Isoliquiritigenin Decreases Bone Resorption and Osteoclast Differentiation

Abstract

Scope:

We performed a dose-ranging study using young estrogen-depleted rats to determine whether dietary isoliquiritigenin (ILQ) alters bone metabolism and if the effects are associated with estrogen receptor signaling.

Methods and Results:

Six-week-old rats (ovariectomized at 4 weeks of age) were fed diets containing 0, 100, 250 or 750 ppm ILQ (n=5/treatment) for 7 days. Gene expression in femur and uterus, blood markers of bone turnover, body composition, and uterine weight and epithelial cell height were determined. Because ILQ lowered bone resorption, the effect of ILQ on in vitro differentiation of osteoclasts from bone marrow of mice was assessed. Treatment resulted in a dose-dependent increase in serum ILQ but no change in serum osteocalcin, a marker of global bone formation. Contrastingly, ILQ administration resulted in reduced serum CTX-1, a marker of global bone resorption, and reduced tartrate resistant acid phosphatase expression in osteoclast culture. ILQ treatment and endogenous estrogen production had limited overlap on gene expression in femur and uterus. However, uterine epithelial cell hyperplasia was observed in 2 of 5 animals treated with 750 ppm.

Conclusions:

In conclusion, dietary ILQ reduced bone resorption in vivo and osteoclast differentiation in vitro, by mechanisms likely differing from actions of ovarian hormones.

Graphical Abstract

Licorice root is consumed by women based on the belief that it attenuates common menopausal symptoms, including bone loss. Administration of dietary isoliquiritigenin (ILQ, compound found in licorice root that contributes to its bioactivity) reduced bone resorption in growing female rats. However, enlargement of epithelial cells in the uterus was observed in some of the animals treated with the highest dose of ILQ. Long duration studies are warranted to establish whether ILQ can attenuate postmenopausal bone loss without detrimental side effects.

Introduction

Licorice root, either in whole root powder or more concentrated extract form, is often used as an ingredient in dietary supplements, homeopathic remedies, and candies. Licorice root is consumed by older women based, in part, on the belief that, as a natural botanical estrogen, it attenuates common menopausal symptoms, including weight gain and bone loss ^[1]. To date, clinical trials designed to evaluate the efficacy of botanical estrogens, including licorice root, to reduce the rate of osteoporotic fractures have not been performed. Therefore, rigorous preclinical studies evaluating candidate botanicals are required to support the rational design of such trials.

We recently reported that dietary licorice root supplementation reduced diet-induced weight gain, lipid deposition, and hepatic steatosis in growing ovariectomized (ovx) mice without stimulating reproductive tissues and mammary gland ^[1,a]. No changes in femur mass or density were detected by densitometry. In contrast, estrogen treatment increased femur mineral content and density in the ovx mice fed both normal and high fat diets, likely due to an osteosclerotic response to exogenous estrogens apparently unique to mice ^{[1b, 2]}. While densitometry is valuable for assessing effects of treatment on whole bone in rodents, it is relatively insensitive to changes in cancellous bone, an important target of estrogen and other gonadal hormones ^[3]. Other studies investigating the effects of licorice root or constituents of licorice root on the skeleton have led to mixed results ^{[1,a, 4]}.

While there is merit in studying whole foods, standardization of mixtures of chemicals with differing biological activities is challenging ^[5]. As such, specific compounds believed to be responsible for bioactivity are often assessed. Isoliquiritigenin is a phenolic compound found in licorice root that contributes to its activity ^[6]. Notably, ILQ mimicked some of the actions of licorice root powder and licorice root extract on weight gain and fat deposition in liver in our ovx mouse studies ^[1,a], suggesting that it is responsible, at least in part, for the reduced fat accrual. The effects of ILQ on the skeleton are poorly characterized, but there is evidence, primarily in vitro, that ILQ suppresses osteoclastogenesis ^{[4d, 7]}.

