Secondary Logo

Journal Logo

SECTION I: SYMPOSIUM: Genetics in Orthopaedics

Functional Modeling of the ACVR1 (R206H) Mutation in FOP

Groppe, Jay C PhD; Shore, Eileen M PhD; Kaplan, Frederick S MD

Editor(s): Dobbs, Matthew B MD

Author Information
Clinical Orthopaedics and Related Research: September 2007 - Volume 462 - Issue - p 87-92
doi: 10.1097/BLO.0b013e318126c049
  • Free

Abstract

Fibrodysplasia ossificans progressiva (FOP), a rare and illustrative genetic disorder of progressive heterotopic ossification, is the most disabling condition of ectopic skeletogenesis. Children who have FOP appear normal at birth except for congenital malformations of the great toes.9 Typically, during the first decade of life, sporadic episodes of painful soft tissue swellings (flareups) occur and are commonly mistaken for tumors.10 These flareups seize the body's skeletal muscles, tendons, ligaments, fascia, and aponeuroses and transform them through an endochondral process into ribbons, sheets, and plates of heterotopic bone that span the joints, lock them in place, and render movement impossible.13 Attempts to remove this heterotopic bone usually lead to episodes of explosive new bone formation. Even minimal trauma such as minor soft tissue injury, muscular stretching, overexertion and fatigue, intramuscular immunizations, injections for dental work, falls, and influenza-like illnesses lead to episodic flareups of the condition.9 Most patients are wheelchair-bound by the end of the second decade of life. The median lifespan is approximately 45 years, and patients commonly die of complications of thoracic insufficiency syndrome.8 The worldwide prevalence is approximately one in 2 million. There is no ethnic, racial, or geographic predilection to FOP. Most cases are spontaneous. Inheritance, when observed, is autosomal-dominant with variable expression.18 At present, there is no definitive treatment.5

A large body of work has supported dysregulated bone morphogenetic protein (BMP) signaling in the pathogenesis of the condition.1-3,17 A single common heterozygous mutation (617G>A; R206H) has recently been identified in the cytoplasmic domain of activin receptor IA/activin-like kinase 2 (ACVR1/ALK2), a BMP type I receptor in affected individuals of five small multigenerational families and in all sporadically affected individuals with the features of classic FOP.19 Identification of the protein, and remarkably a singly substituted amino acid residue, that is altered in FOP lays the groundwork for elucidating the structural basis for the dysregulation of BMP signaling and concomitant formation of extraskeletal bone. Presently however, of the seven TGFβ superfamily type I receptors (ALKs 1-7), only the 3-D structure of the cytoplasmic domain of the TGFβ type I receptor (TβRI) has been determined.6

Nonetheless, because the cytoplasmic domains are highly homologous, the crystal structure of the cytoplasmic domain of TβRI provides a model for interpreting the structural basis of altered ACVR1 function in FOP. We found that, in both the template crystal structures and wild-type homology models, conserved basic residues (arginine or lysine) appear to form salt bridges or ion pairs with an invariant acidic residue (aspartate), linking the glycineserine (GS) activation domain, which includes the FKBP12 interface, with the L45 loop Smad specificity site. We hypothesized that substitution of arginine 206 with histidine, and only histidine, introduces a pH-sensitive switch that can induce a conformational change in the protein backbone at decreased intracellular pH, leading to ligand-independent activation of the receptor in FOP.

MATERIALS AND METHODS

TGF-β superfamily type I receptor (or ALK [activin-like kinase]) sequences were compiled from selected PubMed entries: ALK-1 (CAA80255), ALK-2/ActRIA/ACVR1 (NP 001096), ALK-3/BMPRIA (P36894), ALK-4/ActRIB (P36896), ALK-5/TβRI (NP 004603), ALK-6/BMPRIB (NP 001194), and ALK-7 (Q8NER5). Full-length receptor sequences (extracellular, transmembrane, and cytoplasmic kinase domains) were aligned and ordered as shown by ClustalW through a Web-based server (EMBL European Bioinformatics Institute; www.ebi.ac.uk/clustalw/). TβRI crystal structure coordinate (PDB) and structure factor files were downloaded from the Protein Data Bank (www.rscb.org) and MTZ files generated by a single round of rigid body refinement with Refmac5 in the CCP4 Program Suite. Electron density maps were created automatically and viewed graphically along with their corresponding crystal structure models in Coot (www.ysbl.york.ac.uk/~emsley/coot/). Structure-based homology models of wild-type, R206H, and R206K ACVR1 cytoplasmic domains were calculated through the automated SwissModel routines (Biozentrum, Basel; http://swissmodel.expasy.org//SWISS-MODEL.html) with the FKBP12-bound crystal structure of TβRI (PDB code 1B6C) as a 3-D template. Chemical structures of the ACVR1 R206 and D269 sidechains were prepared with ChemDraw (Cambridge-soft, Cambridge, MA).

