The 199 strains CRAB were isolated from patients in various departments of Affiliated Hospital of Guizhou Medical University, Guizhou, China from 2017 to 2019. Standard A. baumannii ATCC19606, pan-resistant A. baumannii strains are preserved in the Key Laboratory of Modern Pathogenic Biology, School of Basic Medicine, Guizhou Medical University.

The genomic DNA of A. baumannii was extracted using the kit (BIOTEKE, Beijing, China). Seven housekeeping genes were amplified by PCR. The primers used are listed in Table 1. The PCR reactions were performed using PrimeSTAR^® Max DNA Polymerase (Takara, Japan) with the following amplification parameters: 95°C for 5 min; 94°C for 30 s, 50°C for 30 s, and 72°C for 40 s, 30 cycles; 72°C for 10 min.

The amplified products of MLST typing of A. baumannii were sequenced, and the sequence results were compared on the MLST comparison website of A. baumannii. The sequencing results were submitted to the website for comparison to obtain allele number and ST type. The Bio Numerics 7.1 software is used for typing data processing, and generating minimum spanning tree (MST).

The minimal inhibitory concentration (MIC) values of A. baumannii isolated from clinical isolates were measured by the micro-dilution broth method. The antibiotics determined were imipenem, meropenem, ertapenem, and biapenem. There were three repeats for each strain and antibiotic. Except for 5 gradients of biapenem, 6 gradients were done. The concentration of the mother liquid before dilution is shown in Table 2. The results were judged by the Institute of Clinical and Laboratory Standards (CLSI2012).

In this study, the bla_OA–23 and bla_OXA–51 were selected as molecular markers for rapid identification of A. baumannii. The primers are listed in Table 3, and the PCR reaction system is 15 μL, including the premixTaq enzyme 7.5 μL, the primers 1 μL with 10 μM, DNA template 1 μL, and the ddH2O 4.5 μL. Meanwhile, the PCRs were performed at 94°C for 5 min, followed by 35 cycles of 30 s at 94°C, 30 s at 55°C, 30 s at 72°C, and a final extension of 10 min at 72°C. The 5 μL PCR products were sampled in 1% agarose gel and 30 min was electrophoretic at a constant voltage of 100V. The photos were taken by WD-9413B gel imaging analyzer.

The AB77 (weak imipenem resistance) and AB98 (strong imipenem resistance) were selected for transcriptome sequencing with three biological repeats per sample. After resuscitation of CRAB77 and CRAB98, respectively, 1 mL solution was absorbed and cultured in 100 mL LB liquid medium. After overnight incubation at 37°C 220 rpm until logarithmic growth, the organisms were collected by centrifugation, and three biological replicates were performed in each group, totaling six samples. The collected bacterial precipitates were immediately liquid nitrogen snap-frozen and stored in the refrigerator at −80°C. The total RNA of these samples was extracted using Trizol reagent (Sigma-Aldrich, United States) according to the specification described method. The RNA concentration and quality were measured using a Nanodrop 2000 micro-spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and agarose gel electrophoresis. RNA-seq library construction and RNA sequencing were performed by the Meiji Biomedical Technology Co., Ltd. (Shanghai, China).

The raw data was filtered to remove contaminated and low-quality sequences from connectors and to obtain clean reads. Bowtie2-2.5.0 was used to map the genome of the filtered sequence (Langmead and Salzberg, 2012). The reference genome of A. baumannii (ATCC 19606) was downloaded from GenBank (NZ_CP045110.1). The gene expression levels were calculated based on TPM (transcript per million) (Zhao et al., 2021). Differential expression analysis between CRAB77 and CRAB98 was performed using the DEGSeq2 R package. The significance of the DEGs was determined with a P-value < 0.05 found by DEseq2. Transcripts with log 2 (Fold Change) > 1 and P-value < 0.05 were considered significant differential expressions. KOBAS software was used to test the statistical enrichment of differentially expressed genes in KEGG pathways. Significantly enriched KEGG pathways was identified by a P-value < 0.05.

qRT-PCR was performed to confirm the DEGs identified using RNA-seq. The primer sequences of candidate genes and a housekeeping gene (16S rRNA) were designed using Primer Premier 5 (Table 4). The first strand of cDNA was synthesized by PrimeScript TMRT reagent Kit with gDNA Eraser (TaKaRa, Beijing, China) according to the manufacturer’s instructions. The qRT-PCR was carried out with a PikoReal 96 real-time PCR system (Thermofisher Scientific) using PowerUp™ SYBR™ Green Master Mix (Thermo Fisher Scientific, Waltham, MA, USA). The sense and anti-sense primers were designed by Primer Premier 5 and they were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). Sequences of the primers used in this study were shown in detail in Table 4. The reactions were performed with three biological and technical replicates per sample. The data were analyzed using the 2^–ΔCt method (Schmittgen and Livak, 2008).

