Tubular Reabsorption of Albumin

The proximal tubule is a fascinating structure where 66% of the glomerular filtrate is returned to blood together with practically all glucose, amino acids, and vitamins. Our understanding of the mechanisms behind this efficient and selective transport started with the discovery of megalin by Kerjaschki and Farquhar.1,2 It became clear later that the uptake of albumin; vitamins D, A, and B12; iron; and several other solutes, apart from megalin, also involve proteins such as cubilin and amnionless.3,4 The most likely story of tubular uptake has been unveiled in a series of beautiful experiments, some of which are based on genetic disorders in humans and mice.3,5–11 After endocytosis, the megalin–cubilin complex delivers its ligands to lysosomes, where all proteins are degraded and the amino acids and vitamins are returned to the circulation.12 These studies have been highly relevant clinically for several reasons: First, they shed light on the nephrotoxicity and ototoxicity of aminoglycosides.13–15 Uptake of myoglobin and hemoglobin is also mediated by megalin.16,17 Furthermore, the megalin–cubilin complex is recycled in a process requiring chloride channel 5.18 In this issue of JASN, Amsellem et al.19 present new and exciting data on the importance of cubilin using a conditional Cre-LoxP mouse model, resulting in 90% inactivation of tubular cubilin in the kidney. The authors used a similar technique to inactivate megalin and were able to breed some mice that lacked both proteins and lived to adulthood. Amsellem et al. report that cubilin requires amnionless for expression at the brush border and vice versa; cubilin is essential for tubular uptake of albumin, and megalin is needed for the endocytosis of the cubilin–albumin complex; and several ligands known to have high affinity for cubilin, such as apo-A-I, CC16, and transferrin, are actually taken up in mice lacking tubular cubilin. Several important conclusions can be drawn from this study,19 but I focus on one: its relevance for tubular handling of albumin. Human kidneys filter every day 180 L of primary urine with a solute composition similar to plasma. Normally, the final urine contains <30 mg/d albumin. From a theoretical point of view, abnormal proteinuria could be due either to a defective glomerular barrier that increases the filtration of albumin or to reduced tubular uptake of filtered protein. For decades, it was assumed that the cause of proteinuria would be found in the glomerulus because nephritic and nephrotic syndromes are characterized clinically by histologic changes in that structure. Indeed, discoveries during the past decade of several previously unknown components of the glomerular filtration barrier support this view. Today, we know of several proteins in podocytes or glomerular basement membrane that, when mutated, give rise to proteinuria.20 In addition, the surface glycocalyx of glomerular endothelium may also be required for an intact barrier,20,21 although there are no endothelial null phenotypes to prove this unambiguously. The glomerular barrier is highly size and charge selective according to most physiologic measurements, a finding compatible with studies at the molecular level.20 In contrast, other reports suggest that proteinuria in most kidney diseases is due to tubular defects, not glomerular.22–24 According to the latter point of view, the glomerular barrier is normally leaky, resulting in the filtration of >200 g/d albumin. Intact albumin is subsequently retrieved from the urine and returned to the blood according to this albumin-retrieval hypothesis.24 Several physiologic observations seem to refute the latter notion.20,25–28 In particular, Maunsbach29,30 in the mid-1960s found no significant uptake of intact albumin from urine to blood. In this issue of JASN, we now learn that albumin is bound to cubilin at the tubular cell brush border. The cubilin–albumin complex requires megalin to be internalized and transferred to lysosomes for degradation. Most important, inactivation of tubular cubilin, megalin, or both cubilin and megalin only increases the excretion of normal amounts of albumin in urine sixfold.19 Extrapolating from mice to humans suggests that human glomerular filtrate would contain no more than 200 mg/d albumin, suggesting a glomerular sieving coefficient of 10−5 to 10−4, in line with previous physiologic data.20 Please read the paper by Amsellem et al.19; it is an exciting story of cubilin and the tubular uptake of albumin. DISCLOSURES None.