Introduction
The 1st French-Israeli International Conference on B Cells and Therapeutic Antibodies,
organized by Reuven Laskov (Israel), Jean-Luc Teillaud (France) and Claude-Agnès Reynaud
(France) under the auspices of The Hebrew University of Jerusalem and the University
Paris Descartes, was held October 23–25, 2011 in Jerusalem, Israel. The conference
featured 36 presentations and a dozen posters. It included two special guest lectures
given by Michael Neuberger (Laboratory of Molecular Biology, Medical Research Council,
Cambridge, United Kingdom) and by Matthew Scharff (Department of Cell Biology, Albert
Einstein College of Medicine, New York, USA). The total number of delegates exceeded
130.
The goal of the meeting was to allow foreign academic scientists, clinicians and scientists
from biotechnology companies working on B cells and antibodies to meet and exchange
ideas with their Israeli colleagues, and to trigger fruitful collaborations and scientific
exchanges. It was organized to ensure complementarity between the academic and industry
conferences. The meeting also enabled young Israeli scientists and students to freely
attend world-class a conference. It included Israeli, European, and Canadian speakers,
either academic or from companies. A poster session was organized to allow young European
and Israeli scientists to present their work.
One part of the meeting was devoted to fundamental immunology focusing on B cells
and antibodies, including sessions on somatic hypermutation (SHM), gene rearrangement,
FcγR/IgG interactions, B cell subsets, and memory B cells. Another part of the meeting
focused on the engineering of therapeutic antibodies and on their clinical use. It
included sessions on affinity maturation techniques, optimization of antibody effector
functions, and updates on antibodies in the clinic.
Brief summaries of the presentations are included here to give a flavor of the meeting,
which was held in a friendly atmosphere, with the helpful support of many of our Israeli
colleagues from the Hadassah Medical School of the Hebrew University of Jerusalem.
Delegates could enjoy a visit to the Old City of Jerusalem. The success of this first
meeting paves the way for planning the 2nd French-Israeli International Conference
on B Cells and Therapeutic Antibodies in late 2013 or early 2014 in Israel.
October 23, 2011: Day 1
Sessions 1 and 2
Claude-Agnès Reynaud and Sandrine Moutel
“So all that we seem to have acquired is the potential ability to select from an animal
any of the antibodies of his repertoire. It is somewhat like selecting individual
dishes out of a very elaborate menu: antibodies “à la carte.” But surely our “immunological
gourmandizing” cannot be satisfied by the menu that the animals are offering to us,
astonishingly comprehensive and varied as it may be. (...) I am sure that our next
step will be to move away from the dining table, where we order and consume our antibodies
“à la carte,” to the kitchen, where we shall attempt to mess them up!”
(C. Milstein, Proc. R. Soc. Lond. B 1981 211, 393–412, cited by M. Neuberger)
In the first special guest lecture, Michael Neuberger (MRC Laboratory of Molecular
Biology) provided an overview of what makes antibodies so successful from both fundamental
and applied points of view. A specific emphasis was put on the main actor of affinity
maturation, i.e., the activation-induced (cytidine) deaminase (AID) mutator. AID is
a portable mutator, the activity of which can be engineered to generate high-affinity
mutants, and is also tightly controlled in vivo, its actual catalytic activity representing
a compromise between efficient mutagenesis and minimal off-target damage.
1
Examples of recently identified interactors controlling AID stability or subcellular
localization were reported, notably REGgamma, mediating a non-ubiquitin proteasomal
degradation pathway, and eEF1A, a factor delivering tRNA to the ribosome and sequestering
AID in the cytoplasm. Multiple approaches toward high affinity human antibody generation
were presented, spanning from human transloci in the mouse for the generation of heavily
mutated anti-HIV antibodies, to Burkitt lymphoma hypermutating cell lines, or the
use of the easily targetable avian DT40 cell line to select high affinity scFv variants
inserted at the endogenous Ig locus.
An overview of human monoclonal antibodies (mAbs) generated through Epstein-Barr virus
(EBV) immortalization was given by a pioneer of the field, Michael Steinitz (The Hebrew
University-Hadassah Medical School), who discussed passive immunization approaches
with fully tolerable, fully human antibodies. He reminded the audience that the idea
of producing human mAbs was inspired by the finding that EBV efficiently immortalizes
in vitro human B cells. Unlike other in vitro and in vivo methods proposed to generate
fully human or humanized antibodies, the antibodies obtained following EBV transformation
and subsequent molecular cloning result from the in vivo responses of immunized human
donors. Hence, they do not contain T cell epitopes, minimizing the risk of developing
HAGA (human anti-globulin antibody) reactions. The approach of mining the natural
B cell repertoire for auto-antibody specificities of functional relevance was therefore
developed, exemplified through the identification of anti-amyloid β natural antibodies
in almost every healthy individual.
