A vaccine directed to B cells and produced by
cell-free protein synthesis generates potent
Patrick P. Nga, Ming Jiaa, Kedar G. Patelb, Joshua D. Brodya,1, James R. Swartzb,c, Shoshana Levya, and Ronald Levya,2
aDivision of Oncology, Department of Medicine, Stanford University Medical Center, and Departments ofbChemical Engineering andcBioengineering,
Stanford University, Stanford, CA 94305
Contributed by Ronald Levy, July 10, 2012 (sent for review May 26, 2012)
Clinical studies of idiotype (Id) vaccination in patients with
lymphoma have established a correlation between the induced
anti-Id antibody responses and favorable clinical outcomes. To
streamline the production of an Id vaccine, we engineered a small
diabody (Db) molecule containing both a B-cell–targeting moiety
(anti-CD19) and a lymphoma Id. This molecule (αCD19-Id) was
designed to penetrate lymph nodes and bind to noncognate B cells
to form an antigen presentation array. Indeed, the αCD19-Id mol-
ecule accumulated on B cells in vivo after s.c. administration. These
noncognate B cells, decorated with the diabody, could then stim-
ulate the more rare Id-specific B cells. Peptide epitopes present in
the diabody linker augmented the response by activating CD4+
helper T cells. Consequently, the αCD19-Id molecule induced a ro-
bust Id-specific antibody response and protected animals from tu-
mor challenge. Such diabodies are produced in a cell-free protein
expression system within hours of amplification of the specific Ig
genes from the B-cell tumor. This customized product can now be
available to vaccinate patients before they receive other, poten-
tially immunosuppressive, therapies.
immunotherapy|tumor-specific antigen|bispecific antibody fragments
monoclonal antibodies (mAbs) against this target are effective in
therapy (1). Furthermore, studies of Id vaccination had sug-
gested a correlation between induced anti-Id antibody responses
and progression-free survival and overall survival of patients (2–
4). Despite these encouraging results, phase III trials have not
established a clinical benefit from Id vaccination, except for
a possible subset of patients who have prolonged remissions after
initial chemotherapy (5–7). One possible problem may have been
the chemical conjugation of Id to the carrier protein, keyhole
limpet hemocyanin (KLH). Antigenic determinants on the Id
could have been damaged in this process (8). Recombinant vac-
cines that do not require chemical conjugation may lead to im-
proved immunogenicity and clinical outcomes.
Recent studies on antigen (Ag) acquisition by B cells have
provided new insights for vaccine design. The majority of B cells
reside in follicles within secondary lymphoid organs. Foreign Ags
in the form of immune complexes are transported into lymph
node follicles by subcapsular sinus macrophages (9–11), and into
spleen follicles by marginal zone B cells (12). In the follicles,
nonspecific B cells retain immune complexes on their cell sur-
faces. Some complexes are transferred to follicular dendritic cells
(9–11), whereas others may directly cross-link the Ag-specific
receptors (BCRs) on cognate B cells (10, 11). These roles played
by noncognate B cells in the generation of specific antibody
responses were previously not appreciated. In addition to forming
immune complexes that facilitate entering the follicles and pre-
senting on the cell surface, foreign Ags may also be endocytosed,
processed, and presented as peptides that activate CD4+T cells,
which in turn, provide costimulation to cognate B cells. These
attributes argue for the use of foreign carrier proteins such as
diotype (Id), the unique Ig molecule of each lymphoma tumor,
is a good target for the immune system. Passively administered
KLH to help stimulate antibody responses against self-Ags that
do not form immune complexes. However, chemical conjugation
has been shown to reduce vaccine potency (8). Recombinant Id
vaccines may offer distinct advantages because they can be pro-
duced with built-in carrier moieties.
It was recently discovered that small molecules (<70 kDa) can
enter follicles more efficiently through specialized conduits (13,
14). We therefore designed a recombinant vaccine below this
size limit. To provide cell surface anchorage for Ag retention and
presentation, we delivered the vaccine to noncognate B cells
within the follicle by targeting to CD19, a B-cell–specific mole-
cule (15). We created a bispecific diabody (Db), containing the
variable regions of a rat anti-mouse CD19 mAb and those of the
38C13 mouse B-cell lymphoma Id [αCD19-Id, molecular weight:
52 kDa]. We envisioned that the αCD19-Id would form a “lawn”
of Ags on the surface of follicular B cells, where they could cross-
link the BCR of the rare Ag-specific B cell among them. Fur-
thermore, coligation of the BCR with CD19 could result in
synergistic activation of the specific B cells (15). Nonsyngeneic
sequences, such as the rat variable regions in the Db, might help
by activating CD4+T cells (Fig. 1).
