Strontium coupling with sulphur in mouse bone apatites

1 Introduction

Calcium is the fundamental elemental component of bioapatites. Bioapatites are the constituents of both the skeleton and teeth. These natural materials are really wonderful in fulfilling the complicated roles of the organisms, but still not perfect. Numerous diseases, aging, and accidents leading to the weakening or devastation of hard tissues in an organism incline us to look for ways to possibly improve the original material. One of the ideas was the introduction of Sr instead of Ca in relevant positions in bioapatites to weaken Ca release from the hard tissues [1], strengthen teeth and bones [2], increase the bone mineral density, and limit the fragility of bones. Many elements naturally contribute to bioapatites, strontium being one among them. Strontium is heavier than calcium, the element from the second group of the Periodic Table, and is chemically similar to its lighter neighbour. The abundance and availability of strontium in the environment are not as common as calcium and depend on local conditions concerning the soils and associated minerals [3].

The substitution of Sr into apatite is allowed from the steric and stoichiometric points of view [4]. The bones treated with Sr become stiffer, harder, and tougher. While a normal diet has an average of 4 mg level of the element [5,6], which was observed even for foetuses [7], about 99% of attached Sr supplies bones and teeth [8]. It is equivalent to 0.035% Ca included in those organs. Moreover, Sr prefers to be included in cancellous rather than in cortical bones and new and external parts of the tissues in comparison with the old ones [9,10]. Two kinds of mechanisms are considered to act during strontium involvement: the rapid one on a surface with only 0.65% of Ca available from the steric point of view [11] and the long one, allowing the penetration of bulk bones [12], leading to irreversible location [13]. In this situation, one cannot expect to repair the whole damaged old bone but either exchange the old fragment into the new one or only supplement it with a new layer. Nevertheless, the risk of fractures in postmenopausal osteoporosis was diminished by 31% after the systematic administration of strontium ranelate [14]. The mapping of Sr presence in bones mirrors the ossification process location [15]. Most popular mappings of Sr in bones can be obtained using different versions of X-ray fluorescence techniques [16,17].

One of the more interesting options for Sr dosing is the use of novel injectable Sr-containing cements [18]. Probably the greatest effect is attained when Sr is administered locally with Sr-bearing modified materials [19]. Finally, the dosed Sr is rapidly removed after stopping the supply [20], but it concerns only the weakly adsorbed ions. Sr is included relatively nonuniformly in the hard tissue [21,22]. Although Sr can even fully exchange Ca in apatites [23,24,25], it is mineralogical [26] and not from the biological perspective. We must emphasize that the presence of Sr ions introduces some physical and chemical disturbances. The ionic radius of Sr is 1.13 Å, the greater one in comparison with the somewhat smaller Ca ion radius equal to 0.99 Å. This increase lies inside the approximate limit of 15% of radii difference, allowed by the Goldschmidt rule for ionic exchanges in isomorphic substances, and results in the growth of crystallographic parameters a and c for the hexagonal system. Another difference is in the values of Pauling electronegativities [27], which are 1.00 for Ca and 0.95 for Sr, and this means that Sr is slightly more electropositive and perhaps reveals less affinity towards organic matter. It is worth noticing that Sr involvement always increases the solubility of apatites and bioapatites [28,29], which is not necessarily an intuitive effect while observing the tendencies in solubility of some simple compounds of the elements from the second group of Periodic Table, e.g. sulphates.

