| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Article |
Correspondence to: Jere P. Segrest.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Development of a computer program called LOCATE allowed us to show that human apolipoprotein B-100 is composed of five domains, NH2-
1-ß1-2-ß2-3-COOH, enriched, alternately, in amphipathic helixes and amphipathic ß strands. Using updated versions of this program, here we compare the complete sequence of human apolipoprotein B-100 with partial sequences from eight additional species of vertebrates (chicken, frog, hamster, monkey, mouse, pig, rat, and rabbit). The lipid-associating amphipathic helixes cluster in domains 2 (between residues 2075 ± 25 and 2575 ± 25) and 3 (between residues 4100 ± 100 and 4550 ± 50) in all species for which those regions have been sequenced but with little conservation of individual helixes. Lipid-associating amphipathic ß strands cluster in domains ß1 (approximately residues 827-2000) and ß2 (approximately residue 2571 to residue 4000 ± 50) in all species for which these regions have been sequenced, with conservation of several individual amphipathic ß strands. Hydrophobic segments are present in apolipoprotein B-100 sequences of all nine species but the frequency of occurrence is no greater than generally found in ß sheet-containing proteins.
We conclude that four alternating lipid-associating domains, - ß1-2-ß2-3-COOH, are common supramolecular features of apolipoprotein B-100 in nine vertebrate species.— Segrest, J. P., M. K. Jones, V. K. Mishra, V. Pierotti, S. H. Young, J. Borén, T. L. Innerarity, and N. Dashti. Apolipoprotein B-100: conservation of lipid-associating amphipathic secondary structural motifs in nine species of vertebrates. J. Lipid Res. 1998. 39: 85–102.
Supplementary key words:
amphipathic helixes, amphipathic ß strands, hydrophobic amino acid sequences, plasma lipoproteins, computer analysis of amphipathic motifs, low density lipoprotein, protein homology, computer program LOCATE, amphipathic domains, protein–lipid interactions
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Plasma apolipoproteins can be grouped into two general classes, the nonexchangeable apolipoproteins (apoB -100 and apoB -48), and the exchangeable apolipoproteins (apoA-I, apoA-II, apoA-IV, apoC-I, apoC-II, apoC-III, and apoE). ApoB -100 is highly insoluble in aqueous solution and is also one of the largest monomeric proteins known. Because of its size and insoluble nature, it has been difficult to deduce the structural motif(s) responsible for the lipid-associating proper-ties of apoB -100. On the other hand, the exchangeable apolipoproteins are soluble in aqueous solutions, and the secondary structural motif (amphipathic helix) responsible for their lipid association has been extensively studied (1) (2).
The amphipathic helix is a common secondary structural motif in biologically active peptides and proteins. In a previous review article (3) we grouped amphipathic helixes into seven distinct classes, with class A representing the lipid-associating amphipathic helical domains of the exchangeable apolipoproteins. A variation on the class A theme is distinguished by a radial clustering of positive and negative residues into a pattern different from that of class A. Because of the presence of positively charged residues at the polar–nonpolar interface, as well as at the center of the polar face, we term this the class Y motif (4).
An additional class of amphipathic helix, termed G*, was also found in the exchangeable apolipoproteins. This class is distinguished by a random radial arrangement of positively and negatively charged residues. These amphipathic helixes are similar but not identical to the class G amphipathic helixes found in helical globular proteins. Class G* amphipathic helixes differ from those of class G in having both a greater hydrophobic moment and a greater nonpolar face hydrophobicity. The class G* amphipathic helixes are postulated to prefer protein–protein interactions over protein–lipid interactions (4). The best characterized example of this class is the N-terminal 4-helix bundle globular domain of apoE (5) that consists of four class G* amphipathic helixes (4) and, depending upon local conditions, fold as a 4-helix bundle protein or associate with lipid (5).
To identify lipid-associating domains in apoB -100, we developed a computer program called LOCATE that searches amino acid sequences to identify potential amphipathic structural motifs using sets of rules for helix and ß strand termination (6). Two dense clusters of putative lipid-associating amphipathic helixes were located precisely in the middle and the C-terminal end of apoB -100; a cluster of class G* helixes was located at the N-terminus. The two regions between the three amphipathic helix clusters were found to be highly enriched in putative amphipathic ß strands, while the three amphipathic helical domains were largely devoid of this putative lipid-associating motif. We proposed, therefore, that human apoB -100 has a pentapartite structure, NH2-1-ß1-2-ß2-3-COOH, with 1 representing a globular domain (6).