We used an in vitro (cell-free) protein synthesis (CFPS) system
for mammalian proteins that can assemble intrachain disulfide
bonds (16, 17). The reaction contains the DNA template for
each polypeptide chain, an energy source, substrates, and cellular
machinery from Escherichia coli that can carry out both tran-
scription and translation. A small reaction can produce protein
sufficient for vaccination in a matter of hours, as opposed to the
usual methods of mammalian cell protein production that take
several weeks. We produced and screened several structural
variants of αCD19-Id. The most active form was then used for in
Diabody Design, Production, and Initial Characterizations. αCD19-Id
is a heterodimer of noncovalently associated polypeptides con-
taining the variable regions of 38C13 and anti-CD19, separated
by Gly4Ser linkers (Fig. 2A). We produced four different αCD19-
Ids with the respective variable domains in different orientations
(Fig. 2A and Fig. S1). The only polypeptides that incorporate
a radiolabeled amino acid are those encoded by the supplied
templates. This labeling allows quantification and SDS/PAGE
autoradiography without purification, thus expediting screening
Author contributions: P.P.N., J.R.S., S.L., and R.L. designed research; P.P.N., M.J., K.G.P.,
and J.D.B. performed research; P.P.N., J.R.S., S.L., and R.L. analyzed data; and P.P.N., J.R.S.,
S.L., and R.L. wrote the paper.
The authors declare no conflict of interest.
1Present address: Division of Hematology and Medical Oncology, Department of Medi-
cine, Mount Sinai School of Medicine, New York, NY 10029.
2To whom correspondence should be addressed. E-mail: email@example.com.
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.
| September 4, 2012
| vol. 109
| no. 36www.pnas.org/cgi/doi/10.1073/pnas.1211018109
of various constructs. The “open” feature of CFPS also allowed
us to adjust the relative amounts of the two template plasmids to
ensure a 1:1 chain ratio in each Db heterodimer. The Db pro-
teins were screened by flow cytometry for appropriate binding
activities (Fig. 2B). Bispecific binding was determined using
a target cell (A20) that expresses surface CD19 and a detector
consisting of an anti-38C13 Id mAb. As a negative control cell we
used a subclone of the A20 cell line that had lost cell surface
expression of CD19 (A20/CD19NEG) (18). Among these four
products, αCD19-Id-1 showed the best bispecific binding activity
(Fig. 2B). We made a negative control Db that had the variable
regions of a rat mAb of irrelevant specificity (RatFv-Id) (Fig.
S1). Both Dbs were confirmed to be single species heterodimers
by SDS/PAGE and size-exclusion (SE)-HPLC (Fig. S2 A and B),
and were used for subsequent studies.
αCD19-Id Localized to B Cells in Vivo. We injected mice in-
tradermally (i.d.) with fluorophore-labeled Dbs and analyzed
cells from the draining lymph nodes by flow cytometry. αCD19-
Id, but not RatFv-Id, was retained specifically on B cells (B220+
population) but not on T cells (CD3+population) (Fig. 3A), and
this occurred as early as 2.5 h after injection (Fig. S3A). This
rapid accumulation is similar to a report of a small Ag (turkey
egg lysozyme, 14 kDa) that traveled through conduits into fol-
licles (14). The efficiency of B-cell targeting was even more ap-
parent when cells from the spleen and blood were analyzed 2 h
after i.v. injection of αCD19-Id. Again, we found binding of the
specific Db to B cells, but not to T cells (CD3+), monocytes,
macrophages, or granulocytes (CD11b+and F4/80+) (Fig. 3B
and Fig. S3 B and C).
Id-Specific BCR Activation by αCD19-Id–Decorated B Cells. For this
test we constructed an Id-specific B cell (A20/α38BCR) by
transfecting the A20 cell line to express a membrane-anchored
form of the anti-Id antibody (Fig. S4). We demonstrated that
splenic B cells recovered from animals injected with αCD19-Id
(Fig. 4A), or A20 cells decorated with αCD19-Id in vitro (Fig.
S5A), could trigger the phosphorylation of intracellular BCR
pathway signaling molecules in Id-specific B cells. These signals
peaked at 20 min and declined gradually over 30–60 min after
the stimulator and responder cells came in contact (Fig. 4 A and
B). This stimulation did not occur in the negative control cell
line, native A20, that lacked the specific anti-Id BCR. We also
found that A20 cells decorated with αCD19-Id induced a stron-
ger activation signal than that induced by an equal amount of
free αCD19-Id, and reached a level to that induced by the
pentameric 38C13 IgM protein (Fig. S5B). Another way αCD19-
Id could stimulate an Id-specific B cell is by cross-linking its BCR
to its CD19 surface molecule. In fact, αCD19-Id induced phos-
phorylation of phosphatidylinositol 3-kinase (PI3K), a signaling
molecule directly downstream of CD19 (15), as well as the ex-
tracellular signal-regulated protein kinase (ERK) (Fig. 4C).
