Abstract It is well established that motor impairment often occurs alongside healthy aging, leading to problems with fine motor skills and coordination. Although previously thought to be caused by neuronal death accumulating across the lifespan, it is now believed that the source of this impairment instead stems from more subtle changes in neural connectivity. The dendritic spine is a prime target for exploration of this problem because it is the postsynaptic partner of most excitatory synapses received by the pyramidal neuron, a cortical cell that carries much of the information processing load in the cerebral cortex. We repeatedly imaged the same dendrites in young adult and aged mouse motor cortex over the course of 1 month to look for differences in the baseline state of the dendritic spine population. These experiments reveal increased dendritic spine density, without obvious changes in spine clustering, occurring at the aged dendrite. Additionally, aged dendrites exhibit elevated spine turnover and stabilization alongside decreased long-term spine survival. These results suggest that at baseline the aged motor cortex may exist in a perpetual state of relative instability and attempts at compensation. This phenotype of aging may provide clues for future targets of aging-related motor impairment remediation. apical dendrite, cognitive impairment, in vivo imaging, spine clustering, structural plasticity Introduction Healthy aging, even while unassociated with disease or disorder, often coincides with deficits in cognitive processing that manifest as behavioral deficiencies in sensory discrimination (Stevens et al. 1996; Manning and Tremblay 2009), fine motor control (Parikh and Cole 2012; Hoogendam et al. 2014), coordination (Seidler et al. 2002; Hussein et al. 2013), and other similar faculties. Deficits along these same lines have been uncovered in the aged rodent, such as impaired motor coordination (Wallace et al. 1980; Justice et al. 2014). Following years of anatomical study, it is now believed that this consequence of aging is due to changes in neuron connectivity and not neuronal death (Burke and Barnes 2006; Morrison and Baxter 2012). While aging-related motor impairment is a complex, multisystem problem involving many peripheral sensory and motor structures, the health and ability of the central nervous system must not be overlooked. As the main director of voluntary muscle action and sensory-driven behavior, the primary motor cortex (M1) is an important area for examination of this problem. The pyramidal neurons (PNs) in layer 5 (L5) of this cortical region project their axons to the brainstem and spinal cord to directly drive lower motor neurons that control spinal circuits. The action of PNs is significantly influenced by the activity occurring at their apical dendritic tufts (Regehr et al. 1993; Gasparini et al. 2004; Gasparini and Magee 2006). In M1, the axons feeding information to these dendrites most often originate from PNs of the secondary motor cortex (M2) and relay cells of the motor regions of the thalamus (Hooks et al. 2013) as well as from layer 2/3 (L2/3) PNs within M1 itself (Weiler et al. 2008; Anderson et al. 2010). PN dendrites house dynamic structural protrusions from the dendritic shaft known as dendritic spines that constitute the postsynaptic piece of nearly all excitatory connections received by the PN (Uchizono 1965; LeVay 1973; Yuste and Denk 1995). For this reason, the structural plasticity of spines can be used as a proxy for plasticity at excitatory synapses. Until recently, study of this structural plasticity has been limited to snapshots taken from postmortem fixed tissue. However, the advent of two-photon excitation (2PE) microscopy (Denk et al. 1990) coupled with strategies for in vivo experiments (Lendvai et al. 2000) offers an avenue for more complete understanding of these structures and their role in the brain within a temporal context. This approach demonstrated that the spine population is extremely dynamic even in the adult brain (Oray et al. 2004) and that new spines house synapses and can turnover in an activity-dependent manner (Trachtenberg et al. 2002). Furthermore, the role of spines in learning and memory was validated when it was shown that spines formed during learning-induced plasticity persist long after the new experience ends (Hofer et al. 2008) and that survival of these spines can be indicative of retained learning (Yang et al. 2009). Along these same lines, many studies have thoroughly demonstrated the importance of spine dynamics in juvenile and young adult M1 to motor learning and skill performance using in vivo imaging approaches. This includes but is not limited to new spine formation (Xu et al. 2009), branch specificity (Yang et al. 2014), spatial clustering (Fu et al. 2012), stabilization (Clark et al. 2018), and persistence (Hayashi-Takagi et al. 2015). But these dynamic structures, evidently so crucial for learning and memory, are not as well studied in the aged brain in vivo, leaving our understanding of the impact of aging at this postsynaptic site incomplete. Results of studies throughout the aged cortex, in vivo and from postmortem alike, suggest that aging-related changes to dendrites and synaptic density are not generalized and differ between cortical areas. Our results aim to expand this characterization of the aged cortex at the cellular level. The present study examines the baseline behavior of these spines to characterize aging-related changes to the information processing ability of L5 PNs in aged M1 at the morphological level, anticipating that these changes in connectivity may underlie aging-related motor impairment. We examined baseline structural plasticity of dendritic spines on apical dendrites of L5 PNs in M1 of young adult (3–5 months) and aged (20–23 months) mice using high-resolution in vivo 2PE imaging. We found that spine density is increased in aged dendrites, while spine clustering in general appears unaffected. Also, aged dendrites show elevated turnover and short-term stabilization of spines, while the long-term survival of spines is lower. Materials and Methods Animals Male and female transgenic mice expressing eGFP under the Thy-1 promotor (The Jackson Laboratory, 007788, Tg (Thy1-EGFP)MJrs/J) were used. The sparse expression of eGFP by pyramidal tract-type L5 PNs in the cortex (Feng et al. 2000; Popescu et al. 2017) makes this strain well suited for repeated in vivo imaging of apical dendritic tufts of cortical PNs. All animals were virgin and were group housed with cagemates of the same sex. Food and water were available ad libitum, and cages were kept under a 12-h light/dark cycle. The study was carried out in accordance with the recommendations of the NIH Office of Laboratory Animal Welfare’s Public Health Service Policy on Humane Care and Use of Laboratory Animals and Guide for the Care and Use of Laboratory Animals, and all procedures described were approved by the Institutional Animal Care and Use Committee of Tulane University. Cranial Window Procedure Implantation of glass-covered cranial windows was carried out as previously described (Mostany and Portera-Cailliau 2008; Holtmaat et al. 2009). Briefly, animals were anesthetized with isoflurane (5.0% for induction, 1.5–2.0% for maintenance), and the skull was secured into place on a stereotaxic frame. Dexamethasone (0.2 mg/kg bw) and carprofen (5.0 mg/kg bw) were injected subcutaneously. For the occasional animal that experienced cerebral edema, swelling was controlled with a single subcutaneous injection of hypertonic saline (11.7%, 390.0 mg/kg bw). Using a pneumatic drill, a 3-mm diameter circular craniotomy was performed directly above the forepaw area of the primary motor cortex, centered at 0.25 mm anterior and 1.60 mm lateral to bregma (Tennant et al. 2011). A 3-mm glass coverslip was gently placed into the craniotomy, on top of the intact dura, and glued to the surrounding skull with cyanoacrylate-based glue. The remaining exposed skull was covered with dental acrylic. A titanium bar was embedded into the acrylic to secure the mouse onto the imaging stage during later in vivo imaging. Cranial window procedures were scheduled such that, following a 3-week recovery period, animals in the young adult group would begin day 0 of imaging when they were 3–4 months of age and animals in the aged group would begin day 0 of imaging when they were 20–22 months of age. In Vivo Two-Photon Imaging A custom-built two-photon microscope, a Ti:Sapphire laser (Chameleon Ultra II; Coherent Inc.) tuned to 910 nm, and a 40X 0.8 NA water