2
The possible clinical relevance of the infusion of such antibodies is proposed, based
on mouse models of Alzheimer disease.
Session 1: The life of a B cell: diversity, Memory, aging and death
Claude-Agnès Reynaud and Sandrine Moutel
Claude-Agnès Reynaud (INSERM; Paris Descartes University) reported recent data from
a mouse transgenic model based on an AID-controlled, tamoxifen-inducible Cre reporter
system, in which the long-term follow-up of memory B cells can be achieved.
3
This model has allowed the description of IgM and IgG memory subsets with different
functions: an effector function for IgG memory, giving rise to antibody secreting
cells; and a central memory function for IgM memory, reinitiating the germinal center
reaction. The effective role of the IgM subset has been recently questioned, and Professor
Reynaud proposed that its mobilization might require a persistent germinal center
response to override the inhibition mediated through FcγRIIB binding by antigen-specific
IgG.
Doron Melamed (Technion; Israel Institute of Technology) discussed the issue of B
cell immunosenescence, a process occurring through accumulation of long-lived peripheral
B cells, with a decreased efficiency in activation and antibody response to vaccination
and to infectious agents. Interestingly, B cell depletion appears able to rejuvenate
the B cell compartment,
4
indicating that aging for B cells is a peripheral process, not a central one. The
use of B cell depleting therapies as an anti-age treatment is an interesting and provocative
notion, but not yet a therapeutic indication.
Idit Shachar (Weizmann Institute of Science) described the role of CD74 and its interaction
with macrophage migration inhibitory factor (MIF) in the induction of a c-Met operated
signaling cascade that conditions B cell survival. Interestingly, this survival pathway
operates also in chronic lymphocytic leukemia (CLL) in humans, mediating the upregulation
of bcl2 and an activation receptor of the SLAM family (CD84), leading to the production
of IL-8 that promotes cell survival.
5
Interfering with various steps downstream of this signaling pathway could thus represent
a therapeutic option in CLL.
Returning to the question of affinity maturation in the second guest lecture, Matthew
Scharff (Albert Einstein College of Medicine) concluded the meeting with his recent
work on mismatch repair. Professor Scharff discussed the paradox of the diversion
of repair pathways like mismatch repair, which is dedicated in most cells to the repair
of replication errors, but mobilized in hypermutation of immunoglobulin genes in B
cells for fixing and spreading mutations around the initial sites of AID-mediated
deaminations.
6
He reviewed the cascade of mismatch repair complex (MMR) factors and repair signals
that are diverted from their canonical error-free role and hijacked by B cells to
promote genetic diversification of the Ig locus. Now that the process of mutagenesis
is well-described, understanding how mutagenesis takes over repair remains the next
challenge. It has obvious biotechnological consequences if engineering of other actors
of this unique mutagenic process could also be tailored for affinity maturation in
vitro, as discussed at the beginning of the meeting by Professor Neuberger for AID.
Recent advances in the knowledge of human memory B cell compartment are now exploited
to derive fully human antibodies. Esther Breij (AIMM Therapeutics) described an efficient
method for generating antigen-specific B cell lines.
7
Human CD27+ memory B cells are isolated from peripheral blood and modified by introducing
BCL6 and Bcl-xL genes by retrovirus-mediated gene transfer. The culture of these cells
with CD40L and IL-21 converts them to highly proliferating, cell surface B cell receptor-positive,
Ig-secreting B cells with features of germinal center (GC) B cells, including expression
of AID. These cells can then be maintained in culture for several months and used
as a source of antibodies for further engineering. The method was used to select and
engineer powerful neutralizing antibodies against respiratory syncytial virus (RSV).
This method provides a new tool for the rapid generation of high affinity human mAbs.
Another efficient technology to screen, isolate and produce human antibodies was then
presented by Majid Mehtali (Vivalis). This method, developed in Japan, is based on
the use of microwell array chips and allows the screening of thousands of single B
cells by ELISA.
8
After two weeks of culture, RNA of selected cells are retrieved and antibody genes
cloned and produced as recombinant mAbs. The method also allows isolatation of antigen-specific
B cells present at low frequency (< 2x10−8) in peripheral blood of human donors. This
technology has been successfully applied by Vivalis for generatation of high affinity
mAbs against an infectious target. Finally, Dr. Mehtali described the use of the EB66®
cell line, derived from duck embryonic stem cells, to produce low-fucosylated mAbs
with enhanced antibody-dependent cell-mediated cytotoxicity (ADCC).
9
These cells can proliferate in suspension in stirred-tank bioreactors to reach high
cell densities in serum-free media, with a yield of mAb production of about 1 g/L.
