Prevalence of Bovine leukemia virus in Egyptian women’s breast cancer tissues

ABSTRACT Bovine leukemia virus (BLV) infects cattle around the world and presents in commercial bovine products. Several studies have suggested that BLV may play a role in breast oncogenesis. This study aimed to determine the risk factors for BLV infection and the presence of the BLV gene in breast tissue. A total of 100 formalin-fixed and paraffin-embedded breast tissue specimens were collected (50 confirmed cancer samples and 50 negative samples). DNA was extracted from breast tissues, and nested PCR was done to detect the tax gene of BLV. Immunohistochemical staining for ER, PR, Her2/neu, and Ki-67 was applied. Bovine leukemia virus DNA was detected in 16% of breast cancer tissue samples and in only 4% of tissue samples without cancer. No statistically significant association was reported between the presence of BLV and tumor grade, stage, or LN metastasis. A high percentage of positivity (100%, 88%, and 75%) was reported for ER, PR, and Ki-67 in BLV-positive cancer samples, respectively. There was a significant association between the BLV tax gene and breast cancer in Egyptian women. Thus, there is a need to control BLV transmission to humans from cattle, and further investigations about its risk related to public health are highly required.


Introduction
Breast cancer is responsible for 33% of cancerrelated morbidity and mortality in Egypt [1].There are multiple risk factors for developing human breast cancer, and the role of infectious agents in the development of this disease is of great interest [2].About 15-20% of cancers are associated with viral infection [3].Some viruses have been implicated as potential risks for breast cancer development, such as the murine mammary tumor virus (MMTV), the epstein-Barr virus (EBV), and the human papillomavirus (HPV) [4].
Comparative analysis research in breast tissue found a high odd ratio, 4.01, that strongly supported the idea that Bovine leukemia virus (BLV) could be a risk factor for breast cancer [5,6].The viral biomarkers were detected in the human breast tissues, such as gene segments, antibodies against BLV, and different viral proteins [7,8].
Bovine leukemia virus is a delta-retrovirus that infects cattle, and most of the infected cattle remain healthy carriers.Only 1-5% develop malignant B-cell lymphoma [9].The virus is widespread throughout the world, with a prevalence rate between 10 and 90% in cattle [6], whereas in Egypt, the prevalence rate is between 16 and 21% in different governorates [10].BLV is transmitted to humans through the consumption of raw meat and milk products [11,12].
The BLV encodes a tax protein that has the capability to regulate various cellular pathways, act as a cell division activator and inhibit DNA repair that may lead to cellular transformation [13].
Analysing the relationship between the viral existence and breast cancer biomarkers (i.e.estrogen (ER) and progesterone (PR) receptors, cell division marker Ki67, and an excessive expression of human epidermal growth factor receptor 2 (HER2) protein) could give a background of the tumor microenvironment.It may perhaps favor an active viral transcription stage that is probably linked with BLV infection [3].These hormonal and tumor markers can be detected by immunohistochemistry, and they are crucial in determining the prognosis and treatment of breast cancer patients [14].
The aim of the study was to determine the risk factors for BLV infection in humans and to clarify the possible relationship between BLV infection and breast cancer development in Egypt through analyzing its association with hormone markers IHC expression, proliferation index and several clinico-pathological prognostic parameters in breast.

Study design
This retrospective case control study was conducted over a period of 24 months, starting on June 2020 and ending in May 2022, at the pathology laboratory of the Oncology Center, Mansoura University (OCMU) and the Microbiology Department, Faculty of Medicine, Mansoura University, Egypt.

Samples
One hundred formalin-fixed paraffin-embedded (FFPE) breast tissue blocks were collected from the archive of the pathology laboratory at OCMU on the basis of the pathological diagnosis, where the patients were diagnosed according to the results of the routine morphological and immunohistochemical examination of the excised breast tissue samples.
The samples were enrolled in two groups: Group I (n = 50) confirmed cancer mammary tissue samples and Group II (n = 50) benign mammary lesions tissue samples.The selection of the cases was based on the availability of full clinical and pathological data, which were obtained from the institute database for the patients' medical records retrospectively (age of the patients, occupation, living area, educational level, size of tumor, histopathological type of breast cancer, grade, stage, and presence of lymph node metastasis).

Extraction of proviral DNA
From each sample, out of 100, five sections were prepared with a 5-10 μm thickness and a surface area of 250 mm 2 .Then DNA extraction was done by QIAamp DNA FFPE Tissue Kit, Germany, depending on the manufacturer's instructions.

Detection of G3PD gene by conventional PCR
The glyceraldehyde-3-phosphate dehydrogenase housekeeping gene was amplified by (5′ GAGTCAACGGATTTGGTCGT-3′) and (5′ TTGATTT TGGAGGGATCTCG-3′) primers [15].It was used as an internal control for DNA amplification from the samples before amplification of bovine leukemia virus genes.
The agarose gel was examined by using a UV transilluminator for band visualization at 237 bp for the G3PD gene and 113 bp for the tax gene.

