ANTIBODY STRUCTURE AND FUNCTION - bioinf · 2016. 7. 4. · c. The antibody’s antigenic...

58

4ANTIBODY STRUCTURE AND FUNCTION

CHAPTER OUTLINE

OBJECTIVES

1. The structure of an antibody is related to its function.a. Studies by Tiselius and Kabat and later by Edelman,

Porter, and Nisonoff determined the basic structuralcomponents of antibodies.1) Initially, antibodies were identified only by their elec-

trical charge.2) Fragmentation studies showed that antibodies are

made up of two identical light chains and two identicalheavy chains.

b. Further studies explained the fine structure of antibod-ies—the variable and constant regions on the chains.

c. The antibody’s antigenic determinants—called isotypes,allotypes, and idiotypes—determine the variability in an-tibody structure.1) Isotypes are variants present in all members of a

species.2) Allotypes are variants caused by intraspecies genetic

differences.3) Idiotypes are variants caused by structural hetero-

geneity in the antibody V regions.

2. There are five classes of antibodies based on the struc-ture of their heavy-chain C domains.

a. All antibody classes have either l or k light chains.b. IgG is the major antibody in the blood, but it is able to en-

ter tissue spaces and coat antigens, speeding antigenuptake.

c. IgA concentrates in body fluids to guard the entrances ofthe body.

d. IgM is the largest antibody; it tends to remain in theblood, where it can lead to efficient killing of bacteria.

e. IgD remains membrane-bound and somehow regulatesthe cell’s activation.

f. IgE is found in trace amounts in the blood, but it still trig-gers allergies.

3. The biological effector functions of antibodies are me-diated by the C domains.

4. Monoclonal antibodies are pure antibodies with singleantigenic determinant specificities.a. Monoclonal antibody development improves on nature.b. Special procedures are needed to screen for and isolate

monoclonal antibodies.c. Monoclonal antibodies have many applications.

The reader should be able to:

1. Understand how the characterization of antibody structureled to the comprehension of antibody function.

2. Distinguish between the overall structure and the fine struc-ture of antibodies.

3. Describe the variable and constant regions of an antibody’slight and heavy chains.

4. Explain the organization of the variable regions of an anti-body’s light and heavy chains; define hypervariable regionsand domains.

5. Name and compare the biological and chemical character-istics of the five classes of antibodies.

6. Discuss the differences in the biological effector functionsof antibodies.

7. Contrast conventional antibody and monoclonal antibodydevelopment; conceptualize the procedure for monoclonalantibody screening; and discuss hybrid monoclonal anti-bodies.

T H E S T R U C T U R E O F A N A N T I B O DY I S R E L A T E D T O I T S F U N C T I O N 59

n the preceding chapter, antigens were de-scribed.1 This chapter describes the proteins that

bind antigens and mark them for destruction by theimmune system.

The recognition proteins found in the serum2 andother body fluids of vertebrates that react specificallywith the antigens that induced their formation arecalled antibodies. Antibodies belong to a family ofglobular proteins called immunoglobulins. The termsantibody and immunoglobulin are used interchange-ably throughout this text. Immunoglobulins, however,are defined as a family of globular proteins that com-prise antibody molecules and molecules having pat-terns of molecular structure (antigenic determinants)in common with antibodies. The term immunoglobu-lin can be used to refer to any antibody-like molecule,regardless of its antigen-binding specificity.3

THE STRUCTURE OF AN ANTIBODY IS RELATEDTO ITS FUNCTION

The function of an antibody is to bind foreign or non-self molecules. The host can produce a vast array ofantibodies that are structurally similar (all are Y-shaped molecules) yet unique. This variability was astartling finding because all other protein moleculesmade by an individual are identical; they all have thesame amino acid sequence. However, antibodies comein millions of different amino acid sequences and arethe most diverse proteins known.

The chemical structure of antibodies explains threefunctions of antibodies: (1) binding versatility, (2)

binding specificity, and(3) biological activity. Amammal is capable of re-sponding to more than100 million antigenic de-terminants and can evenrespond to artificial anti-gens that do not exist in

nature (Fast Focus 1). Because the amino acid se-quence differs in the arms of various antibody mole-cules, each different antibody can bind specifically to

one unique epitope. Thus, the arms of an antibodymolecule confer the versatility and specificity of re-sponses that a host can mount against antigens. Thestem region of an antibody molecule decides its biolog-ical activity and defines whether the response against aparticular antigen will lead to complement-mediated ly-sis, enhanced phagocytosis, or (in some cases) allergy.These activities start once antibodies bind to antigen.

Studies by Tiselius and Kabat and Later by Edelman, Porter, and NisonoffDetermined the Basic Structural Components of Antibodies

Initially, Antibodies Were Identified Only by TheirElectrical Charge

Since the 1890s, immunologists have known that themolecules of humoral immunity are present in serum.In 1939, Tiselius and Kabat performed electrophoreticstudies with rabbit antiserum specific for ovalbumin.4

They found that electrophoresis of serum from unim-munized rabbits resolved serum proteins into fourdominant families of differing mobility. The albuminswere the fastest-migrating (most negatively charged)fraction, followed by the alpha- (a), beta- (b), andgamma- (g) globulins (Figure 4-1). Tiselius and Kabatthen hyperimmunized rabbits with hen ovalbumin toget a strong antibody response. When they elec-trophoresed the immune serum, they observed a largeincrease in the g-globulin fraction and concluded thatantibodies were g-globulins. After they absorbed outthe anti-ovalbumin antibodies with the antigen ovalbu-min, they observed that the electrophoretic patternwas the same as the pattern for the preimmune serum.This observation thus supported their conclusion thatantibodies are g-globulins. Antibodies can be sepa-rated further by ultracentrifugation. Studies performedin the 1940s and 1950s revealed that antibody mole-cules are heterogeneous in size. They range in molec-ular weight from 150 to 1000 kD.

Fragmentation Studies Showed That Antibodies AreMade Up of Two Identical Light Chains and TwoIdentical Heavy Chains

In the late 1950s and early 1960s, Rodney Porter ofGreat Britain and Gerald Edelman of the United Stateselucidated the chemical structure of antibodies.5 Edel-

I

FAST FOCUS 1

The estimation for an indi-vidual’s antibody repertoire iscalculated from theoretical as-sumptions. Nonetheless, thenumber is huge.

1We use the term antigen, even though immunogen may be moreaccurate at times, as explained in Chapter 3.2Antibody-containing serum is called antiserum, in contrast to nor-mal serum (the clear yellowish fluid collected when whole blood isseparated into its solid and liquid parts) that does not contain anti-body to a specific antigen.3MHC molecules and T-cell antigen receptors are the other twokinds of molecules used by the immune system to bind antigen.Their role in antigen recognition is discussed in Chapters 8 and 9, re-spectively.

4Electrophoresis is the migration of charged molecules after the in-troduction of an electric current.5Porter and Edelman shared the 1972 Nobel Prize in Medicine fortheir structural studies of antibodies.

man structurally studied g-globulins and myelomaproteins, while Porter’s experiments focused on a spe-cific antibody from the g-globulin fraction of rabbitserum called immunoglobulin G or IgG. Edelman’sand Porter’s approaches to the characterization of theIgG molecule were different. Edelman characterizedIgG molecules using chemical solvents, whereas Porterused protein-degrading enzymes. Our understandingof antibody structure draws from the two scientists’ re-sults.

To generate subunits of IgG, Edelman treated rab-bit IgG with dithiothreitol (a reducing agent that dis-rupts disulfide bonds), iodoacetamide (an alkylatingagent that prevents reassociation of the disrupted disul-fide bonds), and a denaturing agent (a substance thatdisrupts noncovalent interactions). When the treatedantibodies were passed through sizing columns, heidentified two subunits in equimolar ratios. He desig-nated the larger subunit (50 kD) as the heavy, or H,chain and the smaller subunit (23 kD) as the light, orL, chain. Because the molecular weight of the originalIgG molecule is 150 kD, he concluded that the IgGmolecule consisted of two heavy and two light chainslinked by disulfide bonds and noncovalent interac-tions.

Porter fragmented rabbit IgG with the proteolyticenzyme papain in the presence of the reducing agentcysteine. Papain hydrolyzed peptide bonds in IgG andproduced three fragments (I, II, and III). The threefragments had similar molecular weights (50 kD) butdifferent charges. Two of the three fragments wereidentical and retained the ability to bind antigen.These two fragments were called the Fab fragments

(for fragments of antigen-binding). Because the intactIgG is bivalent and the two Fab fragments each couldbind antigen, Porter concluded that the Fab fragmentsmust be univalent. The third fragment produced by pa-pain digestion did not bind with antigen and crystal-lized during cold storage. Porter called this piece theFc fragment (for fragment crystallizable). Thus theratio of Fab to Fc is 2:1. Edelman confirmed Porter’sresults by cleaving and electrophoresing human IgGinto two antigenically different fractions equivalent tothe two fragments from rabbit IgG (Figure 4-2).

