Association of the genetic variants (-794 CATT5-8 and -173 G > C) of macrophage migration inhibitory factor (MIF) with higher soluble levels of MIF and TNFα in women with breast cancer.

BACKGROUND
Functional variants -173 G > C (rs755622) and -794CATT5-8 (rs5844572) MIF gene have been associated with the risk in several types of cancer, as well as with the increase of soluble levels of MIF and TNFα. However, in previous studies contradictory and uncertain results have been presented on the implication of MIF polymorphisms with the association in cancer, specifically in breast cancer (BC). We investigated whether the variants are associated with the susceptibility to develop BC and the soluble levels of MIF and TNFα in women with BC from western Mexico.


MATERIALS AND METHODS
A total of 152 women with BC and 182 control subjects (CS) were enrolled in this study. The determination of genotypes -173 G > C and -794 CATT5-8 MIF polymorphisms was performed by PCR-RFLP and PCR, respectively. In addition, the soluble levels of MIF and TNFα in both studied groups were quantified by ELISA and MILLIPLEX assay, respectively.


RESULTS
The most frequent allele found in BC was the G (74.3%) and 6 (54%) in the variants -173G > C and -794 CATT5-8 , respectively, without significant differences in both groups. Nevertheless, the women with BC carriers -173*C and -794CATT7 have higher levels of MIF in comparison with CS. An increase of MIF (BC: 11.1 ng/mL vs CS: 5.2 ng/mL, P < .001) and TNFα (BC: 24.9 ng/mL vs CS: 9.9 pg/mL, P < .001) was found.


CONCLUSION
The functional variants of MIF are not genetic susceptibility markers for BC. Nevertheless, the alleles -173*C and -794CATT7 are associated with the increase of MIF circulating in women with BC.


| INTRODUC TI ON
Breast cancer (BC) is the leading cause of death in women and is defined as a heterogeneous disease of multifactorial origin, 1,2 according to data from the World Health Organization (WHO). There is a classification of BC defined by biomarkers by immunohistochemistry as estrogen receptor (ER), progesterone receptor (PR), receptor 2 of human epidermal growth factor (HER2), and Ki67 have been very useful to define molecular profiles, 3,4 which represent a guide for the diagnosis and treatment of patients. 4 The molecular subtypes are classified in Luminal A, Luminal B, HER2, and triple negative (TN). 5,6 BC is a multifactorial disease in which pathogenesis merges several factors like environmental, genetic, and immunologic, where the proinflammatory cytokines highlight. [7][8][9] Migration inhibitory factor is a pleiotropic cytokine that is involved in tumor pathology, 10 because participates promoting inflammation, angiogenesis, and metastasis, 11 besides MIF can regulate innate and adaptive immunity, it constitutively released by many cells including macrophages, B and T cells, dendritic cells, granulocytes, endocrine, endothelial, epithelial cells, 12,13 and also by tumor cells. 14 Furthermore, it has been observed that MIF has a positive feedback mechanism with the production of TNFα, in vitro. 15 In this regard, there are reports that mention that MIF promotes the tumor microenvironment through inflammation with the production of inflammatory cytokines including IL-1β, IL-6, and TNFα. 16 The MIF gene is encoded on chromosome 22 (22q11.2) in humans; it consists of 3 exons, 12,17,18 and it is composed of 115 amino acids with a molecular weight of 12.5 kDa, and its active form in the conformation homotrimer. 13,17 There are existing reports that establish functional variants associated with the increased expression the MIF (-173G > C and -794 CATT [5][6][7][8] ). 19 The promoter sequence of MIF has several specific binding sites for different transcription factors that regulate its expression, which include SP1, CREB, NFkB, Pit-1, AP-1, ICBP90, and HIF-1α. [20][21][22][23][24] The first functional variant of MIF is a single nucleotide polymorphism at -173 (G > C), which creates a binding site for protein activator-4 (AP4) that increases promoter activity of MIF. 21 The second site consists of the CATT repeats at -794 which repeats 5-8 times, creating a binding site for the transcriptional factor Pit-1, as well as ICBP90, which regulates the activity on this site. 24 Furthermore, two studies suggest the association between MIF polymorphism -173 G > C with the risk of developing BC in the Chinese population. 18,25 MIF protein is related to the development and progression of the tumor through the production of proangiogenic factors such as VEGF. 14 Besides, MIF when forming the complex with CD74 and CD44 (also, interacts with CXCR2, CXCR4, and CXCR7 chemokine receptors) 26 activates different signaling pathways like ERK1/2, MAPK, PI3K/AKT, and NFkB, involved in proliferation, differentiation, survival, and inflammation. 17,26 Likewise, the increased soluble levels of MIF have been associated in patients with prostatic, colon, lung, and skin cancer, as well as BC. 27 However, it is still unknown if this functional variant of the MIF gene and the soluble levels of MIF and TNFα are associated with BC in the western Mexican population. This study aimed to associate the functional variants of the MIF gene (-173G > C and -94 CATT [5][6][7][8] with the soluble levels of MIF and TNFα in women with BC.

