Sei sulla pagina 1di 15

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8023065

Nanobodies as novel agents for cancer therapy

Article  in  Expert opinion on biological therapy · February 2005


DOI: 10.1517/14712598.5.1.111 · Source: PubMed

CITATIONS READS
149 2,524

3 authors, including:

Hilde Revets Serge Muyldermans


Ablynx Vrije Universiteit Brussel
46 PUBLICATIONS   3,299 CITATIONS    366 PUBLICATIONS   15,535 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Ultrasensitive Measurement of Ca(2+) Influx into Lipid Vesicles Induced by Protein Aggregates View project

Wild South American Camelids for the production of nanobodies for biotechnological and pharmaceutical applications. View project

All content following this page was uploaded by Serge Muyldermans on 05 June 2014.

The user has requested enhancement of the downloaded file.


Review
Vaccines & Antibodies

Nanobodies as novel agents for


cancer therapy
Hilde Revets†, Patrick De Baetselier & Serge Muyldermans
1. Introduction †VrijeUniversiteit Brussel, Department of Molecular and Cellular Interactions, Laboratory of Cellular
2. What is a nanobody? and Molecular Immunology, Vlaams Interuniversitair Instituut voor Biotechnologie, Pleinlaan 2,
Building E8, B-1050 Brussels, Belgium
3. Generation and production of
recombinant nanobodies
Nanobodies are the smallest fragments of naturally occurring heavy-chain
4. Unique features of nanobodies antibodies that have evolved to be fully functional in the absence of a light

f
5. Achievements and medical chain. As such, the cloning and selection of antigen-specific nanobodies
opportunities for nanobodies obviate the need for construction and screening of large libraries, and for
lengthy and unpredictable in vitro affinity maturation steps. The unique and

o
6. Nanobodies in cancer diagnosis
and therapy well-characterised properties enable nanobodies to excel conventional ther-
apeutic antibodies in terms of recognising uncommon or hidden epitopes,

o
7. Expert opinion and conclusions
binding into cavities or active sites of protein targets, tailoring of half-life,

r
drug format flexibility, low immunogenic potential and ease of manufac-
ture. Moreover, the favourable biophysical and pharmacological properties
of nanobodies, together with the ease of formatting them into multifunc-

r P
tional protein therapeutics, leaves them ideally placed as a new generation
of antibody-based therapeutics. This review describes the state of the art on
nanobodies and illustrates their potential as cancer therapeutic agents.

Keywords: camels, cancer therapy, heavy-chain antibodies, llama, nanobody,

o
single-domain antibody, VHH

h
Expert Opin. Biol. Ther. (2005) 5(1):xxx–xxx

t
1. Introduction

u
The mouse hybridoma technology described by Köhler and Milstein in 1975 [1] was an
important step in the development of antibody technology and paved the way for the
emergence of therapeutic monoclonal antibodies (mAbs). Today, several antibodies have

A
been approved as therapeutic agents in a wide range of indications, including transplant
rejection [2-4], inflammation [5-7], infectious diseases [8-10] and oncology [11-12] (Table 1).
With many more candidates in late-phase clinical trials, including oncology trials,
antibodies have become well-established therapeutics [13-14].
Unmodified mouse mAbs were the first FDA-approved immunotherapeutic
agents (Table 1), but their in vivo applications in humans were limited because their
xenogeneic nature provoked human antimouse antibody (HAMA) responses when
repetitively injected into patients [15-16]. To mitigate potential immunogenicity, ther-
apeutic mAb technology transitioned from creating completely murine mAbs to
exchanging the murine constant domains for human constant domains to create
‘chimeric’ antibodies [17]. This was followed by ‘implanting’ the complementarity-
determining regions (CDRs) of the mouse antibody into the variable regions of a
human IgG, thus creating humanised antibodies [18]. Modern strategies now allow
the selection of fully human antibodies directly from natural or synthetic repertoires,
including transgenic mice producing purely human antibodies [19-20].
A remaining impediment to the development of new antibody therapeutics is the
cost of producing antibodies in mammalian expression systems. Moreover, a rising
Ashley Publications awareness of antibody effector functions stimulates the efforts towards the modula-
www.ashley-pub.com
tion of Fc-associated functions, such as Fc-receptor activity and complement

10.1517/14712598.5.1.xxx © 2005 Ashley Publications Ltd ISSN 1471-2598 1


Nanobodies as novel agents for cancer therapy

Table 1. Approved monoclonal antibodies.

Product* Specificity Type of molecule Indication Year


OKT®3 (muromonab-CD3; Ortho Biotech) CD3 Murine Transplant rejection 1986
ReoPro® (abciximab; Centocor Inc.) Platelet gpIIb/IIIa Chimeric Cardiovascular disease 1994
Panorex® (edrecolomab; Centocor) EpCAM Murine Colorectal cancer 1995
Rituxan® (rituximab; IDEC) CD20 Chimeric Non-Hodgkin’s lymphoma 1997
Zenapax® (daclizumab; Roche) IL-2 receptor Humanised Transplant rejection 1997
Herceptin® (trastuzumab; Genentech, Inc.) Her-2 Humanised Metastatic breast cancer 1998
Remicade® (infliximab; Centocor) TNF-α Chimeric Crohn’s disease 1998
Simulect® (basiliximab; Novartis AG) IL-2 receptor Chimeric Kidney transplant rejection 1998
Synagis® (palivizumab; MedImmune, Inc.)
Mylotarg® (gemtuzumab
Ozogamicin; Wyeth Research)
CroFab™ (Protherics)
RSV (anti-F)
CD33

Snake venom
Humanised
Humanised

Ovine Fab
RSV

f
Acute myeloid leukaemia

o
Rattlesnake antidote
1998
2000

2000

o
DigiFab® (Protherics) Digoxin Ovine Fab Digoxin overdose 2001

r
Campath® (alemtuzumab; Berlex, Inc.) CD52 Humanised B cell chronic lymphocytic 2001
leukaemia
Zevalin® (ibritumomab tiuxetan; IDEC) CD20 (90yttrium Murine B cell non-Hodgkin’s 2002

P
conjugate) lymphoma
Xolair® (omalizumab; Genentech) IgE-Fc Humanised Allergy 2002

r
Bexxar® (tositumomab; Corixa) CD20 Murine Non-Hodgkin’s lymphoma 2003
(radiolabelled)

o
Humira® (adalimumab; Abbott Laboratories) TNF-α Human Reactive arthritis 2003
Raptiva® (efalizumab; Genentech) CD11a Humanised Psoriasis 2003

h
Avastin® (bevacizumab; Genentech) VEGF Humanised Colorectal cancer 2004

t
Erbitux® (cetuximab; ImClone Systems Inc.) EGFR Chimeric Solid tumours 2004
*Product names are registered trademarks.
EGFR: Epidermal growth factor receptor; EpCAM: Epithelial cell adhesion molecule; RSV: Respiratory syncytial virus; TNF: Tumour necrosis factor;

u
VEGF: Vascular endothelial growth factor.

A
activation [21-23]. Consequently, there has been a growing terms of stability [35], expression yield, protease resistance and
interest in new small therapeutic antibody formats that can aggregation caused by synthetic linkers [36].
be tailored ‘at will’ and manufactured by high-volume liquid Minimising the size of antigen-binding proteins to a single
bacterial or yeast cell culture. Gene technology has succeeded immunoglobulin domain with high affinity to a target anti-
in the cloning and engineering of smaller fragments of anti- gen has been one of the major goals of antibody engineering
body genes [24-25]. The scaffold chosen for recombinant anti- over the past decade. Antigen-binding fragments comprising
bodies has been a simple choice between Fab or single-chain the single variable domain of the conventional heavy chains
variable fragment (scFv) molecules, with most groups favour- (variable heavy [VH]) were generated in the late 1980s [37]
ing scFv due to the higher expression levels compared with Fab and are referred to as single domain antibodies or dAbs. These
in prokaryotic and lower eukaryotic microorganisms [26-27]. In dAbs, however, exhibit some serious shortcomings. The anti-
comparison with whole antibodies, small antibody frag- gen-binding site is matured during the immunisation as a
ments, such as Fab or scFv, exhibit better pharmacokinetics combination of VH and variable light (VL), so that isolated
for tissue penetration, but are monovalent and suffer from VHs will have a lower affinity for their antigen, as compared
fast off-rates and poor retention time on the target [28-29]. To with the original, matured antibody [38]. Moreover, the hydro-
overcome these obstacles, various methods have been devised phobic interface becomes exposed to the aqueous environ-
to tailor the functionality and physicochemical properties of ment following removal of the VL domain and this renders
these fragments, leading to improved stability, functional dAbs ‘sticky’. The poor solubility and reduced affinity for the
affinity, tumour penetration and biodistribution when used antigen indicated that additional protein engineering would
in vivo [30-34]. Despite their beneficial properties, scFvs and be required to successfully generate functional single-domain
conjugates thereof are still amenable for improvement in antibody fragments. A recent paper by Jespers et al. [39]

