J. Dairy Sci. 88:2318-2328
© American Dairy Science Association, 2005.
Contributions of Terminal Peptides to the Associative Behavior of 
s1-Casein*
E. L. Malin,
E. M. Brown,
E. D. Wickham and
H. M. Farrell, Jr.
US Department of Agriculture, Agricultural Research Service, Eastern Regional Research Center, Wyndmoor, PA 19038
Corresponding author: E. L. Malin; e-mail: emalin{at}errc.ars.usda.gov.
 |
ABSTRACT
|
|---|
The N- and C-terminal segments of bovine 
s1-casein-B (f1-23 and f136-196) were characterized under conditions that promoted or inhibited self-association to determine the relative contributions of each fragment to the interaction of s1-casein with itself or with other caseins. In earlier studies of f1-23, nuclear magnetic resonance (NMR) data and circular dichroism (CD) spectra showed that its conformation was thermostable between 10° and 25°C. In contrast, NMR studies of f136-196 indicated temperature sensitivity between 10 and 60°C, as did near-UV and far-UV CD data, suggesting a molten globule-like structure at higher temperatures. To compare the effects of temperature on conformational attributes of s1-casein and its terminal peptides, additional CD studies were conducted over a broader temperature range (10 to 70°C). The far-UV CD spectra indicated little temperature sensitivity for s1-casein, and the N-terminal peptide remained thermostable. During molecular dynamics simulations, the N-terminal peptide conformation did not change significantly, but the conformation of the C-terminal peptide (f136-196) was dramatically altered. These changes are correlated with the thermal instability observed by both CD and NMR in f136-196. Analytical ultracentrifugation studies of the self-association reactions of genetic variants A, B, and C of s1-casein showed that at 37°C the associative state is primarily dimeric; the amounts of higher order polymers significantly decreased when temperature was increased from 20 to 37°C. In all 3 genetic variants, the C-terminal portion of the whole molecule showed thermal instability with respect to aggregation to higher polymers, confirming the predictions of CD data and molecular dynamics simulations. The temperature dependency of these conformational changes suggests a possible function for s1-casein in facilitating casein-casein interactions in casein micelle formation.
Key Words: s1-casein • casein micelle • circular dichroism • molecular modeling
Abbreviation key: CD = circular dichroism, EM = electron microscopy, FTIR = Fourier transform infrared spectroscopy, MGSA = melanoma growth stimulatory activity, µ = ionic strength, NMR = nuclear magnetic resonance, PIPES = piperazine-N,N'-bis(2-ethanesulfonic acid), PPII = polyproline II conformation (a left-handed 310 helix), SAXS = small-angle X-ray scattering.
|
INTRODUCTION
|
|---|
Casein micelles are uniquely designed to provide nourishment for the neonate in several different ways. As carriers of calcium, which is essential for growth and physiological activity, caseins of a putative submicelle contain s2–, s1–, ß-, and 
-caseins in the ratio of 1:4:4:1. For every group of 10 casein chains present in that ratio, there are 65 phosphorylated serine residues with relatively tightly bound calcium, as well as 257 aspartic and glutamic residues that bind calcium more loosely. Moreover, in the association of caseins to form submicelles and micelles, colloidal calcium phosphate acts as a binding medium to maintain micelle integrity (Holt, 1992; Horne, 1998). The casein chains themselves comprise a nutritionally complete protein (Farrell et al., 2004). Finally, the exposure of a unique site on -casein, sensitive to digestive enzymes, ensures coagulation of micelles so that important nutrients of the micelle will remain in the digestive tract long enough for complete breakdown and absorption (Farrell, 1999; Farrell et al., 2003b).
The completeness and efficiency of milk as a food and its evolution as a protein-mineral system that transports additional nutrients are rather well understood, but the process of micelle formation is not well defined. Until recently, only limited speculation was possible, based on ultrastructure research (Farrell, 1999), but sophisticated techniques for protein structure research now make detailed studies a reality.
Caseins are not true globular proteins, having open structures that permit calcium transport and hydration by water molecules (Farrell et al., 2003a). In addition, caseins have specific secondary structural elements including 30% extended (ß-sheet-like) and a similar percentage of turns. The polyproline II conformation (PPII) is recognized as a frequent conformational element in proteins (Adzhubei and Sternberg, 1993), and there is good evidence for the presence of PPII structure in caseins (Farrell et al., 2001, 2002; Malin et al., 2001; Qi et al., 2004). The PPII has been observed in Raman optical activity studies of caseins and amyloid-forming proteins such as synuclein (Smyth et al., 2001; Syme et al., 2002).
Although caseins are not amenable to crystallization, many techniques have been used to gain insight into their structure; as a result, molecular modeling, using sequence-based predictions, has been applied to the structures of all 4 caseins (Kumosinski et al., 1993a, b, 1994a,Kumosinski et al., b; Hoagland et al., 2001). These endeavors were guided by earlier data from nuclear magnetic resonance (NMR) (Huq et al., 1995; Kakalis et al., 1990; Rollema and Brinkhuis, 1989; Tsuda et al., 1991), fluorescence spectroscopy (Clarke and Nakai, 1971), circular dichroism (CD) (Creamer et al., 1981; Chaplin et al., 1988), Fourier transform infrared spectroscopy (FTIR) and Raman spectroscopy (Byler et al., 1988; Curley et al., 1998), electron microscopy (EM) (Kumosinski et al., 1996), and small-angle X-ray scattering (SAXS) (Farrell et al., 1994; Kumosinski et al., 1994c).
