3 - MAP

Transcription

3 - MAP
&CHAPTER 2
Antibody Molecular Structure
ROBYN L. STANFIELD and IAN A. WILSON
2.1 Introduction
51
2.2 General Structural Features
2.3 Canonical Conformations
52
56
2.4 Fab Conformational Changes
56
2.5 Human Anti-HIV-1 Antibodies
2.6 Shark and Camel Antibodies
58
61
2.7 Summary
63
Acknowledgments
63
References
63
ABSTRACT
The structural features of antibodies have been studied extensively over the years by many techniques,
including electron microscopy, x-ray crystallography, and NMR. Consequently, a wealth of structural
information is available for antibodies, alone and in complex with antigens ranging in size from small
haptens to whole viruses. The knowledge gained from these studies has greatly facilitated the engineering of antibodies for use as human therapeutics.
2.1 INTRODUCTION
Antibodies are the key component of the humoral adaptive immune response against foreign pathogens. Their enormous sequence and structural diversity allows for recognition of any foreign antigen
imaginable, with high affinity and specificity. While antibodies have been best studied in mice and
humans, they are also found in animals as evolutionarily distant as the cartilaginous fish, such as
sharks, skates, and rays. An abundance of structural information has accumulated for antibodies,
including over 800 crystal structures that are predominantly Fab, Fab 0 , Fv, or occasionally VH
fragments, many in complex with antigens ranging in size from small haptens, to peptide and DNA
fragments, to proteins. In addition, combined cryo-electron microscope and crystallographic studies
have revealed how Fab and IgG molecules interact with intact viruses. This structural information
Therapeutic Monoclonal Antibodies: From the Bench to the Clinic. Edited by Zhiqiang An
Copyright # 2009 John Wiley & Sons, Inc.
51
52
ANTIBODY MOLECULAR STRUCTURE
has been of great value for our understanding of the antibody-antigen recognition process, and also for
development of antibodies as human therapeutics.
2.2 GENERAL STRUCTURAL FEATURES
Mammalian antibodies are constructed from two types of protein sequences, called the heavy and light
chains. The five classes of antibodies in humans and other placental mammals differ in their heavy
chain sequences, with heavy chain types m, d, g, 1, and a found in IgM, IgD, IgG, IgE, and IgA antibodies, respectively. Each heavy chain can pair with one of two types of light chain, called l or k.
Most antibodies are made up of four chains; two copies of the heavy chain and two copies of the
light chain (Fig. 2.1). However, the IgM class can also exist as a pentamer of these four chains,
with a resultant 10 heavy and 10 light chains, and IgA can exist in a form called secretory IgA,
where an additional J chain stabilizes the dimerization of two antibodies, to give a total of four
heavy and four light chains.
The majority of antibodies found in the serum belong to the IgG class, and most structural information has been derived for this class of antibody; thus, most of our discussion will deal with IgG antibodies. An intact IgG molecule has two heavy (55,000 Da each) and two light chains (24,000 Da
each) that fold into three large domains: two Fab fragments (one light and the N-terminal half of a
heavy chain) and one Fc fragment (two C-terminal heavy chain halves) (Fig. 2.1). Fab and Fc are
abbreviations for fragment-antibody binding and fragment-crystallizable, so called because the Fab
fragment binds antigen, and the first Fc fragments studied were easy to crystallize, although that clearly
now is somewhat of a misnomer. The overall shape of the IgG can be described as a Y, with the Fc
fragment forming the base of the Y, and the two Fab fragments forming the two arms (Fig. 2.2).
Figure 2.1 IgG light and heavy chains. The light chains are shown in light gray, and the heavy chains in dark gray,
with the different Ig domains indicated.
2.2
GENERAL STRUCTURAL FEATURES
53
Figure 2.2 IgG domain organization. The Fc and two Fab domains form an overall Y shape.
Figure 2.3 Structure of IgG b12 determined by x-ray crystallography. Residues in the linker region of the heavy
chain that have no visible electron density are shown as a dotted line. Carbohydrate within the Fc domain is shown in
a gray CPK representation. Rather than a rigid Y shape as depicted in Figure 2.2, in a real IgG, the Fab arms are very
flexible with respect to the Fc domain, allowing for greater flexibility in binding antigen. (See color insert.)
54
ANTIBODY MOLECULAR STRUCTURE
Figure 2.4 Fab fragment. (Left) The light chain is on the left in light gray, and the heavy chain on the right in
darker gray. The CDR loops that contact antigen are labeled. (Right) Looking down onto the Fab antigen-binding
site. This view is the same as on the left, but rotated 90 degrees about a horizontal axis. (See color insert.)
Figure 2.5 An IgG Fc domain. The two heavy chains are colored dark and light gray, and the carbohydrate found
at the center of the Fc in an IgG is shown in a ball-and-stick representation. (See color insert.)
2.2
GENERAL STRUCTURAL FEATURES
55
However, highly flexible hinge regions join the Fc to the two Fabs, and electron microscopy (Roux,
Strelets, and Michaelsen 1997) and crystal structures of IgGs with intact hinge regions (Harris et al.
1992, 1997; Saphire et al. 2001) show that the Fabs have a wide range of flexibility with respect to
the Fc (Fig. 2.3). The Fab fragment can be further subdivided into smaller domains called the variable
domain (the N-terminal half of the Fab; VL and VH) and constant domain (the C-terminal half of the
Fab; CL and CH1) (Fig. 2.4). As suggested by their names, the variable domain contains regions with
high sequence diversity, while the constant domain sequences are highly conserved. The Fc fragment
also has two domains, formed by dimerization of CH2 –CH2 and CH3–CH3 regions (Fig. 2.5).
