Novel TRPV6 mutations in the spectrum of transient neonatal hyperparathyroidism

Maternal–fetal calcium (Ca2+) transport in the placenta plays a critical role in maintaining fetal bone mineralization. Mutations in the gene encoding the transient receptor potential cation channel, subfamily V, member 6 (TRPV6) have been identified as causative mutations of transient neonatal hyperparathyroidism due to insufficient maternal–fetal Ca2+ transport in the placenta. In this study, we found two novel mutations in subjects that have transient neonatal hyperparathyroidism. TRPV6 carrying the mutation p.Arg390His that localizes to the outer edge of the first transmembrane domain (S1) showed impaired trafficking to the plasma membrane, whereas TRPV6 having the mutation p.Gly291Ser in the sixth ankyrin repeat (AR) domain had channel properties that were comparable those of WT channels, although the increases in steady-state intracellular Ca2+ concentration could have led to Ca2+ overload and subsequent death of cells expressing this mutant channel. These results indicate that the AR6 domain contributes to TRPV6-mediated maintenance of intracellular Ca2+ concentrations, and that this region could play a novel role in regulating the activity of TRPV6 Ca2+-selective channels.


Background
Calcium (Ca 2+ ) is essential for many physiological functions. Blood Ca 2+ levels are finely maintained by the parathyroid hormone (PTH)-vitamin D axis that modulates bone resorption and intestinal Ca 2+ absorption. In fetuses and neonates, additional Ca 2+ transport occurs to maintain bone mineralization [1][2][3]. In many mammalian species, including humans, fetal blood Ca 2+ levels are higher than those in mature individuals, suggesting that the transport is mediated by uphill or energy-consuming transport through epithelial, transcellular, rather than paracellular routes [4,5]. Ca 2+ via the transcellular pathway is achieved via three stages: (1) Ca 2+ uptake through a channel driven by an electrochemical Ca 2+ gradient; (2) binding to calbindin that does not increase free calcium concentrations, and (3) basolateral extrusion of Ca 2+ by plasma membrane Ca 2+ -ATPase or Na + /Ca 2+ -exchangers [6,7]. However, the proteins involved in regulating this pathway and the associated regulatory mechanisms for maternal-fetal calcium transport remain unclear.
Transient neonatal hyperparathyroidism is a neonatal bone disorder caused by insufficient fetal bone mineralization due to impaired maternal-fetal Ca 2+ transport across the placenta. Recently, mutations in the gene encoding transient receptor potential cation channel, subfamily V, member 6 (TRPV6) were identified as causative mutations of transient neonatal hyperparathyroidism (TNHP) [8]. There are several hotspots for TRPV6 mutations: the outer edges of the transmembrane domain S2 and S3 that affect trafficking to the plasma membrane; the fourth ankyrin repeat (AR) domain that affects protein stability; and the intracellular S2-S3 loop that affects intracellular Ca 2+ -dependent inactivation.
The different mutations suggest that there are several mechanisms for TNHP disease onset. Moreover, a combination of mutations could also be important for disease pathogenesis since many patients carry compound heterozygous mutations. Here we report novel mutations in the S1 transmembrane domain and the sixth AR domain. The functional significance of these mutations revealed an unexpected role for the AR domain in TRPV6 activity.

