|Home | About | Journals | Submit | Contact Us | Français|
While some reports in humans have shown that nephron number is positively correlated with height, body weight or kidney weight, other studies have not reproduced these findings. To understand the impact of genetic and environmental variation on these relationships, we examined whether nephron number correlates with body weight, kidney planar surface area, or kidney weight in two inbred mouse strains with contrasting kidney sizes but no overt renal pathology: C3H/HeJ and C57BL/6J. C3H/HeJ mice had smaller kidneys at birth and larger kidneys by adulthood, however there was no significant difference in nephron number between the two strains. We did observe a correlation between kidney size and body weight at birth and at adulthood for both strains. However, there was no relationship between nephron number and body weight or between nephron number and kidney size. From other studies, it appears that a greater than two-fold variation is required in each of these parameters in order to demonstrate these relationships, suggesting they are highly dependent on scale. Our results are therefore not surprising since there was a less than two-fold variation in each of the parameters examined. In summary, the relationship between nephron number and body or kidney size is most likely to be demonstrated when there is greater phenotypic variation either from genetic and/or environmental factors.
The functional unit of the kidney is the nephron: a structure that contains vascular loops of the glomerulus at the site of blood filtration and a tubular segment that reabsorbs and excretes solutes and connects to the collecting duct system. Based on autopsy specimens from individuals representing various ethnic groups, a large variation in nephron number exists in the “normal” adult human kidney, such that each kidney contains anywhere from 200,000 to over 1.8 million nephrons.1–3 Recent interest in nephron number has aimed to understand whether individuals with a congenital defect in nephron formation are at an increased risk of developing diseases like hypertension and chronic renal failure.
Nephrogenesis ends by 36 weeks of gestation in humans, such that the final number of nephrons in each kidney is established at birth.4 Infants below the 10th percentile in birth weight have smaller kidneys with fewer nephrons, suggesting that there is a relationship between body weight and kidney size, as well as between body weight and nephron number.2,5–7 In addition to its relationship with birth weight, nephron number has been shown to correlate with other factors, including adult height and body surface area,6 kidney weight or volume3 and glomerular volume.2 A decrease in nephron number has also been found to be associated with susceptibility to developing diseases like hypertension and chronic renal failure.1,2,6,8–12
While the aforementioned studies suggest that there is a relationship between nephron number and kidney or body size and between nephron number and susceptibility to hypertension or chronic renal failure, other studies have not confirmed these relationships.1,9,13 The latter results suggest other possibilities, including that the parameters are independent variables, or alternatively, that they are interdependent, but the relationship is obscured due to excessive genetic and environmental variation.
We examined the relationship between nephron number and kidney or body size in inbred mouse strains. Inbred mice are genetically homogeneous and are exposed to uniform environmental conditions, making them an ideal model to study these relationships.14 We hypothesized that the heritable differences in kidney size observed between inbred strains would be accompanied by a difference in nephron number.15 We identified two inbred mouse strains, C3H/HeJ and C57BL/6J, that differed significantly in their kidney size, measured as planar surface area and as weight. Nephron number counts and body weights were therefore obtained in the newborn period and at adulthood to determine if there is a relationship between nephron number, body weight, or kidney size in these mouse strains.
Kidney weight varies between inbred mouse strains and has been previously demonstrated to be heritable.15 We therefore assessed the variation in kidney size in two inbred mouse strains, C3H/HeJ (C3H) and C57BL/6J (B6) at birth (postnatal day P1). C3H mice had significantly smaller kidneys than B6 mice measured as planar surface area (p = 2 × 10−7, Fig. 1A) or weight (p = 5 × 10−7, data not shown). We also examined kidney size in adult mice at 8 weeks of age (P56), when both nephrogenesis and kidney growth is completed. Unexpectedly, there was a difference in kidney size between the two strains: C3H kidneys had significantly larger planar surface areas (p = 1.8 × 10−7, Fig. 1B) and weights (p = 3.4 × 10−5, data not shown) than B6 kidneys. The increase in kidney size was due to an increase in both the surface area of the medulla and the cortex (mean ± SE): for the medulla, C3H: 11.17 ± 0.19 mm2, n = 24, vs. B6 9.95 ± 0.18 mm2, n = 24, (p = 4 × 10−6), and for the cortex: C3H 27.72 ± 0.51 mm2, n = 24, vs. B6 21.46 ± 0.31 mm2, n = 24, (p = 4 × 10−21). When males and females were compared individually, C3H females and males had larger kidney surface areas and weights (data not shown) than their B6 counterparts (planar surface area for C3H vs. B6 females: p = 0.007, n = 12 and C3H vs. B6 males: p = 1 × 10−5, n = 12, data not shown).
