*Article* **The Effect of Dextrose or Protein Ingestion on Circulating Growth Differentiation Factor 15 and Appetite in Older Compared to Younger Women**

**Catrin Herpich 1,2,3,\*, Stephanie Lehmann 1, Bastian Kochlik 3, Maximilian Kleinert 4,5,6, Susanne Klaus 1,7, Ursula Müller-Werdan 2,8 and Kristina Norman 1,2,3,9**


**Abstract:** Growth differentiation factor 15 (GDF15) is a stress signal that can be induced by protein restriction and is associated with reduced food intake. Anorexia of aging, insufficient protein intake as well as high GDF15 concentrations often occur in older age, but it is unknown whether GDF15 concentrations change acutely after meal ingestion and affect appetite in older individuals. After an overnight fast, appetite was assessed in older (*n* = 20; 73.7 ± 6.30 years) and younger (*n* = 20; 25.7 ± 4.39 years) women with visual analogue scales, and concentrations of circulating GDF15 and glucagon-like peptide-1 (GLP-1) were quantified before and at 1, 2 and 4 h after ingestion of either dextrose (182 kcal) or a mixed protein-rich meal (450 kcal). In response to dextrose ingestion, appetite increased in both older and younger women, whereas GDF15 concentrations increased only in the older group. In older women, appetite response was negatively correlated with the GDF15 response (rho = −0.802, *p* = 0.005). Following high-protein ingestion, appetite increased in younger women, but remained low in the old, while GDF15 concentrations did not change significantly in either age group. GLP-1 concentrations did not differ between age groups or test meals. In summary, acute GDF15 response differed between older and younger women. Associations of postprandial appetite and GDF15 following dextrose ingestion in older women suggest a reduced appetite response when the GDF15 response is high, thus supporting the proposed anorectic effects of high GDF15 concentrations.

**Keywords:** aging; anorexia of aging; GDF15; GLP-1; postprandial

#### **1. Introduction**

Growth differentiation factor 15 (GDF15) is a cellular stress-induced cytokine, and higher circulating concentrations are found in various chronic and acute diseases [1] as well as in older age [2]. The role and effects of higher GDF15 concentrations, during aging in particular, are unclear.

While most cells and tissues are able to secrete GDF15 [3], the expression of the GDF15 receptor, glial cell-derived neurotrophic factor family receptor alpha-like (GFRAL), has been

**Citation:** Herpich, C.; Lehmann, S.; Kochlik, B.; Kleinert, M.; Klaus, S.; Müller-Werdan, U.; Norman, K. The Effect of Dextrose or Protein Ingestion on Circulating Growth Differentiation Factor 15 and Appetite in Older Compared to Younger Women. *Nutrients* **2022**, *14*, 4066. https://doi.org/10.3390/ nu14194066

Academic Editors: Birgit-Christiane Zyriax and Nataliya Makarova

Received: 9 September 2022 Accepted: 26 September 2022 Published: 30 September 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

detected only in the brainstem [4]. Recently, GDF15 has been shown to regulate appetite [5] via GDF15-GFRAL signaling [1]. The activation of this signaling axis is associated with conditioned taste aversion [3,6] and the modulation of the vagal sympathetic nervous system, which controls, e.g., gastric emptying [7]. GDF15 expression is sensitive to various nutritional stimuli such as chronic high-fat overfeeding or lysine-deficient diets, which is mediated by the integrated stress response (ISR) [3].

Aging is frequently accompanied by anorexia of aging, which is characterized by reduced appetite as well as lower food intake, and is associated with an increased risk for malnutrition, since the lower energy intake often coincides with insufficient macro- and micronutrient supply, most importantly of protein [8,9]. The etiology is not clear, but mostly likely involves the interplay of several age-related sensory and metabolic changes [8]. Sensory decline, e.g., the loss of taste and sense of smell, can lead to reduced food palatability. Furthermore, various metabolic alterations contribute to overall increased satiety, and therefore lower appetite and food intake, in the old [8]. For example, postprandial hunger, satiety as well as gastric emptying differ in older compared to younger adults after ingestion of a mixed meal [10]. In this context it is interesting that bariatric surgery, which is known to alter satiety and gastric emptying, increases circulating GDF15 in both men and women [11].

To date, it is not known if the high GDF15 concentrations found in older ages affect appetite in humans, and whether there is an age-dependent difference in the GDF15 response to different meals. In addition, it remains to be elucidated if the GDF15 response to meal ingestion affects postprandial appetite. Overall, there are only a few studies addressing the postprandial response of GDF15 in humans [12–14]. Therefore, we investigated postprandial circulating GDF15 and its association with postprandial appetite in young and older women.

#### **2. Materials and Methods**

This is a sub-analysis of a larger study described elsewhere [15]. In brief, communitydwelling older and younger adults were recruited. In order to obtain a significant age gap between the groups, we pre-specified age ranges of 65 to 85 years for the older group and 18 to 35 years for the younger group. In the older group, we did not recruit adults aged above 85 out of ethical considerations, and in the younger group we did not recruit adults over 35 to preclude any early perimenopausal changes. The study was approved by the ethics committee of the University of Potsdam and registered at drks.de as DRKS00017090. All participants signed a written informed consent.

