Intestinal microorganisms involved in colorectal cancer complicated with dyslipidosis

ABSTRACT

Background: Abnormal lipid metabolism is considered to be one of main promoters of colorectal cancer (CRC), and intestinal microorganisms may be involved in CRC in patients with abnormal lipid metabolism.

Objective: To investigate lipid metabolism in CRC patients and explore the role of intestinal microorganisms in CRC complicated with abnormal lipid metabolism.

Methods: Overall, 150 CRC patients in Huzhou Central Hospital from January 2016 to September 2017 were recruited in the present study. Basic patient information and clinical serological indicators were investigated and analyzed. Twenty-one stool samples were collected from patients after receiving informed consent. Next-generation sequencing technology was used to sequence bacterial 16S ribosomal RNA. Bioinformatics analysis was used to profile the microbial composition and screen distinctive bacteria in patients with CRC complicated with abnormal lipid metabolism.

Results: Apo B and FFA levels were higher in patients with stage I disease than in patients with other stages. HDL, LDL, Apo B and FFA levels were higher in female patients than in male patients. FFA level was higher in rectal cancer patients than in colon cancer patients. These differences were statistically significant (p < 0.05). The proportion of Escherichia/Shigella was increased in CRC patients with hyperlipoidaemia and hypercholesteremia; the abundance of Streptococcus was increased in CRC patients with hyperlipoidaemia; the abundance of Clostridium XIVa was reduced in CRC patients with hyperlipoidaemia and hypercholesteremia; and the abundance of Ruminococcaceae was reduced in CRC patients with hypercholesteremia. Bilophila and Butyricicoccus were closely related to CRC patients without hyperlipoidaemia or hypercholesteremia, and Selenomonas, Clostridium, Bacteroidetes Slackia, Burkholderiales and Veillonellaceae were closely related to CRC patients with hyperlipoidaemia. Some pathways, including secretion system, chaperones and folding catalysts, amino sugar and nucleotide sugar metabolism, arginine and proline metabolism, glycine, serine and threonine metabolism, histidine metabolism, pores and ion channels, nitrogen metabolism and sporulation, may be involved in lipid metabolism abnormality in CRC patients.

Conclusions: Many CRC patients have abnormal lipid metabolism, and the intestinal microbiota is altered in these CRC patients.

Introduction

Colorectal cancer (CRC) is one of the most common malignant tumours in the world, and its morbidity has increased in recent years. The pathogenesis of CRC is not well understood, but numerous studies suggest that CRC seems to be associated with the consumption of a high-fat diet.^1,2 High-fat diet can increase the level of saturated fatty acids in peripheral blood and lead to disturbances in lipid metabolism such as hyperlipoidaemia and hypercholesteremia.³ Epidemiological studies have shown that hyperlipoidaemia and hypercholesteremia can increase the risk of cancer, and hyperlipoidaemia can increase the incidence of CRC.¹ Several investigations have shown that the incidence of lipid metabolism abnormalities in patients with CRC is higher than that in healthy individuals.⁴ In addition, lipid metabolism abnormalities in patients with CRC can increase the risk of recurrence and metastasis and indicate poor prognosis.^5,6 In the present study, we collected clinical data and correlative parameters of lipid metabolism in patients with CRC and analyzed the relationship between lipid metabolism and CRC.

The human body harbours large numbers of microorganisms, particularly in the gastrointestinal tract. Bacteria, bacterial metabolites, body metabolism wastes, exfoliated cells, and indigestible food, among others, constitute a relatively independent microecosystem in the colon and rectum.⁷ Recent research has shown that colorectal cancer is a dysbacteriosis-induced disease.^8,9 Intestinal microorganisms in healthy individuals mainly include Firmicutes, Bacteroidetes, Proteobacterium and Lactobacillus at the phylum level. The relative abundance of Firmicutes such as Ruminococcus, Peptostreptococcus and Clostridium is low,⁹ whereas the relative abundance of Bacteroidetes such as Atopobium and Porphyromonas, Escherichia coli and Desulphovibrio is high.^10–14 In addition, Streptococcus bovis, Helicobacter pylori, Bacteroides fragilis, Enterococcus faecalis, Clostridium septicum and Escherichia coli play a potential role in the occurrence and development of CRC.¹⁵

