and likely overlapping functions. However, there are several regions that have a more selective population of marked interneurons, which may point to a more functionally homogeneous region.

CST terminations target distinctive spinal interneuron populations

To begin to understand whether the CST targets or avoids connections with particular interneuron populations during development and in maturity, we determined the spatial relationships between the CST projection to the cervical enlargement and the locations of ChAT- and CBP-marked interneurons. We reasoned that CST terminations within the territory of a particular marked interneuron class (e.g., CR in lateral laminae 5–7) (Figs. 2, 4) are likely to make connections with those interneurons since the CST terminations are in the immediate vicinity of the marked interneuron cell bodies. In contrast, terminations that spare the region of a particular labeled interneuron class are less likely to make connections with the interneurons. Although, in the case of the large CR interneuron in lateral lamina 7, their dendritic arbors extend the territory over which contacts can be made. Our contour analysis, which revealed distinctive locations of interneurons identified by ChAT and CBPs, provides the basis for this analysis. We conducted two separate analyses one based on composite CST axon labeling (see Materials and Methods) and one based on the distributions of CST axons and varicosities (i.e., boutons; see Materials and Methods) and interneurons in individual animals.

Figure 5 overlays the distributions of ChAT- and CBP-labeled interneurons (black outlines and semitransparent fill; darker fill indicates more overlap between animals) on composite CST axon maps. As predicted from earlier studies (Theriault and Tatton, 1989; Li and Martin, 2000), CST labeling in immature animals covers most of the gray matter, including superficial laminae and the motor pools (Fig. 5, left). Labeling at this age is densest contralaterally, but also present ipsilaterally (Theriault and Tatton, 1989; Alisky et al., 1992; Martin et al., 1999). Moreover, densest labeling was within the deep dorsal horn and intermediate zone. In mature animals, labeling was more restricted dorsoventrally (Fig. 5, right).

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

Topographic relationship between composite CST contralateral termination patterns and interneuron cluster overlays. Each composite map was constructed from four animals, two with rostrally placed BDA tracer injections and two with caudally placed injections. A–D overlay interneuron cell clusters on the composite CST distribution maps. Each cell cluster from each animal is shown as a black outline and 50% transparent fill. The greater the overlap between animals, the darker the fill. As in the analysis for Figure 4, to facilitate spatial comparisons the dorsoventral length of the gray matter was adjusted to be the same in all animals. A, ChAT. B, Calbindin. C, Calretinin. D, Parvalbumin. Scale bars: 1 mm; range for color scale: length, 0–800 μm axon length.

For ChAT, the immature CST projections terminate predominantly within the future location of the more lateral placed interneurons. The composite map in Figure 5A (left) shows labeling located lateral to the territory of the medial ChAT cells in lamina 7. As discussed below, projections to this region were sparse in PW3 and PW4 animals, but were denser by 8 weeks. In the mature animal (Fig. 5A, right), there was a remarkable correspondence between the region of densest contralateral CST labeling and the territory occupied by ChAT cells. For CB, the sparse CST projection to the superficial dorsal horn in immature animals overlaps the high density interneuronal population (Fig. 5B, left). Postnatal withdrawal of the superficial CST projections (Theriault and Tatton, 1989; Alisky et al., 1992; Li and Martin, 2001) occurs in parallel with the reduction in the overall level of CB expression in that region (Fig. 3). Whereas the CST projection is maintained in the deep dorsal horn and intermediate zone, there is a corresponding large-scale reduction in CB expression, so much so in laminae 4–8 that groups dropped out (Fig. 5B, right). Not surprisingly, for the immature CST there was general overlap with CR- and PV-positive cells, since these interneuronal markers are diffuse (Fig. 5C,D). After refinement of CST terminations, the most dorsal and most ventral PV and CR cells remained outside the CST termination territory. Importantly, the regions with densest CST labeling, the medial portions of laminae 5–7 dorsal, target the area containing a high density of CR- and PV-positive neurons. The findings that ChAT immunoreactivity increases and CB immunoreactivity decreased in temporal and spatial association with CST terminations raises the possibility that the CST plays a role in these directing these changes (see below).