Statistical analysis was performed using SPSS21.0. All experiments were performed in triplicate. Statistical differences between groups were assessed by the two-tailed Student’s t-test and one-way analysis of variance (ANOVA) followed by Tukey’s multiple-comparison test or by the nonparametric Mann–Whitney U test and Kruskal-Wallis test, followed by Duncan’s multiple-comparison test. A P-value < 0.05 was considered statistically significant.

The sequences of these 7 housekeeping genes in 199 A. baumannii strains were submitted to the MLST database of the Pasteur Institute¹ to investigate whether there were base substitutions or deletions. The sequences obtained from housekeeping gene sequencing of each strain were submitted to the database, and the corresponding alleles of each sequence were queried, to obtain the ST model of each strain. The results showed that there were 9 different sequence types (Table 5), among which ST136, ST191, and ST208 had high frequency in the hospital.

The minimum spanning tree showed that the ST types of the 199 strains are relatively concentrated, and all of them originated from ST191 (Figure 1). In addition, there is only gpi difference between ST136/ST195/ST368/ST369/ST208/ST540 and ST191. While there are two alleles difference between ST457 and ST191, the genetic relationship is close. It is reported that the mutation rate and recombination rate of gpi in the MLST typing scheme of A. baumannii are higher than those of other housekeeping genes, and it is easy to mutate in the process of evolution. It is found that a similar situation in the study of A. baumannii typing, suggests that A. baumannii MLST scheme has a better resolution on the housekeeping gene gpi. Among the 199 strains of CRAB, there is only one ST, which is widely distributed in various departments. Among these departments, the ICU region is the most distributed. However, the ST2206 does not belong to these ST, and there are six allele differences between ST2206 and ST191, and the ST2206 is isolated in the comprehensive ICU region.

To understand the classification position of these identified strains in the entire A. baumannii database, the ST of all strains were placed in the A. baumannii ST database. The results showed that the ST136 / ST195 / ST368 / ST369 / ST208 / ST540 / ST191 belongs to CC208, and the ST2206 distant relatives (Figure 2).

According to the analysis of the source distribution of 199 CRAB strains, the results showed that most of these clinical strains came from sputum, and a total of 171 strains of each ST type were detected in these sputum samples, while the distribution was less in other samples (Figure 3A). Based on the analysis of sputum samples in various departments, these CRAB were found in all clinical sputum samples in many departments, while there were few types of CRAB samples in the remaining departments (Figure 3B). This result implies that the ST is widely prevalent in all departments in the hospital, and it is necessary to focus on detecting the presence of CRAB in sputum samples to timely control the spread of CRAB.

The result of MIC values for the 199 A. baumannii strains showed that the MIC values for biapenem were concentrated at 128 μg/mL, accounting for 44.5%; the MIC value of meropenem was concentrated at 128 μg/mL, accounting for 39%; the MIC value for ertapenem was concentrated at 512 μg/mL, accounting for 39.5%; as well as the MIC values for imipenem were concentrated at 64 μg/mL, accounting for 51% (Figure 4 and Supplementary Table 1). These results indicate that all experimental strains were highly resistant to the four carbapenems antibiotics, among which the experimental strains had the strongest resistance to ertapenem.

According to the ST type classification of 199 CRAB strains, MIC values of these 199 CRAB strains against ertapenem were calculated. The results showed that ST136, ST191, ST195, and ST208 were more resistant to ertapenem than ST368, ST369, ST457, and ST540. However, the ST2206, which is distantly related to STs strain, also had a MIC value of 1,024 μg/mL. This result indicates that ST2206 with strong antibiotic resistance (Table 6).

In addition, according to the antibiotic resistance detection of these 199 strains to biapenem, the results showed that the MIC values of ST types concentrated at 128 μg/mL. The result showed that these ST types have the same resistance to biapenem. However, the ST2206 showed lower resistance to biapenem than other ST types (Table 7).

According to the detection of the resistance for these 199 CRAB strains to meropenem, it was found that the MIC values of STs to meropenem in these 199 CRAB strains were higher, while the ST2206 showed low resistance to meropenem (Table 8).

