www.elsevier.com/locate/intimp
Abstract
As arbitrators of the immune response, dendritic cells (DC) are uniquely placed to negotiate the balance between the
opposing forces of tolerance and immunity, making them attractive candidates for clinical applications. Accordingly, DC have
been used successfully in the treatment of cancer, enhancing immune responses to tumour-associated antigens (TAA) in
experimental animal models and phase I clinical trials. A novel source of DC that has recently been described is the embryonic
stem (ES) cell whose differentiation in vitro may be directed along multiple lineage pathways. Such pluripotency offers
unparalleled opportunities for the treatment of chronic and degenerative disease states by the replacement of affected tissues, a
vision which has inspired the emerging field of regenerative medicine. By sharing the genotype of therapeutic cell types, such as
cardiomyocytes and dopaminergic neurons derived from the same ES cell line, so-called esDC may offer prospects for
reprogramming the immune system to tolerate the grafted tissues. Here, we describe how the unique properties of esDC and the
ES cells from which they derive, make them eminently suited to clinical applications, overcoming many of the issues that
currently limit the effectiveness of DC-based immune intervention.
D 2004 Elsevier B.V. All rights reserved.
estimates suggest that upwards of 3000 Americans die phenotype and propensity for long-term storage. While
each day of conditions that may ultimately be treated CD34+ stem cells may be stored frozen over prolonged
through the panacea of regenerative medicine [5]. periods of time, access to such progenitors is largely
Although various animal models of disease have constrained by the limited availability of umbilical cord
proven amenable to treatment by such an approach blood; conversely, DC differentiated from the patient’s
[6–8], translation to the clinic will require an effective own peripheral blood monocytes are more readily
strategy for addressing the immunological barriers of available but may show less phenotypic stability with
histoincompatibility between the recipient of CRT and time, displaying an inherent tendency for spontaneous
the ES cell line from which the tissues are derived. maturation in culture. Secondly, an issue which has
A promising approach to this problem lies in the proven to be the rate-limiting step in cancer immuno-
dual capacity of ES cells to act as a source of therapy has been the identification of appropriate target
therapeutic cell types to ameliorate the progression of antigens and the most effective form for their delivery
disease and of leukocyte subsets to ensure their to DC; synthetic peptides, recombinant TAA and
acceptance through the induction of transplantation unfractionated tumour lysates have each been used
tolerance [9]. Given the emerging role of dendritic with varying success. While endogenous expression of
cells (DC) in the maintenance of self-tolerance under TAA is widely believed to be the most appropriate form
steady state conditions and their ability to reestablish of delivery by virtue of the access it provides to both
tolerance in both autoimmune and alloimmune set- MHC class I and II pathways of antigen processing, the
tings [10–12], the description of protocols for their approach is limited by the need for effective protocols
derivation from mouse ES cells [13,14] represents an for genetic modification of the DC inoculum, which
important step towards realising this goal. remains a significant obstacle. Although the develop-
While the tolerogenicity of DC has been appreciated ment of retroviral and adenoviral vectors has begun to
for some years, an understanding of the molecular basis fulfill this need, their use in a clinical context
of this phenomenon remains less well defined than the introduces a significant element of risk which may be
mechanisms that underlie immunogenicity. Conse- difficult to justify on ethical grounds [23].
quently, the use of DC in a therapeutic context has so These considerations help illustrate the pressing
far been largely restricted to strategies for vaccination need for quality control of preparations of DC intended
to tumour-associated antigens (TAA) [15]. Such an for clinical use, so as to ensure both safety and efficacy
approach to cancer immunotherapy has enjoyed con- of the procedure. Clearly, any source of DC that permits
siderable success in animal models of disease [16–18] consistency and reproducibility between batches, while
and has inspired a number of small-scale clinical trials providing ample opportunities for quality control,
for the treatment of melanoma, lymphoma, myeloma, would offer significant advantages over the current
as well as prostate and renal cancer [19,20]. The art. Although efforts are presently underway in a
expansion of tumour-specific cytotoxic T lymphocytes number of laboratories to secure the differentiation of
(CTL) in these studies has correlated with regression of DC from human ES cells, we discuss below how our
established tumours and secondary metastases in a own experiences of esDC in the mouse suggest that
proportion of patients, although clinical outcomes have they fulfil many of these requirements, auguring well
varied considerably between trials. This inconsistency for their ultimate application to the clinic.
