Effect of high-dose methotrexate chemotherapy on intestinal Bifidobacteria,Lactobacillus and Escherichia coli in children with acute lymphoblastic leukemia

Introduction

Pediatric leukemia is a type of malignant tumor with the highest incidence in children. Tumor cells frequently metastasize to the central nervous system and the testes, resulting in extramedullary leukemia. Tumor cells that infiltrate to the central nervous system and testes can avoid being killed by drugs because most chemotherapeutic drugs do not easily pass the blood–brain and blood–testis barriers. This is the key reason that acute lymphoblastic leukemia (ALL) frequently re-occurs. Recently, with the continuous improvement in combined chemotherapy, the life of children with leukemia has been prolonged. High-dose methotrexate (HDMTX) chemotherapy is generally accepted as an effective method for the treatment and prevention of extramedullary leukemia. ^1,2 Owing to the fast proliferation and rapid replacement of human intestinal mucosal cells and enteric cavity bacteria, and the abundant utilization of tetrahydrofolic acid (citrovorum factor [CF]) for DNA synthesis, bacterial flora is easily influenced by the pharmacological action of methotrexate (MTX). Numerous bacteria in the enteric cavity die when tumor cells are killed by MTX, and the intestinal flora lose the interface of colonization and survival because of intestinal mucosa injury, which easily leads to dysbacteriosis and influences the disease condition and prognosis of children with leukemia.

To understand the effects of HDMTX on intestinal bacterial flora in children with ALL, and to provide evidence for the rational use of microbial ecological agents for adjuvant therapy of leukemia, we quantitatively analyzed changes of intestinal Bifidobacteria, Lactobacillus and Escherichia coli in children with ALL before and after HDMTX chemotherapy by application of 16SrRNA/DNA fluorescent quantitative polymerase chain reaction (PCR).

Patients and methods

Methods

The methods are similar to previously described methods, ^4–10 and modified as below.

Preparation of intestinal bacteria DNA

Fecal samples were taken from the −70°C freezer and put in the −20°C freezer for 24 h. They were then transferred to the 4°C refrigerator for another 24 h, and then placed at room temperature. Stool (0.2 g) was weighed and fully mixed with 1.5 mL of phosphate buffer solution (PBS; 0.05 mol/L [pH 7.4]) for five minutes and centrifuged at 1000 rpm for 10 min. The supernatant was collected, and this sequence was repeated three times according to the above method. The entire supernatant was collected and centrifuged at 14,000 rpm for 10 min. The supernatant was then discarded, and the sediment was collected and mixed with 1 mL of PBS, followed by centrifugation at 14,000 rpm for five minutes. Subsequently, the supernatant was discarded and the sediment was washed repeatedly with PBS, and then washed once with water. Bacteria were suspended in 100 μL of distilled water and lysed with 100 μL of 1% Triton X-100. Sample DNA from the control and ALL groups was detected with a spectrophotometer at A ₂₆₀.

Purification of DNA

Bacterial liquid was centrifuged at 14,000 rpm; 200 μL of supernatant was removed and agitated with isovolumetric phenol/chloroform for 10 s, and then centrifuged at 14,000 rpm for 10 s. Subsequently, the supernatant was collected and fully mixed with a 2–2.5-fold volume of 100% iced alcohol and then placed on ice for 20 min. After centrifugation at 14,000 rpm, the supernatant was discarded and the pellet was precipitated once with 70% alcohol. The centrifuge tubes were heated and dried at 100°C, and then 50 μL of TE buffer were added to the tubes at room temperature.

Routine PCR

Bacterial primers were designed with primer-design software and aligned and verified with the BLAST database (www.ncbi.nlm.gov/BLAST). The upstream and downstream primers of the three bacteria were as follows: Bifidobacteria: forward primer, 5′ GATTCTGGCTCAGGATGAACGC 3′; reverse primer, 5′ CTGATAGGACGCGACCCCAT 3′; Lactobacillus: forward primer, 5′ CTGATGAAAGCCCTCG 3′; reverse primer, 5′ GAGCCTCAGCGTCAGTTG 3′; E. coli: forward primer, 5′ CATGCCGCGTGTATGAAGAA 3′; reverse primer, 5′ CGGGTAACGTCAATGAGCAAA 3′. The reaction system included 5 μL of 10× buffer, 1 μL of 4× dNTP, 0.5 μL of forward primer (25 μmmol/L), 0.5 μL of reverse primer (25 μmmol/L), 5 μL of MgCl₂ (25 mol/L), 0.4 μL of Taq polymerase (5 μ/μL), 5 μL of DNA template and 32.6 μL of distilled water, for a total of 50 μL. (The sequences were amplified by a GeneAmp PCR system 9600 [PerkinElmer Inc, Foster city, CA, USA.]) The reaction conditions were as follows: pre-denaturation for five minutes at 95°C, denaturation for 15 s at 95°C, annealing for one minute at 60°C and extension for 45 s at 72°C. A total of 35 cycles were performed, and then the reaction ended with a final extension for 10 min at 72°C. The amplified products showed three bacterial 16SrRNA/DNA fragments (Figures 1 and 2) by agarose gel electrophoresis. OPEN IN VIEWER

