The capillaries in the brain play a crucial role in maintaining local brain perfusion and thus are essential for fulfilling the metabolic needs of normal brain activities. Studies on age-related changes of the capillaries in the brain have included age-related alterations of the microvascular ultrastructure (e.g., atrophy of endothelium, basement membrane thickening, and pericyte degeneration) (Burns et al., 1981; Hicks et al., 1983; Mooradian, 1988; Keuker et al., 2000; Farkas and Luiten, 2001; Alba et al., 2004; Morita et al., 2005), qualitative changes in microvascular structure such as glomerular loops and twisted capillaries in different regions of the aged brain (de Jong et al., 1990; Moody et al., 1997; Wegiel et al., 2002), and quantitative analyses of structural parameters in brain capillaries in the cortical and hippocampal regions (Bell and Ball, 1981; Pawlik et al., 1981; Mann et al., 1986; Amenta et al., 1995; Løkkegaard et al., 2001).
The interest in aged white matter has increased dramatically in recent years. Ample evidence suggests that healthy brain aging is a process affecting predominantly the white matter but not the grey matter (Pakkenberg and Gundersen, 1997; Tang et al., 1997; Piguet et al., 2009). Until now, however, only one study quantified capillary parameters in the corpus callosum of transgenic murine model of Alzheimer's disease (Lee et al., 2005). There were no studies investigating the capillary changes in healthy aged white matter using stereological techniques. In the current study, we investigated the age-related changes of the capillaries in the white matter of female Long-Evans rats by means of immunohistochemistry and stereological techniques.
MATERIALS AND METHODS
Ten young female Long-Evans rats (7-month-old) and 10 aged female Long-Evans rats (27-month-old) were obtained from the Third Military Medical University, P.R. China. The rats were housed three to four per cage at a temperature of 22°C ± 1°C. They were kept under a constant 12-hr light and 12-hr dark cycle. Food and water were available ad libitum. The colony was certified specific pathogen free for the following: mouse pneumonia virus, sendia virus, hepatitis virus, reovirus, lymphocytic choriomeningitis, Theiler Martin encephalomyelitis virus, ectromelia, minute virus of rats, and mucoplasma pulmonitis. Animal care and treatment followed the National Institute of Health Guide for the Care and Use of Laboratory Animals (NIH Publications No. 80–23) revised 1996.
The rats were deeply anaesthetized with 4% chloral hydrate (10 mL/kg) intraperitoneally and perfusion fixed with 4% paraformaldehyde in 0.6 M phosphate buffered saline (pH 7.4). After perfusion, the cerebellum, brain stem, and cranial nerves under the pavimentum cerebri were cut and the cerebral hemispheres were taken out. Each hemisphere was coronally cut into 2-mm thick slabs, starting randomly from the rostral pole. Eight to nine slabs were obtained from each hemisphere. The left or right hemisphere was sampled at random. The white matter volume was estimated according to Cavalieri's principle (Gundersen et al., 1988; Tang and Nyengaard, 1997, 2004).
The slabs of the randomly selected hemisphere were postfixed in 4% paraformaldehyde for at least 2 hr. Then, they were embedded in paraffin with the caudal surface being faced down. To get the isotropic, uniform random (IUR) sections, the embedded slabs were treated with the orientator technique (Gundersen et al., 1988; Mattfeldt et al., 1990), as illustrated in Fig. 1. After the IUR surface was obtained, the tissue blocks were sectioned at 4 μm along the direction parallel to the IUR surface. We called the cut 4-μm sections IUR sections. The orientator technique ensures that the capillaries, in each direction of three-dimensional space, have the same probability of being sampled.
Immunohistochemistry was performed using the Histostain TM-Plus SP/9001 kit from ZYMED (ZSGB; Beijing, China). Briefly, the 4-μm paraffin sections were deparaffinized in xylol and rehydrated in graded alcohol series. The sections were then immersed in citrate buffer (0.01 M, pH 6.0) and heated in a microwave oven for 15 min for antigen retrieval. After being cooled, sections were washed twice in phosphate-buffered saline (PBS, 0.01 M, pH 7.4). Endogenous peroxidase was inhibited by incubation with 3% H2O2 for 10 min and then washed in PBS three times for 5 min. Nonspecific binding sites were blocked with normal goat serum for 20 min at room temperature. Sections were incubated at 4°C overnight and then 37°C for 1 hr with rabbit polyclonal anti-collagen IV primary antibody (ab6586; Abcam, Cambridge, UK) at a dilution of 1:200 in PBS. After three 5-min washes in PBS, sections were incubated with biotinylated goat anti-rabbit IgG for 20 min at 37°C, which was followed by three additional 5-min washes in PBS. Then, the specimens were incubated with S-A/HRP for 20 min at 37°C, which was followed by repeated washes as described previously. Diaminobenzidine (DAB, ZLI-9032, ZSGB; Beijing, China) was used as a chromogen. Then, sections were dehydrated by sequential immersion in gradient ethanol and xylene and then coverslipped.
