THE IMPACT OF THE NRF2 ACTIVATORS ARSENIC TRIOXIDE AND TERT BUTYLHYDROQUINONE ON
                                      B CELL FUNCTION
                                              By
                                     Luca Marius Kaiser
                                      A DISSERTATION
                                         Submitted to
                                Michigan State University
                        in partial fulfillment of the requirements
                                      for the degree of
                   Pharmacology and Toxicology –Doctor of Philosophy
                                            2022


                                             ABSTRACT
B cells produce immunoglobulins and are essential for robust humoral host defense. B cell
function is tightly regulated by interactions with other immune cells and antigens. B cells are
subject to rapid clonal expansion and massive production of antibodies, which subjects them to
oxidative stress. Many environmental toxicants, such as heavy metals and quinone-type food
preservatives, are known to alter the function of other immune cells, such as T cells. The role of
environmental toxicants on B cell function has received little investigation so far. Given the
pivotal role of antibodies in the context of infection, autoimmunity, and allergy, it is gravely
important to understand the impact environmental factors have on B cell function. The food
additive tert-Butylhydroquinone and the heavy metal arsenic trioxide have been shown by our
laboratory to alter T cell responses to infectious agents and alter cytokine production, partially
due to the signaling of nuclear factor erythroid 2-related factor 2 (Nrf2). Due to the reciprocal
relationship of B and T cells, this raises the question of the impact on B cell function. In these
studies, we describe the impact of the Nrf2 activation compounds arsenic trioxide and tBHQ on
human B cell function in vitro and on murine B cell function both in vivo and ex vivo. We isolated
human peripheral blood mononuclear cells and exposed them to arsenic trioxide before a
challenge with influenza A virus (IAV). After 96 hours, cells were harvested and analyzed using
flow cytometry. Cells that were exposed to ATO showed less immunoglobulin surface expression,
activation, and effector function. In rodents, we used an ex vivo activation model mimicking the
T cell- B cell interaction to assess the role of tBHQ on B cell activation in a physiologically
significant manner. In this model, we measured a decrease in IgG1 producing cells and a decrease
in a variety of activation markers, indicating that tBHQ indeed altered B cell activation.


Interestingly, we also observed a tBHQ dependent decrease in B cell clonal expansion and a
diminished inhibition of negative regulators such as CD267. Finally, we measured the effect of
tBHQ in vivo by exposing mice to 0.0014% tBHQ via their diet, a similar amount as found in
commercially used rodent chow. These mice were then sensitized to chicken ovalbumin using a
transdermal sensitization model mimicking the pathogenesis of human food allergies. Cells were
harvested from lymph nodes to determine whether tBHQ had an impact on the development of
allergies. The very early activation marker CD69 was inhibited, while CD25 was not. We also
measured an increase in CD80 and CD138, and most importantly, an increase in the key atopic
immunoglobulin IgE, both in serum and on the surface of B cells. Collectively, these data are the
first descriptions of a change in B cell function upon exposure to NRF2 activating toxicants. Given
the rise of atopic disease and the prevalence of infectious disease requiring immunoglobulin
mediated immunity, further characterization of the impact of the toxiom on B cell function is
urgently needed


                                      ACKNOWLEDGEMENTS
First and foremost, I want to thank Dr. Cheryl Rockwell for her guidance and professional and
collegial support over the last few years. My time in her laboratory coincided with some rather
eventful world events and despite those external challenges we managed to produce the works
presented in this thesis. In addition to the wealth of scientific knowledge, flow cytometric
expertise, and insights about the inner workings of the academic world I received, the many
laughs and jokes shared will not be forgotten. Dr. Yining Jin has supported me both professionally
and in our endeavor to sample all the exotic cuisine that Michigan has to offer. Dr. Rob Freeborn
is likely not the lab mate I deserved, but certainly the one I needed. Where else would I have
learned how to solve “Roblems”? Allison Boss, who received many rants, and returned laughs.
Always available when her unmatched experimental prowess was needed and willing to lend a
hand. Dr. Joe Zagorski, Dr. Alex Turley, Sheng Liu for teaching me valuable skills. Saamera Awali,
Stephanie Brocke, Gloria Yarandi, Elizabeth Mateo Pagan, Kimberly Guzman, and many other
great scientists in the making that crossed my path during the last couple of years. Dr Matt
Bernard and the flow cytometry core for the many hours teaching and optimizing my
experiments. Dr. Karen Liby and her lab for being a safe haven in times of moving stress and
always coming to the rescue with reagents. Dr. Anne Dorrance for guidance and mentorship in
navigating the grad school. Dr. Rick Neubig and Dr. Andy Amalfitano for giving perspectives on
navigating the balance between medical school and graduate school. The entire team of the
Pharmacology and Toxicology department for their invaluable contributions to our successes.
Dr. Brian Schutte and Michelle Volker of the DO PhD office for helping us navigate the
complicated ways in which medical school and graduate school intertwine.
                                                iv


A big thank you to my committee members, Dr. Karen Liby, Dr. Linda Mansfield, and Dr. Norbert
Kaminski for guiding me along the way and helping me develop my research skills.
In addition to all the professional acknowledgments above, I want to thank my parents Heidi and
Erhard for their support in my sometimes-daunting ideas, like moving 4100 miles for an 8-year
doctoral degree program. I also appreciate greatly that they raised me in such a way that I am
inclined to pursue these ideas. My grandma Maria and all the other members of my family who
supported me throughout the years. Mary, Jon, and the rest of my in-law family, who
immediately made me feel welcome and accepted among them. Most of all to my wife Natalie-
professionally, personally, in all aspects of life, she encourages and facilitates me being the best
possible version of myself.
A great thank you to all friends I made along the way, be it at the University, in my neighborhood,
in Barbershops, at rural rest stops, and wherever else you stumbled into my life – thank you for
being there.
Lastly, a solemn nod to those who are not with us anymore. Robert, Atteck, Tim, Mia, Adi – your
presence is missed, and I want to thank you for the impact you had on me.
                                                    v


                                                         TABLE OF CONTENTS
LIST OF FIGURES .............................................................................................................................. viii
LIST OF TABLES .................................................................................................................................. xi
LIST OF ABBREVIATIONS ................................................................................................................... xii
Introduction/Literature review .......................................................................................................... 1
         B cells – a view from 30,000 Feet ................................................................................. 2
         The etymology of the term B cell ................................................................................. 3
         Types of B cells ............................................................................................................ 6
         Overview of immunoglobulins ................................................................................... 16
         The role of B cells in Pathology .................................................................................. 23
         Toxicological studies in B cells ................................................................................... 24
         Nuclear factor erythroid 2-related factor 2 (Nrf2) ...................................................... 28
         Tert-butylhydroquinone ............................................................................................ 35
         The role of Nrf2 in different immune cells ................................................................. 42
Chapter 2: Arsenic trioxide inhibits the response of primary human B cells to influenza virus A
in vitro............................................................................................................................................... 47
         Abstract .................................................................................................................... 48
         Introduction .............................................................................................................. 50
         Methods.................................................................................................................... 52
         Results ...................................................................................................................... 55
         Arsenic trioxide decreases the percentage of IgG+ B cells but does not change the
         number of B cells ....................................................................................................... 57
         Discussion ................................................................................................................. 64
Chapter 3: Nrf2 dependent and independent effects of T-cell dependent B cell activation
ex vivo ............................................................................................................................................... 69
         Abstract .................................................................................................................... 70
         Introduction .............................................................................................................. 72
         Methods.................................................................................................................... 75
         Results ...................................................................................................................... 80
         Discussion ............................................................................................................... 101
Chapter 4: The Food Additive tert-Butylhydroquinone Increases Plasma IgE during transdermal
allergen sensitization in correlation with increased CD40 and CD138 expression on B cells ...... 106
         Abstract .................................................................................................................. 107
         Introduction ............................................................................................................ 108
         Methods and Materials ........................................................................................... 111
         Results .................................................................................................................... 115
         Discussion ............................................................................................................... 126
                                                                           vi


Chapter 5: The impact of age on murine B cell response to transdermal sensitization............... 129
       Abstract .................................................................................................................. 130
       Introduction ............................................................................................................ 131
       Methods.................................................................................................................. 132
       Results .................................................................................................................... 137
       Discussion ............................................................................................................... 150
Chapter 6: Conclusions .................................................................................................................. 153
       Discussion of findings .............................................................................................. 154
       Significance of findings ............................................................................................ 161
       Future directions ..................................................................................................... 163
BIBLIOGRAPHY ................................................................................................................................ 169
                                                                  vii


                                                    LIST OF FIGURES
Figure 1. Schematic overview of B cell development. .............................................................. 4
Figure 2. Schematic of the Germinal center reaction. ............................................................ 13
Figure 3. Antibody structure. ................................................................................................ 17
Figure 4. Simplified scheme of Nrf2 activation by tBHQ. ....................................................... 32
Figure 5. Chemical Structure of tBHQ. ................................................................................... 36
Figure 6. Schematic of the experimental setup. .................................................................... 49
Figure 7. ATO has a minor impact on viability and cell size of human PBMCs. ....................... 56
Figure 8. Surface IgG is decreased in the presence of ATO while B cell populations do not
significantly change. ............................................................................................................. 58
Figure 9. Expression of CD25 decreases with ATO treatment. ............................................... 60
Figure 10. ATO decreases the expression of CD80 but not CD86. ........................................... 61
Figure 11. ATO decreases the expression of CD267 and CD22. .............................................. 63
Figure 12. Viability changes are seen at 1µM and 5µM of tBHQ. ........................................... 81
Figure 13. The number of CD19+ B cells is decreased by increasing concentrations of tBHQ. . 83
Figure 14. The median fluorescent intensity (MFI) of FSC of CD19+ B cells decreases at higher
concentrations of tBHQ. ....................................................................................................... 85
Figure 15. CD19+ cells show decreased expression of the surface markers CD69 and CD25 when
exposed to high concentrations of tBHQ. .............................................................................. 87
Figure 16. The expression of CD86 decreases in the presence of tBHQ while CD80 remains
unchanged. ........................................................................................................................... 89
Figure 17. The number of IgG1-producing cells increases with activation but decreases in
presence of tBHQ in a concentration-dependent manner. .................................................... 91
Figure 18. CD267 (TACI) expression is decreased at 5µM tBHQ. ............................................ 93
Figure 19. The number of cells highly expressing MHC II decreases at 5µM tBHQ. ................. 95
                                                                viii


Figure 20. Genotypical difference in CD23 expression is mitigated by activation and restored at
0.5µM tBHQ. ........................................................................................................................ 97
Figure 21. Production of the B cell cytokine IL-6 increases with activation but decreases at 1µM
and 5µM tBHQ. ..................................................................................................................... 99
Figure 22. tBHQ diet increases Serum IgE after sensitization. .............................................. 116
Figure 23. The percentage of CD19+ cells increases with sensitization. ............................... 117
Figure 24. Sensitization increases CD69 but not CD25. ........................................................ 118
Figure 25. Sensitization does not increase the costimulatory molecule CD80 but does increase
CD86. .................................................................................................................................. 120
Figure 26. The percentage of B cells highly expressing the costimulatory molecule CD40 is
increased with tBHQ consumption, even in unsensitized mice. ........................................... 122
Figure 27. Sensitization has no effect on IgG1 but modestly increases IgE. .......................... 123
Figure 28. The overall percentage of plasmablasts is elevated in mice fed tBHQ, and the
percentage of IgE producing plasma cells increases with both tBHQ and sensitization. ....... 125
Figure 29. Spectral unmixing sample of the in-vivo panel. ................................................... 135
Figure 30. The percentage of CD19+ cells increases more strongly after sensitization in young
mice. .................................................................................................................................. 138
Figure 31. Contrasting patterns of induction for CD69 and CD25 after sensitization. ........... 140
Figure 32. CD80 expression does not change with sensitization independent of age. .......... 142
Figure 33. CD86 is increased with sensitization and the amplitude of this effect is inversely
correlated with age. ........................................................................................................... 143
Figure 34. The percentage of IgE-producing CD19+ cells increases with sensitization in the
younger groups but not in mice who start sensitization at 16 weeks................................... 145
Figure 35. IgG1-producing CD19+ cells increase in mice that start sensitization at 8 weeks, but
not in younger or older mice............................................................................................... 146
Figure 36. The percentage of plasmablasts increases with sensitization and with dietary tBHQ
independently in younger mice, but not in older mice. ....................................................... 148
                                                                    ix


Figure 37. The percentage of plasmablasts producing IgE is increased with both sensitization
and dietary tBHQ in younger mice, and only with sensitization in mice 8 weeks old at first
sensitization. ...................................................................................................................... 149
                                                                  x


                                         LIST OF TABLES
Table 1. Overview of B cell subtypes. ...................................................................................... 7
Table 2. Overview of Immunoglobulin classes. ...................................................................... 22
Table 3. Fluorescent antibodies used in Chapter 2 ................................................................ 54
Table 4. Fluorescent antibodies used in Chapter 3. ............................................................... 78
Table 5. Fluorescent Antibodies used in Chapter 4. ............................................................. 113
Table 6. Fluorescent antibodies used in Chapter 5. ............................................................. 134
                                                    xi


               LIST OF ABBREVIATIONS
15d-PGJ2 15-deoxy-Δ12,14-prostaglandin J2
AD       atopic dermatitis
AID      Activation-induced deaminase
ALI      acute lung injury
AP-1     activator protein 1
APC      antigen-presenting cell
ARE      Antioxidant Response Element
ATO      Arsenic Trioxide
B cell   Bursa cell
BALF     broncheoalveolar lavage fluid
Bcl-6    B cell lymphoma 6
BCR      B cell receptor
BHT      butyl hydroquinone
bTRCP    β-transducin repeat-containing protein
BZIP     basic leucine zipper
CD       Cluster of differentiation
CREB     cAMP response element protein
CXCR5    C-X-C chemokine receptor 5 (CD185)
DAI      daily allowable intake
DC       Dendritic cell
DZ       Dark Zone
                           xii


FDC   Follicular Dendritic cell
FEV1  Forced exhalatory volume 1
FO B  Follicular B cell
GC    Germinal center
GSK3  glycogen synthase kinase 3
GSTP1 glutathione S-transferase Pi-1
HeLa  Helga Lacks derived cells
HIP   hypoxia-induced protein
HSC   Hematopoietic stem cell
HSC   Hematopoietic stem cell
HSC   Hematopoietic stem cells
IAV   Influenza A Virus
IFN-γ Interferon-gamma
IgM   Immunoglobulin M
ILB   Innate like B cell
ILC   innate lymphoid cell
Keap1 Kelch ECH-associated protein 1
LPS   lipopolysaccharide
LZ    Light Zone
MHC   Major Histocompatibility Complex
NAT2  N-acetyltransferase 2
NF-κB Nuclear Factor Kappa light chain enhancer B
                         xiii


NKT    Natural Killer T cell
NOAEL  no observed adverse effect level
NQO1   NAD(P)H-quinone oxidoreductase 1
NRF2   Nuclear factor erythroid 2-related factor 2
PARP1  poly(ADP-ribose)polymerase-1
PC     Plasma cell
PD-1   programmed cell death protein 1 (CD279)
PERK   protein kinase RNA-like endoplasmic reticulum kinase
pIgR   Polyimmunoglobulin-Receptor
PKC    protein kinase c
RAG2   Recombination activating gene 2 protein
ROS    Reactive oxygen species
S1P    sphingosine-1-phosphate
SHM    Somatic Hypermutation
SLE    systemic lupus erythematosus
SLO    secondary lymphoid organ
SNP    single nucleotide polymorphism
SUMO   Small ubiquitin-like modifier
T cell Thymus cell
tBHQ   tert-butylhydroquinone
tBQ    2-tert-butyl-1,4-benzoquinone
TCR    T cell receptor
                          xiv


TFH   T follicular helper cell
TNF-α Tumor necrosis factor alpha
Trx2  thioredoxin 2
VLRA  Variable lymphocyte receptor A
VLRB  Variable lymphocyte receptor B
                         xv


Introduction/Literature review
               1


B cells – a view from 30,000 Feet
B cells are the prime mediator of humoral immunity. Their core function in mammals is the
production of immunoglobulins of different classes. These immunoglobulins, or antibodies, are
small proteins that will bind to specific epitopes on a biological specimen, typically on bacteria,
fungi, viruses, or helminths. Immunoglobulins are also involved in pathological processes, such
as autoimmunity, allergy, or physical dysfunction through the sheer mass of immunoglobulin
fragments. B cells are also professional antigen-presenting cells (APC) that can take up, process,
and present antigen on both MHC I and MHCII molecules.
                                                 2


The etymology of the term B cell
B cells are named after the Bursa of Fabricius, an organ found near the cloaca of birds.
Hieronymus Fabricius ab Aquapente mentioned this organ first in lectures given at Padova in
1621 [1]. The Latin term bursa derives from the Greek word βύρσα [byrsa], which can be
translated as “purse, bag or sac of leathery material”. It was not until 1952 that Bruce Glick
discovered in studies at The Ohio State University that surgical removal of the bursa led to the
loss of immunoglobulin production in the birds that underwent this procedure [2]. The cells
responsible for antibody production were thus dubbed B cells as in “Bursa cells”. Later anatomical
studies indeed revealed that the bursa is indeed a lymphoid organ [3]. Mammals in general, and
humans specifically, do not have any organ that could be considered homologous where B cell
development occurs in the bone marrow. Fortunately, the acronym remains the same, though
considering B cells as “Bone marrow cells” would likely be unacceptably imprecise and confusing
to the wider scientific community.
B cell development
Broadly speaking, all organisms possess some kind of immunological system. Bacteria and
protozoa possess multiple mechanisms that provide defense against other microbes [4], and
some of the applications derived from those mechanisms have revolutionized biotechnology [5].
Whereas these systems are mostly unspecific in lower life forms, vertebral organisms have
evolved to a system that is selective for molecular patterns (antigens) on a cellular level.
Of the cell-based immunities, one can broadly differentiate between the VLRA/VLRB system and
the TCR/BCR system. The divergence of these two systems has been pinpointed to the
                                                3


development of jaws. Our closest relatives that use the VLRA/VLRB system in opposition to the
HLA-based system of mammal adaptive immunity are jawless lampreys and hagfish [6-8] .
B cells in mammals, specifically in humans or mice, are derived from hematopoietic stem cells
(HSC). These are derived from the embryonic aorta-gonado-mesonephros and thus mesodermal
tissue [9], residing in the bone marrow. A broad overview of B cell development is provided in
Fig. 1.
Figure 1. Schematic overview of B cell development. Boxes demarcate surface molecules that
can be used to distinguish B cell state. B cells undergo development in the Bone marrow, starting
as CD34+ hSC, gradually differentiating into immature B cells, which will disseminate to
peripheral tissues and secondary lymphoid organs. Graphic By Mikael Häggström, MD, used with
permission.
 The first step towards B cell fate is the rearrangement of the D and J regions of the heavy chain
section (Dh and Jh) in the immunoglobulin gene, leading to a vastly increased variety of
immunoglobulin gene variants [10]. At this stage, the cell is considered a pre-B cell. The second
                                                   4


step during the pre-B cell period is the rearrangement of the Vh region and the rearrangement of
the light chain V and J regions (VL and JL) This V(D)J recombination process is mediated by the
Recombination activating gene 2 protein (Rag2)[11], which is equally important in T cell
development and used as a genetic target to generate B and T cell-deficient mice. The next step
in development is the rearrangement of the µ-heavy chains and the κ and λ light chains, which
ultimately enables the pro-B cell to express an IgM molecule on its surface [12]. At this stage, the
cell is considered an immature B cell and leaves the bone marrow to circulate into the spleen and
secondary lymphoid organs (SLO)[10]. These immature B cells then reside in SLO as follicular or
marginal zone B Cells (FO B cell and MZB cell, respectively) or recirculate. They will remain in this
state until they are activated.
                                                 5


Types of B cells
B cells can be classified into different subtypes. There are some significant differences between
murine and human cells, and some of those classifications tend to overlap. The subtypes of B
cells remain somewhat debated; the following classification is an attempt to synthesize different
currently held viewpoints in the literature. Table 1 provides an overview of the most common B
cell subtypes.
                                                  6


Table 1. Overview of B cell subtypes. Follicular B cells, Marginal Zone B cells and B-1 B cells differ
in function and surface marker expression. Adapted from: DOI.ORG/10.1016/J.JAU
Marginal zone B cells
Marginal zone B cells derive their name from being present in the marginal zone of the spleen.
This location is particularly well perfused with arterial blood, and thus high concentrations of
sphingosine-1-phosphate (S1P) are found. MZ B cells use their S1P receptors to seed these
particular areas [13]. MZ B cells are the main producer of IgM and natural autoantibodies (see
chapter on Antibodies) [14]. Marginal Zone B cells (MZB) have a strong innate-like function, as
they can become activated independently of T cell help [15]. This activation is dependent on
                                                 7


motifs in the antigens. Murine MZB cells are known to be potently activated by LPS [16-18],
whereas the T cell-independent activation of human MZB cells is more controversial. It has been
reported that human B cells, in contrast to murine B cells, do not react to LPS activation [17].
Newer studies have shown that human B cells, specifically MZB cells, express a wide array of TLR
receptors, namely TLRs 1, 6, 7, 9, and 10 [19, 20], indicating that TLR activation can facilitate a
polyclonal antibody response in MBZ cells [21]. Overall, MZB cells can be considered innate-like
B cells, capable of bridging the innate and adaptive immunity and potentially implicated in
pathology due to autoantibodies [22, 23].
B1 B cells / Innate-like B cells
B1B B cells are commonly found in mice. B1 B cells are somewhat similar to the MZ B cells in
function, though they differ greatly in location – B1 B cells are often found in the periphery, and
they line the peritoneal and pleural cavities [24]. Their localization is largely governed by Cxcl13
[24]. B1 cells do not undergo GC reactions and/or somatic hypermutation (SMH), and their BCR
repertoire is more limited compared to FO B cells [25]. B1B cells are implicated in the defense
against helminths such as nippostrongylus brasiliensis [26, 27], spirochetes such as borrelia
hermsii [28, 29], and Enterobacter such as e. cloacae [30]. The existence of B1 B in human cells
has been somewhat controversial [26], but one can appreciate the great functional similarity
between murine B1B cells and human innate-like, CD5+ B cells [31-33].
Follicular B cells and the germinal center reaction
Follicular B cells are the circulating and lymph node-resident B cells [34]. They are typically found
in the cortex region of lymph nodes, Peyer’s Patches, or the white pulp of the spleen, and are
                                                  8


instrumental to the generation and expansion of SLOs and follicular dendritic clusters via their
secretion of alpha-lymphotoxin [35, 36]. Clusters of naïve FO B cells and follicular dendritic cells
(FDC) are coined primary follicles [37] and may give rise to germinal centers (GCs). These areas
are particularly close to resident T cells, facilitating the T cell interaction on which FO B cells
depend. The interaction between FO B cells and T follicular helper cells (TFH cells) is subject to
tight regulation and results in the formation of high-affinity antibodies and isotype switching.
This process can be repeated sequentially to increase the affinity of immunoglobulins. A
schematic overview of the germinal center reaction is seen in Fig. 3.
 The first step of the germinal center reaction is the recognition of a suitable antigen by the FO B
cell, either presented by an FDC or a soluble antigen that is taken up directly by the FO B cell [38-
41]. This activation sets off a chemotactic sequence: The FO B cells upregulate CD197 (CCR7, C-C
chemokine receptor type 7), which recognizes cysteine motifs in its ligands, CCL19 and CCL21
[42]. These are highly expressed in the adjacent T cell zones, thus the activated FO B cell moves
to the periphery of the emerging GC (hereby constituting the Light zone of the budding GC) and
mingles with the resident TFH cells[43]. This movement happens at surprisingly high speeds of
up to 6µm/min [44]. This movement facilitates the “speed dating” between TFH and B cells – the
B cell will present fragments of the antigen that activated it on its MHC II molecule, and TFH cells
browse through the offered antigens to see if they match the specificity of their TCR. If the
antigens indeed match, the T cell will offer costimulatory signals and form a so-called T-B cell
entanglement [44-46].
                                                   9


The interaction between FOB cell and TFH cell is initiated by TCR-MHCII-antigen peptide
complexes but needs costimulatory signals such as CD40 on the B cell and CD40L on the T cell
[47]. This signal activates non-canonical NF-Kappa B signaling and an array of TRAF signals within
the B cell [48]. These signals subsequently lead to clonal expansion and activation of the FOB cell,
effectively starting the germinal center reaction [46]. At this point, the FOB cells will either
commit to the full germinal center reaction leading to high-affinity antibodies or leave the follicle
and become short-lived, low-affinity extrafollicular plasma cells [49-51].             Additionally,
extrafollicular, T cell-independent memory B cells have been described recently [52]. This
explains the comparably rapid first wave of low-affinity antibodies in response to an infection,
followed by a larger wave of high-affinity antibodies shortly thereafter [52].
FOB cells that commit to the GC now start expressing high amounts of Activation-induced
cytidine deaminase (AID), an enzyme that is essential for the GC process [53]. AID is a cytidine
deaminase that converts cytidine bases in the immunoglobulin gene variable regions to uracil
bases, effectively turning stable guanine-cytidine matches into mismatched uracil-guanine
pairings. These new mismatches are subject to repair processes, which introduce mutations to
the variable regions, at a rate about a million times higher than the baseline mutation rate of the
genome [54, 55].
This process is termed somatic hypermutation (SHM) and introduces great variability to the
variable region of the immunoglobulin gene of each nascent B cell. AID is also implicated in the
process of Isotype switching, in which a B cell changes its immunoglobulin constant region, by
introducing double-stranded DNA breaks. This was generally believed to be part of the GC
                                                  10


reaction, however, newer studies suggest that this may already be happening in pre-GC B cells,
as one of the first actions of AID [56]. Isotype switching changes the BCR immunoglobulin type
from the resting IgD/IgM to an IgG, IgA, or IgE type BCR, depending on the signals given during
the GC.
SHM and clonal expansion take place in the dark zone (DZ) of the GC. B cells that are actively
undergoing these processes are dubbed centroblasts. In the light zone, the B cells present their
newly mutated BCR/antigen complexes to TFH cells, which then assess the affinity of the BCR for
the antigen. This process is described both in mice [57] and humans [58, 59].
All light zone B cells are in a pro-apoptotic state, and only signals provided by the TFH cell will
prevent apoptosis and allow the B cell to either return to the dark zone for further affinity
maturation or to become a plasma cell [60]. This process is dubbed positive selection. SHM can
also lead to autoreactive immunoglobulins and thus potentially cause antibody-dependent
autoimmunity. Considerable debate over the existence of a putative negative selection during
the GC reaction is ongoing, but a consensus emerges that it is likely to be an infrequent or
apoptosis-independent event [46], e.g. through BCR mutation redemption [61]. BCR redemption
is a complex process that involves deactivation of autoreactive BCRs by covering them with
carbohydrates, and subsequent re-entry into the GC reaction and further mutation to change the
VDJ-product to one that does not bind to auto-antigens [62].
GC B cells, if not undergoing apoptosis, are eventually exported from the germinal center, either
as plasmablasts or memory B cells. Plasma cells (PC) tend to have a higher affinity for the antigen,
                                                 11


while memory B cells have a relatively lower affinity [46]. Plasmablasts typically tend to have high
expression levels of the transcription factor Blimp-1[60, 63].
                                                 12


Figure 2. Schematic of the Germinal center reaction. Naïve B cells get activated by T cells specific
to an antigen and enter the dark zone. Somatic hypermutation and clonal expansion introduce
mutations to IG genes. B cells enter the light zone and test their affinity. Potential outcomes are
apoptosis (low affinity), reentering the dark zone for further affinity maturation, or Plasmablast
or Memory B cell fate, which then leave the lymph node. Graphic created in Biorender, own work.
Memory B cells
Memory B cells are antigen-experienced B cells that underwent class switching and SHM during
a germinal center reaction. An exception to that rule are IgM+ memory B cells, which developed
independently of T cells (See B1-B cells). Memory B cells can survive in peripheral tissues for many
                                                 13


years and repeatedly respond to the same antigen if exposure occurs [64]. If they discover an
antigen specific to their BCR, they can either traffic back to a SLO and re-enter the germinal center
reaction or differentiate into a plasma cell [65]. The factors that decide the fate of GC B cells,
once they reach a certain receptor affinity, whether to differentiate into a plasma blast or
memory B cell, have been elusive. A generally higher affinity in the case of plasmablasts has often
been discussed [46]. Other factors that are thought to distinguish between the memory B cell
and plasma cell fates are IL-24 signaling,         NF-κB activity, as well as the Blimp1-related
transcription factor Bach2 [66]. Blimp-1 and Bach2 signaling is a inhibitor for memory B cell fate
and favors plasma cell differentiation, while PAX5 signaling is required for memory B cell
differentiation [67, 68].
Plasmablasts/ Plasma cells
Plasma cells are the major antibody producers among B cells. After affinity maturation in the GC
reaction, B cells with high affinity receptors express high levels of the transcription factor IRF4
and differentiate into plasma cells [66]. Plasma cells leave the lymph node and preferentially
migrate to the bone marrow, where they continuously secrete antibodies. Fully differentiated
plasma cells focus solely on the production of antibodies, while immature plasma cells (called
plasmablasts) also undergo cell division [67].
Associated cells – TFH cells and follicular dendritic cells
TFH cells are a subtype of CD4 T cells and essential partners in the germinal center reaction [59].
They depend on IL-6, IL-21 and ICOS signaling to commit to the TFH fate [69, 70]. These cells are
defined as cells expressing TCR, CD4, CXCR5, Bcl-6, and PD-1. TFH cells participate in the GC
                                                  14


reaction by providing costimulatory signals such as CD40L, CD28, and OX40 [71] to B cells as well
as by producing cytokines that determine the resulting antibody isotype and possibly the fate of
the resulting B cell [65]. TFH have been classified by the secretion of cytokines into functionally-
distinct subtypes including TFH1, TFH2, TFH13 and TFH17 cells [72]. The interaction of TFH with
B cells via PD-1 binding is vital for the generation of long-living plasma cells[73]. CD40L-CD40
interactions are essential for the initial formation of GCs and uncontrolled signaling of this
pathway can lead to a lupus-like disease [74].
Follicular dendritic cells (FDCs) are functionally dendritic cells, but unlike other DCs, are not
derived from HSCs [75]. They form the core of the follicle, where they present antigens to B cells,
as a checkpoint for affinity [76]. FDCs canonically do not carry MHCII molecules, however, they
have been reported to express microvesicles with MHCII that were passively acquired at their
surface [77]. This suggests that their role in phagocytosis and MHCII-restricted antigen
presentation is negligible. Instead, they rely on an intricate mechanism where non-antigen
specific B cells capture blood or lymph borne antigen, which has been opsonized by other innate
immune cells, and transfer these antigens to FDCs [40, 41]. They are also involved in clearing
debris from the follicle and SLOs, a mechanism that in rodents has been shown to prevent
autoimmune processes [78].
                                                 15


Overview of immunoglobulins
Immunoglobulin production is one of the chief functions of B cells. Immunoglobulins are a
considerable part of the protein fraction in serum and are found in most bodily fluids, including
saliva, blood, mucus, and feces [59]. Ig molecules can be found on cell surfaces of B cells or as
soluble proteins. Human and murine antibodies are Y-shaped proteins of approximately 150 kDa
and 10nm in length, consisting of 4 polypeptide chains [79]. The 4 polypeptide chains can be
further classified into 2 light chains and two heavy chains, and each chain has both one or more
constant and one variable section (Fig 3). Functionally, the antibody molecule can be classified
into two regions: The variable region (Fab) and the constant region (Fc).
The variable regions contain paratopes, which are the areas of an antibody that bind to epitopes
on an antigen. The variable region thus determines the binding affinity to epitopes on antigens,
while the constant region determines the isotype of the antibody and provides the structural
integrity of the protein [80]. While there is immense variability in the Fab region, due to VDJ
recombination and SHM, the Fc region has 5 different possible isotypes, of which some have
further subtypes. These isotypes are IgD, IgM, IgG, IgA, and IgE [59]. Each one has a specific
biological function and differs from the other subtypes.
                                                 16


Figure 3. Antibody structure. Simplified cartoon of an immunoglobulin. The heavy chains are
shown in yellow and blue, the light chains in green and pink. The left antigen binding site is
circled. The heavy chains contain a variable domain (VH), followed by a constant domain (CH1),
a hinge region, and two more constant domain (CH2, CH3). The light chains have a variable
domain (VL) and a constant domain (CL). The CH2 and CH3 regions constitute he variable and
CH1, CL regions form a pair of antigen-binding fragments (Fab). Disulfide bonds between the
                                               17


Figure 3 (cont’d)
 chains are drawn as S-S. Their exact number and location vary for different isotypes. Used under
CC BY-SA 4.0, created by User TokenZero
IgD is one of the rarer Ig isotypes, expressed on immature and resting B cells, usually together
with IgM. IgD is also found in serum in very low concentrations, compromising less than 0.25%
of Ig proteins. It also has a comparably short serum half-life of ~2.8 days [81]. IgD does not seem
to be essential to B cell function, as IgD deficient mice have a largely normal B cell phenotype
[82]. IgD seems to have some functions in communication between basophils and B cells [83]. In
addition, serum IgE levels and the number of peripheral B cells are moderately reduced in IgD-
deficient mice [82]. Secreted IgD usually is found as a monomeric molecule[59].
IgM, which is co-expressed with IgD on resting B cells, B1B cells, and marginal zone B cells, is
usually the first antibody produced in large numbers in response to an antigen challenge, making
it a useful clinical indicator to detect recent diseases in contrast to an older resolved or chronic
pathogenic process. IgM can be translocated from the bloodstream into the gut, mucosal
surfaces, or breastmilk using the polyimmunoglobulin receptor (pIgR)[84]. While B cells that
produce IgM have undergone VDJ recombination, they may or may not have been subject to
SHM, and thus the average affinity of IgM molecules to target antigens is lower than other
isotypes, like IgA, IgG, and IgE [85]. IgM is also a potent opsonin, marking antigens tagged by IgM
for phagocytosis by macrophages and other innate immune cells. Uptake of these antigens via Fc
receptors on APCs can also lead to the increased presentation of the antigen on MHC molecules,
                                                    18


further amplifying the immune response to the antigen [59]. IgM typically forms pentamers
connected to each other using the J chain subunit [86].
IgA is the immunoglobulin most often found in mucus, inside the gut canal, and in saliva. It is also
present in breast milk and serum [59]. It is a weak opsonizer[59]. IgA is produced after a B cell
undergoes isotype switching. The interaction with TFH cell with B cells, as described previously,
is essential for IgA class switching.IL-5 and TGF- β1 are the key cytokines produced by TFH cells
that promote IgA production [87]. There is, however, a T-cell independent pathway to IgA
production [88]. IgA molecules are heavily glycosylated to be more resistant in the environments
in which they are often found [89]. IgA often forms a dimer that is held together by a J chain,
similar to IgM, but also by a crucially important secretory component. This protein is part of the
pIgR and cleaved off together with IgA dimers during the translocation of IgA to its destination
compartment [90]. IgA has two subforms, IgA1 and IgA2. They differ structurally, as IgA1 has the
classic structural components shared with the other antibody subtypes, while IgA2 differs as it is
not held together by disulfide bonds, but by non-covalent interactions instead[91]. While IgA is
not the most common Ig molecule class in the bloodstream, it is the most produced overall [59].
IgG is the immunoglobulin class most abundant in serum. IgG is an excellent opsonin and
neutralizing antibody [59]. The key cytokine released by TFH cells to induce an IgG fate is
interferon-gamma (IFN-γ)[92, 93]. IgG has 4 subclasses, IgG1, IgG2, IgG3, and IgG4, which are
biochemically almost homologous, but differ in function. IgG1 is the most common subclass but
can also be an intermediate in the class switching towards an IgE fate [94, 95], a process called
sequential switching. IgG1 is almost exclusively the result of the GC reaction and mostly targets
                                                 19


protein epitopes[92]. IgG2 can be generated in a T cell-independent manner in an extrafollicular
reaction, and often targets bacterial polysaccharides [96-98]. IgG3 antibodies tend to have a
stronger pro-inflammatory effector function and are therefore especially potent in fighting off
infectious pathogens [99]. They also have a remarkably lower half-life than the other IgG
subtypes, perhaps as a biological function to limit the time course of a strong inflammatory
reaction [100]. IgG4 antibodies are considered anti-inflammatory and tolerance-inducing [101].
IgG4 is the least common IgG subtype, comprising only about 5% of all IgG antibodies. It is a non-
opsonin, curiously, it can even prevent the formation of complement immune complexes using
its “Fab-arm exchange” [101, 102]. In this process, the IgG4 heavy chains disassociate from each
other and reassemble with another IgG4 half-molecule that underwent the same process,
yielding a bispecific antibody, which is unable to form complement immune complexes for the
lack of accessible Fc region [102].
IgE is implicated in the immune response to helminths and parasites, but also in allergy and atopic
disease. IgE is mostly a product of the GC reaction, but in rare instances can be generated
extrafollicularly[103]. IgE is found in either a high-affinity or low-affinity form. Low-affinity IgE
antibodies are associated with defense against helminths, as those pathogens do not elicit high-
affinity IgE antibodies [95, 104, 105]. A likely explanation is that helminths and protozoic parasites
often have carbohydrate-covered antigens at their outer surface, and often undergo antigenic
shift, explaining the need for low-affinity, less specific IgE [106]. This has been postulated as the
reason why people infected with helminths suffer less from allergies and anaphylactic shock, one
of the tenets of the hygiene hypothesis [107]. Only high-affinity IgE can prime and subsequently
                                                    20


facilitate the degranulation of mast cells, the crucial step in anaphylaxis [108-110]. The
production of high-affinity IgE is believed to require an intermediate class switch to IgG1, before
switching into IgE [105]. The longstanding dogma of IgE generation states that Th2 cells and IL-4
are essential factors in IgE generation [111-114]. Newer studies paint a more nuanced picture, in
which TFH cells secrete IL-4, IL-5, IL-13, and low amounts of IL-21 to induce IgE fate in the GC
reaction. Loss of IL-13 abrogated the IgE production in this peanut-antigen model [108, 110].
                                                 21