Birth weights between C3H and B6 mice did not differ (mean ± SE): 1.36 ± 0.05 g, n = 14, for C3H vs. 1.32 ± 0.04 g, n = 14, for B6 (p = 0.55), demonstrating that while C3H mice had smaller kidneys at birth, there was no difference in their body weight when compared to B6 mice. Similarly, there was no difference in body weight at adulthood when both strains were compared (mean ± SE): 18.8 ± 0.6 g, n = 12, for C3H vs. 18.6 ± 0.7 g, n = 12, for B6 (p = 0.79), or when males and females were examined separately (C3H vs. B6 females, p = 0.08, n = 12 and C3H vs. B6 males, p = 0.69, n = 12, data not shown).
We hypothesized that the change in kidney size between birth and adulthood could be from a nephron deficit with subsequent compensatory hypertrophy of the remaining glomeruli to maintain renal function. To determine if a nephron deficit in the C3H mouse was responsible for the observed kidney phenotype, newborn and adult kidneys from C3H and B6 mice were examined to assess nephron number.
The relationships between nephron number, kidney size and body weight were examined in newborn C3H and B6 mice. Kidney weight correlated positively with kidney surface area when data from both strains were combined (p = 4 × 10−9, Fig. 2A) and when each strain was examined individually (p < 0.05 for each, Suppl. Table 1), suggesting that these measurements are interchangeable. When kidney weight was compared to birth weight, a positive correlation was seen when both strains were combined (p = 1.2 × 10−4, Fig. 2B) and when each strain was examined separately (p < 0.05 for each, Suppl. Table 1), suggesting that mice with a lower birth weights are born with smaller kidneys. The same correlation was observed when kidney surface area was compared to birth weight when both strains were combined (R2 = 0.385, p = 2.0 × 10−4, n = 28) and when each strain was examined separately (p < 0.05 for each, Suppl. Table 1).
Despite the significant difference in kidney size between the two strains, there was no significant difference in nephron number. C3H (mean ± SE): 2243 ± 79 nephrons (n = 13) vs. B6: 2060 ± 92 nephrons (n = 13, p = 0.14). Importantly, nephron number did not correlate with body weight (p = 0.7, Fig. 2C), with kidney surface area (R2 = 0.0444, p = 0.3, n = 13), or with kidney weight (p = 0.63, Fig. 2D). Similarly, no correlation was found between nephron number and kidney size when either strain was examined separately (p > 0.05 for each, Suppl. Table 1). These results demonstrate that kidney size correlates with birth weight, but nephron number does not correlate with birth weight or kidney size.
Nephrogenesis continues postnatally in mice, therefore, we examined nephron number in C3H and B6 mice at 8 weeks, when both nephrogenesis and kidney growth have ended. The number of nephrons was estimated in 24 mice from C3H and B6 strains (n = 48 kidneys, 6 males and 6 females from each strain). When males and females were pooled together, there was no significant difference in nephron number between C3H and B6 kidneys (p = 0.4, Fig. 3A). When males and females were examined separately, C3H females and both C3H and B6 males had, on average, 13% fewer nephrons when compared to B6 females (p = 0.05, Fig. 3B).
Studies in humans and mice have shown that as the number of nephrons decreases, glomerular volume and size increases.12,13,16 To examine whether the decrease in nephron number in C3H females and C3H and B6 males was accompanied by an increase in glomerular size, glomerular tuft planar surface areas were measured (35 glomeruli were sampled for each of the 48 kidneys).17 C3H females and C3H and B6 males had glomerular tufts that were on average 12% larger when compared to B6 females (p = 0.003, Fig. 3C). However, no relationship was observed between glomerular size and nephron number when all groups were pooled together (R2 = 0.0032, p = 0.7, n = 48, data not shown).