In similar studies, Alfred Nisonoff used pepsin,which hydrolyzes different sites on the IgG moleculethan does papain. IgG treated with pepsin yielded onelarge fragment with a molecular weight (100 kD) dou-ble that of one Fab fragment, and many small frag-ments. Nisonoff called the large fragment F(ab9)2. Thisfragment also could bind antigen, but unlike the Fabfragment, it led to a visible serologic reaction. It hadboth of the antigen-binding sites of IgG (the chains re-mained linked) and could be treated further with re-ducing agents to yield two Fab-like fragments calledFab9. Collectively, the two enzymes cleave at aboutthe same region of the IgG molecule. Papain splits themolecule on one side, and pepsin on the other side, ofthe bond that holds Fab fragments together (see Figure4-2).

Following these studies, Porter showed that eitherFab or Fab9 fragments compose the entire light chainand part of the heavy chain. These data led to the for-mulation of the structure of an antibody and are sum-marized in Table 4-1. IgG is used as the prototype anti-body to explain the basic structure of all antibodies.

60 A N T I B O DY S T R U C T U R E A N D F U N C T I O N

(+) (–)

Preimmune serum

albu

min

α β γ

(+) (–)

albu

min

α β γ

(+) (–)

Immune serum

albu

min

α β γ

Immune serum:antibody removed

FIGURE 4-1 Antibody activity in the g-globulin frac-tion of serum. The electrophoretic pattern of rabbit serum be-fore immunization (top), after immunization (middle), and after ab-sorption of the antibody with specific antigen.

Further Studies Explained the Fine Structureof Antibodies—the Variable and ConstantRegions on the Chains

The primary structure of antibodies was determined byamino acid sequencing of myeloma proteins. Myelomaproteins are molecules that have the characteristicstructure of antibodies but originate from a diseasestate rather than from an immune response to a spe-cific antigen. These pathologic immunoglobulins areformed by individuals with multiple myeloma, a can-cer of antibody-forming (plasma) B cells. Myelomaproteins were invaluable in the resolution of antibodystructure because they are produced in large quantities

and are usually homo-geneous in a patient,whereas in a normal im-mune response the anti-bodies are heterogeneousin composition (Fast Fo-cus 2). The main workinvolved a light-chainmyeloma protein, calledthe Bence-Jones protein(first described in 1847by Henry Bence-Jones).Bence-Jones proteins areimmunoglobulin light


Fraction number

Fab Fab

Fab

Fab

Fc

Fc

F(ab′)2

Ion Exchange chromatographic profile

L

H

H

L

L

H

H

L

L

H

H

L

H

L

Fc

Fab

Heavy Chain

Light chain

Fragment crystallizable

Fragment antigen-binding

Constant portion of chainVariable portion of chain

L

H

H 2 × L

2 × H

LA

bsor

banc

e

Fraction number

Sizing chromatographic profile

H chains

L chains

Low molecularweight

fragments

Pepsin

Papain

IntactIgG

molecule

Reduction withdithiothreitol

ormercaptoethanol

Alkylation andacidification

Abs

orba

nce

S

S

SHSH

SHSH SH

S

COOHNH2

S

SS

SS

SS

SHSH

SS

S S

SH

SS

SS

SS

SS

SS

SS S

S

FIGURE 4-2 Fragmentation of rabbit IgG by the enzymes papain and pepsin.Schematic illustration of the chromatography of reduced and papain-digested (also pepsin-digested) rab-bit IgG.The results show the basic structure of all antibodies. See text for further explanation.

FAST FOCUS 2

Robert A. Good has describedmultiple myeloma as one ofthose “crucial experiments ofnature,” a group of diseasesthat has “contributed maxi-mally to the development of ourconcepts of immunology.” Inpopular parlance, the path ofdiscovery went “from the pa-tient’s bedside to the laboratorybench” and led to the completemapping of an antibody’s struc-ture. (R. A. Good, 1972. In Clin-ical Immunology, Vol. 1. NewYork: Academic Press, p. 23.)

chains excreted in the urine of myeloma patients.Amino acid sequence comparisons of Bence-Jones klight chains led to a startling finding. Each chain con-sisted of a large region that was constant in differenttypes of antibodies (even from unrelated species) anda similar-sized region that was highly variable. Thislarge region, called the constant (C) region, hasamino acid sequences in the carboxyl terminal end(residues 109 to 214) of the k chains that are almostidentical. The opposite end of the k chain, the aminoterminal end (residues 1 to 108), shows great variabil-ity in amino acid sequence among the chains and iscalled the variable (V) region.

The amino acid sequence of antibody heavychains showed similar regions. The first 113 aminoacid residues of the heavy chain6 again showed greatvariability, while the remaining part of the heavy chainconsisted of amino acids that were almost identical(the constant regions). This implies that moleculesbinding different determinants have different V light(VL) and V heavy (VH) regions. A group of words


TABLE 4-1 Mini Summary: Characteristics of theIgG Molecule*

1. IgG (150 kD) consists of two light chains and two heavychains; each light chain pairs with a heavy chain, andeach heavy chain pairs with another heavy chain.a. The chains are linked by covalent interchain disulfide

bonds and noncovalent interactions.b. The two light chains (23 kD each) and two heavy

chains (50 kD each) are identical.2. Papain digestion produces two Fab fragments and one Fc

fragment.a. Fab fragments (50 kD) consist of the entire light chain

and part of the heavy chain.b. Fab fragments contain the antigen-binding sites.c. The Fc fragment (50 kD) crystallizes in cold and does

not bind antigen.3. Pepsin digestion produces one large fragment (100 kD)

called F(ab9)2 and degrades the remainder of the heavychains into small pieces.a. When the large fragment is treated with reducing

agents, the disulfide bridges are disrupted, giving twoFab-like fragments, Fab9.

b. Each of the Fab9 fragments is made up of the entirelight chain and a slightly longer part of the heavy chain.

*Even though characteristics are given for the IgG antibody mole-cule, we will see that all antibody molecules have similar overallstructures and some common properties.

6The length of the heavy-chain V region varies. The V heavy-chainregion is somewhat larger (about 123 amino acid residues) than theV light-chain region (about 108 amino acid residues). V-regionlength depends on CDR length. The CDRs, especially the CDR3s ofIgMs, can be quite large.

Light chains ( chains)κ

κV κCNH2

1 108 109 214

COOH

Heavy chains ( 1 chains)γ

VH γC 1NH2

1 123 446

COOH

CDR1 CDR2 CDR3

Location of CDRs in V regionκ

28-3450-56

89-97

CDR1 CDR2 CDR3

Location of CDRs in VH region

31-3550-65

95-102

FIGURE 4-3 Amino acid sequence of lightand heavy chains—localization of vari-ability. Antibodies have two identical light andheavy chains. Each chain is divided into two re-gions, the variable (V) and constant (C) regions.Theamino-terminal end contains the V region,while thecarboxyl-terminal end contains the C region. Inthe V region are areas of increased variability calledhypervariable regions or complementarity-deter-mining regions (CDRs, marked by arrows); thenumbers below the arrows give the amino acidresidue positions.The CDRs of the light and heavychains form the antigen-binding sites. The moreconserved amino acids between the CDRs arecalled framework residues; they hold the CDRs inplace. See text for further explanation.

can illustrate the concept:P-S-Y-C-H-O-L-O-G-Y, P-H-Y-S-I-O-L-O-G-Y, andI-M-M-U-N-O-L-O-G-Y.All have a constant re-gion (“O-L-O-G-Y”) and avariable region, the firstfive letters of each word(Fast Focus 3). Becauseamino acid sequencedictates the three-dimen-sional structure of pro-tein, the unique sequenceof amino acid residues foreach V region leads to thelarge diversity of struc-

ture, which accounts for antibody specificity.Variability of the amino acid sequences in the V re-

gions is not random but precisely organized. It is local-ized within certain sections of the V region of a chain,and these sections have substantial sequence variationfrom protein to protein. The greatest variability in thelight chain is around residues 30, 55, and 95. These

areas are called hypervariable regions or com-plementarity-determining regions (CDRs). Movingfrom the amino-terminal end of an antibody chain, thethree regions are called CDR1, CDR2, and CDR3. EachCDR is about 10 amino acid residues in length. (The Vregion of the heavy chain has three CDRs near theCDRs of the light chain.) Hypervariable regions arecalled CDRs because these segments line the antibody-combining site. Intervening sequences between theCDRs have restricted variability and show little differ-ence in amino acid sequence between chains. Theseinvariant segments make up the framework residues,which compose about 85% of the V region. Frameworkresidues define the positioning of the CDRs. The V re-gion folds so that the CDRs are exposed on the surfaceof the chain. When the light and heavy chains arejoined, the CDRs of the chains form a cleft that servesas the antigen-binding site of an immunoglobulin. Be-cause the amino acid sequences of the CDRs determinethe shape and ionic properties of the antigen-bindingsite, the CDRs define the specificity of the antibody. Aschematic representation of heavy and light chains isshown in Figure 4-3.


