Cortical area and species differences in dendritic spine morphology - PDF

Please download to get full document.

View again

of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report



Views: 76 | Pages: 10

Extension: PDF | Download: 0

Related documents
Journal of Neurocytology 31, (2002) Cortical area and species differences in dendritic spine morphology RUTH BENAVIDES-PICCIONE 1, INMACULADA BALLESTEROS-YÁ ÑEZ 1, JAVIER DEFELIPE 1 and RAFAEL
Journal of Neurocytology 31, (2002) Cortical area and species differences in dendritic spine morphology RUTH BENAVIDES-PICCIONE 1, INMACULADA BALLESTEROS-YÁ ÑEZ 1, JAVIER DEFELIPE 1 and RAFAEL YUSTE 1,2 1 Instituto Cajal, Madrid, Spain 2 Department of Biological Sciences, Columbia University, New York, USA Received December 7, 2002; accepted January 22, 2003 Abstract Dendritic spines receive most excitatory inputs in the neocortex and are morphologically very diverse. Recent evidence has demonstrated linear relationships between the size and length of dendritic spines and important features of its synaptic junction and time constants for calcium compartmentalisation. Therefore, the morphologies of dendritic spines can be directly interpreted functionally. We sought to explore whether there were potential differences in spine morphologies between areas and species that could reflect potential functional differences. For this purpose, we reconstructed and measured thousands of dendritic spines from basal dendrites of layer III pyramidal neurons from mouse temporal and occipital cortex and from human temporal cortex. We find systematic differences in spine densities, spine head size and spine neck length among areas and species. Human spines are systematically larger and longer and exist at higher densities than those in mouse cortex. Also, mouse temporal spines are larger than mouse occipital spines. We do not encounter any correlations between the size of the spine head and its neck length. Our data suggests that the average synaptic input is modulated according to cortical area and differs among species. We discuss the implications of these findings for common algorithms of cortical processing. Introduction Dendritic spines were discovered by Cajal in 1888 (Ramón y Cajal, 1888), who argued that they were essential structural elements in the nervous system and served to connect axons and dendrites (Ramón y Cajal, 1899). After his early studies, there were no outstanding contributions during the following five decades, until the introduction of electron microscopy confirmed that spines indeed were postsynaptic (Gray, 1959a, b). Renewed interest in the study of pyramidal dendritic spines occurred in the early 1970s, principally as a result of observations indicating that dendritic spines abnormalities were the most consistent anatomopathological correlates of mental retardation (Marín-Padilla, 1972; Purpura, 1974). Another renaissance of spine studies has recently followed the introduction of live imaging techniques to Neuroscience and, in particular, twophoton microscopy (Denk et al., 1994). These studies have demonstrated that spines compartmentalise calcium (Yuste & Denk, 1995), are constantly moving and changing shape (Fischer et al., 1998; Bonhoeffer & Yuste, 2002) and that spine formation, plasticity and maintenance depend on synaptic activity and can be modu- lated by sensory experience (Yuste & Bonhoeffer, 2001). In spite of these recent results, the function of dendritic spines is still somewhat mysterious. Because excitatory inputs can be made on dendritic shafts (Feldman et al., 1984), spines must be serving a specific function, which could range from implementing learning rules to minimising axonal wire (Swindale, 1981; Shepherd, 1996; Yuste & Majewska, 2001). An important aspect of the dendritic spines is the enormous diversity in their morphologies, something which was already noted by Cajal and which could be important to understand their function (Ramón y Cajal, 1899). Indeed, there appears to be a clear relationship between the morphology and function of the spine, particularly with relation to the size of the spine head and the length of the neck. For example, the volume of the spine-head is directly proportional to the size of the postsynaptic density, the number of postsynaptic receptors, to the presynaptic number of docked synaptic vesicles and the ready releasable pool of neurotransmitter (Harris & Stevens, 1989; Nusser et al., 1998; Schikorski & Stevens, 1999, 2001). Also, spines with longer necks To whom correspondence should be addressed C 2003 Kluwer Academic Publishers 338 BENAVIDES-PICCIONE, BALLESTEROS-YÁÑEZ, DEFELIPE and YUSTE show longer time constants of calcium compartmentalisation than spines with shorter necks (Majewska et al., 2000a, b). Therefore, the morphology of dendritic spines has a direct functional relevance since it reveals key characteristics of synaptic inputs and their biochemical compartmentalisation. Most studies on dendritic spines of pyramidal cells have been focused on their density and distribution in a specific