β2-Microglobulin

β2MG is the light chain of class I major histocompatibility complexes and is, therefore, found on most nucleated cells. Its molecular weight is 11,000 Daltons. The rate at which β2MG enters plasma ("influx") is relatively constant within and among healthy subjects, but it may rise in patients with chronic inflammatory conditions or hematologic malignancies [236].

β2MG is eliminated exclusively by the kidneys. A modest fraction of the amount removed is taken up from peritubular capillaries [237], but most elimination results from glomerular filtration, proximal tubular reabsorption, and intracellular degradation. At least 90% of the circulating protein is ultrafilterable [238,239], and 99.9% of the filtered load is ordinarily reabsorbed. When the GFR is normal, the equilibrium between influx and renal processing establishes a plasma concentration ([β2MG]p) between 1.2 and 2.7 mg/<sup>L</sup> [236]. As GFR falls, the filtrate is presented to proximal tubules at a rate that is absolutely reduced but normal or increased per surviving nephron. [β2MG]p rises secondarily, and equilibrium between the influx and the degradation of the protein is maintained [240–244].

In Cd research, it has been customary to declare that proximal tubular toxicity is present at <sup>E</sup>β2MG > 300 μg/g creatinine [41]. At an arbitrary <sup>E</sup>β2MG of 300 μg/d, Ecr of 1 g/d, GFR of 144 L/d (100 mL/min), and filterable [β2MG]p of 2.0 mg/L, fractional excretion of β2MG (FEβ2MG) is 0.1% and fractional reabsorption (FRβ2MG) is 99.9%. Doubling of <sup>E</sup>β2MG to 600 μg/g creatinine, a clearly elevated value, entails an increase in FEβ2MG from 0.1% to 0.2% and a reduction in FRβ2MG to 99.8%. Miniscule Cd-induced reductions in FRβ2MG, therefore, lead to substantial increments in <sup>E</sup>β2MG [245].

The sensitivity of <sup>E</sup>β2MG to slight reductions of FRβ2MG should not be interpreted as evidence that the underlying cellular injury is trivial. Values of ECd at which <sup>E</sup>β2MG exceeds 300μg/g creatinine are at least 10 times higher than in normal populations [246,247]. If ECd itself is a marker of toxicity, then the customary cutoff value of <sup>E</sup>β2MG is not a sensitive metric for detecting tubular injury. For pathophysiologic insight, <sup>E</sup>β2MG is most logically related to the normal *maximal* reabsorptive capacity for the protein—i.e., the tubular maximum (Tmβ2MG)—if such a Tm exists. Hall could not demonstrate one in dogs with an infusion of human β2MG [237], but in rats, Gauthier documented a Tmβ2MG when [β2MG]p was approximately four times the norm [238].

In theory, if a Tmβ2MG existed in humans, a decline in GFR might expose it. In this circumstance, surviving nephrons would be presented with a higher concentration of β2MG in less total filtrate volume, and a normal rate of presentation to a reduced nephron mass could exceed a putative Tmβ2MG. Multiple investigators have argued that this scenario occurs, but it is often possible that the disease lowering GFR has also lowered Tmβ2MG [241,244]. In patients with hepatorenal syndrome, in which the perfusion of normal kidneys is severely limited, a Tmβ2MG was not demonstrable despite extreme reductions in GFR and elevations in [β2MG]p [243]. Similarly, in children with glomerular disease exclusively, on biopsy, FEβ2MG did not correlate with GFR [244].

If some humans can reabsorb all filtered β2MG despite a low GFR and high [β2MG]p, then nephron loss is insufficient to explain excessive <sup>E</sup>β2MG in patients with Cd nephropathy. It appears that Cd imposes a Tmβ2MG or reduces one that already exists, and increased <sup>E</sup>β2MG indicates reduced β2MG reabsorption per nephron at any GFR [248]. Once Cd has established a Tmβ2MG, we expect <sup>E</sup>β2MG to rise substantially as GFR falls. Multiple investigators have documented this phenomenon [53,232,249], but none have quantified the individual contributions of GFR and Tmβ2MG to excessive <sup>E</sup>β2MG.

