The renal nerves contain efferent sympathetic nerve fibers that are important in regulating renal function, and afferent fibers that transmit sensory information from the kidneys to the central nervous system. Changes in multiple sensory inputs and central nervous system activity influence efferent renal sympathetic nerve activity, thereby altering kidney function through changes in renal vascular resistance, renin release, and tubular sodium reabsorption. The innervation within mammalian kidneys (intrinsic innervation) has been extensively described in the literature, particularly for rats (Barajas and Muller, 1973; Barajas, 1978; Barajas et al., 1984, 1985, 1992; Barajas and Powers, 1988, 1989). In contrast, few studies have provided a detailed description of the morphology of the extrinsic renal nerves leading to the kidney. The ex trinsic renal nerves are of interest because they are the site at which sympathetic nerve activity and electrical stimulation are recorded in functional studies. Considerable attention has been given to the extrinsic renal nerves of rats because of the physiological regulation of efferent renal sympathetic nerve activity by the arterial baroreceptor (Judy et al., 1976, 1979; Judy and Farrel, 1979; Kidd et al., 1981), cardiopulmonary receptor (Kidd et al., 1981), arterial chemoreceptor (Kidd et al., 1981; Dorward et al., 1987), somatic receptor (Dorward et al., 1987), and renal mechanoreceptor (Rogenes, 1982; Kopp et al., 1987) reflexes. Furthermore, increased sympathetic nerve activity in the kidneys contributes importantly to sodium and water retention in diseases such as heart failure and cirrhosis, and in people who consume a high-salt diet (DiBona and Kopp, 1997).
Although the electrophysiological characteristics of both afferent and efferent extrinsic nerves innervating the kidney have been described (Ricksten and Thorén, 1981; Lundin et al., 1984; Janssen et al., 1989), knowledge of the morphological characteristics of these nerves is limited (DiBona et al., 1996; DiBona and Kopp, 1997). The ability to genetically manipulate mice has led to the increasing use of this species in physiological studies. Knowledge of the normal physiology and morphology of mice is therefore required to enable interpretation of future studies in genetically manipulated mice. The objectives of the present study were to describe the general morphology of the extrinsic renal nerve in mice and to determine its morphometric parameters, including the number and size of myelinated and unmyelinated fibers contained therein.
MATERIALS AND METHODS
Experiments were performed on five male C57BL/6J mice (20–25 g). The animals were anesthetized with sodium pentobarbital (Nembutal, Abbott Laboratories, Illinois; 50 mg/Kg, i.p.) and the left extrinsic renal nerve was exposed through a left flank incision using a retroperitoneal approach. With the use of a dissecting microscope, the renal nerve branch from the aorticorenal ganglion was isolated and carefully dissected free. Electrophysiological recordings of renal nerve activity were made to confirm the presence of characteristic renal sympathetic nerve activity before the morphological studies were carried out (Ma et al., 2001).
The extrinsic renal nerves were fixed in situ (externally applied solution) for 10 min in 2.5% glutaraldehyde (Polysciences Inc., Warrington, PA). The nerves were then removed and fixed by immersion for 24–48 hr in the same fixative solution. After they were washed in 0.1 M cacodylate buffer, the nerves were post-fixed in 1% OsO4 solution for 2 hr at 4°C, dehydrated with graded ethanol, and embedded in epoxy resin (Polibed 812, Polysciences Inc.). The nerves were oriented to permit semi-thin (0.5–1.0 μm thick) transverse sections of the fascicles, which were stained with 1% toluidine blue and observed under the oil immersion lens of a Leitz diaplan photomicroscope. The images were transmitted via a camera to an IBM PC where the images were digitized. With the use of image analysis software (KS 400, Kontron 2.0, Eching Bei Munchen, Germany), the total number of myelinated fibers was counted and the area and diameter of each nerve fascicle were measured.
For the transmission electron microscopic studies, thin transverse sections (500–600 nm) were mounted on 2 × 1 slot grids covered with formvar 0.5% solution, stained with lead citrate and uranyl acetate, and observed under the transmission electron microscope (Hitachi H7000, Pleasaton, CA). One electron micrograph of the whole fascicle (1,500–2,500 times original magnification) and at least three electron micrographs at 5,000 times original magnification were obtained for each nerve without overlap of the microscopic field. The negatives were then scanned using the Imapro QCS 3200 flatbed scanner (Imapro, Ottawa, Canada), 3200 DPI, and stored on a Power Macintosh for later analysis. Using the image analysis software (KS 400, Kontron 2.0), the unmyelinated fibers were counted and the area and diameter of each fiber were measured. In the case of myelinated fibers, both the axonal (defined by the outer limit of the axolema) and fiber (defined by the outer limit of the myelin lamellae) diameters were measured. Thus, the ratio between the two diameters (G-ratio, a measure of degree of myelination) was obtained (Rushton, 1951; Smith and Koles, 1970). The density of myelinated and unmyelinated fibers was determined and the ratio between myelinated and unmyelinated fibers was calculated. The number of unmyelinated axons associated with each Schwann cell or portion thereof (hereafter called a Schwann cell unit) was determined for all renal nerves studied using the basal lamina to outline the limits of a given Schwann cell unit. The term “Schwann cell body” refers to the portion containing the nucleus (McDonald, 1983a). The number of Schwann cell nuclei present in each transverse section of the renal nerves was counted and their density calculated. A histogram of the frequency distribution of the unmyelinated fibers was constructed and separated into class intervals increasing by 0.1 μm. Data are presented as mean ± standard error of the mean (S.E.M.).