2 Expert Opin. Biol. Ther. (2005) 5(1)


Revets, De Baetselier & Muyldermans

1 VH
VL
CL CH
Fab

Light chain Fv
CH2 Antigen-binding
Fc fragments
scFv
CH3
VH CDR1 CDR2 CDR3

W47
VH

G44
V37

L45
VHH CDR1 CDR2 CDR3
CH2

f
G47
R45
E44
F37
Fc

VHH

o
CH3
nanobody

o
Figure 1. Schematic representation of the conventional (top) and heavy-chain IgG antibodies (bottom) present in camelid
serum. The conventional IgG contains two variable regions (each composed of a VH and VL domain) that confer antigen-binding

r
specificity of the antibody, and an Fc fragment in the constant region that recruits effector functions of the immune system. The entire
light chain and CH1 domain are absent in the HCAb. The antigen-binding domains of conventional antibodies after proteolysis (Fab) or
after cloning and expression of the VH and VL fragments are shown. scFvs are generated by introducing a synthetic linker between the

P
VH and VL domain. The recombinant VHH or nanobody of a heavy-chain of HCAb is obtained after cloning and expression. The
differences between VH and nanobody are also shown. The CDR1 and CDR3 of nanobodies is longer than in VH genes, and they are
often connected by an intersulfide bond (thick line). The hallmark amino acid substitutions in framework-2 and their position according
to the Kabat numbering [57] are given.

r
CDR: Complementarity-determining regions; HCAb: Heavy-chain antibody; scFv: Single-chain variable fragment; VH: Variable heavy; VHH: Camelid variable heavy;

o
VL: Variable light.

described how the aggregation-prone dAbs could be removed The HCAbs of Camelidae have a unique structure consist-

t h
from a repertoire displayed at the tip of filamentous phages by
introducing a heating step. However, by serendipity, it was
discovered that a similar selection for soluble dAbs was already
ing of a single variable domain (VHH), a hinge region and
two constant domains (CH2 and CH3) (Figure 1). These
HCAbs lack the first domain of the constant region (CH1),

u
performed much earlier in nature. Part of the humoral which is present in the genome, but is spliced out during
response of camels, dromedaries and llamas is based on a mRNA processing [42-43]. The absence of the CH1 domain
unique repertoire of fully functional antibodies composed explains the absence of the light chain in the HCAbs, as this

A
solely of heavy chains (Figure 1) [40]. This finding, together domain is the anchoring place for the constant domain of the
with the concurrent improvement of technologies to select light chain [44]. Consequently, HCAbs naturally evolved to
antigen-binding fragments from large libraries and readily confer antigen-binding specificity and high affinity by three
express them in bacteria [25,41], led to the development of a CDRs only, instead of the six naturally occurring CDRs from
new antibody technology platform based on the single varia- conventional antibodies or fragments thereof.
ble domain of these camelid heavy chain antibodies (VHH). The crystal [45-47] and solution [48-49] structures of several
The crystal structure of an isolated VHH indicated that it is nanobodies have been solved and show that their scaffolds
a problate particle of 2.5 nm in diameter and ∼ 4 nm high, consist of two α-sheeted structures similar to a VH immu-
and has been referred to as a nanobody. This review focuses noglobulin fold in a conventional antibody. By contrast, the
on the unique biophysical and pharmacological properties structures of the antigen-binding loops of nanobodies deviate
of nanobodies and will illustrate their potential as cancer importantly from the sets found in mouse and human anti-
therapeutic agents. bodies [50]. This finding provides evidence that the antigen-
binding loops of nanobodies exhibit a much larger structural
2. What is a nanobody? repertoire than observed for conventional VH [51-53]. In addi-
tion, the CDR3 regions of nanobodies are on average longer
A nanobody is the smallest available intact antigen-binding than those of VH. The average CDR3 length in human and
fragment (15 kDa) harbouring the full antigen-binding mouse VH is 12 and 9 amino acids, respectively [54], whereas
capacity of the original heavy chain of naturally occurring in dromedary-derived nanobodies, a length of 16 – 18 amino
heavy-chain antibodies (HCAbs) that have evolved to be fully acids is frequently observed, although in the llama, a consider-
functional in the absence of a light chain. able fraction of the nanobodies seem to have a much shorter

Expert Opin. Biol. Ther. (2005) 5(1) 3


Nanobodies as novel agents for cancer therapy

to generate single domain antigen binders, such as the FN3


domain or the cytotoxic T lymphocyte antigen-4 [72-74]. All of
these single domain variants are stable and monomeric entities
that are well expressed in bacterial expression systems. The
overall dimensions of these soluble particles are similar to those
of a camelid nanobody and, therefore, in this respect, they can
be considered as nanobodies.

3. Generation and production of recombinant


nanobodies
F37,E44 V37,G44
R45,G47 L45,W47 Cloning the repertoire of antigen-binding fragments from
an immunised animal into a phage display vector and selec-

Nanobody Human VH

f
tion of antigen-specific clones by panning has become in the
past decade a routine method for selecting antigen-specific

o
molecules [25,75]. This method is straightforward to adapt to
nanobodies, whereby the single-domain nature of nanobod-

o
Figure 2. The immunoglobulin fold of a nanobody (VHH) ies simplifies the method considerably [76]. Indeed, tapping
compared with a human VH. The CDR3 is shown in black and the antigen-binding repertoire of the HCAbs of an immu-

r
the location of camel mutations on the antibody interface surface
nised dromedary or llama is less complicated with respect to
are indicated. See text for additional explanation.
CDR: Complementarity-determining regions; VH: Variable heavy; the repertoire cloning of conventional antibodies, for exam-

P
VHH: Camelid variable heavy. ple, in the form of scFv, as the intact antigen-binding site is
encoded by a single gene fragment (the nanobody). Gener-
CDR of ∼ 6 amino acids [51,53,55]. Nanobody sequences also ally, cDNA is prepared from peripheral blood lymphocytes,

r
contain amino acid substitutions in the framework regions isolated from an immunised dromedary or llama (Figure 3).
that are not observed in VH domains that pair with VL As all nanobodies belong to one single gene family, they are

o
domains [56]. Hence, these nanobody hallmark substitutions encoded by a single exon with homologous border
might have evolved to compensate for the absence of the VL sequences. Consequently, the complete in vivo matured
domain in the antigen binding site, and they might accom- nanobody repertoire of a single immunised animal can be
modate more flexible loops than a VH.

t
The hydrophobic to hydrophilic amino acid substitutions
(V37F or V37Y, G44E, L45R and W47G) within the frame-
h amplified by a single set of primers. A secondary polymerase
chain reaction with nested primers is then performed to pro-
duce more material and to include restriction enzyme sites

u
work-2 region (the residues in this region of the VH normally for cloning purposes. Following cloning of the amplified
interact with the VL domain and are well conserved throughout nanobody gene fragments in the appropriate expression vec-
evolution [57]) are solvent accessible in the nanobodies (Figure 2). tor, a nanobody library containing the repertoire of the

A
Consequently, they have been allocated to improve the solubility intact in vivo matured antigen-binding sites is obtained [76].
of nanobodies [56,58]. This led to the idea of ‘camelisation’ of Because of the in vivo maturation of the HCAbs, relatively
human VH domains, a technique which involved the replace- small libraries of only 106 – 107 individual nanobody genes
ment of the residues at positions 44, 45 and 47 in human VH have routinely resulted in the isolation of nanobodies with
domains with those found in nanobodies [59]. This process of nanomolar affinity for their antigen [76-78]. Nanobody librar-
camelisation resulted in stable entities where the tendency to ies can be screened for the presence of antigen-specific bind-
aggregate was eliminated [60-63]. Unfortunately, the expression ers either by direct colony screening or by panning. Retrieval
level was sometimes reduced and the original antigen-binding of binders by panning is the preferred method, as panning
activity was lost [61,64]. Further work indicated that besides the allows selection for binders with the highest affinities and
nanobody ‘tetrad’, other mutations together with CDR length those that express better in bacteria. Soluble nanobodies can
and diversity, contribute to the stability and solubility of the then efficiently be produced in bacteria or lower
engineered antibody domains [65-68]. eukaryotes [79]. Despite the success of retrieving high-affinity
Apart from the Camelidae, wobbegong and nurse shark also lead compounds in a short period of time, the method
contain HCAbs [69]. Their variable domain can be cloned and sometimes suffers from the requirement of sufficient
expressed in bacterial systems, and their structure has been amounts of target antigen for immunisation, although good
solved recently [70,71]. Their overall size is similar to the nano- immunisations have been obtained with as little as 100 µg of
bodies from camelids; however, the primary structure, as well antigen (for entire immunisation). The availability of syn-
as the α-strand organisation, is much more divergent to that of thetic nanobody libraries should offer a solution to identify
human or mouse VH or nanobodies. In addition, other scaf- antigen binders in cases where difficulties to immunise are
folds of similar size as camelid nanobodies have been employed encountered (lack of antigen, low immunogenic or toxic

4 Expert Opin. Biol. Ther. (2005) 5(1)


Revets, De Baetselier & Muyldermans

Collect Isolate lymphocytes


blood
Extract mRNA

RT-PCR Select specific


VHHs by panning

Produce soluble
Immunise camel/llama antigen-specific VHH
with immunogens

Generate library

f
of ∼ 107
transformants

for additional explanation.

o o
Figure 3. Schematic overview of the cloning and selection of nanobodies from an immunised dromedary (or llama). See text

