Protein Immunization Induces Memory CD4+ T Cells That Lack TH Lineage Commitment

Protein Immunization Induces Memory CD4
+ T Cells That Lack Th Lineage Commitment

Linda M. Sircy, Malia Harrison-Chau, Camille Leite Novis,
Andrew Baessler, Jacklyn Nguyen and J. Scott Hale
This information is current as
of November 14, 2021. J Immunol 2021; 207:1388-1400; Prepublished online 11
August 2021;
Downloaded from http://www.jimmunol.org/ at Red de Bibliotecas del CSIC on November 14, 2021
doi: 10.4049/jimmunol.2100210
http://www.jimmunol.org/content/207/5/1388
Supplementary http://www.jimmunol.org/content/suppl/2021/08/11/jimmunol.210021
Material 0.DCSupplemental
References This article cites 62 articles, 23 of which you can access for free at:
http://www.jimmunol.org/content/207/5/1388.full#ref-list-1
Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision
• No Triage! Every submission reviewed by practicing scientists
• Fast Publication! 4 weeks from acceptance to publication
*average
Subscription Information about subscribing to The Journal of Immunology is online at:

http://jimmunol.org/subscription
Permissions Submit copyright permission requests at:
http://www.aai.org/About/Publications/JI/copyright.html
Email Alerts Receive free email-alerts when new articles cite this article. Sign up at:
http://jimmunol.org/alerts
The Journal of Immunology is published twice each month by

The American Association of Immunologists, Inc.,
1451 Rockville Pike, Suite 650, Rockville, MD 20852
Copyright © 2021 by The American Association of
Immunologists, Inc. All rights reserved.
Print ISSN: 0022-1767 Online ISSN: 1550-6606.
The Journal of Immunology
Protein Immunization Induces Memory CD4+ T Cells That Lack

Th Lineage Commitment
Linda M. Sircy,1 Malia Harrison-Chau,1 Camille Leite Novis, Andrew Baessler,

