Yu Zhang · Zhi Zhang · Jianzhang Ma · Bo Luo ·Guiquan Zhang · Guocai Zhang · Kexin Yang ·Gang Wei

Abstract This study aimed to investigate the effect of LyPB on the intestinal microf lora of giant pandas with indigestion, using high-throughput sequencing (HTS) technology. The species distribution and microf loral density and diversity before and after administration of the LyPB probiotic agent were analyzed. LyPB evidently has the ability to adjust the f loral imbalance in the panda’s intestine.To test the effects of LyPB on the microf lora of the panda gut, fecal samples were taken from a healthy giant panda(Anan) without administration of LyPB and from a dyspeptic giant panda Yangyang before and after LyPB administration. Compared with the sample obtained from healthy Anan(anan-c) and that obtained from dyspeptic Yangyang before LyPB administration (yangyang1), the sample taken from Yangyang (yangyang2) after LyPB administration displayed a signif icant increase in the operational taxonomic unit index. An increase in the Chao index indicated an increase in the microf loral richness, while an increase in the Shannon index indicated an increase in microf loral diversity. At phylum and genus levels, a signif icant increase was observed in the density of probiotic bacteria of phylum f irmicutes, genus Streptococcus, while a drastic reduction in the density of Escherichia coli / Escherichia coli Shigella/bacteria of genus Shigella was observed. Data obtained in this study shows that LyPB preparations successfully improve the microbial structure within the panda’s intestinal canal by signif icantly increasing the effective microbial community and decreasing the number of pathogenic microbes.

Keywords High-throughput s equencing · Intestinal fl ora ·Pandas · Indigestion · LyPB

Introduction

LyPB is a probiotic with strong cellulose decomposition capacity. It containsPaenibacillus cookiiLS006, a bacterium indigenous to pandas (Zhang et al. 2017), as well asBif idobacteriumandLactobacillus acidophilus(Qu et al.2017; Lin et al. 2017), which are the main probiotic groups(?hman and Simrén 2013) in the stomach and small intestine of pandas and exert mutualistic characteristics. Probiotics promote panda health (Lyra et al. 2016) by maintaining the micro-ecological balance within the stomach and intestine.Bif idobacteriumandL. acidophilusalso produce lactic acid and acetic acid to increase the uptake of calcium,phosphorus, and iron, improve iron and vitamin D digestion,produce vitamins K and B, reduce digestion of cholesterol and reduce radiation effects (Che Othman et al. 2012; Guandalini et al. 2000; Wagner and Balish 1998). A panda’s intestine contains a variety of microorganisms that inf luence the animal’s health (Gaon et al. 2003; Kale et al. 2005; Dahiya et al. 2005). The giant panda (Ailuropoda melanoleuca) is a rare and endangered species (Yan et al. 2017). Giant pandas primarily consume bamboo (Zhu et al. 2018); however, they have short digestive tracts (Wei et al. 2000) and are highly prone to imbalances of the intestinal f lora (He et al. 2017).We previously developed a probiotic agent for giant pandas(LyPB), composed ofBif idobacterium longumBL-01,Bif idobacterium adolescentisBA-10,Lactobacillus acidophilusLP-02,Bif idobacterium infantisBB-02, andPaenibacillus cookiiLS006. The probiotic combination ofDobacterium longumBL-01,Bif idobacterium adolescentisBA-10,Lactobacillus acidophilusLP-02, andBif idobacterium infantisBB-02 has already been approved by the Food and Drug Administration in China to be used as a health product.P.cookiiLS006, the predominant intestinal bacteria extracted from the excreta of the giant panda, was tested for its toxicity through experiments carried out in mice.P. cookiiLS006 was found to be non-toxic and free from side effects.

The results of the strain-specific performance tests showed that different groups of bacteria have varying physiological functions, such as regulating intestinal f lora,facilitating excretion, enhancing oxidation resistance, and improving immune function. In this study, LyPB probiotic preparations were made by culturing, mixing, microencapsulating, and freeze-drying the f ive previously identif ied probiotic agents (LyPB). They were then added to the diet of the giant pandas, in the hope that they could maintain the balance of the intestinal f lora and allow microbial propagation to improve the immune function.

Materials a nd methods

Experimental m aterial a nd animals

The LyPB probiotic agent for administration to giant pandas was stored in the microbiology laboratory of Northeast Forestry University prior to use. Glucoamylase, agar powder,peptones, and MnSO 4 ·7H2Owere purchased from Beijing Ordered Star Biological Technology Co., Ltd.