Obesity and lipid metabolism abnormality are closely related to intestinal microorganisms. The abundance and diversity of intestinal microorganisms are negatively related to the degree of obesity. The abundance and diversity of intestinal microorganisms are low in obese patients.¹⁶ High-fat diet can increase the abundance of Rikenellaceae and Ruminococcaceae and decrease the abundance of Proteobacteria, Archaea, Clostridiales, Prevotellaceae and Bacteroidaceae and destroy the ecological balance of the intestinal tract.¹⁷ Intestinal dysbacteriosis can disturb the secreting rhythm of insulin, increase glucose absorption and cause overexpression of obesity-related transcription factors such as ChREBP and SREBP-1, finally resulting in lipid metabolism abnormality.¹⁸ The disruption of microorganism balance may modify the intestinal barrier function, increase the absorption of lipopolysaccharides and lead to insulin resistance and inflammatory reaction, resulting in lipid metabolism abnormality.^19,20 Moreover, some specific intestinal flora such as Bacteroides thetaiotaomicron can increase the rate of energy absorption and promote free fatty acid entry into the liver.²¹ Overall, there is a significant correlation between intestinal microorganisms and lipid metabolism.

In the present study, intestinal flora in patients with CRC was analyzed by detecting 16S rRNA genes in stool samples. We described and compared the abundance and diversity of intestinal microorganisms in CRC patients with or without hyperlipoidaemia and hypercholesteremia. Our results may provide novel insights into the crosstalk between CRC and lipid metabolism from the perspective of intestinal flora.

Subjects and methods

Subjects

Patients with CRC at Huzhou Central Hospital from January 2016 to January 2018 were studied. Cancer was confirmed by pathologic diagnosis, and the clinical stages were determined according to the American Joint Committee on Cancer (AJCC) staging guidelines. The clinical protocols involving the patients and the informed consent form were approved by the Ethics Committee of Huzhou Central Hospital (No. 2018.31). The inclusion criteria were as follows: patients were diagnosed by pathological examination and volunteered to participate in the study.

The exclusion criteria were as follows: ① Patients with other primary cancer. ② Patients with other intestinal diseases, such as ulcerative colitis and Crohn’s disease. ③ Patients with a history of use of oral microbial agents and lipid-regulatory agents within the last 2 months. ④ Patients with known primary organ failure. All of the subjects signed informed consent under the guidelines approved by the Ethics Committee of Huzhou Central Hospital. The study strategy is shown in Figure 1.

Figure 1.

Study strategy.

In total, 150 CRC patients at Huzhou Central Hospital from January 2016 to January 2018 were recruited in the study. The clinical protocols involving the patients and informed consent forms were approved by the Ethics Committee of Huzhou Central Hospital. Basic information and clinical serological indicators were obtained from the medical record management system of Huzhou Central Hospital after informed consent from patients. Twenty-one stool samples from CRC patients were analyzed from June to October 2017 after elimination of the unqualified specimens.

Collection of clinical data and stool samples

Basic information and clinical serological indicators were obtained from the medical record management system of Huzhou Central Hospital with informed consent from patients. Stool samples were collected in the morning prior to breakfast. Approximated 5–10 g stool sample was after defecation without the use of purgative or lubricant. Within half an hour, the stool samples were stored in an ultra-low temperature freezer. The sample preservation time was not beyond one month. Finally, 21 stool samples from CRC patients collected from June to October 2017 were analyzed after the patients had signed informed consent forms and the unqualified specimens were eliminated. These samples were divided into four groups. C4, C5, C7, C9, C11, C13, C14, C16 and C20 samples were included in Hype 0 group (samples from CRC patients without hyperlipoidaemia or hypercholesteremia). C1, C2 and C17 samples were included in Hype 1 group (samples from CRC patients with hyperlipoidaemia without hypercholesteremia). C3, C6, C8, C10, C15 and C19 samples were included in Hype 2 group (samples from CRC patients with hypercholesteremia without hyperlipoidaemia). C12, C18 and C21 samples were included in Hype 3 group (samples from CRC patients with hyperlipoidaemia and hypercholesteremia).

