Wochenschr. 58, 1217-1231 (1980)
`
`Klinische
`.
`Wochen
`ft
`schn
`© Springer-Verlag 1980
`
`Ubersichten
`
`Spatial Structure of Immunoglobulin Molecules
`
`R. Huber
`Max-Planck-Inst tut fiir Biochemie, Mart nsried
`
`Die raumliehe Struktur der Immunglobulin-Molekiile
`
`Zusammenfassung. Immunglobulin Molekille der
`Klasse G (Antikorper-Molekiile) bestehen aus zwei
`schweren Ketten (50000 dalton Molekulargewicht)
`und zwei leichten Ketten (25000 dalton Molekularge-
`wicht). Ihre Gestalt ist Y-formig, wobei die Anne
`von je einer leichten Kette und der N-terminalen
`Halfte einer schweren Kette in enger Assoziation ge-
`bildet werden. Der Stamm wird von den C-terminalen
`Halften der schweren Ketten aufgebaut.
`Die schweren und die leichten Ketten sind in glo-
`bulare Domanen mit einem Molekulargewicht von
`12 000 dalton gefaltet. Die schweren Ketten bestehen
`aus vier, die leichten Ketten aus zwei Domanen.
`Diese Domanen zeigen eine ahnliche Grundstruktur
`aus zwei fl-Faltblattern, aber erhebliche Unterschiede
`im Detail.
`Die N-terminalen, variablen Domanen der schwe-
`ren und leichten Ketten, spezifisch die hypervariablen
`Polypeptidsegmente der Domanen, die an den Spitzen
`des Y liegen, bauen die Antigen- und Hapten-Bin-
`dungsstelle auf. Die Art der Aminosauren in den hy-
`pervariablen Schleifen bestimmt die Form und die
`Spezifitat des Antikorpers. Alle Domanen mit Aus-
`nahme der CH2 Domane der schweren Kette aggregie-
`ren eng lateral. Die CH2 Doman hat Kohlehydrat
`gebunden, das die laterale Assoziation verhindert.
`Longitudinale Wechselwirkungen zwischen den
`Domanen sind locker und erlauben Flexibilitat in der
`relativen Anordnung der Domanen. Diese Flexibilitat
`ist wahrscheinlich fur die Funktion der Antikorper
`von Bedeutung.
`Arm (Fab) und Stamm (Fc) Teile sind durch ein
`Scharnierpeptide verbunden, das zwei parallelen Poly-
`proline Helizes enthalt.
`Antigenbindung initialisiert die Effektorfunktio-
`nen der Antikorper. Antigen bindet an die Spitzen
`des Y-formigen Molekiils, die Effektorfunktionen
`
`sind im Stammteil lokalisiert. Es ist eine offene Frage,
`ob Konformationsanderungen im Antikorpermolekiil
`bei der Initialisierung eine Rolle spielen.
`
`Schliisselworter: Immunglobulin — Antikorper — Pro-
`teinstruktur — Glykoprotein
`
`Summary. Immunoglobulin molecules of the class G
`(antibody molecules) consist of two heavy chains
`(50,000 dalton molecular weight) and two light chains
`(25,000 dalton). The overall shape is a Y with the
`arms formed by the light chains and the N-terminal
`half of the heavy chains in tight association. The
`stem is formed by the C-terminal halfs of the heavy
`chains.
`The heavy and the light chains fold into globular
`domains of molecular weights of 12,000 dalton. There
`are four domains of the heavy chain and two of the
`light chain. All these domains show a similar fold,
`consisting of two /3-sheets but display considerable
`differences in detail.
`The N-terminal variable domains of heavy and
`light chains and specifically the hypervariable poly-
`peptide segments of the domains, located at the tips
`of the Y, constitute the antigen and hapten binding
`site. The nature of the amino acid residues of the
`hypervariable loops determines the shape and the
`specificity of the antibody.
`All domains pair tightly laterally, except the CH2
`domains of the heavy chain. This domain has carbo-
`hydrate bound which prevents lateral association.
`Longitudinal interaction between the domains is
`loose and allows flexibility in the arrangement. Flexi-
`bility is probably of significance for antibody func-
`tion.
`
`Arm (Fab) and stem (Fc) parts are linked by
`the hinge peptide which contains a segment with a
`unique conformation of two parallel poly-proline he-
`lices.