has highlighted the need to optimise and standardise
protocols for the derivation, culture and administration
of DC; indeed, in the 8 years since the first trial was 2. Materials and methods
performed, various issues have been identified that
must be addressed before DC-based treatment regimes 2.1. ES cell lines and mice
may be offered as reliable alternatives to conventional
therapies [21,22]. ESF116 is an ES cell line derived from CBA/Ca
Firstly, identification of the most appropriate source mice which is karyotypically male and has been
of DC remains to be resolved, the main criteria for their described in detail previously [13]. ESF121 is likewise
use in vivo being their availability, yield, stability of of CBA/Ca origin but is karyotypically female. Both
P.J. Fairchild et al. / International Immunopharmacology 5 (2005) 13–21 15
lines display germline competence and were cultured dry ice before being stored long term under liquid
according to standard protocols [14]. C57Bl/10 mice nitrogen.
were used as a source of allogeneic T cells. To induce the maturation of esDC, replated cells
A1xRAG1 / mice transgenically express a T cell were exposed to 1 Ag/ml of lipopolysaccharide (LPS)
receptor (TCR) specific for the 479–493 epitope of the from E. coli Serotype 0127:B8 (Sigma). After over-
male specific antigen, Dby, in the context of the MHC night culture, the medium was replaced and the cells
class II determinant, H2Ek. Female A1xRAG1 / cultured for a further 24 h: mature esDC were
mice were, therefore, used as a source of T cells harvested the following day as the nonadherent
specific for this well-defined minor histocompatibility population. Populations of control bone marrow
(mH) antigen. derived DC (bmDC) were obtained by culturing bone
marrow cells from CBA/Ca mice in medium supple-
2.2. Directed differentiation of ES cells mented with 25 ng/ml of GM-CSF. Cultures were fed
by replacement of the medium on days 3 and 6 of
Protocols for directing the differentiation of pluri- culture and maturation was specifically induced by
potent ES cells towards the DC lineage have been addition of 1 Ag/ml of LPS on day 6.
described in detail elsewhere [14]. Briefly, ES cells
were cultured on mitotically inactivated embryonic 2.3. Proliferative assays
fibroblast feeder cells in DMEM (Gibco) supple-
mented with 15% FCS, 2 mM l-glutamine, 1 mM Mature or immature esDC (5104 cells per well of a
sodium pyruvate and 510 5 M 2-mercaptoethanol. 96-well round-bottomed plate) were cocultured with
Genetic modification of the parent ES cell lines was 2105 nylon wool purified T cells from either alloge-
achieved by lipofection as described [14]. Stable neic C57Bl/10 or TCR transgenic A1xRAG1 / female
transfectants were selected by neomycin resistance mice. After 2 days of culture, proliferation was assessed
and were cloned at the single-cell level to produce by pulsing with 25 ACi/well of 3H-thymidine deoxy-
individual ES cell lines which were stored frozen. ribose (TdR; Amersham). Cultures were harvested and
3
As a prelude to their differentiation, ES cells were H-TdR incorporation measured using a flat bed
passaged twice in gelatinised flasks in medium scintillation counter.
supplemented with 1000 U/ml of recombinant leukae-
mia inhibitory factor (rLIF; Chemicon) to remove the 2.4. Flow cytometry
feeder cells. ES cells were then cultured in suspension
at low density (4105 cells per 90-mm dish of The surface phenotype of DC was assessed by flow
bacteriological plastic) so as to permit the formation cytometry using FITC-conjugated monoclonal anti-
of embryoid bodies (EB). After 14 days, EB were bodies (mAb) to MHC class II (Clone OX6; mouse
plated onto tissue culture plastic in medium containing IgG1) or biotinylated anti-CD86 (Clone GL-1; Rat
200 U/ml recombinant interleukin-3 (IL-3; R&D IgG2a), followed by ExtrAvidin-FITC (Sigma). Irrel-
Systems) and 25 ng/ml of granulocyte-macrophage evant, species- and isotype-matched mAb were used
colony stimulating factor (GM-CSF) to support the as negative controls. Cells were stained on ice for 45
outgrowth of DC. min, washed and fixed in 1% formaldehyde before
ES cell-derived DC were harvested by gentle being analysed by means of a Becton Dickinson
pipetting to release lightly adherent cells, and were FACSCalibur using CellQuest software.