immersion objective (Olympus) were used for in vivo imaging. Animals were lightly anesthetized with isoflurane (5.0% for induction, 1.0–1.8% for maintenance) throughout each imaging session, which typically lasted 30–60 min. Using ScanImage 3.8 software (Pologruto et al. 2003), the field of view beneath the cranial window was scanned, and apical dendritic tufts within layer 1 (L1) were selected for routine imaging. To verify that dendritic segments belong to a L5 PN, low-magnification image stacks (512 × 512 pixels, 0.72 μm/pixel, 5 μm z-steps) were collected, descending through the cortex, until the cell body was reached, typically at a depth of 500–700 μm from the dura. The same fragments of dendrites were imaged in subsequent sessions at high magnification (512 × 512 pixels, 0.152 μm/pixel, 1.5 μm z-steps) using a coordinate plane built around landmark vasculature and the regions of interest. High-magnification imaging of dendrites occurred at days 0, 4, 8, 30, 34, and 38, and each animal’s imaging sessions took place within a common 4-h window. Image Analysis Analysis of Dendritic Spine Density and Dynamics High-magnification images of fragments of apical dendritic tufts were analyzed using spine analysis software written in MATLAB (provided by T. O’Connor and K. Svoboda, Janelia Research Campus). All visible spines were annotated according to the following criteria: Lateral spines must project from the shaft of the dendrite at a distance equal to or greater than 1/3 the width of surrounding dendritic shaft, and spines projecting along the z-axis must be evident in at least 2 optical slices. Image files were randomly renamed (using Bulk Rename Utility, TGRMN Software) and shuffled to guarantee that analysis was done blind to age group. A total of 4516 distinct dendritic spines on 3.45 mm of dendrite from 14 cells of 8 aged animals and a total of 3781 distinct dendritic spines on 3.56 mm of dendrite from 12 cells of 8 young adult animals were followed through 6 imaging sessions over 38 days. Turnover ratio was defined as the total number of spines gained and lost (since the previous imaging session) divided by double the total number of spines present. A new persistent spine was defined as a gained spine that was still present at the next imaging session, 4 days later, and a transient spine was defined as a gained spine that was not present at the next imaging session. The survival curve was generated by fitting the survival fraction (of all spines present at day 0 that remain at each subsequent day) to a one-phase exponential decay curve. The survival fraction and rate constant are calculated as survival fraction = plateau + unstable fraction X e-t/τ, where t = time in days and τ = time constant; rate constant = 1/τ. Since animals underwent no manipulation or training throughout the experimental timeline, metrics described without specifying imaging time points are expressed as an average across all appropriate time points. For example, spine density was averaged over all 6 time points, whereas new persistent spines per micron was an average of only days 4 and 34 and turnover rate was an average of only days 4, 8, 34, and 38. Analysis of Dendritic Spine Morphology Semiautomated spine subtype classification was completed using ImageJ and a custom-written MATLAB routine, as previously described (Mostany et al. 2013; Alexander et al. 2018). Due to the limited z-resolution of 2PE, our analysis only targets lateral spines whose maximum brightness is seen in the same optical slice as the maximum brightness of the surrounding dendritic shaft. While it is possible that a fraction of the lateral spines categorized as “stubby” are actually mushroom spines with spine necks perpendicular to the imaging plane and outside of our resolution, we expect that this risk is minimized because of our spine selection criteria. This analysis was completed on every fragment of dendrite with a minimum of 20 lateral spines to survey. Using ImageJ, a line was drawn along each spine from the center of the adjacent dendritic shaft through the tip of the spine’s head, and the pixel intensity profile of the image beneath the drawn line was obtained. The MATLAB routine then calculated the first derivative function (Df) of each pixel intensity profile. A morphological classification was automatically assigned to each spine based on how many times Dfx crossed y = 0: a spine whose trace generated no crossings was categorized as stubby, 2 crossings were categorized as mushroom, and 2 or more crossings were categorized as thin. For example, the Dfx of a pixel intensity profile of a mushroom spine would cross y = 0 once at the spine neck and again at the center of the spine head. Analysis of Dendritic Spine Clustering A custom-written MATLAB routine (inspired by Yadav et al. 2012) was used to examine likely spine clustering of each dendritic fragment previously analyzed for spine density and dynamics, as described above. A spine cluster was defined as a group of 3 or more spines in which the distance between any 2 adjacent spines was not greater than an assigned clustered inter-spine distance (CISD) threshold. These putative spine clusters were identified based on proximity to neighboring spines. To detect these clusters, the X and Y coordinates at which the base of each spine contacted the shaft of the dendrite was extracted from the 2PE images (Fig. 2A). The XY coordinates were used to calculate the ``Euclidean'' distance along the dendritic shaft between adjacent dendritic spines and to build a 1-D representation of the dendritic fragment (illustrated in Fig. 2B), based on those distances. We identified the minimum inter-spine distance (MISD = 0.121 μm) from our data set (8297 spines from 129 dendritic fragments from both age groups). For each fragment, we built agglomerative (“bottom up”) hierarchical cluster trees using “Euclidean” distance and the “shortest distance” method for linkage of spines. Trees were built with CISD threshold values that ranged between the MISD and 12 μm with 0.1 μm increments (Fig. 2B). Then, for each inter-spine distance threshold of each fragment, a Monte Carlo simulation was run 5000 times in which the length and spine density of the simulation was equal to that of the observed dendritic fragment, but the positions of the spines along the dendrite were randomized. A cluster tree was built for each iteration, and the frequency distribution of the fraction of spines belonging to a cluster was calculated for all the modeled dendrites (Fig. 2C). The percentile at which the observed dendrite fell was noted as the clustering score (c-score) for that cluster threshold for that dendrite. The lowest CISD threshold at which a dendrite’s c-score equaled or surpassed 0.90 (excluding c-scores of 1.0) was the threshold chosen for further examination of spine clustering, and that dendrite was categorized as a highly clustered fragment (n = 79/129 fragments). If no cluster threshold’s c-score equaled or surpassed 0.90, that dendrite was categorized as a minimally clustered fragment (n = 50/129 fragments) and was excluded from further analysis of spine clustering. Figure 1 View largeDownload slide Increased spine density of the aged dendrite. (A) Upper, illustrated experimental timeline, cranial window (C.W.) implanted 3 weeks prior to first imaging session, imaging sessions (Img.) at 4-day intervals for 3 time points then repeated 1 month later; lower, representative images of apical dendritic segments acquired in vivo from young adult and aged animals; present spines marked with filled arrowheads, absent spines marked with unfilled arrowheads; red, lost; green, new persistent; orange, stable; yellow, transient; scale bar, 2 μm. (B) Density of spines averaged across all imaging time points. (C) Breakdown of major morphological spine types; S, stubby; T, thin; M, mushroom. (D) Density of spines across each time point; hashed lines, average from 1B; solid horizontal line, median; dot, average. n.s. P ≥ 0.05, **P < 0.01. Each data point represents 1 cell. Figure 1 View largeDownload slide Increased spine density of the aged dendrite. (A) Upper, illustrated experimental timeline, cranial window (C.W.) implanted 3 weeks prior to first imaging session, imaging sessions (Img.) at 4-day intervals for 3 time points then repeated 1 month later; lower, representative images of apical dendritic segments acquired in vivo from young adult and aged animals; present spines marked with filled arrowheads, absent spines marked with unfilled arrowheads; red, lost; green, new persistent; orange, stable; yellow, transient; scale