Generation of mice encoding a conditional null allele of Gcm2

Abstract

Glial cells missing homolog 2 (GCM2) is a transcription factor that is expressed predominately in the pharyngeal pouches and, at later stages, in the developing and mature parathyroid glands. In humans, loss of GCM2 function, either through recessive apomorphic mutations or dominant inhibitor mutations in the human GCM2 gene, leads to isolated hypoparathyroidism. In mice, homozygous disruption of Gcm2 by conventional gene targeting results in parathyroid aplasia and hypoparathyroidism. In this study, we report the generation and functional characterization of mice encoding a conditional null allele of Gcm2. We demonstrate the functional integrity of the conditional Gcm2 allele and report successful in vivo deletion of exon 2 using Cre recombinase. The mice with conditional deletion of Gcm2 displayed phenotypes similar to those previously described for a conventional Gcm2 knockout, including perinatal lethality, hypocalemia, low or undetectable serum levels of parathyroid hormone (PTH), and absent parathyroid glands. The production of a conditional mutant allele for Gcm2 represents a valuable resource for the study of the temporal- and spatial-specific roles for Gcm2, and for understanding the postnatal activities of GCM2 protein.

Introduction

Glial cell missing homolog 2 (GCM2) is a homolog of the gcm/glide gene that was originally identified in Drosophila melanogaster. (Hosoya et al. 1995; Wegner and Riethmacher 2001) In the fly, gcm and the related gcm2 are specifically and transiently expressed in the central nervous system where they act as binary switches to promote glial cell fate while simultaneously inhibiting the neuronal fate. These two genes function redundantly and are required for the formation of a subset of glial cells. (Akiyama et al. 1996; Chotard et al. 2005) By contrast, vertebrate GCM2 is expressed predominately in the pharyngeal pouches and, at later stages, in the developing and mature parathyroid glands of tetrapods and in the internal gill buds in fish.(Okabe and Graham 2004; Zajac and Danks 2008) In addition to Gcm2, a second vertebrate gcm homolog, Gcm1, has been identified that is expressed in villous cytotrophoblast cells of the placental and is essential for both syncytiotrophoblast differentiation and formation of chorionic villi. (Hashemolhosseini and Wegner 2004). All members of the GCM family encode nuclear transcription factors that contain a unique DNA binding domain (i.e. the GCM motif) in the amino terminus, one or two transactivation domains, and several potential PEST sequences that are typical of proteins displaying a rapid turnover. (Kim et al. 1998; Jones et al. 1995; Hosoya et al. 1995; Altshuller et al. 1996; Kammerer et al. 1999; Hashemolhosseini and Wegner 2004) Sequence homology between members of the GCM family is greatest within the GCM motif. (Cohen et al. 2003; Cohen et al. 2002; Schreiber et al. 1998)

Conventional deletion of the Gcm2 gene in transgenic mice leads to parathyroid aplasia and hypoparathyroidism, a metabolic condition characterized by hypocalcemia and hyperphosphatemia due to absence of parathyroid hormone (PTH).(Gunther et al. 2000; Kamitani-Kawamoto et al. 2011; Hitoshi et al. 2011) In humans, loss of GCM2 activity, through either recessive amorphic (Ding et al. 2001; Maret et al. 2008; Sticht and Hashemolhosseini 2006; Baumber et al. 2005; Thomee et al. 2005; Doyle et al. 2012) or dominant inhibitor(Mannstadt et al. 2008; Mirczuk et al. 2010; Canaff et al. 2009) mutations in GCM2, is an important cause of idiopathic hypoparathyroidism. Current knowledge about the role of GCM2 in parathyroid homeostasis has come from developmental analyses in mice (Gordon et al. 2001; Liu et al. 2007; Grigorieva et al. 2010), which indicate that GCM2 is not necessary for pouch patterning or establishment of the parathyroid domain (i.e. induction of parathyroid cell precursors), but instead is required for differentiation and subsequent survival of parathyroid cells during embryogenesis. However, GCM2’s role(s) in controlling parathyroid cell proliferation, survival or function during late embryological development and in the postnatal parathyroid gland is still unknown. In addition, recent studies that demonstrate transient expression of Gcm2 in regions of the central nervous system during early development (Hitoshi et al. 2011) suggest that GCM2 may play a wider role that is cell- and context-specific. Hence, a key challenge will be to dissect out the function of GCM2 within various cells at different times during development and in mature tissues. This obviously cannot be accomplished by a simple constitutive knockout approach as performed previously. (Gunther et al. 2000; Hitoshi et al. 2011) These studies would be greatly facilitated or made possible, however, using mice harboring Gcm2 conditional null alleles. We report here the generation and characterization of such mice using DNA homologous recombination in embryonic stem (ES) cells and the Cre-loxP and FLP-FRT technologies.

