Fusobacterium nucleatum in gastroenterological cancer: Evaluation of measurement methods using quantitative polymerase chain reaction and a literature review.

The human microbiome Fusobacterium nucleatum, which primarily inhabits the oral cavity, causes periodontal disease and has also been implicated in the development of colorectal cancer. However, whether F. nucleatum is present in other gastroenterological cancer tissues remains to be elucidated. The present study evaluated whether quantitative polymerase chain reaction (qPCR) assays were able to detect F. nucleatum DNA and measure the quantity of F. nucleatum DNA in esophageal, gastric, pancreatic and liver cancer tissues. The accuracy of the qPCR assay was determined from a calibration curve using DNA extracted from cells from the oral cavity. Formalin-fixed paraffin-embedded (FFPE) tumor tissues from 20 patients with gastroenterological [esophageal (squamous cell carcinoma), gastric, colorectal, pancreatic and liver] cancer and 20 matched normal tissues were evaluated for F. nucleatum DNA content. The cycle threshold values in the qPCR assay for F. nucleatum and solute carrier organic anion transporter family member 2A1 (reference sample) decreased linearly with the quantity of input DNA (r2>0.99). The F. nucleatum detection rate in esophageal, gastric and colorectal cancer tissues were 20% (4/20), 10% (2/20) and 45% (9/20), respectively. F. nucleatum was not detected in liver and pancreatic cancer tissues. The qPCR results from the frozen and FFPE tissues were consistent. Notably, F. nucleatum was detected at a higher level in superficial areas compared with the invasive areas. F. nucleatum in esophageal, gastric and colorectal cancer tissues was evaluated by qPCR using FFPE tissues. F. nucleatum may be involved in the development of esophageal, gastric and colorectal cancer.

Introduction

As a developing research area, the microbiome has been the focus of multiple studies in previous years. The non-spore-forming, anaerobic gram-negative bacterium Fusobacterium nucleatum is part of the normal flora of the human oral cavity and gut mucosa, but is an established opportunistic pathogen in periodontal diseases (1–4) and several inflammatory diseases, including inflammatory bowel disease (5–8), liver abscesses (9,10) and chorioamnionitis (11). Two previous studies have reported an overabundance of F. nucleatum in colorectal cancer tissues compared with adjacent normal tissues (12,13). Following this, a previous study demonstrated that F. nucleatum activates the E cadherin/β-catenin signaling pathway via FadA adhesion, promoting colorectal cancer growth (14). Fusobacterium subspecies (spp.), including F. nucleatum, are also present at increased levels in human colorectal, pancreatic and other types of cancer (12,13,15–20). To the best of our knowledge, there are only five previous studies reporting the presence of Fusobacterium spp. in colorectal and pancreatic cancer tissues and there are no published studies that associate Fusobacterium spp. with esophageal, gastric, hepatocellular and other gastroenterological cancer (Table I) (15,16,19,20,21).

Table I.

Detection rates of Fusobacterium spp. in gastroenterological cancer tissues from previous studies.

					Fusobacterium detection rate, %
Authors	Type of cancer	No. of cases	Tissue fixation	Bacterial strain	Tumor tissue	Normal tissue	(Refs.)
Tahara et al, 2014	Colorectal cancer	149	Frozen tissue	F. nucleatum	52.3 (78/149)	30.3 (27/89)	(15)
Mima et al, 2015	Colorectal cancer	598	FFPE	F. nucleatum	13 (76/598)	3.4 (19/558)	(16)
Ito et al, 2015	Colorectal cancer	511	FFPE	F. nucleatum	56 (286/511)	−	(19)
Mitsuhashi et al, 2015	Pancreatic cancer	283	FFPE	Fusobacterium species	8.8 (25/283)	28 (7/25)	(20)
Viljoen et al, 2015	Colorectal cancer	71	FFPE	F. nucleatum spp. polymorphum	82 (58/71)	81 (48/59)	(21)

spp., subspecies; FFPE, formalin-fixed paraffin-embedded tissues; F., Fusobacterium.

Elevated levels of F. nucleatum DNA in colorectal cancer tissue are associated with certain molecules and cell functions, including microsatellite instability, the CpG island methylator phenotype and hMLH1 (15), and are also associated with a lower density of T cells (16). A number of previous studies have associated high levels of F. nucleatum DNA content with poor patient prognosis (17,18), however, other previous studies have reported that there is no association between the quantity of F. nucleatum DNA and patient survival rate (12,19). In one previous study, the DNA status of Fusobacterium spp. in pancreatic cancer tissue was independently associated with the poor prognosis of patients (20).