Session 2: Molecular events: the making of an antibody molecule
Reuven Laskov, Jean-Luc Teillaud, Sandrine Moutel
Several lectures were devoted to studies on different aspects of the regulation of
antibody diversity, such as allelic exclusion, SHM, class switch recombination and
some of their implications in normal and pathological states. Yehudit Bergman (The
Hebrew University; Hadassah Medical School) summarized her studies on the mechanisms
of allelic exclusion of the murine kappa light chain gene and showed that this important
biological event, which eventually results in the expression of only one B cell receptor
in single B cells, is epigenetically regulated at several stages of B cell development.
It starts at the early embryonic stages where there is an asynchronous replication
of the two kappa light chain alleles in a stochastic manner, where the “early-replicating
allele” serves later as a preferential target for epigenetic modifications of its
histone-chromatin structure in the pro-B cell stage. Later at the pre-B cell stage,
these epigenetic changes dictate the V(J) rearrangement of the modified allele, due
to mono-allelic binding of RAG proteins.
10
Reuven Laskov (The Hebrew University; Hadassah Medical School), described features
of SHM in the IgVH alleles in EBV transformed human B cell line, an experimental model
system that lacks antigenic selection and enables study of the intrinsic properties
of SHM. Similar to the in vivo situation, the mutations were preferentially targeted
to the C/G nucleotides in the classical WRCH/DGYW hotspot motifs, and not to A/T nucleotides,
suggesting that they were generated during the first stage of AID activity. Lineage
tree analysis of the accumulating mutations showed the presence of a significant fraction
of independent repetitive mutations and also of two repetitive “mutational-clusters”
in the productive IgVH allele. Interestingly, 4 out of the 5 hypermutable sites of
these clusters reside in non-hotspot motifs, suggesting that, in addition to the hotspot
targeted sites, some transcriptional dependent secondary DNA structures significantly
contribute to SHM.
11
Bernardo Reina-San-Martin (Institut de Génétique et de Biologie Moléculaire et Cellulaire)
found that specific epigenetic changes are critical for directing the process of class
switch recombination (CSR) and generating different antibody isotypes during the immune
response. AID is already known to be critical in the initiation of this process by
binding to the switch regions (S) and inducing double strand DNA breaks that are a
pre-requisite for CSR. He also showed that, both in vitro and in vivo, AID can form
a tertiary complex with KAP1 (KRAB domain-associated protein 1), a transcription co-repressor
factor and with HP1 (heterochromatin protein 1), a heterochromatin binding protein,
and that this complex preferentially associates during transcription with modified
histone sites (trimethylated histone H3 at lysine 9, H3K9me3) located in the donor
S-regions. It is proposed that this association generates, during transcription of
the S-regions, both double strand DNA breaks and somatic point mutations that promote
CSR. Of note, the use of KAP1 conditional knockout mice indicated that this transcription
factor is required for efficient CSR, but dispensable for SHM.
12
Michel Cogné (CNRS; Limoges University) described the important role of the cis-regulatory
regions located into the constant IgH gene cluster or in its immediate downstream
3′ regulatory region (3′ RR) encompassing the four hs3a, hs1, 2, hs3b, and hs4 transcriptional
enhancers, on immunoglobulin B cell receptor expression in mature B cells. Transgenic
and knockout murine models of far-apart regions of this locus helped to elucidate
the role of their interactions on Ig gene expression, such as shutting-off the expression
of the IgH locus in activated mature B cells. Deletion of the RR region from the mouse
genome demonstrated that this region only controls IgH locus during the late stages
of B cell differentiation.
13
Professor Cogné also described a new recombination mechanism, that takes place in
pseudo-switch sequences located within this 3′RR region and mediates the complete
deletion of the heavy chain locus. This is a new “suicide mechanism” for B cells that
is most probably of physiological relevance because these pseudo-switch elements are
conserved between species.
Bertrand Nadel (CNRS; INSERM; Méditerranée University) presented his studies on the
early stages of follicular lymphoma (FL) in humans, an incurable long-term disease
that originates from the germinal center B cells, and that represents the second most
common adult lymphoid neoplasia. Molecular analyses in healthy donors and FL patients
indicated that subversion of the molecular processes occurs all along B cell differentiation
in vivo, leading to the generation of pre-tumoral clones and to their homing in “niches”
in lymphoid organs. It was emphasized that novel methods of studying B cell molecular
characteristics allow better insights into the earliest stages of FL and may differentiate
between “healthy carriers” and “early patients,” which may eventually aid in the development
of earlier diagnostic tools.
14
Dan Eilat (The Hebrew University; Hadassah Hospital) explored whether receptor editing
of the light chain or allelic inclusion can explain the failure of self-tolerance
and the development of autoimmune systemic lupus erythematosis (SLE) in NZB/NZW mice
(a murine model of the human disease). Using sophisticated “knock in” and double-transgenic
mice generated on normal or NZB/NZW genetic backgrounds and hybridoma generation,
he showed that, in both strains, receptor editing and allelic inclusion occurs to
about the same extent and therefore cannot explain the lack of tolerance and the development
of auto-reactivity and the formation of anti-DNA antibodies in the diseased animals.