Jacklyn Nguyen, and J. Scott Hale
Acute viral infection generates lineage-committed Th1 and T follicular helper (Tfh) memory cells that recall their lineage-specific
functions following secondary challenge with virus. However, the lineage commitment of effector and memory Th cells in vivo following
protein vaccination is poorly understood. In this study, we analyzed effector and memory CD4+ T cell differentiation in mice (Mus
musculus) following adjuvanted glycoprotein immunization compared with acute lymphocytic choriomeningitis virus infection.
Glycoprotein immunization induced CXCR5 non-Tfh effector and memory CD4+ T cells that surprisingly had not undergone
polarization toward any particular Th cell lineage but had undergone memory differentiation. However, upon challenge with virus,
these Th lineagenonpolarized memory CD4+ T cells were able to generate Th1 secondary effector cells, demonstrating their lineage
plasticity. In addition, Tfh and memory Tfh cells were generated in response to protein immunization, and these cells differed from
infection-induced Tfh cells by their lack of the transcription factor Tbet. Rechallenge experiments demonstrated that viral infection,
but not protein immunization, during either the primary or secondary immune response, restricts the recall of Bcl6 expression and the
generation of germinal center Tfh cells. Together, these data demonstrate that protein immunization generates a combination of
nonpolarized memory cells that are highly plastic and memory Tfh cells that can undergo further Th1-like modulation during a
secondary response to viral infection. The Journal of Immunology, 2021, 207: 13881400.
E
ffective immunity to diverse pathogens requires special- differentiation is well characterized, protein immunization-
ized effector immune cell responses. Upon antigenic driven Th cell differentiation, lineage commitment, and memory
challenge, naive CD41 T cells proliferate and differenti- potential in vivo is still largely unclear.
ate into specialized Th cell lineages based on the type of patho- Compared with acute viral infection and live-attenuated vaccines
genic infection and the induced inflammatory cytokine (20), recombinant protein-subunit vaccines vary in their ability to
environment (1). Th cell differentiation is determined in part by stimulate effective adaptive and humoral immune responses on their
lineage-associated transcription factors, including Tbet in Th1 own (20). Therefore, protein vaccinations typically use adjuvants to
cells (2), Gata3 in Th2 cells (3), RORgt in Th17 cells (4), and activate the innate immune response via various mechanisms and
Bcl6 in T follicular helper (Tfh) cells (57). Following acute have been shown to improve both the initial and long-term humoral
intracellular infections, naive CD41 T cells are activated and immune responses (2030). Squalene oil-in-water emulsions, such
differentiate into effector Tfh and Th1 cells that have distinct as MF59 (and its veterinary-grade equivalent Addavax) and AS03,
function and gene expression profiles (5, 811). Tfh cells have been described to induce balanced Th1/Th2 responses (20, 31).
express the follicle homing chemokine receptor CXCR5 and the Class B CpG oligodeoxynucleotides (ODNs), which are TLR9 ago-
transcription factor Bcl6 and provide help to germinal center nists (20, 27, 3236), and Poly(I:C), a TLR3 agonist, are both
(GC) B cells for affinity maturation and memory B cell and described as Th1-driving adjuvants (20, 37). However, previous
long-lived plasma cell differentiation (1216). Th1 cells express studies into the effect of these adjuvants on T cell responses (31, 32,
the transcription factor Tbet and the cytokine IFN-g, which 34, 35, 37) fail to fully elucidate the differentiation of effector Th
attracts adaptive and innate immune cells to sites of infection to cells in vivo through transcription factor profiling. Instead, often, the
aid in the control and clearance of viruses and intracellular bac- relative ratios of induced IgG1 or IgG2 Abs are used to surmise
teria (1, 9, 17, 18). Following pathogen clearance, Th1 and Tfh whether Th1 or Th2 cells are likely present (31, 32, 34, 35, 37). In
cells give rise to their respective memory Th1 and Tfh counter- addition to Th1 or Th2 cells, nonpolarized helper T (Tnp) cells,
parts that maintain lineage commitment and promote their dis- which do not express key Th lineage signature cytokines or tran-
tinct Th1 and Tfh effector functions during a recall response (9, scription factors, have been identified after both in vitro culturing
19). Although acute viral infectioninduced Th and memory and in vivo priming, although some consider these cells to be
Department of Pathology, University of Utah, Salt Lake City, UT M.H.-C., and J.S.H. Wrote the paper: L.M.S. and J.S.H. Supervision and oversight:
1
J.S.H.
L.M.S. and M.H.-C. contributed equally to this work.
Address correspondence and reprint requests to: Dr. J. Scott Hale, Department of
ORCIDs: 0000-0002-1280-5370 (M.H.-C.); 0000-0002-9111-3662 (C.L.N.); 0000- Pathology, University of Utah, School of Medicine, 15 North Medical Drive East, Salt
0002-4572-9001 (A.B.). Lake City, Utah 84112. E-mail address: scott.hale@path.utah.edu
Received for publication March 4, 2021. Accepted for publication July 1, 2021.
The online version of this article contains supplemental material.
This work was supported by the National Institute of Allergy and Infectious Diseases,
National Institutes of Health Career Transition Award K22 AI116803 (to J.S.H.) and Grants Abbreviations used in this article: 1 , primary; 2 , secondary; alum, aluminum hydroxide
R01 AI137238 (to J.S.H.), T32 AI055434 (to L.M.S.), and T32 AI138945 (to A.B.). gel; dLN, draining lymph node; GC, germinal center; LCMV, lymphocytic choriomeningitis
virus; ODN, oligodeoxynucleotide; Tfh, T follicular helper; Tnp, nonpolarized helper T.
Conceived and designed the experiments: L.M.S., M.H.-C., and J.S.H. Performed the
experiments: L.M.S., M.H.-C., C.L.N., A.B., J.N., and J.S.H. Analyzed the data: L.M.S., Copyright © 2021 by The American Association of Immunologists, Inc. 0022-1767/21/$37.50
www.jimmunol.org/cgi/doi/10.4049/jimmunol.2100210
The Journal of Immunology 1389
precursors to Th1 or Th2 committed cells rather than a distinct InvivoGen) at a 1:1 ratio. Recombinant LCMV glycoprotein-expressing 293A
effector cell subset (3842). One study demonstrated that cells were kindly provided by Dr. C. Davis (Emory University), and recombi-
nant glycoprotein was purified from supernatants as described previously (44).
in vitroprimed, nonpolarized CD41 T cells share similarities to
naive CD41 T cells in their migratory capabilities to lymphoid Flow cytometry
organs (39); however, these cells were more effective at responding Single-cell suspensions of spleens and pooled lumbar and inguinal lymph
to Ag and secreting cytokines than naive cells, demonstrating that nodes were prepared, and up to 1 106 cells were stained in 1 PBS sup-
they had undergone some activation-induced changes (39). Although plemented with 2% FBS (FACS buffer) for 1530 min on ice with fluoro-
most of these studies used in vitro activation to generate Tnp CD41 chrome-conjugated Abs CD4 (RM4-5), CD45.1 (A20), CD44 (IM7), CD62L
(MEL-14), CD127 (A7R34), CCR7 (4B12), IFN-g (XMG1.2), TNF-a
T cells, these cells have been shown to arise in vivo after protein (MP6-XT22), IL-2 (JES6-5H4), IL-4 (11B11), IL-17 (eBio17B7), Bcl6
immunization but are not significantly present after viral infection (K112-91), Tbet (4B10), Gata3 (L50-823), RORgt (B2D), Foxp3 (FJK-16s),
(38). Although Tnp cells that are generated in vitro can be subse- and CXCR3 (CXCR3-173) (purchased from BD Biosciences, eBioscience,
quently polarized toward Th1 or Th2 phenotypes (38, 39, 41, 42), it BioLegend, Vector Laboratories, and Invitrogen) for cell surface Ags.
CXCR5 surface staining was performed using a three-step protocol described
is still largely unclear whether these nonpolarized CD41 T cells are in Johnston et al. (2009) (5) using purified rat anti-mouse CXCR5 primary
generated in vivo during an immune response, whether they differ- Ab (2G8; BD Biosciences) in FACS buffer supplemented with 3.4% BSA
entiate into long-lived memory cells, and whether such memory (no. A7284; Sigma-Aldrich) and 2% mouse serum (no. M5905; Sigma-
cells exhibit helper T cell lineage commitment versus plasticity dur- Aldrich) (CXCR5 staining buffer), a secondary Biotin-SPconjugated Affini-
ing recall responses to Ag in vivo. pure F(Ab’)2 Goat Anti-Rat IgG (Jackson ImmunoResearch Laboratories) in
CXCR5 staining buffer and then with a fluorochrome-conjugated streptavidin
To determine the characteristics, lineage differentiation, and com- (APC or PeCy7; Invitrogen) in FACS buffer. For transcription factor stain-
mitment of helper T cells induced by adjuvanted protein immuniza- ing, cells were first stained for surface Ags, followed by permeabilization,
tion, we studied Ag-specific CD41 T cells following viral infection fixation, and staining using the Foxp3 Permeabilization/Fixation kit and pro-
versus protein immunization throughout the primary and secondary tocol (eBioscience). Intracellular cytokine staining was done by standard
techniques following 5-h stimulation of up to 2 106 cells with gp6180
effector responses in vivo. We found that after i.m. immunization
peptide and brefeldin A (GolgiPlug; BD Biosciences) in RPMI medium sup-
with recombinant glycoprotein in adjuvant, Ag-specific naive CD41 plemented with 5% FBS at 37 C with 5% CO2. No peptide controls were
T cells differentiated into effector CXCR51 Tfh cells and Tnp treated under the same conditions supplemented with brefeldin A but without
effector cells that expressed IL-2 and TNF-a but that lacked expres- gp6180 peptide. Cells were then stained for surface Ags, followed by perme-
sion of Th1, Th2, and Th17 lineagespecific transcription factors abilization, fixation, and staining using the Cytofix/Cytoperm kit and protocol
(BD Biosciences). For I-Ab gp6677 tetramer (provided by the National Insti-
and cytokines. Even when using adjuvants previously described as tutes of Health Tetramer Core) staining, cells were incubated with tetramer in
Th1/Th2 or Th1 inducers, only a small fraction induced polarized RPMI medium supplemented with 10% FBS for 2 h at 37 C with 5% CO2.
Th1-like qualities, whereas the vast majority of effector cells did not Flow cytometry data were analyzed using FlowJo v9 and v10 software.
acquire Th1/Th2/Th17 cell traits. In vivo reactivation of memory
Statistical analysis
Tnp cells with protein immunization resulted in similarly neutral
All experiments were analyzed using Prism 8 and 9. Statistically significant
Tnp secondary effector cells, whereas reactivation with acute viral
p values <0.05 are indicated and were determined using a two-tailed
infection drove these memory cells to differentiate into Th1 second- unpaired Student t test.
ary effector cells. In addition, we identified a virus infectionin-
duced programming that limits GC Tfh cell potential during the
secondary Tfh cell response. Based on our findings, we propose that Results
protein immunizationinduced Tnp cells successfully differentiate Protein immunizationinduced, Ag-specific CD41 T cells exhibit
into long-lived memory T cells that remain Th lineage nonpolarized increased memory potential compared with virus infectioninduced
(relative to the Th1, Th2, and Th17 effector lineages), yet possess CD41 T cells
plasticity to undergo Th lineage differentiation during the secondary To determine the kinetics and differentiation of Ag-specific CD41
immune response to infection. T cells following adjuvanted protein immunization compared with
acute viral infection, we adoptively transferred congenically marked
Materials and Methods (CD45.11) naive SMARTA TCR transgenic CD41 T cells specific
for the LCMV gp6180 epitope (43) into C57BL/6J mice. Recipient
Mice and adoptive transfers
mice were then immunized i.m. in both quadriceps with recombi-
Congenically marked (CD45.1) SMARTA CD41 T cell splenocytes specific
nant LCMV glycoprotein in adjuvant [Addavax or Poly(I:C)] or
for the I-Abrestricted lymphocytic choriomeningitis virus (LCMV) gp6180
epitope obtained from naive SMARTA TCR transgenic mice (43) were i.v. infected by i.p. injection with LCMV Armstrong. The i.m. route of
transferred into naive C57BL/6J (CD45.2) mice (The Jackson Laboratory, Bell administration for glycoprotein immunizations was chosen to more
Harbor, ME). For primary infections, we transferred 2 104 CD45.11 closely mimic the administration method of protein vaccinations in
SMARTA cells. For adoptive transfer of memory SMARTA cells, splenocytes clinical settings. Donor SMARTA cells (CD41CD45.11) were ana-