The renal nerves consisted of a single fascicle that was delimited by one or two closely apposed layers of flattened cells, with a prominent basal lamina, between which lay longitudinally oriented collagen fibers that constituted the perineurium (Fig. 1). The endoneurium consisted mainly of longitudinally oriented collagen fibers that occupied much of the space between the myelinated and unmyelinated axons (Figs. 1 and 2). None of the renal nerves presented capillary vessels within the endoneurium.
The renal nerve fascicular area averaged 741 ± 104 μm (range: 475–950 μm) and the fascicular diameter averaged 35 ± 4 μm (range: 27–49 μm). The renal nerves contained an average of 834 ± 170 axons, of which 99.5% were unmyelinated and 0.5% were myelinated (Table 1). Unmyelinated fibers outnumbered myelinated fibers by a factor of 169 to 1 on average. The larger-diameter myelinated fibers occupied 1% of the total fascicular area, on average. Most of the myelinated fibers in the renal nerves were small (total fiber diameter average = 3.1 ± 0.9 μm (Table 1)). Myelinated fibers had axons that averaged 2.2 ± 0.3 μm in diameter and constituted 42% of the total fiber diameter. The ratio of axonal diameter to total fiber diameter (G-ratio) ranged from ∼0.4 to 0.8 (mean value of 0.7 ± 0.01) and tended to increase with increasing axon size (Table 1). The mean diameter of the unmyelinated axons averaged 0.8 ± 0.02 μm, with a unimodal distribution size (Fig. 3). Unmyelinated axons as small as 0.2 μm and as large as 1.8 μm were present in all nerves studied. About 90% of the unmyelinated axons were less than 1.0 μm in diameter. There was considerable overlap in the diameter of myelinated and unmyelinated axons. The larger unmyelinated axons (1.8 μm in diameter) were larger than 52% of the myelinated axons.
Table 1. Morphometric parameters of the extrinsic renal nerves of C57BL/6J mice
Area values are expressed in μm2 and diameter values are expressed in μm. Data are expressed as Mean ± SEM.
741 ± 104
35 ± 4
834 ± 171
4.6 ± 1.7
7.4 ± 1.8
3.1 ± 0.4
0.7 ± 0.01
4.3 ± 0.92
3.1 ± 0.9
2.2 ± 0.3
830 ± 169
0.4 ± 0.02
0.76 ± 0.02
An average of 74 ± 14 Schwann cell units was found within each nerve cross-section. Each Schwann cell enveloped anywhere from one to 20+ axons (mean = 11 ± 1 axons per unit) and no Schwann cell was devoid of axons (Figs. 1 and 2). An average of 11% ± 3% of unmyelinated axons were not accompanied by other axons in their Schwann cell envelopment. Axons of all sizes were represented in this category, but the larger ones predominated. When two or more axons were enveloped by the same Schwann cell, the axons were not in contact with one another. Instead, each axon was located in a separate Schwann cell trough (Figs. 1 and 2). An average of 27 ± 2% of axons were enveloped in large groups of 11–20 axons per unit, and 18% ± 1% of axons were enveloped in groups of 21 or more axons. Smaller proportions of axons were enveloped in smaller groups. Axons of different sizes were often enveloped by the same Schwann cell (Figs. 1 and 2). Within each transverse section of the nerves there were an average of 3 ± 1 Schwann cell nuclei.
The morphology of the extrinsic renal nerves in mice exhibited similarities as well as differences in comparison with renal nerves in other species, such as rat. Each of the mouse renal nerves examined in the present study demonstrated a defined perineurial envelopment that consisted of one or two layers of cells. The lamellar arrangement of flattened cells separated by layers of collagenous connective tissue observed in the renal nerves is the essential structural feature of the perineurium of peripheral mammalian nerves (Thomas, 1963). The number of lamellae varies, depending mainly on the diameter of the fascicle: the larger the fascicle, the greater the number of lamellae (Thomas, 1963). The spinal nerves of rats are often each covered by four layers of lamellae, although very small nerves can be found that are covered by a single layer of lamellae (Gamble, 1964), similar to what we have observed for the renal nerves of mice. In all cases, the basal lamina covered the outer and inner surfaces of the perineurial cells, as described previously for rat nerves (Thomas, 1963; Gamble, 1964). Similar to the renal nerves of mice (see Fig. 2), collagen fibers are larger in the epineurium than in the endoneurium in the peripheral nerves of rats (Gamble, 1964; McDonald, 1983b). Information on the number of lamellae of perineurial cells in a normal nerve is useful for the correct interpretation of perineurium morphological alterations in pathologic conditions in which the perineurium may be thickened.