Recent enhancements to the program LOCATE have been used to compare the complete sequence of human apoB -100 with partial sequences from eight species of vertebrates (chicken, frog, hamster, monkey, mouse, pig, rat, and rabbit). Based on these analyses, we demonstrate in this report the ubiquitous presence of the pentapartite structure in apoB -100 from phylogenetically diverse vertebrate species.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Programs for displaying amphipathic a helixes
The programs WHEEL, COMBO, and CONSENSUS are described in detail elsewhere (7). WHEEL creates a helical wheel diagram of a given sequence of amino acids arranged as an ideal helix seen down the long axis from the amino terminal end. COMBO superimposes and averages a set of rotationally defined helical wheels. Selected residues of the set are projected onto two wheel diagrams. In the default option, the left-hand and right-hand wheels display the counts of all positively and negatively charged residues, respectively. CONSENSUS superimposes a set of helical wheels in the same fashion as COMBO and a single figure classifies the amino acid residues into five physical–chemical groups: positive, negative, polar, neutral, and hydrophobic.
Programs for locating amphipathic helixes, amphipathic b strands, or hydrophobic segments in protein sequences
LOCATE finds and displays potentially lipid-associating motifs in an amino acid sequence where the amino acid file contains a sequence of one letter amino acids and the optional free energy of transfer file contains default settings for amino acid residues. Amino Acid Residue File: This file contains a sequence of one letter amino acid residues in the standard encoding. Free Energy of Transfer File: This file is a modification of the GES hydrophobicity scale (8) and contains a free energy of transfer, hydrocarbon to water, in kcal/mol for each amino acid residue used in the amino acid residue file: Arg -12.3; Asp -9.2; Lys -8.8; Glu -8.2; Asn -4.8; Gln -4.1; His -3.0; Pro -0.6; Ser -0.2; Ala 0.0; Gly 0.1; Thr 0.2; Tyr 0.4; Trp 1.9; Cys 2.0; Val 2.6; Leu 2.8; Ile 3.1; Met 3.4; Phe 3.7.
Display options
/DISPLAY_PS automatically runs LOCATE in X-Windows mode and allows a postscript display of selected single amphipathic helixes, amphipathic ß strands, or hydrophobic segments or a postscript display of selected sets of amphipathic helixes as COMBO or CONSENSUS diagrams. /LINEAR eliminates the hydrophobicity scale and plots the helixes along the same vertical line.
Motif specific options
LOCATE_ALPHA identifies potential amphipathic helixes within a given amino acid sequence using termination rules described elsewhere (7). /CLASS=list specifies the class of helixes that will be reported. A: helix is determined to be class A; Y: helix is determined to be class Y; G*: helix is anything else. The algorithm for identification of class A and class Y amphipathic helixes is diagrammatically illustrated in Figure 1. /LAMBDA_HYB shows the 
lipid affinity of each amphipathic helix on the x axis as calculated by the function. /LAMBDA_HYB_CUTOFF specifies the minimum value that the lipid affinity of the nonpolar face of the candidate helix can be and still be selected.
|
The function for calculating lipid affinity was derived as follows. The snorkeling of basic residues allows for greater penetration of class A amphipathic helixes into the hydrophobic interior of phospholipid monolayers than would otherwise be possible; the greater the angle of the snorkel wedge (largely equivalent to the angle of the hydrophobic face), the greater the lipid penetration (9). Using neutron diffraction, Jacobs and White (10) have measured the gradient that H2O forms from the outside to the inside of a phospholipid monolayer. The hydrocarbon core starts at a depth of approximately 7 Å beneath the center of the phosphatidylcholine head group; at this depth, H2O has a molar concentration approximately 15% of that at the level of the phosphatidylcholine head groups. As the free energy of the hydrophobic effect decreases with a decrease in concentration of H2O, the deeper the penetration of an amphipathic helix into the interior of a phospholipid monolayer, the more effective the hydrophobicity (i.e., the lower the free energy) of its nonpolar face. Therefore, the overall lipid affinity of an amphipathic helix will partially depend upon its depth of lipid penetration (9).