cortical areas and species (Elston, 2002). However, there are not systematic studies regarding possible differences in the morphology of dendritic spines between different species or different cortical areas. This appears important to us, not just to illuminate the potential function of spines, but also to highlight commonalties in the search for general rules of computation of a potentially canonical cortical microcircuit (Douglas et al., 1989). In the present study we have compared the morphology of dendritic spines of pyramidal cells from the occipital and temporal cortex of mice and humans. We report the existence of systematic differences in spine head sizes and neck lengths among these two species and among these cortical areas. We also report the lack of systematic correlations among spine neck and head, indicating that these variables are regulated independently. Methods Preparation of human and animal material In order to compare the size of dendritic spines of human and mouse pyramidal cells, we used intracellular injections in fixed cortical tissue (Elston et al., 1997). Human tissue was obtained from the left hemisphere of two male patients of 28 and 41 years of age, which was removed to gain access to the epileptic focus in the mesial temporal lobe structures during surgical treatment of epilepsy (Department of Neurosurgery, Hospital de la Princesa, Madrid, Spain). Informed consent was obtained from each patient prior to surgery. The neocortical tissue was considered to be normal on the basis of electrophysiological and histopathological examination. Surgically resected tissue was inmediately immersed in cold 4% paraformaldehyde for 24 h. Mice (n = 2, 2 months old) were overdosed by lethal i.p. injection of sodium pentobarbitone and their brain perfused intracardially with 4% paraformaldehyde, then, their brains were removed and further immersed in 4% paraformaldehyde for 24 h. We were worried that differences in fixation method (perfusion vs. immersion) could account for the morphological differences that we report. To test this we fixed a mouse hemisphere by immersion, following exactly the same protocol used with the human tissue. Spines in the immersion-fixed temporal mouse material were indistinguishable in density and size from those in the perfusion-fixed temporal mouse material (head area 0.35 ± 0.01 µ 2, mean ± SEM, n = 376, for immersion vs. x = 0.37 ± 0.01, n = 1306 for perfusion; Bonferroni p = 1; average neck length 0.64 ± 0.02, n = 231 for immersion vs ± 0.01, n = 759 for perfusion; Bonferroni p = 0.19; average spine density 12.5 ± 0.96 spines/10 µm, n = 10 for immersion vs ± 0.46, n = 40 for perfusion; Bonferroni p = 0.68). Intracellular injections For both species, the cerebral cortex was cut tangentially to the cortical surface with the aid of a Vibratome. Our cell injection methodology has been described in detail elsewhere (Elston et al., 1997, 2001). Briefly, cells in the flat portion of the occipital and temporal cortex of mice (approximately corresponding to areas V1M/V1B and A1/S2 of Franklin and Paxinos (Franklin & Paxinos, 1997) and third temporal gyrus (Broadman s area 20) of the human cases) were individually injected with Lucifer Yellow by continuous current. Following injections, the sections were processed with an antibody to Lucifer Yellow as described in Elston et al. (2001) (Fig. 1). Reconstruction and analysis Only neurons whose basal dendritic tree was completely filled were included in this analysis. To preserve a high signal to noise in our analysis we only reconstructed lateral spines, neglecting spines located on the top or bottom surface of the dendrites. To avoid potential differences among neuronal classes and dendritic branches and create a homogeneous sample, we only reconstructed spines from basal dendrites of layer 3 pyramidal neurons. Because spine density (Ruiz-Marcos & Valverde, 1969; Elston & DeFelipe, 2002), and possible also spine size (Konur & Yuste, unpublished observations), changes as a function of distance from the soma, we sought to compare similar segments of dendrites between different cells, by selecting segments of basal dendrites which were located at the same proportional distance from the soma. More specifically, we selected the basal dendrites segments which, according to our previous work (Elston & DeFelipe, 2002) has the highest density of spines. To perform the morphometric analysis of dendritic spines, we studied the same proportional segment of 20 randomly-selected horizontally projecting pyramidal cell basal dendrites of different cells in each area and case. Only one dendrite per cell was analysed. The proportional segments initiated at 45 µm from the soma in mice and 75 µm in humans. These dendritic segments (30 µm long in mice and 50 µm in humans) correspond to the highest density of spines in these two species (Elston et al., 2001; Dierssen et al., 2003 and unpublished obsevations). Images of each portion of dendrite were captured at different focal planes using a BX51 Olympus microscope (100x objective) attached to a Nikon 