Urinary proteins from stone formers promote calcium oxalate crystallization, growth and aggregation via oxidative modifications

Graphical abstract

Highlights

• •Normal urinary proteome inhibited CaOx crystallization, growth and aggregation.
• •Stone formers’ urinary proteome promoted the above CaOx crystal parameters.
• •Oxidatively modified proteins markedly increased (∼2.5-fold) in the stone formers’ urine.
• •Bioinformatics revealed a correlation between the increased proteins and oxidative stress.
• •ELISA confirmed the increased levels of oxidized proteins in the stone formers’ urine.

Abstract

Introduction

Various urinary parameters are used for determining kidney stone risk. However, almost all of the widely used lithogenic indices rely on urinary concentrations of small molecules/ions and pH.

Objective

To address whether urinary macromolecules (especially oxidatively modified proteins) also play a critical role in determining the stone risk.

Methods

Complexed urinary proteins (proteome) were purified from healthy individuals and calcium oxalate (CaOx) stone formers and performed various crystal assays and quantitative proteomics to compare them. Bioinformatic analyses were performed to gain additional insights, and the obtained data were verified by ELISA.

Results

While the normal urinary proteome inhibited CaOx stone-forming mechanisms (i.e., crystallization, growth and aggregation), the stone formers’ urinary proteome promoted all these CaOx crystal parameters. Descriptive proteomics by nanoLC-ESI-LTQ-Orbitrap-MS/MS analysis identified 203 and 381 proteins in the urine of healthy individuals and stone formers, respectively. Analyses of physicochemical properties revealed only molecular mass and isoelectric point that slightly increased in the stone formers’ urine, whereas instability index, grand average of hydrophathicity (GRAVY) and amino acid composition were comparable. Interestingly, proportion of oxidatively modified proteins (particularly those with methionine oxidation, methionine dioxidation and cysteine trioxidation) markedly increased (∼2.5-fold) in the stone formers’ urine. Quantitative proteomics revealed 89 increased and 56 decreased proteins in the stone formers’ urine. The oxidized proteins had a greater proportion (>3-fold) in the increased proteins (77 %) compared with the decreased ones (23 %), whereas the non-oxidized proteins showed comparable proportions (54 % and 46 %, respectively). Functional enrichment analyses revealed a correlation between the increased proteins and oxidative stress biological processes and molecular functions. Finally, ELISA confirmed the significantly increased levels of oxidized proteins in the stone formers’ urine compared with that of healthy individuals.

Conclusion

These data implicate that oxidatively modified proteome serves as a key pathogenic factor or risk for CaOx kidney stone formation.

Introduction

Currently, various arrays of clinical chemistry can be used for determining risk factors of kidney stone disease. Among these, fundamental measures commonly include urine pH, volume and concentrations of electrolytes and other small molecules, e.g., calcium, oxalate, citrate and phosphate [[1], [2], [3], [4]]. Abnormally low urine volume may exacerbate the supersaturation of the stone-forming substances, whereas urine pH serves as a critical determinant of urinary acidity or alkalinity that profoundly impacts the solubility of such substances [5,6]. Additionally, metabolic risk factors, such as hypercalciuria, hyperuricosuria and hyperoxaluria, are well known to promote metabolic stone formation [3,7]. Moreover, urinary citrate serves as a natural inhibitor of kidney stone formation, and its ratio to calcium affects the crystal dynamics [5]. These parameters are frequently used to formulate several popular indices for determining kidney stone risk, including the urinary relative supersaturation (RSS) index, the Tiselius Indices, the Robertson Risk Factor Algorithms (RRFA) and the BONN-Risk Index (BRI) [8].

Kidney stone disease has increasingly been recognized as a chronic condition associated with oxidative stress, inflammation and metabolic imbalances [9,10]. Apart from the aforementioned classical biochemical parameters, oxidative stress has emerged as a key contributor to kidney stone disease [11,12]. Reactive oxygen species (ROS) generated under oxidative stress conditions can induce renal epithelial injury, inflammation and tubular dysfunction, thereby facilitating stone formation [13,14]. In addition to small molecules, non-proteinuric urine contains hundreds (or even greater) of proteins with various molecular and biological functions [15,16]. Oxidative stress can induce alterations in excretion and functions of urinary proteins [[17], [18], [19]] that may influence the stone pathogenesis. Moreover, oxidative stress may affect the activities of urinary proteins, particularly their modulatory activities on calcium oxalate (CaOx) crystals [20,21].

We thus hypothesized that there is a highly oxidative state of the complexed urinary proteins (proteome) in stone formers that may also serve as a key determinant of the stone risk. Herein, we purified urinary proteomes from healthy individuals and calcium oxalate (CaOx) stone formers and performed various crystal assays and quantitative proteomics to compare them.

Materials & methods

Ethics statement

All experiments involving human subjects and clinical samples were approved by Siriraj Institutional Review Board (approval number Si415/2023) and were also conducted following the international guidelines, i.e., the Declaration of Helsinki, the Belmont Report, and ICH Good Clinical Practice with informed consent obtained from all participants.