from chimeric mice (days 4063 postinfection/immunization) were stained
with CD45.2-PE and MACS Anti-PE MicroBeads (Miltenyi Biotec), followed
lyzed at effector (day 7) and memory (up to day 84) time points
by magnetic cell sorting to enrich for CD45.11 SMARTA cells. For the mem- postimmunization or infection (Fig. 1A). Clonal expansion of
ory SMARTA cell transfer experiments, 4 103 to 1.5 104 CD45.11 SMARTA cells was induced following glycoprotein immunization
SMARTA cells ($95% purity) were transferred into naive C57BL/6J recipient but was substantially lower compared with LCMV infection (Fig.
mice. Animal experiments were conducted in accordance with approved Uni- 1B, 1C, Supplemental Fig. 1).
versity of Utah Institutional Animal Care and Use Committee protocols.
Following the peak of clonal expansion at day 7, SMARTA cells
Viral infections and protein immunizations underwent contraction in all groups in the peripheral blood, spleen,
Mice were infected by i.p. injection with 2 105 PFU of LCMV Armstrong and lymph nodes (Fig. 1BD), and memory cells were clearly
or immunized by i.m. (quadriceps) injection with 210 mg recombinant detectable 84 d after both acute LCMV infection and adjuvanted
LCMV glycoprotein with addition of Addavax (InvivoGen) adjuvant at a 1:1 glycoprotein immunization (Fig. 1D). Although the initial clonal
ratio, with 50 mg VacciGrade Poly(I:C) (HMW) (InvivoGen) adjuvant, with
50 mg VacciGrade CpG ODN 1826 (InvivoGen) adjuvant, with aluminum
expansion of protein vaccination-induced SMARTA cells was much
hydroxide gel (alum; Alhydrogel) adjuvant 2% (InvivoGen) at a 1:1 ratio, or lower compared with viral infection (Fig. 1B, 1C), the relative fre-
CFA (1 mg/ml heat-killed dried Mycobacterium tuberculosis strain H37Ra; quency of SMARTA cells that survived the contraction phase to
1390 IMMUNIZATION-INDUCED NONPOLARIZED CD4 T CELLS
FIGURE 1. CD41 T cell clonal expansion and memory formation following glycoprotein immunization versus acute viral infection. A total of 2 104
naive CD45.11 LCMV gp6180-specific TCR transgenic SMARTA CD41 T cells were adoptively transferred into naive CD45.21 C57BL/6J mice. Recipi-
ents were then either infected by i.p. injection with 2 105 PFU of LCMV Armstrong or immunized by i.m. injection with 2 mg recombinant glycoprotein
with Addavax (Adda) or Poly(I:C) adjuvant. Representative FACS plots are gated on CD41CD45.11 SMARTA cells at the indicated time points relative to
infection or immunization. (A) Schematic of experimental design. (B) Kinetics of SMARTA CD41 T cells in peripheral blood. Error bars represent SD. (C)
Number of effector (day 7) SMARTA CD41 T cells in dLN (dLN 5 combined lumbar and inguinal) and spleen. (D) Number of memory (day 84) SMARTA
CD41 T cells in dLN and spleen. (E) Tissue-specific percentage of survival of SMARTA CD41 T cells shown as the ratio of the number of memory (day
84) SMARTA CD41 T cells to the mean of the number of effector (day 7) SMARTA CD41 T cells in dLN and spleen. (F) Representative FACS plots of
CD62L and CD44 analysis of endogenous CD41 T cells (Endog CD4) and day 7 effector SMARTA CD41 T cells in dLN (glycoprotein Adda) and spleen
(LCMV). (G) Histograms of CCR7 analysis of SMARTA CD41 T cells in dLN and spleen comparing immunization with glycoprotein in Adda adjuvant
(green) and LCMV infection (red) and naive CD41 T cells (gray). Graph of CCR71 frequency of day 7 effector SMARTA cells in dLN and spleen. (H) Rep-
resentative CD62L and CD127 analysis FACS plots gated on day 7 effector SMARTA CD41 T cells in dLN. (I) Graph of CD1271 frequency of day 7 effec-
tor SMARTA cells in dLN and spleen. (J) Graph of CD62L1 frequency of day 7 effector SMARTA cells in dLN and spleen. n $ 4 per group per
experiment at each time point. Statistically significant p values are shown and were determined using a two-tailed unpaired Student t test. Error bars represent
mean and SEM. Data shown are from one experiment and are representative of at least two independent experiments.
become memory cells was significantly higher in the draining lymph contraction phase at an 4-fold increased rate compared with
nodes (dLN), 6-fold higher in Addavax-adjuvanted glycoprotein SMARTA cells in LCMV-infected mice (16.12 versus 4.22% mean
immunization compared with LCMV infection (2.130.36% mean), survival to memory) (Fig. 1E). These data show that although adju-
and nearly 3-fold higher in Poly(I:C)-adjuvanted immunization com- vanted glycoprotein protein immunization induces a lower degree of
pared with LCMV infection (0.990.36% mean) (Fig. 1E). Donor clonal expansion compared with acute viral infection, the degree of
SMARTA cells in glycoprotein-immunized mice survived the survival as cells progress to memory is significantly higher.
We predicted that the increased survival of memory SMARTA A large proportion of immunization-induced effector CD41 T cells
CD41 T cells would correlate with the earlier progression of these lack expression of lineage-defining transcription factors
cells to acquire a memory cell phenotype. The overwhelming major- To further determine the identity and characteristics of Th cells that
ity of donor effector T cells at day 7 after adjuvanted glycoprotein respond to adjuvanted glycoprotein immunization versus viral infec-
immunization and acute LCMV infection were CD44high, consistent tion, we analyzed day 7 effector Ag-specific CD41 T cells for
with T cell activation and expansion (Fig. 1F); however, a large pro- expression of Th lineageassociated transcription factors. Both pro-
portion (40%) of SMARTA cells from glycoprotein-immunized
tein immunization and viral infectioninduced CXCR51 Tfh cells
mice expressed the lymphoid homing molecule CD62L (Fig. 1F,
expressed CXCR5 and Bcl6; however, the protein immunization
1J). In addition, these cells expressed significantly higher levels of
(compared with viral infection) generated a much higher proportion
CCR7 (Fig. 1G) and contained a higher proportion of CD1271 and
that were Bcl6high GC Tfh cells (Fig. 3AC). As expected, LCMV
CD62L1 effector cells compared with the acute LCMV-infected
infectioninduced CD41 T cells were comprised of CXCR51 T
group (Fig. 1H, 1I). Together, these data show that although adju-
cells expressing intermediate levels of the Th1 transcription factor
vanted glycoprotein immunization induces a lower degree of clonal
Tbet, and CXCR5- Th1 cells that were either Tbet high or Tbet
expansion compared with acute LCMV infection, protein immuniza-
intermediate (Fig. 3AC, Supplemental Fig. 3A). In contrast, the
tioninduced CD41 T cells exhibit increased relative survival dur-
vast majority of Ag-specific CD41 T cells from glycoprotein-immu-
ing the contraction phase and become memory CD41 T cells.
nized mice did not upregulate expression of Tbet, including both
IFN-c expression is absent in adjuvanted protein CXCR5 non-Tfh cells and CXCR51 Tfh cells (Fig. 3A, 3D, 3E),
immunizationinduced effector CD41 T cells regardless of the adjuvant used (Supplemental Fig. 3B, 3C). These
To determine whether immunization-induced effector CD41 T cells data are consistent with the general lack of IFN-g expression by
have altered expression of cytokines compared with viral infection, Ag-specific CD41 T cells (Fig. 2, Supplemental Fig. 2), further con-
we analyzed day 7 effector SMARTA CD41 T cells isolated from firming that glycoprotein immunizationinduced effector cells do
spleen and dLN (lumbar and inguinal) following ex vivo peptide not undergo Th1 cell polarization in vivo. In addition, Ag-specific
restimulation. Virus-induced effector cells expressed copious CD41 T cells from either Addavax-adjuvanted, glycoprotein-immu-
amounts of the Th1-signature cytokine IFN-g, whereas IFN-g nized or LCMV-infected mice minimally expressed Gata3 or
expression was almost completely absent in Addavax-adjuvanted RORgt (Fig. 3A). We further confirmed that donor SMARTA cells
protein immunizationinduced cells (Fig. 2A, 2B). In contrast, both were not differentiating into regulatory T cells, as demonstrated by
immunization- and viral infectioninduced day 7 effector cells their lack of Foxp3 expression (Fig. 3F). Consistent with the relative
expressed TNF-a (Fig. 2A, 2C, 2D) and IL-2 (Fig. 2A, 2E, 2F). lack of IL-4expressing cells (Supplemental Fig. 2C, 2D), very few
TNF-a was either similar (in dLN) or higher (spleen) in immuniza- effector Ag-specific CD41 T cells (<5%) from alum-adjuvanted
tion-induced effector cells compared with infection-induced cells, mice expressed Gata3 (Supplemental Fig. 3D, 3E), further suggest-
and both the frequency of effector cells and the intensity of IL-2 ing that the majority of effector Ag-specific CXCR5 cells did not
expression were greater in immunization-induced effector CD41 T undergo Th2 differentiation. Effector SMARTA CXCR5 cells
cells (Fig. 2CF). These data suggested that protein immuniza- from CFA-adjuvanted, glycoprotein-immunized mice largely upre-
tioninduced CD41 T cells had undergone effector differentiation, gulated RORgt (Supplemental Fig. 3F, 3G), consistent with their
enabling them to express TNF-a and IL-2, but had not undergone IL-17 expression (Supplemental Fig. 3E, 3F), further confirming
Th1 cell polarization. To evaluate whether other adjuvants could that CFA induced differentiation of Th17 cells as previously shown
induce Th1 differentiation during protein immunization, we used (20, 4648). Together, these data indicate that the majority of adju-
either Poly(I:C) or CpG as adjuvants for i.m. glycoprotein immuni- vants used in this study (with exception of the highly inflammatory
zation. Although glycoprotein immunization with Addavax adjuvant CFA) in combination with protein immunization generated effector
consistently generated negligible IFN-gexpressing effector cells, cells that were not differentiated toward any of the conventional
Poly(I:C) and CpG adjuvants generated only a small fraction of Th1, Th2, or Th17 Th cell lineages.
IFN-g producers (15 and 5%, respectively) that expressed lower We quantified the frequencies of Ag-specific Th cells by their
levels of IFN-g when compared with Th1 cells generated by acute expression of lineage-defining markers and transcription factors
LCMV infection (Fig. 2G, Supplemental Fig 2B). We next evalu- (Fig. 3G). Importantly, 65% of LCMV-induced effector cells were
ated the capability of other adjuvants to induce lineage-associated Th1 cells, combining TbethighCXCR5 and TbetintCXCR5 Th1
cytokines, including alum, which is thought of as a Th2-associated cells (Fig. 3G, Supplemental Fig. 3A), and 30% of cells were
adjuvant (20, 45) and CFA, which contains heat-killed M. tuberculo- CXCR51 Tfh cells (Fig. 3G). In contrast, 47% of the protein
sis and can promote Th17/Th1 cells (20, 4648). As expected, the immunizationinduced generated cells were Tfh cells, whereas the
majority of effector SMARTA cells from alum-adjuvanted, glyco- vast majority (47%) of the remaining CXCR5 cells were nonpolar-
protein-immunized mice did not express IFN-g or IL-17 ized cells, being negative for Tbet, Gata3, and RORgt (Fig. 3G).
(Supplemental Fig. 2A, 2B, 2E, 2F). Surprisingly, IL-4 expressing Together, these data demonstrate that adjuvanted glycoprotein
cells comprised a mere 3% of Ag-specific cells (Supplemental Fig. immunization induces CD41 T cells to differentiate into a heteroge-
2C, 2D), suggesting that the remaining majority of cells have not nous population of Tfh cells and effector CXCR5 cells that have
undergone Th lineage differentiation. CFA-adjuvanted glycoprotein not undergone polarization toward any of the conventional Th effec-
immunization resulted in a relatively high frequency of IL-171 tor lineages (Th1, Th2, and Th17). Therefore, we suggest that these
effector SMARTA cells (mean 22%) (Supplemental Fig. 2E, 2F) effector cells that are Th lineage uncommitted be designated as Tnp
and a smaller fraction of IFN-g1 cells (Supplemental Fig. 2A, 2B), cells, similar to those described in in vitro and in vivo T cell differ-
suggesting a preference toward Th17 differentiation and low fre- entiation studies (3840).
quencies of Th1 cells. Together, these findings suggest that for most
of the adjuvants used in this study, the majority of Ag-specific cells Polyclonal effector cells induced by protein immunization lack Th
are negative for Th1, Th2, or Th17 lineage cytokines, whereas polarization
highly inflammatory adjuvants such as CFA can induce a higher To determine whether polyclonal T cell responses induced by pro-
degree of polarized cytokine-producing cells. tein immunization similarly lack Th lineage differentiation observed
FIGURE 2. Adjuvanted glycoprotein

