In the past twenty years, a growing body of research has unequivocally demonstrated the strong, causal relationship between acidosis and the pathophysiology of bone disease [1
]. Severe acidosis with acidemia occurs when compensatory measures for maintaining the acid–base equilibrium fail and the blood pH value drops below 7.35, while chronic low-grade acidosis is the result of the continual adaptation of the body to a variety of physiological and pathological conditions, including ageing, menopause, excessive dietary acid intake, bowel diseases, excessive/anaerobic exercise, altered cell metabolism, hypoxia, inflammation, infection, diabetes and tumours [7
The prompt skeletal response to acute metabolic acidosis is a physicochemical reaction aimed at buffering hydrogen ions by means of alkali metals (sodium, potassium), carbonate and phosphate groups, thus resulting in a net calcium efflux from the mineralised matrix [1
]. When acidosis follows a more chronic course, it may elicit a cell response. In vitro studies have shown that the bone mineral dissolution due to the ion exchange lasts 48 h on average while, after 48 h, a pivotal role is ascribable to bone cells [2
]. In fact, the resorption activity of osteoclasts increases dramatically when the intra-bone pH drops below 6.9 and, conversely, acidosis significantly inhibits the osteogenic function of osteoblasts, including the production of extracellular matrix, the activity of alkaline phosphatase and the formation of trabecular bone [4
Contextually to the skeletal modifications, metabolic acidosis has a strong impact on citrate homeostasis. The regulation of the acid–base balance depends largely on acid excretion and urinary buffers, and modulation of the citrate excretion in the kidneys plays a pivotal role in driving this function since proton excess reduces the trivalent anion in the divalent form which may be reabsorbed through the sodium-citrate cotransporter [9
]. However, citrate is primarily an intermediate in the tricarboxylic acid cycle (TCA cycle, Krebs cycle), the metabolic pathway which, in humans and all aerobic organisms, is the main adenosine 5′-triphosphate (ATP) provider, i.e., the primary energy source by which living cells accomplish essential functions [10
]. When the cellular ATP is abundant and the energy demand of the cells is low, the excess citrate can be exported outside the mitochondria and used for supporting the lipid biosynthesis of proliferating cells [11
] or the tissue-related functions of specialised cells, i.e., osteogenic cells [12
Basically, the pillars of citrate homeostasis are nutritional intake, renal clearance, cellular metabolism and bone remodelling; however, it is currently well known that the main endogenous bulk of citrate is stored in the bone [6
]. On the one hand, the bone-forming cells are able to synthesise and release citrate into the extracellular matrix so that it may be incorporated into the calcium phosphate–collagen complexes to promote the normal three-dimensional orientation of the apatite nanocrystals in the bone lamellae, thus ensuring the biomechanical properties of bone. On the other hand, the resorption activity by osteoclasts leads to the mobilisation of the citrate incorporated into the mineralised matrix, thus making it the most prominent source for the maintenance of plasma homeostasis [13
]. Hence, all the conditions which affect the balance between bone formation and bone resorption potentially affect citrate homeostasis, of which acidosis is part of.
For the above reasons, urinary citrate excretion has been proposed as a laboratory parameter for monitoring the acid–base balance and bone health status, even in subjects without overt metabolic acidosis [14
]. Similarly, citrate-based supplements have been used as a possible strategy for treating medical conditions related to acid overload and poor citrate bioavailability, and for mitigating the detrimental effect on skeletal homeostasis. Even though several interventional clinical trials have reported encouraging results, to date, there is no consensus regarding the use of citrate supplementation for the management of metabolic bone diseases since the heterogeneity of the studies did not allow identifying precise indications [6
]. Moreover, few studies have dealt with understanding the biological basis of the interaction between extracellular citrate and bone cells, and studies which have investigated the effects under acidic conditions are especially rare. We have recently demonstrated that, independently of its alkalizing capacity, potassium citrate (K citrate) prevented the increase in osteoclast activity induced by the acidic microenvironment while minor effects were observed on bone-forming cells [18
]. Other authors have evaluated the activity of citrate under neutral conditions and have demonstrated that increased citrate bioavailability in the extracellular microenvironment may foster the osteogenic differentiation of the mesenchymal stromal cells (MSCs) and accelerate bone tissue regeneration [19
Based on the background, the aim of this study was to investigate the anabolic properties of the citrate-based supplements most widely used in clinical practice, namely calcium citrate (Ca citrate) and K citrate. To attain this result, an experimental in vitro study which allowed evaluating their capability of fostering the osteogenic capacity of human MSCs (hMSCs) in both acidic and neutral settings was designed. First, the effects of acidosis on the osteogenic properties of hMSCs, in particular, on the citrate release and mineralisation of the extracellular matrix was investigated. Then, we evaluated whether the citrate supplementation allowed restoring the impaired functions resulting from exposure to the acidic conditions.
