Molecular Mechanisms Involved in Craniosynostosis

Embryogenesis and Morphogenesis of the Skull

The human skull consists of three main parts, each of them having different developmental origins. The mesenchyme for these structures derives from neural crest cells, lateral mesodermal plaque and paraxon mesoderm. Recent research has shown that skull genesis takes part in four main stages (26). At the beginning, the undifferentiated cells migrate to the point of development. Afterwards, an epithelial mesenchymal interaction takes place to initiate the process of osteogenesis at the developmental location. Furthermore, through this signal a cell compression and differentiation into chondroblasts or osteoblasts takes place, producing the bone (27).

In the human embryo, the nascent skull is produced from the root parts of the neural tube (the nostril) and the pharynx, encircled by a series of paired aortic arches. Among these structures and the overflowing ectoderm, bulky masses of mesenchymal mesoderm are present (28). Neural crest cells originate in the ectoderm at the margins of the neural tube (the tail end of the forthcoming/posterior brain) and, after a phase of epithelial and mesenchymal transition and extensive migration, settle down in different parts of the body. They then assist in the creation of different tissues and organs. Their derivatives originate from four major segments of the neuraxis: cranial, cardiac, vagal, and trunk neural crest (29, 30). Both the neural crest and the mesoderm cells will create the developing bones of the skull. The human skull is created through two different ossification processes: the facial and cranial bones derive from intramural ossification, whereas the cranial bases derive from endochondral ossification. Nevertheless, both ways of ossifications cooperate to generate a single bone.

In intramembranous ossification, mesenchymal cells and neural crest cells differentiate into osteoblasts and form membranes of ossification centers. Osteoblasts form new bone by generating osteoids, a protein-derived matrix consisting mainly of type I collagen, which will become a bone after mineralization. Osteoblasts also facilitate the deposition of metals (i.e., calcium) into the bone matrix and produce hormones to control bone formation. Differentiation of osteoblasts from cells located on the bone surfaces is regulated by the expression of two genes: Cba-1 that encodes the cobalamin acquisition protein 1, a transcriptional regulator of a number of bone-specific target genes and IHH that encodes the Indian hedgehog homolog protein, a member of the Hedgehog family of signaling factors (31). The first stage of bone formation begins at the genetically-defined osteoporosis centers and is caused by growth factors (FGF, PDGF, TGF-β) and bone morphogenetic proteins (BMPs). Bone development is affected by depositing bone tissue along the peristaltic membrane. Osteoblasts produce bone by always compensating bone resorption by osteoclasts. The entire process of bone remodeling is controlled by hormones (hypercalcemic parathyroid hormone, hypocalcemic calcitonin, and sex steroid hormones). The final bone size is directly related to the onset of osteogenesis: the earlier the onset, the larger the bone (32).

In endochondral osteoporosis, bone formation precedes the configuration of a cartilaginous lattice of glycoproteins, which after metallization are replaced by endochondral bone (32). The process begins with the differentiation of chondroblasts, which mature into chondrocytes that first produce a cartilaginous matrix. Chondrocytes are then subjected to cell apoptosis and are eventually replaced by osteoblasts, which carry blood vessels that penetrate the calcified matrix, which is useful as a template for the construction of endochondral bone (33).

Since for intramembranous ossification a previous step is not required, it is quicker than endochondral ossification. The development of the skull is the result of the combined morphogenesis and development of two major areas (28, 33).

Molecular Mechanisms of Pathogenesis of Craniosynostosis

The cerebral skull consists of the chondrogenic bones of the base of the skull and the broad bones that form the dome of the skull, namely the two frontal, two parietal and occipital. Normally, at birth, the cranial dome bones are separated from each other by slits of several millimeters wide filled with connective tissue, the cranial sutures, which are: the frontal and sagittal sutures along the midline, the coronary and the lambdoid (Figure 1). The closure of the cranial sutures progresses gradually up to the third decade of life or later, except the frontal suture, which usually ends in the first two years after birth (34). When this occurs, the suture is said to “close”. The skull bones can sometimes join together too early (before the brain is fully formed), leading to the birth defect known as craniosynostosis. The skull can become more contorted as the brain grows, often slowing or even limiting the growth of the brain. During craniosynostosis, the head will stop growing in only that part of the skull in which the sutures closed too early, while in the other parts normal growth is observed, leading to an abnormal shape, although the brain inside has grown to its usual size. Sometimes, though, more than one suture closes too early. In these instances, the brain might not have enough room to grow to its usual size, leading to a build-up of pressure inside the skull (33).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Sutures of the human skull.

