help button home button Biophys. J.
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Simon, A.
Right arrow Articles by Simon, I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Simon, A.
Right arrow Articles by Simon, I.

Biophys J, November 2000, p. 2305-2313, Vol. 79, No. 5

Function-Related Regulation of the Stability of MHC Proteins

Á. Simon,*dagger Zs. Dosztányi,* É. Rajnavölgyi,dagger and I. Simon*

 *Institute of Enzymology, Hungarian Academy of Sciences, H-1518 Budapest, P.O. Box 7, Hungary; and  dagger Department of Immunology, Eötvös L. University, H-2131 Göd, Jávorka S. u. 14, Hungary


    ABSTRACT
TOP
ABSTRACT
INTRODUCTION
DATABASE
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES

Proteins must be stable to accomplish their biological function and to avoid enzymatic degradation. Constitutive proteolysis, however, is the main source of free amino acids used for de novo protein synthesis. In this paper the delicate balance of protein stability and degradability is discussed in the context of function of major histocompatibility complex (MHC) encoded protein. Classical MHC proteins are single-use peptide transporters that carry proteolytic degradation products to the cell surface for presenting them to T cells. These proteins fulfill their function as long as they bind their dissociable ligand, the peptide. Ligand-free MHC molecules on the cell surface are practically useless for their primary biological function, but may acquire novel activity or become an important source of amino acids when they lose their compact stable structure, which resists proteolytic attacks. We show in this paper that one or more of the stabilization centers responsible for the stability of MHC-peptide complexes is composed of residues of both the protein and the peptide, therefore missing in the ligand-free protein. This arrangement of stabilization centers provides a simple means of regulation; it makes the useful form of the protein stable, whereas the useless form of the same protein is unstable and therefore degradable.


    INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
DATABASE
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES

Proteins must be stable to ensure their biological activity. However, protein stability must be limited, since degradation products of useless or inactive proteins are the main source of amino acid for de novo protein synthesis. A simple means of regulation, which makes a useful form of a protein stable and the useless form unstable, will be presented. It derives from the study of stabilization centers in an immunologically relevant peptide receptor and transporter.

Cell surface glycoproteins, encoded by genes of the major histocompatibility complex (MHC), begin and end their lifetime as empty molecules. At the cell surface however they are complexed with peptide ligands that are generated within the cell by limited proteolytic degradation and loaded intracellularly (reviewed in Germain and Margulies, 1993; Maenaka and Jones, 1999). A single MHC molecule can bind various, but not all, kinds of peptides (Rammensee, 1995; Rammensee et al., 1995). Peptide loading of MHC class I and class II molecules via the classical antigen presentation pathways occurs in distinct cellular compartments where various peptides compete for the available binding sites (reviewed in Harding and Geuze, 1993; Rajnavölgyi, 1994). The resulting complexes are then transported to the cell surface and presented to T cells (Garcia and Teyton, 1998). The interaction between newly synthesized MHC class I and class II molecules and their ligands is assisted by various chaperones, which support the peptide-accessible conformation of the molecule (Koopmann et al., 1997; Busch et al., 2000). Ligand-free MHC proteins usually do not reach the cell surface, since they are degraded rapidly inside the cell. MHC-peptide complexes expressed on the cell surface, however, are stable for several hours or even for days (Lanzavecchia et al., 1992; Hansen et al., 2000). Peptide ligands bind noncovalently to MHC molecules. Therefore, they dissociate from the heterodimers at a rate determined by the strength of interactions between the two molecules.

As a rule, ligand-free MHC molecules that happen to reach the cell surface are also degraded quickly, and only a few survive, either by reloading their ligand binding site with peptides present in the extracellular environment or by internalization in intact form into peptide-rich endosomes. Fluorescence measurements suggested the presence of a few percent of unliganded and still compact MHC I molecules on the cell surface (Matko et al., 1994; Edidin et al., 1997). Under in vitro conditions, ligand-free MHC class I proteins were demonstrated to exhibit a molten globule state, which resembled what was known as an intermediate in the denaturation process (Bouvier and Wiley, 1998). A molten globule is much less compact than a native protein; therefore, it is more sensitive to proteolytic attacks and it may interact differently with other proteins such as chaperones. Ligand-free MHC class II molecules are found on the surface of professional antigen-presenting cells, such as immature dendritic cells and B cells in association with chaperones (Santambrogio et al., 1999; Arndt et al., 2000).

