Costello syndrome was delineated based on its distinctive phenotype including coarse facial features, severe failure to thrive, intellectual disability, cardiac abnormalities and a high malignancy risk. Costello syndrome is caused by germline mutations in HRAS encoding a small GTPase which cycles between an inactive, GDP-bound and an active, GTP-bound state. Mutations in >90% affect Gly12 or Gly13 and are associated with a relatively homogeneous Costello syndrome phenotype. The same amino acid substitutions occur as somatic changes in malignant tumors and result in constitutive HRAS activation and increased RAF-MEK-ERK and PI3K-AKT signal flow. A few less common germline missense mutations affecting other HRAS codons were reported in patients with a distinctive, usually attenuated or mild, Costello syndrome (p.T58I, p.K117R and p.A146T/V), or in individuals with a predominant muscular phenotype (p.Q22K and p.E63K) . These changes were also suspected or proven to enhance HRAS-dependent signaling.
Here we report a novel heterozygous HRAS alteration, c.266C>G (p.S89C), in a female (Pt. 1) presenting with severe fetal hydrops and pleural effusion, followed by a more benign postnatal course including a renal cyst and strabismus. A brother (Pt. 2) with the same mutation and fetal polyhydramnios showed a Dandy-Walker malformation and his postnatal course was complicated by severe feeding difficulties. Their apparently asymptomatic father is heterozygous for the same change; however, we did not detect this alteration in 488 control alleles. In functional studies of COS-7 cells ectopically expressing various HRAS mutants, the binding domains of the HRAS effector proteins RAF1, PI3K, and RALGEF were used to specifically pull-down active HRAS. Upon growth factor-stimulation, co-precipitation of HRASS89C with any tested effector was decreased compared to wild-type HRAS, indicating a reduced growth factor-dependent activation of HRASS89C. Accordingly, we detected slightly diminished MEK, ERK and AKT phosphorylation in cells overexpressing HRASS89C. Thus, p.S89C appears to reduce downstream signaling, a novel consequence of disease-associated HRAS mutations.
The C. elegans RAS homologue let-60 is essential for vulval induction and activating gain-of-function mutations, such as p.G13E, result in multivulval organisms. Notably, expression of let-60 with the amino acid change p.S89F results in a vulvaless phenotype, thereby supporting our functional data .
The decreased downstream signaling effect of HRASS89C clearly differs from those for typical CS-associated HRAS mutations. Given the Patient 1’s benign postnatal course and presence of this change in her asymptomatic father, its harmful consequences may be time limited, with the late fetal stage being most sensitive. While the physical findings in both patients are not typical for Costello syndrome, they are consistent with a dysregulation of the RAS/MAPK pathway. Together, these data illustrate the wide functional and phenotypic variability of germline HRAS mutations.
 Kerr B. et al., J Med Genet, 43, 401–405 (2006); Zampino G. et al., Hum Mutat, 28, 265–272 (2007); Gripp KW. et al., Am J Med Genet, 146A, 683–690 (2008); Denayer E. et al., Hum Mutat, 29,232–239. (2008); Van der Burgt I. et al., J Med Genet, 44, 459–462 (2007).
 Beitel GJ. et al., Nature, 348, 503-509 (1990); Han M. & Sternberg PW. Genes Dev, 5, 2188-2198 (1991).