Fragments of s1-casein have also been investigated (Malin and Brown, 1995, 1999; Alaimo et al., 1999a,b; Malin et al., 2001). Analytical ultracentrifugation of the N- and C-terminal portions of s1-casein (f1-23 and f136-196, respectively) suggested that f136-196 has a strong propensity for salt-induced self-association (Alaimo et al., 1999a). However, f1-23 was demonstrated to be capable of self-association only at low ionic strength (Malin et al., 2001). There was no evidence for conventional -helix or hydrogen-bonded ß-sheet structures in NMR data for f1-23, and CD spectra showed that the peptide conformation was generally thermostable (Malin et al., 2001). In contrast, a molten globule-like structure at higher temperatures was predicted for f136-196, based on the temperature sensitivity of NMR studies and near-UV and far-UV CD data (Alaimo et al., 1999b).
Molecular dynamics simulations of s1-casein peptides (Malin and Brown, 1995, 1999) indicated wide variations in their ability to adopt backbone conformations that differed from the conformations of the same sequences in the whole protein. Two peptides from the central areas of the protein (f102-142 and f143-164) acquired more secondary structure and a greater degree of folding during molecular dynamics simulations, but small peptides at the N- and C-terminal regions of s1-casein (f1-23 and short sequences between Gln152 and Trp199) resisted major conformational changes.
As noted previously, CD of the C-terminal peptide f136-196 (Alaimo et al., 1999a, b) showed a significant temperature dependence in both the near and far UV ranges, providing strong evidence of a molten globule-like conformation, but molecular dynamics on this fragment are lacking. Therefore, in the work presented here, a further assessment of the f136-196 fragment by molecular dynamics simulations was conducted, and CD studies of whole s1-casein and of the N-terminal peptide were carried out over a broad temperature range to evaluate the thermal stability of their secondary structures. The CD estimates of secondary structure for intact s1-casein are compared with those for f1-23 and f136-196 and with structural assignments for the molecular models. Finally, all of this information is related to potential sites for the self-association of s1-casein and for the binding of calcium and phosphate in casein micelles.
|
MATERIALS AND METHODS
|
|---|
All reagents were of analytical grade or ACS certified from Sigma (St. Louis, MO).
Peptides
Synthetic s1-casein peptide, f1-23, was prepared by the National Food Biotechnology Centre, University College, Cork, Ireland, as described earlier (Malin et al., 2001). Traces of trifluoroacetic acid were removed by treatment with triethylamine and passage through a column of Sephadex G50. Cyanogen bromide cleavage of s1-casein was used to obtain the C-terminal portion, f136-196, resulting in the conversion of methionine 196 to homoserine 196. Cleavage fragments, including f197-199, were separated by passage through a DEAE Sepharose CL-6B column (Alaimo et al., 1999a). The identities of both terminal peptides were confirmed by AA analysis and/or mass spectrometry and sequencing.
Analytical Ultracentrifugation
Analytical ultracentrifugation studies (Beckman Optima XL-A) were described previously for s1-casein fragment f1-23 (Malin et al., 2001) and for f136-196 (Alaimo et al. 1999a). Those studies defined the pH and ionic strengths of buffer systems that promoted or inhibited self-association of those fragments. Centrifugation experiments for the genetic variants of whole s1-casein were conducted at concentrations of 0.7 to 0.8 mg/mL in 25 mM disodium piperazine-N,N'-bis(2-ethanesulfonic acid) (PIPES) (pH = 6.75) with 80 mM KCl, ionic strength (µ) = 0.122 M, at temperatures ranging from 5 to 37°C; the solvent density was adjusted for temperature. The speed was 12,000 rpm, and the partial specific volume was set at that of the B variant, 0.728 cc/g. The data were analyzed using ASSOC 4 (Malin et al., 2001) and statistically enhanced with the Beckman Optima program MULTI.
CD
The CD data in the near-UV and far-UV region were obtained with an AVIV CD spectrometer (Model 60DS; Aviv Associates, Lakewood, NJ), calibrated with d-10-camphorsulfonic acid. Samples were dissolved in buffers at approximately 0.3 mg/mL and filtered through 0.45-µm pore filters into jacketed cylindrical quartz cells. Pathlengths were 1 cm for near-UV and 1 mm for far-UV spectra, respectively. The low-salt buffer was 2 mM dipotassium PIPES-4 mM KCl (pH = 6.75; µ = 0.0074 M), and the high-salt buffer was 66 mM potassium phosphate (pH = 6.75; µ = 0.10 M).
Sample temperatures were maintained by circulating water of specific temperatures through the jacketed cells; successive measurements at increasing temperatures were made after allowing 30 min of equilibration time for a 30°C change in bath temperature. Secondary structures for both peptides and for s1-casein were estimated from the scan data using the CONTIN method of Provencher and Glöckner (1981), which includes polyglutamic acid in the set of proteins on which calculations are based.
Molecular Dynamics
Molecular dynamics simulations of peptides in water were performed using SYBYL (version 6.9; Tripos, St. Louis, MO). The Kollman All-Atom force field and Kollman charges (Weiner et al., 1984, 1986) were used in molecular dynamics simulations and minimizations. Each peptide was excised from the refined structure of s1-casein (Kumosinski et al., 1994a); Protein Data Bank (PDB) files available at www.arserrc.gov/dpp/casein/htm can be viewed with Deep View at www.expasy.org/spdbv (Guex et al., 1997). It should be noted that the molecular model of s1-casein provides a picture of an essentially ab initio structure that could exist, but it was constructed with human intervention using theoretical considerations and does not have the certainty ascribed to structures derived from crystallography.