Possible factors related with a growing risk of craniosynostosis, as shown by epidemiological studies, are: white race, male sex, prolongation of gestational age, mother’s smoking, maternal nitrate salts, and infertility treatments, but the specific mechanism of action of the above factors remains unidentified (35-38). Half a century ago, scientists considered that other factors, such as the limited endometrial space available for the fetal skull in multiple pregnancies play a role in the emergence of early synostosis, especially of the sagittal suture (39, 40). However, according to more recent studies, natural endometrial restriction may lead to partial skull defects, without affecting cranial suture (41). Growth factors, such as fibroblast growth factors (FGFs), transforming growth factor-beta (TGF-β) and insulin-like growth factors (IGFs) have a crucial role in the process of early craniosynostosis (40, 42, 43). This disputes the previous conclusion that the syndrome depends on the space available and attributes its appearance on genetic factors instead.

In recent decades, the progress in mapping human genes has revealed that many of the syndromic craniosynostoses have a genetic origin. For example, mutations in the genes of three of the four fibroblast growth factor receptors (FGFR-1, FGFR-2, FGFR-3) have been associated with the onset of Apert, Crouzon, Jackson-Weiss, Muenke and Pfeiffer syndromes (44, 45). Also, mutations in the MSX and TWIST1 genes have been identified in patients with Boston-type craniosynostosis or Saether-Chotzen syndrome (45, 46). In addition, there are cases of non-syndromic craniosynostoses in which point mutations have been discovered; for example, mutations in the FGFR-3 receptor were found in cases of craniosynostosis of the coronary suture (1, 47). However, despite the identification of the above genetic mutations, the mechanism through which they produce a specific craniosynostosis phenotype remains cloudy, while for the more frequent form of craniosynostosis, i.e., the single synostosis of sagittal suture, no mutation has yet been found.

Fibroblast growth factor receptors (FGFRs) are a subset of receptor tyrosine kinases, which are activated when fibroblast growth factor (FGF), heparin sulfate proteoglycans (HSPGs), and FGFR form a trimolecular complex, resulting in receptor dimerization, autophosphorylation, and downstream signaling. Ligand-independent receptor activation is a feature of many FGFR point mutations involved in craniosynostosis syndromes but certain mutations, in fact, accentuate ligand binding (Table I). FGFs induce cell differentiation and migration playing important roles in bone growth, angiogenesis, wound healing, neuronal development, and tumorigenesis. Regarding craniosynostosis, the common mutations are located in FGFR2, even though there are also plenty in FGFR3 and one on FGFR1 (48). No mutations are known for FGFR4. Many mutations found on FGFR2 and FGFR3 affect cysteine residues and prevent the formation of disulfide bonds (Figure 2). This applies for Crouzon syndrome, Pfeiffer syndrome, and thanatophoric dysplasia type 1. Whereas most of the FGFR2 mutations are located in the two exons encoding the extracellular IgIII domain, namely IIIa/exon 8 and IIIc/exon 10, they can also occur in exons coding the tyrosine kinase region and exon 11, which encodes the transmembrane domain, as well as the IgI and IgII domains. The FGFR2 mutations in Crouzon and Pfeiffer syndrome overlap, most often occurring in extracellular IgIII domain encoded by either exons 8 or 10. Both syndromes may be caused by the Cys278Phe and Cys342Tyr missense mutations. FGFR mutations can also be found in the intracellular domain (49). FGFR may result in gain of function, in contrast to TWIST1 mutations in the Saethre-Chotzen syndrome, which usually result in haploinsufficiency.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

Most mutations for the main craniosynostosis syndromes exist in the FGFR2 gene, while there is only one known mutation in the FGFR1 gene. Depending on the mutation, there is a different syndrome [taken from (2) with modifications].

The TWIST1 gene encodes a transcription factor that is a primary regulator of many genes expressed in specific cell types, resulting in a tissue-specific phenotype (48). TWIST1 mutations in Saethre-Chotzen syndrome may be of the nonsense, missense, insertion or deletion type and are usually located outside or within the coding region (50). Most mutations truncate the TWIST1 protein, resulting in haploinsufficiency. Alternative mechanisms may cause loss of TWIST1 function either by affecting DNA binding or by interaction with regulatory proteins (50). A TWIST1 mutation (340A->G, Asn114Asp) has been noted in a patient with Crouzon-like features (45, 51).

MSX1 and MSX2, belong to the key family of Homeobox (or Box) genes that subdivides the early embryo into fields of cells with the potential of becoming specific tissues and organs. The MSX2 mutant appears to cause Boston type craniosynostosis (Table I) (52). MSX2 over-expression results in enhanced growth, favoring suture closure, whereas MSX2 downregulation or inactivation results in differentiation and delayed suture closure (48).

