Detection of Antibodies against Human and Plant Aquaporins in Patients with Multiple Sclerosis

Abstract

Multiple sclerosis (MS) is an autoimmune disease that affects the body’s central nervous system. Around 90% of MS sufferers are diagnosed with relapsing-remitting MS (RRMS). We used ELISA to measure IgG, IgA, and IgM antibodies against linear epitopes of human and plant aquaporins (AQP4) as well as neural antigens in RRMS patients and controls to determine whether patients suffering from RRMS have simultaneous elevations in antibodies against these peptides and antigens. In comparison to controls, significant elevations in isotype-specific antibodies against human and plant AQP4 and neural antigens such as MBP, MOG, and S100B were detected in RRMS patients, indicating a high correlation in antibody reaction between plant aquaporins and brain antigens. This correlation between the reactivities of RRMS patients with various tested antigens was the most significant for the IgM isotype. We conclude that a subclass of patients with RRMS reacts to both plant and human AQP4 peptides. This immune reaction against different plant aquaporins may help in the development of dietary modifications for patients with MS and other neuroimmune disorders.

Full text

Research Article
Detection of Antibodies against Human and Plant Aquaporins in
Patients with Multiple Sclerosis
Aristo Vojdani,1,2 Partha Sarathi Mukherjee,3Joshua Berookhim,1and Datis Kharrazian2,4
1Immunosciences Lab., Inc., 822 S. Robertson Boulevard, Suite 312, Los Angeles, CA 90035, USA
2Department of Preventive Medicine, Loma Linda University, 24785 Stewart Street, Loma Linda, CA 92354, USA
3DepartmentofMathematics,BoiseStateUniversity,1910UniversityDrive,Boise,ID83725,USA
4Department of Clinical Sciences, Bastyr University California, 4106 Sorrento Valley Boulevard, San Diego, CA 92121, USA
Correspondence should be addressed to Aristo Vojdani; drari@msn.com
Received  April ; Revised  July ; Accepted  July 
Academic Editor: Corrado Betterle
Copyright ©  Aristo Vojdani et al. is is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Multiple sclerosis (MS) is an autoimmune disease that aects the body’s central nervous system. Around % of MS suerers are
diagnosed with relapsing-remitting MS (RRMS). We used ELISA to measure IgG, IgA, and IgM antibodies against linear epitopes
of human and plant aquaporins (AQP) as well as neural antigens in RRMS patients and controls to determine whether patients
suering from RRMS have simultaneous elevations in antibodies against these peptides and antigens. In comparison to controls,
signicant elevations in isotype-specic antibodies against human and plant AQP and neural antigens such as MBP, MOG, and
SB were detected in RRMS patients, indicating a high correlation in antibody reaction between plant aquaporins and brain
antigens. is correlation between the reactivities of RRMS patients with various tested antigens was the most signicant for the
IgM isotype. We conclude that a subclass of patients with RRMS reacts to both plant and human AQP peptides. is immune
reaction against dierent plant aquaporins may help in the development of dietary modications for patients with MS and other
neuroimmune disorders.
1. Introduction
Multiple sclerosis (MS) is characterized by the demyelination
of a nerve’s protective myelin sheaths in the brain and spi-
nal cord, which occurs due to inammation and attack by
the body’s own immune system [,]. is myelin damage
disrupts the communication between the brain and the rest
of the body. Symptoms may include fatigue, vertigo, cog-
nitive impairment, focal cortical decits, unilateral painful
loss of vision, postural and action tremor, dysarthria, limb
incoordination and gait ataxia, diplopia, oscillopsia, pseu-
dobulbar palsy, and bladder dysfunction. In , the United
States National Multiple Sclerosis Society described  clinical
courses of the disease []. In , this set of courses was
reviewed by an international panel [], resulting in the recog-
nition of  main phenotypes of MS. e rst type, relapsing-
remitting multiple sclerosis (RRMS), aects around % of
people who have MS. e dening elements of RRMS are
episodes of acute worsening of neurologic function followed
by a variable degree of recovery, with a stable course between
attacks []. e remaining % have one of these three
progressive forms: secondary progressive (SPMS), primary
progressive (PPMS), and progressive relapsing (PRMS).
Aquaporin  (AQP) is a class of water channels found in
many cells of the body including the stomach, brain, lung, and
skeletal muscle []. AQP is the predominant water channel
in the central nervous system and is expressed in ependy-
mocytes, endothelial cells, and astrocyte foot processes at
the blood-brain barriers (BBB), but not in neurons [,].
In the brain, AQP is believed to have a role in maintaining
homeostasis and water exchange, electrical activity, and
modulation of neuronal transmission and excitability [,].
Neuromyelitis optica (NMO), or Devic’s disease, is a
severe inammatory demyelinating disorder that aects the
white and gray matter in the brain and is classically restricted
totheopticnervesandspinalcord[–]. Studies have
Hindawi Publishing Corporation
Autoimmune Diseases
Volume 2015, Article ID 905208, 10 pages
http://dx.doi.org/10.1155/2015/905208

Autoimmune Diseases
shown that a majority of patients with NMO produce anti-
bodies against the extracellular domain of human AQP [–
]. NMO meets all the formal criteria for an autoimmune
etiology [].