immunization promotes effector Th
cells that do not express IFN-g. Intra-
cellular cytokine analysis of effector
CD41CD45.11 SMARTA cells 7 d
after LCMV infection or immunization
with glycoprotein with the indicated
adjuvants following 5 h restimulation
with gp6180 peptide in the presence
of GolgiPlug. (A) Representative
FACS plots of IFN-g, TNF-a, and IL-
2 analysis of SMARTA CD41 T cells
in dLN comparing immunization with
glycoprotein in Addavax adjuvant and
LCMV infection. Splenocytes from
LCMV-infected mice were used for
the no peptide controls. (B) Percentage
of IFN-g1 of SMARTA cells in dLN
and spleen. (C) Percentage of TNF-a1
of SMARTA cells. (D) TNF-a mean
fluorescence intensity (MFI) of TNF-
apositive SMARTA cells. (E) Per-
centage of IL-21 of SMARTA cells.
(F) IL-2 MFI of IL-2positive
SMARTA cells. (G) Representative
FACS plots of IFN-g and TNF-a anal-
ysis of SMARTA CD41 T cells in
dLN comparing immunization with
glycoprotein in Addavax, Poly(I:C), or
ODN 1826 CpG adjuvants and LCMV
infection. Statistically significant p val-
ues are shown and were determined
using a two-tailed unpaired Student t
test with Welch correction. Error bars
represent mean and SEM. For each
panel, the data shown are from one
experiment (n 5 5 mice per group)
and are representative of at least two
independent experiments.
in SMARTA TCR transgenic T cell responses (Figs. 2, 3), we ana- infected, mice (Fig. 4C, 4D). In contrast, Ag-specific TNF-a and
lyzed day 7 effector CD41 T cells from C57BL/6J mice immunized IL-2expressing cells were relatively proportional to the frequencies
with Addavax-adjuvanted glycoprotein or infected with LCMV. We of tetramer1 cells in immunized mice and infected mice (Fig. 4B,
stained cells isolated from dLN (lumbar and inguinal) with the I-Ab 4EH). In addition, we observed that immunization induced a larger
LCMV gp6677 MHC class II tetramer (Fig. 4A) or analyzed cyto- frequency of tetramer1 CXCR51Bcl6high GC Tfh cells and higher
kine expression by CD4 T cells following restimulation with intensity of Bcl6 expression in Tfh cells compared with viral infec-
gp6180 peptide (Fig. 4B). Ag-specific cells expressed TNF-a and tion (Fig. 4IK), consistent with our previous data analyzing
IL-2, whereas very few of these cells from immunized mice SMARTA cells (Fig. 3AC). Immunization-induced tetramer1 Tfh
expressed IFN-g compared with infected mice (Fig. 4B). When the CXCR51 and non-Tfh CXCR5 cells lacked expression of Tbet
frequency of Ag-specific IFN-g1 cells was compared with the fre- (Fig. 4I, 4L, 4M), RORgt, and Gata3 (data not shown). Together,
quency of I-Ab gp6677 tetramer1 cells, IFN-gexpressing cells these data demonstrate that polyclonal Ag-specific T cells induced
were significantly underrepresented in immunized, but not in by adjuvanted protein immunization are predominantly comprised
FIGURE 3. Glycoprotein immunization generates lineage-unpolarized effector cells that lack Tbet, Gata3, and RORgt. Flow cytometry analysis of effector
CD41CD45.11 SMARTA cells 7 d after LCMV infection or immunization with glycoprotein in Addavax. (A) Representative CXCR5 and Bcl6 FACS analy-
sis of effector SMARTA cells in dLN and Tbet, Gata3, and RORgt analysis in spleen. (B) Percentage of CXCR51Bcl6high GC Tfh cells of effector
SMARTA cells in dLN following glycoprotein in Addavax immunization and LCMV infection. (C) Bcl6 mean fluorescence intensity (MFI) of CXCR51 Tfh
cells of effector SMARTA cells in dLN following glycoprotein in Addavax immunization and LCMV infection. (D) Tbet MFI of CXCR5 SMARTA cells.
(E) Tbet MFI of CXCR51 Tfh SMARTA cells. (F) Representative Foxp3 FACS analysis of endogenous CD41 T cells (Endog CD4) from a glycoprotein-
immunized, SMARTA-recipient and CD45.11 SMARTA CD41 T cells from glycoprotein-immunized (dLN) or LCMV-infected (spleen) mice. (G) Pie
charts indicate the relative fraction of day 7 effector SMARTA CD41 T cells for each of the indicated Th phenotypes. Lack of expression of Tbet, Gata3,
RORgt, and CXCR5 cells (Tnp) represented by green. n $ 4 per group per experiment. For each panel, data shown are from one experiment and are repre-
sentative of at least three independent experiments. Statistically, significant p values are shown and were determined using a two-tailed unpaired Student t
test with Welch correction. Error bars represent mean and SEM.
of Tfh and Tnp effector cells, similar to what was observed using a naive C57BL/6J recipients and then either immunized with glyco-
monoclonal TCR transgenic system (Figs. 2, 3). protein in Addavax adjuvant or infected with 2 105 PFU LCMV
Armstrong, as performed previously, to generate Tnp memory and
Protein immunizationinduced Tnp memory cells maintain plasticity
Th1 memory cells, respectively. Forty days later, we purified mem-
and can differentiate into Th1 cells during recall response to acute
ory CD41CD45.11 SMARTA cells from immunized or infected
viral infection
mice and adoptively transferred these memory cells into new naive
Given that glycoprotein immunizationinduced Tnp effector cells C57BL/6J recipients, followed by protein immunization or LCMV
largely lack Th1, Th2, or Th17 polarization (Figs. 3, 4) and give infection (Fig. 5A). In addition, we also adoptively transferred naive
rise to memory CD4 T cells (Fig. 1BE), we hypothesized that such SMARTA cells to directly compare the primary SMARTA cell
memory cells would display plasticity to undergo Th lineage differ- responses to the secondary responses 7 d following glycoprotein
entiation during a secondary immune response to infection. To test immunization or viral infection (Fig. 5A). Seven days postimmuni-
this, we adoptively transferred naive SMARTA CD41 T cells into zation/infection, glycoprotein-specific CD4 T cell recall responses
FIGURE 4. Polyclonal effector cells induced by protein immunization lack Th polarization. Mice were either infected by i.p. injection with 2 105 PFU
of LCMV Armstrong or immunized by i.m. injection with 2 mg recombinant glycoprotein with Addavax adjuvant. Seven days postinfection or -immuniza-
tion, dLN lymphocytes were stained with I-Ab gp6677 tetramer or restimulated with gp6180 peptide for 5 h in the presence of GolgiPlug followed by
intracellular cytokine staining. (A) CD44 and tetramer analysis of total CD41 T cells. (B) Representative IFN-g, TNF-a, and IL-2 FACS analysis of total
CD41 T cells. (C) Ratio of percentage of IFN-g1 cells to percentage of tetramer1CD44high cells of total CD41 T cells for each individual mouse. (D) Paired
analyses (for each individual mouse) of percentage of tetramer1CD44high cells to percentage of IFN-g1 cells of total CD41 T cells. The p value statistics
shown compare percentage of tetramer1CD44high cells to percentage IFN-g1 cells for immunized mice (green) and infected mice (red). (E) Ratio of percent-
age of TNF-a1 cells to percentage of tetramer1CD44high cells of total CD41 T cells for each individual mouse. (F) Paired analyses (for each individual
mouse) of percentage of tetramer1CD44high cells to percentage of TNF-a1 cells of total CD41 T cells. The p value statistics shown compare percentage of
tetramer1CD44high cells to percentage of TNF-a1 cells for immunized mice (green) and infected mice (red). (G) Ratio of percentage of IL-21 cells of total
CD41 T cells to percentage of tetramer1CD44high cells of total CD41 T cells for each individual mouse. (H) Paired analyses (for each individual mouse) of
percentage of tetramer1CD44high cells to percentage of IL-21 cells of total CD41 T cells. The p value statistics shown compare percentage of tetra-
mer1CD44high cells to percentage of IL-21 cells for immunized mice (green) and infected mice (red). (I) Representative CXCR5, Tbet, and Bcl6 FACS anal-
ysis of effector tetramer1 CD41 T cells. (J) Percentage of CXCR51Bcl6high GC Tfh cells of tetramer1 CD41 T cells. (K) Bcl6 mean fluorescence intensity
(MFI) of tetramer1 CXCR51 cells in dLN. (L) Tbet MFI of tetramer1 CXCR51 cells. (M) Tbet MFI of tetramer1 CXCR5 cells. n $ 4 per group per
experiment. Data shown are from one experiment and are representative of two independent experiments. Statistically significant p values are shown and
were determined using a two-tailed unpaired Student t test with Welch correction. Error bars represent mean and SEM.
were observed in both dLN and spleen (Fig. 5B). However, SMARTA We evaluated the plasticity of the transferred glycoprotein-specific
cell accumulation was generally higher in the spleen in LCMV- memory cells (Fig. 5A) by analyzing the cytokine production and
infected mice, consistent with this being the critical lymphoid organ to transcription factors in secondary effector cells. Importantly, second-
harbor the immune response to LCMV (Fig. 5B). ary LCMV challenge of immunization-induced memory CD41 T
FIGURE 5. Unpolarized Tnp memory cells can differentiate into Th1 cells upon secondary activation during viral infection. CD45.11 SMARTA memory
