Hyponatremia, the phenomenon of serum sodium level falling below 135 mmol/L, is seen frequently in cancer patients and has been correlated with poor prognosis. Hyponatremia has classically been attributed to the “syndrome of inappropriate antidiuretic hormone secretion,” leading to prolonged fluid retention. However, this is unlikely to be the only mechanism. In this study, we advance the hypothesis that upregulation of various sodium-transporting proteins during the cancer process makes a significant contribution to the pathophysiology of cancer-associated hyponatremia. Such sodium-transporting proteins include voltage-gated sodium channels, especially its hypoxia-promoted persistent current, epithelial sodium channels, and transient receptor potential channels. Thus, hyponatremia follows cancer, whereby drop in blood serum level occurs as a result of uptake of sodium from extracellular fluid by cancer cells. Indeed, the sodium content of cancer cells/tissues is higher than normal. In turn, the rise in the intracellular sodium concentration brings about a range of cellular effects, including extracellular acidification that promotes invasiveness and thus leads to poor prognosis. This perspective offers novel therapies for cancer and the associated hyponatremia.
Introduction
Hyponatremia is defined as the condition when the concentration of sodium in blood falls below 135 mmol/L and, in severe cases, below 125 mmol/L. Although the condition has been found to accompany several carcinomas, including lung, pancreatic, prostate, liver, breast, and renal cancers, the cause(s) of cancer-associated hyponatremia (CAHN) remains largely unknown.1–3 From increasing independent evidence for functional involvement of Na+-transporting proteins in the cancer process, a novel perspective has emerged for the pathophysiology of CAHN.
The most common cause of hyponatremia, including CAHN, in all hospitalized patients, is thought to be the “syndrome of inappropriate antidiuretic hormone secretion” (SIADH).4 In healthy individuals, antidiuretic hormone (ADH) is released from the pituitary gland in response to increased plasma osmolality and/or reduced blood pressure. ADH promotes expression of aquaporin-2 in renal collecting ducts and thus enhances water reabsorption and reduces the plasma osmolality. In SIADH, the secretion of ADH is prolonged leading to continued reduction in the concentration of serum sodium by a total increase in body fluid. Consequently, some authors have defined hyponatremia essentially as a disorder of blood osmolality due to excess water retention.2 This is likely to be an oversimplification, however, and other mechanisms, for example, reduced glomerular filtration and enhanced proximal tubular reabsorption of filtrate, may be involved.5 In the event, the relative increase in the quantity of water would affect the concentrations of not just sodium in blood but all the ions (as well as metabolites etc.).
Indeed, some studies have found only limited (∼30%) association between SIADH and CAHN.4
The incidence of CAHN varies widely and is dependent on the type of cancer and its clinical setting. On the whole, the more aggressive a tumor is, the more severe tends to be the level of CAHN. Furthermore, the association between CAHN and SIADH extends to other characteristics of the cancer cells such as histological type.2,3 In fact, CAHN has been observed most consistently and studied most extensively in “neuroendocrine” small-cell lung cancer (SCLC) patients, of whom, up to 44% demonstrate CAHN.6 SCLC is a highly aggressive form of lung cancer that is likely to metastasize early (sometimes, even before an identifiable primary tumor forms), resulting in a median survival time of patients of only months after diagnosis.7
Given that SIADH may be responsible for only a limited number of cases of CAHN, a novel insight into the pathophysiology of CAHN is needed.2 In this study, from increasing independent evidence for functional expression of sodium-transporting proteins in cancer cells and tissues, we propose that there is the following triangular relationship: “cancer progression—sodium-transporting protein expression—CAHN” (Fig. 1A). Na+-permeating channels and transporters are well known to be significantly upregulated, even expressed de novo, in cancer cells and facilitate uptake of Na+ from the tumor microenvironment into the cancer tissue. Ultimately, this will shift the blood-tissue balance of Na+, lowering the extracellular concentration of Na+ and giving rise to hyponatremia (Fig. 1B).
Association of hyponatremia, cancer, and sodium-transporting proteins. (A) The proposed triangular relationship of hyponatremia, cancer, and sodium. Pairwise associations are well known to exist between the three components of the scheme, but it is here for the first time that such a triangular relationship has been proposed. (B) The basic sequence linking cancer to hyponatremia through Na+ transporter expression/activity.