Detection of ER, PR, Her-2/neu, and Ki-67 markers by immunohistochemical staining
IHC was conducted according to the manufacturer's datasheet on approximately 4 µm-thick paraffin sections on heat -fixed positively charged slides.Deparaffinization, rehydration, and epitope exposure were carried out using 0.01 M citrate buffer (pH 6.0) in the microwave for 10 minutes.Then, incubation with 3% hydrogen peroxide for 10 minutes to block the activity of endogenous peroxidase.All sections were rinsed with phosphate buffered saline (PBS) and incubated with the primary antibodies directed against: Dako REAL™ EnVision™kit, Denemark and estrogen (ER), progesterone (PR), human epidermal growth factor 2 (Her-2/neu) receptors, and Ki-67 marker-specific antibodies obtained from Roche Diagnostics K.K., Japan at room temperature for hour.The standard avidin-biotinperoxidase procedure was applied using diaminobenzidine (DAB, incubation for 5 minutes) for visualization.Then, counterstain with hematoxylin (30 seconds), and finally cover with coverslips.
The scoring of ER and PR receptor immunoexpression was determined based on both the intensity and percentage of tumor cells showing positive nuclear staining according to the Allred scoring system [16].Regarding Her2/neu immunoexpression, only membranous staining was considered according to the guidelines of the American Society of Clinical Oncology-College of American Pathologists [17].The Ki-67 score depends on the percentage of the total number of positive tumor cells for nuclear staining [18].

Mansoura Faculty of Medicine Institutional
Research Board (MFM-IRB); code: MD.19.10.236 approved the study protocol.Written executive permission from the manager of the Oncology Center Pathology Laboratory, Faculty of Medicine, Mansoura University.The study was conducted on archived material obtained from paraffin tissue blocks and maintained at the pathology laboratory in the study setting.Confidentiality and personal privacy were respected at all levels of the study.

Statistical analysis
The SPSS program version 26 (SPSS Inc.Chicago, IL, USA) was used to analyze the data.Numbers and percentages were used to describe qualitative data.Chi-square tests were used to determine associations between categorical variables, whereas Fisher exact and Montecarlo tests were used when the predicted cell count was < 5.For normally distributed data, continuous variables were given as mean ±SD (standard deviation), and the two groups were compared using an independent t test.P value of < 0.05 was considered statistically significant.

Results
All breast tissue specimens included were from females.Group (I) enrolled fifty FFPE breast cancer tissue specimens from patients between 29 and 76 years, 51.68 ± 13.16, and group (II) enrolled fifty FFPE breast tissue with benign lesions from patients between 23 and 76 years, with a mean of 48.60 ± 8.26.All extracted DNA samples were amplifiable for glyceraldehyde-3-phosphate dehydrogenase by conventional PCR, which verified the quality of the extracted DNA samples, and these samples were used to detect the BLV Tax gene by a nested PCR assay (Figures 1, 2).

Prevalence of BLV in breast tissue specimens
The bovine leukemia virus tax gene was detected by nested PCR in 8 (16%) and 2 (4%) samples of groups I and II, respectively.There was a statistically significant difference between groups I and II (P value = 0.04).BLV infection can increase the risk of breast cancer development five times (OR = 4.6).

Breast specimens' histopathological types
Group I enrolled 46 specimens with invasive ductal carcinoma (IDC), and the rest four specimens had invasive lobular carcinoma (ILC).
A non-significant difference was detected between BLV DNA and histopathological type in group I.However, six positive BLV samples from group I were IDC, 75%, and only 25% were ILC.Invasive ductal carcinoma showed the highest prevalence of the BLV gene.Group II contains 13 samples of fibroadenoma (two samples were positive for BLV DNA), 13 samples of fibroadenosis, 11 samples of chronic nonspecific inflammation, 6 samples of fibrocystic disease, 5 samples of mammary duct ectasia, and 2 samples of usual ductal hyperplasia.

Effect of bovine leukemia virus infection on cancer breast size, tumor grade, cancer stage, and lymph node metastasis
Although no statistically significant association was detected between BLV DNA and tumor size, six positive BLV DNA samples had tumor sizes between 2 and 5 cm (T2), and only 2 were >5 cm (T3).
There was no statistically significant association between the presence of BLV DNA and tumor grading, cancer staging, or LN metastasis, as shown in Table 1.