bar, 2 μm. (B) Density of spines averaged across all imaging time points. (C) Breakdown of major morphological spine types; S, stubby; T, thin; M, mushroom. (D) Density of spines across each time point; hashed lines, average from 1B; solid horizontal line, median; dot, average. n.s. P ≥ 0.05, **P < 0.01. Each data point represents 1 cell. Analysis of Dendritic Spine Positions A custom-written MATLAB routine was used to determine whether newly formed spines that were present at the final day of imaging had occupied similar positions of spines that were present at the first day of imaging then subsequently lost. The coordinates of all spines on each dendritic fragment at day 0 and day 38 were used to build a 1-D linear representation for each time point based on the ``Euclidean'' distance between adjacent dendritic spines. For each fragment, the shift in positions of stable spines present at each time point was used to determine a maximum distance threshold within which two spines from two time points could be said to occupy the same position on the dendrite (young adult: 3.00 ± 1.00 μm vs. aged: 3.41 ± 0.37 μm). The positions of all nonstable (eventually lost) spines at day 0 were compared with all nonstable (at some point gained) spines at day 38, and any nonstable spine at day 38 that was located within the distance threshold of any nonstable spine at day 0 was categorized as possibly taking a previously occupied position. Results Dendritic Spine Density Is Elevated Across Morphological Subtypes in the Aged Motor Cortex Dendrites and dendritic spines may provide neuroanatomical explanations for the problem of aging-related cognitive impairment. To examine the potential differences in number and behavior of dendritic spines of apical tufts of L5 PNs, we routinely imaged fragments of superficial dendrites belonging to these cells in the forepaw region of M1 of young adult (3–5 months) and aged (20–23 months) thy1-GFP-M mice. To fully explore the dynamic behavior of spines, the same fragments of dendrite were imaged over the course of 38 days (Fig. 1A). Figure 2 View largeDownload slide Method for detecting spine clusters based on proximity between spines. (A) Representative image of young adult, highly clustered dendrite; pink dotted lines, annotated dendritic spines; yellow/gray line, length of fragment analyzed (38.6 μm); scale bar, 2 μm. (B) Upper, illustration of 1-D model based on representative image; lower, single representation of many cluster trees built as CISD increases; each individual tree would end at dotted line labeled at left by CISD and at right by eventual c-score; bold lines of tree indicate formation and growth of clusters; CISD, clustered inter-spine distance; c-score, clustering score; (C) Upper, observed and representations of Monte Carlo simulations ran based on cluster tree built with 2.2 μm CISD; each iteration randomizes position of spines along dendrite; lower, frequency distribution of fraction of spines belonging to clusters; hashed line, benchmark 0.90 c-score for highly clustered fragments; arrow, score of observed fragment from (A), upper (B). (D) Average optimal CISD for highly clustered fragments; each data point represents 1 dendritic fragment. (E) Fractions of all fragments categorized according to spine clustering; light color, minimally clustered; dark color, highly clustered; dark color, exploded, highly clustered (DC); DC, density-controlled. n.s. P ≥ 0.05. Figure 2 View largeDownload slide Method for detecting spine clusters based on proximity between spines. (A) Representative image of young adult, highly clustered dendrite; pink dotted lines, annotated dendritic spines; yellow/gray line, length of fragment analyzed (38.6 μm); scale bar, 2 μm. (B) Upper, illustration of 1-D model based on representative image; lower, single representation of many cluster trees built as CISD increases; each individual tree would end at dotted line labeled at left by CISD and at right by eventual c-score; bold lines of tree indicate formation and growth of clusters; CISD, clustered inter-spine distance; c-score, clustering score; (C) Upper, observed and representations of Monte Carlo simulations ran based on cluster tree built with 2.2 μm CISD; each iteration randomizes position of spines along dendrite; lower, frequency distribution of fraction of spines belonging to clusters; hashed line, benchmark 0.90 c-score for highly clustered fragments; arrow, score of observed fragment from (A), upper (B). (D) Average optimal CISD for highly clustered fragments; each data point represents 1 dendritic fragment. (E) Fractions of all fragments categorized according to spine clustering; light color, minimally clustered; dark color, highly clustered; dark color, exploded, highly clustered (DC); DC, density-controlled. n.s. P ≥ 0.05. Dendrites from aged mice showed increased spine density compared with young adult mice (young adult: 0.53 ± 0.05 spines/μm vs. aged: 0.64 ± 0.06 spines/μm; P = 0.0048, t(23) = 3.109, Welch’s t-test; Fig. 1B), as expected based on our previous work in S1 (Mostany et al. 2013). This increase in spine density was sustained and stable in the aged group throughout the entire imaging timeline (effect of age; P = 0.0058, F(1,24) = 9.152, two-way analysis of variance [ANOVA]; Fig. 1D). Since the morphology of spines is diverse and can be indicative of synaptic strength (Matsuzaki et al. 2004) and stability (Holtmaat et al. 2005), we wanted to know whether any morphological subtype of spine was responsible for this increase in spine density. To this end, we categorized each laterally projecting spine as either stubby, thin, or mushroom using a semiautomated method that classifies each lateral spine based on its pixel intensity profile. The resulting subtype proportions are in line with that reported by other groups for this cortical area using similar methods in vivo (Cui et al. 2016) and in fixed tissue taken from young adult mice (Haas et al. 2013; Li et al. 2015). We found that all 3 classes were present in equal proportion in both age groups (young adult: stubby = 30.67 ± 4.30%, thin = 21.97 ± 2.86%, mushroom = 47.36 ± 4.50% vs. aged: stubby = 36.77 ± 3.98%, thin = 22.46 ± 2.55%, mushroom = 40.77 ± 4.55%; P > 0.99, F(1,72) = −2.219e-13, two-way ANOVA; Fig. 1C). This result suggests that the increase in spine density is equitable across morphological subtypes. Spatial Clustering Does Not Differ Between Age Groups It has been hypothesized that the information processing ability of a neural network is maximized by physical clustering of coactive connections (Poirazi and Mel 2000). Later experiments showed that this optimization is accomplished in vivo through activity-dependent synapse formation and elimination that leads to greater proximity between these connections (McBride et al. 2008; Fu et al. 2012) and that increased clustering is associated with enhanced learning and memory (Frank et al. 2018). This importance of spatial arrangement led us to examine the clustering of spines of the 2 age groups after predicting likely spine clusters based on proximity and identifying highly and minimally clustered dendritic fragments. To accomplish this, each fragment of dendrite that was analyzed for spine density (Fig. 2A) is used to create a 1-D model representing the dendritic fragment analyzed (Fig. 2B, upper). Then, for each CISD, a cluster tree is built (depicted in Fig. 2B as a single, growing tree for simplicity). A Monte Carlo simulation including 5000 iterations is run based on the 1-D model, where each iteration is a unique version of the model in which dendritic length and spine density are unchanged but spine position is randomized (Fig. 2C, upper). Using all the iterations, the frequency distribution of the fraction of spines clustered in each tree is generated, and the tree produced by each CISD is assigned a c-score based on the actual, observed dendrite’s place in the frequency distribution (Fig. 2C, lower). The lowest CISD at which a c-score surpasses or equals 0.90 (excluding c-scores of 1.0) is selected as that fragment’s optimum CISD, and that fragment is categorized as highly clustered. Further analysis of spine clustering on this fragment is based on this CISD. On average, this value did not differ between age groups (young adult: 2.35 ± 0.25 μm vs. aged: 2.40 ± 0.26 μm; P = 0.7472, t(50) = 0.3214, Welch’s t-test; Fig. 2D). If this criterion is not met, the fragment is categorized as minimally clustered and is excluded from further analysis. Highly clustered fragments made up 43/66 young adult dendrites and 36/63 aged dendrites. Before