A blocking group was added to the N- and C-terminal ends, except for Arg1 of f1-23. After a brief initial refinement in vacuo by alternating minimizations and molecular dynamics, each peptide was solvated with a dual layer of water molecules, using the solvent box subroutine, and molecular dynamics simulations were continued through many iterations until energy minima were reached. Periodic boundary conditions for the solvent box were maintained throughout all operations.
The molecular dynamics protocols were chosen to permit changes in each molecule without losing molecular integrity and solvation. Temperatures imposed were limited to 300 K, and simulations were 3 ps in length. These conditions were well below thermal denaturation temperatures (Day et al., 2002) and also prevented the disappearance of water molecules from the solvent box. Each dynamics iteration was followed by an energy minimization. All molecular dynamics simulations and minimizations involved the entire peptide-water assembly, but energies after each operation were calculated for the peptide chain alone to determine the magnitude of change that occurred during simulation. Conformational changes were characterized by the C-C distance between the -carbons of the first and last residue in the peptide and by the appearance or disappearance of defined secondary structure elements, such as -helix, ß-sheet, and turns.
|
RESULTS AND DISCUSSION
|
|---|
CD
The far-UV CD spectra show that s1-casein, the parent molecule, has minimal temperature sensitivity (Figure 1
) between 10 and 70°C, as indicated by the spectra or the calculated percentage of structure (Table 1). Estimates of secondary structure from analysis of the CD data for s1-casein are 13 to 15% -helix; 40 to 46% extended (ß-sheet-like), except at 50°C; and 23 to 25% turns, except at 70°C (Table 1). These estimates differ somewhat from the results of molecular dynamics (Table 2), which indicate smaller percentages of helix and extended conformations and more turns and unspecified structure (Kumosinski et al., 1994a). Overall, the CD estimates reflect an averaging of many kinds of structural features in addition to -helix, extended (ß-sheet-like), and turns, including loops and PPII helix (Malin et al., 2001, Qi et al., 2004); the PPII may be a major component of the unspecified structures of Table 1 (Farrell et al., 2002).
View larger version (20K):
[in this window]
[in a new window]
|
Figure 1. Effect of temperature on the far-UV mean residue ellipticity of s1-casein-B in low ionic strength (µ) buffer (µ = 0.0074 M). 1) 10°C, 2) 27°C, 3) 50°C, and 4) 70°C.
|
|
View this table:
[in this window]
[in a new window]
|
Table 1. Secondary structure predictions from circular dichroism of s1-casein and its N-terminal and C-terminal peptides, f1-23 and f136-196.
|
|
View this table:
[in this window]
[in a new window]
|
Table 2. Effect of molecular dynamics simulations on energies and structural features of s1-casein and its N- and C-terminal peptides.
|
|
In the N-terminal fragment of s1-casein, f1-23, temperature sensitivity was evidenced by less negative minima in the CD spectra at higher temperatures, in both low-salt and high-salt buffers (Figure 2A, B), but no shifts in wavelength were observed, and the apparent sensitivity is not reflected in the calculated percentage of structure (Table 1
). The distinctive spectral changes observed in Figure 2 could arise from increases in peptide mobility as the temperature is increased, rather than any conformational changes. Although the CD data and estimates appear to be insensitive to µ, analytical ultracentrifugation studies previously demonstrated that the peptide dimerizes when µ is low (Malin et al., 2001). Thus, f1-23 was a dimer in low-salt buffer, and the predicted anti-parallel dimerization of s1-casein f1-23 (Malin et al., 2001) has no effect on its secondary structure. The percentage of extended structure estimated by CD is high (55 to 58%) compared with molecular dynamics assignments (13%) (Table 1; Table 2). Both CD and molecular dynamics indicate 30 to 34% turns and no -helix.
View larger version (17K):
[in this window]
[in a new window]
|
Figure 2. Effect of temperature on the far-UV mean residue ellipticity of the N-terminal peptide of s1-casein, f1-23. 1) 2°C, 2) 10°C, 3) 15°C, 4) 25°C, 5) 37°C, 6) 50°C, and 7) 70°C. A: In low ionic strength (µ) buffer (µ = 0.0074 M); B: in high µ buffer (0.10 M).
|
|
The CD studies of the large C-terminal fragment, f136-196, show moderate temperature sensitivity in the CD of the far-UV region (Figure 3B) and a significant temperature sensitivity at near-UV wavelengths (Figure 3A; Table 1). Unlike the parent molecule or the f1-23 portion, the CD data show shifts in the wavelength minima and some loss of extended structure and turns, with a concomitant increase in unspecified structure. The retention of secondary structure and tyrosine dichroism (Figure 3A, B) led to the characterization of this part of the s1-casein molecule as molten globule-like structure (Alaimo et al., 1999a,b). Molecular dynamics simulations are in agreement with this prediction (Table 2). This fragment undergoes higher levels of association at increased µ (Alaimo et al., 1999a), as noted previously.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 3. Effect of temperature on ellipticity of the C-terminal peptide of s1-casein-B, f136-196. 1) 10°C, 2) 27°C, 3) 50°C, and 4) 70°C. A: Molar ellipticity, near UV region in low ionic strength (µ) buffer (µ = 0.0074 M) (Alaimo et al., 1999a, b); B: mean residue ellipticity, far UV region in low µ buffer (0.0074 M) (adapted from Alaimo et al., 1999a).
