The parathyroids are small glands located in the cervical region in close proximity to the thyroids.
The main function of the PGs is the secretion of PTH.
It is on top of a complex hormonal cascade regulating serum calcium concentration. The latter is remarkably constant in diverse organisms under various physiological conditions.
This tight regulation is important since calcium is essential for many functions such as muscle contraction, neuronal excitability, blood coagulation, mineralization of bone and others.
A reduction of the serum calcium concentration to less than 50% will lead to tetany and subsequently to death.
The importance of a strict regulation of the serum calcium is also reflected by the rapid secretion of PTH within seconds, new synthesis of the hormone within minutes and new transcription within hours following a decrease in serum calcium concentration which is detected through the calcium sensing receptor expressed in the PGs.
The overall role of PTH is to increase calcium concentration.
It fulfils this function through three different means.
First it prevents calcium elimination in the urine, second it favors the hydroxylation in one of the 25 hydroxycholecalciferol and as a results it favors indirectly intestinal calcium absorption.
Lastly PTH favors through still poorly understood mechanisms bone resorption and as a result increases the extracellular calcium concentration.
Development of Parathyroid Glands in Vertebrates
The PGs derive from the pharyngeal pouches which are transient structures during embryonic development.
They are evolutionary homologous to gill slits in fish.
The foregut endoderm and cells originating from the neural crest of rhombomere 6 and 7 contribute to the anlage of the PGs.
The neural crest originates at the apposition of neuroectoderm and ectoderm during the formation of the neural tube.
Therefore neural crest cells have to migrate towards the foregut endoderm first before they can add to the anlage of the PGs.
Neural crest of rhombomere 6 migrates towards the third branchial arch while the fourth branchial arch is primarily invaded by neural crest cells from rhombomere 7.
Mice only have one pair of PGs deriving from the third pharyngeal pouch homologous to the inferior PGs in men while the superior ones derive from the fourth pharyngeal pouch.
The anlage of the PGs in mice first becomes visible between embryonic day 11 (E11) and E11. histologically in a very limited area in the dorsal region of the cranial wall of the third endodermal pouch while the caudal portion of the very same pouch develops into the thymus which is involved in the maturation of the immune system.
Both domains are demarkated by the complementary expression of Gcm2 and Foxn1 (the latter mutated in nude mice, lacking a functional thymus), respectively already two days before the anlagen are morphologically visible.
In contrast to thymus development, induction of the ectoderm is not necessary for the formation of the PGs.
In mammals both structures start to migrate shortly thereafter towards the caudal end before at around E14 they seperate. While the thymus moves on further in the direction of the heart the PGs become incorporated to the thyroid gland between E14 and E15.
Pth is expressed already in the anlage of the PGs at E11.5 and contributes to fetal serum calcium regulation to some extent although placental transport involving parathyroid hormone related protein (PTHrP) is more important.
The parathyroid gland is not the only source of PTH. The protein is also synthesized by a few cells in the hypothalamus and in the thymus.
It has been shown in mice that the thymic Pth-expressing cells actually contribute to the circulating hormone keeping the level of serum calcium even in the absence of PGs at a concentration compatible with life.
Genetic Control of Parathyroid Gland Development
Three different steps can be used to separate the formation of the PGs mechanistically.
They !!include!! (I) formation of the PGs, (II) migration towards their final destination and (III) the differentiation towards PTH producing cells.
Mouse mutants that highlight the role of the few genes known to be involved in these different processes have been generated in the last decade.
Both, neural crest cells and the pharyngeal endoderm contribute to the anlage of the PGs.
Neural crest cells possibly already maintain information about their localization along the anterior-posterior axis before they start to migrate ventrally.
They derive this information from a group of evolutionary conserved transcription factors containing a homebox, the Hox genes, organized in four paralogous genomic clusters (Hoxa, b, c and d). Hox genes are expressed in the neural crest prior to, during and after migration into the pharyngeal arches and endodermal epithelia express Hox genes as well.
I. Rae28 is the mouse homologue of the Drosophila polyhomeotic gene which is required for the proper expression of hometic genes along the anterior-posterior axis.
Similar, absence of Rae28 causes an anterior shift of anterior expression boundaries of several genes of the Hox cluster including Hoxa3, Hoxb3 and Hoxb4.
Mice deficient for Rae28 are characterized by malformations of tissues partly derived from neural crest like altered localization of PGs as well as PG and thymic hypoplasia and cardiac anomalies.
How the altered hox expression pattern influences PG formation still needs to be evaluated.
The first reported malformation of PGs caused by a deletion through homologous recombination in mouse embryonic stem cells were represented by Hoxa3-deficient animals.
Among other defects knockout mice are devoid of PGs and thymus and exhibit thyroid hypoplasia.
This coincides very well with Hoxa3 expression in the third and fourth pharyngeal arches and in the pharyngeal endoderm.
The Hoxa3 signal does neither effect the number of neural crest cells nor their migration pattern.
Mutant cells rather lost their capacity to induce differentiation of surrounding tissues.
Absence of the paired box containing transcription factor Pax9 in targeted mice also displays absence of PGs and thymus.
Pax9 is expressed in the pharyngeal endoderm.