cells ($95% purity) were purified by magnetic cell sorting from the spleens of mice 40 d after glycoprotein immunization or after LCMV Armstrong infec-
tion. A total of 4 103 CD45.11-enriched SMARTA memory cells ($95% purity) or naive SMARTA donor cells were transferred into naive C57BL/6J
recipient mice. Recipient mice were then either infected by i.p. injection with 2 105 PFU LCMV Armstrong or immunized by i.m. injection with 10 mg
recombinant glycoprotein with Addavax adjuvant. Primary (1 ) and secondary (2 ) effector SMARTA cell responses were analyzed 7 d later. (A) Schematic
of experimental design. (B) Percentage and number of primary and secondary effector SMARTA cells in spleen and dLN 7 d postinfection/immunization. (C)
IFN-g and IL-2 FACS analysis of effector SMARTA cells in dLN. (D) Frequency of IFN-g1 effector SMARTA cells in dLN. (E) Tbet mean fluorescence
intensity (MFI) of effector CD41CXCR5 SMARTA cells in dLN. (F) CXCR3 MFI of effector CD41CXCR5 SMARTA cells in dLN. (G) Frequency of
IL-21 effector SMARTA CD41 T cells in dLN. Statistically significant p values are shown and were determined using a two-tailed unpaired Student t test with
Welch correction. Error bars represent mean and SEM. Data shown are from one experiment (n 5 5 mice per group) and are representative of three independent
experiments in which memory cells were transferred either 40 d postpriming (data shown in this study) or days 52 or 63 postpriming (data not shown).
cells led to upregulation of IFN-g in secondary effector cells (see pri- being undifferentiated toward the Th1 lineage are able to acquire Th1
mary [1 ] glycoprotein secondary [2 ] LCMV group), but glycopro- cell properties during a secondary response specifically to LCMV, but
tein immunization challenge of immunization-induced memory cells not to adjuvanted protein immunization. Consistent with this idea, Tnp
(1 glycoprotein 2 glycoprotein group) did not result in Th1 differenti- memory cells upregulated both Tbet expression and the Th1-associ-
ation (Fig. 5C, 5D). In addition, glycoprotein immunization challenge ated chemokine receptor CXCR3 during the secondary response to
of LCMV-induced memory CD41 T cells induced a similar frequency LCMV challenge, but not in response to activation with glycoprotein
of IFN-gsecreting cells compared with LCMV challenge of LCMV- immunization in adjuvant (Fig. 5E, 5F). Furthermore, immunization-
induced memory CD41 T cells (Fig. 5C, 5D). Together, these findings induced memory Tnp cells boosted with protein immunization
indicate that protein immunization induced Tnp memory cells while induced only few secondary effector cells that localized to lung tissue,
whereas Tnp memory cells boosted with LCMV infection generated larger in mice that were immunized instead of infected in the most
Th1 secondary effector cells that efficiently localized to the lung and recent challenge (Fig. 5C, 5G). These data suggest that either glycopro-
expressed high levels of Tbet and CXCR3 (Supplemental Fig. 4). tein immunization induces greater IL-2 production from naive and
These findings combine to demonstrate the protein immunizationin- memory CD41 T cells or that LCMV infection restricts IL-2 expression
duced Tnp memory cells maintain a neutral/nonpolarized Th cell line- compared with glycoprotein immunization. These data also correlate
age state but possess plasticity to differentiate into Th1 cells upon with our previous data, suggesting that glycoprotein immunization
secondary challenge with viral infection. induces improved survival of CD41 T cells, as IL-2 has been shown to
In contrast to the recall response by Tnp memory cells, our data be important in CD41 T cell survival and memory generation (49).
show that glycoprotein immunization challenge of LCMV-induced
memory Th1 cells did not significantly alter the frequency of IFN- LCMV infection restricts GC Tfh cell formation compared with
gsecreting cells, which suggests that the Th1 memory cells maintained adjuvanted glycoprotein immunization
their Th1 lineage commitment after the secondary antigenic challenge To determine the differences between protein immunization versus
(Fig. 5C, 5D). The frequency of cells that express IL-2 was significantly virus infectioninduced memory Tfh cells, we first compared the
FIGURE 6. Viral infection induces programming that restricts Bcl6 in Tfh effector and memory cells. CD45.11 SMARTA memory cells ($95% purity)
were purified from the spleens of mice 40 d after glycoprotein immunization or after LCMV Armstrong infection. A total of 4 103 CD45.11-enriched
SMARTA memory cells ($95% purity) or naive SMARTA donor cells were transferred into naive C57BL/6J recipient mice. Recipient mice were then either
infected by i.p. injection with 2 105 PFU LCMV Armstrong or immunized by i.m. injection with 10 mg recombinant glycoprotein with Addavax adjuvant.
Primary (1 ) and secondary (2 ) effector SMARTA cell responses were analyzed 7 d later. Schematic of experimental design is shown in Fig. 4A. (A) Bcl6
and CXCR5 analysis of effector SMARTA CXCR51 Tfh cells in dLN 7 d after glycoprotein immunization, with gate indicating the CXCR5highBcl6high GC
Tfh cell subset. (B) Bcl6 mean fluorescence intensity (MFI) of SMARTA CXCR51 Tfh cells in dLN following glycoprotein immunization. (C) Percentage
of CXCR5highBcl6high GC Tfh cells of SMARTA CXCR51 Tfh cells in dLN following glycoprotein immunization. (D) Bcl6 and CXCR5 analysis of effector
SMARTA CXCR51 Tfh cells in spleen 7 d after LCMV infection, with gate indicating the CXCR5highBcl6high GC Tfh cell subset. (E) Bcl6 MFI of
SMARTA CXCR51 Tfh cells in spleen following LCMV infection. (F) Percentage of CXCR5highBcl6high GC Tfh cells of SMARTA CXCR51 Tfh cells in
spleen following LCMV infection. (G) Tbet MFI of effector SMARTA CXCR51 Tfh cells in dLN after glycoprotein immunization or after LCMV infection.
(H) CXCR3 MFI of effector SMARTA CXCR51 Tfh cells in spleen after glycoprotein immunization or after LCMV infection. Statistically significant p val-
ues are shown and were determined using a two-tailed unpaired Student t test with Welch correction. Error bars represent mean and SEM. Data shown are
from one experiment (n 5 5 mice per group) and are representative of three independent experiments in which memory cells were transferred either 40 d
postpriming (data shown in this study) or days 52 or 63 postpriming (data not shown).
primary versus secondary Tfh cell responses in the dLN 7 d after gly-
coprotein immunization (refer to experimental setup in Fig. 5A). As
observed previously (Fig. 3A), a high proportion of Tfh cells were
comprised of CXCR5highBcl6high GC Tfh cells (37%) in response to
primary glycoprotein protein immunization (1 glycoprotein)
(Fig. 6AC). Similarly, protein immunizationinduced memory cells
generated a high frequency of secondary GC Tfh cells and high Bcl6
expression in response to a second glycoprotein immunization
(1 glycoprotein 2 glycoprotein) (Fig. 6AC). In contrast, viral infec-
tioninduced memory cells were significantly reduced in their ability
to express Bcl6 and generate GC Tfh cells in response to secondary
glycoprotein immunization (1 LCMV 2 glycoprotein) (Fig. 6AC).
These data suggest that acute viral infection induces programming in
Tfh cells, which limits their GC Tfh capacity, and this restrictive pro-
gramming is maintained in subsequent memory Tfh cells, resulting in
reduced GC Tfh cells during recall responses.
We next evaluated the primary versus secondary Tfh cell
responses in the spleen following acute LCMV infection. Compared FIGURE 7. Immunization induces lineage-nonpolarized memory cells
that undergo Th1 polarization upon secondary activation with viral infec-
with the primary Tfh cell response to LCMV infection (1 LCMV),
tion. In response to adjuvanted protein immunization, Ag-specific CD41 T
the secondary responses exhibited significantly lower frequencies of cells undergo clonal expansion and differentiate into CXCR51 Bcl61 Tfh
GC Tfh cells following challenge with LCMV (both 1 glycoprotein and non-Tfh effectornonpolarized (Tnp) cells that do not express lineage-
2 LCMV and 1 LCMV 2 LCMV groups) (Fig. 6DF). Together, defining Tbet, Gata3, or RORgt transcription factors. The subsequent Th
these data suggest that LCMV infection restricts the formation of lineageuncommitted Tnp memory cells are plastic and thus upon second-
Ag-specific GC Tfh cells compared with adjuvanted glycoprotein ary viral infection can differentiate into Th1 secondary effector cells. In
immunization. addition, Tnp-like memory Tfh cells acquire Th1-like properties during
reactivation with a viral infection.
Immunization-induced Tnp-like memory Tfh cells acquire some
Th1-like properties after LCMV challenge
cells lack Tbet expression and exhibit more permissive GC Tfh cell
