Ca2+-binding Motif of βγ-Crystallins*

βγ-Crystallin-type double clamp (N/D)(N/D)XX(S/T)S motif is an established but sparsely investigated motif for Ca2+ binding. A βγ-crystallin domain is formed of two Greek key motifs, accommodating two Ca2+-binding sites. βγ-Crystallins make a separate class of Ca2+-binding proteins (CaBP), apparently a major group of CaBP in bacteria. Paralleling the diversity in βγ-crystallin domains, these motifs also show great diversity, both in structure and in function. Although the expression of some of them has been associated with stress, virulence, and adhesion, the functional implications of Ca2+ binding to βγ-crystallins in mediating biological processes are yet to be elucidated.


␤␥-Crystallin-type double clamp (N/D)(N/D)XX(S/T)S motif is an established but sparsely investigated motif for Ca 2؉ binding. A ␤␥-crystallin domain is formed of two
Greek key motifs, accommodating two Ca 2؉ -binding sites. ␤␥-Crystallins make a separate class of Ca 2؉ -binding proteins (CaBP), apparently a major group of CaBP in bacteria. Paralleling the diversity in ␤␥-crystallin domains, these motifs also show great diversity, both in structure and in function. Although the expression of some of them has been associated with stress, virulence, and adhesion, the functional implications of Ca 2؉ binding to ␤␥-crystallins in mediating biological processes are yet to be elucidated.
The Ca 2ϩ -binding motif of the ␤␥-crystallin type is a recently established motif with a wide prevalence. Here, we lay out our understanding of this motif with respect to the geometry of binding sites, modes of Ca 2ϩ coordination, and prediction of functional, disabled, or degenerate (nonfunctional) motifs. The well characterized founding members of the ␤␥-crystallin superfamily are lens ␤and ␥-crystallins, which are major constituents of the vertebrate eye lens, rendering it with a high refractive index and transparency (for reviews, see Refs. [13][14][15][16][17][18]. As in the case of some other Ca 2ϩ -binding proteins (C2 domains, EGF domains, and cadherins), ␤␥-crystallins possess an all-␤ fold, made of strand exchanged Greek key motifs (19,20).
The ␤␥-crystallin domain is an ancient protein fold, and several proteins across different domains of life are found to have this fold, a majority of them being expressed in bacterial species (15,17,21). Many ␤␥-crystallin domains have been studied in the recent past, and insights obtained from ion binding in these ␤␥-crystallins have led to the proposition that proteins of this superfamily possess a universal Ca 2ϩ -binding motif (22,23). Notwithstanding that ␤␥-crystallins have been established as a superfamily of Ca 2ϩ -binding proteins, their involvement in the cellular Ca 2ϩ metabolism or signaling is far from being recognized yet. We summarize a critical analysis of the Ca 2ϩ -binding motif present in this superfamily of proteins.

␤␥-Crystallins and Ca 2؉ Binding: A Chronology
Although the very first non-lens protein belonging to the ␤␥-crystallin superfamily was Protein S from the bacterium Myxococcus xanthus (24), it took almost two decades to classify these proteins as a separate group of CaBP. In the absence of structural information, Protein S was earlier speculated to possess extensively modified EF-hand motifs (25). Weak binding of Ca 2ϩ to lens ␤-crystallins was reported on the basis of equilibrium dialysis and interaction with the Ca 2ϩ mimic dye Stainsall (26,27). Another decade passed before Spherulin 3a, a protein from slime mold Physarum polycephalum that was predicted to be a single domain ␤␥-crystallin (28), was finally shown to bind Ca 2ϩ , with clues from Ca 2ϩ binding to lens ␤-crystallin (29 -31). Based on Ca 2ϩ binding to a peptide corresponding to a Greek key motif of ␥-crystallin (32), a proposition that ␤␥-crystallin-type Greek key could form the motif for Ca 2ϩ binding was formulated. Concurrent structural studies shed light on the coordination pattern of Ca 2ϩ binding to Spherulin 3a and Protein S (22,33).
The recognition of ␤␥-crystallins as a distinct superfamily of Ca 2ϩ -binding proteins was still in its infancy. The factors responsible for this were: (i) undefined motif of Ca 2ϩ binding, (ii) lack of information about the role of Ca 2ϩ in protein functions, and (iii) no substantial addition of novel members to this superfamily. For a long time, this superfamily was considered a sparsely distributed family with only a few scattered members, although some more proteins (e.g. WmKT, SKLP, and SMPI) were identified based on structural similarity (34 -36); these did not belong to the ␤␥-crystallins lineage and may have arisen from convergent evolution (37). With the advent of genomic sequence information, many members from diverse species (bacteria, archaea, and urochordate) were added to the superfamily and also confirmed to be Ca 2ϩ -binding proteins (22, 23, 38 -43), leading to the recognition of a common motif for ion binding, and thus prompting the organization of these proteins as a separate set of Ca 2ϩ -binding proteins (23,44). based on the pattern and supersecondary features observed in proteins, is a basic theme of many all-␤ proteins (45,46). Greek keys are quite diverse in topology and hence were further classified as (4,0), (3,1), and (2,2) patterns, based on the arrangement and connectivity of strands (46). According to the above classification, proteins of the ␤␥-crystallin superfamily fall into the (3,1) category (47), where one strand (third strand, c and cЈ of the respective Greek key motif) out of the four is shared by a partner motif (Fig. 1).