As the number of men was low and sex differences are known [16], men were excluded from this analysis. One younger woman had to be excluded from postprandial analysis since she did not complete the meal challenge. The postprandial GDF15 response was assessed after dextrose (50 g dextrose, total energy content: 182 kcal; *n* = 10 per age group) or high-protein ingestion (77 energy percent protein, total energy content: 450 kcal; 250 g curd cheese, 50 g protein powder, 100 g raspberries, 100 mL milk 1.5% fat; *n* = 10 older group, *n* = 9 younger group). Blood samples were drawn after an overnight fast and repeated blood samples were taken 30, 60, 120 and 240 min after meal ingestion. EDTA plasma was obtained and stored at −80 ◦C until analysis. Subjective appetite was assessed using a visual analogue scale. Participants were instructed to mark their current feeling of appetite (ranging from 0 = "no appetite" to 10 = "great appetite") every time a blood sample was drawn. Appetite sensation was displayed in cm on the VAS. Fat mass was estimated using bioelectrical impedance analysis and age-appropriate equations [17]. Fat mass index (FMI) was calculated by dividing the fat mass (kg) by height squared (m2).

Plasma GDF15 (intra-assay CV: 6.3–7.2%; inter-assay CV: 2.9–5.6%; BioVendor, Brno, Czech Republic) concentrations were quantified using commercial ELISA assays. As an objective marker for appetite/satiety, we also measured glucagon-like peptide 1 (GLP-1) (intra-assay CV: 4.69–10.7%; inter-assay CV: 9.63–17.6%; Yanaihara Institute Inc, Shizuoka, Japan). As markers of the glucose metabolism, serum insulin (commercial ELISA, intraassay CV: 4.8–6.0%; inter-assay CV: 8.1–9.0%; BioVendor, Brno, Czech Republic) as well as serum glucose concentrations (colorimetric method, ABX Pentra 400, Horiba, Ltd., Kyoto, Japan) were quantified. Homeostasis model assessment was used to estimate insulin resistance (HOMA-IR).

Statistical analyses were performed with SPSS (IBM version 27, SPSS Incorporated, Chicago, IL, USA) and GraphPad Prism (version 7.00 for Windows, GraphPad Software, La Jolla, CA, USA). Data are presented as mean ± standard deviation (SD). Group differences were calculated using as appropriate Student's *t*-test or Mann–Whitney U test and correlations with Pearson's correlation coefficient or Spearman's rho. GLP-1 concentrations were logarithmized for normalization. Changes over time and time × meal interactions were examined with repeated measures ANOVA. GDF15, GLP-1, glucose and insulin response, and increase in appetite after meal ingestion were evaluated using positive incremental area under the curve (iAUC). An acceptable level of statistical significance was established a priori at *p* < 0.05.

#### **3. Results**

A description of the participants is shown in Table 1. The study participants were overall healthy, with self-reported high blood pressure being the most frequent pre-existing condition in older women (50%). Despite having higher fasting glucose concentrations, older women exhibited similar insulin and HOMA-IR values to younger women. Baseline GDF15 concentrations were significantly higher in the older compared to the younger women (802 ± 227 versus 364 ± 125 pg/mL, *p* < 0.001).


**Table 1.** Characteristics of study participants at baseline.

BMI: body mass index; FMI: fat mass index, GDF15: growth differentiation factor 15; GLP-1: glucagon-like peptide 1; HOMA-IR: Homeostasis Model Assessment—Insulin Resistance; SD: standard deviation, differences between groups calculated using Student's *t*-test; <sup>a</sup> differences between groups calculated using Mann–Whitney U test.

BMI was similar between both groups, but older women exhibited a higher fat mass (8.65 versus 7.0 kg/m2, *p* = 0.036). Fat mass index was also positively correlated with fasting GDF15 concentrations (r = 0.346, *p* = 0.029), but not with BMI (Supplementary Material Figure S1). In older women, baseline GDF15 concentrations were negatively correlated with baseline appetite (r = −0.488, *p* = 0.029), whereas fasting GLP-1 concentrations were positively associated with baseline appetite (r = 0.461, *p* = 0.041) (Figure 1). Glycemic parameters were not correlated with GDF15 or GLP-1.

**Figure 1.** Correlation of baseline appetite and baseline (**A**) GDF15 (older: r = −0.488, *p* = 0.029; younger: r = −0.228, *p* = 0.347) and (**B**) GLP-1 concentrations (older: older, r = 0.461, *p* = 0.041; younger: r = 0.227, *p* = 0.350). GLP-1 was logarithmized for normalization. Correlations of GLP-1 were calculated using log-transformed values but are shown as untransformed values for better visualization. Closed circles represent older women, open triangles younger women.

#### *3.1. Appetite*

Overall, in both older and younger women, appetite changed over time (*p* = 0.015 versus *p* < 0.001; Figure 2A,D), but only in older women did postprandial appetite differ between the two test meals (*p* = 0.015 versus *p* = 0.383 in younger women). Following dextrose ingestion, appetite significantly increased in both age groups during the meal challenge. After protein ingestion, only younger women exhibited increasing appetite from 120 to 240 min, whereas appetite did not change over time in the older women. At the end of the meal challenge, appetite was similar for both test meals in younger women, but in older women appetite was higher after dextrose compared to protein ingestion (mean difference: 4 cm, *p* = 0.021).

**Figure 2.** Postprandial appetite (**A**), GDF15 (**B**) and GLP-1 (**C**) concentrations in older and younger women ((**D**,**E**,**F**), respectively) following dextrose (closed circles) or protein (open circles) ingestion. GLP-1 concentrations were logarithmized for normalization. Repeated measures ANOVA, data are shown as mean ± SD. \* indicates significant difference to 240 min, \*\* to 120 min, separately for both test meals. Postprandial changes of GLP-1 concentrations were calculated using log-transformed values but are shown as untransformed values for better visualization. *n* = 10 per group; *n* = 9 in younger high-protein group. Closed circles represent older women, open circles younger women.

