*4.2. Pancreatic Cells*

Of the various elements, arsenic is possibly the most studied with regard to its effects on pancreatic function and the subsequent development of diabetes. Directly comparing arsenic, manganese, and Cd actions, the reduction in glucose-stimulated insulin section after arsenic and manganese exposure appeared to be due to mitochondrial dysfunction. In contrast, the inhibitory effects observed following Cd exposure were due to a mitochondria-independent mechanism [9]. There is evidence that mitochondria can transport Cd via the calcium uniporter, resulting in interference of the K+/H+ exchanger [93]. Nearly forty years ago, a transport-specific deficiency was identified due to changes in Cd sensitivity encoded by the *SLC39A8* gene [94]. The transporter encoded by this gene is highly conserved across species and has been shown to encode a specific element cation transporter referred to

as *ZIP8*. Examination of β-cell function after arsenic, manganese, and Cd exposure revealed a distinct pattern of miRNA expression changes unique to each element, suggesting that biochemical di fferences result in distinct responses. Exposure to inorganic arsenic gave rise to a significant 76% increase in miR-146a, while there was a 60% decrease in miR-195 expression following Cd exposure [95]. Since β-cells are excitable cells and involve cell depolarization, leading to insulin release, several studies have examined the e ffects of metal exposure on voltage-gated calcium channel function. In an early study, ex vivo, isolated, perfused rat pancreas demonstrated that the addition of Cd to the perfusion bu ffer reduced insulin release, possibly due to the blockade of calcium uptake, thus preventing β-cell depolarization [96]. Characterization of the calcium channels suggests that the L-type (long-lasting) calcium channel is the predominant channel on β-cells, and that Cd acts by preventing calcium uptake via L-type channels [97]. Voltage-dependent calcium uptake is dependent on β-cell depolarization via potassium channel activity. Cd-mediated e ffects on potassium channels appear negligible and are mediated by voltage-dependent calcium channel inhibition [98]. In addition to blocking calcium uptake, Cd itself can be transported into β-cells and accumulate within the cells. Interestingly, at relatively low concentrations (5 μM), in the absence of calcium applied for one hour, the non-stimulated insulin release from islets of obese–hyperglycemic mice increased, but not the glucose-stimulated one [99]. Changes in insulin release in the presence of 5 μM Cd appears to be independent of calcium involvement. Lower concentrations of Cd do not increase intracellular calcium, nor inositol 1,4,5-triphosphate (IP3) [100]. When 1 h-applied Cd concentrations exceed approximately 160 μM, insulin release is inhibited [99]. One conclusion from earlier studies is that Cd acts like a "silent killer", being taken up into β-cells and accumulating with time. During the early stages, β-cells would function normally until the Cd detoxification systems are overcome. Cd content from normal, "non-diabetic" human β-cells is approximately 29 nmol/g protein, with significant variability between individuals; that is not high enough to impair normal pancreatic function, as indicated by the lack of diabetic symptomology [43]. Determining intracellular Cd concentration and correlating intracellular values to extracellular concentrations have been challenging. Mathematical modeling data obtained from intestinal cell lines suggests that an external concentration of 10 μM would lead to an intracellular Cd concentration of 5000 amol/cell [101]. Cd continuously accumulates over time, at concentrations that may not significantly alter cell viability or gene expression [43]. Changes in β-cell function at sub-lethal concentrations seem to involve mitochondrial adaptation. As Cd accumulates in β-cells to several hundred-fold over baseline concentrations, mitochondria begin to appear fragmented, with the fusion–fission state shifting towards fission [102]. Studies using the INS-1 human pancreatic β-cell line utilized Cd concentrations 10-fold below the threshold necessary for cell death. Concentrations of Cd that are subtoxic produced no e ffects on mitochondrial function that were assessed by the energy change and the synthesis of adenosine triphosphate (ATP). Yet there were no morphological changes, suggesting a mitochondrial adaptive response to low-level Cd. The authors concluded that if cellular Cd influx continues, impairment of this organelle may contribute to cellular dysfunction and decreased viability of β-cells [102].