The findings described above suggest that acute viral infection indu- programming. Analyses of secondary Tfh cell responses generated
ces cell-intrinsic programming in Tfh memory cells, which limits from adoptively transferred glycoprotein immunizationinduced
their GC Tfh recall potential and Bcl6 re-expression. To determine memory cells demonstrate that secondary infection can alter Tfh cell
if LCMV infection was also able to induce Th1-like properties in programming, resulting in Tbet expression and reduced GC Tfh cell
Ag-specific Tnp-like memory Tfh cells following secondary infec- responses (Fig. 7).
tion, we analyzed effector CXCR51 cells for Tbet and CXCR3 A previous study by Iezzi et al. (39) identified Tnp-like cells that
expression. Importantly, secondary glycoprotein challenge of immu- were generated in vitro and demonstrated that following adoptive
nization-induced memory cells did not induce Tbet or CXCR3 transfer and then boosting recipient mice with peptide, these cells
expression in secondary Tfh cells (Fig. 6G, 6H). In contrast, follow- could acquire IFN-g expression. Unlike our study, the Tnp cells
ing secondary challenge with LCMV, protein immunization described by Iezzi et al. (39) were generated in vitro using a short
induced memory cells generated secondary Tfh cells that had ac- antigenic stimulation (24 h) without exogenous cytokines. In con-
quired Th1-like characteristics, including expression of both Tbet trast, our in vivo study generated day 7 Tnp effector cells and subse-
and CXCR3 (Fig. 6G, 6H). These data, combined with the restric- quent Tnp memory cells up to 84 d after initial immunization with
tion of Bcl6 expression and the GC Tfh phenotype (Fig. 6AF), adjuvanted protein. A study by Divekar et al. (38) first described the
demonstrate that Tnp-like memory Tfh cells can continue to existence of IL-2 and TNF-aproducing Tnp-like CD41 T cells
undergo Th1-like modifications to their Tfh cell programming upon from human PBMCs that were specific for protein-subunit vaccine
secondary responses to viral infection.
Ags and identified that protein immunization of mice led to induc-
tion of IL-21 IFN-gnegative CD41 T cells. However, Divekar
Discussion et al. (38) described these immunization-induced uncommitted IL-
CD41 T cell plasticity and commitment to specific Th lineages in 21 IFN-gnegative cells as precursor CD41 T cells and did not
response to vaccination is not well understood compared with the evaluate their memory potential and recall responses. In contrast,
highly polarizing environments of pathogenic infections. In this our findings suggest that Tnp cells are not necessarily precursors to
study, we investigated the differentiation of Ag-specific Th cells fol- Th1 and Th2 effector cells but instead are effector cells that express
lowing adjuvanted protein immunization versus acute viral infection. lineage-nonspecific cytokines (IL-2 and TNF-a) and lack lineage-
Our data demonstrate that protein immunization can induce effector specific Th differentiation programming. Our findings confirm that
CD41 T cell differentiation, as demonstrated by IL-2 and TNF-a primary protein immunization preferentially promotes the induction
cytokine production, without the polarization and differentiation into of Tnp cells instead of Th1, Th2, or Th17 cells. However, our study,
the Th1, Th2, or Th17 lineages, as the majority of cells did not to our knowledge, is novel in our use of an in vivo secondary anti-
express Tbet, Gata3, or RORgt. These Th lineagenonpolarized genic challenge of memory Tnp cells with protein immunization to
Tnp cells give rise to highly plastic Tnp memory cells that, upon confirm that Tnp cells do not undergo Th1 differentiation because
secondary antigenic challenge with LCMV infection, undergo Th of Ag reactivation alone, as boosting with glycoprotein and adjuvant
cell differentiation and acquire Th1 cell properties, including expres- did not result in Tbet and IFN-g induction. Our findings demon-
sion of IFN-g, Tbet, and CXCR3 (Fig. 7). In addition, although strate that in vivoprimed memory Tnp cells retain their phenotype
both acute LCMV infection and adjuvanted glycoprotein immuniza- throughout their secondary effector response following protein
tion induce differentiation of Tfh cells, immunization-induced Tfh immunization and do not upregulate expression of Th1-associated
IFN-g, Tbet, or CXCR3 unless viral infection is used for secondary response generated from LCMV-induced memory cells generated a
challenge. Thus, our findings indicate that Tnp cells are not neces- much lower frequency of secondary GC Tfh cells. These findings
sarily precursor cells to committed Th1 or Th2 cells. Instead, they indicate that Tfh and memory Tfh cells generated by protein immu-
are a distinct subset of effector and memory cells that are Th lineage nization maintain a higher degree of plasticity and greater propensity
uncommitted yet highly plastic cells that are capable of Th lineage toward the GC Tfh cell phenotype. In addition, as Tfh cells induced
polarization upon reactivation under polarizing conditions such as by viral infection acquire Th1-like properties, including upregulation
viral infection. of Tbet and CXCR3, there are likely cell-intrinsic alterations gener-
The differentiation of CD41 T cells into specialized lineages ated in Tfh cells during viral infection that result in the program-
upon antigenic challenge is controlled by multiple mechanisms, ming of subsequent Tfh memory cells to restrict Bcl6 expression.
including the expression of specific transcription factors and epige- Further studies will be needed to define whether such virus-induced
netic programming (50, 51). Tbet has been identified to regulate and programming in memory Tfh cells is directly regulated by Tbet
maintain Th1 commitment through direct and indirect mechanisms itself or through other genetic or epigenetic means of regulation.
to promote Th1 differentiation and repress other Th cell fates Although the heterogeneity of effector Th cell lineage commit-
(5256). Interestingly, our findings suggest that not only do Tnp ment and maintenance is fairly well described, our study provides
and Tfh cells differentiate into distinct populations following protein evidence that there are still unknown signals and mechanisms that
immunization but Tnp cells also preserve a plastic phenotype that is dictate Th cell commitment versus plasticity in vivo. Our findings
either retained after a secondary protein immunization or can be fur- also suggest that the identity of Th lineage cells during an immune
ther polarized to the Th1 lineage after viral rechallenge. Reactivation response should not simply be surmised or assumed by the relative
with Ag in adjuvant alone was not sufficient to induce Tnp memory ratios of specific IgG subtypes (IgG1, IgG2, etc.) or nominal fre-
cells to become Th1 cells. Although different types of antigenic quencies of Ag-specific cells that express lineage-defining cytokines.
challenges certainly involve variation in Ag quantity or persistence Instead, the majority of Ag-specific helper T cells induced by pro-
and TCR strength of signal, the varying degrees of Th cell plasticity tein vaccination may in fact be Tnp cells that do not actually express
versus lineage commitment and polarization are more likely regu- Th1-, Th2-, or Th17-defining cytokines, transcription factors, or
lated by the presence of pathogen-associated molecular patterns. functions. The degree of Tnp versus Th1/Th2/Th17 differentiation
Pathogen-associated molecular patterns drive a milieu of inflamma- following protein immunization is likely dependent upon the adju-
tory cytokines and conditions that specifically induce Th lineage vant used and whether highly polarizing/inflammatory cytokines are
specific transcription factors in combination with epigenetic pro- present (at sufficient levels) or not during CD4 T cell priming. As
gramming (16, 57). In addition, our study did not specifically ana- our findings indicate that adjuvanted protein vaccination preferen-
lyze the changes in polarizing signals and cytokines from innate and
tially induces plastic memory Tnp cells and less restriction of Tfh
Ag presenting cells, which likely differ in many ways between
cells, elucidating the mechanisms to generate these cellular pheno-
infection and immunization, and likely even differ depending on the
types could provide strategies to engage Tnp cells and further differ-
specific adjuvant formulation. Although most of the adjuvants used
entiate them into desired effector T cell subsets for the development
in our study induced only a minimal proportion of cells to undergo
of better vaccine and therapeutic approaches.
Th polarization, the disparate relative absence of Th1 polarization
following protein vaccination compared with viral infection is clear.
