Wzc and WcaJ mutations in hypervirulent Klebsiella pneumoniae lead to phage resistance at the cost of reduced virulence

Hypervirulent Klebsiella pneumoniae (hvKp) is one of the major community-acquired pathogens, which can cause invasive infections such as liver abscess. In recent years, bacteriophages have been used in the treatment of Klebsiella pneumoniae, but the characteristics of the phage resistant bacteria produced in the process of phage therapy need to be evaluated. In this study, two podoviridae phages, hvKpP1 and hvKpP2, were isolated and characterized. In vitro and in vivo experiments demonstrated that the virulence of the resistant bacteria was significantly reduced compared with that of the wild type. Comparative genomic analysis of monoclonal sequencing showed that nucleotide deletion mutations of wzc and wcaJ genes led to phage resistance, and the electron microscopy and mucoviscosity results showed that mutations led to the loss of the capsule, meanwhile, animal assay indicated that loss of capsule reduced the virulence of hvKp. The findings can contribute to a better understanding of that bacteriophage therapy can not only kill bacteria directly, but also reduce the virulence of bacteria by phage screening. Importance Bacteriophages are considered potential therapeutic alternative to antibiotics; however host-evolved phage resistance has accounted for the resurgences of pathogens, meaning further measures are need to improve the efficacy of phage therapy. This study showed two phages capable of infecting hypervirulent K. pneumoniae and identified phage-resistant mutants whose virulence was significantly reduced. Gene sequencing analysis revealed that mutations of wzc and wcaJ gene, related to capsule synthesis, recovered phage resistance but reduced the virulence of hypervirulent K. pneumoniae.


19
Hypervirulent Klebsiella pneumoniae (hvKp) is one of the major community-acquired 20 pathogens, which can cause invasive infections such as liver abscess. In recent years, 21 bacteriophages have been used in the treatment of Klebsiella pneumoniae, but the 22 characteristics of the phage resistant bacteria produced in the process of phage therapy 23 need to be evaluated. In this study, two podoviridae phages, hvKpP1 and hvKpP2, were 24 isolated and characterized. In vitro and in vivo experiments demonstrated that the 25 virulence of the resistant bacteria was significantly reduced compared with that of the 26 wild type. Comparative genomic analysis of monoclonal sequencing showed that 27 nucleotide deletion mutations of wzc and wcaJ genes led to phage resistance, and the 28 electron microscopy and mucoviscosity results showed that mutations led to the loss of 29 the capsule, meanwhile, animal assay indicated that loss of capsule reduced the 30 virulence of hvKp. The findings can contribute to a better understanding of that 31 bacteriophage therapy can not only kill bacteria directly, but also reduce the virulence 32 of bacteria by phage screening. 33

34
Bacteriophages are considered potential therapeutic alternative to antibiotics; however 35 host-evolved phage resistance has accounted for the resurgences of pathogens, 36 meaning further measures are need to improve the efficacy of phage therapy. This 37 study showed two phages capable of infecting hypervirulent K. pneumoniae and 38 identified phage-resistant mutants whose virulence was significantly reduced. Gene 39 sequencing analysis revealed that mutations of wzc and wcaJ gene, related to capsule 40 synthesis, recovered phage resistance but reduced the virulence of hypervirulent K. 41 pneumoniae. 42

Introduction 45
Klebsiella pneumoniae, a Gram-negative enteric bacterium, is one of the most 46 important opportunistic nosocomial pathogens. Compared with the classic K. 47 pneumoniae (cKp) strains that cause other types of opportunistic infections, K. 48 pneumoniae strains that mainly cause pyogenic liver abscesses often have a 49 substantially higher virulence and are therefore designated hypervirulent K. 50 pneumoniae (hvKp). Hypervirulent K. pneumoniae has the ability to cause 51 life-threatening, community-acquired infections, including liver abscesses 52 complicated by endophthalmitis, meningitis, osteomyelitis, and necrotizing fasciitis, 53 in young and healthy individuals and is therefore associated with high morbidity and 54 mortality. In recent years, multidrug-resistant hypervirulent strains have emerged [ software was used to analyze the contigs obtained by splicing the second and 148 third-generation sequencing data to reconfirm the assembly results and determe the 149 positional relationship between the contigs, and to determine the gap between contigs. 150 To obtain information, the complete genomes of phages were sequenced and 151 analyzed using a variety of bioinformatics tools. Open reading frames (ORFs) 152 prediction were predicted using SoftBerry (http://www.softberry.com). Genome 153 annotations were checked through sequence comparison of protein sequences using the 154 blastn software (https://blast.ncbi.nlm.nih.gov). Genome comparative analysis was 155 performed using Easyfig. 156

Bacteriophage therapy assay and determination of virulence of strains 157
The G. mellonella model was used to evaluate the antibacterial efficacy of 158 bacteriophages in vivo and the virulence of K. pneumoniae strains [24, 25]. The 159 minimum lethal concentration of K. pneumoniae infection by larval caterpillars was 160 determined to be 10 7 CFU/mL within three days. When larvae did not respond to touch, 161 they were considered dead. In the in vivo experiment, the larvae were divided into four 162 groups, sixteen randomly chosen larvae were used for each group: (i) only injected with 163 10ul PBS, (ii) injected with 10ul of 10 6 CFU/mL host bacteria, (iii) only injected with 164 10ul of 10 7 PFU/mL of phage, (iv) injected with 10ul of 10 6 CFU/ml host bacteria, and 165 then injected with 10ul of 10 7 PFU/mL phage within 30 min. All larvae were incubated 166 at 37 °C and the number of dead larvae was counted at 12 h intervals up to 72 h after the 167 incubation. 168 Virulence determination tests were performed according to the bacterial 169 concentration in the previous phage treatment experiment (10 7 CFU/mL). Sixteen G. 170 mellonella larvae were injected with 10ul of the inoculum in every group. Survival was 171 analyzed by Kaplan-Meier analysis with a log-rank test; differences were considered 172 statistically significant at P < 0.05. 173

Screen for phage-resistance strains 174
Phages hvKpP1 and hvKpP2 were mixed with the host bacteria for cultivation, and the 175 mixture was cultured using double-layer soft agar. Plates were incubated overnight at 176 37 °C, and the resulting colonies were picked up and saved for further assays. 177

Bacteria growth curves 178
All strains were cultured as described above. The following day, cultures were 179 incubated in LB at a concentration of 1 × 10 7 CFU/mL and added to individual wells of 180 a 96-well microtiter plate. Plates were incubated for 12 h at 37℃ and absorbance 181 readings at 600 nm were recorded every 30 min using BMG SPECTROstar® Nano. 182 Growth rates of the bacterial strains were calculated using three biological replicates. 183

Mucoviscosity assay 184
The mucoviscosity of the capsule was assessed by low-speed centrifugation of the 185 liquid culture [26]. Various overnight cultures of K. pneumonia were grown to adjust to 186 OD 600 /mL of 1 and centrifuged at 1,000 ×g for 5 min.

Phage isolation and host range 239
Two lytic phages, vB_KpnP_cmc20191 (referred to as hvKpP1) and 240 vB_KpnP_cmc20192 (hvKpP2), were isolated from sewage; microscopic observation 241 of virion morphology by TEM showed that the phages were classified as members of 242 the Podoviridae family (Fig 1A, B). They formed different plaques on the bacterial 243 lawn of K. pneumoniae strain hvKpLS7 (Fig 1C, D)

Phage characterization 256
To determine phage stability, the sensitivity of phages to temperature and pH stability 257 was analyzed (Table 2). Phages were stable in the pH range of 4-11 or under 70℃ (Fig  258  S2AB). These results were in line with those of previous studies [32,33], showing that 259 these two phages could maintain high lytic activity under broad physicochemical 260 conditions. The adsorption rate curves of the two phages showed that more than 90% of 261 bacteriophages were adsorbed within 10 min (Fig S2C). According to the one-step 262 growth curves (Fig S2D), the replication cycle of hvKpP2 was approximately 60 min. 263 However, the latent period of hvKpP1 was relatively short, only approximately 30 min. 264 The burst size of hvKpP1 (149 PFU/Cell) was greater than that of hvKpP2 (96 265 PFU/Cell). The eclipse period and burst size may be part of the reason for the difference 266 in plaque production between the two phages. A summary comparison of these two 267 phages is presented in Table 2. 268 Bioinformatics analysis helps us to better understand and predict biological 269 characteristics of phages. Thus, the complete genomes of the two phages were 270 sequenced and analyzed using bioinformatics tools. bursted more than hvKpP2. The two phages had high homology in the tail packaging 282 region, indicating that their adsorption targets for the host bacteria were the the same. 283 The complete nucleotide sequences of phages vB_KpnP_cmc20191 and 284 vB_KpnP_cmc20192 were determined and deposited in GenBank under the accession 285 numbers MT559526 and MT559527, respectively. 286

In vivo efficiency of bacteriophage treatment 287
The efficacy of phages hvKpP1 and hvKpP2 was evaluated in vivo using the 288 G.mellonella larvae model. For larva infected by the host strain hvKpLS7, the 289 survival rate was only 5 and 10 % in three days, respectively. In the phage treatment 290 groups, survival was significantly superior, with three-day survival rates of 75 and 291 90%. There was a significant difference in the survival rates between the larva 292 infection group and the treatment group (P<0.05). Additionally, the larvae group 293 injected only with phages still had a high survival rate, demonstrating the safety of the 294 phages in this model (Fig 3). The present study reports, for the first time, on phage 295 efficacy against K57 capsular serotype K. pneumoniae in G. mellonella. 296

Virulence declined with phage-resistance K. pneumonia 297
Although phage treatment in the G. mellonella model showed excellent results, 298 phage-resistant K. pneumoniae colonies were easily produced on 299 bacterium-phage-co-cultured LB agar plates. Interestingly, phage-resistant K. 300 pneumonia did not affect therapy. To test the difference between wild type and 301 phage-resistant K. pneumoniae, six mutants were randomly selected from the plates. 302 We then evaluated whether phage-resistant bacteria changed their generation time, 303 and the results indicated that resistance to phage had no effect on growth (Fig 4B). On 304 the LB agar plates, the wild-type host bacteriaium hvKpLS7 was moist, hypermucoid 305 and was reflective when photographed, whereas the phage-resistant bacteria had a 306 translucent appearance and had a reduced ability to produce mucoid ( Fig 4A) grown in LB, diluted to an optical density at 600 nm (OD 600 ) of 1, and then subjected 312 to low-speed centrifugation. During the centrifugation process, hvKpLS7 did not 313 sediment well, the supernatant was still turbid, and the OD 600 of the supernatant was 314 0.35. Contrastingly, the phage-resistant bacteria was well precipitated in the 315 low-speed centrifugation experiment, the supernatant was relatively clear, and the 316 average OD 600 decreased to 1/3, compared to the wild type ( Fig 4CD). This indicated 317 that K. pneumoniae with phage resistance significantly decreased capsular adhesion. 318 Previous studies have demonstrated that K. pneumoniae capsules confer 319 significant phagocytosis resistance to macrophages [37]. These strains were 320 co-incubated with phagocytes RAW264.7, the lysates of washed phagocytes were 321 daubed on agar plates, and the bacterial colonies were counted and recorded. Results 322 showed that amount of the mutant strain (except hvKpP-R1) devoured by the 323 phagocytes increased at least 10 times compared to that of the wild-type, indicating 324 that the phagocytes could effectively eliminate these mutant bacteria. 325 Further, whether the resistance of hvkp to phage affects its virulence during 326 infection was determined in vivo. As expected, data of the infection with G. 327 mellonella showed that the phage-resistant K. pneumoniae had a higher survival rate 328 with the same concentration of bacteria (Fig 4F). The results of these assays indicate 329 that the phage-resistant K. pneumoniae reduced their virulence, which might be due to 330 the absence of a capsule. 331

Identification of mutant genes in phage resistant strains 332
To identify the genes responsible for resistance in bacteria, genomes of wild-type 333 hvKpLS7 and phage-resistant mutants were sequenced using the Illumina Hiseq 334 platform and comparatively analyzed. High-probability mutations (defined as  335 high-frequency, non-silent mutations within an open reading frame) were selected for 336 further validation. These mutant strains had, mutations in two genes associated with 337 the capsule, wzc and wcaJ, which corresponded to hvKpP-R2 and hvKpP-R3 338 respectively. Previous studies have shown that hvkp contains a compact gene cluster, 339 which plays an important role in capsule synthesis. Spot test results indicated that the strain hvKpP-R2 pwzc restored phage 352 sensitivity (Fig 5B). Low-speed centrifugation of the liquid culture showed increased 353 mucoviscosity of the capsule (Fig 5D). TEM analysis showed that the boundaries of 354 hvKpP-R2 were smoother than those of LS7 and hvKpP-R2pwzc, which confirmed 355 that phage resistance selection caused the loss of the hvkp capsule, and the expression 356 of recombinant wzc could significantly restore the polymerization of capsule (Fig 5C). 357 According to the three-day survival rate of G. mellonella, results of the 358 hvKpP-R2pwzc and wild-type infection groups were similar (Fig 5E), indicative that 359 after complementing the wzc gene of the wild-type strain, the virulence of this mutant 360 strain was rescued. 361 Therefore, after a mutation in wzc, the ability of K. pneumoniae to synthesize the 362 capsule was limited, and phages could not adsorb the mutant strain. Similarly, the 363 virulence of the mutant strain decreased due to the loss of the capsule. Previous 364 studies have shown that since wzc plays an indispensable role in capsular 365 polysaccharide surface assembly in K. pneumoniae, it is conceivable that mutations in 366 wzc are dominant in phage-resistant strains [39]. 367 Complementation of strain hvKpP-R3 with wild-Type wcaJ restore phage 368 sensitivity virulence 369 We amplified and cloned wcaJ from wild-type to phage-resistant mutant hvKpP-R3, 370 whose wcaJ, compared with the wild-type, had four bases deleted ( Fig 6A). As 371 described above, the same experiments were carried out on hvKpP-R3 pwcaJ, and 372 results are shown in Fig 6B. The phage-resistant hvKpP-R3 containing the wild-type 373 wcaJ gene can be re-infected by the two phages, and the adhesion of the bacteria was 374 restored to a certain extent. Electron microscopy images (Fig 6C) showed no 375 differences in cell size and shape between the wild-type and phage-resistant mutants. 376 The most evident difference was the loss of capsular in the mutant strain. 377 Replenishment of the wcaJ gene allowed the mutant strain to regenerate the capsule, 378 similar to the wbap gene [12]. In the in vivo infection experiment of G. mellonella, 379 although the lethality rate of the hvKpP-R3 pwcaJ was lower than that of the 380 wild-type in the first 12 h, it was significantly higher than that of the mutant strain 381 hvKpP-R3, and within three days was also comparable to that of the wild type (Fig  382  6E). 383 As showed that the phage therapeutic effect was not affected by these mutants, and 409 instead the virulence of these mutants was significantly reduced. This prompted a 410 further investigation on the mechanisms of phage resistance and virulence reduction. 411 Mutants were randomly selected on bacteria-phage co-cultured plates the colonies 412 produced by these mutants were smaller and rougher than those of the wild type 413