The overall structures of the peptide-binding domains of MHC class I and class II molecules are rather similar. They are built up of a large eight-stranded beta -sheet with two mostly alpha -helical regions on top of the beta -sheets. The MHC class I helices are formed by the alpha 150-alpha 155 and alpha 158-alpha 184 residues on one side of the peptide binding groove and by the alpha 2138-alpha 2148 and alpha 2151-alpha 2180 amino acids on the other side. In MHC class II molecules, two long helices, alpha 54-alpha 76 and beta 52-beta 90, border the two sides of the ligand binding site (Brown et al., 1993; Stern et al., 1994; Madden, 1995). The major difference between the detailed structures of the ligand-binding sites is that they are closed by conserved amino acids at both the N- and C-termini in class I molecules (Madden et al., 1991), whereas they are open at both ends in class II molecules (Brown et al., 1993). Therefore, MHC class I molecules bind peptides of 8 to 10 residues in length, whereas MHC class II molecules can adopt much longer peptides that extend the binding groove (Rammensee, 1995; Rammensee et al., 1995; Rajnavölgyi et al., 1997; Gogolák et al., 2000).

Stabilization centers (SCs), defined as certain clusters of residues involved in cooperative long-range interactions, were described as being primarily responsible for keeping the three-dimensional structure of a protein intact (Dosztányi et al., 1997). Residues involved in SCs, i.e., elements of SCs, are defined by considering the contact maps of proteins of known structure. Residues are considered to be in contact if at least one of their heavy atom distances is less than the sum of the van der Waals radii of the two atoms plus 1.0 Å. Two residues are considered to be in long-range contact if they are at least 10 residues apart in the amino acid sequence or if they belong to separate polypeptide chains of the protein. To identify SC elements, we looked for supporting residues from the flanking tetrapeptides on both sides of the two residues that are involved in long-range interactions. There are 44 = 256 ways to select these two pairs of supporting residues. One central residue and its two selected supporting residues (one on each side) are together called a residue triplet. If there is at least one of the 256 cases in which the two triplets form at least seven interresidue interactions (out of the nine theoretically possible), the central residues of both triplets are considered to be SC elements (Dosztányi et al., 1997). Properties of SCs are discussed in more detail elsewhere (Dosztányi et al., 1997; Dosztányi and Simon, 1999), and a public server is available at http://www.enzim.hu/scpred/scpred.html to identify SC elements of a given protein of known structure and to predict such elements when only the amino acid sequence of the protein is available.

According to our earlier study on a 600-member representative set of unrelated proteins of the Protein Data Bank (PDB), SCs were found in all proteins, and almost 25% of the residues were estimated to be SC elements. This percentage was rather variable in proteins that belonged to different secondary structure subclasses. For example, in all-beta proteins SCs were abundant, whereas in the all-alpha subclass they were rather rare. This suggested that the stability of a certain structure requires a certain number of SC elements. Comparison of proteins of the same function in thermophilic microorganisms and in mammals showed that as few as one or two additional SC elements in a protein results in a denaturation temperature that is higher by 30-50°C. Apparently, protein stability can be adjusted significantly by introducing or removing a few SC elements (Dosztányi, Z., Szirtes, G., Magyar, Cs., Simon, I., manuscript in preparation).

In this work, the stabilization centers of peptide-complexed MHC class I and class II proteins were analyzed. Our studies revealed that residues of the bound peptide could be involved in SCs and thus may contribute to stability of the MHC molecule.


    DATABASE
TOP
ABSTRACT
INTRODUCTION
DATABASE
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES

Peptide binding domains of eight MHC class I and seven MHC class II proteins, complexed with one or several peptides, were studied. The PDB codes, the MHC allotype, and the origin of the bound peptides are listed in Tables 1 and 2.