This hypothesis has several prerequisites and makes a number of predictions. These are discussed in the following sections. In these discussions, generally, we have omitted data from patients receiving chemotherapy since this could complicate the relationship between hyponatremia and cancer due to the well-known side effects of chemotherapy. Such “side effects” would include (1) disruption of the endocrine system, (2) kidney damage, and (3) the fact that Na+ (and K+) would be “dumped” from necrotic dying cells into extracellular space as a possible secondary factor.3,8
Sodium-Transporting Proteins Are Expressed Extensively in Cancer Cells
Plasma membrane ion channels are involved in all major hallmarks of cancer, which has even been described as a “oncochannelopathy.”9 In fact, there are a number of proteins that transport Na+ into cells and these are frequently upregulated in cancer (Fig. 2). On the one hand, primarily Na+-permeant ion channels (see below) facilitate Na+ influx into cancer cells. In particular, functional expression of Na+ channels is associated with metastatic progression.10,11 On the other, proteins that function as “exchangers” use the Na+ electrochemical gradient as an energy source for secondary transport of another ion or metabolite. The activities of both types of transport protein would raise the level of intracellular Na+. In addition, serum Na+ concentration may be regulated indirectly by aquaporins (AQPs), which are also expressed in cancer cells and enhance their aggressive behavior.12 The main cellular Na+ extrusion mechanism, the Na+/K+-ATPase (NKA), has a complex involvement in cancer.13 AQPs and NKA are outside the scope of the current perspective.
A schematic cell showing sodium-transporting proteins in the PM. Various sodium-transporting proteins are responsible for influx of Na+ into cells. These are exemplified by VGSC, ENaC, and Trp channels. The other set of sodium-transporting proteins are exchangers, exemplified in the scheme by the NHE1. ENaC, epithelial sodium channel; NHE1, Na+-H+ exchanger; PM, plasma membrane; Trp, transient receptor potential; VGSC, voltage-gated sodium channel.
Voltage-gated sodium channel
There is growing evidence implicating voltage-gated sodium channels (VGSCs) in cancer progression. Several carcinomas express functional VGSCs de novo.10,11 These include breast cancer, cervical cancer, colon cancer, melanoma, mesothelioma, neuroblastoma, non-SCLC, ovarian cancer, prostate cancer, gastric cancer, and SCLC.11 Recent in vitro work on human strongly metastatic breast cancer cells has shown (1) that elevating the extracellular Na+ concentration indeed leads to corresponding rises in the intracellular Na+ level and (2) that permeation through VGSCs plays a significant role in this process (Fig. 3).14
Data from MDA-MB-231 human breast cancer cells showing effects of raising the extracellular Na+ concentration on intracellular Na+. (A) Fluorescence images of the rise in intracellular Na+ (quantified as the 340/380 ratio). (B) Data quantified from (A). (C) Evidence that the rise in intracellular Na+ is by entry from outside and involves permeation through VGSC. Replacing extracellular Na+ with the large/impermeant cation NMDG or applying the VGSC blocker TTX significantly suppressed the rise in intracellular Na+. (D) Silencing the VGSC (Nav1.5) with shRNA had a similar inhibitory effect as the treatments in (C). Modified from Yang et al.14 NMDG, N-methy-D-glucamine; TTX, tetrodotoxin.
VGSCs are a multigene family and different subtype(s) tend to occur in different carcinomas.11 Cancer-associated VGSCs have two characteristics that could contribute to CAHN. First, where studied, the VGSCs have been found to be embryonic splice variants. This is seen clearly for breast and colon cancers where the dominant VGSC subtype is the neonatal splice variant of Nav1.5 (nNav1.5).11,15,16 Compared with their adult counterparts, these neonatal VGSCs have slower inactivation kinetics and would permeate relatively more Na+ into cells.17 Second, VGSCs generally respond to hypoxia, which occurs naturally in growing tumors, by increasing the amplitude of the “persistent current” (INaP).18 The amplitude of this is only ∼1% of the classic transient component (INaT), but unlike the millisecond duration of the latter, it can last hundreds of milliseconds–seconds and thus bring a significant (mM) amount of Na+ into cancer cells.19
Epithelial sodium channel
Epithelial sodium channels (ENaCs) are well known to occur in epithelia (especially kidney, bladder, lung, colon, skin, and sweat glands) where it regulates Na+ transport and water balance.20 ADH and aldosterone upregulate ENaC expression/activity.21 Indeed, aldosterone-induced increase in cell migration was abolished in cells treated with amiloride or antisense oligonucleotides directed against EnaC.22 Upregulation of ENaC has been seen in several carcinomas.23,24 Upregulation of ENaC has also been seen in a mouse model of colon cancer.25 Also, in glioblastoma, a common and aggressive form of brain tumor was found to express ENaC that is not present in normal astrocytes or low-grade gliomas.26 In such situations, elevated ENaC activity, again, would mediate increased Na+ influx into cells.27
Transient receptor potential channels
Transient receptor potential (Trp) channels comprise a large group of cationic channels that can bring substantial Na+ into cells, especially if expressed in certain heteromeric combinations.28 A variety of Trp channels are expressed/upregulated in cancer cells and subserve a wide range of cellular behaviors, from molecular differentiation to whole-cell behaviors.29,30 These channels and associated cancers include TRPA1 (SCLC), TRPM8 (e.g., prostate and pancreas), TRPV6 (e.g., prostate, breast, thyroid, colon, and ovarian), TRPM1 (e.g., melanoma), and TRPV1 (e.g., prostate, bladder, colon, and pancreas).