Under physiological conditions, the range of pH values measured in the intra-bone blood samples fluctuates from 7.3 to 7.4 with lower values expected in the interstitial fluid around the bone cells [29
]; however, the solubility of bone mineral increases dramatically when the pH drops below 7.0 [1
]. Acidosis dramatically enhances the resorption activity of osteoclasts and, conversely, significantly inhibits the osteogenic function of osteoblasts, including the production of extracellular matrix, the activity of alkaline phosphatase and the formation of trabecular bone [4
As citrate plays a pivotal role in the regulation of the acid–base equilibrium [9
], the maintenance of its homeostasis is a basic requirement for this crucial function to be carried out. Citric acid is naturally contained in fruits and vegetables, particularly in citrus fruits, with concentration ranging from 0.005 mol/L in oranges and grapefruit to 0.30 mol/L in lemons and limes [31
]. Almost the entire citrate intake is absorbed in the small intestine by means of a citrate transporter similar to that described in the kidneys [30
]. Food citrate is rapidly introduced into the circulation, filtered at the glomerular level, and eventually reabsorbed according to physiological needs [30
]. However, the net balance between gastrointestinal absorption and the urinary excretion of citrate suggests that the nutritional intake cannot be solely responsible for the maintenance of plasma homeostasis [13
]. The citrate derived from the Krebs cycle marginally contributes to citrate homeostasis since it is used by cells as an energy source or for supporting specific cell functions [12
]. It is currently well known that the main endogenous source of citrate is bone tissue, and the reasons for which so much citrate is found in bone have largely been clarified [6
]. There is a link connecting citrate and bone as citrate (1) is produced by osteoblasts [12
], (2) may influence their differentiation and functionality [19
] and (3) serves to maintain the integrity of the skeletal nano- and microstructures [35
]. In addition, all the conditions which upset the balance between bone formation and bone resorption may affect citrate homeostasis, including chronic low-grade acidosis [6
]. On the one hand, a low pH stimulates osteoclast resorption and favours the mobilization of citrate stored in bone. Even though the unbound molecules could be used to form new mineral matrix, the main role of citrate in the case of acidosis is to maintain a constant citricemia and ensure renal excretion of the proton excess. On the other hand, the osteogenic function of bone-forming cells is inhibited, including the production of citrate as an essential component for the mineralisation of the extracellular matrix [4
]. To translate the pathophysiology into clinical practice, the lower citrate bioavailability may lead to osteopenia or osteoporosis, i.e., a decrease in bone mass, deterioration of the skeletal microstructure, bone fragility and increased fracture risk [6
]. In this regard, Chen et al. have demonstrated that serum citrate levels of elderly osteoporotic subjects were significantly lower than those of young healthy individuals and positively correlated with the bone mineral density of the lumbar spine and hip [36
However, knowledge regarding the interaction between extracellular citrate and bone cells under acidic conditions is still lacking, and this gap does not allow identifying unequivocal indications for the use of citrate-based supplementation in the management of medical conditions related to acid overload and poor citrate bioavailability, and for mitigating its detrimental effect on skeletal homeostasis.