We next compared focal densities of CST labeling with interneuronal labeling in the same animal (Fig. 6; axon labeling, left panels; varicosity labeling, right panels). We focused primarily on the correspondence between the medial ChAT neurons, which formed a readily identifiable population, and CST terminations. This analysis supports our observation that the medial ChAT neurons in lamina 7 were not targeted by the immature CST projection. Often, there were conspicuous gaps in the distribution of labeling suggesting that the projection avoided this region (Fig. 6, PW4). Rather, CST terminations in the younger animals were consistently located more lateral, but still within the medial portions of laminae 5–7 dorsal. This region contained cells, singly labeled for all four markers (Fig. 4, right columns; supplemental Fig. 2, available at www.jneurosci.org as supplemental material). There was a focus of dense varicosities in the medial portion of laminae 5–6 (Fig. 6B, upper arrow) that was distinct from more ventral projections (Fig. 6B, lower arrow). In more mature animals, we saw more medial lamina 7 labeling, close to the central canal, that was localized within the medial ChAT-positive cells. In the PW8 animal shown, though the densest labeling was in medial laminae 5 and 6, a lower density region of varicosities extends to the medial ChAT cells (Fig. 6C). Tracer injection in the PW12 animal (Fig. 6D) labeled several clusters within lamina 7, including the medial ChAT cell group (white contour) and the lateral CR group (arrows).

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

Targeting of CST projections to particular contralateral spinal subregions. Each panel shows the regional distribution of axons and varicosities for a representative animal. Local density is represented as a color scale. The insets between the axon and varicosity distributions in A show the locations of tracer injection sites for A–D, and the inset in E shows that for tracer injections in E. The location of the medial lamina 7 ChAT cell group for each animal is shown (thick white outline). A, PW3. B, PW4. Note arrows pointing to two distinct CST termination clusters, in lamina 6 and lamina 7. Note that the CST does not target the medial ChAT cell group at PW3 and PW4. C, PW8. CST labeling now extends into the region of the medial ChAT group. Yellow arrows point to labeling associated with the lateral CR interneurons. D, PW12. Yellow arrows point to labeling present among the CR-positive interneuron in lateral laminae 6–7. E, Adult. Arrows in E1 and E2 point to a cluster of labeling present with the caudal injection but not the rostral injection. Diagonal white arrow in E3 and E4 points to a cluster of CST terminations absent in E1 and E2. Horizontal yellow arrows in E3 and E4 point to labeling present among the CR-positive interneuron in lateral laminae 5–7. Scale bars: 1 mm; color scale range: left panels, axon length: 0–400 mm of axon length; right panels, varicosities (number per millimeter): A–C, 0–120; D, 0–60; E2, 0–115; E4, 0–80.

Two different tracers were used in another PW12 animal (Fig. 6E; more caudal M1 injections, E1, E2; more rostral M1 injections, E3, E4). Whereas the distribution of labeling from the adjoining regions were similar—both targeted laminae 5–7—closer examination revealed important differences that only can be interpreted in the context of the locations of labeled interneuron groups: the more rostral injection targeted the medial ChAT group (Fig. 6E3,E4), whereas the more caudal one did not (Fig. 6E1,E2). The lateral CR-positive region received a projection from one (Fig. 6E3,E4) but not the other (Fig. 6E1,E2) M1 region. Overall, the termination patterns from the two adjoining areas were remarkably interdigitated (Fig. 6E1,E3), suggesting a high degree of connectional specificity between M1 and spinal motor circuits. Our findings suggest that the CST targets distinctive populations of spinal interneurons at different times during development.

Activity-dependent changes in CB- and ChAT-positive interneurons

The findings presented above show that CB expression decreases, and ChAT increases, during the period when CST axon terminations are refined. This raises the possibility that these changes depend on a functioning CST input. To test this idea, we inactivated M1 unilaterally during the CST refinement period, PW5–PW7, and examined the distributions of CBPs and ChAT contralateral and ipsilateral to inactivation. Three cats were 8 weeks old and two were 12 weeks old at the time of perfusion. We combined data from these two age groups because no systematic differences were observed. This inactivation produces, without further activity-based intervention, a permanent shift in the distribution of inactivated CST terminations, mostly dorsal to lamina 4 (Friel and Martin, 2005). Inactivation also produces a permanent behavioral impairment, in which cats show movement endpoint biases during reaching (overreaching) and visually guided stepping (overstepping) and grasping impairments (Martin et al., 2000; Friel et al., 2007).

Figure 7A shows the regional density of CB-positive interneurons in a representative animal. There were noticeably more CB-positive interneurons on the inactive (contralateral to inactivation) than the active (ipsilateral to inactivation) side in laminae 1–3. We counted the number of CB-positive interneurons using the optical fractionator, as above, and found more than a doubling of neuron density on the inactive than active side (Fig. 7C). This difference was significant (paired t test, p < 0.01). No side-to-side differences were observed in an animal that received unilateral saline infusion during the same period (contralateral to infusion: 6.78 × 10⁻⁴ cells/μm²; ipsilateral to infusion: 6.83 × 10⁻⁴ cells/μm²). Relative to the active side, it was as if unilateral inactivation prevented the reduction in CB expression in the superficial laminae that normally occurs between PW5 and PW7 (Fig. 2B).