Combining the water gradient determined by Jacobs and White (10) with the free energy of transfer of an amino acid residue from H2O to varying H2O:organic solvent mixtures, we have derived a free energy gradient, 
i(Å), that is a function of depth of penetration in Å, di(Å). Therefore, = 
(
G
i x i(Å)). The method of calculating described in Palgunachari et al. (9) assumes that the lipid surface is always parallel to the helix axis.
A recently developed method of calculating lipid affinity (3) is based upon a three-dimensional helical cylinder analysis. The program, using the assumptions of the algorithm applied to a three-dimensional model, considers each amphipathic helix to be a cylinder 14 Å in diameter with a pitch of 1.5 Å per residue. The 3 algorithm then determines the orientation of the amphipathic helix relative to the plane of the hydrated phospholipids that produces the maximum lipid affinity. From this orientation, four properties are determined: a) angle of rotation of the amphipathic helix relative to the hydrated phospholipid plane, b) long axis tilt of the amphipathic helix relative to the hydrated phospholipid plane, c) depth of penetration of each residue into the hydrated phospholipid in three dimensions, and d) the calculated lipid affinity, 3.
Briefly, the 3 algorithm is based upon the rule that three points determine a plane; here the points are determined by charged residues. For any such plane, all charged residues must lie on the plane or be on one side of the plane, the polar side. Because the maximum value of 3 may not occur for a plane that passes through three charged residues, the plane is allowed to pivot through two charged residues and even one: the plane orientation is adjusted at 1° increments giving multiple rotations and tilts. As in the previous algorithm, Lys, but not Arg, is allowed to snorkel. The present algorithm provides the following output: a) a calculated 3, b) a calculated angle of tilt, c) a diagrammatic representation of the plane of the hydrated phospholipid relative to the N- and C-terminal end of the amphipathic helix (displayed on the helical wheel output), d) depth of penetration (in Å) of each residue, and e) highlighting of all residues buried in the hydrated phospholipid (displayed on both the helical wheel and helical net outputs), and f ) a depth of penetration contour map plotted on the helical net output.
LOCATE_BETA identifies potential amphipathic ß strands within a given amino acid sequence using termination rules also described elsewhere (6). LOCATE_ HYB identifies segments of hydrophobic residues within a given amino acid sequence using charged and/or polar residues for termination.
General options
These programs create either a linear or a two-dimensional plot. The two-dimensional plot displays the location of each sequence with the desired motif on the Y-axis versus a measure of its hydrophobic moment, hydrophobicity of the hydrophobic face (per residue or total), total hydrophobicity, or calculated lipid affinity () on the X-axis. Options allow: 1) selections of amphipathic helixes (negatively or positively charged) or amphipathic ß strands (negatively or positively charged) and 2) minimum allowable cutoff values for any given parameter (e.g., sequence length, hydrophobicity, lipid affinity, etc.).
/MINIMUM_LENGTH specifies the minimum value that the length of the candidate motif can be and still be selected. /NONPOLAR_HYB_CUTOFF specifies the minimum value that the total hydrophobicity of the nonpolar face of the candidate amphipathic motif can be and still be selected. /TOTAL_HYB shows the total (summed) hydrophobicity of each motif selected on the x axis. /TOTAL_HYB_CUTOFF specifies the minimum value that the total hydrophobicity of the candidate motif can be and still be selected.
Sequence alignments and determination of sequence and structural homologies
Partial sequences of apoB -100 from six species of vertebrates, monkey, pig, rat, rabbit, hamster, and chicken, were downloaded in the FASTA format using the ENTREZ Browser (http://atlas.nlm.nih.gov). Partial sequences of apoB -100 from two additional species, mouse and frog, were determined in the laboratories of co-authors, Stephen H. Young and Thomas L. Innerarity, respectively.
