Introduction

Acinetobacter baumannii is a Gram-negative coccobacillus which is troublesome pathogen in hospitalized patients globally. A. baumannii has emerged as a major challenge in infected patients because of its multiple drug resistance [1, pp. 939–951]. Recent data show that A. baumannii is one of the top four most common pathogens isolated in the hospital environment between 2005 and 2017 in China [2, pp. S128–S134]. In global intensive care units (ICUs), the infection rate of A. baumannii accounts for 20% of the total number of infections [3, pp. 25–36]. The resistance profile of Acinetobacter spp. indicates that the resistance rate against the five commonly used antimicrobials, except amikacin, is increasing. The resistance rate of A. baumannii against carbapenems varied from 18 to 25% in different provinces of China [2, pp. S128–S134]. A. baumannii can cause a variety of nosocomial infections, most of which involve the respiratory tract and some involve skin wound infections and bacteremia [4]. This is corroborated by the report that A. baumannii is an increasing threat in burn centers because of its high multidrug resistance based on clinical data obtained from a comprehensive hospital covering 2007–2013 and the CHINET (the surveillance system for bacterial epidemiology and resistance in China) data [5, pp. 1718–1719].

Considerable attention has been focused on exploring the mechanism of drug resistance in A. baumannii in clinical settings [6]. One of the major factors associated with bacterial resistance to antimicrobials is the ability to form biofilms. Previous studies have shown that strains that can form biofilms are more resistant to antibiotics than strains that cannot [7]. Recent evidence shows that biofilm formation is closely related to drug resistance and virulence in A. baumannii [8, pp. 119–127]. It is reported that the biofilm-forming ability of A. baumannii in a solid–liquid interface is approximately three times higher than that of other Acinetobacter species [9, pp. 1–4]. A. baumannii can rapidly produce biofilms while forming antibiotic resistance, and the level of biofilm-specific resistance can vary according to the induction response of the biofilm population and the action mechanism of antibiotics [10, p. 817]. Cold plasma technology, as a new tool for purification of biofilm-contaminated surfaces, has received more and more clinical attention. The increase in the amount of A. baumannii biofilm could reduce the antibacterial effect of cold plasma, thereby improving its tolerance to cold plasma exposure [11, pp. 344–349]. The biofilm formation characteristics of A. baumannii are related to the expression of genes. The genes that have been reported to be associated with the formation of A. baumannii biofilm include csu locus, bap, abaI, adeFGH, and ompA [7]. csuE encodes a fimbriae assembly system, and the formation of fimbriae is an important step in the formation of biofilm on non-living surfaces [12, pp. 3473–3484]. Bap can support the development of mature biofilm structures [13, pp. 1036–1044]. The mutation of abaI will cause the failure of the acyl-homoserine lactone signals to be generated and further damage the maturation of the biofilm [14, pp. 951–957]. adeFGH is involved in the synthesis and transport of autoinducer molecules during the formation of biofilms [15, pp. 4817–25]. ompA encodes a porin protein involved in the formation of biofilms [16, pp. 3150–3160]. Previous reports have implicated the abaR gene in A. baumannii biofilm formation and disruption of abaR expression is associated with a decline in biofilm formation [17, pp. 1802–1805; 18, pp. 200–205; 19, pp. 1339–1346]. Although there have been advances in exploring the mechanism underlying biofilm formation, the role of abaR at the transcriptome level has not been clarified completely.

Therefore, in this study, we aimed to determine the role of abaR in A. baumannii biofilm formation through abaR knockdown and explore the mechanism regulated by abaR expression using high-throughput transcriptional sequencing.

Materials and Methods

Bacterial Strains and abaR Gene Knockouts

A. baumannii ATCC 17978 strain was obtained from Shanghai Santa Biotechnology Co. Ltd., Shanghai, China. The ATCC 17978 strain was cultured in Luria–Bertani (LB) medium at 37 °C overnight. The abaR gene knockdown was performed using homologous recombination method [20, pp. 3743–3751]. A recombinant DNA fragment (ΔabaR::Kn) consisting of kanamycin resistance gene (Kn) sequences (obtained from pKD4 plasmid using PCR) and the upper and lower homologous recombination arms for abaR gene (amplified from ATCC17978 genome using PCR) were obtained using splicing PCR technology. ΔabaR::Kn target fragments were amplified by transferring them into pUC19 vector and tested by PCR analysis. The primers used are listed in Supplementary Table 1. Next, ΔabaR::Kn sequences were linked to pCVD442 vector and transferred to Escherichia coli β2155 by electrotransformation. Conjugation between ATCC17978 strain and the E. coli was performed and the knockdown ATCC17978 strains were selected for kanamycin resistance [21, pp. 71–76].