Table 2. Overview of Immunoglobulin classes. Depicted are 5 common classes of
Immunoglobulins and some of their key functions. Not depicted are subclasses. Graphic taken
with permission under CC-BY-SA 4.0, by unknown user.
                                             22


The role of B cells in Pathology
As the sole producer of antibodies, the role of B cells in pathology cannot be overstated. B cell-
derived antibodies are involved in defense against all infectious diseases, autoimmune processes,
atopic disease, and even in wound healing [115]. While protective against most infectious agents,
autoreactive antibodies and altered B cell phenotypes are found in autoimmune diseases such as
scleroderma [116], lupus [117], multiple sclerosis [118], allergic rhinitis[119], rheumatoid
arthritis [120], celiac disease [121], type 1 diabetes [122] and even potentially in type 2 diabetes
[123], among many other autoimmune processes. B cell inhibitors, such as rituximab (Rituxan),
ocrelizumab (Ocrevus), ofatumumab (Kesimpta), and obinutuzumab (Gazyva) are used for many
autoimmune pathologies[124], often off-label [125] as a last resort. The prominent role of B cells
in cancer immunology is beyond the scope of this thesis and covered in many excellent reviews
[126-129]. B cells themselves can be the source of malignancies, especially for lymphomas,
myelomas, and leukemia subtypes [58].
                                                  23


Toxicological studies in B cells
Given the pivotal role of B cells in physiology and pathology, they have received comparatively
little investigation in toxicological studies. Early studies in a chicken model showed that the
chemotherapeutic and immunosuppressant cyclophosphamide inhibits B cell formation and
humoral immunity even at low doses [130]. A common antihistamine, chlorpheniramine,
conversely induced an expansion of the B cell population in a rat model [131]. Benzo(a)pyrene, a
polycyclic aromatic hydrocarbon amplified the expression of phase two enzymes, but not AHR in
rainbow trout B cells [132]. The heavy metals mercury and lead have been shown to affect IgM
secretion in chicken-derived B cell lines [133, 134]. Interestingly, mercury decreased IgM
production [134], while lead induced an intracellular accumulation of IgM, suggesting a secretion
defect [133]. In human B cells, mercury modified the Lyn pathway [135]. While the physiological
implications are somewhat unclear, this modification has been suggested as a biomarker for
mercury exposure. Toluene diisocyanate exposure can change B cell phenotype with
pathophysiological consequences. B cells exposed to toluene diisocyanate can induce asthma
even if the exposure ceases, as evidenced by a murine adoptive transfer model [136]. Addressing
popular toxicological concerns over environmental radiation, a study examined whether 900mhz
GSM band waves influenced murine lymphocytes and found while there were minor, short-term
alterations in T cell cytokine secretion, B cells were not influenced by this form of radiation [137].
In human cell lines, different translational inhibitors, such as trichothecene mycotoxin, ricin, and
Shiga-toxin were evaluated for their cytotoxicity and found to be generally more potent in
mature B cells than immature ones, though some mycotoxins showed no discrimination between
                                                   24


maturity levels [138]. Studies in the mouse B cell line CH12IκBαAA showed that low doses of
hydrogen peroxide decreased Ig production in an NF-ΚB-dependent manner [139].
Large observational studies, albeit lacking in toxicological specificity, showed that exposure to a
variety of pesticides over 15 years decreased B cells, regulatory B cells, and plasmablasts in the
peripheral blood of Brazilian farmers, concurrent to an increase in serum IL-6 [140]. Another
human observational study in Germany showed that exposure to lowly chlorinated biphenyls
modestly decreases CD19+ B cells [141].
One of the major areas of toxicological investigation into B cell function has been over the
persistent organic pollutant 2,3,7,8-Tetrachlorodibenzodioxin (TCDD), colloquially referred to as
dioxin. Dioxin is a contaminant in the herbicide agent orange, and in concentrated form has been
used for political assassination attempts [142]. TCDD effects on B cells were described in birds
[143] and mice [144], reducing the Ig response [145]. TCDD’s effects on human cells via the aryl-
hydrocarbon receptor (AHR) were also shown. In vitro, most of those donor-derived cells showed
an impaired IgM response to CD40 stimulation [146]. TCDD interruption of the mitogen-activated
protein kinase (MAPK) and protein kinase B (AKT) pathway in primary human B cells led to
decreased expression of CD80, CD86, and CD69 after stimulation with CD40 [147]. Overall,
though AHR signaling is relatively well conserved between species, AHR stimulation by TCDD
shows species-specific effects [148]. While both mice and men have diminished IgM secretion
after TCDD exposure, the human impairment seems to be due to failure in IgM secretion, while
mice have reduced mRNA expression [149], and there is considerable variation between
individual humans regarding the magnitude of these effects [150]. Mechanistically, mouse TCDD
                                                 25


AHR signaling involves the BLIMP-1 pathway in CD40-dependent activation [151] and PAX5 in
TLR-mediated activation [152]. SerpinB has emerged as a key molecule in TCDD injury, with
SerpinB deficient mice having a more profound response to TCDD [153].
In humans, experiments with cells derived from CD34+ umbilical cord-derived cells have shown
impaired B cell lymphopoiesis upon exposure to TCDD with involvement of early B cell factor 1
(EBF1) and paired box 5 (PAX5) [154] and impaired B cell lymphopoiesis upon exposure to TCDD
[155]. Studies with peripheral blood human B cells have identified Src homology phosphatase 1
(SHP-1) as a mediator of B cell repression, potentially due to an interaction with Bcl-6 [156].
These effects seem to involve a reduction of STAT3 phosphorylation and are reversible by the
addition of IFN-γ [157].
Interestingly, CD5+ B cells, which are considered innate-like B cells, seem to be maybe even more
sensitive to TCDD in vitro, as measured by IgM suppression after TCDD challenged and basal AHR
expression [158]. In these cells, lymphocyte tyrosine kinase (LCK) and PD-1 seem to mediate
these effects [159]. In contrast, a recent study in mice suggested that follicular B cells are more
sensitive to TCDD and that B cell suppression by TCDD improves the outcomes in experimental
autoimmune encephalitis, a mouse model of multiple sclerosis, via IgG inhibition [160].
Similarly to the TCDD studies, Delta-9 hydro cannabinol causes a reduction in IgM secretion via a
decrease in STAT3 phosphorylation [161].
Very little is known about the impact of tBHQ on B cell function. Our group showed that tBHQ
increased IgM production by LPS-activated B cells ex vivo in a Nrf2-dependent manner [162]. In
                                                26


contrast, expression of CD69, CD25, CD22, and CD138 were decreased by tBHQ via a Nrf2-
independent mechanism.
                                           27


Nuclear factor erythroid 2-related factor 2 (Nrf2)
Discovered in 1994, Nrf2 is a transcription factor that is activated by oxidative stress [163]. Its
basic structure is part of the cap’n’collar (CNC) basic leucine zipper (BZIP) group. The evolution
of Nrf2 correlates with the rise of atmospheric oxygen levels and its first orthologues were found
in fungi 1.5 GA (~1.5 billion years) ago [164]. Nrf2 is ubiquitously expressed in mammalian cells
and highly conserved among species [165, 166]. In homeostasis, Nrf2 is bound to its repressor
protein, Keap-1, and the total amounts and ratios of Nrf2 and Keap1 differ between tissues [167],
which may hint at differing susceptibility of tissues to oxidative stress. Nrf2 was initially
discovered as a transcription factor binding to a locus that controlled the expression of the
hemoglobin beta-subunit, though it turned out that Nrf2 was not essential for the generation of
hemoglobin or survival in mice [165]. Since its initial discovery, several polymorphisms of Nrf2
have been described in humans [168].
Biochemical properties of Nrf2
Under homeostatic conditions, Nrf2 is bound to Keap-1 [167]. Keap 1 catalyzes the ubiquitination
of bound Nrf2 by E3 ubiquitin ligase (Fig 4A), directing it to the 26S proteasome for degradation
[169-171]. Under conditions of oxidative or electrophilic stress, Nrf2 induces the transcription of
an array of genes in the nucleus [172]. The precise mechanism of activation of Nrf2 remains
somewhat unclear, though it was evident that Nrf2 inducers alter sulfhydryl groups on Keap1
[173]. One mechanistic hypothesis of Nrf2 activation postulates that these changes to Keap1,
alter the binding interactions in the Keap1-Nrf2 complex (Fig 4B). The interaction of the Nrf2 DLG
complex with Keap1 (latch) is released, while the interaction of the Nrf2 ETGE complex (hinge)
                                                 28


remains stable. Thus, Nrf2 is still attached to Keap1, but these changes hinder the ubiquitination
and degradation of bound Nrf2 [174] [175, 176]. In this “hinge and latch” theory, freshly
synthesized Nrf2 does not bind to the already saturated Keap1-Nrf2 complexes, and instead
freely translocates to the nucleus. A competing, but very similar model suggests a conformational
change in Keap1, making the Nrf2-Keap1 complex unsuitable for polyubiquitination [177]. An
alternative mechanism for Nrf2 degradation involves the Neh6 domain of Nrf2, which contains a
redox-insensitive degron [178]. This area of Nrf2 can be phosphorylated by glycogen synthase
kinase 3 (GSK-3) [179], which makes it a binding target for beta-transducin repeat-containing
protein (bTRCP), which then, in turn, ubiquitinates cytosolic Nrf2 and targets it for degradation
[180]. This degradation pathway is Keap1 independent.
However derived, free Nrf2 forms a heterodimeric complex with small Maf proteins, which
induces phase II enzymes through the antioxidant response element (ARE) [181-183]. Complexes
of Nrf2 with jun proteins have also been reported, inducing a similar group of genes [184].
Non-canonical mechanisms of Nrf2 signaling involve defects in autophagy, where p62 proteins
aggregate to Keap1, inhibiting Nrf2 degradation [185-187]. Nrf2 itself can induce the production
of p62 proteins, effectively creating an anti-oxidative positive feedback loop [188].
Nrf2 has been described to undergo post-translational modifications. Particularly important for
this project, Nrf2 is phosphorylated by casein kinase 2 in the Neh4/5 domain, which amplifies
Nrf2 activation by tBHQ [189]. While this phosphorylation of Nrf2 is important for its activation,
high amounts of sequential phosphorylation lead will accelerate the degradation of Nrf2 [190].
Another study has shown that phosphorylation in a different region of Nrf2 by PKC will lead to
                                                  29


diminished Keap1-Nrf2 interaction, enhancing Nrf2 signaling [191, 192]. Nrf2 can also be subject
to an activating phosphorylation by protein kinase RNA-like endoplasmic reticulum kinase (PERK),
which was observed under conditions of endoplasmatic reticulum stress [193, 194]. Nrf2 is also
to some extent regulated via its acetylation status, as Nrf2 deacetylation by SIRT1 activators
stunted Nrf2 target gene expression [195]. One pathway of acetylation involves the Neh4 and
Neh5 domains, which bind to cAMP response element protein (CREB), which then in turn
acetylates Nrf2 [196]. While this leads to increased expression of ARE-regulated genes, the
expression of heme oxygenase 1 seems to be exempt from this activation [195, 197].
Mechanisms that interrupt Nrf2 signaling
The Neh3 domain of Nrf2 is important for the induction of transcription by Nrf2. When small
deletions at the C terminal were introduced, the Nrf2 variant could still bind the antioxidant
response element and maf proteins, but no actual transcription of Nrf2 target genes occurred.
Curiously, this deletion also increased the half-life of Nrf2 [198]. Apart from these mechanisms,
a number of other proteins are involved in the regulation of Nrf2. Insulin has been shown to
decrease the expression of Nrf2 in renal cells by using a specific DNA-responsive element binding
heterogeneous ribonucleoprotein particle F/K [199]. WDR23 is an E3 ligase that can ubiquitinate
Nrf2 completely independent of Keap1 and does not depend on the DLG or ETGE motifs in Nrf2
[200]. Under hypoxic conditions, such as a perfusion reperfusion injury, the hypoxia-induced
protein siah2 abrogated Nrf2 signaling in a Keap1 independent fashion [201]. PAQ3 regulates
Nrf2 signaling by tethering both Nrf2 and Keap1 to the Golgi-apparatus, facilitating the
degradation of Nrf2 [202]. Hrd1, a protein involved in ER stress during liver cirrhosis, has been
                                                 30


shown to negatively regulate Nrf2. Hrd1 itself is an E3 ubiquitin ligase, capable of enhancing Nrf2
ubiquitylation and degradation [203]. There is also evidence for a process that degrades
intranuclear Nrf2 instead of cytosolic Nrf2. Small ubiquitin-like modifiers (SUMO) 1 and 2 target
nuclear Nrf2 for a process called SUMOlyation [204], which then leads to polyubiquitination and
degradation of Nrf2 in nuclear structures called promyelocytic leukemia nuclear bodies. This
process seems to be particularly important in the pathology of viral and cancerous diseases [205].
                                                 31


  A)                                               B)
Figure 4. Simplified scheme of Nrf2 activation by tBHQ. A. In homeostasis, Nrf2 is tethered to a
Keap1 and promptly ubiquitinated and directed to the 26S proteasome. B. Under conditions of
oxidative stress, the binding of Nrf2 to Keap1 is altered and it is no longer directed to the
proteasome. De novo synthesized Nrf2 can translocate to the nucleus, dimerize with small maf
proteins, and activate Nrf2 target gene transcription. Figure taken with permission from Freeborn
et al [206].
Auxiliary Nrf2 activating proteins
Nrf2 binding to the ARE was enhanced by the protein poly(ADP-ribose)polymerase-1 (PARP1),
surprisingly without physically touching or poly(ADP-ribosyl)ating, but instead by forming a
complex with small maf proteins and the ARE, which results in increased Nrf2 target gene
expression [207]. Another unexpected facilitator of Nrf2 was RAC3/AIB1/SRC3, an oncogene that
                                                32


is commonly involved in human breast cancers. Overexpression of RAC3 in HeLa cells induced
Nrf2 and its target gene HMOX1 [208].
Xeno- and endobiotics as Nrf2 activators
Nrf2 activation by xenobiotic compounds is plentiful and has received broad public interest. Nrf2
activation by edible supplements has been deemed as a “hack for health” by prominent
podcasters like Joe Rogan [209] and Dr. Andrew Huberman [210]. These sensationalist claims
usually refer to isothiocyanate sulforaphane, an Nrf2 activating compound found in Brassicaceae
spec and some other plants. Sulforaphane has received great academic and public interest, which
has resulted in over 3000 peer-reviewed publications and 50 clinical trials over the years [211].
However, many of these studies are problematic, as animal studies have used dosages
magnitudes higher than human studies, among other problems. From a toxicological perspective,
Nrf2 activators of interest can be grouped into endogenous compounds such as prostaglandins,
reactive electrophiles like tBHQ [212], dimethyl fumarate [213], and heavy metals, such as
cadmium[214], lead [215] and arsenic [216]. These compounds can also be grouped via their
mechanism of Nrf2 activation – either by inducing oxidative stress, phosphorylation of Nrf2, or
sulfhydryl modifications of Keap1 [217, 218]. Multiple synthetic Nrf2 activators (which often act
as Keap-1 inhibitors), such as CDDO-me, omovalexolone and oltipraz are currently in clinical trials
[219].
tBHQ, an Nrf2 activator of particular interest in this thesis, induces mitochondrial stress via
thioredoxin-2 (Trx2) in HeLa cells [220]. tBHQ also interacts with cysteine residues on Keap1 at
C151, similarly to sulforaphane, diethyl maleate, and arsenic [221, 222]. tBHQ and other
                                               33


hydroquinones need to be oxidated by divalent copper ions to 2-tert-butyl-1,4-benzoquinone
(tBQ) before they can modulate C151 on Keap1 [223]. Arsenic, another model toxicant in this
thesis, can generate H2O2 in cells [224], which triggers Nrf2 to induce expression of target genes,
such as NAD(P)H-quinone oxidoreductase 1 [225]. Like tBHQ, arsenic also modulates C151 on
Keap1, and other reactive cysteine sensors, namely C273 and C288 [222]. However, arsenic
appears to activate Nrf2 through multiple mechanisms, including release of p62 through
autophagy,     promoting    Nrf2   dimerization,    dissociation   of  Nrf2   from     Keap1   and
polyubiqutination/degradation of Keap1 [225]. It is also important to note that arsenic has a
multitude of Nrf2 independent effects on mammalian cells, including epigenetic changes [226],
metabolic shift towards glycolysis [227] and disruption of zinc finger mediated DNA repair [228].
Endogenous activators of Nrf2, like 15-deoxy-Δ12,14-prostaglandin J2 (15d-PGJ2), activate Nrf2 by
altering the sulfhydryl group on C288 on Keap1 [229]. Zinc Ions, released by intracellular damage,
are also sensed by Keap1, leading to a halt of Nrf2 degradation and subsequent activation of Nrf2
signaling as described above [230].
                                                 34


Tert-butylhydroquinone
tBHQ (IUPAC: tert-butyl-1,4-benzenediol) (Fig 7) is a synthetic aromatic phenol derivative. It is
mainly used as an antioxidant type food preservative in vegetable oils and goods containing fats,
such as meats, fish, and baked goods [231, 232]. In addition to foods, it is added to industrial
organic peroxides, varnishes, lacquers, and biodiesel [233]. The World Health Organization’s and
United Nations’ Food and Agriculture Organization’s Joint Expert Committee on Food Additives
(JECFA) established a daily allowable intake (DAI) of 0.7mg/kg/day for humans [234]. Currently,
tBHQ is limited by the FDA to be less than 0.02% of fat or oil (200mg/kg)[235]. The European
Food Safety Authority sets the same regulation and designates tBHQ as food additive E219 [236].
The toxicological studies cited by the opinion of the EFSA rely on unpublished data by BD Astille
from 1968, which was quoted by the WHO JECFA in their meetings [237]. The EFSA justifies their
DAI with a non-peer reviewed feeding study in dogs [238], which showed a NOAEL at
72/mg/kg/day and the LOAEL at 220mg/kg/day, to which they applied a 100-fold safety factor.
Model diets assume that in western diets, 33-55% of all calories are derived from fat, which
translates to 86-194g of fat per day [236]. Adults consuming such a diet, assuming that tBHQ is
present in consumed fats at 200mg/kg would therefore consume up to 38.8mg of tBHQ, or about
0.6mg/kg/day. The JEFCA suggests that in many western countries, such as New Zealand,
Australia, and the United States, consumers may exceed the ADI of 0.7mg/kg/day [239]. These
regulatory decisions are mostly based on studies performed before western authorities
formulated guidelines on immunotoxicological testing [240]. In a proof of concept for a high
throughput immunotoxicity screening, tBHQ was identified as a substance of concern [241]. It is
                                                35


important to note that in the in vivo studies later described in this dissertation, the mice were
exposed to 0.0014% tBHQ, which is roughly tenfold lower than the ADI for humans. This is also
the dose of tBHQ routinely found in commercially available mouse chow.
Figure 5. Chemical Structure of tBHQ. tBHQ is an antioxidant with a basic phenolic structure. It
is used under the code E319 in the European Union. It is typically used as a food preservative in
human and rodent food products.
Nrf2 off-target signaling
The promiscuity of many Nrf2 activators complicates the identification of Nrf2-dependent
effects. When designing studies, it is important to consider indirect effects of Nrf2 on other
                                                36


signaling pathways that may also induce Nrf2 target genes. C-myc negatively regulates Nrf2
signaling by complexing with Nrf2 and c-jun [242]. A cross-activation of Nrf2 and its target gene
NAD(P)H-quinone oxidoreductase 1 (NQO1) has been reported after a challenge with TCDD, a
ligand for AHR [243]. Mice genetically ablated for Nrf2 showed alterations in their NF-κB
pathways [244]. There can be significant overlap between Nrf2, AP-1, and NF-κB signaling. For
example, depending on the mode of activation [245], Nrf2, AP-1 and               NF-κB can each
independently induce expression of the Nrf2 target gene HMOX1 [246]. There is also evidence to
suggest that Nrf2 and the Blimp-1/Bach2 axis negatively regulate each other [247]. While Nrf2 is
generally considered a transcriptional activator, Bach2 is considered a transcriptional repressor.
Because Bach1 and Nrf2 modulate gene expression through the same cis-elements of the ARE,
the two transcription factors antagonize the activity of one another [248]. A similar mutual
repression pattern has been observed between Nrf2 and NF-κB [249]. FOSL/FRA1 is a protein of
the JUN family and part of the AP-1 signaling pathway that exerts some control over Nrf2
signaling. Fra1 ablation leads to abrogated mitochondrially induced apoptosis in response to
oxidative stress and increased Nrf2 target gene products, extended Nrf2 half-life, and
intranuclear Nrf2 accumulation [250]
                                               37


Human Nrf2 polymorphisms
As an area of emerging research, not much is known about Nrf2 polymorphisms in humans. The
central importance of Nrf2 in oxidative homeostasis suggests the potential for polymorphic
impacts on autoimmunity, cancer, and allergy. A study comparing single nucleotide
polymorphisms (SNP) of Nrf2 in juvenile systemic lupus erythematosus (SLE) found that the
G653A SNP did not influence the incidence of SLE itself, but the incidence of lupus-associated
nephritic syndrome [251]. A study looking at patients with lung cancer found a decreased forced
exhalatory volume in 1 second (FEV1) for the intron SNP rs2364723 [252]. Similar findings with
reduced FEV1 were found in current smokers in a Japanese population [253]. The SNP C617A
correlated with a major increase in acute lung injury after trauma in an ethnically diverse US
study [254]. The A653A SNP was correlated with elevated systolic/diastolic blood pressure,
elevated albumin, and higher mortality in female hemodialysis patients [255].
A series of case studies by the University Hospital in Göttingen, Germany described 4 patients
with extremely rare Nrf2 isoforms that were constitutively active. Three of these patients had
SNPs involving the ETGE motif of Nrf2 (the “hinge”), namely G81S, T80K, and E79K. The remaining
patient had SNP involving the DLG motif (the “latch”), G31R. All patients were pediatric and
suffered from hypogammaglobulinemia, recurrent infections, failure to thrive, and
developmental delay. After diagnosing patient one, the other three were detected around the
globe using GeneMatcher [256]. This incredibly rare genetic aberration highlights the wholistic
importance of Nrf2-mediated homeostasis and that uncontrolled Nrf2 signaling has devastating
effects on human health.
                                                38


Nrf2 in Allergy and Autoimmunity
Our group has published two reviews on the role of Nrf2 in autoimmunity [206] and allergy [257]
recently. The primary evidence for a role of Nrf2 in the regulation of the immune system was
found in Nrf2-deficient mice, which developed an SLE-like syndrome in female mice, including
hallmarks like generation of anti-dsDNA antibodies and nephritic syndrome [258, 259]. This
observation in mice dovetails with later observation in human cohorts that Nrf2 SNP increases
the risk for lupus exacerbations[251]. EAE, a model for human MS, is exacerbated in the absence
of Nrf2 [260, 261]. Currently, there are three Nrf2 activators available as FDA-approved
treatments for MS, dimethyl fumarate (Tecfidera) [213], monomethyl fumarate (Bafiertam)
[262], and diroximel fumarate (Vumerity) [263]. Other Nrf2 activators, such as sulforaphane, A-
1396076, 3H-1,2-dithiole 3-thione, and dimethyl itaconate can improve outcomes in a host of
different rodent models of autoimmunity [264-267].
Mouse studies with ablation of Nrf2 typically used Nrf2 -/- mice, which had a genetic insertion of
a lacZ gene into exon 4 and 5 of the Nfe2l2 gene. This disrupted expression of wildtype Nrf2, and
the newly generated fusion gene is not expressed at the protein level in any tissue [165].
In the context of atopic disease and asthma, whole body ablation of Nrf2 in mice increases airway
inflammation and hypersensitivity to OVA, with IL-4 and IL-13 increases in BALF [268]. Nrf2
protected against these pathological changes [269]. Human studies have shown that lower
expression profiles of Nrf2 correlate with asthma severity [270]. Sulforaphane, an Nrf2 activator,
paradoxically relieved asthma-related bronchoconstriction while decreasing Nrf2 target gene
expression in a human study for 60% of patients in the cohort, while 20% of patients experienced
                                                39


clinical exacerbation [271]. In a rodent model of response to the chemical warfare and terrorism
agent chlorine gas (Cl2), Nrf2 was shown to be protective of airway irritation, and the use of GSH
inhibitors and Nrf2 activators elegantly showed a direct Nrf2 effect that was independent of Nrf2-
mediated induction of glutathione synthesis genes [272].
The mainstay of therapy in all airway hypersensitivity pathologies are inhaled glucocorticoids
[58], which can lose efficaciousness over prolonged usage. In a mouse asthma model, response
to inhaled steroids was restored via Nrf2 activation [273], likely due to the signaling of Nrf2 target
gene aldehyde oxidase, which directs the formation of tight junctions and adherent junction in
the epithelium of airways [274]. Human epidemiological studies uncovered a link between
isoforms of the gene encoding for Nrf2, N-acetyltransferase 2 (NAT2), and glutathione S-
transferase Pi-1 (GSTP1) and the use of acetaminophen in the pathogenesis of asthma [275]. This
study leaves the question of whether the Nrf2 SNPs impacted are rooted in the metabolism of
acetaminophen and reactive metabolites, immune cell function, or something entirely different.
Nrf2 has been investigated as a therapeutic target for airway disease [276] and mouse studies
have shown some success in suppressing asthma models by decreasing type two innate lymphoid
cells (ILCs) [277].
The pathogenesis of atopic dermatitis (AD), a skin condition with abnormal epidermal barrier
function and immune cell infiltration [58], is dependent on Nrf2 in a rodent model of AD using
2,4,6-trinitro-1-chlorobenzene as an AD inducer [278]. This study showed that genetically Nrf2
deficient mice have subdued expression of type II cytokines, decreased IgE, and lower dermal
infiltration of immune cells [278]. There have been a lot of tentative studies suggesting a
                                                 40


protective role of Nrf2 activation in atopic dermatological models, however, those studies were
not sufficient to establish a clear causal role of Nrf2 [279-283].
The role of Nrf2 in food allergy has received very little exploration. Our laboratory has shown that
tBHQ in mouse chow exacerbated anaphylactic reactions to OVA in a dermal sensitization model
[284, 285]. This effect was abrogated when mice received Nrf2 deficient T cells in an adoptive
transfer model, indicating a key role of Nrf2 in the pathogenesis of food allergy. The studies in
this dissertation are further characterizing this role, focusing on the contribution of B cells.
                                                   41


The role of Nrf2 in different immune cells
Stem cells
Hematopoietic stem cells (HSC) are to some extent regulated by Nrf2. In Nrf2-deficient mice, the
ratios of different progenitor cell subtypes are altered and Nrf2 slowed the proliferation of cells
in the T cell lineage, skewing the ratio of generated white blood cells towards myelocytes [286].
Nrf2 mRNA was highly expressed in cells developing into granulocytes but was expressed in much
lower levels in cells of the B cell lineage. In addition, genetic activation of Nrf2 led to a decreased
ratio of lymphocytes to granulocytes [287].
Natural Killer cells
Nrf2 activation via tBHQ diminishes the activation and expression of cytotoxic proteins on NK
cells. In a murine ex vivo model [288] and an in vivo model using IAV as a challenge, Nrf2
activation inhibited FasL and CD107a expression [289]. Nrf2 activation induces the production of
IL-27 by myeloid cells, which recruits and activates NK cells in an influenza challenge model [290].
In a mouse model of cancer, topical administration of tBHQ reduced tumor growth in
lymphocyte-deficient mice and increased NK cell infiltration into the tumor. This study also
showed that Nrf2 induced an IL17D response that was required for this infiltration [291]. The
regulation of IL17D by Nrf2 has been suggested as a potential therapeutic target for cancer
therapy [292].
The Nrf2-Keap1 signaling pathway for metabolic control in natural killer T cells (NKT), which have
a constitutively high metabolism generating large amounts of ROS. Interruption of Nrf2-Keap 1
complexes and therefore untethered Nrf2 signaling delayed NKT development and decreased
                                                    42


NKT numbers. They observed both an increase in NKT cell proliferation and NKT cell apoptosis,
and the effect of apoptosis on the NKT cell numbers dominated. A concurrent ablation of Nrf2
rescued the NKT phenotype [293].
Dendritic cells
Dendritic cells (DC) are professional antigen-presenting cells (APC) and thus have a major role in
the pathogenesis of allergy. Nrf2-deficient DCs had higher amounts of ROS than wildtype DC
when exposed to the common allergen ragweed [294]. Nrf2-deficient DCs show increased
expression of MHC-II and CD86 and adapt a Th2-like phenotype when challenged with ambient
particulate matter [295]. Similarly, the loss of Nrf2 in bone marrow-derived DCs led to redox
dysregulation resulting in higher intracellular ROS and lower GSH expression and higher
activation potential for cytotoxic T cells [296]. Additionally, Nrf2-deficient DCs proliferated more,
yielding higher overall cell numbers, and had a stronger IFN-γ - CD8 T cell response in a cancer
model [297]. When TLR agonists such as LPS were used, Nrf2-deficient DCs produced greater
amounts of IFN- γ than the wild-type controls [298]. The direct challenge of dendritic cells with
crude preparations of Helicobacter pylori outer membrane vesicles activates Nrf2 and induces
the Nrf2 target gene HMOX [299]. In a rodent in vivo study where mice were challenged with
non-typeable Haemophilus influenza, absence of Nrf2 in DCs resulted in greater induction of
activation markers and a generally stronger humoral response [300]. Arsenic, one of the
challenge models used in this thesis, can suppress human DC secretion of IL-12 in an Nrf2-
dependent manner [301].
                                                   43


Macrophages
Macrophages are a subtype of APC closely related to DCs. In addition to their APC function,
Macrophages are phagocytic and therefore have a high basal amount of intracellular ROS. Nrf2-
deficient macrophages are more susceptible to oxidative damage [172, 302]. Clearance of
bacteria by phagocytosis is directly dependent on Nrf2 signaling [303, 304]. Macrophages are
highly sensitive to TLR agonists such as bacterial polysaccharides, which typically results in
induction of proinflammatory genes[305]. Nrf2 counteracts the effects of TLR agonists on
macrophages by inhibiting the transcription of IL-6 and Il-1b [306]. Curiously, Nrf2 signaling in
macrophages is protective in atherosclerosis and facilitates the polarization of macrophages to a
phenotype (MOX macrophages) suited to the containment of atherosclerotic lesions [307-309].
Nrf2 has been suggested to interfere with macrophage polarization, favoring the M2 phenotype,
however, the abundance of signaling pathways involved, including Nrf2, NF-κB, PPARg, and
autophagy, make this determination difficult to prove [310].
T cells
T cells have been the primary focus of our lab over the last decade. Seminal was the discovery
that the activation of Nrf2 polarized CD4 T cells ex vivo to a Th2 phenotype via GATA3 while Nrf2-
deficient cells polarized towards a Th1 phenotype [212]. In contrast to this observation, Nrf2-
knockout mice had greater number of Th2 cells than wild-type mice in a model of bleomycin-
induced lung fibrosis [311]. Nrf2 activation in T cells protected from acute kidney injury in mice
by inhibiting IFN-γ and TNF-α secretion by CD4 T cells with a concomitant increase in regulatory
T helper cells [312, 313]. Our group showed impaired activation of Jurkat T cells (a human T cell
                                                 44


line) in the presence of the Nrf2 activators tBHQ and CDDO-Im, though some of the effects were
at least partially Nrf2-independent [314, 315]. Our lab also showed impaired activation of primary
human CD4+ T cells in the presence of tBHQ ex vivo, though the role of Nrf2 was not explored in
this study [316]. A study in human T cells that used CRISPR/CAS9 to knock down Keap1 causing
highly activated Nrf2 signaling, found that the proportion of CD4 T cells expressing activation
markers and T cells expressing IL-10 was increased while the proportion of CD8 and IL-17-
expressing cells was decreased [317]. In stark contrast, in vivo genetic activation of Nrf2
specifically in FOXP3+ T cells reduced the population of this T cell subtype [318]. This model
showed increased inflammatory T cell responses in the lung and liver at baseline, resembling
autoimmune processes. This finding highlights the highly divergent outcomes of increased Nrf2
signaling depending on cell type, compartment, species, and model and the difficulties in
accurately measuring, characterizing, and reporting the effects of Nrf2. One of the current
challenges investigators in the field of Nrf2 in immune populations are facing is the reconciliation
of these complex and sometimes seemingly contradictory observations.
B cells
B cells are understudied in general, and specifically in the context of Nrf2. Studies on a
lymphoblast-derived B cell line have suggested that Bisphenol A is cytotoxic to B cells and
sensitivity depends on the level of Nrf2 expressed [319]. Another study suggested that Nrf2
deficiency increases the immunoglobulin response to the intracellular bacteria Haemophilus
influenza [300]. Sulforaphane, an Nrf2 activator, inhibits B cell function in a mouse model of
arthritis, decreasing IL-6, and IL-17, leading to less collagen-specific IgG1 and IgG2a and inhibiting
                                                   45


the activation of murine splenocytes after LPS challenge [320]. Nrf2 overexpression in vitro
increased CXCR4 expression and increased the survival of a B cell line under hypoxic conditions
[321]. In humans, a cohort study of people suffering from SLE and age-matched healthy controls
showed that in SLE patients, cytosolic ROS were decreased and mitochondrial ROS were
increased in B cells, and expression levels of Nrf2 and Keap1 were significantly increased in B cells
of SLE patients [321].
Our lab has shown recently that B cells, when challenged with TLR agonist LPS, showed
impairment of the activation markers CD69, CD25, CD22, and CD138 when tBHQ was present,
independent of whether the cells had intact Nrf2 signaling. In contrast, tBHQ increased IgM
secretion in wild-type, but not Nrf2-null, mice, indicating this effect is Nrf2-dependent [162]. This
effect on T cell-independent B cell activation underlines the need for studies that investigate T
cell-dependent B cell activation, which are shown in Chapters 3 and 4 of this thesis. I hypothesize
that in addition to these T cell-independent effects, there are effects on the T cell-dependent
activation of B cells. In chapter 2, I will measure the impact of arsenic trioxide on human
peripheral blood B cells after challenge with influenza A. In chapter 3, I will show the effect on B
cell activation in a murine model of T-cell dependent B cell activation, and additionally I will be
able to determine whether these effects are Nrf2 dependent. In chapter 4 and 5 I will show the
impact of dietary tBHQ on the development of allergies against chicken ovalbumin in a
transdermal sensitization model. Collectively, these studies will elucidate the impact of the Nrf2
activators tBHQ and ATO on B cell activation and some elements of B cell function.
                                                  46