Subjects
Subject 1 This male patient was born to a 39-year-old G2P1 mother of Japanese descent; the father was 44 years old and also of Japanese descent. The couple was nonconsanguineous. They had an older daughter who had shorter limbs at 8 months of gestation, but is currently normal. The father had a funnel chest and shorter limbs and there was a family history of similar physical features. Genetic information for the parents and other family members was not available.
Fetal ultrasound at 29 weeks of gestation showed polyhydramnios and thoracic narrowing with rib deformities. Although the mother was treated with uterine contraction inhibitors to prevent premature delivery, uterine contractions began at 36 weeks of gestation and the child was born by a scheduled cesarean section the day after contractions began. The birth weight was 2.351 kg and the length was 41 cm. The occipitofrontal circumference (OFC) was 34.8 cm. Apgar scores were 4, 5, 8 at 1, 5, 10 min, respectively. Artificial invasive ventilation was initiated on day 0. On day 8, the child was taken off the ventilator, but retractive breathing and hyperpnea occurred with normal SpO 2 level. Therefore, neutrally adjusted ventilatory assist was applied on the same day as ventilator removal.
Radiographs showed a narrow thorax, femoral shortening and lower bone mineral density with ossification failure, which was suggestive of hyperparathyroidism. Cord blood levels of intact PTH (iPTH) were extremely high (high sensitive PTH was > 3200 pg/ml and iPTH was 1371 pg/ml), which, together with lower than normal levels of Ca 2+ and vitamin D and maternal hypovitaminosis D, led to a diagnosis of hyperparathyroidism secondary to maternal vitamin D deficiency. I-cell disease was not likely since lysosomal enzyme activity was normal. The subjected was treated with vitamin D, Ca, and P, which improved iPTH, alkaline phosphatase (ALP), Ca, P, and vitamin D levels. Thorax and bone mineral density was also improved by this treatment. After removing the ventilator, the noninvasive positive pressure ventilator was used until the child reached 13 months of age.
Subject 2 This male subject was born to a 29-yearold G0P0 mother of Japanese descent. The father was 30 years old and also of Japanese descent. The couple was healthy and non-consanguineous. They had a younger daughter who had the same genotype as the subject, but she did not have any complications or skeletal abnormalities (Fig. 1b). Fetal ultrasound detected no fetal abnormalities. The subject was delivered vaginally at 38 weeks of gestation. The birth weight was 2.805 kg (43 cm height, 35 cm OFC). Apgar scores were 7, 7 at 1 and 5 min, respectively. The child cried at birth but experienced respiratory distress 2 h after birth. The SpO 2 was < 90% and the subject required nasal continuous positive airway pressure and oxygen for 6 and 8 days, respectively.
Radiographs taken on day 0 showed a bell-shaped thorax, femoral bending and proximal metaphyseal dysplasia that was possibly accompanied by metaphyseal fracture. Blood tests indicated hypocalcemia, extremely high PTH (high sensitive PTH was > 10,000 at day 0, iPTH was 90.8 pg/ml at day 16), and slightly depressed levels of 25(OH)-vitamin D. The subject was treated with calcium gluconate on day 0 and by day 1 the hypocalcemia had improved. Intact PTH levels were normalized on day 72. Bone symptoms had nearly resolved when the subject was 2 months old. Blood ALP levels were increased on day 0, peaked on day 72, and normalized at around 17 months. Femoral bending normalized at around 21 months.

Sanger sequencing
All exonic regions of TRPV6 (RefSeq ID: NM_018646.5, NT_007933.16) were amplified by PCR using QuickTaq HS (Toyobo, Japan) according to the manufacturer's instructions. The PCR conditions were: 94 °C, 2 min followed by 35 cycles of 94 °C, 30 s; 62 °C, 30 s; 68 °C, 30 s (for exons 1-3 and 7-15) or 94 °C followed by 2 min and 35 cycles of 94 °C, 30 s; 55 °C, 30 s; 68 °C, 30 s (for exons [4][5][6]. Primer sequences are listed in Additional file 1: Table S1. TA cloning for subject 1 PCR products were ligated into the TOPO vector using a TOPO TA cloning kit (Invitrogen, USA). After transformation of the vectors into E. coli, the resulting colonies that formed on a selection plate were subjected to colony PCR to screen for the presence of the insert using M13 forward and reverse primers. Colonies having the desired sequence were cultivated in LB medium supplemented with antibiotics. Plasmid isolation was performed using a QIAprep Spin miniprep kit (Qiagen, USA) and sequences were verified by Sanger sequencing.

Mutagenesis
Site-directed mutagenesis was performed as described previously [8]. PCR was performed in a solution

Cell culture and transfection
HEK293T cells were maintained and transfected with a plasmid carrying human TRPV6 as previously described [8,9] . The membrane potential was clamped at − 60 mV. Voltage ramp pulses from − 100 to +100 mV (400 ms) were applied every 5 s. All recordings were performed at room temperature.

Plasma membrane protein biotinylation and western blotting
Biotinylation and western blotting were performed 27-28 h after initiating transfection as previously described [8]. Transfected cells were washed with PBS and incubated with 0.5 mg/ml EZ-link-NHS-LC-biotin (Abcam, USA) for 10 min at 37 °C in a CO 2 incubator. The EZ-link-NHS-LC-biotin solution was again added and incubated for 10 min at 37 °C. The biotinylation reaction was stopped by adding quenching buffer (100 mM glycine in PBS, pH 7.3) before the cells were transferred from the dish by pipetting. Cells were collected by centrifugation (12,000 rpm for 5 min at 4 °C) and stored at − 20 °C until use. Isolation of biotinylated proteins and western blotting was performed as described previously [8].

Fura-2 Ca 2+ -imaging
Measurement of intracellular Ca 2+ concentrations was conducted 20-24 h after starting the transfection as reported previously [8]. Cells were incubated with Fura-2-AM (5 µM, Life Technologies Inc., USA) at 37 °C for 1 h. The intracellular Ca 2+ concentration was analyzed first in the standard bath solution and then in the DVF bath solution before superfusion of the 30 Ca bath solution. Ratiometric imaging was performed at 340 and 380 nm, and emission at 510 nm was recorded with a CCD camera (CoolSnap ES, Roper Scientific/Photometrics, USA) every 4 s. The F340/380 ratio was calculated using IP Lab software (Scanalytics Inc., USA) and data were analyzed using ImageJ (NIH).