However, whether F. nucleatum is present in other types of gastroenterological cancer, including esophageal, gastric or liver cancer, has yet to be investigated.

In the present study, quantitative polymerase chain reaction (qPCR) method was evaluated to determine if it was able to detect the quantity of F. nucleatum DNA from an oral cavity. Subsequently, a qPCR assay was also used to analyze whether it similarly detects the existence of Fusobacterium in formalin-fixed paraffin-embedded (FFPE) tissues and frozen tissues. Finally, the quantity of F. nucleatum DNA in 20 paraffin-embedded digestive cancer specimens and 20 matched normal specimens was evaluated.

Materials and methods

Tissue samples

The test specimens were 20 FFPE tissue samples of esophageal (squamous cell carcinoma), gastric, colorectal, liver and pancreatic cancer, and 20 normal matched paraffin embedded specimens. All specimens were obtained by surgical resection at Kumamoto University Hospital (Kumamoto, Japan). The sampled patients were not administered preoperative treatment. A single pathologist, who was blind to the clinical and molecular data of the patients, evaluated hematoxylin-eosin-stained tissue sections of each cancer case and recorded the pathological features. Tumor staging was conducted as described in the Cancer Staging Manual (7th edition) published by the American Joint Committee on Cancer (22). Written informed consent was obtained from each patient and the present study was approved by the Institutional Review Board of Kumamoto University Hospital (Approval no. 1272).

DNA extraction and qPCR for F. nucleatum DNA content

Genomic DNA in the oral cavity was obtained using a cotton swab. The patients were not allowed to eat or drink 30 min prior to sample collection and the cotton swap was scraped against the inside of each cheek 5–6 times. The collected swab was air-dried for >2 h. The genomic DNA from the oral cavity was extracted using QIAamp DNA Mini kit (Qiagen GmbH, Hilden, Germany). Genomic DNA from the FFPE tissues and from the frozen gastroenterological cancer tissues was extracted using the QIAamp DNA FFPE Tissue kit (Qiagen GmbH) and the QIAamp DNA Mini kit (Qiagen GmbH), respectively. The nusG gene of F. nucleatum and the reference human gene solute carrier organic anion transporter family member 2A1 (SLCO2A1) were amplified using custom-made TaqMan primer/probe sets (Applied Biosystems; Thermo Fisher Scientific, Inc., Waltham, MA, USA) as previously described (18). The primer and probe sequences used for the custom TaqMan Gene Expression assay were as follows: F. nucleatum forward primer, 5′-TGGTGTCATTCTTCCAAAAATATCA-3′; F. nucleatum reverse primer, 5′-AGATCAAGAAGGACAAGTTGCTGAA-3′; F. nucleatum FAM probe, 5′-ACTTTAACTCTACCATGTTCA-3′; SLCO2A1 forward primer, 5′-ATCCCCAAAGCACCTGGTTT-3′; SLCO2A1 reverse primer, 5′-AGAGGCCAAGATAGTCCTGGTAA-3′; SLCO2A1 VIC probe, 5′-CCATCCATGTCCTCATCTC-3′. The PCR mix consisted of 1X TaqMan Environmental Master Mix 2.0 (Applied Biosystems; Thermo Fisher Scientific, Inc.), 0.5 pmol forward and reverse primer, 0.1 pmol probe, nuclease-free water (Invitrogen; Thermo Fisher Scientific, Inc.) and 12.5 ng genomic DNA in a total volume of 10 µl. Assays were performed in a 384-well optical PCR plate. The DNA was amplified and detected with the LightCycler 480 Instrument II (Roche Diagnostics, Basel, Switzerland) under the following reaction conditions: Initial denaturation at 95°C for 10 min, 15 sec at 95°C and 60 sec at 60°C. The quantity of F. nucleatum DNA in each tissue was normalized relative to SLCO2A1 using the 2−ΔΔCq method (where ΔCq is the mean Cq of F. nucleatum minus the mean Cq of SLCO2A1) (16,23). All RT-qPCR reactions were performed in triplicate.

Statistical analysis

All statistical analyses were performed by the JMP program, version 10 (SAS Institute, Inc., Cary, NC, USA). All P-values were 2-sided. The mean quantity of F. nucleatum DNA was compared with paired Student's t-tests for variables with two categories. P<0.05 was considered to indicate a statistically significant difference.