computing the clustering metrics of those fragments, we discarded dendrites whose spine density was less than 0.45 spines/μm and greater than 0.70 spines/μm to control for the effect of the increase in spine density associated with aging, leaving us with 27/66 young adult and 26/63 aged dendrites (Fig. 2E). The fraction of spines from each fragment belonging to clusters (young adult: 0.87 ± 0.04 vs. aged: 0.89 ± 0.03; P = 0.3196, U = 294.5, U′ = 407.5, Mann–Whitney test; Fig. 3A), fraction of dendritic length occupied by clusters (young adult: 0.54 ± 0.07 vs. aged: 0.56 ± 0.06; P = 0.6783, U = 327, U′ = 375, Mann–Whitney test; Fig. 3B), and number of clusters per micron (young adult: 0.09 ± 0.01 clusters/μm vs. aged: 0.10 ± 0.01 clusters/μm; P = 0.7311, t(50) = 0.3455, Welch’s t-test; Fig. 3C) did not differ between age groups. The clusters themselves tend to look alike as well, with similar numbers of spines per cluster (young adult: 5.85 ± 1.06 spines/cluster vs. aged: 5.93 ± 1.12 spines/cluster; P = 0.9076, t(50) = 0.1166, Welch’s t-test; Fig. 3D) and cluster lengths (young adult: 6.86 ± 1.80 μm/cluster vs. aged: 7.08 ± 2.17 μm/cluster; P = 0.8738, t(50) = 0.1596, Welch’s t-test; Fig. 3E). Figure 3 View largeDownload slide Spine clustering appears unaffected by aging. (A) Fraction of all spines belonging to clusters. (B) Fraction of each dendrite’s length that is occupied by clusters. (C) Cluster density, number of clusters per micron of dendrite. (D) Cluster size, number of spines belonging to each cluster. (E) Cluster length in microns. n.s. P ≥ 0.05. Each data point represents 1 dendritic fragment. Figure 3 View largeDownload slide Spine clustering appears unaffected by aging. (A) Fraction of all spines belonging to clusters. (B) Fraction of each dendrite’s length that is occupied by clusters. (C) Cluster density, number of clusters per micron of dendrite. (D) Cluster size, number of spines belonging to each cluster. (E) Cluster length in microns. n.s. P ≥ 0.05. Each data point represents 1 dendritic fragment. Dendritic Spine Dynamics Are Elevated in the Aged Motor Cortex Chronic in vivo imaging of dendrites every 4 days allowed us to determine that spine formation (young adult: 0.10 ± 0.01 spines/μm vs. aged: 0.13 ± 0.02 spines/μm; P = 0.0135, t(22) = 2.683, Welch’s t-test; Fig. 4A) and loss (young adult: 0.10 ± 0.01 spines/μm vs. aged: 0.13 ± 0.02 spines/μm; P = 0.0036, t(23) = 3.24, Welch’s t-test; Fig. 4B) were elevated in aged dendrites. This increase in turnover at baseline (young adult: 0.20 ± 0.03 spines/μm vs. aged: 0.26 ± 0.03 spines/μm; P = 0.0059, t(22) = 3.043, Welch’s t-test; Fig. 4C), balanced by equal amounts of loss and formation, was expected given the stable spine density observed over time (Fig. 1D). To determine whether the increased turnover was simply an effect of the increased density, we calculated turnover ratios for each age group. The turnover ratio is the fraction of all spines present that fall into either the “gained” or “lost” category, and this metric is equal between age groups (young adult: 0.199 vs. aged: 0.205; P = 0.4624, U = 69, U′ = 99, Mann–Whitney test; Fig. 4D). At first glance, this suggested to us that the greater turnover of spines on the aged dendrite could be explained solely by the greater number of spines per length of dendrite; however, a closer look at spine formation led to a different explanation. Figure 4 View largeDownload slide Increased spine dynamics of the aged dendrite, including increased short-term spine stabilization and decreased long-term spine survival. (A) Spines gained per micron. (B) Spines lost per micron. (C) Spine turnover (lost + gained) per micron. (D) Spine turnover ratio. (E) Transient spines per micron. (F) New persistent spines per micron. (G) Fraction of all spines gained that become new persistent spines. (H) Fraction of spines gained between days 0 and 4 that persist to day 8 and survive to day 38. (I) Survival fraction of all spines present at day 0. (J) Lost persistent spines per micron. (K) Fraction of all lost spines that were persistent. (L) Survival fraction of spines present at day 0 and day 4 (Li) and of spines gained between day 0 and day 4 (Lii). (M) Absolute number of spines present at day 38 that were not present at day 0 grouped according to whether or not they occupy space that was previously occupied by a spine that was lost. (N) Fraction of nonstable spines at day 38 that occupy space previously occupied by a nonstable spine at day 0. n.s. P ≥ 0.05, *P < 0.05, **P < 0.01, ****P < 0.0001. Except for (I) and (L), each data point represents 1 cell. For (I) and (L), data points, average survival fraction per age group; solid lines, single phase exponential decay curves fit to data; dotted lines, extrapolated plateaus. Figure 4 View largeDownload slide Increased spine dynamics of the aged dendrite, including increased short-term spine stabilization and decreased long-term spine survival. (A) Spines gained per micron. (B) Spines lost per micron. (C) Spine turnover (lost + gained) per micron. (D) Spine turnover ratio. (E) Transient spines per micron. (F) New persistent spines per micron. (G) Fraction of all spines gained that become new persistent spines. (H) Fraction of spines gained between days 0 and 4 that persist to day 8 and survive to day 38. (I) Survival fraction of all spines present at day 0. (J) Lost persistent spines per micron. (K) Fraction of all lost spines that were persistent. (L) Survival fraction of spines present at day 0 and day 4 (Li) and of spines gained between day 0 and day 4 (Lii). (M) Absolute number of spines present at day 38 that were not present at day 0 grouped according to whether or not they occupy space that was previously occupied by a spine that was lost. (N) Fraction of nonstable spines at day 38 that occupy space previously occupied by a nonstable spine at day 0. n.s. P ≥ 0.05, *P < 0.05, **P < 0.01, ****P < 0.0001. Except for (I) and (L), each data point represents 1 cell. For (I) and (L), data points, average survival fraction per age group; solid lines, single phase exponential decay curves fit to data; dotted lines, extrapolated plateaus. Dendritic Spine Formation and Stabilization Is Increased While Long-Term Survival Is Decreased With Aging The formation of a new spine can fall into one of two categories: (1) transient, in which the newly formed spine is present for a single imaging session then eliminated by the following imaging session (in this case, 4 days later), and (2) new persistent, in which the newly formed spine is present for at least one subsequent imaging session. It has been suggested that new spines are either quickly shed (transient) or maintained (new persistent) based, respectively, on the absence or presence of presynaptic activity (Nikonenko et al. 2003) and misalignment or alignment of trans-synaptic proteins (Kayser et al. 2006; Mao et al. 2018) at the site of the new connection. The density of transient spines is unchanged by aging (young adult: 0.06 ± 0.01 spines/μm vs. aged: 0.06 ± 0.01 spines/μm; P = 0.9525, t(23) = 0.0602, Welch’s t-test; Fig. 4E) while the density of new persistent spines is elevated in the aged group (young adult: 0.04 ± 0.01 spines/μm vs. aged: 0.07 ± 0.01 spines/μm; P = 0.0027, t(22) = 3.373, Welch’s t-test; Fig. 4F). Because gained spines must fall into one of these two categories, this difference in new persistent spine formation suggests that the aging-related increase in spine formation (Fig. 4A) may not simply result in increased spine density, as suggested by the equal turnover ratios (Fig. 4D). The difference in spine formation seems to be a consequence of the formation and subsequent stabilization of more new persistent spines, as is supported by the greater persistent fraction of gained spines in the aged group (young adult: 0.42 vs. aged: 0.54; P = 0.0108, U = 35, U′ = 133, Mann–Whitney test; Fig. 4G). By identifying new persistent spines early in the imaging period (formed between days 0 and 4 and maintained to day 8) then measuring their survival through the final day of imaging, we observed that these new persistent spines are maintained at a greater proportion in the aged group (young adult: 0.19 vs. aged: 0.29; P = 0.0108, U = 35, U′ = 133, Mann–Whitney test; Fig. 4H), further indicating that the formation and stabilization of more new persistent spines is a signature of aging in this region of the cortex. This aging-related increase in spine formation, stabilization, and persistence led us to expect that the aged dendrite would house a more unchanging, less plastic, and