Podometrics in Japanese Living Donor Kidneys: Associations with Nephron Number, Age, and Hypertension

Significance Statement

Podocyte depletion and low nephron number are associated with glomerulosclerosis and CKD. However, the relationship between podometrics and nephron number has not previously been reported. The authors estimated podometric parameters and nephron number in 30 Japanese kidney donors. Their podocyte density and number per glomerulus were similar to values reported for other racial groups, whereas they had fewer nonsclerotic nephrons compared with other races. Total podocyte number per kidney declined at a rate of 5.63 million podocytes per year, with 80% of podocyte loss resulting from glomerulosclerosis-associated glomerular loss, and the remainder occurring in healthy glomeruli. Hypertension was associated with lower podocyte density and larger podocyte volume, independent of age. These approaches could be of value in evaluating the kidney in health and disease.

Visual Abstract

Abstract

Background

Podocyte depletion, low nephron number, aging, and hypertension are associated with glomerulosclerosis and CKD. However, the relationship between podometrics and nephron number has not previously been examined.

Methods

To investigate podometrics and nephron number in healthy Japanese individuals, a population characterized by a relatively low nephron number, we immunostained single paraffin sections from 30 Japanese living-kidney donors (median age, 57 years) with podocyte-specific markers and analyzed images obtained with confocal microscopy. We used model-based stereology to estimate podometrics, and a combined enhanced–computed tomography/biopsy-specimen stereology method to estimate nephron number.

Results

The median number of nonsclerotic nephrons per kidney was 659,000 (interquartile range [IQR], 564,000–825,000). The median podocyte number and podocyte density were 518 (IQR, 428–601) per tuft and 219 (IQR, 180–253) per 10⁶ μm³, respectively; these values are similar to those previously reported for other races. Total podocyte number per kidney (obtained by multiplying the individual number of nonsclerotic glomeruli by podocyte number per glomerulus) was 376 million (IQR, 259–449 million) and ranged 7.4-fold between donors. On average, these healthy kidneys lost 5.63 million podocytes per kidney per year, with most of this loss associated with glomerular loss resulting from global glomerulosclerosis, rather than podocyte loss from healthy glomeruli. Hypertension was associated with lower podocyte density and larger podocyte volume, independent of age.

Conclusions

Estimation of the number of nephrons, podocytes, and other podometric parameters in individual kidneys provides new insights into the relationships between these parameters, age, and hypertension in the kidney. This approach might be of considerable value in evaluating the kidney in health and disease.

Podocyte injury and loss are important early events in the development of glomerulosclerosis, leading to reduced renal function. Podocytes have a limited capacity to regenerate under normal conditions.^1–3 Animal studies have shown that absolute (due to podocyte detachment from glomerular capillaries and apoptosis) and relative (associated with glomerular enlargement) podocyte depletion can cause glomerulosclerosis.^4–7 Importantly, Wharram et al. ⁴ reported a dose-response relationship between the degree of podocyte depletion and the severity of subsequent renal pathologic and physiologic changes in a transgenic rat model of podocyte depletion and FSGS. In humans, podocyte depletion has been reported in patients with type 1 and type 2 diabetes,^8–12 IgA nephropathy,¹³ hypertensive nephrosclerosis,¹⁴ and transplant glomerulopathy.¹⁵ Recently, Naik et al. ¹⁶ reported that higher podocyte stress and detachment are associated with higher mean arterial pressure in the normal range. In many of the above studies, the remaining podocytes underwent compensatory hypertrophy to maintain coverage of glomerular capillaries.^7,17–20 It is now widely accepted that absolute or relative podocyte depletion and compensatory podocyte hypertrophy are important early events in the development of glomerulosclerosis.^21–23

We have previously reported that healthy Japanese subjects have approximately 30% fewer nephrons (glomeruli) than White Americans and Europeans.^24,25 However, the prevalence of kidney disease in Japanese individuals is not as high as in other races.²⁶ Furthermore, in a recent international comparison of CKD cohort studies, Japanese patients with CKD were shown to have a slower progression rate and lower incidence of ESKD than patients in most other countries.²⁷ Therefore, we hypothesized that podocyte number and density in Japanese people is likely to be similar to, or possibly even higher than, that in other races, providing a degree of glomerular protection in the setting of their relatively low nephron number.

Over the past ≥30 years, much research attention has focused on associations between nephron number and adult hypertension and CKD.^28–31 More recently, the podocyte-depletion hypothesis has stimulated efforts to count podocytes and estimate other podocyte metrics, as described above. Interestingly, both nephrogenesis and podocytogenesis cease either shortly before or after birth in humans, so any developmental deficit in nephron or podocyte endowment is permanent; however, evidence for podocyte replacement after loss has been presented.^32,33 Surprisingly, only a handful of studies, to date, have estimated nephron and podocyte number in the same animal model or in humans,^34–36 but associations between nephron number and podometric indices have not previously been reported, to our knowledge. In this study, we report nephron number and podometric parameters for 30 Japanese living-kidney donors. We estimated nephron number using a combined enhanced–computed tomography (CT)/renal biopsy–specimen stereology method,²⁵ whereas podometric parameters were estimated on single paraffin sections obtained during kidney transplant. This approach allowed us to estimate, for the first time, the total number of podocytes in human kidneys and the contributions of glomerulosclerosis and resultant nephron loss, versus loss of podocytes in nonsclerotic glomeruli, to overall podocyte loss. We also report relationships between nephron number; podocyte number and density; and clinical data, including age and hypertension.