Since the pharmacological use of Sr specimens never aims at the total substitution of Ca in apatites, one should consider the question of low and high doses. Additionally, the affinity of Sr towards apatites does not exclude the potential toxicity and harmful side effects (especially rickets [30], sometimes resulting from the perturbation in acidic phospholipid metabolism, and the cardiovascular risks [31,32]). According to some reports, the low level is estimated as ≤4 mmol Sr·kg⁻¹·day⁻¹ [33]. Such concentrations have no consequences for bioapatites [34]. The high doses of Sr lead to hypocalcaemia resulting from the excretion of Ca [35] and to the deficiency of P resulting from forming insoluble strontium phosphates [36]. High doses of Sr on the order of 20–100 µg·mL⁻¹ inhibit the mineralization reactions and mineral formation, and they induce apoptosis [37,38]. The presence of Sr in bioapatites is considered mainly to influence the mechanical features of teeth and bones, especially hardness. Nevertheless, the action of Sr is more universal [19]. Sr promotes bone formation by osteoblasts and suppresses bone decay by osteoclasts [39,40,41,42,43,44,45,46]. Fe is the next and only element exerting similar effects [47,48,49]. The joined introduction and the accumulation of action of these two elements in hydroxyapatite and bioapatite are seriously considered [50,51]. In addition, Sr exerts quite significant antimicrobial influence [52]. As an additional action, the moderately positive influence of Sr administration on the implant osseointegration was observed [53,54]. Atkins et al. [55] rather promoted the opinion about the increased differentiation of osteoblasts as the main effect. Riedel et al. [56] claimed that the positive effects in curing osteoporosis can be obtained if one optionally uses strontium or fluoride medicaments. The action of Sr is even enhanced if the organisms are treated with a low-Ca diet [57]. There is another important question: in which form Sr should be administered – as the simple inorganic salt or the expensive organic complex, e.g., the ranelate? The newest study by Turżańska et al. [58] indicates that there is no statistically significant difference in both treatments.

A relatively small number of publications are devoted to the influences of Sr on other chemical ions. Querido et al. [59], Rossi et al. [60], and Huang et al. [61] have noticed that increased Sr involvement in the crystallographic network of bone results in an increased presence of CO 3 2 − anions in the crystals. On the contrary, Grynpas and Marie [62] determined the loss of carbonates associated with the growth in Sr administration at low doses. The theoretical considerations by Terra et al. [63] indicate the loss of OH^– ions in similar circumstances.

The main aims of this article are the following:

• To check if the mechanisms of involvement of Sr inside bone bioapatite are the same as for the other similar elements, such as Mg or Pb;

• To study the question if the introduction of Sr to bioapatites brings on changes in the forms of apatites and inorganic components constituting a part of bone. According to the earlier papers by Jabłoński et al. [64] and by Neufeld and Boskey [30], there are solid suppositions that Sr influences cells of growing bones and thus bioapatites;

• To estimate if the administration of various Sr compounds brings similar effects.

2 Materials and methods

Our considerations are supported by our studies on mouse bones. Our data concern the experiments with the bones of mice treated with Sr ranelate and Sr chloride.

3 Results

Figure 1 shows the spatial distributions and some inter-elemental correlations in mouse bones. When Sr is administered to mice, the enrichment of bones with Sr is widely scattered, depending on the locations. The important thing is that coupling of Sr with S is quite obvious, and the relationship is nearly linear. The location of Sr is prevalent in the narrow external (periosteum) and internal (endosteum) layers of the cortical bone, with an additional ring structure inside the bone. It is observed in all the studied scans. One must be aware that the directions of scans inside separate bones were selected radially but randomly if it concerns the orientation of the cross-section. As a result, it seems to form pipe-like structures, following the structures of the periosteum (mainly) and endosteum with the additional ring inside, greatly enriched in Sr and S. Astonishingly, the patterns of the distribution of Sr inside the cortical bone are precisely repeatable even if we administer different compounds (chlorides or relates) in the bones of different individuals (Figure 1e). Not only the locations of Sr zones are strictly the same, the same concerns the composition proportions. It is possible to observe in Figure 1e, where the approximately 2-fold increase in Sr invoked by equimolar supply with chloride and ranelate (1Sr : 2Sr) is very well mirrored after quantifying spectral signals in bones. Figure 1a and c shows a worth emphasizing fact that the molar proportion between Sr and S is approximately equal to 1. Finally, Figure 1d suggests that although Sr is supplied from the outside, it is not the case for S. The total amount of S rather does not change and obeys only the variability resulting from the internal re-transportation. The roughly even primary distribution of S in control samples transforms into such a space arrangement that S follows incoming Sr in Sr-treated samples. If we can easily understand the strictly external structures, since they are supplied with Sr by the body fluids from the body or marrow sides, the internal ring can be probably explained as the location of the collision of two diffusive fronts.

Figure 1

(a) Correlation between the changes at molar S and Sr levels detected for mice cortical bones treated with strontium chloride; (b) triple pipe-like structure of Sr and S involvement in mouse Sr-treated bone with the delimitation of bone cross-section by the Ca signal; (c) stricter illustration of the interrelation between Sr and S – see the molar ratio; it is supplemented with the line extracted from the relevant backscattered electron image of the cross-section of bone along the line of the EDX–SEM scan – see the periosteum and endosteum zones; (d) comparison of the sulphur distribution in the original bone and in the Sr-transformed bone (after SrCl₂ treatment), the distances presented in tailored length units; (e) the distance-tailored analogy between the action of chloride and ranelate in two different bones. The left sides of the figures are boundaries with the body volume; the right sides border with the trabecular bone.

4 Discussion

All the proposed chemical equations are related to the apatite formula and are presented with integer coefficients, just for the simplification and better visualization of the results. They present the tendencies in the reactions and not the fully real formulae. The entry of Sr in bone bioapatite results in a rigorous relationship with S amounts. The influence of Sr on S involved in bioapatite is undeniable.

The studies on the interaction of Sr with other chemical entities present in bioapatites bring very interesting results. If it concerns the sulphates, the simplest explanation would be invoking a well-known substitution of carbonates instead of phosphates together with the introduction of Na⁺ instead of Ca²⁺ for the balancing of charges.

where the product on the right side is more similar to bone and enamel than the original mineral-type apatite. We can imagine the analogous variant for sulphates:

It establishes the constant level of sulphates in an inorganic matrix (Figure 1d, for the standard). Next, we have

In fact, tracing carefully Figure 1d, we can observe that the attack of Sr is the main reaction, and only after that we have the rearrangement of the sulphates in Sr-treated samples. We can think about the optional reaction inside bioapatite, where Sr locally attracted sulphate and expelled carbonate:

instead of what would be expected if Sr goes inside the structure in the simplest ion exchange (this mechanism can be called the mineralogical one):

And here, no carbonate or sulphate ions are necessary. Coming back to the compound presented as the product in Eqs. (3) and (4), one can formally split it into the shape well showing affinity for Sr and SO 4 2 − :

The action of Sr is complex. As we know, the introduction of carbonates increases the surface available for reaction in crystallites of bioapatites, and it concerns the situation presented in Eq. (4). Moreover, the introduction of Sr in the structure of the bone leads, on the one hand, to the presence of heavier cations in bones, but on the second hand to the loosening of the structure by the arrival of sulphates. In general, we observe a meaningful increase in the particle asymmetry. It is against the process of enamel maturation, as has been shown in our paper concerning teeth [66], where the product adjacent to the enamel surface was closer and closer to the pure apatite, losing a part of the ionic admixtures

But reaction (5) leads to the formation of the passive layer, as it is in opposition to the situation with the activation of periosteum in the bone. The introduction of Sr, by coupling with sulphates, can be in accordance with the process of bone aging [24,67] and not necessarily with what orthopaedists want to achieve. But the greatest problem with bones is different. As it is obvious from Figure 1c, the proportion between Sr and SO 4 2 − is ∼1:1 at the molar level. At the same time, there is no growth in Na amounts in the locations with the maxima of Sr and S and only a nearly negligible growth in K contents (here not shown). It can testify that Sr enters the material which is already saturated with Na and CO 3 2 − (Eq. (1)) before the action of Sr, and the reaction occurs according to Eq. (4), and those processes are separated in time. Only after the introduction of Sr, the sulphates are rearranged, and we come back to Eq. (4). This reaction pattern explains that we cannot see any increase in Na in the position of Sr and S peaks, and it probably corresponds best to the experimental facts.

Another possibility is that the reaction occurs according to the following equation:

in which the formulation of the product substance emphasizes a huge affinity for sulphur and strontium. This version is in accordance with the fact that the amount of Na does not increase during Sr introduction, the amount of CO 3 2 − does not decrease, and the amount of phosphate group decreases to a greater degree than the amount of calcium, but the latter effect is small. It is indeed observed in our experiments. For the same reasons, we must exclude the version that Sr and SO₄ ^2– arrive strictly at the same time in apatite. The product in Eq. (7) is rather nonapatitic.

Only a small part of the total amount of particles will be in the form presented in Eqs. (3) and (4), but it is noticeable. We suggest that all the changes evidenced in these equations occur in three apatite particles – full hexagon [66] or the neighbouring particles in a very limited space of the crystallite. The greater size of the Sr²⁺ ion enters inside the particle if the available space is widened. It happens when the phosphate ion is exchanged for a carbonate one and the calcium cation for a sodium one, with the liberation of the part of space covered by the original ions. The approximate balance of volumes of ions entering and going out of the crystal (Eqs. (1) + (2)) is given by:

which translates into the following values:

[68] which gives after summarization:

0.0797 < 0.0972.

But the introduction of sulphates instead of carbonates (Eq. (2)) leads to the volume relaxation:

0.0982–0.0972.

In the case of carbonates, the summary volume of the introduced ions is clearly smaller than the volume of the outgoing ions. It would testify to the loosened structure of the apatites after the administration of Sr. However, the replacement of the carbonate ion with the sulphate one with the volume estimated as 0.0611 nm³ leads to the volume (0.0982), which is equivalent to the original value. In such a situation, the improvement of the apatite structure in osteoporosis treatment would result from the induction of osteoblasts and suppression of osteoclasts and not from the physical tightening of the structure.

The compounds NaCaSr(PO₄)(SO₄) · 2Ca₃(PO₄)₂Ca(OH)₂ can also be formally presented as NaCa(PO₄) · SrSO₄ · 2Ca₃(PO₄)₂Ca(OH)₂. The products resulting from Eqs. (1)–(4) are highly asymmetric, in a higher energetical state, and potentially more soluble than the particles of hydroxyapatite.

There is another interesting thread when we consider the involvement of strontium and sulphates. One of the more promising substances applied for bone substitutions is CaSO₄ [69], a compound characterized by relatively good osteoconductivity and the ability to supply Ca ions for the regeneration of hard tissues. The worst side of the application is that α-hemihydrate calcium sulphate with added nano-hydroxyapatite (nHA) can exert an inflammatory effect due to overheating during the chemical reaction. In medical practice, it is counteracted by the addition of metformin [70]. In the newest versions, the strontium-containing α-hemihydrate calcium sulphate (Sr-α-CSH) [71] mixed with added nHA and loaded with aspirin is recommended for its biocompatibility, osteoconductivity, and stabilized degradation rate [72]. Perhaps, the nature, earlier than the researchers cited a moment ago, invented the small admixture of sulphates to the hydroxyapatite to improve the biocompatibility and other advantageous features of synthesized bones [73].

The results concerning the deposition of Sr inside bones are extremely interesting from the geometrical point of view. Although it has been known for years that Sr goes into the external layers of bones, we have shown that the covered zones are relatively very narrow (each up to ∼10% of the whole radius of the mice cortical bone) and located from both sides of the bone (they overlap each other with periosteum and endosteum zones), while the third zone is placed inside the bone. The third zone seems to result from the meeting of the two diffusing waves going from the two ends of the bone. On the transverse cross-sections of bone, it has a pipe-like structure. The only paper of other authors showing this phenomenon is that by Lima et al. [74]. The concentration of Sr in the most external layer of bone (endosteum) is even five-fold greater than the average concentration in bone. The profiles resulting from the supply with SrCl₂ do not differ from those resulting from the action of Sr ranelate. The profiles of Sr are shaped totally by the external supply. Sr distribution is strictly connected with that of S. But the sulphur profile results only from the regrouping of existing amounts of sulphur inside the bone. It seems that strontium stops in restricted areas and attracts the sulphate groups and forces them to relocate. The quantitative comparison suggests that the interactions in a frame of our experiments lead to equimolar equivalence 1:1 Sr/S. Of course, the substitution of the phosphate ion with the sulphate one brings the charge imbalance. Unfortunately, we either did not find the increase in concentrations of Na and K ions or only a very small one, within the limits of statistical error. In that situation, we postulate either the increase in the defect number or the reactions according to Eqs. (4) and (7). These equations postulate the increased asymmetry of new-formed particles. The new particles should be short-living. Such a situation would be in accordance with the increased bone turnover observed during the Sr treatment of bones.

One should recall the basic facts: if Ca↓, P↓, then Sr↑, S↑, and Na ∼const. We explained it by invoking to the ion-exchanges described in Eqs (4) and (7). But we can imagine another explanation. Then, the role of sulphated glycosaminoglycans should be considered.

The first option is that: observe the diagram concerning the standard sample in Figure 1d: if the role of sulphated glycosaminoglycans in the formation of bones is important, then we cannot observe any special role of them in the periosteum and endosteum areas of the standard samples, where most intensive bone formation occurs. This means that the sulphated glycosaminoglycans work evenly in the whole bone, and the surface regions are not more preferred than the interior of the bone. The implication is that the bone should be evenly formed everywhere inside the sample, and it is not true. So, the simple action of the sulphated glycosaminoglycans should be excluded.

In that situation, we must consider another role of the sulphated glycosaminoglycans. In such situation, the introduction of Sr would induce the sulphated glycosaminoglycan activity. Sr would attract the sulphated glycosaminoglycans from the whole bone (see the diagram concerning bone saturated with Sr in Figure 1d). It would suggest some special mechanism of bone formation, important only in the periosteum and endosteum areas and only in the presence of Sr. We do not assume that only the presence of Sr starts the sulphated glycosaminoglycan activity.

What argues in favour of the version of ion exchange? Mainly the strict to equivalent stoichiometry of the relation between Sr and S. Next, the instability of the external layers full of Sr, which can result from the asymmetry inside the particle suggested in our equations. In turn, the great mobility of S observed in Figure 1d and missing Na ions following sulphate ions should suggest the prevalent role of sulphated glycosaminoglycans.

We suppose that the influence of Sr can be even stronger on cartilage, in which the presence of sulphate-containing compounds is greater, but it is outside the scope of our contribution.

5 Conclusions

Strontium exerts some not fully recognized influences on hard tissues. It was detected that the presence of Sr in bioapatite is strictly followed by sulphate ions. The sulphate ions came from the previously present tissue entities, by their regrouping. Both ions are concentrated in the pipe-like structures on the edges of cortical bone (periosteum and endosteum) and some locations in the middle of it. This structure does not depend on the kind of supplied Sr compound. The possible mechanism of involvement of Sr and SO 4 2 − in the crystallographic network of mouse bone apatite is explained by Eqs. (3), (4), and (7) in the text and leads to the idealized structures NaCaSr(PO₄)(SO₄) · 2Ca₃(PO₄)₂Ca(OH)₂ and CaSr(SO₄)(CO₃) · 2Ca₃(PO₄)₂Ca(OH)₂. It is a different mechanism than the one responsible for the introduction of Sr to mineralogical forms of apatites. In our research, the growth of Sr contents is directly related to the growth of sulphur amounts in mouse bones. Several explanations were presented in the text, still the most probable of which is the synthesis of Sr and SO 4 2 − containing apatites, where the admixtures possibly enter one apatite molecule. In general, Ca²⁺, PO 4 3 − , Mg²⁺, Cl^–, and CO 3 2 − ions participate in the main cycle of ion exchanges in bioapatites. Sr²⁺, CO 3 2 − ions, SO 4 2 − , and perhaps OH^– participate in the secondary ion exchange cycle in bioapatites. The asymmetry and better solubility of bioapatites with involved strontium and sulphate ions explain their greater reactivity and greater turnover of bone sulphur role. Alternately, the second mechanism of sulphur role is cited, namely by the action of the sulphated glycosaminoglycans in bone formation. Perhaps, the great mobility of sulphate ions in the presence of strontium is an argument for this option. Astonishingly, the cooperation of Sr²⁺ and SO 4 2 − ions in bones has been overlooked in up-to-date studies.