Sequence alignment and homology analyses of the nine apoB -100 sequences were performed on a Macintosh using the program MACAW (11) (12) downloaded from the Internet (ftp://ncbi.nlm.nih.gov/pub/ schuler/macaw). To analyze structural (amphipathic and hydrophobic motif) homology, aligned sequences were subjected to LOCATE analysis using cutoff parameters described in the figure legends.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
LOCATE analysis of human apoB -100
Before beginning the structural comparisons of apoB -100 from nine vertebrate species, we examined the structural parameters of human apoB -100 predicted by using the most recent version of LOCATE. To aid in the visual identification of clusters of amphipathic motifs with high calculated lipid affinity, each probable amphipathic motif is displayed in the form of a bar graph created by increasing the cutoff incrementally on the X-axis from 5 to 20 kcals/mol for a) calculated lipid affinity () of amphipathic helixes (see Methods), b) the total hydrophobicity of the nonpolar face of amphipathic ß strands, and c) total hydrophobicity of hydrophobic segments.
LOCATE_ALPHA analyses of amphipathic helixes as a function of calculated lipid affinity are shown in Figure 2 A. Three dense clusters of amphipathic helixes with high calculated lipid affinity, residues 58–795 (1), 2250–2584 (2), and 4129–4505 (3), are clearly identified. These same general regions were identified as the three amphipathic helical domains in our original structural analysis of apoB -100 (6). The region 2045–2249 could be considered a portion of 2 containing a lower density of amphipathic helixes with weaker lipid affinity.
|
Figure 2b is a plot of all class A and Y amphipathic helixes identified by LOCATE_ALPHA/CLASS in human apoB -100; for each helix is plotted on the x-axis. Using this option to LOCATE_ALPHA, class A and Y amphipathic helixes are shown to cluster almost exclusively between residues 2052–2560 (2) and 4017–4515 (3); no class A or Y amphipathic helixes are found in the 1 domain. Combining Figure 2a and Figure 1B, we estimate that in human apoB -100 the 2 domain is located between residues 2045–2587 and the 3 domain between residues 4017–4515.
Figure 3 shows COMBO analyses, using LOCATE_ ALPHA/CLASS and LOCATE_ALPHA analyses in the X-Windows mode, of the amphipathic helixes from the 1, 2, and 3 domains of human apoB -100. Plots of the distribution of positively and negatively charged residues and Leu and Phe residues, respectively, in the class A amphipathic helixes in the 2 and 3 domains (Figure 2b) are shown in Figure 3a and Figure b, respectively. These plots illustrate the features common to the class A amphipathic helixes found in the exchangeable apolipoproteins (2) (4): clustering of the positively charged and negatively charged residues at the polar–nonpolar interfaces and the center of the polar face, respectively (Figure 3a), a well-defined clustering of hydrophobic residues in the middle of the nonpolar face (Figure 3b), a high nonpolar face mean hydrophobicity per residue of 2.05 (automatically calculated by both programs) and a mean length of 22.0 ± 9.2 amino acid residues, identical to the length of the tandem repeats found in the exchangeable apolipoproteins.
|
Analyses of amphipathic helixes selected on the basis of a calculated high lipid affinity (defined as 
7, third column from left, Figure 2a), and thus selected independently of amphipathic helix class, are shown in Figure 3c Figure d Figure e Figure f. Figure 3c and Figure d are COMBO plots of the distribution of positively and negatively charged residues and Leu and Phe residues, respectively, in the amphipathic helixes of amphipathic helixes found in the 2 plus 3 domains. These amphipathic helixes show well-defined clustering of the negatively charged residues in the center of the polar face (Figure 3c), a well-defined clustering of the negatively charged residues in the center of the polar face (Figure 3c), a well-defined clustering of hydrophobic residues in the middle of the nonpolar face (Figure 3d), a high nonpolar face mean hydrophobicity per residue (2.23), and a mean length of 19.5 ± 9.0 amino acid residues; the clustering of the positively charged residues (Figure 3c) occurs both at the polar–nonpolar interface and in the center of the polar face, presumably because these amphipathic helixes represent a mixture of class A and class Y (2) (4).
Figure 3e and Figure f are COMBO plots of the distribution of positively and negatively charged residues and Leu and Phe residues, respectively, of amphipathic helixes found in the 1 domain; the amphipathic helixes of this domain are distinctly different from those of the 2 and 3 domains. The most striking difference is that the positions of the positive and negative charge clusters in the amphipathic helixes of the 1 domain are random or perhaps reversed compared to class A amphipathic helixes (Figure 3e). Further, mean length is shorter and less variable (15.5 ± 4.4), and the clustering of hydrophobic residues in the middle of the nonpolar face is poorly defined (Figure 3f), compared to the amphipathic helixes in the 2 and 3 domains.