Collection of urine samples and isolation of urinary proteome

Midstream random urine samples were collected from 40 healthy non-stone individuals (19 males and 21 females, aged 26.1 ± 2.9 years) and 37 CaOx kidney stone formers (patients) (23 males and 14 females; aged 44.9 ± 11.2 years). The healthy individuals had no recent evidence and history of the stone, had no other illnesses, and had not taken any medications within a month before urine collection. The stone formers were limited only to those with CaOx type, whereas those with other stone types were excluded. In addition, patients with renal tubular acidosis, primary hyperparathyroidism and urinary tract infections (by history, urine pH, urinalysis, acid loading test and/or blood chemistry) were also excluded.

After collection, all individual urine samples were immediately subjected to low-speed centrifugation (1,000g for 15 min) to remove cell debris and particulate matter. The clear samples were divided into two parts (5 ml/part for each subject) − one for pooling (at initial screening with crystal assays and mass spectrometric analysis) and the other for individual analysis (ELISA validation). To remove small molecules and to isolate urinary proteome, these pooled and individual urine samples were dialyzed against deionized water at 4 °C overnight, lyophilized and harvested.

Preparation of artificial urine (AU)

The AU was prepared using the AU-Siriraj formula [22,23]. Its components and concentrations were as follows (200 mM urea, 1 mM uric acid, 4 mM creatinine, 5 mM Na₃C₆H₅O₇·2H₂O, 54 mM NaCl, 30 mM KCl, 15 mM NH₄Cl, 3 mM CaCl₂·2H₂O, 2 mM MgSO₄·7H₂O, 2 mM NaHCO₃, 0.1 mM NaC₂O₄, 9 mM Na₂SO₄, 3.6 mM NaH₂PO₄·H₂O, and 0.4 mM Na₂HPO₄; pH = 6.2; specific gravity = 1.010 g/ml; and osmolality = 446 mOsm/kg) [22,23]. The AU was then used as a diluent of the lyophilized proteins to make their final concentration at 1 µg/µl for all the following crystal assays.

CaOx crystallization assay

Crystallization assay [24,25] was performed by adding 500 μl of 10 mM CaCl₂·2H₂O in crystallization buffer (10 mM Tris-HCl and 90 mM NaCl, pH 7.4) into each well of the 24-well plate. Thereafter, 4 µl of 1 µg/µl urinary proteome from healthy individuals or stone formers in AU or 4 µl of AU without proteins (control) was added into the well, followed by 500 μl of 1 mM Na₂C₂O₄ in crystallization buffer. The mixture was incubated at 25 °C for 1 h, and the CaOx crystals formed in each well were imaged under an Eclipse Ti-S inverted phase-contrast light microscope (Nikon; Tokyo, Japan). Crystal size was quantified from 100 crystals/sample, and crystal number was counted from 10 random fields/sample in each experiment using the NIS-Elements D V.4.11 software (Nikon). Crystal mass was then calculated as follows.

$$Crystalmass\left(\mu m^2/field\right)=Averagecrystalsizeineachfield\left(\mu m^2\right)\times Numberofcrystalsineachfield\left(/field\right)$$ (1)

CaOx growth assay

CaOx growth assay [26,27] was performed differently from the crystallization assay. At the initial step, CaOx crystals were allowed to form by mixing 500 μl of 10 mM CaCl₂·2H₂O with 500 μl of 1 mM Na₂C₂O₄ in crystallization buffer in each well of the 24-well plate. After 1-h incubation, the CaOx crystals completely formed. At this time-point (T₀), crystal images were taken under the Eclipse Ti-S inverted phase-contrast light microscope, and 4 µl of 1 µg/µl urinary proteome from healthy individuals or stone formers in AU or 4 µl of AU without proteins (control) was added into the well. The mixture was further incubated for another 60 min. At this latter time-point (T₆₀), crystal images were again taken. Crystal sizes at both T₀ and T₆₀ were quantified from 100 crystals/sample in each experiment using the NIS-Elements D V.4.11 software. Crystal growth, represented by Δ crystal size, and crystal growth inhibitory activity were calculated as follows.

$$\mathrm{\Delta }Crystalsize\left(\mu m^2\right)=Crystalsizeat{T}_{60}\left(\mu m^2\right)-Crystalsizeat{T}_0\left(\mu m^2\right)$$ (2)

$$Crystalgrowthinhibitoryactivity\left(\%\right)=\left[\left(\mathrm{\Delta }Crystalsizeofcontrol-\mathrm{\Delta }Crystalsizeoftest\right)/\mathrm{\Delta }Crystalsizeofcontrol\right]\times 100$$ (3)

Where the control was the condition with only AU (without proteins) and the test was the condition with the tested sample from either healthy individuals or stone formers. A positive value indicated the inhibitory activity, whereas a negative value indicated the promoting activity of the tested sample.

CaOx aggregation assay

Crystal aggregation assay was performed as previously described [28,29]. Briefly, CaOx crystals were allowed to form by mixing 10 ml of 10 mM CaCl₂·2H₂O with 10 ml of 1 mM Na₂C₂O₄ in crystallization buffer. After 1-h incubation, the crystals were harvested by 5-min centrifugation at 2,000g, washed with methanol, and air-dried. Thereafter, 1 ml of the 1,000 µg/ml crystal suspension in crystallization buffer was added into each well of the 6-well plate, followed by 4 µl of 1 µg/µl urinary proteome from healthy individuals or stone formers in AU or 4 µl of AU without proteins (control). The plate was put into a shaking incubator (Zhicheng; Shanghai, China) and incubated with continuous shaking at 150 rpm for 1 h. The crystals were then imaged under the Eclipse Ti-S inverted phase-contrast light microscope. Number of the crystal aggregates (“assembly of three or more individual COM crystals that tightly joined together” [28]) was counted from at least 15 random fields/sample in each experiment using the NIS-Elements D V.4.11 software.