Detection of hormonal receptors and Ki-67 by immunohistochemistry and its relation to bovine leukemia virus infection (Figure 3):
There was no association observed between tumor prognostic biological markers (ER, PR, Her-2/neu oncoprotein) and the presence of BLV DNA in tissues (P values = 0.15, 0.74, and 0.69, respectively).There was no statistical significance observed between the expression of the proliferative marker Ki67 and the presence of BLV DNA (P = 0.28).However, a high percentage of BLV-positive samples were positive for ER (100%), and PR (88%), and had a relatively high proliferative index (75%).

Assessment of risk factors for BLV infection in human
Regarding the risk factors for BLV infection, statistically significant differences correlate the BLV infection and people from rural areas than urban, and those with homebased activities than other occupational activities, as shown in Table 2.

Discussion
In this study, the bovine leukemia virus tax gene was detected through nested PCR in 16% of breast cancer samples and 4% in the control group.There was a statistically significant difference between groups I and II (P value = 0.04).Furthermore, BLV infection is suspected to increase the risk of breast cancer development by five times (OR = 4.6).Tax protein is thought to have an essential role in the leukemogenesis induced by BLV, and nested PCR is an accurate, sensitive, and reliable method for tax gene amplification [11,19].Higher figures of 22.36% in Argentina and 24% in Iran were reported by Lendez et al. [20] and Khalilian et al. [8], respectively.Another study carried out in South Brazil found a significant difference in detection of BLV DNA in breast cancer tissue (30.5%) when compared with the non-neoplastic samples (13.9%) and reported that BLV infection increases the risk of breast cancer development by three times [21].Moreover, in Pakistan, breast cancer tissue samples were positive for BLV in 26.8%, whereas in negative breast cancer samples, they were 12.5% [11].Also in the USA and Australia, BLV has been accused of causing mammary cancer 3 and 4.7 times, respectively [22,23].A meta-analysis study that analyzed nine case-control studies to detect the association between BLV infection and breast cancer showed that BLV infection increased the risk of breast cancer by 2.57 times [24].
Controversially, a study in Japan reported that there was no evidence linking the presence of BLV DNA with the development of human breast cancer [12].This was parallel to a Chinese study that concluded the same [25].
Among the suspected causes of this variability, Amato et al. [26] reported that there are numerous factors that may be responsible for this variation among studies.The most significant of all may be the seroprevalence of BLV infection in the cattle population.In Egypt, several studies have been conducted to detect the seroprevalence of BLV among dairy cattle, which was 16 and 21% [10,27,28].In the USA, 84% and 39% of dairy and beef cattle were infected with BLV, respectively, and in Iran, 35% of dairy cattle were reported to be BLVpositive [8,29].Indeed, there were geographic variations in the association between BLV and breast cancer.Some nations, such as Spain, Switzerland, Finland, the Netherlands, Poland, Denmark, and Estonia, are free of BLV.On the other hand, the USA, Italy, Brazil, Canada, Portugal, Belarus, Egypt, Colombia, Japan, Taiwan, Iran, and Turkey were positive [8].
Other suspected causes of this variability Ramiz et al. [30] reported that there are many factors that might be responsible for this variation between studies.The most significant of all could be the heterogeneity of study populations, with a higher probability for high raw milk consumption as it was proposed that consuming beef and cow's milk can spread BLV to people.Moreover, the different genetic backgrounds of the study population, age, lifestyle, environment, exposure to radiation, and breastfeeding may lead to variability.Also, different techniques with different sensitivities and sample sizes used to detect BLV DNA may lead to incomparable results [8].In this study, the quality of extracted tissue DNA was successfully confirmed by amplification of the gylceraldehyded-3-phosphate dehydrogenase gene.Also, the BLV tax gene was chosen since oncogenesis is known to be mediated by this gene.Additionally, the genomic region occupied by tax is substantially preserved with little mutation in sequence and tends to remain detectable even in advanced stages of the disease when other areas of the viral genome are deleted [31,32].So, the inserted BLV tax gene into the host genome offers a reliable target for determining if mammary cell infection is present [8].
In this study, there was no significant association between BLV DNA and histopathological type; however 75% of positive BLV samples were IDC and 25% were ILC.In correspondence to this result, Khan et al. [11] detected that IDC had the highest prevalence (64% of the BLV positive samples).Also, Olaya-Galán et al. [3] reported that among breast cancer samples that were positive for the BLV, 45% were IDC, 36% were other malignancies (sarcoma, mixed carcinoma, phyllodes tumor, and invasive poorly differentiated carcinoma), and 18% were ILC.On the other hand, Buehring et al. [15] found the virus localized to the secretory epithelium of the breast equally in the ducts and lobules.
The correlation between breast cancer histopathological type and the presence of the virus in that tissue gives a probability that the virus may favor a certain subtype of breast cancer [3].
A BLV gene was found in a study done in Australia both before and after the onset of breast cancer.These findings suggested that viral DNA may have been found in this tissue years before the development of the disease [23].
In the current study, two samples of the control group diagnosed as fibroadenoma lesions were positive for BLV DNA.In agreement with this result, Baltzell et al. [33] and Khan et al. [11] detected that fibroadenoma had the highest prevalence of BLV infection among negative breast cancer tissue samples.
In the present study, although all patients that included their samples showed a high grade and stage, there was no statistically significant association between the presence of BLV DNA and tumor size, grade, stage, or lymph node metastasis, and this was in agreement with Schwingel et al. [21].On the other hand, Michaelson et al. [34] stated that tumor metastasis to lymph nodes increases when the tumor size increases.In the current study, 75% of patients had tumor ranging from 2 to 5 cm, which was correlated to high Ki67, which contributed to high metastasis to the draining lymph nodes.Also, Schwartz et al. [35] reported that 98% of cancer patients with BLV infection were ≥ grade 2 with moderate to high Ki67.
The present study measured the expression of the proliferative marker Ki-67, and there was no statistical significance observed between the proliferative rate and the presence of BLV DNA (P = 0.282); however, it is essential to highlight that the majority of the virus-positive samples (n = 6/8) had a moderate to high proliferative index (n = 6/8).
In Argentina, Lendez et al. [20] found a significant association between BLV and the high Ki67 biomarker (72%), proposing that BLV may have a role in early phases of cancer growth.
In this study, there was no association observed between tumor biological markers such as ER, PR, or HER-2 oncoprotein and the presence of BLV in tissues.However, an in vitro study provided evidence that BLV infection and its transcription rates are accelerated by the presence of progesterone and glucocorticoid hormones and reported that this virus is hormone-responsive [36].So, even if there was a non-significant association between the expressions of tumor markers and the presence of BLV, it is essential to highlight that the majority of the samples that tested positive for BLV also tested positive for hormone receptors with a high percentage (8/8 for ER and 6/8 for PR).
Other studies that analyzed this correlation found no correlation between the cancer prognosis markers (HER2, Ki67, ER, and PR) used for monitoring and treating patients and BLV positivity [15,21,37].
Regarding the risk factors for BLV infection, statistically significant differences correlate the BLV infection and people from rural area than urban and those with home-based activities than other occupational activities.
This may be explained by the fact that people who live in rural areas have direct contact with animals and their products.Besides that, a person's occupation and level of education are reflected in their dietary habits as well as their socioeconomic status and cultural behaviors.Ursin et al. [38] have linked milk utilization to the development of breast cancer.Moreover, Buehring et al. [5] and Corredor-Figueroa et al. [39] found statistically significant results between the human virus infection and the consumption of dairy products, specifically raw milk, and suggested the transmission of BLV to humans based on the evidence of the presence of the virus in the dairy products and its presence in the bovine population.
In the current study, there was no significant difference between age groups, as BLV DNA was detected in 30% of samples for patients <50 years and 13% of samples for patients ≥50 years.This is in agreement with Schwingel et al. [21].
Prospective investigations including a greater sample size, searching other marker of viral presence, performing quantitative analyses of correlation of the amount of virus present with other variables may be necessary to corroborate our findings and establish the presence (or absence) of BLV as a zoonotic agent that significantly affects public health.