replated in fresh medium containing IL-3 and GM-
CSF, to permit the adherence of contaminating
stromal cells. After overnight culture, esDC were 3. Results
again dislodged by gentle pipetting and used in
functional assays or stored frozen for future use. 3.1. Yield and storage of esDC
Cryopreservation was achieved by resuspending
esDC at a density of 5106 cells/ml in medium Protocols developed in our laboratory for the
containing 10% DMSO. Cells were frozen rapidly on differentiation of DC from mouse ES cell lines make
16 P.J. Fairchild et al. / International Immunopharmacology 5 (2005) 13–21
Fig. 2. Comparison of phenotypic stability among esDC and DC derived from bone marrow progenitors. Histograms represent the levels of
expression of MHC class II (A–D) and CD86 (E–H) by bmDC at day 9 (A and E) and day 14 of culture (B and F), showing spontaneous
maturation of a significant proportion of cells over time. For comparison, (C) and (G) depict expression levels of the respective surface markers
by bmDC, induced to mature in response to overnight incubation with LPS. (D) and (H) represent levels of surface expression by esDC
maintained continuously in culture for 35 days, showing a similar phenotypic profile to the most immature populations of bmDC (A and E).
apy. Despite considerable efforts over the past decade same ES cell line. Having differentiated from a
to identify TAA, the number of proteins truly common progenitor, esDC naturally share all relevant
restricted in their distribution to transformed cells is major and minor histocompatibility antigens with the
surprisingly small, even for melanoma, the best therapeutic cell types, their endogenous expression
defined of all tumours at the biochemical level. And obviating the need for the provision of antigen in an
yet, targeting destructive immune responses to anti- appropriate form for uptake by DC. Indeed, the
gens that are more widely disseminated than the molecular identity of the transplantation antigens
tumour itself may have potentially deleterious con- may become little more than academic rather than
sequences. Mice engineered to express a neoantigen representing an insurmountable obstacle to the imple-
in pancreatic h cells that was shared by a tumour cell mentation of immunotherapy. The issues of MHC
line with which they had been inoculated, developed restriction that greatly complicate the use of peptide
fatal autoimmune responses when vaccinated with DC vaccines are also circumvented, since esDC may be
presenting the same antigen in an immunogenic entrusted with the task of selecting appropriate
fashion [27]. Although DC have been generally well epitopes for presentation to the T cell repertoire in
tolerated during clinical trials, autoimmune sequalae the context of both MHC class I and class II
in the form of vitiligo have been reported as a determinants. Furthermore, the risk of eliciting erro-
consequence of vaccination against TAA expressed by neous autoimmune responses is minimised by the lack
melanoma but shared by normal, untransformed of alloantigen expression by the recipient’s own
melanocytes, illustrating the need for caution [21]. somatic tissues and the tolerogenic context in which
Importantly, the issues of antigen identity and the any shared autoantigenic epitopes are presented.
form and vehicle for its delivery to DC are rendered In order to investigate the capacity of esDC to
irrelevant when exploiting esDC for the induction of present endogenously expressed mH antigens, we
transplantation tolerance to tissues derived from the have made use of A1xRAG1 / mice, transgenically
18 P.J. Fairchild et al. / International Immunopharmacology 5 (2005) 13–21
expressing a TCR specific for the immunodominant welcome opportunities to subtly intervene in the
479–493 epitope of the male-restricted antigen, Dby, outcome of the immune response. Nevertheless, all
in the context of H2Ek. NaRve, splenic T cells from sources of DC, described to date, display significant
female A1xRAG1 / mice responded vigorously to barriers to the introduction of heterologous genes and
mature esDC differentiated from a karyotypically constructs. While lipofection is entirely inadequate as
male ES cell line of CBA/Ca origin, but failed to an approach to transfection (Fig. 4A), use of the
respond to female esDC (Fig. 3A). Nevertheless, cationic peptide, CL22, as a vehicle for delivery of
addition of the synthetic Dby peptide to cultures plasmid DNA has enjoyed some measure of success,
restored maximal responses in either case (Fig. 3B). up to 17% of mouse DC having been induced to
These results demonstrate the ability of esDC to express the enhanced green fluorescent protein
process an endogenously expressed mH antigen and (EGFP) reporter gene [28]. The use of mRNA for
present the immunodominant epitopes in the context the purpose of transfection has also proven surpris-
of MHC class II determinants, a necessary require- ingly effective [29], although expression of the
ment for the induction of transplantation tolerance. transcribed gene is necessarily transient and has yet
to exceed a transduction efficiency of 20%. Both of
3.4. Genetic modification of esDC these approaches, while adequate to secure priming to
TAA, are far too inefficient as a strategy for
The advantages that routine genetic modification of modifying cellular phenotype and function. Indeed,
DC would offer to strategies of immune intervention the only approach to genetic modification that comes
extend far beyond the expression of appropriate TAA. close to such a remit is the use of viral vectors, such as
The ability to overexpress or functionally silence adenovirus, through which the stable introduction of
genes involved in immune regulation would provide transgenes may be achieved in up to 80% of cells.