In-solution tryptic digestion, descriptive proteomics and quantitative proteomics of urinary proteome by nanoLC-ESI-LTQ-Orbitrap-MS/MS analyses

An equal amount (50 µg) of urinary proteins from healthy individuals or stone formers was tryptic digested into peptides and subjected to nanoLC-ESI-LTQ-Orbitrap-MS/MS analyses as previously described [30,31]. More details are also provided in Supplementary Methods.

Analyses of physicochemical properties of the identified urinary proteins

All the proteins in urinary proteomes from healthy individuals and stone formers identified by descriptive proteomics were subjected to analyses of their physicochemical properties. These included molecular weight (MW), isoelectric point (pI), instability index, grand average of hydrophathicity (GRAVY) score and amino acid composition. In addition, the proteins with oxidatively modified residues, i.e., oxidation at methionine (Met, M) or cysteine (Cys, C), dioxidation at Met or Cys, and/or trioxidation at Cys were considered as the oxidatively modified proteins. Proteins with these various physicochemical properties and oxidative modifications were counted and compared between healthy individuals and stone formers.

Analyses of biological processes and molecular functions of the differentially excreted urinary proteins

The differentially excreted urinary proteins (those with significantly different levels) identified by quantitative proteomics were subjected to analyses of gene ontology (GO) biological processes and molecular functions using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) tool (https://david.ncifcrf.gov). Only the GO terms deemed significant at a p-value < 0.05 were further analyzed using the Reduce and Visualize Gene Ontology (REVIGO) tool (https://revigo.irb.hr). The latter was used for clustering the data, eliminating redundancies and highlighting the representative or essential GO terms with a dispensability value < 0.2 in a cluster.

Validation of differential levels of the oxidized urinary proteins

ELISA was performed to measure levels of oxidized urinary proteins in individual urine samples obtained from all healthy subjects and stone formers. Briefly, each well of the 96-well ELISA plate (Nunc; Roskilde, Denmark) was coated with an equal amount of urinary proteins from each subject (15 μg/subject/well) at 4 °C overnight. Non-specific bindings were blocked with 5 % bovine serum albumin (BSA) in PBS at 25 °C for 2 h, and the plate was washed three times with 0.05 % Tween-20 (Sigma-Aldrich; St. Louis, MO) in PBS. Thereafter, the proteins were derivatized by adding 2, 4-dinitrophenylhydrazine (DNPH) (Chemicon; Temecula, CA) in each well, whereas the paired non-derivatized samples were used as the negative control to subtract the background signal. The plate was washed three times with 0.05 % Tween-20/PBS and then incubated with rabbit polyclonal anti-DNP antibody (Chemicon) (1:100 in 0.1 % BSA/PBS) for 1 h. The plate was then washed three times with 0.05 % Tween-20/PBS and incubated with swine anti-rabbit IgG conjugated with horseradish peroxidase (Dako; Glostrup, Denmark) (1:1,000 in 0.1 % BSA/PBS) for 1 h. After five washes with 0.05 % Tween-20/PBS, colorization was developed by adding a chromogenic substrate (1.6 mM ortho-phenylenediamine dihydrochloride) (Sigma-Aldrich) in 35 mM citric acid and 0.012 % H₂O₂ (pH 5.5). After 15-min incubation, the colorized reaction was quantified by measuring the absorbance at λ492 nm using an ELISA microplate reader (Biochrom EZ Read 400 ELISA Plus) (Biochrom Ltd.; Cambridge, UK).

Statistical analyses

All quantitative data were derived from three independent experiments and are reported as mean ± SD, unless stated otherwise. Multiple comparisons were performed by ANOVA with Tukey’s post-hoc test, whereas comparisons between two groups were performed by Student’s t-test. P-values less than 0.05 were considered statistically significant.

Results

Effects of urinary proteomes from healthy individuals and stone formers on CaOx crystal formation

The crystallization assay was employed to examine the effects of urinary proteomes isolated from healthy individuals and stone formers on CaOx crystal formation. Microscopic analysis and size measurement of CaOx crystals showed differential patterns of crystal size distribution among the three conditions, i.e., experimental control condition without any proteins, tested condition with urinary proteome from healthy individuals and tested condition with urinary proteome from stone formers (Fig. 1A and B). When the median of the data from the control condition was used as a reference point to discriminate the potential modulatory activity of the tested samples, all of the crystal sizes in the tested condition with urinary proteome from healthy individuals were smaller than this reference value, suggesting its inhibitory activity against CaOx formation (Fig. 1A and B). By contrast, almost all crystal sizes obtained from the tested condition with urinary proteome from stone formers were larger than the reference value, suggesting that it tended to promote CaOx crystal formation (Fig. 1A and B). In concordance, statistical analysis of quantitative comparisons revealed a marked decrease and increase in crystal sizes by urinary proteomes derived from healthy individuals and stone formers, respectively, as compared with the control condition (Fig. 1C). Although crystal number was markedly decreased by urinary proteome from healthy individuals, it was not significantly affected by urinary proteome from stone formers (Fig. 1D). As a result, crystal mass was markedly decreased by urinary proteome derived from healthy individuals but slightly increased by urinary proteome from stone formers (Fig. 1E).