995 camera at a final magnification of 3100x. Thereafter, images were used to make composite projection drawings of the dendritic spines. Only spines arising form the lateral surfaces of the dendrites were included in the study. The analysis was carried out blindly by a different investigator. Spine density was measured by counting the number of spines located in the lateral portion of each dendrite segment. Therefore, the results obtained were presumably an underestimate of the total number of spines present on the mentioned portion of dendrite. The area, major and minor axis of the head of spines and the length of the necks in each portion of the dendrites analysed were determined with the aid of a digitizing tablet (SummaSketch III) and NIH image software (NIH Research Services, Bethesda, MD). Spine necks were measured from the point of attachment of the dendrite to the beginning of the spine head, as estimated by the investigator. Quantitative analysis of spine morphologies 339 Fig. 1. Photomicrographs of pyramidal cells in the human and mouse neocortex. A, B. Low-power photomicrographs of layer III pyramidal cells injected with Lucifer Yellow and processed with DAB in human (A) and mouse (B) temporal cortex. Note the smaller size of mouse cells. Section is parallel to the cortical surface. C, D: Photomicrograph of horizontally projecting dendrites of a human (C) and mouse (D) pyramidal cell. E, F: High-power photomicrographs of the basal dendrite segments of human (E) and mouse (F) pyramidal cells illustrating individual dendritic spines. Note the smaller size of the mouse spines. Scale bar: 425 µm ina,b;45µm inc,d;10µm ine,f. Results SPECIES DIFFERENCES IN SPINE DENSITY In order to explore whether systemic differences in spine morphologies or densities exist among species or among cortical areas, we reconstructed and measured spines from human temporal cortex (n = 2768 spines, 40 cells, 2 patients) and mouse temporal (n = 1306 spines, 40 cells, 2 animals) and occipital (n = 1226 spines, 40 cells, 2 animals) cortex. Spines were labelled using intracellular Lucifer Yellow injections in fixed material and immunocytochemistry (Elston et al., 1997, 2001) (Fig. 1) and they were reconstructed by tracing high magnification digital microphotographs. We first wondered if there were differences in spine densities between species or areas. Quantification of 340 BENAVIDES-PICCIONE, BALLESTEROS-YÁÑEZ, DEFELIPE and YUSTE Table 1. Morphometric values (mean ± sem ) of dendritic spines from layer III pyramidal cells of the occipital and temporal cortex of mice and human. Mouse Mouse Human occipital temporal temporal cortex cortex cortex Spine density ± ± ± 0.43 (per 10 µm) (n = 40 cells) (n = 40 cells) (n = 40 cells) Area of the 0.31 ± ± ± 0.01 head (µm 2 ) (n = 1226) (n = 1306) (n = 2768) Major axis 0.77 ± ± ± 0.01 (µm) (n = 1226) (n = 1306) (n = 2768) Minor axis 0.48 ± ± ± 0.01 (µm) (n = 1226) (n = 1306) (n = 2768) Length of the 0.67 ± ± ± 0.01 neck (µm) (n = 1226) (n = 1306) (n = 2768) the spine density of the corresponding segment of basal dendrites in mouse and human revealed that the mean number (mean ± SEM; for all measurements) of spines per 10 µm segment was 10.2 ± 0.5 and 10.9 ± 0.5 for cells in occipital and temporal cortex of mice, respectively and 14.2 ± 0.4 for the temporal cortex of humans (Table 1). Statistical analysis (one-way ANOVA) revealed differences to be significant between mice and humans, but not between the two areas of the mouse analysed (Table 2; see also Fig. 1E, F). To explore whether there were different classes of neurons with systematic differences in spine density we plotted the individual data for our three samples, finding that different cells covered a continuum of spine densities, although some outliers were evident in the mouse and human temporal data (Fig. 2A). We concluded that human temporal cortical neurons have on average a higher ( 30%) spine density (in the basal dendritic segment with the highest density of spines) than mouse temporal or occipital neurons (see also Elston & DeFelipe, 2002). SPECIES AND AREA DIFFERENCES IN SPINE HEAD AREA We then analyzed whether there were any systematic differences in the size of the spine head. As explained, the volume of the spine is linearly correlated with a variety of pre- and postsynaptic physiological parameters (Harris & Stevens, 1989; Nusser et al., 1998; Schikorski & Stevens, 1999, 2001). As an approximation to the estimation of the spine volume, we measured the maximal cross-sectional area of the spine head, as determined from reconstructions of several images taken at different focal points. The study of the size of spine heads revealed that the mean area in the temporal cortex of mice was smaller than that in humans (mean ± sem: 0.37 ± 0.01 µm 2 and 0.59 ± 0.01 µm 2, respectively; Table 1). Statistical analysis showed the difference to be significant (Table 2; Fig. 1 E, F). Moreover, the area of the head of spines in the occipital cortex of mice was significantly smaller (0.31 ± 0.01 µm 2 ) than that of mouse temporal cortex. To