RESULTS

Residue 206 of ACVR1 is highly conserved among vertebrate orthologs of the receptor and is positioned near the C-terminal end of the GS activation region.19 This residue has also been conserved throughout evolution within the superfamily of TGFβ type I receptors (Fig 1). Five of the seven human type I receptors share an arginine at this position and the remaining two (BMPRIA and BMPRIB) are conservatively substituted with lysine.

F1-15
Fig 1:
Multiple sequence alignment of TGF-β superfamily Type I receptors. (Upper segments) The glycine-serine (GS) loop and α-helix 2 (αGS2) of GS activation regions are depicted. The conservatively substituted basic residues (arginine or lysine) at the penultimate position of the αGS2 helices are highlighted in bold (TβRI R203, ACVR1 R206). Ligand-induced sites of activating phosphorylations in the GS loop of TβRI are shown in bold with a dark gray background. The binding site of the inhibitory FKBP12 protein on TβRI, comprised of a pair of tandem hydrophobic residues at the N-terminal end of αGS2, is boxed with a light gray background. (Lower segments) The β4-β5 hairpin and L45 loop Smad specificity sites are depicted. The invariant acidic residue (aspartate) centered in the β4 strands is highlighted in bold (TβRI D266, ACVR1 D269).

To gain insight into the structural role of this conserved residue, we examined crystal structures of the cytoplasmic domain of the TGFβ type I receptor (TβRI), the only member of the superfamily to be characterized structurally to date. In the structure of the cytoplasmic domain of TβRI in complex with the inhibitory FKBP12 protein,6 this conserved basic residue (TβRI R203) appears to form a salt bridge or ion pair with an invariant acidic residue (TβRI D266). Both sidechains lie within well-defined electron density and share low B-factors indicative of a high degree of order for, and lending confidence to, their positions in the x-ray model. In addition, this interaction was observed between TβRI R203 and D266 in all of the other crystal structures of this domain deposited in the Protein Data Bank (Table 1) with only one exception. The salt bridge was evident in three structures of the cytoplasmic domain in complex with small molecule inhibitors despite substantial truncation of the GS region in some of the TβRI expression constructs crystallized. Truncation of the GS region up to TβRI T200 left only five residues of αGS2 sufficient to form one turn of the helix and to properly orient R203 to pair with D266.16 The only TβRI structure that lacked the R203-D266 interaction was truncated by an additional residue, which abolished formation of the helix and projected the sidechain out into the solvent.4 Importantly, because these complexes contained proteins of different sizes and were crystallized under different conditions and into different crystal forms, the salt bridge observed in the four complexes was unlikely an artifact of crystallization. Thus, the salt bridge between the conserved basic residue corresponding to ACVR1 R206 and an invariant acidic residue is well defined, robust, and recurrent in multiple crystal structures.

T1-15
TABLE 1:
TβRI Cytoplasmic Domain Crystal Structures in the Protein Data Bank

Salt bridge formation is not limited to arginine and aspartate residues. The protonated sidechains of all three basic residues (arginine, lysine, and histidine), which are positively charged, can form ion pairs with the deprotonated sidechains of both of the acidic residues (aspartate and glutamate), which are negative. The sidechain of glutamate, extended by one methylene group (-CH2-) relative to aspartate, forms salt bridges with the protonated sidechains of the basic amino acids in a related fashion. Likewise, the sidechain of lysine mimics that of arginine with respect to the flexibility of the long aliphatic linker between the backbone and the terminal amino group and thus can function interchangeably.