Figure 1

Bands 1 to 4 are fragments of Escherichia coli (95 bp) and bands 5 to 8 are fragments of Bifidobacteria (203 bp)

OPEN IN VIEWER

Figure 2

Band 1 is the marker, bands 2 to 5 are blank, bands 6 and 7 are fragments of Escherichia coli (95 bp) and bands 8 and 9 are fragments of Lactobacillus (166 bp)

Fluorescent quantitative PCR

The reaction included 0.5 μL of forward primer (25 μmmol/L), 0.5 μL of reverse primer (25 μmmol/L), 5 μL of DNA template, 25 μL of SYBR Green (including 10× buffer, 4× dNTP, MgCl₂ and Taq polymerase) and 19 μL of distilled water, for a total of 50 μL. Sequences were amplified by a GeneAmp PCR system 5700 fluorescent quantitative PCR equipment (Applied Biosystems, Inc, Foster City, CA, USA), and a melting curve was made. The reaction conditions were set as follows: pre-denaturation for five minutes at 95°C, denaturation for 15 s at 95°C, annealing for one minute at 60°C and extension for 45°C at 72°C. A total of 40 cycles were performed and then ended with a final extension for 10 min at 72°C. The temperature for removing the primer–dimers was set with the melting curve and added to the cycle program; the fluorescent quantitative amplification was made under the following conditions. After five minutes of pre-denaturation, the conditions were set as follows: Bifidobacteria: 95°C for 15 s, 60°C for  one minute, 72°C for 45 s and 88°C for five seconds; E. coli: 95°C for 15 s, 60°C for one minute, 72°C for 45 s and 82°C for five seconds. A total of 40 cycles were performed and ended with a final extension for 10 min at 72°C. The initial cycle number was obtained from the fluorescent curve when the fluorescent sign appeared.

Drawing standard curve and quantitative analysis

Bifidobacteria and Lactobacillus freeze-dried powder (batch number 11054; The National Strain Management Center of China, Beijing, PR China). E. coli (provided by the laboratory department of our hospital) were inoculated in liquid MRS media and cultured in anaerobic environment for 24 h at 37°C (90% N₂ + I 0% CO₂). One milliliter of medium was removed and centrifuged at 3000 rpm for 15 min, the supernatant was discarded and the sediment was obtained. Fifty microliters of DNA solution dissolved in TE was obtained according to the aforementioned isolation and DNA purification methods. DNA absorbance at 260 nm was detected with a spectrophotometer. The concentration of DNA templates of standard strains was calculated according to the formula, A ₂₆₀ value = 4.74 × 10¹³ molecules. After a 10-fold serial dilution, a concentration gradient from 10⁷ to 10⁴ was obtained. DNA templates with different concentrations were amplified by fluorescent quantitative PCR. The initial cycle number (C_t ) in the PCR procedure was designated as the ordinate when the fluorescent sign appeared, the logarithm of the copy number of different DNA templates was designated as the abscissa, the standard curve was drawn, and then the intercept (b), slope (k) and correlation coefficient were obtained. Finally, the initial cycle number of each sample was put into the regression equation, log C_o = (C_n − b)/k and the corresponding DNA template number was obtained.

Statistical analysis

After logarithmic transformation, the results were displayed in the form of X _lgx ± S _lgx. The bacterial numbers from the ALL and control groups were compared by SPSS, version 13.0 (SPSS Inc, Chicago, IL, USA), and the results from the ALL and control groups were analyzed by one-way analysis of variance (ANOVA). A paired t-test was used in the ALL group for comparison before and after chemotherapy. P < 0.05 indicated a significant difference.