Figure 2c is a LOCATE_BETA bar graph analysis of amphipathic ß strands in human apoB -100. Two dense clusters of amphipathic ß strands with high calculated lipid affinity can be identified: ß1:827–2001 and ß2:2571–4032. These two domains were noted in our previous LOCATE analysis of apoB -100 (6). In addition, there is a lower density cluster of amphipathic ß strands not noted in our previous analysis, residues 24–354 (termed ßN), associated with the N-terminal half of the 1 domain. Even the C-terminal half of this domain, as well as the 2 and 3 domains, contain a few scattered, relatively short amphipathic ß strands.
Finally, it has been suggested that unique hydrophobic segments occur in apoB -100 (13). Figure 2d is a LOCATE_HYB bar graph analysis of hydrophobic segments in human apoB -100 that are 
10 residues in length with a total hydrophobicity 5 to 
20; a total of 17 sequence intervals with a total hydrophobicity 5 were identified. LOCATE_HYB analysis ( Table 1) of the test chimeric protein sequences, used as controls in our previous LOCATE analysis of apoB -100 (6), indicates that apoB -100, in general, is not unique in possessing short hydrophobic segments when compared to the motif B (ß sheet) test protein; proteins rich in ß sheets are also relatively rich in short hydrophobic segments. The cluster of four hydrophobic sequences between residues 3250–3347, however, conceivably could be unique. It will be useful, therefore, to see whether this cluster is conserved in the eight apoB -100 fragments from other vertebrate species.
|
Alignment of the nine apoB -100 sequences
Figure 4 shows the results of the sequence alignment produced by the sequence comparison program MACAW. These results provide two useful pieces of information: sequence alignment for structural homology analysis using the program LOCATE and information concerning sequence identity in aligned regions.
|
All nine apoB -100 molecules have been sequenced between residues 3139–3409 (i.e., overlap in this region), a sequence stretch that covers a portion of the putative LDL receptor-binding domain in ß2 (13). Six sequences overlap in the region of ß2 between residues 3500–3700. Five sequences overlap in the C-terminal 3 domain (residues 4176 to the C-terminus). Five sequences overlap in the region of ß2 between residues 2713–3409. Four sequences overlap the region between residues 2436–2713, representing a small portion of the N-terminal end of ß2 and a small portion of the C-terminal end of 2. Three sequences overlap the entire 2 domain and a small portion of the C-terminal end of ß1. Only two sequences overlap the region of ß1 between residues 1380–2000. Three sequences overlap residues 400–1000, representing the C-terminal half of the 1 domain and a very small portion of ß1. Finally, two sequences overlap the first 100 residues of the N-terminal end of the 1 domain and none between residues 1000–1380 for ß1.
Structural homologies between the nine apoB -100 sequences
Figure 5 represents LOCATE_ALPHA analyses of the nine apoB -100 sequences aligned as shown in Figure 4 and indicates that the 2 and 3 domains are generally conserved. Each of the four partial sequences that overlap in the C-terminal 3 domain (pig, mouse, rat, and chicken) and the two sequences that overlap the entire 2 domain (pig and mouse) show a well-defined cluster of presumed lipid-associating amphipathic helixes ( 7) that correspond closely to the 3 domain and the 2 domain of human apoB -100, respectively (Figure 5a). The additional sequence (frog) that overlaps a small portion of the C-terminal end of 2 also shows the beginning of amphipathic helixes in that portion of 2. In all five species in which the 3 domain has been sequenced, an amphipathic helix more than 50 residues long with high calculated lipid affinity is located in the neighborhood of residue 4480 ( Figure 6 A).
|
|
Figure 5b is a LOCATE_ALPHA/CLASS analysis of all class A and Y amphipathic helixes identified in the nine apoB -100 sequences. The distribution of these amphipathic helixes corresponds closely to that of the presumed lipid-associating amphipathic helixes shown in Figure 5a; however, clustering of class A and Y amphipathic motifs is a more definitive marker of the 2 and 3 domains than calculated high lipid affinity.