### Retinol-Binding Protein 4

RBP4, a small protein with molecular weight 21,000 Daltons, is synthesized in the liver. As its name implies, it binds to retinol (vitamin A) and transports the vitamin to tissues. Most RBP4 in plasma is bound to transthyretin (pre-albumin) in a complex that is too large to be filtered by normal glomeruli, but a minor fraction is unbound and readily filtered [239,250]. The total plasma concentration of RBP4 ([RBP]p) is normally 40–60 μg/mL [251].

As is the case with β2MG, over 99.9% of filtered RBP is normally reabsorbed. Urinary excretion (ERBP) increases markedly in tubulointerstitial disease, and it also rises substantially as GFR falls [233,252]. We are unable to ascertain whether nephron loss per se raises free [RBP]p to a threshold at which the reabsorptive capacity of surviving nephrons is exceeded. Bernard and colleagues reported a threshold [RBP]p of 25 mg/<sup>L</sup> at which [RBP]u rose acutely, but all of the patients with CKD had tubulopathies [253]. Mason and colleagues described a similar finding at high rates of Cd excretion, but we presume that their subjects had sustained Cd-induced tubular injury [233].

RBP4 differs in some respects from β2MG. β2MG is a positive acute-phase reactant (APR), and its plasma concentration rises in association with inflammation. RBP4 is a negative APR, and its concentration falls with inflammation [250]. Whereas β2MG is unstable at pH < 5.5, RBP4 is stable at any physiologic urine pH. Because of this attribute, some have argued that RBP4 should be the reabsorptive marker of choice in studies of Cd tubulopathy [253].

#### 4.3.5. Normalization of Excretion Rates to Creatinine Excretion or Creatinine Clearance

The kidney is the final repository of assimilated Cd and the principal site of persistent toxicity [254]. To study that toxicity, it is reasonable to quantify excretion rates of relevant substances (abbreviated E*x* for a given substance *x*). In practice, however, urine aliquots are more conveniently obtained than timed collections, and concentrations of *x* ([*x*]u) are measured instead of E*<sup>x</sup>*. To nullify the e ffect of urine volume on these concentrations, [*x*]u is usually normalized to the urine creatinine concentration ([cr]u) because volume a ffects [*x*]u and [cr]u proportionately.

This practice may lead to erroneous conclusions. As E*x* and Ecr are biologically unrelated, each excretion rate is influenced by at least one variable that does not a ffect the other. Ecr is determined primarily by muscle mass [255], which has no relationship to E*x*; consequently, in a physically diverse population, normalization of a given [*x*]u to [cr]u alters [*x*]u/[cr]u—in theory, by as much as fourfold—for a reason unrelated to E*x* [256]. Conversely, if substance *x* emanates from tubular cells, E*x* varies directly with the number of cells and the intracellular concentration of *x*, neither of which is related to Ecr. If nephron mass is normal, E*x* may rise as a consequence of cellular injury; if Cd destroys nephrons, E*x* may fall. As Ecr does not change importantly in either circumstance, [*x*]u/[cr]u may overstate tubular injury per nephron when GFR is normal and understate it when GFR is reduced [221,229].

To circumvent these issues, we recently introduced the practice of normalizing E*x* to creatinine clearance (Ccr), a surrogate for GFR, in studies of Cd nephrotoxicity [52,54]. Ccr is the excretion rate divided by the plasma concentration of creatinine; if Vu is the urine flow rate, Ccr = [cr]uVu/[cr]p and E*x*/Ccr = [*x*]uVu/([cr]uVu/[cr]p), which simplifies to [*x*]u[cr]p/[cr]u. Whereas the unit of [*x*]u/[cr]u is mass of *x* per mass of creatinine, the unit of E*x*/Ccr is mass of *x* excreted *per volume of filtrate*. Since Ccr varies directly with the number of nephrons, E*x*/Ccr also depicts excretion of *x* per intact nephron. As the formula for E*x*/Ccr includes the ratio [*x*]u/[cr]u, E*x*/Ccr, like [*x*]u/[cr]u, is una ffected by urine volume; since [cr]p rises with Ecr at a given Ccr, E*x*/Ccr is also una ffected by muscle mass. Most importantly, if substance *x* is released by tubular cells into urine, E*x*/Ccr prevents overstatement of injury per nephron at normal GFR and understatement at reduced GFR.