Although MS and NMO are now recognized as two dis-
tinct illnesses [,] for years similar clinical manifestations
led to one being misdiagnosed as the other or led some to
think that NMO was a severe form of MS. e introduction of
the NMO antibody permitted clearer dierentiation between
the two disorders and increased the accuracy of diagnosis
[].
In NMO lesions, products of complement cascade are
found within astrocytes and macrophages []. Furthermore,
using the immunouorescence method and human AQP
transfected cell lines, a disease-specic antibody against
extracellular domains of human AQP designated as NMO-
IgG has been detected in the blood of patients [,–].
e binding of IgG1to human AQP, in conjunction with
complement activation, leads to a loss of human AQP func-
tionality in lesions through complement-dependent cyto-
toxicity, tissue damage, and demyelination of the spinal
cord and optic nerve, followed by opening of the BBB [].
Since IgG1against human AQP is produced in the blood,
its access to the extracellular space of the CNS is greater
when the BBB is compromised, which allows the antibodies
to reach their target tissue []. is can lead to many
complications, ranging from mild sensory disturbances to
complete transverse myelitis with tetraplegia or paraplegia,
sensory impairments, bladder-bowel dysfunction, and more
[,].
A variety of plant cells contain aquaporins, through
which water can ow more rapidly inside the cells than by
diusing through the phospholipid bilayers []. In fact, 
plant aquaporin families have been structurally and func-
tionally well-studied and characterized [,]. A recent
study showed a signicant similarity between the amino
acid sequences of soy, spinach, corn, tomato, and tobacco
with human aquaporin epitope – []. Furthermore,
using ELISA, the researchers found that, in comparison to
non-NMO samples, the NMO IgG serum reacted to both
human and corn aquaporin peptides. However, that study
was conducted by measuring only IgG in serum collected
from  conrmed NMO patients,  probable NMO patient,
and  non-NMO controls. Previous studies, including our
own, have demonstrated that IgM and IgA antibodies have
been detected against myelin basic protein (MBP), myelin
oligodendrocyte glycoprotein (MOG), and other neural anti-
gens in subgroups of patients suering from MS and other
neurologic disorders [–]. In fact, in a study on the
importance of antibodies against myelin antigens in demyeli-
nation, Egg et al. showed that while IgG antibodies against
MOG were %, IgM antibodies against MOG were the
highest at %, while IgA had a respectable level at % [].
Given the overlapping symptomatologies between NMO and
MS, in this present study we extended the investigation
to IgG, IgM, and IgA isotype antibody reactivity against 
dierent plant sequences using  patients with RRMS. By
measuring antibodies against MBP, MOG, and SB along
with human and plant aquaporins, we wanted to examine
the association between the elevation in antibodies against
plant aquaporins and neural antigens in patients with RRMS
[–]. We hypothesized that, due to exposure to environ-
mental proteins, antibodies to the linear epitopes of AQP
peptides from humans and plants are detected in patients
with RRMS. ese ndings warrant further investigation into
the role of the environment in RRMS.
2. Material and Methods
2.1. Controls and MS Patients’ Sera. Based on MRI scans,
which show focal or conuent abnormalities in the brain’s
white matter, and clinical examinations that show a pattern
of attack, complete or partial remission, and then a relapse
at a future date, patients were classied as having RRMS and
ranged from  to  years of age (male : female,  : ). We
chose only sera taken from patients upon their diagnosis or
not more than  months aer the initial diagnosis. ese
samples were purchased from Sanguine BioSciences, Inc.
(Valencia, CA, USA) and BioServe (Beltsville, MD, USA).
For comparison,  serum samples with matching age and
sex from healthy donors were purchased from Innovative
Research Inc. (Southeld, MI, USA). ese individuals were
qualied to donate blood based on a health questionnaire
provided by the Food and Drug Administration (FDA). Each
individual at the time of blood draw also did not exhibit any
health complaints. Each blood sample was tested according
to FDA guidelines for the detection of hepatitis B surface
antigen, antibodies to HIV, antibodies to hepatitis C, HIV-
RNA,hepatitisCRNA,andsyphilis.Noneofthesamples
were positive for these antibodies or viral RNA.
2.2. Antigens and Peptides. MBP was purchased from Sigma
Aldrich (St. Louis, MO); SB was obtained from EMD
Biosciences (San Diego, CA); human and plant aquaporin
peptides, MOG peptide – with a purity of greater than
%, and ovalbumin peptide – were ordered from
Bio-Synthesis Inc. (Lewisville, TX). Monoclonal antibodies
made against various aquaporin peptides were purchased
from Sigma Aldrich (St. Louis, MO).
2.3. Detection of IgG, IgM, and IgA Antibodies by Enzyme
Linked Immunosorbent Assay. MBP, SB protein, MOG,
and aquaporin peptides at a concentration of . mg/mL were
each diluted  :  in . M carbonate-bicarbonate buer, pH
.;  𝜇L was added to each well of a polystyrene at-bottom
ELISA plate. Plates were incubated overnight at ∘Candthen
washed three times with  𝜇L phosphate-buered saline
(PBS) containing .% Tween , pH .. e nonspecic
binding of immunoglobulins was prevented by adding %
BSA in PBS and incubated overnight at ∘C. Plates were
washed as described above, and then serum samples from
controls and RRMS patients were diluted  :  in . M PBS
Tween containing % BSA, then added to duplicate wells,
and incubated for  hour at room temperature. Plates were
washed, and then alkaline phosphatase goat anti-human IgG,
IgM, or IgA antibody (Jackson ImmunoResearch Laborato-
ries, Inc. (West Grove, PA)) at an optimal dilution of : 

Autoimmune Diseases 
IgA,  :  IgG, and IgM in % BSA-PBS was added to
each well; plates were incubated for an additional  hour at
room temperature. Aer washing six times with PBS-Tween
buer,theenzymereactionwasstartedbyadding𝜇L
of phosphatase substrate in .mL of diethanolamine buer
of  mg/mL containing  mM MgCl2and sodium azide, pH
.. e reaction was stopped  minutes later with  𝜇Lof
N NaOH. e optical density (OD) was read at  nm by
means of a microplate reader. Several control wells containing
human serum albumin or ovalbumin peptide – were
used for detection of nonspecic binding.