The exogenous supplementation of citrate may achieve two main functions: (1) citrate serves as an alkalizing agent because it is metabolised to 3 HCO3− groups which, in turn, could act as a buffer and oppose the detrimental effects of the low pH and (2) the nutritional intake contributes to citrate homeostasis and may save the citrate released through the osteoclast resorption, thus making it reusable for the formation of new bone.
In the first step of the study, the effects of acidosis on the osteogenic properties of hMSCs, in particular on the citrate release and mineralisation of the extracellular matrix, were investigated. In the second step, we aimed to evaluate whether Ca citrate and K citrate, the citrate-based supplements most widely used in clinical practice, restored the impaired functions resulting from exposure to acidic conditions.
Other authors have shown that the detrimental effect of acidosis on bone cells was detectable at pH 6.9 [37
], and therefore, this pH condition was chosen to mimic the acidic milieu for culturing and analysing the osteogenic properties of the hMSCs and the extracellular matrix organisation. The effects were explored at 14 days, in the early phase of matrix maturation and, then again at 21 days, thus providing adequate time for obtaining the formation of mineral nodules which could be microscopically identified using morphology and specific dyes [24
]. The hMSCs used in the experimental plan were selected on the basis of their capability to form mineral nodules after 21 days of culture; their osteogenic commitment was also confirmed in large-scale gene expression profiling as reported in the Gene Expression Omnibus (GEO) dataset repository [38
To the best of our knowledge, for the first time this study demonstrated that citrate excretion was significantly decreased when bone-forming cells lived in an acidic milieu, and that the decline was observed as early as 14 days. The above result could be explained in view of the knowledge that the mitochondrial aconitase and cytoplasmic citrate-lyase may increase during acidosis [39
]; both enzymes favour the citrate consumption in the Krebs cycle and in fatty acid biosynthesis, respectively, thus reducing the release of citrate into the extracellular fluid. As a result, the net loss of the citrate bioavailability could have affected the mineralisation process. In fact, in the early phase, signs of a delay in matrix maturation were observed as the amount of collagen released in the supernatant and the initial deposition of calcium complexes into the extracellular matrix were significantly reduced. These findings did not depend on cell number and viability as these were similar at pH 6.9 and pH 7.4, thus suggesting that the effects induced by a low pH were related to the impairment of the cell function. Other authors have shown that the cell viability of osteogenic precursors was unaffected by the acidic condition if they were in a stationary phase of growth, corresponding to the confluence status of the hMSCs used in our experimental setting [37
]. We verified that collagen released in the supernatant of the cells cultured at pH 6.9 did not derive from the degradation of the extracellular matrix, but was a newly synthesised collagen as the amount of Sirius Red staining correlated significantly with the procollagen type 1 N-terminal propeptide (P1NP). Osteoblasts make collagen in the form of procollagen which is excreted extracellularly, and the cleavage of N-terminal propeptide extensions, precedes the conversion of procollagen to mature collagen [40
]. Even though the collagen release was lower, the amount of fibrils in the extracellular matrix was comparable to that measured at the neutral pH, thus suggesting that the acidic milieu did not hamper the collagen deposition. The literature data on the expression of type I collagen in acidic conditions are conflicting, but confirm the discrepancy between synthesis/release in culture medium and deposition into the extracellular matrix. In a previous study, we found that the expression of type I collagen was highly variable in acidic conditions but not significantly affected [18
], while some authors showed that short-term exposure of mouse calvaria bone cells to acidic pH decreased the type I collagen mRNA [41
], and others demonstrated that collagen deposition in acidic culture was augmented [37
]. After 21 days of culture at pH 6.9, the quantitative differences in the collagen release and calcium deposition were no longer so evident, but the effects observed in the early phase compromised the final outcome as the number and size of the mineral nodules was reduced.