Neither the negative control RatFv-Id (Fig. 4C) nor an anti-
CD19 mAb induced such phosphorylation.
Id-Specific B Cells Captured αCD19-Id from Db-Decorated B Cells and
Internalized the Vaccine Molecule. Ag-specific B cells need to in-
ternalize their cognate Ag for processing and presentation
to receive CD4+T-cell help. Splenic B cells from mice injected
with fluorophore-labeled αCD19-Id (Fig. 5A), or A20 cells dec-
orated with fluorophore-labeled αCD19-Id in vitro (Fig. S6),
could transfer the Db to A20/α38BCR cells, but not to A20
cells lacking the specific BCR. We also confirmed that αCD19-Id
was internalized by A20/α38BCR cells using confocal micros-
copy (Fig. 5B).
αCD19-Id Induced Both an Id-Specific Antibody Response and a Db-
Specific T-Cell Response. αCD19-Id induced a robust Id-specific
IgG response, comparable to that induced by 38C13-KLH. By
contrast, immunization with RatFv-Id, αCD19 + 38C13, or
αCD19-Av-38C13 failed to induce a significant response (Fig.
6A). Whereas both groups of antibody responding mice made
predominantly anti-Id IgG1, αCD19-Id induced a slightly higher
percentage of anti-Id IgG2 than that induced by 38C13-KLH
(33.8 ± 6.2% vs. 22.4 ± 2.8%, as mean ± SEM).
The anti-Id antibody response induced by 38C13-KLH requires
CD4+T cells (8). That was also the case for αCD19-Id. Depletion
of CD4+T cells from animals before vaccination with the Db
dramatically reduced the anti-Id IgG responses (from 77, 50, and
20 μg/mL to 6, 0, and 0 μg/mL serum). Lymphocytes from animals
B cell receptor
T cell receptor
geted to CD19 on B cells in lymphoid follicles.
Proposed model: Id-specific B cells are stimulated by αCD19-Id tar-
plasmids and schematic of one of four αCD19-Ids. Coexpression of both
plasmids in the same CFPS reaction produces two polypeptides that assemble
into a noncovalent heterodimeric Db. Locations of heavy chain variable
domains (α19 VHand 38 VH), and light chain variable domains (α19 VLand
38 VL) of anti-CD19 and 38C13, respectively, T7 promoters (T7), ribosomal
binding sites (rbs), (Gly)4Ser linkers (L), hexahistidine tag (H6), and stop
codons (stop) on the expression plasmids are indicated, as are the 38C13 Id
and the binding site for CD19 on the Db. (B) Flow cytometry analysis of
αCD19-Id bispecific binding. A20 cells were incubated with CFPS products
containing 5 μg of each αCD19-Id variant (—) or with mock CFPS product
(shaded). A20/CD19NEGcells incubated with the same CFPS product (---)
served as a negative target cell control. Cells were then washed and stained
with Alexa Fluor 488-conjugated anti-38C13 mAb.
Design and characterization of αCD19-Id. (A) Design of expression
Ng et al.PNAS
| September 4, 2012
| vol. 109
| no. 36
vaccinated with RatFv-Id and αCD19-Id proliferated (Fig. S7) and
secreted gamma IFN (IFN-γ) (Fig. 6B) to both the specific and
nonspecific Dbs. There was no response to anti-CD19 mAb or to
rat IgG (Fig. 6B). These results indicate that it was a component
other than the Ig variable regions (i.e., the linker) shared by both
Dbs, that provided T-cell responses and help to Id-specific B cells.
Indeed, peptides most likely to bind the major histocompatibility
complex II (MHCII) expressed by C3H/HeN mice (http://imed.
med.ucm.es/Tools/rankpep.html) fall within the V-domain linker
junction (FDYWGQGTTLTVSSGGGGSDIVMTQS) shared by
both Dbs. It is suprising that animals immunized with the vaccine
containing a complex with anti-CD19 and avidin did not have
activated T cells specific for these xenogeneic Ags. However, the
lack of such helper T cells may explain the poor anti-Id antibody
response induced by this complex.
Vaccination with αCD19-Id Protected Mice from a Systemic Tumor
Challenge. Vaccinated mice were challenged with lethal doses
of the aggressive 38C13 lymphoma. Mice vaccinated with RatFv-
Id showed no protection compared with unvaccinated mice. In
contrast, mice that received αCD19-Id were protected to a simi-
lar degree as those vaccinated with 38C13-KLH (Fig. 7). Tumors
from animals vaccinated with αCD19-Id or 38C13-KLH still
bound to immune sera generated by these vaccines. Therefore,
the lack of protection for these animals could not be explained
by the expansion of tumor cells expressing Id variants.