MATERIALS AND METHODS

Generation of a Gcm2-floxed Allele

The C57BL/6 mouse chromosome 13 sequence (n.t.# 41,112,399–41,121,157) was retrieved from the Ensembl database (www.ensembl.org) and used as a reference in this project. The RP23-2H11 bacterial artificial chromosome (BAC) was used to generate the homologous arms and the conditional knockout (CKO) region for the gene targeting vector, as well as the probes for screening targeted events by restriction endonuclease analysis (Figure 1). The 5’ homologous arm (4.2 kb), 3’ homologous arm (4.0 kb) and CKO region that includes exon 2 (252 bp) were generated by PCR. The fragments were cloned in the LoxFtNwCD vector (Taconic Laboratory) or pCR4.0 for sequencing. The final targeting construct was obtained by standard molecular cloning (Figure 1B). Briefly, the CKO region was cloned in BsiWI/SalI sites, upstream from the FRT site in the vector, using PCR (primers used: Cl_GCMB_E2F, 5’-actgcgtacgCCTATTCTGTTTGCCATCTTTTCCC-3’ and Cl_GCMB_E2R, 5’-actggtcgacGGAAGGTAGAGGGACACCAGCTAGG-3’) (Figure 1). The 5’-arm was cloned into LoxFtNwCD at NotI/AvrII sites, upstream from loxP site (Primers: Cl_GCMB_5F, 5’-actggcggccgcCCTTCTGTCTCTTTCTGTCTCAGTCTCC-3’ and Cl_GCMB_5R: 5’-actgcctaggGGAAAAGATGGCAAACAGAATAGGAC-3’). The 3’arm was cloned in LoxFtNwCD in AscI/NruI sites (primers: Cl_GCMB_3F: 5’-actgggcgcgccCTCTGTTCCTAGCTGGTGTCCCTCTACC-3’ and Cl_GCMB_3R: 5’-actgtcgcgaGGCAAGGCTAGTTGACTTGTGTCTAA-3’). Aside from the homologous arms, the final vector also contains loxP sequences flanking the region targeted for deletion (433 bp), FRT sequences flanking the Neo^r expression cassette (for positive selection of the ES cells), and a diphtheria toxin (DTA) expression cassette (for negative selection of the ES cells). The final vector was confirmed by both restriction digestion and end sequencing analysis. The 5’ and 3’ external probes were generated by PCR and were tested by genomic Southern blot analysis for screening of the ES cells. The positions of the probes used for Southern blot analysis are shown in Figure 1C. NotI was used to linearize the final vector prior to electroporation into B6 embryonic stem (ES) cells (Taconic Laboratory). We transfected properly targeted clones with FLP recombinase, and genotyped replicate colonies that had lost resistance to G418 by PCR. In order to confirm that the loxP sites were functional in the floxed Gcm2^tm1Mal2 (i.e. Gcm2^E2fl) alleles, we electroporated Neo^r-deleted clones with Cre recombinase, isolated genomic DNA, and analyzed the Gcm2 gene for proper recombination and loss of exon 2 (i.e. Gcm2^ΔE2). Two ES clones containing properly recombined Gcm2^tm1Mal2 alleles were injected blastocysts from B6(Cg)-Tyr^c-2J/J albino mice (Jackson Laboratory stock #000058) to generate chimeric mice. The resultant chimeras were mated with C57BL/6NTac females and F1 black offspring were genotyped by PCR and Southern blot analyses. Mice carrying the conditional allele were homozygosed by mating Gcm2^E2fl/+ animals, producing Gcm2^E2fl/E2fl mice. Both male and female heterozygous and homozygous animals were subjected to metabolic study. We bred female mice that were heterozygous for the conditional allele with 129S/Sv-Tg(Prm-cre)58Og/J males (Jackson Laboratory stock #003328) to generate mice that were heterozygous for Gcm2 lacking exon 2 (i.e. Gcm2^ΔE2). B6;129Sv-Gcm2^tm1.1Mal2 mice have generalized inactivation of Gcm2 alleles. Mice were maintained on a 129SvE;C57BL/6NTac hybrid genetic background as high lethality had been previously reported for constitutive Gcm2 knockout mice on a non-hybrid background.(Gunther et al. 2000) The mice produced for this project were fed Pico Lab Rodent diet 20–5053 while in holding cages and Pico Lab Mouse diet 20–5058 while in the breeding cages. All animal experiments were conducted according to protocols approved by The Children’s Hospital of Philadelphia and Albert Einstein College of Medicine.