|
|
Molecular Dynamics
There were no significant changes in backbone conformation in the model of s1-casein, f1-23, after molecular dynamics simulations (Figure 4A, B, Table 2). Changes in the orientations of side chains were observed, however, suggesting low-energy interactions and the possible formation of salt bridges. Three turns were expected because of the presence of 3 proline residues; after molecular dynamics simulations, Pro2 and Pro5 participate in turns involving Arg1, Pro2, Lys3 and His4, Pro5, Ile6, His7, respectively. However, the region surrounding Pro12 is not a complete turn because the 
, 
angles of Leu11, Pro12, and Gln13 are not consistent with the criteria for recognized turn conformations.
View larger version (25K):
[in this window]
[in a new window]
|
Figure 4. Stereo view of the effect of molecular dynamics on conformation of the s1-casein N-terminal peptide, f1-23. A: As excised from the parent molecule; B: after molecular dynamics in solvent water.
|
|
The conformation of s1-casein f136-196 changed dramatically during molecular dynamics simulations (Figure 5A, B; Table 2). A small amount of -helix was acquired, and most of the ß-sheet-like conformation disappeared. The percentage of residues involved in turns remained the same (38%), but the number of turns decreased while the number of residues involved in each turn increased. Note that the diminished C-C distance (Table 2) argues for a more compact state following the simulation. The peptide contains 7 Tyr, 3 Phe, and 1 Trp, which can interact hydrophobically but which must be precisely oriented to prevent highly unfavorable van der Waals and steric contacts with each other, with other hydrophobic residues, and with a few charged residues. Changes in the conformation of the sequence following Gln155 were probably greatly influenced by the orientation of Trp164, whose large hydrophobic side-chain had a major effect on backbone changes as dynamic simulations progressed, i.e., as the peptide became more flexible and began to fold, the backbone conformation about Trp164 and the orientations of the side chains were different after each dynamics iteration. It must be noted, however, that the motif centered on Pro147 retained its basic turn conformation.
View larger version (25K):
[in this window]
[in a new window]
|
Figure 5. Stereo view of the effect of molecular dynamics on conformation of the s1-casein C-terminal peptide, f136-196. A: As excised from the parent molecule; B: after molecular dynamics in solvent water. Because of the complexity of the figure, residues may be more easily identified by referring to Figure 5A.
|
|
Interaction Sites
Physical chemical (Horne, 1998) and molecular modeling studies (Kumosinski et al., 1994a) have suggested that the s1-casein molecule may be either di- or tri-functional in its interactions with itself or other proteins. The model proposed by Horne consists of a short hydrophobic sequence (1–23) followed by a hydrophilic loop (30–120), which contains the phosphopeptides, and a longer hydrophobic region (130–199). A combination of physical, chemical, and molecular modeling studies was employed to confirm the possibility of the 1–23 fragment as an associative site (Malin et al., 2001). The earlier molecular modeling studies (Kumosinski et al., 1994a) further subdivided the C-terminal section into 3 possible regions for protein-protein interactions, one centering on Pro147, one centering on Pro168, and the third centering on Pro185. The first 2 of these reactive sites have been confirmed (Alaimo et al., 1999a,b) by studies of the f136-196 peptide, but these concepts may not extend to the whole molecule.
These interacting sites are examined as potential regions for protein self-assembly in light of the data presented previously. The diagram in Figure 6 was constructed using 2 N-terminal sections of the refined structure of s1-casein (Kumosinski et al., 1994a) to show the antiparallel docking of 2 s1-casein chains linked by ion pairs, as proposed by Kumosinski et al. (1994a), possibly consisting of Arg22–Glu18 interactions. This diagram demonstrates the simplicity of an initial dimerization process involving the N-terminal peptide, f1-23, of s1-casein.
Antiparallel docking of 2 s1-casein chains involving segments of the C-terminal region is shown in Figure 7; 2 C-terminal sections of the refined structure of s1-casein (Kumosinski et al., 1994a, b) were used, as in Figure 6, to construct a diagram that shows the possible interactions proposed by Kumosinski et al. (1994a, b). The diagram suggests the possibility of minimal ion pairing of charged residues within the largely hydrophobic environment, such as the interaction of Arg151 of one peptide chain with Asp157 of the other; the possibility of hydrogen bonding between Lys132 of one chain and Tyr144 of a second chain is also suggested. Note that neither of the diagrams in Figures 7 and 8 was treated as a molecular model—i.e., no minimization and/or molecular dynamics operations were conducted; they are shown to illustrate the interaction sites proposed by Kumosinski et al. (1994a).
View larger version (18K):
[in this window]
[in a new window]
|
Figure 8. Analytical ultracentrifugation analysis of s1-casein-B at 20°C (pH = 6.75) in disodium piperazine-N,N'-bis(2-ethanesulfonic acid) (25 mM) with 80 mM KCl; µ (ionic strength) = 0.122 M and speed = 12,000 rpm. The curvature of the plot is indicative of self-association from dimer at the meniscus to higher order polymers. The lower plot shows the fit of the data by the Beckman program ASSOC 4, and the upper plot shows the residuals to the fit. Chi square for all of the fits in this study averaged 1 x 10–4.
|
|
Kumosinski et al. (1994a) suggested that polymers beyond dimer would involve the regions centering on Pro168 and Pro185, with the reporter group for the former interaction being Trp164. Alaimo et al. (1999a) found experimental evidence for the involvement of the region centering on Pro168 for polymer formation beyond dimer for f136-196. However, the significant changes in peptide structure that occurred during molecular dynamics simulations suggest that these areas (Pro168 and Pro185) would not be as stable as the area involving the Pro147 turn.