Methods

Living-Kidney Donors and Clinical Data Collection

This study was conducted in accordance with the Declaration of Helsinki. The clinical and research activities being reported are consistent with the Principles of the Declaration of Istanbul as outlined in the “Declaration of Istanbul on Organ Trafficking and Transplant Tourism.” The study protocol was approved by the Monash University Human Research Ethics Committee (19070) and the Jikei University School of Medicine Ethics Review Board (30-268 [9289]). All kidney donors included in this study provided written informed consent to perform the renal biopsy and to use the biopsy specimen for research. They were also informed they could withdraw permission for the use of their specimen and clinical data at any time.

We used archival biopsy kidney specimens obtained from living-kidney donors at the Jikei University School of Medicine Hospital between 2007 and 2017, as previously described.²⁵ In our institution, we obtained a needle biopsy specimen using an 18-gauge needle at the time of kidney resection and 1 hour after blood reperfusion. In this study, we used the specimens obtained at 1 hour. The exclusion criterion was that the single histologic section used for podometrics contained fewer than eight glomerular profiles.^15,37,38

Clinical data were obtained during admission for renal donation. Body surface area (BSA) for Japanese individuals was determined using the equation:

$$\text{BSA}\left({\text{m}}^2\right){=\text{Weight}}^{\text{0}\text{.}444}\left(\text{kg}\right){\times \text{Height}}^{\text{0}\text{.}663}\left(\text{cm}\right)\times \mathrm{0.008883.}$$

Hypertension was defined as a systolic BP of ≥140 mm Hg and/or a diastolic BP of ≥90 mm Hg, or the use of antihypertensive medications. The eGFR was calculated using a modified three-variable equation for estimating the GFR for Japanese individuals³⁹:

$$\text{eGFR}=194\times {\text{Age}}^{-\text{0}\text{.}287}\times {\text{s}-\text{Cre}}^{-1\text{.}094}\left(\times 0\text{.}739,\hspace{0.17em}\text{if\hspace{0.17em}female}\right),$$

where s-Cre is the serum creatinine level.

Estimation of Nephron Number

Nephron number was estimated using a combined enhanced-CT/biopsy-specimen morphometry method, as previously described.^25,40 In brief, kidney cortical volume was measured from transverse images obtained during the arterial phase of the enhanced CT. Nonsclerotic nephron number and total nephron number (including globally sclerotic glomeruli) were estimated by multiplying total kidney cortical volume by the nonsclerotic or total glomerular numeric density (number per cubed millimeter of cortex), respectively. These values were divided by two (per kidney), by 1.43 (to correct for tissue-volume shrinkage due to paraffin embedding), and by 1.268 (to correct for volume shrinkage due to loss of tissue-perfusion pressure).^25,40

Immunofluorescence for Podocyte Markers

Kidney biopsy specimens were fixed in 10% formalin, embedded in paraffin, and sectioned to a thickness of 3 μm. After deparaffinization and rehydration, the slides were incubated with EDTA buffer at 92°C for 2 hours to unmask antigens. Then, the slides were blocked with 1% BSA for 1 hour. To identify podocyte nuclei, three primary antibodies were initially assessed: dachshund 1 (DACH1, 1:1000 dilution, 10914-1-AP; Proteintech),⁴¹ (preprint) Wilms tumor 1 (WT1, 1:150 dilution, sc-7385; Santa Cruz Biotechnology, Santa Cruz, CA), and transducing-like enhancer of split 4 (TLE4, 1:250 dilution, SC-365406; Santa Cruz Biotechnology). Double labeling with WT1 and TLE4 confirmed that all DACH1-positive nuclei belonged to podocytes (Figure 1). To label podocyte cytoplasm, sections were immunostained with an antibody directed against synaptopodin (SNP, 1:500 dilution, SC-515842; Santa Cruz Biotechnology). Primary antibodies were incubated at 4°C overnight. After washing, slides were incubated with Alexa Fluor 568–conjugated goat anti-mouse IgG antibody (A11004; Invitrogen, Carlsbad, CA) and Alexa Fluor 633–conjugated goat anti-rabbit IgG antibody (A21070; Invitrogen) at 1:1000 dilution for 2 hours at room temperature. A commercially available autofluorescence quenching kit with 4′,6-diamidino-2-phenylindole (DAPI, SP-8500; Vector Laboratories, Burlingame, CA) was used to visualize nuclei.