Fig. 1.

Effects of urinary proteomes from healthy individuals and stone formers on CaOx crystallization. Crystallization assay was performed with an addition of 4 µl of 1 µg/µl urinary proteome from healthy individuals or stone formers in AU or 4 µl of AU without proteins (control). (A): Micrographs of CaOx crystals formed under different conditions. (B): Size distribution of CaOx crystals under different conditions. The vertical line indicates the median of CaOx crystal sizes in the experimental control condition and was used as a reference point to determine inhibitory (less than the reference value) and promoting (greater than the reference value) activities on CaOx crystals. (C) and (D): Crystal size was quantified from 100 crystals/sample, and crystal number was counted from 10 random fields/sample in each experiment. (E): Crystal mass was then calculated using Formula (1) (see Materials and Methods). All quantitative data were derived from three independent experiments. The error bar represents SD, and only significant p-values are indicated.

Furthermore, the morphology of CaOx crystals formed under the two tested conditions exhibited notable differences. While urinary proteome from stone formers caused no significant change in the crystal morphology (mainly monoclinic prismatic shape, which is typical for CaOx monohydrate, the main pathogenic form of CaOx [32,33]), urinary proteome from healthy individuals caused a transformation of CaOx to bipyramidal shape (which is typical for CaOx dihydrate and less pathogenic [32,33]) (Fig. 1A).

Effects of urinary proteomes from healthy individuals and stone formers on CaOx growth

In addition to an initial phase of crystal formation, the effects of urinary proteomes from healthy individuals and stone formers on the later phase, i.e., crystal growth, were examined. Fig. 2A presents fixed-field micrographs of CaOx crystals at T₀ (after initial crystallization was complete) and T₆₀ (after 60-min further incubation to allow crystal enlargement). Quantitative analysis revealed a significant increase in CaOx crystal size (as reflected by Δ crystal size; see Formula (2) in Materials and Methods) in the experimental control condition (Fig. 2A and B). Comparing with the control, urinary proteome from healthy individuals caused a significant decrease in Δ crystal size. By contrast, urinary proteome from stone formers caused a significant increase in Δ crystal size as compared with the experimental control condition and the tested condition with urinary proteome from healthy individuals (Fig. 2A and B). Crystal growth inhibitory activity calculated by using Formula (3) (see Materials and Methods) showed the inhibitory effect of urinary proteome from healthy individuals but the promoting effect of urinary proteome from stone formers (Fig. 2C).

Fig. 2.

Effects of urinary proteomes from healthy individuals and stone formers on CaOx growth. Crystal growth assay was performed with an addition of 4 µl of 1 µg/µl urinary proteome from healthy individuals or stone formers in AU or 4 µl of AU without proteins (control). (A): Fixed-field micrographs of CaOx crystals at T₀ (after initial crystallization was complete) and T₆₀ (after 60-min further incubation to allow crystal enlargement). Crystal sizes at both T₀ and T₆₀ were quantified from 100 crystals/sample in each experiment. (B): Δ crystal size representing crystal growth was calculated using Formula (2) (see Materials and Methods). (C): Crystal growth inhibitory activity was then calculated using Formula (3) (see Materials and Methods). A positive value indicated the inhibitory activity, whereas a negative value indicated the promoting activity of the tested sample. All quantitative data were derived from three independent experiments. The error bar represents SD, and only significant p-values are indicated.

Effects of urinary proteomes from healthy individuals and stone formers on CaOx aggregation

Another critical phase of stone formation, crystal aggregation, was also examined. Under the experimental control condition with an appropriate concentration of CaOx crystals, the crystals could form self-aggregates (Fig. 3A). Comparative analysis of the number of crystal aggregates showed that urinary proteome from healthy individuals caused a significant decrease in crystal aggregation as compared with the control condition (Fig. 3B). By contrast, urinary proteome from stone formers caused a significant increase in the number of crystal aggregates as compared with the control condition and the tested condition with urinary proteome from healthy individuals (Fig. 3B).

Fig. 3.

Effects of urinary proteomes from healthy individuals and stone formers on CaOx aggregation. Crystal aggregation assay was performed with an addition of 4 µl of 1 µg/µl urinary proteome from healthy individuals or stone formers in AU or 4 µl of AU without proteins (control). (A): Micrographs of CaOx crystal aggregates (in the red-dotted circles). (B): Number of the crystal aggregates was counted from at least 15 random fields/sample. All quantitative data were derived from three independent experiments. The error bar represents SD, and only significant p-values are indicated. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Descriptive proteomics to identify and characterize proteins in the urine of healthy individuals and stone formers and their physicochemical properties

To characterize urinary proteomes from healthy individuals and stone formers, we performed descriptive proteomics. In-solution tryptic digestion by filter-aided sample preparation (FASP) method followed by nanoLC-ESI-LTQ-Orbitrap-MS/MS analyses using highly stringent parameters identified 203 and 381 proteins in urine samples from healthy individuals and stone formers, respectively (Supplementary Tables S1 and S2, respectively). Their physicochemical properties were then examined and compared. The results demonstrated that molecular weight (MW) of urinary proteins identified from stone formers was significantly greater than those of healthy controls, but proportions of low-MW (<50 kDa) and high-MW (≥50 kDa) proteins did not significantly differ (Fig. 4A–C). Also, isoelectric point (pI) of urinary proteins identified from stone formers was significantly greater than those of healthy controls (Fig. 4D). As a result, the proportion of acidic proteins was significantly less, whereas the proportion of basic proteins was significantly greater in the urine of stone formers (Fig. 4E and F). Other than these, averaged instability index, proportions of unstable and stable proteins, grand average of hydrophathicity (GRAVY) score, proportions of hydrophobic and hydrophilic proteins, and proportions of amino acid components (aromatic, polar, non-polar, positively charged and negatively charged residues) had no significant differences between the two groups (Fig. 4G–M).

