Patients and group definition
A total of 105 patients over 18 years old admitted to the respiratory intensive care unit, critical care unit and emergency department of Peking University Third Hospital (PUTH) from September 2019 to December 2020 who were experiencing MV for more than 48 h and were suspected of having VAP were collected. The criteria of VAP were defined according to the Guidelines for the diagnosis and treatment of hospital-acquired pneumonia and ventilator-associated pneumonia in Chinese adults (2018 Edition): Chest X-ray or CT showing new or progressive infiltrating, consolidation, or ground glass shadows, accompanied by 2 or more of the following—temperature over 38 °C, purulent airway secretion, and white blood count above 4–10 × 109/L [7]. Those suspected of having VAP were defined as meeting at least one of the 4 criteria. 16 of 105 recruited patients did not have complete clinical data or had insufficient samples; 89 samples were used to extract DNA, while 6 failed to provide enough DNA for further experiments. 83 patients had complete clinical data and sufficient samples and DNA and were finally admitted to this study (Fig. 1).
To optimize the real-time pathogen detection process, we designed our study into 3 parts: Firstly, detect the efficiency of host genome depletion, 12 samples with enough volume were separated into 2 parts respectively, and DNA was extracted with or without host DNA depletion (Fig. 2); secondly, compare the differences in PCR amplification before library construction, 29 low-quality DNA samples (OD value 260/280 or 260/230 were out of 1.8–2.0) were used to build the library with and without PCR amplification (Fig. 3). In the end, all 83 DNA samples were extracted after host genome depletion and sequenced with the non-PCR amplification method (Fig. 1).
This study was approved by the Ethics Committee of Peking University Health Sciences (IRB00001052) and the Ethics Committee of Peking University Third Hospital (M20200352). All patients or families of unconscious patients were informed and agreed to participate in this experiment. The genetic resource management was proved by China Human Genetic Resources Management Office ([2021] GH3154).
Sample collection
2 ETAs from patients suspected of having VAP were collected within 24 h: one was used for microbiology culture in the clinical laboratory of PUTH, and the other was taken to our laboratory for further research. A 4X volume of sterile PBS was added to the ETA sample, pipetted and aliquoted as 1 ml/tube. After centrifugation for 15 min at 8000 rpm, the sediment was collected, snap frozen in liquid nitrogen, and then stored at −80 °C (Fig. 2).
Positive control strains collection
Standard strains of S. aureus, A. baumannii, Stenotrophomonas maltophilia (S. maltophilia), P. aeruginosa, Streptococcus pneumoniae (S. pneumoniae), Escherichia coli (E. coli) and K. pneumoniae were obtained from American Type Culture Collection (ATCC) and the Clinical Culture Department of PUTH (Table 2). Monoclonal colonies were selected after overnight culture, dissolved in bouillon broth, shaken at 37 °C for 8 h, and centrifuged at 13,000 rpm at 4 °C for 1 min. The supernatant was disposed, and the pellet was resuspended in normal saline to make a suspension of 4.5 McFarland (McF). Then, samples were divided into 1 ml/tube and centrifuged at 13,000 rpm at 4 °C for 1 min. The sediment was collected, snap frozen by liquid nitrogen and stored at −80 °C.
Negative control collection
Sterile saline solution was collected by aspiration through the sputum aspirator as a negative control, and the negative control was processed in parallel with the study samples.
Host depletion with saponin
Sediments were resuspended in 250 μl of sterile PBS, and 200 μl of 5% saponin (S0019, Tokyo Chemical Industry, Tokyo, Japan) was added, followed by pipetting. Samples were placed at room temperature for 10 min before 350 μl of nuclease-free water (NF-water) was added and incubated for another 30 s. Then, 12 μl of 5 M NaCl was added and the tubes inverted. Next, the samples were centrifuged at 8000 rpm at 4 ℃ for 5 min, the supernatant was discarded, and the sediment was resuspended in 100 μl of sterile PBS. 100 μl of heat-labile salt active nuclease (HL-SAN) Buffer (100 mM MgCl2 in 5 M NaCl) and 10 μl HL-SAN DNase (25,000 units, 70910-202, Articzymes, Tromso, Norway), were added, and the samples were shaken at 37 °C for 15 min. Finally, the samples were centrifuged at 8000 rpm at 4 °C for 5 min, the supernatant was discarded, and the sediment was washed with 1000 μl of sterile PBS two times. The same procedure was used in the undepleted group, but all reagents were replaced by NF-water (Fig. 2).