The genetic and biochemical information regarding suture formation and closure, combined with detailed clinical studies can lead to a pathophysiological hypothesis for craniosynostosis that will focus on processes playing a critical role on tissue boundary formation and the balance between bone cell proliferation and differentiation. Earlier developmental stages using human stem cells may be used, due to the fact that the relation between genes and pathogenetic mechanisms is not as much assured or since the phenotype of the human disorder cannot always be copied in mice. A recommended signaling pathway is the IL11RA-STAT3 pathway, which may affect remodeling of tissue boundary formation and the balance between proliferation/differentiation and constitutes a surprising link between craniosynostosis and immunity (53).

Molecular studies of Msx2 have revealed that it is necessary for the normal closure of sutures. Msx2 transgenic mice exhibit craniosynostosis with elevated concentrations of osteogenic cells in the skull. Some patients with Msx2 trisomy also exhibit craniosynostosis, indicating that Msx2 function is maintained in humans. Studies of transgenic Msx2 mice showing suture overgrowth and suture-less overlap indicated that the mechanism of premature suture closure in these mice is due to a particular increase in proliferation of osteogenic cells. These results demonstrate that Msx2 is a major modulator of proliferation of osteoporosis-inducing cells expressing Msx2 and subsequent osteoblast differentiation.

The transforming growth factor β (TGF-β) is responsible for osteoblastic differentiation and matrix producing potential. It also stimulates osteoid formation and appears to play a lesser role in premature craniosynostosis (54). While the role of TGF-β in suture regulation is evident, no TGF-β-related primary cause for craniosynostosis has been identified (48, 55, 56).

What is clear, however, is that the bones of the dome of the skull do not develop separately, but in a close and complex interaction with the developing brain and the underlying meninges. Recent studies have concluded that in craniosynostosis the brain architecture differs not only in the cortical regions, but also in several subcortical formations. The shape of the brain and skull is similar; however, differences between the two tissues have been observed. It has been established that the effects of craniosynostosis on the brain are observed throughout its entirety, with subcortical structures being changed according to cortical changes. In conclusion, the lack of direct correspondence in the physical effects of craniosynostosis in the skull and brain and implies that there is a large degree of independence in the developmental stages of the brain and skull, despite their physical and chemical connections between the units of the craniofacial complex (57, 58). This also leads to the hypothesis that perhaps the developing central nervous system plays an equally crucial role in the development of the terminal skull phenotype, so research into mutations that cause early closure of the cranial suture should also be directed to genes and mechanisms involved in the brain development itself (Figure 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

The figure illustrates the signaling pathways of transforming growth factor-β (TGFβ)/bone morphogenetic protein (BMP) and Wnt, which are involved in cell and developmental biology of cranial sutures. Mutations in genes that encode the fibroblast growth factor receptors 1, 2 and 3 along with mutations in genes encoding nuclear transcriptional factors TWIST1 and MSX2 are some of the main causes of craniosynostosis.

Common Syndromes of Craniosynostosis

Apert syndrome. The symptoms of classic Apert syndrome include brachycephaly, a flat nasal bridge and syndactyly of the hands, which are known as hand gloves, although the toes of the feet are also affected in similar ways (44, 45). Nevertheless, in non-classical cases a variable syndactyly is likely to exist.

Crouzon syndrome. Patients often show a long face, hypoplasia of the upper jaw and prominent lower jaw (44, 45). A conductive hearing loss is also common. It is related with the older age of the father. Another Crouzon syndrome variant is Crouzon’s syndrome with acanthosis nigricans; it is genetically different, and the progress of the acanthosis depends on older aging. Most characteristic patients exhibit synostosis with reddish surfaces, acanthosis nigricans with broad and atypical distribution, melanocytic embryos, hydrocephaly, Chiari malformations and mouth anomalies.

Pfeiffer syndrome. Patients exhibit hypertelorism, upper jaw hypoplasia, mandibular protrusion, and brachycephaly (44, 45). A partial syndactyly of the fingers of the hands and feet is also usual. There are three types: type 1 - milder form, type 2 - with clover-shaped head and elbow ankylosis, and type 3-serious cranio-synostosis, no clover-shaped skull, and premature death.

Jackson-Weiss syndrome. Patients exhibit craniosynostosis with facial hypoplasia and skin syndrome of the second and third finger (44, 45).