Fig. 4.

Analyses of physicochemical properties of urinary proteomes from healthy individuals and stone formers identified by descriptive proteomics. All the identified proteins (by descriptive proteomics) in urinary proteomes from healthy individuals and stone formers were subjected to analyses of various physicochemical parameters. (A): Molecular weight (MW). (B) and (C): Proportions of low-MW and high-MW proteins, respectively. (D): Isoelectric point (pI). (E) and (F): Proportions of acidic and basic proteins, respectively. (G): Instability index. (H) and (I): Proportions of unstable and stable proteins, respectively. (J): Grand average of hydrophathicity (GRAVY) score. (K) and (L): Proportions of hydrophobic and proteins, respectively. (M): Amino acid composition. All quantitative data were derived from three independent experiments. The error bar represents SD, and only significant p-values are indicated.

The increased proportion of oxidatively modified proteins in the urine of stone formers

Because physicochemical properties had only mild degrees of differences, we hypothesized that there should be other characteristics of the urinary proteome that obviously differed between healthy individuals and stone formers. Our previous studies have reported that oxidative modifications can modify the CaOx crystal modulatory activities of urinary proteins [20,21]. We, therefore, analyzed oxidative modifications of the urinary proteome from stone formers compared with that of healthy individuals. The data showed that the urinary proteome from stone formers had a much greater proportion (approximately 2.5-fold) of oxidatively modified proteins than that from healthy individuals (Fig. 5A). Regarding the type of oxidative modifications, proportions of proteins with all the oxidative forms (including oxidation, dioxidation and trioxidation) were significantly greater in the urinary proteome from stone formers (Fig. 5B). At the amino acid residue level, oxidation at Met, dioxidation at Met and trioxidation at Cys, but not oxidation at Cys and dioxidation at Cys, had significantly greater proportions in the urinary proteome from stone formers (Fig. 5C).

Fig. 5.

Differential levels of oxidatively modified proteins in urinary proteomes from healthy individuals and stone formers identified by descriptive proteomics. All the identified proteins (by descriptive proteomics) in urinary proteomes from healthy individuals and stone formers were subjected to analyses of their oxidative modifications. (A): Proportion of the oxidatively modified proteins. (B): Proportions of proteins with various types of oxidative modifications. (C): Proportions of proteins with various oxidatively modified amino acid residues. All quantitative data were derived from three independent experiments. The error bar represents SD, and only significant p-values are indicated.

Quantitative proteomics to determine differentially excreted urinary proteins in stone formers compared with healthy individuals

In addition to descriptive proteomics, quantitative proteomics was performed to identify urinary proteins with significantly differential levels between healthy individuals and stone formers. Quantitative analysis revealed a total of 145 urinary proteins with significantly differential levels between the two groups (Supplementary Table S3). Among these, 89 had increased levels, whereas 56 had decreased levels in the urine from stone formers (Supplementary Table S3).

Oxidative modifications in the differentially excreted urinary proteins

To further address the functional significance of these differentially excreted proteins, we analyzed their oxidatively modified status. Among the non-oxidized proteins (without oxidative modifications detected), their proportions in the proteins with increased levels and those with decreased levels in the urine of stone formers were comparable (54 % vs. 46 %, respectively) (Fig. 6A). However, the proportion of the oxidized (oxidatively modified) proteins was much greater (>3-fold) in the proteins with increased levels than those with decreased levels in the urine of stone formers (77 % vs. 23 %, respectively) (Fig. 6B).

Fig. 6.

Proportions of non-oxidized and oxidized proteins in differentially excreted urinary proteins. All the differentially excreted urinary proteins (identified by quantitative proteomics) in stone formers compared with healthy individuals were subjected to analyses of their proportions of non-oxidized and oxidized proteins. (A): Proportions of the non-oxidized proteins among those with increased and decreased levels in stone formers compared with healthy individuals. (B): Proportions of the oxidized proteins among those with increased and decreased levels in stone formers compared with healthy individuals.