                              
View this table:
[in this window]
[in a new window]
 
TABLE 1   MHC I-peptide complexes


                              
View this table:
[in this window]
[in a new window]
 
TABLE 2   MHC II-peptide complexes


    RESULTS AND DISCUSSION
TOP
ABSTRACT
INTRODUCTION
DATABASE
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES

Stabilization centers that contain an SC element, located in the helices that form the sides of the peptide binding grooves, in all of the 31 complexes are listed in Tables 3-6. Representative examples, shown in Fig. 1, include the MHC class I molecules HLA-B*5301 complexed with the HIV-2 gag (182-190) peptide and HLA-B*0801 complexed with the HIV-1 gag (24-31) peptide, as well as the MHC class II molecule HLA-DR1 (composed of HLA-A*0101, HLA-B1*0101) complexed with the human influenza A virus hemagglutinin 306-318 peptide.


                              
View this table:
[in this window]
[in a new window]
 
TABLE 3   Intramolecular stabilization center elements of MHC class I molecules


                              
View this table:
[in this window]
[in a new window]
 
TABLE 4   Intramolecular stabilization center elements of MHC class II molecules


                              
View this table:
[in this window]
[in a new window]
 
TABLE 5   Stabilization center elements in MHCI-peptide complexes connecting the helices of the peptide binding groove to the peptide ligand


                              
View this table:
[in this window]
[in a new window]
 
TABLE 6   Stabilization center elements in MHCII-peptide complexes connecting the helices of the peptide binding groove to the peptide ligand



View larger version (27K):
[in this window]
[in a new window]
 
FIGURE 1   Stabilization centers in peptide binding domains of MHC-peptide complexes. (A) MHC I-peptide complex, PDB code 1a1m. (B) MHC I-peptide complex, PDB code 1agd. (C) MHC II-peptide complex, PDB code 1dlh. Stabilization centers that involve residues from the bound peptide are marked in black. The rest of the stabilization centers are marked in dark grey. Sequence numbers of residues of the helices appearing as stabilization center element in any of the class I and class II complexes are shown in A and C, respectively.

Fig. 1 clearly demonstrates that the eight-stranded beta -sheets are stabilized by a large number of SCs; an average of 27 residues are involved in these interactions. On the other hand, the number of SCs connecting the helices to the eight-stranded beta -sheet plateau is rather limited; these SC element pairs are listed in Tables 3 and 4. There are only a few cases where both helices are linked to the beta -sheet by SCs. This finding is in good agreement with our previous results obtained on a large data set (Dosztányi and Simon, 1999), which indicated that in the alpha /beta subclass of proteins, which include the MHC proteins, 46.5% of the SC elements connect one extended chain to another, whereas only 1.5% of the SC elements connect an extended chain to a helix.

At least one residue of the bound peptide is involved in a SC in every one of the class II protein-peptide complexes analyzed, and also in most of the class I protein-peptide complexes studied (see Tables 5 and 6). In all these cases the complementary residues of the SC elements in the bound peptides were localized in the helices of the MHC molecule and never in the beta -sheet. Because the accommodated peptides adopt an extended conformation in the MHC class I groove (Madden et al., 1991) and a polyproline II-like conformation in the MHC II groove (Jardetzky et al., 1996), the relatively large number of these helix-extended chain-associated SC connections suggests significant biological functions of these connections. Let us consider the two MHC classes separately.

Stabilization centers of MHC class II-peptide complexes

In MHC class II molecules, some of the SC elements of the peptides are identical to the so-called anchor residues, such as those located in relative positions 1 and 9 of the core sequence. These are characterized as strongly interacting peptide residues. When the anchors were not SC elements themselves we often found them as supporting residues in the flanking tetrapeptides of the SC element. Some SC elements are located in the N-terminal flanking region of the peptide, which extends the binding groove of the protein. It is noteworthy that N-terminal elongation of peptide determinants beyond the first primary anchor was recently shown to improve peptide binding to MHC class II molecules (Bartnes et al., 1999).