Construction of Complementary Strains

To complement the knockout mutant, the target gene abaR was amplified from the cDNA of ATCC17978 strain, and the amplification product was cloned into the vector pBAD33-TCR to construct a pBAD33-TCR-abaR. The vector pBAD33-TCR has the origin of replication ori amplified on the pWH1266 plasmid and the tetracycline resistance gene; so, this plasmid can be amplified in A. baumannii. Then, pBAD33-TCR-abaR was transformed to E. coli competent cells DH5α. The plasmids that were positive after tetracycline screening were sequenced. Then, the plasmids with correct sequencing results were transferred to E. coli β2163 competent cells and the positive clone E. coli β2163-[pBAD33-TCR-abaR] was screened. Finally, E. coli β2163-[pBAD33-TCR-abaR] was transformed into knockdown ATCC17978 strains using electrotransformation method [22, pp. 162–168] and the recombinant clones were screened out, followed by PCR verification, and PCR products are sequenced for sequence confirmation (Supplementary Fig. 1). The primers used in this section were listed in Supplementary Table 2.

Growth Curve Analysis

ATCC 17978 and abaR knockdown strains at 1 × 108 CFU/mL were maintained in fresh LB media at 37 °C for consecutive 18 h. The optical density (OD) of cultured strains in each group was evaluated for each hour using a Microplate reader (Bio-Rad, Hercules, CA, USA). There were three repeats in each group.

Biofilm Formation Analysis

Next, 200 μL of strain cultures (1 × 108 CFU/mL) was seeded onto each well of a 96-well plate. After culturing for 8, 24, and 48 h, the cultures were washed thrice with 200 μL PBS. Next, the cultures were fixed in methanol for 15 min followed by staining with 0.2% crystal violet solution for 20 min. After washing, 200 μL of 33% glacial acetic acid was added to each well and then, OD values were measured using a Microplate reader.

RNA Sequencing

After culturing for 8 h, ATCC17978 and abaR knockdown strains were collected for further analysis. Total RNA of the cultures in the two groups were isolated using mirVana miRNA isolation kit (Ambion) following the manufacturer’s protocols, separately [23, p. e72968]. Because the vast majority of RNA in bacteria is rRNA, rRNA was obtained using Ribo-Zero Magnetic Kit (MRZB12424, Illumina, CA, USA) and removed, following which the mRNAs were broken into fragments as templates for cDNA synthesis [24]. An RNA library was constructed using PCR analysis. The quality and size of the DNA fragments were evaluated using Agilent 2100 Bioanalyzer. The whole transcriptome was then sequenced based on the Illumina platform.

Data Processing

The raw data were downloaded and quality control was performed, using Trimmomatic software, including removing adaptor and low-quality reads. Clean reads were then mapped to reference genome using Rockhooper2 software. The gene expression level was measured based on the number of reads mapped to the genome and exome according to the reads per kilobase per million mapped reads (RPKM) method. The formula is provided in Eq. (1) as follows:

$${\text{RPKM}}=\frac{{\text{Total}} \,{\text{exon}} \,{\text{reads}}}{{\text{Mapped}} \,{\text{reads}} \left({\text{millions}}\right)\times {\text{exon}} \,{\text{length}} \left({\text{KB}}\right)}.$$
(1)

Differentially Expressed Gene Analysis

The gene counts were normalized using DESeq software to produce the significant genes with differential expression between the abaR knockdown ATCC 17978 and wild-type ATCC 17978 strains. The fold change (FC) of gene expression was calculated and negative binomial distribution test was applied to evaluate the significance of the differences in gene reads between two samples. The log|FC|> 1.5 and P value < 0.05 were set as the cutoff values.

Function Enrichment Analysis

The Gene Ontology (GO) project provides the possibility of biological function annotation of genes of interest [25, pp. D322–D326]. The differentially expressed genes were subjected to GO function analysis and the significant P values were tested using the hypergeometric distribution test. GO terms with P values < 0.05 were considered significant.

Pathway Enrichment Analysis

The Kyoto Encyclopedia of Genes and Genomes (KEGG) (http://www.genome.ad.jp/kegg/) is a pathway-related database [26, pp. 91–100]. KEGG pathway enrichment analysis was performed to explore the pathways involved in differentially expressed genes using the DAVID online tool. The enrichment P values were calculated using the hypergeometric distribution test. A P value < 0.05 was considered the cutoff value.

Real-Time Reverse Transcription-PCR Verification

RT-PCR was carried out to determine the relative expression of some DEGs. Total RNA was isolated from wild-type strains, mutant strains, and complementary strains after 8 h of culture. The concentration of RNA was detected using NanoDrop 2000. RNA was reverse-transcribed to cDNA using TransScript All-in-One First-Strand cDNA Synthesis SuperMIX for qPCR kit. PerfectStartTM Green qPCR SuperMix kit and LightCycler® 480 system were used together to perform RT-PCR. All assays were performed in triplicate. The relative expression levels of RNAs were calculated using the 2−ΔΔCt method. 16 s was set as internal control. The primer sequences are listed in Supplementary Table 3.

Statistical Analysis

The data for strain growth and biofilm formation are expressed as mean ± standard deviation (SD). Multiple group comparison was performed using one-way ANOVA analysis and the difference between two groups was analyzed using the LSD test. P < 0.05 was considered significant.

Deposit the RNA-Seq Data

The raw sequencing data were deposited in the National Center for Biotechnology Information (NCBI) Sequence ReadArchive (SRA) database under accession number PRJNA720897.

Results

Effective abaR Gene Knockouts in ATCC 17978 Strain

PCR analysis of a recombinant DNA fragment (ΔabaR::Kn) showed that the size of the target sequence was 3029 bp (Fig. 1A). Next, the knockdown ATCC17978 strain was selected for kanamycin resistance. The PCR amplification of the target fragment was about 3300 bp, which indicated that the abaR gene sequence was successfully replaced by ΔabaR::Kn target fragment (Fig. 1B).

Fig. 1
figure 1

Size of the target fragments (ΔabaR::Kn) by PCR analysis (a) and the abaR gene knockdown determined by PCR analysis (b). a: M, marker; 1, target (ΔabaR::Kn) fragment (size: 3029 bp). b: M, DL10000 marker; M2, DL2000 marker; 1, tested fragment after abaR gene knockdown (size: 3300 bp)

Full size image

Growth Curve of ATCC 17978 Strain and abaR Knockdown Strain

The strain growth curve in Fig. 2 showed that the OD values for the ATCC 17978 strain were slightly higher than the abaR knockdown strain at 2, 3, and 4 h of culture. No significant differences were observed in the OD values of the two strain groups from 5 to 12 h of culture. All strains entered the plateau stage after 14 h of culture. ATCC17978 entered the exponential growth stage during 3–4 h of culture in vitro and entered the plateau stage at 14 h, which was similar to the growth curve of the abaR knockdown strain.

Fig. 2
figure 2

Growth curve for ATCC 17978 strain after abaR gene knockdown, compared with wide-type strain. The OD value of strains after abaR gene knockdown was observed at 0–18 h of culture. There were three repeats in each group

Full size image

Biofilm Formation Analysis

As shown in Fig. 3, biofilm formation by the ATCC 17978 strain increased in a time-dependent manner and peaked at 48 h of culture. However, there were no significant changes in the biofilm-forming ability of abaR knockdown strains during 0–8 h of culture. The biofilm-forming ability in the wild type and abaR knockdown strains was similar at 0–4 h of culture. At 8, 24, and 48 h after strain culture, the OD value of ATCC 17978 was 0.63 ± 0.06, 0.522 ± 0.18, and 1.04 ± 0.32, respectively. After abaR gene knockdown, the OD value was 0.41 ± 0.20, 0.45 ± 0.11, and 0.57 ± 0.02 at 8, 24, and 48 h, respectively. There were significant differences in OD values at 8 and 48 h of culture between the groups (all P values < 0.05), indicating that biofilm formation significantly declined in the ATCC 17978 strain after abaR knockdown.

Fig. 3
figure 3

Biofilm formation of ATCC 17978 strain after abaR gene knockdown, compared with wide-type strain. *P < 0.05, compared with wide-type ATCC 17978 strain. There were three repeats in each group

Full size image

Identification of Differentially Expressed Genes

With the cutoff value of log|FC|> 1.5 and P value < 0.05, a total of 137 differentially expressed genes (DEGs) were obtained between the wild type and abaR knockdown strains, including 20 downregulated genes and 117 upregulated genes (Supplementary Table 4). The volcano plot of genes with differential expression is shown in Fig. 4.

Fig. 4
figure 4

Volcano plot for differentially expressed genes in abaR knockdown strain, compared with wide-type strain. Red spots represent genes that were significantly upregulated. Green spots represent genes that were significantly downregulated. Gray spots represent genes that were not significantly different between the two groups (Color figure online)

Full size image

Significant Biological Processes and Pathways for Differentially Expressed Genes

GO analysis was performed for all the genes with differential expression. A total of 79 GO terms were enriched by differentially expressed genes, among which 58 GO terms were significantly enriched with P < 0.05. The differentially expressed genes were closely related to acetoin catabolic process (GO:0045150, BP), glycolytic process (GO:0006096, BP), tartrate dehydrogenase activity (GO:0009027, MF), and viral capsid (GO:0019028, CC). A total of 22 pathways were significantly enriched and the top 15 significant pathways are listed in Table 1. The results showed that pathways such as glycolysis/gluconeogenesis (ko00010), pyruvate metabolism (ko00620), and glucagon signaling pathway (ko04922) were dysregulated in abaR knockdown strains compared with the wild-type ATCC 17978.

Table 1 The top 15 significant GO term and pathways of differentially expressed genes
Full size table

RT-PCR Detection

The relative expression levels of four significantly up-regulated DEGs and four significantly down-regulated DEGs were detected by PT-PCR between wild-type ATCC17978 strains and abaR knockdown strains (Fig. 5). As we predicted, compared in the wild-type ATCC17978 strains, the relative expression of A1S_3544, A1S_0849, A1S_3867, and A1S_3011 showed a significantly higher levels in the abaR knockdown strains (P < 0.001). The expression levels of A1S_3661, A1S_1709, A1S_2237, and A1S_2457 were significantly higher in the wild-type ATCC17978 strains than that in the abaR knockdown strains (P < 0.001).

Fig. 5
figure 5

Relative expression of genes in wide-type ATCC 17978 strains and abaR gene knockdown strains. ***P < 0.001, compared with wide-type ATCC 17978 strain. There were three repeats in each group. Error bar represents SD

Full size image

We also detected the relative expression of A1S_0849 and A1S_2237 between wild-type ATCC17978 strains, abaR knockdown strains, and complementary strains (Fig. 6). The results revealed that there were no significant difference between ATCC17978 strains and complementary strains in the A1S_0849 and A1S_2237 expression. This indicates that our complementation experiment successfully restored gene expression to the wild-type level.

Fig. 6
figure 6

Relative expression of genes in wide-type ATCC 17978 strains, abaR gene knockdown strains, and complemented strains. ***P < 0.001, compared with wide-type ATCC 17978 strain. There were three repeats in each group. Error bar represents SD

Full size image

Discussion

A. baumannii has emerged as an important nosocomial pathogen mainly affecting patients with impaired host defences [27, pp. 640–645]. Now, A. baumannii is one of the most prevalent pathogens in burn-induced infections because of its multidrug-resistance, which can even induce bloodstream infections. A. baumannii is responsible for a vast array of infections, of which ventilator-associated pneumonia and bloodstream infections are the most common, and mortality rates can reach 35% [28, pp. 292–301]. Multidrug resistance may complicate the treatment of serious infections. Biofilm-forming ability has been found to be the major factor affecting drug resistance by A. baumannii [29, pp. 626–633]. Many gram-negative pathogens control their virulence through a quorum sensing (QS) system [30, p. 163]. AbaR is one of the QS-related genes, which is reported to play a regulatory role in biofilm formation [31, pp. 15–27]. In this study, we aimed to explore the mechanism underlying the role of abaR in A. baumannii biofilm formation.

We first determined the regulatory role of abaR in A. baumannii biofilm formation by gene knockout. In A. baumannii, abaR- and abaI-mediated QS plays a regulatory role in biofilm formation and surface motility [32, pp. 1719–1728]. The abaR gene at a size of 717 bp is localized at 1255 bp of the downstream of abaI gene. Additionally, we constructed an abaR gene knockout model in neurons using homologous recombination method. After abaR gene knockdown, the alternative target fragment in ATCC 17978 was determined to be 3300 bp using PCR analysis, which was similar to the size of ΔabaR::Kn target fragment, suggesting that the abaR gene was successfully knocked out.

Data of the growth curve analysis showed that after 4 h of culture, the growth of the wild-type ATCC 17978 entered the exponential growth phase and entered the plateau phase at 14 h after culture. The abaR knockdown strain showed a similar growth curve, which indicates that no significant effect was observed in the growth of ATCC 17978 strain following abaR knockdown.

A previous study suggested that the absorbance value of drug-resistant A. baumannii strain was similar to sensitive strains at all culture time points. Compared with the controls, the biofilm-forming ability of the ATCC 17978 strain exposed to N-Heptanoyl-l-Homoserine lactone significantly declined at 12 h of culture and that of strains exposed to N-(3-Hydroxydodecanoyl)-dl-Homoserine lactone noticeably increased [18, pp. 200–205]. Therefore, inhibition or activation of the abaR gene can influence biofilm formation. Our data showed that biofilm-forming ability significantly increased in the wild-type ATCC 17978 strain compared with the abaR knockdown ATCC 17978 strain at 8 h of culture, which was consistent with the previous findings. All these results suggest that abaR knockdown affected the biofilm-forming ability of A. baumannii without influencing strain growth.

In addition, the mechanism of the declined biofilm-forming ability of A. baumannii induced by abaR knockdown has not been clarified completely. Thus, we performed transcriptome sequencing to identify the changed gene expression profile caused by abaR knockdown. Based on the results, a total of 137 genes were differentially expressed between the wild type and abaR knockdown strains. GO analysis showed that viral procapsid maturation was closely related to the dysregulated expressed genes induced by abaR knockdown. It is reported that viral procapsid maturation is a dynamic process for viral infection [33, pp. 423–439]. The virus capsid assembly and DNA packaging have been proposed to be the target for antiviral therapy [34]. Thus, biofilm formation may share a similar group of involved genes with viral procapsid maturation and biofilm formation may be suggested as the target for antimicrobial treatment.

A previous study suggested that the changes in gene expression were more obvious in the early stage of biofilm formation than the later stages. In our study, biofilm formation was significantly changed at 8 and 24 h of culture in abaR knockdown strains compared with the wild-type strains. Thus, we performed transcriptome sequencing for the strains cultured for 8 h. Previous evidence showed that carbohydrate and amino acid metabolism were the necessary pathways in biofilm formation [35, 36, p. 13160]. Our study suggested that the DEGs in abaR knockdown strains, compared with the wild-type ATCC 17978, were significantly enriched in carbon metabolism and valine, leucine, and isoleucine degradation, which is consistent with previous reports and indicates that our findings are significant. Additionally, glycolysis and gluconeogenesis were the most significant pathways involved in the differentially expressed genes. A recent study reported that genes that play a regulatory role in gluconeogenesis are significantly upregulated in biofilm cells compared with free cells [37, p. 139]. It is reported that carbohydrates are the major composition of biofilms [38]. Carbon sources play a necessary role in the survival, growth, and infection of strains [39, pp. 2034–2045]. Accumulating evidence shows that Candida albicans can utilize glycolysis and gluconeogenesis alternatively for energy production [40]. Glycolysis plays a necessary role in the morphogenesis and virulence of Candida species. The evidence for the significant role of glycolysis and gluconeogenesis pathways in A. Baumannii virulence has been rarely reported. Thus, we suspect that the dysregulation of the glycolysis and gluconeogenesis pathways was induced by abaR knockdown in A. Baumannii.

Conclusion

Although this study lacks certain mechanism experiments, our verification of DEGs through RT-PCR showed that our bioinformatics analysis has a certain degree of credibility. Of course, further research on the protein level of these DEGs is also necessary in the future. In summary, abaR knockdown significantly decreased the biofilm-forming ability of the A. baumannii ATCC 17978 strain without affecting its growth. Carbon metabolism-related pathways, especially the glycolysis and gluconeogenesis pathways, were significantly dysregulated by abaR knockdown in A. Baumannii.