Muenke syndrome. Patients show macrocephaly, brachycephaly, facial hypoplasia, coronal synostosis, developmental delay, and hearing loss. Extreme abnormalities include brachydactyly, broad toes and back pain (44, 45). The affection of females is more frequent compared to males. Hearing loss is another condition observed in most patients, specifically in about 95% of the cases. It can range from mild to moderate, with studies suggesting that all individuals with Muenke syndrome are likely to have at least some degree of hearing loss, usually mild. Individuals can also exhibit hearing loss despite passing the newborn hearing screens, while others develop a progressive type that becomes more severe as they age.

Craniosynostosis 2 Boston type syndrome (MSX2 related). Patients affected by this syndrome exhibit an extended forehead, turribrachycephaly and an anomaly in the skull’s cloverleaf (59). Also, they present with supraorbital ridge hypoplasia, cleft palate, more teeth than normal, hand abnormalities, such as triphalangeal thumb and syndactyly of 3-4 fingers, as well as abnormalities in the feet, such as a short first metatarsal and middle phalangeal agenesis. Other related symptoms contain headaches, vision problems and epilepsy. Their mental capacity is not affected.

Clinical Diagnosis

The phases of diagnosing patients with craniosynostosis include clinical evaluation, molecular genetic investigation, interpretation of genetic test results, genetic counseling and patient management (15).

Clinical evaluation. It is a first step that should precede genetic tests. At this stage, the clinical geneticist evaluates the complete individual history, as well as the family history with special care to craniosynostosis, anomalous head shape, cranial surgery, skeletal or epileptic disorders, delayed development, or additional possible genetic conditions in the family. The evaluation of individual history should include pregnancy history and prenatal reports, birth and development history, possible magnetic resonance imaging of the brain, and X-ray cranial imaging that accurately depicts the bone. Furthermore, at this stage, patients are likely to undergo a complete physical examination to evaluate the involved cranial sutures and head shape, craniofacial and orofacial abnormalities, cerebral malformations, hearing and limb abnormalities, including possible syndactyly in addition to complementary imaging in other body regions that may also be affected, especially in syndromes.

Molecular genetic testing. Based on the data collected by history taking and clinical examination the appropriate genetic testing is performed. Whether the genetic testing is for specific genes, for a craniosynostosis gene panel or for whole exome sequencing depends mostly on the possible diagnosis based on family history and particular suture and other anomalies. Genetic counseling prior to genetic testing should address the needs and wishes of patients and their families and inform them about the test capabilities and limits, as well as what a positive or a negative result mean. Especially in cases showing additional malformations outside craniosynostosis or an unexplained developmental delay, a qualified clinical geneticist-dysmorphologist has to be involved so as to assess if cervical synostosis occurs because of a more complex chromosomal or monogenic disorder. For these patients, the initial series of genetic testing should include a karyotype or a chromosomal microarray examination to search for genomic imbalances (chromosomal overlaps or deletions) (60). The chromosomal array testing seems to be suitable for patients having severe developmental retardation and several abnormalities. The presence of chromosomal abnormalities in patients might require the investigation of the parents as well (61).

In patients that have an isolated coronary suture, genetic tests usually begin with screening for mutations in FGFR3 (Muenke’s syndrome). Testing proceeds to the FGFR2, FGFR1 and TWIST1 gene (for Crouzon syndrome, Pfeiffer’s syndrome, Saethre-Chotzen’s syndrome and Jackson-Weiss syndrome). The final stage of the genetic testing of these patients, if no causal mutation is located, is to estimate copy number variants (for example whole gene deletions) of the above-mentioned genes.

Interpretation of test results and genetic counseling. The genetic testing findings should be communicated to families in a genetic counseling session. The parents of a patient should be informed of the results of all genetic tests, irrespectively of their positive or negative outcome. The families should be advised that even though a specific initial genetic test (such as whole exome sequencing) is negative, that result does not exclude the possibility that the patient has a genetic etiology of craniosynostosis (for example: in the promoter region of a gene).

In incidents of frequently described craniosynostosis syndromes, whole families may need to be tested due to the autosomal dominant pattern of inheritance or varied expression. Families need to realize that testing the first person with a condition can indicate that the parent can also be affected. Patients’ relatives, who are seemingly clinically healthy, usually assume that they are automatically unaffected. Nevertheless, there are cases of reduced penetrance, as in families with Muenke’s syndrome, when a parent who seems to be clinically normal produces a more severely affected offspring.

The finding of a genetic cause of a condition clarifies the diagnosis and offers the best possible approach to accurate prognosis and appropriate therapy. In addition, the best possible genetic counseling may be offered to the family so that new cases can be prevented.