From studies of adults, nephron number correlates with body weight, body surface area, height and kidney weight in humans.2,3,5,6 Kidney weight correlated with kidney surface area in adult C3H and B6 mice (R2 = 0.9088, p = 1.5 × 10−25, n = 48). Importantly, as in the newborn period, kidney weight correlated with body weight (R2 = 0.551, p = 1.6 × 10−9, n = 48). However, we found no relationship between nephron number and body weight (p = 0.97, Fig. 4A). We also did not find a relationship between nephron number and kidney weight (p = 0.58, Fig. 4B) or between nephron number and kidney surface area (R2 = 9 × 10−5, p = 0.95, n = 48) in adult C3H and B6 mice. Similarly, we did not find any correlations between nephron number and body weight or kidney size when either strain was examined separately (p > 0.05 for each, Suppl. Table 1). When data from females and males was examined separately, there was still no correlation between nephron number and kidney size or between nephron number and body weight (p > 0.05 for each, Suppl. Table 1). These results suggest that differences in nephron number cannot be predicted by a change in kidney size or body weight and that nephron number is an independent variable in these mouse strains.
There is great interest in nephron number, its correlation with body and kidney size, and its association with disease susceptibility in humans.3,6,11,12 To understand the impact of genetic and environmental variation on these relationships, we examined the relationship between kidney size and body size and between nephron number and kidney or body size in two inbred mouse strains. We determined that kidney size does correlate with body weight in the newborn and the adult period in both the C3H and B6 mouse strains. However, nephron number at birth and at adulthood did not correlate with kidney size or body weight. At birth, C3H kidneys were 10% smaller than B6 kidneys, although there were no significant differences in nephron number between the two strains. At adulthood, C3H kidneys were 17% larger than B6 kidneys, and there was still no significant difference in nephron number. These results suggest that kidney size is either independent of nephron number in these inbred mouse strains, or alternatively, we may not have been able to discern a relationship between nephron number and kidney or body size because of the lack of variation in these parameters.
This is the first study to examine the relationship between nephron number and kidney size or body weight in inbred mouse strains. Much of our understanding of these relationships has come from studies on rats with naturally occurring low birth weight, and different results have emerged.18,19 One study showed that Wistar rats with naturally occurring low birth weight had normal sized kidneys, a 20% reduction in nephron number, and an increase in glomerular volume,18 while another study showed that Sprague-Dawley rats with naturally occurring low birth weight did not have a decrease in kidney weight or in nephron number.19 This latter study also reported no relationship between nephron number and birth weight.19 Although body weight did correlate with kidney size in C3H and B6 mice, it was not an accurate predictor of nephron number. We also did not observe a relationship between nephron number and kidney size. The C3H mouse is normotensive, does not develop microalbuminuria, and has normal renal function.20–22 Similarly, the B6 mouse is resistant to renal disease.23,24 In contrast, other inbred strains, SWR/J and DBA/2J, have been shown to be susceptible to hypertension and chronic kidney disease, respectively.21,25,26 It is possible that inbred mouse strains that are susceptible to renal disease may exhibit differences in nephron number that correlate with kidney size and body weight, but this was not examined in the current study.
Nephron number may also be influenced by gender. In humans, females generally have smaller kidneys with fewer nephrons than males.1,13 Another study identified a correlation between nephron number and birth weight in males, but not in females.2 Although C3H females had fewer nephrons than C3H males, the inverse relationship was seen in B6 mice in which females had more nephrons than males. Therefore, the impact of gender on nephron number in these strains remains unclear.
Our results demonstrate that at adulthood, C3H females and C3H and B6 males had a 13% decrease in nephron number that was accompanied by a proportional 12% increase in glomerular tuft size when compared to B6 females. Although a 30% decrease in nephron number in Australian Aboriginals was associated with a 27% increase in glomerular volume,6 the relationship between nephron loss and glomerular hypertrophy is poorly understood. It appears that the kidney is able to adapt to maintain an adequate filtration surface area, such that when nephron number decreases, glomerular size increases proportionally.
To determine whether the relationship between nephron number and kidney or body size is an issue of scale, we examined the fold variation in kidney size, body weight and nephron number reported in the literature. In the studies that showed a positive correlation between nephron number and kidney size, at least a 5.6-fold (range 3.5–8.7) variation in nephron number correlated with a three-fold (range 2.4–3.4) variation in kidney size.1,3,6 In one study that identified a correlation between nephron number and birth weight, there was an eight-fold variation in nephron number and a two-fold variation in birth weight.2 In the studies that did not show a correlation, there was an average 2.9-fold (range 2.6–3.2) variation in nephron number and an average 1.9-fold (range 1.7–2.2) variation in kidney size.11,13 In our study, there was a 1.9-fold variation in nephron number, a two-fold variation in kidney size, and a 1.4-fold variation in body weight at birth and at adulthood. We therefore suspect that large variations in these parameters, at least two-fold, may be required before a correlation can be observed between nephron number and kidney or body size.