15
The striking difference between the two strains was that B cells in the disease-prone
NZB/NZW mice used the process of receptor editing to augment, rather than to suppress,
the autoimmune response.
When one thinks of therapeutic antibodies, only IgG isotypes are considered in most
cases. Renato Monteiro (INSERM; Paris Diderot University) discussed the IgA and IgA
Fc receptors as tools for immunotherapy. IgA is the second most abundant Ig isotype
in serum where it is mostly present as monomer, in contrast to mucosal surfaces where
it is present as dimeric secretory IgA (SIgA). Monomeric IgA have an anti-inflammatory
capacity, while being poorly involved in systemic immune responses. Thus, the question
of the underlying mechanisms by which monomeric IgA can down-modulate IgG-mediated
phagocytosis, chemotaxis, oxidative burst and cytokine release in absence of antigen
has been explored. Professor Monteiro showed that the IgA Fc receptor I (FcαRI or
CD89) has inhibitory or activating function depending on whether the receptor is engaged
by monomeric or multimeric ligands.
16
This finding has led to the view that immunoreceptor tyrosine-based activation motifs
(ITAMs) can also be involved in inhibitory activities. Interestingly, the inhibitory
function of ITAM (ITAMi) results from a partial phosphorylation of the ITAM motif
of the associated FcγR subunit. Monovalent targeting of FcαRI favors the recruitment
of the protein tyrosine phosphatase SHP-1 that impairs signaling induced by activating
receptors through the formation of polarized intracellular clusters (“inhibisomes”)
containing these receptors, signaling effectors and SHP-1. The finding underlines
the crucial role of ITAMi signaling in the control of inflammatory responses. Professor
Monteiro and his colleagues also identified an anti-CD89 Fab as a new potential therapeutic
tool to prevent the progression of renal inflammatory diseases.
Over the last decade, a lot of efforts have been devoted to the increase of the effector
functions of therapeutic mAbs exhibiting a human IgG1 Fc domain. Sven Berger (Pierre
Fabre Research Institute) described the engineering of the hinge region and of Fc
domains to increase or decrease the binding of IgG1 to various FcγR, to C1q or to
the neonatal Fc receptor (FcRn). Engineering has also targeted the glycosylation pattern
of human IgG1, allowing the generation of antibodies with potent ADCC activities due
to their increase binding to FcγRIIIa. This binding depends on residues located in
the flexible hinge region and the CH2 domain. Dr. Berger showed that, by replacing
the human hinge with its less flexible mouse counterpart,, one could get an antibody
exhibiting an antagonist activity rather than an agonist one. The pharmaceutical industry
is now attempting to translate these research tools into serious clinical candidates.
17
The Pierre Fabre Research Institute is currently developing therapeutic mAbs in collaboration
with Merck, Sharp and Dohme (dalotuzumab, an anti-IGF-1R mAb now in Phase 2), and
with Abbott (h224G11, an antagonist anti-human c-Met mAb that blocks c-Met dimerization).
Dr. Berger still recommends remaining cautious about the potential immunogenicity
of these new drugs, especially when they are manipulated to increase their half-life.
October 24, 2011: Day 2
Session 3-A: Antibody effector functions: molecules and cells in a complex environment
Jean-Luc Teillaud
Wolf-Herman Fridman (Georges Pompidou European Hospital – Paris Descartes University)
discussed the complexity of tumor microenvironment and the role of immune cells in
controlling cancer spread. It is now established that these factors strongly affect
the clinical outcome of the patients.
18
Professor Fridman noted that the infiltration of T cells within solid tumors such
as colorectal cancers (CRC) and other human malignancies, marked by a Th1/cytotoxic
reaction, is the strongest prognostic factor for the overall survival as shown by
studies of large cohorts of patients, based on integrated biology approaches. He also
reported that CX3CL1, CXCL10, and CXCL9 chemokines are important players in the elaboration
of an efficient immune pattern. The presence of tertiary lymphoid structures adjacent
to the tumors containing mature dendritic cells (DCs), as observed in lung carcinomas,
could polarize naïve T cells toward Th1/cytotoxic subsets. Overall, these immune mechanisms
could control the homing of metastatic cancer cells in new areas. All these observations
raise the question of how mAb-based therapies affect the host immune system in cancer
patients.
The interest of targeting cell surface molecules that supports migration in inflammatory
or cancer conditions has been then emphasized by David Naor (Hadassah Medical School
- The Hebrew University).