Patients with follicular lymphoma can be induced to make anti-
Id antibodies against their tumors. Those who make such a re-
sponse have improved overall survival compared with those who
do not (2, 3). However, randomized controlled trials have failed
to prove a clinical benefit from Id vaccination (5–7). An expla-
nation for this discrepancy may be that the ability to make anti-Id
antibody is simply an indicator for which the patient is destined
to survive longer. An alternative explanation is that anti-Id
antibodies are protective against tumor growth, but only if the
response is robust. In one phase III trial, all of the patients
produced antibodies against the KLH carrier protein, indicating
a certain level of general immune competence, but more than
half of them failed to generate anti-Id antibodies (5). One pos-
sible problem may have been the chemical conjugation to KLH,
a process that is difficult to control, especially by the glutaral-
dehyde method that was used. It has been established that glu-
taraldehyde can damage antigenic determinants of an Id and
abrogate tumor protection in that animal model (8). For patients
with αCD19-Id, RatFv-Id, or 38C13 IgM, each conjugated to Alexa Fluor
488, or with buffer. (A) Draining lymph nodes (LN) were harvested 8 h after
i.d. injections. (B) Spleens (SP) and peripheral blood (PB) were harvested
2 h after i.v. injections. Leukocytes from these organs were stained with
fluorophore-conjugated mAbs specific for B220, CD3, CD11b, and F4/80,
and analyzed by flow cytometry. The percentages of gated total leuko-
cytes in the Upper Right quadrants are indicated. One of two experiments
αCD19-Id targeted specifically to B cells in vivo. Mice were injected
cells. (A) Splenic B cells recovered from mice 2 h after i.v. injection with
αCD19-Id (B-Db) or with PBS (B) were incubated for the indicated times with
responder cells, either A20 or A20/α38BCR that were prelabeled with Cell-
Trace Violet dye. Cells were fixed, permeabilized, stained with PE-conju-
gated antibodies specific for the phosphorylated forms of PLC-γ2 and Syk,
and analyzed by flow cytometry. Responses of gated responder cells are
shown. (B) Kinetics of BCR signaling induced by αCD19-Id–decorated A20
cells (A20-Db). The percentages of A20/α38BCR responder (●) and A20
negative control responder (▲) cells are shown. Data are pooled from three
experiments (an example is shown in Fig. S5A). The percentage of BCR sig-
naling cells for each incubation time was calculated from the corresponding
histograms: [% PE+cells in response to A20-Db (red line)] − [% PE+cells
in response to A20 (black line)]. (C) A20/α38BCR cells were stimulated for
10 min at 37 °C with Dbs, 38C13, or control IgM. Cell lysates were analyzed
by Western blotting using antibodies specific for the phosphorylated forms
of ERK and the p55 subunit of PI3K, and for total ERK and actin. Repre-
sentative results of three experiments are shown.
Id-specific BCR activation by αCD19-Id and αCD19-Id–decorated B
| www.pnas.org/cgi/doi/10.1073/pnas.1211018109Ng et al.
where each Id is unique, the conjugation chemistry may affect
each product to a different degree. Therefore, new vaccines that
do not require chemical conjugation may lead to improved im-
munogenicity and clinical outcomes. To achieve this goal, we
and others have tested various forms of recombinant Id vaccines
(17, 19–23). A common approach is to produce fusions of Id
sequences to targeting moieties that direct the construct to cy-
tokine receptors or to other activating receptors on dendritic
cells, macrophages, and other antigen-presenting cells (APCs)
(19, 22, 23). The peptides derived from Id proteins would then
be presented to T cells (24, 25).
Herein, we report an alternative strategy designed to activate
Id-specific B cells. This approach targets Id to the surface of non-
cognate B cells where they can be presented as intact molecules
to cognate B cells. Vaccines targeted to the complement receptor
2 (CD21) expressedon a variety ofimmunecells,including B cells,
have been constructed by several groups. Some showed enhance-
ment of Ag-specific immunity (26, 27), whereas others reported
unexpected suppression of antibody responses (28, 29). We chose
to target Id to the CD19 molecule expressed exclusively on B
cells. There is no competing ligand for CD19 as there is for CD21.
Importantly, it is known that the majority of CD19-antibody com-
plexes remain on or recycle to the surface of B cells even after
extended periods (30, 31). Furthermore, coligation of CD19 tothe
BCR lowers the activation threshold of B cells (15).
Syngeneic Ig are poor immunogens. However, Id mixed with
complete Freud’s adjuvant can generate anti-Id antibody to
protective levels in several tumor models (32, 33). Interestingly,
no anti-Id antibody can be induced this way in the 38C13 model
(34). The unusually poor immunogenicity of this Id may be due
to the lack of somatic mutation in its VHand VLgenes (35),
resulting in a paucity of CD4+T-cell epitopes. The 38C13 Id can
be made immunogenic by coupling it to KLH. KLH binds to
natural antibodies and complement (36), and has been shown to
be transported into the follicles of lymph nodes (14). KLH also
contains peptide epitopes that activate CD4+T cells. These
properties make KLH an effective carrier.