Figure 1.

Representative images of double immunofluorescence labeling for podocyte nuclear markers. (A) Representative images for double immunofluorescence for DACH1 and WT1. (B) Representative images for double immunofluorescence for DACH1 and TLE4. Dotted lines indicate the outer edge of the glomerular tuft. Markers: DACH1, magenta; WT1, green; TLE4, green; DAPI, blue. Overlay images (DACH1+WT1 and DACH1+TLE4) appear gray due to the colocation of magenta and green markers.

Confocal Imaging

Images of glomeruli were obtained using a Leica SP5 laser confocal microscope (Leica Microsystems, Mannheim, Germany). A ×40 oil-immersion objective lens (numeric aperture [NA]=1.25) and a ×1.6 set zoom was used. All images were obtained using the same laser power throughout the study.

Estimation of Glomerular Volume

We used Fiji⁴² for all image analysis. Each glomerular-tuft area was defined as the area of the outer side of the capillary loops of the glomerular tuft on the merged image of DACH1, SNP, and DAPI immunofluorescence (Figure 2A). Mean glomerular area for each biopsy specimen was calculated by averaging all measured areas of the glomerular tufts.^43,44 Glomerular volume was calculated from the mean glomerular area using the following Weibel and Gomez equation:

$$\text{Glomerular\hspace{0.17em}volume}=\frac{\mathit{\beta }}{\text{d}}\times {\left(\text{mean\hspace{0.17em}glomerular\hspace{0.17em}area}\right)}^{3\text{/}2},$$

where β is a dimensionless shape coefficient (1.382 for spheres), and d is a size distribution coefficient used to adjust for variations in glomerular size. We used a value of d=1.01 in this study.^40,45 Total glomerular volume per kidney was calculated by multiplying mean glomerular volume by the number of nonsclerotic glomeruli.

Figure 2.

Representative images used for quantifying glomerular area, podocyte nuclei, and podocyte cytoplasm. (A) A trace of a glomerular-tuft area (white line) on the merged image of DACH1 (magenta), SNP (green), and DAPI (blue). (B) Representative images for double immunofluorescence for DACH1 (magenta) and SNP (green). (C) DACH1-positive podocyte nuclei are surrounded by SNP-positive podocyte cytoplasm and are located on the outer side of glomerular capillaries. (D) Arrows indicate parietal epithelial cells on Bowman’s capsule that were DACH1 positive. Such cells were not classified as podocytes. (E) Identification of podocyte nuclei immunostained for DACH1 (upper left); DAPI (upper right); merged image of DACH1 and DAPI (bottom left); and merged image of DACH1, SNP, and DAPI (bottom right). (F) A binary image for SNP-positive area.

Identification of Podocyte Nuclei and Estimation of Podocyte Density and Number

Podocyte nuclei were identified using the following criteria: (1) nuclei were positive for DACH1 and DAPI (Figure 2B), (2) nuclei were located outside glomerular capillaries (Figure 2C), and (3) nuclei were not located on the Bowman’s capsule. Nuclei located on the Bowman’s capsule were defined as belonging to parietal epithelial cells (Figure 2D) and were not counted as podocytes.³⁵

All nuclei within a glomerulus were marked by regions of interest on a DAPI image using a region-of-interest manager tool in Fiji. Regions of interest were then edited according to the criteria for podocyte nuclei (provided above) using reference images of DACH1, DAPI, the merged image of DACH1 and DAPI, and the merged image of all previous images (Figure 2E). The x-y axis caliper diameters of all identified podocyte nuclei were measured and averaged for each biopsy specimen.

Podocyte density was calculated on the basis of the number of podocyte nuclei per section, apparent podocyte nuclear caliper diameter, total glomerular area, and optical-section thickness, according to the method previously reported by Venkatareddy et al. ³⁷ Optical-section thickness was calculated as 0.541 μm using the formula for axial resolution on confocal microscopy:

$$\text{Axial\hspace{0.17em}resolution}=\frac{0.88\times \text{λ}}{n-\sqrt{n^2-{\left(NA\right)}^2}},$$

where λ is the excitation wavelength for DAPI (405 nm), n is the refractive index of the immersion oil (1.515), and NA is 1.25 (NA of the objective lens). Podocyte number per tuft was calculated by multiplying podocyte density by glomerular volume. Total podocyte number per kidney was obtained by multiplying podocyte number per tuft by the number of nonsclerotic glomeruli.