Fig. 3. Presentation of the endogenously expressed male mH antigen, Dby, by mature esDC. (A) Primary proliferative responses of naRve,
female T cells from A1xRAG1 / mice, transgenically expressing a TCR specific for the 479–493 epitope of Dby in the context of H2Ek. T
.
cells were cultured with mature esDC differentiated from either karyotypically male ( ) or female (E) ES cells. (B) Stimulation of T cell
proliferation by both male and female esDC pulsed with 5 nM Dby479–493 peptide.
P.J. Fairchild et al. / International Immunopharmacology 5 (2005) 13–21 19
Fig. 4. Comparison of bmDC (A) and esDC (B) genetically modified to express the EGFP reporter gene (filled histograms) under control of
the EF1a promoter. Open histograms show levels of autofluorescence among control, nontransfected cells. Lipofection was used in both
cases, either to modify terminally differentiated bmDC directly (A) or esDC indirectly by modification and differentiation of a standard ES
cell line (B).
Nevertheless, despite the opportunities that adenoviral deficient in both alleles of a target gene through
vectors may provide, their use exacts a heavy price by homologous recombination. Such an approach has
provoking the premature maturation of DC, thereby already been successfully pursued for the production
sabotaging aspects of their normal physiology [30]. of lines of esDC deficient in expression of Notch 1
Furthermore, the introduction of transgenes and their [34], an achievement which is far beyond the scope of
subsequent expression is accompanied by the syn- traditional strategies for genetic modification of
thesis of viral proteins whose presentation may elicit terminally differentiated DC.
vigorous CTL responses, rendering the DC them-
selves vulnerable to depletion [29,31,32].
Whereas esDC share with conventional sources an 4. Discussion
intrinsic resilience to genetic modification, we have
shown that ES cells may serve as an efficient conduit ES cells display a number of properties that make
through which genes may be introduced into DC them unique, their pluripotency in vitro, capacity for
without corrupting their function [33]. Since ES cells self-renewal and tractability for genetic modification
are comparatively amenable to genetic manipulation readily distinguishing them from fully differentiated
and their capacity for self-renewal permits cloning at cell types. By exploiting ES cells as a novel source of
the single-cell level, lines may be produced that DC, these features may be harnessed in addition to
express a desired mutant phenotype which is faithfully those of the DC themselves, to enhance their
transmitted to esDC, downstream of the differentia- suitability as candidates for immunotherapy. Accord-
tion pathway [14,33]. Furthermore, the capacity for ingly, esDC may be obtained in unparalleled abun-
cloning and expansion of individual cells as a means dance and with a stability of phenotype that eludes
of rescuing rare, mutant phenotypes helps counteract DC differentiated from conventional sources, such as
the low transduction efficiency of lipid-based bone marrow progenitors or peripheral blood mono-
approaches to transfection, thereby avoiding the need cytes. Furthermore, the potential for genetic modifi-
for viral vectors and their far-reaching influence on cation of the parental ES cell line provides
DC function. Fig. 4 illustrates the efficiency with opportunities for the rational design of esDC,
which esDC may be genetically modified through uniformly expressing a desired mutant phenotype,
lipofection of the parental ES cell line (Fig. 4B), aimed at directing the outcome of the immune
compared to terminally differentiated bmDC (Fig. response upon readministration in vivo. The capacity
4A). In addition to paving the way for the rational to obtain such lines without recourse to the use of
design of esDC uniformly expressing a desired mutant viral vectors offers a significant advantage over
phenotype, the properties of ES cells may be further current practice, since the corrupting influence of
exploited for the production of lines functionally viral infection on the physiology of DC is avoided, as
20 P.J. Fairchild et al. / International Immunopharmacology 5 (2005) 13–21
antigen gene are therapeutically effective against established [28] Irvine AS, Trinder PKE, Laughton DL, Ketteringham H,
pulmonary metastases. J Exp Med 1997;186:1213 – 21. McDermott RH, Reid SC, et al. Efficient nonviral transfection
[19] Nestle FO, Alijagic S, Gilliet M, Sun Y, Grabbe S, Drummer R, of dendritic cells and their use for in vivo immunization. Nat
et al. Vaccination of melanoma patients with peptide- or tumor Biotechnol 2000;18:1273 – 8.