19
He showed that the targeting CD44 or RHAMM (receptor hyaluronic acid mediated motility
or CD168) by anti-CD44 or anti-RHAMM mAbs is efficient in reducing or eradicating
the inflammatory activity in different mouse models [rheumatoid arthritis, multiple
sclerosis and type 1 diabetes (T1D)], even when the treatment was initiated after
the onset of the disease as is the case with T1D. Interestingly, Professor Naor showed
that RHAMM treatment could compensate CD44 in supporting in vitro cell migration and
in vivo invasion of inflammatory cells into inflamed joints in a collagen-induced
arthritis (CIA) mouse model where CD44 is genetically deleted. The detailed study
of the NOD mouse model of T1D also demonstrated that RHAMM and CD44 cooperate in supporting
inflammatory cell invasion into the pancreatic islets, although the presence of CD44
is not mandatory. However, the expression of CD44 by insulin secreting cells is needed
for killing by invading cells. Professor Naor concluded that RHAMM is a new therapeutic
marker in T1D.
The importance of modulating the immune system of cancer patients was then reported
by Rinat Rotem-Yehudar (CureTech Ltd).
20
She showed that, in experimental tumor models, treatment with an antibody (CT-011)
that targets PD-1 (Program Death-1), an inhibitory receptor present on effector lymphocytes
and whose ligands (PD-L1 and PD-L2) trigger apoptosis, not only induces immediate
anti-tumor effect but also facilitates the generation of a tumor-specific memory control.
Interim results of a Phase 2 study in patients with diffuse large B cell lymphoma
(DLBCL) after autologous stem cell transplantation showed a positive clinical effect
as compared with an average of relevant historical controls. A peripheral increase
of specific subsets of memory T cells was also observed. Thus, Dr. Rotem-Yehudar concluded
that PD-1 blockade by CT-011 could be related to clinical efficacy in aggressive lymphoma
patients. Of note, a Phase 2 clinical trial of the combination of rituximab and CT-011
is in progress.
The idea that combinations of mAbs with different molecular targets could enhance
the overall response in cancer patients is currently being explored in a number of
laboratories and clinical departments. Hélène Haegel (Transgène S.A.) proposed an
alternative to this approach. She reported the development of a humanized mAb directed
to CD115, TG3003, with multiple intervention points for cancer therapy. CD115, the
CSF-1 receptor is expressed on many epithelial cancers (breast, ovary…), on tumor-associated
macrophages that control tumor progression, as well as on osteoclasts that are responsible
for metastasis-induced bone destruction. She showed that treatment with anti-CD115
mAb decreases tumor growth, expression of F4/80 macrophage marker in tumors, and inhibits
bone destruction in mouse tumor models. Analysis of TG3003 indicated that it exerts
selectively ADCC against CD115+ tumor cells and inhibits CD115 function through a
non-ligand competitive mode of action. Thus, Dr. Haegel concluded that TG3003 is a
promising candidate for the treatment of solid tumors associated with bone metastases.
The engineering and production of new antibody formats that allow the recruitment
of cytotoxic effector mechanisms remain a major challenge for antibody-based cancer
therapies. Zelig Eshhar (The Weizmann Institute of Science) described his efforts
to develop antibody-based chimeric receptors for adoptive cell therapy.
21
He showed that “T-bodies,” which are T cells engineered to express an antibody-based
chimeric receptor (CAR) made of an anti-tumor (HER2/Neu, CEA, CD24) scFv linked to
a hinge region, a transmembrane motif and intracellular signaling (CD3zeta or FcRγ
chain) moieties, can be used in adoptive cell therapy (ACT) of cancer. Experimental
models using immuno-deficient mice showed that medium to high affinity CARs are the
best candidates for tumor elimination. However, Professor Eshhar noted that very high
affinity CARs can bind target molecules expressed at a low density on normal cells,
hence leading to severe side effects. Exploration of T-bodies anti-tumor effect showed
that intra-tumor inoculation is an efficient therapeutic method. Current ACT approaches
employ the use of autologous lymphocytes. Using allogeneic lymphocytes in ACT is problematic
because it causes severe immunological reaction against HLA-mismatched individual.
Professor Eshhar and his team have recently set up a new method to overcome this issue.
It is based on a mild pre-conditioning of the recipient animal before the injection
of allogeneic T cells redirected with a human Her2/neu-specific CAR in a mouse model
of metastatic disease. A transient lympho-depletion is achieved by irradiation or
cyclophosphamide injection. It delays the rejection of the allogeneic T-bodies that
have enough time to destroy the tumor, but not enough to cause significant damage
to the host. Professor Eshhar proposed that ex vivo generated HLA-mismatched allogeneic
T-bodies could be used as a universal tool for ACT, while their graft-vs.-host reactivity
can be controlled by pre-conditioning treatment.
