Article Text

Download PDFPDF
When chromatin organisation floats astray: the Srcap gene and Floating–Harbor syndrome
  1. Giovanni Messina,
  2. Maria Teresa Atterrato,
  3. Patrizio Dimitri
  1. Istituto Pasteur Italia-Fondazione Cenci Bolognetti, Dipartimento di Biologia e Biotecnologie “Charles Darwin”, Sapienza Università di Roma, Italy
  1. Correspondence to Professor Patrizio Dimitri, Istituto Pasteur Italia-Fondazione Cenci Bolognetti, Dipartimento di Biologia e Biotecnologie “Charles Darwin”, Sapienza Università di Roma Via dei Sardi 70, Roma 00185, Italy; patrizio.dimitri{at}


Floating–Harbor syndrome (FHS) is a rare human disease characterised by delayed bone mineralisation and growth deficiency, often associated with mental retardation and skeletal and craniofacial abnormalities. FHS was first described at Boston's Floating Hospital 42 years ago, but the causative gene, called Srcap, was identified only recently. Truncated SRCAP protein variants have been implicated in the mechanism of FHS, but the molecular bases underlying the disease must still be elucidated and investigating the molecular defects leading to the onset of FHS remains a challenge. Here we comprehensively review recent work and provide alterative hypotheses to explain how the Srcap truncating mutations lead to the onset of FHS.

  • Genetic diseases
  • Chromatin organization
  • SRCAP chromatin-remodelling complex
  • Epigenetics

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Chromatin and remodelling

In eukaryotic cells, genomic DNA interacts with specific proteins to form a compact nucleoprotein structure called chromatin. The term ‘chromatin’ was introduced in 1882 by Walther Flemming, a famous German anatomist. During his pioneering cytological studies on cell division, using aniline dye, he discovered a network of ‘fibrous’ material within the cell nucleus that could undergo structural changes during cell division. He called this material ‘chromatin’ to refer to a ‘stainable material’ (figure 1).1 ,2

Figure 1

Walter Flemming chromatin. The beautiful illustrations are drawn by Walter Flemming for his book titled Zellsubstanz, Kern und Zelltheilung (‘Cell substance, nucleus and cell division’).

The basic structural and functional unit of present-day chromatin is the nucleosome,3 ,4 an octameric disc consisting of two copies of each of the four histone proteins (H3, H4, H2A and H2B) with 147 bp of DNA wrapped around it. Chromatin organisation is highly dynamic and regulated by a variety of epigenetic changes mediated by histone-modifying enzymes and ATP-dependent chromatin remodelling complexes.5 These complexes are specialised multiprotein machines capable of sliding or displacing nucleosomes, thus altering histone–DNA interactions and making nucleosomal DNA more accessible to specific binding proteins during transcription, replication and DNA repair.5

ATP-dependent chromatin remodelling complexes generally consist up to 17 subunits that are highly conserved within eukaryotes. They are characterised by the presence of an ATPase subunit belonging to the superfamily II helicase-related proteins which carries an ATPase domain that contains two different subdomains, the DExx and HELICc regions, separated by a linker portion.6 This class has been further subdivided into at least four different families (SWI/SNF, ISWI, NURD/Mi-2/CHD and INO80) based on the presence of additional domains.

In particular, the INO80 (inositol requiring 80) family of remodellers, initially isolated from Saccharomyces cerevisiae, contains more than 10 subunits.7 In addition to INO80 complex, it includes the yeast SWR1, Drosophila Tip-60 and human SRCAP and P400 chromatin remodelling complexes that share many conserved subunits.5 The main function of INO80 family is to promote the exchange of canonical H2A with the H2A variant forms into nucleosomes in diverse animal and plant species.5 ,8 ,9 H2A variants are important epigenetic factors modulating chromatin architecture and function. They are involved in the formation of higher-order chromatin structure and implicated in many essential processes, including transcriptional regulation, DNA repair and chromosome segregation.5 ,7–11 Recently, the exchange of H2A with variant H2A.Z in cells of the nervous system has been implicated in the molecular basis of cognitive functions.12