We show that a robust anti-Id antibody response can also be
induced by fusing 38C13 Id to anti-CD19. Although different
from Id-KLH, αCD19-Id may achieve similar immune-stimula-
tory functions by alternative strategies. Being small, Db can enter
follicles through conduits, as inferred from the speed that
αCD19-Id reached B cells in the lymph node (Fig. S3A). Similar
conduit systems have been found to channel small molecules into
the T-cell areas of a lymph node, and into the white pulp of the
spleen (37). αCD19-Id may also be actively transported into
spleen follicles by marginal zone B cells expressing CD19.
αCD19-Id anchored to CD19 on abundant noncognate B cells
provided cross-linking of BCRs on the cognate B cell. Indeed,
αCD19-Id bound to B cells induced a stronger BCR signal than
free αCD19-Id, and reached the level induced by the pentameric
38C13 IgM (Fig. 4 A and B and Fig. S5).
CD4+T cells were required for the anti-Id response gener-
ated by αCD19-Id. The rat variable regions of anti-CD19 might
have been expected to be the source of CD4+T-cell epitopes.
However, instead, our data indicate that the nonnatural
Gly4Ser linker provided such epitopes (Fig. 6B and Fig. S7).
The potential to generate immune-stimulatory epitopes is an-
other advantage of recombinant Id vaccines over native Ig Id
vaccines, in addition to avoiding the regulatory T-cell epitopes
found on Ig constant regions (38). Ding et al. reported that B
cells targeted by an anti–CD19-Ag conjugate could prime
CD4+T cells (39). We have no evidence for this because the
nontargeting RatFv-Id was as effective as αCD19-Id in acti-
vating T cells. It is likely that some molecules of both Dbs were
internalized and presented to T cells by macrophages or den-
dritic cells. However, in addition, some αCD19-Id targeted to
noncognate B cells where they formed an array to present the Id
to cognate B cells. By contrast, the nontargeting RatFv-Id in-
duced no anti-Id antibody response, nor did the 38C13 IgM,
a good cross-linker of Id-specific BCR but lacking T-cell epitopes.
Together, these results underscore the importance of vaccines such
as αCD19-Id that are designed to activate both cognate B cells and
Rituximab is now a part of the standard therapy for follicular
lymphoma, therefore, therapeutic vaccine strategies for lym-
phoma will need to be used in conjunction with this mAb that
depletes normal B cells. Rituximab can blunt antibody responses
to new Ags but it does not ablate an existing response once it is
established by prior vaccination (40, 41). Id vaccines produced
rapidly by cell-free protein synthesis, as tested here, can be
available before rituximab is used. This strategy may have the
additional benefit of delaying the use of rituximab, and there-
fore, the development of rituximab resistance.
Materials and Methods
Plasmids. To construct expression plasmids for Dbs, RNAs were extracted from
hybridomas producing the anti-CD19 rat IgG2a/κ (1D3) (18) and a rat IgG2a/κ
of irrelevant specificity (H22-15-5) (RNeasy; Qiagen). The VH and VL
sequences were isolated using the SMART RACE kit (Clontech) and primers
specific to rat IgG2a constant region 1 (5′-ggaaatagcccttgaccaggcatcc-3′) and
injected i.v. with PBS (B) or with αCD19-Id conjugated to Alexa Fluor 488
(B-Db). Splenic B cells recovered after 2 h were incubated for 1 h at 37 °C
with A20 or A20/α38BCR cells prelabeled with Violet dye, then fixed and
analyzed by flow cytometry. One of two experiments is presented. (B)
A20/α38BCR cells were incubated at 0 °C or 37 °C for 30 min with Alexa Fluor
488-conjugated αCD19-Id. Cells were washed, fixed, and analyzed by con-
focal microscopy. Representative images of cells are shown with a z-section
thickness of 2.4 μm. (Scale bar, 10 μm.)
Id-specific B cells captured and internalized αCD19-Id. (A) Mice were
Ng et al. PNAS
| September 4, 2012
| vol. 109
| no. 36
κ constant region (5′-gactgaggcacctccagttgctaactg-3′). These sequences and
those of the 38C13 cells (35) were codon optimized for expression in E. coli
with the online resource, DNAworks. The pY71 expression vector (42) con-
tains T7 promoter and termination sequences. The coding region, flanked by
the 5′ NdeI and 3′ SalI sites, contains two V sequences separated by a linker.
An analysis of potential secondary structures in the upstream 58 nucleotides
and the codons of the first nine amino acids was performed using the online
resource, Mfold. Silent codon changes were made to eliminate G:C pairings
that stabilize secondary structures, which may impede translation. Over-
lapping oligonucleotides of the coding regions were designed (DNAworks),
purchased (IDT), assembled by PCR, and cloned into pY71. The plasmid
expressing a membrane-bound anti-38C13 IgM, created for the present
work, has been described (42).