In contrast, however, histidine is not structurally analogous to the other basic residues and differs in three important respects. First, the amino groups of the sidechain of histidine are constrained within the imidazole ring and extended from the backbone by a single methylene linker, greatly restricting the conformational freedom of this basic sidechain relative to those of arginine and lysine. Second, the single trigonal N-H bond that participates in ion pair formation imposes strict restraints on the geometry of the interaction. Finally, and perhaps most importantly, unlike those of arginine and lysine, which are fully ionized at physiological pH, the sidechain amino group of histidine is largely unprotonated, becoming only half protonated at approximately pH 6, the lower end of the physiological range. Because lysine can substitute functionally for arginine in salt bridge formation, presumably the conservatively substituted lysine in BMPRIA and BMPRIB (Fig 1) can form ion pairs with the invariant aspartate in the same manner as the corresponding ion pair in TβRI. However, histidine at this position (like in ACVR1 R206H) would participate in a salt bridge only at decreased intracellular pH and with extensive structural rearrangement, providing a hypothetical role and suggesting physiological implications for this mutation in altered ACVR1 function in FOP.

To test this hypothesis, structure-based homology models of wild-type and mutant ACVR1 kinase domains were produced (Fig 2). As a result of the extensive homology between the cytoplasmic domains (Fig 1), the models could be calculated with considerable confidence and provided a credible basis for interpreting the consequences of evolutionary divergence or mutation. As in the TβRI crystal structures, we observed a clearly defined salt bridge in the wild-type model with arginine at position 206 (Figs 2A, B). Lysine, the conservative substitution in BMPRIA and BMPRIB, could be accommodated in a fashion similar to arginine without structural rearrangement as anticipated (not shown). The histidine-substituted mutant adopted the same orientation (Fig 2C) but would require protonation of the sidechain imidazole group and major rearrangement to properly orient the N-H group with the carboxylate of aspartate to form a salt bridge (Fig 3). Based on the results of our protein modeling, we propose that substitution of arginine 206 with histidine creates a pH-sensitive switch within the activation domain of the receptor that leads to ligand-independent activation of ACVR1 in fibrodysplasia ossificans progressiva.

F2-15
Fig 2A:
C. Structure-based homology model of human ACVR1. (A) The overall structure of the cytoplasmic domain is portrayed. The juxtamembrane regulatory region (foreground) is composed of the helix-loop-helix glycine-serine (GS) activation region (αGS1, GS loop, αGS2) and β-strands 1-5 and α-helix C of the N-terminal lobe (including the β4-β5 hairpin and L45 loop Smad specificity site). The catalytic domain of the receptor kinase is shown in the background. (B) The salt bridge between arginine 206 and aspartate 269 in wild-type ACVR1 is depicted. Ionic interactions between the sidechain nitrogen (blue) and oxygen (red) atoms are represented by dots and bond lengths given in angstroms. The view is from the perspective of behind the overall structure. (C) The relative positions of the histidine 206 and the aspartate 269 sidechains in the model of the ACVR1 mutant are portrayed. A nitrogen atom of the histidine sidechain lies proximal to the carboxylate oxygen atoms of the aspartate sidechain yet lacks the orientation needed to interact ionically and form a salt bridge.
F3-15
Fig 3:
pH-induced salt bridge between histidine 206 and aspartate 269. (left) The orientations of the histidine 206 and the aspartate 269 sidechains in the model of the ACVR1 mutant at physiological pH, as in Figure 2C, are shown. (right) At lower pH, the basic sidechain of histidine is titrated by H+, inducing the formation of a salt bridge and structural rearrangement on one or both sides.

DISCUSSION

pH-sensitive switches have been postulated for numerous proteins regulated by conformational change.12 On protonation or deprotonation, these switches can produce small but important shifts in their relative positions. Consisting of only pairs or small clusters of residues, these switches induce major rearrangements in the tertiary or quaternary structure of proteins as a result of amplification of the small displacements by secondary structure elements (α-helices, β-strands) acting as hinges and levers.14

Histidine is a common element of pH-sensitive molecular switches because the pKa of its imidazole group (approximately 6.0) allows the sidechain to protonate and deprotonate within the physiological pH range.14 As a result, the histidine sidechain becomes partially ionized as the pH decreases from the midpoint (approximately 7.4) to the lower end of the range, going from an uncharged to a positively charged state. Histidine is unique in this respect, because the four other sidechain groups (two other basic, two acidic) that participate in ion pairs remain fully charged over the entire range of physiological pH.