␤␥-Crystallin-type
Greek key motif has a distinct signature sequence of residues "(F/Y/W)XXXX(F/Y)XG" (28) in the ␤-hairpin loop between the first and the second strands (19) (Fig. 1). Along with this signature sequence, ϳ24 -30 residues downstream, a conserved Ser is located (on the fourth ␤-strand), which plays a structural role by stabilizing this ␤-hairpin (19,20,48). The third strand of each Greek key is swapped and becomes a part of the partner Greek key motif ((3,1) arrangement of ␤-strands) (Fig. 1b). The fourth strand folds back to the previous sheet via a connecting loop, which is variable in length and occasionally has a small helical segment. This loop (named loop1 in the first Greek key and loop2 in the second Greek key motif) occupies the top of the ␤␥-crystallin domain (Fig. 1c).
A ␤␥ domain has two juxtaposed Ca 2ϩ binding sites, which are mainly formed of the loops (loop1 and loop2) with an (N/D)(N/D)XX(T/S)S sequence stretch, running in opposite directions in three-dimensional space, along with the residues from ␤-hairpins (22, 23, 33, 40) (Fig. 1, c and d). Thus, each Ca 2ϩ -binding site is formed by four residues located at three different regions in the primary sequence. The first Ca 2ϩ -binding site is formed via one residue from ␤-hairpin1, two residues from loop1, and one residue from loop2, and in a similar way, the second site is formed by one residue from ␤-hairpin2, two residues from loop2, and one residue from loop1 (22,23) (Fig.  1d). This arrangement is common in the ␤␥ domains studied structurally: Protein S (33), Spherulin 3a (22), Ci-␤␥-crystallin (40), Clostrillin, Flavollin, M-crystallin (23,41), and Geodin (42).

Coordination Geometry of the Ca 2؉ -binding Site
In ␤␥-crystallins, the Ca 2ϩ coordination number varies from five to eight, four being provided by protein ligands, the rest being satisfied by water molecules (22,23,33,40,42). The ϩx position of the coordination sphere is provided by the main chain carbonyl of the first X residue of (F/Y)XXXX(Y/F)XG stretch, which is next to the first conserved aromatic residue of ␤-hairpin1 (Fig. 2). The second coordination (ϩy position) is provided by the main chain carbonyl oxygen of the third residue and the third coordination (ϩz position) by the side chain oxygen of the fifth residue of the (N/D)(N/D)XX(T/S)S stretch of the same Greek key motif (22,23,40) (Fig. 2). The involvement of the Ser or Thr hydroxyl group to coordinate Ca 2ϩ is distinctive to this motif as most Ca 2ϩ -binding sites do not involve these residues in Ca 2ϩ ion coordination (49). The fourth coordination (Ϫx position) is provided by the side chain oxygen of the second residue (mostly Asp or sometime Asn) of Both sites of a domain exhibit similar or slightly altered coordination geometry. The coordination number is seven, with pentagonal bipyramidal geometry, but octahedral geometry (coordination number: 6) and square anti-prismatic (coordination number: 8) are also seen. A coordination number of five has also been observed in the first Ca 2ϩ binding site of Protein S (33). The average coordination radius at each of these sites varies from 2.4 to 2.8 Å.