Flow Cytometry. Db in vitro binding assay. Cells (106) were incubated with CFPS
products for 1 h on ice, then with 1 μg Alexa Fluor 488-conjugated S1C5 for
30 min. Cells were washed after each incubation, fixed with 2% (wt/vol)
paraformaldehyde, and analyzed on a FACScalibur (Becton Dickinson).
Db in vivo trafficking studies. Mice were injected i.d. on the abdomen or i.v.
in the tail with 10 μg AlexaFluor 488 conjugates. Cells isolated from
the indicated body compartments were incubated with Fc blocker, stained
with fluorophore-conjugated mAbs, washed, fixed, and analyzed as
described above. The animal study protocol was approved by the Stanford
University Institutional Animal Care and Use Committee.
Db transfer studies. A20 or A20/α38BCR cells at 106cells/mL in PBS were in-
cubated with 5 μM CellTrace Violetdye (Invitrogen) for 20min at 37 °C, washed,
and cultured overnight in full media before use. Splenocytes were harvested
from mice injected i.v. with buffer or 20 μg Alexa Fluor-488 conjugated αCD19-
Id. A total of 3 × 106splenic B cells were mixed with 3 × 105Violet dye-labeled
cells, centrifuged for 30 s, and incubated for 1 h at 37 °C. Cells were washed,
fixed, and analyzed on an LSR II cytometer (Becton Dickinson).
Signal transduction assays. Splenic B cells and Violet dye-labeled cells were
mixed, incubated at 37 °C for 15 s before adding 3.3 mM hydrogen peroxide.
The cells were immediately vortexed, centrifuged for 30 s, and incubated
for the indicated time. Cells were washed with cold PBS, fixed for 30 min
in BD CytoFix/CytoPerm solution, washed with BD Perm/Wash buffer, and
incubated for 30 min with PE-conjugated mAbs. After washes, cells were
fixed and analyzed.
ELISAs. Serum anti-Id IgGs were quantified as described previously (17).
To quantify IFN-γ, splenocytes were seeded (5 × 105cells per well) in
96-well U-bottom plates in 100 μL media (5% FBS, 100 μg/mL gentamy-
cin). A final concentration of 50 μg/mL of 38C13 IgM, anti-CD19, re-
spective isotype control antibodies, 10 μg/mL of KLH, avidin, or 2 μg/mL
of Dbs was added. Culture supernatants were tested with an IFN-γ ELISA
kit (Thermo Scientific).
ACKNOWLEDGMENTS. We thank D. Czerwinski and R. Rajapaksa for
technical expertise in flow cytometry; A. Virrueta for technical assistance;
C.-C. Kuo for advice on molecular biology; and R. Houot and H. Kohrt for
producing the anti-CD4 mAb. This work was supported by a Leukemia and
αCD19 + 38C13
αCD19 + 38C13
on consecutive days. Vaccines consisted either of 6 μg Dbs or an Id molar equivalent of 38C13 IgM, either chemically conjugated to KLH (38C13-KLH), mixed
with anti-CD19 mAb (αCD19 + 38C13), or conjugated to anti-CD19 mAb by avidin (αCD19-Av-38C13). Sera collected a week after the last immunization were
tested by ELISA for antibodies against the 38C13 Id. Data were combined from four studies. The number of animals in each group is indicated. Each bar on the
graph represents the mean serum anti-Id IgG concentration ± SEM of each group. None of the sera reacted with a mouse IgM/κ isotype control. (B) IFN-γ
production by splenocytes from immunized mice. Mice (two to three per group) were vaccinated as in A. Spleens from each group were harvested and pooled
a week later. Splenocytes were cultured for 4 d with Ags listed in the legend in hexaplicate wells each. IFN-γ in culture supernatants was measured by ELISA.
Immune responses to αCD19-Id. (A) αCD19-Id induced a robust Id-specific antibody response. Mice received four biweekly i.d. vaccinations given twice
n. s. (P = 0.70)
0 2040 6080100
Days since tumor challenge
n. s. (P = 0.27)
P = 0.015
n. s. (P = 0.91)
0 2040 60 80100
Days since tumor challenge
per group) received αCD19-Id, RatFv-Id, 38C13-KLH, or buffer as described in
Fig. 6A. Ten days later, mice were challenged with 100 38C13 cells by i.v.
injection. (B) Mice (10 per group) were vaccinated with αCD19-Id, 38C13-
KLH, or buffer and challenged with 400 cells. Survival was analyzed by the
Kaplan–Meier method and the log-rank statistical test.