Chromatin remodelling and genetic disorders in humans

Over the past decade, a growing body of experimental evidence has shown that mutations in genes encoding chromatin factors and epigenetic regulators, such as histone deacetylases, methyltransferases and members of ATP-dependent chromatin remodelling complexes, are crucial players in cancer and human developmental disorders.13–18

Among chromatin remodelling diseases, an emblematic case is represented by the human CHARGE syndrome, a genetic disorder that severely perturbs human development.19 ,20

The chromodomain helicase DNA-binding (CHD) 7 gene, which encodes a member of the CHD family of ATP-dependent chromatin remodelling enzymes, was found to be mutated in patients affected by the CHARGE syndrome.21–23 Most mutations identified in patients with CHARGE syndrome are due to nonsense or a frameshift mutations in the CHD7 coding sequence, which may results in CHD7 truncated variants, if the mutated transcript escapes nonsense-mediated mRNA decay. Whatever is the effect of CHD7 mutations (formation of truncated variant or loss of the protein due to transcript degradation) haploinsufficiency is likely to explain the dominant effect of about 250 CHARGE patient mutations.

The CHD7 protein can exert both positive or negative controls on gene expression by interacting with enhancer/promoter elements.24 In vitro studies suggest that the ATP-dependent chromatin remodelling activity of CHD7 is affected by CHARGE mutations,23 leading to a disruption of global gene expression. Moreover, CHD7 is required for the maintenance of chromatin accessibility at the promoters of CHD7 target genes in neural stem cells.25 Clearly, characterising the factors required for chromatin organisation and function would give a foothold for the comprehension of the mechanisms underpinning human developmental diseases.

Floating–Harbor syndrome

Floating–Harbor syndrome (FHS), also known as Pelletier–Leisti syndrome, is a rare human disease characterised by delayed bone mineralisation and growth deficiency, often associated with mental retardation and skeletal and craniofacial abnormalities.

FHS was first described at Boston's Floating Hospital in 197319 and 2 years later at Harbor General Hospital in Torrance, California.20 Since then about 50 cases have been reported in the medical literature. It appears to have no link with specific ethnic groups, since the above cited cases have been found in different countries, such as USA, Canada, Germany, Japan, Italy and the Netherlands. Common diagnostic features of FHS are growth deficiency that becomes apparent in the first year of life; delay of bone age in early childhood; typical facial features which include a triangular face, low hairline, deep-set eyes, long eyelashes and a large, distinctive nose that becomes more prominent with age, a shortened distance between the nose and upper lip and thin lips. Some affected individuals also exhibit various defects at the level of fingers and toes, in particular, brachydactyly, short broad thumbs and big toes and broad fingertips.

The majority of FHS cases present in the literature are sporadic, but reported cases of parent-to-child transmission suggested that, in at least some instances, the pattern of inheritance of the disease is autosomal-dominant21–23 in that one copy of the altered gene in each cell is sufficient to cause the disorder.

Although the FHS condition was first described 42 years ago, the causative gene, called Snf2-related cAMP response element-binding protein (CREB)-binding protein (CREBBP) activator protein (Srcap), was identified only recently by the exome sequencing of a cohort of 13 unrelated individuals with the classic features of FHS.24 This conclusion was confirmed by genomic studies carried out in an additional group of patients.25 ,26

Truncating mutations of Srcap gene are responsible for FHS

The Srcap gene maps to chromosome 16p11.2. It comprises a 40 989 bps long coding region with 34 exons (figure 2A) and encodes an SNF2-related chromatin-remodelling ATPase, the SRCAP protein (figure 2B), which is highly abundant in the nucleus of human cells (figure 2C).26 ,27

Figure 2

The Scrap gene and its encoded protein. (A) The wild-type Srcap gene maps to the short arm (p) of chromosome 16, band p11.2. It is composed of a 40 989 bps long coding region with 34 exons. (B) The SRCAP protein contains different domains. The ATPase domain is divided into two subdomains, one carrying the conserved motifs I–IV and the other containing motifs V–VI. (C) Indirect immunofluorescence experiments showing the subcellular localisation of SRCAP protein in HeLa cells. The SRCAP protein (red) exhibits a clear nuclear localisation (G. Messina, M.T. Atterrato and P Dimitri, unpublished). (A) and (B) were modified from figure 2 of Hood et al31. DAPI, 4’,6-diamidino-2-phenylindole; HSA, helicase-SANT-associated domain.