longer-lasting population of spines than the young dendrite. However, the survival fraction of the total population of spines calculated from the first to the final day of imaging reveals the opposite: the aged dendrites maintain a smaller stable fraction of spines than the young adult over the long-term (plateau: young adult: 0.59 ± 0.004 vs. aged: 0.49 ± 0.004; P < 0.0001, F(1,152) = 64.09, extra sum-of-squares F-test; rate constant: young adult: 0.01648 vs. aged: 0.02273; P < 0.0001, F(1,154) = 25.98, extra sum-of-squares F-test; Fig. 4I). The survival fraction reported here for the young group is comparable to that reported by others in S1 and primary visual cortex of similarly aged animals (Grutzendler et al. 2002; Holtmaat et al. 2005). This apparent opposition between the aged dendrite’s increased stabilization of new spines and the decreased stable fraction of spines led us to expect that the aged dendrite housed more persistent spines that would be lost than the young dendrite. Spines of this condition, present for at least 8 days before being lost, were more numerous in the aged group (young adult: 0.05 ± 0.01 spines/μm vs. aged: 0.07 ± 0.02 spines/μm; P = 0.0052, t(20) = 3.135, Welch’s t-test; Fig. 4J). Furthermore, the fraction of all loss events that occur at persistent spines is also increased in the aged group (young adult: 0.46 ± 006. vs. aged: 0.55 ± 0.04; P = 0.002; U = 26, U′ = 142, Mann–Whitney test; Fig. 4K). To more precisely appreciate how these two populations of spines contributed to spine turnover, we next calculated the survival fractions of spines present at day 0 and day 4 (to approximate stable spines) and of spines absent at day 0 and present at day 4 (to isolate gained spines). This analysis was centered on these day 4 spines because it maximized the number of subsequent imaging sessions available to calculate survival fractions and because we assume that if it were possible to calculate analogous metrics for other time points, the outcome would be comparable because the steady-state condition at each timepoint is unchanged. We found that the fraction of “stable” spines maintained by the young dendrite is greater than that maintained by the aged dendrite (plateau: young adult: 0.72 ± 0.006 vs. aged: 0.57 ± 0.002; P = 0.0254, F(1,126) = 5.117, extra sum-of-squares F-test; rate constant: young adult: 0.01075 vs. aged: 0.01545; P < 0.0001, F(1,128) = 51.85, extra sum-of-squares F-test; Fig. 4Li). Conversely, the fraction of gained spines maintained by the young dendrite is lower than that of the aged dendrite (plateau: young adult: 0.21 ± 0.009 vs. aged: 0.30 ± 0.007; P < 0.0001, F(1,126) = 17.16, extra sum-of-squares F-test; rate constant: young adult: 0.05194 vs. aged: 0.03836; P = 0.0019, F(1,128) = 10.07, extra sum-of-squares F-test; Fig. 4Lii), and the young animal value is closely aligned with that reported by other groups in M1 (Yang et al. 2009). The opposing action of the survival of these two spine subpopulations on the aged dendrite led us to wonder if the increased stabilization of new spines (Fig. 4Lii) might compensate for the decreased survival of persistent spines (Fig. 4Li). To test this hypothesis, we created a 1-D model of each dendritic fragment at day 0 and at day 38. Each observation-based model included the combination of stable spines that were present at both time points and nonstable spines that were either present at day 0 and lost by day 38 or not present at day 0 and gained by day 38. We compared the positions of the two nonstable spine populations and classified each new spine as either forming in space on the dendrite that was previously occupied by a lost spine or forming in space that was previously unoccupied by any spine. New spines of aged dendrites were more likely to form in previously occupied space than those of young adult dendrites (young adult, 879 spines: previously occupied = 0.709, previously unoccupied = 0.291; aged, 1196 spines: previously occupied = 0.919, previously unoccupied = 0.081; P < 0.0001, Fisher’s exact test; Fig. 4M). This difference was also evident when we compared new spine placement on a per cell basis (young adult: 0.71 ± 0.06 vs. aged: 0.91 ± 0.04; P < 0.0001; U = 7, U′ = 161, Mann–Whitney test; Fig. 4N). While we cannot definitively state that newly formed spines on the aged dendrite are replacing previously lost connections, we do have evidence that the physical positioning of these gain and loss events is such that this phenomenon could occur. Therefore, it is possible that the increased formation and survival of spines on the aged dendrite is occurring to compensate for the decreased lifetime of previously stabilized spines. Discussion Using longitudinal, high-resolution in vivo imaging of dendritic spines on apical dendrites of L5 PNs, we examined the impact of aging on steady-state structural plasticity in M1. We found that in the absence of overt manipulation, dendrites in aged M1 house more spines per length of dendrite and that these spines tend to be more dynamic, particularly in the formation of new persistent spines. In opposition to this higher rate of short-term spine stabilization, aged dendrites endure a reduction in long-term spine survival. The increase in short-term stabilization could be a compensatory mechanism for the loss of important connections over the long term. Coincident with these results, there are no apparent differences between age groups in spine clustering when controlling for differences in spine density. These findings suggest that at baseline the circuitry in the aged cortex may be in a perpetual state of relative instability and attempted compensation, and this suboptimal connectivity may play a part in aging-related cognitive impairment. These findings provide context for similar in vivo studies that examine structural plasticity and healthy aging or aging-related diseases. Currently, the information available from in vivo study of dendritic spines in the cortex with healthy aging is limited. Experiments in S1 provide evidence of the stability of mushroom spines over the lifespan (Zuo et al. 2005), in which animals were imaged once as young adults (approximately 5 months of age) and once at old age (approximately 23 months), revealing that roughly 70% of mushroom spines present at the first imaging session are present at the second. This long-term stability seems to differ from the interpretation of our data until noting that these mushroom spines would account for just 40–50% of all spines analyzed by our criteria. It is also possible that some of these spines were lost and re-formed at some point over the 18 months between imaging sessions, a possibility supported by our data. Since mushroom spines tend to have large heads, corresponding with stronger synapses (Matsuzaki et al. 2004), it should be expected that any structural plasticity-inducing event involving these spines could ultimately settle back into a conformation similar to that preceding the circuit rearrangement. Manual skill learning has been shown to induce this type of short-lived structural plasticity in which skill retention is successful even without the persistence of spines formed in young adult M1 during practice (Clark et al. 2018). Other experiments in aged S1 report evidence for both stable spine density and dynamics (Spires-Jones et al. 2007) and increased spine density and dynamics (Mostany et al. 2013). Although, these differences may be explained by the facts that the former study followed spines (labeled by viral injection at a 1.2 mm depth) over 1 h while the latter followed spines (labeled by fluorophores endogenous to only L5 PNs at 0.5–0.7 mm depth of the cortex of the transgenic animal) over many weeks, as in this study. Along these same lines, in vivo evidence from the other side of the synapse also hints toward instability in the aged cortex. Axons in aged S1 have been reported to show increased en passant bouton (EPB) density and size (Mostany et al. 2013) as well as elevated turnover, destabilization, and fluctuation in persistent bouton size (Grillo et al. 2013). The effects of normal aging on dendrites and dendritic spines throughout the cortex are wide ranging, and the efforts to characterize them often lead to contrasting and difficult-to-reconcile results. This is likely due to regional and even cell type-specific signatures of aging, and it may be that there is no universal pattern to describe these anatomical changes across the cortex. It has been reported that dendrites of L5 PNs (Feldman and Dowd 1975) and L2/3 PNs (Wallace et al. 2007) of the rodent prefrontal