This singular role in ion pair formation has been well documented in hemoglobin, which provides a classic example of a pH-sensitive switch between histidine and aspartate, the two residues postulated as comprising a switch in the ACVR1 mutant. In his crystal structures of hemoglobin, Perutz12 observed that the C-terminal histidine residue of each β-chain (Hisβ146) forms an ion pair with an internal aspartate (Aspβ94) that would further stabilize the deoxy (tight) form of hemoglobin in tissues at low pH. This linkage has been implicated in the regulation of oxygen binding by hydrogen ion, which plays a key role in the Bohr Effect. In actively metabolizing tissues such as muscle that need oxygen and produce lactic acid, thereby lowering the pH (approximately 7.2), protons promote the release of bound oxygen from hemoglobin. Conversely, in the lungs, the pH of blood is elevated (approximately 7.4), disrupting the linkage stabilizing the tight conformation and allowing hemoglobin to become oxygenated.

Because of its proximity to important regulatory elements within the GS activation domain (Fig 2A), the histidine-aspartate ion pair in the ACVR1 mutant could induce ligand-independent activation of the receptor in FOP by several potential mechanisms. Histidine 206 of the mutant (Fig 2C) extends from a rigid secondary structure element (αGS2) to form a putative pH-sensitive switch with aspartate 269, which also protrudes from a fixed element of secondary structure (β4-β5 hairpin). This pair of residues is juxtaposed between two functionally important sites, the GS loop and the L45 Smad specificity loop, and may act as a hinge that alters the position of the αGS2 helix. As a result of its rigidity, the αGS2 helix may act as a lever, amplifying small displacements of the histidine and aspartate residues into large conformational changes within the GS loop, the site of activating phosphorylations. Lever-like movement of the αGS2 helix could also have dramatic effects on the structural integrity of the binding site for FKBP12, an inhibitory protein that stabilizes the inactive conformation of the receptor.7 Structural alterations affecting the recruitment of the R-Smad effector proteins at the adjacent L45 loop Smad specificity site are also conceivable.

Consistent with the key role of the C-terminal end of the αGS2 helix in activation of the receptor, substitution of the terminal polar residue (threonine or glutamine) flanking the penultimate basic residue (Figs 1, 2) with an acidic one (aspartate or glutamate) results in ligand-independent, constitutive activation of the receptor.21 Because of the separation between this residue and the GS loop and the FKBP12-binding site, the structural basis for the deregulating effect of this substitution has not been apparent.6 However, in light of the bordering ion pair identified in this study, it is tantalizing to speculate that an acidic residue immediately adjacent to the conserved basic residue on the αGS2 helix might outcompete the invariant aspartate on the distal β4-β5 hairpin in ion pair formation, leading to destabilization of the secondary structure of the helix and ultimately the tertiary structure of the remote regulatory elements of the GS region.

A fundamental aspect of the ligand-independent activation mechanism is that the molecular switch is only predicted to be responsive in the R206H mutant context, ie, when arginine is substituted with the pH-sensitive residue histidine. Because the mutation results in loss of regulation, and not activity of the receptor, we hypothesize that only histidine can have an activating effect at this position and that this substitution is revealed through the unambiguous clinical features of FOP. Apparently, substitution with any of the 17 nonbasic residues at codon 206 has no prominent effect, rendering such changes largely silent. Nonetheless, because the residues on both sides of the salt bridge have been maintained throughout evolution, this structural element must endow the receptor with added stability or play an important functional role. Perhaps substitution with any of the nonbasic amino acids, which cannot participate in a salt bridge, or even with histidine, which cannot near neutrality, does in fact cause loss of function of the mutant receptor. In most contexts in a heterozygote, this loss of ACVR1 function may be compensated for by the wild-type gene product and only produce phenotypes resulting from haploinsufficiency in a subset of cellular or developmental processes. Hence, the congenital malformation of the great toe, one of the characteristic features of FOP, might arise from loss of function of the R206H mutant receptor at neutrality during embryogenesis, whereas heterotopic ossification would result from ligand-independent activation at low pH. An FOP-like toe malformation is indeed seen in the BMPRIB knockout mouse and in human BMPRIB heterozygotes, which exhibit a form of brachydactyly.11,22