4. Unique

r
features of nanobodies
antigen). Synthetic phage or phagemid libraries, containing
a randomised CDR3 on a ‘camelised’ human VH or on a

P
mouse VH, have been constructed [59,80]. From such a reper- The single domain nature of nanobodies endows them with
toire, camelised VH domains specific for hapten, peptide several unique features as compared with antigen-binding
and protein antigens have been selected [61,62]. However, as derivatives of conventional antibodies. Besides the easy cloning

r
the binders obtained from this repertoire are not matured and selection of high-affinity binders from an in vivo matured
against the antigen, the affinity of these camelised human library, nanobodies have other technological and biophysical

o
VH domains were only in the submicromolar range. Like- advantages that enable them to outperform conventional
wise, the availability of a naive, llama nanobody library antibodies in several areas and are discussed below.
seems to be a good source to retrieve antigen-specific binders

t h
against all possible targets [81]. However, here as well, the
affinities seem to be inferior to those of nanobodies retrieved
from ‘immune’ libraries. Several methods have been devel-
4.1 Ease of manufacture and formulation
The expression of the recombinant nanobodies produces sin-
gle-domain proteins of 15 kDa. In contrast to VH domains of

u
oped for the in vitro affinity maturation of recombinant conventional antibodies, nanobodies are efficiently expressed
antibody fragments from synthetic or naive libraries to as soluble, non-aggregating recombinant proteins due to their
improve the affinity of these weakly binding antibody frag- framework substitutions in framework-2: V37Y or V37F;

A
ments [82-84]. This approach involves, again, the randomisa- G44E; L45R or L45C and W47, most often substituted by
tion of some residues of the selected camelised VH domain G. Without optimisation, recombinant nanobodies are rou-
and the construction of a second library that should be as tinely obtained at levels up to 10 mg/l when expressed in
large as possible to have a fair chance of success. Alterna- E. coli grown in shake-flask cultures [76]. The production
tively, the ribosome display method has been employed for process is scalable and expression systems other than bacteria
the same purpose of affinity improvement [81]. Considering can be used. The high-level secretion of nanobodies from a
the time that is spent to increase the first generation binders range of fungal and yeast species has been shown [85], with
from synthetic and naive libraries to bring their affinity to secretion by Saccharomyces cerevisiae at levels > 100 mg l-1
the required level, it might in the end be much faster to from shake-flask cultures and > 1 g l-1 from a 10 l fed-batch
retrieve high-affinity nanobodies from an immunised drom- fermentation [86]. Scaled-up yields of > 1 kg of nanobody were
edary or llama. This strategy obviates the need for construc- obtained from a 15-m3 fermentation [87]. Performing the final
tion and screening of large libraries, and for lengthy and round of nanobody selection in the production organism, for
unpredictable affinity maturation steps. Starting from example, by yeast surface display, further improves maximum
immune animals, one can also apply an alternative, even more yields [88]. Due to their superior stability compared to conven-
simple method, avoiding library construction altogether. This tional antibodies, nanobodies can be formulated as a long
involves the direct cloning of VHH genes from purified, anti- shelf-life, ready-to-use solution.
gen-specific B cells. Combined with the ease of manufacture,
high-affinity nanobodies can be generated within a few 4.2 Nanobodies are highly soluble and stable
weeks from accessing the target, although some purified Nanobodies are naturally soluble in aqueous solution, proba-
antigen is required for the immunisation (and selections). bly due to the substitution of hydrophobic by hydrophilic

Expert Opin. Biol. Ther. (2005) 5(1) 5


Nanobodies as novel agents for cancer therapy

residues in the framework-2 region. VH from conventional receptor–ligand [99]. The antigen-binding surface of conven-
antibodies interact through hydrophobic areas with the CH1 tional antibodies forms a concave or flat surface, depending
and VL domains. As a consequence, the separate expression on the antigen [100]. Large protruding flexible loops are not
of the VH domain only leads to the formation of inclusion encountered frequently at the antigen-binding surface of
bodies or to folded domains exposing hydrophobic patches conventional antibodies because these would be immobi-
which render them sticky. Replacement of the hydrophobic lised following antigen interaction and would have a nega-
framework-2 residues in these VH domains by the nano- tive influence (entropic effect) on the binding energy [101].
body-specific sequences results in increased solubility and In nanobodies, the long CDR3 loop usually accounts for
lack of aggregation [61]. most of the binding interaction with the antigen. The crys-
A striking property of several nanobodies is their high ther- tallographic structure of a recombinant dromedary nanobody
mal and conformational stability. It has been shown that in complex with lysozyme illustrated how the N-terminal part
nanobodies retain > 80% of their binding activity after of the long CDR3 loop forms a loop that extends from the
1 week of incubation at 37°C [76], indicating that nanobodies remaining antigen-binding site and inserts into the active site
have a very good shelf-life. In addition, nanobodies regain
antigen-binding specificity after prolonged incubation at
temperatures up to 90°C, a property that has been attributed
to a reversible unfolding behaviour [89-91]. Melting points of

o f
of the enzyme [45]. Biochemical analysis further demonstrated
that the antibody fragment inhibits the enzymatic activity of
lysozyme in a competitive manner [102]. From the cloned
nanobody repertoire of a dromedary, immunised with small

o
nanobodies are in the range of 60 – 78°C [66,91], whereas anti- quantities of enzymes, several inhibitory binders were
bodies and their fragments derived from other species often retrieved against the different enzyme targets. Interestingly,

r
aggregate irreversibly following thermal denaturation [89]. the inhibitory nanobodies for each given enzyme target used
Several approaches, such as molecular evolution and site- totally different CDR sequences, indicating that the drome-

P
directed mutagenesis, have been developed to optimise the dary finds different solutions to interact with the same
thermostability of antibody fragments [92-93]. epitope [77]. The careful reader might question how the
Besides their thermal resistance, nanobodies were shown to camelid nanobodies with a long CDR3 loop avoid the nega-

r
be stable against the denaturing effect of chaotropic tive influence on the binding energy following fixation of the
agents [91], in the presence of proteases and to extremes of pH loop during antigen pairing. The crystal structures of nano-

o
(SM, unpublished results). Therefore, nanobodies are bodies reveal the solution to this apparent paradox. Appar-
expected to be able to survive harsh conditions, such as those ently, a large part of the long CDR3 loop folds over the
found in the stomach, and remain biologically active in the framework-2 region, where it shields the residual hydrophobic

to treat gastrointestinal diseases.

t h
gut, creating opportunities for the oral delivery of nanobodies patches of this area from contact with the aqueous environ-
ment [45]. In addition, the long CDR3 loops seem to contain
a cysteine that forms a disulfide bond with a cysteine in the

u
4.3 Nanobodies recognise unique conformational CDR1 (or at position 45) [53,103].
epitopes
The nanobody repertoire contains only three of the six natu- 5. Achievements and medical opportunities for

A
rally occurring CDRs from an antibody. Although surprising nanobodies
at first glance that three CDRs are sufficient to confer anti-
gen-binding specificity and high affinity, evolution has by Nanobodies are distinguished from other conventional
itself arrived at that very solution in camelids and some shark antibody formats by their unique properties of size, solubil-
species. Therefore, irrespective of the precise driving force ity, intrinsic stability, easy tailoring into pluripotent con-
behind the emergence of HCAbs in camelids, their variable structs, recognition of uncommon or hidden epitopes,
domains bind a wide range of antigens, ranging from small binding into cavities or active sites of enzyme targets, ease
molecules, such as haptens [94-95] and peptides [96], to large and speed of drug discovery, and ease of manufacture.
antigens, such as proteins [45-46] or viruses [97]. These features should lead to a number of biotechnological
The small size of the nanobody allows the epitope recogni- and medical applications in which nanobodies should excel
tion even when the target proteins are densely packed so that other antibody formats.
the epitopes become cryptic for conventional antibodies [98].
Thus, it can be argued that the nanobodies have access to a 5.1 Nanobodies as modular building blocks for
wider range of epitopes than conventional antibodies. This manifold constructs
was also concluded from a study where the immune responses Bispecific and bifunctional antibodies have many practical
of heavy-chain and conventional antibodies were compared applications in immunodiagnosis and therapy. The structural
for enzymatic interactions [77]. requirement for multispecificity is to fuse two or more bind-
Clefts and cavities play a critical role in multiple biologi- ing domains together, with sufficient flexibility to allow simul-
cal activities, as these often form the interaction site taneous binding to different target epitopes. The simplest
between two molecules, for example, enzyme–substrate and bispecific antibody is one that binds two adjacent epitopes on