Intestinal microorganism detection

DNA extraction and PCR amplification

A E.Z.N.A.® Soil DNA Kit (Omega Bio-Tek, Norcross, GA, U.S.) was used to extract total DNA from the stool samples according to manufacturer’s protocols. PCR (95°C for 3 min, followed by 25 cycles at 95°C for 30 s, 55°C for 30 s, and 72°C for 45 s and a final extension at 72°C for 5 min) was used to amplify the V3-V4 region of the bacterial 16S ribosomal RNA gene (the primers of 16S V3-V4 rDNA are as follows: forward, CCTACGGGNGGCWGCAG and reverse, GACTACHVGGGTATCTAATCC). PCR amplifications were performed in triplicate in a 25 μl mixture containing 5 μl of DNA template, 2 μl of Nextera XT Index Primer 1 (10 M), 2 μl of Nextera XT Index Primer 2 (10 M) and 16 μl ddH₂O. Amplicons were extracted from 2% agarose gels, purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, U.S.) and quantified using QuantiFluor™-ST according to the manufacturer’s instructions.

Library construction and sequencing

The MiSeq library was constructed as follows: purified PCR products were quantified by Qubit® 3.0 (Life Technologies, Invitrogen), and twenty-four amplicons whose barcodes were different were mixed equally. An Illumina Pair-End library was constructed using pooled DNA products, and the amplicon library was pair-end sequenced (2 × 250) on an Illumina MiSeq platform (Shanghai BIOZERON Co., Ltd) according to standard protocols.

Sequencing data bioinformatics analysis

Sequencing data processing and optimization were performed according to the criteria at http://en.wikipedia.org/wiki/Fastq. Cutadapt (version 1.11) was used to demultiplex and quality filter the raw fastq files. Pandaseq (version 2.9) was used to assemble PE reads; only sequences that overlapped longer than 10 bp were assembled according to their overlap sequence. Mosaic sequences and sequences longer than 300–480 bp were discarded. In addition, the reads that receiving an average quality score< 20 were discarded.

The operational taxonomic unit (OTU) is an artificial classification unit in phylogenetic and population genetics studies. Vsearch and RDP-classifier software were used to cluster the OTUs with a 97% similarity and annotate species.

Results

Basic characteristics of the patients

Lipid metabolism abnormality mainly includes hyperlipoidaemia and hypercholesteremia. According to the Guidelines for the Prevention and Treatment of Dyslipidaemia in Chinese Adults (2016 revision), hyperlipoidaemia and hypercholesteremia are respectively defined as triglyceride> 2.3 mmol/l (200 mg/dl) and total cholesterol (TC)> 6.2 mmol/l (240 mg/dl) in peripheral blood.²² The characteristics of study participants are shown in Table 1 There was no significant difference between the patients with colon cancer and rectal cancer in age, sex, pathological stages, hyperlipoidaemia or hypercholesteremia.

Table 1.

Characteristics of study participants.

		Colon cancer	Rectal cancer	t/X2 value	p value
Sex	Male	28	65	1.48	0.26
	Female	12	45
Age	–	62.75 ± 12.11	65.91 ± 8.35	1.80	0.073
Pathological stages	Stage I	1	3	1.99	0.58
	Stage II	18	59
	Stage II-III	10	29
	Stage III	11	19
Lipid metabolism	Complicated without hyperlipoidemia or hypercholesteremia	29	86	0.54	0.91
	Complicated with hyperlipoidemia	5	11
	Complicated with hypercholesteremia	5	11
	Complicated with hyperlipoidemia and hypercholesteremia	1	2

Overall, 150 CRC patients in Huzhou Central Hospital from January 2016 to September 2017 were recruited in the present study. Hyperlipoidaemia and hypercholesteremia are respectively defined as triglyceride> 2.3 mmol/l (200 mg/dl) and total cholesterol (TC)> 6.2 mmol/l (240 mg/dl) in peripheral blood according to the Guidelines for the Prevention and Treatment of Dyslipidaemia in Chinese Adults (2016 revision). There was no significant difference between the patients with colon cancer and rectal cancer in age, sex, pathological stages, hyperlipoidaemia or hypercholesteremia.

