Associations between proteins and heavy metals in urine at low environmental exposures: Evidence of reverse causality

Abstract

Heavy metals can cause renal effects on vulnerable populations but it is uncertain whether these metals still pose health risks at the low exposure levels now prevailing in most industrialized countries.

In a cross-sectional study performed on 736 adolescents, we assessed the associations between the concentrations of cadmium and lead in blood and urine and the urinary concentrations of albumin and of low-molecular-weight (LMW) proteins, retinol-binding protein (RBP) and β₂-microglobulin. Multiple regression analyses were tested using urinary markers normalized to urinary creatinine or specific gravity.

Median metal concentrations were in blood (μg/L): lead, 15.1, cadmium, 0.18 and in urine (μg/g creatinine): cadmium, 0.09 and lead, 0.82. Multivariate analyses revealed significant associations in urine between RBP and cadmium as well as between β₂-microglobulin and lead whereas no associations were seen with metals in blood. These associations were completely abolished in subjects with increased urinary albumin, which may be explained by the competitive inhibition of LMW protein reabsorption by albumin.

Given the evidence that cadmium and lead circulate mainly bound to LMW proteins, these associations observed at low exposure might simply reflect the interindividual variations in the renal uptake of proteins sharing the same affinity for tubular binding sites.

Introduction

Cadmium (Cd) and lead (Pb) are among the most nephrotoxic agents (Nordberg et al., 2007). Like for most nephrotoxicants, one of the earliest signs of renal damage caused by these metals is an increased urinary excretion of proteins (Bernard, 2004). The determination in urine of albumin and of low-molecular-weight (LMW) proteins such as retinol-binding protein (RBP) or β₂-microglobulin (β₂m) is now used for screening defects in the glomerular filtration or in the tubular reabsorption of proteins in populations exposed to heavy metals (Bernard, 2004). These studies have clearly demonstrated that heavy metals exert their nephrotoxicity in a dose-dependent manner, the risk of renal dysfunction appearing only from a threshold of metal accumulation as reflected by the concentration of the metal in kidney, urine or in blood (Bernard, 2008a). In the case of Cd, studies in industrial workers and in populations living in highly polluted areas have derived thresholds of urinary Cd (U-Cd) for the increased LMW proteinuria in the range of 4–10 μg/g creatinine (Chaumont et al., 2011, Jarup et al., 1998). Similarly, threshold value for the renal effects of Pb has been estimated for blood Pb in the range of 300–400 μg/L (Cardenas et al., 1993).

More recently, studies conducted in the general population with low environmental exposure have reported increasingly lower threshold values for the renal effects of heavy metals, in particular for Cd (Buchet et al., 1990, Ferraro et al., 2010). Some of these studies focused on vulnerable populations such as children or the elderly, have even derived U-Cd threshold levels within the range of values now prevailing in most industrialized countries (Akesson et al., 2005, de Burbure et al., 2006). Assuming that these relationships are causal, this would suggest that a substantial fraction of the human population might be at risk of Cd nephrotoxicity. The causal nature of the associations found in these studies has however not been proven (Bernard, 2008a, Bernard, 2008b). A recent study suggests that these associations might be to a significant extent driven by factors unrelated to Cd toxicity such as a residual influence of diuresis, the renal impairment due to chronic smoking or the co-excretion of Cd with urinary proteins (Haddam et al., 2011). While confounding by diuresis or smoking can be easily overcome by further adjustment or stratification, the issue of metal–protein co-excretion is much more difficult to resolve. Indeed, the processes governing the renal handling of proteins, i.e. their glomerular filtration and tubular reabsorption, are basically the same for metal-transport proteins as for proteins used as biomarkers of metal toxicity.

This is particularly true for Cd, which circulates in plasma bound to metallothionein (MT), a LMW protein that follows the same glomerular filtration–tubular reabsorption pathway as RBP and other LMW proteins used for screening Cd-induced renal dysfunction (Nordberg, 1984). This co-excretion mechanism between MT and LMW proteins, which has been demonstrated in experimental animals (Bernard et al., 1987), is strongly supported by basic research. Studies that have explored the mechanisms of the tubular reabsorption of proteins have indeed demonstrated that proteins are removed from the tubular fluid by endocytosis mediated by two multiligand receptors megalin and cubilin. This endocytic complex, expressed in the proximal tubule, binds a large variety of proteins, including MT and the LMW proteins involved in the tubular proteinuria (Christensen et al., 2009).

The aim of this study was to address all these critical issues in order to assess the health significance of the associations between proteins and heavy metals in the urine of subjects with a low environmental exposure to Cd and Pb.