#### 4.3.6. A Pathophysiologic Synopsis of Cadmium Nephropathy

We recently published an analysis of cross-sectional data from Thai subjects living in areas of low, moderate, and high intensity of environmental Cd exposure. The patients were clinically well and were not hemodynamically predisposed to reductions in GFR [54]. In each subset and in the entire sample, we examined linear and quadratic regressions of eGFR on ECd/Ccr and ENAG/Ccr, regressions of ENAG/Ccr on ECd/Ccr, and regressions of <sup>E</sup>β2MG on ECd/Ccr and ENAG/Ccr. All regressions were statistically significant except those of <sup>E</sup>β2MG/Ccr on ECd/Ccr and ENAG/Ccr in the low-exposure subset. In general, e ffect size (standardized β) and coe fficients of determination (R2) rose with exposure intensity. A minority of subjects was found to have eGFR < 60 mL/min/1.73 m2; in the absence of renal hypoperfusion, which would have been accompanied by disqualifying signs and symptoms, the only plausible explanation for subnormal glomerular filtration was a reduction in the number of intact nephrons. "Nephron loss" is a widely used term to describe this state [257].

Our goals in the analysis of these data were to explain the correlation of ECd/Ccr with ENAG/Ccr, identify the source of excreted Cd, and elucidate the inverse relationship of eGFR to ECd/Ccr and ENAG/Ccr. Although we recognized that tubular injury might interfere with reabsorption of filtered CdMT, we doubted that *this interference* would lead to a statistically significant relationship of ENAG/Ccr to ECd/Ccr. As multiple lines of evidence suggested that excreted Cd, like NAG, emanates from proximal tubular cells (see Section 4.3.2), we reasoned that a common origin of the two substances would account for the relationship between ENAG and ECd.

Whereas we consider both ECd/Ccr and ENAG/Ccr to be parameters of cellular injury at the time of testing, eGFR reflects progressive nephron loss due to continuous accrual of Cd in proximal tubules. To explain why eGFR varied inversely with ECd/Ccr and ENAG/Ccr despite these temporal di fferences, we argued that all three parameters were either current or historical functions of the same intracellular Cd content. We concluded that Cd-induced injury had led to tubular cell death and a reduction of GFR. Relationships of eGFR to ECd/Ccr and ENAG/Ccr suggested that the severity of cellular injury had determined the extent of nephron loss.

In the literature on Cd toxicity, we occasionally encounter the concept that a reduction of GFR implies injury to glomeruli by the metal. Ample evidence suggests that this concept is both erroneous and unnecessary. CKD is a common sequela of ischemic acute tubular necrosis and numerous acute and chronic tubulointerstitial (TI) diseases that do not a ffect glomeruli [257–262]. Moreover, primary glomerular disease also leads to TI inflammation and fibrosis, presumably because reabsorbed, inappropriately filtered proteins are toxic to tubular cells [263]. Whether glomeruli or tubules are injured initially, the extent of TI fibrosis is the histologic finding that correlates best with GFR in CKD [264,265]. Possible filtration-reducing e ffects of TI fibrosis include the destruction of post-glomerular peritubular capillaries, amputation of glomeruli from tubules, and obstruction of nephrons with cellular debris [257,266].

We have not found English-language reports relating histopathology to GFR in asymptomatic humans exposed to Cd. However, in 61 autopsied subjects with itai-itai disease (IID), a syndrome of painful osteomalacia and proximal tubular dysfunction associated with severe Cd toxicity, Baba and colleagues showed that the most extreme osteomalacia was associated with the most advanced renal shrinkage [267]. In autopsies of 15 patients with IID, Yasuda and colleagues found that low kidney weight correlated with loss of tubules on microscopy; severely a fflicted kidneys showed interstitial fibrosis and widespread atrophy of tubular epithelium [268]. Although reduced kidney weight and disrupted cortical architecture sugges<sup>t</sup> that GFRs were reduced in these studies, neither Baba nor Yasuda provided relevant quantitative information. In contrast, Saito and colleagues performed extensive renal function studies (but no histopathology) in 13 patients with IID; endogenous creatinine clearance, a surrogate for GFR, was reduced in 12 [269]. Nogawa and associates measured serum creatinine concentrations in 4 of 5 patients with severe skeletal manifestations of IID, and the concentrations were substantially increased in each case [270]. Yasuda mentions that numerous Japanese patients with IID required chronic dialysis [268]; this choice of treatment suggests that extreme Cd toxicity reduced the GFR to levels that could not sustain life.

