Department of Food Science and Human Nutrition, Washington State ... Supported by a grant from the Washington State Attorney General's Office; soymilk. 15.
1 1
Soy isoflavones modulate immune function in healthy postmenopausal women
2 3
Tracy A. Ryan-Borchers, Jean Soon Park, Boon P. Chew, Michelle K. McGuire, Lisa R.
4
Fournier, Kathy A. Beerman
5 6 7
From the Department of Food Science and Human Nutrition (TAR-B, JSP, BPC, MKM,
8
KAB), and Department of Psychology (LRF), Washington State University, Pullman,
9
WA 99164-6376.
10 11
Address correspondence about the manuscript and reprint requests to B. P. Chew,
12
Department of Food Science and Human Nutrition, Washington State University,
13
Pullman, WA 99164-6376.
14 15
Supported by a grant from the Washington State Attorney General’s Office; soymilk
16
provided by White Wave, Inc. (Boulder, CO); and isoflavone (Novasoy®) and placebo
17
supplements provided by the Archer Daniels Midland Co. (Decatur, IL).
18 19 20
Running head: Isoflavones modulate postmenopausal immunity
2 1
ABSTRACT
2
Background: After menopause, the immune system may be compromised due to effects
3
of aging and diminishing concentrations of estrogen, an immune-modulating hormone.
4
Isoflavones, plant-derived compounds with estrogenic and antioxidant properties, may
5
offer immunological benefits to women in this stage of life.
6
Objective: The objective of this study was to evaluate the effects of soy isoflavones,
7
both soymilk and supplement form, on markers of immunity and oxidative stress in
8
postmenopausal women.
9
Design: Postmenopausal women aged 50-65 y (n = 52) enrolled in this 16 wk double-
10
blind, placebo-controlled trial were randomized to one of three experimental groups: 1)
11
Control – 706 mL/d cow’s milk plus a placebo supplement; 2) Soymilk – 71.6 mg
12
isoflavones derived from 706 mL/d soymilk plus a placebo supplement; and 3)
13
Supplement – 70 mg isoflavones in a supplement plus 706 mL/d cow’s milk. Plasma and
14
24-h urine samples were obtained at baseline and at 16 wk. Immune variables included
15
lymphocyte subsets, cytokine production, as well as markers of inflammation and
16
oxidative damage.
17
Results: Baseline immune variables did not differ among groups. Both isoflavone
18
interventions increased (p < 0.05) B cell populations and concentrations of plasma
19
interferon-gamma (IFN-γ) and decreased (p < 0.05) concentrations of 8-hydroxy-2-
20
deoxy-guanosine (8-OHdg), an oxidative marker of DNA damage. Isoflavone treatments
21
did not influence concentrations of urinary 8-isoprostane (8-iso), plasma interleukin-2
22
(IL-2), tumor necrosis factor-alpha (TNF-α) or C-reactive protein (CRP).
3 1
Conclusions: Soymilk and supplemental isoflavones modulate B cell populations and
2
IFN-γ concentrations, and appear protective against DNA damage in postmenopausal
3
women.
4 5
KEY WORDS
6
soymilk
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Isoflavones, immune function, postmenopausal women, DNA damage,
4 1
INTRODUCTION
2
Isoflavones are a subclass of flavonoids, a broad group of polyphenolic compounds
3
widely distributed in foods of plant origin. Biological effects attributed to flavonoids
4
include anti-estrogenic and pro-estrogenic, as well as antioxidant and antiproliferative
5
actions (1). The three main isoflavone derivatives are daidzein, genistein and glycitein,
6
which are found predominantly in soybeans and other legumes (2). During digestive and
7
absorptive processes, isoflavones often undergo further metabolic transformations (3).
8
For example, daidzein can be converted to equol by intestinal microflora. The biological
9
activities of this mammalian derived isoflavone are thought to be stronger than its parent
10 11
compound, daidzein (4). Isoflavones have been classically defined as phytoestrogens, compounds that exert
12
estrogenic effects. The presence of the phenolic ring enables isoflavones to bind to both
13
types of estrogen receptors, Erα and Erβ (5). Unlike endogenous estrogen, however,
14
isoflavones bind with a much higher affinity to Erβ, which suggests they may be more
15
accurately defined as selective estrogen receptor modulators (SERMs) capable of both
16
pro-estrogenic and anti-estrogenic effects (6). This hormonal mechanism explains, in
17
part, how isoflavones protect against age-related diseases, including cardiovascular
18
disease (7, 8), osteoporosis (9), and cancers of the breast (10) and prostate (11, 12).
19
Antioxidant and antiproliferative properties of isoflavones offer additional, important
20
mechanisms for their protection against many prevalent chronic diseases (13, 14).
21
Cellular damage resulting from oxidative stress is believed to be a major contributor to
22
the etiology of cardiovascular disease through the oxidation of LDL, and cancer by
23
causing DNA strand breaks which may lead to mutations (15-17). In humans,
5 1
isoflavones have been found to prevent LDL oxidation in vivo (16), as well as inhibit
2
DNA damage in vitro (17).
3
The immune system encompasses an array of defenses, which help guard against the
4
development of age-related diseases. However, like other bodily systems, its functions
5
can be adversely affected by factors such as oxidative damage and hormonal changes
6
(18). Therefore, the immune system may benefit from the various biological properties
7
of isoflavones. Enhanced immune responses have been found in animals fed genistein,
8
the isoflavone most abundant in soy foods (19, 20). In humans, consumption of
9
isoflavone-containing soy foods modulated cytokine production (21), while daidzein and
10
genistein glucuronides enhanced the activity of natural killer (NK) cells in vitro (22).
11
Postmenopausal women, in particular, are susceptible to chronic diseases associated
12
with aging, as well as major shifts in hormonal status. It is possible that isoflavones may
13
improve immune function in this population. Also, it is important to determine whether
14
dietary or supplement forms of isoflavones affect immunological variables similarly.
15
Therefore, the purpose of this study was to evaluate the effects of soy isoflavones,
16
obtained from either soymilk or supplement, on immune and oxidative markers in
17
postmenopausal women.
18 19 20 21 22 23
6 1
SUBJECTS AND METHODS
2
Subjects and study design
3
This study was conducted as part of a larger research project designed to investigate the
4
effects of soy isoflavones on cognitive function. Thus, the sample sizes were determined
5
based on expected changes and variability in cognitive, rather than immunologic
6
variables. Healthy postmenopausal women between 50 and 65 y of age were recruited
7
regionally from the states of Idaho and Washington. Potential subjects (n = 300) were
8
screened by telephone to determine study eligibility. Women free of major health
9
conditions, with a body mass index (BMI) between 18 and 40 kg/m2, and who had not
10
menstruated for at least one year were eligible to participate in the study. Additional
11
exclusion criteria included legume allergies, smoking, kidney stones and antibiotic
12
therapy within the past six months.
13
Telephone screening resulted in 150 women who were both eligible and interested in
14
enrollment; a total of 117 of these women completed the enrollment process. Of these
15
women, 5 dropped out of the study, all for personal reasons. After exclusion of women
16
using hormone replacement therapy (HRT), the first 52 women completing the cognitive
17
study composed the subset for this secondary investigation of immune and antioxidant
18
variables. All study procedures were approved by the Institutional Review Board (IRB)
19
of Washington State University.
20
Prior to intervention, subjects providing informed consent adhered to a 4 wk adjustment
21
diet that minimized intake of foods containing isoflavones. Following this washout
22
period, the 16 wk intervention period began. Subjects were enrolled in groups of 20 to 28
23
women; these groups were then considered to be treatment blocks. Within each block,
7 1
subjects were randomly assigned to receive: 1) cow’s milk and placebo supplement
2
(Control group); 2) soymilk (White Wave, Inc.; Boulder, CO) and placebo supplement
3
(Soymilk group, 71.6 + 3.1 mg isoflavones/d based on 30 samples); or 3) cow’s milk and
4
isoflavone tablets (Supplement group, 70 mg isoflavones/d). Randomization was
5
performed using random number generation. Researchers and subjects remained blinded
6
to group assignment throughout the study.
7
Nutrient composition and caloric value of the two types of milk was nearly identical
8
(Table 1). The isoflavone content and composition of the supplements (30 mg daidzein,
9
33 mg genistein and 7 mg glycetin) were formulated to match those of the soymilk and
10
values are verified by Archer Daniels Midland Co. (Novasoy®; Decatur, IL). Placebos
11
were composed of maltodextrin with 10% caramel color. The isoflavone content of the
12
soymilk was determined by HPLC (P. Murphy, Iowa State University) and is presented in
13
Table 1. Milk products were delivered to each subject weekly.
14
Subjects were instructed to consume a total of 706 mL milk/d (soymilk or cow’s milk)
15
for a total of 16 wk. Milk was consumed in the morning (353 mL) and in the evening
16
(353 mL). The cow’s milk was treated with flavoring and color agents to resemble the
17
flavor and appearance of soymilk. Supplements (isoflavone or placebo tablet) were taken
18
with the milk, twice per day, throughout the duration of the intervention period. The
19
research team’s dietitian provided individual instruction, along with written guidelines, to
20
the women regarding replacement of usual dairy or other similar food items with the
21
study milk to allow for consistent macronutrient intake and prevention of weight gain
22
during the trial. Compliance was assessed via personal communication, dietary records
23
and appearance of isoflavone metabolites in plasma and urine.
8 1
Blood and urine samples were obtained at baseline (0 wk) and post-intervention
2
(16 wk). Blood was drawn into evacuated collection tubes (10 mL) coated with EDTA.
3
An aliquot of whole blood from each subject was used for flow cytometry analysis
4
immediately after the collection. The remainder was centrifuged for 30 min and aliquots
5
of the plasma were frozen at -80 °C until analyzed for interleukin-2 (IL-2), interferon-
6
gamma (IFN-γ) and tumor necrosis factor-alpha (TNF-α); the acute-phase protein, C-
7
reactive protein (CRP), the oxidative marker, 8-hydroxy-2-deoxy-guanosine (8-OHdg)
8
and isoflavone concentrations. Two, 24-h urine collections (0 wk and 16 wk) were
9
obtained from participants. Total urine volume was recorded, and aliquots frozen at
10
-20 °C until analyzed for the oxidative marker, 8-isoprostane (8-iso), as well as urinary
11
isoflavone concentrations.
12
Analytical procedures
13
Isoflavone concentrations
14
The isoflavone concentrations in plasma and urine were measured by HPLC as
15
previously described (23, 24). Briefly, β-glucuronidase-sulfatase H-2 (EC 3.2.1.31, from
16
Helix pomatia, Sigma Chemical, St. Louis, MO) was added to plasma (1 mL) and urine
17
(5 mL), and the mixtures were incubated at 37 °C for 20 h to release isoflavone
18
aglycones. Samples were then loaded onto Extrelut QE columns (EM Science,
19
Gibbstown, NJ) and extracted with 8 mL (plasma) or 12 mL (urine) of ethyl acetate. The
20
eluent was collected and dried under nitrogen gas. Extracted isoflavones were dissolved
21
in 80% methanol in water for HPLC analysis. An (Waters, Milford, MA) HPLC system
22
equipped with a photodiode array detector (PDA 996) was used to distinguish the
23
ultraviolet spectra of the specific isoflavone compounds: daidzein, equol and genistein.
9 1
Isoflavones were separated in a µ-Bondapak C18 reverse-phase column (Waters, Milford,
2
MA; 3.9 mm x 30 cm). The samples were eluted using a linear gradient of 40% to 70%
3
methanol in 30 min at a flow rate of 1 mL/min. Retention times for the isoflavones were
4
as follows: daidzein at 14.9 min, equol at 16.8 min, and genistein at 19.5 min. Daidzein,
5
equol and genistein were detected at 248, 280 and 260 nm, respectively. The identity of
6
each isoflavone was confirmed by comparing its absorption spectrum with that of a
7
corresponding standard. Plasma and urine isoflavone concentrations were calculated
8
based on standard curves constructed for daidzein (Valeant Pharmaceuticals, USA),
9
genistein (Calbiochem, USA) and equol (Sigma-Aldrich, USA).
10
Lymphocyte subsets
11
Specific populations of lymphocytes were measured using dual color flow cytometric
12
analysis (FACScalibur equipped with Cell Quest software, BD Biosciences, San Diego,
13
CA): Total T cells (CD3+CD19-), T cytotoxic cells (Tc; CD3+CD8+), T helper cells (Th;
14
CD3+CD4+), Th/Tc ratio (CD3+CD4+/CD3+CD8+), B cells (CD3-CD19+) and NK cells
15
(CD3-/CD16+56+). Whole blood aliquots obtained from subjects were treated with a lysis
16
buffer to avoid contamination by erythrocytes (25). After centrifugation, the resultant
17
cell pellet was washed, and cells were labeled with monoclonal antibodies conjugated
18
with fluorescein isothiocyanate (FITC) or phycoerythrin (PE): anti-CD3 conjugated to
19
FITC, and anti-CD8, anti-CD4 or anti-CD19 conjugated to PE (Caltag Laboratories,
20
Burlingame, CA). Cell suspensions were fixed with paraformaldehyde and analyzed
21
using a leucogate, or automatically established lymphocyte analysis gate. The leucogate
22
helps define and distinguish the lymphocytes from other blood cell types. A total of 2000
23
events were counted for each sample.
10 1
Cytokine production
2
Plasma samples were analyzed by enzyme-linked immunosorbent assays (ELISA) for
3
IFN-γ (Human IFN-γ BD OptEIA™ ELISA Set, BD Biosciences, San Diego, CA), TNF-
4
α (BD OptEIA™ Set Human TNF, BD Biosciences), and IL-2 (BD OptEIA™ Set
5
Human IL-2, BD Biosciences). In these assays, wells were coated with capture antibody,
6
diluted in coating buffer and incubated overnight at 4 °C. Samples were pipetted into
7
appropriate wells and incubated for 2 h at room temperature. Detection antibody with
8
avidin-horseradish peroxidase was added to plates for color development and incubated
9
for 30 min. A stop solution was then added and absorbance read on a plate reader (Bio-
10
Tek Instruments, Inc., Winooski, VT) at 450 nm.
11
C-reactive protein
12
Plasma samples were analyzed for CRP by ELISA (Human CRP ELISA kit, Alpha
13
Diagnostic International, San Antonio, TX). The human CRP kit is a sandwich-type
14
ELISA based on the CRP in samples binding simultaneously to two antibodies, one fixed
15
to the microtiter well plates and the other conjugated to horseradish peroxidase. After
16
color development, absorbance was read at 450 nm. Concentration of CRP in samples
17
was based on the standard curve.
18
Lipid peroxidation
19
8-Isoprostane was measured in purified urine by an enzyme immunoassay (EIA)
20
method (8-Isoprostane EIA kit, Cayman Chemical Company, Ann Arbor, MI). This
21
assay is based on the competition between 8-iso and an 8-iso-acetylcholinesterase
22
(AChE) tracer. Concentration of tracer is held constant, while the amount of 8-iso in
23
urine samples varies. The 8-iso antibody binds to the IgG mouse monoclonal antibody
11 1
that has been previously attached to the wells. After washing, color is developed with
2
Ellman’s Reagent and absorbance measured at 410 nm.
3
DNA damage
4
Plasma concentrations of 8-OHdg were measured by ELISA (Bioxytech 8-OHdg-EIA™
5
kit, OXIS Health Products, Inc., Portland, OR). Samples and standards were added to
6
microtiter plate wells that had been precoated with 8-OHdg. The 8-OHdg present in both
7
the wells and samples competes for binding with the monoclonal antibody to 8-OHdg.
8
Plates were washed, leaving only the antibody that had bound to the 8-OHdg coating the
9
wells. Secondary antibody conjugated to horseradish peroxidase was then added and
10
bound to the monoclonal antibody remaining on the plate. Chromogen was added
11
following another wash step, color was developed, and absorbance measured at 450 nm.
12
Statistical analysis
13
Data were analyzed by ANOVA using the General Linear Model procedure of SAS
14
statistical software (version 8; 26). The initial statistical model included the effects of
15
block, treatment, period, and the interaction of treatment by period. Upon preliminary
16
statistical analyses, it was determined that there were no effects of block. Thus, this
17
variable was removed from the statistical models prior to further analyses. Treatment
18
means were compared using Tukey test; group BMI means (0 wk and 16 wk) were
19
evaluated using paired t tests, and equol production among the groups was analyzed by
20
Chi-square analysis. Statistical significance was set at p < 0.05.
21 22 23
12 1
RESULTS
2
Subjects
3
Demographic characteristics are presented in Table 2. A total of 52 (95%)
4
postmenopausal women completed the study. Mean age of study participants did not
5
differ by group (Table 2). There was no significant group difference in body mass index
6
(BMI) at baseline or post-intervention. Ninety-four percent of the study participants were
7
Caucasian and the majority (88%) received education beyond the high school level.
8
Study treatments, including cow’s milk, soymilk and supplemental tablets, were tolerated
9
well overall. Compliance with study protocol was very good as indicated by verbal
10
reports and written dietary records from the women, as well as by the absence (control) or
11
presence (isoflavone interventions) of isoflavone metabolites in plasma and urine samples
12
of the participants.
13
Isoflavone concentrations
14
At baseline, plasma and urinary isoflavone concentrations were non-detectable among
15
the women in all experimental groups, confirming dietary compliance during the
16
adjustment period. Following the 16 wk intervention, plasma concentrations of genistein
17
and equol were significantly higher (p < 0.05) in both isoflavone treatment groups, but
18
remained non-detectable in the control group (Figure 1). Plasma daidzein was detectable
19
in all women and was significantly higher (p < 0.05) in the supplement group compared
20
to control. Urinary concentrations of daidzein, genistein and equol were higher (p