Lipid metabolism-related indexes such as high-density lipoprotein (HDL), low-density lipoprotein (LDL), apolipoprotein A (Apo A), apolipoprotein B (Apo B) and free fatty acids (FFAs) were investigated to get more detailed information on lipid metabolism status. As shown in Table 2, Apo B and FFA levels were higher in patients with stage I disease than in patients with other stages. HDL, LDL, Apo B and FFA levels were higher in female patients than in male patients. FFA level was higher in patients with rectal cancer than in patients with colon cancer. These differences were statistically significant (p < 0.05). TG and TC were not significantly different depending on sex, pathological stage and cancer location.

Table 2.

Dyslipidosis in patinets with CRC.

Groups	Subgroups	TG(mmol/L)	TC(mmol/L)	HDL	LDL(mg/dL)	Apo A (mg/dL)	Apo B	FFA(μmol/L)
				(mg/dL)			(g/L)
Reference ranges		0.45–1.69	2.58–5.69	>35	<50	0–30	0.6–1.2	129–769
Colorectal cancer		1.26 ± 0.54	4.68 ± 2.37	43.67 ± 12.11	100.41 ± 27.99	29.62 ± 30.66	0.86 ± 0.21	485.15 ± 263.12
Pathological stages	Stage I	1.00 ± 0.46	5.52 ± 0.77	55.81 ± 17.54	124.65 ± 23.402	31.10 ± 30.76	1.00 ± 0.29	801.78 ± 464.055
	Stage II	1.19 ± 0.506	4.35 ± 0.85	43.89 ± 11.62	95.25 ± 23.46	27.79 ± 26.92	0.80 ± 0.16	451.18 ± 280.60
	Stage II-III	1.37 ± 0.64	4.55 ± 1.041	41.67 ± 12.74	102.54 ± 35.67	28.72 ± 28.11	0.92 ± 0.26	520.303 ± 222.95
	Stage III	1.34 ± 0.49	5.59 ± 4.91	44.088 ± 11.37	107.67 ± 25.42	39.61 ± 43.65	0.94 ± 0.21	484.45 ± 230.30
	F value	1.48	2.22	1.74	2.71	1.081	5.701	2.68
	p value	0.222	0.088	0.161	0.047	0.359	0.001	0.049
Sex	Male	1.26 ± 0.57	4.59 ± 2.91	41.57 ± 11.99	95.56 ± 25.65	28.38 ± 29.67	0.83 ± 0.204	427.76 ± 215.91
	Female	1.26 ± 0.51	4.83 ± 0.99	47.096 ± 11.605	108.33 ± 30.0073	33.90 ± 33.88	0.92 ± 0.21	578.80 ± 305.33
	t value	0.021	−0.62	−2.77	−2.77	−1.047	−2.60	−3.54
	p value	0.983	0.536	0.006	0.006	0.297	0.01	0.001
Cancer location	Colon cancer	1.28 ± 0.49	5.15 ± 4.31	41.91 ± 11.32	102.54 ± 25.21	36.99 ± 31.27	0.89 ± 0.19	412.24 ± 218.79
	Rectal cancer	1.25 ± 0.56	4.51 ± 0.94	44.32 ± 12.37	99.64 ± 29.0029	28.11 ± 31.17	0.85 ± 0.22	511.67 ± 273.59
	t value	−0.20	−1.46	1.079	−0.56	−1.54	−1.022	2.069
	p value	0.843	0.147	1.125	0.576	0.125	0.308	0.04

Basic patient information and clinical serological indicators were investigated and analyzed. Apo B and FFA levels were higher in patients with stage I disease than in patients with other stages. HDL, LDL, Apo B and FFA levels were higher in female patients than in male patients. FFA level was higher in patients with rectal cancer than in patients with colon cancer. These differences were statistically significant (p < 0.05). TG and TC were not significantly different depending on sex, pathological stage and cancer location.