The final, smallest level of domain architecture is the immunoglobulin or Ig domain. Each individual VL, VH, CL, CH1, CH2, or CH3 domain is an Ig domain, consisting of a characteristic Ig fold with
either seven or nine b strands, in two b sheets, that form a Greek-key b-barrel, with a highly conserved
disulfide connecting the two sheets at the core of the barrel (Fig. 2.6). A constant-type Ig domain has
seven strands, three in one sheet and four in the other, while a variable-type Ig domain adds two extra
strands (C0 and C00 ) to the three-stranded sheet, so that the barrel is composed of five- and four-stranded
b-sheets (Fig. 2.6). While first identified in immunoglobulins, the Ig fold is found in many other types
of molecules and, in addition to the variable- and constant-type Ig domain, other different but related
types of Ig domains exist, some without the canonical disulfide bridge (Bork, Holm, and Sander 1994;
Halaby, Poupon, and Mornon 1999).
Figure 2.6 A constant-type Ig domain (left) and a variable-type Ig domain (right). The variable domain has two
extra strands (C0 and C00 ) that form the CDR2 loop. This domain is from a light chain; however, a heavy chain variable domain has the same domain architecture. (See color insert.)
56
ANTIBODY MOLECULAR STRUCTURE
The Fab fragment recognizes antigen at a site formed by six linear polypeptide segments called
complementarity determining region (CDR) loops. Three CDRs are contributed by the VL (L1, L2,
and L3) and three by the VH domain (H1, H2, and H3) (Fig. 2.4). These loops are hypervariable in
sequence, and were predicted to be the interaction site for antigen recognition long before any structural information existed for the Fab fragment (Wu and Kabat 1970; Kabat and Wu 1971).
However, the structure of the first Fab-hapten complex (McPC603-phosphocholine) determined by
x-ray crystallography in 1973 (Padlan et al. 1973; Segal et al. 1974) verified that the antigen interacts
mainly with the CDR loop regions of the Fab. While recognition of protein antigens by Fab generally,
but not always, utilizes all six of the CDR loops, recognition of smaller antigens may use a more limited
set of these CDRs; in fact, some isolated VH fragments containing CDRs H1-3 are sufficient
for recognition of their antigen (Davies and Riechmann 1996). Typical VH-VL antibody-antigen interface sizes (the combined buried molecular surface area for both the antigen and antibody) range from
´
´
´
around 300 –600 Å2 for small haptens, to 800–900 Å2 for peptides, to 1500– 1800 Å2 for proteins
(Wilson et al. 1991).
2.3 CANONICAL CONFORMATIONS
While the CDR loops are the most variable part of the Fab (in both their sequence and structure), five of
the six loops have been shown to have a limited number of conformations, termed canonical structures
(Chothia and Lesk 1987). Based on their sequence, and a set of rules derived from known structures,
the conformations of these loops can usually be predicted with reasonable accuracy. However, CDR H3
has been far more difficult to classify. H3 is the most variable of the CDR loops in sequence and length,
and structures of many Fab fragments have shown that H3 is also the most structurally variable and
flexible of the CDRs. The base of CDR H3 can usually be found in one of two different conformations
that can often be predicted from the sequence (Shirai, Kidera, and Nakamura 1996; Al-Lazikani, Lesk,
and Chothia 1997; Koliasnikov et al. 2006), but the remainder of the loop has no conserved structure, is
frequently seen to undergo large changes in conformation upon binding of antigen, and is often poorly
ordered in crystal structures of unliganded Fabs. More recent studies have emphasized important
functional differences in the lengths of the CDR H3s in humans versus mice that can be exploited
for antigen recognition (Collis, Brouwer, and Martin 2003).
2.4 Fab CONFORMATIONAL CHANGES
In addition to hinge flexibility between the Fab and Fc fragments, flexibility also occurs between the
domains that make up the Fab fragment. The variable and constant domains can move with respect with
each other around what is termed the elbow angle. This angle has been seen to vary between 117
degrees and 227 degrees (Stanfield et al. 2006), with larger values seen more often with antibodies
that contain l light chains (Stanfield et al. 2006). The VL and VH domains can also rotate with respect
to each other by as much as 16 degrees (Stanfield et al. 1993), and these movements can serve to
dramatically reconfigure the antigen binding site.
Currently about 100 Fabs have been determined in both the unliganded and ligand-bound form.
Comparisons of these liganded and ligand-free structures have shown that conformational changes
often occur in the antigen-binding site to enable greater complementarity of fit to the antigen
(Stanfield and Wilson 1994; Wilson and Stanfield 1994). These changes usually take place in the
CDR loops and, as mentioned earlier, the largest changes are often seen in CDR H3. While often
termed induced-fit binding, studies of antibodies in solution indicate that the conformational changes
are probably not completely induced by binding of the antigen, but rather the antigen may in some
cases bind to one of several preexisting antibody conformations found in solution, where small structural changes then allow the antibody to adapt and custom-fit its binding site to the antigen in question
(Foote and Milstein 1994; James, Roversi, and Tawfik 2003).