6 Expert Opin. Biol. Ther. (2005) 5(1)


Revets, De Baetselier & Muyldermans

a single target antigen, thereby gaining an avidity advantage to systemic non-neuropathic amyloid disease. These results
the antigen [104]. However, most bispecific antibodies are provide compelling evidence that reducing the ability of an
designed as crosslinking reagents that bind to different target amyloidogenic protein to form partially unfolded species
antigens. This has been a major focus in cancer therapy, par- can be a highly effective method of preventing its aggrega-
ticularly for the recruitment of cytotoxic T cells to tumour tion, and suggest new and potentially general approaches to
cells [13,105]. Besides bispecific antibodies, bifunctional anti- the rational design of therapeutic agents directed against
bodies are particularly attractive, as they can deliver a ‘toxic protein deposition diseases [112].
payload’, such as radionuclides, cytotoxic drugs, peptides or In a similar application, a nanobody against an intrinsi-
enzymes, to the target antigen, including cancer cells [13,106]. cally flexible protein, the addiction antidote MazE, was
The ideal characteristics of such a multispecific molecule shown to stabilise the antigen and to keep it in a structured
are: small size, absence of linkers prone to aggregation or pro- form for crystallisation [113].
teolytic cleavage, high solubility, high stability and good Finally, a llama nanobody binding to the poly-A binding
expression yields. The construction of bispecific and bifunc- protein nuclear 1, containing an expanded alanine region
tional scFvs as well as innovative alternatives, including dia-
bodies [107], seemed a promising approach, but large-scale
clinical development can be hampered by low expression lev-
els, sensitivity to aggregation and proteolytic degradation at 5.3 Nanobodies

o f
causing a protein aggregation disorder, was retrieved from an
‘immune’ library and used for diagnostic screenings [114].

as intrabodies

o
the level of the linkers, or unproductive association between Intracellular expression of recombinant antibodies (intrabod-
the VH and VL chains in the case of diabodies (due to ies) is a powerful tool to inhibit in vivo functions of selected

r
promiscuous bindings between various VHs and VLs). molecules. The expression level, correct folding and appropri-
The strict single-domain nature of nanobodies promotes ate subcellular targeting of the antibody are important factors

P
them as ideal molecular recognition units into more complex for successful immunomodulation. Indeed, most antibodies
entities, such as bispecific, multivalent and bifunctional con- do not function intracellularly because critical disulfide bonds
structs. Moreover, nanobodies are highly soluble and remarka- cannot form in the reducing environment of the cytoplasm or

r
bly robust, allowing a high versatility as building blocks for because of difficulties in targeting to subcellular compart-
manifold constructs. Nanobodies have been tandemly cloned, ments. Moreover, few antibodies bind to the active site of

o
spaced by the natural hinge of camel HCAbs [108]. The result- enzymes and, thus, they generally do not neutralise the cata-
ing bifunctional or bivalent proteins were highly expressed in lytic activity of enzymes. Jobling and co-workers [105] demon-
bacteria in shake-flasks and retained fully the original func- strated that nanobodies can be correctly targeted to

t h
tionality. Immunotoxins and other immunoconjugates have
also been generated by fusing the gene of a nanobody to that
of a toxic protein or of an enzyme [77,98,108]. Recently, pentam-
subcellular organelles and inhibit enzyme function in plants
more efficiently than antisense approaches. Because nanobod-
ies are in general more stable than scFvs, they better preserve

u
erisation of single-domain antibodies was possible by fusion of their antigen-binding capacity even in the reducing environ-
a single-domain monomer to the B-subunit of Escherichia coli ment of the cytoplasm. For this approach, it might be prefera-
verotoxin, or shiga-like toxin, which self-assembles into a ble to use llama nanobodies, as these possess much less

A
homopentamer. A strong avidity effect was noticed for the frequent interloop disulfide bonds compared with nanobodies
pentavalent nanobodies, and from structural considerations, it from dromedaries [116]. In a recent publication, it has been
should be feasible to generate a bispecific pentavalent con- indicated that the intracellularly expressed llama nanobody
struct by cloning one nanobody at the N terminal end and against the p15 matrix protein of the porcine endogenous ret-
another at the C terminal end of the verotoxin scaffold [109]. rovirus Gag polyprotein prevents the production of porcine
The genetic fusion between the genes for the S-layer pro- retroviruses [117]. This approach may be further developed to
tein and a nanobody offers the possibility to produce fusion combat viral infections of human organs after gene therapy
proteins that crystallise spontaneously into a monomeric pro- with such intrabodies. Alternatively, intrabodies may be used
tein lattice that can be used as a nanopatterned sensing layer to block intracellular signalling pathways in cells [118]. Inter-
to detect low concentrations of antigen [110-111]. estingly, efforts are being made at present to increase the sta-
bility of scFv antibodies and to try to endow them with
5.2 Nanobodies as a format to stabilise intrinsically properties of nanobodies [119,120].
flexible or unstable proteins
Amyloid diseases are characterised by an aberrant assembly 6. Nanobodies in cancer diagnosis and therapy
of specific proteins or protein fragments into fibrils and
plaques that are deposited in various organs and tissues, Based on their unique biophysical and pharmacological prop-
often with serious pathological consequences. A nanobody erties, nanobodies should be ideally placed to become a new
with specificity for human lysozyme was shown to inhibit class of cancer therapeutics. In addition to therapy, nanobod-
dramatically the in vitro aggregation of D67H variant, a nat- ies are also expected to have a future as a tool for the diagnosis
urally occurring mutational variant that is responsible for of cancer.

Expert Opin. Biol. Ther. (2005) 5(1) 7


Nanobodies as novel agents for cancer therapy

6.1 Nanobodies as a cancer diagnostic tool 6.2 Nanobodies in cancer therapeutic tools
Fast and reliable in vivo diagnosis at an early stage of the disease, Given their high affinity and specificity, the small size of
or monitoring treatment efficiency, remains a major challenge nanobodies makes them particularly suitable for targeting
in cancer diagnosis. The ideal cancer-imaging agent should antigen in obstructed locations, such as tumours, where pene-
deliver an amount of label sufficient to detect the smallest tration into poorly vascularised tissue is crucial to the success
metastases against a low-level of nonspecific noise background. of the drug [124]. In addition, nanobodies have an extremely
In addition, it should localise promptly by penetrating the low immunogenic potential. In animal studies, repetitive
tumour, and unbound conjugate should clear rapidly from the administration of nanobodies does not yield any detectable
system to reduce target-to-background ratios. Prolonged clear- humoral or cellular immune response [124].
ance kinetics have hampered the development of intact anti- As a versatile building block for manifold constructs, nano-
bodies as imaging agents, despite their ability to effectively bodies can easily be used as a delivery vehicle for a range of
deliver radionuclides to tumour targets in vivo [121-123]. Small molecules that provide ancillary functions after target binding.
antibody fragments have potential for good tumour targeting, This ‘magic bullet’ would be delivering the toxic payload to the
as they penetrate rapidly and give high tumour-to-background
ratios at very early time points after administration [30].
The superior penetration potential of nanobodies, due to
their high-affinity target binding and fast clearance from the

o f
tumour while minimising the length of time that the toxic
compound could cause damage to healthy cells. In a separate
experiment, a bifunctional fusion was successfully constructed
between a nanobody directed against carcinoembryonic anti-
gen and a bacterial β-lactamase. Use of this conjugate in an

o
circulation of the excess of non-targeted nanobodies, repre-
sents an ideal basis for imaging purposes. A nanobody against antibody-dependent enzyme prodrug therapy setting led to a

r
lysozyme, which was previously shown to inhibit lysozyme complete cure of established LS 174T human adenocarcinoma
activity in vitro [76,108], effectively targeted bulky tumours and xenografts without any adverse effect [78].

P
metastatic lesions transgenic for hen egg white lysozyme, Today, with the discovery of bacterial strains that specifi-
whereas excess of this nanobody was rapidly cleared from the cally target tumours, and aided by genetic engineering, there
circulation. A few hours after administration, tumour-to- is a new interest in the use of bacteria as tumour vectors.

r
blood ratios of up to 10 could be demonstrated with radiola- Clostridium, Salmonella and Bifidobacterium have been shown
belled nanobodies [124]. Nanobody uptake by tumours to preferentially replicate in solid tumours, and can be manip-

o
matches well with the half-life of radionuclides, including ulated to transport and amplify genes encoding factors such as
99mTc and 123I. Recently, a stable one-step 99mTc labelling of toxins, angiogenesis inhibitors and cytokines [128]. To improve
His-tagged recombinant proteins has been developed [125]. the tropism of these bacteria for tumours, retargeting mole-

t h
Application of this technology to His-tagged nanobodies may
lead to the development of a new class of radioimmunophar-
maceuticals for scintigraphic imaging of tumours. Initial
cules, such as antibodies, may be expressed on their surface. In
this context, it has been shown that nanobody-based manifold
constructs can be efficiently translocated to the bacterial cell

u
experiments have indicated that nanobodies are stably labelled surface, whereas scFv constructs failed to do so, probably
with 99mTc using this one-step labelling method. The radiola- reflecting the high stability of nanobodies and their low
belled nanobodies are stable in vitro and in vivo, and allowed tendency to form aggregates [129].