Mitochondrial respiration (oxygen consumption) and energy state (adenosine triphosphate production) appear unchanged during this process, until the cells commit to the death pathways [93]. Disruptions of mitochondrial morphology and energy state are linked to the onset of apoptosis. The intracellular mechanisms associated with Cd-mediated apoptosis have not been completely elucidated. The intracellular apoptotic mechanisms usually co-exist with necrosis, with a proportion of each depending on the cadmium dose and other conditions [102]. Exposure to Cd has been shown to elevate oxidative stress in pancreatic cells [103]. Increases in oxidative stress are linked to increased levels of malondialdehyde, free cytochrome c, p53, extracellular-regulated kinases 1/2, p38-mitogen-activated kinase, and *c-jun* N-terminal kinase (JNK), but decreases in mitochondrial membrane potential and Bcl-2 have also been observed [104]. Of the changes observed following Cd exposure, the increase in JNK activity by increased oxidative stress has been postulated to be one trigger for apoptosis. Not only is the phosphorylation of JNK upregulated after just one hour of exposure to 10 μM Cd, the expression of CCAAT-enhancer binding protein homologous protein, CHOP, is significantly upregulated [105]. Mitogen-activated kinases are vital for the normal function of the cell through proliferation, differentiation, and apoptosis. CHOP has been linked as an apoptotic response to oxidative stress in the endoplasmic reticulum. Additionally, Cd-mediated effects have been demonstrated directly via activation of the extracellular, signal-regulated kinases (ERK1/2) [106,107]. In general, ERKs activate numerous downstream pathways and are involved in Cd-induced carcinogenesis [108,109]. Together, reported effects of Cd on β-cells provide different pathways for Cd-mediated responses, ultimately leading to apoptosis, cell death, and a lack of β-cell responsiveness to elevated glucose, leading to stimulated insulin release.

#### **5. Insulin Resistance and Cadmium: Is There a Threshold?**

Considering Cd-mediated IR, the critical question is as follows: is there is a threshold for this effect? As recently reviewed [110], uncertainties in the mechanisms of low-level metal toxicity for humans and the demonstration of the existence of a safe threshold remains a rather challenging issue in toxicology. The comparable issue of whether endocrine disruption (ED) occurs at a threshold value has been holding for years, and the scientific community still expresses controversial views. The Endocrine Disrupters Expert Advisory Group (ED EAG) of the European Commission published a report in 2013 [111] summarizing experts' opinions regarding the existence of ED thresholds of adversity and the possibility of estimating such thresholds from existing experimental data. Most experts expressed the view that a threshold is likely to exist, but it might be exceptionally low. The basis of this argumen<sup>t</sup> is that one molecule, bound to only one receptor, would not be enough to activate the cascade of events needed to lead to apical adversity. However, in the case of fetal development, a threshold might not exist, due to the immaturity of the endocrine system [112]. Furthermore, with regard to the possibility of non-monotonic dose-response curves (NMDRCs), the available assays and methodologies most probably are not adequate for estimating a threshold with sufficient accuracy and sensitivity [113]. Other experts have expressed the view that a threshold does not exist, as endogenous hormones are already present in the body, and only one molecule of a xenobiotic might be enough to overwhelm the homeostatic system [114,115]. Based on these arguments, the EDs' risk assessment in most cases in the European Union is based on a hazard-based (no-threshold) approach, although a different, case-by-case approach for the Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) regulation has been proposed [116]. Other countries, such as the United States, Canada, and Australia, have adopted a risk-based threshold approach to overcome the difficulties discussed previously [117,118].

The estimation of a threshold under which Cd does not precipitate IR (further than the theoretical discussion of existence or not of such a threshold) would necessitate the existence of appropriate validated assays. For the moment, the globally existing recognized and accepted testing approaches cover modalities related only to the estrogen, androgen, thyroid, and steroidogenesis (EATS) pathways, as described in the OECD's Conceptual Framework (CF) for Testing and Assessment of Endocrine Disrupters [119]. Assays designed explicitly for the IR are not available in clinical practice or the field of toxicology. However, we know that IR, at least as a syndrome, generates obesity, hypertension, high glucose, triglycerides, and an increased LDL/HDL ratio in the blood. These endpoints should be determined using validated assays, like the repeated dose 90-day study (OECD Test Guidelines (TG) 408) and chronic toxicity and carcinogenicity studies (OECD TGs 451, 452, and 453).

In a regulatory context, for setting an experimental threshold for Cd causing IR, the "no observed adverse effect level" (NOAEL) or, better, a benchmark dose (BMD) should be determined. To set a BMD, it should be defined (a) if adversity is considered only as the apical effect(s) or even earlier signs of effect, and (b) the exact values for the various endpoints and conditions that compose the IR as a syndrome. In relation to the apical adversity, as mentioned above, a practical approach could be to consider as such the combination of obesity, hypertension, high glucose, triglycerides, and increased LDL/HDL ratio in blood, as in clinical practice. In vivo experiments would be necessary, and consequently specific

values that consider adverse e ffects should be set for animal studies (considering that an appropriate model exists). The available studies with rats and mice mentioned in the review of Tinkov et al. [84] have various shortcomings (such as very low duration, single dose used, use of engineered animals, and lack of adequate endpoints) to support the establishment of a NOAEL.