FAST FOCUS 3

Antibodies are large proteinsshaped to form a Y. The sec-tions that make up the tips ofthe Y’s arms, or variable (V) re-gions, vary greatly from one an-tibody to another, creating apocket uniquely shaped to en-fold a specific antigen epitope.The fork in the Y is the hinge re-gion, which connects the armsto the stem and allows them toswing. The stem, or constant(C) region, of the Y serves tolink the antibody to other partic-ipants in the immune response.The stem is identical in all anti-bodies of the same class.

VL

CLVH

Heavychain

Lightchain

CH2

CH3

CH1

S SS S

S S

S SS S

S S

S S

S S

S S

S S

S S

S S

S S

S S

S S

S S

Antigen-binding clefts

Hinge region

CarbohydrateFIGURE 4-4 Schematic rep-resentation of IgG domains.Areas of variability and constancy,divided into segments of 110amino acid residues, within an IgGmolecule (and for all antibodies)are known as domains. The IgGmolecule has a V and a C domainfor each light chain and one V andthree C domains for each heavychain.

Although antibodies are complex globular proteins,they do possess some structural repetitiveness. Edel-man showed that each antibody chain has a tandemseries of repeating homology units roughly 110 aminoacid residues in length called immunoglobulin do-mains, which fold independently into a compact glob-ular structure (Figure 4-4, see page 63). All proteinsthat exhibit this structural motif belong to the immu-noglobulin gene superfamily (see Chapter 9). Do-mains are separated by less ordered regions of the pep-tide chains. The light chain of IgG has two domainscalled VL and CL. The heavy chain of IgG has four do-mains: one VH and three in the CH region (CH1, CH2,

and CH3 or Cg1, C

g2, and C

g3). The heavy-chain V unit

shows similarity to the V part of the light chain, whilethe three C-region units show strong homology to eachother and to the C region of the light chain. Each do-main has a characteristic tertiary structure consisting oftwo b-pleated sheet structures. Each domain has asandwich-like structure with a hydrophobic interior.Layers are covalently linked by a disulfide bridge nearthe center of the domain. When the VL chain frame-work residues assume this three-dimensional arrange-ment, loops are held out from the sandwich. Theseloops are the CDRs. The size and shape of the loops aredefined by their amino acid residues. When heavy and


VH domain

VL domain

VL domain

VH domainCL domain CH1

CH2

CH2

S S

CH3(a)

(b)

Carbohydrate chain

Heavy chains

Carbohydrate

Antigen-binding site

Antigen-binding

site

Antigen-binding site

VH

VL

CL

CH2

CH3

VL

Antigen-binding

site

CLVH

FIGURE 4-5 Three-dimensional model of IgG. From studies by X-ray crystallography andelectron microscopy, (a) the three-dimensional structure of IgG can be depicted as a Y-shaped molecule.Each sphere represents an amino acid residue.The two light chains are shades of blue and the two heavychains are shades of gray. (b) The schematic cartoon shows how the domains interact. (Adapted fromE.W. Silverton, M.A. Navia, and D.R. Davies. 1977.Proc.Natl.Acad. Sci.USA 74:5140.)

light chains are brought together to form an IgG mole-cule, extensive noncovalent interactions occur betweenVL and VH, CL and CH1, CH2 and CH2, and CH3 and CH3.These regions of interaction are shown by the “three-dimensional” model in Figure 4-5 (see page 64). Theassociation of the CDRs and the antigen-binding site ofan antibody was confirmed by crystallographic studies.An example of such a study is the binding of antibodyspecific for the hapten phosphorylcholine (Figure 4-6).

Antibodies also demonstrate segmental flexibil-ity, which means that the two Fab portions can moverelative to one another on antigen binding. The anglevaries from 60 to 180 degrees. This flexible regionwhere the arms meet the stem of the Y is called thehinge region and is located between the CH1 and CH2domains. Only IgG, IgA, and IgD antibody molecules7

have hinge regions. The heavy chains of IgM, IgA, andIgD antibody molecules possess additional amino acidresidues on the carboxyl-terminal end of the last CH do-main. These areas, called tail pieces, permit IgM andIgA to interact with like antibodies and form multimericmolecules. Multimeric IgM and IgA also have a poly-peptide called the joining ( J) chain, which is disul-fide-linked to the tail pieces and stabilizes the multi-meric structure. Antibodies also contain carbohydrates;the percentage and location of the carbohydrate mole-cules differ according to the class of antibody.


M I N I S U M M A RY

Amino acid sequencing of light chains shows thatvariability is localized at the amino terminal end ofthe chains in the variable, or V, regions. The other(carboxyl) ends are identical and are called theconstant, or C, regions. This variability arrange-ment is the same for heavy chains. Variability isconcentrated in segments called complementarity-determining regions (CDRs) or hypervariable re-gions. The CDRs line the walls of the antigen-bind-ing sites in antibody molecules. Segments betweenthe CDRs are highly conserved regions calledframework residues. Light chains also possess tworepeated homology units (domains), and heavychains possess four to five domains. The three-dimensional structure of antibodies shows that do-mains are folded in similar configurations, with theV-light and V-heavy units lying together to formthe V domains of the complete antibody mole-cule. V domains contain the antigen-binding cleft.Changes in amino acid residues at the position ofthis cavity change its shape and thus its specificity.The amino acid sequencing of antibody light andheavy chains divides them into V regions for anti-gen recognition and C regions for biological effec-tor functions.

Phosphorylcholine

H3CN+ CH2 CH2 O O–

CH3H3C

P

O

O

==

The cleft formed by the amino acids of theheavy (H) and light (L) chain complementarydetermining region (CDR) "holds" the antigenicdeterminant, in this case the haptenphoshorylcholine.

L chainAsp 96 96

H chainTry 104

H chain

Try 33

Glu(-)35

Glu(-)58N+

O

O

PO

O–

CH3

CH2

CH2

CH3

CH3

(+) Lys 58

(+) Arg 52

7The five antibody classes—IgG, IgA, IgM, IgD, and IgE—will be dis-cussed shortly.

FIGURE 4-6 Antigen-com-bining site of an antibodyfor the hapten phosphoryl-choline showing the com-plementarity between anepitope and an antibody-combining site. The aminoacids lining the CDRs bind to theepitope through weak, noncova-lent interactions.The three CDRsof the heavy chain and one CDRof the light chain contribute tothe combining site.

The Antibody’s Antigenic Determinants—Called Isotypes, Allotypes, and Idiotypes—Determine the Variability in Antibody Structure

Humans express five groups of antibodies, calledimmunoglobulin (Ig), or antibody, classes. The

five classes of antibodies, designated IgG, IgA, IgM,IgD, and IgE, differ in their physicochemical (charge,size, and solubility) and serologic (in vitro reac-tions with antigens) properties, and in their behav-ior as antigens. The latter characteristic usually is theone used to divide immunoglobulins into the fiveclasses.


Genes for isotypic markersare present in all members

of a species.

Oz markers of L chainsλ andκ λ

s

s s

s

s s

ISOTYPIC PORTIONS

γ 1, γ 2, γ 3, γ 4, µ,α 1, α 2, δ, and ε

Genetic variation due toindividuals having different

alleles for particularallotypic markers.

Gm markers of IgGAm markers of IgA

Km markers of L chainsκs ss s

ALLOTYPIC PORTIONS

s s s s

s s

s s

s ss s

IDIOTYPIC PORTIONS

Unique idiotypic marker foreach antigen determinant(generally associated with

antigen-binding site ofantibody), thus localized in

variable region of chain

Concept: Because immunoglobulins (lg) arecomplex globular proteins, they are goodantigens in the appropriate recipient. Theresulting antibodies (anti-lg) can be used tocharacterize lg. The three different markersthat can be recognized are illustrated here.

s s

FIGURE 4-7 Nomenclature of antibody variability. Three overall variants of antibody exist.Going from general to specific variation, the first, known as isotypic variation, refers to the differentheavy and light chain classes (IgG, IgA, IgM, IgD, and IgE) and subclasses (such as human IgG1 throughIgG4). Most allotypic variation occurs in the C region of H and L chains, but some has been found in theframework residues of V regions.The last variation, called idiotypic, is found only in the V regions of theheavy and light chains.

Antibodies are divided into classes by the anti-genic determinants on their heavy chains. Antibodiesthemselves can be immunogenic. For example, if oneimmunizes a rabbit with human antibodies, the rabbit’simmune system does not see an antibody molecule butrather a foreign complex glycoprotein. The rabbit re-sponds by making antibodies to the human antibodies,and this antiserum can be used to distinguish at leastthree types of determinants (isotypic, allotypic, and id-iotypic) that can then be used to classify antibodiesinto isotypes, allotypes, and idiotypes (Figure 4-7).

Isotypes Are Variants Present in All Members of a Species

The heavy-chain antigenic determinants that definethe antibody class are called isotypic determinants.The antibody molecule is an isotype (G. iso, thesame). Isotypic determinants distinguish C-region sites.Thus, an antibody specific for the heavy chain of IgGreacts with only IgG molecules. Antigenic determi-nants that define the heavy chain of IgG, IgA, IgM,IgD, and IgE are given the Greek letters gamma (g), al-pha (a), mu (m), delta (d), and epsilon (ε), respec-tively. Other isotypic heavy-chain determinants definedifferences within a class and thus are called antibodysubclasses. IgG has four subclasses, called g1, g2, g3,and g4 or IgG1, IgG2, IgG3, and IgG4. In humans,IgG1, IgG2, IgG3, and IgG4 are found in normal serumin the approximate proportions of 65, 25, 5, and 5%,respectively. Two subclasses have also been found inIgA, designated a1 and a2 or IgA1 and IgA2.

Antigenic determinants on light chains classifythem as either k or l chains. Humans have only onegene for the k C region, but the l chain has at leastfour functional C-region genes. The four l chain iso-types are designated l1, l2, l3, and l6. Light-chainantigenic determinants are not useful in determiningantibody class or subclass because k and l chains areassociated with all classes and subclasses. Within anantibody molecule, both heavy chains and both lightchains are identical. All of the genes for isotypic deter-minants are expressed in a normal individual; there-fore, all isotypic determinants are present on some an-tibodies of all members of a species.

Allotypes Are Variants Caused by IntraspeciesGenetic Differences

Allotypic determinants are encoded by one allele(variation) of a given antibody gene and are presenton the antibodies of some members of a species. Theseantibodies are called allotypes (G. allos, other). Alter-nate alleles code for antigenically distinct allotypicmarkers. The genes encoding allotypic determinantsare inherited in a Mendelian fashion and illustrates thegenetic diversity within a species (polymorphism). Al-lotypic nomenclature consists of the right class and

subclass marker followed by each allele in paren-theses—for example, G2m(23), G4m(4a), A2m(1), orKm(3). The g, a, and k chains have allotypic determi-nants called Gm, Am, and Km. All four g chain iso-types have allotypes (humans have more than 20 Gmmarkers). IgA2 has allotypes designated A2m(1) andA2m(2). The k-light chain allotypes are called Km(1),Km(1,2), and Km(3). No allotypes have been de-scribed for IgM, IgA1, IgD, IgE, or the l chain.

Idiotypes Are Variants Caused by StructuralHeterogeneity in the Antibody V Regions

Idiotypic determinants are found in the V regionnear the antigen-binding site of the antibody molecule.These determinants classify antibodies into idiotypes(G. idios, one’s own) and are common to antibodieshaving specificity for the same foreign antigenic deter-minant. Each individual has as many different idio-types as it has different antibodies. These determinantsare individual-specific. Idiotypic determinants reflectthe antibody-combining site’s structure and usually de-pend on the arrangement of heavy and light chains.

THERE ARE FIVE CLASSES OF ANTIBODIESBASED ON THE STRUCTURE OF THEIR HEAVY-CHAIN C DOMAINS

All Antibody Classes Have Either l or k Light Chains

As mentioned earlier, the two determinants used to de-fine antibody light chains are called kappa (k) andlambda (l). Each chain has a molecular weight ofabout 23 kD and exists as one polypeptide chain ofabout 214 amino acid residues. The first 108 aminoacids make up the V region, followed by roughly 110

T H E R E A R E F I V E C L A S S E S O F A N T I B O D I E S 67


The antigenic determinants of antibodies showthree levels of variability. The determinants (fromgeneral to specific) are called isotypic, allotypic, oridiotypic and classify antibody molecules into iso-types, allotypes, and idiotypes, respectively. Iso-typic determinants, found on the C-region part ofheavy chains, divide antibodies into five classes(or isotypes) called IgG, IgA, IgM, IgD, and IgE.Isotypic determinants also can be used to classifyantibody light chains into k and l chains. Isotypicdifferences can be used to partition classes intosubclasses. Allotypic determinants are carried byonly some individuals within a given species andare inherited in a Mendelian fashion. Idiotypic de-terminants are individual-specific and representthe antigen-combining site of an antibody.

amino acids that make up the C region. These two re-gions compose the VL and CL chain domains. The ra-tios of k to l light chains varies greatly within mam-malian species. In mice, it is 95:5; in humans, it isabout 60:40. However, there is no difference in eitherchain’s ability to pair with a heavy chain.

IgG Is the Major Antibody in the Blood, but It Is Able to Enter Tissue Spaces and CoatAntigens, Speeding Antigen Uptake

IgG, primarily induced by protein antigens, constitutesabout 80% (12.5 mg/ml) of the antibody in serum. TheIgG (150 kD) is composed of two light chains (either kor l) and two heavy chains (g). The four polypeptidechains are covalently held together by disulfide bonds.Human IgG consists of four subclasses (isotypes),which are numbered in order of their serum concen-trations (IgG1, IgG2, IgG3, and IgG4). The four sub-classes have 90 to 95% identity with each other in theC-region domains. The g chain is made up of four do-mains, one in the V portion and three in the C portionof the chain. The g1 chain is the shortest heavy chain,with 446 amino acid residues. On the CH2 domain (atposition 297) of all g chains is attached one carbohy-drate group that controls the quaternary structure ofthis domain. The chief distinguishing characteristicamong the four IgG subclasses is the pattern of inter-chain linkages in the hinge region.

IgA Concentrates in Body Fluids to Guard theEntrances of the Body

Human IgA constitutes only 13% (2.1 mg/ml) of theantibody in human serum, but it is the predominantclass of antibody in extravascular secretions. The IgApresent in secretions (tears, saliva, nasal secretions,

bronchial and digestive tract mucus, and mammarygland secretions) is secretory IgA. While the preciseorganization of secretory IgA is unknown, the modelof current choice is depicted in Figure 4-8.

The J chain is a 15-kD polypeptide consisting of129 amino acid residues and one carbohydrate group.It is synthesized by plasma cells and attaches to IgA(or IgM) either before or at the time of secretion. TheJ chain attaches to the carboxyl-terminal penultimatecysteine of either the a or the m chain. Dimeric IgAbinds to the blood side of the epithelial cells throughFc receptors (Figure 4-9). These receptors are alsocalled secretory, or S, proteins. Bound IgA is inter-nalized and moves through the cytoplasm of the ep-ithelial cells. IgA is detached from the cell followingcleavage of S protein. The remaining peptide, calledsecretory component or piece, attaches to dimeric IgA.Depending on the species, it may or may not bedisulfide-linked to the IgA dimer. It also gives resis-tance to enzymatic cleavage while in mucosal secre-tions.

The a chain is made up of one V domain andthree C domains. IgA1 is the most prevalent form inserum, but IgA2 is slightly more prevalent in secre-tions. Only IgA2 has allotypic determinants, and onlythe A2m(1) uniquely lacks interchain disulfide bridgesbetween light and heavy chains. Instead, chains arelinked to their own counterparts (one light chain to theother light chain). Another difference between IgA al-lotypes is the size of their hinge regions.

IgM Is the Largest Antibody; It Tends toRemain in the Blood, Where It Can Lead to Efficient Killing of Bacteria

IgM, primarily induced by polysaccharide antigens, isa 950-kD pentamer that makes up about 8% (1.25


Secretory componentDisulfide bridgefor secretory

component

J chain

DIMERIC IgAMONOMERIC IgA

J chainbinding

FIGURE 4-8 Human secretory IgA. This is made up of two monomeric IgA molecules, a joining(J) molecule and a secretory component (SC), with molecular weights of 160,000 (each monomer),15,000, and 75,000, respectively. Intact secretory IgA has a molecular weight of 390,000. The twomonomers of IgA are joined by their Fc ends by the J chain.They are attached by disulfide bridges to theCa2 domain of each IgA monomer. The secretory piece is probably wrapped around the two IgAmonomers.

mg/ml) of the antibody in the serum. The fivemonomeric IgM molecules are arranged radially, theFab fragments pointing outward and the Fc fragmentspointing to the center of the circle (Figure 4-10). IgM isthe first antibody to appear during an immune re-sponse and the first formed by a developing fetus. Be-cause of its many antigen-binding sites, IgM canquickly clump antigen and efficiently activate comple-ment. IgM acts as one of the main receptors on the sur-face of mature B cells, along with IgD. When IgM is asurface receptor, it is in its monomeric form.

The IgM m chain consists of 576 amino acidresidues, with 452 making up the C region. Unlike gand a chains, which have three C-region domains, them chain has four. The five carbohydrate groups are inthe CH1 and CH3 domains and in the part of the mchain where the J chain binds. The CH2 domain of them chain is equivalent to the hinge regions of the g anda chains. The m chain has two interchain disulfidebonds. The membrane form of IgM has a different car-boxyl-terminal end. The membrane form of IgM ismade up of 41 additional amino acid residues, ofwhich 25 form a transmembrane segment of hydro-phobic (nonpolar) amino acids followed by hydro-philic (polar) amino acids.

T H E R E A R E F I V E C L A S S E S O F A N T I B O D I E S 69

IgA monomer

IgA dimer

J chainaddition

Plasma cell

IgA dimer

Receptor for IgA(poly lg receptor) Epithelial cells

Endocyticvacuole Complete

secretoryIgA

molecule

Secretory IgA

Secretorycomponent

SUBMUCOSA MUCOSA LUMEN

FIGURE 4-9 Synthesis and transport of secretory IgA. IgA molecules are synthesizedas monomers, but plasma cells secrete them as dimers linked by the J chain.These IgA molecules bindto the poly immunoglobulin receptor on the interior surface of the epithelial cells.The complex is en-docytosed and passed through the epithelial cells lining the lumens of the body, where the complexacquires the secretory component by cleavage of the receptor and finally enters the lumen as secre-tory IgA.

J chain

s-s bridges

FIGURE 4-10 Human IgM. This is a pentameric (fivemonomeric IgM molecules) polypeptide chain that has four domainsin each of the heavy chains. Disulfide bonds cross-link adjacent C

m3

and Cm4 domains of individual monomeric molecules.The possible lo-

cation of the J chain is given.

IgD Remains Membrane-Bound andSomehow Regulates the Cell’s Activation

IgD (175 kD) constitutes less than 1% (40 mg/ml) ofthe antibody in human serum. IgD is an antibodywhose function remains unknown, even though it isone of the main receptors on mature B cells. As B cellsmature, IgD is replaced by other antibodies. IgD maybe a regulator of immune responses through its role inantigen internalization. The d-chain C region is dividedinto three domains and consists of 383 amino acidresidues. The hinge region of IgD consists of 64 aminoacid residues, longer than any other antibody class.

IgE Is Found in Trace Amounts in the Blood,but It Still Triggers Allergies

Human IgE (190 kD) makes up less than 0.003% (0.4mg/ml) of the antibody in serum. IgE binds through itsFc part to mast cells or basophils. On later exposure tothe same antigen, mast cells and basophils bind anti-gen with membrane-bound IgE and trigger allergic re-actions. IgE protects against parasites by releasing me-diators that attract eosinophils. Like the m chain, the εchain contains four C-region domains. IgE is made upof about 13% carbohydrate. The ε chains are similar insize to m chains, except that ε chains lack the 18 aminoacid residues for J-chain binding. For further discus-sion of IgE, see Chapter 15.

THE BIOLOGICAL EFFECTOR FUNCTIONS OF ANTIBODIES ARE MEDIATED BY THE C DOMAINS

Whereas a small part of the V region on an antibodydetermines the antigen specificity, single domains inthe C region of heavy chains determine the effectorfunctions. Biologic activities of antibodies divide intothree general areas: (1) protection, (2) placental trans-fer, and (3) cytophilic (literally, “cell-loving”) proper-ties. (The biologic function of membrane antibody asthe B-cell receptor for antigen is discussed in Chapter

6.) Although antibody binding blocks the attachmentof toxins or viruses (called neutralization), antibodiesalone cannot directly destroy a foreign organism. In-stead, antibodies mark them for destruction by otherdefense systems. When IgM or IgG (except IgG4)binds to antigen, the complement system is activatedand promotes bacterial lysis or accelerated phagocyticuptake. IgM or IgG molecules that have not reactedwith antigen do not activate complement. IgM alsomediates agglutination reactions. The coating (opson-ization) of organisms with primarily IgG antibodiesleads to enhanced phagocytosis by macrophages andneutrophils. Antibodies allow for the interaction ofseveral cell types with antigen–antibody complexesthrough the cells’ Fc receptors. Three groups of humanFc receptors for IgG have been described (Table 4-2).Because of their characteristic immunoglobulin-likeextracellular domains, Fc receptors belong to the im-munoglobulin gene superfamily (see Figure 9-4 inChapter 9). Multiple biologic functions can be trig-gered through the crosslinking of any of the three Fcreceptor classes. Macrophages have enhanced engulf-ment of antigen–antibody complexes through Fc re-ceptors. B-cell Fc receptor engagement by antigen–antibody complexes regulates B-cell activation. Othercell types expressing Fc receptors (CD16) can use anti-body-dependent cell-mediated cytotoxicity (ADCC), tolyse target cells coated with IgG. Certain antibodies,like IgA, can be localized to the lumens of mucosa-lined organs to provide mucosal immunity.

The second area of biologic activity associatedwith the antibodies is the movement of maternal anti-body across the placenta to the fetus. The human fetusand newborns have limited immune responses. Mech-anisms of acquired immunity are not at full strengthuntil some time after birth. Most of the protection fora fetus or newborn comes from maternal IgG thatcrosses the placenta during pregnancy. Only IgG cancross the placenta because only the g-chain CH1 andCH3 domains can bind to placental cells. Intact IgG orFc fragments from IgG can cross the placenta, but Fabor F(ab9)2 cannot. Maternal IgA, secreted in breastmilk, neutralizes pathogens in the infant’s gut.


TABLE 4-2 Human IgG Fc Receptors

Molecular AffinityCD Name Common Name weight (21000) (Kd) Distribution

CD16 FcRIII 50–65 2 2 1026 M Granulocytes, macrophages, NK cells, neutrophilsCD32 FcRII 40 5 2 1027 M B cells, eosinophils, granulocytes, macrophagesCD64 FcRI 75 1 2 1028 M Monocytes, macrophages, cytokine-activated neutrophils

Note: Three classes of human Fc receptor for IgG have been defined. The high-affinity FcRI is constitutively expressed on monocytes, macro-phages, and cytokine-activated neutrophils, while the lower-affinity FcRII is widely distributed (found on immune and nonimmune cells). FcRIIIhas the lowest affinity of the three Fc receptors and is the main Fc receptor on neutrophils.

The third area of biologic activity is the binding ofIgE to mast cells and basophil receptors through theirFc regions. Because of this “stickiness,” IgE antibodiesare called cytophilic antibodies. The reaction followingthe second exposure of specific antigen with IgE mole-cules bound to mast cells triggers allergic responses.The IgE molecule and allergic responses are detailedin Chapter 15. Lymphocytes have Fc receptors for IgGthat regulate IgG–antigen complex-mediated antibodyfeedback (see Chapter 12).

MONOCLONAL ANTIBODIES ARE PUREANTIBODIES WITH SINGLE ANTIGENICDETERMINANT SPECIFICITIES

Normal serum contains 1016 antibody molecules permilliliter. These antibodies can be collected from exper-imental animals and have long been an important toolof investigators, who have used them to identify or la-bel molecules or cells and to separate molecules or cellsfrom mixtures. A concern, however, has always beenthe variability of antisera (Figure 4-11, see page 72).The development of hybridization techniques allowed

for the production of one kind of specific antibody byimmortalized antibody-producing cells. Methods werethen developed to screen for and produce large num-bers of these antibodies for use in many applications.

Antibodies of a single idiotype produced by im-mortalized B cells are called monoclonal antibodies.Regrettably, normal antibody-secreting cells are endcells of a differentiation series; thus they cannot bemaintained in culture. In contrast, myeloma cells areimmortal. Therefore: “Why not use immunoglobulinsproduced by myeloma cells?” The reason myeloma im-munoglobulins cannot be used is twofold: (1) theirantigen specificity is usually unknown and (2) it is dif-ficult to tailor-make antigen-specific myelomas. Thesecond problem is overshadowed by the question “Isthe combining site of the myeloma protein an accuraterepresentation of the antibody produced during animmune response?” What would happen if one com-bined the characteristics of each cell into one? GeorgesKohler and Cesar Milstein fused cells secreting anti-body of one specificity with myeloma cells. The result-ing clone of cells is a hybrid-myeloma or a hybrid-oma. The hybridoma cells inherit the lymphocyte’sproperty of specific-antibody production and the im-mortality of the myeloma cell. In 1975, Kohler and Mil-stein published a short paper in Nature detailing howcontinuous cultures of fused cells secreting a mono-clonal antibody of predefined specificity were pro-duced. They were awarded the 1984 Nobel Prize inPhysiology and Medicine for their work.

Monoclonal Antibody Development Improveson Nature

How are hybridomas produced? Mice are immunizedwith specific antigens (Figure 4-12, see page 73), andspleens from hosts with the highest titers are collected.The separated spleen cells are mixed with myelomacells that cannot produce immunoglobulins of theirown; thus, the myeloma cells do not interfere with theproduction of normal antibody by the fused B cells.Mouse spleen cells are mixed with the myeloma cellline in a ratio of roughly 10:1 in the presence of poly-ethylene glycol to change membrane permeability andallow cell fusion.

Because cell fusion is random, the cell culture con-tains a mixture of myeloma–spleen cell fusions, my-eloma–myeloma fusions, spleen–spleen cell fusions,single myeloma cells, and single spleen cells. Selectionfor only myeloma–spleen cell fusions is accomplishedby culturing the cell mixture in hypoxanthine-aminopterin-thymidine (HAT) medium (see Figure4-12). Aminopterin is a folic acid analog that blocks thede novo biosynthesis of purines and pyrimidines vitalfor DNA synthesis. Myeloma cells used in hybridoma

M O N O C L O N A L A N T I B O D I E S 71


Antibodies are divided, based on their heavy chainstructure, into six chemically distinct classes—fourkinds of IgG and two kinds of IgA, plus IgM, IgD,and IgE. Each class plays a different role in the im-mune system’s defense strategy. IgG, the major an-tibody in the blood, coats microorganisms tospeed up their uptake by phagocytic cells. Thedimeric IgA concentrates in body fluids to guardthe entrances of the body. IgM, a pentamer, is themost effective activator of complement (and there-fore, is the best indirect killer of blood-borne bac-teria). IgD is found almost exclusively on the sur-face of B cells, where it may regulate the cell’sactivation. IgE, found in trace amounts in theblood, attaches itself to specialized cells, where ittriggers the symptoms of allergies.

Antibodies have many biologic activities. Theycan prevent toxins and viruses from entering cells,whereas other antibodies coat bacteria or targetcells to make them more palatable or vulnerable toattack by certain immune cells. Antigen–antibodycomplexes also can activate complement. Someantibodies, like IgG, cross the placenta and pro-vide protection to the fetus. IgE can promote aller-gies. IgG-containing antigen-antibody complexesattach to Fc receptors on B cells, and inhibit B-cellactivation.

production lack the enzyme hypoxanthine-guaninephosphoribosyl transferase (HGPRT2), so they cannotuse the exogenous hypoxanthine to synthesize purinesby the salvage pathway. Aminopterin also blocks theirendogenous synthesis of purines and pyrimidines. InHAT medium, the myeloma–myeloma cells and freemyeloma cells die in the first week. Although singlespleen cells and spleen–spleen cell fusions expressHGPRT and therefore are not selected against by HATmedium, these cells have limited growth in cultureand die in 2 weeks. Myeloma cells (HGPRT2) thathave fused with HGPRT+ spleen cells now have the en-zyme and can grow in HAT medium. About 500 hy-brids are formed per mouse spleen, and 20 to 30 ofthem produce specific antibody. The correct antibody-producing hybridomas are then identified.

Special Procedures Are Needed to Screen forand Isolate Monoclonal Antibodies

One commonly used method for identifying specifichybridomas is the enzyme-linked immunosorbent as-

say (ELISA) (see Figure 5-9 in Chapter 5). Antigen usedto immunize the mice is bound to microculture wells,and supernatants from each of the clones are placed inthe antigen-coated wells. After incubation, an enzyme-coupled antibody with specificity for mouse antibodiesis added to the wells. After further incubation, thewells are washed to remove unbound antibody, and asubstrate for the enzyme is added. If the wells containantibodies bound to the test antigen, a brightly coloredreaction product develops.

Two general methods may be used to isolate indi-vidual clones of cells. To clone by limiting dilution, thehybridoma is diluted to a density of 1 to 10 cells permilliliter and dispensed into microculture wells. Seed-ing cells at such low concentrations implies that sta-tistically each well contains only one clone. Once theclone covers 10 to 50% of the well’s surface, themedium can be tested for the presence of antibody. Toclone in soft agarose with either fibroblasts or thymo-cytes added as a feeder layer, hybridoma cells areplated over agarose to immobilize them. Within 10days, the individual clones are picked out and placed


CONVENTIONAL ANTIBODYPRODUCTION

MONOCLONAL ANTIBODYPRODUCTION

Antigenicdeterminants

The mouse produces B cellsspecific for each

antigenic determinant.

Repeated antigeninoculation

Removespleen

Spleen cells

Antiserum(contains a mixture of antibodies

specific for each antigenicdeterminant)

FUSION

Antigen

Hybrid-myeloma cells

Myeloma cells

Grow upindividual cells

into clones

Individual populations of antibodies

Clone 1 Clone 2 Clone 3

FIGURE 4-11 Conventional antisera compared with monoclonal antibody produc-tion. The main difference between the two ways of producing antibodies is the resulting prepara-tions;one is heterogeneous,while the other is homogeneous.B-cell-derived hybridomas can be separatedinto individual clones and grown indefinitely. One clone gives rise to identical daughter cells, all produc-ing one antibody idiotype (a monoclonal antibody) directed against a specific antigenic determinant. Seetext for further explanation.

in microculture wells with medium. After a week, theclones can be tested for antibody production.

Hybridomas can be grown in large quantities us-ing roller bottles. One liter of spent medium can yield10 to 100 mg per milliliter of antibody. Alternatively,hybridomas can be grown in the peritoneal cavity of amouse as an ascites (a tumor that usually grows as asingle cell and produces a fluid). Ascitic fluid containsfrom 1 to 25 mg per milliliter of antibody.

The production of monoclonal antibodies is notas easy as it sounds. From immunization to the estab-lishment of a cloned cell line can take 3 to 6 months of

continuous bench work. The procedure is expensive,time-consuming, labor-intensive, and gives no guar-antee of success (Fast Fo-cus 4). Immunologists aretrying a new method ofmonoclonal antibody pro-duction that uses a clon-ing system in Escherichiacoli. If the entire repertoireof an animal can be ex-pressed easily in bacteria, the current methods of pro-ducing monoclonal antibodies may become obsolete.

M O N O C L O N A L A N T I B O D I E S 73

Fusion

Repeated antigeninoculation

Remove spleen

Culture in HATmedium

Nonsecretor myeloma cells(HAT sensitive)

Culture may contain:1 2

3

1

2

3

4

5

Spleen cells (end cells, thuswill die with time)Myeloma cells (killed by HAT medium)Fused spleen cells (sameas no. 1)Fused myeloma cells (same as no. 2)Fused spleen cells—myelomacells (hybrid cell = hybridoma;lives forever [myeloma cells]and produces antibodies[spleen cells or plasmacells])

12

De novo pathway (endogenous sources)Salvage pathway (exogenous sources)

5

4

Test for positive cultures(contain antibody against antigen)

Clone antibody producers

CELLULAR DNA SYNTHESIS

Spleen cells grown inHAT medium

(HAT resistant)

De novo

Ribonucleotide

Ribonucleotide

H

TTK

HGPRT

DNA

FIGURE 4-12 Standard procedure for the development of monoclonal antibodies.Hypoxanthine (H) and thymidine (T) are required for the salvage pathway. Aminopterin (A) blocks denovo DNA synthesis; thus,myeloma cells deficient in HGPRT enzyme cannot grow in HAT medium.How-ever, B-cell–myeloma cell hybrids have the HGPRT enzyme; thus, they can grow because they use the sal-vage pathway for DNA synthesis.TK, thymidine kinase. See text for further explanation.

FAST FOCUS 4

Kohler and Milstein could notduplicate their first experi-ments and only after 6 monthsof ruined experiments did theydiscover that the problem wascaused by a toxic batch ofreagent—the “fun” of R Esearch!

Monoclonal Antibodies Have Many Applications

Monoclonal antibodies ushered in a new era in im-munologic research and in the application of immuno-logic assays to basic and clinical questions. The impactof hybridoma technology is obvious when we begin tolook at the many situations in which it has been suc-cessfully used (Table 4-3).

Monoclonal antibodies are not without problems:(1) the affinity for antigen may be low, (2) individualspecies of monoclonal antibodies do not readily activatecomplement or precipitate or agglutinate antigens invitro, and (3) monoclonals are difficult to use in vivo inhumans because of the difficulty of producing human,as opposed to mouse, hybridomas. To avoid these prob-lems, antibody–gene DNA that has been manipulated in


TABLE 4-3 Applications of HybridomaTechnology

1. Generation of routine serologic reagents2. Monoclonal antibodies against viruses3. Monoclonal antibodies against parasites4. Monoclonal antibodies to define mouse and human

differentiation and tumor antigens5. Monoclonal antibodies to the human MHC-encoded

antigens6. In clinical pharmacology, immunosuppressants in

transplantation and autoimmune reactions and a targetingvehicle for treatment of cancers

7. In cell biology studies, such as anti-tubulin and anti-actinantibodies

8. In plant physiology studies

IN VITRO MANIPULATIONS

3.

1.Mouse H

chain gene

Mouse–humanrecombinantheavy chain

gene

HUMAN–MOUSERECOMBINANT

ANTIBODY

LeaderVDJ

CH1

Hinge

CH2

CH3

LeaderVDJ

CH1

Hinge

CH2

CH3

Human Hchain gene

4.

+

2. INTRODUCTION INTOE. coli plasmid

Proliferating E. coli containing the plasmid

E. coli spheroplast

Myeloma cell

Tissue culture

Ascites

SelectionTRANSFECTANT

TRANSFECTION

EXPANSION

+L chain gene

FIGURE 4-13 Transfectomas—the production of a hybrid antibody molecule. The invitro manipulations include ligating the gene containing the sequence of interest (for an antibody, en-zyme, toxin, and so on; in this case, combining mouse and human antibody genes into one molecule) intoan expression vector such as a eukaryotic cell plasmid.The plasmid is introduced into E.coli, the bacteriaare grown, and the E. coli containing the plasmid with integrated DNA are selected for.These cells aretreated with lysozyme to remove the bacterial cell walls, and the resulting spheroplasts are fused (trans-fection) with myeloma cells. After fusion, the stable cells (transfectants) are identified and selected.These cells represent the transfectomas. Last, the transfectomas are amplified by growing them as as-cites or in tissue culture, and the resulting recombinant antibodies are isolated.

vitro is introduced into myeloma cells. The process iscalled transfection, and the resulting transfected cellsare called transfectomas. Transfection is the integra-tion of donor DNA into a cell’s chromosomes (Figure 4-13, see page 74). If the recipient cell lacks the genetictrait encoded by the donor DNA, the recipient cell ac-quires that trait.8 Transfection uses several recombinantDNA techniques to produce an entire new family ofnovel, tailor-made monoclonal antibodies called eitherhybrid, chimeric, or recombinant, monoclonal an-tibodies. The goal is to humanize antibody.

Custom-made antibodies can be molecules withvariable regions joined to different isotypic constantregions. These antibodies could enhance the bindingof antigen-specific antibodies to protein A (a cell wallcomponent of staphylococci that binds specifically tothe Fc portion of IgG molecules) by changing the Fcpart, or they could change the antibody’s ability tobind complement (keep the same antigen reactivitybut change effector function). Custom-made antibod-ies also can be generated that have unique heavy- andlight-chain combinations. This allows one antibodymolecule to bind two different antigenic determinants.Furthermore, molecules with Fab antibody sequencescan be fused with nonantibody sequences (such as en-zyme or toxin sequences). This approach could makechimeric antibodies that are part antibody and part en-zyme (which has use in immunoassays) or part toxin(antibody directs toxin to a specific site). The applica-tions of these recombinant antibodies seem limitless.

Why not bypass hybridoma technology and evenimmunization? You can. Lymphocyte V-region generepertoires are harvested or assembled in vitro andcloned for display of heavy- and light-chain Fab frag-ments on the surface of bacteriophage, thus the namephage display antibodies. Rare phage display antibodyis selected by binding to antigen, and soluble Fab frag-ments are expressed from infected bacteria. Ironically,we are using the bacteria that our immune system isdesigned to defend against to make antibodies.

SUMMARY

Humoral immunity involves a class of immunoglobu-lins called antibodies. The electrophoretic mobilityof antibodies usually places them in the g-globulinfraction of serum. Antibodies can directly inactivateantigens and indirectly lead to their destructionthrough enhanced phagocytosis and complement acti-vation. Antibody classes differ in antigenicity, charge,function, and size. These differences explain antibodyfunctions such as (1) versatility in antigen binding, (2)binding specificity, and (3) biologic activities.

A typical antibody molecule (the prototype classof immunoglobulin is IgG) is made up of four poly-peptide chains with a molecular weight of about 150kD. The four chains are divided into two identicallight chains and two identical heavy chains. An an-tibody molecule is Y-shaped, with two identical anti-gen-binding sites at the ends of the arms of the Y.The light and heavy chains contribute to the antigen-binding sites. Each antibody molecule can bind totwo identical antigenic determinants. Where the armsmeet, the stem of the Y is known as the hinge region.The hinge region allows segmental flexibility of theantibody molecule. The two antigen-binding ends (oramino-terminal ends) of the antibody molecule arecalled the Fab fragments (for fragment antigen bind-ing), whereas the stem (or carboxyl-terminal end) ofthe Y is considered to be the Fc fragment (for frag-ment crystallizable). The Fc region of the antibodymolecule is responsible for its biologic properties,which include activation of the complement system(leading to enhanced phagocytosis), placental transfer,and binding to cell-surface receptors. Myeloma pro-teins were important in early characterization studieson antibodies.

The amino-terminal end of an antibody is calledthe variable, or V, region and the carboxyl-terminalend is called the constant, or C, region. The C regionis about the same size as the V region in the light chainand three to four times larger than the V region in theheavy chain. The V regions of light and heavy chainsform the antigen-binding sites. Light and heavy chainsconsist of repeating, similarly folded homology unitsor domains. The light chain has one V-region (VL) do-main and one C-region (CL) domain, whereas theheavy chain has one V region (VH) and three or four C-region (CH) domains. The most variable parts of the Vregions are limited to several small hypervariable re-gions or complementarity-determining regions(CDRs). The light- and heavy-chain V regions containthree CDRs. The CDRs come together at the amino-terminal end of the antibody molecule to form theantigen-binding site, which determines specificity. Theinvariant regions of amino acids, between the CDRs,

S U M M A R Y 75


Antigen-specific B cells from an immunized mousecan be fused with immortal B-cell myeloma cells toform hybridomas. Hybridomas have both the spe-cific antibody-producing ability of the B cell andthe immortality of the tumor cell. Transfectomas(cells resulting from the integration of donor DNAinto the recipient cell’s chromosomes) can be gen-erated to produce tailor-made antibodies, calledhybrid, chimeric, or recombinant monoclonals.

8Transfection is the mammalian counterpart of bacterial transforma-tion.

make up about 85% of the V regions and are desig-nated framework residues.

Antigenic determinants that define antibody classesare called isotypic determinants, and the classes arereferred to as isotypes. Isotypic determinants also de-fine subclasses of antibodies. The IgG class consistsof four subclasses, while the IgA class has two sub-classes. Light chains are made up of two classes, k andl. Either k or l chains can be associated with any classof heavy chain. A kl chain combination never occurson the same antibody molecule. Only one isotype ofthe k chain exists, while there are four isotypes of lchains. Allotypic determinants define antibodiescalled allotypes and are encoded by genes that haveseveral alleles. Allotypes have been characterized forall four subclass chains of IgG, the IgA2 subclass chain,and the k chain. They are called Gm, Am, and Km, re-spectively. Idiotypic determinants define antibodymolecules called idiotypes. These determinants are lo-cated at the antigen-binding site.

The five different classes of immunoglobulinsare called IgG, IgA, IgM, IgD, and IgE, each with adistinctive heavy chain designated g, a, m, d, and ε, re-spectively. IgG is the major antibody in the blood, butit is capable of entering tissue spaces, where it coatsantigens, speeding their uptake. IgA concentrates inbody fluids to guard the entrances of the body. Theextracirculatory dimeric IgA and the circulating IgMare polymerized by a glycopeptide J chain. A glyco-protein secretory component is added to dimeric IgAby glandular epithelial cells. This antibody molecule iscalled secretory IgA. IgM is the largest antibody; ittends to remain in the blood, where it can efficientlykill bacteria. IgD remains membrane-bound and some-how regulates the cell’s activation. IgE is found in traceamounts in the blood and triggers allergies.

When antibody-forming cells are fused with my-eloma cells, the resulting clone of cells is a hybrid-oma. Hybridoma cultures are the source of mono-clonal antibodies of predefined antigen specificity.Use of monoclonal antibodies has revolutionized mod-ern biology. Transfectomas (transfected cells) haveled to the creation of novel, tailor-made monoclonalantibodies called either hybrid, chimeric, or recom-binant monoclonal antibodies.

SELF-EVALUATION

RECALLING KEY TERMS

MULTIPLE-CHOICE QUESTIONS

(An answer key is not provided, but the page[s] location of the answer is.)

1. A human myeloma protein (IgM k) is used to immunize a rabbit.The resulting antiserum is then absorbed with a large pool ofIgM purified from normal human serum. Following this absorp-tion, the antiserum is found to react only with the particular IgMmyeloma protein used for immunization; it is now defined as ananti-idiotypic antiserum. With what specific portion(s) of theIgM myeloma protein would the antiserum react? (a) constantregion of the k chain, (b) variable regions of the m and k chains,(c) constant region of the m chain, (d) J chain, (e) none of these.(See pages 66 and 67)

2. Which of the following statements concerning papain cleavageof IgG is incorrect? (a) the Fc fragment contains part of theheavy chain, (b) the Fab fragment contains both heavy-chainand light-chain determinants, (c) the Fc fragment interactsspecifically with certain lymphoid cell receptors, (d) the Fabfragment is divalent, (e) antisera prepared against the Fc frag-ment are specific for the IgG class of immunoglobulins, (f) noneof these. (See pages 60 and 61)

3. The antigen-binding sites of an antibody molecule are located inthe (a) C region of light chains, (b) V region of light chains, (c)C region of heavy chains, (d) V region of heavy chains, (e) re-gions of both light and heavy chains, (f) none of these. (Seepages 63, 64, and 67)

4. The class-specific antigenic determinants of immunoglobulinsare associated with the (a) light chain, (b) heavy chain, (c) Jchain, (d) secretory component, (e) none of these. (See pages 66and 67)

5. The greater resistance of secretory IgA to proteolytic enzymes isassumed to be a consequence of the (a) binding of the secretorycomponent, (b) predominantly dimeric nature of secretory IgA,(c) antiprotease activity of the heavy chain, (d) presence of the Jchain, (e) none of these. (See page 68)

6. IgG is not (a) the major antibody that confers passive immu-nity on the fetus, (b) divided into five subclasses, (c) made up of