Results

Discussion

The human gut is the natural habitat for a large and dynamic bacterial community, but recently developed molecular biology tools discovered significant inter-subject variability and differences between the composition of stool and mucosal communities. Characterization of this immensely diverse ecosystem is the first step in elucidating its role in health and disease. The human endogenous intestinal microflora is an essential ‘organ’ in providing nourishment, regulating epithelial development, preventing mucosal barrier dysfunction, and instructing innate immunity. Intestinal microflora might also be an essential factor in certain pathological conditions, including multisystem organ failure, colon cancer and inflammatory bowel diseases. Nevertheless, bacteria are also useful in promoting human health. Probiotics and prebiotics are known to have roles in the prevention or treatment of some diseases. ^11–13

The intestinal microflora is the largest and most important microecosystem in the body. Normal flora is formed with long-term evolution of the host. Owing to the poor equilibrium stability, the intestinal microflora is influenced easily by various factors, such as feeding patterns, age, living habits, hygienic environment, radiotherapy, chemotherapy and antibiotics, which could result in dysbacteriosis, ¹⁴ increasing intestinal permeability, ^15–20 intestinal barrier dysfunction ^21–23 and intestinal bacteria and/or endotoxin (including bacterial cell wall components, such as lipopolysaccharide and peptidoglycan polysaccharide translocation). ^24–28 It is possible that enteric bacteria, their cell wall components, or both, may disseminate throughout the body, causing sepsis, shock, multisystem organ dysfunction or death of the host. ²⁷ These disturbances in critically ill patients may influence their outcome and prognosis.

A better understanding of the factors that influence translocation could result in the implementation of interventions that contribute to improved patient outcomes. Glutamine supplementation and targeted nutritional intervention (including Bifidobacteria, maintaining splanchnic flow, the judicious use of antibiotics and directed selective gut decontamination regimens), hold some promise in limiting bacterial translocation. ^{15,20,22,23,27} It is hoped that improving our understanding of the role of the gut in disease processes will lead to the development of novel therapeutic strategies in the future.

To determine whether dysbacteriosis or the graduation of dysbacteriosis exists in clinical practice, concrete quantitative analysis on the intestinal flora is needed. In this study, Bifidobacteria, Lactobacillus and E. coli in stool among a control group and children with ALL were quantified before and after HDMTX chemotherapy by 16SrRNA/DNA fluorescent quantitative PCR. The results showed that the amount of Bifidobacteria, Lactobacillus and E. coli in the stools of the ALL group were much less than in the control group. After chemotherapy, the amount of Bifidobacteria, Lactobacillus and E. coli in the stools from the ALL group decreased significantly compared with the amount before chemotherapy, especially on the third day after chemotherapy; however, there was partial recovery by the seventh day after chemotherapy. The amount of E. coli was greater after chemotherapy than before chemotherapy, indicating that HDMTX chemotherapy could lead to a further disorder of intestinal flora in children with ALL. This reduction and disorder gradually improved with drug metabolism. In general, it takes two weeks or more to completely recover from this disorder.

MTX was toxic to normal cells that rapidly proliferate and differentiate, such as intestinal epithelial cells and bone marrow cells. Sufficient hydration, alkalization of urine and rational CF rescue was used to ensure the safety of HDMTX application (12 g/m²), but animal experiments and long-term clinical observation showed that HDMTX still caused notable intestinal mucosal injury and toxic side-effects, such as diarrhea and hematochezia. ²⁹ The intestinal flora lost the interface of colonization because of intestinal mucosal injury, and numerous bacteria that acted as cell bodies were probably killed, leading to a reduction in probiotics and dysbacteriosis. It was discovered in mice fed in a bacteria-free environment that the growth rate of pit cells, which differentiate and produce intestinal epithelial cells, decreased and the crypt cell number was significantly less than in mice with normal flora in the colon. ³⁰ Experimental data have also shown that Bifidobacteria and Lactobacillus, as well as their metabolites (mainly short-chain fatty acids), are able to not only promote the synthesis of intestinal epithelial cells (IEC-6) DNA to quicken the proliferation of intestinal epithelial cells, ³¹ but can also prevent intestinal epithelial cells from generating tumor necrosis factor (TNF), which can relieve inflammatory injury and protect the mucosal barrier. ³² This suggests that loss of nutrition, protection and repair of intestinal normal flora of intestinal mucosa was probably the reason that hematochezia, constipation and other symptoms occurred and developed.