Preferred Residues in the Binding Motif and Their Significance
An analysis of Ͼ100 sequences of canonical motifs available demonstrated that the amino acid residues forming the (N/D)(N/D)XX(S/T)S fingerprint vary in various proteins of the superfamily (44). The first residue of the fingerprint is involved in stabilizing the pocket through hydrogen bonding with the hydroxyl side chain of Ca 2ϩ coordinating Ser/Thr and in some cases supports a water molecule present at the Ϫy position. The second residue, also a polar amino acid (mostly Asp), directly coordinates Ca 2ϩ by providing monodentate ligation at the Ϫx position. These first two residues act in trans, i.e. the first two residues of loop1 of the first Greek key, become part of the second Ca 2ϩ -binding site and vice versa; thus, the motif gets its name as a "double clamp" motif (23) (Figs. 1 and 2). The third residue (X1) coordinates via main chain carbonyl oxygen and is generally a polar residue, but is variable in nature, thus providing variations in the microenvironment around the Ca 2ϩ -binding site. The nonligating fourth residue (X2) is generally hydrophobic. It forms a part of the hydrophobic core and may be an important player in relaying signals to the core; local conformational changes in the loop upon Ca 2ϩ binding can thus affect the stability of the protein (23,44). The fifth residue is usually Ser (or sometimes Thr), which coordinates via the hydroxyl oxygen and is a determinant for domain properties. The sixth nonligating residue is highly conserved structural Ser, which stabilizes the ␤-hairpin loop (48). The crystal structures of ␤␥-crystallin domains available in the Ca 2ϩbound form display a common theme of Ca 2ϩ coordination that led to the proposition of a distinct motif for Ca 2ϩ binding in the ␤␥-crystallin superfamily (Table 1), although some of these structures carry minor variations in this theme, which are discussed below.

Deviations within the Common Theme of the Binding Sites
The second Ca 2ϩ -binding site of Ci-␤␥-crystallin (Protein Data Bank (PDB) ID: 2BV2) from Ciona intestinalis is a minor exception to the theme described above (40). In this site, the protein molecule provides five coordinations, where the third residue (Glu-76) of the (N/D)(N/D)XX(T/S)S fingerprint coordinates through the side chain oxygen at the Ϫz position, in addition to the usual coordination by the main chain carbonyl oxygen at ϩy (Fig. 3a). Similarly, in the W39D mutant of Clostrillin, Asp-39 coordinates via the main chain carbonyl oxygen in addition to the side chain carboxyl oxygen, whereas in the wild type, only the main chain carbonyl of Trp-39 coordinates at the ϩy position (44). The pentagonal bipyramidal geometry   with coordination number 7 is more prevalent in the ␤␥-crystallin superfamily, where three positions corresponding to Ϫy1, Ϫy2, and Ϫz positions of the coordination sphere are occupied by water molecules (Fig. 2). However, the water molecules coordinating with Ca 2ϩ vary in number, leading to different coordination geometries as well. In Geodin (from a marine sponge), only one Ca 2ϩ bound to a C-terminal domain is observed as only one site is canonical (42), attesting to the prediction made by Clout et al. (22) applying fingerprint comparison. In selenomethionine derivatized crystal, the water molecules corresponding to Ϫy1 and Ϫy2 positions, as seen in underivatized crystals, have been replaced by hydroxyl oxygens of glycerol molecule without affecting the overall coordination geometry (Fig. 3b). Ci-␤␥-crystallin is another example, where a water molecule at the first Ca 2ϩ -binding site (Ϫz position) is replaced by the side chain carboxyl oxygen of Glu-7 from a symmetry-related molecule as an effect of crystal packing (Fig.  3c) (40). These examples suggest that the water molecules participating in Ca 2ϩ coordination are not very strongly bound in ␤␥-crystallins and can be replaced by other suitable ligands, if conditions are favorable for their binding.

