Fecal expression of Escherichia coli lysine decarboxylase (LdcC) is downregulated in E-cadherin negative lobular breast carcinoma

Breast cancer is characterized by oncobiosis, the abnormal composition of the microbiome in neoplastic diseases. The biosynthetic capacity of the oncobiotic flora in breast cancer is suppressed, as suggested by metagenomic studies. The microbiome synthesizes a set of cytostatic and antimetastatic metabolites that are downregulated in breast cancer, including cadaverine, a microbiome metabolite with cytostatic properties. We set out to assess how the protein expression of constitutive lysine decarboxylase (LdcC), a key enzyme for cadaverine production, changes in the feces of human breast cancer patients ( n 5 35). We found that the fecal expression of Escherichia coli LdcC is downregulated in lobular cases as compared to invasive carcinoma of no special type (NST) cases. Lobular breast carcinoma is characterized by low or absent expression of E-cadherin. Fecal E. coli LdcC protein expression is downregulated in E-cadherin negative breast cancer cases as compared to positive ones. Receiver operating characteristic (ROC) analysis of LdcC expression in lobular and NST cases revealed that fecal E. coli LdcC protein expression might have predictive values. These data suggest that the oncobiotic transformation of the microbiome indeed leads to the downregulation of the production of cytostatic and antimetastatic metabolites. In E-cadherin negative lobular carcinoma that has a higher potential for metastasis formation, the protein levels of enzymes producing antimetastatic metabolites are downregulated. This ﬁ nding represents a new route that renders lobular cases permissive for metastasis formation. Furthermore, our ﬁ ndings underline the role of oncobiosis in regulating metastasis formation in breast cancer. metabolites in suppressing or preventing breast cancer metastasis.

Changes to the microbiome can affect the immune functions of the host [15] and the microbes can modulate the behavior of the host by secreting pleiotropic metabolites. These metabolites get to cancer cells through the systemic circulation [10]. The majority of the studies report lower diversity in the breast cancer oncobiome as compared to the healthy eubiome, resulting in a restricted biosynthetic capacity [3,6,7]. This lower biosynthetic capacity probably leads to lower production of bacterial metabolites in both the serum and the tumor. A number of antineoplastic bacterial metabolites were identified in breast cancer, including short chain fatty acids [10], lithocholic acid [7,16] and cadaverine [6].
There is an ample set of metagenomic studies in breast cancer cases [10]. However, to our best knowledge, no protein studies are available due to difficult sample preparation and lack of antibodies for analysis. In the current study, we assessed the expression of a bacterial enzyme, LdcC, which produces the cytostatic bioactive metabolite, cadaverine [6]. The aim of this study was to assess the fecal protein expression of E. coli LdcC in breast cancer patients.

Patient cohort
Fecal protein expression was assessed in a female breast cancer patient cohort composed of 35 participants with a mean age of 57 years. Samples from stage I-III and Nottingham grade 1-3 patients were used in the study. The average stage and grade were not statistically different amongst the groups. Fecal samples were obtained from patients with no special type (NST) and lobular tumors; tumors of other histological types were excluded.
The collection and biobanking of feces were authorized by the Hungarian national authority (ETT). Patients and healthy volunteers meeting the following criteria were excluded from the study: 1) has a previous history of breast cancer or had been operated due to neoplasia, 2) has a disease of unknown origin, 3) has a chronic contagious disease, 4) had contagious diarrhea 6 months prior to enrollment, 5) taken antibiotics within the 6 months prior to enrollment, 6) had chemotherapy, biological therapy, or immunosuppressive therapy 6 months prior to enrollment, 7) used intravenous drugs 12 months prior to enrollment, 8) had piercing, tattooing, acupuncture, or other endangering behavior or action 12 months prior to enrollment, 9) exposure to an allergen to which the enrolled individual had been sensitized to, or 10) underwent colonoscopy 12 months prior to enrollment. The first morning feces was sampled; samples were frozen and deposited in the biobank within 2 h after defecation. Samples were stored at À70 8C until subsequent use. We obtained informed consent from study participants. The patient's routine pathological data were assessed in the study and were compared to the fecal expression of E. coli LdcC.

Statistical analysis
All data are represented as average ± SEM. Statistical tests are mentioned in figure captions. For statistical analysis, Graphpad 8.1 software was used. The receiver operating characteristic (ROC) curve was re-calculated using R (version v 3.6.3 (2020-02-29)) [23] and the pROC package [24].

Fecal expression of LdcC is low in the E-cadherin negative, lobular subtype of breast cancer
First, we assessed the fecal protein expression of E. coli LdcC as a function of the histological subtype of breast cancer. We found that the fecal expression of E. coli LdcC enzyme was lower in the lobular subtype compared to the invasive carcinoma of no special type (NST) of breast cancer (Fig. 1A). ** indicates statistically significant differences between groups at P < 0.01 Lobular breast carcinoma is characterized by lower or absent E-cadherin expression [25]. Furthermore, the absence of E-cadherin is a predictive factor for metastasis formation, and therefore, predicts poorer clinical outcomes [25][26][27]. We compared the LdcC expression in Ecadherin positive and negative tumors. The fecal expression of E. coli LdcC was significantly lower in E-cadherin negative cases as compared to E-cadherin positive cases (Fig. 1B).
The striking difference in the fecal expression of E. coli LdcC in E-cadherinþ versus E-cadherinÀ cases prompted us to assess whether it would be possible to discriminate between E-cadherinþ and E-cadherinÀ cases using LdcC expression. To that end, we performed a ROC analysis. According to the ROC analysis, fecal E. coli LdcC expression is a good assay to classify E-cadherinþ and E-cadherinÀ cases (Fig. 2). The area under the curve for the ROC analysis was 0.8276 (Std error: 0.0864, 95% confidence interval [CI]: 0.6583-0.9969). The threshold of the curve was 0.642 (corresponding specificity 83.33% (CI: 50-100) and sensitivity was 72.41% (CI: 55.17-86.21)). The P value for the curve was P50.0126.