#### *3.2. GDF15*

GDF15 concentrations significantly changed over time in both older (*p* < 0.001) and younger women (*p* = 0.019; Figure 2B,E). Only in older women did the test meal have an effect on postprandial GDF15 concentrations (*p* = 0.026). Following dextrose ingestion, GDF15 concentrations significantly increased in older women from 60 to 240 min, but not in younger women. After protein ingestion, GDF15 concentrations slightly increased in older women from 120 to 240 min and from 60 to 120 min in younger women. However, in older women, GDF15 concentrations were higher in response to dextrose compared to protein at 120 min (mean difference: 202 pg/mL, *p* = 0.023) and 240 min (mean difference: 320 pg/mL, *p* = 0.017) after meal ingestion. GDF15 concentrations during the meal challenge were not different between dextrose or high-protein ingestion in younger women.

#### *3.3. GLP-1*

In both age groups, GLP-1 concentrations increased after meal ingestion (older: *p* < 0.001, younger *p* < 0.001; Figure 2C,F), but were not different between the meals.

#### *3.4. Glucose Metabolism*

Postprandial glucose and insulin concentrations after meal ingestion are depicted in Figure 3. Glucose concentrations significantly changed over time in both older (*p* < 0.001, Figure 3A) as well as younger women (*p* = 0.012, Figure 3C) and were different between test meals (old: *p* < 0.001, young: *p* = 0.001). Postprandial insulin concentrations also significantly changed over time in both older and younger women (*p* < 0.001 in both age groups), but only in older women, the insulin concentration changes over time were different between the test meals (*p* = 0.027; Figure 3B,D).

**Figure 3.** Postprandial glucose (**A**) and insulin (**B**) concentrations in older and younger women ((**C**,**D**) respectively) following dextrose (closed circles) and high protein (open triangles ingestion. Repeated measures ANOVA, data are shown as mean ± SD. *n* = 10 per group; *n* = 9 in the younger high-protein group. Closed circles represent older women, open triangles younger women.

#### *3.5. Association of Appetite, GDF15, GLP-1, Glucose and Insulin Response*

To evaluate associations among appetite, GDF15 and GLP-1 response, iAUCs were calculated. There was no age difference regarding appetite, GDF15, GLP-1, glucose and insulin iAUCs (Supplementary Material Figure S2). However, following dextrose ingestion in older women only, appetite iAUC was negatively correlated with GDF15 iAUC

(rho = −0.802, *p* = 0.005), which suggests that the increase in appetite was less prominent when the postprandial increase in GDF15 was high (Figure 4). In younger women after protein ingestion, GDF15 iAUC was positively correlated with GLP-1 iAUC (rho = 0.729, *p* = 0.026) and insulin (rho = 0.949, *p* < 0.001), but this was not seen in older women. This might indicate that GDF15 behaves similarly to GLP-1 in the young but not in older women.

**Figure 4.** Correlation of appetite iAUC and GDF15 iAUC after dextrose ingestion in older (rho = −0.802, *p* = 0.005) and younger women (rho = −0.215, *p* = 0.550). iAUC: incremental area under the curve. Closed circles represent older women, open triangles younger women.

#### **4. Discussion**

Due to its many functions, GDF15 has been a target of pharmacological research to treat, e.g., cachexia and diabetes [18]. Moreover, GDF15 has recently gained attention as a potential biomarker for cellular senescence [19] and a key player in the aging process [2], but also as an important regulator of weight homeostasis and appetite [1,20]. To our knowledge, to date there are no studies investigating postprandial appetite and GDF15 concentrations in older adults compared to younger. In this analysis, we show that acute GDF15 response was different between older and younger women and dependent on the type of test meal in older women. Only in the older group and after dextrose ingestion was an increase in GDF15 concentrations found. Postprandial appetite increased over time in both age groups following dextrose ingestion; however, following protein ingestion, appetite remained low in the old, while increasing in the young. In older women, a higher GDF15 response to dextrose was associated with a lower increase in appetite.

The age difference regarding appetite after high protein ingestion might be due to the slower gastric emptying time in older age [10]. In addition, our results imply that GDF15 affects appetite in older women, since fasting GDF15 concentrations as well as their responses were negatively associated with appetite. However, these results also suggest that in younger women, GDF15 does not influence appetite. Studies in mice imply that the anorectic and even nauseating properties of GDF15 unfold only at high/pharmacological concentrations [21,22]. Yet, it is unclear what levels are required in humans for GDF15 to affect energy balance and exert anorexic effects. The association between GDF15 and appetite might therefore be even more prominent in older adults with strongly elevated concentrations, as seen in disease [16].

The regulation of GDF15 secretion is complex as various organs and tissues produce GDF15 [2]. Moreover, multiple stressors and stimuli are able to regulate its expression. One prominent regulator is the ISR, which is induced in response to, among other things, protein restriction [3]. As the dextrose meal was free of protein, the GDF15 increase after dextrose intake might also be interpreted as a response to a lack of protein. This is supported by the observation that GDF15 concentrations after protein ingestion do not change to the same extent as after dextrose intake. Furthermore, the different amounts of calories ingested (180 versus 450 kcal) might also have affected postprandial GDF15 concentrations. However, as short-term overfeeding studies in humans and rodents did not result in altered GDF15 concentrations [3], this effect appears to be negligible.