Gene ontology (GO) biological processes and molecular functions of the differentially excreted urinary proteins

All of the differentially excreted proteins that had significantly differential levels identified by quantitative proteomics were subjected to analyses of GO biological processes and molecular functions using the DAVID and REVIGO tools. Such analyses revealed multiple GO biological processes and molecular functions related to the differentially excreted proteins (Fig. 7, Fig. 8, Fig. 9). Interestingly, analyses of the GO biological processes revealed that the proteins with increased levels in the urine of stone formers were associated with protein oxidation, oxygen transport, nitric oxide transport, hydrogen peroxide catabolic processes, negative regulation of hydrogen peroxide catabolic processes, response to oxidative stress, response to hydrogen peroxide, response to carbon dioxide, response to ozone, negative regulation of oxidoreductase activity, and cellular oxidant detoxification (Fig. 7, Fig. 9A and 9A), all of which are related to oxidative stress and oxidative modifications of proteins. However, none of the GO biological functions among the proteins with decreased levels in the urine of stone formers were associated with oxidative stress (Fig. 7, Fig. 9B and 9C). In concordance, analyses of the GO molecular functions demonstrated that the proteins with increased levels in the urine of stone formers, but not those with decreased levels, were associated with oxygen binding, oxygen carrier activity, and antioxidant activity, all of which are related to oxidative stress and oxidative modifications of proteins (Fig. 8, Fig. 9B and D).

Fig. 7.

GO biological processes of differentially excreted urinary proteins. All the differentially excreted urinary proteins (identified by quantitative proteomics) in stone formers compared with healthy individuals were subjected to analyses of their GO biological processes. (A): GO biological processes of urinary proteins with significantly increased levels in stone formers compared with healthy individuals. (B): GO biological processes of urinary proteins with significantly decreased levels in stone formers compared with healthy individuals. Those related to oxidative stress are localized in the purple-dashed box. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 8.

GO molecular functions of differentially excreted urinary proteins. All the differentially excreted urinary proteins (identified by quantitative proteomics) in stone formers compared with healthy individuals were subjected to analyses of their GO molecular functions. (A): GO molecular functions of urinary proteins with significantly increased levels in stone formers compared with healthy individuals. (B): GO molecular functions of urinary proteins with significantly decreased levels in stone formers compared with healthy individuals. Those related to oxidative stress are localized in the purple-dashed box. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 9.

Cluster analysis of GO biological processes and molecular functions of the differentially excreted urinary proteins. All the differentially excreted urinary proteins (identified by quantitative proteomics) in stone formers compared with healthy individuals were subjected to cluster analyses of their GO biological and molecular functions using the REVIGO tool. Only the clusters with dispensability value < 0.2 were considered as the cluster representatives. (A) and (B): Representative clusters of GO biological processes and molecular functions, respectively, in urinary proteins with significantly increased levels in stone formers compared with healthy individuals. (C) and (D): Representative clusters of GO biological processes and molecular functions, respectively, in urinary proteins with significantly decreased levels in stone formers compared with healthy individuals. Those related to oxidative stress are indicated with red fonts and underlined. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Validation of the elevated levels of oxidized proteins in the urine of stone formers

Based on descriptive proteomics, quantitative proteomics, and analyses of biological processes and molecular functions of the differentially excreted proteins, differential levels of oxidatively modified proteins were likely to determine the contradictory modulating activities of urinary proteomes from healthy individuals and stone formers on CaOx stone formation. We thus performed ELISA to validate the differential levels of oxidatively modified urinary proteins in the two groups. Quantitative analysis of individual urine samples from 40 normal healthy non-stone individuals and 37 CaOx kidney stone formers confirmed that levels of the oxidatively modified proteins were significantly greater in the urine of stone formers compared with healthy individuals (Fig. 10).

Fig. 10.

Validation of differential levels of the oxidized urinary proteins in stone formers compared with healthy individuals. DNPH-based ELISA assay was performed to measure levels of oxidized urinary proteins in individual urine samples obtained from all healthy subjects (n = 40) and stone formers (n = 37). All quantitative data were derived from three independent experiments, and the error bar represents SD.

Discussion

Oxidative modifications of proteins are the dynamic biochemical processes that are crucial for determining physiological and pathological conditions of various cells and tissues [34,35]. Oxidation can induce alterations in both structural conformation and function of the modified proteins. The oxidative modification processes primarily occur following exposure to ROS with unpaired electrons that make them highly reactive [34,35]. As a consequence, the oxidized proteins or amino acids are commonly considered abnormal and are frequently linked to many pathological conditions, including neurodegenerative disorders [36], cardiovascular diseases [37,38], aging [39,40] and many others. While the cells possess intricate mechanisms to prevent, repair and eliminate the damaged proteins, an imbalance when the rate of protein oxidation exceeds the cellular reparative capability can result in the accumulation of dysfunctional proteins and the initiation and progression of various diseases [34].

Urine is a compelling biofluid in the field of clinical proteomics studies due to its non-invasive accessibility and availability in most patients to obtain large volume. The urine of a healthy individual, but not that of a patient with renal pathology, contains a trace amount of proteins [41]. Advances and progress in proteomics technologies have greatly improved the characterization of proteins and peptides in the urine, rendering them well-suited for various clinical applications [[42], [43], [44]]. Proteinuria is often an indicator of underlying renal pathologies caused by various diseases, such as diabetes, hypertension, glomerulonephritis, and other kidney diseases [[45], [46], [47], [48], [49], [50], [51]]. Although proteinuria is not a common feature of kidney stone disease, our recent in vitro studies have shown that chemically induced oxidative modifications of urinary proteins by performic acid can change their CaOx crystal modulatory activities [20,21]. The oxidized forms of a major abundant urinary protein called Tamm-Horsfall protein (THP) or uromodulin (UMOD) significantly promote CaOx crystallization and growth, in contrast to its unmodified form that inhibits CaOx crystallization and growth [21]. In addition to the evidence on a single urinary protein, oxidative modifications of the whole urinary proteome cause the transformation of its inhibitory activities to promote CaOx crystallization, growth and aggregation [20]. These findings serve as the first pieces of evidence that a single urinary protein or the whole urinary proteome can switch their modulatory activities on CaOx crystals, from inhibition to promotion [20,21].