A
clear-cut imaging of tumours at 5-h postadministration with Besides targeting to obstructed lesions, the delivery of thera-
tumour-to-background ratios > 9:1 (Cortez-Retamozo, peutics across the blood–brain barrier (BBB) remains a major
unpublished results). challenge in developing treatments for neurological diseases
Besides their potential as imaging agents, nanobodies may and cancers of the brain. Macromolecule delivery across the
offer added value to cancer diagnostic tests used at present. For BBB using antibodies takes advantage of transporters or recep-
example, early detection and staging of prostate cancer is based tors selectively expressed on brain capillary endothelium, such
on the detection of prostate-specific antigen (PSA) in the blood as transferring-receptor, that undergo receptor-mediated trans-
circulation. However, different isoforms of PSA are present in cytosis. Nanobodies have been isolated that target the brain
the blood, of which some correlate better with prostatic disorders in vivo and transmigrate across human BBB endothelium [130].
than others [126]. Recently, nanobodies have been generated that The potential of nanobodies as intrabodies provides an
can discriminate between different isoforms of PSA. Remarkably, attractive means for manipulating antitumour triggering
these nanobodies seem to sense or induce conformational agents when expressed in tumour cells. Such antibodies can
changes on different PSA isoforms, a feature that may be neutralise or modify the activity of oncogenic molecules when
exploited to discriminate different stages of prostate cancer [127]. addressed in specific subcellular compartments and/or they
A fusion of one of these PSA-specific binders with the S-layer can be used to trigger antitumour immunity when expressed
protein was used for direct detection of low PSA concentrations on the tumour cell surface.
by surface plasmon resonance [110]. Moreover, llama-derived Besides applications where rapid clearance of the therapeu-
nanobodies were shown to be excellent tools in immunocyto- tic agent is wanted, there are therapeutic situations where a
chemical, immunohistochemical and immunoblot analysis to prolonged serum half-life is desirable to increase the time that
detect oculopharyngeal muscular dystrophy [114]. the antibody can interact with its target and to reduce the fre-

8 Expert Opin. Biol. Ther. (2005) 5(1)


Revets, De Baetselier & Muyldermans

quency of dosing. Several methods have been used to increase includes protruding loops. As a consequence, the surface of the
the half-life of antibody fragments, including PEGylation paratope of nanobodies can be as large as that of conventional
[131], conjugation or fusion to albumin [132], and fusion to the antibodies. However, the footprint it leaves on the antigen is
Fc fragment of an antibody [133]. It is known that the neonatal smaller due to the large concave paratope architecture. There-
receptor FcRn regulates the serum half-lifes of both IgG [134] fore, hidden or convex epitopes can still be targeted by nano-
and albumin [135]. Hence, fusion of antibody fragments to bodies. As such, nanobodies can interact with receptor clefts
albumin or Fc probably confers a prolonged serum half-life and the active site of enzymes where they actually act as inhibi-
and also reduces kidney clearance (due to an increase in tors. Hence, nanobodies may represent an interesting lead to
molecular size). Alternatively, peptides with high affinity for design small enzyme inhibitors from which peptide mimetic
serum albumin can be fused to an antibody fragment and drugs can eventually be derived.
extend the half-life [136]. All of these strategies can also be eas- The immunogenicity of nanobodies is a concern that
ily applied to increase the serum half-life of nanobodies. Con- should be thoroughly investigated. However, the large degree
necting the antigen-specific nanobody with a serum albumin of sequence identity between a nanobody and human VH of
binding entity or with the Fc of human IgG increases the
serum half-life of a nanobody in a mouse from 45 min to 2
weeks. This versatility will obviously increase the range of
therapeutic options available for nanobodies.

o f
family III is a good sign. The lack of any B or T cell response
against nanobodies in mice is another. Furthermore, the
retrieval of nanobodies with the framework-2 imprint of a
human VH [137] might be the ultimate solution to obtain

o
non-immunogenic nanobodies.
7. Expert opinion and conclusions The nanobody technology platform that has been devel-

r
oped can quickly deliver therapeutic leads for a wide range of
Since their discovery a decade ago, nanobodies have progressed targets within a few weeks. Based on their heavy-chain-only

P
from a ‘Darwinian evolution curiosity’ to a highly versatile origin, nanobodies obviate the need for construction and
molecule with a wide variety of potential biotechnological and screening of large libraries and for subsequent affinity matura-
medical applications. Nanobodies are the smallest functional tion techniques. Moreover, nanobodies are endowed with

r
antigen-binding fragments derived from in vivo matured camel unique properties such as a high stability against extremes of
or llama HCAbs. These entities are extremely stable and bind temperature and pH, and the possibility to produce large

o
antigen with nanomolar affinity. The subtle substitutions and quantities at low cost. Because of these unique properties and
structural rearrangements that occur between the VL-interact- the ease of generating manifold constructs, it is clear that
ing surface of a VH domain and the corresponding site of a nanobodies possess advantageous characteristics for therapy

t h
nanobody provide an example of natural protein engineering
converting a heterodimeric molecule into a monomeric mole-
cule. The absence of a VL domain is compensated in nanobod-
and diagnostics. The possibility of engineering multiple and
robust antibody-based formats positions them far ahead of
conventional antibodies. Future studies in humans will be

u
ies by an extended CDR1 and a more exposed CDR3 that required to confirm their performance, benefits and their effi-
often folds over the framework-2 region. The structural reper- cacy in cancer immunodiagnosis and/or therapy. Bringing to
toire of nanobodies is thus sufficiently diverse. Therefore, the the field nanobodies with specificities to other targets of bio-

A
CDR grafting from an antigen-binding human or mouse logical relevance will be another future challenge, not only to
VH domain to a nanobody scaffold has little chance for address fundamental understandings about their molecular
success. Besides the standard architecture of planar surfaces characteristics and structural stabilities, but also to position
and cavities, the antigen-binding site of nanobodies also them as clinically useful pharmaceuticals.

Bibliography Expert Opin. Pharmacother. (2002) 7. PANACCIONE R, FEDORAK RN,


Papers of special note have been highlighted as 3:1657-1663. AUMAIS G et al.: Canadian Association of
either of interest (•) or of considerable interest 4. PANKHURST T, ADU D. Antibodies in Gastroenterology Clinical Practice
(••) to readers. the prevention of renal allograft rejection. Guidelines: the use of infliximab in Crohn’s
Expert Opin. Biol. Ther. (2004) 4:243-252. disease. Can. J. Gastroenterol. (2004)
1. KÖHLER G, MILSTEIN C: Continuous
18:503-508.
cultures of fused cells secreting antibody of 5. SANDBOM WJ, HANAUER SB:
predefined specificity. Nature (1975) Antitumor necrosis factor therapy for 8. SAEZ-LLORENS X, CASTANO E,
256:495-497. inflammatory bowel disease: a review of NULL D: Safety and pharmacokinetics of
agents, pharmacology, clinical results and an intramuscular humanised antibody
2. BERARD JL: A review of interleukin-2
safety. Inflamm. Bowel Dis. (1999) monoclonal antibody to respiratory syncitial
receptor antagonists in solid organ
5:119-133. virus in premature infants with
transplantations. Pharmacotherapy (1999)
bronchopulmonary dysplasia. Pediatr. Infect.
19:1127-1137. 6. O’DELL JR: Therapeutic strategies for
Dis. (1998) 17:787-791.
3. HENRY ML, RAJAB A: The use of rheumatoid arthritis. N. Engl. J. Med.
basiliximab in solid organ transplantation. (2004) 350:2591-2602.

Expert Opin. Biol. Ther. (2005) 5(1) 9


Nanobodies as novel agents for cancer therapy

9. GROOTHUIS JR, NISHIDA H: 21. PRESTA LG: Engineering antibodies for 33. HUDSON PJ, SOURIAU C: Recombinant
Prevention of respiratory syncitial virus therapy. Curr. Pharm. Biotechnol. (2002) antibodies for cancer diagnosis and therapy.
infections in high-risk infants by 3:237-256. Expert Opin. Biol. Ther. (2001) 1:845-855.
monoclonal antibody (palivizumab). 22. WEIR ANC, NESBITT A, 34. TODOROVSKA A, ROOVERS RC,
Pediatr. Int. (2002) 44:235-241. CHAPMAN AP: Formatting antibody DOLEZAL O: Design and application of
10. BURTON DR: Antibodies, viruses and fragments to mediate specific therapeutic diabodies, triabodies and tetrabodies for
vaccines. Nat. Rev. Immunol. (2002) functions. Biochem. Soc. Trans. (2002) cancer targeting. J. Immunol. Methods
2:706-713. 30:512-516. (2001) 248:47-66.
11. HULS GA, HEIJNEN IA, CUOMO ME 23. CHOY EH, HAZLEMAN B, SMITH M: 35. WILLUDA J, HONEGGER A,
et al.: A recombinant, fully human Efficacy of a novel PEGylated humanized WAIBEL R et al.: High thermal stability is
monoclonal antibody with antitumor anti-TNF fragment (CDP870) in patients essential for tumor targeting of antibody
activity constructed from phage-displayed with rheumatoid arthritis: a Phase II fragments: engineering of a humanized anti-
antibody fragments. Nat. Biotechnol. (1999) double-blinded, randomized, dose- epithelial glycoprotein-2 (epithelial cell

f
17:276-280. escalating trial. Rheumatology (2002) adhesion molecule) single-chain Fv
12. FARAH RA, CLINCHY B, HERRERA L, 41:1133-1137. fragment. Cancer Res. (1999)
59:5758-5767.