Chapter 2: Arsenic trioxide inhibits the response of primary human B cells to
                          influenza virus A in vitro
                                       47


Abstract
Arsenic compounds are common environmental toxicants worldwide and particularly enriched
in the Northeast and the Southwestern United States, the Alps, and Bangladesh. Exposure to
arsenic is linked with various detrimental health outcomes, including cancer, cognitive decline,
and kidney damage. Our group has previously shown that arsenic trioxides alter T cell cytokine
production. In this study, we demonstrate that exposure to arsenic compounds alters B cell
function in an in vitro influenza model. Human peripheral blood mononuclear cells (PBMCs) were
isolated from blood and cultured with arsenic trioxide (As3O2, ATO) and subsequently challenged
with Influenza A virus. B cells showed a decreased expression level of CD267, surface IgG and
CD80 when treated with arsenic trioxide. Taken together, the data suggest that arsenic trioxide
affects the activation and surface antibody expression of human peripheral B cells. Overall, this
suggests that arsenic trioxide exposure could cause impaired humoral immunity.
                                               48


Figure 6. Schematic of the experimental setup. Human PBMCs were isolated and treated with
either VEH (0.04% PBS) or 0.5µM or 1µM of As2O3. 30 min later, PBMCs were challenged with 0.5
HAU of influenza virus A. Cells were collected at 96 h, washed, and labeled for FACS analysis.
                                                49


Introduction
Arsenic is a naturally occurring contaminant in aquiferous water reservoirs, on which a significant
minority of US households rely [322]. In the United States, the Southwest, Northeast, and upper
Midwest are prone to high levels of arsenic in well water. The EPA recommends arsenic
mitigation or alternative water sources if the arsenic concentration in drinking water exceeds
10mg/l [322]. Globally, Bangladesh and the Austrian Styr have chronically elevated arsenic
contamination in water [323, 324]. Most arsenic compounds found in water are inorganic AS(III)
or AS (V) species [325], which are the most acutely toxic variants of arsenic [326]. Acute arsenic
toxicity, rarely seen due to environmental exposure, usually manifests as fulminant multi-organ
failure [327]. The gastrointestinal system is typically affected first within hours of oral or
intravenous administration of high amounts of inorganic arsenic, followed by the liver, lungs, and
kidneys. Death often ensues within hours to days following exposure [328]. Chronic exposure to
arsenic is much more common and is associated with an increased lifetime risk for multiple
cancers, cardiovascular disease, intellectual decline, dermatological, endocrinological, and
neurological impairment, and miscarriages [329]. Our group has previously shown that arsenic
trioxide impairs the release of cytokines from T cells [330]. These findings raised the question of
whether exposure to arsenic changes the response of immune cells to pathogens, such as viruses
and bacteria. In this study, we are using an in vitro model of influenza A exposure to measure the
response of B lymphocytes exposed to arsenic trioxide. We have previously shown that T cells
respond differently and produce fewer type 1 cytokines in the same model [331].
                                                  50


Viral infections are subject both to the cellular and humoral response of the immune system. The
cellular response is driven by T cells, while the humoral response relies on antibodies produced
by B cells. Both pathways are intimately intertwined. On the one hand, B cells require T cells for
isotype switching and antibody production but conversely can serve as antigen-presenting cells
to activate T cells.
Successful clearing of viral infections often requires the generation of antigen-specific
immunoglobulins. B cells produce the low-affinity pentamer immunoglobulin M (IgM) first,
before undergoing somatic hypermutation, affinity maturation, and isotype switching [332]. In
an antiviral scenario, the result is a high-affinity immunoglobulin G (IgG) which binds with great
specificity and affinity to viral antigens [332, 333]. B cells also undergo clonal expansion during
this time, an energy-intensive process resulting in a marked increase in the number of active,
antigen-specific B cells [334].
Arsenic trioxide is also used therapeutically to treat cancer, Specifically, it is used together with
all-trans retinoic acid as the first-line therapy for promyelocytic leukemia [335]. New oral
formulations are currently in trial to replace the inconvenient intravenous administration [335].
However, most human exposure to arsenic trioxide occurs in the context of naturally occurring
or inadvertent environmental contamination of drinking water.
In this study, we report a direct effect of arsenic trioxide on B cell activation, using our unique in
vitro influenza infection model by which we can monitor the response of primary human immune
cells (Fig. 6).
                                                   51


Methods
Materials
Arsenic trioxide was purchased from Sigma Aldrich (St. Louis, MO). All fluorescent antibodies
used in this study were purchased from Biolegend, (San Diego, CA). All other reagents were
purchased from Sigma Aldrich (St. Louis, MO) unless otherwise indicated.
Human Peripheral Mononuclear Cell Preparation
PBMCs were isolated from Leukapheresis packs (Gulf Coast Blood Center, Houston, TX) using
Lymphoprep gradient separation media (Stem Cell Technologies, BC). Cells were cultured in
complete RPMI (1640 RPMI, 10% Fetal Bovine Serum (Biowest, MO), 25mM HEPES, 1mM Sodium
Pyruvate, 1x nonessential amino Acids, 100 U/ml Penicillin, 100 U/ml streptomycin) at 2 x 105
cells/well in a 96-well round bottom plate. Cells were either incubated without further treatment
(BKG) or treated with vehicle (VEH, 0.004% PBS), 0.5µM, 1µM, or 2µM arsenic trioxide. After 30
min, cells in the VEH and arsenic groups were exposed to 0.5 hemagglutination units (HAU) of
influenza A/PR/8/34 (H1N1). The number of hemagglutination units influenza virus achieving
optimal PBMC activation was determined in pilot studies and found to be 0.5 HAU. No virus was
added to the BKG cells, BKG cells were neither treated with ATO nor IAV, and corresponding
amounts of cell culture media were added instead. Cells were cultured for 96 h at 37˚C and 5%
CO2 and subsequently harvested for FACS analysis and supernatant collection (Fig 6).
Flow Cytometry
96 h after treatment, supernatants were collected, and cells were prepared for flow cytometry.
Cells were washed with phosphate-buffered saline, labeled with Zombie Aqua Fixable Viability
                                                 52


Dye (BioLegend, San Diego, CA), washed in FACS buffer (1% FBS in PBS), labeled with surface
marker antibodies (Table 3), washed in FACS buffer, fixed and permeabilized with
Foxp3/Transcription factor staining kit (Invitrogen, Waltham, MA), then labeled with antibodies
against Immunoglobin M (IgM) and Immunoglobulin G (IgG). After labeling, cells were washed
twice in permeabilization buffer, resuspended in FACS buffer, then immediately analyzed on a 4-
laser Attune NxT (ThermoFisher, Waltham, MA). UltraComp eBeads (Invitrogen, Waltham, MA)
were used to determine laser gains and compensation matrices.
                                                53


Table 3. Fluorescent antibodies used in Chapter 2.
    Target               Label            Clone            Manufacturer
     CD22               BV421            S-HCL-1            Biolegend
     CD86               PE-Cy7            BU63              Biolegend
     CD40               AF-700             5C3              Biolegend
     CD25               BV711            M-A251             Biolegend
     CD80               PE-Cy5            2D10              Biolegend
     CD19                  PE             HIB19             Biolegend
      IgG                 APC          M1310G05             Biolegend
    CD267           PE-Dazzle 594          1A1              Biolegend
     CD69               BV-605            FN50              Biolegend
      IgM               BV-605           RMM-1              Biolegend
    CD122            PerCP Cy5.5          TU27              Biolegend
     B220                 FITC           RA36B2             Biolegend
  Live/Dead          Fixable Aqua           /             Thermo Fisher
Statistical Analysis
The data were compiled from three different donors and are presented as the mean ± standard
error of the mean. One-way ANOVA was used to determine statistical differences between
treatment groups, and Dunnett’s multiple comparisons test was used to compare each treatment
group to VEH. Calculations and graphical visualization of the results were done using PRISM
Graphpad 9.20 software (La Jolla, Ca).
                                              54


Results
Arsenic Trioxide decreases B cell viability and size increase following activation
In flow cytometry, the forward scatter (FSC) parameter provides an estimate of the size of cells
being measured. The FCS is a function of the voltage recorded of a photodiode located behind,
but slightly transposed to the travel path of the blue laser (408nm wavelength). Larger cells
with diffract the blue laser beam more so than smaller cells and thus increase the fraction of
light hitting the transposed photodiodes, increasing the measurement of FCS. B cells upon
activation will expand in cell size and accordingly, the FSC intensity will increase
proportionately. To facilitate interpretation of FCS data, which typically ranges from 1x10^6 to
3x10^6 , we normalized the measured FCS values to that of the vehicle group, thus comparing
the all other groups to it. We observed a statistically significant increase in FSC intensity 96 h
after exposure to influenza (Fig 7B). At 1µM ATO this effect was diminished. As ATO is known to
be cytotoxic, we wanted to control for potential changes in cell viability. Using an amine-
reactive dye, we measured the percentage of amine-nonreactive cells (live cells) of total cells
after debris and doublet exclusion. We observed a non-statistically relevant decrease, but no
significant changes in the number of viable cells with ATO in concentrations up to 1 mM
compared to the vehicle group (Fig. 7a). Significant changes were seen at 2µM. The data at 2µM
is presented in the following figures to show a concentration response, however the results at
2µM may be partially due to unspecific cytotoxicity.
                                                 55


Figure 7. ATO has a minor impact on viability and cell size of human PBMCs. Human PBMCs
were isolated and treated with either VEH (0.04% PBS) or 0.5µM, 1µM or 2µM of As2O3. 30 min
later, PBMCs were challenged with 0.5 HAU of influenza virus A. Cells were collected at 96 h,
washed, and labeled for FACS analysis. A: Viability was quantified by use of a fixable, viability dye
and detected by flow cytometry. Flow cytometric gates were set on FCS/SSC and doublet
exclusion prior to live/dead analysis. B: Cell size was ascertained by measuring forward scatter
by flow cytometry. Data were normalized to the VEH group. Data are presented as the mean ±
SE of three donors. ** represents p < 0.01, *** represents p < 0.001? and ns = not significant.
                                                  56


Arsenic trioxide decreases the percentage of IgG+ B cells but does not change the
number of B cells
Among the many roles a B cell can play during an infection, the production of high-affinity
immunoglobulins is arguably the most important. Mature but resting B cells express IgM and IgD
type B cell receptors on their surface. After activation, B cells undergo affinity maturation and
isotype switching. Typically, during a viral infection, IgM-producing B cells undergo class switch
recombination to produce IgG antibodies. Circulating B cells that express IgG on their surface are
antigen-experienced cells that have previously undergone isotype switching.
In this study, to determine B cell phenotype we used CD45R/B220 as a marker for B cells. In Fig
3, we observe no statistically significant change in the percentage of B cells among PBMCs after
exposure to either ATO, influenza virus, or both.
Among B cells, the number of cells that expressed IgG on the cell surface was significantly
increased upon exposure to influenza. The addition of ATO diminished this effect in a
concentration-dependent manner, with statistically significant decreases in the 1µM group
(P=0.0069) (Fig 8B).
                                                 57


Figure 8. Surface IgG is decreased in the presence of ATO while B cell populations do not
significantly change. Human PBMCs were isolated and treated with either VEH (0.04% PBS) or
0.5µM, 1µM or 2µM of As2O3. 30 min later, PBMCs were challenged with 0.5 HAU of influenza
virus A. Cells were collected at 96 h, washed, and labeled for FACS analysis. Viability dyes for
live/dead cell discrimination as well as fluorescently conjugated antibodies for B cell
identification and surface IgG, among other targets, were used to identify and quantify protein
expression on B cells using an Attune NxT 4 Laser cytometer. A. B cells were
                                               58


Figure 8 (cont’d)
identified by expression of B-220 and are represented as a percentage of the total PBMC
population. B. IgG-expressing B cells (B-220+) were identified and quantified by flow cytometry.
Data was normalized to VEH group due to inherent physiological variation in PBMC donors. C.
Representative pseudocolor plots for single samples in the BKG, VEH and 1µM As2O3 groups.
Data are presented as the mean ± SE of three donors. (ns indicates not significant, * indicates
significance relative to the VEH group, p<0.05, ** indicates significance relative to the VEH group,
p<0.01, *** indicates significance relative to the VEH group, p<0.001, **** indicates significance
relative to the VEH group, p<0.0001)
Arsenic trioxide decreases the expression of CD25 and CD80 but does not affect CD86
expression
CD25 is a marker for mature B cells with increased antigen-presenting capacity and a marker of
activation. The addition of influenza virus increased the percentage of CD25-expressing B cells in
the PBMC population. The addition of ATO caused a concentration-dependent decrease in the
relative amounts of CD25+ B cells (Fig. 9). CD80 and CD86 are costimulatory proteins expressed
on professional antigen-presenting cells that interact with CD28 and CTLA-4 on T cells. Both
molecules are typically expressed simultaneously with overlapping functions; however, CD80
shows a much higher affinity for CD28 [336]. CD86 co-stimulation is implicated in a skew towards
a TH2 response in T cells, while CD80 has been associated with TH1 responses [337]. In our assay,
PBMCs increased the expression of both CD80 and CD86 in response to challenge with Influenza
A virus (Fig 10). We observed a concentration-dependent decrease in CD80 expression in the
                                                 59


presence of arsenic trioxide. CD86 on the other hand did not show a comparable effect. (Fig 10).
Collectively, the data suggest differential effects on markers of activation and co-stimulation.
Figure 9. Expression of CD25 decreases with ATO treatment.Human PBMCs were isolated and
treated with either VEH (0.04% PBS) or 0.5µM, 1µM or 2µM of As2O3. 30 min later, PBMCs were
challenged with 0.5 HAU of influenza virus A. Cells were collected at 96 h, washed, and labeled
for FACS analysis. Viability dyes for live/dead cell discrimination as well as fluorescently
conjugated antibodies for B cell identification and CD25, were used to identify and quantify
protein expression on B cells using an Attune NxT 4 Laser cytometer. Data are presented as the
mean ± SE of three donors. (ns indicates not significant, * indicates significance relative to the
VEH group, p<0.05, ** indicates significance relative to the VEH group, p<0.01, *** indicates
                                                60


significance relative to the VEH group, p<0.001, **** indicates significance relative to the VEH
group, p<0.0001). The figure depicts the compiled data from 4 donors.
Figure 10. ATO decreases the expression of CD80 but not CD86. Human PBMCs were isolated
and treated with either VEH (0.04% PBS) or 0.5µM, 1µM or 2µM of As2O3. Cells were collected
at 96 h, washed, and labeled for FACS analysis. Viability dyes for live/dead cell discrimination as
well as fluorescently conjugated antibodies for B cell identification, CD80 and CD86, were used
to identify and quantify protein expression on B cells using an Attune NxT 4 Laser cytometer. Data
are presented as the mean ± SE of three donors. (ns indicates not significant, * indicates
significance relative to the VEH group, p<0.05, ** indicates significance relative to the VEH group,
p<0.01, *** indicates significance relative to the VEH group, p<0.001, **** indicates significance
relative to the VEH group, p<0.0001). The figure depicts the compiled data from 4 donors.
                                                 61


Arsenic trioxide decreases the expression of CD22 and CD267 on B cells in a concentration-
dependent manner.
B cell differentiation and activation are tightly regulated. Overactive B cells and autoantibodies
are implicated in most autoimmune diseases, chiefly in SLE (lupus), Mixed Connective Tissue
Disease, Sjogren’s Disease, and autoimmune hepatic disorders [338]. Synchronizing affinity
maturation, clonal expansion, and isotype switching is a delicate balance regulated by many
different proteins [332]. We next tested the hypothesis that arsenic trioxide alters the expression
of some of these regulatory proteins, such as CD267 and CD22.
CD267, also known as Transmembrane activator and CAML interactor (TACI) is a transmembrane
protein associated with T-cell independent B cell activation and negative regulation of B cell
proliferation [339]. CD267/TACI was decreased by influenza infection in our model and was
further decreased in the ATO-treated groups (Fig. 11). CD22, a member of the immunoglobulin
superfamily and has a similar physiological role to CD267 in B cell activation. It works as an
autoregulatory protein, preventing hyperactivation of the B cell receptor (BCR). In our in vitro
model, we saw an induction of CD22 during the influenza challenge, likely to prevent an
overreaction by B cells. This induction was abrogated in a concentration-dependent manner by
arsenic trioxide (Fig 11). The results suggest that ATO has a suppressive effect on the expression
of regulatory proteins on the cell surface of B cells.
                                                 62


Figure 11. ATO decreases the expression of CD267 and CD22. Human PBMCs were isolated and
treated with either VEH (0.04% PBS) or 0.5µM, 1µM or 2µM of As2O3. 30 min later, PBMCs were
challenged with 0.5 HAU of influenza virus A. Cells were collected at 96 h, washed, and labeled
for FACS analysis. Viability dyes for live/dead cell discrimination as well as fluorescently
conjugated antibodies for B cell identification, CD267 and CD22, were used to identify and
quantify protein expression on B cells using an Attune NxT 4 Laser cytometer. Data are presented
as the mean ± SE of three donors. (ns indicates not significant, * indicates significance relative
to the VEH group, p<0.05, ** indicates significance relative to the VEH group, p<0.01, ***
indicates significance relative to the VEH group, p<0.001, **** indicates significance relative to
the VEH group, p<0.0001)
                                                63


Discussion
We used an in vitro infection model to determine whether the presence of arsenic trioxide
affected the activation of human B cells in response to influenza in isolated PBMCs. We
demonstrated a decrease in activation markers, a decrease in surface IgG+ B cells and a change
in the expression pattern of the costimulatory molecules CD80 and CD86. The effects
demonstrated here were observed in a concentration-dependent fashion. I acknowledge that
arsenic trioxide started to show a cytotoxic effect at 2µM and thus, nonspecific cytotoxic effects
may confound results. However, even if the 2µM point were omitted, there is a significant effect
in most markers at 1µM. Taken together, the data suggest that arsenic trioxide affects the
activation and surface antibody expression of human peripheral B cells.
The activation marker CD25 is expressed in humans after activation. It’s expression levels
correlate with a larger cell size and indicate readiness to expand clonally, as CD25 the high-affinity
subunit of the IL-2 receptor [340] and is also a marker for memory B cells [341]. Though it would
be premature to infer from the markers used in this study to say that the memory B cell pool
expanded, overall the data suggest the B cell population as a whole showed a higher activation
status, which was diminished in the presence of ATO.
CD80 and CD86 are costimulatory molecules that interact with T cells and can send stimulatory
signals to T cells. Although CD80 and CD86 are often considered to be immunologically
equivalent, they share only 26% sequence homology and have different affinities for their
receptors CD28 and CTLA-4. CTLA-4 is considered canonically to be an inhibitory signal while
CD28 canonically is a stimulatory signal. It has been hypothesized in the past that CD80 signaling
                                                   64


preferentially polarizes T cells to a Th1 phenotype, while CD86 polarizes towards a Th2 phenotype
[342]. This is particularly interesting in the present study, as I would expect immune cells to react
to an intracellular pathogen such as IAV with a Th1 polarized, IgG-producing response. Indeed,
we observed a marked increase in CD80-expressing cells, and a less pronounced increase in CD86-
expressing cells after activation with IAV. Interestingly, the presence of tBHQ decreased the
percentage of CD80 expressing B cells, while it did not decrease the percentage of CD86
expressing B cells. The decrease CD80 is consistent with our previously published study, which
showed that ATO decreased induction of the Th1 cytokine IFNg by IAV-activated T cells. Likewise,
this is consistent with the other observations in the current study, e.g. the expression of surface
IgG, which was suppressed in the presence of ATO.
While there are many strengths to this in vitro infection model, which uses primary human
immune cells, there are also some limitations. Influenza A virus, a respiratory pathogen, infects
cells of the nasal, oropharyngeal, and bronchial epithelium in vivo, which is obviously not
accounted for in this in vitro model. Another limitation is that local responses of cells and in the
secondary lymphoid organs are not necessarily modeled well using PBMCs. However, this in-vitro
infection model is scalable and thus suitable for medium throughput assays, which could be
useful for both investigation of mechanisms and chemical screening—the data from which could
potentially be used for predictive toxicity modeling.
The use of PBMC samples from unknown donors also likely introduces some level of variance into
these studies. In particular, the unknown vaccination and previous exposure status could
certainly introduce some variability in the results. It is important to note, however, that in PBMCs
                                                   65


from every individual we have tested so far, we have observed a rapid and robust T cell and B cell
response to this viral strain, indicating that the PR8 influenza strain consistently produces a
memory response from most healthy donors. Overall, this suggests broad community immunity
against this particular virus strain (VEH group in Fig 8-11). However, it also suggests that this
approach models a secondary response to IAV and a different peptide or virus would be needed
to assess a primary response.
Overall, our results suggest that ATO has an inhibitory effect on the B cell response to IAV, which
is consistent with other published studies. Other groups have shown that arsenic trioxide reduces
T and B cell activation, numbers, and transplant rejection [343, 344]. There have also been
reports that arsenic trioxide decreases glutathione (GSH) in T and B cells in vitro [345]. Decreasing
GSH would predispose B cells potentially to oxidative stress, which they are prone to during
activation and proliferation. A murine inhalation study of arsenic trioxide found no changes in
LPS-mediated B cell activation, but major changes in the T cell-dependent humoral response to
sheep red blood cell vaccination [346].
A potential confounding factor in these studies may be the presence of other immune cells, which
themselves react to influenza virus. We measured the impact of ATO on T cell function [331] and
on NK cell function (manuscript in preparation at the time of writing). To eliminate potential
confounding factors, future studies may use magnetic or flow cytometric isolation or enrichment
of B cells prior to the assay. This strategy could also be used to differentiate between B cell
subtypes in peripheral blood, and subtype specific effects could be illuminated. Peripheral blood
mononuclear cells contain both antigen-experienced B cells, such as memory B cells, and naïve B
                                                66


cells. I hypothesize that enriching the assay for antigen-experienced B cells may amplify the effect
of IAV challenge. It would be interesting to quantify the effect of ATO on these antigen-
experienced B cells, as there may be a differential effect of ATO on more differentiated memory
B cells compared to the less differentiated naïve B cells.
Exposure to arsenic trioxide occurs most commonly as an environmental contaminant through
drinking water, however, exposure may also occur through other routes. Clinically, ATO is used
to treat promyelocytic leukemia. The therapeutic IV dosage for a 70kg human with 5.5L blood
volume is 0.15mg/kg, which is equivalent to ~9µM plasma concentration [347]. This is a much
higher concentration than used in these studies, as we adjusted for a higher susceptibility of cells
in vitro. Exposure of humans to arsenic trioxide or other arsenite compounds through occupation
(as in pesticides, wood preservatives, iron working) or environmental (as in oral ingestion or
inhalation) is much harder to quantify. Arsenite compounds, including arsenic trioxide, As(III),
and As(V) are found commonly found in drinking water, residually in foods, and rarely airborne
in wildly varying concentrations globally [348]. Studies estimate that humans in western society
are exposed to up to 50µg of arsenic species per day through ingestion, most of which is of the
less bioactive organic variant. About 10µg are expected to be inorganic, and up to 4µg of these
are consumed through drinking water [348-350]. Inorganic arsenites are converted to organic,
biomethylated arsenic compounds and excreted through the urine, though some amount of
bioaccumulation occurs [348]. It is important to note that the amount of exposure can vary
markedly depending on location, dietary choices, and other lifestyle factors. Certain regions of
Taiwan, Bangladesh, India, and southern South America have groundwater arsenic
                                                  67


concentrations in excess of 50µg/L [351-354]. In the US, the United States Geological Survey
estimates that about 10% of people that rely on well water are exposed to inorganic arsenic levels
exceeding 10µg/L, with clustering in the Upper Northeast, Southern Michigan/Northern Ohio,
and the Southwest [355]. While exact numbers are impossible to extrapolate from these findings,
these data indicate a clear need to characterize the impact of low to moderate arsenic exposure
on immunity.
In the present study, we have shown that arsenic trioxide impacts the activation and antibody
response of B cells to influenza virus. Perhaps more importantly, however, we have developed a
novel in vitro model that can be used to characterize the impact of toxicants on the response of
primary human B cells to influenza virus, which prior to the Covid-19 pandemic, was the 8th
leading cause of death in the U.S. Overall, this study demonstrates that arsenic trioxide strongly
impairs the B cell response to influenza virus, while also establishing the use of a novel, scalable
model for immunotoxicity screening.
                                                 68


Chapter 3: Nrf2 dependent and independent effects of T-cell dependent B
                        cell activation ex vivo
                                   69


Abstract
B cells provide humoral immunity via the production of antibodies. The activation of B cells and
generation of high-specificity antibody is tightly regulated via an intricate interaction with T
lymphocytes in vivo, both in mice and men. This process is crucial for immunity against intra-and
extracellular pathogens, long-lasting immunological memory, and vaccination strategies.
Conversely, exaggerated antibody responses can be equally impactful and lead to autoimmunity
and allergy. Our group has previously shown that the synthetic food additive, tert-
butylhydroquinone (tBHQ) alters the response of T cells to both allergic and infectious challenges
and increases the production of IgE- antibodies in vivo. tBHQ is an activator of nuclear factor
erythroid 2-related factor 2 (Nrf2), which is a master regulator of antioxidant and detoxification
pathways. To measure a potential Nrf2-dependent effect of tBHQ on B cell activation, we
designed an ex vivo polyclonal activation assay that mimics T cell signaling via the B cell receptor,
CD40, and cytokine signaling. We isolated splenocytes from wild-type and Nrf2-deficient mice,
which were then treated with increasing concentrations of tBHQ and then activated. After 48
hours, the cells were harvested and analyzed using high dimensional intracellular spectral flow
cytometry. We saw a concentration-dependent increase in the expression of the costimulatory
molecule CD80 by tBHQ in activated B cells while decreasing the expression of the closely related
costimulatory molecule CD86. tBHQ caused a decrease in the expression of the early activation
markers CD69 and CD25 in activated B cells. Interestingly, IgG1 surface expression was reduced
in both wildtype and Nrf2-deficient cells, suggesting an Nrf2- independent effect of tBHQ.
Collectively, the data suggest disparate effects of the food additive tBHQ on the expression of
                                                 70


cell surface proteins and antibody production by activated splenic B cells, which may play a role
in the effects of tBHQ in allergy and host immunity.
                                               71


Introduction
B cells are the main mediator of humoral immunity, producing a range of antibodies against
foreign and autoantigens. In addition to their immunoglobulin (Ig) producing role, B cells are
professional antigen-presenting cells (APC), able to phagocytose, trim and present antigens to
both CD8 and CD4 T cells [356].
The production of antibodies is tightly regulated. Naïve, inactive B cells will express
immunoglobulin D (IgD) receptor transiently when they exit the bone marrow [357]. IgD is mostly
replaced by the most prevalent Ig on resting B cells, Immunoglobulin M (IgM). Upon activation
of B cells by T follicular helper cells (TFH), the B cell begins to secrete IgM antibodies. Depending
on the signals provided by TFH cells, B cells can change their isotype from IgM or IgD to one of
the late-phase, high specificity immunoglobulins, such as Immunoglobulin A (IgA),
Immunoglobulin E (IgE), or Immunoglobulin G (IgG). This process happens concurrently with
somatic hypermutation, which greatly enhances the specificity of the antibodies. B cells then
either undergo apoptosis upon cessation of TFH signaling or develop into memory B cells or
plasma cells. During these processes, B cells are subject to considerable oxidative stress, and the
transcription factor Nrf2 presumably plays a role in mitigating this stress [358].
Nrf2 is a cytoprotective transcription factor that primarily responds to nucleophilic and
electrophilic stress. Structurally it is a cap’n’collar transcription factor with a prominent basic
leucine zipper domain [165, 359]. Inactive Nrf2 is typically located in the cytosol, where it is
anchored to its repressor protein, Keap1. Nrf2 bound by Keap1 is ubiquitinated and directed to
                                                    72


the proteasome for degradation [360]. Under oxidative or electrophilic stress conditions though,
Nrf2 translocates to the nucleus and promotes an array of cytoprotective genes[361-363].
Nrf2 activators have been shown to be protective in many different toxicant models, however,
the role of Nrf2 has not been verified in all of these (important due to the promiscuity of Nrf2
activators). Our group has published numerous studies in immune cells using Nrf2 activators,
such as cadmium[364], CDDO-Im [365], arsenic trioxide [330], and tert-butylhydroquinone
(tBHQ) [212, 288, 314, 316, 365]. The latter is a common phenolic food preservative that prevents
the oxidation of fatty acids in hydrocarbons [232]. This is particularly important in shelved
products containing vegetable oils, margarine, and lards. Exposure to tBHQ is increased in a
western style diet.
Nrf2 has been shown to orchestrate several effects within the immune system. Nrf2 alters the
cytokine response of activated T cells and alters T cell differentiation [212, 365]. NK cells have
been shown to have a less pronounced effector function in vitro when exposed to tBHQ [288].
Exposure to tBHQ in murine LPS-stimulated B cells increases IgM secretion and decreases
expression of surface activation markers in an Nrf2 dependent-and independent fashion [162].
Whereas the previous study of tBHQ in B cells focused on a T cell-independent stimulus (LPS),
the present study was designed to understand the effects of tBHQ on T cell-dependent activation.
Since our previous studies demonstrate that tBHQ strongly impacts T cell function, we used an
activation cocktail that models T cell-dependent B cell activation. This allowed us to investigate
the effect of tBHQ on B cells specifically.
                                                  73


In this study, we demonstrate that the food additive tBHQ has disparate effect on the antibody
production by splenic B cells and their expression of surface markers, which may indicate a role
for tBHQ in allergy and host immunity.
                                               74


Methods
Materials
The sources of fluorescent antibodies are shown in table 4. tBHQ was purchased from Sigma
Aldrich (St. Louis, MO) and kept in an airtight, desiccated container. Media for cell culture were
purchased from Gibco (Thermo-Fisher, Waltham, MA). All other reagents were purchased from
Sigma Aldrich (St. Louis, MO) unless otherwise indicated.
Splenocyte isolation
All animal protocols are in compliance with the Guide for the Care and Use of Animals and were
approved by the Institutional Animal Care and Use Committee (IACUC) at Michigan State
University. Splenocytes were isolated using aseptic technique from C57BL/6 female mice. The
mice used were either wild-type or Nrf2-deficient. Spleens were mechanically dissociated and
single cell suspension of splenocytes was centrifuged at 300g/5min, washed, underwent ACK
lysis, and were washed twice again. Cells were resuspended in complete Gibco Roswell Park
Memorial Institute (cRPMI) 1640 Medium. Cells were then seeded out in a 96-well U-bottom
plate (Greiner Bio-One) at 1x10^6 cells per well. Cells were treated with either 2µl cRPMI
(Background), 0.005% Ethanol (Vehicle), or an escalating concentration of tBHQ ranging from
0.1µM to 5µM. Cells were incubated at 37C/5% CO2 for 30 minutes and then treated with either
6µl cRPMI (Background) or 6µl of our activation cocktail (10µg/ml F(ab')2-Goat-Anti-Mouse IgM,
1µg/ml Anti-Mo CD40 and 10ng/ml recombinant mouse IL-4) (Vehicle and all tBHQ treated
groups). Cells were incubated for 48 hours and then harvested for analysis. Background cells are
defined as cells that received neither activation cocktail nor tBHQ treatment, but equal amounts
                                                 75


of cRPMI instead. Vehicle cells received activation cocktail and 0.005% ethanol in cRPMI, which
is the same amount of ethanol as received by cells exposed to tBHQ. tBHQ groups received stated
concentration of tBHQ, 0.005% ethanol in RPMI and activation cocktail.
Activation cocktail
To simulate a T cell-dependent B activation ex vivo, we designed an activation cocktail consisting
of 10µg/ml F(ab')2-Goat-Anti-Mouse IgM (Invitrogen), 1µg/ml Anti-Mo CD40 (eBioscience, clone
1C10) and 10ng/ml recombinant mouse IL-4 (Peprotech, 214-14).
Fluorescent labeling
Cells were labeled both on the surface and intracellularly using the following protocol: Cells were
centrifuged at 300g/5min, the supernatant was discarded, and cells were washed with Calcium
and Magnesium deficient Phosphate Buffered Saline (PBS). This step was repeated twice, then
100µl of amine-reactive LIVE/DEAD™ Fixable Blue Dead Cell Stain Kit (Invitrogen) was added. Cells
were incubated for 30 min at room temperature in the dark, then washed/centrifuged at
300g/5min with PBS containing 1% fetal bovine serum (FACS Buffer).
Cells were then treated with 20µl of Fc Receptor Binding Inhibitor Polyclonal Antibody (“FC
block”, Invitrogen) and incubated at 4C for 10 minutes. Fluorescent surface antibodies (see Table
2) were centrifuged at 10^4G for 10 minutes then added to Brilliant Stain Buffer (BD Biosciences)
and added to the cell/FC block suspension. Cells were incubated for 30min at 4C and then
washed/centrifuged at 300g/5min. Cells were then permeabilized using Foxp3 Transcription
Factor Staining Buffer Set (eBioscience) for 1 hour at 4C. Cells were then centrifuged/washed
twice with Perm/Wash buffer (eBioscience) at 700g/5min. Intracellular antibodies (Table 4) were
                                                76


centrifuged at 10^4G for 10 minutes and then added to Perm/Wash Buffer and cells. Cells were
incubated for 30min at 4C and then centrifuged/washed twice with Perm/Wash buffer. 100µl of
Cytofix (BD Biosciences) was added to the cells and incubated for 15min at 4C. Cells were
centrifuged at 700g/5min, the supernatant discarded, and then resuspended in FACS buffer and
subsequently analyzed on a 5 Laser Cytek Aurora (Cytek Biosciences, Fremont, CA).
                                             77


Table 4. Fluorescent antibodies used in Chapter 3.
        Target            Fluorophore          Clone  Manufacturer
        CD 45               BUV 615            I3/2.3      BD
         B220               BUV 805          RA3-6B2       BD
        OX40L                BV421            RM134L       BD
         CD19                BV570              6D5    Biolegend
         CD40                SB600              1C10   Invitrogen
         IgG1                BV650            RMG1-1   Biolegend
       GATA-3                BV711            L50-823      BD
         CD86                BV785              GL-1   Biolegend
   I-A/I-E (MHC II)           FITC              2G9    Biolegend
         CD80              PerCPCy5.5         16-10A1      BD
         CD23           PerCP-eFluor 710        B3B4   Invitrogen
         CD25                Pe-Cy7              3C7   Biolegend
        CD138                AF647             281-2   Biolegend
         CD69             APC-Fire 750         H1.2F3  Biolegend
        CD267                  PE               8F10   Biolegend
         CD27                 APC             LG.3A10  Biolegend
                                             78


Statistical Analysis
A total of 3 experiments was conducted and one wildtype mouse and one Nrf2 -/- mouse was
used in each experiment, for a total n=3 per group. 3 technical replicates were included in each
experiment. Unmixed FCS files from Cytek Aurora were analyzed using FlowJo (BD Biosciences,
Franklin Lakes, NJ). Calculations and graphical visualization of the results were done using PRISM
Graphpad 9.20 software (La Jolla, Ca) using 2-Way ANOVA and Tukey’s test for post hoc multiple
comparisons. Statistical comparisons were made against the vehicle (VEH) treatment group and
the wild-type genotype.
                                                 79


Results
tBHQ shows cytotoxicity at higher concentrations
At 48 hours, the percentage of living cells is generally greater in the activated groups compared
to the unactivated background group, indicating that treatment cell activation increases survival
(Fig. 12). At 1µM and 5µM tBHQ, there is a significant decrease in viability in both wildtype and
Nrf2-deficient splenocytes, suggesting either an Nrf2 independent cytotoxic effect, or a
significant percentage of cells that were not activate, which made them more susceptible to
death. Decreased viability at these concentrations was unexpected, as we did not observe similar
issues in other in vitro models using different cell types, e.g., T cell activation [365].
                                                 80