Results

Literature review

An online search of MEDLINE (PubMed) was performed for all articles published in the English language. The following Medical Subject Headings terms were used in combination: ‘Fusobacterium esophageal cancer’, ‘Fusobacterium gastric cancer’, ‘Fusobacterium colorectal cancer’, ‘Fusobacterium hepatocellular carcinoma’, and ‘Fusobacterium pancreatic cancer’. The latest search was performed on December 2015. Among them, five studies which had detection rates of Fusobacterium spp. in cancer tissues were identified. In total, four previous studies have reported detectable levels of F. nucleatum in colorectal cancer tissues (15,16,19,21). The F. nucleatum detection rate was 13–82% in colorectal tumor tissue and 3.4–81% in adjacent normal tissue (Table I). A single previous study detected F. nucleatum in pancreatic cancer (the detection rate was 8.8% in tumor tissue and 28% in adjacent normal tissue) (20). However, the expression status of Fusobacterium DNA in esophageal, gastric and liver cancer remains to be elucidated.

Validation of qPCR for F. nucleatum

A cheek swab from a healthy researcher (Dr Kensuke Yamamura, Department of Gastroenterology, Kumamoto University, Kumamoto, Japan) was submitted for genomic DNA determination of the oral cavity. F. nucleatum and SLCO2A1 in the oral cavity were evaluated using qPCR in a 2-fold dilution series (5, 10, 12.5, 20 and 40 ng). The assays were quantified using the coefficient of determination (r2) between 5 and 40 ng. In the qPCR assays of oral F. nucleatum and SLCO2A1, the cycle threshold (Cq) values linearly decreased with the quantity of input DNA (on a linear-log scale, r2>0.99; Fig. 1). These results demonstrated that qPCR has the ability to quantify F. nucleatum DNA in the oral cavity.

Figure 1.

Analysis of linearity in quantitative polymerase chain reaction assay of F. nucleatum from the oral cavity. Levels of F. nucleatum and the human reference gene (SLCO2A1) from the oral cavity obtained by oral swab, demonstrated in a 2-fold dilution series (5, 10, 12.5, 20 and 40 ng). Data points and error bars denote the mean ± standard deviation of the cycle thresholds, respectively. Results were performed with triplicate runs in three separate experiments. F., Fusobacterium; R2, coefficients of determination; SLCO2A1, solute carrier organic anion transporter family member 2A1.

qPCR of F. nucleatum in frozen tissue and FFPE

F. nucleatum DNA in FFPE and frozen tissues of 10 esophageal squamous cell carcinoma (ESCC) cases were investigated. In the 5 tissues that were positive for F. nucleatum, the organism was also detected in the matched FFPE tissues. Similarly, in the 5 Fusobacterium-negative ESCC cases, F. nucleatum was not detected in the matched FFPE tissue (Table II). Therefore, the qPCR results were consistent between the frozen tissues and FFPE tissues. These results suggested that F. nucleatum may be accurately assayed in FFPE tissues.

Table II.

Consistency of quantitative polymerase chain reaction detection of Fusobacterium nucleatum in tumor FFPEs and frozen tissues of esophageal cancer.

Variable	Case 1	Case 2	Case 3	Case 4	Case 5	Case 6	Case 7	Case 8	Case 9	Case 10
FFPE	−	−	−	−	−	+	+	+	+	+
Frozen tissue	−	−	−	−	−	+	+	+	+	+
Concordance	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes

FFPE, formalin-fixed paraffin-embedded tissues.

F. nucleatum in gastroenterological cancer tissue

20 FFPE tumors and their adjacent non-tumorous tissues in each cancer were analyzed using qPCR assays. F. nucleatum was detected in 4 (20%) cases of esophageal cancer, 2 (10%) cases of gastric cancer and 9 (45%) cases of colorectal cancer (Fig. 2; Table III). In esophageal and colorectal cancer, F. nucleatum was also detected in adjacent non-tumor tissue, whereas F. nucleatum was not detected in the liver and pancreatic cancer tissues and their adjacent non-tumor tissues. Among all cancer cases that were positive for F. nucleatum, the level of F. nucleatum DNA content in esophageal and colorectal cancer ranged from 2.68×10−3 to 365.2×10−3 (median, 101.3×10−3) and from 2.10×10−3 to 166.7×10−3 (median, 5.08×10−3), respectively.

Figure 2.

Relative quantity of F. nucleatum DNA in gastroenterological cancer. Quantitative polymerase chain reaction assays of F. nucleatum in the tumor and adjacent non-tumor tissues of 20 patients with (A) esophageal cancer, (B) gastric cancer and (C) colorectal cancer. The cancer stage of each tumor is also presented. F., Fusobacterium.

Table III.