DISCUSSION
All previous work on the breast cancer oncobiome applied nucleic acid-based techniques [10]. The bacterial metabolome was not assessed in dedicated studies, only tangentially in general metabolomic studies [7,28,29]. General bacterial proteomic studies are missing. Due to these caveats, we do not have a comprehensive picture of the metabolic function of the oncobiome and its pathophysiological role in breast cancer. Our aim was to provide an initial, rudimentary study on fecal bacterial protein expression in breast cancer patients, as a proof of principle study. The major technical drawback we faced was the lack of antibodies against relevant bacterial enzymes in diverse species. In this study, we used an antibody specific to the LdcC enzyme that produces cadaverine, which has cytostatic properties in breast cancer [6]. The anti-LdcC antibody was specific for E. coli and there were no available antibodies against the same enzyme in other species. LdcC was downregulated at the DNA level, when comparing breast cancer patients to healthy controls [6]. In support of this observation, breast cancer is characterized by suppressed diversity [10] and suppressed biosynthetic capacity of the gut microbiome [3,6,7]. Fecal E. coli LdcC enzyme expression was downregulated in lobular E-cadherin negative breast cancer cases. E-cadherin is a cell adhesion molecule, which is crucial in the suppression of invasion [26]. Epigenetic reprogramming or mutations that downregulate E-cadherin expression support tissue invasion and metastasis formation [25][26][27]30]. Re-expression of E-cadherin reprograms mesenchymal cells towards epithelial morphology [30]. Lobular breast cancer has a higher potential for metastasis formation, marked by a higher frequency of peritoneal metastases (e.g., Krukenberg tumors). Increased peritoneal metastases is probably linked to low or absent E-cadherin expression. The finding that fecal E. coli LdcC expression was lower in E-cadherin negative cases is in good agreement with the observation that oncobiosis supports metastasis formation in breast cancer [6,7,15]. Cadaverine, produced by LdcC, suppresses metastasis formation [6].
Earlier, metagenome-based reports have suggested that the breast cancer oncobiome changes as a function of stage [6,7], estrogen, and HER2 receptor status [1,2,10]. Therefore, the association of oncobiome changes with lobular breast cancer or with E-cadherin status is not surprising. However, to the best of our knowledge, breast cancer oncobiome changes have not associated with E-cadherin expression in breast cancer.
What do these data suggest to us from a functional perspective? The literature suggests that the biosynthetic or "bioconversion" capacity of the oncobiome in breast cancer is drastically suppressed [10,[31][32][33][34][35]. Bacterial estrogen reactivation, through the bacterial estrobolome, is upregulated in breast cancer patients. Increased estrogen reactivation and reuptake supports the proliferation of breast cancer cells [3,4,33,35]. Cytostatic bacterial metabolites were also identified, such as short chain fatty acids [10], lithocholic acid [7,16,36] and cadaverine [6]. LdcC plays role in cadaverine biosynthesis. LdcC is downregulated in E-cadherin negative cases that have a higher risk for metastasis formation. These observations support the idea that oncobiotic transformation in breast cancer downregulates the production of cytostatic and antimetastatic metabolites and, hence, supports metastasis formation in humans [6,7,15]. The bacterial metabolome-driven metastatic events are likely to be parallel mechanisms to the pathways regulated by the loss of E-cadherin or other adhesion proteins.
We are approaching a better understanding of the pathological role of the oncobiome and the oncobiotic transformation in breast cancer. This study provides evidence that the low bacterial production of cytostatic/antimetastatic metabolites in human breast cancer supports metastasis formation in breast cancer. Furthermore, the results of the DNA-based studies suggesting differences in the microbiome in subtypes of breast cancer [1,2,6,7,35] can be translated into changes in the proteome and, most probably, to changes in the bacterial metabolome as well. From a therapeutic perspective, these data support the applicability of bacterial metabolites in suppressing or preventing breast cancer metastasis.

PERSPECTIVES
Oncobiosis characterizes most cancers [37,38]. Breast cancer oncobiosis affects the distal gut [1][2][3][4][5][6][7], oral [9], urinary [9] and the breast's own microbiome [8]. Apparently, changes to these microbiome compartments can increase the risk for developing breast cancer. Breast cancer cases in women peak around the menopause, hence, aging is a risk factor for breast cancer. Aging comes along with dysbiosys [39,40] suggesting a likely causative relationship between changes to the microbiome and increased risk to breast cancer. Aging-associated, low grade inflammation is held responsible for the deterioration of multiple organs and organ systems [39][40][41][42][43]. Inflammation may in one hand damage cells and lead to mutations and cellular dysfunction, eventually to cell death [44]. Furthermore, constant low grade inflammation may be a pathway for the aging-associated simplification of the microbiome [39]. These changes are then translated to changes to the microbial metabolome [45] that may in turn give an agerelated pathway for breast carcinogenesis.