2.4 Fab CONFORMATIONAL CHANGES
57
Conformational changes in antibodies include changes in side-chain rotamers, movements of the
CDR loops as either a rigid unit or by more extensive structural rearrangements, and through changes
in the relative disposition of their VH/VL domains. An excellent illustration of a key side-chain
rearrangement is found in the anti-progesterone antibody DB3 (Arevalo et al. 1993; Arevalo,
Taussig, and Wilson 1993; Arevalo et al. 1994). When the unliganded and steroid-bound structures
Figure 2.7 Examples of conformational changes in Fabs after binding antigen. Top left: The anti-progesterone
Fab DB3 in its unliganded form, with TrpH100 filling its antigen-binding pocket. Top right: DB3 with progesterone
(ball-and-stick) bound, and TrpH100 moved away from its position in the unliganded Fab. (Middle left) The antiHIV-1 Fab 50.1 in its unliganded form. Middle right: 50.1 with bound peptide (ball-and-stick). The H3 and H1
CDRs are labeled. H3 undergoes a structural rearrangement, while H1 moves away from the binding site while maintaining its overall shape. Bottom left: The unliganded anti-HIV-1 Fab X5, with its long CDR H3 labeled. Bottom
right: Fab X5 in its gp120-bound conformation, with H3 labeled. A large conformational change occurs in the CDR
H3 between the unliganded and liganded forms. (See color insert.)
58
ANTIBODY MOLECULAR STRUCTURE
for DB3 are compared, a substantial conformational change in the position of the TrpH100 side chain is
observed (Fig. 2.7). Interestingly, this side chain fills the antigen-binding site in the unliganded Fab
structure, and then moves out of the way to allow binding of steroid (Fig. 2.7). Thus, TrpH100 acts
as a surrogate ligand for the antibody in the absence of steroid. One of the most extensive conformational changes has been seen for antibody 50.1, an anti-HIV-1 neutralizing antibody that recognizes the
HIV-1 gp120 V3 loop (Rini et al. 1993; Stanfield et al. 1993). A structural comparison of the unliganded and V3 peptide-bound Fab showed that the H3 CDR loop undergoes a large structural
´
rearrangement, with the side chain of Tyr H97 moving by about 6 Å. The CDR H1 also moves, in a
rigid body or segmental fashion, with a root-mean square deviation for the H1 main-chain atoms of
´
1.1 to 1.4 Å (for multiple copies of both the native and peptide-bound Fab in the crystals). In addition,
the VL and VH domains of 50.1 move with respect to each other by about 16 degrees (Fig. 2.7). The
combination of loop rearrangement and domain rotation result in a drastically differently shaped and
configured binding pocket for the peptide ligand (Fig. 2.7). Another example of an extremely large
H3 CDR movement is seen in crystal structures of the anti-HIV-1 antibody X5 that recognizes
the recessed CD4-binding site of gp120. The long X5 CDR loop undergoes an extensive rearrangement
´
and moves by as much as 17 Å (Darbha et al. 2004; Huang et al. 2005) between the unbound and gp120
bound states (Fig. 2.7). Unusually long H3 CDR loops are not uncommon in human antiviral antibodies, and it has been proposed that long CDR loops may help antibodies target recessed clefts in
viral antigens (Burton et al. 2005).
2.5 HUMAN ANTI-HIV-1 ANTIBODIES
While most of the early crystal structures for antibodies were obtained for human myeloma antibodies,
the advent of the monoclonal antibody technology made it very easy to produce large amounts of
mouse monoclonal antibodies to known antigens. Thus, many more crystal structures have been determined for antibody fragments from mice than from humans. However, subsequent developments, such
as Epstein-Barr virus immortalized human B-cells (Steinitz et al. 1977) and human-mice heterohybridomas (Cole et al. 1984), production of human antibodies and antibody fragments in bacteria (Skerra
and Pluckthun 1988), and phage-display of antibody fragments (Huse et al. 1989), have made it possible to produce human monoclonal antibodies against known antigens. In addition, techniques, such
as humanization of mouse or rat antibodies (Jones et al. 1986) or the production of human/mouse
chimeric antibodies (Boulianne, Hozumi, and Shulman 1984; Morrison et al. 1984; Neuberger et al.
1985; Better et al. 1988), have been made possible by now standard molecular biology techniques.
Currently about 40 fully human Fab or Fv fragments have crystal structures deposited in the PDB, Q1
although many of these are for the same fragment with multiple, related antigens.
Many of the human antibodies that have been studied by x-ray crystallography are against viral antigens, including the HIV-1 virus. The HIV-1 virus is able to rapidly evolve to evade the host immune
system, but only a handful of antibodies have been discovered that are able to effectively neutralize a
wide variety of the different strains of the virus (Burton, Stanfield, and Wilson 2005). A successful
vaccine would induce similar antibodies that are able to neutralize any strain of the virus that an individual might encounter. Structures of these rare antibodies have shown that the antibodies outwit the
virus by using novel structural features in their mechanisms of antigen recognition (Burton, Stanfield,
and Wilson 2005).
One of the most interesting of the anti-HIV-1 antibodies is 2G12. 2G12 was isolated from an HIV-1
infected patient, and is one of the most potent, broadly neutralizing antibodies known. While initial
studies indicated that 2G12 recognized a carbohydrate epitope on the HIV-1 surface (Sanders et al.
