Oral supplementation of inorganic pyrophosphate in pseudoxanthoma elasticum

Abstract

Pseudoxanthoma elasticum (PXE; OMIM 264800) is a rare heritable multisystem disorder, characterized by ectopic mineralization affecting elastic fibers in the skin, eyes, and the cardiovascular system. Skin findings often lead to early diagnosis of PXE, but currently no specific treatment exists to counteract the progression of symptoms.

PXE belongs to a group of Mendelian calcification disorders linked to pyrophosphate metabolism, which also includes generalized arterial calcification of infancy (GACI), and arterial calcification due to CD73 deficiency (ACDC). Inactivating mutations in ABCC6, ENPP1 and NT5E are the genetic cause of these diseases, respectively, and all of them result in reduced inorganic pyrophosphate (PP_i) concentration in the circulation. Although PP_i is a strong inhibitor of ectopic calcification, oral supplementation therapy was initially not considered because of its low bioavailability. Our earlier work however demonstrated that orally administered pyrophosphate inhibits ectopic calcification in the animal models of PXE and GACI, and that orally given Na₄P₂O₇ is absorbed in humans.

Here we report that gelatin encapsulated Na₂H₂P₂O₇ has similar absorption properties in healthy volunteers and people affected by PXE. The sodium-free K₂H₂P₂O₇ form resulted in similar uptake in healthy volunteers, and inhibited calcification in Abcc6^−/− mice as effectively as its sodium counterpart. Novel pyrophosphate compounds showing higher bioavailability in mice were also identified. Our results provide an important step toward testing oral PP_i in clinical trials in PXE, or potentially any condition accompanied by ectopic calcification including diabetes, chronic kidney disease or ageing.

INTRODUCTION

Pyrophosphate (PP_i) is an endogenous metabolite which not only has a crucial role in bone and teeth formation, but is also among the strongest inhibitors of ectopic calcification (pathological hydroxyapatite deposition). Two major mechanisms responsible for extracellular or systemic presence of PP_i have been identified. First, it was found that release of ATP from the liver is the main source of PP_i in the circulation¹. ABCC6, a transporter expressed in the basolateral membrane of hepatocytes facilitates ATP release to sinusoid capillaries, and ENPP1, a membrane-bound ectonucleotide pyrophosphatase/phosphodiesterase, effectively hydrolyzes ATP to produce PP_i that appears in the circulation. Mutations in the ABCC6 gene are associated with pseudoxanthoma elasticum (PXE, OMIM 177850), a late onset, slowly progressing monogenic mineralization disease^2–4. In PXE, calcification of the elastic fibers in the skin, clinically manifesting as skin folds and yellowish papules, represent the first sign of the disease. Mineralization of the Bruch’s membrane of the eyes is accompanied with peau d’orange, angioid streaks, and sometimes choroidal neovascularization and retinal bleeding resulting in macular atrophy. Arterial calcification affecting the media layer of arteries leads to peripheral arterial disease manifesting with intermittent claudication, hypertension and rarely myocardial infarcts. Mutations in the ENPP1 gene are responsible for generalized arterial calcification of infancy type 1 (GACI1, OMIM 208000⁵), a more severe disorder characterized by prenatal calcification of the internal elastic lamina of large- and medium-sized arteries and the joins. Both pathological conditions are characterized by low plasma PP_i concentration: ~40% of the control in PXE and virtually zero in GACI, and this is the common causal metabolic alteration in the two diseases. There are overlapping phenotypic features between the two disorders, and in a few cases mutations in the ABCC6 gene are also causative in GACI type 2 (OMIM 614473)^6,7.

Secondly, another mechanism regulating PP_i metabolism involves ANKH, which was identified earlier as a pyrophosphate transporter releasing the metabolite from the cytosol to the extracellular space⁸. A recent study, however, clearly proved that ANKH is an ATP transporter and its conversion to PP_i is due to ENPP1 hydrolytic activity⁹. Mutations of ANKH cause craniometaphyseal dysplasia, a bone dysplasia characterized by overgrowth of craniofacial bones and abnormal metaphyses of long bones¹⁰. Other proteins, such as tissue non-specific alkaline phosphatase (TNAP) and CD73, are also involved in the regulation of local or systemic PP_i levels [for reviews, see: ¹¹ and ¹²]. Similar calcification as that observed in PXE is present in patients with diabetes¹³, chronic kidney disease¹⁴ and can also occur as a result of ageing¹⁵. Thus, PXE can be considered a model to study age-associated medial arterial calcification and cardiovascular disease¹⁵.

The role of PP_i as an inhibitor of ectopic mineralization was established in the 1960’s¹⁶. The inhibitory effect of PP_i is based on blocking crystal growth centers on forming hydroxyapatite¹⁷. It was demonstrated in various animal studies that supplementation treatment by injecting PP_i effectively counteracts ectopic calcification: uremic vascular calcification was prevented in rats¹⁸ and the PXE-like spontaneous calcification was inhibited in Abcc6^−/− mice¹⁹. Daily transient increase of PP_i in the circulation was sufficient to inhibit mineralization in these animal models consistent with its proposed mechanism of action.

It was claimed that orally administrated PP_i is not effective in inhibiting ectopic calcification²⁰, and it was suggested that PP_i cannot appear in the circulation because of the hydrolytic pyrophosphatase activities in the gut²¹. On the other hand, oral administration would be preferred to treat patients lifelong in diseases such as PXE.

We have demonstrated in our previous work that in spite of the consideration cited above, PP_i given orally to Abcc6^−/− and Enpp1^−/− mice does inhibit ectopic calcification²². The two animal models recapitulate the calcification syndromes of the corresponding human diseases. We have also shown that orally given PP_i is absorbed in humans to an extent that it can elevate plasma PP_i level from pathologically low to above normal. Importantly, it was also shown that progression of ectopic calcification in PXE patients can be halted by treatment with a non-hydrolysable pyrophosphate analog, etidronate^23,24, and early administration of etidronate to Abcc6^−/− mice completely prevents the ectopic mineralization²⁵. Collectively, these results argue for the potential effectiveness of oral pyrophosphate therapy.

In our previous experiments we used the Na₄P₂O₇ form of PP_i and delivered it dissolved in drinking water. In spite of its clear inhibitory effect, this form has disadvantages: high ion load which results in gastro-intestinal (GI) discomfort and unpleasant bitter taste. In addition, it results in an unwanted excess of sodium intake. In the present study our aim was to overcome the above downside of administration, and to find the optimal chemical form and delivery method of orally given PP_i and to reduce or completely avoid the presence of sodium. We also asked the question if oral absorption of PP_i in patients with PXE is similar to that in normal volunteers.

METHODS

Approval of the human studies:

Oral uptake study involving healthy human volunteers (female and male) was approved by the National Review Board of the Ministry of Health, Hungary (ETT TUKEB). The permit based on the above approval has been issued by National Public Health and Medical Officer Service (NNK; authorization number: 16412–5/2021). Informed consent was obtained from each volunteer prior to the study and experiments conformed to the principles of Declaration of Helsinki what is indicated in the above document. All patient samples were handled in an anonymized form also approved by the above document.

Oral uptake study involving PXE patients was approved by the Genetic Alliance Review Board, and mutation analysis confirming the diagnosis in PXE patients was approved by the Institutional Review Board of Thomas Jefferson University. The diagnosis of PXE was based on characteristic skin findings, supported by characteristic histopathology, and ocular and vascular involvement.

Animals and animal studies:

The animal studies have been approved by the Ethical Committee of Animal Experiments, Governmental Office of Pest County, Hungary; permit No. PE/EA/748–2/2021 and XVI-I-001/707–4/2012, and were conducted according to the national guidelines.

All animals were housed in approved animal facilities at the Research Centre for Natural Sciences. Mice were kept under routine laboratory conditions with a 12-hour light-dark cycle and ad libitum access to water and chow. Anesthesia was carried out by intraperitoneal injection of the mixture of Zoletil (30 mg/kg, Virbac, France), Xylazine (12.5 mg/kg, Produlab Pharma, The Netherlands) and Butorphanol (3 mg/kg, Richterpharma, Austria) during all procedures.

Cryo-injury was performed as described previously^26,22, with 10 mM PP_i treatment provided as Na₂H₂P₂O₇ or K₂H₂P₂O₇ or no PP_i added (control) to the drinking water of female Abcc6^−/− mice²⁷.

For PP_i uptake studies in mice, three-month-old C57/BL6 female mice were fasted for 4 hours prior to PP_i administration. PP_i solution was given under anesthesia via gastric gavage (the dose equivalent to 39 mg/kg PP_i) in a volume of 5 μl/g. Blood was taken by cardiac puncture prior and after 10, 30 and 60 minutes (n≥3 mice at each timepoint) of administration, and collected into tubes containing 50 μl CTAD (BD, Franklin Lakes, NJ). Following filtration through Centrisart I 300,000 MWCO filters (Sartorius, Frankfurt, Germany), plasma PP_i was assayed as described¹.

Synthesis of pyrophosphate compounds

Detailed description of the synthesis of K₂H₂P₂O₇, (NH₄)₂P₂O₇, monoarginine-H₂P₂O₇, monolysine-H₂P₂O₇ and bisethanolamine-H₂P₂O₇ is given in the Supplementary Material.

Human uptake and plasma pyrophosphate assay

Na₄P₂O₇ (anhydrous, Code 118) and Na₂H₂P₂O₇ (both food grade) were purchased from ICL Food Specialist (St. Louis, MO, USA) and from Fosfa (Breclav, Czech Republic). The K₂H₂P₂O₇ form in Good Manufacturing Practice (GMP) quality was from ERAS Labo (Grenoble, France). Gelatin capsules were from Molar Chemicals (Halásztelek, Hungary), acid resistant HPMCP (hydroxypropylmethyl cellulose phthalate) type capsules were from Capsuline (Davie, FL, USA). Capsules were filled manually with the pyrophosphate salts to be tested the day before the uptake experiment, and were stored in a sealed container at room temperature until use.

Volunteers and PXE patients (after overnight fasting) were involved in the studies. The duration of ingestion was less than one minute. Blood samples were collected from the vena cubiti before ingestion (0 min) and after 30, 60, 120 and 240 min into CTAD anticoagulated tubes (BD, Franklin Lakes, NJ, USA), and filtered through Centrisart I 300,000 MWCO filters (Sartorius, Frankfurt, Germany). Plasma PP_i was assayed as described¹.

Plasma inorganic phosphate assay

Plasma inorganic phosphate was measured from ultrafiltered plasma used for PP_i assay since about 85 to 90% of serum phosphate is free and is ultrafilterable²⁸. We have tested that the anticoagulant mixture in the CTAD blood collection tubes does not interfere with the assay. Plasma inorganic phosphate was measured by ammonium molybdate method: 10 μl of plasma sample or calibration standard was added to a mixture of 0.3 ml reagent containing 2.5 M H₂S0₄, 1% ammonium molybdate, and 0.014% antimony potassium tartrate and 0.7 ml 20% acetic acid. For the reduction of the complex 0.15 ml of 1% ascorbic acid (freshly prepared) was added. Optical density was determined after 15 minutes at 880 nm.

RESULTS and DISCUSSION

To reduce the ion load and the amount of sodium in PP_i, we first turned to the Na₂H₂P₂O₇ salt form which contains half as much sodium as the previously utilized tetrasodium (Na₄P₂O₇) variant²². A case report has been published describing oral pyrophosphate treatment of a PXE patient utilizing Na₂H₂P₂O₇ dissolved in water²⁹. To avoid the unpleasant taste we used capsulated PP_i salts instead of water-based solution as the form of delivery. Two different capsules were tested: gelatin and HPMCP-type cellulose capsules, which can resist dissolution at the acidic pH of gastric fluid³⁰.

The human absorption experiments were performed and plasma PP_i levels were monitored as described in the Methods Section. We first recorded the absorption curves of three different delivery methods: Na₄P₂O₇ dissolved in drinking water (40 mg pyrophosphate/kg), and Na₂H₂P₂O₇ loaded into gelatin or HPMCP capsules (39 mg pyrophosphate/kg each), as shown in Figure1. While there are individual differences in the extent of absorption among the healthy volunteers in the case of each formulation, similar to our previous observations²², the absorption from the two different types of capsules was very different (compare panels B, C and D in Figure1), with gelatin capsules being more effective than HPMCP. Furthermore, both capsule-Na₂H₂P₂O₇ combinations were superior when compared to the Na₄P₂O₇-water solution (compare panels a and d). The major outcome of this experiment was that the introduction of gelatin capsule-Na₂H₂P₂O₇ formulation not only provided a solution for the problem of the unpleasant taste, but proved to be the most effective delivery method resulting in the highest absorption. In addition, the GI discomfort experienced by volunteers earlier with tetrasodium salt was reduced to the minimum. The two types of capsules release their cargo in different compartments of the GI system: gelatin capsules dissolve within minutes of reaching the stomach³¹, in contrast, HPMCP capsules can withstand the acidic pH of the gastric fluid, and release their cargo at a higher pH in the small intestine³⁰. We hypothesize that the significantly more effective absorption from gelatin capsule is, at least partly, due to the acidic environment of the stomach. At low pH the pyrophosphate anion carries less negative charge which makes passive diffusion through the organ barrier of the GI system more probable.

Figure 1.: Oral uptake of pyrophosphate in humans delivering different forms and formulations.

Plasma PP_i levels were determined by the luminescent method¹. Error bars represent SD on panels (A), (B) and (C), while mean and SEM are displayed on (D).

Panel A: oral delivery of Na₄P₂O₇-water solution, 40 mg/kg pyrophosphate, n=10.

Panel B: oral delivery of Na₂H₂P₂O₇ in gelatin capsule formula, 39 mg/kg pyrophosphate, n=9.

Panel C: oral delivery of Na₂H₂P₂O₇ in HPMCP capsule formula, 39 mg/kg pyrophosphate, n=7.

Panel D: comparison of uptake curves presented on Panels a, b and c, n= 10, 9 and 7, respectively.

After establishing the Na₂H₂P₂O₇-gelatin capsule as a more effective delivery method we asked the question: How is PP_i absorbed in individuals affected by PXE compared to its absorption in non-PXE volunteers? We tested the exact same experimental setup in the case of PXE patients using the same batch of Na₂H₂P₂O₇ and the same batch of gelatin capsules. Figure 2A displays PP_i absorption in PXE patients, showing individual differences, while on panel b the compared absorption curves of patients versus healthy volunteers confirm very similar uptake in the two groups. The peak PP_i concentrations are 5.5 ± 0.9 μM in the non-PXE group and 5.7±1.4 μM in the PXE group. Baseline concentrations, as expected, are different, 1.3±0.1 and 0.5±0.1 μM in the non-PXE and PXE group, respectively (Figure 2C). Gastro-intestinal discomfort was not reported.

Figure 2.: Absorption of orally delivered pyrophosphate in PXE patients.

Plasma PP_i levels were determined by the luminescent method ¹. Error bars represent SD on panel A and C, and SEM on panel B.

Panel A: oral uptake of Na₂H₂P₂O₇ in gelatin capsule formula by PXE patients, 39 mg/kg pyrophosphate, n=9.

Panel B: comparison of uptake curves of healthy volunteers and PXE patients taking 39 mg/kg Na₂H₂P₂O₇ in gelatin capsules, n=9.

Panel C: baseline plasma pyrophosphate levels of healthy volunteers and PXE patients involved in the study (n= 9, p-value calculated by Mann-Whitney U test).

Next, we have embarked upon the development of sodium-free pyrophosphate forms to avoid excessive delivery of sodium. Figure 3A shows the chemical structures of compounds synthesized for this study (K₂H₂P₂O₇, (NH₄)₂P₂O₇, monoarginine-H₂ P₂O₇, monolysine-H₂P₂O₇ and bisethanolamine-H₂P₂O₇) along with Na₂H₂P₂O₇. The synthesis and purification of the above compounds are detailed in Supplementary Materials.

Figure 3.: Absorption of sodium-free pyrophosphate compounds and Na₂H₂P₂O₇ in mice.

Panel A: the chemical structure of the pyrophosphate compounds studied.

Panel B: Plasma PP_i levels of mice after dosing with various pyrophosphate compounds via gastric gavage. Dose was 39 mg/kg pyrophosphate in each case. Mean ± SEM are displayed, nγ3 at each data point.

We have tested the absorption of the novel PP_i salts in mice. Oral delivery for the absorption experiments was carried out as follows: three-month-old female C57/BL6 mice were fasted for 4 hours prior to PP_i administration. PP_i solution was given under anesthesia via gastric gavage (the dose equivalent to 39 mg/kg PP_i). Blood was taken by cardiac puncture before and after 10, 30 and 60 minutes of delivery (n≥3 mice at each timepoint), and plasma PP_i was assayed as described in the Methods section. We have observed PP_i absorption in the case of each form, however the two amino acid derivatives, the monoarginine-H₂P₂O₇ and the monolysine-H₂P₂O₇ resulted in the highest concentration at the maximum of the absorption curve (15.6±1.5 and 13.5±1.3 μM, respectively). The potassium and sodium forms showed approximately same maximum (9.4±0.9 and 10.4±2.7 μM, respectively), while administration of the bisethanolamine and the (NH₄)₂PP_i derivatives yielded only 5.3±0.5 and 3.4±0.6 μM peak plasma concentrations, respectively.

Currently, only the K₂H₂P₂O₇ salt is available in GMP quality (see Material and Methods), and therefore, following our protocol utilized in the human studies (oral administration of a 39 mg PP_i/kg dose loaded in gelatin capsules) we investigated the properties of the potassium salt, K₂H₂P₂O₇. We designed this experiment with the same healthy volunteers in the Na₂H₂P₂O₇ and K₂H₂P₂O₇ treatment groups, as we wanted to establish a clear comparison of the two salt forms.

Figure 4A summarizes the uptake curves of all volunteers involved in the study while the individual absorption plots of six healthy volunteers are presented on Figure 4C. Figure 4A and C panels again indicate large individual differences similar to those observed in the case of the sodium forms (see above). The individual curves demonstrate better PP_i uptake in two cases when K₂H₂P₂O₇ was taken and approximately the same extent of absorption as with Na₂H₂P₂O₇ in the case of the other four volunteers.

Figure 4.: The potassium pyrophosphate K₂H₂P₂O₇ shows similar absorption characteristics in humans as Na₂H₂P₂O₇, and plasma inorganic phosphate remains in the normal range. K₂H₂P₂O₇ and Na₂H₂P₂O₇ inhibit dystrophic cardiac calcification of Abcc6^−/− mice when given orally.

Panel A: oral uptake of K₂H₂P₂O₇ in gelatin capsule formula by healthy volunteers, 39 mg/kg pyrophosphate. Mean ± SD values are displayed, n=6.

Panel B: Plasma inorganic phosphate (P_i) levels of volunteers after dosing with K₂H₂P₂O₇ remain within the normal range of 0.81–1.45 mmol/L, indicated by shaded area. Mean ± SD values are plotted, n=6.

Panel C: Individual plasma PP_i (pyrophosphate) curves to compare the absorption in the same healthy volunteers from K₂H₂P₂O₇ (solid line) versus Na₂H₂P₂O₇ (dashed line). Mean ± SD values are displayed.

Panel D: total Ca²⁺-content of the heart tissue was measured; mean ± SEM are graphed, p-values indicated were calculated by Mann-Whitney U test, animal numbers in each treatment group (n) are displayed on the figure.

Pyrophosphate is hydrolyzed in the gut to inorganic phosphate (P_i) and excreted via urine and feces as indicated by a study using rats³². We have determined P_i concentration in the blood of volunteers during the time course of K₂H₂P₂O₇ uptake (Figure 4B).

An increase of plasma P_i was detected from 0.96 to 1.34 mmol/L and with much smaller individual differences than in case of PP_i in plasma. These values are within the normal range of 0.81–1.45 mmol/L³³.

Next, the capacity of the K₂H₂P₂O₇ salt in inhibiting ectopic calcification in Abcc6^−/− mice was tested by the cryo-injury based DCC-method as described in our earlier papers^19,22. In this experiment we induce calcification in the myocardium and five days after we determine the Ca²⁺-content of the heart. As it is demonstrated in Figure 4D, this treatment does not induce mineralization in wild type mice but a massive calcification develops in the Abcc6^−/− animals. Both Na₂H₂P₂O₇ and K₂H₂P₂O₇, when given orally (10 mM in drinking water), reduce calcification to the same extent, indicating that the newly synthesized potassium salt is as effective as the sodium form. Earlier we have demonstrated a similar inhibitory effect using the Na₄P₂O₇ form²².

In summary, we have shown that a sodium-free PP_i derivative, K₂H₂P₂O₇, is absorbed in humans when given orally, similar to Na₂H₂P₂O₇. This overcomes the problem of the excess sodium intake.

Importantly, we demonstrated that K₂H₂P₂O₇ effectively inhibits calcification in Abcc6^−/− mice. Our results prove that formulation in gelatin capsules provides the highest uptake. These findings suggest that the K₂H₂P₂O₇-gelatin capsule formulation is a suitable candidate for therapeutic trials. Furthermore, we have identified two novel pyrophosphate forms (monolysine- and monoarginine derivatives), showing high absorption in animal experiments when given orally.

Large individual differences in uptake of pyrophosphate both in the group of healthy volunteers and in the group of PXE patients were observed. The dose we have applied was chosen to provide at least 2.5 fold elevation of plasma pyrophosphate concentration at the peak level even in the case of the lowest absorption efficacy (see Figure 4). It is notable that a daily transient increase of PP_i in the circulation was sufficient to inhibit mineralization in Abcc6^−/− mice¹⁹. The observed large individual variation in absorption may indicate a necessity of personalized dosing established prior therapy or a clinical trial. Gelatin capsules represent a suitable formulation for oral administration of PP_i in double-blinded clinical trials, as both the therapeutic compound and the placebo can be delivered by the same type of vehicle, masking the characteristic salty-bitter taste of pyrophosphate salts.

Medial arterial calcification observed in PXE is also present in patients with diabetes¹³, chronic kidney disease¹⁴ and as a result of ageing¹⁵. PXE can be considered as a model to study medial arterial calcification in cardiovascular disease, and research can also lead to novel therapeutic interventions which would be relevant not only in PXE but could also reduce calcification and related cardiovascular risk in the above mentioned populations³⁴. The intervention based on oral administration of pyrophosphate is one of the promising candidate approaches. Our study, in order to find the best chemical form and administration method of pyrophosphate, is an important step toward this direction. Indeed, a clinical trial in PXE based on the oral administration of pyrophosphate³⁵ has already been registered.

Abbreviations

GACI

generalized arterial calcification of infancy

HPMCP

hydroxypropylmethyl cellulose phthalate

PXE

pseudoxanthoma elasticum

PP_i

inorganic pyrophosphate

Data Availability Statement

No datasets were generated or analyzed during the current study.