Signature Sequence and Identification of New Ca 2؉ -binding ␤␥-Crystallins
With the growing number of protein sequences in databases, the (N/D)(N/D)XX(S/T)S sequence fingerprint can be used to predict Ca 2ϩ -binding ␤␥-crystallins as with the consensus sequence of the EF-hand motif. The (N/D)(N/D)XX(S/T)S fingerprint occurs twice in a domain due to the pairing of the two Greek key motifs. As two Greek key motifs participate in constructing a site, both should be taken into consideration for prediction (Fig. 3d). Only two residues use their side chains in coordinating Ca 2ϩ , and any change at these positions leads to either complete loss or reduction in Ca 2ϩ binding ability. The first critical residue is Ser/Thr at the fifth position, and the second is Asp or Asn from the juxtaposed motif at the second position in the (N/D)(N/D)XX(S/T)S fingerprint (Fig. 3d). As an example, in Clostrillin, at the first Ca 2ϩ -binding site, Thr from the motif NDWMTS ligates via the side chain hydroxyl oxygen at the ϩz position and Asp residue from the second motif NDKMTS coordinates via the side chain carboxyl oxygen at the Ϫx position (Table 1). Similarly, for the second site, Thr from the motif NDKMTS would provide its side chain for coordination at the ϩz position, and the site would be completed by the side chain of the Asp residue coming from NDWMTS (23). Thus, both sites are well matched for binding Ca 2ϩ (Table 1).
In the case of the first site of the C-terminal domain of Geodin, Lys from the first ␤-hairpin loop and Gly from the motif IGGVSS would coordinate via main chain, besides Ser at the fifth position in the motif (Table 1) (42). This site would bind Ca 2ϩ only if the second residue of the second motif NDALKS is also favorable for ligation. It is Asp, and hence the first site in this protein is suitable for Ca 2ϩ binding. At the second site, Lys occupies the fifth position in the motif NDALKS, which is not compatible. Also, the second residue of the partner motif IGGVSS is Gly, which, again, cannot provide its side chain for completing the site, and thus the second site of this domain would not bind Ca 2ϩ . This consensus sequence has aided into the identification of new members of this family.

Degeneracy of Motif: Constraints and Gains
Not all ␤␥-crystallins possess the canonical (N/D)(N/ D)XX(S/T)S fingerprint of ion binding. Most proteins of this family with functional Ca 2ϩ -binding motifs are from bacterial sources or lower eukaryotes ( Table 1). The vertebrate homologues, such as lens ␤␥-crystallins, AIM1, Crybg3, and a few more diverged domains do not have the canonical sequence, and their binding sites are degenerated to an extent that Ca 2ϩ binding is either very poor or lost (23, 43, 50 -54). It is proposed that although Ca 2ϩ binding might have been a vital need for ancestors of lens crystallins, most binding sites in lens ␤␥-crystallins were evolutionarily disabled to gain extra stability (52,54) because high domain stability is an indispensable requirement of lenticular proteins (15,55). It is, however, speculated that these ␤␥-crystallins may be involved in Ca 2ϩ -dependent functions in non-lenticular tissues as shown in brain and testes (56,57).
The C-terminal domain of Geodin with one naturally disabled Ca 2ϩ -binding site (42) is unusual as most ␤␥ domains have two juxtaposed Ca 2ϩ -binding sites. Nitrollin, a ␤␥-crystallin from the bacterium Nitrosospira multiformis, which is peculiar for its mode of domain swapping, does not bind Ca 2ϩ due to the presence of Arg in place of conserved Ser along with other unfavorable residues (43). Thus, even bacterial homologues can have disabled sites. Substitution of Ser/Thr with Arg leads to the loss of Ca 2ϩ binding in many homologous proteins, such as Protein S, Flavollin, and Clostrillin (24,44). This leads to enhanced stability of the protein, which, in some cases, is comparable with stability gained upon Ca 2ϩ binding (52).

Comparisons with Other Ca 2؉ -binding Motifs
Multiple modules, EF-hands, C2 domains, and EGF domains, have purposes such as sensing and sequestration of Ca 2ϩ (5,7,58,59), and Ca 2ϩ binding takes place at a specific motif forming continuous or discontinuous Ca 2ϩ -binding sites (60,61). The EF-hand motif, an example of a continuous site, is the predominant class in CaBP. The EF-hand loop between the E and the F helices coordinates Ca 2ϩ ion, usually with coordination numbers of 7 or 8 (6,(62)(63)(64). Non-EF-hand proteins utilize fewer protein ligands for coordinating Ca 2ϩ (65). Unlike ␤␥-crystallins in which the ion is exposed to the solvent with 1-4 water molecules in direct coordination with Ca 2ϩ , the EF-hand bound Ca 2ϩ is exposed to a single water molecule (Table 2). Similarly, in ␤␥-crystallins, fewer protein ligands (4) and more water (1-4) molecules are seen. The affinities of Ca 2ϩ toward EF-hand proteins are moderate to very high with K d values in the nanomolar range (6), whereas ␤␥-crystallins display affinities in the micromolar range (Table 1).
Unlike the EF-hand motifs, C2 domains are all-␤ conformational Ca 2ϩ sensors and share the ␤-sandwich arrangement not akin to ␤␥-crystallins (Table 2) (7). C2 domain-containing synaptotagmins are major players in neurotransmitter exocytosis and are involved in Ca 2ϩ -dependent phospholipid binding (66). The domains are incorporated in multiple proteins (such as phospholipases) for their Ca 2ϩ sensing roles (67). The Ca 2ϩ -binding site is discontinuous and formed by mutually distant regions in the primary sequence of this domain (68) as is the case with ␤␥-crystallins. In the C2 domains, the Ca 2ϩ ions are coordinated by a network of aspartate side chains in mono-or bidentate fashion with one aspartate involved in coordination with two Ca 2ϩ ions (58, 69 -71).