Genomic analyses indicate that FHS is indeed caused by a dominant mutant allele, as suggested by reported cases of parent-to-child transmission.21–23 First, 52 identified patients with FHS characterised thus far carry heterozygous truncating mutations, either nonsense or frameshift, in the Srcap gene, which is consistent with a dominant pattern of inheritance.28–30 Mutation hotspots were found in exon 34, located in the terminal portion of the Srcap gene, while in two FHS cases mutations occurred in exon 33.31–35 Nonsense or frameshift mutations may result in the formation of C-terminal-truncated SRCAP protein variants missing functional domains,28–30 which would then be responsible for a dominant negative effect. Second, two individuals carrying a 208 kb deletion of a region containing Srcap and nine other genes have no reported phenotype suggesting that Srcap deletion is haplosufficient.31 Finally, Gerundino et al36 described a patient carrying a 186 kb de novo microdeletion on 16p11.2, which encompasses nine genes including Srcap. This patient has a speech impairment and global developmental delays and only few subtle phenotypic features resembling FHS. However, further evidence is required to prove that haploinsufficiency of Srcap is indeed implicated in the generation of the above-mentioned defects.

In conclusion, the data available indicate that the Srcap deletion can be haplosufficient with no consequences on the phenotype or at most may have developmental defects different from those found in patients with FHS.

While it is clear that Srcap-truncating mutations trigger FHS, at present there is no experimental evidence showing the presence of truncated SRCAP proteins in patients with FHS. This critical aspect needs further investigation. For example, western blots could be done on protein extracts from primary cell lines derived from patients with FHS, using an anti-SRCAP antibody directed against the N-terminal portion of the SRCAP protein. This could permit the detection of both wild type and truncated variants.

Worth noting, truncating mutations have a crucial role in both FHS and CHARGE syndrome, determining their dominant transmission, but the underlying genetic consequences of these mutations appear to be different between the two syndromes. In fact, dominant negative effect due the Srcap-encoded truncated protein variants trigger FHS,28–30 while haploinsufficiency due to CHD7 null alleles explain a large number of CHARGE cases.21–23

The function of SRCAP protein

The SRCAP ATPase is one of the major subunits of the SRCAP chromatin remodelling complex that promotes the exchange of H2A with variant H2A.Z in humans.37–39 SRCAP-related proteins are widely distributed in the animal and plant kingdoms and therefore represent ancient members of the evolutionarily conserved SWI2/SNF2 family of ATPases/helicases.5

In particular, the yeast orthologue, SWR1, provides a platform for the assembly of the SWR1 complex that promotes the exchange of H2A with variant Htz1. The Drosophila melanogaster gene orthologue, domino,5 encodes two isoforms of the Domino (DOM) protein: DOM-A composed of 3202 amino acids (aa) and DOM-B composed of 2498 aa. Similar to SRCAP and p400, DOM-A can be isolated as part of a TRRAP/TIP60 histone–acetylase complex.9 The domino gene is essential for viability and has been implicated in several aspects of fly development, including haematopoiesis, wing development and female fertility.40 Genetic studies also suggest that domino may play a role in homeotic gene silencing and in heterochromatin-mediated silencing.40

Independently of its role in chromatin remodelling, the SRCAP protein can function as a transcriptional activator via binding to the cAMP response element-binding protein (CREB)-binding protein, called CREBBP or CBP.41 CBP regulates gene expression by mediating interactions between transcription factors and the basal transcriptional machinery.41 Notably, the CBP encoding gene was found to be mutated in half of patients affected by the Rubinstein–Taybi syndrome (RTS).42 It has been suggested that disrupted interaction between these two proteins likely explains some of the clinical overlap between FHS and RTS.33 As shown in figure 2B, in addition to the CBP-binding domain (1639-1988), SRCAP carries other functional domains, located at the C-terminal end (2316–2971), which activate transcription independently of CBP;41 one such domain may correspond to the portion containing the AT-hook motifs. The AT-hook is a small DNA-binding motif of about nine amino acids that was first described in the high mobility group non-histone chromosomal proteins called HMG-I and later was found in many other DNA-binding proteins that play a key role in chromatin organisation and gene expression controlling fundamental cellular processes.43 ,44