Figure 1. Targeting vector strategy.

A. the mouse Gcm2 gene is located on chromosome 13 and is organized into five exons, with exon 1 containing the start codon and the first 30 amino acids of the protein, exon 2 (E2) containing the entire GCM motif, and exons 3–5 encoding the C-terminal two thirds of the protein. B. the initial targeting vector (LoxFTNwCD/mGcm2) was approximately 14.2 kb, and contained a 5’ arm of homology (approximately 4.2 kb), a 3’ arm of homology (approximately 4.0 kb) and a neo^r cassette for positive selection that was flanked by FRT sites. Exon 2 (E2) is flanked by loxP sites. A DTA cassette encoding diphtheria toxin for negative selection was placed after the 3’ arm. C. the upper panel shows the targeting construct that was electroporated into B6-3 ES (C57BL/6) cells with selection by G418. The positions of the three hybridization probes are noted. D. Southern blot of genomic DNA from four G418-resistant clones (C5, E8, H3 and H11) that were identified by primary screening by PCR and B63, a negative control. Restriction analysis of DNA from clones H11 and E8 generated hybridizing fragments of the predicted sizes (arrows) with all three probes, thus confirming proper homologous recombination. E. Schematic of the wild type (upper) and floxed (middle) Gcm2 allele and the final Gcm2 allele (lower) after excision of the exon 2 region by Cre recombinase. Diagram shows sizes of PCR products and positions of primers used for genotyping by PCR. F. Panel shows PCR genotyping using genomic DNA from F1 offspring of both E8 and H11 chimera mice. The conditional knockout allele is predicted to generate a 0.9 kb fragment (upper arrow) and the wild type allele is predicted to generate a 0.6 kb fragment (lower arrow). Lane MW, 100 bp ladder; lane 5 is water control template and lane 14 is intentionally blank. Mice that are wild type, heterozygous and homozygous for the conditional allele can be readily distinguished. G. Panel shows PCR genotyping using genomic DNA from offspring of matings between E8 founder mice and Prm-Cre mice. The conditional knockout allele (Gcm2^E2fl) is predicted to generate a 0.9 kb fragment (lane 3) and the wild type allele is predicted to generate a 0.6 kb fragment (lane 2), and the recombined allele (Gcm2^ΔE2) is predicted to generate a 0.4 kb fragment (lane 4). Lane MW, 100 bp ladder and lane 5 is a water control template. Mice that are heterozygous (lanes 6, 8, 9, 10 and 13) or homozygous (lane 12) for the exon 2-deleted Gcm2 allele are readily distinguished from wild type mice (lanes 7, 11 and 14). Denotes loxP site and denotes Frt site.

Genotyping

Genotypes were determined by Southern blot both with 5’ and 3’ external probes (Figure 1, panels C and D). For initial PCR screening of floxed alleles, we used forward primer 5’-GGTTTCCAAATGTGTCAGTTTCATAGCCT-3’ and reverse primer 5’-GGAAGAACTACTAAACCCACCAGACTGC-3’ (PCR primers shown in Figure 1, panel C, with the forward primer annealing in the Neo^r gene approximately 100 bp from 3’ end and reverse primer annealing 120 bp downstream of the 3’ homologous arm), with expected size of the amplified product approximately 4.1 kb (not shown). Diagnostic PCR for in vivo Cre recombinase-induced deletion of the CKO region used forward primer 5’-GATGCAGCTGGAGATTGACATTGTATAC-3’ and reverse primer 5’-GGCAAGGAACCAGCTCCAAAG-3 (Figure 1, panels E-G).