`
`Genzyme Ex. 1011, pg 264
`
`

`
`1218
`
`R. Huber: Structure of Immunoglobulin Molecules
`
`Antigen binding triggers effector functions of anti-
`bodies. Antigen binding is at the tips of the Y-shaped
`antibody, but effector functions are displayed by the
`stem part. It is an open question whether conforma-
`tional changes of the antibody molecule play a signifi-
`cant role in the trigger mechanism.
`
`Key words: Immunoglobulin — Antibody — Protein
`structure — Glycoprotein
`
`Antibody molecules (immunoglobulins) form the ba-
`sis of the humoral immune defence reactions in prob-
`ably all vertebrate species. They recognize foreign
`macromolecules or cells (better: antigens on cell sur-
`faces) and, by binding these antigens, initiate their
`elimination. One route of elimination utilizes comple-
`ment components in a complicated cascade of reac-
`tions, which is intensively studied (for reviews see:
`[1, 2]). Immunoglobulin-like molecules also occur as
`membrane-bound receptors on the surface of bone-
`marrow derived B-lymphocytes. Recognition of the
`corresponding antigen leads to proliferation and anti-
`body secretion (for reviews see: [3-5]).
`Our present understanding of the molecular basis
`of antigen antibody recognition and complement acti-
`vation began with the elucidation of the chemical
`nature of immunoglobulins, their covalent structure
`[6-8] and culminated in the analysis of their spatial
`structure.
`These studies were almost exclusively based on
`myeloma and Bence-Jones proteins, which are found
`in large quantities and homogenous form in patients
`with multiple myeloma or Waldenstroem's macroglo-
`bulinemia. In most cases the corresponding antigens
`are unknown. Recently large amounts of homogenous
`antibodies elicited against streptococcal or pneumo-
`coccal polysaccharides became available from certain
`rabbit and mouse strains [9, 10]. These antibodies
`and the use of hybrids obtained from myeloma and
`spleen cells have offered the possibility to obtain ho-
`mogenous antibodies of predefined specificity [11, 12].
`Biochemical studies with these materials fully confirm
`the notion that there is no basic difference between
`the structures of myeloma proteins and induced anti-
`bodies [13].
`In this article I shall describe our present under-
`standing of the spatial structure of immunoglobulins
`and its functional implications. The detailed picture
`which we have today is based on crystal structure
`analyses of a number of immunoglobulin molecules
`and their fragments performed during the last seven
`years (for recent reviews see [14, 15]).
`Immunoglobulins are divided in a number of
`classes and sub-classes according to differences in
`
`I GG
`Fig. 1. Schematic drawing of an IgG1 immunoglobulin molecule.
`The arms and the stem of the Y-shaped molecule are formed
`by the Fab parts and the Fe part, respectively. The light chains
`are linked to the heavy chains by a disulfide bond close to the
`C-terminus. The two heavy chains are covalently connected by
`two disulphide linkages located in the hinge region
`
`heavy chains : IgG, IgM, IgA, IgD, IgE. There are
`two light chain classes : kappa and lambda (K, A), which
`are shared by all Ig classes.
`The schematic drawing in Fig. 1 represents an im-
`munoglobulin molecule of the most abundant class
`(IgG) as it was obtained from the structural studies
`described below. It is composed of two identical heavy
`and light chains with molecular weights of about
`50,000 and 25,000 daltons, respectively. These are
`held together by non-covalent forces and disulphide
`linkages. Limited proteolytic digestion of IgG yields
`stable and functional fragments. The Fab fragment
`comprises the light chain and the N-terminal half
`of the heavy chain. It binds antigen. The C-terminal
`Fe part of the heavy chain is involved in effector
`functions such as complement activation and binding
`to Fc receptors on certain cell types [6, 7].
`The polypeptide chains are folded into compact
`domains: four domains of the heavy chains and two
`of the light chains. These domains are designated
`VH, CH I, CH2, CH3 in the heavy chain VL and CL
`in the light chain. V stands for variable and C for
`
`Genzyme Ex. 1011, pg 265
`
`

`
`R. Huber: Structure of Immunoglobulin Molecules
`
`1219
`
`constant amino acid sequence Amino acid sequence
`analysis has shown that the N-terminal domains, with
`a molecular weight of about 12,000 daltons, are highly
`variable, while the constant domains show identical
`amino acid sequences in a given sub-class and species
`except for a few allotypic variations due to allelic
`genes [8]. The V domains bind antigen while C do-
`mains exhibit other functions. The view, that these
`domains are under separate genetic control, was ex-
`perimentally confirmed for the light chains by chemi-
`cal analysis of the corresponding genes of embryonic
`and mature antibody forming cells. In addition, it
`was found that part of the third hypervariable seg-
`ment and the switch peptide connecting V and C
`domains is controlled by a third gene [16, 17].