The LOCATE_ALPHA analysis is Figure 5a, in addition to marking the 2 and 3 domains, also marks the 1 domain. Although a number of the amphipathic helixes in the 1 domain, unlike the 2 and 3 domains, may be conserved (arrows), more sequences that overlap the 1 domain of human apoB -100 are needed to confirm this impression.
Figure 7 is a LOCATE_BETA analysis of all amphipathic ß strands identified in the nine apoB -100 sequences that have a nonpolar face hydrophobicity 7. It is clear from this figure that amphipathic ß strands cluster in domains ß1 (approximately residues 827–2000) and ß2 (approximately residues 2571–4000) in all species sequenced in these regions. In contrast to the presumed lipid-associating amphipathic helixes, a number of the amphipathic ß strands of human apoB -100 are conserved (Figure 7, arrowheads); this is particularly true in the putative LDL receptor-binding domain (13) located between residues 3100–3500 in ß2. Another feature of the amphipathic ß strands that appears to be conserved are the several gap regions that occur in the ß1 and ß2 domains of human apoB -100 (e.g., between residues 900–1000 and 3200–3300). It is worthy of note that a long amphipathic ß strand is found within the first 60 N-terminal residues of both human and mouse apoB -100 (Figure 6b).
|
Finally, Figure 8 is a LOCATE_HYB analysis of all hydrophobic segments identified in the nine apoB -100 sequences that have a total hydrophobicity 7.5. The important point to make from this figure is that there is little conservation of the cluster of hydrophobic segments noted earlier in the ß2 domain of human apoB -100 between residues 3250–3347; these residues are represented in all 9 sequences.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Conservation of amphipathic motifs
In this study we have used the computer programs LOCATE_ALPHA, LOCATE_BETA and LOCATE_HYB to compare the distribution of lipid-associating motifs (amphipathic helixes, amphipathic ß strands, and hydrophobic segments, respectively) in human apoB -100 to their distribution in partial sequences from eight additional species of vertebrates (monkey, pig, mouse, rat, hamster, rabbit, chicken, and frog). Figure 9 is schematic map of the general distribution in the apoB -100 molecule of amphipathic helixes and amphipathic ß strands as determined here.
|
The LOCATE_ALPHA/CLASS algorithm that produced Figure 2b and Figure 5b used the radial position of charged residues only to identify class A and Y amphipathic helixes; hydrophobicity of the hydrophobic face was not a consideration. It therefore seems remarkable that: a) the hydrophobic residues of the class A and Y amphipathic show clustering of hydrophobic residues that is as well defined as that found in amphipathic helixes selected on the basis of high calculated lipid affinity (Figure 3b vs. Figure 3d); and b) class A and Y amphipathic helixes are confined almost exclusively to the 2 and 3 domains. These results suggest that class A and Y motifs are confined almost exclusively to amphipathic helixes possessing high lipid affinity.
Lipid-associating amphipathic ß strands cluster in two domains in all species for which these regions have been sequenced (Figure 7). Unlike the lipid-associating amphipathic helixes, several individual amphipathic ß strands within these domains, indicated by arrows in Figure 7, are conserved. The ß domains show considerably fewer regions of "homology" to the exchangeable apolipoproteins than do the domains (MACAW analyses, data not shown).
The ß1 domain is represented only partially in the sequences of pig, mouse, and rat. In these species, the amphipathic ß strands of this domain are distributed, on average, between residues 827–2001. There are a number of obvious gaps in amphipathic ß strand distribution in the ß1 domain and several of these gaps may be conserved (e.g., residues 878–966); however, the paucity of overlapping sequences in this domain precludes a firm conclusion.
The amphipathic ß strands of the ß2 domain are distributed, on average, between residue 2571 to approximately residue 4000. There is a variability of 100 residues or so in the C-terminal end of this domain in the four sequences in which this region is present (human, pig, mouse, rat); there appears to be more conservation of individual amphipathic ß strands in the N-terminal two-thirds of the ß2 domain than in the C-terminal one-third (Figure 7). As is true for the ß1 domain, a number of obvious gaps in amphipathic ß strand distribution are noted in the ß2 domain. The gap in the region of residue 3200 is conserved across all 9 sequences. The gap at residue 2700 is conserved in the three sequences, human, pig, and mouse, that overlap this region; there may also be a conserved gap of variable length just beyond residue 3500.