Figure 12. Viability changes are seen at 1µM and 5µM of tBHQ. Splenocytes were treated with
either Vehicle (0.005% Ethanol) or an escalating concentration of tBHQ. After 30 min, Splenocytes
were activated using our activation cocktail. After 48 hours, cells were harvested and labeled
with amine-reactive live/dead discriminator and fluorescent antibodies and analyzed using flow
cytometry. Representative dot plots of Flow cytometric data are shown. Data are presented as
the mean ± SE. N=3 per group (ns indicates not significant, * indicates significance relative to the
VEH group, p<0.05, ** indicates significance relative to the VEH group, p<0.01, *** indicates
significance relative to the VEH group, p<0.001, **** indicates significance relative to the VEH
group, p<0.0001)
                                                81


tBHQ decreases the percentage of CD19+ B cells within the splenocyte population
Excluding all amine-reactive (dead) cells, we next evaluated the percentage of CD19-expressing
cells. CD19 is a surface marker expressed during all phases of B cell development, except for
terminally differentiated antibody-producing plasma cells [366]. With the increase in viability of
all cells observed in Fig.12, we expected to see a concurrent increase in CD19+ cells,
hypothesizing that the increased viability was due to the expansion of the B cell population. We
indeed observed an increase in the percentage of CD19+ cells in the activated VEH group as
compared to the unactivated background group (Fig. 13). There was a marked decrease in CD19+
cells in the presence of tBHQ, staring at 0.1µM, which became more pronounced with increasing
concentrations. At 1µM, the percentage of living CD19+ cells was comparable to the background
group. Paradoxically, there was a reversal of that trend at 5µM, however the minor increase
compared to 1µM is likely due to the small numbers of cells in that group, since the analysis
quantifies the percentage of living CD19+ cells. Given that viability was measurably affected at
5µM, the cell populations became small, and the biological significance of this observation is
                                                82


doubtful.
Figure 13. The number of CD19+ B cells is decreased by increasing concentrations of tBHQ.
Splenocytes were treated with either Vehicle (0.005% Ethanol) or an escalating concentration of
tBHQ. After 30 min, Splenocytes were activated using our activation cocktail. After 48 hours, cells
were harvested and labeled with amine-reactive live/dead discriminator and fluorescent
antibodies and analyzed using flow cytometry. Representative dot plots of flow cytometric data
are shown. Data are presented as the mean ± SE. N=3 per group. (ns indicates not significant, *
indicates significance relative to the VEH group, p<0.05, ** indicates significance relative to the
VEH group, p<0.01, *** indicates significance relative to the VEH group, p<0.001, **** indicates
significance relative to the VEH group, p<0.0001)
                                               83


Changes in forward scatter (FCS) are seen at higher concentrations of tBHQ
FCS measures the dispersion of the blue laser along the laser beam's direction of travel, thereby
measuring the size of cells. For lymphocytes, a modest increase in size and therefore median
fluorescent intensity of FCS is common after activation. We observed a small, nonsignificant
increase in FCS 48hrs after activation and a significant decrease at 1µM and 5µM (Fig 14). This
decrease was robust in both wildtype and Nrf2 deficient cells, suggesting a Nrf2-independent
effect on the size of B cells ex vivo.
                                                84


                                                                     ✱✱✱✱
                                                                      ✱✱✱✱
                                                            ✱✱✱
                                                                     ✱✱✱
                       1000000
                                                                                                           Wildtype
                       800000                                                                              NRF2 -/-
        MFI of CD19+
                       600000
                       400000
                       200000
                            0
                                 Background             0.1µM TBHQ     0.5 µM TBHQ
                                                                                     1µM tBHQ   5µM tBHQ
                                              Vehicle
                                                        Activation cocktail
Figure 14. The median fluorescent intensity (MFI) of FSC of CD19+ B cells decreases at higher
concentrations of tBHQ. Splenocytes were treated with either Vehicle (0.005% Ethanol) or an
escalating concentration of tBHQ. After 30 min, Splenocytes were activated using our activation
cocktail. After 48 hours, cells were harvested and labeled with amine-reactive live/dead
discriminator and fluorescent antibodies and analyzed using flow cytometry. Data represent the
MFI of FSC channel and are presented as the mean ± SE. N=3 per group (ns indicates not
significant, * indicates significance
                                                                       85


Figure 14 (cont’d)
relative to the VEH group, p<0.05, ** indicates significance relative to the VEH group, p<0.01, ***
indicates significance relative to the VEH group, p<0.001, **** indicates significance relative to
the VEH group, p<0.0001)
The early activation markers CD69 and CD25 are decreased by tBHQ
CD69 is the one of the first activation markers upregulated on B lymphocytes following activation.
CD25 is rapidly activated after CD69 induction. Our activation cocktail robustly induced
expression of both markers (Fig 15 A and B.). CD69 expression trended downwards with
escalating concentrations of tBHQ, reaching statistical significance at 1µM and a more
pronounced effect at 5µM (Fig 15 A). There was no difference between genotypes, suggesting a
tBHQ-independent effect. In contrast to CD69 which did not show any genotype-specific
differences, CD25 was more highly expressed by Nrf2-deficient B cells in the background group
(Fig 15B). After a robust induction, the expression of CD25 was decreased at 1µM of tBHQ. This
effect was concentration dependent.
                                                 86


Figure 15. CD19+ cells show decreased expression of the surface markers CD69 and CD25 when
exposed to high concentrations of tBHQ. Splenocytes were treated with either Vehicle (0.005%
Ethanol) or an escalating concentration of tBHQ. After 30 min, Splenocytes were activated using
our activation cocktail. After 48 hours, cells were harvested and labeled with amine-reactive
live/dead discriminator and fluorescent antibodies and analyzed using flow cytometry. Data are
presented as the mean ± SE. N= 3 per group. (ns indicates not significant, * indicates significance
relative to the VEH group, p<0.05, ** indicates significance relative to the VEH group, p<0.01, ***
indicates significance relative to the VEH group, p<0.001, **** indicates significance relative to
the VEH group, p<0.0001)
                                                 87


CD80 expression increases while CD86 expression decreases following exposure to tBHQ
The costimulatory molecules CD80 and CD86 are implicated in the T-B cell interaction, where
they bind to the T cell costimulatory molecules CD28 and CTLA-4 [342]. CD80 did not increase in
response to activation by the cocktail, but showed an unexpected increase at 5µM, potentially
due to low cell numbers (Fig. 16B). CD86 conversely showed the opposite effect: While markedly
induced by activation, CD86 expression was diminished at 1µM and 5µM (Fig 16A). While there
was a trend towards wildtype cells being more susceptible to inhibition of CD86 expression, this
did not reach statistical significance.
                                               88


Figure 16. The expression of CD86 decreases in the presence of tBHQ while CD80 remains
unchanged. Splenocytes were treated with either Vehicle (0.005% Ethanol) or an escalating
concentration of tBHQ. After 30 min, Splenocytes were activated using our activation cocktail.
After 48 hours, cells were harvested and labeled with amine-reactive live/dead discriminator and
fluorescent antibodies and analyzed using flow cytometry. Data are presented as the mean ± SE.
N= 3 per group. (ns indicates not significant, * indicates significance relative to the VEH group,
p<0.05, ** indicates significance relative to the VEH group, p<0.01, *** indicates significance
relative to the VEH group, p<0.001, **** indicates significance relative to the VEH group,
p<0.0001)
                                                89


The number of cells expressing IgG1 intracellularly is decreased in the presence of tBHQ
IgG1 is an intermediate immunoglobulin that B cells produce early in the process of isotype
switching and thus can be a proxy measurement for B cells that are class switching. We
hypothesized that the number of B cells expressing intracellular IgG1 would increase with
activation and indeed we saw a robust increase in cell counts (Fig 17). This increase was
diminished starting at 0.5µM tBHQ and decreased to baseline at 1µM and 5µM tBHQ. While there
was a trend towards increased IgG1 expression in Nrf2-deficient cells, this did not reach statistical
significance.
                                              90


                                                                             ✱✱✱✱
                                                                              ✱✱✱✱
                                                                 ✱✱✱✱
                                                                ✱✱
                                                    ✱✱✱                      ✱✱✱✱
                                           ✱✱✱                    ✱✱
                                 600
                                                                                                                    Wildtype
           IgG1+ of CD19 count
                                                                                                                    NRF2 -/-
                                 400
                                 200
                                   0
                                       Background               0.1µM TBHQ     0.5 µM TBHQ
                                                                                              1µM tBHQ   5µM tBHQ
                                                      Vehicle
                                                                Activation cocktail
Figure 17. The number of IgG1-producing cells increases with activation but decreases in
presence of tBHQ in a concentration-dependent manner. There is a non-significant trend
towards a Nrf2 dependent, tBHQ mediated decrease in IgG1 counts. Splenocytes were treated
with either Vehicle (0.005% Ethanol) or an escalating concentration of tBHQ. After 30 min,
Splenocytes were activated using our activation cocktail. After 48 hours, cells were harvested and
labeled with amine-reactive live/dead discriminator and
                                                                                         91


Fig 17 (cont’d)
fluorescent antibodies and analyzed using flow cytometry. Graphs show the cell count of IgG1+
positive cells of all CD19+ cells. Data are presented as the mean ± SE. N= 3 per group (ns indicates
not significant, * indicates significance relative to the VEH group, p<0.05, ** indicates significance
relative to the VEH group, p<0.01, *** indicates significance relative to the VEH group, p<0.001,
**** indicates significance relative to the VEH group, p<0.0001)
CD267 (TACI) expression in activated cells is inhibited by tBHQ at high concentrations
CD267 is a negative regulator of splenic B cell expansion and mediates maturation and survival
[367]. We hypothesized that CD267 expression would increase with activation. Indeed, the
percentage of CD267 expressing CD19+ cells increased with activation but was non-significantly
decreased at 1µM tBHQ and ablated at 5µM (Fig 18). There was no statistically relevant
difference between genotypes.
                                                    92


                                                                      ✱✱✱✱
                             50       ✱✱✱                              ✱✱✱✱
                                                                                                                Wildtype
           %CD267 of CD19+
                             40                                                                                 NRF2 -/-
                             30
                             20
                             10
                              0
                                  Background             0.1µM TBHQ    0.5 µM TBHQ
                                                                                          1µM tBHQ   5µM tBHQ
                                               Vehicle
                                                         Activation cocktail
Figure 18. CD267 (TACI) expression is decreased at 5µM tBHQ. Splenocytes were treated with
either Vehicle (0.005% Ethanol) or an escalating concentration of tBHQ. After 30 min,
Splenocytes were activated using our activation cocktail. After 48 hours, cells were harvested
and labeled with amine-reactive live/dead discriminator and fluorescent antibodies and
analyzed using flow cytometry. Data are presented as the mean ± SE. N=3 per group. (ns
indicates not significant, * indicates significance relative to the VEH group, p<0.05, ** indicates
significance relative to the VEH group, p<0.01, *** indicates significance relative to the VEH
group, p<0.001, **** indicates significance relative to the VEH group, p<0.0001)
                                                                                     93


The percentage of cells expressing high levels of MHCII (I-A/I-E) is decreased by exposure to
tBHQ
MHCII I-A/I-E is expressed on the surface of antigen-presenting cells, such as dendritic cells, B
cells, and macrophages. Antigen-derived peptides are presented on these molecules to CD4 T
cells. Activated B cells typically express higher densities of MHC II, while virtually all B cells express
this receptor to some degree. To assess changes in our assay, we gated on CD19+ cells with very
high expression of MHCII (Fig. 19A). CD19+ cells with high expression of MHCII were indeed
increased upon activation with our cocktail (Fig 19B). At 1µM and 5µM we saw a significant
decrease in the MHCII high population, in both the wildtype and Nrf2 deficient cells, suggesting
an Nrf2 independent effect.
                                                   94


Figure 19. The number of cells highly expressing MHC II decreases at 5µM tBHQ. A shows
representative gating strategy for MHC II High B cells on CD19+ living cells. B shows that
activation significantly increased the percentage of MHCII high cells, which decrease at 1µM
and 5µM. Splenocytes were treated with either Vehicle (0.005% Ethanol) or an escalating
concentration of tBHQ. After 30 min, Splenocytes were activated using our activation cocktail.
After 48 hours, cells were harvested and labeled with amine-reactive live/dead discriminator
and fluorescent antibodies and analyzed using flow cytometry. Data are presented as the mean
± SE. N=3 per group. (ns indicates not significant, * indicates significance relative to the VEH
group, p<0.05, ** indicates significance relative to the VEH group, p<0.01, *** indicates
significance relative to the VEH group, p<0.001, **** indicates significance relative to the VEH
group, p<0.0001)
                                                 95


Expression of CD23, an autoregulatory low-affinity IgE receptor, is decreased by tBHQ in a
partially Nrf2-dependent manner
CD23 is an autoregulatory surface protein that binds to low-affinity IgE, supposedly to capture
IgE-coated antigens in the lymphatic system or plasma and process them within the B cell [368].
CD23, in humans, can regulate the production of IgE antibody [369]. In mice, CD23 was found to
be essential to enhance CD4 T cell responses and antibody production by B cells in a mechanism
that also depends on CD11c cells [370]. CD23 is typically expressed in higher levels on activated
B cells. In the background cells, we appreciated a genotype difference, with Nrf2-deficient cells
having a higher baseline expression of CD23 (Fig 20). This Nrf-2 dependent trend was seen in all
activated groups too, reaching statistical significance at 0.5µM. At 5µM, we observed an Nrf2
independent decrease in CD23 expression.
                                                96


                                                                        ✱✱✱✱
                                                                         ✱✱✱✱
                                               ✱✱✱                      ✱
                                  ✱✱✱✱                          ✱✱✱
                             50    ✱                                     ✱✱                                       Wildtype
           %CD23+ of CD19+
                             40                                                                                   NRF2 -/-
                             30
                             20
                             10
                              0
                                  Background               0.1µM TBHQ    0.5 µM TBHQ
                                                                                            1µM tBHQ   5µM tBHQ
                                                 Vehicle
                                                           Activation cocktail
Figure 20. Genotypical difference in CD23 expression is mitigated by activation and restored at
0.5µM tBHQ. The effects are both Nrf2 dependent and at higher concentrations Nrf2
independent. Splenocytes were treated with either Vehicle (0.005% Ethanol) or an escalating
concentration of tBHQ. After 30 min, Splenocytes were activated using our activation cocktail.
After 48 hours, cells were harvested and labeled with amine-reactive live/dead discriminator and
fluorescent antibodies and analyzed using flow cytometry. Data are presented as the mean ± SE.
N=3 per group. (ns indicates not significant, * indicates significance relative to the VEH group,
                                                                                       97


Figure 20 (cont’d)
p<0.05, ** indicates significance relative to the VEH group, p<0.01, *** indicates significance
relative to the VEH group, p<0.001, **** indicates significance relative to the VEH group,
p<0.0001)
Induction of the B cell cytokine IL-6 induction is impaired in the presence of 1µM and 5µM
tBHQ
IL-6 is a master regulator of cytokine networks and is secreted by activated B cells. IL-6 is
canonically considered to be a pro-inflammatory cytokine [371] and can modulate the generation
of new TFH cells [70]. In our ex vivo activation system, we expected a rise in IL-6 secretion after
activation and a concentration-dependent decrease with tBHQ. Indeed, we saw an increase from
~10pg/ml to 60pg/ml when comparing the unactivated background to activated vehicle groups
in Fig. 21. At 0.5µM and 0.1µM tBHQ, IL-6 secretion was stable, but we saw a decrease at 1µM
and 5µM. At these concentrations, the suppression trended towards a greater decrease in the
wild-type splenocytes, but this was not statistically significant.
                                                 98


                                                                            ✱✱✱✱
                                                                             ✱✱✱✱
                                                 ✱✱✱✱                       ✱✱✱✱
                                   80   ✱✱✱✱                    ✱✱✱✱
                                   60
                       IL6 pg/ml
                                   40
                                   20
                                    0
                                        Background             0.1µM TBHQ     0.5 µM TBHQ
                                                                                            1µM tBHQ   5µM tBHQ
                                                     Vehicle
                                                               Activation cocktail
Figure 21. Production of the B cell cytokine IL-6 increases with activation but decreases at 1µM
and 5µM tBHQ. Splenocytes were treated with either Vehicle (0.005% Ethanol) or an escalating
concentration of tBHQ. After 30 min, Splenocytes were activated using our activation cocktail.
After 48 hours, cell supernatant was harvested and IL-6 content was measured using IL-6 ELISA
(Thermo Fisher, Waltham, MA). Data are presented as the mean ± SE. N=1 per group. (ns
indicates not significant, * indicates significance relative to the VEH group, p<0.05, ** indicates
significance relative to the
                                                               99


Figure 21 (cont’d)
VEH group, p<0.01, *** indicates significance relative to the VEH group, p<0.001, **** indicates
significance relative to the VEH group, p<0.0001)
                                              100


Discussion
This study demonstrates both Nrf2-dependent and -independent effects of tBHQ on B cell
activation and surface marker expression. While the results are remarkable and reproducible,
this ex vivo study simplifies the intricate T and B cell interaction, and it is impossible to predict if
the tBHQ effects on T cells will have an additive, synergistic or inhibitory effect on B cells. tBHQ
decreased the induction of the early activation marker CD69, which is rapidly expressed after
activation of lymphocytes. Emerging studies indicate CD69 plays a metabolic role in T cells [372],
though no work has been done on potential roles beyond a marker for early activation in B cells.
A secondary activation marker, CD25, is a subunit of the high-affinity IL-2 receptor with a role in
lymphocyte expansion. This is well characterized in T cells and there is emerging evidence that
there is a similar role in B cells [340].
Both CD25 and CD69 are intended in our assay as markers of B cell activation, and we saw a
robust increase with activation, and a subsequent decrease at 1µM tBHQ, indicating that B cells
are less readily activated in presence of tBHQ. This could either work through an inhibition of
activation pathways or a temporal delay in activation. To discriminate between these two
possibilities, we could implement further mechanistic studies or repeat the study at different
timepoints to establish a kinetic time course might be interesting.
We also compared the activation of the costimulatory molecules CD80 and CD86, which both
interact with either CTLA-4 or CD28 expressed on the cell surface of T cells. CD80 and CD86 are
structurally similar, but there have been indications that they convey different signals to a T cell,
where CD80 promotes a Th1 fate and CD86 signaling promotes Th2 differentiation [336, 342].
                                                 101


In this chapter, we saw that our activation cocktail strongly increased the percentage of cells
expressing CD86 but did not increase the percentage of cells expressing CD80. This may be due
to limitations of this model.
While we included the major signals required for B cell stimulation, it is impossible to fully mimic
the B cell niche required for T cell dependent activation, and it may be that there are additional
signals, e.g. cytokines, required for robust induction of CD80 expression. Treatment with tBHQ
significant decreased CD86 induction at 1µM. There was no significant change in CD80 expression
when tBHQ was added. These observations contrast with the effect of ATO on CD80/CD86
expression in IAV-activated B cells in chapter 2, where the addition of IAV challenge induced both
CD80 and CD86, but the Nrf2 activator ATO decreased only CD80 and did not change CD86. It is
important to consider the substantial differences in experimental design between the two
studies. The studies in chapter 2 were conducted on primary human immune cells derived from
peripheral blood and a virus was the B cell activator, whereas in this chapter we used mouse
splenocytes and challenged them with a B cell specific activation cocktail. Most importantly,
different toxicants were used. Nonetheless, this observation is interesting because of the
aforementioned possibility of the CD80/86 ratio modulating T cell polarization.
In addition, it is important not to draw early clinical implications out of those studies, as there is
a significant difference between murine and human B cells. Unfortunately, sources for human B
cells are somewhat limited. Peripheral blood-derived B cells are often antigen-experienced,
which limits their susceptibility to an activation cocktail as proposed here. Ideal human tissue
would be a lymph node or a spleen, but these are almost impossible to access without surgical
                                                 102


intervention. Some groups have used tonsil tissue discarded after elective tonsillectomy with
great success, but that requires an otolaryngology clinic that performs non-emergent
tonsillectomies as a collaborator to which we currently do not have access to. Furthermore, this
study focused on splenic B cells, and while they are considered a good model, populations in the
spleen differ from B cells in other compartments, as they are relatively enriched in marginal zone
B cells (359). Marginal zone B cells are unlikely to be optimally activated by our activation cocktail,
as it mimics the activation of follicular B cells by T cells. Some of the results described above were
primarily seen at concentrations that coincided with alterations in cell viability. It is impossible to
rule out that the changes might partially be due to enhanced cytotoxicity. The decreased viability
may also be due to the signal we are providing with activation being very specific to FO B cells,
and other B cell types as well as non- B cells may die due to the lack of survival signals. During
analysis of flow cytometric data, we ensured that we only gated on cells that were neither amine
reactive nor had FSC/SSC abnormalities that could be indicative of cell death. With these
measures, we can be reasonably sure that the cell populations analyzed were only viable B cells.
However, the effects of tBHQ were unexpected in magnitude, given the relatively low
concentrations used, and are relevant to human exposure (226).
This raises an important question – does environmental tBHQ exposure alter the efficacy of
antibody production in humans? At the time of writing, around 610.000.000 million doses of
various vaccinations against COV-SARS-2 were administered in the United States (360), and
around 93.000.000 cases were registered (361). Every single one of these immunological
exposures to an antigen found in either the COV-SARS-2 virus or a vaccination triggers the
                                                   103


generation of neutralizing antibodies by B cells as outlined in the introduction. Considering the
ubiquitous presence of tBHQ and other similar compounds in our environment, and the results
of this ex vivo rodent study, it would be important to explore the magnitude of these effects in
human populations.
Interestingly, the tBHQ effects observed here were independent of Nrf2. While we know that
Nrf2 is strongly activated by tBHQ and that this specifically has implications on T cell activation
[212, 314], we also described Nrf2-independent effects of tBHQ on immune cell function [162,
315]. This may be in part due to tBHQ’s propensity inhibit NF-kB signaling [316]. Our lab has
shown tBHQ effects on calcium signaling which may be Nrf2-independent [314], and potentially
a calcium-dependent mechanism may be facilitating the effects seen here. Furthermore, tBHQ
has been shown to activate the AHR [373], which could lead to Nrf2 independent effects.
In future studies, we would like to determine the effect of tBHQ on B cells at different time points,
e.g., 120 hours to be able to measure immunoglobulin secretion or 24 hours for the very early
activation markers. It would also be feasible to adapt this system to a human ex vivo system, in
which we could use human PBMC-derived B cells ex vivo and change the activation cocktail to
human orthologues of our activators, e.g. anti-human IgM F(ab), human recombinant IL-4 and
human recombinant CD40L. This would allow us to screen different toxicants quickly and
efficiently for their effects on B cell activation. We would not be able to assess Nrf2 specificity
initially for the lack of human Nrf2-deficient cells, however, we have explored the possibility of
using the CRISPR-CAS9 system to generate Nrf2deficient PMBCs.
                                                  104


In this study, we have shown in a novel ex vivo model that the common food preservative tBHQ
impacts T cell-dependent B cell activation in a partially Nrf2-dependent manner. B cell activation
and antibody production are critical to humoral immunity and therefore essential to the well-
being and host defense of mammals. We demonstrate the marked effects of tBHQ on this process
in a rodent model, which warrants further studies. We also established a rapid and easy model
to screen potential toxicants for their effect on B cell activation.
                                                105


 Chapter 4: The Food Additive tert-Butylhydroquinone Increases Plasma IgE
during transdermal allergen sensitization in correlation with increased CD40
                      and CD138 expression on B cells
                                     106


Abstract
Atopic diseases, such as allergies, are a global public health problem with concerning
epidemiological dynamics. Both the severity and incidence of food allergies are increasing in the
western world. Congruently, the use of synthetic food additives such as tert-butylhydroquinone
(tBHQ) has also been increasing. Our lab has previously described a change in T cell polarization
towards a T helper cell type 2 (TH2) phenotype when mammalian cells were exposed to tBHQ in
vitro. These effects were mostly mediated by the stress-activated transcription factor Nuclear
Factor erythroid-2 related factor 2 (Nrf2). Due to the close relationship of T and B cells in the
production of high-affinity immunoglobulins, the question of the role of tBHQ in the generation
of atopic disease arose. In this study, we demonstrate that exposure to low concentrations of
tBHQ increases the concentration of plasma IgE and changes the differentiation status and
phenotype of B cells isolated from the lymph nodes in a mouse allergy model of transdermal
sensitization. Specifically, we found that dietary tBHQ increased expression of CD40, a molecule
associated with T cell signaling and activation in B cells, and CD138, a marker of plasma cell
differentiation. Furthermore, tBHQ increased the percentage of IgE+ CD138+ cells, which are
known to be causative in allergic sensitization. In contrast, other molecules associated with B cell
activation were not different between the tBHQ and control groups. Taken together, the data
suggest that tBHQ promotes T cell signaling and plasma cell differentiation in B cells, which may
contribute to the increased plasma concentrations of IgE during transdermal sensitization.
                                                107


Introduction
Food allergies are a significant public health problem with rising concern. The incidence of
allergies to common foods, such as peanuts, shellfish, eggs, and stone fruit has been steadily
rising over the last decades [374]. Current estimations are that about 10% of children suffer from
some form of food allergy, which results in 300,000 emergency room visits and 100-150 deaths
annually [375, 376]. This phenomenon initially observed in the western world is also seen in
developing countries and correlates with their rate of economic growth [377]. With asthma being
the most prevalent atopic disease, food allergies are the second most common and have been
called the “second wave of the allergic pandemic” [374].
 Although food allergies often present as mild nonspecific discomfort, gastroenteric issues such
as diarrhea or constipation and/or airway restriction, the most serious cases results in life-
threatening acute anaphylactic shock [338]. Hospitalizations for anaphylactic shock have
increased by 13.4% between 1994 and 2004 in Australia, which keeps emergency room admission
statistics publically available at a centralized location, unlike other western countries [378].
The hygiene hypothesis has been postulated as an explanation for the rise in the incidence in
allergies, arguing that the lack of exposure to certain pathogens hampers the development of the
functional immune system [379, 380]. While these mechanisms of immune dysfunction may
contribute to the rising prevalence of atopic disease, the epidemiological dynamics suggest that
there may be other factors which remain unidentified. Notably, there has been an increase in the
exposure to synthetic food additives congruent with the increased severity and incidence of food
allergies in the same time frame. While it is important to note that this correlation does not in
                                                   108


itself mean that there is a causal relationship, I want to emphasize little work has been done to
identify the effects of food additives on the pathology of food allergies.
Our lab has demonstrated that the synthetic food additive tBHQ activates the nuclear factor
erythroid 2-related factor 2 (Nrf2) pathway and polarizes T helper cells to a type II phenotype,
which is a key step in the development of atopic disease [212]. tBHQ is widely used as a
preservative to stabilize vegetable oils in processed foods, is rapidly absorbed upon digestion and
can reach blood concentrations exceeding 200µM [234]. The Food and Agriculture
Organization/World Health Organization meeting in 1999 reported that in a typical western diet,
folks may be exposed to 1100% of the allowed daily intake of tBHQ (0.7mg/kg body weight) [239].
Research from our lab suggests Nrf2 as a possible mediator of many of the immune effects of
tBHQ. Nrf2 is a member of the cap’n’collar subfamily and has a highly conserved CNC domain
[381]. Nrf2 can be activated by cell stress, including oxidative and electrophilic stressors [359].
tBHQ is also a robust Nrf2 activator through direct interaction with thiol groups of cysteine
molecules on the Keap1 protein, which disrupt the binding of Keap1 to Nrf2 [382].Under
homeostatic conditions, most Nrf2 is bound by its repressor protein Keap1, which
polyubiquitinates Nrf2 and thus directs it to the proteasome forl degradation [360, 383].
Food allergy, by definition, is an IgE-dependent reaction to food antigen. The most severe
manifestation of food allergies, anaphylactic shock, occurs when IgE bound on mast cell Fcε
receptors binds to an antigen, crosslinks, and causes mast cell degranulation[384]. IgE is classified
into two subforms, high-affinity IgE and low-affinity IgE. Low-affinity IgE is involved in the defense
against helminths and large multicellular parasites[95, 105], while high-affinity IgE is the product
                                                 109


of the germinal center reaction and somatic hypermutation [108, 385]. In serum, IgE has a very
short half-life of ~2 days [100], but once bound to Fc ε receptors on mast cells, IgE is extremely
stable [59]. The decision as to whether a B cell produces IgE or another immunoglobulin subclass
depends heavily on the interaction of the B cell with TFH cells in the germinal cente reaction and
the signals provided to the B cell [385]. As our lab has previously described polarization changes
in CD4 T cells upon exposure to tBHQ in vitro [212, 314, 316] and changes in B cell activation in
vitro [162], the purpose of the present study is to determine the effects of low-dose oral tBHQ
on B cell activation in an experimental mouse food allergy model.
                                                110


Methods and Materials
Materials
The sources of fluorescent antibodies are shown in table 5. Rodent diets were purchased from
Dyets (Bethlehem, PA) and stored at -20C. PBS for cell preparation was purchased from Lonza
(Thermo-Fisher, Waltham, MA). All other reagents were purchased from Sigma Aldrich (St. Louis,
MO) unless otherwise indicated.
Sensitization
All animal protocols are in compliance with the Guide for the Care and Use of Animals and were
approved by the Institutional Animal Care and Use Committee (IACUC) at Michigan State
University. 4-week-old female BALB/CJ were acquired from the Jackson Laboratory and
maintained with either a normal AIN-93G (0.0014% tBHQ) diet or modified tBHQ-free AIN-93G.
Diets were procured from Dyets Inc, Bethlehem PA. After 1 week, mice were subjected to a
thoracodorsal shave. One day later, 100µl of chicken ovalbumin (Sigma Aldrich, Burlington, MA)
(100mg/ml) in sterile normal saline (allergy group, n=5) or sterile normal saline (control group,
n=5) were applied to their exposed thoracodorsal skin and covered with a commercially available
adhesive wound dressing (Meijer, Grand Rapids, MI). Sensitization was repeated 7 days and 14
days later and mice were sacrificed three days post last sensitization. The total number of mice
was 5 per group.
                                               111


OVA-specific IgE Quantification
Plasma was isolated from cardiac blood and analyzed for IgE using OVA-specific IgE ELISA
(Biolegend, LaJolla, CA). ELISA was read using a Tecan Infinite M1000 (Tecan Trading AG,
Hombrechtikon, Switzerland).
Cell preparation and flow cytometry labeling
Lymph nodes were placed between glass slides and mechanically dissociated by rubbing the
slides together. Single cell suspension was washed twice and plated out in 96 well U bottom
plates (Greiner, Kremsmuenster, Austria). The cells were then resuspended in FACS buffer
(1%FBS in DPBS) and stained with amine-reactive live/dead blue (Invitrogen, Waltham MA). After
30 min, cells were washed twice with FACS buffer and anti-Fc-gamma receptor I, II, III (anti-CD16,
anti-CD32, anti-CD64, Invitrogen, Waltham, MA) was added to diminish non-specific binding.
After 15 minutes, cells were labeled with fluorescent surface antibodies (see Table 1) and
incubated for 30 min. Cells were then permeabilized using the mouse FOXP3 intracellular staining
kit (BD Pharmingen, San Diego, CA) according to the manufacturer’s instructions and labeled with
intracellular antibodies (see Table 2). After 30 min, cells were washed twice with Perm/Wash
Buffer and fixated using Cytofix (BD Pharmingen, San Diego, CA). Cells were then resuspended in
FACS buffer and analyzed on a Cytek Aurora (Cytek Biosciences, Fremont, CA). The data
presented herein were obtained using instrumentation in the MSU Flow Cytometry Core Facility.
The facility is funded in part through financial support from Michigan State University’s Office of
Research & Innovation, College of Osteopathic Medicine, and College of Human Medicine. Flow
                                                 112


cytometric data were analyzed using FlowJo (BD Bioscience, Waltham, MA) and exported to
Graph Pad Prism (GraphPad, LaJolla, CA).
Table 5. Fluorescent Antibodies used in Chapter 4.
         Target            Fluorophore          Clone    Manufacturer
          CD 4               BUV 496            GK1.5          BD
         OX 40               BUV 563            OX-40          BD
         CD 45               BUV 615            I3/2.3         BD
          B220               BUV 805           RA3-6B2         BD
         OX40L                BV421            RM134L          BD
           IgE                BV480             R35-72         BD
          CD19                BV570              6D5       Biolegend
          CD40                SB600              1C10      Invitrogen
          IgG1                BV650            RMG1-1      Biolegend
        GATA-3                BV711            L50-823         BD
          CD86                BV785              GL-1      Biolegend
    I-A/I-E (MHC II)           FITC              2G9       Biolegend
          PD-1                BB-700              J43          BD
          CD80              PerCPCy5.5         16-10A1         BD
          CD23          PerCP-eFluor 710         B3B4      Invitrogen
         CXCR5                  PE             L138D7      Biolegend
                                            113


 Table 5 (cont’d)
           IL-4            PE-Dazzle 594           11B11         Biolegend
         FOXP3                 Pe-Cy5             FJK-16s        Invitrogen
          CD25                 Pe-Cy7                3C7         Biolegend
          Bcl-6                 APC                 7D1          Biolegend
         CD138                 AF647               281-2         Biolegend
           CD3                 AF700                17A2         Biolegend
          CD69              APC-Fire 750          H1.2F3         Biolegend
         CD40L                 SB 436               MR1              BD
         CXCR5                BUV737                2G8              BD
           IgM                 BV480              R6-60.2            BD
Analysis
Flow cytometric data were analyzed using FlowJo (BD Bioscience, Waltham, MA) and exported
to Graph Pad Prism (GraphPad, LaJolla, CA). ELISA was read using a Tecan Infinite M1000 (Tecan
Trading AG, Hombrechtikon, Switzerland) and exported to GraphPad Prism (GraphPad, LaJolla,
CA).
Statistical analysis was done using a two-way ANOVA and Tukey’s post hoc test comparing
average of the different diet groups and sensitization groups (row and column effect). Results
show average ± SEM. ns indicates not significant, * indicates significance p<0.05, ** indicates
significance p<0.01, *** indicates significance p<0.001, **** indicates significance p<0.0001
                                                114


Results
Serum IgE is increased upon exposure to tBHQ
An increase in plasma IgE is one of the hallmarks of food allergy. While being a very rare
immunoglobulin isotype, IgE plasma concentrations rise considerably during allergic sensitization
or parasitic infections. In this study, we measured serum IgE specific for OVA on day 17 after the
first exposure to OVA. In Fig 22, we show that at baseline, both mice on the non-tBHQ diet and
mice on the tBHQ diet have comparably low levels of OVA-specific IgE. Our sensitization protocol
increases IgE about 7-fold in the non-tBHQ mice, indicating a robust response to the sensitization.
In the non-tBHQ group, however, serum IgE is increased 16fold, indicating that dietary tBHQ
markedly increases the plasma OVA-specific IgE levels in our transdermal sensitization model
(p<0.01).
                                                115


                                                   ✱✱✱✱
                                    25       ✱✱✱✱      ✱✱
                                                                   No TBHQ
               IgE in Serum ng/ml
                                    20                             0.0014% TBHQ
                                    15
                                    10
                                     5
                                     0
                                         Control     Allergic
Figure 22. tBHQ diet increases Serum IgE after sensitization. 5-week-old mice where sensitized
via their thoracodorsal skin against OVA 3 times and harvested 3 days after the third sensitization.
Blood was collected using cardiac puncture and serum isolated using heparinated vacuettes
(Greiner, Kremsmünster, Austria). Serum was analyzed for OVA-specific IgE using ELISA
(Biolegend, La Jolla, CA). Data are presented as the mean ± SE. N=5 per group. (ns indicates not
significant, * indicates significance p<0.05, ** indicates significance p<0.01, *** indicates
significance p<0.001, **** indicates significance p<0.0001)
CD19-expressing cells increase with sensitization but show little difference in treatment
groups
To further characterize the B cell response to tBHQ, we used high-dimensional spectral flow
cytometry on lymph node cells to phenotype the B cell response in this model. In Fig 23, we show
that the percentage of CD19-expressing cells within lymph nodes increased significantly with the
                                                          116


sensitization protocol (p<0.0001) with no statistically significant difference between the tBHQ
and control groups.
                                                  ✱✱✱✱
                                            ✱✱✱✱
                                   40
                                                                 No TBHQ
               % CD19+ of living
                                   30
                                                                 0.0014% TBHQ
                                   20
                                   10
                                    0
                                        Control     Allergic
Figure 23. The percentage of CD19+ cells increases with sensitization. 5-week-old mice where
sensitized via their thoracodorsal skin against OVA 3 times and harvested 3 days after the third
sensitization. Axillary and brachial lymph nodes were isolated, mechanically dissociated, and
single cell suspension of lymphocytes was labeled with fluorescent antibodies and analyzed on a
Cytek Aurora spectral flow cytometer. Gating strategy involved FCS/SSC selection, doublet
exclusion, exclusion of dead cells and gating on CD19+ cells. DATA ARE PRESENTED AS THE MEAN
± SE. N=5 per group(ns indicates not significant, * indicates significance p<0.05, ** indicates
significance p<0.01, *** indicates significance p<0.001, **** indicates significance p<0.0001)
Induction of the early activation marker CD69, but not CD25, after second sensitization
Expression of both CD69 and CD25 are induced shortly after B cell activation and thus these
molecules are sometimes used as activation markers. Three days after the second sensitization,
                                                         117


we observed a robust increase in CD69 expression in the sensitized groups with a statistically
non-significant increase in the tBHQ group (Fig 24A). Neither sensitization nor tBHQ had any
effect on the expression of the early-activation marker CD25 (Fig 24B). Given that this time point
is early in the development of allergy, we hypothesize that at later time points there may be an
increase in the expression of CD25.
Figure 24. Sensitization increases CD69 but not CD25.5-week-old mice where sensitized via their
thoracodorsal skin against OVA 3 times and harvested 3 days after the third sensitization. Axillary
and brachial lymph nodes were isolated, mechanically dissociated, and single cell suspension of
lymphocytes was labeled with fluorescent antibodiesand analyzed on a Cytek Aurora spectral
flow cytometer. Gating strategy involved FCS/SSC selection, doublet exclusion, exclusion of dead
cells and gating on CD19+ cells. Data are presented as the mean ± SE. N=5 per group. (ns indicates
                                                118