2.4.DeterminationofSpecicityofAntibodyAssay. For the
determination of the specicity of the AQP antibody reac-
tion, serial dilutions of sera as well as inhibition studies were
conducted using specific and nonspecic antigens.
Dierent sera with high levels of IgG, IgM, or IgA
antibodies against each aquaporin were diluted serially from
 :  to  :  and then applied to ELISA plates coated with
the same peptide. Aer completion of the ELISA procedure,
the recorded ODs were used for the generation of curves.
For inhibition,  dierent sera with a very high titer
of IgG, IgA, or IgM antibody against human AQP were
used in the inhibition study. In dierent test tubes, mL of
 :  diluted sera sample was preincubated with  𝜇Lof
diluent containing either 𝜇gHSAorhumanAQPor
spinach, tomato, soy, or corn aquaporins. Aer mixing, the
tubes were kept for  hour at ∘Cwaterbathfollowedby
-hourincubationat
∘C and then centrifuged at  g for
 mins. e supernatant was used for measuring IgG, IgA,
or IgM antibody level against human AQP, before and aer
absorption with dierent aquaporins.
2.5. Coecients of Intra- and Interassay Variation. Coe-
cients of intra-assay variation were calculated by running
ve samples eight times within a single assay. Coecients of
interassay variation were determined by measuring the same
samples in six consecutive assays. is replicate testing estab-
lished the validity of the ELISA, determined the appropriate
dilution with minimal background, and detected serum IgG,
IgM, and IgA against dierent aquaporins. Coecients of
intra- and interassay variations for IgG, IgM, and IgA against
all tested aquaporins were less than %.
2.6. Reaction of Antibody against AQP4 Peptides with Various
AQP4. For measuring anti-AQP reactivity with dierent
AQPpeptides,weusedELISAsimilartoIgG,IgM,andIgA
detection. Aside from the fact that mouse serum was used
instead of human serum and the secondary antibody was
enzyme-labeled anti-mouse IgG, all the other steps were the
same.
2.7. Statistical Analysis. We rst calc u l a t e d P e a r s o n’s corre-
lation coecient between each isotype (lgG, lgA,. and lgM)
of the food proteins (soy aquaporin, corn aquaporin, tomato
aquaporin, and spinach aquaporin) and similar isotype of
brainprotein(MBP,MOG,SB,andhumanaquaporin)
in RRMS patients. Next, we performed simple regression
analysis between each of those combinations and calculated
their 𝑝values. If a 𝑝valueislessthan.,wecon-
clude that particular isotype of food protein signicantly
elevates similar isotype of that particular brain protein.
Finally, we performed a two-way cluster analysis of Pearson’s
correlation coecients between the antibody against food
peptides and the brain proteins in RRMS patients. We
performed all statistical analyses in the statistical soware “R”
(http://www.r-project.org/).
3. Results
3.1. Detection of Antibodies. Sera from  patients with RRMS
and  healthy controls were evaluated by ELISA to measure
IgG, IgA, and IgM antibodies against both plant and human
aquaporins, MBP, MOG, and SB. Results presented as low
and high OD values with the mean ±standard deviation
(SD) are summarized in Tab l e  .eODsforIgGantibody
values obtained with  :  dilution of healthy control sera
ranged from . to ., varying according to subjects and
antigens (Figures –). e mean ±SD of these values ranged
from 0.52 ± 0.26 to 0.81 ± 0.36. e corresponding IgG OD
values for the sera from RRMS patients ranged from .
to ., while the mean ±SD ranged from 0.91 ± 0.45 to
1.15 ± 0.48 (Table ). For the levels of IgG antibody against
almost all the eight tested antigens, the dierences in mean ±
SD in patients versus controls were highly signicant (𝑝<
0.0001)(Figure ).eresultsforIgAantibodiesagainst
neural antigens and various aquaporins in the sera of controls
and RRMS patients are also shown in Tab l e  ,aswellasin
Figure . e levels of serum IgA antibodies against all tested
antigens were also signicantly higher in patients than in
controls. e mean ±SD for controls ranged from 0.49±0.20
to 0.67 ± 0.34.emean±SD for patients ranged from
0.76 ± 0.54 to 1.11 ± 0.68. e levels of IgM antibody against
the neural antigens and plant aquaporins were also examined
in both groups. e individual test results shown in Figure ,
as well as the mean ±SD depicted in Tabl e ,showedeven
more signicant dierences between the control and patient
groups, with all eight antigens showing 𝑝 < 0.0001.We
examinedthedatabasedonmaleandfemalesubjectsandalso
found a signicant dierence between male patients versus
male controls and female patients versus female controls
(𝑝 < 0.0001). However, no signicant dierence was detected
between male and female patients (𝑝 > 0.05).