`Amino acid sequence analyses has shown that
`there is homology between all domains suggesting
`a similar chain folding [6]. There are also close rela-
`tions between amino acid sequences of the various
`Ig classes. The differences between the Ig classes and
`sub-classes reside predominantly in the hinge seg-
`ment, in the interchain disulphide linkages, in the
`bound carbohydrate, and in the state of aggregation.
`A close relationship in amino acid sequence is also
`found when immunoglobulins from different species
`are compared.
`There is no doubt therefore that the basic structural
`principles found for IgG are valid for other classes.
`Class specific structural variations are of course im-
`portant; they alter functional properties of the mole-
`cule considerably and certainly need to be analysed
`in detail in the future.
`
`Domain Structure
`
`The folding pattern is very similar in all immunoglob-
`ulin domains. It is shown schematically in Fig. 2 for
`a V domain, looking along the polypeptide strands.
`The folding is characterized by two pleated sheets
`connected by an internal disulphide bridge linking
`strands B and G. The two sheets cover a large number
`of hydrophobic amino acid sidechains.
`Figure 3 compares V and C domains seen in the
`intact IgG1 (2) molecule Kol and the V (K) chain
`of Rei [18-26]. The domain structures are represented
`by the positions of the C" atoms of the amino acids.
`It is clear that the topology of the strands is identi-
`cal in all domains. There are only minor differences
`between members of the V family and the C family
`with one another but substantial differences when
`we compare V and C domains: The number of
`strands and the length of the loop regions is different,
`changing the overall shape considerably.
`
`A
`
`X
`
`X
`
`H
`
`ARRANGEMENT OF STRANDS IN IMMUNOGLOBULIN DOMAINS
`X N-TERMINUS UP, • C-TERMINUS UP
`
`Fig. 2. Topology of strands in a V domain looking along the
`strands. (x) and (o) indicate N- and C-termini pointing towards
`the observer
`
`VH,
`and VL, ), form a family of closely related
`structures as do CL, CH1, and CH3.
`CH2 represents yet a third type, differentiated from
`the other C domains mainly by the branched carbohy-
`drate chain linked to it. It will be discussed in more
`detail below.
`
`Domain Domain Interactions
`
`Lateral Interactions
`
`Immunoglobulin domains other than CH2 interact
`strongly in a lateral fashion to form modules VH —VL,
`CL — CH1, CH3 —CH3. Large parts of the domain sur-
`faces are in contact. In V modules VH may be replaced
`by VL to form a light chain V dimer as seen in the
`Bence Jones protein fragments Rei or Au [18-20].
`In Bence Jones proteins, which are light chain dimers,
`one of the light chains simulates the heavy chain in
`Fab parts, as described for Mcg [27].
`Figure 4 shows the Fab parts of Kol [21, 26]. It
`is obvious that V and C pairings are entirely different.
`In a V pairing the HGCD faces of the domains and
`in a C pairing the opposite ABFE sides are in contact.
`CH3 exhibits C pairing, as shown below for the Fc
`part (Fig. 6).
`The basis of the different aggregation characteris-
`tics of V and C domains resides in the amino acid
`sequence. Residues important for lateral contact for-
`mation are conserved in all Ig classes and subclasses.
`The lateral pairing buries hydrophobic residues which
`would be exposed in isolated domains. The distribu-
`tion of these residues is different in V and C domains.
`There are hydrophobic patches on the HGCD face
`of V domains and the ABFE face of C domains.
`CH2 is an exception, as it forms a single unit with-
`out lateral domain domain interaction. Instead it in-
`
`Genzyme Ex. 1011, pg 266
`
`

`
`1220
`
`L
`
`a o
`
`O
`
`L
`
`R. Huber: Structure of Immunoglobulin Molecules
`
`0
`
`R
`
`0
`
`R
`
`0
`
`O
`
`b o
`
`0
`Fig. 3a—g. Polypeptide chain folding of V and C domains oriented approximately in the same way. VH of Kol (a), VL, a of Kol
`„ of Rei (c), CH1 of Kol (d), CL of Kol (e), CH2 (f) and CH3 (g) of IgG from pooled serum. In CH2 the carbohydrate
`(b),
`has been omitted. Light chains are numbered from 1 to 214 and heavy chains from 300 to make differentiation easier; the Fc fragment
`is numbered in the usual way with the unique hinge sequence Cys 226 — Pro 227—Pro 228 — Cys 229 [18-26]
`
`teracts with bound carbohydrate, which covers a large
`proportion of the ABFE face normally involved in
`a C type interaction, and there are amino acid ex-
`changes within the ABFE face not compatible with
`a C type aggregation (Fig. 5) [22, 25].