Therapeutical Approaches

Depending on the seriousness of the incident, treatments of craniosynostosis can vary. Non urgent and non-syndromic cases can follow the elective surgical management while other syndromic cases require urgent interventions. In severe cases, the conservation of the airway, support of nutrition, eye protection and normal intracranial pressure ought to be considered (62). For instance, patients having positional plagiocephaly may not require treatment. Most of the time, having a flattened area will not have a negative impact on brain growth or mental development. Deformities begin to improve when the child is able to stand and sit, even though they may not completely disappear. A total resolution of the deformities has also been observed with the progress of time. In some severe cases, surgery may be necessary, usually when unresolved flattening causes facial abnormalities, problems with chewing, eating or vision. In addition, patients may have difficulties with socialization due to their appearance. In other cases, a minimally invasive approach, such as endoscopic suturectomy, can accurately correct the positional irregularities. The use of remodeling helmets is also helpful for very young patients, followed or not by suturectomy (63-65). In addition, children with craniosynostosis usually show a growing need for dental care due to hypodontia or ill-shaped teeth, dental caries or tendency for periodontal disease, thus oral and facial surgeries for jaw and dental reconstructions might be also required (66).

To date, the strategies employed to treat craniosynostosis are almost absolutely surgical. The aim of these surgical procedures is to correct the premature fusion so as to increase the intracranial volume, decrease intracranial pressure and let the brain develop (13, 67-69). Existing surgical methods comprise cranial vault reconstruction, minimally invasive strip craniectomy with the use of postoperative molding helmet, minimally invasive strip craniectomy with spring implantation, and cranial distraction (4). However, these strategies carry risks for infection, bleeding, venous air emboli and damage to the unprotected brain (70). Even after surgical correction, a range of abnormal symptoms (such as post-operative edema) and increased intracranial pressure might happen. These problems occur with larger severity and more frequently in patients with syndromic craniosynostosis (16, 36). Therefore, these patients should be followed-up postoperatively up to the age of 12 years old when brain development is completed. Postoperative problems may also require repeated surgeries for secondary defects improvement, like midface hypoplasia, especially in patients with syndromic craniosynostosis conditions like Apert and Crouzon, with the same risks as previously mentioned (36). Therefore, there is a need not only to enhance the known strategies but at the same time to develop new biological therapies that would possibly prevent the need for surgery and could cure or even prevent premature suturing.

Research about the genetic basis of craniosynostosis syndromes and their underlying etiology has provided insights for their nonsurgical treatment. The identification of signaling pathways pathologically stimulated by the cranial suture and identification of proteins involved in the development of craniosynostosis will aid the selection of upcoming adjuvant medical therapies (71). According to recent studies, the fibroblast growth factor receptors (FGFRs), transforming growth factor beta (TGF-β), bone morphogenetic proteins (BMPs), as well as the transcription factors TWIST1 and MSX2 are the most important molecules involved in the development of both syndromic and nonsyndromic craniosynostosis (72, 73). Specific medication that is capable of controlling the expression of these signaling pathways could be a possible in the future (72, 73).

Important Issues of Proper Therapy

Early diagnosis and suitable treatment are very important in craniosynostosis, regardless its genetic or nongenetic etiology. Patients are routinely examined by a care team formed between different specialists, such as an orthodontist, a neurosurgeon, an oral and maxillofacial surgeon, an ophthalmologist, a psychologist, a pediatrician, a plastic surgeon, a speech and language pathologist, a clinical geneticist, and a molecular geneticist. The evaluation of the patients by this team is crucial to settle the diagnosis, recognize the level of the deformity and set the basis for a treatment plan and a surgical strategy for reconstruction (70). The identification of the genetic defect also permits recognition of the persons’ genetic risks and appropriate genetic counseling for the patients’ families (71). The results of genetic testing should be reported to families, taking into account that a negative result does not rule a genetic cause that might not be covered by the analysis (33).

The main critical factors that influence successful correction of craniosynostosis include age and correct timing of surgery. The best period for the operation is still controversial and is thought to be between 6 and 12 months of age in cases without evidence of increased intracranial pressure (74). Delay in craniosynostosis treatment can exacerbate associated facial skeletal abnormalities, such as facial asymmetry, strabismus, or malocclusions. The operations needed are usually performed in the young infant, a time when they are most susceptible to physiologic insults (75). Even though these ages are considered to be optimal for a surgical improvement of the head shape and re-ossification of the bone defects, it should be noted that events, such as infections, optic nerve ischemias, seizures, bleeding, and the need for blood transfusions are not insignificant in the infant or young child (76, 77). In addition, especially in nonsyndromic craniosynostosis cases, waiting for surgical intervention until the age of 6 months is related to the lowest risk for a surgery repetition (78). Patients operated on before 6 months old have higher probability of needing surgical repetitions, in contrast to those operated after 6 months old, which show decreased probability (79).