We did observe that the use of CFA, a highly inflammatory adju- Acknowledgments
vant that contains killed mycobacteria that is used in murine studies,
Flow cytometry data collection for this publication were supported by the
did induce strong Th17 and some Th1 cell differentiation; however, University of Utah Flow Cytometry Core Facility. We thank Dr. Carl Davis
CFA is highly reactogenic, making it unsuitable as a clinically appli- at Emory University for providing the 293A-sGP cell line and glycoprotein
cable adjuvant for human vaccination (58). This finding demon- purification protocol. We thank the National Institutes of Health Tetramer
strates that protein immunization, when combined with a highly Core Facility (Emory University) for providing I-Ab gp6677 tetramer.
inflammatory adjuvant, can indeed generate polarized effector helper
T cells, but that for most of the adjuvants used in our study, Tnp
cells were the predominant subset, and only a small fraction of Ag- Disclosures
specific cells were polarized Th1, Th2, or Th17 cells. The authors have no financial conflicts of interest.
In response to various types of infections, Tfh cells are able to
express low amounts of Th-specific transcription factors, including
Tbet, Gata3, or RORgt (9, 59). However, even despite their Tbet
References
1. Zhu, J., and W. E. Paul. 2008. CD4 T cells: fates, functions, and faults. Blood
expression, virus-induced memory Tfh cells are able to preferen- 112: 15571569.
tially recall Tfh-specific functions, with only a fraction of cells that 2. Szabo, S. J., S. T. Kim, G. L. Costa, X. Zhang, C. G. Fathman, and L. H.
gained partial plasticity toward the Th1 phenotype (9). Although Glimcher. 2000. A novel transcription factor, T-bet, directs Th1 lineage commit-
ment. Cell 100: 655669.
Bcl6 is critical for the maintenance of Tfh cell commitment (60), 3. Zheng, W., and R. A. Flavell. 1997. The transcription factor GATA-3 is neces-
Tbet may be important early in the differentiation of Tfh cells, par- sary and sufficient for Th2 cytokine gene expression in CD4 T cells. Cell 89:
587596.
ticularly in response to specific infections (54, 61, 62). Tbet and 4. Ivanov, I. I., B. S. McKenzie, L. Zhou, C. E. Tadokoro, A. Lepelley, J. J. Lafaille,
Bcl6 share a reciprocal relationship, as Tbet represses Tfh differenti- D. J. Cua, and D. R. Littman. 2006. The orphan nuclear receptor RORgammat
ation through recruitment of Bcl6 to upstream Tfh-associated genes directs the differentiation program of proinflammatory IL-171 T helper cells.
Cell 126: 11211133.
to promote Th1 commitment (54, 61), whereas Tbet can recruit 5. Johnston, R. J., A. C. Poholek, D. DiToro, I. Yusuf, D. Eto, B. Barnett, A. L.
Bcl6 to the Ifng locus to repress gene expression (61). In this study, Dent, J. Craft, and S. Crotty. 2009. Bcl6 and Blimp-1 are reciprocal and antago-
we observed that the Tnp-like Tfh cells generated by protein immu- nistic regulators of T follicular helper cell differentiation. Science 325:
10061010.
nization did not express Tbet but did exhibit a higher degree of 6. Nurieva, R. I., Y. Chung, G. J. Martinez, X. O. Yang, S. Tanaka, T. D. Matske-
Bcl6 expression and increased frequency of GC Tfh cells compared vitch, Y. H. Wang, and C. Dong. 2009. Bcl6 mediates the development of T fol-
with LCMV infection. Interestingly, secondary Tfh responses to gly- licular helper cells. Science 325: 10011005.
7. Yu, D., S. Rao, L. M. Tsai, S. K. Lee, Y. He, E. L. Sutcliffe, M. Srivastava, M.
coprotein immunization generated secondary effector Tfh cells con- Linterman, L. Zheng, N. Simpson, et al. 2009. The transcriptional repressor Bcl-6
taining a high frequency of GC Tfh cells, whereas the secondary directs T follicular helper cell lineage commitment. Immunity 31: 457468.
8. Crotty, S., R. J. Johnston, and S. P. Schoenberger. 2010. Effectors and memories: strong and persistent humoral and T cell responses against hepatitis B surface
Bcl-6 and Blimp-1 in T and B lymphocyte differentiation. Nat. Immunol. 11: antigen in healthy adult volunteers. Vaccine 26: 13751386.
114120. 31. Ott, G., G. L. Barchfeld, D. Chernoff, R. Radhakrishnan, P. van Hoogevest, and
9. Hale, J. S., B. Youngblood, D. R. Latner, A. U. Mohammed, L. Ye, R. S. G. Van Nest. 1995. MF59. Design and evaluation of a safe and potent adjuvant
Akondy, T. Wu, S. S. Iyer, and R. Ahmed. 2013. Distinct memory CD41 T cells for human vaccines. Pharm. Biotechnol. 6: 277296.
with commitment to T follicular helper- and T helper 1-cell lineages are gener- 32. Chu, R. S., T. McCool, N. S. Greenspan, J. R. Schreiber, and C. V. Harding.
ated after acute viral infection. Immunity 38: 805817. 2000. CpG oligodeoxynucleotides act as adjuvants for pneumococcal polysaccha-
10. Marshall, H. D., A. Chandele, Y. W. Jung, H. Meng, A. C. Poholek, I. A. Parish, ride-protein conjugate vaccines and enhance antipolysaccharide immunoglobulin
R. Rutishauser, W. Cui, S. H. Kleinstein, J. Craft, and S. M. Kaech. 2011. Differ- G2a (IgG2a) and IgG3 antibodies. Infect. Immun. 68: 14501456.
ential expression of Ly6C and T-bet distinguish effector and memory Th1 33. Kobayashi, H., A. A. Horner, K. Takabayashi, M. D. Nguyen, E. Huang, N. Cin-
CD4(1) cell properties during viral infection. Immunity 35: 633646. man, and E. Raz. 1999. Immunostimulatory DNA pre-priming: a novel approach
11. Pepper, M., A. J. Pagan, B. Z. Igyarto, J. J. Taylor, and M. K. Jenkins. 2011. for prolonged Th1-biased immunity. Cell. Immunol. 198: 6975.
Opposing signals from the Bcl6 transcription factor and the interleukin-2 receptor 34. Kulis, M., B. Gorentla, A. W. Burks, and X. P. Zhong. 2013. Type B CpG oligo-
generate T helper 1 central and effector memory cells. Immunity 35: 583595. deoxynucleotides induce Th1 responses to peanut antigens: modulation of sensiti-
12. Ansel, K. M., L. J. McHeyzer-Williams, V. N. Ngo, M. G. McHeyzer-Williams, zation and utility in a truncated immunotherapy regimen in mice. Mol. Nutr.
and J. G. Cyster. 1999. In vivo-activated CD4 T cells upregulate CXC chemokine Food Res. 57: 906915.
receptor 5 and reprogram their response to lymphoid chemokines. J. Exp. Med. 35. Martins, K. A. O., C. L. Cooper, S. M. Stronsky, S. L. W. Norris, S. A. Kwilas,
190: 11231134. J. T. Steffens, J. G. Benko, S. A. van Tongeren, and S. Bavari. 2015. Adjuvant-
13. Breitfeld, D., L. Ohl, E. Kremmer, J. Ellwart, F. Sallusto, M. Lipp, and R. enhanced CD4 T cell responses are critical to durable vaccine immunity. EBio-
F€orster. 2000. Follicular B helper T cells express CXC chemokine receptor 5, Medicine 3: 6778.
localize to B cell follicles, and support immunoglobulin production. J. Exp. Med. 36. Tighe, H., K. Takabayashi, D. Schwartz, G. Van Nest, S. Tuck, J. J. Eiden, A.
192: 15451552. Kagey-Sobotka, P. S. Creticos, L. M. Lichtenstein, H. L. Spiegelberg, and E.
14. Kim, C. H., L. S. Rott, I. Clark-Lewis, D. J. Campbell, L. Wu, and E. C. Butcher. Raz. 2000. Conjugation of immunostimulatory DNA to the short ragweed aller-
2001. Subspecialization of CXCR51 T cells: B helper activity is focused in a gen amb a 1 enhances its immunogenicity and reduces its allergenicity. J. Allergy
germinal center-localized subset of CXCR51 T cells. J. Exp. Med. 193: Clin. Immunol. 106: 124134.
13731381. 37. Longhi, M. P., C. Trumpfheller, J. Idoyaga, M. Caskey, I. Matos, C. Kluger,
15. Schaerli, P., K. Willimann, A. B. Lang, M. Lipp, P. Loetscher, and B. Moser. A. M. Salazar, M. Colonna, and R. M. Steinman. 2009. Dendritic cells require a
2000. CXC chemokine receptor 5 expression defines follicular homing T cells systemic type I interferon response to mature and induce CD41 Th1 immunity
with B cell helper function. J. Exp. Med. 192: 15531562. with poly IC as adjuvant. J. Exp. Med. 206: 15891602.
16. Crotty, S. 2014. T follicular helper cell differentiation, function, and roles in dis- 38. Divekar, A. A., D. M. Zaiss, F. E. Lee, D. Liu, D. J. Topham, A. J. Sijts, and
ease. Immunity 41: 529542. T. R. Mosmann. 2006. Protein vaccines induce uncommitted IL-2-secreting
17. Kaech, S. M., E. J. Wherry, and R. Ahmed. 2002. Effector and memory T-cell human and mouse CD4 T cells, whereas infections induce more IFN-gamma-
differentiation: implications for vaccine development. Nat. Rev. Immunol. 2: secreting cells. J. Immunol. 176: 14651473.
251262. 39. Iezzi, G., D. Scheidegger, and A. Lanzavecchia. 2001. Migration and function of
18. Seder, R. A., and R. Ahmed. 2003. Similarities and differences in CD41 and antigen-primed nonpolarized T lymphocytes in vivo. J. Exp. Med. 193: 987993.
CD81 effector and memory T cell generation. Nat. Immunol. 4: 835842. 40. Kim, C. H., D. J. Campbell, and E. C. Butcher. 2001. Nonpolarized memory T
19. Harrington, L. E., K. M. Janowski, J. R. Oliver, A. J. Zajac, and C. T. Weaver. cells. Trends Immunol. 22: 527530.
2008. Memory CD4 T cells emerge from effector T-cell progenitors. Nature 452: 41. Sad, S., and T. R. Mosmann. 1994. Single IL-2-secreting precursor CD4 T cell