3.2. Statistical Analysis of the Data for Investigating Association
between the Food Proteins and the Brain Proteins in RRMS
Patients. We tested whether there were signicant associa-
tions between the elevations of each antibody isotype (IgG,
IgA, and IgM) of the brain proteins (MBP, MOG, SB, and
human aquaporin) with the corresponding antibody isotype
ofthefoodproteins(soy,corn,tomato,andspinachaqua-
porins) in RRMS patients. We tted simple linear regression
models between each such pair, and calculated the 𝑅2values
and the 𝑝values. e summary of the results is presented in

Autoimmune Diseases
T : Measurement of antibodies against MBP, MOG, SB, human AQP, soy AQP, corn AQP, tomato AQP, and spinach AQP in controls (C) and patients (P) with RRMS expressed
by ELISA optical densities and mean ±SD.
MBP (OD) MOG (OD) SB (OD) Human AQP (OD) Soy AQP (OD) Corn AQP (OD) Tomato AQP (OD) Spinach AQP (OD)
CPCPCP C P CPCP C P CP
IgG
Low . . . . . . . . . . . . . . . .
High . . . . . . . . . . . . . . . .
Mean ±. . . . . . . . . . . . . . . .
±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±.
𝑝value <. <. <. . <. <. <. <.
IgA
Low . . . . . . . . . . . . . . . .
High . . . . . . . . . . . . . . . .
Mean ±. . . . . . . . . . . . . . . .
±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±.
𝑝value . <. . <. . . . .
IgM
Low . . . . . . . . . . . . . . . .
High . . . . . . . . . . . . . . . .
Mean ±. . . . . . . . . . . . . . . .
±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±. ±.
𝑝value <. <. <. <. <. <. <. <.

Autoimmune Diseases 
MS versus control MBP, MOG, S100 B, human AQP4,
and spinach aquaporin, IgG
soy aquaporin, corn aquaporin, tomato aquaporin,
Optical density at 450nm
0
0.5
1
1.5
2
2.5
3
0
MBP
MOG
S100B
HAQP4
Soy AQP4
Corn AQP4
Tom at o AQ P4
Spinach AQP4
F : IgG antibody values of RRMS patients (blue diamond)
versus controls (red circle): MBP, MOG, SB, human AQP, soy
aquaporin, corn aquaporin, tomato aquaporin, and spinach aqua-
porin. e levels of serum IgG antibodies against almost all tested
antigens were signicantly higher in patients than in controls.
3.5
MS versus control MBP, MOG, S100 B, human AQP4,
and spinach aquaporin, IgA
soy aquaporin, corn aquaporin, tomato aquaporin,
0
Optical density at 450nm
0.5
1
1.5
2
2.5
3
0
MBP
MOG
S100B
HAQP4
Soy AQP4
Corn AQP4
Tom at o AQ P4
Spinach AQP4
F : IgA antibody values of RRMS patients (blue diamond)
versus controls (red circle): MBP, MOG, SB, human AQP, soy
aquaporin, corn aquaporin, tomato aquaporin, and spinach aqua-
porin. e levels of serum IgA antibodies against all tested antigens
were signicantly higher in patients than in controls.
Tables ,and . From the tables we see that all of those
food proteins signicantly elevate similar isotypes of those
four brain proteins or peptides in RRMS patients. Figure 
presents the result of a two-way cluster analysis of Pearson’s
correlation coecients between the food and brain proteins
or peptides where we see that IgG, IgA, and IgM isotypes are
3.5
MS versus control MBP, MOG, S100 B, human AQP4,
and spinach aquaporin, IgM
soy aquaporin, corn aquaporin, tomato aquaporin,
0
Optical density at 450nm
0.5
1
1.5
2
2.5
3
0
MBP
MOG
S100B
HAQP4
Soy AQP4
Corn AQP4
Tom at o AQ P4
Spinach AQP4
F : IgM antibody values of RRMS patients (blue diamond)
versus controls (red circle): MBP, MOG, SB, human AQP, soy
aquaporin, corn aquaporin, tomato aquaporin, and spinach aqua-
porin. e levels of serum IgM antibodies against all tested antigens
showed even more signicant dierences between the control and
patient groups, with all eight antigens showing 𝑝 < 0.0001.
clustered together with high correlations among the food and
brain proteins or peptides in each isotype. While there was
a correlation between the reactivities of the MS patients’ sera
to neural cell antigens and plant aquaporins, the dierences
in IgM antibody reactivity between the two groups was the
most signicant (Figure ).
3.3. Specicity of Antibodies. In order to demonstrate speci-
city of detected antibody and to rule out nonspecic
reaction, in addition to neural cell antigens and aquaporins,
all sera were reacted with wells coated with HSA and OVA
peptide –, followed by the addition of all reagents in
the ELISA. ODs for all tested sera, aer reaction with HSA or
OVA peptide, were less than .. Additionally, serial dilution
of  : – :  of sera with high levels of antibodies against
each aquaporin was performed. Results depicted in Figures
–showed that, in proportion to dilution, a signicant
decline in antibody reactivity was observed.
In addition, inhibition by specic and nonspecic antigen
was conducted by the addition of either HSA, human AQP,
or each plant aquaporin to three dierent sera with a very
high level of IgG antibody against human AQP. e data
summarized in Figure  show that while HSA did not cause
any inhibition of human anti-AQP binding to ELISA wells
coatedwithhumanAQP,theadditionofhumanAQPand
corn, spinach, tomato, and soy aquaporins to the same sera
resulted in inhibition of antibody-antigen reaction by %,
%, %, %, and %, respectively (Figure ).

Autoimmune Diseases
S100B IgA
MOG IgA
Human AQP IgA
MBP IgA
S100B IgG
MOG IgG
Human AQP IgG
MBP IgG
S100B IgM
MOG IgM
Human AQP IgM
MBP IgM
Soy AQP IgM
Corn AQP IgM
Spinach AQP IgM
Tomato AQP IgM
Soy AQP IgG
Corn AQP IgG
Spinach AQP IgG
Tomato AQP IgG
Soy AQP IgA
Corn AQP IgA
Spinach AQP IgA
Tomato AQP IgA
−0.95 0 0.95
F : IgA, IgM, and IgG isotypes are clustered together with
high correlations among the aquaporin peptides and brain proteins
in each isotype.
0
0.5
1
1.5
2
2.5
1: 100 1 : 200 1 : 400 1 : 800 1 : 1600 1 : 3200
Dilution factor
Antibody levels expressed as
optical density at 405nm
F : Serial dilution of IgG (blue diamond), IgA (green tri-
angle), and IgM (red square) antibody against AQP.
4. Discussion
In an earlier study [], it was shown that several proteins in
nature have a signicant similarity in sequence and structure
to human AQP. e researchers found that IgG from the
sera of patients with NMO cross-reacted with a sequence
found in plant aquaporins and that this reactivity was much
higher in NMO patients than in controls. However, only
patientsandcontrolswereinvolvedinthisstudy[],
and no IgM or IgA antibody measurements were done. For
this reason, we wanted to examine whether or not this
0
0.5
1
1.5
2
2.5
1: 100 1 : 200 1 : 400 1 : 800 1 : 1600 1 : 3200
Dilution factor
Antibody levels expressed as
optical density at 405nm
F : Serial dilution of IgG (blue diamond), IgA (green trian-
gle), and IgM (red square) antibody against tomato.
0
0.5
1
1.5
2
2.5
1: 100 1 : 200 1: 400 1: 800 1 : 1600 1 : 3200
Dilution factor
Antibody levels expressed as
optical density at 405nm
F : Serial dilution of IgG (blue diamond), IgA (green trian-
gle), and IgM (red square) antibody against soy.
3
0
0.5
1
1.5
2
2.5
1: 100 1 : 200 1: 400 1 : 800 1 : 1600 1: 3200
Dilution factor
Antibody levels expressed as
optical density at 405nm
F : Serial dilution of IgG (blue diamond), IgA (green trian-
gle), and IgM (red square) antibody against corn.

Autoimmune Diseases 
0
0.5
1
1.5
2
2.5
1: 100 1: 200 1 : 400 1 : 800 1 : 1600 1 : 3200
Dilution factor
Antibody levels expressed as
optical density at 405nm
F : Serial dilution of IgG (blue diamond), IgA (green trian-
gle), and IgM (red square) antibody against spinach.
0
0.5
1
1.5
2
2.5
HAQP4 Spinach
AQP4 AQP4
Tom a t o
AQP4
Soy
AQP4
Corn HSA
Optical density 405 nm
F : Inhibition of human AQP IgG antibody with human
AQP, spinach AQP, tomato AQP, soy AQP, corn AQP, and
HSA. Controls = blue column; patients = red column.
immunoreactivity to aquaporins is unique to NMO or could
also be detected in patients with RRMS. erefore, we studied
IgG, IgM, and IgA isotype antibodies in the sera of  patients
withRRMSagainsthumanAQP,againstplantAQPfrom
soy,corn,spinach,andtomato,andagainstneuralantigens
such as MBP, MOG, and SB.
Elevation in antibodies against MBP, MOG, and alpha-
B-crystallin have been shown as an aid in the diagnosis and
prognosis of MS [,].
SB and AQP are both astrocytic proteins that enter
the bloodstream when there is a disruption of the BBB. is
entry of SB and AQP into the bloodstream can result in
the production of antibodies against them [–].
Based on these studies, we tested the presence of antibod-
ies against human and plant aquaporins in RRMS patients
and examined their correlation with other brain-specic
antibodies detected in a subgroup of MS. As shown in Figures
–and Table , a signicant percentage of RRMS patients
showed elevation not only in antibodies against human AQP
and the aquaporins of soy, corn, tomato, and spinach, but also
against MBP, MOG, and SB. At this point, it is not known
whether the antibodies are reacting rst to human AQP
and then cross-reacting with plant AQP or vice versa. ere
is always a possibility that this reaction against the specic
AQP peptides used in this study is an epiphenomenon. We
think, however, that this probability is slight, not only because
we detected elevations against human and plant aquaporins
and also against MBP, MOG, and SB, but because we also
detected very high correlations between the aquaporins and
the neural antigens.
To test this association between the elevations of antibod-
ies against MBP, MOG, SB, human AQP, and plant aqua-
porins, we used a simple linear regression model between
each such pair and calculated 𝑅2values and 𝑝values (Tables
–). Data presented in these tables show 𝑅2of . to
. for IgG, . to . for IgA, and . to . for
IgM.isregressionanalysissuggestsarelationshipbetween
antibodies against the food and brain proteins and peptides
with the highest correlation between SB and soy AQP.
e relationship between each antibody is also shown in
a -way cluster analysis of Pearson’s correlation coecients
between the food and brain proteins. Each isotype antibody
is clustered together with high correlations between the
food and brain antibodies, with the IgM antibody reactivity
betweenthegroupsbeingthemostsignicant(Figure ).
To support the importance of these AQP antigenic epi-
topes in immunoreactivity, we used  dierent commercially
available monoclonal antibodies made against human AQP
aa –, AQP aa –, and AQP aa –, and
reacted them with human and plant AQP peptides described
in the Vaishnav study []. Only antibody made against
peptide aa – reacted strongly against both human and
all  plant aquaporins. Antibody against peptide aa –
resultedinweakreactivityagainsthuman,corn,andsoy
aquaporins, while antibody made against peptide aa – did
not react at all with any of the AQP peptides used in our
study. is shows heterogeneity in antibody reaction against
various AQP epitopes.
Another study by Iorio et al. []foundthatantibody
against AQP extracellular loop peptide aa – was also
restricted to patients with NMO. is peptide was not used
in either the Vaishnav et al. study []orthispresentstudy.
Using antibody bound to live M and M cells, Iorio’s group
found that while NMO serum bound to % of the AQP in
live cell membranes, only % of the NMO sera reacted with
peptides originating from loops A, C, and E using ELISA and
Western Blot, with the detected loop C antibody being highly
specic to NMO.
erefore, for the dierentiation of NMO from classic
MS, RRMS, and other neuroimmune disorders, it is crucial
to use live cell or tissue-based assays employing native AQP,
rather than assays utilizing peptides selected from extracel-
lular or intracellular loops. However, these assays could not
be used for cross-reactivity studies between plant proteins
and human tissue antigens and as aids in possible dietary
manipulation in autoimmune disorder treatment protocols.
In sum, while cell-based assays for AQP IgG antibodies
are more specic to NMO, our present study shows that IgG,

Autoimmune Diseases
T : Results of the simple linear regression between each pair of lgG isotypes of the food proteins and brain proteins in RRMS patients.
e rst number in each cell presents corresponding Pearson’s correlation coecient and the second number in parentheses presents its 𝑝
value. Small 𝑝values (less than .) are marked in bold. Note that 𝑅2values of these regressions are the squares of Pearson’s correlation
coecients.
IgG MBP (OD) MOG (OD) SB (OD) Human AQP (OD)
Soy AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
Corn AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
Tomato AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
Spinach AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
T : Results of the simple linear regression between each pair of lgA isotypes of the food proteins and brain proteins in RRMS patients.
e rst number in each cell presents corresponding Pearson’s correlation coecient and the second number in parentheses presents its 𝑝
value. Small 𝑝values (less than .) are marked in bold. Note that 𝑅2values of these regressions are the squares of Pearson’s correlation
coecients.
IgA MBP (OD) MOG (OD) SB (OD) Human AQP (OD)
Soy AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
Corn AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
Tomato AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
Spinach AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
T : Results of the simple linear regression between each pair of lgM isotypes of the food proteins and brain proteins in MS patients. e
rst number in each cell presents corresponding Pearson’s correlation coecient and the second number in parentheses presents its 𝑝value.
Small 𝑝values (less than .) are marked in bold. Note that 𝑅2values of these regressions are the squares of Pearson’s correlation coecients.
IgM MBP (OD) MOG (OD) SB (OD) Human AQP (OD)
Soy AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
Corn AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
Tomato AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
Spinach AQP . (<0.0001) . (<0.0001) . (<0.0001) . (<0.0001)
IgA, and, in particular, IgM antibodies against AQP peptides
canalsobedetectedinpatientswithMS.Itispossible
that exposure to epitopes that resemble human AQP from
exogenous sources such as plants may play a role in the
etiology of RRMS and possibly other autoimmune disor-
ders. Although an association between plant antigens and
autoimmune diseases has been previously suggested for celiac
disease, lupus, scleroderma, type I diabetes, and MS [–
], there is need for further evaluation of the role of plant
proteins in the generation of cross-reactive antibodies against
human AQP, SB, MOG, and MBP and the consequent
development of RRMS and other neuroimmune disorders.
is may help in the development of dietary guidelines
for dietary modications for patients with neuroimmune
disorders.
Conflict of Interests
e authors declare that there is no conict of interests
regarding the publication of this paper.
Authors’ Contribution
Aristo Vojdani designed the study, performed some of the
experiments, and wrote the paper. Datis Kharrazian per-
formedsomeoftheimmunologicalassaysandhelpedindata
interpretation. Partha Sarathi Mukherjee did the statistical
analysis and helped in writing the paper. Joshua Berookhim
helped in the performance of the ELISA.
Acknowledgment
e authors would like to thank Joel Bautista for his work with
some of the gures and in typing and editing the paper.
References
[] A. Compston and A. Coles, “Multiple sclerosis,” e Lancet,vol.
, no. , pp. –, .