Vaccination with αCD19-Id protected mice from tumor. (A) Mice (10
| www.pnas.org/cgi/doi/10.1073/pnas.1211018109Ng et al.
Lymphoma Society Specialized Center of Research (SCOR) program grant,
Ruth L. Kirschstein Grant 5 T32 AI07290 (to P.P.N.), and a Lymphoma
Research Foundation fellowship (to P.P.N.). R.L. is an American Cancer Soci-
ety clinical research professor.
1. Davis TA, Maloney DG, Czerwinski DK, Liles TM, Levy R (1998) Anti-idiotype antibodies
can induce long-term complete remissions in non-Hodgkin’s lymphoma without
eradicating the malignant clone. Blood 92:1184–1190.
2. Weng WK, Czerwinski D, Timmerman J, Hsu FJ, Levy R (2004) Clinical outcome of
lymphoma patients after idiotype vaccination is correlated with humoral immune
response and immunoglobulin G Fc receptor genotype. J Clin Oncol 22:4717–4724.
3. Ai WZ, Tibshirani R, Taidi B, Czerwinski D, Levy R (2009) Anti-idiotype antibody re-
sponse after vaccination correlates with better overall survival in follicular lymphoma.
4. Navarrete MA, et al. (2011) Upfront immunization with autologous recombinant
idiotype Fab fragment without prior cytoreduction in indolent B-cell lymphoma.
5. Levy R, Robertson M, Leonard J, Vose J, Danney D (2008) Results of a phase 3 trial
evaluating safety and efficacy of specific immunotherapy, recombinant idiotype (Id)
conjugated to KLH (Id-KLH) with GM-CSF, compared to non-specific immunotherapy,
KLH with GM-CSF, in patients with follicular non Hodgkin’s lymphoma (FNHL). Ann
Oncol 19(suppl 4):101–102.
6. Freedman A, et al. (2009) Placebo-controlled phase III trial of patient-specific immu-
notherapy with mitumprotimut-T and granulocyte-macrophage colony-stimulating
factor after rituximab in patients with follicular lymphoma. J Clin Oncol 27:
7. Schuster SJ, et al. (2011) Vaccination with patient-specific tumor-derived antigen in
first remission improves disease-free survival in follicular lymphoma. J Clin Oncol 29:
8. Betting DJ, Kafi K, Abdollahi-Fard A, Hurvitz SA, Timmerman JM (2008) Sulfhydryl-
based tumor antigen-carrier protein conjugates stimulate superior antitumor immu-
nity against B cell lymphomas. J Immunol 181:4131–4140.
9. Carrasco YR, Batista FD (2007) B cells acquire particulate antigen in a macrophage-rich
area at the boundary between the follicle and the subcapsular sinus of the lymph
node. Immunity 27:160–171.
10. Phan TG, Grigorova I, Okada T, Cyster JG (2007) Subcapsular encounter and com-
plement-dependent transport of immune complexes by lymph node B cells. Nat Im-
11. Phan TG, Green JA, Gray EE, Xu Y, Cyster JG (2009) Immune complex relay by sub-
capsular sinus macrophages and noncognate B cells drives antibody affinity matu-
ration. Nat Immunol 10:786–793.
12. Ferguson AR, Youd ME, Corley RB (2004) Marginal zone B cells transport and deposit
IgM-containing immune complexes onto follicular dendritic cells. Int Immunol 16:
13. Gretz JE, Norbury CC, Anderson AO, Proudfoot AE, Shaw S (2000) Lymph-borne
chemokines and other low molecular weight molecules reach high endothelial ven-
ules via specialized conduits while a functional barrier limits access to the lymphocyte
microenvironments in lymph node cortex. J Exp Med 192:1425–1440.
14. Roozendaal R, et al. (2009) Conduits mediate transport of low-molecular-weight
antigen to lymph node follicles. Immunity 30:264–276.
15. Fearon DT, Carroll MC (2000) Regulation of B lymphocyte responses to foreign and
self-antigens by the CD19/CD21 complex. Annu Rev Immunol 18:393–422.
16. Goerke AR, Swartz JR (2008) Development of cell-free protein synthesis platforms for
disulfide bonded proteins. Biotechnol Bioeng 99:351–367.
17. Kanter G, et al. (2007) Cell-free production of scFv fusion proteins: an efficient ap-
proach for personalized lymphoma vaccines. Blood 109:3393–3399.
18. Shoham T, et al. (2003) The tetraspanin CD81 regulates the expression of CD19 during
B cell development in a postendoplasmic reticulum compartment. J Immunol 171:
19. Tao MH, Levy R (1993) Idiotype/granulocyte-macrophage colony-stimulating factor
fusion protein as a vaccine for B-cell lymphoma. Nature 362:755–758.
20. Syrengelas AD, Chen TT, Levy R (1996) DNA immunization induces protective im-
munity against B-cell lymphoma. Nat Med 2:1038–1041.