The number of reads and corresponding OTUs were used to construct rarefaction curves by random sampling of all reads. The rarefaction curve (Figure 2) suggested that the sequencing depth was adequate. The box chart (Figure 2) describes the average level and variation degree among the four groups.

Figure 2.

Rarefaction curve.

Rarefaction analysis was conducted using Mothur v.1.21.1. The number of reads and corresponding OTUs were used to construct rarefaction curves by random sampling of all reads. The rarefaction curve (Figure 1) suggests that the sequencing depth was adequate. The box chart (Figure 1) describes the average level and variation degree among the four groups.

Figure 3 shows the composition and proportion of microorganisms at different taxonomic levels including phylum, class, order, family, genus, and species. The taxonomic tree heatmap is shown in panel A. The innermost layer shows the taxonomic tree. The phylum levels are marked in background colour, and red and white in the intermediate layer represents the average abundance in the form of a heatmap. The deeper the colour, the higher is the species abundance. Orange circles represent statistical difference of species among the different groups (SNK test). The outermost layer indicates annotated species. A bar chart (panel B) and error bars (panel C) were respectively used to describe the proportion and abundance of the microorganisms among the four groups at the genus level. The bar chart (panel B) shows that the proportion of Escherichia/Shigella was increased in CRC patients with hyperlipoidaemia and hypercholesteremia. The error charts (panel C) shows that abundance of the Streptococcus was increased in CRC patients with hyperlipoidaemia (Hype 1 and Hype 3), the abundance of the Clostridium XIVa was reduced in CRC patients with hyperlipoidaemia and hypercholesteremia (Hype 3), and the abundance of Ruminococcaceae was reduced in CRC patients with hypercholesteremia (Hype 2 and Hype 3).

Figure 3.

Microbial community composition.

The figure represents the composition and proportion of microorganisms at different taxonomic levels including phylum, class, order, family, genus and species. Panel A is the taxonomic tree heatmap, and it represents the taxonomy of microorganisms by the taxonomic tree and shows the abundance of microorganisms by a heatmap. A bar chart (panel B) and error bars (panel C) were respectively used to represent the proportion and abundance of microorganisms among the four groups at the genus level.

As shown in Figure 4, the enterotype map represents the correlation and contribution of the microorganisms to different groups. The different colours represent the different groups. The closer the microorganisms and the groups, the greater are the correlation and contribution. The figure showed that Bilophila and Butyricicoccus are closely related to Hype 0 group, and Selenomonas, Clostridium, Bacteroidetes Slackia, Burkholderiales and Veillonellaceae are closely related to Hype 1 group.

Figure 4.

Species correlation.

The enterotype map describes the correlation and contribution of the microorganisms to different groups. The different colours represent the different groups. The closer the microorganisms and the groups, the greater are the correlation and contribution.

Microorganisms can adjust and modify their surroundings by their metabolic products. A series of enzymes and genes participate in the process of metabolism. Functional analysis of these enzymes and genes would be conducive to know the whole microecological environment. The gene and copy number changes associated with the corresponding 16S sequences have been confirmed. The metabolism of the microorganisms can be speculated by referencing related databases, such as KEGG and NOG. As shown in Figure 5, each bar represents the abundance of the gene at the pathway level. The significant differences in functional metabolism are in shown in red font. These pathways, including secretion system, chaperones and folding catalysts, amino sugar and nucleotide sugar metabolism, arginine and proline metabolism, glycine, serine and threonine metabolism, histidine metabolism, pores and ion channels, nitrogen metabolism and sporulation, may be involved in lipid metabolism abnormality in CRC patients.

Figure 5.

Functional analysis.

It has been confirmed that many genes are associated with corresponding 16S sequences. Functional analysis of enzymes and genes would be conducive to know the whole microecological environment. The figure shows the results of functional analysis of 50 pathways. Each bar represents the abundance of the gene at the pathway level. The differences in functional metabolism are shown in red font.