This study was performed in China Conservation and Research Center for the Giant Panda (Ya’an, Sichuan Province). The experimental animals included two adult male giant pandas: (A) Yangyang (16 years old), with indigestion and disorder in the intestinal f lora under natural conditions and (B) Anan (9 years old), a healthy specimen that provided the feces used as the healthy samples control.

Drug a dministration

Preparation of the giant panda probiotic preparation (LyPB):ThePaenibacillus cookiiLS006 strain, which has a strong cellulose-degrading capacity, was previously screened from giant panda feces, and identif ied using 16sRNA as a species underPaenibacillus. After enrichment through in vitro fermentation, the strain was freeze-dried and used to prepare the probiotic agent. Then, LS006Bifi dobacterium longumBL-01,B. adolescentisBA-10,Lactobacillus acidophilusLP-01, andB. infantisBB-02 were mixed at a ratio of 1:1:1:1:1 to produce the giant panda probiotic preparation,known as LyPB, which has a probiotic content of 10 9 CFU/g(Zhang et al. 2017). The preparation was stored in the microbiology laboratory of Northeast Forestry University prior to use.

Yangyang was administered 10 g LyPB per day; other food was provided ad libitum, and his living conditions were unchanged. During the 7-d pre-experimental and 30-d experimental periods, a fecal sample was collected under sterile conditions every day, freeze-dried, and stored.

During the 7-d pre-experimental period, Yangyang was kept on his normal diet and a fecal sample was collected under sterile conditions every day, freeze-dried and stored with the sample named “yangyang1”. After this period,Yangyang was administered 10 g LyPB per day for 30 days,other food was provided ad libitum, and his living conditions were unchanged. During a 30-day period, a fecal sample named “yangyang2” was collected under sterile conditions every day, freeze-dried and stored. A normal fecal sample was obtained from Anan under sterile conditions, freezedried, and stored as control sample 2 with the sample name“anan-c”.

DNA e xtraction a nd sequencing

After the experimental period, the fecal sample yangyang2 was collected under sterile conditions and freeze-dried.anan-c and yangyang1 samples were also obtained. Total DNA was extracted from each sample and amplif ied via polymerase chain reaction and sequenced at Sangon Biotech(Shanghai) Co., Ltd. The bacterial and archaeal 16S rRNA sequence database and functional gene database were used for data comparison. After sample sequences had been distinguished through barcodes and subjected to quality control and f iltering, alpha diversity and bacterial f lora distribution analyses of the sequences were performed.

Bioinformatics anal ysis

High-throughput technology can directly identify genes and simultaneously sequence millions of DNA molecules.In addition, it does not require isolation and culture of bacteria and enables quick and comprehensive identif ication of the microorganism genes. The new generation of sequencing technology mainly includes 454 sequencing technology, Solexa sequencing technology, and SOLiD sequencing technology, which can rapidly carry out deep sequencing of samples. Vargas-Albores et al. ( 2017) used a high-throughput sequencing method to examine the intestinal microbial community of farmed shrimp to detect whether the farm sewage had been treated or not. Standen et al. ( 2015) studied the intestinal f lora structure of tilapia after introducingLactobacillus reuteri,Bacillus subtilis,and other probiotics into the feed by the high-throughput sequencing method.By comparing the sequencing data, researchers can evaluate the genetic information, composition, and richness of the f lora in the samples. The similarities between samples can be analyzed to uncover links between intestinal f lora and diseases at the genetic level (Sek et al. 2003; Li et al.2004). High-throughput sequencing technology improves the understanding of intestinal f lora (Gaon et al. 2003).

Sample sequences were distinguished using barcodes, and quality control analysis was performed to eliminate nonspecif ic amplif ied sequences and chimeras. Data analysis was performed using Mothur, Usearch, and Cytoscape, and rarefaction analysis was performed using Mothur (McFarland et al. 1994). Bacterial diversity indices were determined using the Shannon, Simpson, and ACE (abundance-based coverage estimator) indices, and Venn diagrams were generated using Excel and Mothur.

Results and di scussion

Shannon

The Shannon—Wiener curve ref lects the microbiological diversity of the sample (Fig. 1). When the sequencing quantities occur at various sequencing depths, they provide an indication of the biological diversity. The curve tends to be f lat when the sequencing data is sufficient and ref lects most of the microbiological information in the samples.

The Shannon—Wiener diversity index was used to analyze the bacterial diversity of the samples. The number of sequences randomly selected from the samples formed the horizontal coordinate, the corresponding Shannon—Wiener diversity index was the vertical coordinate, and the result was a sparse curve. The calculation for Shannon—Wiener diversity index (H ?) is as Eq. 1:

whereH= Shannon—Wiener diversity index;Pi= The percentage of the individual number of the species compared to the total species;S= the total number of different species in a population.