21. King CA, et al. (1998) DNA vaccines with single-chain Fv fused to fragment C of tet-
anus toxin induce protective immunity against lymphoma and myeloma. Nat Med 4:
22. Biragyn A, Tani K, Grimm MC, Weeks S, Kwak LW (1999) Genetic fusion of chemokines
to a self tumor antigen induces protective, T-cell dependent antitumor immunity. Nat
23. Fredriksen AB, Bogen B (2007) Chemokine-idiotype fusion DNA vaccines are poten-
tiated by bivalency and xenogeneic sequences. Blood 110:1797–1805.
24. Biragyn A, et al. (2004) Chemokine receptor-mediated delivery directs self-tumor
antigen efficiently into the class II processing pathway in vitro and induces protective
immunity in vivo. Blood 104:1961–1969.
25. Schiavo R, et al. (2006) Chemokine receptor targeting efficiently directs antigens to
MHC class I pathways and elicits antigen-specific CD8+ T-cell responses. Blood 107:
26. Dempsey PW, Allison ME, Akkaraju S, Goodnow CC, Fearon DT (1996) C3d of com-
plement as a molecular adjuvant: Bridging innate and acquired immunity. Science
27. Lou D, Kohler H (1998) Enhanced molecular mimicry of CEA using photoaffinity
crosslinked C3d peptide. Nat Biotechnol 16:458–462.
28. Suradhat S, et al. (2001) Fusion of C3d molecule with bovine rotavirus VP7 or bovine
herpesvirus type 1 glycoprotein D inhibits immune responses following DNA immu-
nization. Vet Immunol Immunopathol 83:79–92.
29. Gor DO, Ding X, Li Q, Greenspan NS (2006) Genetic fusion of three tandem copies of
murine C3d sequences to diphtheria toxin fragment B elicits a decreased fragment B-
specific antibody response. Immunol Lett 102:38–49.
30. Pulczynski S, Boesen AM, Jensen OM (1993) Antibody-induced modulation and in-
tracellular transport of CD10 and CD19 antigens in human B-cell lines: An immuno-
fluorescence and immunoelectron microscopy study. Blood 81:1549–1557.
31. Du X, Beers R, Fitzgerald DJ, Pastan I (2008) Differential cellular internalization of
anti-CD19 and -CD22 immunotoxins results in different cytotoxic activity. Cancer Res
32. Lynch RG, Graff RJ, Sirisinha S, Simms ES, Eisen HN (1972) Myeloma proteins as tumor-
specific transplantation antigens. Proc Natl Acad Sci USA 69:1540–1544.
33. George AJ, Tutt AL, Stevenson FK (1987) Anti-idiotypic mechanisms involved in sup-
pression of a mouse B cell lymphoma, BCL1. J Immunol 138:628–634.
34. Kaminski MS, Kitamura K, Maloney DG, Levy R (1987) Idiotype vaccination against
murine B cell lymphoma. Inhibition of tumor immunity by free idiotype protein.
J Immunol 138:1289–1296.
35. Carroll WL, Starnes CO, Levy R, Levy S (1988) Alternative V kappa gene rearrange-
ments in a murine B cell lymphoma. An explantation for idiotypic heterogeneity.
J Exp Med 168:1607–1620.
36. Thornton BP, Vĕtvicka V, Ross GD (1994) Natural antibody and complement-mediated
antigen processing and presentation by B lymphocytes. J Immunol 152:1727–1737.
37. Sixt M, et al. (2005) The conduit system transports soluble antigens from the afferent
lymphto resident dendritic cells inthe T cell area ofthe lymphnode. Immunity 22:19–29.
38. De Groot AS, et al. (2008) Activation of natural regulatory T cells by IgG Fc-derived
peptide “Tregitopes”. Blood 112:3303–3311.
39. Ding C, Wang L, Marroquin J, Yan J (2008) Targeting of antigens to B cells augments
antigen-specific T-cell responses and breaks immune tolerance to tumor-associated
antigen MUC1. Blood 112:2817–2825.
40. DiLillo DJ, et al. (2008) Maintenance of long-lived plasma cells and serological
memory despite mature and memory B cell depletion during CD20 immunotherapy in
mice. J Immunol 180:361–371.
41. Takata T, et al. (2009) Attenuated antibody reaction for the primary antigen but not
for the recall antigen of influenza vaccination in patients with non-Hodgkin B-cell
lymphoma after the administration of rituximab-CHOP. J Clin Exp Hematop 49:9–13.
42. Patel KG, et al. (2009) Cell-free production of Gaussia princeps luciferase—antibody
fragment bioconjugates for ex vivo detection of tumor cells. Biochem Biophys
Res Commun 390:971–976.
Ng et al.PNAS
| September 4, 2012
| vol. 109
| no. 36