How to cite this article: Gripp KW, Innes AM, Axelrad ME, Gillan TL, Parboosingh JS, Davies C, Leonard NJ, Lapointe M, Doyle D, Catalano S, Nicholson L, Stabley DL, Sol-Church K. 2008. Costello syndrome associated with novel germline HRAS mutations: An attenuated phenotype? Am J Med Genet Part A 146A:683–690.
Costello syndrome associated with novel germline HRAS mutations: An attenuated phenotype?†
Article first published online: 4 FEB 2008
Copyright © 2008 Wiley-Liss, Inc.
American Journal of Medical Genetics Part A
Volume 146A, Issue 6, pages 683–690, 15 March 2008
How to Cite
Gripp, K. W., Innes, A. M., Axelrad, M. E., Gillan, T. L., Parboosingh, J. S., Davies, C., Leonard, N. J., Lapointe, M., Doyle, D., Catalano, S., Nicholson, L., Stabley, D. L. and Sol-Church, K. (2008), Costello syndrome associated with novel germline HRAS mutations: An attenuated phenotype?. Am. J. Med. Genet., 146A: 683–690. doi: 10.1002/ajmg.a.32227
- Issue published online: 25 FEB 2008
- Article first published online: 4 FEB 2008
- Manuscript Accepted: 2 NOV 2007
- Manuscript Received: 14 SEP 2007
- The Nemours Foundation
- NIH. Grant Number: 4P20 RR020173-01
- National Center for Research Resources
- mitral valve prolapse;
- pyloric stenosis;
Costello syndrome is a rare congenital disorder typically characterized by severe failure-to-thrive, cardiac abnormalities including tachyarrhythmia and hypertrophic cardiomyopathy, distinctive facial features, a predisposition to papillomata and malignant tumors, neurologic abnormalities, developmental delay, and mental retardation. Its underlying cause is de novo germline mutations in the oncogene HRAS. Almost all Costello syndrome mutations affect one of the glycine residues in position 12 or 13 of the protein product. More than 80% of patients with Costello syndrome share the same underlying mutation, resulting in a G12S amino acid change. We report on two patients with novel HRAS mutations affecting amino acids 58 (T58I) and 146 (A146V), respectively. Despite facial features that appear less coarse than those typically seen in Costello patients, both patients show many of the physical and developmental problems characteristic for Costello syndrome. These novel HRAS mutations may be less common than the frequently reported G12S change, or patients with these changes may be undiagnosed due to their less coarse facial features. In addition to the findings previously known to occur in Costello syndrome, one of our patients had hypertrophic pyloric stenosis. This led us to review the medical histories on a cohort of proven HRAS mutation positive Costello syndrome patients, and we found a statistically significantly (P < 0.001) increased frequency of pyloric stenosis in Costello syndrome (5/58) compared to the general population frequency of 2–3/1,000. Thus we add hypertrophic pyloric stenosis to the abnormalities seen with increased frequency in Costello syndrome. © 2008 Wiley-Liss, Inc.
Costello syndrome typically presents with a characteristic phenotype encompassing severe failure-to-thrive, cardiac abnormalities including tachyarrhythmia and hypertrophic cardiomyopathy (HCM), a predisposition to papillomata and malignant tumors, and neurologic abnormalities including nystagmus, hypotonia, developmental delay, and mental retardation [for review see Gripp and Lin, 2006]. Facial features become coarser with age. More than 80% of patients with Costello syndrome share the same underlying HRAS mutation, resulting in a G12S amino acid change; and almost all Costello syndrome causing changes affect one of the glycine residues in position 12 or 13 [Aoki et al., 2005; Estep et al., 2006; Gripp et al., 2006a,b; Kerr et al., 2006; van Steensel et al., 2006; Zampino et al., 2007]. We now report on two patients with novel HRAS mutations affecting amino acids 58 and 146, respectively.
Patients and families were identified through physician referral for Patient 1, and through self-referral for Patient 2. Informed consent was obtained based on protocols approved by the Institutional Review Boards at Texas Children's Hospital (H-17456) for Patient 1; and the A. I. duPont Hospital for Children (#2005-051) for both cases. Results were compared to data from a cohort of HRAS mutation positive patients enrolled in an ongoing study of Costello syndrome (A. I. duPont Hospital IRB #2003-006 and #2005-051).