The structure of the segment centering on Pro147 is important to casein chemistry for another reason. Among the structural motifs found in the caseins, few structures homologous with similar regions in protein crystal structures have been found (Sawyer et al., 2002). The region centering on Pro 147 scores 32 on the T-Coffee scale (Notredame et al., 2000; www.expasy.org; Guex et al., 1997) with residues 17–33 of GRO/melanoma growth stimulatory activity (MGSA) and 72 with residues 13–33 of IL-8 (Kim et al., 1994) (Table 3). In the former protein, these residues constitute a sheet-
-turn-sheet motif similar to the predicted model as judged by NMR results, and this motif represents one of the sites of dimerization for the MGSA protein (Kim et al., 1994). The MGSA protein can readily be structurally aligned with IL-8, which has a crystal structure demonstrating a similar sheet-turn-sheet motif (T-Coffee = 94).
View this table:
[in this window]
[in a new window]
|
Table 3. T-Coffee1 scores for selected residues of s1-casein, melanoma growth stimulatory activity (MGSA), and IL-8.
|
|
The questions to be resolved in the context of protein self-association are whether one type of association is more likely to occur first in the process or whether the processes are random and not directed by structure. The molecular dynamics studies presented previously would argue that as the temperature is increased, the 1–23 portion would remain stable, and the CONTIN analysis of the CD data is in agreement with this conclusion. For the 136–196 peptide, both the CD analysis and, significantly, the molecular dynamics studies suggest changes in structure. However, the molecular dynamics data for f136-196 also predict that the Pro147 section would remain stable, while the 2 other portions of the large C-terminal region would unwind. Additionally, the near-UV CD data would argue for stability of the Tyr conformation in the molecule and, therefore, for stability of the Pro147 region (Alaimo et al., 1999a). Comparison of the self-associative properties of genetic variants of the whole molecule could help to clarify this issue.
Analytical Ultracentrifugation of s1-Casein
A significant amount of data has been accumulated on the self-association of s1-casein under many conditions of pH and µ (Alaimo et al., 1999a; Swaisgood, 2003). In summary, at very low µ, the protein is monomeric and dimerizes with increased µ. However, at pH and µ conditions approximating physiological, the molecule forms dimers, trimers, and higher associated species, as reviewed by Swaisgood (2003). It is important to note that most of these studies were conducted at 20°C. Figure 8 shows the plot for analytical ultracentrifugation studies at 20°C for s1-casein-B; the curvature at the right clearly indicates the presence of polymers, and the analysis of the meniscus area yields the dimer as the reactive species. Studies of the A, B, and C variants of s1-casein yielded weight average molecular weights commensurate with the literature for the B and C variants (Table 4). Although the mass of the A variant is somewhat smaller because of the absence of 13 residues (Glu14-Ala26) per monomer, it behaved similarly to the B and C variants. This is important because the A variant lacks a significant portion of the N-terminal hydrophobic region. These data, considered alone, seem to suggest that the f1-23 segment cannot be a primary interaction site, but other factors may point to a role for an N-terminal interaction, as outlined subsequently.
When the temperature was raised to 25°C, the weight average molecular weight decreased for all 3 variants, and on heating to 37°C, further decreases occurred (Table 5). This dissociation at elevated temperatures has not been reported previously. When the temperature was lowered to 5°C, the weight average molecular weight returned to the higher value (Table 5), indicating that the process is reversible. The ultracentrifugation data, as shown in Figure 8, were analyzed with the ASSOC 4 protocol (Malin et al., 2001) to yield the reactive protomer and a weight-based association constant, ka (Table 5). No monomeric species were detected with this model; the reactive species was dimeric, and the percent dimer can be calculated from the constants (Table 5). For the 3 genetic variants at 37°C, and at conditions approximating physiological, the percentage dimer for the A, B, and C variants was 94, 86, and 94%, respectively. Thus, of the 4 potential hydrophobic interaction sites in the molecule, only one predominates at near physiological conditions.
View this table:
[in this window]
[in a new window]
|
Table 5. Effect of temperature on association constants and degree of association of s1-casein genetic variants.1
|
|
Four lines of argument can be presented to identify the specific reactive region:
- Because the A variant lacks the N-terminal reactive region and still forms dimers, this area would seem to be ruled out. However, as will be shown later, the interaction of N-terminal regions could have a role in s1-casein self-association.
- The molecular dynamics studies suggest that regions centering on Pro168 and Pro185 would be less structurally stable than the region centering on Pro147. Although these studies were conducted at a relatively low constant temperature (300°K), the molecular dynamics conditions provided enough energy to produce structural changes. Thus, the energy added to the molecule during dynamics simulation was sufficient to unfold some of the structures and alter the sheet-turn-sheet motifs. As a result, the reactive region centering on Pro147 is predicted to be more thermally stable than the regions centering on Pro residues 168 and 185.
- Near-UV CD studies showed that tyrosine dichroism is maintained at elevated temperatures, and this residue is a reporter group for the Pro147 region (Alaimo et al., 1999a).