Figure 24 (cont’d)
not significant, * indicates significance p<0.05, ** indicates significance p<0.01, *** indicates
significance p<0.001, **** indicates significance p<0.0001)
Increased expression of the costimulatory molecule CD86, but not CD80, following sensitization
CD80 and CD86 are co-stimulatory molecules that interact with CD28 on T cells during B cell
activation [59]. CD80 and CD86 are considered immunologically similar with respect to T cell
activation but have different signaling properties. CD80 has been shown to promote Th1
polarization in T cells, while CD86 has been associated with Th2 polarization [337]. Th2
polarization is particularly important for this study as allergies are traditionally associated with
Th2 polarized T cells [108] and our previous studies showed tBHQ promotes Th2 polarization
[212, 315]. In Fig 25, we show that our sensitization protocol does not alter the surface expression
of CD80 (Fig 25A) but increases the expression of CD86 (Fig 25B), which is consistent with the
hypothesis that CD86 promotes a Th2-like response. There was no difference between the
control and tBHQ groups.
                                                 119


Figure 25. Sensitization does not increase the costimulatory molecule CD80 but does increase
CD86. 5-week-old mice were sensitized via their thoracodorsal skin against OVA 3 times and
harvested 3 days after the third sensitization. Axillary and brachial lymph nodes were isolated,
mechanically dissociated, and single cell suspension of lymphocytes was labeled with fluorescent
antibodies and analyzed on a Cytek Aurora spectral flow cytometer. Gating strategy involved
FCS/SSC selection, doublet exclusion, exclusion of dead cells and gating on CD19+ cells. A shows
the percentage of CD80+ cells of CD19+ cells. B shows the percentage of CD86+ of CD19+. Data
are presented as the mean ± SE .N=5 per group. (ns indicates not significant, * indicates
significance p<0.05, ** indicates significance p<0.01, *** indicates significance p<0.001, ****
indicates significance p<0.0001)
                                               120


Expression of the costimulatory marker CD40 is increased by tBHQ at baseline and further
increases with allergic sensitization
CD40 is a costimulatory molecule of the tumor necrosis factor receptor superfamily. It is
expressed on APCs and interacts with CD154 (CD40-L) on T cells. This interaction is required for
the germinal center reaction, isotype switching, and formation of plasma cells. FO B cells express
CD40 constitutively; however, expression of CD40 is increased following FO B cell activation. We
defined these cells as CD40 high. In our study, we found that even at baseline, mice on a tBHQ
diet had a somewhat higher percentage of CD40-high cells (Fig 26). Sensitization increased the
percentage of CD40+ high cells, but this increase was much more pronounced in mice in the tBHQ
group compared to those on control diet (p<0.01).
                                                121


                                                         ✱
                                                     ✱
                                     2.0
                                            ✱✱               ✱✱     No TBHQ
              % CD40 high of CD19+
                                     1.5
                                                                    0.0014% TBHQ
                                     1.0
                                     0.5
                                     0.0
                                           Control       Allergic
Figure 26. The percentage of B cells highly expressing the costimulatory molecule CD40 is
increased with tBHQ consumption, even in unsensitized mice. 5-week-old mice where
sensitized via their thoracodorsal skin against OVA 3 times and harvested 3 days after the third
sensitization. Axillary and brachial lymph nodes were isolated, mechanically dissociated, and
single cell suspension of lymphocytes was labeled with fluorescent antibodies and analyzed on a
Cytek Aurora spectral flow cytometer. Gating strategy involved FCS/SSC selection, doublet
exclusion, exclusion of dead cells and gating on CD19+ cells. Shown is the percentage of CD40+
cells of CD19+ cells. Data are presented as the mean ± SE. N=5 per group. (ns indicates not
significant, * indicates significance p<0.05, ** indicates significance p<0.01, *** indicates
significance p<0.001, **** indicates significance p<0.0001)
Increase in IgE following sensitization was not different between tBHQ and control groups
IgG1 is an intermediate in the isotype switching process to IgE[105]. We used intracellular
labeling to detect whether IgG1 or IgE were expressed by CD19+ cells . Unexpectedly, there was
                                                             122


no difference in intracellular IgG1 (Fig 27A) in our study, independent of sensitization status or
tBHQ exposure. In contrast, IgE expression increased with allergic sensitization in both diet
groups (Fig 27B). While we saw a sizeable 40% increase in IgE+ expressing cells (non-tBHQ mice=
2.47%, tBHQ mice = 3.54%) in sensitized mice on the tBHQ diet as compared to those on control
diet, this effect was not statistically significant (p=0.21)
Figure 27. Sensitization has no effect on IgG1 but modestly increases IgE.4-week-old mice where
sensitized via their thoracodorsal skin against OVA 3 times and harvested 3 days after the third
sensitization. Axillary and brachial lymph nodes were isolated, mechanically dissociated, and
single cell suspension of lymphocytes was labeled with fluorescent antibodies and analyzed on a
Cytek Aurora spectral flow cytometer. Gating strategy involved FCS/SSC selection, doublet
exclusion, exclusion of dead cells and gating on CD19+ cells. A shows the percentage of CD19+
cells expressing intracellular IgG1. B shows the percentage of CD19+ cells expressing intracellular
                                                   123


Figure 27 (cont’d)
IgE. Data are presented as the mean ± SE. N=5 per group. (ns indicates not significant, * indicates
significance p<0.05, ** indicates significance p<0.01, *** indicates significance p<0.001, ****
indicates significance p<0.0001)
The percentage of CD138 expressing cells among all lymph node-derived cells is increased
with both sensitization and tBHQ exposure
CD138 (Syndecan-1) is a marker of plasma cell differentiation and indicates that cells have started
producing antibodies [386]. At this stage, B cells typically cease to express traditional B cell
markers such as CD19 [387]. We evaluated the percentage of CD138-expressing cells in the lymph
nodes and found that there was a baseline increase in CD138-expressing cells when mice were
on the tBHQ diet (Fig 28A). Sensitization increased CD138 expression in both the non-tBHQ and
the tBHQ groups, but the increase in the tBHQ group was more marked. This was somewhat
unexpected, as typically cells exit the lymph node following isotype switching and home to either
the bone marrow or peripheral tissues where they produce antibodies. We then measured the
number of CD138+ cells that expressed IgE intracellularly, which represents a rare population
among lymphocytes. Again, we saw a minuscule, but statistically significant increase in IgE+
plasma cells in unsensitized mice on the tBHQ diet (Fig 30B). Sensitization enhanced the
difference in CD138 expression between treatment groups significantly (Fig 28B). This
observation dovetails with Figure 22, the plasma IgE data. Overall, the data suggest that tBHQ
may promote plasma cell differentiation, and specifically favor IgE-producing plasma cells.
                                                124


Figure 28. The overall percentage of plasmablasts is elevated in mice fed tBHQ, and the
percentage of IgE producing plasma cells increases with both tBHQ and sensitization. 5-week-
old mice where sensitized via their thoracodorsal skin against OVA 3 times and harvested 3 days
after the third sensitization. Axillary and brachial lymph nodes were isolated, mechanically
dissociated, and single cell suspension of lymphocytes was labeled with fluorescent antibodies
and analyzed on a Cytek Aurora spectral flow cytometer. Gating strategy involved FCS/SSC
selection, doublet exclusion and exclusion of dead cells. A: In the remaining live lymphocyte
population, CD138 surface expression was measured. B: Among CD138+ cells, presence
intracellular IgE was measured. Data are presented as the mean ± SE. N=5 per group. (ns indicates
not significant, * indicates significance p<0.05, ** indicates significance p<0.01, *** indicates
significance p<0.001, **** indicates significance p<0.0001)
                                               125


Discussion
The studies presented here show that exposure to tBHQ via ingestion increases serum IgE and
alters B cell phenotypes in this murine model of allergic sensitization to ovalbumin. Allergic
sensitization increased the expression of CD69, CD86, CD40 and intracellular IgE on CD19-
expressing B cells and increased the percentage of plasma (CD138-positive) cells within the lymph
node. IgE is the key molecule in allergic and anaphylactic reactions. In this study, we show a direct
impact by the food additive tBHQ on IgE blood concentrations specific to the model antigen. It is
important to acknowledge that the dosage of tBHQ used here in this study is very low – it is the
standard dose used in rodent chow and less than what model diets estimate for the exposure of
a typical western consumer. While this is a specific allergy model in mouse, it suggests the need
to consider the impact of tBHQ in humans and raises the question whether tBHQ may alter or
impair humoral immunity.
Similarly to the chapters before, we measured the percentage of cells expressing CD80 and CD86,
as they are implicated in modulating T cell-B cell interactions and in potentially influence Th1 or
Th2 polarization [336, 342]. There is evidence to suggest that CD86 signaling promotes
polarization of Th2 cells, a cell type that is known to be causative in allergy, and thus highly
relevant to this chapter. Indeed, we saw that mice that were sensitized had a higher percentage
of CD86-expressing cells. The percentage of CD80-expressing cells showed a large amount of
variation and neither sensitization nor oral tBHQ had an impact. These observations are
consistent with the idea that increased CD86 signaling is associated with a Th2 response, as
                                                 126


allergies are typically considered to be Th2-mediated. The hypothesis following this, namely that
oral tBHQ increases CD86 expressing B cells, is not supported by the data shown in this chapter.
This study focused on 4-week old, female Balb/cj mice. This reductionist study was designed to
exclude confounding factors such as age (which we assess in chapter 5), sex, and inbred mouse
strain. There are important sex differences in mice, specifically with respect to mast cell function,
a downstream effector cell type in food allergies [388]. In general, sex differences in the immune
response are well described in mouse models [389] Clinically, in humans, there is a slight skew
towards females having a higher incidence of atopic disease and allergies [390].
A discrepancy we observed in this study was the very robust increase in plasma OVA-specific IgE,
while the number of IgE-producing B cells was not significantly altered by tBHQ (Fig27B). One of
the key things to consider here is that this is a very dynamic process in which B cells develop into
either memory cells, plasmablasts or plasma cells. Fig 28B shows that among CD138+ cells of
lymphoid lineage, IgE expression is significantly increased when tBHQ is present. CD138+ CD45+
cells signify either plasmablasts or plasma cells, which are committed to the antibody-producing
role. The ELISA results shown in Fig 22 show a more marked increase in OVA-IgE plasma
concentrations; however, this may be reflective of differences in the parameters as OVA-IgE was
quantified in Fig. 22, whereas total IgE+ 138+ cells were quantified in Fig. 27. In addition, protein
quantification by ELISA is more sensitive than intracellular labeling and flow cytometry.
 The ELISA shown In Fig 22 shows a more marked increase, but we must remember that not only
are ELISA typically more sensitive than flow cytometric data, but also that this ELISA measured
                                                  127


OVA-specific IgE compared to general IgE, which is a more precise tool to quantify the response
to sensitization.
This study is limited in that it only evaluates cells in the lymph node and therefore in only one
compartment of the body. We selected the lymph node because it is the site of the germinal
center reaction and somatic hypermutation, which arguably have the largest impact on antibody
production and humoral immunity. Future studies into other compartments, such as peripheral
blood and bone marrow, would likely also be informative to include—particularly with respect to
characterization of the effect of tBHQ on plasma cell differentiation. These findings are important
as they indicate that tBHQ promotes allergic sensitization in this model and suggest the need to
investigate the effect of tBHQ on allergy in humans. The data also suggest tBHQ may also impact
humoral immunity and could potentially alter the efficacy of vaccinations and the generation of
high affinity antibodies as a response to seasonal, epidemic, or pandemic infections. Further, the
role of tBHQ in the global rise of atopic diseases should be considered. The increasing exposure
of the average western consumer to synthetic food additives such as tBHQ may contribute to the
prevalence and increasing incidence of food allergies and related diseases, such as asthma,
eczema, and allergic rhinitis, in industrialized countries. The significant morbidity and burden of
allergic disease on patients warrants further investigation into the toxicological implications of
these findings.
Taken together, in this study, we have shown that a very low dose of the synthetic food additive
tBHQ, as found in standard mouse chow and standard western diets, promotes the development
of allergy in a mouse model closely mimicking human allergy pathogenesis.
                                                 128


Chapter 5: The impact of age on murine B cell response to transdermal
                            sensitization
                                 129


Abstract
Food allergies are a common and sometimes life-threatening problem encountered in up to 8%
of the pediatric population. The incidence of food allergies is increasing in the U.S. and globally.
To study food allergies, robust and physiologically relevant animal sensitization models are
required. Many models use adjuvants such as cholera toxin and physiologically-irrelevant
delivery methods of the potential allergen, such as intraperitoneal injection.
In contrast, transdermal sensitization without an adjuvant closely mimics dermal absorption of
food-derived molecules in humans. We have previously shown that dietary exposure of 5-week-
old mice to the synthetic food additive tert-butylhydroquinone promotes allergic sensitization
resulting in increased plasma concentrations of IgE and increased numbers of plasmablasts. In
this study, we investigated the impact of age on the increased allergic sensitization in mice on a
tBHQ diet. We found that 5-week-old mice on a tBHQ diet did not show a statistically significant
difference in the expression of the activation markers, CD69, CD25, CD80 or CD86, in comparison
to those on control diet. Nor did tBHQ impact the number of IgE-producing B cells in 5-week-old
mice. However, tBHQ increased the number of plasmablasts in 5-week-old mice, but this effect
was diminished with age. Furthermore, while transdermal sensitization to OVA increased the
number of IgE+ plasmablasts in 5-week-old mice, this increase was diminished in 8-week-old and
16-week-old mice, suggesting that older animals may be less susceptible to allergic sensitization
in this model. Overall, this study shows that age significantly blunts transdermal sensitization in
this model and diminishes the enhancing effect of tBHQ on allergic sensitization.
                                                130


Introduction
Food allergies are a significant public health concern [391]. About 8% of children and 4% of adults
have some form of allergic reaction to food-derived antigens during their lifetime [392, 393] and
the incidence has been rising over the past few years [392]. In some western states, the rate of
the most severe outcome of food allergies, anaphylactic shock, increases by 10% every 10 years
[374, 377]. In the previous study we have shown that the synthetic food additive tBHQ can
increase the production of IgE and the number of plasma cells following sensitization. tBHQ is an
activator of Nrf2 and has been shown to promote Th2, while inhibiting Th1, polarization [212].
Nrf2 belongs to the subfamily of Cap’n’collar proteins and has a well-conserved CNC domain
[381]. As stated above, Nrf2 can be activated by tBHQ, but also by a host of cellular stressors,
such as heavy metals, food additives, and nucleophilic compounds [359]. Under non-stress
conditions, Nrf2 is bound to its repressor protein Keap-1. Upon activation by the aforementioned
compounds, Nrf2 translocates to the nucleus and activates the antioxidant response element,
triggering the expression of cytoprotective genes [360, 383, 394]. Under non-stress conditions,
Keap-1 polyubiquitinates Nrf2 and directs it thus to proteasomal degradation [360]. In this study,
we explore the impact of age at the start of sensitization by comparing mice that started
sensitization at 5 weeks of age, 9 weeks of age and 16 weeks of age. We hypothesize that similar
to epidemiological observations in humans [391], older mice are less susceptible to the
development of food allergies.
                                                 131


Methods
Materials
The sources of fluorescent antibodies are shown in table 5. Rodent diets were purchased from
Dyets (Bethlehem, PA) and stored at -20C. PBS for cell preparation was purchased from Lonza
(Thermo-Fisher, Waltham, MA). All other reagents were purchased from Sigma Aldrich (St. Louis,
MO) unless otherwise indicated.
Sensitization
All animal protocols are in compliance with the Guide for the Care and Use of Animals and were
approved by the Institutional Animal Care and Use Committee (IACUC) at Michigan State
University. BALB/cj mice were acquired from Jackson Laboratories and housed in a specified
germ-free environment. Mice were fed either AIN-93G rodent chow (containing 0.014% tBHQ)
or tBHQ-free AIN-93G. At 4, 8, or 16 weeks of age, mice were shaved thoracodorsically to expose
the epidermis. One day later, 100µl of 100mg/ml chicken ovalbumin (Merck, Darmstadt,
Germany) in 0.9% saline or just 0.9% saline were applied to the backs of the mice and
subsequently covered with a commercially available flexible fabric bandage (Johnson & Johnson,
New Brunswick, NJ). The procedure was repeated every 7 days for 3 weeks for mice in the 5-week
age group and for 4 weeks in the 9 and 16-week age group. Lymph nodes and blood were
harvested three days after their last sensitization. Lymph nodes were mechanically dissociated
to collect lymphocytes. Brachial and axial lymph nodes of each mouse were processed together
and pooled, to account for anatomic variation in lymph node size and grouping. Each sample
represents lymphocytes from a single mouse, pooled from axillary and brachial lymph nodes,
                                               132


which we have previously determined to be the primary draining lymph nodes. Lymphocyte
single cell suspension in PBS was washed twice prior to culturing the cells in 96 well U bottom
microplates (Greiner, Kremsmünster, Austria).
Flow cytometry
Cells were washed twice in Phosphate buffered saline and stained with amine-reactive live/dead
blue (Invitrogen, Waltham MA). After 30 min, cells were washed twice with FACS buffer, and anti-
Fc-gamma receptor I, II, III (anti-CD16, anti-CD32, anti-CD64, Invitrogen, Waltham, MA) was
added to diminish non-specific binding. After 15 minutes, fluorescent surface antibodies (see
Table 1) were added and incubated for 30 min. After a FACS wash, cells were subjected to
permeabilization using the mouse FOXP3 intracellular staining kit (BD Pharmingen, San Diego,
CA). After 1 hour, cells were washed twice using Perm/Wash Buffer (BD Pharmingen, San Diego,
CA) and incubated with intracellular antibodies (see Table 2). After 30 minutes, cells were
washed twice with Perm/Wash Buffer and fixated using Cytofix (BD Pharmingen, San Diego, CA).
Cells were then resuspended in FACS buffer and analyzed on a Cytek Aurora (Cytek Biosciences,
Fremont, CA). The data presented herein were obtained using instrumentation in the MSU Flow
Cytometry Core Facility. The facility is funded in part through the financial support of Michigan
State University’s Office of Research & Innovation, College of Osteopathic Medicine, and College
of Human Medicine. FCS files were analyzed using FlowJo (BD Bioscience, Waltham, MA).
                                               133


Table 6. Fluorescent antibodies used in Chapter 5.
        Target             Fluorophore           Clone  Manufacturer
          CD 4               BUV 496             GK1.5       BD
         OX 40               BUV 563             OX-40       BD
         CD 45               BUV 615             I3/2.3      BD
         B220                BUV 805           RA3-6B2       BD
         OX40L                BV421            RM134L        BD
            IgE               BV480             R35-72       BD
         CD19                 BV570               6D5    Biolegend
         CD40                 SB600               1C10   Invitrogen
          IgG1                BV650            RMG1-1    Biolegend
        GATA-3                BV711            L50-823       BD
         CD86                 BV785               GL-1   Biolegend
   I-A/I-E (MHC II)            FITC               2G9    Biolegend
          PD-1                BB-700               J43       BD
         CD80               PerCPCy5.5         16-10A1       BD
         CD23            PerCP-eFluor 710         B3B4   Invitrogen
        CXCR5                   PE              L138D7   Biolegend
           IL-4           PE-Dazzle 594          11B11   Biolegend
        FOXP3                 Pe-Cy5            FJK-16s  Invitrogen
                                             134


 Table 6 (cont’d)
         CD25                 Pe-Cy7               3C7         Biolegend
         Bcl-6                 APC                 7D1         Biolegend
        CD138                 AF647               281-2        Biolegend
          CD3                 AF700               17A2         Biolegend
         CD69              APC-Fire 750          H1.2F3        Biolegend
        CD40L                 SB 436              MR1              BD
        CXCR5                BUV737                2G8             BD
          IgM                 BV480              R6-60.2           BD
Figure 29. Spectral unmixing sample of the in-vivo panel. Exemplary spectral analysis of a model
sample in which all fluorophores were present.
                                              135


Analysis
The data were compiled from 5 mice per group and are presented as the mean ± standard error
of the mean. Each data sample represents measurement from one individual mouse. Two-way
ANOVA was used to determine statistical differences between treatment groups, and Šídák’s post
hoc test was used to compare the diet groups and sensitization groups (row and column effects).
Calculations and graphical visualization of the results were done using PRISM Graphpad 9.20
software (La Jolla, Ca). ns indicates not significant, * indicates significance p<0.05, ** indicates
significance p<0.01, *** indicates significance p<0.001, **** indicates significance p<0.0001
                                                 136


Results
Transdermal sensitization results in CD19+ expansion in 5-week-old and 9-week-old mice
After sensitization, we expect an increase in the B cell population due to clonal expansion and
the germinal center reaction. The number of B cells increases during this process. In this study,
we defined B cells as CD19+ cells, and CD3- cells. Indeed, we saw a statistically significant increase
in the 5-week-old mice and the 9-week-old mice, and a diminished, non-significant trend in the
16-week-old mice (Fig. 30). The increase in the 5-week-old mice was much more robust than in
the older subjects.
                                                137


Figure 30. The percentage of CD19+ cells increases more strongly after sensitization in young
mice. Mice at different ages were sensitized via their thoracodorsal skin against OVA 3 times and
harvested 3 days after the third sensitization for the 5-week-old mice and 3 days after the 4th
sensitization for older mice. Axillary and brachial lymph nodes were isolated, and mechanically
dissociated, and single-cell suspension of lymphocytes was labeled with fluorescent antibodies
and analyzed on a Cytek Aurora spectral flow cytometer. The gating strategy involved FCS/SSC
selection, doublet exclusion, and exclusion of dead cells. Data are presented as the mean ± SE.
N=5 per group. (ns indicates not significant, * indicates significance p<0.05, ** indicates
significance p<0.01, *** indicates significance p<0.001, **** indicates significance p<0.0001)
                                                138


CD69 is a C-type lectin protein that is expressed early in lymphocyte activation.
We hypothesized that sensitization would increase the percentage of CD69-expressing B cells.
Approximately 4% of B cells express CD69 in naïve 5-week-old mice, and sensitization significantly
increases this percentage (Fig 31A). In 9-week-old mice, we see a comparably higher baseline of
CD69 expression, but no increase after sensitization. In mice aged 16 weeks at the start of the
treatment, the baseline of CD69 expression is very low with considerable variability, and there is
no significant increase in expression following allergic sensitization. tBHQ does not show any
effect in either naïve nor sensitized mice of any age group. CD25 was not affected by allergic
sensitization or tBHQ treatment (Fig 31B). Of note, there was an odd and unexpected high
percentage of CD25 expressing cells in the unsensitized, tBHQ-fed 16-week-old mice that was
statistically significant. Currently, it is not clear at this time why this spike occurred and whether
it is biologically significant.
                                                     139


Figure 31. Contrasting patterns of induction for CD69 and CD25 after sensitization. Mice at
different ages were sensitized via their thoracodorsal skin against OVA 3 times and harvested 3
days after the third sensitization for the 5-week-old mice and 3 days after the 4th sensitization
for older mice. Axillary and brachial lymph nodes were isolated, and mechanically dissociated,
and single-cell suspension of lymphocytes was labeled with fluorescent antibodies and analyzed
on a Cytek Aurora spectral flow cytometer. The gating strategy involved FCS/SSC selection,
doublet exclusion, and exclusion of dead cells. A shows the percentage of CD19+ cells expressing
CD69. B shows the percentage of CD19+ cells expressing CD25. Data are presented as the mean
± SE. N=5 per group. (ns indicates not significant, * indicates significance p<0.05, **
                                                140


Fig 31 (cont’d)
indicates significance p<0.01, *** indicates significance p<0.001, **** indicates significance
p<0.0001)
Expression of CD86, but not CD80, is induced on CD19+ B cells following allergic sensitization
The costimulatory molecules CD80 and CD86 interact with receptors at the T cell surface and are
considered robust activation markers of B cell activation. There is some evidence that the
CD80:CD86 ratio can influence T cell cytokine secretion and therefore serve as a signaling
pathway in the adaptive immune response [395, 396]. In this study, we observed some variability,
but no effect of either tBHQ nor sensitization status on CD80 expression. Interestingly, the
baseline expression of CD80 increased in the 9-week-old mice compared to 5-week-old mice and
was greatest in the 16-week-old mice (Fig 32). In contrast, CD86 expression increased with
sensitization in all age groups, though the overall induction was somewhat blunted in the 16-
week-old animals. The induction of CD86 in 16-week-old mice following sensitization was further
dampened in mice exposed to tBHQ (Fig 33).
                                              141


Figure 32. CD80 expression does not change with sensitization independent of age. Mice at
different ages were sensitized via their thoracodorsal skin against OVA 3 times and harvested 3
days after the third sensitization for the 5-week-old mice and 3 days after the 4th sensitization for
older mice. Axillary and brachial lymph nodes were isolated, and mechanically dissociated, and
single-cell suspension of lymphocytes was labeled with fluorescent antibodies and analyzed on a
Cytek Aurora spectral flow cytometer. The gating strategy involved FCS/SSC selection, doublet
exclusion, and exclusion of dead cells. Data are presented as the mean ± SE. N=5 per group. (ns
indicates not significant, * indicates significance p<0.05, ** indicates significance p<0.01, ***
indicates significance p<0.001, **** indicates significance p<0.0001)
                                                142


Figure 33. CD86 is increased with sensitization and the amplitude of this effect is inversely
correlated with age. Mice at different ages were sensitized via their thoracodorsal skin against
OVA 3 times and harvested 3 days after the third sensitization for the 5-week-old mice and 3 days
after the 4th sensitization for older mice. Axillary and brachial lymph nodes were isolated, and
mechanically dissociated, and single-cell suspension of lymphocytes was labeled with fluorescent
antibodies and analyzed on a Cytek Aurora spectral flow cytometer. The gating strategy involved
FCS/SSC selection, doublet exclusion, and exclusion of dead cells. Data are presented as the mean
± SE. N=5 per group. (ns indicates not significant, * indicates significance p<0.05, ** indicates
significance p<0.01, *** indicates significance p<0.001, **** indicates significance p<0.0001)
Exposure to tBHQ does not impact the percentage of IgE+ or IgG1+ B cells following
sensitization in any age group
Following activation and in certain cellular microenvironments, B cells may undergo antibody
class switching in which they transition away from production of IgM to producing one of the
other classes of antibody. IgE and IgG1 are classes of antibodies that are associated with type 2
immune responses and allergies. Thus, we measured whether tBHQ and allergic sensitization
                                                 143


increased the percentage of B cells expressing IgE and IgG1. In Fig 34, we found that the
percentage of IgE-expressing B cells increased robustly with sensitization in 5-week-old mice and
9-week-old mice, but that tBHQ did not have a significant effect. In 16-week-old mice, there was
greater variability with a trend towards an increase due to sensitization in the control diet group,
but not in the tBHQ group. IgG1, an immunoglobulin often seen as an intermediate precursor to
IgE, is also sometimes associated with allergy. In contrast to IgE, there was no effect of
sensitization on IgG1 in the 5-week-old mice at this time point, and generally a low baseline
expression. In the 9-week-old group, however, there was a robust increase in sensitization in
both the control diet and tBHQ groups. In 16-week-old mice, the baseline expression was very
low again, and considerable variability was observed (Fig. 35)
                                                144


Figure 34. The percentage of IgE-producing CD19+ cells increases with sensitization in the
younger groups but not in mice who start sensitization at 16 weeks. Mice at different ages were
sensitized via their thoracodorsal skin against OVA 3 times and harvested 3 days after the third
sensitization for the 5-week-old mice and 3 days after the 4th sensitization for older mice. Axillary
and brachial lymph nodes were isolated, and mechanically dissociated, and single-cell suspension
of lymphocytes was labeled with fluorescent antibodies and analyzed on a Cytek Aurora spectral
flow cytometer. The gating strategy involved FCS/SSC selection, doublet exclusion, and exclusion
of dead cells. Data are presented as the mean ± SE. N=5 per group. (ns indicates not significant,
* indicates significance relative to the VEH group, p<0.05, ** indicates significance relative to the
VEH group, p<0.01, *** indicates significance relative to the VEH group, p<0.001, **** indicates
significance relative to the VEH group, p<0.0001)
                                                145


Figure 35. IgG1-producing CD19+ cells increase in mice that start sensitization at 8 weeks, but
not in younger or older mice. Mice at different ages were sensitized via their thoracodorsal skin
against OVA 3 times and harvested 3 days after the third sensitization for the 5-week-old mice
and 3 days after the 4th sensitization for older mice. Axillary and brachial lymph nodes were
isolated, and mechanically dissociated, and single-cell suspension of lymphocytes was labeled
with fluorescent antibodies and analyzed on a Cytek Aurora spectral flow cytometer. The gating
strategy involved FCS/SSC selection, doublet exclusion, and exclusion of dead cells. Data are
presented as the mean ± SE. N=5 per group (ns indicates not significant, * indicates significance
p<0.05, ** indicates significance p<0.01, *** indicates significance p<0.001, **** indicates
significance p<0.0001)
Dietary exposure to tBHQ promotes plasmablast differentiation in 4-week-old, but not older,
mice
In addition to class-switching, B cells are also able to differentiate into plasma cells, which
produce large amounts of antibodies and are thought to be the largest producers of IgE in allergy.
The expression of CD138 can be used to identify plasma cells. Accordingly, we assessed the
                                              146


percentage of CD138+ plasma cells among all living cells (Fig 36). In the 5-week-old mice, both
the presence of tBHQ and sensitization status significantly increased the number of plasma cells
(defined as CD138+ CD45+ cells). 9-week-old mice showed a similar trend, but much less
pronounced and not statistically significant. Among 16-week-old mice, there was considerable
variability, but no observed effect of either sensitization or tBHQ.
Taking a closer look at these plasmablasts, we measured the percentage of CD138+ cells that
expressed intracellular IgE (Fig 37). Interestingly, in mice aged 5 weeks at the start of
sensitization, both dietary tBHQ and allergic sensitization increased the percentage of IgE-
producing plasmablasts. In mice 9 weeks at the start of sensitization, only sensitization increased
the percentage of plasma cells, whereas tBHQ had no effect. In mice 16 weeks at the start of
sensitization, no effect was seen with either tBHQ or sensitization. Overall, concerning IgE-
producing plasma cells, the data suggest that younger mice are more sensitive to both
transdermal sensitization and dietary tBHQ exposure.
                                                147


Figure 36. The percentage of plasmablasts increases with sensitization and with dietary tBHQ
independently in younger mice, but not in older mice. Mice at different ages were sensitized via
their thoracodorsal skin against OVA 3 times and harvested 3 days after the third sensitization
for the 5-week-old mice and 3 days after the 4th sensitization for older mice. Axillary and brachial
lymph nodes were isolated, and mechanically dissociated, and single-cell suspension of
lymphocytes was labeled with fluorescent antibodies and analyzed on a Cytek Aurora spectral
flow cytometer. The gating strategy involved FCS/SSC selection, doublet exclusion, and exclusion
of dead cells. Data are presented as the mean ± SE. N=5 per group. (ns indicates not significant,
* indicates significance p<0.05, ** indicates significance p<0.01, *** indicates significance
p<0.001, **** indicates significance p<0.0001)
                                              148


Figure 37. The percentage of plasmablasts producing IgE is increased with both sensitization
and dietary tBHQ in younger mice, and only with sensitization in mice 8 weeks old at first
sensitization. Mice at different ages were sensitized via their thoracodorsal skin against OVA 3
times and harvested 3 days after the third sensitization for the 5-week-old mice and 3 days after
the 4th sensitization for older mice. Axillary and brachial lymph nodes were isolated, and
mechanically dissociated, and single-cell suspension of lymphocytes was labeled with fluorescent
antibodies and analyzed on a Cytek Aurora spectral flow cytometer. The gating strategy involved
FCS/SSC selection, doublet exclusion, and exclusion of dead cells. Data are presented as the mean
± SE. N=5 per group. (ns indicates not significant, * indicates significance p<0.05, ** indicates
significance p<0.01, *** indicates significance p<0.001, **** indicates significance p<0.0001)
                                                149


Discussion
In this study, we demonstrate that there is a considerable effect of age at the start of sensitization
in our transdermal sensitization model. Mice that are 5 weeks at the time of first sensitization
have a much stronger response than 9-week-old mice, which in turn have a stronger response
than 16-week-old mice. This tracks with human epidemiological data, where we see that young
children are much more likely to develop new food allergies compared to adolescents, adults, or
geriatric populations, in that order [391].
From previous, unpublished data and based on the clinical data in humans, we were concerned
that older animals, such as in the 9-week-old and 16-week-old groups, would have diminished
immune response to the allergen. Thus, we used a more robust sensitization regimen that
included 4 sensitizations. The younger mice, which we knew had a robust IgE response, were only
sensitized 3 times, as previous studies indicated that this was the optimal timepoint to detect
both B cells and TFH cells in the lymph node. On the one hand, this slightly diminishes the
meaningfulness of the comparisons made in this chapter, but on the other hand, 4 weeks of
sensitization should, in theory, enhance the effect of sensitization allowing for better
comparisons of the impact of tBHQ on allergic sensitization in animals of different ages.
Interestingly, only the 9-week-old mice showed an increase in IgG1, an immunoglobulin
considered an intermediate in IgE class switching, compared to the 5-week-old, which showed
an increase in IgE. There are multiple possible explanations for this observation. In this study, we
only observe one immunological compartment, the lymph node. B cells that differentiate into
plasma cells will eventually leave the lymph node and either infiltrate antigen-rich tissues or take
                                                150


up residence in the bone marrow, where they perpetually secrete immunoglobulins into the
bloodstream [338]. There is a possibility that the timepoint of migration is different depending
on the age of the mouse. Evaluating other compartments such as peripheral blood and bone
marrow might reveal more about the plasma cell population. A murine study where all these
immunological compartments are harvested simultaneously might shed some light on this. Also
of interest would be a time-course study, evaluating the increase in plasma IgE levels throughout
the course of sensitization at different ages. Interestingly, different allergens seem to affect
different age groups in humans – toddlers and infants suffer more often from allergy to milk and
egg, while adolescents react more often to nuts and shellfish [390, 397]. Considering this, a
mouse study evaluating the induction of food allergies to different antigens in different age
groups would further optimize this model. Once an antigen-specific optimal age is determined,
we would be able to determine the impact of tBHQ on food allergy in the context of different,
common food antigens.
Compared to adults, B cell subsets in peripheral blood, bone marrow and secondary lymphoid
organs are not well characterized. Despite a dearth of information in this area; however, it is clear
significant changes occur in the B cell population during infancy and childhood. Newborns have
relatively fewer circulating lymphocytes overall, including B cells, in blood as compared to older
infants and children [398]. Human infants have the greatest percentage of naïve B cells as
compared to older children and adults where young infants have the highest percentage of naïve
B cells which gradually decrease to adult levels by around 5 - 10 years of age. There is a
concordant increase in the memory B cell population that is inversely proportional to the naïve
                                               151