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Development of CB and ChAT interneurons depends on M1 activity. A, Representative regional distribution of CB interneurons in the cervical spinal cord contralateral (right) and ipsilateral (left) to M1 inactivation. Cell density is represented as a color scale. B, Similar to A, but for ChAT. C, Unbiased measurement of CB interneuron density, using the optical fractionator method. Note that there is more than a doubling of neuron density on the inactive over the active side. The bar graph plots mean. Data from individual animals are shown. D, As in C, but for ChAT interneuron density. E, Ratio of ChAT to CB, calculated for each animal. Scale bar: (in B) A, B, 1 mm. Color scale range: in (B) A, 0–75 cells/mm; B, 0–30 cells/mm.

Surprisingly, CB interneuron density on the inactive side was not significantly different from age-matched controls (means: inactive, 5.47 × 10⁻⁴ ± 2.85 × 10⁻⁵; controls, 6.21 × 10⁻⁴ ± 6.55 × 10⁻⁵; unpaired t test, p = 0.2951, df = 7); CB density on the active side (2.33 × 10⁻⁴ ± 3.09 × 10⁻⁵) was significantly less than the inactive side (paired t test, p < 0.0001, df = 8). However, the relative difference in CB interneuron counts between the inactive and active sides (i.e., high and low numbers, respectively) was the same as the relative difference for immature (i.e., PW3 and PW4) and mature (i.e., PW8 and adult) cats.

An asymmetry in ChAT-positive interneurons also was produced by unilateral inactivation. Figure 7B shows the regional density of ChAT-positive interneurons in a representative animal. The inactive side (contralateral to inactivation) contained significantly fewer ChAT-positive interneurons than the active side, as shown in Figure 7D (paired t test, p < 0.01). Side-to-side asymmetry was not present in the animal subjected to saline infusion (contralateral to infusion: 3.45 × 10⁻⁴; ipsilateral to infusion: 3.26 × 10⁻⁴). Thus, as with CB, inactivation produced less mature (i.e., lower) regional density of ChAT-positive interneurons in the contralateral spinal gray matter.

ChAT interneuron density was significantly less on the inactive side than in age-matched controls (means: inactive, 2.76 × 10⁻⁴ ± 6.09 × 10⁻⁵; controls, 7.37 × 10⁻⁴ ± 4.82 × 10⁻⁵, unpaired t test p < 0.01). ChAT neuron density was significantly higher on the active side (5.01 × 10⁻⁴ ± 7.61 × 10⁻⁵; paired t test, p = 0.05, df = 8). While this is different from CB, where inactive and age-matched controls were the same, ChAT also showed the same relative difference with respect to younger animals. ChAT interneuron counts were lower on the inactive sides, which is the same as immature cats, and higher on the active side, as in mature cats.

In contrast to affecting CB and ChAT interneuron development, CS system inactivation had no effect on PV and CR interneuron development (PV, PW3 and PW4: 1.67 × 10⁻⁴, controls: 1.54 × 10⁻⁴, paired t test, p = 0.57, CR, PW3 and PW4: 1.94 × 10⁻⁴, controls: 1.49 × 10⁻⁴, paired t test, p = 0.44). These findings support our hypothesis that CS system activity is necessary for the postnatal changes in ChAT and CB.

On the inactive side, the density of CB-positive interneurons was high and the density of ChAT interneurons was low. This is the general pattern of CB and ChAT in immature animals. In contrast, on the active side the density of CB-positive interneurons was low and the density of ChAT interneurons was high. This pattern was like that in mature animals. Our hypothesis is that the ChAT–CB ratio on the inactive side, rather than the absolute numbers of cells, will be similar to that in immature animals and that the ratio on the active side, will be similar to that in mature animals. We next compared the ratio of ChAT-positive to CB-positive interneuron density for animals subjected to unilateral inactivation, with that in normally developing animals (Fig. 7E). As predicted, PW3 and PW4 animals have an immature CST topography and they also have a ChAT–CB ratio close to zero. PW8 and older animals have a CST with a mature termination pattern and they have a ChAT–CB ratio of one and greater (Fig. 7E, black dots). After unilateral inactivation, the inactive side (gray dots) had ratios that were less than one, similar to the immature animals, and the active side (gray dots), greater than one, as with the mature animals. These results support our hypothesis that interneurons within the CST spinal target zone contralateral to CS system inactivation have a more immature ChAT–CB pattern, than the active side.