- The B and C genetic variants have differing degrees of self-association at lower temperatures, presumably because of the Glu/Gly change at residue 192, and both species give similar dimeric structures at 37°C, ruling out the Pro185 region.
|
CONCLUSIONS
|
|---|
The process of casein micelle assembly in mammary gland is thought to proceed in a progressive fashion, based on EM observations (Farrell, 1999). Protein-protein interactions may occur first in the endoplasmic reticulum, bringing preformed protein aggregates about the size of the putative submicelles (as visualized by positive stain in EM) to the Golgi region (Farrell, 1999). In the Golgi region, ATP-mediated pumps deliver calcium and initiate casein micelle formation. If the initial µ is low, it is not likely that f136-196 could provide a nucleation site for the earliest stages of s1-casein association, as explained previously, but with higher degrees of association at increased µ, it can indeed provide a stronger lattice for a casein micelle framework. It would also have a role in stabilizing the entry of other caseins (e.g., ß-, -, and s2-caseins) into the assembly.
As suggested earlier (Malin et al., 2001), the evidence points to the N-terminal, f1-23, as a likely locus for initial self-association of s1-casein because it forms dimmers readily at low ionic strength and has a stable secondary structure that is not likely to be affected by interactions with other caseins of the micelle. However, at elevated µ and temperature only, the Pro147 region can form stable self-association reactions. It must be kept in mind that these conditions mimic the lumen of the endoplasmic reticulum, prior to the introduction of calcium by the ATP-driven pumps of the Golgi. During the process of casein micelle assembly in Golgi vesicles, the introduction of calcium into preformed protein complexes would provide new interaction sites. Thus, f1-23 would be important in this latter process, as the s1-casein-A variant lacks this region, and its calcium complexes are much less stable (Farrell et al., 2004).
An interesting conjecture may be made at this point concerning the N-terminal section that contains f1-23. This fragment is quite basic compared with the rest of the molecule, and 5 of the first 7 residues are positively charged. This is one of the few areas in any of the caseins that could present a positive surface charge (Swaisgood, 2003). This can be dramatically shown by a pseudo-charge surface map of the molecule (Figure 9), where only the N-terminal segment is coded blue for a basic surface charge. The large red lobe on the right and above the blue N-terminal corresponds to the phosphopeptide region. It may be speculated that, because at physiological µ the N-terminal segment does not interact with other caseins, it could readily react with inorganic phosphate. This would provide a bidentate region for attachment of calcium phosphate; as calcium binds to the phosphopeptide region, phosphate may bind to the N-terminal portion of the molecule. Experimentally, it has been shown that the presence of calcium enhances phosphate binding (Visser et al., 1979) and that modification of amino groups decreases calcium phosphate binding to s1-casein (Zhang and Aoki, 1995). The presence and stability of these 2 oppositely charged regions on the same hydrophilic surface could anchor the developing colloidal calcium phosphate during micelle accumulation in the Golgi apparatus.
View larger version (79K):
[in this window]
[in a new window]
|
Figure 9. Pseudo-charge surface representation of the energy-minimized putative 3 D molecular model of s1-casein-B (Kumosinski et al., 1994a). The C-terminal is on the left. Note the preponderance of acidic charges (red) throughout the molecule, except for the N-terminal region, which is basic (blue). The latter region corresponds to the f1-23 studied in this work. This figure was produced using the molecular surface and electrostatics (Kumosinski et al., 1994a) and calculations from Deep View/Swiss PDB Viewer 3.7 (www.expasy.org/spdbv/; Guex et al., 1997). For this calculation, the dielectric constant of the solvent was set at 150, that of the protein at 10; the red potential is –1.8, white is 0, and blue is +1.8. It should be noted that the molecular model of s1-casein (Kumosinski et al., 1994a) provides a picture of an essentially ab initio structure that could possibly exist, but it was constructed with human intervention using theoretical considerations and does not have the certainty ascribed to structures derived from crystallography.
|
|
In summary, these arguments suggest a stepwise progression from stable interactions at the C-terminal region, mainly involving the residues surrounding Pro147, to N-terminal interactions that help to anchor the assembly. Subsequent associations of s1-casein dimers into a micelle framework might involve other mechanisms that have not been discussed here, e. g., dimer-monomer or multimer-multimer associations mediated by calcium ion concentration.
|
ACKNOWLEDGEMENTS
|
|---|
The authors thank Paul L. H. McSweeney and Patrick F. Fox, Department of Food Chemistry, University College, Cork, Ireland, for assistance in obtaining the s1-casein N-terminal peptide, f1-23, which was synthesized at the National Food Biotechnology Centre (Cork, Ireland).
|
FOOTNOTES
|
|---|
* Mention of trade names or commercial products in this article is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the US Department of Agriculture. 
Received for publication October 8, 2004.