To date, only a few studies have investigated post-meal GDF15 concentrations. One study found increasing GDF15 concentrations after an oral glucose tolerance test (75 g dextrose) in younger to middle-aged adults with obesity [13]. In addition, it was shown that the ingestion of a high-carbohydrate or a high-fat mixed meal (protein content 12 E%) did not result in postprandial changes of GDF15. This was also seen in another study using mixed meals (protein content 15 E%) in a younger cohort [14]. Possibly, the protein content of the meals in these studies was sufficient, and therefore GDF15 expression was not induced. However, in none of these studies was postprandial appetite assessed.

Additionally, insulin resistance might play a role in the acute regulation of GDF15. In response to dextrose ingestion, glucose concentrations rose to higher levels at 60 min in older compared to younger women (8.27 versus 5.93 mmol/L), whereas postprandial insulin concentrations were not different between the age groups. This indicates that the older women in this analysis were more insulin-resistant than the younger women, which is a known age-associated effect. GDF15 and insulin iAUC were strongly positively correlated in younger women after protein ingestion, which is in line with the literature, wherein GDF15 was found to regulate insulin secretion [18]. This suggests a complex interplay between nutritional stimuli, GDF15 and insulin, as this is neither seen in response to dextrose nor in older women.

Our study is subject to limitations, such as the number of subjects and the subjective evaluation of the appetite. As a metabolic indicator of energy status as opposed to the subjective rating of appetite, we also analyzed postprandial GLP-1 concentrations. GLP-1 is known to enhance satiety and reduce energy intake [23], and overall exhibits similar actions to GDF15 (Figure 5) [24]. However, GLP-1 is associated differently with fasting and postprandial appetite compared to GDF15. This might imply a dissociation between the expected actions of GLP-1 on satiety and the subjective rating of appetite. A reason for this dissociation might be that humans can be more sensitive to external factors (such as meal size, company while eating, time of day) than internal biological stimuli [25]. Moreover, we were not able to control for all confounders that might have an effect on the findings.

**Figure 5.** Overview of selected shared functions between GDF-15 and GLP-1. Arrows pointing up indicate enhancing of, arrows pointing down refer to a downregulation. Green refers to GDF15 functions, blue to GLP-1. Created with https://biorender.com/ (accessed on 23 September 2022).

In conclusion, we showed age-specific differences in the appetite's response to protein intake, and in the GDF15 response following dextrose ingestion. Possibly, studies on subjects with GDF15 concentrations higher than 1200 pg/mL, such as in clinical settings, might reveal a more prominent anorectic effect of GDF15 after meal ingestion.

**Supplementary Materials:** The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/nu14194066/s1, Figure S1: Correlation of baseline GDF15 concentrations and (A) FMI and (B) BMI. BMI: body mass index, FMI: fat mass index; Figure S2: Incremental area under the curve (iAUC) for appetite (A), GDF15 (B), GLP-1 (C), glucose (D) and insulin (E) for each test meal and age group. OW: older women, YW: younger women, Dex: dextrose, HP: high protein.

**Author Contributions:** Data curation, C.H., S.L. and B.K.; writing—original draft preparation, C.H.; supervision, K.N.; writing—review and editing, B.K., M.K., S.K., U.M.-W. and K.N. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted in accordance with the Declaration of Helsinki, and was approved by the ethics committee of the University of Potsdam and registered at drks.de as DRKS00017090.

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** Data sharing not applicable. Data cannot be shared due to national data protection laws.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


## *Article* **The Associations between Metalloestrogens, GSTP1, and SLC11A2 Polymorphism and the Risk of Endometrial Cancer**

**Kaja Michalczyk 1,\*, Patrycja Kapczuk 2, Grzegorz Witczak 1, Mateusz Bosiacki 3, Mateusz Kurzawski 4, Dariusz Chlubek <sup>2</sup> and Aneta Cymbaluk-Płoska <sup>1</sup>**


**Abstract:** Background: The incidence of endometrial cancer (EC) is still rising. Numerous risk factors including patient characteristics and molecular instability have been identified for EC. The presence of specific molecular markers allows specific diagnostic and prognostic approaches. Several single nucleotide polymorphisms (SNPs) have been identified to influence endometrial cancer risk. Metalloestrogens are metal ions which can mimic estrogen activity; however, their role in uterine pathologies remains unknown. This study aimed to investigate total blood trace elements levels and evaluate the distribution of selected genotypes in GSTP1 and SLC11A2 genes. Methods: This retrospective case-control analysis was carried out in peripheral blood samples of 110 women with endometrial cancer (EC; *n* = 21), uterine fibroma (*n* = 25), endometrial polyp (*n* = 48), and normal endometrium (*n* = 16). Analysis included measurement of metals and phosphor in serum, and of genetic polymorphisms in GST (rs1695) and SLC11A2 (rs224589) in DNA from white blood cells. Serum trace elements were measured using ICP-OES spectrometry. SNPs were identified using Taq Man real-time PCR genotyping assays. Results: The study confirmed higher age (OR 2.19, 95% CI 1.69–2.24), post-menopausal status (OR 1.89, 95% CI 1.36–1.94), and diabetes type 2 (OR 1.54; 95% CI 0.97–1.72) as independent risk factors for EC. We also found a high level of Cd (OR 1.49; 95% CI 1.31–1.63) and a low level of Co (OR 0.76; 95% CI 0.53–0.59) to be independent risk factors of EC. None of the tested polymorphisms of GSTP1 and SLC11A2 were associated with EC risk. However, high Cd (OR 1.21, 95% CI 1.15–1.29) and Ni (OR 1.07, 95% CI 1.05–1.18) serum levels were significantly associated with a SLC1A2 TG genotype, and high Cd levels with GSTP1 (OR 1.05, 95% CI 1.01–1.13).