Sodium leak channel
Sodium leak channel (NALCN) is a voltage-independent Na+ channel that is expressed in some cancers, including SCLC and pancreatic cancer.31,32 Most work has focused on mutations of its gene, as in non-SCLC.32 It is not yet clear how these mutations may affect the Na+ permeation characteristics of the channel and/or its contribution to the membrane potential that may impact other Na+-transporting proteins. Nevertheless, as a “leak” channel, NALCN activity can significantly raise tissue Na+ concentration.33
Sodium-dependent transporters
A number of transporters utilize the Na+ electrochemical gradient to co-transport other ions and metabolites (e.g., amino acids) in and out of cells, including importantly, cancer cells. Such co-transported chemical entities include H+ (out), Ca2+ (out), HCO3− (in), glutamate (in), glucose (in), and monocarboxylates (in). These co-transporters are also often upregulated during carcinogenesis.34 Again, the upregulation would raise the level of intracellular Na+ indirectly, and thus could contribute to depletion of extracellular Na+.
Sodium Content of Tumor Cells and Tissues
Our hypothesis would predict that the Na+ content, including the intracellular Na+ concentration ([Na+]i), of tumor cells and tissues should be higher than normal. This is generally the case (for recent reviews, see Djamgoz et al.11 and Leslie et al.35). In human non-SCLC cells in vitro, [Na+]i was indeed found to be significantly (∼100%) higher than normal lung cells (but the in vivo situation has not been studied).36 In rodent models, also, [Na+]i in cancerous tissues in vivo was similarly higher than corresponding normal tissues.37,38 Using X-ray microanalysis, Smith et al. showed that [Na+]i in cancerous liver was 159% higher with normal hepatocytes in young adult mice.38 This result was confirmed and extended to mammary adenocarcinomas where [Na+]i was found to be 77–334% higher than normal mammary epithelium in both mice and rats.37,39 These results complement the earlier observations on human glial and glioma cells.40 In more recent clinical work, tissue sodium concentrations of human brain, breast, and ovarian cancers have been measured noninvasively using 23Na-MRI and, compared with surrounding tissues, increases of 50–60% were observed (Fig. 4).41–44 In agreement with these findings, Barrett et al. used “magnetic resonance spectroscopic imaging” to show that human prostate cancer tissue Na+ levels were significantly higher than in adjacent normal tissue.45
Accumulation of sodium in breast tissue during cancer initiation and progression, measured with 23Na-magnetic resonance imagining. Images from three different breast cancer patients are shown. (A) A benign lesion (proliferative fibrocystosis), indicated by the arrow, aligned from gadolinium-enhanced image. (B) Infiltrating poorly differentiated ductal carcinoma (outlined in blue). Below (indicated by the green arrow) is a region with edema. (C) A large locally advanced breast cancer (outline in middle blue). The arrows indicate positioning landmarks. The intensity scale on the far right indicates approximate sodium concentrations (relative). Modified from Ouwerkerk et al.42
In conclusion, in accordance with our hypothesis, the available in vitro, animal and human studies are highly consistent is showing that Na+ content of tumors is elevated relative to corresponding normal tissues.