The stability of MHC class II heterodimers as well as superdimers (Schafer et al., 1998) also found in the cell surface depends on the bound peptide (Germain and Margulies, 1993; Germain and Rinker, 1993), although apparently to a lesser extent than for class I molecules (Stern and Wiley, 1992; Reich et al., 1997). Peptides facilitate the assembly of the two chains in the endoplasmic reticulum, and determine their longevity in endosomal compartments and at the cell surface (Germain, 1995). Hydrogen bonds between conserved amino acids of MHC class II molecules and the backbone of the peptide make profound contributions to stabilization of the complex and to increasing enzyme susceptibility (Ceman et al., 1998). Stability of MHC class II-peptide complexes can be so delicately balanced that disruption of a single solvent-exposed hydrogen bond (such as substitution of histidine by asparagine at position beta 81) can lead to defects in complex assembly and speed up dissociation of the peptide, which is the triggering step in protein degradation (McFarland et al., 1999). It is noteworthy, that the conserved beta 81 histidine is located between residue beta 80 which forms an SC with another protein residue and beta 82 which forms an SC with peptide residues in many MHC class II proteins (see Tables 3 and 6). His beta 81 was found to be a supporting residue in all these cases. However, not all residues play the same role in stabilization, since, for example, loss of intramolecular interactions such as the salt bridge between alpha 76 arginine and beta 57 aspartic acid in the resistant allotypes against insulin-dependent diabetes mellitus, can initiate local rearrangement of the peptide binding groove without altering global tertiary structure (Sato et al., 1999).

Empty MHC class II molecules were shown to have two peptide-free isomers; the active, peptide-receptive form is unstable and rapidly converts to an inactive one (Rabinowitz et al., 1998), which aggregates, becomes sensitive to proteases, and is degraded in endosomes (Germain and Rinker 1993). Under in vivo conditions, the folding, intracellular transport and continuous groove occupancy of MHC class II molecules is directed by the invariant chain (Ii), which acts as a chaperone for newly synthesized alpha beta dimers, targets them to the endosomal compartments, and retards them from the cell surface in an inactive state. Removal of Ii is mediated by multistep proteolytic degradation, which results in the generation of class II associated invariant chain peptide (CLIP), peptides covering the 81-104 sequence of Ii, which bind to various MHC class II molecules and occupy the peptide binding groove until it is exchanged by a self- or an antigenic peptide. This process takes place in acidic MHC-rich compartments and only rarely on the cell surface, and it is catalyzed by the MHC-like HLA-DM/H-2M chaperone molecule (Koopmann et al., 1997). A large number of peptidic SC-elements were found in the CLIP peptide complexed with the HLA-DR3 molecule that is expressed on the surface of HLA-DM deficient cells (Table 6; Ghosh et al., 1995).

In their immature differentiation state, dendritic cells, the most potent of the professional antigen presenting cells, express a large fraction of empty MHC class II molecules. These are in a peptide-receptive state, bind exogenous protein fragments, and present them for T cells (Santambrogio et al., 1999). Immature dendritic cells also express MHC-Ii complexes, which traffic to the cell surface before internalization into the endosomal peptide-loading compartment. Peptide uptake by these unusual MHC class II molecules is assisted by functional HLA-DM molecules expressed on the cell surface (Arndt et al., 2000). Thus they acquire a novel extracellular receptor function, which broadens the spectrum of peptides captured and saved by MHC class II molecules of dendritic cells, which in turn play a pivotal role in priming immune responses.