356360. can develop into either Th1 or Th2 cytokine secretion phenotype. J. Immunol.
20. Coffman, R. L., A. Sher, and R. A. Seder. 2010. Vaccine adjuvants: putting 153: 35143522.
innate immunity to work. Immunity 33: 492503. 42. Wang, X., and T. Mosmann. 2001. In vivo priming of CD4 T cells that produce
21. Ellebedy, A. H., R. Nachbagauer, K. J. L. Jackson, Y. N. Dai, J. Han, W. B. interleukin (IL)-2 but not IL-4 or interferon (IFN)-gamma, and can subsequently
Alsoussi, C. W. Davis, D. Stadlbauer, N. Rouphael, V. Chromikova, et al. 2020. differentiate into IL-4- or IFN-gamma-secreting cells. J. Exp. Med. 194:
10691080.
Adjuvanted H5N1 influenza vaccine enhances both cross-reactive memory B cell
43. Oxenius, A., M. F. Bachmann, R. M. Zinkernagel, and H. Hengartner. 1998.
and strain-specific naive B cell responses in humans. Proc. Natl. Acad. Sci. USA
Virus-specific MHC-class II-restricted TCR-transgenic mice: effects on humoral
117: 1795717964.
and cellular immune responses after viral infection. Eur. J. Immunol. 28:
22. Galli, G., K. Hancock, K. Hoschler, J. DeVos, M. Praus, M. Bardelli, C. Mal-
390400.
zone, F. Castellino, C. Gentile, T. McNally, et al. 2009. Fast rise of broadly
44. Kulkarni, R. R., M. A. Rasheed, S. K. Bhaumik, P. Ranjan, W. Cao, C. Davis, K.
cross-reactive antibodies after boosting long-lived human memory B cells primed
Marisetti, S. Thomas, S. Gangappa, S. Sambhara, and K. Murali-Krishna. 2014.
by an MF59 adjuvanted prepandemic vaccine. Proc. Natl. Acad. Sci. USA 106:
Activation of the RIG-I pathway during influenza vaccination enhances the ger-
79627967.
minal center reaction, promotes T follicular helper cell induction, and provides a
23. Galli, G., D. Medini, E. Borgogni, L. Zedda, M. Bardelli, C. Malzone, S. Nuti, S.
dose-sparing effect and protective immunity. J. Virol. 88: 1399014001.
Tavarini, C. Sammicheli, A. K. Hilbert, et al. 2009. Adjuvanted H5N1 vaccine
45. Serre, K., E. Mohr, K. M. Toellner, A. F. Cunningham, S. Granjeaud, R. Bird,
induces early CD41 T cell response that predicts long-term persistence of protec- and I. C. MacLennan. 2008. Molecular differences between the divergent
tive antibody levels. Proc. Natl. Acad. Sci. USA 106: 38773882. responses of ovalbumin-specific CD4 T cells to alum-precipitated ovalbumin
24. Huleatt, J. W., A. R. Jacobs, J. Tang, P. Desai, E. B. Kopp, Y. Huang, L. Song, compared to ovalbumin expressed by Salmonella. Mol. Immunol. 45: 35583566.
V. Nakaar, and T. J. Powell. 2007. Vaccination with recombinant fusion proteins 46. Bing, S. J., C. Lyu, B. Xu, W. S. Wandu, S. J. Hinshaw, Y. Furumoto, R. R.
incorporating Toll-like receptor ligands induces rapid cellular and humoral Caspi, M. Gadina, and I. Gery. 2020. Tofacitinib inhibits the development of
immunity. Vaccine 25: 763775. experimental autoimmune uveitis and reduces the proportions of Th1 but not of
25. Kasturi, S. P., M. A. U. Rasheed, C. Havenar-Daughton, M. Pham, T. Legere, Th17 cells. Mol. Vis. 26: 641651.
Z. J. Sher, Y. Kovalenkov, S. Gumber, J. Y. Huang, R. Gottardo, et al. 2020. 3M- 47. Dube, J. Y., F. McIntosh, J. G. Zarruk, S. David, J. Nigou, and M. A. Behr. 2020.
052, a synthetic TLR-7/8 agonist, induces durable HIV-1 envelope-specific Synthetic mycobacterial molecular patterns partially complete Freund’s adjuvant.
plasma cells and humoral immunity in nonhuman primates. Sci. Immunol. 5: Sci. Rep. 10: 5874.
eabb1025. 48. Skepner, J., M. Trocha, R. Ramesh, X. A. Qu, D. Schmidt, E. Baloglu, M.
26. Khurana, S., W. Chearwae, F. Castellino, J. Manischewitz, L. R. King, A. Honor- Lobera, S. Davis, M. A. Nolan, T. J. Carlson, et al. 2015. In vivo regulation of
kiewicz, M. T. Rock, K. M. Edwards, G. Del Giudice, R. Rappuoli, and H. Gol- gene expression and T helper type 17 differentiation by RORgt inverse agonists.
ding. 2010. Vaccines with MF59 adjuvant expand the antibody repertoire to Immunology 145: 347356.
target protective sites of pandemic avian H5N1 influenza virus. Sci. Transl. Med. 49. McKinstry, K. K., T. M. Strutt, B. Bautista, W. Zhang, Y. Kuang, A. M. Cooper,
2: 15ra5. and S. L. Swain. 2014. Effector CD4 T-cell transition to memory requires late
27. Lefeber, D. J., B. Benaissa-Trouw, J. F. Vliegenthart, J. P. Kamerling, W. T. Jan- cognate interactions that induce autocrine IL-2. Nat. Commun. 5: 5377.
sen, K. Kraaijeveld, and H. Snippe. 2003. Th1-directing adjuvants increase the 50. Murphy, K. M., and S. L. Reiner. 2002. The lineage decisions of helper T cells.
immunogenicity of oligosaccharide-protein conjugate vaccines related to Strepto- Nat. Rev. Immunol. 2: 933944.
coccus pneumoniae type 3. Infect. Immun. 71: 69156920. 51. Wilson, C. B., E. Rowell, and M. Sekimata. 2009. Epigenetic control of T-helper-
28. Leroux-Roels, I., F. Roman, S. Forgus, C. Maes, F. De Boever, M. Drame, P. Gil- cell differentiation. Nat. Rev. Immunol. 9: 91105.
lard, R. van der Most, M. Van Mechelen, E. Hanon, and G. Leroux-Roels. 2010. 52. Hwang, E. S., S. J. Szabo, P. L. Schwartzberg, and L. H. Glimcher. 2005. T
Priming with AS03 A-adjuvanted H5N1 influenza vaccine improves the kinetics, helper cell fate specified by kinase-mediated interaction of T-bet with GATA-3.
magnitude and durability of the immune response after a heterologous booster Science 307: 430433.
vaccination: an open non-randomised extension of a double-blind randomised 53. Lazarevic, V., X. Chen, J. H. Shim, E. S. Hwang, E. Jang, A. N. Bolm, M.
primary study. Vaccine 28: 849857. Oukka, V. K. Kuchroo, and L. H. Glimcher. 2011. T-bet represses T(H)17 differ-
29. Moyer, T. J., Y. Kato, W. Abraham, J. Y. H. Chang, D. W. Kulp, N. Watson, entiation by preventing Runx1-mediated activation of the gene encoding RORgt.
H. L. Turner, S. Menis, R. K. Abbott, J. N. Bhiman, et al. 2020. Engineered Nat. Immunol. 12: 96104.
immunogen binding to alum adjuvant enhances humoral immunity. [Published 54. Nakayamada, S., Y. Kanno, H. Takahashi, D. Jankovic, K. T. Lu, T. A. Johnson,
erratum appears in 2020 Nat. Med. 26: 804.] Nat. Med. 26: 430440. H. W. Sun, G. Vahedi, O. Hakim, R. Handon, et al. 2011. Early Th1 cell differen-
30. Vandepapeliere, P., Y. Horsmans, P. Moris, M. Van Mechelen, M. Janssens, M. tiation is marked by a Tfh cell-like transition. Immunity 35: 919931.
Koutsoukos, P. Van Belle, F. Clement, E. Hanon, M. Wettendorff, et al. 2008. 55. Oestreich, K. J., and A. S. Weinmann. 2012. T-bet employs diverse regulatory
Vaccine adjuvant systems containing monophosphoryl lipid A and QS21 induce mechanisms to repress transcription. Trends Immunol. 33: 7883.
56. Zhu, J., D. Jankovic, A. J. Oler, G. Wei, S. Sharma, G. Hu, L. Guo, R. Yagi, H. 60. Alterauge, D., J. W. Bagnoli, F. Dahlstr€om, B. M. Bradford, N. A. Mabbott, T.
Yamane, G. Punkosdy, et al. 2012. The transcription factor T-bet is induced by Buch, W. Enard, and D. Baumjohann. 2020. Continued Bcl6 expression prevents
multiple pathways and prevents an endogenous Th2 cell program during Th1 cell the transdifferentiation of established Tfh cells into Th1 cells during acute viral
responses. Immunity 37: 660673. infection. Cell Rep. 33: 108232.
57. Kaiko, G. E., J. C. Horvat, K. W. Beagley, and P. M. Hansbro. 2008. Immunolog- 61. Oestreich, K. J., A. C. Huang, and A. S. Weinmann. 2011. The lineage-defining
ical decision-making: how does the immune system decide to mount a helper T- factors T-bet and Bcl-6 collaborate to regulate Th1 gene expression patterns. J.
cell response? Immunology 123: 326338. Exp. Med. 208: 10011013.
58. Vogel, F. R. 2000. Improving vaccine performance with adjuvants. Clin. Infect. 62. Weinstein, J. S., B. J. Laidlaw, Y. Lu, J. K. Wang, V. P. Schulz, N. Li, E. I. Her-
Dis. 30(Suppl 3): S266S270. man, S. M. Kaech, P. G. Gallagher, and J. Craft. 2018. STAT4 and T-bet control
59. Crotty, S. 2011. Follicular helper CD4 T cells (TFH). Annu. Rev. Immunol. 29: follicular helper T cell development in viral infections. [Published erratum
621663. appears in 2018 J. Exp. Med. 215: 999.] J. Exp. Med. 215: 337355.

Protein Immunization Induces Memory CD4+ T Cells That Lack TH Lineage Commitment

Uploaded by

Copyright:

Available Formats

Protein Immunization Induces Memory CD4+ T Cells That Lack TH Lineage Commitment

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Protein Immunization Induces Memory CD4+ T Cells That Lack TH Lineage Commitment

Uploaded by

Copyright:

Available Formats

Protein Immunization Induces Memory CD4

+ T Cells That Lack Th Lineage Commitment

Why The JI? Submit online.

Subscription Information about subscribing to The Journal of Immunology is online at:

The Journal of Immunology is published twice each month by

Protein Immunization Induces Memory CD4+ T Cells That Lack

Linda M. Sircy,1 Malia Harrison-Chau,1 Camille Leite Novis, Andrew Baessler,

FIGURE 2. Adjuvanted glycoprotein

You might also like