Areas in which the mineralisation nodules were not clearly visible were also evaluated by using TEM and FT-IR spectroscopy, with the aim of highlighting the submicroscopic changes in the organisation and composition of the matrix. Under the acidic conditions, the ultrastructural TEM images showed an impaired arrangement of the collagen and the lack of matrix vesicles aligned to the fibres. Fibrillar collagen provides a template for the mineral deposition since spindle- or plate-shaped crystals of hydroxyapatite tend to be oriented in the same direction as the fibres [43
]. Mutations in the aminoacid sequence of collagen, defects as disorganised collagen fibres, as well as pathological accumulation of unfolded collagen triple helices, impair fibrillar collagen functions and, ultimately, tissue mineralisation [44
]. In human cells, the damage seemed to be more evident than that observed in the murine model as the osteoblast cultures derived from rat calvaria did not exhibit notable differences in the organisation of collagen fibrils at pH 6.9 and pH 7.4 [37
FT-IR spectroscopy has been proposed as a powerful technique for the characterisation of proteins and collagen-based materials, and we focused on Amide I and Amide II which are also considered to be the two main markers of collagen structure [45
]. Differences in the intensity of the Amide II band suggested that collagen orientation and organisation could be affected by pH and supported the TEM observation of a disorder in the fibril arrangement [45
To assess the presence of mineralisation nuclei in areas where mineral nodules were not microscopically visible was a challenging and scarcely investigated application in the field of FT-IR spectroscopy. We were expecting to find a low amount of mineralised phases and close to the detectability threshold of the instrument. Furthermore, the heterogeneity of cell culture components, i.e., cells, matrix and culture medium, generated several bands through the spectrum which tended to conceal those object of study. Nevertheless, we were able to identify the phases of interest and to highlight the qualitative differences among the curves related to the culture conditions [27
]. The FT-IR analysis demonstrated that the bands corresponding to calcium phosphates were detectable even when the mineral nodules were not microscopically visible, thus suggesting that submicrometric mineralisation was beginning [50
]. In agreement with the result obtained by measuring the amount of Alizarin Red bound to the calcium complexes, the extent of the mineralisation did not change at pH 6.9, but the the marked presence of the octacalcium-phosphate phase was consistent with a deviation of the mineralisation process. Studies regarding mineral formation have demonstrated that the least soluble calcium-phosphate phase, hydroxyapatite, was preferentially formed under neutral or basic conditions while a low pH favoured the less stable phases, i.e., octacalcium phosphate which is a precursor of bone apatite formation [54
]. Our findings showed that FT-IR analysis provided relevant information by qualitative comparison of the curves. Based on these encouraging results, we could further exploit the FT-IR spectroscopy to determine in a larger number of samples the exact number of phases formed and to evaluate the differences among samples statistically.
To better support the results observed with biochemical, morphological and ultrastructural analyses, we also evaluated the expression of two genes that play an essential role in mineralization of the extracellular matrix [55
]. Osteonectin/SPARC is a secreted protein acidic and rich in cysteine that is required for the regulation of procollagen processing and assembly in the bone matrix, mineral incorporation and cross-linking. There is a close association between SPARC and collagen I expression, and its capacity to bind to collagen is a critical step of the mineralization process. Indeed, SPARC-null osteoblasts show similar levels of osteoblast differentiation markers, including bone sialoprotein, but the formation of mineralized nodules is impaired [56
]. Bone sialoprotein/IBSP belongs to the “small integrin-binding ligand N-linked glycoproteins” (SIBLING) family, which is an extracellular matrix protein family playing a critical role in the mineralisation process. Bone sialoprotein binds to calcium, induces nucleation of hydroxyapatite crystals in vitro, and is crucial for the structure of the mineralised matrix. Furthermore, IBSP can bind to collagen fibrils, especially to their hole zones that are the site of early mineral deposition [57
The acidic milieu determined a reduction in osteonectin/SPARC expression, thus providing a molecular basis for a possible explanation of the observed events, i.e., the decrease in collagen release, the ultrastructural disorganization in the extracellular matrix, the prevalence of precursor of bone apatite, and the reduced formation of mineralization nodules. Conversely, IBSP transcription was slightly increased, but the upregulation, while ensuring the precipitation of amorphous calcium-phosphate, was not sufficient to compensate for the altered phenotype, probably due to the excessive disorganization of the extracellular matrix.