`
`The complex, branched carbohydrate chain bound
`to CH2 forms a few hydrogen bonds with the protein
`moiety, but the dominant interactions are of hydro-
`phobic nature. The carbohydrate covers a hydrophob-
`ic patch of the protein made up of Phe 241,
`
`Genzyme Ex. 1011, pg 267
`
`

`
`R. Huber: Structure of Immunoglobulin Molecules
`
`0
`
`L
`
`1221
`
`O
`
`V, RE 1
`
`0
`
`0
`
`Co
`
`O
`
`C
`
`d o
`
`Fig. 3 c—d.
`
`243 Val 262, 264 Tyr 296 Thr 260 Arg 301. Removal
`of the carbohydrate would probably destabilize the
`compact three-dimensional conformation of the C 2
`domain, since these residues would then be exposed.
`The functional relevance of carbohydrate in anti-
`
`bodies is unclear. It might be involved in intracel-
`lular movements of the glycoproteins and in secretion
`[28-30]. It may well be that the origin of the altered
`functional properties of carbohydrate-free antibody
`variants is structural destabilization.
`
`Genzyme Ex. 1011, pg 268
`
`

`
`1222
`
`0
`
`R. Huber: Structure of Immunoglobulin Molecules
`
`70
`
`170
`
`CL KoL
`
`e o
`
`0
`L
`
`0
`
`0
`
`R
`
`CH2
`
`f 0
`Fig. 3 e—f.
`
`0
`
`0
`
`0
`
`Genzyme Ex. 1011, pg 269
`
`

`
`R. Huber: Structure of Immunoglobulin Molecules
`
`0
`
`L
`
`0
`
`R
`
`400
`
`1223
`
`0
`
`g
`
`Fig. 3g.
`
`0
`
`0
`
`The features described point to an important
`structural role of bound carbohydrate in antibodies
`and possibly in glycoproteins in general.
`
`Longitudinal Interactions
`
`In contrast to the extensive lateral interactions, non-
`bonded longitudinal interactions along the heavy
`chain or light chain are much weaker or do not exist
`at all. They are interesting, however, because confor-
`mational changes in antibodies affect such interac-
`tions in specific ways.
`The CH3 —CH2 interaction is shown in Fig. 6,
`which represents the Fc part of an IgG1 molecule.
`The contact surface is small, as only the tips of the
`domains touch each other. Residues participating in
`this contact are conserved in all Ig classes suggesting
`that this contact is preserved [22].
`We note that the CH3 —CH2 orientation is some-
`what variable and influenced by external forces. In
`the Fc fragment crystals the two chemically identical
`chains are in different environment. As a consequence
`th CH3 —CH2 orientation varies by about 6°. In Fc-
`protein A complex crystals this arrangement also
`differs by a small amount from that of Fe crystals
`[25].
`
`More drastic changes are observed in VH —CH1
`and VL —CL longitudinal contacts, if we compare
`chemically different Fab fragments. This arrangement
`is most conveniently described by an elbow angle
`which is enclosed by the pseudo diads relating VH
`to VL and ClH to CL, respectively. The elbow angle
`may vary from more than 170° to 120° when we
`compare Kol Fab with McPc Fab (Fig. 7) [21, 24,
`26, 31].
`In two instances elbow angles of the same mole-
`cule in different crystal lattices were compared and
`found to differ by 8° to 17°, respectively [24, 32].
`It is obvious from Fig. 7, that there is no non-bonded
`longitudinal contact in Kol, a molecule characterized
`by an open elbow. We interpret these observations
`to mean that in Kol the V C arrangement is flexible
`in solution. In the crystal, the molecule is stabilized
`by packing interactions; these will be discussed from
`a different point of view later.