This boy is the third child born to his nonconsanguineous 34-year-old mother and 45-year-old father by repeat cesarean at 37 weeks gestation after an uncomplicated pregnancy. His birth weight of 2.95 kg was at the 50–75th centile; length and head circumference were not documented. Apgars were 7 and 9, at 1 and 5 min, respectively. Transient tachypnea resolved after several days. Inability to breastfeed resulted in bottle feeding. Despite surgical correction of pyloric stenosis at age 6 weeks, failure-to-thrive persisted and nasal-gastric tube feeding was initiated. Severe constipation suggested Hirschsprung disease, but this diagnosis was not confirmed on biopsy.
During an evaluation for failure-to-thrive and developmental delay at age 9 months, his head circumference measured 46.3 cm (75th centile); length 67 cm (<3rd centile; 50th centile for 6 months); and weight 7.19 kg (<3rd; 50th centile for 4.5 months). A high and prominent forehead, dolichocephaly and minimally coarse facial features were noted (Fig. 1). A flat nasal bridge and posteriorly rotated ears with fleshy and dimpled lobes were seen. Lax skin was present on hands and feet. Deep palmar creases and prominent fingertip pads were noted. Physical findings commented on at varying ages included hypotonia, broad chest, mild ligamentous laxity, sparse hair and a skin complexion darker than his family members. An ophthalmologic evaluation revealed subtle horizontal nystagmus and intermittent exotropia. Surgical procedures included strabismus repair and adenoid and tonsillectomy at age 3 years. A brain MRI and skeletal survey were normal. On echocardiography at age years, mild mitral valve prolapse was the only abnormality noted.
The patient's development remained delayed with walking independently at around 24 months. He began to use spoken words at age 3 years. At age 6 years he prefers to play by himself and his behavior is challenging due to temper tantrums. He attends school with the help of an aide and continues to receive speech therapy. At age 6 years his height was 104.3 cm (−2 to −3 SD for age); weight 19.2 kg (25th centile) and OFC 55.0 cm (>98th centile; 50th centile for age 15 years). Facial features are minimally coarse (Figs. 2 and 3), and palmar creases slightly deep to normal (Fig. 4).
This female was born as the third child of a 37-year-old mother and a 35-year-old father after a pregnancy complicated by drug and alcohol exposure and lack of prenatal care. Gestational age at delivery was reported as 36 weeks. Her birth weight was 2.4 kg (25–50th centile); length was 48.3 cm (50–75th centile). Severe failure-to-thrive (Fig. 5) and gastro-esophageal reflux resulted in gastrostomy placement and fundoplication at age 7 months. She began to eat by mouth at age 15 months and the gastrostomy tube was removed at 17 months. Treatment resistant tachycardia was noted, and a VSD and PDA resolved without medical intervention. Deep palmar creases were seen. Arching was attributed to neurologic abnormalities. Growth hormone (GH) stimulation testing showed her to be GH deficient, and replacement therapy was given from age years. After GH therapy was started, tachycardia became an increasing problem and HCM was noted. The HCM required surgical myectomy at age years. Strabismus was treated with vision therapy and glasses. A bone density study resulted in a Z-score of −3.1, suggesting osteopenia.
At age years her height was 87.9 cm (below −5 SD); her weight 13 kg (<3rd centile) and her OFC 49.7 cm (within 2 SD of mean for age). Her facial features are mildly coarse with a short nasal tip and a long philtrum (Figs. 6 and 7) and her earlobes are mildly prominent. Her palmar creases are not strikingly deep (Fig. 8).
Her development was delayed with hypotonia noted in infancy, walking independently at age years, and using single words at 3 years. She spoke in sentences at around 4 years. At her current age of years she performs well in a regular kindergarten class with little additional help from her teacher.
Genomic DNA was extracted from buccal cells from Patient 1 and his parents, and Patient 2, using Pure gene DNA Isolation Kit (Gentra Systems, Minneapolis, MN; www.gentra.com). Patient 2's biological parents were unavailable for testing.
Sequencing of HRAS exons 2, 3, and 4 was performed using primers previously described in Gripp et al., 2006b.
Patient 1 has a heterozygous HRAS mutation consisting of 173C > T in exon 3, resulting in a predicted substitution of the threonine in position 58 by isoleucine (T58I). This result was consistent with that identified in blood cell derived DNA by a diagnostic laboratory. Neither parent carries the mutation. Biologic relationships were confirmed as previously reported [Sol-Church et al., 2006] using microsatellite markers. The family was genotyped around the mutation site in order to identify polymorphic markers to determine the germline origin of the mutation. A complete description of the polymorphic markers found in this region is published elsewhere [Sol-Church et al., 2006]. The family was informative at the P-region (Fig. 9A, Table I). The proband is heterozygous P1/P2 at this site, with the P1 allele being inherited from the mother, and P2 allele from the father.