B cell population in humans. It seems likely there may be an activated B cell state that occurs
during the transition of naïve B cells to memory B cells, however this has not been characterized.
Our study shows a gradual increase in markers of B cell activation (CD25 and CD80) in mice with
age. While it is difficult to directly correlate the ages of mice with those of humans, there are
some developmental milestones that can be compared. Mice begin eating solid food at
approximately 12 days of age, which occurs in humans at around 6 months of age. Mice begin to
become reproductively mature at around 6 weeks of age, which could roughly approximate
puberty. Thus, the mice in our youngest group (5 weeks at first sensitization) could be considered
somewhat equivalent to pre-pubescent humans at the beginning of sensitization and adolescents
at the end of sensitization. Although we did not directly assess B cell memory in this study, the
increase in B cell activation markers over time suggests that mice, like humans, may have an
increase in B cell memory as juveniles develop into adults.
In this present study, we demonstrated that there is a significant effect of age on transdermal
sensitization in a rodent model. To our knowledge, we are the first group to evaluate this
parameter in this very physiologically applicable model of food allergy. This informs the design
of subsequent studies and makes a valuable argument for the enrollment of younger mice into
allergy studies.
                                                  152


Chapter 6: Conclusions
         153


Discussion of findings
In the chapters above, we investigated the impact of Nrf2 activators on various aspects of B cell
function, including polyclonal in vitro activation, antiviral immunity, and allergic sensitization.
Chapter 2 demonstrated that in an in vivo infectious disease model, the heavy metal arsenic
trioxide impairs B cell activation and function. In chapter 3, an ex vivo rodent model modeling T
cell-dependent B cell activation showed that tBHQ impacts this activation in a partially Nrf2-
dependent fashion. Chapter 4 took this finding in vivo and showed that in a murine model of
allergic sensitization, tBHQ exacerbates the allergic phenotype, including higher serum IgE, a
higher percentage of IgE-producing B cells, and higher activation status of B cells. In Chapter 5,
we evaluated the impact of age on the efficacy of the allergic sensitization model, finding that
younger mice are much more susceptible to the induction of allergy. This tracks with human
epidemiological studies. Taken together, these findings suggest that common environmental
toxicants that stimulate Nrf2 signaling can impact B cell activation and function as well as the
ratios of Immunoglobulins produced.
A common thread in these studies was that the activation markers CD69 and CD25 were impaired
in the presence of ATO and tBHQ, respectively. In chapter 2, I found that CD25 expressing cells
were increased after IAV challenge and decreased with ATO. In chapter 3, I showed that both
CD25 and CD69 increase after activation ex vivo and decrease with 1µM of tBHQ. In chapter 4
and 5, I observed that during allergic sensitization in vivo, CD69 increased, but not CD25 in mice
that are 5 weeks of age at the start of sensitization, and that effect disappeared when mice were
older at the beginning of sensitization. The common thread in between different species and
                                                 154


toxicants is that there is a decrease in activation markers when Nrf2-activating compounds are
present. With the data collected thus far; however, I cannot distinguish between an overall
inhibition versus a delay in B cell activation or some combination of these two possibilities.
Similarly, we saw disparate results regarding the expression of the costimulatory molecules CD80
and CD86. As mentioned above, CD80 and CD86 bind to their ligands CD28 and CTLA-4 on T cells
to convey stimulatory and inhibitory signals, respectively. CD86 signaling through CD28 can
stimulate T cells towards a Th2 phenotype, while CD80 signaling via CD28 can stimulate towards
a Th1 phenotype [336, 342]. In chapter 2, we saw that CD80 expression was decreased by ATO,
while CD86 was not. In chapter 3, we saw that there was no increase in CD80-expressing cells in
the activated groups, and tBHQ did not change the percentage of CD80-expressing cells. CD86 on
the other hand was increased with activation and decreased at 1µM tBHQ. In Chapter 4 and 5,
we saw that allergic sensitization in vivo did not impact levels of CD80-expressing cells, but
increased CD86-expressing cells in sensitized mice independent of their diet. Considering the
immunology of B cells in these models, this makes sense, as CD86 is the Th2-associated
costimulatory molecule. In characterizing the role of Nrf2, the image is less clear – ATO decreases
CD80 but not CD86 in a human in vitro model, while CD86 is decreased by tBHQ in a murine ex
vivo model. With the data accumulated in this thesis, it is not possible to reconcile these disparate
findings in a meaningful way. While the models and toxicants differ too much to draw
conclusions, it is important to note that ATO and tBHQ activate Nrf2 via different molecular
mechanisms, which may have functionally distinct outcomes. These findings, disparate as they
may be, are highly interesting though, as the role of CD80 and CD86 in stimulating T cells is
                                                  155


underexplored in the context of infectious disease and allergy. The results presented here
warrant more research into the role and impact of costimulation and how different signals
change the ratio in which those costimulatory molecules are expressed. For toxicologists, a
potential impact of environmental toxins on the CD80/86 expression could be a promising lead
to explain immunological alterations.
Overall, the main function of B cells is the production of immunoglobulins. Antibodies need to
have the right antigen affinity and the right isotype for optimal function. These studies have, with
the exception of chapter 4, not directly measured antibody production, antibody specificity and
antibody isotypes. Instead, I took proxy measurements, such as Ig subclasses expressed on the
surface of B cells as part of the B cell receptor or intracellular Ig subclasses. In Chapter 2, I
measured surface IgG as a response to IAV challenge. IgG is a later immunoglobulin class and only
expressed on cells that have undergone isotype switching. Measuring the percentage of cells with
IgG on their surface, therefore, served as a proxy for cells that have undergone isotype switching,
or their daughter cells as this population expands. The measured increase of IgG+ cells in
response to IAV is likely due to the expansion of a population of antigen-experienced B cells that
recognized IAV antigen.
Chapter 3 used a much earlier time point to measure cell status, and labeled IgG1 specifically,
one of the earliest immunoglobulins produced during isoclass switching in mice. In this assay, we
used intracellular labeling, and were able to detect IgG1 inside the cell, which is a more sensitive
assay than surface labeling. IgG1 was of particular interest because it can be a precursor to both
IgG and IgE. There was a tBHQ concentration-dependent decrease in the counts of IgG1-
                                                 156


producing cells, which may signify either inhibition or delay of immunoglobulin production. An
extension of this assay, with later collection timepoints and an added supernatant analysis for
immunoglobulin content may shed light on this question. In chapters 4 and 5, I used an allergic
model, and IgE was of particular interest. I used both intracellular staining to assess the
percentage of B cells and plasmablasts actively producing IgE and OVA-specific IgE ELISA on
plasma, which measures IgE that specifically binds to the model allergen. Taken together, the
experimental data suggests that in the context of influenza in humans, ATO suppresses the
percentage of IgG+ B cells, which likely would lead to lower overall levels of IgG specific to IAV.
In our generic T cell-dependent B cell activation model, tBHQ lowered the number of IgG1-
producing cells among mouse splenocytes. In contrast to those findings, dietary tBHQ in vivo
increased the percentage of IgE-producing B cells, plasma cells and plasma IgE in the allergy
model. Does this mean that tBHQ decreases IgG production and favors IgE production? It would
be premature at this point to draw this conclusion; however, future studies could be designed to
test this hypothesis.
There are several limitations to these studies. In chapter 2, we used human PBMCs, which are a
specific subtype of B cells and differ from B cells found in the spleen and lymph nodes, where the
germinal center reaction is located and where isotype switching and clonal expansion occur. This
study also only used H1N1 influenza, a single antigenic target, to which only a subset of PBMCs
responds. This limitation was addressed in Chapter 3, where we used a polyclonal activation
cocktail that is not limited to a specific antigen but activates B cells in a way that closely mimics
T cell-dependent activation without the need for an antigen. As we were able to use cells from
                                                 157


both wild-type mice and mice genetically ablated for Nrf2, we could demonstrate that these
effects are at least partially Nrf2-dependent. In Chapter 4, we evaluated the impact of dietary
tBHQ on the development of food allergy in young mice by inducing atopy using a transdermal
sensitization model. This study did not use Nrf2-deficient mice and thus, the role of Nrf2 in the
effects of tBHQ could not be evaluated. Chapter 5 evaluated the impact of age in the same allergic
sensitization model.
Relative to other immune cell types, comparably less work has been done on the toxicology of B
cells, and most of this work has focused on AHR signaling [149-151, 153, 155, 157]. The studies
described in this dissertation add to this understudied research area and add a class of ubiquitous
toxicants. The generation of diverse antibodies is of utmost importance as deficiency leads to
severe clinical outcomes, such as hyper-IgM syndrome, which is an immunodeficiency defined by
a lack of CD40-CD40L signaling. This is a key process to the B cell activation model used in
chapters 3 and 4 [58].
In general, alterations in these signaling and activation pathways have the potential to greatly
impact humoral immunity. It is important to note that for the average, immunocompetent
western subject, humoral immunity operates on a high level, and standard vaccinations routinely
elicit high titers of high-affinity antibodies, sometimes lasting for a lifetime, while xenobiotics like
tBHQ are ubiquitous in everyday life. In that context, it seems unlikely that tBHQ has a major
impact on B cell function related to humoral immunity. More likely is a small, insidious effect that
subtly changes outcomes in immunoglobulin production, which is what I hypothesize based on
the data shown here. However, small changes may be more dramatic and consequential if a
                                                   158


person is, due to genetic or pharmaceutical reasons, immunosuppressed or deficient.
Furthermore, small, insidious processes are hard to measure in human populations due to the
inherent variability between subjects, both genetically and behaviorally. B cell activation happens
in secondary lymphoid organs, such as the spleen and the lymph node, and access to these tissues
is almost impossible in human subjects. If they are excised, it is usually in the context of
malignancy or other severe pathological processes, which change the dynamic in these organs,
and pathology usually requires these organs for grading of disease processes. Biopsies like fine
needle aspirations are theoretically possible but must be done under local or general anesthesia
and are major procedures and thus unlikely to be available to basic scientists. PBMCs, even with
the limitations stated above, are our best tool for immunotoxicity testing in human subjects. Our
rodent models come with their limitations too, as rodent immune systems are only an
approximation of human immune systems. The applicability of observations in mouse models to
humans has been hotly debated in the scientific community [399]. Nonetheless, the results
reported in chapters 2,3, and 4 are reproducible and indicate that these compounds alter B cell
activation and effector function.
Allergy and atopic disease can be described, in a very simplified way, as overactivation of the
immune system resulting in an inappropriate response to innocuous antigens. The discussion in
the paragraphs above focused mainly on extrapolating the effects presented in this thesis in the
context of humoral immunity. Looking at the results, specifically in chapter 4, through the lens of
allergy, paints a different picture: the aberrant activation of the immune system is further
exacerbated, and the immunoglobulin IgE is increased when tBHQ is present. Studies in our lab
                                                159


that are yet to be published have shown that more advanced endpoints, such as anaphylactic
shock, are also more severe in tBHQ-fed mice compared to non-tBHQ-fed mice. As mentioned in
the introduction, the incidences of food allergies and anaphylactic shock are rising in the western
world [378, 393, 400]. The data presented herein is not sufficient to establish a causal
relationship between the increased prevalence of food allergy in humans and tBHQ exposure but
is congruent with the hypothesis that dietary tBHQ promotes IgE sensitization in mice. B cells of
sensitized mice that were fed tBHQ show higher expression of CD40, a higher percentage of IgE-
expressing CD138 cells, and an increase in plasma IgE concentrations. The data suggest the need
for further investigation into the impact of tBHQ on allergy in humans.
                                                160


Significance of findings
Further characterization of these impairments is necessary, and these and further studies could
be used to inform regulatory decision-making. For the public, these results may help personal
decision-making regarding the consumption of highly processed foods that contain tBHQ. As
mentioned above, it is important to strike a healthy balance in interpreting these results. On the
one hand, the results are real, significant, and reproducible, on the other hand, we can estimate
that most humans in the western world are exposed to Nrf2 activating compounds such as tBHQ
daily, and the vast majority of humans in the western world have normal, effective humoral
immunity. It is reasonable to recommend the avoidance of processed foods if possible and
regular testing of well water for heavy metals, but we must be careful not to slip into alarmism
and over-exaggeration of results. On the other hand, the number of people perishing from
infectious diseases such as influenza, SARS-COV-2, and others, is still staggering. Infectious
disease was the third leading cause of death in 2020, with over 400.000 deaths [401]. If there
was a way to mitigate human exposure to ATO and tBHQ, and if this hypothetical change would
cause a small but noticeable change in B cell function, would there be lives saved by this
intervention? The answer to this is, of course, impossible to formulate without straying into
speculation and sensationalism, but the data presented herein raises this very insurmountable
question.
Chapter 5 answers a question we asked ourselves when we designed these studies, namely if age
was a factor in this allergy model, and publication of our results may help other groups working
on similar models fine-tune their methodology. We show that younger mice are more susceptible
                                                 161


to immunization, which correlates with clinical observations in humans. In Chapter 4, we show
that the number of CD40 expressing cells and the number of IgE-producing plasma cells increases
when tBHQ is present during the development of allergy, and that the concentration of plasma
IgE is increased. These observations are limited to the mouse model but could be used as a basis
to inform current food allergy patients and parents that try to avoid the development of food
allergies in their infants through their dietary choices. Avoiding foods that contain tBHQ is likely
neither curative nor ultimately preventative in the development of human allergy but could
potentially mitigate the severity of food allergies and empower parents and patients. A possible
caveat to this is socioeconomic – often foods that are rich in tBHQ are cheaper and more easily
accessible, making the above recommendation selective in nature, as some families simply would
not be able to afford a diet that stringently avoids tBHQ-containing food products. Adding a
potential stigma in telling patients' families that they may be to blame for the development of
their loved one’s condition because they were not able to financially provide a diet would be an
unfortunate outcome of this recommendation, which itself is based on evidence from a mouse
model.
                                                 162


Future directions
In chapter 2, we described a novel human PBMC-based assay for immunotoxicity testing using
influenza A as a challenge. This model is easily adapted to other toxicants and other challenges.
We have started gathering preliminary data on other related compounds like sodium arsenite,
which could be another way to model human exposure to As(III), which is a common exposure
that is particularly of concern with rice-heavy diets [402]. It would be easy to use the same model
to evaluate the effects of other heavy metals with relevant environmental or occupational
exposure, such as cadmium, mercury, or lead compounds. In 2022, Washtenaw county in
Michigan had a spill in hexavalent chromium (Cr (VI)), which is another metal that could be
evaluated using this assay. Cr (VI) is also a known Nrf2 activator [403, 404], similarly to cadmium
[214, 364, 405], lead [215, 406], and mercury [407]. To determine whether the ATO-dependent
changes we have documented in Chapter 2 or any potential effects in follow-up experiments are
Nrf2-dependent, we may explore the use of CRISPR/CAS9 to ablate Nrf2 in PBMC-derived human
B cells [408]. Our lab has previous, yet unpublished experience in ablating Nrf2 in human PBMC-
derived CD4 T cells, and protocols could be adapted to facilitate B cell-specific Nrf2 ablation. This
experiment, with isolated and Nrf2-deficient B cells, would give us greater resolution and more
precise data about the involvement of Nrf2.
precise data about the involvement of Nrf2.
The studies in chapter 2 exclusively used H1/N1 A/PR/8/34. Influenza is a very important model
but considering the ongoing SARS-COV-2 pandemic, we are actively exploring the expansion of
our model to include SARS-CoV-2 and other coronaviruses. Given that there is a significant
                                                 163


portion of the population that has received immunizations against SARS-COV-2, and a high
number of subjects that recovered from active disease, I would love to pursue a bigger, IRB-
requiring study that enrolls subjects with information regarding their vaccination status, previous
disease history, age, and sex. Blood draws from this populace would be analyzed for the presence
of anti-SARS-COV-2 antibodies and recombinant SARS-COV-2 spike protein, which has been used
by others in in-vitro assays [409], as an infectious challenge. This modified assay could be used to
assess the effect of different toxicants (such as the ones described above) on B cell function
following a challenge by SARS-COV-2 protein, and we could differentiate between naïve and
antigen-experienced B cells based on vaccination/previous disease status of the subject. This
study would be highly relevant to the current global pandemic and has the potential to be
immensely impactful.
In a similar vein, Chapter 3 is limited in its scope by using mouse splenocytes ex vivo for the
studies. The scope could be expanded by using mouse cells from other compartments, such as
lymph nodes or peripheral blood, thus changing the ratio of B cell subtypes in the preparation.
This could also be achieved by using flow cytometry-assisted cell sorting or magnetic isolation of
a particular B cell subtype, e.g. FO B cells, which in theory should respond more strongly to our
activation cocktail. Similar to the studies in chapter 2, this study is easily scalable and could be
performed with most toxicants. One possibility is even to use Nur77-GFP reporter mice, which
exhibit a bright GFP signal when turning into a GC-reaction-like phenotype [410]. Hypothesizing
that our activation cocktail activates this Nur77-GFP signaling, this model could even be used in
high-throughput toxicology screening, where the GFP signal is the primary readout and a
                                                  164


reduction in GFP signaling would signify a reduction in GC-reaction-like phenotype in B cells. This
would be an excellent starting point to screen toxicants on a large scale for inhibition of T cell-
dependent B cell activation. Having a simple fluorescent read-out would enable that study to be
used in a high throughput screen.
Chapter 3 used splenocytes from mice and a mouse-specific antibody cocktail. A very similar
cocktail would be feasible to activate human PBMC-derived B cells. After verification of the
activation by an anti-human IgM antibody fragment, anti-human CD40 or recombinant human
CD40L, and recombinant IL-4, this could be a useful tool for further screening of the impact of
toxicants on T cell-dependent B cell activation, though it would converge somewhat with the
future directions outlined for chapter 2.
Regarding the in vivo allergy studies we describe in Chapters 4 & 5, the use of large-scale unbiased
approaches, such as scRNAseq, could elucidate pathways that are differentially activated during
sensitization when tBHQ is present, and thus direct further studies. It would be of particular
interest to see whether there is differential activity of BLIMP-1 controlled genes, which control
both B cell and plasma cell fate [63, 68]. The role of tBHQ on BLIMP-1 activity overall would be
interesting to investigate as there is a mutually-antagonistic relationship between BLIMP-1 and
Nrf2. Furthermore, BLIMP-1 is involved in many different immune processes, many of which are
relevant to the present studies.
In all our previous allergy studies and the ones presented in this thesis, a tetramer detecting T or
B cells specific for OVA would have been a great addition. MHC II tetramers that can detect OVA-
specific T cells are readily available but were considered unreliable, though new, improved
                                                165


tetramers have recently become available. Specifically for B cells, new protocols emerged in the
last few years that effectively use tetramers to study antigen-specific B cells [411]. Measuring T
and B cells that are specific to OVA would be an excellent addition to the studies in chapters 4
and 5.
Taking a step back from toxicology, a knowledge gap that my literature review revealed was the
role of Nrf2 in B cells. B cells, once activated, have high metabolic turnover and are subject to
significant oxidative stress [412]. Further investigation of the role of Nrf2 in B cells is an
unexplored field with the potential to reveal important therapeutic targets. This is underscored
by the findings in chapter 4, where we saw that even without sensitization, dietary tBHQ
increased the percentage of CD40-expressing B cells and the number of plasmablasts in mice.
Nrf2-deficient mice develop a lupus-like autoimmune disease with aberrant IgM, IgG, and C3
complexes leading to a nephritic syndrome, which hints that there is some B cell involvement
[413]. Conversely, the data presented in chapter 3 suggest that many (though not all) of the
effects observed in these studies may be Nrf2-independent. Nonetheless, studies confirming the
effect (or absence of an effect!) of Nrf2 activation on B cell function would add to our
understanding of B cell biology.
One question that arose in the discussion of limitations of chapters 4 and 5 was the location of
IgE-secreting cells. While we were able to capture some in the lymph node at the timepoint
chosen, plasma cells and memory B cells often do not dwell in secondary lymphoid organs [414].
While the consensus is that plasma cells relocate to the bone marrow and use its relatively high
circulation as a base for antibody secretion, infiltrating B cells and plasma cells are often observed
                                                  166


at sites of infection, inflammation, or malignancy [127, 415-417]. The use of mice expressing IgE
with a fluorescent VeriGem reporter construct [418] on an in-vivo imaging system like the Perkin-
Elmer IVIS Spectrum, located in Michigan State’s advanced molecular imaging facility, would
allow us to measure multiple independent relevant data. Primarily, we would be able to track
IgE+ cells and their movement after leaving the lymph node, potentially even after another
sensitization or an oral or inhalatory antigen challenge. Secondly, using frequent imaging would
us to better understand the kinetics of the B cell response and plasma cell differentiation
following allergic sensitization in our model.
This remarkable mouse data would best be followed up by something that is more clinically
applicable. Humanized mouse models for allergy research are possible, but complex. This is part
due to the nature of allergy itself, which is a complicated mechanism involving many cell types
and the recognition of an innocuous antigen [419]. XenoGraft vs Host disease is also a problem
in these models, though newer protocols mitigate these issues [420]. Given these limitations,
studies on humans are an attractive option. One idea is to use a large academic center with a
dedicated allergy unit, such as Mott’s Children’s Hospital (Mary H Weiser Center for Food
Allergies) in Ann Arbor or Detroit Children’s Hospital and enroll patients for a correlative tBHQ
study. While they visit the clinic for things like specific antigen testing, desensitization
treatments, or educational events, urine samples could be collected and analyzed for tBHQ. At
the same time, patients and loved ones may be asked to fill out a questionnaire regarding allergy
symptom severity. Additionally, blood work, such as total IgE titers or antigen specific IgE titers,
may be assessed. The analysis could compare plasma tBHQ levels, questionnaire results, and
                                                 167


potential lab values for correlation. Food allergy centers, such as Mary Weiser in Ann Arbor,
already provide uncountable educational events, and another study could assess whether
counseling and education towards a tBHQ-free diet have a measurable impact on urine tBHQ
levels and food allergy symptoms.
Access to blood samples from food allergy patients could also provide an avenue for allergen-
specific ex vivo assays. A prospective study may enroll patients with a specific allergy and collect
PBMCs from patients via phlebotomy, and culture those cells as described in chapter 2. Instead
of using influenza as a challenge model, we could expose the cells to the specific allergen (e.g.,
OVA for patients with egg allergies or crude peanut extract for patients with peanut allergy) to
which the patient is allergic, and thus specifically screen circulating B and T cells for their activity.
As described in Chapter 2, we could use different environmental toxicants to measure whether
they have an impact on this allergen-specific activation. This would take the model described in
chapter 2 directly from bench to bedside and provide a very clinically applicable measurement.
Ultimately, the studies presented here give us a starting point in assessing the impact of
environmental toxicants and Nrf2 activators on B cell function, highlighting both Nrf2-dependent
and -independent effects. The future directions suggested here would allow us on one hand to
further quantify other toxicants for their effect on B cell activation, but also expand our
understanding on whether xenobiotics present in our everyday life influence the activity and
function of B cells. Finally, the human studies could determine whether or not there is a
correlation between dietary tBHQ exposure and allergy severity.
                                                 168


                                        BIBLIOGRAPHY
1.  Smith, S.B., et al., Hieronymus [corrected] Fabricius Ab Acquapendente (1533-1619). Clin
    Anat, 2004. 17(7): p. 540-3.
2.  Ribatti, D., E. Crivellato, and A. Vacca, The contribution of Bruce Glick to the definition of
    the role played by the bursa of Fabricius in the development of the B cell lineage. Clinical
    and experimental immunology, 2006. 145(1): p. 1-4.
3.  Sternberg, S.S., Bottoms up to a Nobel-worthy chicken's bottom. Am J Surg Pathol, 2003.
    27(11): p. 1471-2.
4.  Bikard, D. and L.A. Marraffini, Innate and adaptive immunity in bacteria: mechanisms of
    programmed genetic variation to fight bacteriophages. Curr Opin Immunol, 2012. 24(1):
    p. 15-20.
5.  Bikard, D., et al., CRISPR interference can prevent natural transformation and virulence
    acquisition during in vivo bacterial infection. Cell Host Microbe, 2012. 12(2): p. 177-86.
6.  Boehm, T., N. Iwanami, and I. Hess, Evolution of the immune system in the lower
    vertebrates. Annu Rev Genomics Hum Genet, 2012. 13: p. 127-49.
7.  Matsunaga, T. and A. Rahman, What brought the adaptive immune system to
    vertebrates?--The jaw hypothesis and the seahorse. Immunol Rev, 1998. 166: p. 177-86.
8.  Boehm, T., et al., VLR-based adaptive immunity. Annu Rev Immunol, 2012. 30: p. 203-20.
9.  Muller, A.M., et al., Development of hematopoietic stem cell activity in the mouse
    embryo. Immunity, 1994. 1(4): p. 291-301.
10. Pieper, K., B. Grimbacher, and H. Eibel, B-cell biology and development. J Allergy Clin
    Immunol, 2013. 131(4): p. 959-71.
11. Shinkai, Y., et al., RAG-2-deficient mice lack mature lymphocytes owing to inability to
    initiate V(D)J rearrangement. Cell, 1992. 68(5): p. 855-867.
12. van Zelm, M.C., et al., Replication history of B lymphocytes reveals homeostatic
    proliferation and extensive antigen-induced B cell expansion. J Exp Med, 2007. 204(3): p.
    645-55.
13. Cinamon, G., et al., Sphingosine 1-phosphate receptor 1 promotes B cell localization in
    the splenic marginal zone. Nature Immunology, 2004. 5(7): p. 713-720.
                                               169


14. Ichikawa, D., et al., Natural anti-intestinal goblet cell autoantibody production from
    marginal zone B cells. Journal of immunology (Baltimore, Md. : 1950), 2015. 194(2): p.
    606-614.
15. Appelgren, D., et al., Marginal-Zone B-Cells Are Main Producers of IgM in Humans, and
    Are Reduced in Patients With Autoimmune Vasculitis. Frontiers in immunology, 2018. 9:
    p. 2242-2242.
16. Jaworski, M.A., C. Shiozawa, and E. Diener, Triggering of affinity-enriched B cells.
    Analysis of B cell stimulation by antigen-specific helper factor or lipopolysaccharide. I.
    Dissection into proliferative and differentiative signals. J Exp Med, 1982. 155(1): p. 248-
    63.
17. Coutinho, A., et al., Mechanism of thymus-independent immunocyte triggering.
    Mitogenic activation of B cells results in specific immune responses. The Journal of
    experimental medicine, 1974. 139(1): p. 74-92.
18. Pasare, C. and R. Medzhitov, Control of B-cell responses by Toll-like receptors. Nature,
    2005. 438(7066): p. 364-368.
19. Hornung, V., et al., Quantitative expression of toll-like receptor 1-10 mRNA in cellular
    subsets of human peripheral blood mononuclear cells and sensitivity to CpG
    oligodeoxynucleotides. J Immunol, 2002. 168(9): p. 4531-7.
20. Botos, I., D.M. Segal, and D.R. Davies, The structural biology of Toll-like receptors.
    Structure, 2011. 19(4): p. 447-59.
21. Dumont, N., et al., Increased secretion of hyperimmune antibodies following
    lipopolysaccharide stimulation of CD40-activated human B cells in vitro. Immunology,
    2009. 126(4): p. 588-595.
22. Reap, E.A., et al., Conventional B cells, not B-1 cells, are responsible for producing
    autoantibodies in lpr mice. J Exp Med, 1993. 177(1): p. 69-78.
23. Bendelac, A., M. Bonneville, and J.F. Kearney, Autoreactivity by design: innate B and T
    lymphocytes. Nature Reviews Immunology, 2001. 1(3): p. 177-186.
24. Colombo, M.J., G. Sun, and K.R. Alugupalli, T-Cell-Independent Immune Responses Do
    Not Require Cxc Ligand 13-Mediated B1 Cell Migration. Infection and Immunity, 2010.
    78(9): p. 3950-3956.
25. Kantor, A.B., et al., An unbiased analysis of V (H)-DJ (H) sequences from B-1a, B-1b, and
    conventional B cells. The Journal of Immunology, 1997. 158(3): p. 1175-1186.
                                              170


26. Cunningham, A.F., et al., B1b Cells Recognize Protective Antigens after Natural Infection
    and Vaccination. Frontiers in Immunology, 2014. 5.
27. Horsnell, W.G., et al., IL-4Rα-associated antigen processing by B cells promotes immunity
    in Nippostrongylus brasiliensis infection. PLoS pathogens, 2013. 9(10): p. e1003662.
28. Alugupalli, K.R., et al., The Resolution of Relapsing Fever Borreliosis Requires IgM and Is
    Concurrent with Expansion of B1b Lymphocytes. The Journal of Immunology, 2003.
    170(7): p. 3819-3827.
29. Alugupalli, K.R., A Distinct Role for B1b Lymphocytes in T Cell-Independent Immunity, in
    Specialization and Complementation of Humoral Immune Responses to Infection, T.
    Manser, Editor. 2008, Springer Berlin Heidelberg: Berlin, Heidelberg. p. 105-130.
30. Foote, J.B. and J.F. Kearney, Generation of B Cell Memory to the Bacterial Polysaccharide
    α-1,3 Dextran. The Journal of Immunology, 2009. 183(10): p. 6359-6368.
31. Baumgarth, N., Innate-like B cells and their rules of engagement. Crossroads Between
    Innate and Adaptive Immunity IV, 2013: p. 57-66.
32. Baumgarth, N., A hard (y) look at B-1 cell development and function. The Journal of
    Immunology, 2017. 199(10): p. 3387-3394.
33. Romero‐Ramírez, S., et al., Innate‐like B cell subsets during immune responses:
    beyond antibody production. Journal of leukocyte biology, 2019. 105(5): p. 843-856.
34. Pillai, S. and A. Cariappa, The follicular versus marginal zone B lymphocyte cell fate
    decision. Nature Reviews Immunology, 2009. 9(11): p. 767-777.
35. Tumanov, A.V., et al., Distinct role of surface lymphotoxin expressed by B cells in the
    organization of secondary lymphoid tissues. Immunity, 2002. 17(3): p. 239-250.
36. Fu, Y.-X., et al., B lymphocytes induce the formation of follicular dendritic cell clusters in
    a lymphotoxin α–dependent fashion. The Journal of experimental medicine, 1998.
    187(7): p. 1009-1018.
37. Asano, S., et al., Immunohistologic detection of the primary follicle (PF) in human fetal
    and newborn lymph node anlages. Pathol Res Pract, 1993. 189(8): p. 921-7.
38. Allen, C.D. and J.G. Cyster, Follicular dendritic cell networks of primary follicles and
    germinal centers: phenotype and function. Semin Immunol, 2008. 20(1): p. 14-25.
39. Batista, F.D. and N.E. Harwood, The who, how and where of antigen presentation to B
    cells. Nature Reviews Immunology, 2009. 9(1): p. 15-27.
                                               171


40. Phan, T.G., et al., Subcapsular encounter and complement-dependent transport of
    immune complexes by lymph node B cells. Nature immunology, 2007. 8(9): p. 992-1000.
41. Carrasco, Y.R. and F.D. Batista, 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, 2007. 27(1): p. 160-171.
42. Okada, T., et al., Antigen-Engaged B Cells Undergo Chemotaxis toward the T Zone and
    Form Motile Conjugates with Helper T Cells. PLOS Biology, 2005. 3(6): p. e150.
43. Allen Christopher, D.C., et al., Imaging of Germinal Center Selection Events During
    Affinity Maturation. Science, 2007. 315(5811): p. 528-531.
44. Miller, M., et al., Miller, M.J. , Hejazi, A.S. , Wei, S.H. , Cahalan, M.D. & Parker, I. T cell
    repertoire scanning is promoted by dynamic dendritic cell behavior and random T cell
    motility in the lymph node. Proc. Natl. Acad. Sci. USA 101, 998-1003. Proceedings of the
    National Academy of Sciences of the United States of America, 2004. 101: p. 998-1003.
45. Garside, P., et al., Visualization of specific B and T lymphocyte interactions in the lymph
    node. Science, 1998. 281(5373): p. 96-9.
46. Victora, G.D. and M.C. Nussenzweig, Germinal centers. Annu Rev Immunol, 2012. 30: p.
    429-57.
47. Elgueta, R., et al., Molecular mechanism and function of CD40/CD40L engagement in the
    immune system. Immunological reviews, 2009. 229(1): p. 152-172.
48. Bishop, G.A., et al., TRAF proteins in CD40 signaling. Adv Exp Med Biol, 2007. 597: p.
    131-51.
49. Chan, T.D., et al., Antigen affinity controls rapid T-dependent antibody production by
    driving the expansion rather than the differentiation or extrafollicular migration of early
    plasmablasts. The Journal of Immunology, 2009. 183(5): p. 3139-3149.
50. MacLennan, I.C.M., et al., Extrafollicular antibody responses. Immunological Reviews,
    2003. 194(1): p. 8-18.
51. Coffey, F., B. Alabyev, and T. Manser, Initial Clonal Expansion of Germinal Center B Cells
    Takes Place at the Perimeter of Follicles. Immunity, 2009. 30(4): p. 599-609.
52. Takemori, T., et al., Generation of memory B cells inside and outside germinal centers.
    European Journal of Immunology, 2014. 44(5): p. 1258-1264.
                                                172


53. Muramatsu, M., et al., Class switch recombination and hypermutation require
    activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell, 2000.
    102(5): p. 553-563.
54. Pavri, R. and M.C. Nussenzweig, AID targeting in antibody diversity. Adv Immunol, 2011.
    110: p. 1-26.
55. Peled, J.U., et al., The biochemistry of somatic hypermutation. Annu Rev Immunol, 2008.
    26: p. 481-511.
56. Roco, J.A., et al., Class-Switch Recombination Occurs Infrequently in Germinal Centers.
    Immunity, 2019. 51(2): p. 337-350.e7.
57. Cozine, C.L., K.L. Wolniak, and T.J. Waldschmidt, The primary germinal center response
    in mice. Curr Opin Immunol, 2005. 17(3): p. 298-302.
58. Jameson, J.L., et al., Harrison's principles of internal medicine. Twentieth edition / ǂb
    editors, J. Larry Jameson, Anthony S. Fauci, Dennis 1. Kasper, Stephen L. Hauser, Dan L.
    Longo, Joseph Loscalzo. ed. Principles of internal medicine. 2018, New York: McGraw-
    Hill Education.
59. Kenneth Murphy, C.W., Janeway’s Immunobiology, Ninth Edition. . Vol. 9. 2016, New
    York: Garland Science: New York. 928.
60. Kräutler, N.J., et al., Differentiation of germinal center B cells into plasma cells is initiated
    by high-affinity antigen and completed by Tfh cells. The Journal of experimental
    medicine, 2017. 214(5): p. 1259-1267.
61. Reed, J.H., et al., Clonal redemption of autoantibodies by somatic hypermutation away
    from self-reactivity during human immunization. The Journal of experimental medicine,
    2016. 213(7): p. 1255-1265.
62. Sabouri, Z., et al., Redemption of autoantibodies on anergic B cells by variable-region
    glycosylation and mutation away from self-reactivity. Proceedings of the National
    Academy of Sciences, 2014. 111(25): p. E2567-E2575.
63. Radtke, D. and O. Bannard, Expression of the Plasma Cell Transcriptional Regulator
    Blimp-1 by Dark Zone Germinal Center B Cells During Periods of Proliferation. Frontiers in
    immunology, 2019. 9: p. 3106-3106.
64. Seifert, M. and R. Küppers, Human memory B cells. Leukemia, 2016. 30(12): p. 2283-
    2292.
65. Kurosaki, T., K. Kometani, and W. Ise, Memory B cells. Nature Reviews Immunology,
    2015. 15(3): p. 149-159.
                                                173