2002; Scanlan et al. 2002), it was difficult to understand how the antibody might recognize carbohydrate with high affinity. The carbohydrate on the HIV-1 viral surface is transferred onto the viral
coat proteins by the human host cellular machinery. Thus, these sugars should appear as self to our
immune system and, hence, give rise to tolerance and not lead to a strong antigenic response. Such
2.5
HUMAN ANTI-HIV-1 ANTIBODIES
59
anti-carbohydrate antibodies should be exceedingly rare and, as for other anti-carbohydrate antibodies
against non-self sugars, be of low affinity. The crystal structure of the 2G12 Fab fragment was, therefore, a huge surprise (Calarese et al. 2003), and revealed the Fab had dimerized via a domain swap of
its VH domains (Fig. 2.8). This domain-swapped dimer has two closely spaced antigen-binding sites
and a potential third binding region at the interface of the two newly associated VH domains (VHVH0 ). These intertwined Fab regions give rise to the unusual linear shape of the intact IgG that differs
from the more typical Y configuration seen for other IgGs. Electron microscope (EM) studies have also
clearly shown this unusual linear structure in the intact 2G12 IgG (Roux et al. 2004), and analytical
ultracentrifugation also confirmed that the Fab is an obligate dimer in solution (Calarese et al.
2003). Crystal structures have been determined for the dimeric 2G12 Fab in complex with carbohydrates (Calarese et al. 2003, 2005) ranging in size from Mana1,2Man to Man9GlcNac2, and these
structures indicate that the Fab can bind to the terminal Mana1-2Man in either the D1 or D3 arms
of Man9GlcNac2. Mutagenesis and modeling studies of 2G12 have pointed to a conserved cluster of
high mannose moieties, especially those linked to gp120 residues 332, 339, and 392, as being the
likely epitope for this antibody (Sanders et al. 2002; Scanlan et al. 2002; Calarese et al. 2003). In
addition, glycan array and solution-phase ELISA analyses have helped to define the carbohydrate
specificity of 2G12 (Blixt et al. 2004; Bryan et al. 2004; Calarese et al. 2005). Modeling studies
show that the distances between these groups on gp120 carbohydrates are compatible with the distance
between the two antigen-binding sites on the dimeric Fab (Calarese et al. 2003) and that a third, novel
antigen-binding site may exist at the VH-VH0 interface. The inherent high avidity (nM) of the dimeric
2G12 has thus given the antibody high affinity for carbohydrate on the virus due to the multivalency.
Analysis of the 2G12 sequence uncovers several unusual residues that may favor the domain swap
event. One rare residue is ProH113, located in the linker between VL and VH. Other unusual residues
include the hydrophobic IleH19 and PheH77 that may help to stabilize the novel VH-VH0 interface,
Figure 2.8 The broadly neutralizing, anti-HIV-1 Fab 2G12. 2G12 forms an unusual domain swapped dimer, with
the VH domains from each heavy chain (labeled VH-VH0 ) swapping to the other Fab, resulting in a tightly linked
dimer of Fabs. Bound Man9GlcNac2 is shown in a CPK representation. (See color insert.)
60
ANTIBODY MOLECULAR STRUCTURE
whereas an Arg at position H39 disrupts a highly conserved GlnL38-GlnH39 interaction that is found in
all antibodies, as well as in T cell receptors.
Two other unusual anti-HIV-1 antibodies that recognize neighboring epitopes are 2F5 and 4E10.
Both of these antibodies recognize epitopes in the membrane proximal epitope region (MPER) of
gp41. The MPER region, as its name implies, is proximal to the membrane spanning region of
gp41, and is thought to be only transiently exposed during the viral-cell fusion process. The 2F5
and 4E10 Fab structures include complexes with gp41 peptides (Ofek et al. 2004; Cardoso et al.
2005; Cardoso et al. 2007), with the 2F5 peptide forming a b-turn, and the 4E10 peptide forming
an a-helix. Estimations of the distances of these gp41 epitopes to the membrane spanning regions
of gp41 indicate that the Fabs must come very close to or even contact the host cell membrane in
order to bind these epitopes. Interestingly, the Fab CDR residues surrounding the peptide-binding
site include a large number of hydrophobic residues, including a number of Trp and Gly residues in
4E10 that may facilitate interaction with the membrane.
An interesting class of anti-HIV-1 antibodies has been found that only bind to gp120 after it binds
to its primary receptor, CD4. Thus, these antibodies are called CD4-induced, or CD4i. The binding
of CD4 to gp120 induces conformational changes in the gp120 that are necessary for binding to the
co-receptor (usually CCR5 or CXCR4) and, hence, required for viral entry into the cell. The CCR5
co-receptor has sulfated tyrosine residues in its N terminus, and this N-terminal region is proposed
to interact with gp120 during the binding process. Surprisingly, some of the CD4i antibodies were
also found to have acquired sulfated tyrosine residues in their highly acidic CDR H3 loops through
posttranslational modification, suggesting that they may be structural mimics of the co-receptor
(Choe et al. 2003; Huang et al. 2004, 2007). A recent structure of gp120 in complex with CD4 and
sulfated antibody 412D (Fig. 2.9) shows how these two sulfated tyrosine residues in the CDR
Figure 2.9 The anti-HIV-1 Fab 412D has sulfated tyrosine residues in CDR H3. Left: The complex of Fab 412d
with CD4 and gp120. Right: enlargement of the sulfated tyrosine residues and their interaction with gp120. It is
thought that the gp120 co-receptor (CCR5) binds to the bridging sheet and V3 regions of the gp120 molecule.
CCR5 has four sulfated tyrosine residues in its N-terminal region, and at least two of these are thought to take
part in the interaction with gp120. Several antibodies that recognize the same region of gp120 have also evolved
to have sulfated tyrosine residues. (See color insert.)
2.6
SHARK AND CAMEL ANTIBODIES
61
Figure 2.10 Broadly neutralizing anti-HIV-1 antibody b12 interactions with gp120. Left: The CD4 (tubes) binding site on gp120 (solid surface) is shown. Right: b12 binds to the CD4-binding site, accessing a deep cleft with its
long CDR loop (H2) and clasping the CD4-binding loop between CDRs H2 and H1 on one side and H3 on the other
side. (See color insert.)