cortex experience a decrease in spine density with aging. Similar decreases in spine density of L2/3 PNs in aged nonhuman primate prefrontal cortex have been observed, and, importantly, correlate with cognitive impairment (Dumitriu et al. 2010). There is less consensus in the aged nonhuman primate visual cortex where there is evidence for unchanged spine density of L2/3 PNs (Young et al. 2014) and evidence for decreased spine density of L2/3 PNs due to a loss of mushroom spines (Luebke et al. 2015). Spine loss has similarly been reported in the aged rodent visual cortex (Feldman and Dowd 1975). There is also evidence for opposite shifts in spine density in other cortical areas. Increased spine density was reported in L2/3 PNs of the aged rodent occipital cortex (Connor et al. 1980), and increased immunoreactivity to a common presynaptic marker, synaptophysin, was reported in layers 2–4 of the aged rodent entorhinal cortex (Benice et al. 2006). The results reported in this study serve to expand our understanding of the influence of normal aging across the cortex. The increased spine density reported here is a striking anatomical phenotype of aging that is difficult to interpret without further functional data, either in the form of cellular physiology or animal behavior experiments. Even so, it is well established that there is a shift in the excitation–inhibition balance in M1 associated with aging (Hortobágyi et al. 2006; Oliviero et al. 2006; Heise et al. 2013), particularly within the corticospinal pathway (Fujiyama et al. 2012; Baudry et al. 2014), and hyperexcitability has been observed to coincide with increased spine density (Chen et al. 2012; Afroz et al. 2017). Therefore, we predict that the increase in spine density reported here occurs alongside the shift toward excess excitation through the late life. The associated decrease in the signal-to-noise ratio within cortical circuits may drive the density increase through spurious initiation of activity-dependent de novo spine formation (Kwon and Sabatini 2011). It is easy to imagine that such an event would lead to a homeostatic increase in spine elimination through which the cell attempts to re-balance its volume of connections. This response would lead to a sustained increase in spine turnover, as we observed. An alternative explanation for increased density and turnover is that the aged dendrite struggles to maintain previously stabilized, active spines. Many factors determine the lifetime of a synapse, so there is no shortage of targets for aging-related influence. The Eph/Ephrin synaptic signaling components are involved at each step of synaptic formation: sampling of the local microenvironment by filopodia (Mao et al. 2018), spine stabilization (Henkemeyer et al. 2003), and spine elimination (Murai et al. 2002; Fu et al. 2006). There is also evidence that EphB2 expression, among others in the Eph/Ephrin pathway, is reduced with age in hippocampal PNs (Mohammed et al. 2016). If this pathway or others like it are somehow impaired with aging, leading to reduced spine persistence, increased spine turnover could be interpreted as the aged dendrite’s attempt to combat the neuron’s inability to maintain important connections. The formation of new persistent spines may occur to replace the apparently untenable but seemingly important connections compromised by this loss of previously stabilized spines. Our analysis of the relative positions of gained and lost spines suggests that this may be occurring. A similar phenomenon has been reported in the gain and loss of EPBs of axons within aged S1 (Grillo et al. 2013). The sites of EPB turnover tend to overlap in aged axons, suggesting the same, and ostensibly weak, synapses are subject to repeated cycles of elimination and formation. Our comparable finding of overlapping spine turnover may indicate a similar postsynaptic instability or perhaps even the postsynaptic response to presynaptic instability. Presynaptic inputs with correlated activity, indicating functional similarity, are likely to synapse in a common dendritic compartment (McBride et al. 2008; Kleindienst et al. 2011). Additionally, nonlinear dendritic integration models predict that the predominant factor determining the information processing capacity of a cell is the spatial organization of structural plasticity events (Poirazi and Mel 2000). This importance of synapse arrangement led us to explore the potential for altered spine clustering at the aged dendrite. Approaches for anatomically identifying spine clusters are widely varied. Strategies are often built around “fixed windows” of a decided distance within which spines must coexist or clustered structural plasticity events must take place. This is typically combined with a comparison between actual, observed clustering and clustering that would occur if spine arrangement were random. Our approach attempts to repurpose this general design while minimizing the influence of observer input and arbitrary criteria, being inspired by similar work (Yadav et al. 2012). The results of our method are in agreement with many physiological studies exploring molecular events at the dendrite thought to facilitate synaptic clustering, such as the spread of long-term potentiation (LTP)-dependent factors that prime nearby synapses for subsequent LTP (Harvey and Svoboda 2007; Harvey et al. 2008) or the sharing between spines of products of LTP-dependent protein synthesis (Govindarajan et al. 2011). Consideration of these spatial arrangement metrics is important for our understanding of the morphology of information processing. The state of young adult and aged dendrites at baseline is significant because the morphological signature of learning is written on top of this physical dendritic framework. In fact, it has been shown that high pre-learning spine turnover is predictive of enhanced spine clustering and performance in both contextual fear conditioning and Morris water maze models (Frank et al. 2018). While it is possible that the increased spine turnover associated with aging in the present study serves to enhance learning and memory, we do not anticipate this benefit (intended as compensation or otherwise) to endure because of the reduced long-term spine survival associated with aging. In summary, even when unchallenged by new experience or motor skill training, the circuitry within M1 seems to become less stable with aging. This aging-related instability may be met with efforts at compensation in the form of increased spine density and turnover as well as an increased short-term retention of newly formed spines. These phenotypes of aging may be a piece of the neuroanatomical basis for aging-related cognitive impairment. If that is the case, we expect that this aging-related shift in spine density and dynamics has some effect on motor learning or skill performance, but this remains to be seen. To that end, our future work will explore effects of motor learning on spine dynamics of these dendrites in aged M1. Funding National Institute on Aging at the National Institutes of Health (RO1AG047296); Louisiana Board of Regents Research Competitiveness Subprogram (LEQSF(2016-19)-RD-A-24); Center of Biomedical Research Excellence on Aging and Regenerative Medicine of the Tulane School of Medicine (P20GM103629). Notes We would like to thank Dr Rebecca Voglewede for her contribution to the analysis of spine clustering and helpful comments on the manuscript. Address correspondence to Ricardo Mostany, Tulane University School of Medicine, Department of Pharmacology, 1430 Tulane Avenue, Mail Code 8683, New Orleans, LA 70112, USA. Email: firstname.lastname@example.org. References Afroz S , Shen H , Smith SS . 2017 . α4βδ GABAA receptors reduce dendritic spine density in CA1 hippocampus and impair relearning ability of adolescent female mice: effects of a GABA agonist and a stress steroid . Neuroscience. 347 : 22 – 35 . Google Scholar Crossref Search ADS PubMed WorldCat Alexander BH , Barnes HM , Trimmer E , Da vidson A , Ogola BO , Lindsey SH , Mostany R . 2018 . Stable density and dynamics of dendritic spines of cortical neurons across the estrous cycle while expressing differential levels of sensory-evoked plasticity . Front Mol Neurosci. 11 : 83 . Google Scholar Crossref Search ADS PubMed WorldCat Anderson CT , Sheets PL , Kiritani T , Shepherd GM . 2010 . Sublayer-specific microcircuits of corticospinal and corticostriatal neurons in motor cortex . Nat Neurosci. 13 : 739 . Google Scholar Crossref Search ADS PubMed WorldCat Baudry S , Penzer F , Duchateau J . 