The epithelial buds separating from the third pharyngeal pouch did not form in the mutant mice.
This phenotype could be traced back to delayed development of the third pouch already at E11.5 and coincides with the expression of Pax9 in the pharyngeal endoderm.
II. PGs develop normally in mice deficient for the paralogous Hoxb3 and Hoxd3.
However further removal of a single Hoxa3 allele leads to the inability of the normally formed anlge of the PGs to migrate to their position next to the thyroid gland.
Therefore, development and migration of the PGs are separable events which is consistent with the fact that in other vertebrates like fish and birds PGs do not migrate from location of their origination.
III. Glial cell missing (Gcm2) is the homoloug of the Drosophila GCM transcription factor.
Unlike its glia cell fate determining function in fruit flies implies, mouse Gcm2 exclusively characterizes parathyroid cells and starts to be expressed around E10 in the pharyngeal endoderm.
The pattern rapidly becomes restricted to the cranial portion of the third pharyngeal pouch.
Mice deficient for Gcm2 revealed that PTH is never expressed in the PG anlage although parathyroid like cells characterized by Pax9 expression are still present at E14.5.
This clearly points out that Gcm2 is essential for the specification of precursors to become Pth-expressing cells rather than for the induction of the precursors itself. Interestingly, Pth-positive cells still could be detected in the thymus of mutant mice indicating that at least 2 pathways for the specification of Pth-expressing cells exist.
Gcm1 expressed in the thymus is the most likely candidate to compensate for Gcm2 function.
It will be compelling to determine if a ‚backup mechanism‘ for the parathyroid gland also exists in man.
In this direction it is very interesting to note that the first human homozygous mutation for GCM2 has been identified in hypoparathyroidic patients.
It has been discovered just recently that newborn Pax1-deficient mice exhibit severely reduced PGs.
The reduction in size could be traced back to the beginning of PG development at E11.5.
The hypoplasia of the anlage was even more severe in Hoxa3+/-Pax1-/- embryous and PGs were absent at late gestational stages. Interestingly, Gcm2 expression although properly initiated at E10.5 was reduced at E11.5 in Pax1-deficient embryos while the reduction was even more severe in the compound mutant.
Hoxa3-deficient embryos exhibit no Gcm2 signal at all. Therefore, Hoxa3 is necessary for Gcm2 induction while both Hoxa3 and Pax1 are substantial for the proper maintenance of Gcm2 expression.
Pax1 expression in the PG primordium on the other hand is reduced in Hoxa3-deficient mice. This would place Hoxa3 genetically upstream of Pax1 and both upstream of Gcm2 which in turn is required for PTH expression in PGs.
A long time known conglomerate of congenital malformations in humans including dysplasia or absence of the PGs and thymus as well as malformations of the heart outflow is the DiGeorge syndrome.
The organs affected derive in part from neural crest so that mutations in one or several genes influencing these cells have been suspected to be the cause for the disease.
It could be shown that most patients are hemizygous for a megabase deletion on chromosome 22q11.
Recently, two groups came up with a good candidate gene for several of the features in DiGeorge syndrome including PG defects simultaneously.
Both laboratories generated hemizygous megabase deletions comprising more than a dozen genes on the synthenic mouse chromosome 16 that reflected the human malformations including PG abnormalities.
TBX1 was among them and it could be shown that the gene is expressed in the pharyngeal endoderm and mesoderm-derived core but not in neural crest-derived mesenchyme.
Tbx1 expression in the pharyngeal arches is possibly induced through the morphogen Sonic hedgehog.
Mice heterozygous for a Tbx1 deletion by homologous recombination reflected the pharyngeal arch artery malformations while homozygous-deficient mice exhibited PG hypoplasia.
DiGeorge syndrome patients resemble hemizygous deletions.
This suggests that other genes of this region may contribute to the PG phenotype.
Indeed, Guris and colleagues could demonstrate that mice homozygous for a targeted null mutation for Crkol dysplay cardiovascular, PG and thymus defects.
The migration and early proliferation of neural crest cells was not altered pointing out that Crkol influences the function of neural crest during later stages.
CRKL (homolog human gene name) also maps within the common deletion region for the DiGeorg syndrome.
Deletions on chromosome 10p also cause DiGeorge like malformations.
The locus includs a subregion that encodes for the hypoparathyroidism, sensorineural deafness, renal anomaly (HDR) syndrome.
Van Esch and her colleagues could demonstrate that two heterozygous patients exhibit loss of function mutations in GATA3.
The transcription factor is indeed expressed in the affected organs during human and mouse embryonic development. Surprisingly though, heterozygous knockout mice have been reported to be normal while homoyzgous mice die around E12.
The understanding of the contribution from several gene products to the development of PGs from these critical regions still awaits further analysis.
Concluding Remark
Clinical studies indicate that multiple mutations can account for the malfunction of serum calcium regulation through PTH in humans.
These !!include!! the synthesis of PTH, sensing of the calcium content in the blood stream as well as the development PGs and proper specification of PTH translating cells.
It is astonishing how rather little is known so far on the molecular level in comparison to the formation of other organs. Surely, the genome sequencing projects for mice and man and the use of microarrays to compare different cDNA pools will shed new light on this issue in the near future.