H3 region interact with the gp120 bridging sheet/V3 regions, and mimic the interaction with CCR5
(Choe et al. 2003; Huang et al. 2004, 2007).
Antibody b12 was discovered by phage-display technology using a library developed from blood
marrow taken from an HIV-1 infected, long-term nonprogressing patient. Crystal structures of the
intact b12 IgG (Saphire et al. 2001) and the recently determined structure of b12 Fab in complex
with gp120 (Zhou et al. 2007) revealed that the Fab recognizes the highly conserved, but deeply
recessed CD4-binding site (Fig. 2.10). The Fab CDR H2 loop accesses the CD4-binding site, with
TyrH53 binding in the same pocket as Phe43 of CD4. The long CDR H3 unexpectedly was found to
bind on the outside of the binding site, so that the CD4-binding loop is sandwiched between the H3
CDR on one side, and H1 and H2 on the other side. Interestingly, the antibody light chain makes
no contact with the gp120 monomer, but may possibly contact part of the intact gp120 trimer on the
viral surface. Thus, this antibody has managed to find a site of vulnerability on the virus, its
Achilles’ heel, and interacts primarily with the structurally conserved outer domain that is the primary
site of attachment for CD4. But unlike CD4, it can remain bound to the outer domain by itself with
high affinity, whereas CD4 reorganizes the highly flexible inner domain of gp120 to decrease its
off-rate and assemble the bridging sheet that constitutes a major portion of the CCR5 co-receptor
binding site (Zhou et al. 2007).
2.6 SHARK AND CAMEL ANTIBODIES
Crystal structures have also been determined for antibodies from rats, hamsters, camels, and sharks.
While the rodent and human antibodies are very similar, camels and sharks both have, in addition
to a conventional antibody repertoire, unusual antibodies that exist as dimers of heavy chains with
62
ANTIBODY MOLECULAR STRUCTURE
no associated light chains. In the case of the camel (and llama), these antibodies are the result of a gene
deletion of the IgG CH1 domain, resulting in a heavy chain with VH, CH2, and CH3. Light chains do not
associate with the CH1-free heavy chain; however, the free heavy chains still pair up through their two
associated constant domains. The camel heavy chain antibodies are closely related in sequence to their
normal IgG antibodies. Sharks and other cartilaginous fish, such as skates and rays, also have heavy
chain antibodies, called new antigen receptors or IgNAR, that contain two heavy chains and no
light chains. However, unlike the camel heavy chain dimers, the IgNAR antibodies are not closely
related by sequence to the other, more typical antibodies found in the shark (IgM and IgX). The
five constant domains on each heavy chain dimerize to form a long stalk, leaving the two variable
domains free to bind antigen independently. Although the shark and camel VH domains have very
low sequence homology, they have co-evolved many unique structural features, including mutations
in the interface that would normally be involved in the VL-VH dimerization, to make their VH domains
more soluble. Both camel and shark VH domains also have long CDR3 regions that usually contain
noncanonical disulfides to tether their long CDR H3s to the body of the VH domain. While the
Figure 2.11 Shark IgNAR single domain antibodies bound to lysozyme. Left: a Type I IgNar (bottom) is shown
bound to lysozyme (top). The Type I IgNAR is characterized by a long CDR3 region stabilized by two noncanonical
disulfide bonds tethering the CDR to the IgNAr framework. Other regions that have a large number of somatic
mutations and may interact with antigen are HV2 and HV4. Right: a Type II IgNAR is characterized by a long
CDR3 region with one noncanonical disulfide that tethers it to the CDR1 loop. (See color insert.)
REFERENCES
63
camel VH domain has the three CDR regions found in a typical VH domain (Desmyter et al. 1996;
Spinelli et al. 1996), the shark IgNAR domains have only CDR1 and 3, and are missing the C0 and
C00 strands that make up a typical CDR2 (Stanfield et al. 2004) (Fig. 2.11). Despite having only
three (camel) or two (shark) CDR loops, these small, independent VH domains are able to bind antigen
with very high affinity. The production of camel and shark VH domains has been adapted to phage-display technology (Arbabi Ghahroudi et al. 1997; Dooley, Flajnik, and Porter 2003) and the resulting
molecules show great promise as small antigen-binding fragments for many different uses. Several
studies have shown that the shark and camel single-domain antibodies are able to access deeply
recessed clefts better than their Fab counterparts (Stanfield et al. 2007). In addition, the solubilityenhancing mutations in these single-domain fragments have been adapted into engineered, human
single-domain VH fragments to enhance their solubility (Davies and Riechmann 1996).
2.7 SUMMARY
An enormous wealth of structural information is available for antibodies and their fragments, making
the antibody molecule one of the most highly studied in the entire protein universe. This information
has proven to be invaluable for scientists working towards the development and production of antibodies and antibody fragments for human therapeutic products. Nevertheless, although we know a
great deal about antibody structure, recent structures of several human antibodies have revealed surprisingly novel and diverse structural features, such as Fab domain swapping, long but relatively rigid CDR
H3s, and posttranslational modifications, that were totally unexpected from the study of an enormous
number of mouse antibodies. Given the tremendous diversity of the antibody immune response, we
expect to continue to uncover further novel aspects of antibody-antigen recognition that will provide
new ideas and inspiration for the development of antibody-based therapeutics.