Minutiae of Ca 2؉ Binding and Origin of Domain Diversity
The affinity of Ca 2ϩ for Ca 2ϩ -binding proteins ranges from intracellular nM to extracellular mM, depending on their spatiotemporal localization, which in turn reflects their functions. With the limited information available, it appears that most ␤␥-crystallins are extracellular or secretory proteins (72)(73)(74)(75)(76)(77)(78).  (Table 1) (23,24,79). Ca 2ϩ binding affinities of Ci-␤␥-crystallin from C. intestinalis, DdCad-1, and Geodin have not been reported. Comparatively weak binding affinity is reported for ␤␥-crystallins from eukaryotic species (9 and 200 M for Spherulin 3a, and 260 M for amphibian EP37) (23,30,31). Although the affinity of these proteins ranges in M, a clear-cut variation has been seen not only in different ␤␥-crystallin domains but also between two juxtaposed sites of the same domain.
The molecular determinants of affinity are not well understood in this domain. Either of the sites in a domain is interdependent as abrogating binding in one site attenuates the affinity of the juxtaposed site (44,80). Nevertheless, it is not clear whether filling of Ca 2ϩ at the two sites is simultaneous or sequential, cooperative or noncooperative. The microenvironment around a Ca 2ϩ -binding site affects its affinity toward Ca 2ϩ . Even a homologous replacement of the fifth residue of the motif between structurally very similar domains changes its affinity as well as domain properties (44,52). Consequently, this motif acts as a tuning knob for such a high diversity in domain properties (52). All the proteins studied so far show little or no binding toward divalent Mg 2ϩ , confirming the Ca 2ϩ specificity of these binding sites. Similar observations for ion selectivity have been made in C2 domains (81), whereas Mg 2ϩ is known to bind functionally, at least in the resting state, to Ca 2ϩ /Mg 2ϩbinding EF-hands (6). Mg 2ϩ may bind even the Ca 2ϩ -specific (regulatory) EF-hands, although with very weak affinity, without affecting the overall conformation of the protein (6).

Prevalence of the ␤␥-Crystallin-type Ca 2؉ -binding Protein Family
The advent of more genomic information has revealed a widespread occurrence of this superfamily and recruitment of ␤␥-crystallin domains in conjunction with domains of different functions. An analysis demonstrates that it exists in several hundred species (source: Pfam database, accession number PF00030). Three out of four protein sequences of ␤␥-crystallins from an invertebrate species Branchiostoma floridae (amphioxus) also possess Ca 2ϩ binding fingerprints probably similar to that seen in Ci-␤␥-crystallin from Ciona intestinalis (21). In a limited analysis of sequences, it was observed that the domain is recruited in serine proteases (Sorangium cellulosum), aspartate metalloproteases (Saccharophagus degradans), carbohydratebinding glycosyl hydrolases (Flavobacterium johnsoniae and other bacterial species), and cell adhesion molecules (Dictyostelium discoideum). Isolated ␤␥-crystallins are also found as part of proteins from Vibrio cholerae, Maricaulis maris, Oceanicaulis alexandrii, and M. xanthus and in the archaeal species Methanosarcina acetivorans. Such extensive recruitment as modules in proteins of diverse functions clearly indicates some significant and widespread roles of these domains in protein function.