The sections were viewed using a modified Olympus BX51 microscope (Olympus, Tokyo, Japan). A DP-70 video camera mounted on the top of the microscope was connected to a computer system. Under an oil objective lens (100×), the entire white matter region on each section was photographed. Three to five fields of vision were captured from each section. For each section, vessels with luminal diameter of <10 μm were defined as components of the capillary net (Villena et al., 2003; Alba et al., 2004).
Estimation of White Matter Volume
On each slab of the randomly selected hemisphere, a transparent counting grid with an area of 0.4 mm2 associated with each point was placed at random on the caudal surface. The points hitting the white matter were counted under an optical microscope (Fig. 2). The white matter volume, V (wm), was calculated according to the Cavalieri's principle (Gundersen et al., 1988; Tang and Nyengaard, 1997, 2004):
where t equals the slab thickness, 2 mm, a(p) equals the area associated with each grid point, 0.4 mm2, and ΣP(wm) is the total number of grid points hitting the white matter per rat hemisphere.
To estimate the volume shrinkage of white matter induced by tissue processing, two tissue blocks were randomly taken from each animal. The white matter volume of each tissue block before the tissue processing, Vbefore, was estimated according to the Cavalieri's principle (Gundersen et al., 1988; Tang and Nyengaard, 1997, 2004) as described above. After a series of tissue processing (dehydrating, embedding, and sectioning), the white matter volume of each tissue block was estimated again in the way illustrated in Fig. 3. First, a 4-μm section was cut along the z-axis direction (or the direction of the slab's thickness) from the embedded tissue block and stained with hematoxylin. A sliding caliper was used to measure the slab's thickness after tissue processing, t′. Then, another 4-μm section was cut along the slab's coronal surface direction and was stained with hematoxylin and eosin. The point grid was randomly put on the stained coronal section. The points hitting white matter were counted (P′WM) under microscope using a 4× objective lens. The white matter volume of tissue block after tissue processing, Vafter, was calculated, again, according to the Cavalieri's principle (Gundersen et al., 1988; Tang and Nyengaard, 1997, 2004):
where a(p) equals the area associated with each grid point, 0.4 mm2.
The processing-induced volume shrinkage for one brain was calculated as an average of the two tissue blocks (Tang et al., 2001):
where Vbefore is the white matter volume of the selected tissue block before tissue processing, Vafter is the white matter volume of the same tissue block after tissue processing, and mean is the simple arithmetic mean over the two tissue blocks.
Estimation of the Length Density, Volume Density, and Surface Area Density of the Capillaries in White Matter
An unbiased counting frame (Gundersen, 1977) was randomly superimposed onto the captured photographs. The capillary profiles inside the counting frame or touching the top line and right line (inclusion lines) were included for counting and the capillary profiles touching the left line, bottom line and the extensions of the right line and left line (exclusion lines) were excluded for counting (Fig. 4A). The length density of the capillaries in white matter, Lv (cap/wm), was estimated as (Gundersen et al., 1988; Tang and Nyengaard, 1997, 2004; Tang et al., 1999):
where ΣQ(cap) denotes the total number of the capillary profiles counted per rat white matter, a(frame) equals the area associated with a counting frame, and Σframes is the total number of counting frames counted.