Biochemical Analyses

Whole blood and serum were collected using methods described previously (Germain-Lee et al. 2005). Total calcium and phosphate were measured using colorimetric assays (QuantiChrome Calcium Assay Kit DICA-500 and QuantiChrome Phosphate Assay Kit DIPA-500) from BioAssay Systems, Hayward, CA, USA. PTH was measured with an ELISA for rodent PTH 1–34 that has a detection limit of 1.6 pg/mL (Immutopics, San Clemente, CA, USA). Levels of intact FGF23 were measured using a mouse FGF23 ELISA kit that detects intact hormone (Kainos Laboratories, Inc., Tokyo, Japan). We measured serum 1,25(OH)₂D by ELISA using a kit from Immunodiagnostic Systems (Fountainhead, AZ, USA).

Histology

Mice were euthanized at postnatal 30 days, and tracheal blocks and other organs were dissected, formalin-fixed and embedded in paraffin and 5-µm sections were prepared for subsequent analyses. To determine parathyroid gland size, appropriate tissue sections were digitalized and maximal parathyroid cross-sectional area was measured by computerized pixel counting. For immunofluorescence (IF) staining, paraffin sections were incubated with primary antibodies to calcium-sensing receptor (CASR) (1:50, rabbit polyclonal antiserum #33821, Santa Cruz Biotechnology), PTH (1:200, mouse monoclonal antiserum AB63993, Abcam), GCM1 (1:50, rabbit polyclonal antiserum 98811, Santa Cruz Biotechnology) or GCM2 (1:50, goat polyclonal antiserum 79495, Santa Cruz Biotechnology) at 4° C overnight. Antibody binding was detected by incubation with appropriate secondary antibodies (anti-mouse Alexa Fluor 647, anti-rabbit Alexa Fluor 488, and anti-goat Alexa Fluor 488, Invitrogen) at 1:200 dilution for 45 minutes in the dark. Finally, the sections were stained with 4′6-diamidino-2-phenylinodole (DAPI) (Vector Laboratories, Burlingame, CA) and imaged by fluorescence microscopy.

Statistical analyses

Data were analyzed using GraphPad InStat v3.06 for Windows (GraphPad Software). Means were compared using Student’s t-test. All results are expressed as the mean ± SD. Statistical significance was accepted at P < 0.05.

Results

Generation of the Conditional Allele of Gcm2

The mouse Gcm2 gene is organized into five exons, with exon 1 containing the start codon and the first 30 amino acids of the protein, exon 2 containing the entire GCM motif, and exons 3–5 encoding the C-terminal two thirds of the protein (Figure 1). Two groups have developed Gcm2 knockout mice with constitutive deletions. Günther et al. used a targeting vector that replaced exons 2–5 of Gcm2 (>5 kb) with the gene encoding neomycin resistance (Neo^r) (Gcm2^tm1Kry) to create conventional Gcm2 knockout mice (129S7-Gcm2^tm1Kry), (Gunther et al. 2000), whereas Hitoshi et al. replaced exons 2–5 with the LacZ gene.(Hitoshi et al. 2011) To generate a conditional allele for Gcm2, we inserted two loxP sites in the same orientation to flox exon 2 (Gcm2^E2fl), such that Cre-mediated recombination would result in excision of exon 2 (Gcm2^ΔE2)(Figure 1E). Splicing of E1 into E3 was predicted based on the gene sequence to create a frameshift rapidly followed by a stop codon, and thus to result in production of a short peptide lacking the two main functional domains of Gcm2. Hence, the Gcm2^ΔE2 allele (Gcm2^tm1.1Mal2) would be a null allele of Gcm2.

The Gcm2 targeting construct was electroporated into ES cells and 4 clones (C5, H3, H11, E8) were positively identified by PCR and were expanded. Two clones, H11 and E8, were confirmed by Southern analysis (Figure 1D) with all 3 probes to contain the properly targeted Gcm2^E2fl-Neo allele.

We transfected clones E8 and H11 with FLP recombinase and selected 96 colonies from each clone for replicate testing for G418 sensitivity. A number of G418-sensitive clones were selected and analyzed by PCR, which confirmed deletion of the Neo^r gene and the presence of 5’ loxP sites (Figure 1F). To assure that the loxP sites were functional in the Gcm2^E2fl alleles, we electroporated Neo^r-deleted clones with Cre recombinase. These cells were grown for several days and then batch-harvested for isolation of genomic DNA, and PCR confirmed the excision of E2 (data not shown). We used both clones E8 and H11 to generate mouse male chimeras. These mice transmitted the Gcm2^tm1Mal2 floxed allele to their progeny. Both mouse lines were maintained and one is described here.

