The genetic renal Fanconi syndromes have attracted clinical and scientific interest out of all proportion to their clinical importance. Why is this? Probably the main reason is that individual patients have a fairly consistent “knockout” of a variety of renal tubular functions. Although there are large differences between patients even for the same form of the Fanconi syndrome, each patient is a relatively stable subject for rewarding study over several years.
What are the renal Fanconi syndromes?1 The first description of a Fanconi syndrome was probably by Abderhalden in 1903; this was a patient with cystinosis. Abderhalden's name has been eclipsed by those of Lignac, de Toni, Debré, and Fanconi and instead of this quartet of eponyms, the syndrome is now generally associated with Fanconi alone. Although there is burgeoning interest in several acquired forms of Fanconi syndrome, such as that associated with HIV disease and its treatment, it is the genetic forms which have attracted most research. The major primary Mendelian forms are cystinosis (specifically nephropathic cystinosis), the oculocerebrorenal syndrome of Lowe, and Dent Disease types 1 and 2.
The most consistent feature of these syndromes appears to be tubular proteinuria with variable expression of amino acid, glucose, phosphate, bicarbonate, calcium, uric acid, and potassium tubular leaks. Decreased glomerular filtration rate ensues with very variable age of onset and rate of progression, but is probably most severe in nephropathic cystinosis. There is abundant evidence that the two giant endocytic receptors on the tubular brush border, megalin and cubilin, function together to reabsorb proteins from the glomerular filtrate. Their functional loss is then likely to underlie the consistent finding of tubular proteinuria.
If that is true, what abnormalities of these two receptors might be present in cystinosis? Wilmer et al have conducted a careful study of a cystinotic kidney and have used a variety of clever controls to try to dissect information about pathogenesis of the proteinuria.2 Since this is an observational study, the number of experimental variables is limited and careful choice of controls is vital. Taking the lead from previous studies of Dent Disease type 1 and Lowe syndrome, Wilmer et al hypothesized that there might be deficient brush-border expression of megalin and cubilin and that this might also manifest itself as a reduction in these receptors in urine.3, 4, 5 In spite of a meticulous approach, they did not find either the renal or urinary abnormalities which were predicted.
To answer why this is, we need better to understand the previous observations of megalin deficiency in urine in Dent Disease type 1 in humans and in animal models.4, 5 The theory underlying these is that defective recycling of megalin and cubilin leads to reduced apical expression and that this source of urinary megalin and cubilin is then cut off or reduced.
As Wilmer et al point out, exceptions were immediately apparent: in a very rare family with the autosomal dominant form of Fanconi and tubular proteinuria, no deficiency of urinary megalin was found.5 So, Wilmer et al's work, as well as previous work, suggests that even if megalin and cubilin recycling are part of a final common pathway for protein reabsorption, failure of apical expression is not a necessary feature when this pathway is defective.
The difficulty with many of these experimental approaches is an attempt to understand the highly dynamic process of receptor and ligand recycling from snapshots of either kidney immunohistochemistry or urine composition. We are trying to determine the throughput of one or more cellular pathways from static observations. Positive observations, that is marked abnormalities, may be interpretable, albeit they could be epiphenomena. Negative observations, as made here and previously, are more difficult to interpret. One needs little imagination to see that a very defective recycling system may be compatible with only slight changes in steady-state levels of receptors at certain sites.
It may be important that most previous studies were not undertaken in patients or animals in end-stage renal failure. The nonspecific abnormalities, including interstitial fibrosis and cyst formation seen in end-stage Fanconi syndrome, will not ease interpretation. Did this confound Wilmer et al's findings? We do not know for certain but it is unlikely. After all, a reasonable number of the urine samples which they studied were obtained from patients well before end-stage disease. One might even hypothesize that previous reports of megalin deficiency in kidney from Dent type 1 serve as a historical “control” for Wilmer et al and strengthen the reliability of what they have found in cystinotic kidney and in their own controls.6
If the above, slightly depressing view is correct, what approaches are likely to be fruitful? Coauthors of the Wilmer et al paper and several others have already transformed our knowledge of these syndromes by transgenic experiments.3, 4 But even these have a nagging worry: does an accessible animal model of the human renal Fanconi syndrome really exist? How good are the rodent models? Undoubtedly the animal models have features of the human condition but there are also significant differences—hence the importance of Wilmer et al's study in humans. We still have little idea of the molecular basis of the variation of the tubular knockout among amino acids, bicarbonate, and so on within these syndromes. The rodent models themselves display many of these variations depending on the diet and the exact type of genetic modification.
New imaging techniques, still in their infancy and with their own problems of calibration and interpretation, offer tantalizing views of protein processing by the renal tubule.7 They offer views of pathways in animals but will they ever be applied to humans? Another approach is more directly molecular and is to gather a more precise understanding of megalin and cubilin cooperativity. We still do not know how these two receptors cooperate, even in a model cell-free system.
Although nephropathic cystinosis was first described over 100 years ago, it is unlikely we shall have to wait a further 100 years for a molecular interpretation of Wilmer et al's findings.
FULL TEXT