Recruitment of effector cytotoxic cells can also be potentiated by selection of structurally
well-defined Fc region of IgG therapeutic antibodies, as discussed by Christophe de
Romeuf (Laboratoire français du Fractionnement et des Biotechnologies). He presented
data showing that improvement of IgG Fc interaction with FcγRIIIa expressed on NK
cells and macrophages by a careful selection of the Fc glycosylation pattern (low
fucosylation) leads to a much stronger ADCC in vitro. Dr. de Romeuf showed as an example
that a fully human anti-D antibody (LFB-R593) that strongly binds FcγRIIIa exhibits
excellent ADCC in vitro, even when elevated amounts of IVIg (i.e., polyclonal IgG)
are present in the assay. Clinical trials showed that LFB-R593 has the same ability
to clear RhD+ red blood cells in human volunteers as the reference polyclonal anti-D
antibodies currently used for the prevention of fœto-maternal allo-immunization. Similarly,
he presented data showing that a low-fucosylated FcγRIII anti-CD20 mAb (LFB-R603)
is able to efficiently kill CLL cells that express low density of CD20 in vitro and
depletes rapidly peripheral B cells in all CLL patients medicated with methyl-prednisone.
The anti-tumor effect of another low-fucosylated mAb, directed against the human Müllerian
Inhibiting Substance type II receptor (MISRII) expressed on most ovarian cancer subtypes
is currently investigated in mouse models.
Itai Benhar (The George S. Wise Faculty of Life Sciences - Tel Aviv University) then
presented the production of a new method to produce IgG and IgG-cargo fusion molecule
in bacteria, named “Inclonals” technology.
23
It is based on expert biochemical procedures starting with the different proteins
(heavy and light chains of IgG, toxins) produced in bacterial cultures and present
in insoluble inclusion bodies. The method has the advantage of leading to a high yield
of highly purified full-length antibodies and fusion molecules that exhibit the same
affinity and stability properties as their mammalian cell culture-produced counterparts.
Professor Benhar stressed that the method also allows the efficient production of
toxic proteins as opposed to mammalian cells. Examples of the production and use of
fluorescent IgG-based fusion molecules were given. It was concluded that the technology
is an attractive option for antibody production.
October 25, 2011: Day 3
Session 4: Engineering Antibodies
Sandrine Moutel
Irit Sagi (The Weizmann Institute of Science) presented data on the rational design
of antibodies directed to the active sites of metalloenzymes.
24
Among these are the zinc-dependent endopeptidases belonging to the matrix metalloproteinases
(MMPs) family, enzymes that are involved in vivo in cell migration and that have been
strongly associated with tumor metastasis, inflammation, tissue degeneration and cell
death. The catalytic metal centers of zinc-dependent endopeptidases are therefore
attractive targets. Molecular inhibitors have already been designed and used in vivo,
but these molecules inhibit the entire MMPs family and cause severe side effects.
The laboratory of Professor Sagi has recently developed antibodies, termed metallobodies,
targeting the MMPs catalytic sites. An original immunization strategy based on molecular
mimicry was set up. A synthetic molecule mimicking the conserved structure of the
metalloenzyme catalytic zinc-histidine complex residing within the active site of
gelatinases (MMP2 and 9) was used. Hence, the antibodies generated appear to be similar
to the naturally occurring MMPs inhibitors [endogenous tissue inhibitors of metalloproteinases
(TIMPs)].
24
These antibodies bind the catalytic metal ion in the enzyme active conformation and
compete with zinc binding, exhibiting an exquisite specificity. The therapeutic potential
of these antibodies has been demonstrated in models of inflammatory bowel disease
(IBD) using knockout MMP2/9 mouse.
Serge Muyldermans (Vrije Universiteit Brussel) presented an overview on single variable
VHH domains (Nanobody®) derived from Camelidae. Owing to their small size, VHH can
be easily manipulated, formatted as bispecific antibodies or fused to Fc domains or
other polypeptides. They offer many advantages for research, industrial and clinical
applications over conventional mAbs. They can be produced with a high yield in bacteria,
they exhibit excellent stability and good solubility, and their production cost is
lower. VHH are usually isolated after immunization of llama or dromedary, cloning
of the VHH repertoire and selection by phage display. Dr. Muyldermans reported the
selection of VHH that modulate the conformation and spectral properties of green fluorescent
protein (GFP) and their use as intrabodies to track molecules within living cells.
25
Also, a VHH recognizing a conserved carbohydrate of the variant surface coat glycoprotein
that covers the entire surface of Trypanosoma brucei rhodesiense responsible for sleeping
sickness in East Africa has been developed. Once conjugated with a truncated form
of apoL-I (Tr-apoL-I) that is derived from apoL-I-neutralizing serum resistance-associated
(SRA) protein expressed by the parasite and lytic for T. b. rhodesiense, the molecule
could eliminate the parasites from blood. Professor Muyldermans also indicated that,
due to their short half-life in blood, VHH can also be used for in vivo imaging and
radioimmunotherapy (RIT). An anti-HER2/Neu VHH conjugated either to technetium-99
or to lutetium-177 radionuclides is currently developed for imaging and RIT of HER2-positive
breast cancer, respectively.