Our hypothesis for the mechanism of ligand-independent activation of ACVR1 in FOP, supported by the structure-based homology modeling described here, is amenable to further testing by functional analyses in cultured mammalian cells and possibly model organisms such as the mouse and zebrafish. The most crucial test is to determine whether the ligand-independent activity of the ACVR1 R206H mutant kinase is sufficiently increased at the lower range of physiological pH (histidine sidechain partially ionized) relative to its activity at the higher range, where the uncharged sidechain cannot participate in for mation of a salt bridge.

In the event that the His206-Asp269 salt bridge is indeed shown to form in ACVR1 R206H and function as an aberrant switch that induces ligand-independent BMP signaling, the pH sensitivity of the histidine sidechain could then serve as a tractable therapeutic target. Flareups of FOP might often result from physiological changes such as trauma-associated hypoxia that lower intracellular pH and turn the switch on. If so, perhaps the intracellular pH of affected connective tissues could be modulated to suppress flareups of FOP or diminish the extent of heterotopic ossification. Interestingly, generation of a hypoxic environment triggered by BMP-2 in muscle has recently been shown to be a critical step in the formation of heterotopic bone in a murine model.16 Thus, development of therapeutic approaches that eliminate hypoxic stress may serve to block heterotopic bone formation in general, whether initiated by BMP ligand or by the aberrant ACVR1 R206H receptor hypothesized to activate under hypoxic conditions.