A transparent counting grid with a total of 1680 points was randomly placed on the photographs. The points hitting the capillaries, ΣP (cap), and the points hitting the white matter, ΣP (wm), were counted (Fig. 4B). The volume density of the capillaries in white matter, Vv (cap/wm), was estimated as (Gundersen et al., 1988; Tang and Nyengaard, 1997, 2004; Tang et al., 1999):
The test lines with length of 5 mm associated with each line were randomly placed on the photographs. The number of intersections between the test lines and capillary luminal surface, ΣI (cap), and the total length of test lines hitting the white matter, ΣL (wm), were recorded (Fig. 4C). The surface area density of the capillaries in white matter, Sv (cap/wm), was estimated as (Gundersen et al., 1988; Tang et al., 1999):
Estimation of the Total length, Total Volume, and Total Surface Area of the Capillaries in White Matter
Because of the volume shrinkage induced by the histological processing, the total volume of white matter was corrected for shrinkage by multiplying the white matter volume before tissue processing by (1 − volume shrinkage) (Tang et al., 2001). The total length, total volume, and total surface area of the capillaries in white matter were estimated by multiplying the length density, volume density, and the surface area density of the capillaries in white matter by the corrected white matter volume (Gundersen et al., 1988; Tang and Nyengaard, 1997, 2004; Tang et al., 1999, 2001).
Unpaired, two-tailed Student t test was used. A significant difference was considered when P < 0.05.
White Matter Volume
The volume shrinkage of white matter induced by tissue processing was 46.4 ± 8.8% in the young female group and 46.1 ± 7.3% in the aged female group. The mean white matter volume before tissue processing in the young rats (91.7 ± 12.1 mm3, mean ± SD) was significantly larger than that of aged rats (78.0 ± 16.4 mm3) (P < 0.05; Fig. 5). After corrected for tissue shrinkage, the mean white matter volume was 49.7 ± 12.6 mm3 in the young female group and 42.1 ± 9.9 mm3 in the aged female group.
Capillaries in White Matter
The length density and total length of the capillaries in the white matter of young female rats (0.7 ± 0.07 m/mm3 and 34.4 ± 7.8 m, respectively) were significantly higher than those of aged female rats (0.57 ± 0.09 m/mm3, P < 0.01 and 23.3 ± 5.1 m, P < 0.01, respectively) (Fig. 6).
The volume density and total volume of the capillaries in the white matter of young female rats (1.7 ± 0.42 × 10−2 and 0.81 ± 0.13 mm3, respectively) was significantly higher than those of aged female rats (1.3 ± 0.34 × 10−2, P < 0.05 and 0.54 ± 0.14 mm3, P < 0.01, respectively) (Fig. 7).
There was no significant difference in the surface area density of the capillaries in white matter between the two groups (11.6 ± 2.8 mm2/mm3 in the young group, and 9.3 ± 2.1 mm2/mm3 in the aged group, respectively, P > 0.05). The total surface area of the capillaries in the white matter of young female rats (5.6 ± 1.0 cm2) was significantly higher than that of aged female rats (3.8 ± 0.89 cm2) (P < 0.01) (Fig. 8).
In the past, Morita et al. (2005) detected a slight decrease in laminin immunolabeling in the basement membrane of the capillaries in the white matter of old dogs, as compared with that in the white matter of young dogs. Farkas et al. (2006) found that the number of intact microvessels in white matter decreased with age. In contrast with those qualitative studies or semiquantitative studies of capillaries, we investigated, for the first time, age-related changes of capillaries in white matter with design-based stereological technology and immunohistochemistry. The present study described efficient and unbiased methods of quantitatively investigating the capillaries of white matter. Methodologically, the methods described in the current study have several advantages over the previously used methods.
Previously, researchers selected typical parts of a brain region of interest. Farkas et al. (2006) examined the capillary density of the human white matter of 14 subjects. They did not sample the white matter uniformly but selected the subcortical white matter of the frontal, parietal, and occipital regions for investigation. The conclusions that can be drawn from the analysis of a certain typical portion of the region of interest, a standard section or sections in which the objects of interest are best identified can only apply to that part of tissue and those sections. The “ideal tissue” and “ideal fields” are not representative of the entire tissue and thus introduce a bias. A uniform random sampling of the white matter was used in this study so that all parts of the white matter were sampled equally. When an unbiased estimate of the length and surface area of capillaries is obtained from two-dimensional sections, the capillary profiles must be either from isotropic, uniform, and random sections or the capillary profiles must be isotropic themselves. In this study, isotropic, uniform random sections were ensured by the orientator technique so that all the capillaries in three-dimensional space had an equal probability of being sampled (Gundersen et al., 1988; Mattfeldt et al., 1990).
Besides the aforementioned problem associated with sampling in previously used methods, another methodological problem in previous studies was that they measured the densities of microvessels rather than the total quantities of microvessels (Moody et al., 2004; Farkas et al., 2006). Biological conclusions based on density measurements are very difficult to interpret because it will never be known if changes in density are due to an alteration of total quantity and/or an alteration in the reference volume (Braendgaard and Gundersen, 1986). In the present study, the problem associated with the density estimate has been solved by the use of a design-based stereological technique, the Cavalieri principle, to estimate the reference volume. The total length, total volume, and total surface area of the capillaries were obtained by multiplying the density estimates with the volume of white matter. Therefore, our results are the estimates of total quantities of the capillaries in white matter and can be interpreted unambiguously.