On the basis of the above results, it was reasonable to hypothesise that the decreased citrate release induced by the acidic microenvironment could have been involved in a decreased ability for mineralisation. The role of citrate in driving the mineralisation process has been well recognised as it is an integral part of the apatite-collagen nanocomposite and contributes to controling the size, longitudinal growth and thickness of the apatite nanocrystals in achieving the typical plate-like morphology which ensures the biomechanical properties of bone, including stability, strength, and resistance to fracture [58
]. That the “osteoblast citration” is a fundamental step of bone formation has been conceptualised by Costello et al. (2012) who argued that “mineralization without citration will not result in the formation of normal bone, i.e., bone that exhibits its important properties, such as stability, strength, and resistance to fracture” [61
The aim of the next step of the study was to demonstrate that the impairment of the mineralisation process due to an acidic milieu could have been opposed by supplying the missing citrate, thus resulting in the same amount released by the hMSCs cultured under neutral conditions. For this purpose, the citrate supplementation used in the present experimental plan took into account the real deficit found under acidic conditions. Two different citrate sources were evaluated, namely Ca citrate and K citrate, which are also the citrate-based compounds commonly used in clinical practice.
When comparing the effect of the citrate supplementation under acidic conditions with what was observed under neutral conditions, the result could be considered satisfactory where significant differences were no longer observed, thus responding to the objective.
Both compounds exhibited alkalizing properties, but they were not able to fully restore the neutral pH as all the concentrations used raised the pH value of the culture medium to over 7.0 but the maximum increase, pH 7.27, was observed only with the highest dose of K citrate. Nevertheless, the functions for the most part impaired under the acidic conditions, i.e., collagen release and mineralisation, were partially or fully recovered after citrate supplementation, having different effects according to the compound used.
Overall, the intermediate doses were the most effective while the highest concentration, which was non-cytotoxic in the short-term preliminary test, showed an inhibitory effect on some parameters in the long term, and it cannot be excluded that this was due to the excess of Ca2+
in the culture medium. Ca2+
may influence not only the ion balance and the extracellular pH but also influence multiple cellular functions, including molecular pathways, ion channels and membrane transporter. Regarding potential cytotoxic effects, in different experimental settings, other authors demonstrated that elevated concentrations of extracellular Ca2+
, until 5 mM, promoted cell viability and late-stage osteogenic differentiation of MSCs, but may suppress early-stage osteogenic differentiation [62
]. A recent study of Gao et al. (2018) showed that MSCs could endure K+
concentrations ranging from 5 to 130 mM, but high concentrations may impair proliferation and induce apoptosis [63
]. In the above studies, Ca2+
concentrations were much higher than those employed in our experimental plan.
In the first phase of the culture, the supplementation was not able to correct the notable decrease in collagen release which was found under the acidic conditions. However, after 21 days, both compounds increased the collagen bioavailability, even more than in the control cultures, thus suggesting that the positive effect was ascribable to citrate irrespective of the source, although the presence of calcium induced a faster response.
Instead, the first 14 days were sufficient for the recovery of mineralisation and to make it similar to that observed under the neutral conditions. The Ca citrate was slightly more effective than the K citrate, but the difference became more evident in the late phase. In fact, at 21 days, the Ca citrate showed notable pro-osteogenic activity and, in spite of the unfavourable microenvironment, fostered the mineralisation process even exceeding what was observed at pH 7.4 while the initial positive effect exhibited by the K citrate seemed to be exhausted. This result confirmed what we demonstrated in previous studies in which was that, although it showed significant anti-osteoclastogenic activity, the K citrate was not as effective in promoting mineralisation [18
]. We cannot exclude that the decreased mineralization ability observed with the highest concentration of citrate-based compounds under acidic condition could depend on the decalcifying properties of citric acid. The capability to dissolve calcium compounds is also exploited for medical purposes, especially in dentistry, but the decalcifying properties strongly depend on pH value and concentration, usually lower than pH 6.9 and higher than 0.56 mg/mL, respectively [64
]. Both the above conditions differ significantly from those applied in our experimental setting.