The molecular bases of FHS

A crucial question is how the Srcap truncating mutations trigger the onset of FHS. It is conceivable that the dominant effect underlying FHS is due to the loss of the C-terminal portion of SRCAP protein, including the AT-hook motifs.31 ,33 In Srcap+/ΔSrcap heterozygous FHS patients, where both the wild-type SRCAP and the truncated SRCAP variant (ΔSRCAP) are expressed, ΔSRCAP may disrupt the binding of wild-type SRCAP to both DNA and chromatin targets, thus affecting the expression levels of genes controlling the onset of differentiation and developmental processes.31 ,33 However, how could this effect be produced? At least two alternative situations can be considered, both resulting in a dominant negative effect. Perhaps in heterozygous patients, Δ-SRCAP competes with wild-type SRCAP in binding to chromatin partners (subunits of SRCAP complex, CBP and others) that in turn would be sequestered and mislocalised due to the absence of the AT-hook motifs and/or other C-terminal domains. Alternatively, Δ-SRCAP may physically interact with wild-type SRCAP to form inactive heterodimers unable to bind DNA elements and/or chromatin partners, which would result in the overall loss of functional SRCAP protein. As is well known, dominant negative effects can frequently be the consequence of the titration of endogenous proteins by the mutant form, giving rise to inactive heterodimers. At present, however, there is no evidence that SRCAP physiologically undergoes self-association to form homodimers.

Like yeast SWR1, human SRCAP protein may provide an assembly platform for other subunits of the SRCAP complex. Thus, apart from the primary cause of the dominant negative effect (competition or the formation of inactive heterodimers), truncated SRCAP variants might affect the assembly of a functional SRCAP complex, perturbing gene expression and leading to the developmental defects found in patients with FHS. If this is true, one would expect mutations in other SRCAP subunits implicated in the same regulatory networks to give rise to similar molecular defects, leading to overlapping aberrant developmental phenotypes. This conjecture is supported by studies on CFDP1, a member of the evolutionarily conserved BCNT protein family that participates in the exchange of H2A with variant H2A in different species.45 In fact, CFDP1 interacts with subunits of the SRCAP complex and has been also implicated in craniofacial development.46 ,47


Although Srcap was found to be the candidate gene for FHS, there are a number of questions that remain unanswered, and understanding the underlying molecular bases of FHS remains a challenge. In this regard, it is important to remind that SRCAP is a multifaceted protein. In fact, in addition to its various crucial roles in chromatin remodelling and gene expression, SRCAP is also important in DNA damage response48 and cell division (our unpublished data). Thus, different processes involving the SRCAP protein may be impaired in patients with FHS having SRCAP-truncating mutations and contribute to the origin of the developmental deficits found in FHS.

In-depth multidisciplinary functional approaches are required to explore the nature of FHS onset.

In particular, cellular models could be generated by expressing Δ-SRCAP variants in human cultured cells to reproduce the genetic defect found in patients with FHS. These models could be then used to investigate in details the effects of Δ-SRCAP on chromatin organisation and global gene expression, using the following approaches: (1) electrophoretic mobility shift assay to characterise DNA binding ability of Δ-SRCAP, (2) immunostaining and chromatin fractionation experiments to study the recruitment of SRCAP complex subunits to chromatin and (3) RNA seq approaches to perform comparative transcriptome analysis of wild-type SRCAP versus Δ-SRCAP expressing cells. The results from in vitro studies would be then validated on primary cell cultures derived from patients with FHS. A better understanding of molecular mechanisms leading to the onset of FHS will also be crucial to test epigenetic-based therapeutic approaches.


We thank Patrizia Lavia for helpful comments and suggestions.


View Abstract


  • Contributors PD is the leader of the group, he wrote the review article. GM and MTA are actually working on the molecular bases of the Floating–Harbor syndrome.

  • Funding Grant from Istituto Pasteur Italia- Fondazione Cenci Bolognetti; grant from Ministero Università e Ricerca, Progetti Universitari, Sapienza Università di Roma.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.