Previous studies on kidney stone disease have consistently indicated its association with oxidative stress [14,52,53] and renal tissue injury [14,53,54]. Generally, an overproduction of ROS is considered a “result” or “effect” of the kidney stones or crystals formed inside the stone matrix, leading to oxidative stress, cellular injury, and the release of various inflammatory mediators [55,56]. As such, oxidative stress generally refers to a downstream event after a kidney stone or intrarenal crystal deposition is developed. Herein, we provide direct evidence demonstrating the upstream effects of the oxidative modifications of urinary proteome on kidney stone formation. Descriptive and quantitative proteomics analyses of the pooled urine followed by ELISA validation in individual samples revealed significantly elevated levels of oxidized urinary proteins in the stone formers as compared with healthy individuals. Moreover, the urinary proteome from stone formers promoted CaOx crystallization, growth and aggregation, whereas that from healthy individuals inhibited these CaOx stone-forming processes. These findings suggest that oxidative modifications of urinary proteins play a pivotal role in augmenting stone formation, thereby increasing the risk of kidney stone disease.

CaOx is the predominant stone type in the kidney [[57], [58], [59]]. Not only the inorganic small molecular constituents (such as chemical elements) but also organic macromolecules (such as proteins) are found inside the stone matrix and are believed to play roles in regulating the stone formation [[60], [61], [62]]. Several proteins have been identified in the stone matrix, e.g., S100A8, S100A9, UMOD, albumin, osteopontin, and lactotransferrin [60,63,64]. Among these, UMOD is the most abundant urinary protein. Our previous study has verified that the increases in oxidized Cys, Met and/or tryptophan (Trp) residues in UMOD promote crystallization and growth of CaOx crystals [21]. Structurally, a full-length UMOD contains 48 Cys residues [21]. This native form exhibits an inhibitory effect on CaOx crystals and stone formation [21,65]. However, alterations or modifications of Cys side-chain, especially those located in the calcium-binding domain, of UMOD can impair its calcium-binding activity, leading to diminished affinity with free calcium ions. As UMOD is the most abundant urinary protein, when this condition occurs, the urine may easily reach a supersaturated state of calcium, thereby enhancing CaOx crystallization and growth [21]. This pathogenic mechanism may also offer an explanatory framework for the alterations in the stone modulatory activities of many other calcium-binding urinary proteins.

Descriptive proteomics in our present study also provided details for the oxidative modifications, i.e., proportions of oxidized proteins, type of oxidative modifications (e.g., oxidation, dioxidation, and trioxidation), and oxidized amino acid residues. Met oxidation was the most abundant form of oxidative modifications found in urinary proteome from stone formers, constituting over 30 %. The second and third most abundant forms increased in the urinary proteome from stone formers were Met dioxidation and Cys trioxidation, respectively. Met and Cys are both sulfur-containing amino acids that are easily modified by ROS, leading to structural alterations and protein damage [[66], [67], [68]]. The sulfur atom in the thioether side chain of Met can be oxidized to Met sulfoxide (Met-O). This Met sulfoxide can be further oxidized to Met sulfone (Met-O₂), which is a highly oxidized state of Met and is irreversible. The side chain of Cys possesses potent nucleophilic properties according to its thiol (-SH) group. Normally, thiol exhibits mild acidity, due to the relatively weak S-H bond. The pKa of Cys may decrease when the other positively charged residues are present close to its thiol group. As a result, deprotonation can occur, leading to thiolate (-S-) formation. Compared with other forms, the nucleophilicity of thiolate is even higher, making it easily interact with oxidants and/or other electrophilic species. When Cys reacts with two-electron oxidants (i.e., peroxides and peroxynitrite), it forms sulfenic acid (-SOH). However, when Cys reacts with one-electron oxidants (i.e., superoxide anionic radical (O₂^•–) and hydroxyl radical (OH^•)), it can form thiyl radical (RS^•). The reaction between thiyl and hydroxyl radicals results in the formation of sulfenic acid. Sulfenic acid is a highly reactive entity capable of engaging in two distinct reactions, forming a disulfide bond (S-S) with a neighboring Cys or undergoing an irreversible oxidation process, leading to the formation of sulfinic acid (−SO₂H) and sulfonic acid (−SO₃H) [63,66].

As oxidative modifications can be reversible and irreversible chemical reactions, the formation of oxidized proteins in diverse oxidation states may induce a wide variety of alterations in protein properties such as molecular mass, folding, stability and hydropathicity. Therefore, we also evaluated these protein parameters. Analyses of physicochemical properties revealed that urinary proteome from stone formers had (slightly) higher MW as compared with that of healthy individuals. This might be because most oxidative modifications in protein side chains can affect molecular mass by the addition of 16 and 32 Da after the conversion of Met to Met sulfoxide and the formation of Met sulfone, respectively [69,70]. Also, sulfenic, sulfinic and sulfonic acids in Cys residues can create increases of 16, 32 and 48 Da, respectively [69,70].