ACKNOWLEDGMENTS
This is manuscript number MB-19299 from the Scripps Research Institute. The authors acknowledge
support from National Institutes of Health grants GM-46192 and the Neutralizing Antibody
Consortium of the International Aids Vaccine Initiative.
REFERENCES
Al-Lazikani, B., A.M. Lesk, and C. Chothia. 1997. Standard conformations for the canonical structures of immunoglobulins. J. Mol. Biol. 273:927– 948.
Arbabi Ghahroudi, M., A. Desmyter, L. Wyns, R. Hamers, and S. Muyldermans. 1997. Selection and identification
of single domain antibody fragments from camel heavy-chain antibodies. FEBS Lett. 414:521– 526.
Arevalo, J.H., C.A. Hassig, E.A. Stura, M.J. Sims, M.J. Taussig, and I.A. Wilson. 1994. Structural analysis of
antibody specificity. Detailed comparison of five Fab’-steroid complexes. J. Mol. Biol. 241:663– 690.
Arevalo, J.H., E.A. Stura, M.J. Taussig, and I.A. Wilson. 1993. Three-dimensional structure of an anti-steroid
Fab’ and progesterone-Fab’ complex. J. Mol. Biol. 231:103–118.
Arevalo, J.H., M.J. Taussig, and I.A. Wilson. 1993. Molecular basis of crossreactivity and the limits of antibodyantigen complementarity. Nature 365:859–863.
Better, M., C.P. Chang, R.R. Robinson, and A.H. Horwitz. 1988. Escherichia coli secretion of an active chimeric
antibody fragment. Science 240:1041–1043.
Blixt, O., S. Head, T. Mondala, C. Scanlan, M.E. Huflejt, R. Alvarez, M.C. Bryan, F. Fazio, D. Calarese, J. Stevens,
N. Razi, D.J. Stevens, J.J. Skehel, I. van Die, D.R. Burton, I.A. Wilson, R. Cummings, N. Bovin, C.H. Wong,
and J.C. Paulson. 2004. Printed covalent glycan array for ligand profiling of diverse glycan binding proteins.
Proc. Natl. Acad. Sci. USA 101:17033– 17038.
64
ANTIBODY MOLECULAR STRUCTURE
Bork, P., L. Holm, and C. Sander. 1994. The immunoglobulin fold. Structural classification, sequence patterns and
common core. J. Mol. Biol. 242:309– 320.
Boulianne, G.L., N. Hozumi, and M.J. Shulman. 1984. Production of functional chimaeric mouse/human antibody.
Nature 312:643– 646.
Bryan, M.C., F. Fazio, H.K. Lee, C.Y. Huang, A. Chang, M.D. Best, D.A. Calarese, O. Blixt, J.C. Paulson,
D. Burton, I.A. Wilson, and C.H. Wong. 2004. Covalent display of oligosaccharide arrays in microtiter
plates. J. Am. Chem. Soc. 126:8640–8641.
Burton, D.R., R.L. Stanfield, and I.A. Wilson. 2005. Antibody vs. HIV in a clash of evolutionary titans. Proc. Natl.
Acad. Sci. USA 102:14943 –14948.
Calarese, D.A., H.K. Lee, C.Y. Huang, M.D. Best, R.D. Astronomo, R.L. Stanfield, H. Katinger, D.R. Burton,
C.H. Wong, and I.A. Wilson. 2005. Dissection of the carbohydrate specificity of the broadly neutralizing
anti-HIV-1 antibody 2G12. Proc. Natl. Acad. Sci. USA 102:13372– 13377.
Calarese, D.A., C.N. Scanlan, M.B. Zwick, S. Deechongkit, Y. Mimura, R. Kunert, P. Zhu, M.R. Wormald,
R.L. Stanfield, K.H. Roux, J.W. Kelly, P.M. Rudd, R.A. Dwek, H. Katinger, D.R. Burton, and I.A. Wilson.
2003. Antibody domain exchange is an immunological solution to carbohydrate cluster recognition. Science
300:2065–2071.
Cardoso, R.M., F.M. Brunel, S. Ferguson, M. Zwick, D.R. Burton, P.E. Dawson, and I.A. Wilson. 2007. Structural
basis of enhanced binding of extended and helically constrained peptide epitopes of the broadly neutralizing
HIV-1 antibody 4E10. J. Mol. Biol. 365:1533– 1544.
Cardoso, R.M., M.B. Zwick, R.L. Stanfield, R. Kunert, J.M. Binley, H. Katinger, D.R. Burton, and I.A. Wilson.
2005. Broadly neutralizing anti-HIV antibody 4E10 recognizes a helical conformation of a highly conserved
fusion-associated motif in gp41. Immunity 22:163–173.
Choe, H., W. Li, P.L. Wright, N. Vasilieva, M. Venturi, C.C. Huang, C. Grundner, T. Dorfman, M.B. Zwick,
L. Wang, E.S. Rosenberg, P.D. Kwong, D.R. Burton, J.E. Robinson, J.G. Sodroski, and M. Farzan. 2003.
Tyrosine sulfation of human antibodies contributes to recognition of the CCR5 binding region of HIV-1
gp120. Cell 114:161–170.
Chothia, C., and A.M. Lesk. 1987. Canonical structures for the hypervariable regions of immunoglobulins. J. Mol.
Biol. 196:901–917.
Cole, S.P., B.G. Campling, T. Atlaw, D. Kozbor, and J.C. Roder. 1984. Human monoclonal antibodies. Mol. Cell
Biochem. 62:109–120.