DNA extraction
BSCC45S1E kits and GenePure Pro (Bioer Technology, Hangzhou, Zhejiang, China) were used for DNA extraction. Lysozyme was dissolved in TET buffer and mixed by shaking. A 180 μl mixture was added to each sample and incubated for 30 min at 37 °C after shaking. Then, 20 μl of Proteinase K and sample were added to columns 1 and 7 of the kit and placed into the machine. DNA concentration and purity were determined by a NanoDrop after extraction.
Library construction, sequencing and data analysis
The undepleted DNA library construction and depleted DNA non-PCR-amplified library construction were performed using a rapid sequencing kit (SQK-RAD004, ONT, Oxford, UK) and rapid barcode kit (SQK-RBK004, ONT, Oxford, UK), while the depleted DNA PCR- amplified library construction was performed using a rapid PCR barcode kit (SQK-RPB004, ONT, Oxford, UK). The Non-PCR-amplified library construction method was performed following the instructions, and 400 ng of DNA from each sample was used (when the maximum amount of 7.5 μl of DNA was less than 400 ng, then 7.5 μl was used) for sequencing. The PCR amplification library construction method was carried out according to the instructions. 5 ng of DNA was used for each sample, the extension time was shortened from 6 to 4 min, and the amplification cycle was increased from 14 to 25 cycles [18]. Sequencing was performed using MinION (ONT, Oxford, UK) and R9.4 flowcellls (FLO-MIN106D, ONT, Oxford, UK). Raw data collection and base-calling were performed using MinKNOW (v.19.12.5, ONT, Oxford, UK) and Guppy (v.3.2.10, ONT, Oxford, UK) software. The data were collected and analysed in real time. Sequencing was continued for 1–2 h after the pathogens may cause VAP (the pathogens leading to VAP was determined according to the previous research results and Chinese VAP Guideline [7, 18, 19] and the isolation reports of VAP infection pathogens in the PUTH, the possible pathogens of VAP were determined by clinicians from Respiratory Department of PUTH) were identified. If no more pathogenic bacteria were detected, sequencing was stopped (Figs. 3 and 4).
The raw data generated by sequencing were filtered using NanoFilt (v.2.7.1) for joint sequence resection (–headcrop 150 –tailcrop 50) and low-quality segments (-q 7 -l 500) and NanoPlot (v.1.32.1) for filtration quality statistics and visualization. Minimap2 (v.2.17) was used to align the filtered clean FASTQ file with the Human GRCh38 Genome (NCBI). SamTools (v.1.11) was used to extract the unaligned sequence (-f 4) and to convert the generated data to FASTQ format. Kraken2 (v.2.1.1) was used for sequence classification (Fig. 3).
qRT-PCR
qRT-PCR was used to confirm 7 identified pathogens in this study. In each sample, 10 μl of SYBR Master Mix (11184ES08, Yasen, Shanghai, China), 7.2 μl of NF-water, 0.4 μl of forward and reverse primer (synthesized by Beijing Ruibio Biotech Co., Ltd) (Additional file 1: Table S1), and 2 μl of DNA samples were added. Bacteria from ATCC were extracted as a positive control group of pathogens (Additional file 1: Table S1). DNA of the A549 cell line was extracted and used as a human genomic positive control group. The PCR cycling conditions were set as pre-incubation at 95 °C for 2 min; amplification for 40 cycles at 95 °C for 10 s and 60 °C for 30 s; and the final melting curve was 95 °C for 15 s, 60 °C for 60 s, and 95 °C for 15 s. The results were analysed using CT values.
Statistical analysis
The qRT-PCR results were analysed using a T test, and the sensitivity and specificity used a binomial distribution. P value less than 0.05 was considered to indicate a significant difference. R (v.4.0.3) and SPSS (v.19) were used for statistical analysis, and the tool http://vassarstats.net/ was used for the sensitivity and specificity calculation. The images were produced using OriginPro 2017C (b8.4.2.380), Microsoft office PowerPoint 2019, R and Adobe Photoshop CC 2018.
Data availability
All clinical samples metagenomic sequencing datum are available via China National Center for Bioinformation (www.cncb.ac.cn) under Project PRJCA006892.