Accepted for publication April 4, 2005.
|
REFERENCES
|
|---|
Adzhubei, A. A., and M. J. E. Sternberg. 1993. Left-handed polyproline II helices commonly occur in globular proteins. J. Mol. Biol. 229:472–493.[Medline]
Alaimo, M. H., H. M. Farrell, Jr., and M. W. Germann. 1999b. Conformational analysis of the hydrophobic peptide s1-casein(136–196). Biochim. Biophys. Acta 1431:410–420.[Medline]
Alaimo, M. H., E. D. Wickham, and H. M. Farrell, Jr. 1999a. Effect of self-association of s1-casein and its cleavage fractions s1-casein(136–196) and s1-casein(1–197) on aromatic circular dichroic spectra: Comparison with predicted models. Biochim. Biophys. Acta 1431:395–409.[Medline]
Byler, D. M., H. M. Farrell, Jr., and H. Susi. 1988. Raman spectroscopic study of casein structure. J. Dairy Sci. 71:2622–2629.[Abstract/Free Full Text]
Chaplin, L. C., D. C. Clark, and L. J. Smith. 1988. The secondary structure of peptides derived from caseins: A circular dichroism study. Biochim. Biophys. Acta 956:162–172.[Medline]
Clarke, R., and S. Nakai. 1971. Investigation of -– s1-casein interaction by fluorescence polarization. Biochemistry 10:3353–3357.[Medline]
Creamer, L. K., T. Richardson, and D. A. D. Parry. 1981. Secondary structure of bovine s1- and ß-casein in solution. Arch. Biochem. Biophys. 211:689–696.[Medline]
Curley, D. M., T. F. Kumosinski, J. J. Unruh, and H. M. Farrell, Jr. 1998. Changes in the secondary structure of bovine casein by FTIR: Effects of calcium and temperature. J. Dairy Sci. 81:3154–3162.[Abstract]
Day, R., B. J. Bennion, S. Ham, and V. Daggett. 2002. Increasing temperature accelerates protein unfolding without changing the pathway of unfolding. J. Mol. Biol. 322:189–203.[Medline]
Farrell, H. M., Jr. 1999. Milk composition and synthesis. Pages 256–263 in Encyclopedia of Reproduction. Vol. 3. E. Knobil and J. D. Neill, ed. Academic Press, New York.
Farrell, H. M., Jr., E. M. Brown, P. D. Hoagland, and E. L. Malin. 2003a. Higher order structures in casein: A paradox. Pages 96–116 in Advanced Dairy Chemistry: Proteins. Vol. 1. 3rd ed. P. F. Fox and P. L. H. McSweeny, ed. Kluwer Publishers, New York.
Farrell, H. M., Jr., P. H. Cooke, E. D. Wickham, E. G. Piotrowski, and P. D. Hoagland. 2003b. Environmental influences on -casein: Reduction and conversion to fibrillar (amyloid) structures. J. Protein Chem. 22:259–273.[Medline]
Farrell, H. M., Jr., R. Jimenez-Flores, G. T. Bleck, E. M. Brown, J. E. Butler, L. K. Creamer, C. L. Hicks, C. M. Hollar, K. F. Ng-Kwai-Hang, and H. E. Swaisgood. 2004. Nomenclature of the proteins of cows milk—Sixth revision. J. Dairy Sci. 87:1641–1674.[Abstract/Free Full Text]
Farrell, H. M., Jr., T. F. Kumosinski, and G. King. 1994. Three-dimensional molecular modeling of bovine caseins. Pages 392–419 in Molecular Modeling: From Virtual Tools to Real Problems. T. F. Kumosinski and M. N. Liebman, ed. American Chemical Society, Washington, DC.
Farrell, H. M., Jr., P. X. Qi, E. D. Wickham, and J. J. Unruh. 2002. Secondary structural studies of bovine caseins: Structure and temperature dependence of ß-casein phosphopeptide (1–25) as analyzed by circular dichroism, FTIR spectroscopy, and analytical centrifugation. J. Protein Chem. 21:307–321.[Medline]
Farrell, H. M., Jr., E. D. Wickham, J. J. Unruh, P. X. Qi, and P. D. Hoagland. 2001. Secondary structural studies of bovine caseins: Temperature dependence of ß-casein structure as analyzed by circular dichroism and FTIR spectroscopy and correlation with micellization. Food Hydrocoll. 15:341–354.
Guex, N., and M. C. Peitsch. 1997. SWISS-MODEL and the Swiss-PdbViewer: An environment for comparative protein modeling. Electrophoresis 18:2714–2723.[Medline]
Hoagland, P. D., J. J. Unruh, E. D. Wickham, and H. M. Farrell, Jr. 2001. Secondary structure of bovine s2-casein: Theoretical and experimental approaches. J. Dairy Sci. 84:1944–1949.[Abstract]
Holt, C. 1992. Structure and stability of bovine casein micelles. Adv. Protein Chem. 43:63–113.[Medline]
Horne, D. 1998. Casein interactions: Casting light on the black boxes, the structure in dairy products. Int. Dairy J. 8:171–177.
Huq, N. L., K. J. Cross, and E. C. Reynolds. 1995. A 1H-NMR study of the casein phosphopeptide s2-casein (59–79). Biochim. Biophys. Acta 1247:201–208.[Medline]
Kakalis, L. T., T. F. Kumosinski, and H. M. Farrell, Jr. 1990. A multinuclear, high-resolution NMR study of bovine casein micelles and submicelles. Biophys. Chem. 38:87–98.[Medline]
Kim, K.-S., I. Clark-Lewis, and B. D. Sykes. 1994. Solution structure of GRO/melanoma growth stimulatory activity determined by 1H NMR spectroscopy. J. Biol. Chem. 269:32909–32915.[Abstract/Free Full Text]
Kumosinski, T. F., E. M. Brown, and H. M. Farrell, Jr. 1993a. Three-dimensional molecular modeling of bovine caseins: An energy-minimized ß-casein structure. J. Dairy Sci. 76:931–945.[Abstract/Free Full Text]
Kumosinski, T. F., E. M. Brown, and H. M. Farrell, Jr. 1993b. Three-dimensional molecular modeling of bovine caseins: A refined, energy-minimized -casein structure. J. Dairy Sci. 76:2507–2520.[Abstract]
Kumosinski, T. F., E. M. Brown, and H. M. Farrell, Jr. 1994a. Predicted energy-minimized s1-casein working model. Pages 368–389 in Molecular Modeling: From Virtual Tools to Real Problems. T. F. Kumosinski and M. N. Liebman, ed. American Chemical Society, Washington, DC.