Ca 2؉ Binding and Domain Stabilization
CaBP, mostly those of the EF-hand family, undergo large conformational changes upon binding Ca 2ϩ . Contrary to this, ␤␥-crystallins generally do not undergo a drastic change in conformation upon binding Ca 2ϩ , suggesting that apo forms are conformationally not very flexible (e.g. Protein S, Spherulin 3a) (30,82). This is attributed to thermodynamically robust domain architecture (reviewed in Ref. 17). Some individual domains of larger proteins from the pathogenic bacterium Yersinia pestis and the extremophilic Caulobacter crescentus and Hahella chejuensis are intrinsically unstructured (or partly unstructured) in the apo form and gain significant structure upon binding Ca 2ϩ (38,39,83).
Although ␤␥ domains do not undergo major structural change upon binding Ca 2ϩ , they assume a reduced hydrodynamic size and thermodynamically drift to a state of higher structural stabilization. Therefore, in some ␤␥-crystallins, Ca 2ϩ plays the role of an extrinsic stabilizer. Within the superfamily, there exists a stability gradient across the domains and differences in the extent of gain in stability upon Ca 2ϩ binding from very low (in Vibrillin) to very high (in Centillin) (Table 1) (52,84). As noted in Y. pestis, ␤␥ domain-containing proteins are differentially expressed in Ca 2ϩ -depleted avirulent strains (85,86) and are unstructured in apo form (38), raising the prospective of their role in virulence via low calcium response.

Ca 2؉ -dependent Functions: An Underexplored Arena
It is clear that the ␤␥-crystallin domain is widely spread. However, it has been difficult to assign functions in many ␤␥-crystallins studied so far (22,23,38,41,40,83,87,88). Protein S of M. xanthus is expressed as a soluble protein and selfassembles as a multilayer spore coat in a Ca 2ϩ -dependent manner. Based on the observations that adjacent spores were fused to each other via a common Protein S layer, it was proposed that Protein S is involved in spore-spore interaction in the fruiting body (76). Spherulin 3a is another Ca 2ϩ -binding ␤␥-crystallin induced during stressful situations, but just like Protein S, its Ca 2ϩ -dependent functional implications in the physiological context have not yet emerged.
Ci-␤␥-crystallin from C. intestinalis is localized in the otolith and thus might be involved in a primordial sensory system (40). It binds Ca 2ϩ at both binding sites, but the Ca 2ϩ -dependent role of this single domain protein in the urochordate remains an enigma.
␤␥-CAT from the skin of the frog Bombina maxima is implicated in several in vivo toxic effects on mammals (72,89,90). The ␤␥-Crystallin domain of this protein binds Ca 2ϩ (its affinity has not been reported), and when red blood cells were treated with this protein, increased Ca 2ϩ flux was observed that eventually resulted in hemolysis (73).
The cell adhesion molecule DdCad-1 from D. discoideum is involved in cell-cell adhesion in a Ca 2ϩ -dependent manner via dimer formation through two ␤␥-crystallin domains (91). During its transport, the protein is internalized in vacuoles in a Ca 2ϩ -and conformation-dependent manner (92,93). The available functional information, although limited, suggests important roles played by this domain that remain yet to be explored.

Perspective
Although ␤␥-crystallins are distinct in terms of their Ca 2ϩ binding properties with a well defined Ca 2ϩ -binding motif, the functions of many proteins in the protein sequence/structure databases are either uncertain or unknown. The loss of Ca 2ϩ binding ability in recent homologues of ␤␥-crystallins (as in mammals) with persistence of their structural features is a useful paradigm for understanding the evolution of new protein functions. The functional aspect of these domains in isolation and with respect to Ca 2ϩ binding still needs to be worked out in detail. This is particularly relevant in prokaryotes where the study of calcium sensing is still in the primitive stages, and except for a few proteins of the EF-hand superfamily (3,94) and some proteins with orphan motifs for Ca 2ϩ binding, there are no major classes of Ca 2ϩ -binding proteins identified. Recently, all-␤-sheet bacterial immunoglobulin-like (Big) domains were shown to have Ca 2ϩ binding ability; however, their Ca 2ϩ coordination pattern is yet to be deciphered (95,96).
In summary, the Ca 2ϩ binding at the (N/D)(N/D)XX(S/T)S motif may serve as a domain stabilizer, and evolutionary imperatives may have replaced this stabilizing function of Ca 2ϩ with an intrinsic stability of the domain. Although the motif has established its identity, the functional analysis of its members still remains to be addressed, which is a challenging yet imminent task considering the appearance of numerous new members.