|Individual||Genotypes at polymorphic sitesa|
The germline C > T mutation created a de novo recognition site for endonuclease EcoRV (5′-GATATC-3′), used to differentiate the mutant from the wild-type allele and to determine the parental origin of the mutation (Fig. 9B). An 862-bp region of the HRAS gene, containing the first and most of the second translated exons (exons 2 and 3), was amplified by PCR under standard conditions as recommended by the manufacturer, using Q-solution and Taq Polymerase with Int1F 5′-ATTTGGGTGCGTGGTTGA-3′ and Ex3R2 5′-TTGGTGTTGTTGATGGCAA-3′. Initial denaturation was performed for 5 min at 95°C, followed by 40 cycles of 95°C for 30 sec, 56°C for 30 sec, and 72°C for 90 sec. A final extension of 72°C for 10 min was added to complete the reaction. The unpurified product was digested with 10 units of restriction endonuclease EcoRV (New England Biolabs, Beverly MA; www.neb.com) for 1 hr at 37°C. The EcoRV digestion products (Fig. 9C) were separated by gel electrophoresis on a 1% agarose gel in TAE, and the 770-bp DNA fragment of the mutant allele was purified using the Qiagen Gel Extraction kit (Qiagen, Valencia, CA; www.qiagen.com). This fragment generated from the mutated allele was subjected to a second round of amplification using Int1F and Int2R 5′-CCTCTAGAGGAAGCAGGAGACA-3′ and sequenced in both directions using the ABI BigDye Terminator Cycle Sequencing Ready Reaction kit v3.1, with a 1/8 dilution of the terminator mix under standard conditions. The sequences were processed on an Applied Biosystems 3130xl Genetic Analyzer and analyzed using MacVector 7.1.1 (Accelrys, San Diego, CA; www.accelrys.com). Presence of the paternal P2 allele indicates the paternal origin of the mutation.
Patient 2 was found to have a heterozygous 437C > T in exon 4, resulting in a predicted alanine to valine change at position 146 of the protein product. This result is consistent with that reported on blood cell derived DNA by a diagnostic laboratory. Though no parental samples were available for testing, this DNA variant was not detected in any of the 89 other DNA samples analyzed. Furthermore, the mutation has not been reported as a polymorphic variant in the SNP database (www.snpper.chip.org), and is thus likely to be disease causing.
Patient 1 participated in cognitive testing of patients with Costello syndrome. His results on cognitive, language, and developmental tests were compared to the data obtained from other HRAS mutation positive patients as published previously [Axelrad et al., 2004; Axelrad et al., 2007]. His results fell within the 1 SD range of the mean for Costello syndrome patients in the Peabody Picture Vocabulary Test IV, a measure of receptive vocabulary; the Leiter International Performance Scale—Revised Brief IQ; and the Adaptive Behavior Composite of the Vineland Adaptive Behavior Scales—Second Edition. Patient 1's test scores for the Fluid Reasoning scale from the Leiter International Performance Scale—Revised were within +2 SD of the mean for Costello syndrome patients. Overall, his cognitive profile, as measured by the test described, is not significantly different from the cohort of patients with Costello syndrome.
Patient 2 does not speak English, and we were thus unable to perform these cognitive studies.
Pyloric Stenosis in Costello syndrome
We reviewed medical records and asked parents as available for a history of myotomy for hypertrophic pyloric stenosis. Amongst 56 HRAS mutation positive patients enrolled in our study on Costello syndrome, excluding the patients reported here, we identified four additional cases, one female and three males. One of these patients was previously published [Patient 4 in Johnson et al., 1998] and his pyloric stenosis was mentioned in the original report. This patient has since been found to carry the common G12S mutation, as do the other three newly identified patients with pyloric stenosis in our cohort. Including our Patient 1, pyloric stenosis requiring surgical repair occurred in 5/58 patients with Costello syndrome (8.6%). The population frequency is 2–3 cases per 1,000 live births, with a male predominance of 4:1 to 6:1 [To et al., 2005]. Statistical comparison by Fisher exact test using the ratio of 5/58 for pyloric stenosis in Costello syndrome, and 2.5/1,000 for pyloric stenosis in the general population, results in a P-value of <0.001.