o
VITETTA ES: The development of 24. WINTER G, MILSTEIN C: Man-made
monoclonal antibodies for the therapy of antibodies. Nature (1991) 349:293-299. 36. WHITLOW M, BELL BA, FENG SL
cancer. Crit. Rev. Eukaryot. Gene Expr. 25. WINTER G, GRIFFITHS AD, et al.: An improved linker for single-chain

o
(1998) 8:321-356. HAWKINS RE, HOOGENBOOM HR: Fv with reduced aggregation and enhanced
proteolytic stability. Protein Eng. (1993)

r
13. CARTER P: Improving the efficacy of Making antibodies by phage display
antibody-based cancer therapies. Nat. Rev. technology. Annu. Rev. Immunol. (1994) 6:989-995.
Cancer (2001) 1:118-129. 12:433-455. 37. WARD ES, GÜSSOW DH,

P
14. GURA T: Therapeutic antibodies: magic 26. BETTER M, CHANG CP, GRIFFITHS AD, JONES PT, WINTER
bullets hit the target. Nature (2002) ROBINSON RR, HORWITZ AH: G: Binding activities of a single
417:584-586. Escherichia coli secretion of an active immunoglobulin variable domain secreted

r
chimeric antibody. Science (1988) from E. coli. Nature (1989) 341:544-546.
15. KHAZAELI MB, CONRY RM,
240:1041-1043. •• First report on the possibility of obtaining
LOBUGLIO AF: Human response to
single domain antibody fragments from

o
monoclonal antibodies. J. Immunother. 27. FRENKEN L, HESSING JG,
immunised mice.
(1994) 15:42-52. VAN DEN HONDEL CA, VERRIPS CT:
Recent advances in the large-scale 38. BORREBAECK CAK, MALMBORG AC,

h
16. PENDLEY G, SCHANTZ A,
production of antibody fragments using FUREBRING C et al.: Kinetic analysis of
WAGNER C: Immunogenicity of

t
lower eukaryotic microorganisms. recombinant antibody-antigen interactions:
therapeutic monoclonal antibodies.
Res. Immunol. (1998) 149:589-598. relation between structural domains and
Curr. Opin. Mol. Ther. (2003) 5:172-179.
antigen binding. Biotechnology (1992)

u
17. MORRISON SL, JOHNSON MJ, 28. ADAMS GP, SCHIER R, MCCALL AM:
10:697-698.
HERZENBERG LA, OI VT: Chimeric High affinity restricts the localisation and
tumor penetration of single-chain Fv 39. JESPERS L, SCHON O, FAMM K,
human antibody molecules: mouse antigen-
WINTER G: Aggregation-resistant domain

A
binding domains with human constant antibody molecules. Cancer Res. (2001)
61:4750-4755. antibodies selected on phage by heat
region domains. Proc. Natl. Acad. Sci. USA
denaturation. Nat. Biotechnol. (2004)
(1984) 21:6851-6855. 29. WU AM, YAZAKI PJ: Designer genes:
22:1161-1165.
18. RIECHMANN L, CLARK M, recombinant antibody fragments for
biological imaging. Q. J. Nucl. Med. (2000) 40. HAMERS-CASTERMAN C,
WALDMANN H et al.: Reshaping human
44:268-283. ATARHOUCH T, MUYLDERMANS S
antibodies for therapy. Nature (1988)
et al.: Naturally occurring antibodies devoid
332:323-327. 30. YOKOTA T, MILENIC DE,
of light chains. Nature (1993) 363:446-448.
19. GREEN LL: Antibody engineering via WHITLOW M, SCHLOM J: Rapid tumor
•• First report on the occurrence of HCAbs
genetic engineering of the mouse: penetration of a single-chain Fv and
in camelids.
XenoMouse strains are a vehicle for the comparison with other immunoglobulin
forms. Cancer Res. (1990) 52:3402-3408. 41. ORLANDI R, GÜSSOW DH, JONES PT,
facile generation of therapeutic human
WINTER G: Cloning immunoglobulin
monoclonal antibodies. J. Immunol. 31. HUDSON PJ, SOURIAU C: Engineered
variable domains for expression by the
Methods (1999) 231:11-23. antibodies. Nat. Med. (2003) 9:139-134.
polymerase chain reaction. Proc. Natl. Acad.
20. HULS G, HEIJNEN IA, CUOMO E: 32. NIELSEN UB, ADAMS GP, Sci. USA (1989) 86:3833-3837.
Antitumor immune effector mechanisms WEINER LM, MARKS JD: Targeting of
42. NGUYEN VK, HAMERS R, WYNS L,
recruited by phage display-derived fully bivalent anti-ErbB2 diabody antibody
MUYLDERMANS S: Loss of splice
human IgG1 and IgA1 monoclonal fragments to tumor cells is independent of
consensus signal is responsible for the
antibodies. Cancer Res. (1999) the intrinsic antibody affinity. Cancer Res.
removal of the entire CH1 domain of the
59:5778-5784. (2000) 60:6434-6440.
functional camel IgG2a heavy-chain
antibodies. Mol. Immunol. (1999)
36:515-524.

10 Expert Opin. Biol. Ther. (2005) 5(1)


Revets, De Baetselier & Muyldermans

43. WOOLVEN BP, FRENKEN LG, 53. MUYLDERMANS S, CAMBILLAU C, suggestion for converting a murine VH
VAN DER LOGT P, NICHOLLS PJ: The WYNS L: Recognition of antigens by domain into a more tractable species. J. Mol.
structure of the llama heavy-chain constant single-domain antibody fragments: the Biol. (2000) 300:963-973.
genes reveals a mechanism for heavy-chain superfluous luxury of paired domains. 64. RIECHMANN L: Rearrangement of the
antibody formation. Immunogenetics (1999) Trends Biochem. Sci. (2001) 26:230-235. former VL interface in the solution
50:98-101. 54. WU TT, JOHNSON G, KABAT EA: structure of a camelised, single antibody
44. PADLAN EA: Anatomy of the antibody Length distribution of CDRH3 in VH domain. J. Mol. Biol. (1996)
molecule. Mol. Immunol. (1994) antibodies. Prot. Struct. Funct. Genet. (1993) 259:957-969.
31:169-217. 16:1-7. 65. TANHA J, XU P, CHEN Z et al.: Optimal
45. DESMYTER A, TRANSUE TR, 55. HARMSEN MM, RUULS RC, design features of camelised human single-
ARBABI-GHAHROUDI M et al.: NIJMAN IJ, NIEWOLD TA, domain antibody libraries. J. Biol. Chem.
Crystal structure of a camel single-domain FRENKEN LG, DE GEUS B: Llama (2001) 276:24774-24780.
VH antibody fragment in complex with heavy-chain V regions consist of at least four 66. EWERT S, CAMBILLAU C,

f
lysozyme. Nat. Struct. Biol. (1996) distinct subfamilies revealing novel sequence CONRATH K, PLUCKTHUN A:
3:803-811. features. Mol. Immunol. (2000) 37:579-590. Biophysical properties of camelid VHH
• First description of the structure of a

o
56. MUYLDERMANS S, ATARHOUCH T, domains compared to those of human VH3
nanobody in complex with its target. SALDHANA J, BARBOSA JARG, domains. Biochemistry (2002)
46. SPINELLI S, FRENKEN L, HAMERS R: Sequence and structure of 41:3628-3636.

o
BOURGEOIS D et al.: The crystal structure VH domain from naturally occurring camel 67. EWERT S, HUBER T, HONEGGER A,
of the llama heavy chain variable domain. heavy chain immunoglobulins lacking light

r
PLUCKTHUN A: Biophysical properties of
Nat. Struct. Biol. (1996) 3:752-757. chains. Protein Eng. (1994) 7:1129-1135. human antibody variable domains. J. Mol.
47. DECANNIERE K, DESMYTER A, 57. KABAT E, WU TT, PERRY HM, Biol. (2003) 325:531-553.

P
LAUWEREYS M, GHAROUDI MA, GOTTESMAN KS, FOELLER C: 68. AIRES DA SILVA F, SANTA-MARTA M,
MUYLDERMANS S, WYNS L: A single- Sequence of Proteins of Immunological FREITAS-VIEIRA A et al.: Camelized
domain antibody fragment in complex with Interest. (1991). US Public Health Services, rabbit-derived VH single-domain

r
RNase A: non-canonical loop structures and NIH Bethesda, MD, Publication No. 91. intrabodies against Vif strongly neutralize
nanomolar affinity using two CDR loops. 58. RIECHMANN L, MUYLDERMANS S: HIV-1 infectivity. J. Mol. Biol. (2004)
Struct. Fold. Des. (1999) 7:361-370.

o
Single-domain antibodies: comparison of 340:525-542.
48. RENISIO JG, ROMI-LEBRUN R, camel VH and camelised human VH 69. GREENBERG AS, AVILA D,
BLANC E: Solution structure and backbone domains. J. Immunol. Methods (1999) HUGHES M, HUGHES A,

h
dynamics of an antigen-free heavy chain 231:25-38. MCKINNEY EC, FLAJNIK MF: A new

t
variable domain (VHH) from llama. 59. DAVIES J, RIECHMANN L: ‘Camelising’ antigen receptor gene family that undergoes
Prot. Struct. Funct. Genet. (2002) human antibody fragments: NMR studies rearrangement and extensive somatic
47:546-555. on VH domains. FEBS Lett. (1994) diversification in sharks. Nature (1995)
49.