Stabilization centers of MHC class I-peptide complexes

In five out of the 21 MHC class I-peptide complexes, no peptide-protein SC could be identified. In these five cases, one of the helices was not stabilized by any SC. Binding experiments with various peptides indicated that terminal residues of the peptide, which interacted not only with residues of the helices but also with conserved amino acids, at the closed ends of the MHC class I groove, contributed significantly to the binding energy (Madden, 1995). These promising candidates could not be recognized as SC elements because of the special definition of SC that requires flanking segments to occur on both sides of an SC element (Dosztányi et al., 1997). Based on our earlier study of a large number of proteins, SC elements made 1.77 times more long range interatomic interactions than did those residues involved in long range interactions that were not SC elements (Dosztányi et al., 1997). Therefore, the extremely large number of inter-atomic interactions identifies SC element-like residues. In all 31 complexes studied, all 323 residues of the bound peptides had interactions with MHC-derived atoms; therefore, all of them were involved in long range interactions. Fifty-two were SC elements, and 42 were terminal residues of the bound peptides. A comparison of the numbers of long-range atomic contacts shows that the average number is larger for residues, at the termini of bound peptides than for the same type of SC residues and almost twice that of the rest of the residues (Table 7). This suggests that although the terminal residues in question can not be recognized as SC elements, their interactions with the MHC class I protein are so pronounced that in fact they play the same role as do SC elements in structure stabilization, so they could be called quasi-SC elements.


                              
View this table:
[in this window]
[in a new window]
 
TABLE 7   Average number of stabilization centers and long range atomic contacts in globular proteins, in MHC molecules and of peptide termini in MHC I complexes

Differences in serology and thermal lability indicated that MHC class I molecules lacking peptide have a structure significantly different from those complexed with peptide ligands. Empty MHC class I molecules possess partially folded alpha 1 and alpha 2 domains and many of their physical properties are typical of molten globules (Bouvier et al., 1998). A small proportion (3-5%) of native alpha -chains was detected on the surface of T cells (Edidin et al., 1997), but empty MHC class I molecules can be transported to the cell surface only at sub-physiological temperatures (Ljunggren et al., 1990). These results demonstrate that the MHC class I ligand binding site is inherently unstable in the absence of peptide. To select peptide under in vivo conditions, the highly polymorphic MHC class I alpha -chains interact with beta 2-microglobulin. Analyzing SC elements in the peptide binding domains we found that the polymorphic alpha 76 residue and the tryptophan at position 60 in beta 2-microglobulin form an SC in almost all studied MHC-peptide complexes except one of the HLA-A2 complexes (Table 1, no.2).

Intracellular loading of MHC class I molecules is also assisted by chaperones, which maintain the fragile peptide-binding sites in a peptide-receptive conformation, and retain the class I molecules in the endoplasmic reticulum until peptide binding occurs. MHC class I molecules, transported to the cell surface, retain their folded peptide binding site and are stable for hours or even days (Micheletti et al., 1999). However, some MHC class I molecules, such as the murine H-2Ld protein, exhibit weak interaction with peptides and with beta 2-microglobulin (Balendiran et al., 1997). As a consequence, their peptide loading within the cell is insufficient and results in low level cell surface expression; the Ld molecules have an unusually short half-life and are able to exchange their peptides for exogenous ligands (Hansen et al., 2000). In accord with these data, our analysis failed to detect SCs in the alpha -helices of the H-2Ld molecule (Table 3), but SC elements were detected in the peptide ligand (Table 5). Thus, MHC class I molecules, which display quantitative differences in their interactions with peptides and other molecules involved in intracellular peptide loading, may also acquire novel functions that provide alternative pathways for antigen presentation (Jondal et al., 1996; Khare et al., 1996; Hansen et al., 2000).

Our survey suggests that SCs and quasi-SCs composed of residues derived from both the protein and the complexed peptide can be vital for ensuring the compact structure of MHC proteins by fixing their alpha -helices. These highly ordered structures play a pivotal role in stabilizing the molecular surface which contacts the T cell antigen receptor and also support a stable MHC conformation resistant to proteolytic attack (Willcox et al., 1999).