Elevated [Na+]i Can Drive Cancer Progression
There is mechanistic information suggesting that the raised level of [Na+]i in cancer cells makes a significant contribution to the cancer process. Invasiveness can be promoted by elevated [Na+]i through the activities of Na+-H+ exchanger (NHE1) and/or Na+-Ca2+ exchanger (NCX). The former is well known to be upregulated in a wide range of cancers (for a review, see Stock and Pedersen46). Overall, the activity of NHE1 acidifies the pericellular space, thereby initiating the proteolysis necessary for migration of the invading cells out of their primary location.46 Increased [Na+]i (especially by INaP under the hypoxic conditions in growing tumors) would slow down or even reverse NCX activity, thereby raising the level of intracellular Ca2+, which could promote various components of the cancer process from gene expression to cytoskeletal remodeling, motility, and secretion. Indeed, the available evidence suggests that NCX expression/activity is increased in cancer cells.47 Furthermore, epithelial-mesenchymal transition (EMT) that classically precedes the invasiveness is driven by VGSC activity.48
Hyponatremia Would Precede Metastatic Progression
The hypothesis presented in this study would predict implicitly that hyponatremia would closely precede metastasis, since it has been proposed that it is the various Na+-transporting proteins that are expressed during cancer progression that gradually deplete blood Na+. It has indeed been noted that hyponatremia can precede tumor diagnosis.2 Nevertheless, the evidence overall that hyponatremia precedes metastatic progression is not clear cut. One reason is that studies of CAHN are often carried out on hospitalized cancer patients undergoing treatment. Consequently, with differing conditions of patients and differing treatment regimens and so on, there are too many variables to confidently conclude either way that CAHN precedes or follows progressive cancer (or both). Furthermore, the time taken to diagnose metastatic disease may make it look as if hyponatremia follows it. Nevertheless, mechanistically, as well known, hyponatremia would precede the invasive stage of cancer. Indeed, House et al. showed for human colon cancer cells that expression of the VGSC (SCN5A) is upstream of a set of canonical genes that promote invasiveness, including Wnt, MAPK, calcium signaling and protease activity, and membrane remodelling.49 Also, as already noted, EMT, one of the earliest events in invasiveness, is controlled by VGSC activity.48
Hyponatremia, Cancer Progression, and Patient Prognosis
Reduced rates of survival of cancer patients with CAHN have been observed. Several studies have followed the survival of SCLC patients, comparing the prognosis of those with and without hyponatremia. One such study monitored 453 SCLC cases in Denmark over a period of 10 years.50 It was thus discovered that the survival time for SCLC patients with hyponatremia was some 50% shorter than those without. Similar findings were reported earlier for patients of non-SCLC, renal cell carcinoma, gastric cancer, and non-Hodgkin's lymphoma.6 A study of 3,886 patients found that patients with hematologic malignancies tended to have mild hyponatremia, whereas moderate and severe cases of hyponatremia occurred in carcinoma patients.1 In all cases, the length of stay in hospital and mortality were significantly higher in hyponatremic patients compared to eunatremic. A higher incidence of hyponatremia was found in cancer patients than those hospitalized for general medical conditions, and patients whose hyponatremia normalized had better prognosis than those who did not.1
Concluding Remarks: Clinical Implications and Future Perspectives
In overall conclusion, Na+-transporting proteins expressed/upregulated during the cancer progress can make a contribution to CAHN. However, the pathophysiology of CAHN is likely to involve additional mechanisms dependent upon the type and/or stage of cancer. Furthermore, it is unlikely that the ca. 10 mM loss of Na+ in blood can simply be “packed” into a solid tumor of finite size. Nevertheless, in principle, our perspective raises novel clinical possibilities. As regard to diagnosis, hyponatremia is already recognized as a predictor of clinical outcome. Thus, serum level of Na+ level may provide a convenient, progressive measure for diagnosis of cancers with metastatic potential. However, the serum Na+ level is already high (ca. 145 mM), so hyponatremia-associated changes may not be readily detectable, at least at early stages of the cancer. Interestingly, it was the serum level of K+ rather than Na+ that was found to be linearly correlated with risk of cancers of pancreas, stomach, lung, and prostate.51 As regard to Na+, 23Na-MRI would be more direct and sensitive than serum-based measurements.35,37 In fact, 23Na-MRI has already been proposed both as an early diagnostic tool and for monitoring treatment efficacy.52–54 These approaches can help predict the possible course of the disease especially, as regard to metastatic progression, and thus help optimize treatment. There are also several therapeutic possibilities. VGSC blockers may serve as potential antimetastatic drugs.10,11,15 In particular, the hypoxia-driven persistent current of the VGSCs can be blocked by drugs like ranolazine without significantly affecting the transient component INaT, which is essential for normal nerve and muscle function.15 Since the hypoxia-driven persistent current is the possible main carrier of sodium into tumors, its blockers are also potential antihyponatremia agents. The ENaC/NHE1 blocker amiloride is also a potential anticancer agent.55,56 Indeed, a recent study argued for metastatic BRAF V600-mutant melanoma that BRAF/MEK inhibitors lead to hyponatremia (including water retention), in part, by activating ENaCs.57 Furthermore, it may be possible to combine inhibitors of sodium-transporting proteins with other drugs to generate novel combination therapies.58 These may be better options than attempting to counteract hyponatremia by giving patients excess Na+.
Finally, we should note that there are other electrolyte disorders that accompany the cancer process, including hypokalemia, hypomagnesemia, and hypocalcemia, affecting serum levels of potassium, magnesium, and calcium.3 Equally, there are several other ion transport mechanisms (channels, transporters, and pumps) involved in cancer.9,34 It would be interesting, therefore, to determine the possible functional association of these pathologies. Such investigations could both improve our understanding of the cancer process and generate novel treatment modalities.
Post a Comment