Estimating Podocyte-Volume Indices

Podocyte cytoplasm was defined as being positive for SNP. The binary image of SNP staining was created using the IsoData algorithm⁴⁶ and measured the SNP-positive area within a glomerulus (Figure 2F). The percentage SNP (%SNP) was defined as the sum of SNP-positive areas in glomerular tufts divided by the sum of all glomerular tuft areas in the biopsy specimen. The average volume of cytoplasm in a podocyte was calculated as follows:

$$\text{Podocyte\hspace{0.17em}cytoplasmic\hspace{0.17em}volume}=\frac{\text{Glomerular\hspace{0.17em}volume}\times \left(\text{\%SNP}\right)}{\text{Podocyte\hspace{0.17em}number\hspace{0.17em}per\hspace{0.17em}tuft}}.$$

Average podocyte nuclear volume was estimated on the basis of the mean apparent caliper diameter of podocyte nuclei using the following formula^47,48:

$$\text{Podocyte\hspace{0.17em}nuclear\hspace{0.17em}volume}=\frac{4\mathit{\pi }}{3} \times {\left(\frac{2}{\mathit{\pi }}\times \text{apparent\hspace{0.17em}caliper\hspace{0.17em}diameter\hspace{0.17em}of\hspace{0.17em}podocyte\hspace{0.17em}nuclei}\right)}^3.$$

Average podocyte volume was calculated as the sum of podocyte cytoplasmic volume and podocyte nuclear volume.⁴⁹ Podocyte volumetric density (the proportion of glomerular-tuft volume comprised by podocytes; V_V[Pod/Glom]) was estimated using

$${\text{V}}_{\text{V}\left(\text{Pod/Glom}\right)}=\frac{\text{Podocyte\hspace{0.17em}volume}\times \text{podocyte\hspace{0.17em}number\hspace{0.17em}per\hspace{0.17em}tuft}}{\text{Glomerular\hspace{0.17em}volume}}.$$

Statistical Analyses

Due to the relatively small sample size in this study, we used nonparametric statistical methods. Continuous variables are expressed as median (interquartile range; IQR). For comparisons between two groups, Mann–Whitney U tests or chi-squared tests were used. The Spearman correlation test was used to evaluate correlations between two variables. Multiple regression analyses were used to examine the effect of age and hypertension on podometrics, nephron number, and glomerular volume. To estimate the annual rate of decline in nephron number and total podocyte number per kidney, the slope of the linear regression was calculated. A value of P<0.05 was considered significant for all statistical tests. Data were analyzed using SPSS software version 23 (IBM Inc., Armonk, NY) and GraphPad PRISM 8 (GraphPad Software, La Jolla, CA).

Results

Clinical Characteristics of Living-Kidney Donors

Biopsy specimens from 50 Japanese living-kidney donors were obtained. Biopsy specimens from 30 of these donors met the inclusion criteria for the study. The age, sex, body-size indices, hypertensive status, and renal function of these 30 included donors were very similar to the 20 excluded donors (Table 1). Of the 30 included donors, 21 were female (70%). Age ranged from 36 to 72 years with a median of 57 years. BSA was 1.59 (IQR, 1.45–1.72) m², and significantly smaller in females than males (1.53 [IQR, 1.44–1.62] m² versus 1.73 [IQR, 1.66–1.87] m²; P<0.001). Five donors (17%) were hypertensive, with their BP controlled using antihypertensive medications. The median number of glomerular profiles evaluated on the 30 included biopsy specimens was 9.5 (IQR, 8.0–12.25).

Table 1.

Characteristics of the living-kidney donors enrolled in the study

Variables	Eight or More Glomeruli (n=30)	Less than Eight glomeruli (n=20)	P Value
Age, yr	57.0 (48.0–63.25)	59.0 (55.5–66.25)	0.12
Male sex, n (%)	9 (30)	10 (50)	0.15
Height, cm	161 (155–166)	161 (154–167)	0.81
Body weight, kg	59.0 (53.1–67.3)	62.2 (55.0–66.4)	0.77
Body mass index, kg/m2	22.2 (20.8–26.1)	24.3 (21.4–25.6)	0.48
BSA, m2	1.59 (1.45–1.72)	1.61 (1.47–1.69)	0.92
Hypertension, n (%)	5 (17)	7 (35)	0.14
Serum creatinine, mg/dl	0.71 (0.59–0.78)	0.70 (0.61–0.82)	0.74
eGFR, ml/min per 1.73 m2	77.7 (67.5–85.3)	77.0 (67.6–86.6)	0.96

Continuous variables are shown as median (IQR). Mann–Whitney U tests or chi-squared tests were used.

Nephron Number and Podometrics

The nephron number and podometric data are shown in Table 2. The median number of nonsclerotic glomeruli per kidney was 659,000 (IQR, 564,000–825,000), with values ranging from 291,000 to 1,216,000. The median number of total glomeruli per kidney was 720,000 (IQR, 579,000–988,000), with values ranging from 291,000 to 1,340,000. Mean glomerular volume and total glomerular volume per kidney varied by 4.4- and 9.6-fold, respectively. Podocyte density and number were 219 (IQR, 180–253) per 10⁶ μm³ of glomerular tuft volume and 518 (IQR, 428–601) per tuft, respectively. Total podocyte number per kidney (the product of podocyte number per tuft and the number of nonsclerotic glomeruli) was 376 million (IQR, 259–449×10⁶), with a 7.4-fold difference (range, 90×10⁶–673×10⁶). Podocyte volume (combined nucleus and cytoplasm) was 1347 (IQR, 1086–1590) μm³. Podocytes comprised 27.7% (IQR, 24.7%–33.1%) of total glomerular volume (V_V[Pod/Glom]).