References

1. Ahn J. Serrano de la Pena L, Shore EM, Kaplan FS. Paresis of a bone morphogenetic protein-antagonist response in a genetic disorder of heterotopic skeletogenesis. J Bone Joint Surg Am. 2003;85:667-674.
2. de la Pena LS, Billings PC, Fiori JL, Ahn J, Kaplan FS, Shore EM. Fibrodysplasia ossificans progressiva (FOP), a disorder of ectopic osteogenesis, misregulates cell surface expression and trafficking of BMPRIA. J Bone Miner Res. 2005;20:1168-1176.
3. Fiori JL, Billings PC, de la Pena LS, Kaplan FS, Shore EM. Dysregulation of the BMP-p38 MAPK signaling pathway in cells from patients with fibrodysplasia ossificans progressiva (FOP). J Bone Miner Res. 2006;21:902-909.
4. Gellibert F, de Gouville AC, Woolven J, Mathews N, Nguyen VL, Bertho-Ruault C, Patikis A, Grygielko ET, Laping NJ, Huet S. Discovery of 4-{4-[3-(pyridin-2-yl)-1H-pyrazol-4-yl]pyridin-2-yl}-N-(tetrahydro-2H- pyran-4-yl)benzamide (GW788388): a potent, selective, and orally active transforming growth factor-β type I receptor inhibitor. J Med Chem. 2006;49:2210-2221.
5. Glaser DL, Kaplan FS. Treatment considerations for the management of Fibrodysplasia Ossificans Progressiva. Clin Rev Bone Miner Metab. 2005;3:243-250.
6. Huse M, Chen YG, Massagué J, Kuriyan J. Crystal structure of the cytoplasmic domain of the type I TGF-β receptor in complex with FKBP12. Cell. 1999;96:425-436.
7. Huse M, Muir TW, Xu L, Chen YG, Kuriyan J, Massagué J. The TGF- β receptor activation process: an inhibitor- to substrate-binding switch. Mol Cell. 2001;8:671-682.
8. Kaplan FS, Glaser DL. Thoracic insufficiency syndrome in patients with Fibrodysplasia Ossificans Progressiva. Clin Rev Bone Miner Metab. 2005;3:213-216.
9. Kaplan FS, Glaser DL, Shore EM, Deirmengian GK, Gupta R, Delai P, Morhart R, Smith R, Le Merrer M, Rogers JG, Connor JM, Kitterman JA. The phenotype of Fibrodysplasia Ossificans Progressiva. Clin Rev Bone Miner Metab. 2005;3:183-188.
10. Kitterman JA, Kantanie S, Rocke DM, Kaplan FS. Iatrogenic harm caused by diagnostic errors in Fibrodysplasia Ossificans Progressiva. Pediatrics. 2005;116:654-661.
11. Lehmann K, Seemann P, Stricker S, Sammar M, Meyer B, Suring K, Majewski F, Tinschert S, Grzeschik KH, Muller D, Knaus P, Nurnberg P, Mundlos S. Mutations in bone morphogenetic protein receptor 1B cause brachydactyly type A2. Proc Natl Acad Sci USA. 2003;100:12277-12282.
12. Perutz MF, Muirhead H, Mazzarella L, Crowther RA, Greer J, Kilmartin JV. Identification of residues responsible for the alkaline Bohr effect in haemoglobin. Nature. 1969;222:1240-1243.
13. Pignolo RJ, Suda RK, Kaplan FS. The Fibrodysplasia Ossificans Progressiva Lesion. Clin Rev Bone Miner Metab. 2005;3:195-200.
14. Rötzschke O, Lau JM, Hofstatter M, Falk K, Strominger JL. A pH-sensitive histidine residue as control element for ligand release from HLA-DR molecules. Proc Natl Acad Sci USA. 2002;99:16946-16950.
15. Sawyer JS, Anderson BD, Beight DW, Campbell RM, Jones ML, Herron DK, Lampe JW, McCowan JR, McMillen WT, Mort N, Parsons S, Smith EC, Vieth M, Weir LC, Yan L, Zhang F, Yingling JM. Synthesis and activity of new aryl- and heteroaryl-substituted pyrazole inhibitors of the transforming growth factor-β type I receptor kinase domain. J Med Chem. 2003;46:3953-3956.
16. Sawyer JS, Beight DW, Britt KS, Anderson BD, Campbell RM, Goodson T Jr, Herron DK, Li HY, McMillen WT, Mort N, Parsons S, Smith EC, Wagner JR, Yan L, Zhang F, Yingling JM. Synthesis and activity of new aryl- and heteroaryl-substituted 5,6-dihydro-4H-pyrrolo[1,2-b]pyrazole inhibitors of the transforming growth factor-β type I receptor kinase domain. Bioorg Med Chem Lett. 2004;14:3581-3584.
17. Shafritz AB, Shore EM, Gannon FH, Zasloff MA, Taub R, Muenke M, Kaplan FS. Overexpression of an osteogenic morphogen in Fibrodysplasia Ossificans Progressiva. N Engl J Med. 1996;335:555-561.
18. Shore EM, Feldman GJ, Xu M, Kaplan FS. The genetics of Fibrodysplasia Ossificans Progressiva. Clin Rev Bone Miner Metab. 2005;3:201-204.
19. Shore EM, Xu M, Feldman GJ, Fenstermacher DA, Brown MA, Kaplan FS. A recurrent mutation in the BMP type I receptor ACVR1 causes inherited and sporadic fibrodysplasia ossificans progressiva. Nat Genet. 2006;38:525-527.
20. Singh J, Chuaqui CE, Boriack-Sjodin PA, Lee WC, Pontz T, Corbley MJ, Cheung HK, Arduini RM, Mead JN, Newman MN, Papadatos JL, Bowes S, Josiah S, Ling LE. Successful shape-based virtual screening: the discovery of a potent inhibitor of the type I TGFβ receptor kinase (TβRI). Bioorg Med Chem Lett. 2003;13:4355-4359.
21. Weiser R, Wrana JL, Massagué J. GS domain mutations that constitutively activate TβR-I, the downstream signaling component in the TGF-β receptor complex. EMBO J. 1995;14:2199-2208.
22. Yi SE, Daluiski A, Pederson R, Rosen V, Lyons KM. The type I BMP receptor BMPRIB is required for chondrogenesis in the mouse limb. Development. 2000;127:621-630.
© 2007 Lippincott Williams & Wilkins, Inc.