Table 1.
Biomarkers and related metabolic pathways in human calcium metabolic disorders discovered using metabolomics approaches.
| Disorders | Approaches | Study Groups | Sample Type | Metabolite Change | Authors |
|---|---|---|---|---|---|
| PHPT SHPT THPT |
HRMAS NMR spectroscopy |
-18 patients with PHPT; -11 patients with SHPT; -3 patients with THPT. |
Parathyroid tissue | Single adenoma and multiglandular disease could be distinguished with metabolomic profiling. Single adenoma: ↑ fumarate, choline, β-glucose, myo-inositol, ascorbate, glycine, scyllo-inositol. Multiglandular disease: ↑ glutamate, glutamine, taurine, aspartate, lactate, GSH. PHPT and SHPT could be distinguished with metabolomic profiling. PHPT: ↑ β-glucose, GSH, ascorbate, myo-inositol glutamate, phosphorylcholine, taurine. SHPT: ↑ fumarate, serine, choline, aspartate, glycerophosphocholine glutamine. |
Battini et al., 2016 [23] |
| PHPT SHPT |
GC-MS LC-MS |
-40 patients with PHPT; -40 patients with SHPT; -4 patients with normal parathyroid gland. |
Parathyroid tissue |
Polychlorinated biphenyls, dichloro-diphenyl-trichloroethane derivatives and polybrominated diphenyl ethers were detected in parathyroid tumors. Polychlorinated biphenyl-49 and polychlorinated biphenyl-28 and levels were positively correlated with parathyroid tumor mass. The levels of calcium serum were inversely correlated with concentrations of polybrominated diphenyl ether-47. Polychlorinated biphenyl-49 and p,p’-dichlorodiphenyldichloroethylene were not detected in normal parathyroid gland tissue. |
Hu et al., 2021 [26] |
| SHPT | UPLC-Q- TOF/MS |
Maintenance- peritoneal-dialysis patients: -19 disease controls with PTH 150–300 pg/mL; -19 patients with PTH > 300 pg/mL. |
Serum | 32 unique metabolites were identified in the high-PTH group: ↑ N-(1-Deoxy-1-fructosyl)-tryptophan, N-acetylserotonin glucuronide, dopamine glucuronide, prolyl-tyrosine, glycylprolylhydroxyproline, aminohippuric acid, 2-phenylglycine, 4-carboxyphenylglycine, N-(3-Indolylacetyl)- L-isoleucine, glutaminyl-hydroxyproline, diethyl fumarate, isopropyl citrate, (R)-2-methylmalate, glutamyl-glutamate, N-acetylaspartylglutamic acid. Only 2 of 32 found metabolites were downregulated: ↓ cytidine and L-phenylalanine. Most identified metabolites were not primary metabolites. |
Wu et al., 2015 [27] |
| SHPT | UPLC-MS | -15 patients in the preoperative group with SHPT with PTH level > 600 pg/mL; -15 patients in postoperative group, after parathyroidectomy plus forearm transplantation due to SHPT with PTH level < 150 pg/mL; -5 healthy controls. |
Serum | 5 metabolites were highly correlated with SHPT. Biomarker group with SHPT: ↑ allyl isothiocyanate, D-aspartic acid, L-phenylalanine. ↓ D-galactose, indoleacetaldehyde. -preoperative group vs. healthy controls (AUC = 0.947); -postoperative group vs. healthy controls (AUC = 0.6). |
Shen et al., 2019 [28] |
| Calcium deficiency | UPLC-Q-TOF MS/MS | Phase I: male rats. Phase II: 70 postmenopausal women. |
Urine | Biomarkers (Phase I): glycine, sebacic acid, oxoglutaric acid, pyrophosphoric acid, pseudouridine, taurine, phenylacetylglycine, indoxyl sulfate. 2 biomarkers (pseudouridine and citrate) were further confirmed in 70 women. |
Wang et al., 2013 [29] |
| Nutritional rickets | UPLC−MS/MS | -115 children with rickets; -85 healthy children. |
Urine | 31 biomarkers of nutritional rickets were identified. 5 candidate biomarkers for clinical diagnosis were screened (phosphate, pyrophosphate, citric acid, cAMP, sebacic acid). The combination of sebacic acid and phosphate was selected as the candidate biomarker with high sensitivity (94.0%) and specificity (71.2%) (AUC = 0.85). |
Wang et al., 2014 [30] |
| Osteopenia and osteoporosis | NMR spectroscopy | 56 postmenopausal women: -36 with low bone mass (T-score < −1); -20 with normal bone mass. |
Serum and Urine | Disparities between individuals’ responses to vitamin D and calcium supplementation in patients with genetic variations in oestrogen receptor 1 gene and vitamin D receptor gene were found. NMR studies indicated unique patterns of metabolites, separating responders from non-responders and controls. |
Elnenaei et al., 2010 [31] |
| Osteopenia and osteoporosis | LC-MS | 69 patients: -25 patients with osteoporosis; -22 patients with osteopenia; -22 patients with normal bone mineral density. |
Serum | 116 metabolites were associated with low bone mineral density compared to controls (94 metabolites were dysregulated: ↑52, ↑42). The most frequently dysregulated metabolic pathways in low bone mineral density: histidine metabolism, glyoxylate, aminoacyl-tRNA biosynthesis, dicarboxylate metabolism and unsaturated fatty acid biosynthesis. 35 metabolites were dysregulated between patients with osteopenia and osteoporosis: ↑11(3-carboxy-4-methyl-5-propyl-2-2furanopropionic acid, carnitine derivatives) and ↓24(phosphatidylcholine, sphingomyelin, palmitic acid) in patients with osteopenia compared to patients with osteoporosis). |
Aleidi et al., 2021 [32] |