Collis, A.V., A.P. Brouwer, and A.C. Martin. 2003. Analysis of the antigen combining site: Correlations between
length and sequence composition of the hypervariable loops and the nature of the antigen. J. Mol. Biol.
325:337– 354.
Darbha, R., S. Phogat, A.F. Labrijn, Y. Shu, Y. Gu, M. Andrykovitch, M.Y. Zhang, R. Pantophlet, L. Martin,
C. Vita, D.R. Burton, D.S. Dimitrov, and X. Ji. 2004. Crystal structure of the broadly cross-reactive HIV-1neutralizing Fab X5 and fine mapping of its epitope. Biochemistry 43:1410– 1417.
Davies, J., and L. Riechmann. 1996. Single antibody domains as small recognition units: Design and in vitro antigen
selection of camelized, human VH domains with improved protein stability. Protein Eng. 9:531– 537.
Desmyter, A., T.R. Transue, M.A. Ghahroudi, M.H. Thi, F. Poortmans, R. Hamers, S. Muyldermans, and L. Wyns.
1996. Crystal structure of a camel single-domain VH antibody fragment in complex with lysozyme. Nat. Struct.
Biol. 3:803–811.
Dooley, H., M.F. Flajnik, and A.J. Porter. 2003. Selection and characterization of naturally occurring single-domain
(IgNAR) antibody fragments from immunized sharks by phage display. Mol. Immunol. 40:25–33.
Foote, J., and C. Milstein. 1994. Conformational isomerism and the diversity of antibodies. Proc. Natl. Acad. Sci.
USA 91:10370–10374.
Halaby, D.M., A. Poupon, and J. Mornon. 1999. The immunoglobulin fold family: Sequence analysis and 3D
structure comparisons. Protein Eng. 12:563–571.
Harris, L.J., S.B. Larson, K.W. Hasel, J. Day, A. Greenwood, and A. McPherson. 1992. The three-dimensional
structure of an intact monoclonal antibody for canine lymphoma. Nature 360:369–372.
Harris, L.J., S.B. Larson, K.W. Hasel, and A. McPherson. 1997. Refined structure of an intact IgG2a monoclonal
antibody. Biochemistry 36:1581– 1597.
REFERENCES
65
Huang, C.C., S.N. Lam, P. Acharya, M. Tang, S.H. Xiang, S.S. Hussan, R.L. Stanfield, J. Robinson, J. Sodroski,
I.A. Wilson, R. Wyatt, C.A. Bewley, and P.D. Kwong. 2007. Structures of the CCR5 N terminus and of a
tyrosine-sulfated antibody with HIV-1 gp120 and CD4. Science 317:1930–1934.
Huang, C.C., M. Tang, M.Y. Zhang, S. Majeed, E. Montabana, R.L. Stanfield, D.S. Dimitrov, B. Korber,
J. Sodroski, I.A. Wilson, R. Wyatt, and P.D. Kwong. 2005. Structure of a V3-containing HIV-1 gp120 core.
Science 310:1025– 1028.
Huang, C.C., M. Venturi, S. Majeed, M.J. Moore, S. Phogat, M.Y. Zhang, D.S. Dimitrov, W.A. Hendrickson,
J. Robinson, J. Sodroski, R. Wyatt, H. Choe, M. Farzan, and P.D. Kwong. 2004. Structural basis of tyrosine
sulfation and VH-gene usage in antibodies that recognize the HIV type 1 coreceptor-binding site on gp120.
Proc. Natl. Acad. Sci. USA 101:2706–2711.
Huse, W.D., L. Sastry, S.A. Iverson, A.S. Kang, M. Alting-Mees, D.R. Burton, S.J. Benkovic, and R.A. Lerner.
1989. Generation of a large combinatorial library of the immunoglobulin repertoire in phage lambda. Science
246:1275–1281.
James, L.C., P. Roversi, and D.S. Tawfik. 2003. Antibody multispecificity mediated by conformational diversity.
Science 299:1362– 1367.
Jones, P.T., P.H. Dear, J. Foote, M.S. Neuberger, and G. Winter. 1986. Replacing the complementarity-determining
regions in a human antibody with those from a mouse. Nature 321:522–525.
Kabat, E.A., and T.T. Wu. 1971. Attempts to locate complementarity-determining residues in the variable positions
of light and heavy chains. Ann. NY Acad. Sci. 190:382– 393.
Koliasnikov, O.V., M.O. Kiral, V.G. Grigorenko, and A.M. Egorov. 2006. Antibody CDR H3 modeling rules:
Extension for the case of absence of Arg H94 and Asp H101. J. Bioinform. Comput. Biol. 4:415–424.
Morrison, S.L., M.J. Johnson, L.A. Herzenberg, and V.T. Oi. 1984. Chimeric human antibody molecules:
Mouse antigen-binding domains with human constant region domains. Proc. Natl. Acad. Sci. USA
81:6851– 6855.
Neuberger, M.S., G.T. Williams, E.B. Mitchell, S.S. Jouhal, J.G. Flanagan, and T.H. Rabbitts. 1985. A haptenspecific chimaeric IgE antibody with human physiological effector function. Nature 314:268–270.
Ofek, G., M. Tang, A. Sambor, H. Katinger, J.R. Mascola, R. Wyatt, and P.D. Kwong. 2004. Structure and mechanistic analysis of the anti-human immunodeficiency virus type 1 antibody 2F5 in complex with its gp41 epitope.
