reply: In aging male Wistar rats, renal injury develops spontaneously (2). It is gratifying that Dr. Schreuder is intrigued by our recent finding of a protective effect of early postnatal protein restriction (i.e., during lactation) on the spontaneous development of renal injury in male Wistar rats followed up to 12 mo of age (10). In his Letter to the Editor, in this issue of AJP-Regulatory, Integrative and Comparative Physiology, Dr. Schreuder (7a) postulates that protein restriction during lactation should result in a lower glomerular number, and presumably increased glomerular volume. This would conform to a finding documented recently by Dr. Schreuder's group in 75-day-old rats that were food restricted by doubling litter size directly after birth. Thus, these pups had a low calorie and protein intake. They found that the glomerular number was 25% lower and the glomerular volume 35% higher in the food-restricted group (8).
Intrauterine protein or food restriction usually reduces glomerular number and increases albuminuria [reviewed by Schreuder et al. (9)]. However, the setup in our experiment was totally different. We cross-fostered male pups born to control dams to dams on low-protein (8%) diets during lactation. Control pups were cross-fostered to dams on normal protein (20%) diets. Note that the diets were isocaloric (10). All offspring were weaned onto normal chow and studied at 3 and 12 mo of age. This approach allowed us to study the isolated effect of postnatal protein restriction on adult renal function and injury. We also measured telomere length and antioxidant enzyme expression.
Our main finding was less susceptibility to the normal process of renal injury that accompanies aging in many laboratory rat strains (5), including the Wistar rats that we used. Reduced albuminuria was accompanied by longer telomeres and increased protein levels of antioxidant enzymes. Of course, these data do not allow us to dissect cause and effect. However, they do extend the well-known protective effect of protein restriction on the progression of renal disease (4) into an earlier phase of development than that studied previously.
Dr. Schreuder, by extrapolation from observations in studies utilizing prenatal dietary or protein restriction or postnatal calorie restriction suggests that our intervention should have reduced nephron number (7a). This would be unexpected in the presence of less renal injury according to the hyperfiltration hypothesis of Zandi-Nejad et al. (12). This is indeed an interesting question to which we have now attempted to provide an answer.
Although our samples were not collected under ideal conditions for glomerular morphometry and counting (3), we measured glomerular dimensions and assessed glomerular density in a single section. Paraffin sections were stained with periodic acid-Schiff. Glomeruli were counted at magnification ×40 by applying a grid on randomly chosen fields and expressed as number of glomeruli/mm3 (n), calculated by the formula n = G/(F × A × (D + T)), where G is the number glomeruli counted in 400 fields, F is the number of fields counted, A is the grid area, D is the average glomerular tuft diameter, and T is the section thickness (0.003 mm) (6). Tuft diameter was assessed by tracing tuft edge and assuming a spherical shape.
The results, listed in Table 1 are quite clear. Glomeruli in the protein-restricted group were smaller, and glomerular density was about 30% higher. However, because kidneys in the protein-restricted group weighed about 23% less, the estimated glomerular number will not be significantly different. On the other hand, by 12 mo, body weight of the postnatal low-protein rats was also ∼21% less. Thus, the estimated relative nephron number was indeed increased. Of course, we acknowledge that this extrapolation is uncertain. First, the absolute product of glomerular density and kidney weight will grossly overestimate nephron number because kidney weight included the medulla. Second, absolute glomerular numbers can only be assessed using the dissection-fractionater technique. Finally, in the absence of standardized perfusion conditions, absolute morphometry is not reliable (7). However, these limitations apply to both groups to the same extent, because all kidneys were collected in an identical fashion.
Thus, contrary to the expectation of Dr. Schreuder, the estimated nephron number of the protein-restricted group was not lower than in the control group. Furthermore, the lower glomerular diameter does not suggest glomerular hypertension. In fact, the differences were in the opposite direction. The latter may well be secondary to more injury-related nephron loss in the control group. Thus, ideally nephron number should be assessed prior to the development of renal injury. N-acetyl-glucosaminidase excretion was already higher in the control rats at 1 mo of age, suggesting that development of injury starts at a very young age. Thus, for nephron counting in our model, kidneys should ideally be sampled at 21 days of age, i.e., at weaning, directly after exposure to low-protein intake, be perfusion-fixed, and be analyzed with state of the art technology. These samples are currently not available but will be collected in a future study.
We wish to point out that a recent study directed at lactational environment challenges the low nephron number hypothesis (11). A dissociation of blood pressure and nephron number was also documented by using cross-breeding studies in spontaneously hypertensive rats and normotensive Wistar-Kyoto rats (3). Both studies utilized state of the art counting technology. Dr. Bagby recently discussed the issue of causality in relation to the low nephron number hypothesis (1).
- Copyright © 2008 the American Physiological Society