Discussion

A meta-analysis on the relationship between dyslipidaemia and colorectal cancer risk²³ suggested that hyperlipoidaemia and hypercholesteremia increase the risk of CRC, and HDL decreases the risk of CRC. Some investigations have shown that high levels of LDL and Apo B were correlated with a high incidence of CRC. ²⁴ Although there is a lack of epidemiological data that can reflect triglyceride and total cholesterol levels in Chinese populations, the data in the published literature can be used for reference information. The serum triglyceride levels are 1.3 ± 0.8 mmol/l in individuals with no prediabetes and no diabetes mellitus.²⁵ The total cholesterol level in big cities and intermediate-size cities are (4.66 ± 0.04)mmol/l and (4.57 ± 0.06) mmol/l, respectively.²⁶ There is no obvious difference of triglyceride and total cholesterol between healthy individuals and patients with CRC according to these published data. Moreover, there is no significant difference in TG and TC levels based on sex, pathological stage or cancer location. However, there are significant differences in lipid metabolism indexes including Apo B and FFA levels based on pathological stage, in HDL, LDL, Apo B and FFA levels based on sex and in FFA level based on cancer location. These data indicate disorders in lipid metabolism, glucose metabolism and endocrine function among the different groups.²⁷ Serum lipid levels in CRC patients remain to be studied more in-depth through expanding sample number and extending the time of investigation.

The abundance of microbial populations is relatively stable in human intestinal microecosystem, a complex ecosystem. Mutual competition and mutual dependence of these microbial populations exist in the microecosystem.⁸ Imbalances in the intestinal microecosystem may lead to multiple diseases such as CRC, obesity, and diabetes.²⁸ The abundance and composition microbial populations are altered in CRC patients.^29–31 Additionally, some specific bacteria and microbial metabolites may drive the occurrence and development of colorectal cancer.^32,33

Intestinal microecosystem and lipid metabolism are interrelated and interact with each other.³⁴ Intestinal dysbacteriosis, such as changes in abundance and composition of Bacillus bifidus, Lactobacillus and Enterococcus can lead to lipid metabolic disturbance^35,36 by regulating and influencing the activity of cholesterol oxidase and hepatic fat synthase, redistribution of triglycerides and cholesterol in liver and blood, and enterohepatic circulation of cholate.^37–39 On the other hand, high-fat diet and abnormal lipid metabolism can disrupt the balance in intestinal microorganisms by decreasing the intestinal energy and nutrient absorption, changing the redox state and influencing the microenvironment of intestinal microorganisms.^40,41

More clinical and experimental studies will provide stronger evidence to elucidate the interrelationships and interactions between colorectal cancer, lipid metabolism and intestinal microorganisms. In the present study, we described and compared the abundance and diversity of intestinal microorganisms in CRC patients with or without hyperlipoidaemia and hypercholesteremia. We identified increased proportions some specific microorganisms including Escherichia/Shigella in CRC patients with hyperlipoidaemia and hypercholesteremia, increased abundance of Streptococcus in CRC patients with hyperlipoidaemia, reduced abundance of Clostridium XIVa in CRC patients with hyperlipoidaemia and hypercholesteremia and reduced abundance of Ruminococcaceae in CRC patients with hypercholesteremia. However, we cannot draw define conclusions because of sample size limitation, and larger studies may investigate this issue further. We also found that some intestinal microorganisms are closely related to CRC complicated with lipid metabolism disorder, and some pathways including protein structure, amino acid and nucleotide metabolism, pores and ion channels and microbial secondary metabolism may participate in molecular regulation in CRC patients with lipid metabolism disorder. These microorganisms and pathways can provide potential molecular targets and potential signalling pathways for future experimental studies.

Abbreviation

CRC

Colorectal cancer

AJCC

American Joint Committee on Cancer

OUT

Operational taxonomic unit

TC

Total cholesterol

HDL

High-density lipoprotein

LDL

Low-density lipoprotein

Apo A

Apolipoprotein A

Apo B

Apolipoprotein B

FFA

Free fatty acid

Funding Statement

This work was supported by the Public Welfare Technology Application Research Program of Huzhou (No. 2016GYB14).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgements

We thank the patients and volunteers for their contributions to the imaging and sample collection.

Consent

Written informed consent was obtained from the patients for the publication of this paper and the accompanying images.