Comparison of our results with those of DiBona et al. (1996) suggests some differences between mice and rats in the relative percentage of myelinated vs. unmyelinated nerve fibers, and in the diameters of unmyelinated fibers. We found an unmyelinated : myelinated fiber ratio of 99.5:0.5, indicating a relatively smaller percentage of myelinated fibers in mice compared with rats. DiBona et al. (1996) reported that the extrinsic renal nerves of rats contain an unmyelinated : myelinated fiber ratio of 96:4. Myelinated fibers in the kidney are generally considered to be sensory afferents (DiBona and Kopp, 1997) suggesting that this category of afferent may be less prevalent in mice than in rats. A limitation of the present study is the inability to distinguish between afferent and efferent unmyelinated fibers. Although the majority of unmyelinated fibers are sympathetic efferents, the afferent innervation is also primarily unmyelinated (Barajas and Muller, 1973; Barajas and Powers, 1988; Barajas et al., 1984, 1985; DiBona and Kopp, 1997). In addition, the diameter of unmyelinated renal nerves was smaller in mice (0.8 ± 0.02 μm) than in rats (1.26 ± 0.01 μm) (DiBona et al., 1996). The diameter of myelinated renal nerve fibers was similar in rats (3.14 ± 0.02 μm) (DiBona et al., 1996) and mice (3.1 ± 0.4 μm) (present study), suggesting that the difference in unmyelinated fiber diameter between species was not the result of differences in the methods of nerve preparation and analysis in the two studies.
Another difference between the sympathetic nerve fibers of mice and rats relates to the size distribution of the unmyelinated fibers. The size distribution histogram for the unmyelinated fibers of mice is unimodal (see Fig. 3), in contrast to the bimodal distribution in rats (DiBona et al., 1996). DiBona et al. (1996) suggested that the bimodal distribution of fiber size in rats may reflect two subgroups of renal sympathetic nerve fibers with possible different functional roles. The physiological significance of the unimodal size distribution of C57/BL mouse unmyelinated fibers remains to be determined. It is important to realize that the morphology of the fibers is not the only parameter one should consider in relation to function. Immunolocalization of growth factors in mice kidneys has been reported (Salido et al., 1986; Barajas et al., 1987). These and other techniques may enable investigation of possible differences in neurotransmitters and other regulatory molecules in putative subtypes of renal nerve fibers.
Despite some differences in renal nerve morphology between mice and rats, the regulation of efferent renal sympathetic nerve activity and the effects of changes in sympathetic activity on renal function appear to be at least qualitatively similar in mice and rats. For example, changes in arterial blood pressure evoke powerful baroreflex-mediated changes in renal sympathetic nerve activity in mice, as has also been observed in rats and other species (Ling et al., 1998; Ma et al., 2001).
The quantitative morphometric analysis of peripheral nerves in different species has implications for functional studies of the regulation of sympathetic nerve activity. Such studies commonly involve direct recording of neural activity in fibers traveling in the extrinsic renal nerves (Judy et al., 1976, 1979; Judy and Farrel, 1979; Ricksten and Thorén, 1981; Lundin et al., 1984; DiBona and Kopp, 1997; Ling et al. 1998; Ma et al., 2001). The absolute level of activity recorded is dependent not only on the activity in single nerve fibers but also on the number of nerve fibers near the recording electrode and the recording conditions. Ling et al. (1998) recently reported that the absolute levels of resting and maximal renal sympathetic nerve activity, and the range of baroreflex control of sympathetic activity are less in mice than in rats. The authors acknowledged that the lack of information on the number of fibers and fiber diameter in the renal nerves of mice prevented evaluation of the underlying cause of the lower nerve activity in mice.
The absolute level of sympathetic nerve activity is considered to be increased in certain pathologic states and in the elderly. Increased sympathetic activity occurs with aging (Hajduczok et al., 1991a, b; Davy et al., 1998), despite evidence suggesting that the number of sympathetic nerve fibers decreases with age (Low et al., 1977). Clearly, knowledge of the number of nerve fibers within the sympathetic nerves should enable a more robust interpretation of the results of nerve recording experiments.
The results of the present study provide baseline data for future studies in genetically modified mice. The use of mouse models to advance knowledge of physiology, pathology, and development is exploding in all areas, including studies of the kidney (Moore et al., 1996; Sanchez et al., 1996; Sweetser et al., 1999). The addition of morphometric analysis of the extrinsic renal nerves to functional studies in mouse models should provide a powerful approach to investigate the structural basis of altered autonomic reflexes, genes affecting the development and structure of the sympathetic and sensory innervation of the kidney, and the influence of renal nerves on kidney structure and function.
Valéria Paula Sassoli Fazan was the recipient of fellowships from the Fulbright Foundation and CAPES, Brasilia, Brazil, during the time the study was carried out. The authors thank the Central Microscopy Research Facility, University of Iowa, Iowa City, Iowa, for excellent technical support.