explore whether these differences were due to differences in spine heterogeneity, we plotted histograms of the relative frequency of spines in the three samples, as a function of head area (Fig. 2B). For each of the three populations of spines, these histograms showed a unimodal α-type function, without any clear indications of a separate classes of spines with different head sizes. The peak of this function was clearly shifted in the three samples, with the mouse occipital spines distribution peaking around 0.2 µm 2, the mouse temporal around 0.3 µm 2 and the human temporal around 0.4 µm 2. We concluded that the distribution of heterogeneous spine head areas is systematically different between human and mouse temporal cortex and even between mouse temporal and occipital cortices. SPECIES DIFFERENCES IN SPINE NECK LENGTH We then searched for potential differences in the length of the spine neck. While the diameter of the spine neck appears relatively constant among spines (Harris & Table 2. Statistical comparisons among dendritic spines from layer III pyramidal cells of occipital and temporal cortex of human and mouse. Mouse occipital Mouse occipital Mouse temporal Mouse temporal Human temporal Human temporal Spine density (per 10 µm) * * one-way ANOVA, F (3.129) = 14.0, p Area of the head (µm 2 ) * * * one-way ANOVA, F (3.5675) = 343.7, p Length of the neck (µm) * * one-way ANOVA, F (3.3489) = 58.9, p Post-hoc Bonferroni analysis, p Quantitative analysis of spine morphologies 341 Fig. 2. Species and area differences in spine densities and morphologies. Distribution of spine density (A), area of spine heads (B) and length of necks (C) for the occipital and temporal cortex of the mouse and the human. *includes stubby spines and spines whose neck were not distinguishable from their head. 342 BENAVIDES-PICCIONE, BALLESTEROS-YÁÑEZ, DEFELIPE and YUSTE Stevens, 1989), there is a large variability in its length (Ramón y Cajal, 1899). Moreover, it is likely that the length of the spine neck controls the time constant of calcium compartmentalization, which could be the major function of the spine (Majewska et al., 2000a, b). Specifically, the diffusion of calcium out of the spine can be substantial in stubby spines with short necks but is prolonged or perhaps even non-existent in spines with long necks (Majewska et al., 2000a; Sabatini et al., 2002). We found that temporal cortex of mice showed shorter neck lengths (0.73 ± 0.01 µm) than that of humans (0.94 ± 0.01 µm). A one-way ANOVA demonstrated that these differences were significant between the two groups (see Tables 1 and 2). However, this was not the case when comparing the values from the occipital cortex of mice (0.67 ± 0.01 µm) and those from the temporal region. There was a wide distribution of lengths of necks in both mice and human (Fig. 2C). These distributions were clearly bimodal for our three samples, with a peak at zero and a second peak around 0.5 µm. Indeed, almost half of the spines analyzed are represented as no neck. This bin included stubby spines and also those spines whose head was not distinguishable from the neck. This bimodal distribution suggest that there are two populations of spines: a population with no neck, and a population with substantial necks. The differences between human and mouse spines that we encountered appeared due both to a reduction of the no neck spines in humans, as well as the systematic shift in the rest of the human spines towards longer neck lengths. CORRELATIONS AMONG SPINE DENSITY, HEAD AREA AND NECK LENGTH We finally studied the possible correlation of spine density, head area and neck length (Fig. 3). This appeared important to us in order to uncover potential causal links between the regulation of these three variables. For every pair of variables in the three samples, we plotted the average for each cell and fitted linear regression equations to the data. We found that none of the correlations (head area vs. neck length, length vs. density and area vs. density in the three samples) were statistically significant from the null hypothesis of zero slope in the linear fits. Moreover, the correlation coefficients were low, all below 0.34 and most hovering around zero. Within each pair of variables, no systematic trends was observed, with the exception of the negative correlation found between head area and spine density in the three samples of cells. Albeit not significant (p = 0.15), we cannot rule out that a larger sample could uncover a potential relation between the size of the spine head and spine density, whereby larger spines are spaced farther away from each other than smaller spines. Discussion METHODOLOGICAL CONSIDERATIONS The morphological diversity of dendritic spines has baffled neuroscientist since their first description by Cajal (Ramón y Cajal, 1888). Spines must be playing a fundamental role in the nervous system, and it is likely that their morphological diversity reflects the diversity of this elusive function of the spine. In this work we are
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks