|Home | About | Journals | Submit | Contact Us | Français|
The FXYD family proteins are auxiliary subunits of the Na,K-ATPase, expressed primarily in tissues that specialize in fluid or solute transport, or that are electrically excitable. These proteins range in size from about 60 to 160 amino acid residues, and share a core homology of 35 amino acid residues in and around a single transmembrane segment. Despite their relatively small sizes, they are all encoded by genes with six to nine small exons. We show that the helical secondary structures of three FXYD family members, FXYD1, FXYD3, and FXYD4, determined in micelles by NMR spectroscopy, reflect the structures of their corresponding genes. The coincidence of helical regions, and connecting segments, with the positions of intron–exon junctions in the genes, support the hypothesis that the FXYD proteins may have been assembled from discrete structural modules through exon shuffling.
The FXYD family proteins are tissue-specific and physiological-state-specific auxiliary subunits of the Na, K-ATPase, the primary enzyme responsible for maintaining the equilibrium of Na and K ion concentrations across animal cell membranes.1-3 All of the FXYD genes are expressed in the early stages of fetal life, with prevalence in tissues that specialize in fluid or solute transport, or that are electrically excitable. In keeping with their distribution in tissues, several FXYD family members have been shown to regulate ion transport by binding and modulating the activity of Na, K-ATPase molecules.2-11 In addition, certain FXYD family members can induce ionic currents in Xenopus oocytes, or in phospholipid bilayers, although the direct formation of ion channels has not been demonstrated in vivo.12-16
Seven FXYD family members have been identified in mammals. FXYD1 (PLM; phospholemman) is the principal substrate of hormone-stimulated phosphorylation by cAMP-dependent protein kinase A and C in sarcolemma heart.17 FXYD2 (gamma), unique in the family for having two alternative splice variants (FXYD2a and FXYD2b), and FXYD4 (CHIF; channel-inducing factor; corticosteroid hormone induced factor), are each expressed in distinct, specialized segments of the kidney, with unique expression patterns that help explain the physiological differences in Na,K-ATPase activity among the nephron segments.5,6,13,18,19 Two family members, FXYD3 (Mat8; mammary tumor protein 8 kDa) and FXYD5 (RIC; resembles ion channel; dyshaderin), are expressed in cancers and play a role in tumor progression.20-23 FXYD6 and FXYD7 are expressed exclusively in the brain, and are believed to play a role in neuron excitability during postnatal development and in the adult brain.8,24
Membrane proteins belonging to what is now recognized as the FXYD family were first discovered, in association with the Na,K-ATPase,4 or as a major substrate of hormone-stimulated phosphorylation in sarcolemma.25 The subsequent sequencing and identification of additional proteins, with similar structural and functional features, pointed to the existence of a family of homologous membrane proteins involved in ion transport regulation.5,13,17,20,21 The complete gene family was established recently by expressed sequence tags (EST) database analysis, in a detailed genetic study that has provided complete cDNA sequences, protein signature sequences, and expression patterns for the genes.1 A surprising finding of this study was that the proteins, despite their relatively small sizes, ranging from about 60 to 160 amino acid residues, are all encoded by genes with six to nine small exons. In contrast, phospho-lamban, another small, single-spanning, membrane protein that regulates the Ca-ATPase of cardiac tissue, and has no sequence homology to the FXYD proteins, is encoded by a single exon.26 Here, we report that the secondary structures of the FXYD proteins reflect the structures of their corresponding genes, suggesting that they might have been assembled from modules through exon shuffling.
The FXYD family members share a core homology of 35 invariant and conserved amino acid residues, in and around a single transmembrane segment (Figure 1). The short signature motif PFXYD (Pro, Phe, X, Tyr, Asp), from which the family takes its name, is invariant in all seven known mammalian examples, and identical in those of other vertebrates, except for Pro, and residue X is typically Tyr, but can also be substituted by Thr, Glu, or His. In all the proteins, conserved basic residues flank the transmembrane domain, the extracellular N termini are acidic, and the cytoplasmic C termini are basic; however, outside the signature homology motif there is little sequence conservation among the family members. The presence of two conserved Gly residues in the transmembrane domain of all FXYD proteins (G30 and G41 in FXYD2a), suggests that it is involved in specific intramembrane helix–helix interactions. Indeed, in a family with primary hypomagnesemia, a disease characterized by renal or intestinal Mg loss, the dominant-negative mutation, G41R, in FXYD2a causes misrouting of the protein by inhibiting its association with the Na,K-ATPase, suggesting that G41 is important for mediating helix–helix interactions between the FXYD transmembrane helix and the Na,K-ATPase.27
The first step in our structural studies of the FXYD family proteins was to find a detergent that would be suitable for high-resolution NMR spectroscopy. We examined the 1H/15N heteronuclear single quantum correlation (HSQC) spectra of the proteins in the best-characterized micelle-forming detergents (DHPC, di-heptanoyl-phosphocholine; DPC, dodecylphosphocholine; LPPG, lyso-palmitoylphosphoglycerol; OG, octyl-glucopyranoside; SDS, sodium dodecylsulfate), at various conditions (protein, detergent and salt concentration; pH; temperature). The highest quality spectra were obtained for the proteins in SDS micelles, at 40°C (Figure 2). Although SDS is widely assumed to be a universal protein-denaturing detergent because of its common use in protein electrophoresis, many hydrophobic membrane proteins actually retain their structures in SDS micelles.28 Furthermore, since all functional studies of Na,K-ATPase have been done on enzymes purified in the presence of SDS, and the non-covalent associations of α, β and FXYD subunits are maintained through the SDS purification process,29-31 we reasoned that this detergent would also be a good choice for FXYD structural studies.
Each of the three FXYD proteins that we studied adopts a unique structure in SDS, reflected in the HSQC spectra shown in Figure 2, and the 1.5 ppm dispersion of the amide 1H chemical shift frequency is typical of native helical membrane proteins in micelles. Furthermore, the HSQC spectra obtained in 2H2O, revealed a core region in each protein, with amide protons that exchange very slowly with the surrounding aqueous solvent (Figure 3(b)). These regions, spanning V22 to S37 for FXYD1, I23 to S38 for FXYD3, and I23 to L37 for FXYD4, match the hydrophobic transmembrane segments of the proteins identified by the hydropathy plots (Figure 3(a)), and reflect the strong intra-molecular hydrogen bonds of transmembrane helices in the low dielectric environment of the membrane or, in this case, the micelle interior.
To determine the secondary structures of the three proteins, we relied primarily on 1H–15N residual dipolar couplings (RDCs), measured from weakly aligned samples and analyzed in terms of dipolar waves.32 These were supplemented with chemical shifts, analyzed in terms of chemical shift indices (CSI),33 and with TALOS.34 The RDCs measured in solution NMR experiments provide direct orientation restraints relative to an intra-molecular reference frame,35-37 and are very useful for evaluating helical residues, helix regularity, and the relative orientations of helical segments.38,39 This is particularly useful for membrane proteins where structure determination is usually hindered by the shortage of reliable long-range nuclear Overhauser effects (NOEs) as distance restraints.
In the transmembrane helix of each of the three FXYD proteins, the fits of experimental RDCs to a sinusoid with the signature α-helical periodicity of 3.6 residues demonstrate that the transmembrane helices are close to ideal (Figure 3(c)). In each protein, the boundaries of the transmembrane helix were identified from the changes in sinusoid periodicity, and correlate well with 13Cα CSI profiles (Figure 3(d)), as well as with dihedral angles obtained from TALOS analysis. In all three proteins, the transmembrane helix is preceded by a short helical stretch of approximately five residues, beginning at the aspartate residue of the FXYD signature sequence. This helix-break-helix motif spans exactly the entire length of exon 4 in FXYD1, exon 6 in FXYD3, and exon 5 in FXYD4, and is followed by a third short helix, spanning the stop-transfer signal sequence of each protein, and encoded by the next exon (exon 5 in FXYD1; exon 7 in FXYD3; exon 6 in FXYD4). In all the three proteins, these three helices, plus the preceding PFXYD signature motif, constitute a common core module, which spans the entire 35 residue FXYD protein homology region, over three exons.
The heteronuclear 1H–15N NOE measurements are excellent indicators of local backbone motions in proteins, and have been used to characterize the dynamics of helical membrane proteins in micelles (reviewed by Opella & Marassi40). Throughout the core helical modules of FXYD1, FXYD3, and FXYD4, all residues have similar positive values of the 1H–15N NOE, reflecting similar values of the rotational correlation time, and indicating that the three helices are rigidly connected (Figure 3(e)). Lower negative values of the heteronuclear NOE, reflecting additional backbone motions, are present in residues near the N and C termini, and at the boundaries of the core modules, marked by the junction between exons 5 and 6 in FXYD1, 7 and 8 in FXYD3, and 6 and 7 in FXYD4.
Resonance line-widths also reflect local backbone dynamics, and correlate with resonance intensities, which can be easily measured as peak heights in the HSQC spectra of proteins. Large-amplitude backbone motions that are rapid compared to the overall reorientation rate of the protein reduce the line-widths, and increase the resonance intensities. All of the resonances from amino acids within the core helical modules of the three proteins have similar peak intensities that plateau at minimum values (Figure 3(f)), indicating that the three helices are rigidly connected. In contrast, residues at the core module boundaries, and at the terminal regions of the proteins have greater intensities reflecting increased dynamics in these regions.
Beyond the core helical module, FXYD1 has a fourth helix, spanning from F60 to S68 near the C terminus, which is well defined by the RDCs, as well as by CSI and TALOS analyses of the chemical shifts. This helix contains the S63 and S68 consensus sites for phosphorylation by PKA and PKC, and it is separated from the core module by a flexible linker, with significantly smaller values of the 1H–15N NOE, and significantly greater peak intensities. The position of the flexible linker coincides with the junction between exons 5 and 6, and the fourth helix ends at the end of exon 6.
The addition of MnCl2 to FXYD1 in micelles caused significant line broadening and disappearance of the 1H/15N HSQC peaks from amino acids in helix 1, in the flexible connecting segment between the helices 3 and 4, and in the N and C-terminal regions of FXYD1. In contrast, the peaks from amino acids in helix 2 (the transmembrane helix), in helix 3, and in helix 4, were largely unaffected. The paramagnetic electrons in the Mn2+ induce distance-dependent broadening of peaks from protein sites that are solvent-exposed, while residues in the hydrophobic interior of the micelle, as those in the transmembrane helix, are mainly unaffected. This result suggests that helices 3 and 4 are tightly associated with the micelle, and is consistent with the 15N chemical shift solid-state NMR spectra of FXYD1 in lipid bilayers, which indicate the presence of a helical segment associated with the membrane surface.41
FXYD3 also appears to form a fourth helix at the start of exon 8, although it is very short and not as well defined by the smaller values of the RDCs. A short and somewhat more flexible stretch of amino acids is evident from the small, but significant, increase in peak intensities at the junction between exons 7 and 8 connecting helix 3 to helix 4. FXYD4, on the other hand, has a very well defined fourth helix, which begins at the start of exon 7 and spans its entire length, before reaching the mobile C terminus, which coincides with exon 8.
For FXYD1, FXYD3, and FXYD4, solid-state NMR studies of uniformly and selectively 15N-labeled proteins in lipid bilayers, indicate the presence of transmembrane and C-terminal helical segments similar to those found in micelles.41 For FXYD4, the emerging data indicate that the transmembrane helix exhibits the helix-break-helix motif that is observed in micelles, hinting that the structures in bilayers are similar to those in SDS micelles. This has been observed for other membrane polypeptides studied by both methods.40
Many proteins are composed of discrete domains that are characterized by their ability to fold independently and to confer specific biological functions. These modular proteins are often generated by exon shuffling, where copies of individual exons are recruited from different unrelated genes, to create new combinations of exons.42 Because the modules can be easily moved around, they have played a dominant role in protein evolution. Each module tends to be encoded by one, or a combination of exons, that begin and end in the same splice frame, and the module organization in the proteins is strongly correlated with the intron–exon structure of their corresponding genes. According to the rules of exon shuffling, introns may interrupt the reading frame of a gene between two consecutive codons (phase 0 introns), between the first and second nucleotide of a codon (phase 1 introns), or between the second and the third nucleotide (phase 2 introns). Exons are then classified into symmetric (0-0; 1-1; 2-2) or asymmetric (0-1; 0-2; 1-0; 1-2; 2-0; 2-1) classes, depending on the phase of the flanking introns.43,44 Symmetric exons, or exon sets, are the only ones that can be inserted into introns of the same phase, undergo tandem duplication, or be deleted without disturbing the reading frame.
To examine the relationship between gene structure and protein structure in the FXYD family, we mapped the splice frame diagrams of the seven mammalian genes onto the organization of the structured modules determined by NMR, and of the protein homology domains (Figure 4). The gene data, which had been determined by EST analysis and experimentally,1,45-47 were obtained from the NCBI databank (National Center for Biotechnology Information†). The majority of mobile modules are encoded by symmetric exons, or exon sets, of class 1-1, while only a few class 0-0 and class 2-2 domains are known.43,44 Similarly, in the FXYD family proteins, many of the intron–exon junctions are conserved across the family, and the majority of introns (64%) have phase 1, and separate symmetric exons, or exon sets, of class 1-1.
In all the mammalian FXYD proteins, the FXYD signature motif (Figure 4, red boxes) is encoded by a symmetric exon of class 1-1. It is followed by the transmembrane domain (Figure 4, green boxes), also encoded by a class 1-1 exon, which forms a helix-break-helix structure that is conserved in FXYD1, FXYD3, and FXYD4, with helix breaks that exactly match the intron–exon junctions in these three family members. The stop-transfer signal (Figure 4, white boxes) is encoded by an asymmetric class 1-2 exon, which is interrupted by a phase 2 intron, and in FXYD1, FXYD3, and FXYD4, it forms a short helix followed by a more mobile linker. In these three proteins, the neighboring exon belongs to class 2-1, and encodes a fourth helix. Taken together, this asymmetric pair begins and ends in the same splice frame as the preceding sequence, and forms a helix-link-helix structure that is a good candidate for a mobile module.
The large number of exons in the genes of the relatively small FXYD proteins is intriguing and has been suggested to reflect modular gene assembly.1,45 The presence of conserved modules in different FXYD family members suggests that they are derived from a common ancestor gene, from which FXYD5 appears to have been the first to diverge.3 The coincidence of intron–exon junctions with helical structures and flexible connecting segments, in the three family members, FXYD1, FXYD3, and FXYD4, further supports this hypothesis, and indicates that the proteins in this family may have been assembled from discrete structural modules through exon shuffling. The multiple exon organization of the FXYD genes could serve to confer high levels of structural and functional diversity among the family members.
This research was supported by a grant from the National Institutes of Health (R01 CA082864). The NMR studies utilized the Burnham Institute NMR Facility, supported by a grant from the National Institutes of Health (P30 CA030199).