In summary, while kidney size appears to be tightly correlated with body weight at birth and at adulthood, nephron number does not correlate with kidney size or body weight in C3H and B6 inbred mice. Our results suggest that within the normal, relatively small, range of kidney and body size seen in these inbred mouse strains, these parameters are independent of nephron number. However, we speculate that positive correlations between nephron number and kidney or body size may be observed when there are large differences, greater than two-fold in these variables. This is most likely to be seen when there is a greater range of phenotypic variability either from a developmental insult or from an acquired disease. Future studies clearly need to continue to examine the relationship between nephron number, kidney size, body weight and disease susceptibility to determine if these predictions in regards to the issue of scale are correct.
C3H/HeJ and C57BL/6J (Jackson Laboratories) kidneys were collected at postnatal day 1 (P1) for the newborn studies. C3H/HeJ and C57BL/6J kidneys were collected at 8 weeks of age (P56) for the adult studies. Body weights and kidney weights were obtained for all mice studied. Kidney planar surface areas were measured using SPOT (v.3.5.9) to assess kidney size, as previously described.27 Surface areas and weight measurements were normalized for body weight to account for any differences in gestational age in the newborn mice, and to account for any differences in body size in the adult mice.20 Kidneys were fixed in 4% PFA, processed for paraffin-embedding and cut into 7 µm serial sections. Hematoxylin and eosin staining was performed according to standard protocol. Animal studies were performed in accordance with the rules and the McGill University UACC approved regulations of the Canadian Council of Animal Care (CCAC) and animal protocols.
An absolute nephron number count was obtained from nine newborn kidneys from C3H/HeJ and C57BL/6J mice. Male and female mice as well as left and right kidneys were sampled. Images of each section were taken using a Canon PowerShot S50 attached to a Zeiss Axiophot by a Leica adaptor. The serial sections were printed and “write-on” transparency films were applied to each image. Glomeruli were identified by the presence of a glomerular tuft, circled on each transparency, and numbered in ascending order. The transparency was then compared to the adjacent section. Overlapping glomeruli were assigned the same number while all new glomeruli were assigned a new number, ensuring that glomeruli that appeared in multiple sections were only counted once. All glomeruli in a newborn kidney were tracked, and an absolute nephron number count was obtained for these nine kidneys. The same kidneys were then used to develop and validate a formula28,29 to estimate the total number of nephrons: Nglom = f × 0.4 × NN. Where f is the fraction of the kidney sampled, 0.4 is a constant derived from the absolute counts that corrects for overlapping glomeruli when all glomeruli on one section are counted, and NN is the total number of nephrons counted from the sampling. For the newborn studies, 10% of each kidney was sampled and sections were chosen at least 70 µm apart to avoid the possibility of double-counting glomeruli.
Serial sections from adult C3H/HeJ and C57BL/6J kidneys were imaged using a Canon PowerShot S50 attached to a Zeiss Axiophot by a Leica adaptor. Whole kidney sections were imaged in parts and merged using Adobe Photoshop CS. Adult kidneys were analyzed using Image J (v1.36b) and individual glomeruli, defined by the presence of a glomerular tuft, were counted using the point selection tool with auto measure. The estimation method described for newborn kidneys was used to estimate nephron number in all adult kidneys. Two independent counts were obtained for each kidney and the average nephron count is reported.
Glomerular tuft sizes, measured as planar surface area, were obtained using SPOT (v.3.5.9) as previously described.17 To obtain cortex and medulla planar surface area measurements, the kidney section with the largest surface area was identified, and the surface area from six adjacent sections was measured and used to obtain an average. Student’s t-test and regression analysis was used for statistical analysis of the data.
We thank C.G. Goodyer and N. Haddad for critical reading of the manuscript and discussions; X. He for assistance with the paraffin sections; and X. Zhang for help with statistical analysis. This work was supported by an operating grant from the Kidney Foundation of Canada. I.R.G. is a recipient of an FRSQ chercheur-boursier clinicien salary award. I.J.M. is a recipient of an FRSQ doctoral award.
Source of support: Kidney Foundation of Canada.
Previously published online: www.landesbioscience.com/journals/organogenesis/article/12125