| MATERIAL S AND ME THODS
For this cross-sectional study, 152 women were diagnosed histopathologically with BC belonging to the "Ella" Binational Breast Cancer Study and 182 women as control subjects (CS); all the subjects were recruited from Mexican-mestizo population. All women with BC and CS participants received a written informed consent, approved by the ethics committee from the Universidad de Guadalajara (CI-9708). The participation of CS in this study was voluntary, and both groups signed a written informed consent. In addition, this study was performed according to the ethical principles for experiments involving humans stated on the Declaration of Helsinki. The women with BC, under pharmacology treatment with corticoids or chemotherapy, were not included in the present study.

| Identification of molecular subtypes
The intrinsic molecular-like subtypes were identified by immunohis-

| Quantification of MIF and TNFα plasma levels
The plasma was collected following centrifugation and stored immediately at −80°C until analysis. MIF soluble levels were measured by commercial ELISA kits (BioLegend® Legend MaxTM, Human active MIF, Cat. No. 438408). The MIF assay limit detection was 8 pg/mL.
The optical density was immediately determined, using a microplate reader set to 450 and 570 nm, and TNFα was measured in a multiplex assay (MILLIPLEX®MAP Cat. No. HCYTOMAG-60K) according to the manufacturer´s instructions.

| Statistical analysis
The statistical analysis was performed using STATA software (v9.2) and GraphPad Prism (v5.0) software. Differences in characteristics between groups were analyzed with chi-square test, Student's t test (parametric variables), and Mann-Whitney U test and Kruskal-Wallis test (nonparametric variables) (data presented as median and 25th to 75th percentiles) were used. The Hardy-Weinberg equilibrium test and genotype and allele frequencies were calculated by chi-square test or Fisher's exact test, when it was necessary. Odds ratios (OR) and 95% confidence intervals (95% CI). In this study, the sample size was calculated according to the minor allele frequency of the -173 G > C and -794 CATT 5-8 MIF gene polymorphism, using the Kelsey formula. A P value < .05 was considered to be statistically significant.

| RE SULTS
The clinical and demographic characteristics of BC are summarized in Table 1. The women with BC had a mean of 55 years and postmenopausal state (70.6%), and also, around 83% of the women with BC were classified in clinical stage II-IV, in agreement with the distribution present in Mexico. According to the classification with molecular subtypes, the Luminal A and Luminal B are the most predominant in this study with 70.2% of the cases, while HER2 and TN presented the 16.7% and 13.2%, respectively. In addition, the body mass index (BMI) of the majority of women with BC was overweight and obese with 37.3% and 33.3%, respectively, and also presented high Ki67 (>14%).

| The soluble levels of MIF and TNFα in women with BC and CS
The soluble levels of MIF were increased in BC (11.1 ng/mL) compared to CS (5.

| MIF gene -173 G > C polymorphism in the women with BC and CS
The functional variant analyzed was in Hardy-Weinberg equilibrium in the CS (P = .94). The distribution of -173 G > C MIF polymorphism in women with BC and CS is shown in Table 2. We found that when comparing the genotypes of women with BC and CS, the significant differences were not shown. We observed a higher frequency of the G/G homozygous genotypes, followed by G/C heterozygote genotype.
We did not find any statistical association with -173 G > C MIF gene polymorphism in BC. Also, by applying a genetic model of dominant inheritance, the significant association between MIF -173 G > C polymorphism and the risk of BC for Mexican population (G/G vs GC + CC: OR = 1.13, 95% CI = 0.72-1.78 P = .56) has not found. In addition, we found that the most frequent allele for women with BC was G (74.3%), followed by C (25.7%) ( Table 2).
On the other hand, to determine whether this polymorphism is associated with soluble levels in women with BC, we compared the soluble levels of MIF and TNFα in women with BC according to -173G > C MIF. Nevertheless, significant differences were not observed, as well as by applying the dominant model. But, a tendency to the increase of soluble levels of MIF in C/C genotype (20.0 ng/mL) ( Figure 2A) was observed, rather than G/C and G/G genotypes. Also, to compare the levels of MIF stratified by -173G > C MIF genotypes between women with BC and CS, we found that MIF is increased for each genotype in BC, while in CS shown similar levels in all the genotypes ( Figure 2B)