| Osteoporosis | HRMAS NMR spectroscopy | 601 healthy Taiwanese women (40–55 years old). | Plasma | 7 metabolites characterizing low BMD were identified. Elevated glutamine was significantly associated with low BMD. Elevated lactate, lipids, acetone and very-low-density lipoprotein protected against low BMD. Metabolomic profiling may improve the risk prediction of osteoporosis. |
You et al., 2014 [33] |
| Osteoporosis | GC-MS | 364 women: -Premenopausal women with normal BMD; -Postmenopausal women with normal BMD; -Postmenopausal women with osteopenia; -Postmenopausal women with osteoporosis. |
Serum | 12 metabolites were able to differentiate low-BMD groups from normal-BMD groups. 5 free fatty acids (11,14-eicosadienoic acid, oleic acid, LA and AA) had the greatest potential to be used as osteoporosis biomarkers. ↑ Arachidonic acid, eicosadienoic acid, lysine, linoleic acid, tryptophan, allose, oleic acid. ↓ 3-hydroxy-l-proline, homoserine, pyruvic acid. |
Qi et al., 2016 [34] |
| Osteoporosis | CE-TOFMS | Women (39–64 years old). | Serum | 57 metabolites differed significantly among various groups. Diglycine and cystine were lower in the low-BMD group. Hydroxyproline was higher in the low-BMD group. Metabolomic profiling may improve the risk prediction of osteoporosis. |
Miyamoto et al., 2017 [35] |
| Osteoporosis | CE-TOFMS | Women (31–69 years old): -30 premenopausal and normal BMD; -46 postmenopausal and normal BMD; -33 postmenopausal and low BMD. |
Serum | 52 metabolites differed significantly among various groups. Metabolomic profiling may improve the risk prediction of osteoporosis. Ornithine, arginine, citrulline, creatine and urea levels were increased in postmenopausal groups. The level of guanidinoacetate was decreased in postmenopausal groups. The levels of most amino acids, except branched-chain amino acids, were increased in postmenopausal women with low BMD. |
Miyamoto et al., 2018 [36] |
| Osteoporosis | LC-MS | 136 women (20–40 years old): -65 with low hip BMD; -71 with high hip BMD. |
Serum | 14 metabolites (7 amino acids and amino acid derivatives and 5 lipids: 3 bile acids and 2 organic acids) were associated with a risk of low BMD. Glutamic acid, threonine, taurine, GABA and cysteine were significantly associated with BMD. Metabolomic profiling may improve the risk prediction of osteoporosis. |
Zhao et al., 2018 [37] |
| Osteoporosis | (1)H-NMR | -18 healthy volunteers; -18 diabetic patients with disordered bone metabolism. |
Plasma | Metabolomic profiling may improve the risk prediction of diabetic osteoporosis. ↑ Isoleucine valine, glutamine, alanine, inositol, proline, leucine, glucose, 1-methyl-histidine, tyrosine, N-acetylglycoprotein. ↓ O-acetylglycoprotein, creatine, α-ketoglutaric acid, citrate. |
Liang et al., 2020 [38] |
| Osteoporosis | LC-MS/MS | Discovery cohort: 1552 participants. Replication cohort: 634 participants. |
Serum | 27 metabolites were associated with femoral-neck BMD or lumbar-spine BMD. Glycine, triacylglycerol and phosphatidylcholine were negatively associated with femoral-neck BMD. Phosphatidylcholine and triacylglycerol were negatively associated with lumbar-spine BMD. Authors replicated improvement of fracture prediction with selected metabolites in 634 participants. Metabolomic profiling may improve the risk prediction of osteoporosis. |
Zhang et al., 2021 [39] |
| Osteoporosis | UPLC−MS/MS | 971 adults. | Serum and feces | Isoleucine, valine and leucine degradation was associated with osteoporosis. Strong evidence linking gut dysbiosis, fecal metabolomics and serum metabolomics with osteoporosis was reported. |
Ling et al., 2021 [40] |
| Vitamin D3 deficiency | DI-LC/MS/MS GC-MS |
30 healthy adults were given 600, 4000 or 10,000 IUs of vitamin D3/day for 6 months. | Serum and urine | Statistically significant changes in 11 metabolites (7 from serum and 4 from urine) after 6 months of vitamin D3 supplementation were found. There was a distinct difference in the targeted metabolites between the more and less vitamin D3 responsive participants. Targeted analysis included 83 metabolites from serum and 36 metabolites from urine. |
Shirvani et al., 2020 [41] |
| Vitamin D3 deficiency | (1)H-NMR | 76 postmenopausal women with Vitamin D insufficiency (<50 nmol/l) were given 2800 IUs of vitamin D3/day or placebo for 12 weeks | Serum | Supplementation of vitamin D significantly increased serum levels of carnitine, choline and urea and trimethylamine-N-oxide tendency to rise. | Bislev et al., 2020 [42] |
PHPT = primary hyperparathyroidism; SHPT = secondary hyperparathyroidism; THPT = tertiary hyperparathyroidism; AUC = area under the curve; HRMAS= high-resolution magic angle spinning; NMR = nuclear magnetic resonance; UPLC = ultra-performance liquid chromatography; Q-TOF = quadrupole time of flight; MS= mass spectrometry; GC = gas chromatography; CE = capillary electrophoresis; LC = liquid chromatography; DI-LC/MS/MS = direct-flow-injection mass spectrometry; BMD = bone mineral density; (↑) = increased metabolite levels; (↓) = decreased metabolite levels.