Conclusion
This study supports the hypothesis that the presence of BLV in human breast tissue may be a risk factor for human breast cancer.It was found that people from rural areas who deal more with animals and home-made dairy products and practice household activities are at greater risk of developing BLV infection than others.So, control of the infection in cattle and screening of dairy products may aid in decreasing the transmission of the virus to humans, and this knowledge may open the way for the development of new preventive and therapeutic approaches for human breast cancer.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Figure 3 .
Figure 3. (a and b) H&E-stained sections of breast tissue showing infiltration of fibrofatty breast tissue by sheets and nests of malignant epithelial spheroidal cells showing nuclear atypia and pleomorphism with attempts for tubular differentiation (x100 and x400, respectively).(c) immunohistochemically stained section of invasive breast carcinoma showing strong nuclear staining of more than 70% of tumor cells for ER antibodies (score 3+5 = 8/8) (x200).(d) immunohistochemically stained section of invasive breast carcinoma showing strong nuclear staining of 30-70% of tumor cells for PR antibodies (score 2+5 = 7/8) (x400).(e) immunohistochemically stained section of invasive breast carcinoma showing strong complete membranous staining of more than 10% of tumor cells for anti-HER2/neu (score +3 = positive HER2/neu result) (x400).(f) immunohistochemically stained section of invasive breast carcinoma showing nuclear staining of about 40% of tumor cells for the Ki-67 proliferation marker (high proliferation index) (x100).

Table 1 .
Comparison between BLV positive and negative cancer breast tissue samples in tumor grade, stage, and lymph node metastasis.

Table 2 .
Assessment of risk factors for BLV infection in humans.