The white matter volume was estimated before the tissue processing, and the capillary density parameters were estimated from the tissue sections that were dehydrated, embedded and sectioned. If the tissue shrinkage happened during the tissue processing, the total quantity of the capillaries in white matter could not be obtained by multiplying the white matter volume before the tissue processing by the density parameters after the tissue processing. In the current study, the tissue was embedded in paraffin, which is the embedding medium that induces the largest shrinkage (Iwadare et al., 1984), and the shrinkage of brains of young individuals may not be the same as the brains of old individuals (Haug and Eggers, 1991). Therefore, the tissue shrinkage induced by tissue processing was estimated in the current study. The volume shrinkage induced by histological processing was 46.4% in young rats and 46.1% in aged rats. These changes were significant. Therefore, the white matter volume was corrected for the tissue shrinkage. In this way, we estimated the white matter volume and the capillary density under the same conditions, and thus the estimation of the total parameters of capillaries in white matter would not be affected by shrinkage.
In the present study, the total length, total volume, and total surface area of the capillaries in white matter of aged rats were all significantly lower than those of young rats, which indicated that there was significant loss of the capillaries in aged white matter. Farkas and Luiten (2001) reported that there was the ultrastructural degeneration in cerebral microvessels with aging. Moreover, it has been reported that angiogenesis was substantially impaired in aged brains (Black et al., 1989; Riddle et al., 2003). These factors might induce the age-related loss of the capillaries in white matter. It was reported that there was significantly lower cerebral blood flow in aged brains (Tachibana et al., 1984; Reich and Rusinek, 1989; Schultz et al., 1999). The results in the current study might provide the morphological basis for the decline of cerebral blood flow in aged brains.
Our research team found that there were age-related changes of white matter and the nerve fibers in the white matter (Tang et al., 1997; Li et al., 2009; Yang et al., 2009). The current study further confirmed that there was age-related reduction of white matter volume. This study also found that there was age-related decline of the capillaries in white matter. Is there any relationship between the age-related change of white matter and age-related change of the capillaries in white matter? Wender et al. (1991) investigated white matter of temporal, parietal and occipital lobes of 13 subjects using histological methods and biochemical techniques. They found that there were significantly age-related change of the expression of myelin-associated protein in both the brains with vascular changes only and brains with senile atrophy of the Alzheimer type. They thought that the degeneration of vessels might be the decisive factor in the pathogenetic mechanism of myelin lesions in the aged brains. Some researchers also found that experimental cerebral hypoperfusion had a deleterious impact on the neural tissue in the white matter, such as vacuoles and irregular myelin sheaths (Wakita et al., 2002; Farkas et al., 2004). Brown et al. (2007) investigated the vessel density in 12 subjects with leukoaraiosis (LA) and 9 age-matched normal subjects. They found that both the lesion and nonlesion areas of white matter showed lower vascular density in LA subjects when compared to normal subjects. They postulated that vascular loss precedes parenchymal cell loss. According to these studies, we speculate that the age-related loss of capillaries identified in the present study may have important implications for the degeneration of the myelinated fibers in white matter. However, the exact relationship between the morphometric parameter changes of capillaries and myelinated fiber changes in aged white matter needs to be investigated further. What are the likely functional implications of the age-related loss of capillaries in the white matter? Adequate oxygen and substrate supply are all dependent on the integrity of capillary network. Dysregulated nutrient and/or oxygen transport caused by the rarefaction of the capillary network would upset the normal functioning of the surrounding neural tissue, which might represent a precondition for the development of cognitive impairment. Therefore, we presumed that the capillary loss in white matter might be one of the reasons for cognition decline during aging.
In conclusion, the present study, for the first time, investigated the age-related changes of the capillaries in rat white matter using stereological techniques. The total length, total volume, and total surface area of the capillaries in the white matter of aged rats were significantly decreased as compared to young rats. The age-related changes of the capillaries in white matter may have important implications for the age-related white matter atrophy and cognitive impairments. An important goal of future research will be to relate these structural changes to myelinated fiber changes in white matter and brain function changes that accompany normal aging.