Recombination of the floxed Gcm2^tm1Mal2 Allele in Vivo

The floxed Gcm2^E2fl was converted into the inactive Gcm2^ΔE2 by breeding homozygous Gcm2^tm1Mal2 females with 129S/Sv-Tg(Prm-cre) 580g/J males (Jackson Labs stock# 003328) that are homozygous for the Protamine-Cre (Prm-cre) trans-gene, which is expressed exclusively in the male germ line. (O'Gorman and Wahl 1997; O'Gorman et al. 1997) The cloned “CKO” fragment is 347 bp (95 bp longer than E2), and FLP treatment left an additional piece (frt + linker), which made the actual deleted piece 433 bp long. Subsequent matings of Gcm2^E2fl/+;Prm-cre male progeny with wildtype females yielded Gcm2^ΔE2/+ progeny. Those mice that did not carry the Prm-cre transgene were intercrossed for analysis (see below). All mice were genotyped by PCR using strategies designed to readily identify the Gcm2, Gcm2^E2fl, and Gcm2^ΔE2 alleles (Figure 1G).

Phenotype of the Cre-recombined Gcm2 Mice

Gcm2^E2fl/+, Gcm2^E2fl/E2fl, and Gcm2^+/+ mice were viable and obtained at the expected Mendelian ratios. Serum levels of calcium, phosphate and PTH were normal in all genotypes (data not shown). Subsequent matings of Gcm2^ΔE2/+ mice generated Gcm2^ΔE2/ΔE2 progeny that were viable but with an apparent 30% lethality, similar to that previously reported in the B6;129S7-Gcm2^tm1Kry mice. (Liu et al. 2010; Gunther et al. 2000) Wild type, floxed and knockout mice were of similar weight and length from weaning through 20 months; data at six months of age are shown in Table 1. By contrast, biochemical analyses demonstrated significant differences between the knockout mice and the floxed and wild type mice (Table 1): serum calcium was lower and serum phosphate was greater in knockout mice than in wild type or floxed mice and serum levels of PTH were undetectable in sixteen and extremely low in seven of twenty-three Gcm2^ΔE2/ΔE2 mice. Moreover, serum 1,25(OH)₂D levels were lower (120.3± 33.9 pmol/L) in knockout mice compared to those in wildtype mice (160.6 ± 31.8 pmol/L). And finally, serum levels of intact FGF23 were significantly (P <0.05) lower in knockout mice (63.4 ± 35 pg/mL) than in wildtype mice (137.8 ± 35 pg/mL).

Table 1.

Physical and biochemical measurements from mice at 6 months of age.

	Wild type	Floxed	Knockout
Number	22	9	23
Weight, gram	36.6 ± 7	36.2 ± 6	35.3 ± 7.1
Length, cm	3.7 ± 0.1	3.7 ± 0.2	3.6 ± 0.2
sCalcium, mg/dL	9.9 ± 1.0	10.3 ± 0.9	7.0 ± 1.5*
sPhosphorus, mg/dL	6.1 ± 1.0	6.8 ± 0.6	9.6 ± 2.1*
PTH intact, pg/mL	158.7 ± 177	156.3 ± 50.4	32 ± 27*

Data are ± SD.

^*p <0.001 versus wild type.

Analysis of the thyroidal area indicated that knockout mice lacked parathyroid tissue, whereas floxed mice had parathyroid tissue of normal size (Figure 2A). Knockout mice also showed no expression of PTH, CASR (Figure 2B) and GCM2 (GCM2) in the neck. The knockout mice showed normal development of other major organs, including, thymus, brain, lung, heart, liver, pancreas, testis, and ovary (data not shown). Wild type and knockout mice showed equivalent expression of CASR (Figure 3) and GCM1 (Figure 4) in the thymus and kidney.

Figure 2. Parathyroid and neck anatomy.