`
`The Hinge Segment
`
`Fab and Fc parts are covalently linked by the hinge
`segment, which has a unique primary and spatial
`structure. The central region of the hinge consists
`of two parallel disulphide-linked, poly-L-proline he-
`
`Genzyme Ex. 1011, pg 270
`
`

`
`1224
`
`L
`
`0
`
`R
`
`R. Huber: Structure of Immunoglobulin Molecules
`
`O
`
`KOL IGGi, (FAB)2
`
`-
`0
`0
`0
`Fig. 4. IgG 1 molecule Kol. In the crystal the Fab parts and the hinge segment are well ordered, but the Fc part is disordered and
`not visible. — light chain, = heavy chain. The heavy chain is numbered from 300 [21, 26]
`
`lices formed by the hinge sequence — Cys — Pro —
`Pro — Cys — (Fig. 8) [21, 26]. In the IgG 1 sub-class
`represented by the Kol molecule shown in Fig. 8,
`the poly-praline double helix is short. IgG 3, however,
`has a quadruplicated hinge sequence [33] and model
`building suggests that the poly-proline segment of
`this molecule may be more than 100 A long. The
`poly-proline segment, a relatively rigid structure, is
`flanked on both sides by flexible segments : The seg-
`ment on the N-terminal side is well defined in the
`crystal lattice of Kol, due to crystal packing interac-
`tions, but it lacks internal interactions that would
`provide stability in solution. The C-terminal segment
`is disordered and flexible in Kol crystals and in the
`Fe crystal structure [22]. The hinge segment allows
`independent movement of the Fab arms and the Fc
`
`part. There is direct evidence for flexibility in the
`crystal lattice of Kol [21, 26] and Zie [34]. Both pro-
`teins have their Fc parts disordered in the crystalline
`state. This is in contrast to an abnormal IgG protein
`Dob, which lacks a hinge region. Here Fab and Fc
`are rigidly arranged [35]. The significance of the hinge
`for Fab Fc flexibility is obvious.
`
`The Antigen Binding Area
`
`Comparison of amino acid sequences of variable parts
`has demonstrated hypervariability of some segments.
`These were considered to be involved in antigen bind-
`ing [36]. Crystal structure analyses of Ig fragment-
`hapten complexes indeed show that haptens bind in
`
`Genzyme Ex. 1011, pg 271
`
`

`
`R. Huber: Structure of Immunoglobulin Molecules
`
`0
`
`L
`
`0
`
`R
`
`1225
`
`O
`
`O
`
`O
`
`L
`
`CH2, CARBOHYDRATE AND AMINOACID RESIDUES
`COVERED BY THE CARBOHYDRATE
`
`0
`
`O
`
`R
`
`0
`
`O
`
`CH2 AND CARBOHYDRATE
`
`o
`O
`Fig. 5 a—c. The C142 domain and its complex carbohydrate which is linked to Asn 297. The polypeptide chain is drawn with thick
`lines, the carbohydrate with thin lines. Amino acid residues covered by the carbohydrate are also shown in a [22, 25]. Two different views
`of CH2 (a, b) and the isolated carbohydrate (c) are plotted
`
`a cleft or depression formed by the hypervariable
`segments. A representative example is shown in
`Fig. 9. A tri-nitrophenyl group binds to the Rei frag-
`ment, a VL, „ dieter [18, 20]. This and other examples
`demonstrate that structural complementarity is the
`
`basis of antibody hapten recognition. Structural
`changes upon hapten binding seen so far are quite
`localized, involving amino acid side chains in contact
`with the hapten. Antigens are usually macromolecules
`which cover a much larger part of the antibody than
`
`Genzyme Ex. 1011, pg 272
`
`

`
`1226
`
`R. Huber: Structure of Immunoglobulin Molecules
`
`0
`
`Asti 297
`
`Co
`
`CARBOHYDRATE CHAIN LINKED TO ASN 297 IN CH2
`
`c
`
`•
`
`L
`
`Fc FRAGMENT WITH BOUND CARIOHYDRATE
`
`0
`Fig. 6. The Fc fragment with the bound carbohydrate. The head is formed by the CH3 module, the ears by CH2. The hinge segment
`is disordered in Fc fragment crystals and not shown [22]
`
`haptens do. The lattice contact found in the Kol crys-
`tals might be an instructive model. Here, the hyper-
`variable segments of one molecule touch residues of
`the hinge and adjacent parts of another molecule
`(Fig. 10) [21, 24, 26].
`
`Complement Binding Site
`
`The first step in the classical pathway of complement
`activation involves the binding of C 1 complex, or
`more specifically, the C 1 q component, to antigen an-
`tibody complexes [1, 2]. Binding is between the C 1 q
`
`head pieces and the Fc part of the antibodies. The
`CH2 part must be involved, as a fragment Facb, which
`lacks the CH3 domains fixes complement [37]. In addi-
`tion a fragment derived from CH2 shows activity [38].