[] J. Nakahara, M. Maeda, S. Aiso, and N. Suzuki, “Current
concepts in multiple sclerosis: autoimmunity versus oligoden-
drogliopathy,” Clinical Reviews in Allergy &Immunology,vol.,
no. , pp. –, .
[] F. D. Lublin and S. C. Reingold, “Dening the clinical course of
multiple sclerosis: results of an international survey,” Neurology,
vol. , no. , pp. –, .
[] F. D. Lublin, S. C. Reingold, J. A. Cohen et al., “Dening
the clinical course of multiple sclerosis: the  revisions,”
Neurology,vol.,no.,pp.–,.
[] T. Misaka, K. Abe, K. Iwabuchi et al., “A water channel closely
related to rat brain aquaporin  is expressed in acid- and

Autoimmune Diseases 
pepsinogen-secretory cells of human stomach,” FEBS Letters,
vol. , no. , pp. –, .
[] E. A. Nagelhus, T. M. Mathiisen, and O. P. Ottersen,
“Aquaporin- in the central ner vous system: cellular and subcel-
lular distribution and coexpression with KIR.,” Neuroscience,
vol.,no.,pp.–,.
[]O.Nesic,J.Lee,Z.Yeetal.,“Acuteandchronicchangesin
aquaporin  expression aer spinal cord injury,” Neuroscience,
vol. , no. , pp. –, .
[] C. Iacovetta, E. Rudlo, and R. Kirby, “e role of aquaporin
 in the brain,” Ve t e r i nary Clin i cal Pa t holog y,vol.,no.,pp.
–, .
[] P. Fallier-Becker, J. P. Vollmer, H.-C. Bauer, S. Noell, H.
Wolburg, and A. F. Mack, “Onset of aquaporin- expression in
the developing mouse brain,” International Journal of Develop-
mental Neuroscience, vol. , pp. –, .
[] S.-H. Kim, W. Kim, X. F. Li, I.-J. Jung, and H. J. Kim, “Clinical
spectrum of CNS aquaporin- autoimmunity,” Neurology,vol.
,no.,pp.–,.
[] S. Jarius, F. Paul, D. Franciotta et al., “Mechanisms of disease:
aquaporin- antibodies in neuromyelitis optica,” Nature Clini-
cal Practice Neurology,vol.,no.,pp.–,.
[] D. J. Graber, M. Levy, D. Kerr, and W. F. Wade, “Neuromyelitis
optica pathogenesis and aquaporin ,” Journal of Neuroinam-
mation,vol.,no.,pp.–,.
[] V.A.Lennon,T.J.Kryzer,S.J.Pittock,A.S.Verkman,andS.
R. Hinson, “IgG marker of optic-spinal multiple sclerosis binds
to the aquaporin- water channel,” e Journal of Experimental
Medicine,vol.,no.,pp.–,.
[] S. Jarius and B. Wildemann, “AQP antibodies in neuromyelitis
optica: diagnostic and pathogenetic relevance,” Nature Reviews
Neurology,vol.,no.,pp.–,.
[]W.Kim,M.S.Park,S.H.Leeetal.,“Characteristicbrain
magnetic resonance imaging abnormalities in central nervous
system aquaporin- autoimmunity,” Multiple Sclerosis,vol.,
no. , pp. –, .
[] A. Petzold, S. Pittock, V. Lennon, C. Maggiore, B. G.
Weinshenker, and G. T. Plant, “Neuromyelitis optica-IgG
(aquaporin-) autoantibodies in immune mediated optic neu-
ritis,” Journal of Neurology, Neurosurgery &Psychiatry,vol.,
no. , pp. –, .
[] I. D ujmovic, S. Mader, K. Schanda et al., “Temporal dynamics of
cerebrospinal uid anti-aquaporin- antibodies in patients with
neuromyelitis optica spectrum disorders,” Journal of Neuroim-
munology,vol.,no.-,pp.–,.
[] B. G. Weinshenker, “Neuromyelitis optica is distinct from
multiple sclerosis,” Archives of Neurology,vol.,no.,pp.–
, .
[] S. Lalan, M. Khan, B. Schlakman, A. Penman, J. Gatlin, and
R. Herndon, “Dierentiation of neuromyelitis optica from
multiple sclerosis on spinal magnetic resonance imaging,”
International Journal of MS Care,vol.,no.,pp.–,
.
[] P.-W.Phuan,J.Ratelade,A.Rossi,L.Tradtrantip,andA.S.Verk-
man, “Complement-dependent cytotoxicity in neuromyelitis
optica requires aquaporin- protein assembly in orthogonal
arrays,” eJournalofBiologicalChemistry,vol.,no.,pp.
–, .
[] P. V. A. Lennon, D. M. Wingerchuk, T. J. Kryzer et al., “A serum
autoantibody marker of neuromyelitis optica: distinction from
multiple sclerosis,” e Lancet,vol.,no.,pp.–,
.
[] S. Viswanathan, M. Arip, N. Mustafa et al., “e frequency of
anti-aquaporin- Ig G antibody in neuromyelitis optica and
its spectrum disorders at a single tertiary referral center in
Malaysia,” Multiple Sclerosis International,vol.,ArticleID
,  pages, .
[] T. Takahashi, K. Fujihara, I. Nakashima et al., “Anti-aquaporin-
 antibody is involved in the pathogenesis of NMO: a study on
antibody titre,” Brain,vol.,no.,pp.–,.