2014 . Vision and proprioception do not influence the excitability of the corticomotoneuronal pathway during upright standing in young and elderly adults . Neuroscience. 268 : 247 – 254 . Google Scholar Crossref Search ADS PubMed WorldCat Benice TS , Rizk A , Kohama S , Pfankuch T , Raber J . 2006 . Sex-differences in age-related cognitive decline in C57BL/6J mice associated with increased brain microtubule-associated protein 2 and synaptophysin immunoreactivity . Neuroscience. 137 : 413 – 423 . Google Scholar Crossref Search ADS PubMed WorldCat Burke SN , Barnes CA . 2006 . Neural plasticity in the ageing brain . Nat Rev Neurosci. 7 : nrn1809 . Google Scholar Crossref Search ADS WorldCat Chen C-C , Lu H-C , Brumberg JC . 2012 . mGluR5 knockout mice display increased dendritic spine densities . Neurosci Lett. 524 : 65 – 68 . Google Scholar Crossref Search ADS PubMed WorldCat Clark TA , Fu M , Dunn AK , Zuo Y , Jones TA . 2018 . Preferential stabilization of newly formed dendritic spines in motor cortex during manual skill learning predicts performance gains, but not memory endurance . Neurobiol Learn Mem. 152 : 50 – 60 . Google Scholar Crossref Search ADS PubMed WorldCat Connor JR , Diamond MC , Johnson RE . 1980 . Aging and environmental influences on two types of dendritic spines in the rat occipital cortex . Exp Neurol. 70 : 371 – 379 . Google Scholar Crossref Search ADS PubMed WorldCat Cui L , Wang D , McGillis S , Kyle M , Zhao L-R . 2016 . Repairing the brain by SCF+G-CSF treatment at 6 months postexperimental stroke . ASN Neuro. 8 : 1759091416655010 . Google Scholar Crossref Search ADS PubMed WorldCat Denk W , Strickler J , Webb W . 1990 . Two-photon laser scanning fluorescence microscopy . Science. 248 : 73 – 76 . Google Scholar Crossref Search ADS PubMed WorldCat Dumitriu D , Hao J , Hara Y , Kaufmann J , Janssen WG , Lou W , Rapp PR , Morrison JH . 2010 . Selective changes in thin spine density and morphology in monkey prefrontal cortex correlate with aging-related cognitive impairment . J Neurosci. 30 : 7507 – 7515 . Google Scholar Crossref Search ADS PubMed WorldCat Feldman ML , Dowd C . 1975 . Loss of dendritic spines in aging cerebral cortex . Anat Embryol. 148 : 279 – 301 . Google Scholar Crossref Search ADS PubMed WorldCat Feng G , Mellor RH , Bernstein M , Keller-Peck C , Nguyen QT , Wallace M , Nerbonne JM , Lichtman JW , Sanes JR . 2000 . Imaging neuronal subsets in transgenic mice expressing multiple spectral variants of GFP . Neuron. 28 : 41 – 51 . Google Scholar Crossref Search ADS PubMed WorldCat Frank AC , Huang S , Zhou M , Gdalyahu A , Kastellakis G , Silva TK , Lu E , Wen X , Poirazi P , Trachtenberg JT et al. 2018 . Hotspots of dendritic spine turnover facilitate clustered spine addition and learning and memory . Nat Comm. 9 : 422 . Google Scholar Crossref Search ADS WorldCat Fujiyama H , Hinder MR , Schmidt MW , Garry MI , Summers JJ . 2012 . Age-related differences in corticospinal excitability and inhibition during coordination of upper and lower limbs . Neurobiol Aging. 33 : 1484.e1 – 1484.e14 . Google Scholar Crossref Search ADS WorldCat Fu M , Yu X , Lu J , Zuo Y . 2012 . Repetitive motor learning induces coordinated formation of clustered dendritic spines in vivo . Nature. 483 : 92 . Google Scholar Crossref Search ADS PubMed WorldCat Fu W-Y , Chen Y , Sahin M , Zhao X-S , Shi L , Bikoff JB , Lai K-O , Yung W-H , Fu AK , Greenberg ME et al. 2006 . Cdk5 regulates EphA4-mediated dendritic spine retraction through an ephexin1-dependent mechanism . Nat Neurosci. 10 : 67 – 76 . Google Scholar Crossref Search ADS PubMed WorldCat Gasparini S , Magee JC . 2006 . State-dependent dendritic computation in hippocampal CA1 pyramidal neurons . J Neurosci. 26 : 2088 – 2100 . Google Scholar Crossref Search ADS PubMed WorldCat Gasparini S , Migliore M , Magee JC . 2004 . On the initiation and propagation of dendritic spikes in CA1 pyramidal neurons . J Neurosci. 24 : 11046 – 11056 . Google Scholar Crossref Search ADS PubMed WorldCat Govindarajan A , Israely I , Huang S-Y , Tonegawa S . 2011 . The dendritic branch is the preferred integrative unit for protein synthesis-dependent LTP . Neuron. 69 : 132 – 146 . Google Scholar Crossref Search ADS PubMed WorldCat Grillo FW , Song S , Ruivo LM , Huang L , Gao G , Knott GW , Maco B , Ferretti V , Thompson D , Little GE et al. 2013 . Increased axonal bouton dynamics in the aging mouse cortex . Proc Natl Acad Sci U S A. 110 : E1514 – E1523 . Google Scholar Crossref Search ADS PubMed WorldCat Grutzendler J , Kasthuri N , Gan W-B . 2002 . Long-term dendritic spine stability in the adult cortex . Nature. 420 : 812 . Google Scholar Crossref Search ADS PubMed WorldCat Haas MA , Bell D , Slender A , Lana-Elola E , Watson-Scales S , Fisher EM , Tybulewicz VL , Guillemot F . 2013 . Alterations to dendritic spine morphology, but not dendrite patterning, of cortical projection neurons in Tc1 and Ts1Rhr mouse models of down syndrome . PLoS One. 8 :e78561. WorldCat Harvey CD , Svoboda K . 2007 . Locally dynamic synaptic learning rules in pyramidal neuron dendrites . Nature. 450 : 1195 . Google Scholar Crossref Search ADS PubMed WorldCat Harvey CD , Yasuda R , Zhong H , Svoboda K . 2008 . The spread of Ras activity triggered by activation of a single dendritic spine . Science. 321 : 136 – 140 . Google Scholar Crossref Search ADS PubMed WorldCat Hayashi-Takagi A , Yagishita S , Nakamura M , Shirai F , Wu YI , Loshbaugh AL , Kuhlman B , Hahn KM , Kasai H . 2015 . Labelling and optical erasure of synaptic memory traces in the motor cortex . Nature. 525 : 333 . Google Scholar Crossref Search ADS PubMed WorldCat Heise K-F , Zimerman M , Hoppe J , Gerloff C , Wegscheider K , Hummel FC . 2013 . The aging motor system as a model for plastic changes of GABA-mediated intracortical inhibition and their behavioral relevance . J Neurosci. 33 : 9039 – 9049 . Google Scholar Crossref Search ADS PubMed WorldCat Henkemeyer M , Itkis OS , Ngo M , Hickmott PW , Ethell IM . 2003 . Multiple EphB receptor tyrosine kinases shape dendritic spines in the hippocampus . J Cell Biol. 163 : 1313 – 1326 . Google Scholar Crossref Search ADS PubMed WorldCat Hofer SB , Mrsic-Flogel TD , Bonhoeffer T , Hübener M . 2008 . Experience leaves a lasting structural trace in cortical circuits . Nature. 457 : 313 . Google Scholar Crossref Search ADS PubMed WorldCat Holtmaat A , Bonhoeffer T , Chow DK , Chuckowree J , Paola V , Hofer SB , Hübener M , Keck T , Knott G , Lee W-CA et al. 2009 . Long-term, high-resolution imaging in the mouse neocortex through a chronic cranial window . Nat Protoc. 4 : 1128 – 1144 . Google Scholar Crossref Search ADS PubMed WorldCat Holtmaat A , Trachtenberg JT , Wilbrecht L , Sh epherd G , Zhang X , Knott GW , Svoboda K . 2005 . Transient and persistent dendritic spines in the neocortex in vivo . Neuron. 45 : 279 – 291 . Google Scholar Crossref Search ADS PubMed WorldCat Hoogendam Y , van der Lijn F , Vernooij MW , Hofman A , Niessen WJ , van der Lugt A , Ikram AM , van der Geest JN . 2014 . Older age relates to worsening of fine motor skills: a population-based study of middle-aged and elderly persons . Front Aging Neurosci. 6 : 259 . Google Scholar Crossref Search ADS PubMed WorldCat Hooks BM , Mao T , Gutnisky DA , Yamawaki N , Svoboda K , Shepherd GM . 2013 . Organization of cortical and thalamic input to pyramidal neurons in mouse motor cortex . J Neurosci. 33 : 748 – 760 . Google Scholar Crossref Search ADS PubMed WorldCat Hortobágyi T , del Olmo FM , Rothwell JC . 2006 . Age reduces cortical reciprocal inhibition in humans . Exp Brain Res. 171 : 322 – 329 . Google Scholar Crossref Search ADS PubMed WorldCat Hussein T , Yiou E , Larue J . 2013 . Age-related differences in motor coordination during simultaneous leg flexion and finger extension: influence of temporal pressure . PLoS One. 8 :e83064. WorldCat Justice JN , Carter CS , Beck HJ , Gioscia-Ryan RA , McQueen M , Enoka RM , Seals DR . 2014 . Battery of behavioral tests in mice that models age-associated changes in human motor function . Age. 36 : 583 – 595 . Google Scholar Crossref Search ADS WorldCat Kayser MS , McClelland AC , Hughes EG , Dalva MB . 2006 . Intracellular and trans-synaptic regulation of glutamatergic synaptogenesis by EphB receptors . J Neurosci. 26 : 12152 – 12164 . Google Scholar Crossref Search ADS PubMed WorldCat Kleindienst T , Winnubst J , Roth-Alpermann C , Bonhoeffer T , Lohmann C . 2011 . Activity-dependent clustering of functional synaptic inputs on developing hippocampal dendrites . Neuron. 72 : 1012 – 1024 . Google Scholar Crossref Search ADS PubMed WorldCat Kwon H-B , Sabatini BL . 