50.
VRANKEN W, TOLKATCHEV D, XU P:

u
Solution structure of a llama single-domain
antibody with hydrophobic residues typical

A
of the VH/VL interface. Biochemistry
(2002) 41:8570-8579.
DECANNIERE K, MUYLDERMANS S,
WYNS L: Canonical antigen-binding loop

60.
339:285-290.
First description of ‘camelisation’ of
human antibody fragments.
KORTT AA, GUTHRIE RE,
HINDS MG: Solution properties of
Escherichia coli-expressed VH domain of
anti-neuraminidase antibody NC41.
70.
374:168-173.
STANFIELD RL, DOOLEY H,
FLAJNIK MF, WILSON IA: Crystal
structure of a shark single-domain antibody
V region in complex with lysozyme. Science
(2004) 305:1770-1773.
71. STRELTSOV VA, VARGHESE JN,
structures in immunoglobulins: more J. Protein Chem. (1995) 14:167-178. CARMICHAEL JA, IRVING RA,
structures, more canonical classes? J. Mol. 61. DAVIES J, RIECHMANN L: Single HUDSON PJ, NUTTALL SD: Structural
Biol. (2000) 300:83-91. antibody domains as small recognition evidence for evolution of shark Ig new
51. VU KB, GHARHOUDI MA, WYNS L, units: design and in vitro antigen selection antigen receptor variable domain antibodies
MUYLDERMANS S: Comparison of llama of camelized, human VH domains with from a cell-surface receptor. Proc. Natl.
VH sequences from conventional and heavy improved protein stability. Protein Eng. Acad. Sci. USA (2004) 101:12444-12449.
chain antibodies. Mol. Immunol. (1997) (1996) 9:531-537. 72. HUFTON SE, VAN NEER N, VAN DEN
34:1121-1131. 62. MARTIN F, VOLPARI C, BEUKEN T, DESMET J, SABLON E,
52. NGUYEN VK, HAMERS R, WYNS L, STEIKÜHLER C et al.: Affinity selection HOOGENBOOM HR: Development and
MUYLDERMANS S: Camel heavy-chain of a camelised VH domain antibody application of cytotoxic T lymphocyte-
antibodies: diverse germline VHH and inhibitor of hepatitis C virus NS3 protease. associated antigen 4 as a protein scaffold for
specific mechanisms enlarge the antigen- Protein Eng. (1997) 10:607-614. the generation of novel binding ligands.
binding repertoire. EMBO J. (2000) 63. VOORDIJK S, HANSSON T, FEBS Lett. (2000) 475:225-231.
19:921-931. HILVERT D, VAN GUNSTEREN WF: 73. KARATAN E, MERGUERIAN M,
• Description of the mechanisms involved Molecular dynamics simulations highlight HAN Z, SCHOLLE MD, KOIDE S,
in enlarging the antigen-binding repertoire mobile regions in proteins: a novel KAY BK: Molecular recognition properties
of nanobodies.

Expert Opin. Biol. Ther. (2005) 5(1) 11


Nanobodies as novel agents for cancer therapy

of FN3 monobodies that bind the Src SH3 human antibodies by chain shuffling. combining sites to haptens. Biochemistry
domain. Chem. Biol. (2004) 11:835-844. Biotechnology (1992) 10:779-783. (2000) 39:1217-1222.
74. KOIDE A, BAILEY CW, HUANG X, 84. LOW NM, HOLLIGER PH, WINTER G: • First crystal structure of a nanobody in
KOIDE S: The fibronectin type III domain Mimicking somatic hypermutation: affinity complex with its cognate hapten.
as a scaffold for novel binding proteins. maturation of antibodies displayed on 95. SPINELLI S, TEGONI M, FRENKEN L,
J. Mol. Biol. (1998) 284:1141-1151. bacteriophage using a bacterial mutator VAN VLIET C, CAMBILLAU C: Lateral
75. HOOGENBOOM HR, DE BRUINE A, strain. J. Mol. Biol. (1996) 260:359-368. recognition of a dye hapten by a llama
HUFTON SE, HOET RM, ARENDS JW, 85. JOOSTEN V, LOKMAN C, VHH domain. J. Mol. Biol. (2001)
ROOVERS RC: Antibody phage display VAN DEN HONDEL CA, PUNT PJ: The 311:123-129.
technology and its applications. production of antibody fragments and 96. RAHBARIZADEH F, RASAEE MJ,
Immunotechnology (1998) 4:309-318. antibody fusion proteins by yeasts and MOGHADAM FM, ALLAMEH AA,
76. GHARHOUDI MA, DESMYTER A, filamentous fungi. Microb. Cell Fact. SADRODDINY E: Production of novel
WYNS L, HAMERS R, (2003) 2:1. recombinant single-domain antibodies

f
MUYLDERMANS S: Selection and 86. FRENKEN L, VAN DER LINDEN RH, against tandem repeat region of MUC1
identification of single-domain antibody HERMANS PW et al.: Isolation of antigen- mucin. Hybrid. Hybridomics (2004)
23:151-159.

o
fragments from camel heavy-chain specific llama VHH antibody fragments and
antibodies. FEBS Lett. (1997) 414:521-526. their high level secretion by Saccharomyces 97. LEDEBOER AM, BEZEMER S,
77. LAUWEREYS M, GHARHOUDI MA, cerevisiae. J. Biotechnol. (2000) 78:11-21. DE HAARD JJ et al.: Preventing phage lysis

o
DESMYTER A et al.: Potent enzyme 87. VAN DER VAART JM: Expression of of Lactococcus lactis in cheese production
using a neutralizing heavy-chain antibody

r
inhibitors derived from dromedary heavy- VHH antibody fragments in Saccharomyces
chain antibodies. EMBO J. (1998) cerevisiae. Methods Mol. Biol. (2002) fragment from llama. J. Dairy Sci. (2002)
17:3512-3520. 178:359-366. 85:1376-1382.

P
•• First report of nanobodies as 88. VAN DER LINDEN RH, DE GEUS B, 98. STIJLEMANS B, CONRATH K,
enzyme inhibitors. FRENKEN GJ, PETERS H, CORTEZ-RETAMOZO V et al.: Efficient
78. CORTEZ-RETAMOZO V, VERRIPS CT: Improved production and targeting of conserved cryptic epitopes of

r
BACKMANN N, SENTER P et al.: function of llama heavy chain antibody infectious agents by single-domain
Efficient cancer therapy with a nanobody- fragments by molecular evolution. antibodies. African trypanosomes as a
paradigm. J. Biol. Chem. (2004)

o
based conjugate. Cancer Res. (2004) J. Biotechnol. (2000) 80:261-270.
64:2853-2857. 279:1256-1261.
89. VAN DER LINDEN RH, FRENKEN LG,
• First report describing the efficiency of • First report on nanobodies recognising
DE GEUS B et al.: Comparison of physical

h
nanobodies as cancer therapeutics. cryptic epitopes.
chemical properties of llama VHH antibody

t
79. FRENKEN L, HESSING JG, fragments and mouse monoclonal 99. LASKOWSKI RA, LUSCOMBE NM,
VAN DEN HONDEL CA, VERRIPS CT: antibodies. Biochem. Biophys. Acta (1999) SWINDELLS MB, THORNTON JM:
Recent advances in the large-scale 1431:37-46. Protein clefts in molecular recognition and

u
production of antibody fragments using function. Protein Sci. (1996) 5:2438-2452.
90. PEREZ JM, ENISIO JG, PROMPERS JJ
lower eukaryotic microorganisms. Res. et al.: Thermal unfolding of a llama 100. WEBSTER DM, HENRY AH, REES AR:
Immunol. (1998) 149:589-598. Antibody-antigen interactions. Curr. Opin.

A
antibody fragment: a two-state reversible
80. REITER Y, SCHUCK P, BOYD LF, process. Biochemistry (2001) 40:74-83. Struct. Biol. (1994) 4:123-129.
PLASKIN D: An antibody single-domain 91. DUMOULIN M, CONRATH K, VAN 101. PADLAN EA: X-ray crystallography of
phage display library of a native heavy chain MEIRHAEGHE A et al.: Single-domain antibodies. Adv. Protein Chem. (1996)
variable region: isolation of functional antibody fragments with high 49:57-133.
single-domain VH molecules with a unique conformational stability. Protein Sci. (2002) 102. TRANSUE TR, DE GENST E,
interface. J. Mol. Biol. (1999) 290:685-698. 11:500-515. GHARHOUDI MA, WYNS L,
81. YAU KY, GROVES MA, LI S et al.: 92. WORN A, PLÜCKTHUN A: Stability MUYLDERMANS S: Camel single-
Selection of hapten-specific single domain engineering of antibody single-chain Fv domain antibody inhibits enzyme by
antibodies from a non-immunized llama fragments. J. Mol. Biol. (2001) mimicking carbohydrate substrate.
ribosome display library. J. Immunol. 305:989-1010. Prot. Struct. Funct. Genet. (1998)
Methods (2003) 281:161-175. 32:515-522.
93. EWERT S, HONEGGER A,
82. HAWKINS RE, RUSSELL SJ, PLUCKTHUN A: Structure-based 103. DESMYTER A, SPINELLI S, PAYAN F,
WINTER G: Selection of phage antibodies improvement of the biophysical properties LAUWEREYS M, WYNS L,
by binding affinity. Mimicking affinity of immunoglobulin VH domains with a MUYLDERMANS S: Three camelid VHH
maturation. J. Mol. Biol. (1992) generalisable approach. Biochemistry (2003) domains in complex with porcine pancreatic
226:889-896. 42:1517-1528. alpha amylase. Inhibition and versatility of
83. MARKS JD, GRIFFITHS AD, binding topology. J. Biol. Chem. (2002)
94. SPINELLI S, FRENKEN LG,
MALMQVIST M, CLACKSON TP, 277:23645-23650.
HERMANS P et al.: Camelid heavy-chain
BYE JM, WINTER G: By-passing variable domains provide efficient 104. ROBERT B, DORVILLIUS M,
immunisation: building high affinity BUCHEGGER F: Tumor targeting with