`The crystal structure of the complex formed by
`the Fc fragment and protein A, a small protein from
`the cell wall of staphylococcus aureus, provides us
`with a further hint [25]. Protein A binds specifically
`to the Fc part of antibody molecules of certain classes
`and sub-classes, but does not interfere with comple-
`ment binding. It forms a small globular domain made
`up of three helices, which binds to segments of CH3
`and CH2. Figure 11 shows the complex. The area of
`
`Genzyme Ex. 1011, pg 273
`
`

`
`R. Huber: Structure of Immunoglobulin Molecules
`
`1227
`
`H
`
`CL
`
`KOL
`
`NH
`
`C
`
`McPC 603
`
`CL
`
`Fig. 7. Fab Kol and McPC 603 seen along an axis
`through the switch peptides. Kol and McPC are
`characterized by open and closed elbow angles,
`respectively. (V variable module, C constant module,
`NH, NL N-termini of heavy and light chain, CH, CL
`C-termini [24, 26, 31]
`
`CH2 not covered by protein A must contain the C 1 q
`binding site. This includes the tips of the CH2 do-
`mains, the hinge and the surface between the CH2
``ears' of the Fc fragment. The surface between the
`ears is partly covered by the carbohydrate and certain-
`ly less accessible in general due to the proximity of
`the two ears.
`A plausible Cl q interacting site is therefore at
`the tips of the C2 domains and the hinge. Interestingly
`these segments are disordered in protein A — Fc com-
`plex crystals. Possibly, flexibility is required in anti-
`body C 1 q interaction.
`
`Conformational Changes
`
`Antigen antibody complex formation triggers fixation
`of complement. Antigen binds at the tips of the Y-
`shaped molecule while C 1 attaches to the Fc part.
`C 1 q has a weak intrinsic affinity to free antibody
`molecules, which is strongly enhanced in antigen anti-
`body complexes. The molecular basis of this enhance-
`ment is not clear and two mechanisms may be consid-
`ered : (for a review see [39]).
`Aggregate formation through crosslinking of the
`antibody molecules by antigen may enhance C 1 q
`
`Genzyme Ex. 1011, pg 274
`
`

`
`1228
`
`0
`
`0
`
`R
`
`HINGE IGG1 KOL ••••.....mm LIGHT CHAIN
`HEAVY CHAIN
`
`0
`
`0
`
`R. Huber: Structure of Immunoglobulin Molecules
`
`0
`
`0
`
`Fig. 8. The hinge peptide conformation as seen in IgG1 Kol. It forms a loosely folded segment from residues 520 to 526 and a
`poly-L-proline double helix Cys—Pro—Pro—Cys (residue 527-530) [21, 26]
`
`Tyr 232
`
`L
`
`REI (VK )0 WITH TNP GROUP BOUND
`Fig.9. A TNP (tri-nitrophenyl) group bound to the (Vi., K )2 dimer Rei. One of the protomers is numbered from 1 on, the second
`from 200 on [20]
`
`Genzyme Ex. 1011, pg 275
`
`

`
`R. Huber: Structure of Immunoglobulin Molecules
`
`1229
`
`60
`
`R
`
`60
`
`0
`CRYSTAL PACKING IN KOL
`— HINGE REGION AND ADJACENT SEGMENTS
`— HYPERVARIABLE REGIONS OF NEIGHBOURING MOLECULES
`
`0
`0
`0
`Fig. 10. The crystal packing in Kol, The hinge peptide and adjacent segments are in close contact with the hypervariable segments
`of a neighbouring molecule. This arrangement may serve as a model for an antibody antigen complex. The heavy chain is numbered
`from 300. — hinge segment and adjacent residues, = hypervariable region of a crystallographically related molecule [21. 26]
`
`FC-PROTEIN A COMPLEX, FLEXIBLE SEGMENTS
`INDICATED BY THIN LINES
`
`Fig. 11. Protein A —Fc complex. Protein A forms a small globular domain consisting of three helices, that binds to segments of the
`C.2 and C.3 domain. The upper segments of C.2 and part of the carbohydrate are flexible in the complex. These show no defined
`electrondensity and are indicated by thin lines in the model [25]
`
`Genzyme Ex. 1011, pg 276
`
`

`
`1230
`
`R. Huber: Structure of Immunoglobulin Molecules
`
`binding as C 1 q is multimeric with six or more binding
`sites for antibody molecules. C 1 q binding to an anti-
`body aggregate could therefore be potentiated. Alter-
`natively, antigen binding might induce a conforma-
`tion change in the Fc part which enhances affinity
`for C 1 q. A mixed mechanism involving both aggre-
`gation and conformation change cannot be excluded,
`of course.