Quantitative polymerase chain reaction results of Fusobacterium nucleatum in gastroenterological cancer and adjacent normal tissues.

	Fusobacterium detection rate, %
Type of cancer	Tumor tissue	Normal tissue	Tumor and normal tissues
Esophageal cancer	20 (4/20)	5 (1/20)	0
Gastric cancer	10 (2/20)	0	0
Colorectal cancer	45 (9/20)	40 (8/20)	25 (5/20)
Liver cancer	0	0	0
Pancreatic cancer	0	0	0

Heterogeneity of F. nucleatum in esophageal cancer tissue

To evaluate the heterogeneity of the F. nucleatum DNA within tumor tissues, the F. nucleatum DNA in the superficial and invasive areas of the 5 esophageal cancer tissues that were positive for F. nucleatum were evaluated (Fig. 3A). High levels of Fusobacterium nucleatum DNA was observed in superficial areas, but low levels were observed in invasive areas. In the superficial areas, the quantity of F. nucleatum DNA ranged from 30.1×10−3 to 200.3×10−3, whereas in invasive areas it was 12.4×10−3 at its highest (P=0.02; Fig. 3B). Therefore, the F. nucleatum DNA may distribute heterogeneously within a single tumor.

Figure 3.

F. nucleatum DNA in superficial and invasive tissue areas. (A) Macrodissection of the superficial and invasive areas in tissues that were positive (cases 1–5) and negative (cases 6–10) for F. nucleatum. (B) F. nucleatum DNA was observed at high levels in superficial areas (range 30.1×10−3-200.3×10−3) and at low levels in invasive areas (0–12.4 ×10−3; P=0.02). F., Fusobacterium.

Discussion

F. nucleatum has received increased recognition as an opportunistic pathogen in periodontal diseases, but also in human cancer. As the microbiome has a number of important effects on the functions of the human body, the clinical significance of the discovery of F. nucleatum cannot be overemphasized. To the best of our knowledge, the present study has reported the first detection of F. nucleatum DNA in esophageal, gastric and liver cancer tissues. The present study has demonstrated that the qPCR assay may reliably detect F. nucleatum DNA from oral swabs, as F. nucleatum is among the most prevalent species in the oral cavity (1,2,24). The association between cycle threshold and input DNA in the qPCR assay of F. nucleatum was linear (r2>0.99). Furthermore, the FFPE and frozen tissues prepared from the same esophageal tumor yielded a similar level of detection accuracy. Typically, the fixation process chemically alters the nucleic acids in a sample by inducing covalent DNA cross-linking and fragmentation. These alterations may reduce the efficacy of analysis using PCR and DNA sequencing methods (25,26). In the present study, the results of the FFPE and frozen tissues were concordant, which suggested that in the two types of tissue preparations, qPCR accurately detected F. nucleatum DNA in gastroenterological cancer tissues.

F. nucleatum DNA was successfully detected in gastrointestinal tract cancer tissues (esophageal, gastric and colorectal cancer), but F. nucleatum was not detected in pancreatic and liver cancer tissues. In previous studies, the detection rates of F. nucleatum were 13–82% in colorectal cancer (15,16,19,21) and 8.8% in pancreatic cancer (20). These previous studies are concordant with the data from the present study that uses colorectal cancer tissues, but these results contradict the pancreatic cancer results in the current study. Although F. nucleatum is part of the normal flora of the oropharyngeal and gastrointestinal tracts, F. nucleatum also expresses FadA, a bacterial cell surface adhesion protein that activates the WNT signaling pathway in colorectal cancer cells, and consequently promotes tumor growth (14). Therefore, it may reasonably be expected that the detection rate of F. nucleatum is higher in gastroenterological cancer compared with liver and pancreatic cancer. However, the presence of F. nucleatum in esophageal and gastric cancer tissues remains to be investigated.

In addition, the F. nucleatum expression levels in superficial and invasive areas of esophageal cancer tissues were compared, and an increased level was observed in superficial areas. This result suggested that F. nucleatum may not be able to infiltrate into the invasive area and may only contribute to the tumor growth through the side of the gastrointestinal tract. As the differential distribution of F. nucleatum has not been previously reported, the low level of F. nucleatum in invasive areas remains to be fully elucidated. The involvement of F. nucleatum in tumor growth requires further investigation.

In conclusion, F. nucleatum DNA was detected in esophageal, gastric and colorectal cancer, but not in pancreatic and liver cancer. This suggested that F. nucleatum may be associated with the progression of gastroenterological tract cancer, but not the progression of pancreatic and liver cancer.