Allotypes

Allotypic determinants

Am

Gm

Km

Antibodies

Antibody subclasses

Constant (C) region

Fab fragment

Fab9 fragment

F(ab9)2 fragment

Fc fragment

Framework residues

Gm markers

Heavy (H) chain

Hinge region

Hybrid, chimeric, orrecombinant, monoclonalantibodies

Hybridoma

Hypervariable regions (or complementarity-determining regions[CDRs])

Hypoxanthine-aminopterin-thymidine (HAT) medium

Idiotypes

Idiotypic determinants

Immunoglobulin (orantibody) classes

Immunoglobulin domains

Immunoglobulin genesuperfamily

Immunoglobulins

Isotype

Isotypic determinants

Joining (J) chain

Km markers

Light (L) chain

Monoclonal antibodies

Multiple myeloma

Secretory IgA

Segmental flexibility

Tail pieces

Transfection

Transfectomas

Variable (V) region

two antigen-binding sites, (d) any of these. (See pages 68 and70)

7. IgM is (a) not involved in agglutination (clumping) reactions,(b) found as a pentamer on the B lymphocyte’s surface, (c) thepredominant antibody in the serum, (d) a pentamer in theblood, (e) none of these. (See page 69)

8. IgD (a) is the predominant antibody in colostrum, (b) is presenton the surface of all mature B cells, (c) has a molecular weightof 150,000, (d) mediates precipitation reactions, (e) is none ofthese. (See page 70)

9. IgE (a) increases in individuals with parasitic infections, (b) isthe major agglutinating antibody, (c) is a and b, (d) is none ofthese. (See page 70)

10. Which of the following represent allotypes? (a) l light chains,(b) IgA2, (c) Gm, (d) Km, (e) c and d, (f) none of these. (Seepages 66 and 67)

11. Idiotypes are (a) located in the C region of the light chain, (b)located in the C region of the heavy chain, (c) located in andaround the antigen-binding region of an antibody molecule, (d)none of these. (See pages 66 and 67)

12. The fusion of what cells leads to a hybridoma cell line that se-crets monoclonal antibodies? (a) plasma cells and myelomacells, (b) monocytes and lymphocytes, (c) lymphocytes andlymphoma cells, (d) plasma cells and activated T cells, (e) noneof these. (See pages 71 and 72)

13. Transfectomas are (a) antibodies infected by viruses, (b) hy-bridomas resulting from fusion induced by viruses, (c) mono-clonal antibodies that stop bacterial transfection, (d) cells thatproduce hybrid antibodies resulting from transfection of bacte-ria with antibody genes, (e) none of these. (See pages 74 and75)

SHORT-ANSWER ESSAY QUESTIONS

1. The structure of an antibody molecule is related to its function.Explain.

2. The analysis of antibody molecules by papain and pepsin frag-mentation led to similar yet different results. Explain.

3. What is the difference between an immunoglobulin and a my-eloma protein? Why were myeloma proteins critical to the earlystudies on antibody structure?

4. The variability of the amino acid sequence in the antibody’s Vregion was not random but precisely organized. Explain. Differ-entiate between complementarity-determining regions andframework regions.

5. What are antibody domains?

6. Explain the statements “antibodies can be antigens” or “antibod-ies are used to characterize antibodies.”

7. Briefly discuss immunoglobulin isotypic, allotypic, and idiotypicdeterminants. Give examples of each.

8. Why can’t light chains be used to classify antibodies?

9. List the five classes of antibodies and give a key biologic prop-erty of each.

10. Compare conventional antibody production with monoclonalantibody production.

11. In the development of monoclonal antibodies, five different cel-lular fusion combinations can result. What are they? How canyou isolate the appropriate combination?

12. The use of transfectomas to create novel, tailor-made mono-clonal antibodies has reduced some of the limitations of mono-clonal antibodies. Explain.

FURTHER READINGS

Alzari, P. M., M.-B. Lascombe, and R. J. Poljak. 1988. Three-dimen-sional structure of antibodies. Annu. Rev. Immunol. 6:555–580. Asummary of the relationship between the structure (antigenicity of anti-body molecules) and function (antigen recognition by antibody mole-cules).

Burton, D. R. 1987. Structure and function of antibodies. In Molecu-lar Genetics of Immunoglobulins, F. Calabi and M. S. Neuberger,eds. Amsterdam: Elsevier. A more recent reference text on the correla-tions between antibody structure and function.

Burton, D. R., and J. M. Woof. 1992. Human antibody effector func-tion. Adv. Immunol. 51:1–84. An exhaustive, detailed review of anti-bodies once they interact with antigen.

Campbell, A. M. 1984. Monoclonal Antibody Technology. New York:Elsevier. A cookbook covering all aspects of monoclonal antibody pro-duction.

Davies, D. R., and H. Metzger. 1983. Structural basis of antibodyfunction. Annu. Rev. Immunol. 1:87–117. An examination of thestructure of antibodies according to high-resolution data on antibody-combining sites and specificity of binding to hapten.

Harris, L. J., S. B. Larson, K. W. Hasel, J. Day, A. Greenwood, and A.McPherston. 1992. The three-dimensional structure of an intactmonoclonal antibody for canine lymphoma. Nature 360:369–372.A recent look at the crystal structure of an IgG2a anti-tumor monoclonalantibody using X-ray diffraction.

Kabat, E. A. 1976. Structural Concepts in Immunology and Immuno-chemistry, 2nd ed. New York: Holt, Rinehart and Winston. Chapter9 gives a concise summary of the structural basis of antibody specificity.

Kindt, T. J., and J. D. Capra. 1984. The Antibody Enigma. New York:Plenum Press. A personal view of the antibody diversity question pre-sented in a chronological and methodological story form. Chapters 1–4detail how structure-function studies presented some answers to thequestion of antibody diversity.

Kohler, G. 1986. Derivation and diversification of monoclonal anti-bodies. Science 233:1281–1286. An adaptation of the author’s lecturedelivered when he received the 1984 Nobel Prize in Physiology andMedicine.

Kohler, G., and C. Milstein. 1975. Continuous cultures of fused cellssecreting antibody of predefined specificity. Nature 256:495–497.The benchmark paper by Kohler and Milstein that started the monoclonalantibody revolution through the use of hybridoma technology. The au-thors received the 1984 Nobel Prize in Physiology and Medicine for theirwork.

Mestecky, J., and J. R. McGhee. 1987. Immunoglobulin A (IgA): Mol-ecular and cellular interactions involved in IgA biosynthesis andimmune response. Adv. Immunol. 40:153–246. A comprehensivereview of the structural, functional, and cellular aspects of IgA, concen-trating on new information on the effect of IgA physiology on the immunesystem.

Milstein, C. 1980. Monoclonal antibodies. Sci. Am. 243:66–74. De-scription in a nutshell of the development of monoclonal antibodies byone of the originators.

Milstein, C. 1986. From antibody structure to immunological diversi-fication of immune response. Science 231:1261–1268. An adapta-tion of the author’s lecture delivered when he received the 1984 NobelPrize in Physiology and Medicine.

Morrison, S. L. 1985. Transfectomas provide novel chimeric antibod-ies. Science 229:1202–1207. A review of the new technology for theproduction of custom-made recombinant antibodies.

Nisonoff, A. 1985. Introduction to Molecular Immunology, 2nd ed.Sunderland, Mass.: Sinauer Associates. Presents immunology with amolecular biology flavor.

S U M M A R Y 77

Poljak, R. J. 1975. X-ray diffraction studies of immunoglobulins.Adv. Immunol. 21:1–33. This landmark study on the X-ray diffrac-tion analysis of the Fab fragment of a human myeloma protein providedthe first insight into the three-dimensional structure of the antibody mole-cule.

Spiegelberg, H. L. 1974. Biological activities of immunoglobulins ofdifferent classes and subclasses. Adv. Immunol. 19:259–294. Anearly, but still pertinent, review of the many specific biologic activities ofthe different immunoglobulin classes and subclasses.

Springer, T. A., ed. 1985. Hybridoma Technology in the Biosciencesand Medicine. New York: Plenum Press. Exhaustive coverage of theuse of hybridoma techniques by scientists with the right expertise.

Steward, M. W. 1984. Antibodies: Their Structure and Functions.London: Chapman and Hall. A concise book on the immunochemistryof antibodies.

Tonegawa, S. 1985. The molecules of the immune system. Sci. Am.253:122–131. A review of antibody and T-cell antigen receptor specificityin a nicely illustrated format by the 1987 Nobel Laureate in Medicine.

Wu, T. T., and E. A. Kabat. 1970. An analysis of the sequences of thevariable regions of Bence-Jones proteins and myeloma lightchains and their implications for antibody complementarity. J.Exp. Med. 132:211–250. The report that led to the development of theWu and Kabat plots, which show the presence and location of the hyper-variable regions within the variable regions.


ANTIBODY STRUCTURE AND FUNCTION - bioinf · 2016. 7. 4. · c. The antibody’s antigenic...

Documents

Transcript of ANTIBODY STRUCTURE AND FUNCTION - bioinf · 2016. 7. 4. · c. The antibody’s antigenic...