Table 2.

Nephron-related and podometric data for 30 living-donor kidneys

Variables	Median (IQR)
Kidney cortical volume (cm3)	90.5 (74.5–104.0)
Nonsclerotic nephron number (×103)	659 (564–825)
Total nephron number (×103)	720 (579–988)
Glomerular volume (×106 μm3)	2.32 (1.79–3.07)
Total glomerular volume per kidney (mm3)	1540 (1125–2261)
Number of glomerular profiles evaluated per biopsy specimen	9.5 (8.0–12.25)
Apparent caliper diameter of podocyte nuclei (μm)	6.09 (5.74–6.35)
Estimated true diameter of podocyte nuclei (μm)	8.26 (7.78–8.61)
Podocyte density (per 106 μm3)	219 (180–253)
Podocyte number per glomerular tuft	518 (428–601)
Total podocyte number per kidney (×106)	376 (259–449)
Percentage of glomerulus stained for SNP	23.0 (19.1–27.5)
Podocyte nuclear volume (μm3)	243 (204–276)
Podocyte cytoplasmic volume (μm3)	1105 (869–1342)
Podocyte volume (μm3)	1347 (1086–1590)
Podocyte nuclear/cytoplasmic ratio	0.21 (0.18–0.27)
VV(Pod/Glom) (%)	27.7 (24.7–33.1)

Continuous variables are shown as median (IQR).

Relationship between Glomerular Size and Podometrics

As expected, podocyte number per tuft was directly correlated with glomerular volume (r _s coefficient=0.48, P<0.01; Supplemental Figure 1A), whereas podocyte density was inversely correlated with glomerular volume (r _s=−0.79, P<0.0001; Supplemental Figure 1B). Larger glomerular volume (r _s=0.50, P<0.01; Supplemental Figure 1C) and lower podocyte density (r _s=−0.75, P<0.0001; Supplemental Figure 1D) were both associated with larger podocyte size. Larger glomerular volume was inversely correlated with V_V(Pod/Glom) (r _s=−0.65, P<0.0001; Supplemental Figure 1E). Podocyte density was directly correlated with V_V(Pod/Glom) (r _s=0.62, P<0.001; Supplemental Figure 1F), whereas podocyte number per tuft was not significantly associated with either podocyte volume or V_V(Pod/Glom). These associations remained statistically significant after removal of the two donors with glomerular volumes >5.0×10⁶ μm³ (data not shown).

Associations between Podometric Parameters and Nephron Number

The number of nonsclerotic glomeruli per kidney was not associated with podocyte density (Figure 3A). However, a direct correlation between the number of nonsclerotic glomeruli per kidney and podocyte number per glomerulus approached statistical significance (r _s=0.35, P=0.06; Figure 3B). Podocyte volume and V_V(Pod/Glom) were not associated with the number of nonsclerotic glomeruli per kidney (Figure 3, C and D).

Figure 3.

Associations between podometrics and the number of nonsclerotic glomeruli. (A) Podocyte density, (B) podocyte number per tuft, (C) podocyte volume, and (D) V_V(Pod/Glom). Spearman correlation test was used.

Podometrics and Nephron Number in Relation to Age and Hypertension

Multiple linear-regression analyses were performed to determine how age and hypertension associated with podometrics and nephron number. An inverse correlation between the number of nonsclerotic glomeruli and age approached statistical significance (standardized β=−0.40, P=0.05; Supplemental Table 1A). Total nephron number and glomerular volume were not associated with either age or hypertension (Supplemental Table 1, B and C). In contrast, hypertension was significantly associated with lower podocyte density, independent of age (standardized β=−0.41, P=0.04; Table 3). Neither age nor hypertension was significantly associated with podocyte number per tuft (Table 3), whereas increasing age was associated with reduced podocyte number per kidney, independent of hypertension (standardized β=−0.40, P=0.05; Table 3). Both increased age and hypertension were associated with larger podocyte volume in univariate analysis. In a multivariate model, hypertension was associated with increased podocyte volume, independent of age (standardized β=0.39, P=0.04; Table 3). If glomerular volume was used as an independent variable, hypertension was significantly associated with larger podocyte volume, independent of glomerular volume (standardized β=0.39, P=0.02; Supplemental Table 2). Neither age nor hypertension was associated with V_V(Pod/Glom) (Table 3).

Table 3.