There is a close relationship between the occurrence, development and prognosis of ALL, and the immune status of patients. Lymphocyte carcinogenesis and immune inhibition of chemical drugs have always been the main reasons underlying decreased immune function and secondary infection in children with ALL. The reduction of intestinal probiotics and dysbacteriosis during chemotherapy affects immune function directly or indirectly. A study showed that lymphocytes in associated lymphoid tissues of germ-free animals were decreased, ³³ whereas after bacterial inoculation in the intestinal tract, lymphocytes in the mesentery increased, the speed and strength of lymphocytes against antigen increased, the clonal diversity of peripheral hyperglobulinemia increased and the function of phagocytes was enhanced, suggesting that bacterial inoculation of germ-free cells in the intestinal tract enhances the local and systemic immune functions of the intestinal tract. Liang et al. ³⁴ discovered in mouse experiments that dysbacteriosis leads to a reduction of humoral and cellular immunity and non-specific immunity, as well as hematopoietic function. Chemotherapy can damage the intestinal flora of leukemia patients, and dysbacteriosis can lead to a reduction in immunity function. Chen and Wen ³⁵ analyzed the intestinal flora of leukemia patients before chemotherapy and found that the amount of Bifidobacteria and E. coli decreased significantly compared with healthy children. The degree of reduction was closely related to the incidence of infection and septicemia after chemotherapy. Therefore, the effect of intestinal flora on immunity in chemotherapy should not be negligible. As proposed by Delaloye et al., ³⁶ protection of body immunity and the killing of tumor cells are equally important in the process of tumor therapy. It is known that Bifidobacteria promote the secretion of cytokines, such as interleukin 26 and TNF2α, by activating macrophages, which play an indirect role in tumor inhibition. ³⁷ In addition, peptidoglycan, teichoic acids and polysaccharides in the cell walls of Bacillus bifidus also have antitumor effects. A study by Tan and Yao ³⁸ showed that after tumor therapy with Bifidobiogen–Livzon, tumors were significantly smaller than in the control group, which is evidence of the view mentioned above. Data from our experiments showed that a notable reduction in Bifidobacteria led to declines in antitumor ability. Therefore, the active prevention of dysbacteriosis after chemotherapy in children with ALL will not only enhance body immunity and reduce the incidence of complications but may also profoundly improve long-term prognoses.

E. coli, the main physiological symbiotic bacteria, forms an intestinal biological barrier with other anaerobic bacteria, which is resistant to the colonization and invasion of pathogenic bacteria, and plays a role in inhibiting the growth of opportunistic pathogens. ³⁹ In addition, it has a stimulatory function that is similar to antigens, and mediates specific and non-specific immunity. ⁴⁰ This study showed that chemotherapy also resulted in a significant reduction of E. coli. However, on the seventh day after chemotherapy, E. coli levels recovered greatly and exceeded that after chemotherapy. We postulate that MTX, especially at high doses, can damage the gastrointestinal structure, which may then affect anaerobic bacteria including Bifidobacteria colonization, because Bifidobacteria must have tight conjunctions with the intestinal epithelial cell to survive and form the biological barrier in physiological conditions. However, E. coli do not depend on intestinal epithelial cells and can grow in the intestinal tract. The quicker recovery of E. coli colonies compared with Bifidobacteria after influence by external factors, may be because E. coli is a kind of facultative anaerobe and can grow in aerobic and anaerobic environments. Owing to the great reduction of other aerobic and anaerobic bacteria during chemotherapy, the peripheral biological antagonism obviously weakens, and thus provides a favorable microhabitat for the propagation of E. coli. The quick proliferation of E. coli might overgrow and easily result in strain imbalance, and then cause dysbacteriosis or dual infection, as well as septicemia.

Our study also showed that there was a significant difference of intestinal flora between the ALL group and the healthy control group before therapy, which was likely caused by children with ALL receiving other therapies before admission. During the treatment process, various kinds of chemotherapeutics that destroy DNA and RNA, or inhibit the synthesis of DNA, RNA and proteins, were used in the children. In addition, repeated application of antibiotics for concurrent infections destroyed the ecological balance of the intestinal flora. The relationship between the prevalence, improvement rate and prognosis of children with leukemia deserves further study and investigation. Specifically, our data indicate that due attention should be paid to intestinal microecology and the microenvironment with rational prevention against dysbacteriosis and proper use of antibiotics to regulate the intestinal colony balance.

Author contributions: All authors participated in the design, patient screening, experiments, interpretation of the studies, analysis of the data and review of the manuscript. YH, WY, JD, YZ, ML and HL performed patient diagnosis, treatment and patient screening; HL and ZH conducted the experiments; and YH and KKW wrote the manuscript and made the figures.