Coagulase-negative Staphylococci (CoNS) are low-virulence commensal bacteria in human. However, Staphylococcus lugdunensis, as a distinctive CoNS, virulence is similar to that of Staphylococcus aureus and stronger than other CoNS species [1]. As an opportunistic pathogen, S. lugdunensis colonizes skin surfaces with high humidity such as the groin, perineum, breast, and armpit in 30%-50% of patients [2]. The clinical infection characteristics are likely to that of S.aureus, including infective endocarditis, bacteremia, meningitis, and bone and joint infections [3, 4]. However, the incidence and outcome of S. lugdunensis are largely unknown, with only a few small case studies [5]. S. lugdunensis was first reported by Freney et al. [6] in 1988, which did not attract clinical attention at the beginning. However, with the increasing number of reported cases of S. lugdunensis infection, people gradually realized that S. lugdunensis was an important cause of serious infection [7]. In the systematic review by Non et al. [8], S. lugdunensis is associated with a similar proportion of infective endocarditis as S. aureus. A study by Shah et al. [9] showed that bone and joint infections due to S. lugdunensis tend to be more invasive than other CoNS species and are more similar to those due to S. aureus. The high pathogenicity of S. lugdunensis is closely related to some virulence factors, such as the fibrinogen binding protein (FBL), VWF-binding protein (vWbl) and autolysis (AtlL). Some studies suggest that the virulence factors of S. lugdunensis may be related to infectious diseases [10, 11]. Therefore, we attempted to describe the relationship between the virulence genes of S. lugdunensis and different infectious diseases, so as to explore the relationship between virulence genes and pathogenic phenotypes, more to describe the epidemiological characteristics of S. lugdunensis in this hospital using an appropriate molecular biological method.

In addition, the early pathogens typing is decisive to initiate appropriate infection control measures for different pathogenic phenotypes [12]. We searched for suitable typing methods to distinguish different pathogenic phenotypes of S. lugdunensis, to take effective infection control. At present, the most widely used methods for typing S. lugdunensis are pulsed-field gel electrophoresis (PFGE) and multilocus sequence typing (MLST) [11, 13, 14]. However, these molecular typing methods have the disadvantages of being time-consuming and high cost. Matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) is an emerging technology that has been increasingly used in clinical laboratories for rapid bacterial identification. Moreover, it has also shown the potential for typing applications among Staphylococcus species [15]. Therefore, we attempted to create a MALDI-TOF MS dendrogram to observe whether strains from different infection sites presented cluster distribution. Besides, fbl-typing is a special typing method for S. lugdunensis. The protein FBL of S. lugdunensis is encoded by fbl gene and belongs to the Sdr family, which has about 60% homology with clumping factor A (ClfA) of S. aureus [16, 17]. The fbl gene has three regions, namely N1, N2, N3, and a repeat region (R-region). More and more studies have confirmed that the fbl gene may be specific in S. lugdunensis, and its sequence is relatively conservative. Therefore, the fbl gene has been considered as a specific target of S. lugdunensis [18, 19]. Meanwhile, Dahyot et al. [14] proposed that the R-region be used as a genotyping method, which is similar to the S. aureus spa molecular typing method. The fbl-typing was based on the presence of gene polymorphism in a variable number of the 18 bp repeat sequences in the R-region of the fbl gene. Moreover, they confirmed that the fbl-typing was highly consistent with the results of MLST typing and had good identification ability [14].

Therefore, this study first described the relationship between the virulence genes of S. lugdunensis and different infectious diseases, so as to explore the relationship between virulence genes and pathogenic phenotypes. Subsequently, the MALDI-TOF MS dendrogram and fbl-typing were performed using BioNumerics software. We tried to compare the correlation between different methods and the different infectious diseases, antimicrobial resistance of the strains, in order to obtain the epidemic characteristics and antimicrobial resistance information of S. lugdunensis.

S. lugdunensis is a skin-colonizing bacterium, once the body is damaged, it is more likely to cause skin and soft tissue infection. However, in this hospital, the bone and joint infection type was the common type (25.4%, 19/65), which may be caused by the increase in artificial joint replacement. It is worth noting that reproductive system infection also accounted for a large proportion (21.5%, 14/65), most of which were isolated from semen, cervical or vaginal secretions, suggesting that reproductive system infection was also an important type of infection caused by S. lugdunensis in this hospital. The results of virulence genes showed that the seven virulence genes mentioned above existed in most strains, and there was no significant correlation between virulence gene distribution and infection type. The atlL gene encodes the major autolysin of S. lugdunensis, which plays a key role in cell division, biofilm formation, adhesion, and invasion [24, 25, 26]. Some studies have proposed that vWf-binding protein (vWbl) encoded by the vwbl gene can directly bind to von Willebrand factor (vWf), which is the key to resisting shear force and cause endocarditis, but other studies have pointed out that vWbl may not be the ligand of vWf [21, 27]. The vwbl gene is widely present in various infection types, and whether it plays a key role in causing infective endocarditis still needs to be studied which are similar to Giormezis et al. [28]. The genes encoding beta-hemolysin and hemolysin III of S. lugdunensis were widely present in isolates. In addition, three open reading frames, slush-A, slush-B and slush-C, exist outside the AGR genetic locus, encoding S. lugdunensis synergistic haemolysin (SLUSH). Rautenberg et al. [29] proposed that SLUSH was a phenol-soluble modulins like peptide (PSM-like peptide), which acted similarly to S. aureus PSM and can bind to formylpeptide receptor 2 (FPR2) of human leukocytes, which may be associated with its ability to cause invasive infections. The detection rate of slush gene was significantly different. Giormezis et al. [28] proposed that the vwbl and slush were more common in isolates of deep infection, but Szabados et al. [20] reported that the detection rate of the slush gene was low in all infection types, and there was no difference in distribution among all infection types. In this study, the detection rate of slush was high, showing no difference in distribution among different infection types. Therefore, the distribution of each virulence gene and its pathogenesis still need to be further studied. Due to the lack of research, it is not clear whether the distribution of virulence genes varies across regions and infection types.