Multiple linear regression analyses for podometric data

Variable	Univariable		Multivariable
	R	P Value	Standardized β	P Value
Podocyte density
Age (yr)	−0.14	0.47	0.02	0.91
Hypertensive	−0.40	0.03	−0.41	0.04
Podocyte number per tuft
Age (yr)	−0.19	0.32	−0.16	0.43
Hypertensive	−0.13	0.49	−0.07	0.75
Total podocyte number per kidney
Age (yr)	−0.37	0.05	−0.40	0.05
Hypertensive	−0.06	0.74	0.10	0.63
Podocyte volume
Age (yr)	0.39	0.03	0.24	0.19
Hypertensive	0.48	0.007	0.39	0.04
VV(Pod/Glom)
Age (yr)	0.24	0.21	0.33	0.10
Hypertensive	−0.12	0.53	−0.25	0.22

Age-Related Glomerular and Podocyte Loss

To estimate the annual rate of glomerular and podocyte loss per kidney, we performed regression analyses on the number of nonsclerotic glomeruli and total podocyte number per kidney. The rates of decline (regression slope) for nonsclerotic glomeruli and podocyte number per kidney were estimated using individual values for the 30 donors (Figure 4). The decline in number of nonsclerotic glomeruli with age approached statistical significance (P=0.06), indicating an average loss of 8320 glomeruli per year over the 36-year age range of the donors (Figure 4A). The total number of podocytes per kidney declined significantly with age (P<0.05) at a rate of 5.63 million podocytes per kidney per year over the 36-year period studied (Figure 4B). This equates to a loss of 643 podocytes per kidney per hour.

Figure 4.

Correlations between age, number of nonsclerotic glomeruli, and total podocyte number per kidney. (A) Correlation between age and number of nonsclerotic glomeruli. (B) Correlation between age and total podocyte number per kidney. The solid lines show the regression slope from 30 individual donors. The dotted curves show 95% confidence intervals for the regression slope.

Next, we calculated how many podocytes were lost as a result of glomerular loss due to glomerulosclerosis compared with podocyte loss in nonsclerotic glomeruli. Knowing that 8320 nonsclerotic nephrons were lost each year and knowing average podocyte number per tuft, the glomerulosclerosis-associated podocyte loss was estimated for each donor. A median of 4.31 million (IQR, 3.55×10⁶–5.00×10⁶) podocytes were lost per kidney per year as the result of glomerulosclerosis-associated nephron loss. We then subtracted glomerulosclerosis-associated podocyte loss from total podocyte loss to obtain the number of podocytes lost in nonsclerotic glomeruli for each donor. A median of 1.32 million (IQR, 0.63×10⁶–2.08×10⁶) podocytes were lost per kidney per year in nonsclerotic glomeruli. Finally, knowing the number of nonsclerotic glomeruli in each donor allowed us to estimate that a median of 1.90 (IQR, 0.98–3.63) podocytes were lost per nonsclerotic glomerulus per year.

Discussion

In this study, we estimated nephron number and podometric parameters, including podocyte number and density, in 30 Japanese living-donor kidneys. The main findings were: (1) mean podocyte density and podocyte number per glomerulus in Japanese individuals were similar to values previously reported for other racial groups obtained using similar techniques; (2) the number of nonsclerotic glomeruli per kidney was not significantly associated with any podometric values, although a direct correlation between podocyte number per glomerulus and the number of nonsclerotic glomeruli approached statistical significance; (3) median podocyte number per kidney was 376 million with values ranging by 7.4-fold between donors; (4) total podocyte number declined with age at a rate of 5.63 million podocytes per kidney per year; (5) the majority of podocytes were lost through glomerulosclerosis-associated glomerular loss, with only about 20% of podocyte loss occurring in nonsclerotic glomeruli; and (6) hypertension was associated with lower podocyte density and larger podocyte volume.

This is the first study to report podometrics in healthy Japanese kidneys. Table 4 provides a comparison of the podocyte number and density values observed during studies in adult populations without apparent kidney disease. Results from studies that used the Weibel and Gomez method for podometric estimation showed wide variability in podocyte number, with values ranging from 575 to 878 podocytes per tuft.^{8,10,12,14,50,51} However, in the two studies in which design-based stereology was used to estimate podocyte number in White Americans and Europeans, podocyte numbers were 558 and 580, respectively.^11,35 In studies in which the method of Venkatareddy et al. ³⁷ was used to estimate mean podocyte number per tuft in White and Black individuals, values ranged from 422 to 527,^15,38 with an average value of 479 podocytes. In this study, in which we used a modified version of the Venkatareddy et al. ³⁷ method, average podocyte number was 518 per tuft, which is similar to results from other studies that used either a design-based strategy or the Venkatareddy et al. method. Our results thus suggest that podocyte number per glomerulus in healthy Japanese individuals is similar to that in healthy European, American White, and American Black individuals.

Table 4.