    CONCLUSION
TOP
ABSTRACT
INTRODUCTION
DATABASE
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES

We suggest a new view of function-related regulation of the stability of MHC proteins by SCs, which were analyzed in 31 MHC-peptide complexes of known three-dimensional structure. The primary biological function of MHC proteins is to bind intracellular peptides, generated by limited proteolysis and loaded onto nascent or recycling MHC class I or class II proteins inside the cell, to transport them to the cell surface and hold them there until they are recognized by T cells. To accomplish this job MHC molecules must acquire a stable conformation. Although MHC molecules can bind a wide array of peptides, almost every one of them carries only one ligand to the cell surface in its lifetime; only a few molecules lose and exchange their ligands at the cell surface or after their recycling to endosomes. The peptide binds to the protein noncovalently; without the dissociable peptide ligand, the MHC molecule loses its stable structure very rapidly and becomes sensitive to proteolytic degradation.

We suggest in this paper that regulation of protein stability is based on the formation of SCs composed of residues from both the peptide binding groove of the MHC protein and the small dissociable ligand. These SCs stabilize the helices bordering the peptide binding groove and the remainder of the MHC protein. Thus, these interactions may also be essential for capturing the peptide and so generating a novel molecular surface, recognizable by the T cell receptor (Willcox et al., 1999). Because these SCs are missing in the ligand-free protein, the empty MHC molecules are fragile. The lifetime of MHC class I-peptide complexes at the cell surface determines the efficiency of cytotoxic T lymphocytes, which play a pivotal role in virus- and tumor-specific responses. Stable MHC class II-peptide complexes trigger helper T cell responses, which regulate immune responses against foreign and self-antigens and may cause autoimmune diseases. Ligand-free MHC molecules are immunologically incompetent and have a short half life, but they may acquire novel biological functions or become a source of amino acids for de novo protein synthesis.

    ACKNOWLEDGMENTS

This work was supported by grants AKP 98-13 3,3 (to I. S. and É. R.), OTKA T017157 (to É. R.), OTKA T030566 (to I. S.), and FKFP 0186/1999 (to É. R.) by the Hungarian-Israeli Intergovernmental S&T Cooperation Programmes and by the scientific collaboration project between the Hungarian Academy of Sciences and the U.S. National Science Foundation.

We thank Csaba Magyar and Gábor E. Tusnády for their help in programming and Nicholas E. Dixon for his critical comments on the manuscript.

    FOOTNOTES

Received for publication 1 May 2000 and in final form 31 July 2000.

Address reprint requests to I. Simon, Institute of Enzymology, Hungarian Academy of Sciences, H-1518 Budapest, P.O. Box 7, Hungary. Tel.: 36-1-466-9276; Fax: 36-1-466-5465; E-mail: simon{at}enzim.hu.

É. Rajnavölgyi's present address: Institute of Immunology, Faculty of Medicine, Medical and Health Science Center, University of Debrecen, 98 Nagyerdei krt. Debrecen H-4012, Hungary.


    REFERENCES
TOP
ABSTRACT
INTRODUCTION
DATABASE
RESULTS AND DISCUSSION
CONCLUSION
REFERENCES

Biophys J, November 2000, p. 2305-2313, Vol. 79, No. 5
© 2000 by the Biophysical Society   0006-3495/00/11/2305/09  $2.00



This article has been cited by other articles:


Home page
Nucleic Acids ResHome page
Z. Dosztanyi, C. Magyar, G. E. Tusnady, M. Cserzo, A. Fiser, and I. Simon
Servers for sequence-structure relationship analysis and prediction
Nucleic Acids Res., July 1, 2003; 31(13): 3359 - 3363.
[Abstract] [Full Text] [PDF]


Home page
J. Biol. Chem.Home page
K. Saito, A. Sarai, M. Oda, T. Azuma, and H. Kozono
Thermodynamic Analysis of the Increased Stability of Major Histocompatibility Complex Class II Molecule I-Ek Complexed with an Antigenic Peptide at an Acidic pH
J. Biol. Chem., April 18, 2003; 278(17): 14732 - 14738.
[Abstract] [Full Text] [PDF]


This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Simon, A.
Right arrow Articles by Simon, I.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Simon, A.
Right arrow Articles by Simon, I.


HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Copyright © 2000 by the Biophysical Society.