[] M. Bradl and H. Lassmann, “Anti-aquaporin- antibodies in
neuromyelitis optica: how to prove their pathogenetic rele-
vance?” International MS Journal,vol.,no.,pp.–,.
[] C. Maurel, “Plant aquaporins: novel functions and regulation
properties,” FEBS Letters, vol. , no. , pp. –, .
[] R. M. Dean, R. L. Rivers, M. L. Zeidel, and D. M. Roberts,
“Purication and functional reconstitution of soybean nodulin
. An aquaporin with water and glycerol transport properties,”
Biochemistry,vol.,no.,pp.–,.
[] P. Fleurat-Lessard, P. Michonneau, M. Maeshima, J.-J. Drevon,
and R. Serraj, “e distribution of aquaporin subtypes (PIP,
PIP and 𝛾-TIP) is tissue dependent in soybean (Glycine max)
root nodules,” Annals of Botany,vol.,no.,pp.–,
.
[] R. A. Vaishnav, R. Liu, J. Chapman et al., “Aquaporin 
molecular mimicry and implications for neuromyelitis optica,”
Journal of Neuroimmunology,vol.,no.-,pp.–,.
[] R. Egg, M. Reindl, F. Deisenhammer, C. Linington, and T.
Berger, “Anti-MOG and anti-MBP antibody subclasses in mul-
tiple sclerosis,” Multiple Sclerosis,vol.,no.,pp.–,.
[] T. Berger, P. Rubner, F. Schautzer et al., “Antimyelin antibodies
as a predictor of clinically denite multiple sclerosis aer a rst
demyelinating event,” e New England Journal of Medicine,vol.
, no. , pp. –, .
[] A. Vojdani, E. Vojdani, and E. Cooper, “Antibodies to myelin
basic protein, myelin oligodendrocytes peptides, 𝛼-𝛽-crystallin,
lymphocyte activation and cytokine production in patients with
multiple sclerosis,” Journal of Internal Medicine,vol.,no.,
pp. –, .
[] M. Reindl, C. Linington, U. Brehm et al., “Antibodies against
the myelin oligodendrocyte glycoprotein and the myelin basic
protein in multiple sclerosis and other neurological diseases: a
comparative study,” Brain,vol.,no.,pp.–,.
[] G. Isoardo, B. Ferrero, P. Barbero et al., “Anti-GM and anti-
sulfatide antibodies in polyneuropathies: threshold titers and
accuracy,” Acta Neurologica Scandinavica,vol.,no.,pp.
–, .
[] A. M. Connolly, A. Pestronk, J. L. Trotter, E. L. Feldman, D. R.
Cornblath,andR.K.Olney,“High-titerselectiveserumanti-
𝛽-tubulin antibodies in chronic inammatory demyelinating
polyneuropathy,” Neurology,vol.,no.,pp.–,.
[] H.Choi,V.Puvenna,C.Brennan,M.Beukemann,D.Janigro,
and P. Mazzone, “SB as a serum marker for early detection
of brain metastasis in lung cancer,” Chest,vol.,p.A,.
[] N. Marchi, J. J. Bazarian, V. Puvenna et al., “Consequences
of repeated blood-brain barrier disruption in football players,”
PLoS ONE,vol.,no.,ArticleIDe,.
[] H. Alexopoulos, E. I. Kampylaa, I. Chatzi et al., “Reactivity
to AQP epitopes in relapsing-remitting multiple sclerosis,”
Journal of Neuroimmunology,vol.,no.-,pp.–,.
[] J.-I. Kira, “Autoimmunity in neuromyelitis optica and opti-
cospinal multiple sclerosis: astrocytopathy as a common
denominator in demyelinating disorders,” JournaloftheNeu-
rological Sciences, vol. , no. -, pp. –, .

 Autoimmune Diseases
[] R. Iorio, J. P. Fryer, S. R. Hinson et al., “Astrocytic autoanti-
body of neuromyelitis optica (NMO-IgG) binds to aquaporin-
 extracellular loops, monomers, tetramers and high order
arrays,” Journal of Autoimmunity,vol.,no.,pp.–,.
[] A. Vojdani and I. Tarash, “Cross-reaction between gliadin and
dierent food and tissue antigens,” Food and Nutrition Sciences,
vol. , no. , pp. –, .
[] P. F. Agris, R. Parks, L. Bowman, R. H. Guenther, S. A. Kovacs,
and S. Pelsue, “Plant DMA topoisomerase I is recognized and
inhibited by human Scl- sera autoantibodies,” Experimental
Cell Research,vol.,no.,pp.–,.
[] R.Bullard-Dillard,J.Chen,S.Pelsue,V.Dao,andP.F.Agris,
“Anti-sm autoantibodies of systemic lupus erythematosus cross
react with dietary plant proteins,” Immunological Investigations,
vol.,no.,pp.–,.
[] A. Steerl, A. Schubart, M. Storch et al., “Butyrophilin, a milk
protein, modulates the encephalitogenic T cell response to
myelin oligodendrocyte glycoprotein in experimental autoim-
mune Encephalomyelitis,” Journal of Immunology,vol.,no.
, pp. –, .
[] M. Knip, S. M. Virtanen, K. Sepp¨
a et al., “Dietary intervention
in infancy and later signs of beta-cell autoimmunity,” e New
England Journal of Medicine,vol.,no.,pp.–,
.