It is likely that bioactive levels of ILQ in serum are achievable with a dietary intervention ^[8]. Thus, dietary ILQ should be evaluated for potential beneficial actions on bone. The primary goal of this dose range study was to establish dose response effects of dietary ILQ on serum levels of ILQ and bioactive metabolites, serum biochemical markers of bone formation and resorption, and differential expression of a panel of genes related to bone metabolism. While ILQ binds to estrogen receptor (ER) αand β in vitro, it has ER-independent as well as ER-dependent actions ^[9]. Therefore, an additional goal of the study was to determine whether the actions of ILQ delivered via the diet are similar to those mediated via endogenous estrogens. As a well-established estrogen target tissue, we measured uterine weight, histology, and gene expression. ILQ was administered to growing estrogen-depleted female rats. Rats were ovx at 4 weeks of age (preadolescent) to prevent normal puberty-related increases in ovarian secretion of estrogens. Thus, in contrast to adult women, the rats in this study have not been exposed to high levels of gonadal hormones and, as a consequence, are highly sensitive to exogenous estrogens, providing an excellent in vivo assay for evaluating the effects of putative estrogenic compounds ^[10]. Based on our finding that ILQ decreased a serum marker of bone resorption, we additionally performed an in vitro study with bone cells differentiated from mice to determine if ILQ has a direct effect on osteoclastogenesis.

Materials and Methods

Animals and study design

The animals were maintained in accordance with the NIH Guide for the Care and the Use of Laboratory Animals and the experimental protocol was approved by the Institutional Animal Care and Use Committee at Oregon State University – ACUP #4836, approved July 23, 2016.

Six-week-old intact (n=5) and ovx (n=20, ovx at 4 weeks of age) female Sprague Dawley rats were obtained from Envigo (Madison, WI) and housed in plastic shoebox cages (1 rat/cage) in a temperature- and humidity-controlled room with a 12/12 hour light/dark cycle. Upon arrival, the rats were acclimated for 3 days to a standard powdered AIN-93G purified diet; the diet contains 57.4% (w/w) carbohydrates and 17.7% (w/w) protein and is designed to provide all nutrients required for optimal rodent growth and health ^[11]. Following acclimation, the ovx rats were randomized by weight into one of 4 dietary treatments: powdered AIN-93G diet containing either 0, 100, 250 or 750 ppm ILQ, equivalent to 0, 100, 250, and 750 mg ILQ/kg AIN93G diet, respectively. At the start of treatment, there were no difference in body weight among groups. Food intake and body weight were measured daily. The rats were maintained on their respective diets for 7 days and sacrificed at 7 weeks of age.

Tissue collection

For tissue collection, rats were anesthetized with 2–3% isoflurane delivered in oxygen. Whole blood was collected from non-fasted rats by cardiac puncture. Serum was stored at −80°C for measurement of ILQ and liquiritigenin (LIQ) and markers of global bone turnover. Abdominal white adipose tissue was removed and weighed. Successful ovx (lack of ovarian tissue) was confirmed at necropsy. Uteri were excised, weighed, and cut in half along the sagittal plane. The right half of each uterus was stored in 10% formalin for histological evaluation while the left half was stored in RNA later for RNA extraction and gene expression analysis. Tibiae were excised, fixed overnight in 10% formalin, and stored in 70% ethanol for microcomputed tomography (μCT) and histomorphometric evaluation. Femora were removed, frozen in liquid nitrogen, and stored at −80°C for mRNA analysis. Total cellular RNA was isolated from the distal femur, which includes cortical bone, cancellous bone, growth plate, and bone marrow.

Blood concentrations of isoliquiritigenin and liquiritigenin

Total ILQ and LIQ (the major bioactive metabolite of ILQ) concentrations were evaluated in the serum. A validated LC/MS/MS method with internal standard quantification for ILQ and LIQ was used to quantify total conjugated and aglycone forms of ILQ and LIQ (limits of detection 0.0002 – 0.006 μM depending on volume analyzed), without and with enzymatic hydrolysis, respectively ^[8].

Blood concentrations of markers of bone turnover

Serum carboxyterminal cross-linking telopeptide of type 1 collagen (CTX-1, a marker of global bone resorption) was measured using rat CTX-1 ELISA kit (Novatein Bioscience, Woburn, MA). Serum osteocalcin (a marker of global bone formation) was measured using rat Gla-Osteocalcin High Sensitive EIA Kit (TaKaRa Bio Inc, San Jose, CA). Both assays were performed following the specific manufacturer’s protocol.

Microcomputed tomography

Microcomputed tomography (μCT) was used for 3-dimensional evaluation of cancellous and cortical bone in tibia. Tibiae were scanned in 70% ethanol at a voxel size of 12 × 12 × 12 μm (55 kVp, 145 μA, and 200 ms integration time) using a Scanco μCT40 scanner (Scanco Medical AG, Basserdorf, Switzerland). For evaluation, filtering parameters sigma and support were set to 0.8 and 1, respectively. Tibia length was measured in ScoutView as the distance between the intercondyloid eminence and the medial malleolus. Eighty-three slices (996 μm) of cancellous bone were analyzed in the tibial metaphysis (volume of interest, 9.4 ± 0.2 mm³) (Figure 1). A similar number of slices was analyzed for cortical bone in the tibial diaphysis (midshaft) (Figure 1). Cancellous bone measurements included cancellous bone volume fraction (bone volume/tissue volume, %), connectivity density (mm⁻³), trabecular thickness (μm), trabecular number (mm⁻¹), and trabecular separation (μm). Cortical bone measurements included cross-sectional volume (cortical bone volume + marrow volume, mm³), cortical bone volume (mm³), marrow volume (mm³), and cortical thickness (μm). Polar moment of inertia (I_polar, mm⁴) was calculated as a surrogate measure of bone strength in torsion.

Figure 1.

Regions of interest (shaded) with cancellous bone evaluated in the proximal tibial metaphysis and cortical bone evaluated in the tibial diaphysis.

Quantitative bone and uterine histomorphometry

Bone:

Specimens were prepared for histomorphometric evaluation as previously described ^[12]. In brief, proximal tibiae were dehydrated in a graded series of ethanol and xylene, and embedded undecalcified in modified methyl methacrylate. 4 μm thick longitudinal sections were cut using a vertical bed microtome (Leica 2065) and affixed to gel slides (pre-coated with 1% gelatin solution). One histological section/animal was stained for tartrate-resistant acid phosphatase (TRAP) and counterstained with toluidine blue (Sigma, St Louis, MO) and used for cell-based measurements in metaphysis and growth plate. Cell-based measurements in the tibial metaphysis included osteoblast perimeter (osteoblast perimeter/bone perimeter, %) and osteoclast perimeter (osteoclast perimeter/bone perimeter, %). Growth plate morphology was measured as described ^[13]; measurements included proliferative zone width (μm), hypertrophic zone width (μm), and growth plate width (μm).

Uterus:

Uteri were fixed in 10% formalin and embedded in paraffin. Tissue sections were prepared and stained with hematoxylin and eosin by the Oregon Veterinary Diagnostic Laboratory (Oregon State University). Epithelial cell height (μm) was measured as described ^[14].

Gene expression

Total RNA was isolated using TRIzol reagent according to the manufacturer’s protocol (Thermo Fisher, Waltham, MA). cDNA was prepared using SuperScript^® III First-Strand Synthesis SuperMix for qRT-PCR (Thermo Fisher). For rat femoral samples, the expression of genes related to bone metabolism were determined using RT2 Profiler™ Rat Osteoporosis PCR Array (PARN-170ZE) according to the manufacturer’s protocol (Qiagen, Germantown, MD). Gene expression was normalized to B2m and relative quantification was determined by ΔΔCt method using RT2 Profiler PCR Array Data Analysis software (Qiagen). A Venn diagram was generated using BioVenn ^[15]. For rat uterine samples, the expression of Esr1, Esr2, Igf1, and Pgr were determined using the following primers: Esr1-for (TCCGGCACATGAGTAACAAA), Esr1-rev (TGAAGACGATGAGCATCCAG), Esr2-for (AAAGTAGCCGGAAGCTGACA), Esr2-rev (CTGCTGCTGGGAGGAGATAC), Igf1-for (CCGGACCAGAGACCCTTTG), Igf1-rev (CCTGTGGGCTTGTTGAAGTAAAA), Pgr-for (GAGAGGCAGCTGCTTTCAGT), Pgr-rev (AAACACCATCAGGCTCATCC). For mouse bone marrow-derived osteoclast samples, the expression of Acp5 and Tnfrsf11a were determined by using the following primers: Acp5-for (GACAAGAGGTTCCAGGAGACC), Acp5-rev (GGGCTGGGGAAGTTCCAG), Tnfrsf11a-for (TGCCTACAGCATGGGCTTT), Tnfrsf11a-rev (AGAGATGAACGTGGAGTTACTGTTT). Uterus and osteoclast gene expression were normalized to Rn18S using the following primers: rat Rn18S-for (GGACCAGAGCGAAAGCATTTGC), rat Rn18S-rev (CGCCAGTCGGCATCGTTTATG), mouse Rn18S-for (CCGCAGCTAGGAATAATGGAAT), mouse Rn18S-rev (CGAACCTCCGACTTTCGTTCT). Relative gene expression was determined using SDS RQ Manager (Applied Biosystems, Carlsbad, CA).

Ex vivo differentiation of bone marrow-derived osteoclasts

To evaluate bioactivity of ILQ using a different model (in vitro) and a different species (mouse), bone marrow was collected from the tibia and femur of C57Bl/6 female mice (n=3). The marrow was pooled from the 3 mice to obtain sufficient numbers of cells for the cell culture experiments. The samples were made into a single cell suspension. Red blood cells were removed by incubating bone marrow cells with red cell lysis buffer (150 mM NH₄Cl, 1 mM KHCO₃, 0.1 mM EDTA, pH 7.2). After red blood cell removal, bone marrow cells were counted and cell density adjusted to 1×10⁶ cells/ml in osteoclast differentiation media (α-MEM culture media containing 10% fetal bovine serum, 50 ng/ml MCSF and 50 ng/ml RANKL). Bone marrow cells were seeded in replicate (n=5) in 6-well tissue culture plates, in the presence or absence of differing concentrations of ILQ (0, 0.2, 1, 5, 25 μM). ILQ stock solutions were prepared in DMSO. To control for potential vehicle effect, a vehicle control was included, where DMSO concentration matched that present in the highest ILQ dose (0.025% DMSO). On day 4, non-adherent cells were removed and media were replaced, and cells were cultured for an additional 3 days. A sample of cell culture media was collected on days 4 and 7 to determine the production of TRAP by bone marrow-derived osteoclasts. Levels of TRAP in culture supernatants were measured using a TRAP staining kit (Kamiya Biomedical, Tukwila, WA) per manufacturer’s protocol and read in a microplate reader at 540 nm. Osteoclasts were harvested for RNA isolation on day 7. Data represent cell culture ILQ treatment of 5 replicates.

Statistical analysis

One-way analysis of variance was used to compare group means (intact control, ovx control, ovx 100 ppm ILQ, ovx 250 ppm ILQ, and ovx 750 ppm ILQ for in vivo study; and vehicle control, 0.2 μM ILQ, 1 μM ILQ, 5 μM ILQ, and 25 μM ILQ for in vitro study). The nonparametric Kruskal-Wallis test was used for non-normal data for which there was no statistical evidence to support unequal variance. A linear model with distinct variance parameters for each group was used for heteroscedastic data. Pairwise comparisons were made using two-sample t-tests or the Wilcoxon-Mann-Whitney nonparametric test. Model fit was assessed using residual analysis (normal quantile plots, Anderson-Darling test for normality, plots of residuals versus fitted values, and Levene’s test for equal variance). The false discovery rate was capped at 5% to adjust for multiple comparisons ^[16]. Data analysis was conducted in R version 3.6.3.

Results

Effects of dietary ILQ supplementation on serum ILQ, food intake, body weight, and white adipose tissue weight

The effects of ovx and dietary treatment with ILQ on serum ILQ and LIQ concentrations and on food intake, terminal body weight, body weight gain and abdominal white adipose tissue weight are shown in Figure 2. ILQ and LIQ were not detected in serum of untreated intact or ovx rats (Figures 2A and B, respectively). Dietary supplementation with ILQ at 100, 250 and 750 ppm resulted in serum total ILQ levels of 0.588 μM, 0.998 μM, and 2.363 μM, respectively and serum total LIQ levels of 0.052 μM, 0.083 μM, and 0.198 μM, respectively. Ovx resulted in a trend (P = 0.09) for higher food consumption (Figure 2C), had no effect on terminal body weight (Figure 2D), increased percent body weight gain (Figure 2E), and had no effect on abdominal white adipose tissue weight (Figure 2F). ILQ treatment had no significant effect on food consumption, terminal body weight, or abdominal white adipose tissue weight. Body weight gain was higher in in ovx rats and ovx rats treated with 100 and 250 ppm ILQ compared to intact controls; there were no differences in weight gain among the ovx groups.

Figure 2.

Effects of ovariectomy (ovx) and treatment with isolquiritigenin (ILQ), administered in diet at 100, 250, or 750 ppm for 7 days, on A) serum ILQ levels, B) serum LIQ levels, C) food intake, D) terminal body weight, E) percent body weight change, and F) abdominal white adipose tissue weight, in rapidly growing ovx rats. N = 5/group.

^aDifferent from intact control, P ≤ 0.05, ^a* Different from intact control, P ≤ 0.1.

^cDifferent from ovx + 100 ppm ILQ, P ≤ 0.05.

^dDifferent from ovx + 250 ppm ILQ, P ≤ 0.05.

Effects of dietary ILQ supplementation on bone and serum markers of bone turnover

The effects of ovx and dietary ILQ treatment on cancellous bone microarchitecture, cellular indices of bone turnover, and growth plate morphology in the proximal tibia metaphysis are shown in Figure 3. Ovx resulted in lower cancellous bone volume fraction (Figure 3A) and connectivity density (Figure 3B) but had no effect on other indices of cancellous bone architecture [trabecular thickness (Figure 3C), trabecular number (Figure 3D), and trabecular separation (Figure 3E)], cellular indices of cancellous bone turnover [osteoblast perimeter (Figure 3F) and osteoclast perimeter (Figure 3G)] or growth plate morphology [proliferative zone width (Figure 3H), hypertrophic zone width (Figure 3I), and growth plate width (Figure 3J). ILQ supplementation had no notable or consistent impact on either cancellous bone architecture, cellular indices of bone turnover, or growth plate.

Figure 3.

Effects of ovariectomy (ovx) and treatment with isolquiritigenin (ILQ), administered in diet at 100, 250, or 750 ppm for 7 days, on A) cancellous bone volume fraction, B) connectivity density, C) trabecular thickness, D) trabecular number, E) trabecular separation, F) osteoblast perimeter, G) osteoclast perimeter, H) proliferative zone width, I) hypertrophic zone width, and J) growth plate width in the proximal tibial metaphysis, in rapidly growing ovx rats. N = 5/group.

^aDifferent from intact control, P ≤ 0.05, ^a*Different from intact control, P ≤ 0.1.

^bDifferent from ovx control, P ≤ 0.05, ^b*Different from ovx control, P ≤ 0.1.

^c*Different from ovx + 100 ppm ILQ, P ≤ 0.1.

^d*Different from ovx + 250 ppm ILQ, P ≤ 0.1.

The effects of ovx and dietary ILQ treatment on tibia length and cortical bone microarchitecture are shown in Figure 4. Ovx resulted in longer tibiae (Figure 4A) but had no effect on cross-sectional volume (Figure 4B), cortical volume (Figure 4C), marrow volume (Figure 4D), cortical thickness (Figure 4E), or polar moment of inertia (Figure 4F). ILQ treatment had no effect on any of the cortical endpoints evaluated.

Figure 4.

Effects of ovariectomy (ovx) and treatment with isolquiritigenin (ILQ), administered in diet at 100, 250, or 750 ppm for 7 days, on A) tibia length and B) cross-sectional volume, C) cortical volume, D) marrow volume, E) cortical thickness, and F) polar moment of inertia in the tibial diaphysis, in rapidly growing ovx rats. N = 5/group.

^aDifferent from intact control, P ≤ 0.05, ^a*Different from intact control, P ≤ 0.1.

The effects of ovx and dietary ILQ treatment on serum CTX, a marker of global bone resorption, and serum osteocalcin, a marker of global bone formation, are shown in Figure 5. Ovx had no significant effect on serum CTX (Figure 5A) or osteocalcin (Figure 5B). Dietary supplementation with ILQ resulted in lower serum CTX in all supplemented groups in comparison to intact and ovx controls. In contrast, ILQ had no significant effect on serum osteocalcin.

Figure 5.

Effects of ovariectomy (ovx) and treatment with isolquiritigenin (ILQ), administered in diet at 100, 250, or 750 ppm for 7 days, on A) CTX and B) osteocalcin in serum, in rapidly growing ovx rats. N = 5/group.

^aDifferent from intact control, P ≤ 0.05.

^bDifferent from ovx control, P ≤ 0.05.

^dDifferent from ovx + 250 ppm ILQ, P ≤ 0.05.

The effects of ovx and dietary ILQ treatment on expression of genes related to bone metabolism in femur are shown in Figure 6. Compared to intact, ovx resulted in differential expression of 19/84 genes evaluated (Figure 6A). Administration of 100, 250 and 750 ppm ILQ to ovx rats resulted in differential expression of 7, 14, and 17 genes, compared to ovx control rats, respectively. As shown in the Venn diagram (Figure 6B), there was little overlap between genes that were differentially expressed following ovx and genes that were differentially expressed following treatment with 250 and 750 ppm ILQ.

Figure 6.

Gene array showing fold changes for differentially expressed genes in femur in ovariectomized (ovx) rapidly growing rats administered isoliquiritigenin (ILQ) in diet at 100 ppm, 250 ppm or 750 ppm for 7 days.

Effects of ILQ on osteoclasts in vitro

The effects of ILQ treatment on in vitro osteoclast differentiation are shown in Figure 7. Vehicle treatment has no effect on TRAP levels on either day 4 or day 7 of culture, or on Tnfrs11a or Acp5 expression. On day 4, TRAP levels were higher at 0.2 μM and 1 μM ILQ and lower at 5 μM and 25 μM compared to vehicle control. On day 7, TRAP levels were lower at all doses above 0.2 μM compared to vehicle control. Significant differences in Tnfrs11a expression were not detected with ILQ treatment. Acp5 expression was lower in cultures treated with 25 μM ILQ compared to vehicle control.

Figure 7.

Effects of treatment with isolquiritigenin (ILQ), at a concentration of 0, 0.2, 1, 5, or 25 μM, on production of TRAP by marrow-derived osteoclasts following A) 4 days and B) 7 days in culture, and on expression of C) Tnfrsf1a and D) Acp5 by osteoclasts following 7 days in culture. N = 5 replicates/treatment.

^aDifferent from vehicle control, P ≤ 0.05, ^a*Different from vehicle control, P ≤ 0.1.

^bDifferent from 0.2 μM, P ≤ 0.05, ^b*Different from 0.2 μM, P ≤ 0.1.

^cDifferent from 1 μM, P ≤ 0.05.

^dDifferent from 5 μM, P ≤ 0.05, ^d*Different from ovx + 250 ppm ILQ, P ≤ 0.1.

Effects of dietary ILQ supplementation on uterus

The effects of ovx and dietary ILQ treatment on the uterus are shown in Figure 8. As expected, uterine wet weight (Figure 8A) and epithelial cell height (Figure 8B) were lower in ovx rats compared to intact control rats. Dietary supplementation with ILQ had no significant effect on uterine weight or epithelial cell height compared to ovx rats. However, 2/5 rats in the 750 ppm group exhibited uterine hyperplasia. Representative histological images of uterus from an ovary intact, ovx and hyperplastic ILQ-treated rat (Figure 8C) clearly illustrate the pathological response of this subgroup to ILQ treatment. Ovx resulted in higher expression of Esr1 (Figure 8D), no change in expression of Esr2 (Figure 8E), lower expression of Igf1 (Figure 8F) and no change in expression of Pgr (Figure 8G). ILQ-supplemented rats did not differ significantly in gene expression from ovx rats and there was no clear dose-response effect.

Figure 8.

Effects of ovariectomy (ovx) and treatment with isolquiritigenin (ILQ), administered in diet at 100, 250, or 750 ppm for 7 days, on A) uterine weight and B) uterine epithelial cell height, and expression of D) Esr1, E) Esr2, F) Igf1, and G) Pgr in uterus, in rapidly growing ovx rats. Representative images of uterus from an intact, an ovx, and an ovx + 750 ppm ILQ rat with uterine hyperplasia are shown in panel C. N = 5/group.

^aDifferent from intact control, P ≤ 0.05, ^a*Different from intact control, P ≤ 0.1.

^b*Different from ovx control, P ≤ 0.1.

^dDifferent from ovx + 250 ppm ILQ, P ≤ 0.05.

Discussion

We evaluated the dose response effects of one week of dietary ILQ administration in rapidly growing ovx rats. ILQ administration resulted in serum levels of ILQ ranging from 0.6 to 2.4 μM. This dose range had no effect on food intake, terminal body weight, or white adipose tissue weight. Ovx resulted in higher percent body weight gain, lower uterine weight, epithelial cell height, and Igf1 expression and higher Esr1 expression. Ovx also resulted in lower cancellous bone volume fraction in proximal tibia and differential expression of numerous genes in distal femur. Whereas ILQ treatment had no effect on serum osteocalcin levels, it resulted in a dose-independent decrease in serum CTX and differential expression of genes in femur associated with bone resorption, including Acp5. Additionally, ILQ suppressed TRAP expression in murine osteoclasts generated in vitro. Finally, ovx and ILQ treatment had limited overlap on differential expression of genes in femur. ILQ treatment had no significant effect on uterine weight, epithelial cell height, or gene expression. However, the highest dose of ILQ resulted in severe epithelial cell hyperplasia in two of the five treated rats.

The absence of effect of ILQ on food intake, body weight, or white adipose tissue weight is in general agreement with a study performed in growing ovx mice fed normal diet ^[1,a]. Because of the short duration of the present study, bone mass and architecture were not considered primary endpoints. Nevertheless, we detected changes in ovx rats, including lower cancellous bone volume fraction in the proximal tibial metaphysis and increased tibial length. Both findings are consistent with the well-established role of gonadal hormones in contributing to sexual dimorphism of the rat skeleton ^[17], where estrogen antagonizes bone elongation but promotes a net increase in accrual of cancellous bone that is largely due to a reduction in bone resorption ^[18].

The absence of effect on either serum osteocalcin levels or osteoblast number suggests that ILQ did not influence bone formation. In contrast, ILQ lowered serum CTX, implying a global reduction in bone resorption. We did not detect a decrease in osteoclast number in proximal tibia, suggesting a location-specific, rather than universal, decrease in bone resorption. Alternatively, ILQ may have inhibited osteoclast activity rather than number. This latter possibility is supported by our finding that ILQ lowered expression levels of Acp5 (tartrate resistant acid phosphatase, TRAP) in femur and osteoclast culture and TRAP levels in osteoclast culture. The in vitro findings suggest that ILQ or a metabolite, such as LIQ, has direct effects on osteoclasts. Our findings are in general agreement with prior work demonstrating inhibitory effects of ILQ and LIQ on osteoclast differentiation in vitro ^{[7, 19]}.

There is evidence that ILQ can bind to estrogen receptors and activate estrogen receptor signaling ^[20]. We therefore evaluated the effects of ILQ treatment on expression of genes related to bone metabolism in femur of ovx rats. We found little overlap in genes that were differentially expressed following ovx and treatment of ovx rats with ILQ. These null findings imply that the actions of ILQ on bone differ from endogenously produced gonadal hormones.

Two of five rats exhibited pathological uterine epithelial cell hypertrophy, where uterine luminal cell height in ILQ-treated rats greatly exceeded luminal cell height in ovary-intact rats. The mechanism for this dramatic response by a subgroup of animals is unclear. Unopposed estrogen treatment increases epithelial cell height in growing ovx rats ^[21] by an estrogen receptor-mediated mechanism ^[5]. Whereas ovx resulted in higher expression of Esr1 and lower expression of IGF-1 in uterus, this effect was not reversed by ILQ treatment, suggesting that ILQ was not acting on the uterus as an estrogen receptor agonist. These findings are in general agreement with our earlier work in mice ^[1,a].

ILQ and LIQ are reported to have estrogenic activities through both ERα and ERβ, with a 20-fold higher affinity for ERβ ^[22]. Previous studies have shown ILQ and LIQ bind directly to the estrogen receptor in vitro, compete with estradiol for binding, and are estradiol agonists for both forms of the estrogen receptor ^{[22a, 23]}. A range of 0.1 – 1 μM of ILQ in vitro is needed for a half-maximal response, compared to 0.2 nM for 17β-estradiol ^[24] but, in addition, the response can depend on concentration. At low concentrations, ILQ stimulates the proliferation of ERα-dependent MCF7 breast cancer cells but is suppressive at high levels ^[25]. The present in vivo study achieved serum levels of ILQ sufficient to bind to and activate ER. This finding implies that total serum ILQ does not reflect levels available for estrogen receptor binding in target tissues. Additional research will be required to understand the precise mechanism.

ILQ has been shown to also act via multiple non-estrogen receptor pathways. A recent report described a life threatening case of pseudo-aldosteronism secondary to excessive licorice ingestion ^[26]. Furthermore, ILQ and LIQ can act on the uterus through voltage Ca2+ channel blockade, altering uterine contractions and inducing an analgesic effect in an estrogen-independent manner ^[27]. This latter study revealed that licorice root exerted a uterine relaxant effect by inhibiting the phosphorylation of HSP27 to alter the interaction of this protein and actin and, based on this finding, LIQ and ILQ were preliminarily classified as PKCδ inhibitors. The varied estrogen receptor independent effects of ILQ warrant further investigation for potential therapeutic benefits.

Bone turnover markers typically increase following ovx in sexually mature rats. In this study, growing rats underwent ovx prior to sexual maturity. This was done as a strategy ^[10] to prevent the rats from being exposed to high levels of gonadal hormones prior to treatment with ILQ. Osteocalcin and CTX levels in serum and osteoblast-lined and osteoclast-lined bone perimeters were quite high, indicating rapid bone growth and high turnover. While high bone turnover due to young age of rats, combined with the small sample size, might present a challenge to observing an increase in turnover following ovx, it enhanced our ability to detect effects of weak estrogens.

Limitations of the current study include short duration and the fact that the study does not distinguish effects of the parent compound ILQ from metabolites such as LIQ, which may contribute to bioactivity ^[28]. The short duration of the study precludes definitive assessment of the effects of ILQ treatment on bone mass, density, and architecture. However, important causality-associated changes in gene expression and biochemical markers of bone turnover are anticipated to occur prior to changes in bone mass or architecture. The young age of the rats at ovx may also be perceived as a limitation because it does not model menopause, which is important because botanicals believed to have estrogenic activity are often consumed as a strategy to prevent or attenuate post-menopausal symptoms, including bone loss ^[1b]. However, as indicated above, the goal of this study was not to evaluate ILQ as a postmenopausal treatment but to determine whether blood levels of ILQ and metabolites that could be reasonably achieved with a dietary intervention influence bone turnover. For this purpose, the growing ovx rat is exquisitely sensitive and data generated from this study can be used to inform the design of studies to provide preclinical evidence for an intervention trial in postmenopausal women.

Regulators, such as the Food and Drug Administration (FDA) in the United States, the European Union, and the World Health Organization, have established guidelines for preclinical studies assessing treatment for postmenopausal osteoporosis. The FDA, for example, recommends performing dose-ranging experiments to facilitate design of rigorous preclinical studies intended to provide evidence justifying a clinical trial (FDA-2016-D-1273). Furthermore, the FDA guidelines suggest a target dose and a dose 10x the target dose, the latter being important for evaluating safety. Based on the present dose-ranging study, a long duration experiment in skeletally mature (≥ 8 months old) ovx rats ^[29] designed to achieve blood levels of ILQ of 1 μM (efficacy) and 10 μM (safety) is warranted, the latter being especially important in view of the potential negative effects on the uterus.

In summary, dietary ILQ reduced a biochemical marker of bone resorption in vivo and osteoclast differentiation in vitro, by mechanisms likely differing from those mediating the antiresorptive actions of estrogen. Concentrations of total ILQ in the μM range were present in serum without impacting normal food consumption or weight gain. Long duration studies in skeletally mature ovx rats are warranted to establish whether ILQ can attenuate bone loss associated with gonadal hormone deficiency without detrimental side effects.