Roger MacKenzie (National Research Council Canada) also reported the development of
therapeutic VHH. AFAI is a VHH targeting carcinoembryonic antigen-related cell adhesion
molecule 6 (CEACAM6) that was isolated from a non-immune camelid. He noted that AFAI
has been formatted as a pentamer. The molecule strongly stains several tumor types
and, unlike conventional antibodies to CEACAM6, discriminates between tumor and normal
tissues. Dr. MacKenzie hypothesized that this VHH binds a hidden epitope otherwise
inaccessible to conventional antibodies due to its small size and a long complementarity
determining region 3 (CDR3). The company HelixBiopharma has fused the pentameric AFAI
with a plant enzyme, urease (DOS47), which converts urea to ammonia, a toxic metabolic
product. The resulting drug candidate, L-DOS47, is designed to act in a targeted manner
by selectively recognizing non-small cell lung cancer cells to produce a potent anti-tumor
effect based on the local production of ammonia at the tumor sites.
The presentation of Yoram Reiter (Technion - Israel Institute of Technology) detailed
two strategies for cancer immunotherapies that involve combining high affinity antibody
fragments with the recruitment of cytotoxic CD8+ T cells. The first consists of fusing
a scFv specific for tumor cell surface antigens to a human single-chain HLA-A2 molecule
covalently linked to a tumor or viral-derived peptide to recruit specific cytotoxic
T cells that are able to kill the scFv-decorated target cells.
26
The second strategy is developed in collaboration with Applied Immune Technologies
(AIT). AIT has created a platform technology for the development of human recombinant
T-Cell Receptor-Like (TCRL™) antibodies capable of binding to intracellular-derived
peptides presented by MHC molecules. Novel therapeutic scFv mimicking the T cell receptor,
binding to various peptide-MHC complexes, have been generated for therapeutic and
diagnostic applications in a variety of cancer, viral and autoimmune diseases. The
identification of candidate peptides derived from the proteasome of patients has been
accelerated by a new tool that combines bioinformatic analysis and mass spectroscopy
(EpiTarget™ Discovery). This made it possible to identify more than 2000 new peptides
presented on MHC molecules from patient cells, from which 33 have been selected as
peptide candidates for novel MHC/peptide based targets.
Session 5: Learning from treatment with therapeutic antibodies
Marie-Caroline Dieu-Nosjean
Josée Golay (Laboratory of Cellular therapy «G. Lanzani», Division of Hematology,
Ospedali Riuniti) reported on the role of macrophages on the therapeutic activity
of the anti-CD20 antibody rituximab. In vivo, macrophages can have a dual role with
a negative impact on tumor growth and a positive impact on rituximab-mediated therapeutic
activity. Dr. Golay showed that M2 type tumor-infiltrating macrophages, although producing
IL-10 and VEGF that negatively impact anti-tumor response and angiogenesis and lead
to tumor growth, exhibit a strong phagocytic activity against rituximab-opsonized
CLL targets.
27
In vitro experiments showed that addition of IL-10 increased phagocytosis by both
M-CSF- and GM-CSF-differentiated macrophages. These data demonstrate that future therapeutic
antibody-based strategies combined with cellular therapies should also aim at improving
antibody-dependent phagocytosis (ADP), as well as complement-dependent cytotoxicity
(CDC) and ADCC.
Recent reports have suggested that antibody treatment in oncology not only involves
cells from the innate immunity but also cells from the adaptive immunity. Jean-Luc
Teillaud (INSERM – Paris Descartes University) first reminded the conference participants
that many advances have been achieved over the last decade to improve the effector
function of therapeutic antibodies via better recruitment of FcγR+ cells of the innate
immunity. In particular, he showed that ADCC against CLL cells can be strongly increased
in vitro when using a low-fucosylated antibody directed to CD20. Dr. Teillaud then
showed that antibody treatment of CD20+ tumor-bearing mice induces the recruitment
of CD4+ cells that are required to induce long-term protection.
28
The presence of CD4+ cells is also required when mAb-treated surviving mice are challenged
with tumor cells. Interestingly, the presence of CD8+ cells was not required at the
initiation of the treatment, but was necessary after tumor challenge. Thus, these
findings indicate that mAb-based treatment of tumors induces an adaptive cellular
immunity. It paves the way to the development of new therapeutic strategies aimed
at strengthening this adaptive response, as exemplified by Dr. Teillaud who showed
that IL-2 improves the overall survival rate when given after tumor challenge. The
induction of a long-lasting anti-tumor response following anti-CD20 treatment has
been also explored by Nurit Hollander (Sackler School of Medicine - Tel Aviv University).
29
She discussed the use of anti-CD20 mAbs in combination with dendritic cells (DCs)
in order to supplement the anti-CD20-induced depletion of normal B cells, another
subset of antigen-presenting cells. In a therapeutic model of disseminated tumor,
Professor Hollander reported a synergistic effect of tumor cell-loaded DC injected
subcutaneously with chemotherapy and anti-CD20 antibody. Alternatively, intra-tumoral
injection of naïve DCs in addition to mAb treatment and chemotherapy (cyclophosphamide)
was also shown to be effective. All these observations suggest that the clinical efficacy
of anti-CD20 mAb treatment can be increased by inducing T cell-mediated antitumor
immune responses with long-term memory.