Results
Here, we report on two Japanese subjects with transient neonatal hyperparathyroidism (TNHP) with bone abnormalities. Both had pre-and post-natal history of skeletal abnormalities as well as elevated parathyroid hormone (PTH) and alkaline phosphatase levels ( Table 1); these abnormalities gradually improved over time as previously reported [8].  (Fig. 1b). A novel mutation, pGly291Ser, was found in the sixth ankyrin repeat (AR) domain, whereas p.Arg390His localized to the outer edge of the first transmembrane (S1) domain [10]. These mutations were found far from previously reported mutation hotspots (Fig. 2). Meanwhile, the mutation p.Lys285ArgfsTer9 caused a frameshift that generates a truncation protein that lacks the transmembrane domains. As no genetic analysis was available for the parents, we performed TA cloning and confirmed that these were in fact compound heterozygous mutations (Fig. 1a).
To analyze the functional significance of the p.Gly291Ser and p.Arg390His mutations, we transfected HEK293T cells with expression vectors carrying mutated TRPV6 and performed whole-cell patch-clamp recordings 20-24 h after transfection. HEK293T cells expressing the p.Gly291Ser mutant exhibited currents evoked in response to both divalent cation-free (DVF) and high Ca 2 bath solution that were similar to that seen for WT (Fig. 3a, c). The amplitudes of the Ca 2+ -evoked currents were larger than those previously reported [8], most likely because of differences in the NMDG solution used. Here, we used Ca 2+ -and Mg 2+ -free NMDG solution in order to observe Ca 2+ currents more clearly. Neither the current-voltage relationship nor reversal potential differed between WT and Gly291Ser (Fig. 3b), indicating a similar ion selectivity. We also found no significant difference in intracellular Ca 2+ -dependent inactivation, at least in our patch-clamp recordings within the time range of 20-30 s (Fig. 3d) [11]. These results suggested that apparent channel properties were preserved for the Gly291Ser mutation. In contrast, in cells expressing p.Arg390His no currents were evoked by DVF-or high Ca 2+ bath solutions (Fig. 3c, Additional file 2: Fig. S1). Thus, we hypothesized that trafficking of p.Arg390His proteins was impaired. Indeed, similar to the TRPV6 S2 transmembrane domain mutant p.Gly428Arg that was previously shown to have deficient trafficking, p.Arg390His did not localize to the plasma membrane, but was present in total cell lysates (Fig. 4a). Meanwhile, for the TRPV6 AR6 mutant p.Gly291Ser, we could not detect the presence of this protein even with longer transfection time, which is similar to that previously reported for p.Gly451Glu. With a shorter transfection time, we could detect it in the plasma membrane (Additional file 3: Fig. S2). We thus hypothesized that the p.Gly291Ser mutation affects intracellular Ca 2+ -dependent inactivation over a longer time range as was observed previously [8]. To examine the effects of Ca 2+ overload and cell death, we performed conventional western blotting of cell lysates from p.Gly291Sertransfected cells with or without treatment with the TRP channel blocker ruthenium red (RuR). The pGly291Ser proteins were detectable in RuR-treated cells, suggesting that impaired [Ca 2+ ] i -dependent inactivation over a longer time range leads to Ca 2+ overload and cell death (Fig. 4b). In cells without RuR treatment the band for p.Gly291Ser was faint, possibly because this mutant could elicit Ca 2+ overload leading to detachment of cells from the culture dish. In the presence of RuR, p.Gly291Ser TRPV6-derived Ca 2+ influx is blocked and as such no Ca 2+ overload would be elicited. Upon measurement of intracellular Ca 2+ levels with Fura-2, we observed a significantly higher Ca 2+ concentration in cells expressing p.Gly291Ser under a steady-state condition (p < 0.05, n = 39-49, Mann-Whitney test, Fig. 5a, b). Moreover, the intracellular Ca 2+ in these cells did not change significantly following superfusion of divalent cation-free solution (DVF) or 30 Ca bath solution, likely because the Ca 2+ concentration was nearly saturated. These results supported our hypothesis that the Gly291Ser mutation causes impaired [Ca 2+ ] i -dependent inactivation across a longer time range that may be too long to allow detection in patch-clamp recordings.