The facial features of our two patients (Figs. 2, 3, 6, and 7) are less coarse than those of other patients with Costello syndrome at comparable ages. Neither patient has developed papillomata, however, these typically arise in late childhood or adolescence. Neither patient shows the very deep palmar (Figs. 4 and 8) or plantar creases, or the ulnar deviation of wrists and fingers often seen in Costello syndrome. Both patients walk well independently. While our patients present with an apparently mild facial expression of Costello syndrome at their current ages of 6 and 5 years, a review of their medical history reveals many issues typical for Costello syndrome (Table II). Severe failure-to-thrive led to use of feeding tubes; short stature and relative macrocephaly is present; and hypotonia with developmental delay was seen in both. Patient 1 was diagnosed with nystagmus, a finding common in Costello syndrome. His current cognitive abilities are not significantly different from that of other patients with Costello syndrome. Patient 1's history of pyloric stenosis led us to review our patient cohort for this finding. Pyloric stenosis, ascertained in 5 of 58 Costello patients, was statistically significantly increased, compared to a population frequency of 2–3 per 1,000. It is tempting to speculate that the pyloric hypertrophy results from similar mechanisms as the HCM frequently seen in Costello syndrome. Patient 1 had mitral valve prolapse, a finding seen in other patients with Costello syndrome [Gripp et al., 2006a]. Because valve prolapse was not considered a cardiovascular malformation (CVM) in the published cohort of Costello patients [Gripp et al., 2006b], for consistency of the data used in Table II, we do not consider this a CVM.
|Patient 1||Patient 2||Costello (%) [Gripp et al., 2006b]|
|Ulnar deviation||−||−||25/33 (75)|
|CNS abnormality||−||−||9/33 (27)|
|GH deficiency||NA||+||15/33 (45)|
|Any cardiac abnormality||+MVP||+||30/40 (75)|
|Cardiac hypertrophy||−||+||19/40 (47)|
|Pyloric stenosis||+||−||4/56a (7)|
Patient 2 has growth hormone deficiency, a finding common in Costello syndrome. She required surgical myectomy for severe HCM and treatment resistant tachycardia, cardiac anomalies typical for Costello syndrome. Thus, both patients show findings characteristic for Costello syndrome, despite their less striking facial features. We previously reported somatic mosaicism for an HRAS mutation in a patient with some findings unusual for Costello syndrome [Gripp et al., 2006a]. Therefore it is noteworthy that in both patients reported here identical mutations were identified on cheek swab derived DNA, and on blood cell derived DNA analyzed in clinical laboratories. Thus we have no evidence suggesting somatic mosaicism as the reason for their less striking facial findings.
Because the HRAS T58I mutation seen in Patient 1 is a novel change, it was important to perform parental studies in order to show that this sequence change is not a familial polymorphism. The mutation occurred in the paternal germline, as was the case in 14/16 informative families reported by Sol-Church et al. 2006 and 9/9 reported by Zampino et al. 2007. In contrast to the glycines in positions 12 and 13 of the HRAS protein, which are often affected by mutations in sporadic tumors, the threonine in position 58 is not known to be a mutation hot spot in malignancies. Functional studies are needed to prove that the predicted amino acid change results in gain-of-function and increased activation of the MAPK signaling pathway. Threonine 58 is a highly conserved amino acid within the Switch II region of small GTPases. This domain is the gamma phosphate-sensing domain of the RAS protein, which drives the conformational switch of RAS during the GDP/GTP cycle [Pai et al., 1990]. The close sequence and functional homology of the HRAS and KRAS proteins may allow inference of information derived from a T58I change in KRAS. A female with KRAS T58I presented with cardiac findings including pulmonic valve stenosis, ASD and VSD; severe developmental delay; sagittal suture synostosis, macrocephaly, and juvenile myelomonocytic leukemia. She was clinically diagnosed with Noonan syndrome [Schubbert et al., 2006]. Functional studies showed defective intrinsic GTP hydrolysis and impaired responsiveness to GTPase activating proteins in recombinant T58I KRAS, resulting in hypersensitivity to growth factors [Schubbert et al., 2006].
The A146V HRAS mutation seen in Patient 2 is also novel, and unfortunately we could not prove the de novo nature of this mutation because the biologic parents are unavailable. However, an amino acid change affecting the alanine in position 146 (A146T) was previously reported in one patient clinically diagnosed with Costello syndrome [Zampino et al., 2007]. That patient was reported to require a feeding tube until age 6 years, but her growth was “less compromised;” minor involvement of skin and joints was observed. She was microcephalic, her hair was sparse and thin, but not curly, and her ears lacked the “distinctive fleshy and forward-cocked lobes” [Zampino et al., 2007]. This description may suggest a milder facial phenotype, similar to our Patient 2. The Ala146 is positioned in the guanine binding pocket of the RAS protein, thus substitution by a more reactive amino acid, such as threonine, is likely to change the GTP binding affinity of the mutated protein. Upon review of amino acid changes affecting the alanine in position 146 of the highly homologous KRAS protein, no patient with a germline mutation has been reported. However, the alanine in position 146 of KRAS is a mutation hot spot in isolated malignancies, with threonine and valine substitutions [Sanger Institute, Cosmic database, 2007] suggesting an activating, oncogenic effect of these mutations.