J. Virol. 78:10724– 10737.
´
Padlan, E.A., D.M. Segal, T.F. Spande, D.R. Davies, S. Rudikoff, and M. Potter. 1973. Structure at 4.5 Å resolution
of a phosphorylcholine-binding Fab. Nat. New Biol. 245:165–167.
Rini, J.M., R.L. Stanfield, E.A. Stura, P.A. Salinas, A.T. Profy, and I.A. Wilson. 1993. Crystal structure of a human
immunodeficiency virus type 1 neutralizing antibody, 50.1, in complex with its V3 loop peptide antigen. Proc.
Natl. Acad. Sci. USA 90:6325–6329.
Roux, K.H., L. Strelets, and T.E. Michaelsen. 1997. Flexibility of human IgG subclasses. J. Immunol. 159:
3372–3382.
Roux, K.H., P. Zhu, M. Seavy, H. Katinger, R. Kunert, and V. Seamon. 2004. Electron microscopic and immunochemical analysis of the broadly neutralizing HIV-1-specific, anti-carbohydrate antibody, 2G12. Mol. Immunol.
41:1001– 1011.
Sanders, R.W., M. Venturi, L. Schiffner, R. Kalyanaraman, H. Katinger, K.O. Lloyd, P.D. Kwong, and J.P. Moore.
2002. The mannose-dependent epitope for neutralizing antibody 2G12 on human immunodeficiency virus type
1 glycoprotein gp120. J. Virol. 76:7293–7305.
Saphire, E.O., P.W. Parren, R. Pantophlet, M.B. Zwick, G.M. Morris, P.M. Rudd, R.A. Dwek, R.L. Stanfield,
D.R. Burton, and I.A. Wilson. 2001. Crystal structure of a neutralizing human IgG against HIV-1: A template
for vaccine design. Science 293:1155–1159.
Scanlan, C.N., R. Pantophlet, M.R. Wormald, E. Ollmann Saphire, R. Stanfield, I.A. Wilson, H. Katinger,
R.A. Dwek, P.M. Rudd, and D.R. Burton. 2002. The broadly neutralizing anti-human immunodeficiency
virus type 1 antibody 2G12 recognizes a cluster of alpha1– – .2 mannose residues on the outer face of
gp120. J. Virol. 76:7306– 7321.
Segal, D.M., E.A. Padlan, G.H. Cohen, S. Rudikoff, M. Potter, and D.R. Davies. 1974. The three-dimensional
structure of a phosphorylcholine-binding mouse immunoglobulin Fab and the nature of the antigen binding
site. Proc. Natl. Acad. Sci. USA 71:4298– 4302.
66
ANTIBODY MOLECULAR STRUCTURE
Shirai, H., A. Kidera, and H. Nakamura. 1996. Structural classification of CDR-H3 in antibodies. FEBS Lett.
399:1–8.
Skerra, A., and A. Pluckthun. 1988. Assembly of a functional immunoglobulin Fv fragment in Escherichia coli.
Science 240:1038– 1041.
Spinelli, S., L. Frenken, D. Bourgeois, L. de Ron, W. Bos, T. Verrips, C. Anguille, C. Cambillau, and M. Tegoni.
1996. The crystal structure of a llama heavy chain variable domain. Nat. Struct. Biol. 3:752–757.
Stanfield, R.L., H. Dooley, M.F. Flajnik, and I.A. Wilson. 2004. Crystal structure of a shark single-domain antibody
V region in complex with lysozyme. Science 305:1770– 1773.
Stanfield, R.L., H. Dooley, P. Verdino, M.F. Flajnik, and I.A. Wilson. 2007. Maturation of shark single-domain
(IgNAR) antibodies: Evidence for induced-fit binding. J. Mol. Biol. 367:358–372.
Stanfield, R.L., M. Takimoto-Kamimura, J.M. Rini, A.T. Profy, and I.A. Wilson. 1993. Major antigen-induced
domain rearrangements in an antibody. Structure 1:83– 93.
Stanfield, R.L., and I.A. Wilson. 1994. Antigen-induced conformational changes in antibodies: A problem for
structural prediction and design. Trends Biotechnol. 12:275– 279.
Stanfield, R.L., A. Zemla, I.A. Wilson, and B. Rupp. 2006. Antibody elbow angles are influenced by their light
chain class. J. Mol. Biol. 357:1566–1574.
Steinitz, M., G. Klein, S. Koskimies, and O. Makel. 1977. EB virus-induced B lymphocyte cell lines producing
specific antibody. Nature 269:420–422.
Wilson, I.A., and R.L. Stanfield. 1994. Antibody-antigen interactions: New structures and new conformational
changes. Curr. Opin. Struct. Biol. 4:857– 867.
Wilson, I.A., R.L. Stanfield, J.M. Rini, J.H. Arevalo, U. Schulze-Gahmen, D.H. Fremont, and E.A. Stura. 1991.
Structural aspects of antibodies and antibody-antigen complexes. Ciba Found. Symp. 159:13–28; discussion
28–39.
Wu, T.T., and E.A. Kabat. 1970. An analysis of the sequences of the variable regions of Bence Jones proteins and
myeloma light chains and their implications for antibody complementarity. J. Exp. Med. 132:211–250.
Zhou, T., L. Xu, B. Dey, A.J. Hessell, D. Van Ryk, S.H. Xiang, X. Yang, M.Y. Zhang, M.B. Zwick, J. Arthos,
D.R. Burton, D.S. Dimitrov, J. Sodroski, R. Wyatt, G.J. Nabel, and P.D. Kwong. 2007. Structural definition
of a conserved neutralization epitope on HIV-1 gp120. Nature 445:732– 737.