Discussion
Transient receptor potential vanilloid 6 (TRPV6) is a member of the TRP channel family that has very high Ca 2+ -selectivity [12]. TRPV6 is reported to play a role in Ca 2+ absorption in the intestine [13,14] and the epididymis [15], as well as in maternal-fetal Ca 2+ transport across the placenta [16], suggesting a role for this channel in transcellular Ca 2+ transport in particular organs and tissues [12,17]. Despite its physiological significance, the molecular mechanisms that regulate this channel remain largely undefined. For example, how TRPV6 maintains intracellular Ca 2+ concentrations that in turn maintain continuous transcellular Ca 2+ transport is unclear.
Previously, we found that mutations in TRPV6 cause transient neonatal hyperparathyroidism (TNHP) with bone abnormalities [8]. These mutations localize to hotspots on the extracellular side of S2 and S3 transmembrane domains, as well as to the intracellular ankyrin repeat domain 4 (AR4). Functional analysis indicated that: (1) TRPV6 with mutations in the transmembrane domains exhibit impaired trafficking to the plasma membrane; (2) mutations in AR4 are associated with partial loss-of-function, likely due to protein instability; and (3) the pGly451Glu mutation leads to intracellular Ca 2+ overload due to impaired [Ca 2+ ] i -dependent inactivation [8]. Another study identified the C-terminal TRPV6 mutation p.Gly660Arg, although the functional significance of this mutation is unclear [18].
In this study, we identified novel TRPV6 mutations in two patients with TNHP: p.Gly291Ser (subject 1), and p.Arg390His (subject 2). The results of the present study suggested that pGly291Ser might cause intracellular Ca 2+ overload. In terms of [Ca 2+ ] i -dependent inactivation, in our patch-clamp recordings we found no significant difference between the mutant and WT upon repetitive application of calcium over 20-30 s. This outcome is a general indicator for the [Ca 2+ ] i -dependent fast/intermediate inactivation. Meanwhile, the intracellular Ca 2+ concentration was significantly higher in the mutant-expressing cells than in cells expressing WT. Although the reason for this difference is unclear, Ca 2+ -selective channels can be inhibited via multiple pathways, including those having fast and slow inactivation kinetics [11]. Based on our results, we propose that p.Gly291Ser has slower inactivation than can be observed in the repetitive application time range.
In subject 1, another allele carried a frameshift mutation (Lys285ArgfsTer9), suggesting that pGly291Ser expression could be upregulated by a decrease in maternal-fetal Ca 2+ transport caused by a frameshift mutation into another allele, thus leading to further Ca 2+ overload and cell death as suggested in the previous report [8]. Subject 2 had the mutation p.Arg390His and Red, gray and black dots in the traces indicate the points at which I-V curves were generated. c Statistical analysis of the peak amplitude of Ca 2+ -evoked currents. The currents for p.Arg390His were significantly smaller compared to those for WT (p = 0.013, n = 4, Mann-Whitney rank-sum test). Currents for p.Gly291Ser did not significantly differ from that for the wild type placental calcium transport, strongly suggesting that Trpv6 is critical for total body Ca 2+ homeostasis [13,14,16]. However, in these same Trpv6 knockout mice, results of several reports indicated that Trpv6 is not crucial for 1,25(OH) 2 -vitamin D 3 -dependent intestinal Ca 2+ absorption [19]. We propose that these different phenotypes could be due to different Ca 2+ environments such as differences in the Ca 2+ content of drinking water or maternal vitamin D levels. In the case of calcium intake, in a 2017 study Japanese subjects were reported to have an average daily Ca 2+ intake of 517 mg. Women aged 20-29 and those aged 30-39 consumed an average of 420 mg/ day and 421 mg/day, respectively. These amounts were smaller than those for Western countries where the consumption of dairy products is generally higher [20]. Based on these findings, for subject 2 who carried p.Arg390His and had THNP, whereas a sister having the same genotype did not, we hypothesize that the Ca 2+ status of the mother might have differed between pregnancies. If this is indeed the case, there could be other potential mutations/polymorphisms that have milder effects than p.Arg390His, and would lead to TNHP depending on the Ca 2+ status of the mother. In other words, there may be a spectrum of transient neonatal hyperparathyroidism that is affected by both genotype and environment. Rare variants as well as common SNPs in the TRPV6 gene should be further analyzed in different human populations to determine the relationship between TRPV6 variants and disease severity, and reveal the molecular mechanism by which TRPV6 activity is regulated as well as the significance of molecular evolution of the TRPV6 gene [21][22][23].

Conclusion
Mutations in the gene encoding the TRP channel TRPV6 cause transient neonatal hyperparathyroidism (TNHP). In this study, we report novel mutations in the TRPV6 S1 and AR6 domains that occurred outside of domains recognized as being mutation hotspots (e.g., S2, S3, AR4 and the S2-S3 loop). Unexpectedly, we found that the mutation in AR6 was associated with increased intracellular Ca 2+ concentrations. This result suggests that the AR6 domain of TRPV6 is involved in the maintenance of the [Ca 2+ ] i homeostasis that is crucial for sustained, unidirectional Ca 2+ transport.