2011 . Glutamate induces de novo growth of functional spines in developing cortex . Nature. 474 : 100 . Google Scholar Crossref Search ADS PubMed WorldCat Lendvai B , Stern E , Chen B , Svoboda K . 2000 . Experience-dependent plasticity of dendritic spines in the developing rat barrel cortex in vivo . Nature. 404 : 876 – 881 . Google Scholar Crossref Search ADS PubMed WorldCat LeVay S . 1973 . Synaptic patterns in the visual cortex of the cat and monkey. Electron microscopy of Golgi preparations . J Comp Neurol. 150 : 53 – 85 . Google Scholar Crossref Search ADS PubMed WorldCat Li W , Lee M-H , Henderson L , Tyagi R , Bachani M , Steiner J , Campanac E , Hoffman DA , von Geldern G , Johnson K et al. 2015 . Human endogenous retrovirus-K contributes to motor neuron disease . Sci Transl Med. 7 : 307ra153 – 307ra153 . Google Scholar Crossref Search ADS PubMed WorldCat Luebke JI , Medalla M , Amatrudo JM , Weaver CM , Crimins JL , Hunt B , Hof PR , Peters A . 2015 . Age-related changes to layer 3 pyramidal cells in the rhesus monkey visual cortex . Cereb Cortex. 25 : 1454 – 1468 . Google Scholar Crossref Search ADS PubMed WorldCat Manning H , Tremblay F . 2009 . Age differences in tactile pattern recognition at the fingertip . Somatosens Mot Res. 23 : 147 – 155 . Google Scholar Crossref Search ADS WorldCat Mao Y-T , Zhu JX , Hanamura K , Iurilli G , Datta S , Dalva MB . 2018 . Filopodia conduct target selection in cortical neurons using differences in signal kinetics of a single kinase . Neuron. 98 : 767 – 782.e8 . Google Scholar Crossref Search ADS PubMed WorldCat Matsuzaki M , Honkura N , Ellis-Davies GC , Kasai H . 2004 . Structural basis of long-term potentiation in single dendritic spines . Nature. 429 : 761 . Google Scholar Crossref Search ADS PubMed WorldCat McBride TJ , Rodriguez-Contreras A , Trinh A , Bailey R , Bello W . 2008 . Learning drives differential clustering of axodendritic contacts in the barn owl auditory system . J Neurosci. 28 : 6960 – 6973 . Google Scholar Crossref Search ADS PubMed WorldCat Mohammed C , Rhee H , Phee B , Kim K , Kim H , Lee H , Park J , Jung J , Kim J , Kim H et al. 2016 . miR-204 downregulates EphB2 in aging mouse hippocampal neurons . Aging Cell. 15 : 380 – 388 . Google Scholar Crossref Search ADS PubMed WorldCat Morrison JH , Baxter MG . 2012 . The ageing cortical synapse: hallmarks and implications for cognitive decline . Nat Rev Neurosci. 13 : 240 . Google Scholar Crossref Search ADS PubMed WorldCat Mostany R , Anstey JE , Crump KL , Maco B , Knott G , Portera-Cailliau C . 2013 . Altered synaptic dynamics during normal brain aging . J Neurosci. 33 : 4094 – 4104 . Google Scholar Crossref Search ADS PubMed WorldCat Mostany R , Portera-Cailliau C . 2008 . A craniotomy surgery procedure for chronic brain imaging . J Vis Exp . 12 : e680 . WorldCat Murai KK , Nguyen LN , Irie F , Yamaguchi Y , Pasquale EB . 2002 . Control of hippocampal dendritic spine morphology through ephrin-A3/EphA4 signaling . Nat Neurosci. 6 : 153 – 160 . Google Scholar Crossref Search ADS WorldCat Nikonenko I , Jourdain P , Muller D . 2003 . Presynaptic remodeling contributes to activity-dependent synaptogenesis . J Neurosci. 23 : 8498 – 8505 . Google Scholar Crossref Search ADS PubMed WorldCat Oliviero A , Prof ice P , Tonali PA , Pilato F , Saturno E , Dileone M , Ranieri F , Lazzaro DV . 2006 . Effects of aging on motor cortex excitability . Neurosci Res. 55 : 74 – 77 . Google Scholar Crossref Search ADS PubMed WorldCat Oray S , Majewska A , Sur M . 2004 . Dendritic spine dynamics are regulated by monocular deprivation and extracellular matrix degradation . Neuron. 44 : 1021 – 1030 . Google Scholar Crossref Search ADS PubMed WorldCat Parikh PJ , Cole KJ . 2012 . Handling objects in old age: forces and moments acting on the object . J Appl Physiol. 112 : 1095 – 1104 . Google Scholar Crossref Search ADS PubMed WorldCat Poirazi P , Mel BW . 2000 . Impact of active dendrites and structural plasticity on the memory capacity of neural tissue . Neuron. 29 : 779 – 796 . Google Scholar Crossref Search ADS WorldCat Pologruto TA , Sabatini BL , Svoboda K . 2003 . ScanImage: flexible software for operating laser scanning microscopes . Biomed Eng Online. 2 : 13 . Google Scholar Crossref Search ADS PubMed WorldCat Popescu IR , Le KQ , Palenzuela R , Voglewede R , Mostany R . 2017 . Marked bias towards spontaneous synaptic inhibition distinguishes non-adapting from adapting layer 5 pyramidal neurons in the barrel cortex . Sci Rep. 7 : 14959 . Google Scholar Crossref Search ADS PubMed WorldCat Regehr W , Kehoe J , Ascher P , Armstrong C . 1993 . Synaptically triggered action potentials in dendrites . Neuron. 11 : 145 – 151 . Google Scholar Crossref Search ADS PubMed WorldCat Seidler RD , Alberts JL , Stelmach GE . 2002 . Changes in multi-joint performance with age . Motor Control. 6 : 19 – 31 . Google Scholar Crossref Search ADS PubMed WorldCat Spires-Jones TL , Meyer-Luehmann M , Osetek JD , Jones PB , Stern EA , Bacskai BJ , Hyman BT . 2007 . Impaired spine stability underlies plaque-related spine loss in an Alzheimer’s disease mouse model . Am J Pathol. 171 : 1304 – 1311 . Google Scholar Crossref Search ADS PubMed WorldCat Stevens JC , Foulke E , Patterson MQ . 1996 . Tactile acuity, aging, and braille reading in long-term blindness . J Exp Psychol Appl. 2 : 91 . Google Scholar Crossref Search ADS WorldCat Tennant KA , Adkins DL , Donlan NA , Asay AL , Thomas N , Kleim JA , Jones TA . 2011 . The organization of the forelimb representation of the C57BL/6 mouse motor cortex as defined by intracortical microstimulation and cytoarchitecture . Cereb Cortex. 21 : 865 – 876 . Google Scholar Crossref Search ADS PubMed WorldCat Trachtenberg JT , Chen BE , Knott GW , Feng G , Sanes JR , Welker E , Svoboda K . 2002 . Long-term in vivo imaging of experience-dependent synaptic plasticity in adult cortex . Nature. 420 : 788 . Google Scholar Crossref Search ADS PubMed WorldCat Uchizono K . 1965 . Characteristics of excitatory and inhibitory synapses in the central nervous system of the cat . Nature. 207 : 207642a0 . Google Scholar Crossref Search ADS WorldCat Wallace J , Krauter E , Campbell B . 1980 . Motor and reflexive behavior in the aging rat . J Gerontol. 35 : 364 – 370 . Google Scholar Crossref Search ADS PubMed WorldCat Wallace M , Frankfurt M , Arellanos A , Inagaki T , Luine V . 2007 . Impaired recognition memory and decreased prefrontal cortex spine density in aged female rats . Ann NY Acad Sci. 1097 : 54 – 57 . Google Scholar Crossref Search ADS WorldCat Weiler N , Wood L , Yu J , Solla SA , Shepherd GM . 2008 . Top-down laminar organization of the excitatory network in motor cortex . Nat Neurosci. 11 : 360 – 366 . Google Scholar Crossref Search ADS PubMed WorldCat Xu T , Yu X , Perlik AJ , Tobin WF , Zweig JA , Tennant K , Jones T , Zuo Y . 2009 . Rapid formation and selective stabilization of synapses for enduring motor memories . Nature. 462 : 915 . Google Scholar Crossref Search ADS PubMed WorldCat Yadav A , Gao YZ , Rodriguez A , Dickstein DL , Wearne SL , Luebke JI , Hof PR , Weaver CM . 2012 . Morphologic evidence for spatially clustered spines in apical dendrites of monkey neocortical pyramidal cells . J Comp Neurol. 520 : 2888 – 2902 . Google Scholar Crossref Search ADS PubMed WorldCat Yang G , Lai C , Cichon J , Ma L , Li W , Gan W-B . 2014 . Sleep promotes branch-specific formation of dendritic spines after learning . Science. 344 : 1173 – 1178 . Google Scholar Crossref Search ADS PubMed WorldCat Yang G , Pan F , Gan W-B . 2009 . Stably maintained dendritic spines are associated with lifelong memories . Nature. 462 : 920 . Google Scholar Crossref Search ADS PubMed WorldCat Young ME , Ohm DT , Dumitriu D , Rapp PR , Morrison JH . 2014 . Differential effects of aging on dendritic spines in visual cortex and prefrontal cortex of the rhesus monkey . Neuroscience. 274 : 33 – 43 . Google Scholar Crossref Search ADS PubMed WorldCat Yuste R , Denk W . 1995 . Dendritic spines as basic functional units of neuronal integration . Nature. 375 : 682 – 684 . Google Scholar Crossref Search ADS PubMed WorldCat Zuo Y , Lin A , Chang P , Gan W-B . 2005 . Development of long-term dendritic spine stability in diverse regions of cerebral cortex . Neuron. 46 : 181 – 189 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2019. Published by Oxford University Press. All rights reserved. For permissions, please e-mail: email@example.com. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
Cerebral Cortex – Oxford University Press
Published: Jan 9, 18
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera