When bacteria, viruses or parasites invade the human body, the immune system responds by producing proteins called antibodies. Antibodies recognize and bind to molecules (known as antigens) on the surface of the invaders. This binding can either neutralize the invader directly or trigger signals that cause other parts of the immune system to destroy it.

Our blood contains a huge range of different antibody molecules that each bind to a different antigen. This is despite most human antibodies having the same basic shape and structure. Six loops, known as complementarity determining regions (CDRs), emerge from the surface of the antibody to form the surface that recognizes the antigen. However, variations in the structure of the loops alter this surface enough to allow different antibodies to recognize completely different molecules.

In 2016, a new class of antibodies was identified. Unlike previously identified antibodies, these molecules had an entire human protein, called LAIR1, inserted into one of their CDR loops. Members of this group of antibodies bind to a molecule, known as a RIFIN, that is found on the surface of human red blood cells that are infected with the parasite that causes malaria.

How do LAIR1-containing antibodies bind to their RIFIN targets? Hsieh and Higgins investigated this question by using a technique called X-ray crystallography to determine the structure of the antibody. This revealed that instead of binding directly to an antigen, all of the six CDR loops in the LAIR1-containing antibody bind to the LAIR1 insert. By doing so, LAIR1 is oriented in a manner that enables it to bind to the RIFIN molecule from the parasite.

This is the first known example of an antibody that recruits another protein to bind to an antigen rather than binding directly to the pathogen itself. A future challenge will be to see if other antibodies exist that use this mechanism and whether it can be employed to design new therapeutic antibodies.

The antigen-binding sites of human antibodies commonly adopt similar structures, with the light and heavy chains each providing three hypervariable loops that combine to form a surface that is complementary to the epitope. While the sequences of these complementarity determining regions (CDRs) are highly variable, five of the six CDRs (L1, L2, L3, H1 and H2) can be classified into a number of relatively small sets, with similar lengths and architectures, and their structures are predictable from sequence (Chothia et al., 1989; North et al., 2011). In contrast, the third CDR loop of the heavy chain (CDR H3) is more structurally diverse, most likely due to its location close to the V(D)J recombination site (Weitzner et al., 2015). Human antibodies typically have CDR H3 lengths of 8–16 residues (Zemlin et al., 2008) while mouse antibodies have CDR H3 lengths of 5–26 residues (Zemlin et al., 2003).

However, recent years have seen the discovery of antibodies with major differences from the norm, in particular due to changes in the length of the third CDR of the heavy chain. A set of antibodies with broadly neutralizing potential against HIV is one such example. Here, the third CDR loop of the heavy chain is elongated, allowing it to reach through the glycan shield that surrounds the gp120 protein to bind an otherwise concealed epitope (McLellan et al., 2011; Pancera et al., 2013; Pejchal et al., 2010). Such antibodies are rare, making the induction of a broadly inhibitory response against HIV a major challenge (Corti and Lanzavecchia, 2013).

In a more extreme example, while the majority of bovine antibodies have CDR H3 loops of around 23 residues, around 10% contain a highly elongated third CDR loop of up to 69 residues, containing a small disulphide rich domain (Saini et al., 1999; Wang et al., 2013). These domains adopt a conserved β-sheet structure that displays variable loops and are each presented on an elongated, but rigid β-hairpin (Stanfield et al., 2016; Wang et al., 2013). While it is clear that the additional domains play an important role in ligand binding, the remaining five CDR loops are also exposed and further studies are needed to see the contribution that they make (Wang et al., 2013).

A recent study identified a group of even more unusual human antibodies in malaria endemic regions of Africa (Tan et al., 2016). These antibodies were discovered through their capacity to agglutinate human erythrocytes infected with different strains of Plasmodium falciparum, and they bind to a subset of RIFIN proteins. These RIFINs are displayed by the parasite on infected erythrocyte surfaces and are of uncertain function (Chan et al., 2014; Gardner et al., 1998; Kyes et al., 1999). The antibodies show a remarkable adaptation with an intact 96 residue protein, LAIR1, inserted into the third CDR loop of the antibody heavy chain. Indeed, LAIR1 was shown to be essential for the antibody to interact with RIFINs (Tan et al., 2016). In this study, we reveal the structure of the Fab fragment of one of these antibodies, showing how LAIR1 is presented on the antibody surface and drawing conclusions about how this class of antibody can recognize its ligand.