From Fig. 1, the number of operational taxonomic units(OTUs) in the samples increased with an increase in the number of sequences. However, the extent of this increase gradually declined, indicating a fairly high quality and a certain depth of sequencing. As shown in Fig. 1, for each sample, there was a sharp initial increase in the Shannon index, followed by a rapid decline in the gradient of the curve as it approached saturation. This indicates that, even with a large number of sequences, the microf loral diversity in the samples was adequately measured and that most of the microbiological information was ref lected.

Analysis o f bacterial fl ora diversity

Fig. 1 Shannon index has the largest increase in the number of sequences in a certain range at the beginning, and its curve quickly declines and becomes saturated, indicating that the quantity of sequencing data is sufficient

Table 1 Improvements in the number of OTUs and alpha diversity of fecal microf lora before and after LyPB administration

The number of OTUs (at the similarity level of ≥ 97%)in each sample, which reflects species abundance, was calculated using Mothur. Table 1 shows the results of the analysis of the bacterial f lora diversity, performed using samples anan-c, yangyang1, and yangyang2.

As shown in Table 1, the OTUs of yangyang2 were 4536 after administering LyPB. This was twice the value of yangyang1, suggesting that the biodiversity of the intestinal f lora had substantially improved. Based on the analysis of the phyla and genera of the intestinal f lora, yangyang2 had increased intestinal f loral growth after the administration of probiotics.

ACE and Chao values evaluate the total number of bacterial species in the sample. The values increase with the abundance of microbial species. From the samples obtained, yangyang2 had the highest ACE and Chao values, 161,896.00 and 52,715.69, respectively. The ACE and Chao values of yangyang1 were the lowest, 63,058.02 and 24,930.59, respectively.

Analysis of the alpha diversity also revealed that the administration of LyPB resulted in a higher microf loral diversity in yangyang2 as compared to yangyang1 and ananc. Furthermore, the coverage values for all three samples exceeded 0.9, thereby indicating high coverage rates and high reliability of the experimental data.

Venn diagrams for operational taxonomic unit (OTU)distribution

Venn diagrams were generated to determine the numbers of shared and unique OTUs between the samples and illustrate similarity and overlap. VennDiagram (package R) was used to analyze and draw the Venn diagram of OTU distribution between the three samples, as shown in Fig. 2.

As shown in Fig. 2, 2362 unique OTUs were identif ied in anan-c, while 2154 unique OTUs were identif ied in yangyang1; There were only 160 shared OTUs between anan-c and yangyang1, which accounts for 7% of the average, indicating the presence of a relatively serious microf loral disorder in Yangyang. Comparing yangyang1 and yangyang2 stool samples, collected before and after LyPB administration, 4536 unique OTUs were identif ied in yangyang2. This was more than twice that in yangyang1, indicating that the diversity of intestinal f lora in Yangyang was signif icantly enhanced by LyPB ingestion.

Fig. 2 Venn diagrams of OTUs show that LyPB promoted the diversity of intestinal f lora

Analysis of mic rof loral distribution at the phylum and genus levels

Bacterial microf loral distribution at the phylum level

Bacterial microf loral distribution in the samples at the phylum level is shown in Table 2 where four phyla with relatively high abundance are listed. The species distribution of yangyang1 and yangyang2 fecal samples were analyzed at the phylum level, and the results indicated changes in the proportions of the various intestinal f lora in Yangyang after LyPB administration. The proportions of Firmicutes andActinobacteria, comprising numerous probioticspecies, increased from 53.55% to 83.44% and from 0.02%to 0.079%, respectively. Meanwhile, the proportions ofProteobacteriaandCyanobacteria, which primarily comprise pathogenic bacterial species, decreased from 42.96%to 15.26% and from 3.45% to 0.15%, respectively.Lactobacillus acidophiluslp-02 in LyPB belongs to the phylum Firmicutes, which can adjust the balance of intestinal f lora and inhibit the proliferation of undesirable intestinal microorganisms (Sanders and Kleanhammer 2001).Bif idobacterium longumbl-01,Bif idobacterium youthba-10, andBif idobacterium infantisbb-02, all belong to order actinomycetes,which have biological barriers, nutritional effects, anti-tumor effects, immune enhancement effects, gastrointestinal function improvement, among other functions (Muralidhara et al.1977; Yang et al. 2004).