Comparison of podometrics in adults without apparent kidney diseases

Methods	Subjects	Cohort	Age	Race/Country	Glomerular Volume (×106 μm3)	Podocyte Number Per Tuft	Podocyte Density (n/106 μm3)	Ref.
Design-based method	10	N/A	38	United Kingdom, Italy	3.16 a	580 a	193 a	11
Design-based method	12	Autopsy	40	White	2.42 (1.54–3.27)	558 (431–746)	N/A	35
Weibel and Gomez method	8	L	34±3	Three Asian, Three European, Two Hispanic	2.6±0.3	575±45	235±25	8
Weibel and Gomez method	24	L, C	41 (23–69)	United States	3.9±1.3	878±220	N/A	10
Weibel and Gomez method	20	L	N/A	United States	3.5±1.0	833±184	263±110	12
Weibel and Gomez method	10	N/A	35.6±9.23	China	2.94±0.43	714±103	245±48.4	50
Weibel and Gomez method	10	N/A	37.3±13.5	China	N/A	773±296	N/A	14
Weibel and Gomez method	10	N/A	N/A	China	2.47±0.02	825±118	365±45.9	51
Venkatareddy et al.	10	N/A	N/A	United States	2.8 a	527 a	194 a	37
Venkatareddy et al.	20	L, C, N	N/A	65% White, 13% Black	2.5±0.7	498±111	212±67	15
Venkatareddy et al.	19	L, C, N	35.5±6.8	United States	2.5±0.7	494±107	211±69	38
Venkatareddy et al.	27	L, C, N	52.4±4.6	United States	3.3±1.2	454±99	151±60	38
Venkatareddy et al.	23	L, C, N	72.2±9.1	United States	3.8±1.3	422±82	121±34	38
Modified Venkatareddy et al.	30	L	57	Asian (Japanese)	2.32 (1.79–3.07)	518 (428–601)	219 (180–253)	This study

Value are shown as median (IQR) or mean±SD. Ref., reference; N/A, not available; L, living-kidney donors; C, cadaver-kidney donors; N, subjects underwent nephrectomy.

^aIndicates mean value.

Our estimate of podocyte density in Japanese individuals of 219 (IQR, 180–253) podocytes per 10⁶ μm³ is also remarkably similar to previous estimates for White American, European, and Black individuals obtained using either design-based stereology or the Venkatareddy et al. method. In these four studies, the reported podocyte densities were 193,¹¹ 194,³⁷ 212¹⁵ and 121–211³⁸ podocytes per 10⁶ μm³ (Table 4). This indicates that, despite their generally smaller glomerular volumes, Japanese individuals have similar podocyte densities as other racial groups studied to date. This suggests that healthy human glomeruli, regardless of their size, have a fairly constant podocyte density which is optimal for glomerular function and health. This is despite the fact that BSA in White and Black individuals tends to be much greater than that in Japanese individuals. For example, in this study, BSA in male donors was 1.73 (IQR, 1.66–1.87) m², whereas the White American males studied by Puelles³⁵ had a BSA of 1.9 (IQR, 1.7–2.3) m². Given their smaller body size, our finding of similar podocyte number per glomerulus and density in Japanese individuals, compared with racial groups that have a larger BSA, suggests this may contribute to better renal outcomes in Japanese individuals, despite their relatively low nephron number.

To our knowledge, this is the first study to report associations between nephron number and podometrics in the same kidneys. The number of nonsclerotic glomeruli per kidney tended to be directly correlated with podocyte number per glomerulus, although this did not reach statistical significance (P=0.06). With a greater sample size, this relationship may have become statistically significant and, if this is the case, it would raise many new questions about the regulation of nephron endowment, glomerular size, and podocyte endowment. It is well known that a range of genetic factors and perturbations to the fetal-maternal environment^52,53 can influence nephron endowment at birth. Fetal-maternal factors include maternal nutrition,^54–56 gestational diabetes,⁵⁷ exposure to alcohol,^58,59 natural and synthetic glucocorticoids,⁶⁰ and hypoxia.^61,62 Little is known about regulation of podocytogenesis and podocyte endowment at birth, although we recently reported that mild hypoxia in mice during gestation gives rise to male offspring with low podocyte endowment, although female offspring were not affected.⁶³ Moreover, both nephron number and podocyte number decrease with healthy aging, so this may also contribute to the correlation between number of glomeruli and podocyte number.^{29,30,38,40,64}

Estimation of both nephron number and podocyte number per glomerulus allowed us to estimate the total number of podocytes in human kidneys. To our knowledge, this is the first report of total podocyte number in human kidneys. Median podocyte number per kidney in the 30 Japanese donors was 376 million, with values ranging 7.4-fold from 90 to 673 million. This surprisingly large variation in total podocyte number was the result of variations in both the number of nonsclerotic glomeruli per kidney (4.2-fold range) and the number of podocytes per glomerulus (2.7-fold range). The lowest total podocyte count per kidney was in a 67-year-old woman who had only 348,623 nonsclerotic glomeruli, and just 259 podocytes per tuft. In contrast, the highest total podocyte count per kidney was in a 42-year-old woman who had 1,162,663 nonsclerotic glomeruli and 579 podocytes per tuft. Knowledge of this wide variability in total podocyte number per kidney should be borne in mind when using urinalysis to assess podocyte loss or damage. For example, absolute values such as micrograms per milliliter of podocin protein or mRNA in urine could be expected to vary widely in healthy kidneys, simply due to the very different numbers of podocytes per kidney, and these values have little or nothing to do with current or subsequent glomerular disease.^16,65,66 For this reason, strategies that rely on ratio data to estimate podocyte stress (e.g., ratio of podocin to nephrin mRNAs) or detachment (e.g., podocin mRNA normalized to urine creatinine) are to be preferred over absolute values.^16,67