We expressed the two chains that make up the Fab fragment of antibody MGD21 (Tan et al., 2016) in a secreted form from HEK293 cells. This antibody has a kappa light chain (VK1-8/JK5) and a heavy chain in which LAIR1 has been inserted into CDR H3. This fragment was purified and crystallised, allowing a dataset to be collected to 2.52 Å resolution. The structure was determined by molecular replacement using LAIR1 (Brondijk et al., 2010) and the Fab fragment of antibody OX117 (Nettleship et al., 2008) as search models. This identified two copies of the MGD21 Fab fragment in the asymmetric unit of the crystal. A model was built for residues 2–211 of the light chain and 1–351 (with 214–219 and 264–270 disordered) of the heavy chain (Figure 1, Figure 1—figure supplement 1, Figure 1—figure supplement 2, Table 1). The two Fab fragments adopt the same structure with a root mean square deviation of 0.26 Å (calculated over 475 Cα atoms) suggesting a highly ordered linkage between the variable domains of the antibody and the LAIR1 insert (Figure 1—figure supplement 3). The antibody sequence has three putative N-linked glycosylation sites, but of these (light chain N30; heavy chain N61 and N242) only N242 shows electron density corresponding to an Asn-linked N-acetyl glucosamine, in a position distant from the LAIR1 insert.

Figure 1 with 3 supplements see all

Download asset Open asset

The structure shows LAIR1 emerging from the CDR3 loop of the heavy chain and lying across the antigen-binding surface of the variable domains of the Fab fragment (Figure 1A). The long axis of the LAIR1 insert is positioned with the β-strands aligned approximately perpendicular to the groove between the heavy and light chain CDRs and the insertion and linkers interact with, and occlude all five of the remaining CDR loops. The N- and C-termini of LAIR1 lie at opposite ends of its structure, necessitating long linkers between the sites from which CDR3 emerges from the antibody heavy chain and each terminus of the LAIR1 insert (Figure 1A). The N-terminal linker (linker 1) is 10 residues long and adopts a simple loop structure that joins the antibody variable domain to the N-terminus of the LAIR1 insert. The C-terminal linker (linker 2) is longer at 34 residues and is more complex in structure. It extends out from the C-terminus of the LAIR1 insert before zigzagging back towards the insertion site in the heavy chain variable domain. It is stabilized by hydrogen bonds to the LAIR1 insert and to the antibody heavy chain as well as by a disulphide bond to C93 of the antibody light chain. The linkers of the LAIR1-containing antibodies sequenced to date are variable both in length and content, involving different parts of the intronic regions of the LAIR1 gene, or intergenic sequences of chromosome 13 (Tan et al., 2016). The arrangement of these linkers, which radiate away from the remainder of the antibody, will in theory accommodate almost limitless variation in both length and sequence without disturbing the packing of LAIR1 against the variable domains of the antibody.

The five CDR loops lacking the LAIR1 insertion are representatives of previously identified canonical classes (Figure 1—figure supplement 2) (Martin and Thornton, 1996). However, a search using the Abcheck server (Martin, 1996) identified seven unusual residues within the antibody structure; C91, C93, D97 and I106 from the light chain and Y28, R34 and Q54 from the heavy chain, all within the CDR loops. In particular, C91, C93 and D97 all lie in CDR3 of the light chain, perhaps facilitating its interaction with linker 2. Indeed, the most unusual residue is C93, which is found in only 0.096% of light chains, and is the residue that forms a disulphide bond with linker 2 (Figure 1B). The heavy chain CDR H3 loop has a base that adopts the ‘kinked’ conformation (Shirai et al., 1999), with the loop rapidly spreading to the two termini of LAIR1.