Table 2 Distribution of improvements in sample microf lora by LyPB at the phylum level

Bacterial microf loral distribution at the genus level

Bacterial microfloral distribution in the samples at the genus level is shown in Table 3. From Table 3,Streptococcus, which includes many species of probiotic bacteria, was found to be the most prevalent microbe present in the sample obtained from Anan. The amount ofStreptococcusin yangyang1 was signif icantly lower than that in anan-c, while the number of species of pathogenic bacteria was higher in yangyang1. After LyPB administration, the amount ofStreptococcusin the fecal sample of yangyang2 increased signif icantly from 14,194 to 57,102. In addition, bacteria of the generaEscherichiareduced signif icantly from 22,862 to 7378. The amounts ofClostridiumsensu stricto andStreptophyta,comprising numerous pathogenic bacterial species,andStreptococcusin yangyang2 was comparable to that of anan-c.

The aforementioned data indicate that LyPB administration can increase the proportion of benef icial microf loraand decrease the number of pathogenic microf lora in giant pandas with indigestion. Thus, LyPB is effective in treating indigestion caused by intestinal f lora imbalance and improves immune function. As a 30-day experimental period is considered relatively short for giant pandas, intestinal f lora imbalance in giant pandas can be further rectif ied upon longterm LyPB administration.

Table 3 Distribution of improvements in sample microf lora by LyPB at the genus level

Microf loral composition

Taxonomic analyses reveal comparative results between one or more samples at various levels. Two types of information can be obtained from the results: (1) the types of microbial species in the sample and (2) the number of sequences in each microorganism in the sample, i.e., the relative abundance of microorganisms. We analyzed the composition of each sample group at the genus level and plotted pie charts(Figs. 3 A—C) of relative species abundance at the genus level for all samples, based on the 50 bacterial genera with the highest relative abundance. From the charts, changes in bacterial f lora and microf loral distribution in anan-c, yangyang1, and yangyang2 at the genus level could be directly observed.

It can be seen from Fig. 3 that the f lora of each sample has its own distribution characteristics at the genus level.

As shown in Fig. 3,Streptococcus(98.5%) was the dominant intestinal microbe in Anan, the healthy giant panda.A signif icant disorder in intestinal f lora was observed in yangyang1, asStreptococcusonly accounted for 25.8% of the intestinal f lora, whileEscherichia/Shigella(41.63%)andClostridiumsensu stricto (24.08%) accounted for relatively large proportions. After LyPB administration, a signif icant improvement was observed in the intestinal f lora of yangyang2, asStreptococcusaccounted for 75.85% of the intestinal f lora, thereby dominating the intestinal f lora once again, while the proportions ofEscherichia/ShigellaandClostridiumsensu stricto decreased to 9.77 and 5.99%,respectively.

High throughput gene sequencing data ref lected that the intestinal canal f lora OTU of Yangyang (with dyspepsia)was 2154, while that of the healthy panda, Anan, was 2362.Although the OTU values are similar, the intersection was only 160, accounting for 7% of the mean value. This indicated that the normal f lora in Yangyang’s intestinal canal was very limited and severely disturbed.

After administering LyPB in the diet for 30 days, the intestinal canal f lora OTU value of the panda with dyspepsia increased to 4536 and doubled in value. This conf irmed that LyPB effectively improved the diversity of the intestinal canal f lora of Yangyang.

Fig. 3 2D pie chart of improvements in sample species abundance by LyPB

The intestinal f lora of the healthy panda Anan, had probioticStreptococcus(98.5%) as the dominant bacteria,while in Yangyang, it only accounted for 25.8%. Further,pathogenic bacteriaEscherichia/Shigella(41.63%) andClostridiumsensu stricto (24.08%) were both present in large amounts in the intestine of Yangyang. The excessive proliferation of these bacteria produces detrimental substances such as ammonia, amine, hydrogen sulf ide, skatole,indole, nitrite, and bacterial toxins that will cause further deterioration in health (Bose et al. 2013). After the administration of LyPB, the intestinal f lora of Yangyang improved substantially. ProbioticStreptococcusaccounted for 75.85%of the f lora and was re-established as the dominant bacteria.The presence of pathogenic bacteriaEscherichia/ShigellaandClostridiumsensu stricto decreased to 9.77 and 5.99%,respectively.

Conclusion

From the above study, it can be seen that the probiotic preparation LyPB can signif icantly limit the production of harmful bacteria, prevent pathogenic bacterial infections,restrain intestinal canal putrefaction, promote the proliferation of probioticStreptococcusin large amounts, and clean the intestinal canal.

As the duration of this experiment was comparatively short, more tests will be arranged in the future that span a longer period of time. In the meantime, the research and development of the product series will be focused on and the probiotics fodder will be developed for experiments in other herbivores. The production of the LyPB product series will be implemented and then promoted for large scale application.