Finally, our analysis indicates that hydrophobic segments are present in apoB -100 but at a frequency no greater than in other ß sheet-containing proteins (Table 1). The cluster of hydrophobic segments located in the neighborhood of residue 3300 in human apoB -100 is poorly conserved (Figure 8).
Amphipathic a helical domains
The 2 domain in human apoB -100 is "homologous" to certain amphipathic helix-containing regions of six of the exchangeable apolipoproteins A-I, A-II, A-IV, C-I, C-III, and E (MACAW analyses, data not shown). This domain is represented in its entirety in the sequences of human, pig, and mouse and partially represented in the frog sequence. In these species, the amphipathic helixes of the 2 domain cluster between residues 2075 ± 25 and 2575 ± 25 (Figure 5); the highest concentration of amphipathic helixes occurs in the C-terminal half of this domain.
The 3 domain in human apoB -100 is "homologous" to specific amphipathic helix-containing regions of all seven of the exchangeable apolipoproteins A-I, A-II, A-IV, C-I, C-II, C-III, and E; this homology is especially striking with the C-terminal two-thirds of human apoA-I (MACAW analyses, data not shown). The 3 domain is represented in its entirety in the sequences of human, pig, mouse, rat, and chicken. In these species, the amphipathic helixes of the 3 domain cluster between residues 4100 ± 100 and 4550 ± 50; they are distributed more evenly than the amphipathic helixes of the 2 domain (Figure 5).
The 2 and 3 domains of apoB -100 correspond closely to the two major apoB -100 lipid-associating domains reported by Yang, et al. (13) at residues 2100–2700 and 4100–4500 using the principle of non-releasability of tryptic peptides from trypsin-treated intact LDL. Others have also suggested the presence of amphipathic helixes in these regions of apoB -100 (14) (15).
What might be the function of the lipid-associating amphipathic helical domains (2 and 3) of apoB -100? We have proposed that these domains represent the flexible, hinged domains of apoB -100, perhaps playing a role in LDL particle size variability (6).
It is unfortunate that only limited amino acid sequence data on the 1 domain of apoB -100 from species other than human is currently available, as comparative analyses of 1 sequence data from multiple species may be useful in the elucidation of mechanisms regulating the early steps in assembly of apoB -containing lipoproteins (see discussion below).
Amphipathic b strand domains
The presence of ß structure in apoB -100 has been suggested in the past using spectroscopic methods (16) (17) (18) (19) (20). The first suggestion of the presence of amphipathic ß strands in apoB -100 was made by Osterman et al. (21). While the amphipathic ß strand is now accepted by many investigators as a major lipid-associating motif in apoB (15), there is little published direct evidence for the existence of this motif and little is known about its mode of association with phospholipid monolayers; even less is known about its biological functions.
There are a number of unanswered questions concerning amphipathic ß strands in apoB -100. a) Do amphipathic ß strands associate with lipid as individual strands or, as suggested recently by Don Small (personal communication), as antiparallel sheets or both? We favor the possibility that both strands and sheets can interact depending upon the local conditions. b) What is the lipid affinity of amphipathic ß strands compared to class A amphipathic helixes from apoB -100 as well as from exchangeable apolipoproteins? c) Assuming they exist, what is the lipid affinity of antiparallel amphipathic ß sheets compared to amphipathic ß strands? We hypothesize that large antiparallel amphipathic ß sheets have what amounts to an infinitely high lipid affinity. d) What is the biological function of the amphipathic ß strands in apoB beyond determining lipid affinity? e) What is the function of the gap regions between amphipathic ß strands of the ß1 and ß2 domains?
Relationship of pentapartite model to truncated apoB -100 constructs
What experimental evidence is there to support the NH2-1-ß1-2-ß2-3-COOH pentapartite model for apoB -100 shown in Figure 9? Several studies have shown that the diameters of secreted lipoprotein particles are proportional to the length of C-terminal truncated apoB -100 fragments until the N-terminal end of the ß1 domain is approached (residues 1100–1300), at which point the linearity begins to fall off (22) (23) (24).