Additionally, the Cd mode of action or the related adverse outcome pathway should be elucidated, and adequate in vitro mechanistic assays should be developed and performed to prove causality. The exact cause(s) of IR is not ye<sup>t</sup> known. Still, there are data supporting perturbations of the enzymes and modulatory proteins involved in insulin signal transduction. Oxidative stress, inflammation, insulin receptor mutations, endoplasmic reticulum stress, and mitochondrial dysfunction a ffecting the insulin-dependent cells of skeletal muscle and adipocytes have been put forward [120,121]. The study of which of the above mechanisms are the most sensitive to Cd and the possible existence of NMDRC through appropriate methodology are essential steps for substantiating the role of Cd in IR, and eventually proposing a scientifically robust and reliable threshold if any can be validated.

#### **6. Conclusions and Remarks for Future**

Even though evidence abounds suggesting the damaging role of Cd on glucose homeostasis, the debate continues over the importance of Cd toxicity in the increasing occurrence of diabetes. Conversations are ongoing about the preventive or curative measures that should be taken within the susceptible populations. Cd's biological harm has been documented at di fferent levels, from cells to human populations. However, the ample literature on cadmium toxicity (more than 15,000 articles are referenced in PubMed) covers such a variety of experimental conditions and formats that it is di fficult to draw convincing strong conclusions from the available data.

The main topic of the present review, namely IR, is no exception. Cd doses, origins of cells, and detailed information on the exposed populations are a few of the numerous variables that change and impair comparison between studies from di fferent groups. At the time of writing this, we are still unable to propose a range of possible toxicity mechanisms, without knowing whether any level of exposure to environmental Cd can be accepted for human populations. Figure 1 briefly summarizes the role of Cd in impaired glucose metabolism in various organs.

Forthcoming studies should thus focus on the application of environmentally realistic doses in experimental studies since chronic low-level exposure of humans to Cd is seemingly inescapable. For its contribution to IR, parallel data would have to be obtained with optimized cellular models for each insulin target tissue. Meanwhile, cells and tissues involved in insulin turnover, secretion, and withdrawal would have to be similarly studied. The data at hand should now be enough to reach an international consensus on which concentration, (bio)chemical form, and duration of exposure should be implemented. In this process, negative results, i.e., a lack of observed e ffects as long as state-of-the-art methods are used, are as useful as positive ones. Since prospective population studies are precluded for obvious ethical reasons, and in the face of the skyrocketing diabetes development worldwide, it may be hoped that the above reductionist approach can define the most sensitive and useful markers of minute exposure for humans. This way, reliable monitoring of Cd-associated IR may be reached, and the degree of usefulness of preventive or corrective measures can be knowledgeably addressed.

**Figure 1.** Cadmium's (Cd's) role in insulin resistance and glucose tolerance/impaired glucose metabolism in various organs: summary of the results obtained from human, animal, and cell studies. Insulin resistance appears first, followed by impairment of glucose metabolism. Cd's effects in the kidneys do contribute to IR development to a certain point. In the pancreas, Cd accumulates in insulin-producing β-cells, leading to a decrease of insulin release and an increase in blood glucose in blood. Besides the direct effect of Cd on the pancreas, this toxic metal affects glucose transport in insulin-independent (liver) and insulin-dependent tissues (skeletal muscle and adipose tissue).

**Author Contributions:** The role for each of the authors is as follows: conceptualization, A.B. and D.W.; investigation, M.C., M.G., Z.B., and B.A.; writing—original draft preparation, A.B., D.Ð.- ´ C., J.-M.M., M.G., and ´ D.W.; writing—review and editing, A.B., J.-M.M., Z.B., and D.W.; visualization, D.Ð.-C.; supervision, A.B., Z.B., ´ and B.A., funding acquisition, A.B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the Ministry of Education, Science and Technological Development, Republic of Serbia (No. 451-03-68/2020-14/200161; A.B., D.Ð.-C., M. ´ C., Z.B., and B.A.). Oklahoma State University ´ Center for Health Sciences Pilot/Seed Grant (1-54357; D.W.), and the French Agence Nationale de la Recherche (ANR-13-CESA-008-Cadmidia; J.-M.M.).

**Conflicts of Interest:** The authors declare no conflict of interest. No funding source had a role in the design of the study; in the writing of the manuscript, or in the decision to publish the results.