**Keywords:** metalloestrogens; nutrients; trace elements; GSTP1; SLC11A2; endometrial cancer

#### **1. Introduction**

Endometrial cancer (EC) is the world's most common gynecological malignancy and one of a few cancers with an increasing incidence [1]. Numerous factors have been identified to increase the risk of endometrial cancer, including patients' age, obesity, late menopause onset, and unbalanced estrogen [2,3]. The new molecular classification, based on the Cancer Genome Atlas Research Network, distinguishes four types of EC by their molecular characteristics: Polymerase Epsilon Mutation (POLE) ultramutated, microsatellite instability hypermutated (MSI), copy-number low (CNL), and copy-number high (CNH) [4]. The presence of specific molecular markers allows new, more specific diagnostic and prognostic

**Citation:** Michalczyk, K.; Kapczuk, P.; Witczak, G.; Bosiacki, M.; Kurzawski, M.; Chlubek, D.; Cymbaluk-Płoska, A. The Associations between Metalloestrogens, GSTP1, and SLC11A2 Polymorphism and the Risk of Endometrial Cancer. *Nutrients* **2022**, *14*, 3079. https://doi.org/ 10.3390/nu14153079

Academic Editors: Birgit-Christiane Zyriax and Nataliya Makarova

Received: 24 June 2022 Accepted: 25 July 2022 Published: 27 July 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

possibilities and the use of targeted therapy. The most common alternations were found to occur in TP53, PTEN, PI3KCA, CTNNB1, ARID1A, and KRAS pathways [4]. Multiple studies have investigated the associations between diverse single nucleotide polymorphisms (SNPs) and endometrial cancer risk. Strong associations were found for HNF1B, KLF, EIF2AK, CYP19A1, SOX4, and MYC [5].

Endometrial hyperplasia is a term describing an excessive proliferation of endometrial cells. It is a benign condition; however, it is considered as a precursor state for endometrial carcinomas [6]. Endometrial hyperplasia is usually correlated with unopposed estrogen, which in the absence of abnormal endometrial proliferation in the progesterone, stimulates abnormal endometrial proliferation [7]. Additionally, other benign uterine lesions, including uterine fibroids, were found to be hormone-dependent and estrogen is considered to be the major mitogenic factor in the uterus [8].

There is limited information on the influence of trace elements on human endometrium and their role in carcinogenesis. Metals including Zn, Cu, and Fe are essential chemicals for human well-being in trace amounts. However, when in excess, they may cause adverse effects including the induction of cellular damage, alternation in cellular homeostasis, inflammation, production of ROS (reactive oxygen species), and finally carcinogenic activity [9,10]. Additionally, certain metals such as Aluminum, Cadmium, Copper, Cobalt, Nickel, Lead, Tin, and Chromium have been found to have the ability to mimic estrogen activity and to activate estrogen receptors and were therefore named metalloestrogens [11–13]. This is why we decided to measure serum trace elements concentration in patients diagnosed with malignant and benign uterine lesions.

Glutathione S-transferase (GST) is an enzyme that catalyzes the conjugation of glutathione into electrophilic compounds. The enzyme not only detoxifies endogenous and exogenous species but also participates in the activation of oxidative metabolites participating in carcinogenesis, including ROS, and regulates stress-induced signaling pathways [14,15]. Some studies have presented an association between GSTP1 gene polymorphisms and increased risk of cancers, including endometrial cancer [16–18]. A similar association was found for endometrial hyperplasia [19].

SLC11 is a family of integral membrane proteins that are divalent metal ions transporters that use H+ -electrochemical gradient as a driving force to transport metal ions [20,21]. SLC11A2, also known as DMT1, and Nramp2, is widely distributed and expressed in the duodenum, erythroid cells, kidney, lung, brain, testis, thymus, and placenta [22]. The predominant substrates of SLC11A2 are Fe2+, Cd2+, Co2+, Cu1+, Mn2+, Ni2+, Pb2+, and Zn2+ [22].

SLC11A2 has an important role in iron homeostasis and transport. Mutations in the SLC11A2 gene were found in patients suffering from hypochromic microcytic anemia with serum and liver iron overload [23,24], while its activation was found to lead to severe pathologies including autophagy and cell death in Parkinson's disease [25]. Additionally, overexpression of SLC11A2 was found to be associated with several cancers including esophageal, colorectal, and breast carcinomas [22,26,27]. They also correlated with an invasive form of the disease.

Single nucleotide polymorphism (SNP) are forms of DNA variation among individuals either caused by nucleotide transition or transversion. They may change the encoded amino acids into nonsynonymous, replacing one nucleotide with a different one, and can be silent (synonymous) or occur in the noncoding region. SNPs may result in gene expression changes, mRNA stability and protein coding. The identification of gene variations and their effect analysis may allow a betting understanding of their impact on gene function as they may be responsible for characteristics causing population diversity, genome evolution, familial or interindividual traits, and differences in disease prevalence and treatment response.

Genes and the genetic polymorphism of genes involved in metalloestrogen homeostasis (i.e., glutathione S-transferase P1 gene (GSTP1) and metal ions transport (i.e., the Solute Carrier 11 group A member 2 gene (SLC11A2) may be closely related to estrogen

overstimulation and therefore serve as a potential risk factor of endometrial cancer. Genetic variation may explain the heterogeneity of patients and help identify those more susceptible to metalloestrogen stimulation. GSTP1 rs1695 and SLC11A2 rs224589 are widely studied polymorphisms of the genes coding regions of the mentioned enzymes that may alter enzyme activity among different genotypes.

This study aimed to investigate the association of the selected polymorphisms: rs1695 in GSTP1 and rs224589 in SLC11A2, together with serum and blood trace elements in different endometrial pathologies.

#### **2. Materials and Methods**

#### *2.1. Study Participants*

The study included 140 patients consecutively admitted to the Department of Gynecological Surgery and Gynecological Oncology of Adults and Adolescents, Pomeranian Medical University. This case-control study included patients with a confirmed diagnosis of endometrial cancer based on histopathological evaluation who were admitted for radical surgery. The control group consisted of patients admitted for hysteroscopy or laparoscopy/laparotomy with histopathologically confirmed benign uterine conditions or normal endometrium. The exclusion criteria included recurrence of endometrial cancer, previous cancer treatment or other types of primary care, and presence of unbalanced/untreated chronic diseases. In addition, patients with lost or incomplete data were removed from the study group. Finally, a total of 110 patients were included in the study analysis. The research was conducted in accordance with the Helsinki Declaration and with the consent of the Ethics Committee of Pomeranian Medical University in Szczecin under the number KB-0012/27/2020 on 9 March 2020. Patient characteristics are demonstrated in Table 1.


**Table 1.** Group characteristics.

#### *2.2. Laboratory Analyses*

From each patient, two peripheral blood samples were collected for the study purpose: one was used to obtain serum for serum trace elements analysis, while the other was used

for blood sample trace elements analysis and isolation of genomic DNA. The specimens were stored at a temperature of −80 degrees Celsius. The samples were obtained at the time of hospital admission for hysteroscopy/laparoscopy or laparotomy. Informed consent to participate in the study was obtained from all patients.

#### *2.3. Trace Elements Analysis*

For the purpose of elemental analysis, serum and whole blood samples underwent a microwave decomposition procedure using a microwave digestion system. After defrost and sample preparation, 65% HNO3 was added to the samples, which were then transferred into Teflon vessels and placed in the microwave. The process of sample digestion was composed of two stages: an initial of 15 min, at which the samples were gradually heated up to 180 ◦C, and the second of 20 min, at which the temperature was maintained at 180 ◦C. Samples were analyzed using inductively coupled plasma optical emission spectrometry (ICP-OES, ICAP 7400 Duo, Thermo Scientific, Waltham, MA, USA) equipped with a concentric nebulizer and cyclonic spray chamber to determine Zinc (Zn), Copper (Cu), Iron (Fe), Chromium (Cr), Cobalt (Co), Strontium (Sr), Phosphor (P), Magnesium (Mg), Cadmium (Cd), Nickel (Ni), and Manganese (Mn) content. The digested samples were further diluted 20-fold. For the analysis, 500 μL of yttrium was added with the final standard sample concentration at 0.5 mg/L and 1 mL of 1% Triton (Triton X-100, Sigma-Aldrich, Poland). The samples were further diluted with 0.075% HNO3 (Suprapur, Merck, Poland) up to the volume of 10 mL and stored in the fridge at 4–8 ◦C final until analysis. The calibration curve was constructed using multielement standard solutions (ICP multielement standard solution IV, IX and XVI, Merck, Kenilworth, NJ, USA).

#### *2.4. Molecular Analysis*

For the study purpose, genomic DNA was isolated from 0.2 mL of a whole blood sample (all with blood cells) using a commercial kit for genomic DNA isolation using Genomic Mini AX Blood 1000 Spin (A&A Biotechnology). The genotyping of the selected SNPs was performed using pre-designed Genotyping Assays (TaqMan real-time PCR genotyping assays, Thermo Fisher Scientific, Assay IDs: C\_\_\_2967992\_1\_, C\_\_\_3237198\_20). The following SNPs were genotyped: GSTP1 rs1695 A > G and SLC11A2 rs224589 T > G.

#### *2.5. Statistical Analysis*

The analysis was conducted using Statistica 10, StataSoft, Poland The comparison of patient characteristics between the groups was performed using the U-Mann Whitney test. The associations between the tested SNPs and cancer risk were calculated by the comparison of the frequencies of selected genotypes among the EC and control group. Odds ratios (OR) and the corresponding confidence intervals (95% CI) for each SNP were calculated using univariable regression models. *p*-value < 0.05 was adopted as the statistical significance threshold.

#### **3. Results**

We found no differences between total blood/serum trace elements concentration between endometrial cancer patients and the control group of patients diagnosed with benign uterine conditions or normal endometrium. When analyzed separately, we found a significant difference for Cd levels between patients diagnosed with endometrial cancer vs endometrial polyps as patients with endometrial polyps tended to have lower Cd concentrations (*p* = 0.002). We also found a significant difference in Cu levels between patients diagnosed with uterine fibromas vs. endometrial polyps (*p* = 0.042). A similar trend was noticed for Cd levels. Another association was found for Fe concentration as patients with endometrial polyps showed lower Fe expression than patients with normal endometrium (*p* = 0.038). All of the associations are demonstrated in Table 2.



Our study showed no correlation between patients' menopausal status and blood concentration of any of the investigated metalloestrogens (Table 3).

**Table 3.** Serum metalloestrogen concentration in pre- and postmenopausal patients.


As a part of the study, we checked for any correlations between patients' characteristics and serum trace element. Patients' BMI and age did not influence either serum trace element concentration. Yet, we found some significant intracorrelations between selected trace elements levels. Serum Zn levels positively correlated with Cu, Fe, P, and Mg concentration. All of the correlations are presented in Table 4.


**Table 4.** Spearman correlation.

The underlined variables are significant with *p* < 0.05.

We conducted a univariate logistic regression model to assess the risk factors for endometrial cancer. We found patients' age, BMI, menopausal status, and history of diabetes mellitus type 2 to influence the risk for endometrial cancer. However, trace elements concentration did not seem to influence the probability of cancer occurrence. The results are presented in Table 5.


**Table 5.** Univariate logistic regression model.

However, upon multivariate analysis, Cadmium and Cobalt levels were found to be associated with endometrial cancer risk, while above median Co levels were found to correlate with a decreased the risk of EC. The results are displayed in Table 6.

**Table 6.** Multivariate logistic regression model.


None of the tested polymorphisms revealed a correlation with endometrial cancer risk (Table 7). However, we found some non-significant differences in genotypes frequencies among endometrial cancer patients and controls as the TG genotype was more frequently expressed among EC patients compared with controls (35% vs. 25.5%, respectively).


**Table 7.** The associations between the analyzed SNPs and endometrial cancer risk.

In our study, patients with high Cd (OR 1.21, 95% CI 1.15–1.29) and Ni (OR 1.07, 95% CI 1.05–1.18) serum levels were significantly associated with a SLC1A2 TG genotype, and high Cd levels with GSTP1 (OR 1.05, 95% CI 1.01–1.13). The prevalence of other genotypes did not seem to correlate with blood metalloestrogen levels (see Table 8).


6

 0.83 0.80–0.86

 0.274

 0

 -

 -

 -

 8

 1.27 1.22–1.28

 0.058

 7

 0.86 0.85–0.93

 0.680

 3

 1.16 1.15–1.19

 0.564

#### **4. Discussion**

The etiology of most benign and malignant uterine pathologies is hormone-dependent. Estrogen is the major mitogenic factor in the uterus and is responsible for tissue remodeling during the menstrual cycle. However, abnormalities to the endometrial tissue including hormonal disbalance, changes in tissue microenvironment, abnormalities in cytokine or growth factors expression, or increased production of ROS can result in carcinogenesis. Metalloestrogens may mimic estrogen activity and activate estrogen receptors and thus can also be increased in various endometrial pathologies.

In the presented study, we analyzed the trace elements concentrations in different uterine pathologies. We did not find any significant correlations between the selected microelements and endometrial cancer when compared to the control group of benign uterine pathologies/normal endometrium as a whole; however, we discovered some significant correlations only when we divided the control group into separate categories for each histopathological diagnosis. We found some correlations for Cd, Cu, and Fe serum concentration. Higher Cd serum concentrations were observed in patients diagnosed with endometrial cancer when compared with patients diagnosed with endometrial polyps (*p* = 0.002). Additionally, higher Cd and Cu concentrations were found in patients diagnosed with uterine fibromas when compared with endometrial polyps; however, the median concentrations of Cd and Cu were still lower in patients diagnosed with uterine fibromas than in EC patients. Moreover, patients diagnosed with endometrial polyps also had significantly higher Fe serum expression when compared with patients with normal endometrium. Elevated copper levels (both serum and tissue) have been found in multiple cancers [28]. As copper is a co-factor in redox reactions of enzymes participating in basic biological reactions required for cell growth and development including superoxide dismutase and cytochrome c oxidase, its increased amount may correlate with an increased need for tissue proliferation. As uterine fibromas are benign tumors which are also characterized with a rapid growth, this may be the reason for an increased trace elements concentration. Further studies are required to deepen the knowledge on the role of trace elements and their distribution in different uterine pathologies.

So far, only a few studies have evaluated metalloestrogen concentrations in endometrial cancer and resulted in conflicting reports. Due to the high importance of metalloestrogens and their unexplained role in uterine pathologies, we decided to conduct a study to further evaluate their distribution in patients with different uterine conditions. Atakul et al. [29] found associations between serum Cu and Zn. In accordance with their study, lower Cu, Zn, and Cu/Zn ratio was found in patients diagnosed with endometrial cancer than in control group. Cu concentration also inversely corelated with myometrial invasion. Yaman et al. [30] investigated trace metal concentration in different cancerous and noncancerous endometrial, ovary, and cervical tissue samples. The authors found increased Fe, similar levels of Cu, and lower levels of Zn in endometrial cancer patients when compared with controls. Additionally, Rzymski et al. [31] investigated metal accumulation in uterine tissue samples. Compared with normal endometrium, endometrial cancer, hyperplasia, and CIN samples revealed significantly increased levels of Cd, Pb, Cu, Mn, and Cu/Zn ratio. Both current and former smoking status were associated with significantly higher Cd and Pb levels. Additionally, endometrial polyps, when compared with histologically normal endometrium, showed increased median concentrations of Al, Cd, Ni, and Pb. There was no significant association for Cu/Zn ratio. The study showed no correlation between patients' age, menopausal status and the concentration of any of the investigated elements in endometrial tissue sample. Additionally, in our study, the menopausal status did not influence any of the assessed metalloestrogen levels. As there are still few data on the role of metalloestrogens and their levels in patients with different gynecological conditions, further research is needed to evaluate their significance. It would also be interesting to assess endometrium/uterine tissue metalloestrogen levels and compare their expression with serum levels.

The limitation of our study is that we did not ask the patients about any recent use of dietary supplements. As the supplements containing elements such as Zn, Cu, or Fe are widely accessible, their use might have influenced their blood concentration. However, as the supplements in Eastern Europe are still not as popular and accessible as in the western countries, we believe that their use was very limited. There is no obvious explanation for the differences observed between the studies in serum microelement concentration; however, the discrepancies may be caused based on the study size and patient characteristics. The differences may be also caused by the methods used for serum microelement analyzes, e.g., colorimetry, spectrophotometry, ICP, which each has different sensitivity and selectivity. In this study, we used ICP-OES—a very accurate method to analyze the trace elements concentration, which was also previously described in other trace elements analyses.

As a part of the study, we also conducted univariate analysis to evaluate the influence of patient characteristics and the assessed variables on the endometrial cancer risk. We confirmed that a high patient age, BMI, post-menopausal status and diabetes type 2 were significant risk factors of EC (see Table 5). Upon multivariate analysis, we found cadmium and cobalt levels to be associated with endometrial cancer risk. Higher Cd levels were found to be associated with increased endometrial cancer risk (OR 1.49, *p* = 0.0367), while above median Co levels were found to correlate with lower EC risk (OR 0.76, *p* = 0.0423). In a study by Rzymski et al. [31], the author found both current and former smoking status to be associated with significantly higher Cd and Pb levels. In our study, we did not further evaluate the correlation between cigarette use and trace element levels, as smoking was found not to be associated with endometrial cancer risk in our analysis; however, the study findings by Rzymski may be a potential explanation for this observation and require further analysis. Cobalt is an essential component of vitamin B12. Vitamin B12 is essential for the maintenance of DNA methylation, repair, synthesis, and thus for cell development and proper function. Accumulating evidence suggests the role of increased levels of folate and B-vitamins to have a role in cancer formation. Supplemental use of vitamin B12 intake was found to be associated with type 2 EC. However, the study did not include multivariable analysis. There is limited and still conflicting evidence regarding the influence of vitamin B12 and folic acid on cancer incidence. Further studies are required to explain their effect on cancer risk and cancer formation.

GSTs have a particularly important detoxification capability that protects against environmental and oxidative stress. Recent studies reported alternated GST expression to be associated with increased cancer risk [32]. As genetic polymorphisms can alternate enzyme expression, we decided to measure selected SNPs. We analyzed whether polymorphisms rs1695 in GSTP1 and rs224589 in SLC11A2 genes are associated with the risk of endometrial cancer. In our study, we found no significant association between GSTP1 and SLC11A2 polymorphisms and endometrial cancer prevalence.

A meta-analysis by Zhao et al. [33] tried to evaluate the associations between GSTP1 Ile105Val polymorphism and gynecological cancer susceptibility; however, the researchers found no significant associations with any genetic model even when accounting for cancer type, ethnicity, and smoking status. The study included a limited population as only two studies discussed endometrial cancer patients. A study by Ozerkan et al. [34] found no associations between GSTP1 polymorphism and EC risk in Caucasian population. On the other hand, a study by Chan et al. [35] revealed GSTP1 Ile(105)Val polymorphism to be associated with an increased risk of endometrial cancer. In our study, we found no associations between GTSP1 polymorphism and EC risk; however, as there are contradictory results, further research is needed.

This was the first study to analyze functional polymorphism in SLC11A2 gene in endometrial cancer patients. Previous studies described SLCA11A2 overexpression in breast carcinomas [22,26]. As endometrial cancer is also hormone dependent, we wanted to determine if its gene polymorphism affects the risk of EC. Even though we found no correlation between SLC11A2 polymorphisms and endometrial cancer risk, we found a positive correlation between SCL11A2 TT genotype and Cu concentration (*p* = 0.037). On the other hand, patients with TG genotype demonstrated lower Cd and Ni levels (*p* = 0.023 and *p* = 0.040, respectively). Further research is needed to evaluate the correlations between selected genotypes and trace elements levels due to the limited number of patients included in the analysis.

#### **5. Conclusions**

The study confirmed higher patient age, post-menopausal status, presence of diabetes type 2, and higher BMI as independent risk factors for endometrial cancer. Menopausal status did not influence metalloestrogen levels. High serum cadmium and low cobalt concentrations were found to influence endometrial cancer risk. None of the tested genetic polymorphisms (rs1695 in GSTP1 and rs224589 in SLC11A2) were found to be associated with endometrial cancer risk. However, GTSP1 and SLC11A2 SNPs may correlate with selected trace elements concentrations as high Cd and Ni serum levels were significantly associated with the SLC1A2 TG genotype and high Cd levels with GSTP1. Analyses of a more extensive study group should be performed to confirm our findings.

**Author Contributions:** Conceptualization, K.M., A.C.-P. and D.C.; methodology, D.C. and M.K.; software, D.C. and M.B.; validation, M.K., D.C. and A.C.-P.; formal analysis, K.M. and P.K.; investigation, K.M. and P.K; resources, K.M.; data curation, K.M., G.W. and P.K.; writing—original draft preparation, K.M.; writing—review and editing, K.M., A.C.-P. and D.C.; visualization, K.M.; supervision, A.C.-P., M.K. and D.C.; project administration, A.C.-P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of Pomeranian Medical University in Szczecin (protocol code KB-0012/27/2020 of 9 March 2020).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Data Availability Statement:** The data presented in this study are available on request from the corresponding author. The data are not publicly available due to ethical restrictions.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