Moreover, we observed a decrease in the proportion of acidic proteins and an increase in the proportion of basic proteins in the urinary proteome of stone formers. Previous studies have suggested that the oxidative modification processes can induce alterations in the net charge and surface characteristics of the proteins [71,72]. However, our current knowledge of this phenomenon is limited, necessitating further comprehensive investigations in this area. A plausible explanation may involve negative charges of acidic proteins, which normally facilitate their interactions with calcium ions [73] or crystal surfaces [74], thereby influencing nucleation, growth and cell adhesiveness of individual CaOx crystals. A reduction in acidic protein levels may weaken the natural inhibitory processes against crystal formation, leading to stone development. Conversely, an increased abundance of basic urinary proteins, which carry positive charges, in stone formers may enhance their adsorption onto the negatively charged oxalate moieties on CaOx crystal surfaces, thereby influencing the dynamics of crystal formation [75]. The rest of the physicochemical parameters showed no significant differences between the two groups, suggesting that oxidative modifications did not affect protein stability and hydrophobicity/hydrophilicity.

In addition to Met oxidation, Met dioxidation and Cys trioxidation, carbonylation represents another (irreversible) form of protein oxidation characterized by the presence of ketone and aldehyde in amino acids such as lysine (Lys), arginine (Arg), proline (Pro) or threonine (Thr). ROS is a common inducer for this type of oxidative modification [[76], [77], [78]]. The generation of reactive ketones or aldehydes can be readily detected by ELISA using an anti-DNP antibody after protein samples are conjugated with DNPH [79]. Our result indicated that the stone formers had significantly greater levels of carbonylated proteins in their urine, consistent with the results obtained by descriptive and quantitative proteomics. Similar to other forms of oxidative modifications, carbonylation can cause slight changes in molecular mass due to the formation of aminoadipic semialdehyde, glutamic semialdehyde and hydroxyproline [80,81]. These data indicated that multiple forms of oxidative modifications were triggered in the urinary proteome of stone formers.

Finally, we have proposed the mechanisms by which the normal urinary proteome and the stone formers’ urinary proteome modulate CaOx crystallization, growth and aggregation (Fig. 11). For crystallization and growth, normal urinary proteins tend to bind free calcium ions, leading to a reduction of free calcium ions in the urine and decreased urinary supersaturation, thereby inhibiting CaOx crystallization and growth [27,[82], [83], [84], [85]]. However, oxidatively modified urinary proteins in the stone formers potentially have structural and functional defects that handicap their calcium-binding abilities, leading to abundant free calcium ions and supersaturation in the urine, thereby enhancing CaOx crystallization and growth [20,21]. For crystal aggregation, normal urinary proteins tend to bind CaOx crystal surfaces [27,82,83]. Because of their high calcium-binding abilities, the crystal surfaces are full of coating proteins and calcium ions on top, resulting in a reduction in adhesive force among adjacent crystals, thereby inhibiting CaOx aggregation. On the other hand, oxidative modifications can induce conformational changes in redox-sensitive proteins, potentially exposing (protein–protein) aggregation-prone motifs [[86], [87], [88]]. Together with their reduced calcium-binding abilities, oxidatively modified proteins bound on the crystal surfaces are full of the (protein–protein) aggregation-prone motifs and thus tend to possess protein–protein aggregation, leading to increased adhesive force among adjacent crystals and enhanced CaOx crystal aggregation (Fig. 11).

Fig. 11.

Mechanisms underlying inhibitory activities of the normal urinary proteome and promoting activities of the stone formers’ urinary proteome on CaOx stone formation. Left: The CaOx-inhibitory activities of the normal urinary proteins. Right: The CaOx-promoting activities of the stone formers’ urinary proteome.

Conclusions

We present herein solid evidence demonstrating that the CaOx stone formers harbor a set of urinary proteins (proteome) with promoting effects on CaOx crystallization, growth and aggregation. By contrast, urinary proteome from healthy individuals exerts inhibitory effects on CaOx crystallization, growth and aggregation. Descriptive and quantitative proteomics as well as GO functional enrichment analyses of biological processes and molecular functions revealed significantly greater levels of oxidatively modified proteins in the urinary proteome derived from stone formers as the most striking difference among other physicochemical properties. Of the increased oxidative modifications in the urinary proteome of stone formers identified by proteomics, Met oxidation is the most abundant form, whereas Met dioxidation and Cys trioxidation are the second and third most abundant forms, respectively. ELISA confirmed that other forms of oxidative modifications (particularly carbonylation) also increase in the urinary proteome of stone formers. These data implicate that oxidatively modified proteome serves as the key pathogenic factor or risk for CaOx kidney stone formation.

Ethical statement

This study was reviewed and approved by the Siriraj Institutional Review Board (approval no. Si415/2023), and signed consent forms were obtained from all participants. Experiments using clinical samples were set following various standard guidelines, i.e., the Declaration of Helsinki, the Belmont Report, and ICH Good Clinical Practice.

Data availability

The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium (https://www.proteomexchange.org/) via the PRIDE (https://www.ebi.ac.uk/pride/) partner repository with the dataset identifier PXD050609 and https://doi.org/10.6019/PXD050609. (Username: reviewer_pxd050609@ebi.ac.uk/Pass: jtSHVrcl).

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article: