Initial screening of hvKP clinical isolates

Based on previous studies [20], we set clinical inclusion criteria for initial screening of hvKP, and used clinical data to preliminarily screen 94 hvKP strains (Table 1). Possible primary infections (number of cases) are liver abscess (4), renal abscess (4), lung abscess (3), abdominal wall abscess (1), testicular abscess (1), perianal abscess (1), acute cholecystitis (23), and necrotizing fasciitis (2).

Table 1 Clinical data of hvKPIdentification of hvKP clinical isolates

It has been found that peg-344, iroB, iucA, prmpA, prmpA2 and SP yields greater than 30 μg/ml have been shown to accurately distinguish hvKP from cKp strains [10]. Therefore, this study established hvKP identification criteria: initial screening of hvKP strains that are positive for any gene such as peg-344, iroB, iucA, prmpA, prmpA2 or SP yield greater than 30 μg/ml, were identified as hvKP, and vice versa, cKP.

String test of initial screening hvKP strains

In this study, 13 of the 108 primary hvKP strains were positive for string test, and the positive string results were shown in Fig. 1B.

Fig. 1

Results of each diagnostic index of hvKP (A), Southern blot analysis of virulence genes of some primary screened hvKP isolates (B), string test of initial screening hvKP strains (C), SP qualitative testing

Virulence gene detection of initial screening hvKP strains

In this study, PCR was used to detect five virulence genes (peg-344, iroB, iucA, prmpA, prmpA2) in 108 primary screening hvKP strains. Among them, 74 strains were positive for any of these five virulence genes, and the part of positive results were shown in Fig. 1A, of which 41 cases were detected with peg-344-positive strains, 18 cases with iroB-positive strains, 47 cases with iucA-positive strains, 10 cases with prmpA-positive strains and 47 cases with prmpA2-positive strains. The details of southern blots were shown in Figure S1.

SP detection of initial screening hvKP strains SP qualitative testing

The production of SP of primary screening hvKP strains was detected by CAS medium. 59 SP-positive strains were detected in 108 hvKP strains, with an orange halo around the colonies (Fig. 1C). However, the diameter and size of the rings vary, indicating that the strain can produce SP, but the amount of production varies.

SP quantitative detection

The production amount of SP was determined in 59 strains that tested positive in the qualitative SP testing using a quantitative SP test. Deferoxamine mesylate was employed as the standard, and a quantitative standard curve for SP was constructed, as depicted in Fig. 2. The calculated SP yield exceeded 30 μg/ml (Table S1).

Fig. 2

SP quantitative standard curve

Therefore, according to the hvKP identification criteria set by this study, a total of 94 hvKP strains were screened, and the detection rate of each diagnostic index of hvKP was analyzed, as shown in Table 2, the positive rates of prmpA, iroB, peg-344, prmpA2, iucA were 10.64%, 19.15%, 43.62%, 50%, and 50%, respectively. The positive rate of string test was 13.83%, and the positive rate of SP test was 62.77%. It can be seen that the positive rate of string test is low, and the missed rate of hvKP is high, which cannot be used for the identification of hvKP. The detection rate of peg-344, prmpA2, iucA and SP tests in virulence genes was high, all > 40%, which was a key indicator for the diagnosis of hvKP.

Table 2 Positive rate of hvKP diagnostic index (%)Experiment on virulence detection of Galleria mellonella Galleria mellonella health index score

In this study, a combination of PCR analysis of virulence genes and SP detection was employed.. If either of the two tests yielded a positive result, the high virulence characteristics of hvKP were verified by the virulence test of Galleria mellonella. Therefore, 94 hvKP strains were screened out from the above experiments for the virulence detection of Galleria mellonella, 94 hvKP strains were prepared into 1 × 105、1 × 106、1 × 107、1 × 108 CFU/ml bacterial solutions by double dilution method, respectively. The Galleria mellonella were inoculated with phosphate buffer saline (PBS) buffer and blank group as the parallel controls. There were 6 groups with 10 larvae in each group. The health of larvae was assessed at 6 h, 12 h, 24 h, 48 h, 72 h, 96 h, 120 h and 144 h after inoculation. The assessment involved recording the activity level, cocoon formation, blackening, and overall survival of the Galleria mellonella larvae. The scoring criteria for the specific larval health index were adopted from a previous study [21].

As shown in Fig. 3A, the Galleria mellonella used in this study were 2 ~ 3 cm long, and were divided into 5 situations according to the formation of larval melanin (Fig. 3B), with a score of 0 ~ 4. According to the larval cocoon formation, it was divided into no cocoon, partial cocoon and whole cocoon, with a score of 0 ~ 1 (Fig. 3C). According to the survival of the larvae, the score was 0 and 2, in which the death of the larvae was black and no response to touch (Fig. 3D).

Fig. 3

Performance of Galleria mellonella at different ratings (A), The mean length of Galleria mellonella was 2.5cm (B), melanization of Galleria mellonella (C), cocoon formation of Galleria mellonella (D), survival of Galleria mellonella

The Galleria mellonella were divided into six groups. The experimental group was the larvae injected with four different concentrations of hvKP bacterial solution (1 × 105–1 × 108 CFU/ml), the control group was the blank group without treatment and the PBS group was injected with PBS. The average of 10 larvae in each group was taken and compared according to the x ± s score of each group. As shown in Fig. 4, it was found that after injection of hvKP strains with different concentrations, the larvae of the Galleria mellonella showed a sharp decrease in health index score within 6 to 24 h, and the number of surviving larvae decreased significantly after 24 h. The larval scores of the high-concentration group were significantly lower compared to those of the low-concentration group. Additionally, as the concentration of the bacterial solution increased, the larval scores decreased within the same time period. This concentration-dependent trend in the larval scores indicated that the 94 strains of hvKP exhibited high toxicity characteristics. On the other hand, the scores of the blank group and the PBS group did not show significant changes over the observation period, suggesting that these groups did not experience any adverse effects.

Fig. 4

Health index scores of hvKP strains in Galleria mellonella at different concentrations and different time

Time of death of Galleria mellonella

After inoculation of 94 Galleria mellonella larvae with different concentrations (1 × 105, 1 × 106, 1 × 107, 1 × 108 CFU/ml), the time when the larvae reached 80% death (LT80) during 6 to 144 h was recorded (Fig. 5). Among them, the survival time of larvae inoculated by the low concentration group (1 × 105 and 1 × 106 groups) was > 144 h, which was labeled as *144 h, while most of the larvae inoculated in the high concentration group (1 × 107, 1 × 108 groups) reached 80% death within 48 h. With the increase of the concentration of bacterial solution, the death time of 80% larvae was shorter and showed a concentration dependence. Compared with the low concentration group (1 × 105 group), the 1 × 106, 1 × 107 and 1 × 108 groups had significant statistical differences (P < 0.0001). However, 80% of the larvae in the blank group and PBS group did not die with the increase in observation time, so LT80 at different times was expressed as *144 h. Therefore, it was further verified that all 94 hvKP strains had high toxicity characteristics.

Fig. 5

Time to 80% Galleria mellonella death of hvKP strains at different concentrations (LT80)

Survival curve of Galleria mellonella

Figure 6 illustrates the survival curves of Galleria mellonella larvae that were injected with different concentrations of hvKP strains, as well as the PBS and blank control groups. The survival rate of the larvae inoculated with different concentrations of hvKP strains was significantly lower compared to the control group. This difference in survival rates was found to be statistically significant(P < 0.0001). It was further verified that all 94 hvKP strains had high virulence characteristics.

Fig. 6

Survival curves of Galleria mellonella in different concentrations of hvKP strains

Molecular characteristics of the CR-hvKP strain

Worldwide, there are increasing reports of antibiotic-resistant hvKP isolates, mainly in countries with endemic hvKP transmission [22]. Studies have found that the hvKp strain obtains a carbapenem-resistant plasmid and evolves into CR-hvKP [23], and the characteristics of high virulence and multidrug resistance have attracted more and more clinical attention. Therefore, while observing the clinical data of patients, according to the drug susceptibility report of the isolates, 36 strains of carbapenem-sensitive hypervirulent Klebsiella pneumoniae (CS-hvKP) and 58 strains of CR-hvKP were found in 94 hvKP strains, and their molecular characteristics were comparatively analyzed. The comparison of various diagnostic indicators is shown in Table 3. We found that compared with CS-hvKP, the detection rate of peg-344, prmpA and iroB virulence genes in CR-hvKP was lower with a statistically significant difference (P < 0.00), and the detection rate of prmpA2, iucA was higher with a statistical difference (P < 0.05), while the string test and SP experiment did not have a statistical difference (P > 0.05). The virulence genes prmpA2 and iucA may be involved in plasmid integration with carbapenems resistance genes, carried by hvKp or CRKP clonal strains, and thus evolved into CR-hvKP.

Table 3 Comparison of the positive rate of each diagnostic index in CS-hvKP group and CR-hvKP group (%)

The SP of 59 SP-positive hvKP strains were quantitatively detected. According to the drug resistance of carbapenems, the strains were divided into CS-hvKP and CR-hvKP, and the concentration characteristics of SP were analyzed. As shown in Fig. 7, the CR-hvKP strain was found to have a higher concentration of SP compared to CS-hvKP, and the difference was statistically significant (P < 0.01). The results show that CR-hvKP has a stronger iron-bearing capacity.

Fig. 7

Comparison of SP concentrations of CS-hvKP and CR-hvKP strains