In another study, expression of C-terminal truncated forms of apoB -100 in HepG2 cells demonstrated that, with stepwise truncation, progressively more apoB -100 was recovered free of lipid in the cell culture medium (25); recovery of truncated apoB -100 was 100% at residue 1860, 48% at residue 1315, 20% at residue 1040, 18% at residue 770, and 0% at residue 590. Thus, beyond the N-terminal end of the ß1 domain (approximately residue 827, Figure 9), the lipid affinity of truncated and secreted apoB -100 approaches zero.
In a separate study using the McA-Rh777 cell line, apoB -100 truncated at residues 1040 and 815 were recovered in the lipid-free fraction, whereas apoB -100 truncated at or beyond residue 1270 formed discrete lipoprotein particles (23). Again the ability of apoB -100 to form lipoprotein particles disappears as the N-terminal end of the ß1 domain is approached.
Finally, when apoB -100 was truncated at residue 771 (apoB -17) and expressed in murine C127 cells, 60% was secreted as soluble, lipid-poor protein that formed discoidal complexes with dimyristoyl phosphatidylcholine (26). Circular dichroic spectra of this complex suggested lipid-associated apoB -17 was 39% helix and 36% ß sheet. LOCATE analyses of the 1 "globular" domain of human apoB -100 (residues 1–800) predicts that 36% of this domain is lipid-associating amphipathic helix ( 7) and 26% is lipid-associating amphipathic ß strand (total nonpolar face hydrophobicity 5).
We conclude that four alternating lipid-associating domains, -ß1-2-ß2-3-COOH, enriched in amphipathic ß strands and amphipathic helixes, respectively, are ubiquitous to the structure of apoB -100 from vertebrate species. While the 1 "globular" domain of apoB -100 seems almost certain to be conserved in apoB -100 from all species of vertebrates, the degree of this conservation must await more overlapping sequences with this domain in human apoB -100. We speculate that, because of the apparent role of this domain in the assembly of apoB -containing lipoprotein particle, the 1 "globular" domain will prove to be the most conserved of the and ß domains in apoB -100.
|
ACKNOWLEDGMENTS |
|---|
This work was supported in part by NIH grants HL-34343 to JPS, HL-41633 to SHY and HL-41633 to TLI. SHY would like to thank Dr. Brian J. McCarthy for advice and assistance during the sequencing of mouse apoB -100.
Manuscript received July 23, 1997; and in revised form September 17, 1997.
Abbreviations:
apo, apolipoprotein; , calculated affinity; 1, 2, 3, amphipathic helical domains of apolipoprotein B-100; ß1, ß2, amphipathic ß strand domains of apolipoprotein B-100
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
helix: a multifunctional structural motif in plasma apolipoproteins. Adv. Protein Chem. 45:303-369[Medline]. helical domains alternating with two amphipathic ß strand domains: detection by the computer program LOCATE. Arterioscler. Thromb. 14:1674-1685 helical domains. J. Lipid Res. 33:287-296[Abstract].
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  M. M. W. Smolenaars, A. de Morree, J. Kerver, D. J. Van der Horst, and K. W. Rodenburg Insect lipoprotein biogenesis depends on an amphipathic {beta} cluster in apolipophorin II/I and is stimulated by microsomal triglyceride transfer protein J. Lipid Res., September 1, 2007; 48(9): 1955 - 1965. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. M. W. Smolenaars, O. Madsen, K. W. Rodenburg, and D. J. Van der Horst Molecular diversity and evolution of the large lipid transfer protein superfamily J. Lipid Res., March 1, 2007; 48(3): 489 - 502. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Johs, M. Hammel, I. Waldner, R. P. May, P. Laggner, and R. Prassl Modular Structure of Solubilized Human Apolipoprotein B-100: LOW RESOLUTION MODEL REVEALED BY SMALL ANGLE NEUTRON SCATTERING J. Biol. Chem., July 14, 2006; 281(28): 19732 - 19739. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Rava, H. Athar, C. Johnson, and M. M. Hussain Transfer of cholesteryl esters and phospholipids as well as net deposition by microsomal triglyceride transfer protein J. Lipid Res., August 1, 2005; 46(8): 1779 - 1785. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. Boyle-Roden and R. L. Walzem Integral apolipoproteins increase surface-located triacylglycerol in intact native apoB-100-containing lipoproteins J. Lipid Res., August 1, 2005; 46(8): 1624 - 1632. [Abstract] [Full Text] [PDF]
|
|
|
|