Kumosinski, T. F., G. King., and H. M. Farrell, Jr. 1994b. An energy-minimized casein submicelle working model. J. Protein Chem. 13:681–700.[Medline]
Kumosinski, T. F., G. King., and H. M. Farrell, Jr. 1994c. Comparison of the three-dimensional molecular models of bovine submicellar caseins with small-angle X-Ray scattering—Influence of protein hydration. J. Protein Chem. 13:701–714.[Medline]
Kumosinski, T. F., J. Uknalis, P. H. Cooke, and H. M. Farrell, Jr. 1996. Correlation of refined models for casein submicelles with electron microscopic studies of casein. Lebensm.-Wiss. u.-Technol. 29:326–333.
Malin, E. L., M. H. Alaimo, E. M. Brown, J. M. Aramini, M. W. Germann, H. M. Farrell, Jr., P. L. H. McSweeney, and P. F. Fox. 2001. Solution structures of casein peptides: NMR, FTIR, CD, and molecular modeling studies of s1-casein, 1–23. J. Protein Chem. 20:391–404.[Medline]
Malin, E. L., and E. M. Brown. 1995. Influence of casein peptide conformations on textural properties of cheese. Pages 303–310 in Chemistry of Structure-Function Relationships in Cheese. E. L. Malin and M. H. Tunick, ed. Plenum Press, New York.
Malin, E. L., and E. M. Brown. 1999. Molecular modeling of conformational changes in solvated s1-casein peptides. Int. Dairy J. 9:207–213.
Notredame, C., D. G. Higgins, and J. Heringa. 2000. T-Coffee: A novel method for fast and accurate multiple sequence alignment. J. Mol. Biol. 302:205–217.[Medline]
Provencher, S. W., and J. Glöckner. 1981. Estimation of globular protein secondary structure from circular dichroism. Biochemistry 20:33–37.[Medline]
Qi, P. X., E. D. Wickham, and H. M. Farrell, Jr. 2004. Thermal and alkaline denaturation of bovine ß-casein. J. Protein Chem. 23:389–402.
Rollema, H. S., and J. A. Brinkhuis. 1989. A 1H-NMR study of bovine casein micelles; Influence of pH, temperature and calcium ions on micellar structure. J. Dairy Res. 56:417–425.[Medline]
Sawyer, L., P. N. Barlow, M. J. Boland, L. K. Creamer, H. Denton, P. J. B. Edwards, C. Holt, G. B. Jameson, G. Kontopidis, G. E. Norris, S. Uhrinova, and S-Y Wu. 2002. Milk protein structure—What can it tell the dairy industry? Int. Dairy J. 12:299–310.
Smyth, E., C. D. Syme, E. W. Blanch, L. Hecht, M. Vasak, and L. D. Barron. 2001. Solution structure of native proteins with irregular folds from Raman optical activity. Biopolymers 58:138–151.[Medline]
Swaisgood, H. E. 2003. Chemistry of the caseins. Pages 138–201 in Advanced Dairy Chemistry: Proteins. Vol. 1. 3rd ed. P. F. Fox and P. L. H. McSweeny, ed. Kluwer Publishers, New York.
Syme, C. D., E. W. Blanch, R. Jakes, J. Ross, M. Goedert, L. Hecht, and L. D. Barron. 2002. A Raman optical activity study of rheomorphism in caseins, synucleins and tau: New insights into the structure and behavior of natively unfolded proteins. Eur. J. Biochem. 269:148–156.[Medline]
Tsuda, S., R. Niki, T. Kuwata, I. Tanaka, and K. Hikichi. 1991. 1H-NMR study of casein phosphopeptide (1–25): Assignment and conformation. Magnetic Resonance Chem. 28:1097–1102.
Visser, J., R. W. Schaier, and M. van Gorkom. 1979. The role of calcium, phosphate and citrate ions in the stabilization of casein micelles. J. Dairy Res. 46:333–335.[Medline]
Weiner, S. J., P. A. Kollman, D. A. Case, U. C. Singh, C. Ghio, G. Algona, S. Proeta, Jr., and P. Weiner. 1984. A new force field for molecular mechanical simulation of nucleic acids and proteins. J. Am. Chem. Soc. 106:765–784.
Weiner, S. J., P. A. Kollman, D. T. Nguyen, and D. A. Case. 1986. An all atom force field for simulations of proteins and nucleic acids. J. Computational Chem. 2:230–252.
Zhang, Z. P., and T. Aoki. 1995. Effect of modification of amino groups on crosslinking of casein by micellar calcium phosphate. J. Dairy Sci. 78:36–43.[Abstract]
This article has been cited by other articles:
|
|
|
|
 
D. J. McMahon and B. S. Oommen
Supramolecular Structure of the Casein Micelle
J Dairy Sci,
May 1, 2008;
91(5):
1709 - 1721.
[Abstract]
[Full Text]
[PDF]
|
|
|
TOP
ABSTRACT
INTRODUCTION