#### 4.3.7. Assessment of Cadmium Nephrotoxicity: Summary

Cd is a cumulative toxin to proximal tubular cells. Ample evidence suggests that the metal inflicts injury by promoting the creation of reactive oxygen species. The injury commences at a low intracellular concentration of Cd and intensifies as the concentration rises. Excretion of KIM1 is the first identifiable manifestation of toxicity, and NAG and Cd are subsequently released from injured or apoptotic cells. Inflammation and fibrosis follow, nephrons are lost, and GFR falls. After significant proximal tubular injury has occurred, reabsorption of small filtered proteins decreases and excretion of these proteins exceeds the normal limit. Once a Tmβ2MG is established, <sup>E</sup>β2MG rises rapidly as GFR falls.

ENAG and EKIM1 correlate with ECd. This observation and many others sugges<sup>t</sup> that Cd excretion results from the cellular release of the metal rather than filtration without reabsorption. If this conclusion is accepted, then increased ECd is itself a manifestation of Cd toxicity, and the concept of a threshold ECd at which <sup>E</sup>β2MG becomes excessive loses its pathophysiologic relevance. In subjects with low environmental exposure to Cd, eGFR is statistically related to ECd/Ccr when <sup>E</sup>β2MG/Ccr is not, and the relationship grows stronger as exposure increases. Markers of cellular injury at the time of testing, e.g., ENAG/Ccr and ECd/Ccr, correlate with eGFR, an indicator of historical injury, because all three parameters are determined by the intracellular concentration of Cd. Tubular injury inflicted by Cd is su fficient to explain reductions in GFR and progression of CKD.

#### **5. Environmental Exposure to Cd and Pb, Toxic Kidney Burden, CKD, and Other Common Ailments**

Environmental exposures are estimated to account for 70–90% of the risk of acquiring chronic ailments such as diabetes type 2, CKD and cancer [271,272]. The kidney is particularly at risk of injury from long-term use of therapeutic drugs and chronic exposure to environmental toxicants, especially when they are present in the diet [273–275]. The increased risk of kidney injury is attributed to its

large blood flow (20–25% of cardiac output) and exposure to high solute concentrations as the primary glomerular filtrate is concentrated [276]. In the following sections, we discuss cross-sectional studies that sugges<sup>t</sup> that Cd and Pb are synergistic CKD risk factors and longitudinal studies that implicate combined Cd and Pb exposure in enhanced mortality risk.

#### *5.1. The Increased Risk of CKD Associated with Cadmium and Lead Exposure*

CKD a fflicts 8% to 16% of the world population. Diabetes and hypertension are the most common risk factors universally, while obesity is an additional risk factor, especially in industrialized countries [272–279]. CKD is a cause of morbidity and mortality as it is an important predictor of end-stage kidney disease (ESKD), stroke and cardiovascular disease (CVD) [280–285]. CKD is characterized by albuminuria (a urinary albumin to creatinine ratio, uACR, above 30 μg/g) and/or a decrease of GFR to ≤60 mL/min/1.73 m<sup>2</sup> that persists for at least three months [48–50].

GFR is considered the best indicator of overall kidney function because it reflects the number of functioning nephrons at any given time [50]. In practice, the GFR is estimated from equations, notably, the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equations [47–49], and is reported as eGFR. The CKD-EPI equations, which have been validated using inulin clearance, are considered as the most accurate approximation of GFR [286]. CKD in its early stage is asymptomatic, and CKD staging is vital to evaluate nephron loss. Accordingly, CKD stages 1, 2, 3a, 3b, 4, and 5 correspond to eGFR of 90–119, 60–89, 45–59, 30−44, 15–29, and <15 mL/min/1.73 m2, respectively [48,287]. For simplicity, a low eGFR refers to an eGFR of <60 mL/min/1.73 m2, and albuminuria refers to uACRs above 30 μg/g.