lysate-pulsed dendritic cells. Nat Med 1998;4:328 – 32. [29] Mitchell DA, Nair SK. RNA-transfected dendritic cells in
[20] Thurner B, Haendle I, Roder C, Dieckmann D, Keikavoussi P, cancer immunotherapy. J Clin Invest 2000;106:1065 – 9.
Jonuleit N, et al. Vaccination with Mage-3A1 peptide-pulsed [30] Morelli AE, Larregina AT, Ganster RW, Zahorchak AF,
mature, monocyte-derived dendritic cells expands specific Plowey JM, Takayama T, et al. Recombinant adenovirus
cytotoxic T cells and induces regression of some metastases in induces maturation of dendritic cells via an NF-nB-dependent
advanced stage IV melanoma. J Exp Med 1999;190:1669 – 78. pathway. J Virol 2000;74:9617 – 28.
[21] Nestle FO, Banchereau J, Hart D. Dendritic cells: on the move [31] Yang Y, Li Q, Ertl HCJ, Wilson JM. Cellular and humoral
from bench to bedside. Nat Med 2001;7:761 – 5. immune responses to viral antigens create barriers to lung-
[22] Cerundolo V, Hermans IF, Salio M. Dendritic cells: a journey directed gene therapy with recombinant adenoviruses. J Virol
from the laboratory to the clinic. Nat Immunol 2004;5:7 – 10. 1995;69:2004 – 15.
[23] Somia N, Verma IM. Gene therapy: trials and tribulations. Nat [32] Roth MD, Cheng Q, Harui A, Basak SK, Mitani K, Low TA,
Rev, Genet 2000;1:91 – 9. et al. Helper-dependent adenoviral vectors efficiently express
[24] Dhodapkar MW, Steinman RM, Krasovsky J, Munz C, transgenes in human dendritic cells but still stimulate antiviral
Bhardwaj N. Antigen-specific inhibition of effector T cell immune responses. J Immunol 2002;169:4651 – 6.
function in humans after injection of immature dendritic cells. [33] Fairchild PJ, Nolan KF, Cartland S, Graça L, Waldmann H.
J Exp Med 2001;193:233 – 8. Stable lines of genetically-modified dendritic cells from mouse
[25] Jonuleit H, Schmitt E, Schuler G, Knop J, Enk AH. Induction embryonic stem cells. Transplantation 2003;76:606 – 8.
of interleukin 10-producing, nonproliferating CD4+ T cells [34] Cheng P, Nefedova Y, Miele L, Osborne BA, Gabrilovich D.
with regulatory properties by repetitive stimulation with Notch signaling is necessary but not sufficient for differ-
allogeneic immature human dendritic cells. J Exp Med 2000; entiation of dendritic cells. Blood 2003;102:3980 – 8.
192:1213 – 22. [35] Matsuyoshi H, Senju S, Hirata S, Yoshitake Y, Uemura Y,
[26] Hawiger D, Inaba K, Drosett Y, Guo M, Mahnke K, Rivera M, Nishimura Y, et al. Enhanced priming of antigen-specific
et al. Dendritic cells induce peripheral T cell unresponsiveness CTLs in vivo by embryonic stem cell-derived dendritic
under steady state conditions in vivo. J Exp Med 2001;194: cells expressing chemokine along with antigenic protein:
769 – 79. application to antitumor vaccination. J Immunol 2004;172:
[27] Ludewig B, Ochsenbein AF, Odermatt B, Paulin D, 776 – 86.
Hengartner H, Zinkernagel RM, et al. Immunotherapy with [36] Fairchild PJ, Cartland S, Nolan KF, Waldmann H. Embryonic
dendritic cells directed against tumor antigens shared with stem cells and the challenge of transplantation tolerance.
normal host cells results in severe autoimmune disease. J Exp Trends Immunol 2004;25:465 – 70.
Med 2000;191:795 – 803.