In previous structures of antibodies with extended heavy chain CDR3 loops, the remaining five CDRs of the antibody are exposed, with the potential to engage in antigen binding (McLellan et al., 2011; Stanfield et al., 2016, Wang et al., 2013). One of the remarkable features of the LAIR1-containing antibody is therefore the occlusion of large parts of each of the remaining five CDRs, with these loops each interacting directly with the LAIR1 insert and/or linkers (Figure 1B, Table 2). The degree of occlusion of the CDRs by LAIR1 was determined by accessibility to a 1.4 Å probe in the presence and absence of LAIR1 and the linkers. Each of these five CDR loops was partly occluded by the presence of LAIR1 or the linkers (occluding 12.7% of the accessible surface area of CDR L1, 18.3% of CDR L2, 47.3% of CDR L3, 34.7% of CDR H1 and 16.0% of CDR H2). Indeed, both the first and second CDRs of the heavy chain directly contact the LAIR1 insert (Table 2). In addition, each of the three CDR loops of the light chain interacts with one of the two linkers, with interactions including a disulphide bond between C93 of the light chain and C223 of linker 2 (Figure 1C, Table 2). These interactions are replicated in both copies of the molecule in the asymmetric unit of the crystal.

The structure of MGD21 argues for a rigid association of the LAIR1 insert with the remainder of the antibody. Firstly, the structures of the two molecules of the antibody in the asymmetric unit of the crystal superimpose closely (Figure 1—figure supplement 2). It is unlikely that this is due solely to constraints from crystal packing as LAIR1 is anchored to the variable domains of the antibody through three fixed positions: the attachment sites of the two linkers, and the disulphide bond between light chain C93 and heavy chain C223 (Figure 1C). In addition, each of the five CDR loops not baring a LAIR1 insertion makes direct interactions with either LAIR1 or the linker, through contacts found in both copies of the antibody in the asymmetric unit of the crystal. This will stabilize a tight association between LAIR1 and the antibody variable domains. As these antibodies can include multiple different light chains, and very different linkers (Tan et al., 2016), these interaction will not be replicated precisely across the antibody family, but some variant of interaction between light chain CDRs and the linkers is likely.

A comparison of the LAIR1 insert with that of the chromosomal copy of LAIR1 (referred to below as germ line) (Brondijk et al., 2010) reveals that no global structural changes have taken place (root mean square deviation 0.43 Å for the 82 Cα residues) (Figure 2A,F). Indeed, the LAIR1 insertion in the MDG21 antibody differs in only 13 positions relative to the germ line sequence. Mapping these sites onto the structure reveals that they do not alter residues through which the LAIR1 insert interacts with the rest of the antibody (Figure 2B). Presumably, the stable interaction between LAIR1 and the antibody has therefore come instead from adaptations to the CDR loops. In contrast, polymorphisms are mostly located on the surface of LAIR1 distal to the rest of the antibody and have the potential to alter its interaction with its original ligand, collagen, and with the RIFIN proteins which are the target of these antibodies.

Figure 2

Download asset Open asset

The normal function of LAIR1 is to interact with collagen (Meyaard, 2008). The structure of germ line LAIR1, together with NMR analysis and mutagenesis, allowed the mapping of residues critical for the collagen interaction onto a LAIR1 crystal structure (Brondijk et al., 2010). In particular, mutations in residues R59, E61 and R65 have a significant impact on collagen binding (Brondijk et al., 2010). These residues map onto the surface of LAIR1 (Figure 2C) that is most exposed in the context of the antibody (Figure 2D). Indeed, mapping of the polymorphisms found in the 27 LAIR1-containing antibodies sequenced to date shows that large parts of this surface are mutable (Figure 2E). The polymorphisms in LAIR1 include R149N, which is in the position equivalent to R65 in germ line LAIR1 and this change may impact collagen binding. A second polymorphism, found in 7/27 of the antibodies (including MGD21) alters the N-linked glycosyation site at residue 69 of LAIR1 (Wollscheid et al., 2009), which may alter collagen binding and/or increase RIFIN binding, but is not conserved across the antibody family. Indeed 11 of the 27 sequenced antibodies have mutations in at least one of the residues implicated in collagen binding, or other polymorphisms that reduce the interaction (Tan et al., 2016).

The LAIR1-containing antibodies are a remarkable variant of the standard immunoglobulin fold. While the majority of mammalian antibodies have predicable and short CDRs, the third CDR of the heavy chain can accommodate usual diversity (Figure 3) (McLellan et al., 2011; Wang et al., 2013; Weitzner et al., 2015). This is seen in the elongated CDR3 of the broadly neutralizing antibodies that interact with HIV surface proteins and in the insertion of a β-hairpin and disulphide-rich domain in a fraction of bovine antibodies. However, in both of these cases, only the heavy chain CDR3 is altered and the remaining CDR loops remain exposed for antigen binding. The LAIR1-containing antibodies are an exception to this, with the LAIR1-insert interacting with, and partly occluding, all five of the remaining CDR loops. In many ways, the structure resembles an antibody with CDR loops adapted for LAIR1 binding, into which LAIR1 has also been inserted.

Figure 3

Download asset Open asset

This occlusion of large parts of the CDR loops by the LAIR1 insert has major consequences for its role in antigen recognition, as the majority of the antigen-binding surface will be contributed by LAIR1. Indeed, it has been shown that the LAIR1 insert alone can bind to infected erythrocytes, as can a LAIR1-containing antibody with the heavy and light chain regions exchanged (Tan et al., 2016). Surprisingly an antibody in which the LAIR1 insert has been exchanged for the unaltered germ line LAIR1 did not bind to erythrocytes, although the folding of this chimera was not tested (Tan et al., 2016). In addition, the capacity of RIFINs to bind to unaltered LAIR1 alone has not yet been reported. Indeed, it seems most likely that LAIR1, or a highly related homologue, is the physiological ligand of the group of RIFINs that are recognized by these antibodies and that its insertion into the Fab fragment of an antibody allows it to be affinity matured to mobilise it for immune recognition and recruitment of immune cells. This remarkable LAIR1-containing antibody therefore uses the classical hypervariable loops for a novel function: to position an inserted auxillary domain for antigen recognition. The classical Fab fragment therefore now acts as a link between a ligand for a pathogen surface receptor and the Fc region of the antibody with its immune recruitment capability. It will be fascinating to see if this is a paradigm that is repeated in other novel antibodies, as yet undiscovered.

1. 1. Aricescu AR
   2. Lu W
   3. Jones EY

   (2006) A time- and cost-efficient system for high-level protein production in mammalian cells

   Acta Crystallographica Section D Biological Crystallography 62:1243–1250.

   https://doi.org/10.1107/S0907444906029799

   • PubMed
   • Google Scholar
2. 1. Blanc E
   2. Roversi P
   3. Vonrhein C
   4. Flensburg C
   5. Lea SM
   6. Bricogne G

   (2004) Refinement of severely incomplete structures with maximum likelihood in BUSTER-TNT

   Acta Crystallographica Section D Biological Crystallography 60:2210–2221.

   https://doi.org/10.1107/S0907444904016427

   • PubMed
   • Google Scholar
3. 1. Brondijk TH
   2. de Ruiter T
   3. Ballering J
   4. Wienk H
   5. Lebbink RJ
   6. van Ingen H
   7. Boelens R
   8. Farndale RW
   9. Meyaard L
   10. Huizinga EG

   (2010) Crystal structure and collagen-binding site of immune inhibitory receptor LAIR-1: unexpected implications for collagen binding by platelet receptor GPVI

   Blood 115:1364–1373.

   https://doi.org/10.1182/blood-2009-10-246322

   • PubMed
   • Google Scholar
4. 1. Chan JA
   2. Fowkes FJ
   3. Beeson JG

   (2014) Surface antigens of plasmodium falciparum-infected erythrocytes as immune targets and malaria vaccine candidates

   Cellular and Molecular Life Sciences 71:3633–3657.

   https://doi.org/10.1007/s00018-014-1614-3

   • PubMed
   • Google Scholar
5. 1. Chothia C
   2. Lesk AM
   3. Tramontano A
   4. Levitt M
   5. Smith-Gill SJ
   6. Air G
   7. Sheriff S
   8. Padlan EA
   9. Davies D
   10. Tulip WR

   (1989) Conformations of immunoglobulin hypervariable regions

   Nature 342:877–883.

   https://doi.org/10.1038/342877a0

   • PubMed
   • Google Scholar
6. 1. Corti D
   2. Lanzavecchia A

   (2013) Broadly neutralizing antiviral antibodies

   Annual Review of Immunology 31:705–742.

   https://doi.org/10.1146/annurev-immunol-032712-095916

   • PubMed
   • Google Scholar
7. 1. Dunbar J
   2. Deane CM

   (2016) ANARCI: antigen receptor numbering and receptor classification

   Bioinformatics 32:298–300.

   https://doi.org/10.1093/bioinformatics/btv552

   • PubMed
   • Google Scholar
8. 1. Emsley P
   2. Lohkamp B
   3. Scott WG
   4. Cowtan K

   (2010) Features and development of coot

   Acta Crystallographica Section D Biological Crystallography 66:486–501.

   https://doi.org/10.1107/S0907444910007493

   • PubMed
   • Google Scholar
9. 1. Gardner MJ
   2. Tettelin H
   3. Carucci DJ
   4. Cummings LM
   5. Aravind L
   6. Koonin EV
   7. Shallom S
   8. Mason T
   9. Yu K
   10. Fujii C
   11. Pederson J
   12. Shen K
   13. Jing J
   14. Aston C
   15. Lai Z
   16. Schwartz DC
   17. Pertea M
   18. Salzberg S
   19. Zhou L
   20. Sutton GG
   21. Clayton R
   22. White O
   23. Smith HO
   24. Fraser CM
   25. Adams MD
   26. Venter JC
   27. Hoffman SL

   (1998) Chromosome 2 sequence of the human malaria parasite plasmodium falciparum

   Science 282:1126–1132.

   https://doi.org/10.1126/science.282.5391.1126

   • PubMed
   • Google Scholar
10. 1. Hsieh FL
    2. Turner L
    3. Bolla JR
    4. Robinson CV
    5. Lavstsen T
    6. Higgins MK

    (2016) The structural basis for CD36 binding by the malaria parasite

    Nature Communications 7:12837.

    https://doi.org/10.1038/ncomms12837

    • PubMed
    • Google Scholar
11. 1. Kabsch W

    (2010) XDS

    Acta Crystallographica. Section D, Biological Crystallography 66:125–132.

    https://doi.org/10.1107/S0907444909047337

    • PubMed
    • Google Scholar
12. 1. Kyes SA
    2. Rowe JA
    3. Kriek N
    4. Newbold CI

    (1999) Rifins: a second family of clonally variant proteins expressed on the surface of red cells infected with plasmodium falciparum

    PNAS 96:9333–9338.

    https://doi.org/10.1073/pnas.96.16.9333

    • PubMed
    • Google Scholar
13. 1. Martin AC

    (1996) Accessing the kabat antibody sequence database by computer

    Proteins: Structure, Function, and Genetics 25:130–133.

    https://doi.org/10.1002/(SICI)1097-0134(199605)25:1<130::AID-PROT11>3.3.CO;2-Y

    • PubMed
    • Google Scholar
14. 1. Martin AC
    2. Thornton JM

    (1996) Structural families in loops of homologous proteins: automatic classification, modelling and application to antibodies

    Journal of Molecular Biology 263:800–815.

    https://doi.org/10.1006/jmbi.1996.0617

    • PubMed
    • Google Scholar
15. 1. McCoy AJ
    2. Grosse-Kunstleve RW
    3. Adams PD
    4. Winn MD
    5. Storoni LC
    6. Read RJ

    (2007) Phaser crystallographic software

    Journal of Applied Crystallography 40:658–674.

    https://doi.org/10.1107/S0021889807021206

    • PubMed
    • Google Scholar
16. 1. McLellan JS
    2. Pancera M
    3. Carrico C
    4. Gorman J
    5. Julien JP
    6. Khayat R
    7. Louder R
    8. Pejchal R
    9. Sastry M
    10. Dai K
    11. O'Dell S
    12. Patel N
    13. Shahzad-ul-Hussan S
    14. Yang Y
    15. Zhang B
    16. Zhou T
    17. Zhu J
    18. Boyington JC
    19. Chuang GY
    20. Diwanji D
    21. Georgiev I
    22. Kwon YD
    23. Lee D
    24. Louder MK
    25. Moquin S
    26. Schmidt SD
    27. Yang ZY
    28. Bonsignori M
    29. Crump JA
    30. Kapiga SH
    31. Sam NE
    32. Haynes BF
    33. Burton DR
    34. Koff WC
    35. Walker LM
    36. Phogat S
    37. Wyatt R
    38. Orwenyo J
    39. Wang LX
    40. Arthos J
    41. Bewley CA
    42. Mascola JR
    43. Nabel GJ
    44. Schief WR
    45. Ward AB
    46. Wilson IA
    47. Kwong PD

    (2011) Structure of HIV-1 gp120 V1/V2 domain with broadly neutralizing antibody PG9

    Nature 480:336–343.

    https://doi.org/10.1038/nature10696

    • PubMed
    • Google Scholar
17. 1. Meyaard L

    (2008) The inhibitory collagen receptor LAIR-1 (CD305)

    Journal of Leukocyte Biology 83:799–803.

    https://doi.org/10.1189/jlb.0907609

    • PubMed
    • Google Scholar
18. 1. Nettleship JE
    2. Ren J
    3. Rahman N
    4. Berrow NS
    5. Hatherley D
    6. Barclay AN
    7. Owens RJ

    (2008) A pipeline for the production of antibody fragments for structural studies using transient expression in HEK 293T cells

    Protein Expression and Purification 62:83–89.

    https://doi.org/10.1016/j.pep.2008.06.017

    • PubMed
    • Google Scholar
19. 1. North B
    2. Lehmann A
    3. Dunbrack RL

    (2011) A new clustering of antibody CDR loop conformations

    Journal of Molecular Biology 406:228–256.

    https://doi.org/10.1016/j.jmb.2010.10.030

    • PubMed
    • Google Scholar
20. 1. Pancera M
    2. Shahzad-Ul-Hussan S
    3. Doria-Rose NA
    4. McLellan JS
    5. Bailer RT
    6. Dai K
    7. Loesgen S
    8. Louder MK
    9. Staupe RP
    10. Yang Y
    11. Zhang B
    12. Parks R
    13. Eudailey J
    14. Lloyd KE
    15. Blinn J
    16. Alam SM
    17. Haynes BF
    18. Amin MN
    19. Wang LX
    20. Burton DR
    21. Koff WC
    22. Nabel GJ
    23. Mascola JR
    24. Bewley CA
    25. Kwong PD

    (2013) Structural basis for diverse N-glycan recognition by HIV-1-neutralizing V1-V2-directed antibody PG16

    Nature Structural & Molecular Biology 20:804–813.

    https://doi.org/10.1038/nsmb.2600

    • PubMed
    • Google Scholar
21. 1. Pejchal R
    2. Walker LM
    3. Stanfield RL
    4. Phogat SK
    5. Koff WC
    6. Poignard P
    7. Burton DR
    8. Wilson IA

    (2010) Structure and function of broadly reactive antibody PG16 reveal an H3 subdomain that mediates potent neutralization of HIV-1

    PNAS 107:11483–11488.

    https://doi.org/10.1073/pnas.1004600107

    • PubMed
    • Google Scholar
22. 1. Saini SS
    2. Allore B
    3. Jacobs RM
    4. Kaushik A

    (1999) Exceptionally long CDR3H region with multiple cysteine residues in functional bovine IgM antibodies

    European Journal of Immunology 29:2420–2426.

    https://doi.org/10.1002/(SICI)1521-4141(199908)29:08<2420::AID-IMMU2420>3.0.CO;2-A

    • PubMed
    • Google Scholar
23. 1. Shirai H
    2. Kidera A
    3. Nakamura H

    (1999) H3-rules: identification of CDR-H3 structures in antibodies

    FEBS Letters 455:188–197.

    https://doi.org/10.1016/S0014-5793(99)00821-2

    • PubMed
    • Google Scholar
24. 1. Stanfield RL
    2. Wilson IA
    3. Smider VV

    (2016) Conservation and diversity in the ultralong third heavy-chain complementarity-determining region of bovine antibodies

    Science Immunology 1:aaf7962.

    https://doi.org/10.1126/sciimmunol.aaf7962

    • PubMed
    • Google Scholar
25. 1. Tan J
    2. Pieper K
    3. Piccoli L
    4. Abdi A
    5. Foglierini M
    6. Geiger R
    7. Tully CM
    8. Jarrossay D
    9. Ndungu FM
    10. Wambua J
    11. Bejon P
    12. Fregni CS
    13. Fernandez-Rodriguez B
    14. Barbieri S
    15. Bianchi S
    16. Marsh K
    17. Thathy V
    18. Corti D
    19. Sallusto F
    20. Bull P
    21. Lanzavecchia A

    (2016) A LAIR1 insertion generates broadly reactive antibodies against malaria variant antigens

    Nature 529:105–109.

    https://doi.org/10.1038/nature16450

    • PubMed
    • Google Scholar
26. 1. Wang F
    2. Ekiert DC
    3. Ahmad I
    4. Yu W
    5. Zhang Y
    6. Bazirgan O
    7. Torkamani A
    8. Raudsepp T
    9. Mwangi W
    10. Criscitiello MF
    11. Wilson IA
    12. Schultz PG
    13. Smider VV

    (2013) Reshaping antibody diversity

    Cell 153:1379–1393.

    https://doi.org/10.1016/j.cell.2013.04.049

    • PubMed
    • Google Scholar
27. 1. Weitzner BD
    2. Dunbrack RL
    3. Gray JJ

    (2015) The origin of CDR H3 structural diversity

    Structure 23:302–311.

    https://doi.org/10.1016/j.str.2014.11.010

    • PubMed
    • Google Scholar
28. 1. Winn MD
    2. Ballard CC
    3. Cowtan KD
    4. Dodson EJ
    5. Emsley P
    6. Evans PR
    7. Keegan RM
    8. Krissinel EB
    9. Leslie AG
    10. McCoy A
    11. McNicholas SJ
    12. Murshudov GN
    13. Pannu NS
    14. Potterton EA
    15. Powell HR
    16. Read RJ
    17. Vagin A
    18. Wilson KS

    (2011) Overview of the CCP4 suite and current developments

    Acta Crystallographica Section D Biological Crystallography 67:235–242.

    https://doi.org/10.1107/S0907444910045749

    • PubMed
    • Google Scholar
29. 1. Wollscheid B
    2. Bausch-Fluck D
    3. Henderson C
    4. O'Brien R
    5. Bibel M
    6. Schiess R
    7. Aebersold R
    8. Watts JD

    (2009) Mass-spectrometric identification and relative quantification of N-linked cell surface glycoproteins

    Nature Biotechnology 27:378–386.

    https://doi.org/10.1038/nbt.1532

    • PubMed
    • Google Scholar
30. 1. Zemlin M
    2. Klinger M
    3. Link J
    4. Zemlin C
    5. Bauer K
    6. Engler JA
    7. Schroeder HW
    8. Kirkham PM

    (2003) Expressed murine and human CDR-H3 intervals of equal length exhibit distinct repertoires that differ in their amino acid composition and predicted range of structures

    Journal of Molecular Biology 334:733–749.

    https://doi.org/10.1016/j.jmb.2003.10.007

    • PubMed
    • Google Scholar
31. 1. Zemlin M
    2. Schelonka RL
    3. Ippolito GC
    4. Zemlin C
    5. Zhuang Y
    6. Gartland GL
    7. Nitschke L
    8. Pelkonen J
    9. Rajewsky K
    10. Schroeder HW

    (2008) Regulation of repertoire development through genetic control of DH reading frame preference

    The Journal of Immunology 181:8416–8424.

    https://doi.org/10.4049/jimmunol.181.12.8416

    • PubMed
    • Google Scholar

Author details

1. Fu-Lien Hsieh

   Department of Biochemistry, University of Oxford, Oxford, United Kingdom

   Contribution

   F-LH, Conceptualization, Formal analysis, Investigation, Writing—original draft

   Competing interests

   The authors declare that no competing interests exist.

2. Matthew K Higgins

   Department of Biochemistry, University of Oxford, Oxford, United Kingdom

   Contribution

   MKH, Conceptualization, Formal analysis, Supervision, Funding acquisition, Investigation, Methodology, Writing—original draft, Project administration

   For correspondence

   matthew.higgins@bioch.ox.ac.uk

   Competing interests

   The authors declare that no competing interests exist.

   "This ORCID iD identifies the author of this article:" 0000-0002-2870-1955

• 2,597

  views

• 594

  downloads

• 12

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.