While the release of collagen into the culture supernatant increased after citrate supplementation, the amount deposited in the extracellular matrix tended to decrease; however, this reduction did not correlate with the ability to accomplish the mineralisation process. Rather, the mineralisation observed at the final time point faithfully reflected the ability to release the collagen which was observed in the early phase of the culture, thus suggesting a sequential connection of events: the better the collagen bioavailability at 14 days, the better the mineralisation capability observed at 21 days; conversely, the higher the extent of the mineralisation observed at 21 days, the lower the need of new collagen deposition into the extracellular matrix. After treatment with citrate-based supplements, the expression osteonectin/SPARC and IBSP expression was highly variable and no significant changes were detectable.
Positive effects on the collagen release and mineralisation were also observed under neutral conditions, thus testifying that citrate concentrations exceeding what is available in the physiological microenvironment do not hamper the osteogenic potential of bone-forming cells. However, Ca citrate seemed more effective than K citrate in promoting the mineralisation, also favouring the collagen release and the IBSP expression. Additional evidence of the pro-osteogenic effect of the citrate supplementation emerged from the spectroscopy analysis which revealed that the extent of the submicrometric mineralisation activity in areas lacking in large mineral nodules was promoted by both supplements. This was correlated with the citrate concentration rather than with the type of compound, although in this setting the K citrate exhibited a slightly better effect than the Ca citrate. We had previously found that K citrate did not have a considerable osteopromotive activity and these results were herein confirmed [18
]. However, by making an in-depth study of the extracellular matrix, some aspects which had been undetected earlier were highlighted.
This study enhances the knowledge regarding the pro-osteogenic activity of citrate supplements in an acidic microenvironment, while previously published data were obtained mainly under neutral conditions. Costello et al. (2015) have demonstrated that the BMP2 induction of osteoblast differentiation and mineralisation was related to the citrate released from osteogenic precursors [19
]. Ma et al. (2018) have shown that extracellular citrate fostered a “metabonegenic” regulation of intracellular events in preparation for the osteogenic differentiation of the hMSCs. The citrate uptake affected downstream osteophenotype progression and favoured the metabolic switch from glycolysis to oxidative respiration to generate more ATP and meet the high energy demands required for the production of matrix proteins [20
For many years now, citrate-based supplements have been proposed in the clinical setting for treating patients affected by disorders of bone-remodelling, such as osteopenia and osteoporosis. In particular, Ca citrate is used when calcium supplementation is required for preventing bone loss [65
], and K citrate has been proposed as a strategy capable of opposing the acid overload and the deleterious effects on bone health status [66
]. However, the numerous clinical trials did not consider in their rationale that citrate, by itself, plays an essential role in maintaining bone health, as proved by the previously cited experimental studies and also supported by clinical evidence. Low citrate excretion has been found in a considerable proportion of osteopenic women [16
] and there is a strong relationship between urinary citrate excretion and the prevalence of fragility fracture in postmenopausal women [30
]; plasma citrate levels correlate with the bone mineral density of the lumbar spine and hip [36
]. By comparing the results obtained in this study with the results of interventional clinical trials, a “common ground” emerges. Overall, in the clinical setting, an adequate calcium intake is fundamental for preventing bone loss; however, Ca citrate seems to be more effective than calcium carbonate [67
]. Potassium citrate limits bone loss in postmenopausal women and elderly subjects, with or without osteoporosis [70
] and also anabolic effects with increased circulating levels of bone formation markers, i.e., PINP, have been reported [16
]; combined treatment K citrate and Ca citrate allows obtaining an additional decrease in bone turnover [76
]. Finally, positive effects of citrate supplementation have also been observed in the absence of an excessive acid load [77