Based on the detection of the resistance STs to imipenem in 199 CRAB strains, the results showed that MIC values were concentrated in 32 μg/mL-128 μg/mL, indicating strong resistance. Moreover, the ST2206 also showed resistance to imipenem, but the resistance was low, with a MIC value of 32 μg/mL (Table 9). According to this data, it was be defined weakly resistant strain with MIC values ≤ 32 to imipenem, while strains with MIC value > 32 was defined as strongly resistant strain. By this definition, ST136 and ST208, have quite different resistance to imipenem; In this study, weakly resistant strains accounted for 18% in ST136, and 47.22% were weakly resistant strains in ST208. This result indicates that different STs may have different resistance mechanisms to imipenem.

To understand the distribution of antibiotic resistance genes, the bla_OXA–23, and bla_OXA–51 were detected in 199 CRAB, standard A. baumannii (ATCC19606), pan-resistant A. baumannii (PRAB) and SJZ24 by using PCR amplification. The results showed that the target sequence of bla_OXA–23 (Figure 5A) and bla_OXA–51 (Figure 5B) genes were found in these strains. The BLAST analysis was conducted after sequencing these target bands, and it was found that these target bands were all target antibiotic resistance genes. Statistical data showed that the detection rate of the bla_OXA–23 gene and bla_OXA–51 gene in CRAB was 92.5 and 100%, respectively (Figure 5C). These results suggest that the class D carbapenemases genes are widely distributed in these 199 CRAB strains, which may be one of the reasons for its high antibiotic resistance to carbapenems antibiotics.

To understand the resistance mechanisms of different STs to imipenem, the transcriptome sequencing analysis between ST208 (AB77) and ST136 (AB98) was performed in this study. The results showed that a sum of 159.60 million reads was generated and each sample provided an average production of 26.60 million. A mean of 26.60 million clean reads was obtained from each library with an adequate read ratio of 94.63% after removing adaptor sequences, N-containing reads, and low-quality. A mean of 20.44 million CDS-mapped reads was obtained from each library with an adequate read ratio of 80.11% (Supplementary Table 2). A total of 3,829 genes were mapped from all the samples. The resistance strain and sensitive strain gene expression levels were analyzed, and 390 differentially expressed genes (DEGs) were recognized (Supplementary Table 3). Among such DEGs, the resistance strain had 256 DEGs up-regulated and 134 DEGs down-regulated (Figure 6A). It is worth pointing out that some genes related to antibiotic resistance, efflux pump, and biofilm formation were up-regulated in the resistance strain (Figure 6B). The increased expression level of these genes may increase the antibiotic resistance of A. baumannii.

Among these DEGs, 19 genes may be involved the carbapenem resistance, including biofilm formation, efflux pump, peptidoglycan biosynthesis, chaperonin synthesis, and drug transporter. In the biofilm formation process, there were pilus assembly protein (Pap, FQU82_RS03605), urease accessory protein (UreD, FQU82_RS05390), OmpA family protein (FQU82_RS07900), NADP-specific glutamate dehydrogenase (gdhA), phosphoenolpyruvate–protein phosphotransferase (pstP), and amido phosphoribosyl transferase (purF, FQU82_RS13115). In addition, there were 5 genes involved in efflux pump, including LysR family transcriptional regulator (FQU82_RS08650), PACE efflux transporter (FQU82_RS09170), chlorhexidine efflux PACE transporter (aceI, FQU82_RS12055), MacB family efflux pump subunit (FQU82_RS03105), MacA family efflux pump subunit (FQU82_RS03110). Meanwhile, there were 3 genes may contribute to peptidoglycan biosynthesis, including D-alanyl-D-alanine carboxypeptidase PBP5/6 (dacC, FQU82_RS14285), endocytic transglycosylase (MltG, FQU82_RS14770), and UDP-glucose 4-epimerase (GalE, FQU82_RS00890). Additionally, there were 3 chaperonins, including DnaK, GroEL, and GroES may be involved in antibiotic resistance. It is worth mentioning that the MFS transporter and ABC transporter substrate-binding protein may participate in drug transport (Table 10).

To confirm findings of gene expression obtained from transcriptome data, 6 DEGs concerned with antibiotic resistance were chosen for qRT-PCR. These DEGs are mainly involved in biofilm formation, efflux pump, peptidoglycan biosynthesis, and chaperonin synthesis. As shown in, the 6 DEGs had a very similar expression pattern based on the transcriptome data and qRT-PCR, which indicates the trustworthiness of the transcriptomic analysis (Figure 7).