12 Expert Opin. Biol. Ther. (2005) 5(1)


Revets, De Baetselier & Muyldermans

newly designed biparatopic antibodies 115. JOBLING AS, JARMAN C, TEH MM, with a novel Tc(I)-carbonyl complex.
directed against two different epitopes of HOLMBERG N, BLAKE C, Nat. Biotechnol. (1999) 17:897-901.
the carcinoembryonic antigen (CEA). Int. J. VERHOEYEN ME: Immunomodulation 126. NIKOLAJCZYK SD, CATALONA WJ,
Cancer (1999) 81:285-291. of enzyme function in plants by single- EVANS CL et al.: Proenzyme forms of
105. SEGAL DM, WEINER GJ, WEINER LM: domain antibodies. Nat. Biotechnol. (2003) prostate-specific antigen in serum improve
Bispecific antibodies in cancer therapy. 21:77-80. the detection of prostate cancer. Clin. Chem.
Curr. Opin. Immunol. (1999) 11:558-562. • First description of a nanobody as (2004) 50:1017-1025.
an intrabody.
106. HUDSON P, SOURIAU C: Engineered 127. SAERENS D, KINNE J, BOSMANS E,
antibodies. Nat. Med. (2003) 1:129-134. 116. CONRATH KE, WERNERY U, WERNERY U, MUYLDERMANS S,
MUYLDERMANS S, NGUYEN VK: CONRATH K: Single-domain antibodies
107. KIPRIYANOV SM: Generation of
Emergence and evolution of functional derived from dromedary lymph node and
bispecific and tandem diabodies. Methods
heavy-chain antibodies in Camelidae. Dev. peripheral blood lymphocytes sensing
Mol. Biol. (2002) 178:317-331.
Comp. Immunol. (2003) 27:87-103. conformational variants of prostate-specific
108. CONRATH K, LAUWEREYS M,

f
117. DEKKER S, TOUSSAINT W, antigen. J. Biol. Chem. (2004) (In Press).
WYNS L, MUYLDERMANS S: Camel
PANAYOTOU G et al.: Intracellularly 128. PAWELEK JM, BROOKS KL,
single-domain antibodies as modular
expressed single domain antibody against

o
building units in bispecific and bivalent BERMUDES D: Bacteria as tumor-
p15 matrix protein prevents the production targeting vectors. Lancet Oncol. (2003)
antibody constructs. J. Biol. Chem. (2001)
of porcine retroviruses. J. Virol. (2003) 4:548-556.
276:7346-7350.

o
77:12132-12139.
109. ZHANG J, TANHA J, HIRAMA T et al.: 129. VEIGA E, DE LORENZO V,
118. KONTERMAN RE: Intrabodies as

r
Pentamerization of single-domain FERNANDEZ LA: Structural tolerance of
therapeutic agents. Methods (2004) bacterial autotransporters for folded
antibodies from phage libraries: a novel
34:163-170. passenger protein domains. Mol. Microbiol.
strategy for the rapid generation of high-

P
avidity antibody reagents. J. Mol. Biol. 119. AUF DER MAUR A, ESCHER D, (2004) 52:1069-1080.
(2004) 335:49-56. BARBERIS A: Antigen-independent 130. MURUGANANDAM A, TANHA J,
selection of stable intracellular single-chain NARANG S, STANIMIROVIC D:

r
110. PLESCHBERGER M, SAERENS D,
antibodies. FEBS Lett. (2001) 508:407-412. Selection of phage-displayed llama single-
WEIGERT S, SLEYTR UB,
MUYLDERMANS S, SARA M: An S-layer 120. VISINTIN M, QUONDAM M, chain domain antibodies that transmigrate

o
heavy chain camel antibody fusion protein CATTANEO A: The intracellular antibody across human blood-brain barrier
for generating of a nanopatterned sensing capture technology (IACT): towards a endothelium. FASEB J. (2001)
layer to detect the prostate specific antigen consensus sequence for intracellular 16(2):240-242.

h
by surface plasmon resonance technique. antibodies. J. Mol. Biol. (2002) 317:73-83. • First description of a nanobody crossing

t
Bioconjug. Chem. (2004) 15:664-671. 121. ROSEBROUGH SF: Two-step the BBB.

111. PLESCHBERGER M, NEUBAUER A, immunological approaches for imaging and 131. CHAPMAN AP: PEGylated antibodies and
therapy. Q. J. Nucl. Med. (1996) antibody fragments for improved therapy: a

u
EGELSEER EM et al.: Generation of a
functional mononuclear protein lattice 40:234-251. review. Adv. Drug Deliv. Rev. (2002)
consisting of an S-layer fusion protein 122. ZUCKIER LS, DENARDO GL: Trials and 54:531-545.

A
comprising the variable domain of a camel tribulations: oncological antibody imaging 132. BENDER E, WOOF JM, ATKIN JD,
heavy-chain antibody. Bioconjug. Chem. comes to the fore. Semin. Nucl. Med. (1997) BARKER MD, BEBBINGTON CR,
(2003) 14:440-448. 27:10-29. BURTON DR: Recombinant human
112. DUMOULIN M, LAST AM, 123. SUNDARESAN G, YAZAKI PJ, antibodies: linkage of a Fab fragment from a
DESMYTER A et al.: A camelid antibody SHIVELY JE et al.: 124I-labeled engineered combinatorial library to an Fc fragment for
fragment inhibits the formation of amyloid anti-CEA minibodies and diabodies allow expression in mammalian cell culture.
fibrils by human lysozyme. Nature (2003) high contrast, antigen-specific small-animal Hum. Antibodies Hybridomas (1993)
424:783-788. PET imaging of xenografts in mice. J. Nucl. 4:74-79.
• First report of nanobodies interfering with Med. (2003) 44:1962-1969. 133. SMITH BJ, POPPLEWELL A,
fibril formation. 124. CORTEZ-RETAMOZO V, ATHWAL D et al.: Prolonged in vivo
113. LORIS R, MARIANOVSKY I, LAH J LAUWEREYS M, HASSANZADEH GH residence times of antibody fragments
et al.: Crystal structure of the intrinsically et al.: Efficient tumor targeting by single- associated with albumin. Bioconjug. Chem.
flexible addiction antidote MazE. J. Biol. domain antibody fragments of camels. (2001) 12:750-756.
Chem. (2003) 278:28252-28257. Int. J. Cancer (2002) 98:456-462. 134. JUNGHANS RP: Finally! The Brambell

114. VAN KONINGSBRUGGEN S, •• First report describing the receptor (FcRB): mediator of transmission
DE HAARD H, DE KIEVIT P et al.: tumour-targeting properties of immunity and protection from
Llama derived phage display antibodies in of nanobodies. catabolism for IgG. Immunol. Res. (1997)
the dissection of the human disease 125. WAIBEL R, ALBERTO R, WILLUDA J 16:29-57.
oculapharyngeal muscular dystrophy. et al.: Stable one-step technetium-99m 135. CHAUDHURY C, MEHNAZ S,
J. Immunol. Methods (2003) 279:149-161. labeling of His-tagged recombinant proteins ROBINSON JM et al.: The major
histocompatibility complex-related Fc

Expert Opin. Biol. Ther. (2005) 5(1) 13


Nanobodies as novel agents for cancer therapy

receptor for IgG (FcRn) binds albumin and Affiliation


prolongs its lifespan. J. Exp. Med. (2003) Hilde Revets†, Patrick De Baetselier &
197:315-322. Serge Muyldermans
†Author for correspondence
136. DENNIS MS, ZHANG M, MENG YG
et al.: Albumin binding as a general strategy Vrije Universiteit Brussel, Department of
for improving the pharmacokinetics of Molecular and Cellular Interactions, Laboratory
proteins. J. Biol. Chem. (2002) of Cellular and Molecular Immunology, Vlaams
277:35035-35043. Interuniversitair Instituut voor Biotechnologie,
Pleinlaan 2, Building E8, B-1050 Brussels,
137. TANHA J, DUBUC G, HIRAMA T,
Belgium
NARANG SA, MACKENZIE CR:
Tel: +32 2 629 19 76; Fax: +32 2 629 19 81;
Selection by phage display of llama
E-mail: harevets@vub.ac.be
conventional V(H) fragments with heavy
chain antibody V(H)H properties.

f
J. Immunol. Methods (2002) 263:97-109.

o o
P r
o r
t h
A u

14 Expert Opin. Biol. Ther. (2005) 5(1)

View publication stats

Potrebbero piacerti anche