66. Shinnakasu, R. and T. Kurosaki, Regulation of memory B and plasma cell differentiation.
    Current Opinion in Immunology, 2017. 45: p. 126-131.
67. Nutt, S.L., et al., The generation of antibody-secreting plasma cells. Nature Reviews
    Immunology, 2015. 15(3): p. 160-171.
68. Minnich, M., et al., Multifunctional role of the transcription factor Blimp-1 in
    coordinating plasma cell differentiation. Nat Immunol, 2016. 17(3): p. 331-43.
69. Akiba, H., et al., The Role of ICOS in the CXCR5<sup>+</sup> Follicular B Helper T Cell
    Maintenance In Vivo. The Journal of Immunology, 2005. 175(4): p. 2340-2348.
70. Eto, D., et al., IL-21 and IL-6 are critical for different aspects of B cell immunity and
    redundantly induce optimal follicular helper CD4 T cell (Tfh) differentiation. PLoS One,
    2011. 6(3): p. e17739.
71. Tahiliani, V., et al., OX40 Cooperates with ICOS To Amplify Follicular Th Cell Development
    and Germinal Center Reactions during Infection. J Immunol, 2017. 198(1): p. 218-228.
72. Kurata, I., I. Matsumoto, and T. Sumida, T follicular helper cell subsets: a potential key
    player in autoimmunity. Immunological Medicine, 2021. 44(1): p. 1-9.
73. Good-Jacobson, K.L., et al., PD-1 regulates germinal center B cell survival and the
    formation and affinity of long-lived plasma cells. Nature Immunology, 2010. 11(6): p.
    535-542.
74. Higuchi, T., et al., Cutting Edge: Ectopic Expression of CD40 Ligand on B Cells Induces
    Lupus-Like Autoimmune Disease. The Journal of Immunology, 2002. 168(1): p. 9-12.
75. van Nierop, K. and C. de Groot, Human follicular dendritic cells: function, origin and
    development. Seminars in Immunology, 2002. 14(4): p. 251-257.
76. Banchereau, J., et al., Immunobiology of dendritic cells. Annu Rev Immunol, 2000. 18: p.
    767-811.
77. Denzer, K., et al., Follicular dendritic cells carry MHC class II-expressing microvesicles at
    their surface. J Immunol, 2000. 165(3): p. 1259-65.
78. Aguzzi, A. and N.J. Krautler, Characterizing follicular dendritic cells: A progress report.
    European Journal of Immunology, 2010. 40(8): p. 2134-2138.
79. Reth, M., Matching cellular dimensions with molecular sizes. Nat Immunol, 2013. 14(8):
    p. 765-7.
80. Woof, J.M. and D.R. Burton, Human antibody–Fc receptor interactions illuminated by
    crystal structures. Nature Reviews Immunology, 2004. 4(2): p. 89-99.
                                                174


81. Rogentine, G.N., Jr., et al., Metabolism of human immunoglobulin D (IgD). The Journal of
    clinical investigation, 1966. 45(9): p. 1467-1478.
82. Nitschke, L., et al., Immunoglobulin D-deficient mice can mount normal immune
    responses to thymus-independent and -dependent antigens. Proceedings of the National
    Academy of Sciences, 1993. 90(5): p. 1887-1891.
83. Chen, K., et al., Immunoglobulin D enhances immune surveillance by activating
    antimicrobial, proinflammatory and B cell-stimulating programs in basophils. Nature
    immunology, 2009. 10(8): p. 889-898.
84. Johansen, Braathen, and Brandtzaeg, Role of J Chain in Secretory Immunoglobulin
    Formation. Scandinavian Journal of Immunology, 2000. 52(3): p. 240-248.
85. MacLennan, I.C., et al., Extrafollicular antibody responses. Immunol Rev, 2003. 194: p. 8-
    18.
86. Halpern, M.S. and M.E. Koshland, Novel Subunit in Secretory IgA. Nature, 1970.
    228(5278): p. 1276-1278.
87. Cerutti, A., The regulation of IgA class switching. Nature reviews. Immunology, 2008.
    8(6): p. 421-434.
88. Macpherson, A.J., et al., A primitive T cell-independent mechanism of intestinal mucosal
    IgA responses to commensal bacteria. Science, 2000. 288(5474): p. 2222-6.
89. Maverakis, E., et al., Glycans in the immune system and The Altered Glycan Theory of
    Autoimmunity: A critical review. Journal of Autoimmunity, 2015. 57: p. 1-13.
90. Kaetzel, C.S., et al., The polymeric immunoglobulin receptor (secretory component)
    mediates transport of immune complexes across epithelial cells: a local defense function
    for IgA. Proc Natl Acad Sci U S A, 1991. 88(19): p. 8796-800.
91. de Sousa-Pereira, P. and J.M. Woof, IgA: Structure, Function, and Developability.
    Antibodies (Basel), 2019. 8(4).
92. Vidarsson, G., G. Dekkers, and T. Rispens, IgG subclasses and allotypes: from structure to
    effector functions. Frontiers in immunology, 2014. 5: p. 520-520.
93. Xu, Z., et al., Immunoglobulin class-switch DNA recombination: induction, targeting and
    beyond. Nature Reviews Immunology, 2012. 12(7): p. 517-531.
94. Mandler, R., et al., IL-4 induction of IgE class switching by lipopolysaccharide-activated
    murine B cells occurs predominantly through sequential switching. The Journal of
    Immunology, 1993. 150(2): p. 407-418.
                                              175


95.  He, J.S., et al., Biology of IgE production: IgE cell differentiation and the memory of IgE
     responses. Curr Top Microbiol Immunol, 2015. 388: p. 1-19.
96.  Kuijpers, T.W., R.S. Weening, and T.A. Out, IgG subclass deficiencies and recurrent
     pyogenic infections, unresponsiveness against bacterial polysaccharide antigens. Allergol
     Immunopathol (Madr), 1992. 20(1): p. 28-34.
97.  Schauer, U., et al., Levels of antibodies specific to tetanus toxoid, Haemophilus
     influenzae type b, and pneumococcal capsular polysaccharide in healthy children and
     adults. Clin Diagn Lab Immunol, 2003. 10(2): p. 202-7.
98.  Ferrante, A., L.J. Beard, and R.G. Feldman, IgG subclass distribution of antibodies to
     bacterial and viral antigens. Pediatr Infect Dis J, 1990. 9(8 Suppl): p. S16-24.
99.  Stapleton, N.M., et al., Competition for FcRn-mediated transport gives rise to short half-
     life of human IgG3 and offers therapeutic potential. Nat Commun, 2011. 2: p. 599.
100. Mankarious, S., et al., The half-lives of IgG subclasses and specific antibodies in patients
     with primary immunodeficiency who are receiving intravenously administered
     immunoglobulin. J Lab Clin Med, 1988. 112(5): p. 634-40.
101. Trampert, D.C., et al., On the role of IgG4 in inflammatory conditions: lessons for IgG4-
     related disease. Biochimica et Biophysica Acta (BBA) - Molecular Basis of Disease, 2018.
     1864(4, Part B): p. 1401-1409.
102. van der Neut Kolfschoten, M., et al., Anti-Inflammatory Activity of Human IgG4
     Antibodies by Dynamic Fab Arm Exchange. Science, 2007. 317(5844): p. 1554-1557.
103. Corrado, A., et al., Extrafollicular IgD+ B cells generate IgE antibody secreting cells in the
     nasal mucosa. Mucosal Immunology, 2021. 14(5): p. 1144-1159.
104. Hamid, F., et al., Helminth-induced IgE and protection against allergic disorders. Curr
     Top Microbiol Immunol, 2015. 388: p. 91-108.
105. Xiong, H., et al., Sequential class switching is required for the generation of high affinity
     IgE antibodies. J Exp Med, 2012. 209(2): p. 353-64.
106. Fitzsimmons, C.M., F.H. Falcone, and D.W. Dunne, Helminth Allergens, Parasite-Specific
     IgE, and Its Protective Role in Human Immunity. Front Immunol, 2014. 5: p. 61.
107. Yazdanbakhsh, M., P.G. Kremsner, and R. van Ree, Allergy, parasites, and the hygiene
     hypothesis. Science, 2002. 296(5567): p. 490-4.
108. Gowthaman, U., et al., Identification of a T follicular helper cell subset that drives
     anaphylactic IgE. Science, 2019. 365(6456).
                                                176


109. He, J.S., et al., IgG1 memory B cells keep the memory of IgE responses. Nat Commun,
     2017. 8(1): p. 641.
110. Gowthaman, U., J.S. Chen, and S.C. Eisenbarth, Regulation of IgE by T follicular helper
     cells. Journal of Leukocyte Biology, 2020. 107(3): p. 409-418.
111. Mosmann, T.R. and R.L. Coffman, TH1 and TH2 cells: different patterns of lymphokine
     secretion lead to different functional properties. Annu Rev Immunol, 1989. 7: p. 145-73.
112. Kopf, M., et al., Disruption of the murine IL-4 gene blocks Th2 cytokine responses.
     Nature, 1993. 362(6417): p. 245-8.
113. Finkelman, F.D., et al., IL-4 is required to generate and sustain in vivo IgE responses. J
     Immunol, 1988. 141(7): p. 2335-41.
114. Stevens, T.L., et al., Regulation of antibody isotype secretion by subsets of antigen-
     specific helper T cells. Nature, 1988. 334(6179): p. 255-8.
115. Nishio, N., et al., Antibodies to wounded tissue enhance cutaneous wound healing.
     Immunology, 2009. 128(3): p. 369-380.
116. Yoshizaki, A., Pathogenic roles of B lymphocytes in systemic sclerosis. Immunol Lett,
     2018. 195: p. 76-82.
117. Karrar, S. and D.S. Cunninghame Graham, Abnormal B Cell Development in Systemic
     Lupus Erythematosus: What the Genetics Tell Us. Arthritis Rheumatol, 2018. 70(4): p.
     496-507.
118. Cencioni, M.T., et al., B cells in multiple sclerosis — from targeted depletion to immune
     reconstitution therapies. Nature Reviews Neurology, 2021. 17(7): p. 399-414.
119. Pawankar, R., et al., Overview on the pathomechanisms of allergic rhinitis. Asia Pac
     Allergy, 2011. 1(3): p. 157-67.
120. Silverman, G.J. and D.A. Carson, Roles of B cells in rheumatoid arthritis. Arthritis Res
     Ther, 2003. 5(4): p. S1.
121. Lejeune, T., C. Meyer, and V. Abadie, B Lymphocytes Contribute to Celiac Disease
     Pathogenesis. Gastroenterology, 2021. 160(7): p. 2608-2610.e4.
122. Smith, M.J., K.M. Simmons, and J.C. Cambier, B cells in type 1 diabetes mellitus and
     diabetic kidney disease. Nature Reviews Nephrology, 2017. 13(11): p. 712-720.
123. Zhai, X., et al., Elevated B Cell Activation is Associated with Type 2 Diabetes Development
     in Obese Subjects. Cellular Physiology and Biochemistry, 2016. 38(3): p. 1257-1266.
                                                177


124. Pierpont, T.M., C.B. Limper, and K.L. Richards, Past, Present, and Future of Rituximab—
     The World’s First Oncology Monoclonal Antibody Therapy. Frontiers in Oncology, 2018.
     8.
125. Grech, P. and M. Khamashta, Targeted therapies in systemic lupus erythematosus.
     Lupus, 2013. 22(10): p. 978-86.
126. Sharonov, G.V., et al., B cells, plasma cells and antibody repertoires in the tumour
     microenvironment. Nature Reviews Immunology, 2020. 20(5): p. 294-307.
127. Guo, F.F. and J.W. Cui, The Role of Tumor-Infiltrating B Cells in Tumor Immunity. Journal
     of Oncology, 2019. 2019: p. 2592419.
128. Largeot, A., et al., The B-side of Cancer Immunity: The Underrated Tune. Cells, 2019. 8(5).
129. Yuen, G.J., E. Demissie, and S. Pillai, B lymphocytes and cancer: a love-hate relationship.
     Trends in cancer, 2016. 2(12): p. 747-757.
130. Misra, R.R. and S.E. Bloom, Roles of dosage, pharmacokinetics, and cellular sensitivity to
     damage in the selective toxicity of cyclophosphamide towards B and T cells in
     development. Toxicology, 1991. 66(3): p. 239-56.
131. Jung, K.J., et al., Enhancement of B cell and monocyte populations in rats exposed to
     chlorpheniramine. Arch Pharm Res, 2012. 35(12): p. 2183-9.
132. Phalen, L.J., et al., Transcriptional response in rainbow trout (Oncorhynchus mykiss) B
     cells and thrombocytes following in vivo exposure to benzo[a]pyrene. Environ Toxicol
     Pharmacol, 2017. 53: p. 212-218.
133. Han, B., et al., Modulatory Effects of Pb(2+) on Virally Challenged Chicken Macrophage
     (HD-11) and B-Lymphocyte (DT40) Cell Lines In Vitro. Environ Toxicol Chem, 2020. 39(5):
     p. 1060-1070.
134. Han, B., et al., Modulatory Effects of Mercury (II) Chloride (HgCl(2) ) on Chicken
     Macrophage and B-Lymphocyte Cell Lines with Viral-Like Challenges In Vitro. Environ
     Toxicol Chem, 2021. 40(10): p. 2813-2824.
135. Caruso, J.A., et al., A systems toxicology approach identifies Lyn as a key signaling
     phosphoprotein modulated by mercury in a B lymphocyte cell model. Toxicol Appl
     Pharmacol, 2014. 276(1): p. 47-54.
136. De Vooght, V., et al., B-lymphocytes as key players in chemical-induced asthma. PLoS
     One, 2013. 8(12): p. e83228.
                                                178


137. Gatta, L., et al., Effects of in vivo exposure to GSM-modulated 900 MHz radiation on
     mouse peripheral lymphocytes. Radiat Res, 2003. 160(5): p. 600-5.
138. Uzarski, R.L. and J.J. Pestka, Comparative susceptibility of B cells with different lineages
     to cytotoxicity and apoptosis induction by translational inhibitors. J Toxicol Environ
     Health A, 2003. 66(22): p. 2105-18.
139. Romer, E.J. and C.E. Sulentic, Hydrogen peroxide modulates immunoglobulin expression
     by targeting the 3'Igh regulatory region through an NFκB-dependent mechanism. Free
     Radic Res, 2011. 45(7): p. 796-809.
140. Jacobsen-Pereira, C.H., et al., Immune response of Brazilian farmers exposed to multiple
     pesticides. Ecotoxicol Environ Saf, 2020. 202: p. 110912.
141. Haase, H., et al., Immunotoxicity Monitoring in a Population Exposed to Polychlorinated
     Biphenyls. Int J Environ Res Public Health, 2016. 13(3).
142. Schecter, A., et al., Dioxins: An overview. Environmental Research, 2006. 101(3): p. 419-
     428.
143. Nikolaidis, E., et al., TCDD inhibits the support of B-cell development by the bursa of
     Fabricius. Pharmacol Toxicol, 1990. 67(1): p. 22-6.
144. Chastain, J.E., Jr. and T.L. Pazdernik, 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD)-induced
     immunotoxicity. Int J Immunopharmacol, 1985. 7(6): p. 849-56.
145. Karras, J.G., et al., 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) elevates basal B-cell
     intracellular calcium concentration and suppresses surface Ig- but not CD40-induced
     antibody secretion. Toxicol Appl Pharmacol, 1996. 137(2): p. 275-84.
146. Lu, H., et al., Induction of the aryl hydrocarbon receptor-responsive genes and
     modulation of the immunoglobulin M response by 2,3,7,8-tetrachlorodibenzo-p-dioxin in
     primary human B cells. Toxicol Sci, 2010. 118(1): p. 86-97.
147. Lu, H., et al., 2,3,7,8-Tetrachlorodibenzo-p-dioxin-mediated disruption of the CD40
     ligand-induced activation of primary human B cells. Toxicol Appl Pharmacol, 2011.
     255(3): p. 251-60.
148. Kovalova, N., et al., Comparative analysis of TCDD-induced AhR-mediated gene
     expression in human, mouse and rat primary B cells. Toxicol Appl Pharmacol, 2017. 316:
     p. 95-106.
149. Zhou, J., et al., Suppression of the IgM Response by Aryl Hydrocarbon Receptor
     Activation in Human Primary B Cells Involves Impairment of Immunoglobulin Secretory
     Processes. Toxicol Sci, 2018. 163(1): p. 319-329.
                                                179


150. Dornbos, P., et al., The Influence of Human Interindividual Variability on the Low-Dose
     Region of Dose-Response Curve Induced by 2,3,7,8-Tetrachlorodibenzo-p-Dioxin in
     Primary B Cells. Toxicol Sci, 2016. 153(2): p. 352-60.
151. Schneider, D., et al., Involvement of Blimp-1 and AP-1 dysregulation in the 2,3,7,8-
     Tetrachlorodibenzo-p-dioxin-mediated suppression of the IgM response by B cells.
     Toxicol Sci, 2009. 108(2): p. 377-88.
152. Yoo, B.S., et al., 2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) alters the regulation of Pax5
     in lipopolysaccharide-activated B cells. Toxicol Sci, 2004. 77(2): p. 272-9.
153. Dornbos, P., et al., Characterizing Serpinb2 as a Modulator of TCDD-Induced Suppression
     of the B Cell. Chem Res Toxicol, 2018. 31(11): p. 1248-1259.
154. Li, J., et al., Aryl Hydrocarbon Receptor Activation Suppresses EBF1 and PAX5 and Impairs
     Human B Lymphopoiesis. J Immunol, 2017. 199(10): p. 3504-3515.
155. Li, J., et al., Aryl hydrocarbon receptor activation by 2,3,7,8-tetrachlorodibenzo-p-dioxin
     impairs human B lymphopoiesis. Toxicology, 2017. 378: p. 17-24.
156. Phadnis-Moghe, A.S., et al., SHP-1 is directly activated by the aryl hydrocarbon receptor
     and regulates BCL-6 in the presence of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD).
     Toxicol Appl Pharmacol, 2016. 310: p. 41-50.
157. Blevins, L.K., et al., TCDD-mediated suppression of naïve human B cell IgM secretion
     involves aryl hydrocarbon receptor-mediated reduction in STAT3 serine 727
     phosphorylation and is restored by interferon-γ. Cell Signal, 2020. 65: p. 109447.
158. Blevins, L.K., et al., Identification of a Sensitive Human Immunological Target of Aryl
     Hydrocarbon Receptor Activation: CD5(+) Innate-Like B Cells. Front Immunol, 2021. 12:
     p. 635748.
159. Zhou, J., et al., Role of Programmed Cell Death Protein-1 and Lymphocyte Specific
     Protein Tyrosine Kinase in the Aryl Hydrocarbon Receptor- Mediated Impairment of the
     IgM Response in Human CD5(+) Innate-Like B Cells. Front Immunol, 2022. 13: p. 884203.
160. Kummari, E., et al., TCDD attenuates EAE through induction of FasL on B cells and
     inhibition of IgG production. Toxicology, 2021. 448: p. 152646.
161. Ngaotepprutaram, T., et al., Suppression by Δ(9)-tetrahydrocannabinol of the primary
     immunoglobulin M response by human peripheral blood B cells is associated with
     impaired STAT3 activation. Toxicology, 2013. 310: p. 84-91.
162. Bursley, J.K. and C.E. Rockwell, Nrf2-dependent and -independent effects of tBHQ in
     activated murine B cells. Food Chem Toxicol, 2020. 145: p. 111595.
                                                 180


163. Moi, P., et al., Isolation of NF-E2-related factor 2 (Nrf2), a NF-E2-like basic leucine zipper
     transcriptional activator that binds to the tandem NF-E2/AP1 repeat of the beta-globin
     locus control region. Proc Natl Acad Sci U S A, 1994. 91(21): p. 9926-30.
164. Gacesa, R., et al., Rising levels of atmospheric oxygen and evolution of Nrf2. Scientific
     reports, 2016. 6: p. 27740-27740.
165. Chan, K., et al., NRF2, a member of the NFE2 family of transcription factors, is not
     essential for murine erythropoiesis, growth, and development. Proc Natl Acad Sci U S A,
     1996. 93(24): p. 13943-8.
166. Lacher, S.E., et al., Beyond antioxidant genes in the ancient Nrf2 regulatory network.
     Free Radic Biol Med, 2015. 88(Pt B): p. 452-465.
167. Iso, T., et al., Absolute Amounts and Status of the Nrf2-Keap1-Cul3 Complex within Cells.
     Mol Cell Biol, 2016. 36(24): p. 3100-3112.
168. Yamamoto, T., et al., Identification of polymorphisms in the promoter region of the
     human NRF2 gene. Biochem Biophys Res Commun, 2004. 321(1): p. 72-9.
169. McMahon, M., et al., Keap1-dependent proteasomal degradation of transcription factor
     Nrf2 contributes to the negative regulation of antioxidant response element-driven gene
     expression. J Biol Chem, 2003. 278(24): p. 21592-600.
170. Zhang, D.D., et al., Keap1 is a redox-regulated substrate adaptor protein for a Cul3-
     dependent ubiquitin ligase complex. Mol Cell Biol, 2004. 24(24): p. 10941-53.
171. McMahon, M., et al., Dimerization of substrate adaptors can facilitate cullin-mediated
     ubiquitylation of proteins by a "tethering" mechanism: a two-site interaction model for
     the Nrf2-Keap1 complex. J Biol Chem, 2006. 281(34): p. 24756-68.
172. Ishii, T., et al., Transcription factor Nrf2 coordinately regulates a group of oxidative
     stress-inducible genes in macrophages. J Biol Chem, 2000. 275(21): p. 16023-9.
173. Dinkova-Kostova, A.T., et al., Direct evidence that sulfhydryl groups of Keap1 are the
     sensors regulating induction of phase 2 enzymes that protect against carcinogens and
     oxidants. Proc Natl Acad Sci U S A, 2002. 99(18): p. 11908-13.
174. Horie, Y., et al., Molecular basis for the disruption of Keap1-Nrf2 interaction via Hinge &
     Latch mechanism. Commun Biol, 2021. 4(1): p. 576.
175. Tong, K.I., et al., Two-site substrate recognition model for the Keap1-Nrf2 system: a
     hinge and latch mechanism. Biol Chem, 2006. 387(10-11): p. 1311-20.
                                                 181


176. Tong, K.I., et al., Different electrostatic potentials define ETGE and DLG motifs as hinge
     and latch in oxidative stress response. Mol Cell Biol, 2007. 27(21): p. 7511-21.
177. Baird, L., et al., Regulatory flexibility in the Nrf2-mediated stress response is conferred by
     conformational cycling of the Keap1-Nrf2 protein complex. Proc Natl Acad Sci U S A,
     2013. 110(38): p. 15259-64.
178. McMahon, M., et al., Redox-regulated turnover of Nrf2 is determined by at least two
     separate protein domains, the redox-sensitive Neh2 degron and the redox-insensitive
     Neh6 degron. J Biol Chem, 2004. 279(30): p. 31556-67.
179. Rada, P., et al., SCF/{beta}-TrCP promotes glycogen synthase kinase 3-dependent
     degradation of the Nrf2 transcription factor in a Keap1-independent manner. Mol Cell
     Biol, 2011. 31(6): p. 1121-33.
180. Chowdhry, S., et al., Nrf2 is controlled by two distinct beta-TrCP recognition motifs in its
     Neh6 domain, one of which can be modulated by GSK-3 activity. Oncogene, 2013.
     32(32): p. 3765-81.
181. Venugopal, R. and A.K. Jaiswal, Nrf1 and Nrf2 positively and c-Fos and Fra1 negatively
     regulate the human antioxidant response element-mediated expression of
     NAD(P)H:quinone oxidoreductase1 gene. Proc Natl Acad Sci U S A, 1996. 93(25): p.
     14960-5.
182. Itoh, K., et al., Cloning and characterization of a novel erythroid cell-derived CNC family
     transcription factor heterodimerizing with the small Maf family proteins. Mol Cell Biol,
     1995. 15(8): p. 4184-93.
183. Kataoka, K., et al., Small Maf proteins heterodimerize with Fos and may act as
     competitive repressors of the NF-E2 transcription factor. Mol Cell Biol, 1995. 15(4): p.
     2180-90.
184. Venugopal, R. and A.K. Jaiswal, Nrf2 and Nrf1 in association with Jun proteins regulate
     antioxidant response element-mediated expression and coordinated induction of genes
     encoding detoxifying enzymes. Oncogene, 1998. 17(24): p. 3145-56.
185. Fan, W., et al., Keap1 facilitates p62-mediated ubiquitin aggregate clearance via
     autophagy. Autophagy, 2010. 6(5): p. 614-21.
186. Villeneuve, N.F., A. Lau, and D.D. Zhang, Regulation of the Nrf2-Keap1 antioxidant
     response by the ubiquitin proteasome system: an insight into cullin-ring ubiquitin ligases.
     Antioxid Redox Signal, 2010. 13(11): p. 1699-712.
187. Lau, A., et al., A noncanonical mechanism of Nrf2 activation by autophagy deficiency:
     direct interaction between Keap1 and p62. Mol Cell Biol, 2010. 30(13): p. 3275-85.
                                                  182


188. Jain, A., et al., p62/SQSTM1 is a target gene for transcription factor NRF2 and creates a
     positive feedback loop by inducing antioxidant response element-driven gene
     transcription. J Biol Chem, 2010. 285(29): p. 22576-91.
189. Apopa, P.L., X. He, and Q. Ma, Phosphorylation of Nrf2 in the transcription activation
     domain by casein kinase 2 (CK2) is critical for the nuclear translocation and transcription
     activation function of Nrf2 in IMR-32 neuroblastoma cells. J Biochem Mol Toxicol, 2008.
     22(1): p. 63-76.
190. Pi, J., et al., Molecular mechanism of human Nrf2 activation and degradation: role of
     sequential phosphorylation by protein kinase CK2. Free Radic Biol Med, 2007. 42(12): p.
     1797-806.
191. Huang, H.C., T. Nguyen, and C.B. Pickett, Phosphorylation of Nrf2 at Ser-40 by protein
     kinase C regulates antioxidant response element-mediated transcription. J Biol Chem,
     2002. 277(45): p. 42769-74.
192. Huang, H.C., T. Nguyen, and C.B. Pickett, Regulation of the antioxidant response element
     by protein kinase C-mediated phosphorylation of NF-E2-related factor 2. Proc Natl Acad
     Sci U S A, 2000. 97(23): p. 12475-80.
193. Cullinan, S.B., et al., Nrf2 is a direct PERK substrate and effector of PERK-dependent cell
     survival. Mol Cell Biol, 2003. 23(20): p. 7198-209.
194. Zimmermann, K., et al., Activated AMPK boosts the Nrf2/HO-1 signaling axis--A role for
     the unfolded protein response. Free Radic Biol Med, 2015. 88(Pt B): p. 417-426.
195. Kawai, Y., et al., Acetylation-deacetylation of the transcription factor Nrf2 (nuclear factor
     erythroid 2-related factor 2) regulates its transcriptional activity and nucleocytoplasmic
     localization. J Biol Chem, 2011. 286(9): p. 7629-40.
196. Katoh, Y., et al., Two domains of Nrf2 cooperatively bind CBP, a CREB binding protein,
     and synergistically activate transcription. Genes Cells, 2001. 6(10): p. 857-68.
197. Sun, Z., Y.E. Chin, and D.D. Zhang, Acetylation of Nrf2 by p300/CBP augments promoter-
     specific DNA binding of Nrf2 during the antioxidant response. Mol Cell Biol, 2009. 29(10):
     p. 2658-72.
198. Nioi, P., et al., The carboxy-terminal Neh3 domain of Nrf2 is required for transcriptional
     activation. Mol Cell Biol, 2005. 25(24): p. 10895-906.
199. Ghosh, A., et al., Insulin Inhibits Nrf2 Gene Expression via Heterogeneous Nuclear
     Ribonucleoprotein F/K in Diabetic Mice. Endocrinology, 2017. 158(4): p. 903-919.
                                                 183


200. Lo, J.Y., B.N. Spatola, and S.P. Curran, WDR23 regulates NRF2 independently of KEAP1.
     PLoS Genet, 2017. 13(4): p. e1006762.
201. Baba, K., H. Morimoto, and S. Imaoka, Seven in absentia homolog 2 (Siah2) protein is a
     regulator of NF-E2-related factor 2 (Nrf2). J Biol Chem, 2013. 288(25): p. 18393-405.
202. Zhang, Y., et al., Identification of an adaptor protein that facilitates Nrf2-Keap1 complex
     formation and modulates antioxidant response. Free Radic Biol Med, 2016. 97: p. 38-49.
203. Wu, T., et al., Hrd1 suppresses Nrf2-mediated cellular protection during liver cirrhosis.
     Genes Dev, 2014. 28(7): p. 708-22.
204. Malloy, M.T., et al., Trafficking of the transcription factor Nrf2 to promyelocytic
     leukemia-nuclear bodies: implications for degradation of NRF2 in the nucleus. J Biol
     Chem, 2013. 288(20): p. 14569-14583.
205. Komaravelli, N., et al., Respiratory syncytial virus induces NRF2 degradation through a
     promyelocytic leukemia protein - ring finger protein 4 dependent pathway. Free Radic
     Biol Med, 2017. 113: p. 494-504.
206. Freeborn, R.A. and C.E. Rockwell, The role of Nrf2 in autoimmunity and infectious
     disease: Therapeutic possibilities. Adv Pharmacol, 2021. 91: p. 61-110.
207. Wu, T., et al., Poly(ADP-ribose) polymerase-1 modulates Nrf2-dependent transcription.
     Free Radic Biol Med, 2014. 67: p. 69-80.
208. Kim, J.H., et al., The nuclear cofactor RAC3/AIB1/SRC-3 enhances Nrf2 signaling by
     interacting with transactivation domains. Oncogene, 2013. 32(4): p. 514-27.
209. Joe Rogan, R.P., The Joe Rogan Experience, in JRE Episode 901: Rhonda Patrick, J. Rogan,
     Editor. 2017, Spotify: Austin, Texas.
210. Huberman, A., Dr. Rhonda Patrick: Micronutrients for Health & Longevity, in Huberman
     Lab, R.P. Andrew Huberman, Editor. 2022: Palo Alto, California.
211. Yagishita, Y., et al., Broccoli or Sulforaphane: Is It the Source or Dose That Matters?
     Molecules, 2019. 24(19).
212. Rockwell, C.E., et al., Th2 skewing by activation of Nrf2 in CD4(+) T cells. J Immunol,
     2012. 188(4): p. 1630-7.
213. Yan, N., et al., Dimethyl fumarate improves cognitive deficits in chronic cerebral
     hypoperfusion rats by alleviating inflammation, oxidative stress, and ferroptosis via
     NRF2/ARE/NF-κB signal pathway. Int Immunopharmacol, 2021. 98: p. 107844.
                                                184


214. Buha, A., et al., The Role of Toxic Metals and Metalloids in Nrf2 Signaling. Antioxidants
     (Basel), 2021. 10(5).
215. Jiang, X., et al., Lead exposure activates the Nrf2/Keap1 pathway, aggravates oxidative
     stress, and induces reproductive damage in female mice. Ecotoxicol Environ Saf, 2021.
     207: p. 111231.
216. Liu, D., et al., Differential effects of arsenic species on Nrf2 and Bach1 nuclear
     localization in cultured hepatocytes. Toxicol Appl Pharmacol, 2021. 413: p. 115404.
217. Takaya, K., et al., Validation of the multiple sensor mechanism of the Keap1-Nrf2 system.
     Free Radic Biol Med, 2012. 53(4): p. 817-27.
218. Kobayashi, M., et al., The antioxidant defense system Keap1-Nrf2 comprises a multiple
     sensing mechanism for responding to a wide range of chemical compounds. Mol Cell
     Biol, 2009. 29(2): p. 493-502.
219. Robledinos-Antón, N., et al., Activators and Inhibitors of NRF2: A Review of Their
     Potential for Clinical Development. Oxid Med Cell Longev, 2019. 2019: p. 9372182.
220. Imhoff, B.R. and J.M. Hansen, Tert-butylhydroquinone induces mitochondrial oxidative
     stress causing Nrf2 activation. Cell Biol Toxicol, 2010. 26(6): p. 541-51.
221. Matsunaga, T., et al., Expression of MRP and cMOAT in childhood neuroblastomas and
     malignant liver tumors and its relevance to clinical behavior. Jpn J Cancer Res, 1998.
     89(12): p. 1276-83.
222. Saito, R., et al., Characterizations of Three Major Cysteine Sensors of Keap1 in Stress
     Response. Mol Cell Biol, 2016. 36(2): p. 271-84.
223. Wang, X.J., et al., Activation of the NRF2 signaling pathway by copper-mediated redox
     cycling of para- and ortho-hydroquinones. Chem Biol, 2010. 17(1): p. 75-85.
224. Pi, J., et al., Transcription factor Nrf2 activation by inorganic arsenic in cultured
     keratinocytes: involvement of hydrogen peroxide. Exp Cell Res, 2003. 290(2): p. 234-45.
225. He, X., et al., Arsenic induces NAD(P)H-quinone oxidoreductase I by disrupting the Nrf2 x
     Keap1 x Cul3 complex and recruiting Nrf2 x Maf to the antioxidant response element
     enhancer. J Biol Chem, 2006. 281(33): p. 23620-31.
226. Reichard, J.F. and A. Puga, Effects of arsenic exposure on DNA methylation and
     epigenetic gene regulation. Epigenomics, 2010. 2(1): p. 87-104.
227. Ruan, Y., et al., Metabolic reprogramming in the arsenic carcinogenesis. Ecotoxicol
     Environ Saf, 2022. 229: p. 113098.
                                                 185


228. Zhou, X., et al., Arsenic co-carcinogenesis: Inhibition of DNA repair and interaction with
     zinc finger proteins. Semin Cancer Biol, 2021. 76: p. 86-98.
229. Itoh, K., et al., Transcription factor Nrf2 regulates inflammation by mediating the effect
     of 15-deoxy-Delta(12,14)-prostaglandin j(2). Mol Cell Biol, 2004. 24(1): p. 36-45.
230. McMahon, M., S.R. Swift, and J.D. Hayes, Zinc-binding triggers a conformational-switch
     in the cullin-3 substrate adaptor protein KEAP1 that controls transcription factor NRF2.
     Toxicol Appl Pharmacol, 2018. 360: p. 45-57.
231. Shahidi, F., Antioxidant factors in plant foods and selected oilseeds. Biofactors, 2000.
     13(1-4): p. 179-85.
232. Shahidi, F., Antioxidants in food and food antioxidants. Nahrung, 2000. 44(3): p. 158-63.
233. Almeida, E.S., et al., Behaviour of the antioxidant tert-butylhydroquinone on the storage
     stability and corrosive character of biodiesel. Fuel, 2011. 90(11): p. 3480-3484.
234. FAO/WHO, 19th JECFA Meeting. Toxicological Evaluation of Some Food Colours,
     Thickening Agents, and Certain Other Substances., in 19th JECFA Meeting. 1975: Geneva.
235. Administration, F.a.D. Food additive status List. 2021 08/26/2021 [cited 2022 07-15-
     2022]; Available from: https://www.fda.gov/food/food-additives-petitions/food-
     additive-status-list.
236. EFSA, Opinion of the Scientific Panel on food additives, flavourings, processing aids and
     materials in contact with food (AFC) on a request from the Commission related to
     tertiary-Butylhydroquinone (TBHQ). 2004, EFSA: EFSA Journal.
237. Astill, B.D., Blakeley, R.V., Cantor, E.E., Tischer, K.S., Walton, D.A., McEwan, D.B., Jones,
     W.H., and Ely, T.S., Biochemical studies on tert-butylhydroquinone (TBHQ): A summary.
     Unpublished report from the Laboratory of Industrial Medicine, Eastman Kodak. , 1968.
238. Products, E.C., Two-year chronic feeding studies with tertiary butyl hydroquinone (TBHQ)
     in dogs. Unpublished report from the Food and Drug Research Laboratories, Inc., 1968.
239. FAO/WHO, Evaluation of National Assessments of Intake of tert-Butylhydroquinone., in
     51st JECFA Meeting. 1999.
240. House, R., Regulatory Guidance in Immunotoxicology. Encyclopedia of
     Immunotoxicology. 2005.
241. Naidenko, O.V., et al., Investigating Molecular Mechanisms of Immunotoxicity and the
     Utility of ToxCast for Immunotoxicity Screening of Chemicals Added to Food. Int J Environ
     Res Public Health, 2021. 18(7).
                                                 186