Despite a suggestion by Kerr et al. 2006 that the malignancy risk is increased in patients with the G12A mutation compared to those with the G12S change, no statistically significant data are available proving a mutation dependent variation of the cancer risk in Costello syndrome. Given the lack of statistically valid data for the more common mutations, we cannot speculate on how the malignancy risk for our two patients with novel mutations compares to that for patients with the G12S change.
In summary, we report on two patients with novel HRAS mutations, whose facial findings are less coarse than those of other patients with Costello syndrome. Despite the milder facial features, both show numerous findings typical for Costello syndrome. Interestingly, Patient 1's history of hypertrophic pyloric stenosis led us to recognize this as a finding occurring with increased frequency in Costello syndrome. It is impossible to draw phenotype–genotype conclusions based on single patients, therefore we cannot know if our patients' apparent milder facial phenotype is typical for patients with these particular germline mutations. Functional studies on these two novel presumed Costello syndrome causing mutations have yet to be completed. Furthermore, we do not know if these mutations are rare compared to the more common Costello syndrome causing mutations affecting the glycine residues in position 12 or 13, or if these patients remain unidentified due to their less characteristic facial findings.
We thank the patients and their families for allowing us to share this information. This report was supported by The Nemours Foundation and by funds to KSC from NIH grant number 4P20 RR020173-01 from the National Center for Research Resources.
- 2005. Germline mutations in HRAS proto-oncogene cause Costello syndrome. Nat Genet 37: 1038–1040. , , , , , , , , , , .
- 2004. Adaptive skills, cognitive, and behavioral characteristics of Costello syndrome. Am J Med Genet Part A 128A: 396–400. , , , .
- 2007. Longitudinal assessment of cognitive characteristics in Costello syndrome. Am J Med Genet Part A 143A: 3185–3193. , , , , .
- 2006. HRAS mutations in Costello syndrome: Detection of constitutional activating mutations in codon 12 and 13 and loss of wild-type allele in malignancy. Am J Med Genet Part A 140A: 8–16. , , , , .
- 2006. Costello Syndrome in: GeneReviews at GeneTests: Medical Genetics Information Resource [database online]. Seattle: Copyright, University of Washington. 1997-2006. Available at http://www.genetests.org. , .
- 2006a. Somatic mosaicism for an HRAS mutation causes Costello syndrome. Am J Med Genet Part A 140A: 2163–2169. , , , , .
- 2006b. HRAS mutation analysis in Costello syndrome: Genotype and phenotype correlation. Am J Med Genet Part A 140A: 1–7. , , , , , , , , , , , , , , , .
- 1998. Costello syndrome: Phenotype, natural history, differential diagnosis, and possible cause. J Pediatr 133: 441–448. , , , , , , , , .
- 2006. Genotype-phenotype correlation in Costello syndrome; HRAS mutation analysis in 43 cases. J Med Genet 43: 401–405. , , , , , , , , , , , , , , , , , , , , , , , , , .
- 1990. Refined crystal structure of the triphosphate conformation of H-ras p21 at 1.35 A resolution: Implications for the mechanism of GTP hydrolysis. EMBO J 8: 2351–2359. , , , , , .
- Sanger Institute: Catalogue of Somatic Mutations in Cancer. 2007. Distribution of somatic mutations in HRAS. www.sanger.ac.uk.
- 2006. Germline KRAS mutations cause Noonan syndrome. Nat Genet 38: 331–336. , , , , , , , , , , , , , , , , , .
- 2006. Paternal bias in parental origin of HRAS mutations in Costello Syndrome. Hum Mutat 27: 736–741. , , , , .
- 2005. Population demographic indicators associated with incidence of pyloric stenosis. Arch Pediatr Adolesc Med 159: 520–525. , , , .
- 2006. Recurring HRAS mutation G12S in Dutch patients with Costello syndrome. Exp Dermatol 15: 731–734. , , , , , , .
- 2007. Diversity, parental germline origin, and phenotypic spectrum of de novo HRAS missense changes in Costello syndrome. Hum Mutat 28: 265–272. , , , , , , , , , , , , , , , , , , , .