`There is unequivocal evidence for the importance
`of aggregation, but only controversial indications of
`conformation changes induced by antigen binding.
`Nevertheless, it may be rewarding to sketch a pos-
`sible conformation change originating at the antigen
`binding site and transmitted to the Fc part on the
`basis of structural features of the IgG molecule eluci-
`dated to date. There are structural and functional
`studies indicating cooperativity between the two ends
`of Fab fragments [41, 42]. In the absence of a direct
`structural comparison of the same antibody molecule
`with and without antigen bound, we must employ
`a comparison of the various crystal structures. These
`should provide an idea of what conformational states
`are accessible to antibody molecules:
`Antibody molecules and their domains resemble
`pearls on a string with weak longitudinal non-cova-
`lent interactions between the domains, but strong lat-
`eral interactions (except for the CH2 domains). This
`may be deduced from the large changes in elbow
`angle, when we compare different Fab molecules, but
`also the same molecules in different crystal environ-
`ments. Flexibility of Fc in Kol and Zie crystals em-
`phasizes this aspect in a most obvious way. There
`is also some variability in CH3 — CH2 arrangement.
`This obvious lack of longitudinal non-bonded interac-
`tions presents a major problem for any hypothesis
`about a signal transmitted along the antibody chain.
`It appears that a rigid conformer of the molecule
`must be involved, presumably with the elbow angle
`closed and Fab and Fc in contact [23]. Lateral interac-
`tions of VH VL, ClH CL, CH3 — CH3 are extensive.
`They seem to be conserved in the various crystal struc-
`tures, but there are small differences observed be-
`tween Kol and New [40] which may be important,
`and could be caused by the antigen-antibody-like
`crystal lattice contacts in Kol [26]. Small changes in
`lateral associations might trigger conformation
`changes in the surfaces involved in the longitudinal
`contacts producing large changes there.
`Equilibria between flexible and rigid conformers
`and their functional significance has been demon-
`strated for trypsin and its precursor trypsinogen [43,
`44]. The structural features of antibody molecules sug-
`gest conformation changes between a flexible and a
`rigid conformer as a pathway for signal transfer in
`antibodies as well, if such a signal exists at all.
`
`Acknowledgement. The help of Dr. W.S. Bennett in preparing the
`manuscript is gratefully acknowledged. Dr. J. Deisenhofer and
`Dr. M. Marquart helped in preparing the figures.
`
`References
`
`I Mueller-Eberhard HJ (1975) Complement. Ann Rev Biochem
`44:697-724
`2 Porter RR, Reid KBM (1979) Activation of the complement
`system by antibody-antigen complexes: The classical pathway.
`Adv Prot Chem 33:1-71
`3 Warr GW (1980) Membrane immunoglobulins of vertebrate
`lymphocytes. Cont Top Immunobiol 9:141-162
`4 Marchalonis JJ (1976) Surface immunoglobulins of B and T
`lymphocytes: Molecular properties, association with the cell
`membrane, and a unified model of antigen recognition. Cont
`Top Mol Immunol 5:125-160
`5 Loor F, Roelants GE (eds) (1977) B and T Cells in immune
`recognition. J Wiley & Sons, London New York Sydney Toron-
`to
`6 Edelman GM (1970) The structure and function of antibodies.