Although the antibiotic resistance of CoNS is serious at present, S. lugdunensis in this study is sensitive to most antibiotics in the clinic. There were significant differences in penicillin resistance rates among different reports, and penicillin resistance was serious in this study [30]. The oxacillin-resistant rate of S. lugdunensis was 11.1%, and the MDR was 25.4%, which should be paid more attention to clinically. Comparing the antibiotic resistance rates of S. lugdunensis isolates from different infection sites, it was found that the antibiotic resistance rates of bone and joint infections and ear infections isolates were generally higher than those of other infection isolates. Three isolates of four ear infections S. lugdunensis (antibiotic resistance spectrum: OXC-ERY-CLI---------PEN----CFX--) were oxacillin-resistant isolates, which was higher than the oxacillin-resistance rate for non-ear infections (7.1%, 4/56), and the difference was statistically significant (p 0.05). These results suggest that the majority of ear infections caused by S. lugdunensis are multiple resistance isolates, and the antibiotic should be guided by the results of the antibiotic sensitivity test before the empirical use of antibiotics. After the isolates were divided into the oxacillin-resistant group and oxacillin-sensitive group, the antibiotic resistance rates of ERY, CLI and PEN in the oxacillin-resistant group were higher than those in the oxacillin-sensitive group, the difference was not statistically significant (p $>$ 0.05). Therefore, although all isolates in this study were sensitive to most antibiotics, the antibiotic resistant isolates of oxacillin still need to be concerned.

In order to explore an appropriate typing method to distinguish different pathogenic phenotypes of S. lugdunensis, we created a MALDI-TOF MS dendrogram and performed fbl-typing by BioNumerics software. Compared with the two methods, we found that the isolates from different infection sites were difficulty having the tendency of cluster distribution on the MALDI-TOF MS dendrogram, and the recognition and reproducibility were poor. With some reports, MALDI -TOF MS excepts for identification has also been explored for typing of diverse microorganisms (MALDI-typing) [31, 32]. But due to the high overall similarity of MS fingerprints within species and the difficulty of reproducibly detecting sufficient numbers of biomarkers below species-level specificity. It was proposed by researchers that MALDI-TOF MS should be established and performed as a robust typing tool under standard culture conditions, i.e., using a single medium type, incubation time and sample preparation procedure [15, 33]. In this study, 65 strains of S. lugdunensis were classified by fbl-typing, among which 30 strains were of novel type and New type3 was the dominant type. The comparison between the two dendrograms showed that the clustering trend of isolates from different infection sites in the two dendrograms had low similarity. The fbl-typing of 240 S. lugdunensis by Dahyot et al. [14], the dominant type was fbl47b. The new type 3 collected in our study was the dominant type, followed by fbl47b, indicating that the epidemic types of S. lugdunensis were different in different regions, and it was necessary to continuously monitor it. Phylogenetic tree construction showed that isolates from different infection types had a cluster distribution trend, suggesting that fbl-typing could indicate the pathogenicity of S. lugdunensis in this hospital.

In conclusion, the virulence genes of S. lugdunensis in our study were not significantly correlated with the infection sites. And the isolates were sensitive to most antibiotics, but attention should be paid to the oxacillin-resistance isolates. Compared with the MALDI-TOF MS dendrogram, fbl-typing could better correspond to the pathogenic phenotype. fbl-typing is a method of molecular typing based on the polymorphism of fbl gene repeats, which has the advantages of convenient operation, low cost, and reproducide, it is worth promoting and supplementing. Based on fbl-typing, the antibiotics resistance and infection sites of 65 S. lugdunensis were analyzed, it was found that the antibiotics resistance spectrum and infection sites of isolates had a certain distribution pattern in the phylogenetic tree. Our study provided basic data for the epidemic characteristics of S. lugdunensis in this hospital and supplementary data for fbl-typing. We concluded that fbl-typing is a excellent method for rapid genotyping of S. lugdunensis and can be used as an important tool in molecular epidemiological investigation, have been developed to analyze the genetic characteristics and phylogeny of S. lugdunensis from different sources.

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

GC—Conceptualization, Methodology, Data curation. YH—Conceptualization, Methodology, Writing- Original draft preparation. LY—Methodology, Data curation. QZ—Software, Validation. JL—Conceptualization, Methodology, Writing - Reviewing and Editing. ZW—Conceptualization, Methodology, Writing - Reviewing and Editing. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

Not applicable.

Not applicable.

This research received no external funding.

The authors declare no conflict of interest.