| MIF gene -794 CATT 5-8 polymorphism in the women with BC and CS
With respect to the functional variant -794 CATT 5-8 , the Hardy-Weinberg equilibrium in the CS was also performed (P = .99). The distribution of -794CATT 5-8 MIF polymorphism in women with BC and CS is shown in Table 3. We found no significant differences when comparing the genotypes of women with BC and CS. We ]. Furthermore, we found that the most frequent allele for women with BC was 6 (54%), followed by 7 (24%) ( Table 3). We compared the genotypic and allelic frequencies of -794 CATT 5-8 MIF polymorphisms; however, we did not find significant differences.
Regarding to the association in this polymorphism with the soluble levels of MIF and TNFα in women with BC significant differences by genotype were not shown, but we observed higher MIF levels in homozygous 7,7 carriers (18.8 ng/mL) in women with BC ( Figure 3A). These results show that women with BC with allele 7 have higher soluble levels of MIF, which suggest that the allele -794 CATT 7 similarly to 173*C could increase the MIF circulating in BC.
Subsequently, we observed that the soluble levels of TNFα were higher in the women with BC in comparison with CS (P < .001).
Likewise, no significant differences were observed with the soluble levels of TNFα according to genotypes (data not shown).
Furthermore, the soluble levels of MIF and TNFα were analyzed according to the molecular subtype; however, significant differences were not observed (data not shown).

| D ISCUSS I ON
In the present study, we have investigated the association of the functional variants of the MIF gene (-173G > C and -794 CATT [5][6][7][8] with the soluble levels of MIF and TNFα in women with BC of western Mexico. BC is a major lethal cancer in female. 30  repeat (-794 CATT) both in the promoter region. 11 The involvement of MIF alleles in the development of cancer was first associated and described in prostate cancer. 19 As far as we were concerned, this is the first study that evaluates the allelic and genotypic frequencies of -173 G > C and -794   increases the risk of BC especially in older women of Chinese population. 25 Also, other study suggests that MIF -173G > C polymorphism was associated with the participation of developing cancer, preponderantly in prostate cancer. 23 For the frequencies of -794 CATT 5-8 , MIF polymorphism was similar to other studies that evaluated autoimmune diseases in patients from western Mexico, 32 but there is no existing report that determines the distribution of this polymorphism in cancer from this population.
In a study of Wu et al, evaluating the -794 CATT 5-8 MIF polymorphism in early stage cervical cancer (ESCC), they found that -794 CATT 7 allele is associated with ESCC. 27 Nonetheless, in our results we did not find association between the functional -173*C and -794 CATT 7 MIF gene and women with BC in from western Mexico. These differences could be attributed to the criteria for inclusion and the sample size, and they can be influenced by the racial differences between populations. 32 On the other hand, it has been reported that the presence of the C allele creates a binding site for the AP-4 transcriptional factor and its presence is associated with greater expression of the gene, 37 similarly to the -794 CATT 7 which generates a binding site for the transcriptional factor Pit-1, with the consequently increased expression of MIF gene. 19 In addition, another transcriptional factor involved in the regulation of MIF gene was ICBP90, a protein that interacts directly with MIF -794 CATT [5][6][7][8] and that favors transcription of the MIF. 24 Additionally, the functional variants of MIF are associated with overexpression of MIF that participates in tumor growth promoting angiogenesis and metastasis; 10,26 moreover, MIF can induce secretion of bFGF, VEGF, and IL-8 when activates the MAPK, PI3K signaling pathways, which lead to the increased secretion of these proangiogenic factors that are necessary for blood vessel formation. Also, under hypoxic conditions, that is a characteristic of the tumor, HIF-1α is expressed and induces the production of MIF playing a role in tumor angiogenesis. 11 In addition, HIF-1α downregulates the expression of e-cadherin, responsible for the formation of focal adhesion complex, which favors the metastasis. Similarly, MIF induces the expression of matrix metalloproteinase which degrade the basal membrane which enhances the metastatic process. 11 Furthermore, our results confirm previous reports that establish that TNFα is increased in women with BC. 16,35,38 The high BMI could be a risk factor that contributes to inflammatory conditions as previously was reported in BC. 25 Some reports establish that overweight and obesity are states of chronic inflammation, which contributes to the promotion of the production of inflammatory cytokines that maintain an inflammatory tumor microenvironment. 16 We have tested the possible association of MIF and TNFα, but we have not found significant correlation. A possible explanation for this could be the lack of association of MIF and TNFα which play a role in the site of tumor where the inflammatory process is exacerbated; therefore, more studies are necessary for the determination of these cytokines in situ.
MIF is considered a pleiotropic cytokine that participates in innate and adaptive immunity regulating the inflammation, also considered as the link between chronic inflammation and cancer. 26 However, the association of MIF with the development of BC is not conclusive.
To our knowledge, the results presented in this study provide

| CON CLUS IONS
The functional variants of the MIF gene (-173 G > C and -794 CATT 5-8 ) did not shown to be genetic susceptibility markers in BC in women of western Mexico population. Furthermore, the soluble levels of MIF and TNFα in women with BC were increased in comparison with CS, while the presence of the alleles -173*C and -794CATT 7 shown a tendency of increase of circulating levels of MIF in women with BC.

E TH I C A L A PPROVA L
This study was conducted conforming to the Declaration of Helsinki, and the research was approved by the ethical investigation, committee from each of the hospitals, and Universidad de Guadalajara (CI-9708).

CO N S ENT TO PA RTI CI PATE
Informed consent was obtained from each of the participant before being enrolled in this study.