242. Levy, S. and H.J. Forman, C-Myc is a Nrf2-interacting protein that negatively regulates
     phase II genes through their electrophile responsive elements. IUBMB Life, 2010. 62(3):
     p. 237-46.
243. Ma, Q., et al., Induction of murine NAD(P)H:quinone oxidoreductase by 2,3,7,8-
     tetrachlorodibenzo-p-dioxin requires the CNC (cap 'n' collar) basic leucine zipper
     transcription factor Nrf2 (nuclear factor erythroid 2-related factor 2): cross-interaction
     between AhR (aryl hydrocarbon receptor) and Nrf2 signal transduction. Biochem J, 2004.
     377(Pt 1): p. 205-13.
244. Yang, H., et al., Nrf1 and Nrf2 regulate rat glutamate-cysteine ligase catalytic subunit
     transcription indirectly via NF-kappaB and AP-1. Mol Cell Biol, 2005. 25(14): p. 5933-46.
245. Rossler, O.G. and G. Thiel, Specificity of Stress-Responsive Transcription Factors Nrf2,
     ATF4, and AP-1. J Cell Biochem, 2017. 118(1): p. 127-140.
246. Pronk, T.E., et al., Comparison of the molecular topologies of stress-activated
     transcription factors HSF1, AP-1, NRF2, and NF-kappaB in their induction kinetics of
     HMOX1. Biosystems, 2014. 124: p. 75-85.
247. Waxman, E.A., Bach2 is a potent repressor of Nrf2-mediated antioxidant enzyme
     expression in dopaminergic neurons. bioRxiv, 2019: p. 687590.
248. Igarashi, K. and J. Sun, The heme-Bach1 pathway in the regulation of oxidative stress
     response and erythroid differentiation. Antioxid Redox Signal, 2006. 8(1-2): p. 107-18.
249. Cuadrado, A., et al., Transcription factors NRF2 and NF-kappaB are coordinated effectors
     of the Rho family, GTP-binding protein RAC1 during inflammation. J Biol Chem, 2014.
     289(22): p. 15244-58.
250. Vaz, M., et al., Oxidant-induced cell death and Nrf2-dependent antioxidative response
     are controlled by Fra-1/AP-1. Mol Cell Biol, 2012. 32(9): p. 1694-709.
251. Cordova, E.J., et al., The NRF2 gene variant, -653G/A, is associated with nephritis in
     childhood-onset systemic lupus erythematosus. Lupus, 2010. 19(10): p. 1237-42.
252. Sasaki, H., et al., Polymorphisms of NRF2 gene correlated with decreased FEV1 in lung
     cancers of smokers. Biomed Rep, 2013. 1(3): p. 484-488.
253. Masuko, H., et al., An interaction between Nrf2 polymorphisms and smoking status
     affects annual decline in FEV1: a longitudinal retrospective cohort study. BMC Med
     Genet, 2011. 12: p. 97.
254. Marzec, J.M., et al., Functional polymorphisms in the transcription factor NRF2 in
     humans increase the risk of acute lung injury. Faseb j, 2007. 21(9): p. 2237-46.
                                               187


255. Shimoyama, Y., et al., Polymorphism of Nrf2, an antioxidative gene, is associated with
     blood pressure and cardiovascular mortality in hemodialysis patients. Int J Med Sci,
     2014. 11(7): p. 726-31.
256. Huppke, P., et al., Activating de novo mutations in NFE2L2 encoding NRF2 cause a
     multisystem disorder. Nature Communications, 2017. 8(1): p. 818.
257. Rockwell, C.E., et al., The Complicated Role of Nuclear Factor Erythroid-Derived 2-Like 2
     in Allergy and Asthma. Drug Metabolism and Disposition, 2022. 50(4): p. 500-507.
258. Ma, Q., L. Battelli, and A.F. Hubbs, Multiorgan Autoimmune Inflammation, Enhanced
     Lymphoproliferation, and Impaired Homeostasis of Reactive Oxygen Species in Mice
     Lacking the Antioxidant-Activated Transcription Factor <em>Nrf2</em>. The American
     Journal of Pathology, 2006. 168(6): p. 1960-1974.
259. Li, J., T.D. Stein, and J.A. Johnson, Genetic dissection of systemic autoimmune disease in
     Nrf2-deficient mice. Physiological Genomics, 2004. 18(3): p. 261-272.
260. Johnson, D.A., et al., The Absence of the Pro-antioxidant Transcription Factor Nrf2
     Exacerbates Experimental Autoimmune Encephalomyelitis. Toxicological Sciences, 2009.
     114(2): p. 237-246.
261. Larabee, C.M., et al., Loss of Nrf2 exacerbates the visual deficits and optic neuritis
     elicited by experimental autoimmune encephalomyelitis. Molecular vision, 2016. 22: p.
     1503.
262. Berger, A.A., et al., Monomethyl Fumarate (MMF, Bafiertam) for the Treatment of
     Relapsing Forms of Multiple Sclerosis (MS). Neurol Int, 2021. 13(2): p. 207-223.
263. Wang, Y. and P. Bhargava, Diroximel fumarate to treat multiple sclerosis. Drugs Today
     (Barc), 2020. 56(7): p. 431-437.
264. Goess, C., et al., NRF2 activator A-1396076 ameliorates inflammation in autoimmune
     disease models by inhibiting antigen dependent T cell activation. J Transl Autoimmun,
     2021. 4: p. 100079.
265. Kuo, P.-C., et al., 3H-1, 2-dithiole-3-thione as a novel therapeutic agent for the treatment
     of experimental autoimmune encephalomyelitis. Brain, Behavior, and Immunity, 2016.
     57: p. 173-186.
266. Geisel, J., et al., Sulforaphane Protects from T Cell–Mediated Autoimmune Disease by
     Inhibition of IL-23 and IL-12 in Dendritic Cells. The Journal of Immunology, 2014. 192(8):
     p. 3530-3539.
                                                188


267. Kuo, P.-C., et al., Dimethyl itaconate, an itaconate derivative, exhibits
     immunomodulatory effects on neuroinflammation in experimental autoimmune
     encephalomyelitis. Journal of neuroinflammation, 2020. 17(1): p. 1-15.
268. Rangasamy, T., et al., Disruption of Nrf2 enhances susceptibility to severe airway
     inflammation and asthma in mice. J Exp Med, 2005. 202(1): p. 47-59.
269. Sussan, T.E., et al., Nrf2 reduces allergic asthma in mice through enhanced airway
     epithelial cytoprotective function. Am J Physiol Lung Cell Mol Physiol, 2015. 309(1): p.
     L27-36.
270. Hirai, K., et al., Impact of gene expression associated with glucocorticoid-induced
     transcript 1 (GLCCI1) on severe asthma and future exacerbation. Biological and
     Pharmaceutical Bulletin, 2019. 42(10): p. 1746-1752.
271. Brown, R.H., et al., Sulforaphane improves the bronchoprotective response in asthmatics
     through Nrf2-mediated gene pathways. Respiratory research, 2015. 16(1): p. 1-12.
272. Ano, S., et al., Inflammation and airway hyperresponsiveness after chlorine exposure are
     prolonged by Nrf2 deficiency in mice. Free Radical Biology and Medicine, 2017. 102: p. 1-
     15.
273. Sakurai, H., et al., Sulforaphane ameliorates steroid insensitivity through an Nrf2-
     dependent pathway in cigarette smoke-exposed asthmatic mice. Free Radical Biology
     and Medicine, 2018. 129: p. 473-485.
274. Shintani, Y., et al., Nuclear factor erythroid 2-related factor 2 (Nrf2) regulates airway
     epithelial barrier integrity. Allergology International, 2015. 64(Supplement. 1): p. 54-63.
275. Kang, S.H., et al., Effect of paracetamol use on the modification of the development of
     asthma by reactive oxygen species genes. Ann Allergy Asthma Immunol, 2013. 110(5): p.
     364-369 e1.
276. Wang, C., et al., Aloperine suppresses allergic airway inflammation through NF-kappaB,
     MAPK, and Nrf2/HO-1 signaling pathways in mice. Int Immunopharmacol, 2018. 65: p.
     571-579.
277. Nagashima, R., et al., Nrf2 Suppresses Allergic Lung Inflammation by Attenuating the
     Type 2 Innate Lymphoid Cell Response. J Immunol, 2019. 202(5): p. 1331-1339.
278. Ogawa, T., et al., NRF2 Augments Epidermal Antioxidant Defenses and Promotes Atopy.
     The Journal of Immunology, 2020. 205(4): p. 907-914.
                                               189


279. Choi, J.H., et al., Platycodon grandiflorum root-derived saponins attenuate atopic
     dermatitis-like skin lesions via suppression of NF-κB and STAT1 and activation of
     Nrf2/ARE-mediated heme oxygenase-1. Phytomedicine, 2014. 21(8-9): p. 1053-1061.
280. Karuppagounder, V., et al., Modulation of HMGB 1 translocation and RAGE/NF κB
     cascade by quercetin treatment mitigates atopic dermatitis in NC/Nga transgenic mice.
     Experimental dermatology, 2015. 24(6): p. 418-423.
281. Wu, W., et al., Sulforaphane has a therapeutic effect in an atopic dermatitis murine
     model and activates the Nrf2/HO-1 axis. Molecular medicine reports, 2019. 20(2): p.
     1761-1771.
282. Akram, M., et al., A newly synthesized macakurzin C-derivative attenuates acute and
     chronic skin inflammation: The Nrf2/heme oxygenase signaling as a potential target.
     Toxicology and applied pharmacology, 2016. 307: p. 62-71.
283. Yu, C.-H., et al., Inhibitory effects of a novel chrysin-derivative, CPD 6, on acute and
     chronic skin inflammation. International journal of molecular sciences, 2019. 20(11): p.
     2607.
284. Jin, Y., et al., The role of the stress-activated transcription factor Nrf2 in OVA-induced
     food allergy. The Journal of Immunology, 2021. 206(1 Supplement): p. 94.08-94.08.
285. Rockwell, C., et al., The Nrf2 activator, tBHQ, exacerbates immediate hypersensitivity
     response to food allergen (HYP7P.315). The Journal of Immunology, 2014. 192(1
     Supplement): p. 119.30-119.30.
286. Tsai, J.J., et al., Nrf2 regulates haematopoietic stem cell function. Nat Cell Biol, 2013.
     15(3): p. 309-16.
287. Murakami, S., M. Yamamoto, and H. Motohashi, Hematopoietic stem and progenitor cell
     activation during chronic dermatitis provoked by constitutively active aryl-hydrocarbon
     receptor driven by Keratin 14 promoter. Toxicol Sci, 2014. 138(1): p. 47-58.
288. Boss, A.P., et al., The Nrf2 activator tBHQ inhibits the activation of primary murine
     natural killer cells. Food Chem Toxicol, 2018. 121: p. 231-236.
289. Boss, A., et al., The synthetic food additive tert-butylhydroquinone impairs early NK cell
     responses to influenza. The FASEB Journal, 2021. 35(S1).
290. Kumar, P., et al., IL-27 promotes NK cell effector functions via Maf-Nrf2 pathway during
     influenza infection. Scientific Reports, 2019. 9(1): p. 4984.
291. Saddawi-Konefka, R., et al., Nrf2 Induces IL-17D to Mediate Tumor and Virus
     Surveillance. Cell Rep, 2016. 16(9): p. 2348-58.
                                                  190


292. Seelige, R., A. Washington, and J.D. Bui, The ancient cytokine IL-17D is regulated by Nrf2
     and mediates tumor and virus surveillance. Cytokine, 2017. 91: p. 10-12.
293. Pyaram, K., et al., Keap1-Nrf2 System Plays an Important Role in Invariant Natural Killer
     T Cell Development and Homeostasis. Cell Reports, 2019. 27(3): p. 699-707.e4.
294. Rangasamy, T., et al., Nuclear erythroid 2 p45-related factor 2 inhibits the maturation of
     murine dendritic cells by ragweed extract. Am J Respir Cell Mol Biol, 2010. 43(3): p. 276-
     85.
295. Williams, M.A., et al., Disruption of the transcription factor Nrf2 promotes pro-oxidative
     dendritic cells that stimulate Th2-like immunoresponsiveness upon activation by ambient
     particulate matter. J Immunol, 2008. 181(7): p. 4545-59.
296. Yeang, H.X.A., et al., Loss of transcription factor nuclear factor-erythroid 2 (NF-E2) p45-
     related factor-2 (Nrf2) leads to dysregulation of immune functions, redox
     homeostasis,and intracellular signaling in dendritic cells. Journal of Biological Chemistry,
     2012.
297. Wang, J., et al., Nrf2 suppresses the function of dendritic cells to facilitate the immune
     escape of glioma cells. Experimental Cell Research, 2017. 360(2): p. 66-73.
298. Wei, H.J., et al., Nrf2-mediated metabolic reprogramming of tolerogenic dendritic cells is
     protective against aplastic anemia. J Autoimmun, 2018. 94: p. 33-44.
299. Ko, S.H., et al., Crude Preparations of Helicobacter pylori Outer Membrane Vesicles
     Induce Upregulation of Heme Oxygenase-1 via Activating Akt-Nrf2 and mTOR-IkappaB
     Kinase-NF-kappaB Pathways in Dendritic Cells. Infect Immun, 2016. 84(8): p. 2162-2174.
300. Lugade, A.A., et al., Nrf2 regulates chronic lung inflammation and B-cell responses to
     nontypeable Haemophilus influenzae. Am J Respir Cell Mol Biol, 2011. 45(3): p. 557-65.
301. Macoch, M., et al., Nrf2-dependent repression of interleukin-12 expression in human
     dendritic cells exposed to inorganic arsenic. Free Radic Biol Med, 2015. 88(Pt B): p. 381-
     390.
302. Cho, H.Y., et al., Role of NRF2 in protection against hyperoxic lung injury in mice. Am J
     Respir Cell Mol Biol, 2002. 26(2): p. 175-82.
303. Harvey, C.J., et al., Targeting Nrf2 signaling improves bacterial clearance by alveolar
     macrophages in patients with COPD and in a mouse model. Sci Transl Med, 2011. 3(78):
     p. 78ra32.
304. Wang, Z., et al., Tim-3 inhibits macrophage control of Listeria monocytogenes by
     inhibiting Nrf2. Sci Rep, 2017. 7: p. 42095.
                                               191


305. Orecchioni, M., et al., Macrophage Polarization: Different Gene Signatures in M1(LPS+)
     vs. Classically and M2(LPS-) vs. Alternatively Activated Macrophages. Front Immunol,
     2019. 10: p. 1084.
306. Kobayashi, E.H., et al., Nrf2 suppresses macrophage inflammatory response by blocking
     proinflammatory cytokine transcription. Nat Commun, 2016. 7: p. 11624.
307. Kadl, A., et al., Identification of a novel macrophage phenotype that develops in response
     to atherogenic phospholipids via Nrf2. Circ Res, 2010. 107(6): p. 737-46.
308. Ruotsalainen, A.K., et al., The absence of macrophage Nrf2 promotes early
     atherogenesis. Cardiovasc Res, 2013. 98(1): p. 107-15.
309. Mimura, J. and K. Itoh, Role of Nrf2 in the pathogenesis of atherosclerosis. Free Radic
     Biol Med, 2015. 88(Pt B): p. 221-232.
310. Wang, Y.-R., et al., Targeting macrophage polarization by Nrf2 agonists for treating
     various xenobiotics-induced toxic responses. Toxicology Mechanisms and Methods,
     2021. 31(5): p. 334-342.
311. Kikuchi, N., et al., Nrf2 protects against pulmonary fibrosis by regulating the lung
     oxidant level and Th1/Th2 balance. Respir Res, 2010. 11: p. 31.
312. Noel, S., et al., T Lymphocyte-Specific Activation of Nrf2 Protects from AKI. J Am Soc
     Nephrol, 2015. 26(12): p. 2989-3000.
313. Noel, S., A.R. Hamad, and H. Rabb, Reviving the promise of transcription factor Nrf2-
     based therapeutics for kidney diseases. Kidney Int, 2015. 88(6): p. 1217-1218.
314. Zagorski, J.W., et al., The Nrf2 activator, tBHQ, differentially affects early events
     following stimulation of Jurkat cells. Toxicol Sci, 2013. 136(1): p. 63-71.
315. Zagorski, J.W., et al., Nrf2-Dependent and -Independent Effects of tert-
     Butylhydroquinone, CDDO-Im, and H2O2 in Human Jurkat T Cells as Determined by
     CRISPR/Cas9 Gene Editing. J Pharmacol Exp Ther, 2017. 361(2): p. 259-267.
316. Turley, A.E., J.W. Zagorski, and C.E. Rockwell, The Nrf2 activator tBHQ inhibits T cell
     activation of primary human CD4 T cells. Cytokine, 2015. 71(2): p. 289-95.
317. Noel, S., et al., KEAP1 Editing Using CRISPR/Cas9 for Therapeutic NRF2 Activation in
     Primary Human T Lymphocytes. J Immunol, 2018. 200(5): p. 1929-1936.
318. Klemm, P., et al., Nrf2 expression driven by Foxp3 specific deletion of Keap1 results in
     loss of immune tolerance in mice. Eur J Immunol, 2020. 50(4): p. 515-524.
                                                 192


319. Jang, J.-W., et al., Bisphenol A and its substitutes regulate human B cell survival via Nrf2
     expression. Environmental Pollution, 2020. 259: p. 113907.
320. Moon, S.-J., et al., The anti-arthritis effect of sulforaphane, an activator of Nrf2, is
     associated with inhibition of both B cell differentiation and the production of
     inflammatory cytokines. PLOS ONE, 2021. 16(2): p. e0245986.
321. Gautam, P., et al., Proportion of B cell subsets and Nrf2 mediated redox regulation in
     systemic lupus erythematosus patients. Immunobiology, 2022. 227(2): p. 152180.
322. Kim, M.J., J. Nriagu, and S. Haack, Arsenic species and chemistry in groundwater of
     southeast Michigan. Environ Pollut, 2002. 120(2): p. 379-90.
323. Hossain, M.F., Arsenic contamination in Bangladesh—An overview. Agriculture,
     Ecosystems & Environment, 2006. 113(1): p. 1-16.
324. Ayotte, J.D., B.T. Nolan, and J.A. Gronberg, Predicting Arsenic in Drinking Water Wells of
     the Central Valley, California. Environ Sci Technol, 2016. 50(14): p. 7555-63.
325. Sorg, T.J., A.S. Chen, and L. Wang, Arsenic species in drinking water wells in the USA with
     high arsenic concentrations. Water Res, 2014. 48: p. 156-69.
326. Lievremont, D., P.N. Bertin, and M.C. Lett, Arsenic in contaminated waters:
     biogeochemical cycle, microbial metabolism and biotreatment processes. Biochimie,
     2009. 91(10): p. 1229-37.
327. Tournel, G., et al., Acute arsenic poisoning: clinical, toxicological, histopathological, and
     forensic features. J Forensic Sci, 2011. 56 Suppl 1: p. S275-9.
328. Lech, T. and F. Trela, Massive acute arsenic poisonings. Forensic Sci Int, 2005. 151(2-3):
     p. 273-7.
329. Kapaj, S., et al., Human health effects from chronic arsenic poisoning--a review. J Environ
     Sci Health A Tox Hazard Subst Environ Eng, 2006. 41(10): p. 2399-428.
330. VanDenBerg, K.R., et al., Inhibition of early T cell cytokine production by arsenic trioxide
     occurs independently of Nrf2. PLoS One, 2017. 12(10): p. e0185579.
331. Freeborn, R.A., et al., Trivalent arsenic impairs the effector response of human CD4(+)
     and CD8(+) T cells to influenza A virus ex vivo. Food Chem Toxicol, 2022. 165: p. 113122.
332. Casali, P., L.E. Silberstein, and New York Academy of Sciences., Immunoglobulin gene
     expression in development and disease. Annals of the New York Academy of Sciences,.
     1995, New York: New York Academy of Sciences. xvi, 586 p.
                                                193


333. Loffert, D., et al., Early B-cell development in the mouse: insights from mutations
     introduced by gene targeting. Immunol Rev, 1994. 137: p. 135-53.
334. Rajewsky, K., Clonal selection and learning in the antibody system. Nature, 1996.
     381(6585): p. 751-8.
335. Coombs, C.C., M. Tavakkoli, and M.S. Tallman, Acute promyelocytic leukemia: where did
     we start, where are we now, and the future. Blood Cancer J, 2015. 5: p. e304.
336. Mir, M.A., Developing Costimulatory Molecules for Immunotherapy of Diseases.
     Academic Press. 2015.
337. Kuchroo, V.K., et al., B7-1 and B7-2 costimulatory molecules activate differentially the
     Th1/Th2 developmental pathways: application to autoimmune disease therapy. Cell,
     1995. 80(5): p. 707-18.
338. Kasper, D.L., editor, Harrison's principles of internal medicine. 19th edition / editors,
     Dennis L. Kasper, Anthony S. Fauci, Stephen L. Hauser, Dan L. Longo, J. Larry Jameson,
     Joseph Loscalzo. ed. 2015, United States: New York : McGraw Hill Education Medical,
     [2015].
339. Xu, S. and K.P. Lam, Transmembrane Activator and CAML Interactor (TACI): Another
     Potential Target for Immunotherapy of Multiple Myeloma? Cancers (Basel), 2020. 12(4).
340. Brisslert, M., et al., Phenotypic and functional characterization of human CD25+ B cells.
     Immunology, 2006. 117(4): p. 548-57.
341. Amu, S., et al., The human immunomodulatory CD25+ B cell population belongs to the
     memory B cell pool. Scand J Immunol, 2007. 66(1): p. 77-86.
342. Mir, M.A., Chapter 1 - Introduction to Costimulation and Costimulatory Molecules, in
     Developing Costimulatory Molecules for Immunotherapy of Diseases, M.A. Mir, Editor.
     2015, Academic Press. p. 1-43.
343. Zhao, B., et al., Immunosuppressive effect of arsenic trioxide on islet xenotransplantation
     prolongs xenograft survival in mice. Cell Death Dis, 2018. 9(3): p. 408.
344. Jiao, Z.X., et al., As2 O3 combined with leflunomide prolongs heart xenograft survival via
     suppressing the response of Th1, Th2, and B cells in a rat model. Xenotransplantation,
     2016. 23(3): p. 237-48.
345. Ullah, N., et al., Study of the effect of inorganic and organic complexes of arsenic metal
     on the status of GSH in T. cells and B. cells of blood. Pak J Pharm Sci, 2015. 28(2): p. 457-
     64.
                                                 194


346. Burchiel, S.W., et al., Immunotoxicity and biodistribution analysis of arsenic trioxide in
     C57Bl/6 mice following a 2-week inhalation exposure. Toxicol Appl Pharmacol, 2009.
     241(3): p. 253-9.
347. Cui, W., et al., Arsenic trioxide at conventional dosage does not aggravate hemorrhage
     in the first-line treatment of adult acute promyelocytic leukemia. Eur J Haematol, 2018.
     100(4): p. 344-350.
348. Abernathy, C.O., D.J. Thomas, and R.L. Calderon, Health effects and risk assessment of
     arsenic. J Nutr, 2003. 133(5 Suppl 1): p. 1536S-8S.
349. Abernathy, C.O., Calderon, R.L. and Chappell, W.R. eds., , Arsenic: exposure and health
     effects. . 2012: Springer Science & Business Media.
350. Borum, D.R.A., C. O, Human oral exposure to inorganic arsenic. Arsenic: Exposure and
     Health(Chappell W. R., Abernathy, C. O.& Cothern, C. R., eds.), 1994. Science and
     Technology Letters, Northwood, England. p. pp. 21–29.
351. Kurokawa, M., et al., Investigation of skin manifestations of arsenicism due to intake of
     arsenic-contaminated groundwater in residents of Samta, Jessore, Bangladesh. Arch
     Dermatol, 2001. 137(1): p. 102-3.
352. Guha Mazumder, D.N., et al., Arsenic levels in drinking water and the prevalence of skin
     lesions in West Bengal, India. Int J Epidemiol, 1998. 27(5): p. 871-7.
353. Smith, A.H., et al., Marked increase in bladder and lung cancer mortality in a region of
     Northern Chile due to arsenic in drinking water. Am J Epidemiol, 1998. 147(7): p. 660-9.
354. Chen, C.L., et al., Ingested arsenic, characteristics of well water consumption and risk of
     different histological types of lung cancer in northeastern Taiwan. Environ Res, 2010.
     110(5): p. 455-62.
355. Leslie A. DeSimone, P.B.M., and Michael R. Rosen, Water Quality in Principal Aquifers of
     the United States, 1991–2010, in The Quality of Our Nation’s Waters, U.S.G.S.C. 1360,
     Editor. 2015. p. 151.
356. Ghosh, D., et al., New insights into B cells as antigen presenting cells. Curr Opin
     Immunol, 2021. 70: p. 129-137.
357. Chen, K. and A. Cerutti, New insights into the enigma of immunoglobulin D. Immunol
     Rev, 2010. 237(1): p. 160-79.
358. Bertolotti, M., et al., B- to plasma-cell terminal differentiation entails oxidative stress
     and profound reshaping of the antioxidant responses. Antioxid Redox Signal, 2010.
     13(8): p. 1133-44.
                                                195


359. Ma, Q., Role of nrf2 in oxidative stress and toxicity. Annual review of pharmacology and
     toxicology, 2013. 53: p. 401-426.
360. Itoh, K., et al., Keap1 represses nuclear activation of antioxidant responsive elements by
     Nrf2 through binding to the amino-terminal Neh2 domain. Genes Dev, 1999. 13(1): p.
     76-86.
361. Tonelli, C., I.I.C. Chio, and D.A. Tuveson, Transcriptional Regulation by Nrf2. Antioxid
     Redox Signal, 2018. 29(17): p. 1727-1745.
362. Moinova, H.R. and R.T. Mulcahy, Up-regulation of the human gamma-glutamylcysteine
     synthetase regulatory subunit gene involves binding of Nrf-2 to an electrophile
     responsive element. Biochem Biophys Res Commun, 1999. 261(3): p. 661-8.
363. Alam, J., et al., Nrf2, a Cap'n'Collar transcription factor, regulates induction of the heme
     oxygenase-1 gene. J Biol Chem, 1999. 274(37): p. 26071-8.
364. Turley, A.E., et al., Chronic low-level cadmium exposure in rats affects cytokine
     production by activated T cells. Toxicol Res (Camb), 2019. 8(2): p. 227-237.
365. Zagorski, J.W., et al., Differential effects of the Nrf2 activators tBHQ and CDDO-Im on the
     early events of T cell activation. Biochem Pharmacol, 2018. 147: p. 67-76.
366. Wang, K., G. Wei, and D. Liu, CD19: a biomarker for B cell development, lymphoma
     diagnosis and therapy. Experimental hematology & oncology, 2012. 1(1): p. 36-36.
367. Xia, X.-Z., et al., Taci Is a Traf-Interacting Receptor for Tall-1, a Tumor Necrosis Factor
     Family Member Involved in B Cell Regulation. Journal of Experimental Medicine, 2000.
     192(1): p. 137-144.
368. Hjelm, F., M.C. Karlsson, and B. Heyman, A novel B cell-mediated transport of IgE-
     immune complexes to the follicle of the spleen. J Immunol, 2008. 180(10): p. 6604-10.
369. Delespesse, G., H. Hofstetter, and M. Sarfati, Low-affinity receptor for IgE (FcERII, CD23)
     and its soluble fragments. Int Arch Allergy Appl Immunol, 1989. 90 Suppl 1: p. 41-4.
370. Henningsson, F., et al., IgE-mediated enhancement of CD4+ T cell responses in mice
     requires antigen presentation by CD11c+ cells and not by B cells. PLoS One, 2011. 6(7): p.
     e21760.
371. Uciechowski, P. and W.C.M. Dempke, Interleukin-6: A Masterplayer in the Cytokine
     Network. Oncology, 2020. 98(3): p. 131-137.
372. Cibrián, D. and F. Sánchez-Madrid, CD69: from activation marker to metabolic
     gatekeeper. Eur J Immunol, 2017. 47(6): p. 946-953.
                                                  196


373. Gharavi, N. and A.O. El-Kadi, tert-Butylhydroquinone is a novel aryl hydrocarbon receptor
     ligand. Drug Metab Dispos, 2005. 33(3): p. 365-72.
374. Prescott, S. and K.J. Allen, Food allergy: riding the second wave of the allergy epidemic.
     Pediatr Allergy Immunol, 2011. 22(2): p. 155-60.
375. Muñoz-Furlong, A. and C.C. Weiss, Characteristics of food-allergic patients placing them
     at risk for a fatal anaphylactic episode. Curr Allergy Asthma Rep, 2009. 9(1): p. 57-63.
376. Clark, S. and C.A. Camargo, Jr., Emergency management of food allergy: systems
     perspective. Curr Opin Allergy Clin Immunol, 2005. 5(3): p. 293-8.
377. Leung, A.S.Y., G.W.K. Wong, and M.L.K. Tang, Food allergy in the developing world. J
     Allergy Clin Immunol, 2018. 141(1): p. 76-78 e1.
378. Poulos, L.M., et al., Trends in hospitalizations for anaphylaxis, angioedema, and urticaria
     in Australia, 1993-1994 to 2004-2005. J Allergy Clin Immunol, 2007. 120(4): p. 878-84.
379. Brooks, C., N. Pearce, and J. Douwes, The hygiene hypothesis in allergy and asthma: an
     update. Curr Opin Allergy Clin Immunol, 2013. 13(1): p. 70-7.
380. Stiemsma, L.T., et al., The hygiene hypothesis: current perspectives and future therapies.
     ImmunoTargets and therapy, 2015. 4: p. 143-157.
381. Kensler, T.W., N. Wakabayashi, and S. Biswal, Cell survival responses to environmental
     stresses via the Keap1-Nrf2-ARE pathway. Annu Rev Pharmacol Toxicol, 2007. 47: p. 89-
     116.
382. Gharavi, N., S. Haggarty, and A.O. El-Kadi, Chemoprotective and carcinogenic effects of
     tert-butylhydroquinone and its metabolites. Curr Drug Metab, 2007. 8(1): p. 1-7.
383. Itoh, K., et al., An Nrf2/small Maf heterodimer mediates the induction of phase II
     detoxifying enzyme genes through antioxidant response elements. Biochem Biophys Res
     Commun, 1997. 236(2): p. 313-22.
384. Oettgen, H.C. and R.S. Geha, IgE regulation and roles in asthma pathogenesis. J Allergy
     Clin Immunol, 2001. 107(3): p. 429-40.
385. Gowthaman, U., J.S. Chen, and S.C. Eisenbarth, Regulation of IgE by T follicular helper
     cells. J Leukoc Biol, 2020. 107(3): p. 409-418.
386. Martinez-Murillo, P., et al., CD138 and CD31 Double-Positive Cells Comprise the
     Functional Antibody-Secreting Plasma Cell Compartment in Primate Bone Marrow. Front
     Immunol, 2016. 7: p. 242.
                                               197


387. Halliley, J.L., et al., Long-Lived Plasma Cells Are Contained within the CD19(-
     )CD38(hi)CD138(+) Subset in Human Bone Marrow. Immunity, 2015. 43(1): p. 132-45.
388. Mackey, E. and A.J. Moeser, Sex Differences in Mast Cell–Associated Disorders: A Life
     Span Perspective. Cold Spring Harbor Perspectives in Biology, 2022: p. a039172.
389. Klein, S.L. and K.L. Flanagan, Sex differences in immune responses. Nat Rev Immunol,
     2016. 16(10): p. 626-38.
390. Sicherer, S.H. and H.A. Sampson, Food allergy: A review and update on epidemiology,
     pathogenesis, diagnosis, prevention, and management. J Allergy Clin Immunol, 2018.
     141(1): p. 41-58.
391. Savage, J. and C.B. Johns, Food allergy: epidemiology and natural history. Immunology
     and allergy clinics of North America, 2015. 35(1): p. 45-59.
392. Osborne, N.J., et al., Prevalence of challenge-proven IgE-mediated food allergy using
     population-based sampling and predetermined challenge criteria in infants. J Allergy Clin
     Immunol, 2011. 127(3): p. 668-76 e1-2.
393. Loh, W. and M.L.K. Tang, The Epidemiology of Food Allergy in the Global Context.
     International journal of environmental research and public health, 2018. 15(9): p. 2043.
394. Itoh, K., et al., Regulatory mechanisms of cellular response to oxidative stress. Free Radic
     Res, 1999. 31(4): p. 319-24.
395. Bhatia, S., et al., Different cell surface oligomeric states of B7-1 and B7-2: implications
     for signaling. Proc Natl Acad Sci U S A, 2005. 102(43): p. 15569-74.
396. Engel, P., B Cell–Associated Surface Molecules and B Cell Responses, in Encyclopedia of
     Immunobiology, M.J.H. Ratcliffe, Editor. 2016, Academic Press: Oxford. p. 253-258.
397. Sampson, H.A., Update on food allergy. J Allergy Clin Immunol, 2004. 113(5): p. 805-19;
     quiz 820.
398. Piątosa, B., et al., B cell subsets in healthy children: Reference values for evaluation of B
     cell maturation process in peripheral blood. Cytometry Part B: Clinical Cytometry, 2010.
     78B(6): p. 372-381.
399. Justice, M.J. and P. Dhillon, Using the mouse to model human disease: increasing validity
     and reproducibility. Dis Model Mech, 2016. 9(2): p. 101-3.
400. Gupta, R.S., et al., The prevalence, severity, and distribution of childhood food allergy in
     the United States. Pediatrics, 2011. 128(1).
                                                 198


401. Murphy, S.L., et al., Mortality in the United States, 2020. NCHS Data Brief, 2021(427): p.
     1-8.
402. Mawia, A.M., et al., Inorganic arsenic toxicity and alleviation strategies in rice. J Hazard
     Mater, 2021. 408: p. 124751.
403. Myers, C.R., The effects of chromium(VI) on the thioredoxin system: implications for
     redox regulation. Free Radic Biol Med, 2012. 52(10): p. 2091-107.
404. Li, J., et al., Melatonin protects against chromium(VI)-induced cardiac injury via
     activating the AMPK/Nrf2 pathway. J Inorg Biochem, 2019. 197: p. 110698.
405. Ren, X., et al., The role of Nrf2 in mitigating cadmium-induced oxidative stress of
     Marsupenaeus japonicus. Environ Pollut, 2021. 269: p. 116112.
406. Cao, Y., et al., Activation of Nrf2 by lead sulfide nanoparticles induces impairment of
     learning and memory. Metallomics, 2020. 12(1): p. 34-41.
407. Wataha, J.C., et al., Effect of mercury(II) on Nrf2, thioredoxin reductase-1 and
     thioredoxin-1 in human monocytes. Dent Mater, 2008. 24(6): p. 765-72.
408. Wu, C.M., et al., Genetic engineering in primary human B cells with CRISPR-Cas9
     ribonucleoproteins. J Immunol Methods, 2018. 457: p. 33-40.
409. Corpetti, C., et al., Cannabidiol inhibits SARS-Cov-2 spike (S) protein-induced cytotoxicity
     and inflammation through a PPARγ-dependent TLR4/NLRP3/Caspase-1 signaling
     suppression in Caco-2 cell line. Phytother Res, 2021. 35(12): p. 6893-6903.
410. Mueller, J., M. Matloubian, and J. Zikherman, Cutting edge: An in vivo reporter reveals
     active B cell receptor signaling in the germinal center. Journal of immunology
     (Baltimore, Md. : 1950), 2015. 194(7): p. 2993-2997.
411. Boonyaratanakornkit, J. and J.J. Taylor, Techniques to Study Antigen-Specific B Cell
     Responses. Front Immunol, 2019. 10: p. 1694.
412. Tohyama, Y., T. Takano, and H. Yamamura, B cell responses to oxidative stress. Curr
     Pharm Des, 2004. 10(8): p. 835-9.
413. Yoh, K., et al., Nrf2-deficient female mice develop lupus-like autoimmune nephritis.
     Kidney Int, 2001. 60(4): p. 1343-53.
414. Roth, K., et al., Tracking plasma cell differentiation and survival. Cytometry A, 2014.
     85(1): p. 15-24.
415. Wang, S.S., et al., Tumor-infiltrating B cells: their role and application in anti-tumor
     immunity in lung cancer. Cell Mol Immunol, 2019. 16(1): p. 6-18.
                                                199


416. van Langelaar, J., et al., B and T Cells Driving Multiple Sclerosis: Identity, Mechanisms
     and Potential Triggers. Front Immunol, 2020. 11: p. 760.
417. Asano, Y., et al., Innate-like self-reactive B cells infiltrate human renal allografts during
     transplant rejection. Nat Commun, 2021. 12(1): p. 4372.
418. Yang, Z., J.B. Jung, and C.D.C. Allen, Study of IgE-Producing B Cells Using the Verigem
     Fluorescent Reporter Mouse. Methods Mol Biol, 2018. 1799: p. 247-264.
419. Ito, R., et al., Recent Advances in Allergy Research Using Humanized Mice. Int J Mol Sci,
     2019. 20(11).
420. Stripecke, R., et al., Innovations, challenges, and minimal information for
     standardization of humanized mice. EMBO Mol Med, 2020. 12(7): p. e8662.
                                                200