Session 6: Monoclonal antibodies in the treatment of auto-immune and viral diseases
Marie-Caroline Dieu-Nosjean and Sandrine Moutel
Luc Mouthon (Cochin Hospital - Paris Descartes University) discussed the clinical
use of intravenous immunoglobulin (IVIg) preparations and mAbs as therapeutic antibodies.
Professor Mouthon presented an overview of the efficacy, as well as the limitation,
of antibody-based treatment through many examples of inflammatory and auto-immune
diseases.
30
He pointed out that the consumption of IVIg increases each year due to their use in
a large number of these diseases. The interest of different mAbs and fusion proteins
in the treatment of inflammatory and auto-immune diseases such as anti-TNF mAbs, the
TNFαRII-Fc IgG1 fusion protein etanercept,, and abatacept, an IL-1RA recombinant molecule
in rheumatoid arthritis, as well as rituximab in systemic vasculitis, was also discussed.
Yaakov Naparstek (Hadassah - Hebrew University Medical Center) described an IL-10-mediated
mechanism involved in the resistance to adjuvant arthritis that has been exploited
to develop a new therapeutic mAb. He noted that the serum of rheumatoid arthritis
patients contains anti-heat shock protein (HSP) antibodies directed against peptide
6, a 16 amino-acid surface epitope. These antibodies cross-react with an epitope of
adenyl cyclase associated protein (CAP1) present on the surface of monocytes. Professor
Naparstek showed that the binding of a humanized anti-peptide 6 mAb to monocytes results
in the secretion of the immunosuppressive cytokine IL-10 and the suppression of the
disease in different animal models such as collagen-induced arthritis and 2,4,6-trinitrobenzene
sulfonic acid (TNBS)-induced colitis. He concluded his presentation by indicating
that this humanized antibody may be used as a therapeutic tool in human auto-immune
and inflammatory diseases.
Yves Poulain (Hôtel Dieu Hospital of Quebec - Laval University) presented the physiopathology
of the psoriasis and the limitation of the existing therapies. Professor Poulain reported
new data obtained in clinical Phase 2 and 3 studies in patients treated with antibodies
that target cytokines and receptors or block the DC-T cell interaction. He noted that
a significant improvement of the quality of life of patients have been obtained with
the advent of biotherapies based on the use of anti-TNF molecules. Professor Poulain
also stressed that the use of antibodies directed against the IL-12/IL-23 common p40
subunit such as the human mAb ustekinumab, had led to major improvement in patients
with moderate or severe plaque psoriasis.
31
This approach has been elaborated because IL-23 is essential for the differentiation
of Th17 lymphocytes, a T cell subset involved in chronic inflammatory and autoimmune
diseases. However, as already observed with anti-TNF treatments, the use of neutralizing
anti-IL-12/IL-23 antibody has been associated with serious adverse events, including
an increased risk of infection. Long-term follow-up of treated patients will give
a more accurate insight in the benefit-risk ratio of this treatment, although no evidence
of cumulative toxicity has been observed to date with up to four years of exposure.
Roberto Mancini (Università Vita Salute San Raffaele) presented studies on mAbs isolated
from human antibody repertoires directed to influenza
32
and hepatitis C (HCV) viruses. Two antibodies were obtained from a patient exposed
to H1N1 influenza A strains with a negative clinical history for influenza infection
over the last years, but with a detectable serum neutralizing activity against a 1934
influenza A strain. These two mAbs exhibited a strong neutralization activity against
both swine-origin influenza virus (S-OIV) and a broad range of human and swine H1N1
isolates. Thus, these two mAbs could constitute the basis, alone or in a combination
with other mAbs, for a new class of drugs to be used to design a vaccine and in the
prophylaxis of this disease. Similarly, Dr. Mancini and his colleagues developed from
a chronically infected patient an “intrabody” mAb that inhibits the activity and replication
of HCV. This mAb could be an attractive tool when a gene delivery method of antibody
to liver cells will be effective.
Jamie Scott discussed the properties of antibodies produced during acute vs. chronic
viral infections.
33
She showed that neutralizing anti-HIV antibodies are raised only after protracted
rounds of viral escape from the antibody response. Interestingly, antibodies derived
from chronically infected patients demonstrate peculiar structural features in terms
of CDR-H3 length, somatic mutations, likely due to persistent antigen selection during
chronic infection, and favor distal germline V(H)-genes usage. Thus, Professor Scott
raised the issue of whether the generation of antibodies sharing the characteristics
of those elicited during chronic infections should be a major goal when developing
a protective vaccine, in particular in the case of HIV infection.