A. Histological analysis of tracheal sections from 30-day wild type, floxed, and knockout mice. Panels show H&E staining and 2.5× magnification. No parathyroid tissue is present in the neck of knockout mice. B. Immunofluorescence analysis of the expression of PTH and CASR (red) in wild type and knockout mice. C. Immunofluorescence analysis of the expression of GCM2 (green) in wild type and knockout mice.

Figure 3. Expression of CASR in thymus and kidney.

Genotype of mice is noted above each panel. Immunofluorescence (IF) analysis of the expression of CASR in thymus and kidney tissues from wild type (left) and knockout (right) mice at 30 days of age. CASR was stained red with anti-CASR Ab (Alexa Fluro 647).

Figure 4. Expression of GCM1 in thymus and kidney.

Genotype of mice is noted above each panel. Immunofluorescence (IF) analysis of the expression of GCM1 in thymus and kidney tissues from wild type (left) and knockout (right) mice at 30 days of age. GCM1 was stained green with anti-GCM1 Ab (Alexa Fluro 488).

Discussion

We have generated a conditional Gcm2 allele that is fully active, and shown that generalized deletion of exon 2 is associated with parathyroid aplasia and hypoparathyroidism. Our knockout model differs from previous mouse models of constitutive ablation of Gcm2 in which the approach led to insertion of either the Neo^r (Gunther et al. 2000) or lacZ (Hitoshi et al. 2011) genes into the Gcm2 locus.

Serum levels of PTH were initially reported to be normal in B6.129S7-Gcm2^tm1Kry Gcm2 knockout mice, which led to the proposal that thymic epithelial cells that co-expressed GCM1 and PTH represented an auxiliary source of circulating PTH. (Gunther et al. 2000) No information about mineral metabolism was reported in a second constitutive Gcm2 knockout model.(Hitoshi et al. 2011) Subsequent reports demonstrated low or undetectable levels of serum PTH in the B6.129S7-Gcm2^tm1Kry mouse line, however (Simmonds et al. 2010; Liu et al. 2010), similar to the data that we present here. In the present work we identified immunoreactive PTH within a few thymus cells in wild type and homozygous knockout mice, consistent with the proposal that expression of PTH in medullary thymic epithelial cells provides a source of self-antigen for negative selection. (Liu et al. 2010) Recent work has excluded the thymus as a source of circulating PTH in humans and mice (Maret et al. 2004; Liu et al. 2010). Similarly, the minimal expression of PTH in kidney cells that we identified, as well as previous demonstrations of PTH in the hypothalamus (Nutley et al. 1995; Fraser et al. 1990; Gunther et al. 2000), are unlikely to be physiologically significant. Although we have no certain explanation for the presence of low circulating levels of PTH in some of our knockout mice, this is similar to what we and others have described in humans with hypoparathyroidism due to inactivating GCM2 mutations (Canaff et al. 2009; Mannstadt et al. 2008). As one proposed role for GCM2 is to ensure parathyroid cell survival after initial specification of the parathyroid domain within the pouch endoderm (Liu et al. 2007), it is possible that small numbers of parathyroid cells can escape apoptosis in the absence of GCM2 and secrete measurable if physiologically insignificant amounts of PTH.

Finally, we found that serum levels of FGF23 were not elevated in our homozygous knockout Gcm2 mice, which lack parathyroid tissue, similar to that described in mice in which the Pth gene had been deleted (Yuan et al. 2013; Quinn et al. 2013). These observations are consistent with emerging evidence for a complex regulatory mechanism for FGF23 secretion that involves PTH (Silver and Naveh-Many 2012) as well as serum concentrations of calcium and phosphorus to ensure a normal calcium × phosphorus product (Quinn et al. 2013). By contrast, previous studies of patients with chronic (Gupta et al. 2004) or acute and transient hypoparathyroidism (Yamashita et al. 2007) had shown that levels of serum FGF23 were increased, and were correlated with hyperphosphatemia. But in these cases the patients were also receiving calcitriol or 1α-hydroxyvitamin D, which would increase FGF23 (Collins et al. 2005; Gupta et al. 2004).

In conclusion, the Gcm2 conditional null allele that we have generated will constitute a very valuable tool to block expression of Gcm2 conditionally, in specific cell types and times in vivo. It will allow or greatly facilitate studies on the roles of Gcm2 in multiple developmental, physiological, and even pathological processes in the mouse from embryogenesis onto adulthood, and will further elucidate the regulation of parathyroid development and extracellular calcium homeostasis.