`Sci Am Aug:34-42
`7 Porter RR (1967) The Structure of Antibodies. Sci Am Oct: 81-
`87
`8 Hilschmann N (1969) Die molekularen Grundlagen der Anti-
`koerperbildung. Naturwissenschaften 56:195-205
`9 Jaton J-C, Huser H, Braun DG, Givol D, Pecht J, Schlessinger
`JC (1975) Conformational changes induced in a homogenous
`anti-Type HI pneumococcal antibody by oligosaccharides of
`increasing size, Biochem 14:5312-5315
`10 Braun DG, Huser H (1977) Rabbit anti-polysaccharide Anti-
`bodies: structure and genetics. In: Mandel TE, Cheers CH,
`Hosking CS, McKenzie IFC, Nossal GJV (eds) Progress in
`immunology III. North Holland Publ Comp, Amsterdam New
`York Oxford, pp 255-264
`11 Koehler G. Milstein C (1975) Continuous cultures of fused
`cells secreting antibody of predefined specificity. Nature
`256: 495-497
`12 Melchers F, Potter BM, Bethesda NW (eds) (1978) Lymphocyte
`hybridomas. Curr Top Microbiol Immunol 81:
`13 Nisonoff A, Hopper JE, Spring SB (1975) The antibody mole-
`cule. Academic Press New York San Francisco London (1975)
`14 Edmundson AB, Ely KR, Abola EE (1978) Conformational
`flexibility in immunoglobulins. Cont Top in Mol Immunol
`7:95-118
`15 Amzel LM, Poljak RJ (1979) Three-dimensional structure of
`Immunoglobulins. Ann Rev Biochem 48:961-997
`16 Tonegawa S, Maxam AM, Tizard R, Bernard 0, Gilbert W
`(1978) Sequence of a mouse germ-line gene for a variable region
`of an immunoglobulin light chain. Proc Nat Acad Sci USA
`75:1485-1489
`17 Nishioka Y, Leder P (1980) Organization and complete se-
`quence of identical embryonic and plasmacytoma K V-region
`genes. J Biol Chem 255:3691-3694
`18 Epp 0, Colman P, Fehlhammer H, Bode W, Schiffer M, Huber
`R, Palm W (1974) Crystal and molecular structure of a dimer
`composed of'the variable portions of the Bence-Jones protein
`Rei. Eur J Biochem 45:513-524
`19 Fehlhammer H, Schiffer M, Epp 0, Colman PM, Lattman
`EE, Schwager P, Steigemann W, Schramm HJ (1975) The struc-
`ture determination of the variable portion of the Bence-Jones
`Protein Au. Biophys Struct Mechanism 1:139-146
`20 Epp 0, Lattman EE. Schiffer M, Huber R, Palm W (1975)
`The molecular structure of a 'dimer composed of the variable
`portions of the Bence-Jones protein Rei. Biochem 14:4943-
`4952
`
`Genzyme Ex. 1011, pg 277
`
`

`
`R. Huber: Structure of Immunoglobulin Molecules
`
`1231
`
`21 Colman PM, Deisenhofer J, Huber R, Palm W (1976) Structure
`of the human antibody molecule Kol (Immunglobulin G 1):
`An electron density map at 5 A Resolution. J Mol Biol
`100:257-282
`22 Deisenhofer J. Colman PM. Epp 0. Huber R (1976) Crystallo-
`graphic structural studies of a human Fc fragment. II. A com-
`plete model based on a fourier map at 3.5 A resolution. Hoppe-
`Seyler's Z Physiol Chem 357:1421-1434
`23 Huber R, Deisenhofer J, Colman PM, Matsushima M, Palm
`W (1976) Crystallographic structure studies of an IgG molecule
`and an Fc fragment. Nature 264:415-420
`24 Matsushima M. Marquart M, Jones TA, Colman PM. Bartels
`K. Huber R. Palm W (1978) Crystal structure of the human
`Fab fragment Kol and its comparison with the intact Kol mole-
`cule. J Mol Biol 121:441-459
`25 Deisenhofer J. Jones TA. Huber R, Sjoedahl J. Sjoequist J,
`(1978) Crystallization, crystal structure analysis and atomic
`model of the complex formed by a human Fc fragment and
`fragment B of protein A from staphylococcus aureus. Z Physiol
`Chem 359:975-985
`26 Marquart M. Deisenhofer J, Huber R, Palm W (1980) Crystal-
`lographic refinement and atomic models of the intact immuno-
`globulin molecule Kol and its Fab fragment at 3.0 and 1.9 A
`resolution. J Mol Biol 141:369-392
`27 Edmundson AB, Ely KR, Abola RR, Schiffer M, Panagioto-
`poulos N (1975) Rotational Allomerism and Divergent Evolu-
`tion of Domains in Immunoglobulin light Chains. Biochemistry
`14 : 3953-3961
`28 Melchers F (1973) Biosynthesis. intracellular transport. and se-
`cretion of immunoglobulins. Effect of 2-deoxy-D-glucose in
`tumor plasma cells producing and secreting immunoglobulin
`G 1. Biochem 12:1471-1476
`29 Weitzman S. Scharft MD (1976) Mouse myeloma mutants
`blocked in the assembly, glycosylation and secretion of immu-
`noglobulin. J Mol Biol 102:237-252
`30 Hickman S. Kulczycki A Jr, Lynch RG. Kornfeld S (1977)
`Studies of the mechanism of tunicamycin inhibition of IgA
`and IgE secretion by plasma cells. J Biol Chem 252:4402-4408
`31 Segal DM, Padlan EA, Cohen GIL Rudikoff S, Potter M,
`Davies DR (1974) The three-dimensional structure of a phos-
`phorylcholine-binding mouse immunoglobulin Fab and the na-
`ture of the antigen binding site. P