In this study, the number of nonsclerotic glomeruli per kidney declined at a rate of 8320 per year. This value is similar to previous estimates from autopsy studies.^29,30,64 In this study, total podocyte number per kidney declined at a rate of 5.63 million per year, with the majority (approximately 80%) of this loss due to glomerulosclerosis-associated nephron loss (4.31 million podocytes per year). A further 1.32 million podocytes (approximately 20% of lost podocytes) were lost in glomeruli that, in single histologic sections, appeared to be nonsclerotic. To our knowledge, this is the first report to differentiate between podocyte loss associated with glomerular loss due to glomerulosclerosis and loss of podocytes in glomeruli that appear to be nonsclerotic. Of course, these two forms of podocyte loss likely occur at different ends of a temporal continuum, with podocyte loss in apparently healthy glomeruli ultimately leading to destabilization of the glomerular tuft and glomerular loss.

Knowing the number of nonsclerotic glomeruli per kidney and podocyte number per nonsclerotic glomerulus enabled us to calculate the rate of podocyte loss per year in individual nonsclerotic glomeruli. Our estimate of 1.90 (IQR, 0.98–3.63) podocytes lost per nonsclerotic glomerulus per year is similar to the value of 2.3 podocytes per tuft per year for all age groups (birth to 85 years) that was previously reported.⁶⁶ The younger cohort in this study may explain the apparently lower rate of podocyte loss. However, it should be noted that our estimates of podocyte loss are determined on the basis of data obtained at a single time point from biopsy specimens obtained at the time of transplantation. The use of repeated biopsy samples to assess the decline in podocyte number per glomerulus with age may provide additional insights.

In this study, hypertension was associated with lower podocyte density, but not podocyte number per tuft. This finding is consistent with that of our previous autopsy study of White Americans, which showed that hypertension was associated with lower podocyte density, but not podocyte number per tuft.³⁶ In this study, we also found that hypertension was associated with larger podocyte volume. Previous animal studies reported that average podocyte volume increased with increasing glomerular volume, which was interpreted as podocyte hypertrophy to cover the amplified glomerular capillary surface.^17,68 To the best of our knowledge, there was no previous report regarding increased podocyte volume in hypertensive humans or animals. Of note, hypertension was a significant factor associated with increased podocyte volume, independent of age or glomerular volume in this study. A recent report showed that levels of urinary markers of podocyte stress and detachment were directly correlated with mean BP within the normal range in living-kidney donors.¹⁶ This finding, together with our present results, suggests that significant changes in podometrics occur in individuals who are prehypertensive, and in those who are hypertensive with controlled BP, before the development of apparent kidney diseases, because the subjects in both studies were living-kidney donors. However, because our present cohort included only five subjects who were hypertensive, studies on a larger cohort should be conducted to confirm these findings.

In conclusion, this is the first study to report both podometric data and nephron number in the same healthy human subjects. We studied healthy Japanese kidney donors who have a lower nephron number than most previously studied racial groups. Our findings show that podocyte density and podocyte number per tuft in Japanese individuals are similar to values reported in other races, and that total podocyte number per kidney, a parameter not previously reported in human kidneys, varies widely. We also found that the majority of age-related podocyte loss per kidney is associated with loss of glomeruli due to glomerulosclerosis, rather than podocyte loss in glomeruli that are nonsclerotic. Finally, hypertension was associated with lower podocyte density and larger podocyte volume, suggesting podometric changes occurred before the development of kidney disease. Our findings indicate the value of estimating both nephron and podocyte number in healthy kidney donors. We expect these approaches may be of even greater value in assessing nephron and podocyte loss in patients, and will thereby assist in guiding clinical care.

Disclosures

J.F. Bertram reports being a scientific advisor for, or member of, Kidney International. K. Haruhara reports receiving research funding from Japan Arteriosclerosis Prevention Fund, and reports receiving funds for study abroad from The Uehara Memorial Foundation. T. Yokoo reports being an editorial board member of Human Cell. All remaining authors have nothing to disclose.

Funding

This work was supported by Japan Society for the Promotion of Science grant JP16K0936 and the Japan Arteriosclerosis Prevention Fund.

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2020101486/-/DCSupplemental.

Supplemental Figure 1. Correlations between glomerular volume and podometrics in 30 Japanese living kidney donors.

Supplemental Table 1. Multiple linear regression analyses for nephron number and glomerular volume.

Supplemental Table 2. Multiple linear regression analyses for podocyte volume.