key: cord-0979270-olbxaczd
authors: Qu, Jiuxin; Cai, Zhao; Duan, Xiangke; Zhang, Han; Han, Shuhong; Yu, Kaiwei; Jiang, Zhaofang; Zhang, Yingdan; Liu, Yang; Liu, Yingxia; Liu, Lei; Yang, Liang
title: Nosocomial Pseudomonas aeruginosa regulates alginate biosynthesis and Type VI secretion system during adaptive and convergent evolution for coinfection in critically ill COVID-19 patients
date: 2021-04-11
journal: bioRxiv
DOI: 10.1101/2021.04.09.439260
sha: a4d77628760c1d80296d0f8a067d0b4af509cd56
doc_id: 979270
cord_uid: olbxaczd

COVID-19 pandemic has caused millions of death globally and caused huge impact on the health of infected patients. Shift in the lung microbial ecology upon such viral infection often worsens the disease and increases host susceptibility to secondary infections. Recent studies have indicated that bacterial coinfection is an unignorable factor contributing to the aggravation of COVID-19 and posing great challenge to clinical treatments. However, there is still a lack of in-depth investigation on the coinfecting bacteria in COVID-19 patients for better treatment of bacterial coinfection. With the knowledge that Pseudomonas aeruginosa is one of the top coinfecting pathogens, we analyzed the adaptation and convergent evolution of nosocomial P. aeruginosa isolated from two critical COVID-19 patients in this study. We sequenced and compared the genomes and transcriptomes of P. aeruginosa isolates longitudinally and parallelly for its evolutionary traits. P. aeruginosa overexpressed alginate and attenuated Type VI secretion system (T6SS) during coinfection for excessive biofilm formation and suppressed virulence. Results of bacterial competition assay and macrophage cytotoxicity test indicated that P. aeruginosa reduced its virulence towards both prokaryotic competitors and eukaryotic host through inhibiting its T6SS during evolution. P. aeuginosa T6SS is thus one of the reasons for its advantage to cause coinfection in COVID-19 patients while the attenuation of T6SS could cause a shift in the microecological composition in the lung. Our study will contribute to the development of therapeutic measures and the discovery of novel drug target to eliminate P. aeruginosa coinfection in COVID-19 patient.

Four isolates of Pseudomonas aeruginosa were collected at two different time points The isolates were cultured in LB broth at 37°C for overnight. The overnight cultures were diluted 112 to OD 600nm 0.01 in fresh LB broth. 100µL of the diluted cultures were loaded into 96-well plate 113 in triplicates and incubated for 24h at 37°C statically allowing the formation of biofilm. After 114 removing spent media, biofilms were washed carefully with ddH 2 O for two times. Biofilms were 115 then stained by 125 µL of 0.1% crystal violet (CV) with 15 min incubation at room temperature. 116 CV stain in the wells was discarded while stained biofilms were washed twice thoroughly with 117 ddH 2 O and air-dried. Biofilms were then dissolved into 125 µL of 30% acetic acid and quantified 118 relatively by measuring OD 550nm values on a Tecan infinity pro200 microplate reader.

Genome extraction and sequencing 120 The isolates were cultured in LB broth at 37°C to early stationary phase. For Illumina 121 sequencing, genomic DNA of each isolates was extracted using AxyPerp Bacterial Genomic Transcriptome extraction and sequencing 134 The isolates were cultured in triplicates in LB broth at 37°C to early stationary phase. Magen 135 HiPure Universal RNA Mini kits (MCBio, China) was used to extract total RNA following the 136 manufacturer's protocol. Extracted RNA was quantified using Qubit 2.0 (Thermo Fisher 137 Scientific, MA, USA) and Nanodrop One (Thermo Fisher Scientific, MA, USA). Quality of 138 RNA samples was assessed by Agilent 2100 system (Agilent Technologies, Waldbron, 139 Germany). RNA libraries were constructed according to standard protocol using NEB Next® 

Hospital-acquired P. aeruginosa in the respiratory systems of COVID-19 patients 203 During routine screening of respiratory samples of COVID-19 patients, we discovered that two 204 patients were colonized by P. aeruginosa with the same ST type. We have identified and isolated 205 P. aeruginosa from 5 different critically ill patients in total. To investigate genetic adaptation and 206 epidemiological link between these P. aeruginosa isolates, genomes of all isolates were 207 sequences by Illumina HiSeq platform. Multi-Locus Sequence Typing (MLST) analysis indicated 208 that four isolates collected from 2 patients are of the same type, P. aeruginosa ST1074, showing 209 that these P. aeruginosa isolated from these two patients were hospital-acquired. We thus focused 210 on these isolates to analyze their competitive advantage during colonization in COVID-19 211 environment. These four P. aeruginosa isolates were collected from respiratory samples of the (Table S1 ). No obvious 236 change in resistance to these drugs was observed between LYSZa2 and LYSZa3 probably due to 237 the short evolving time (Table S1 ).

Genomic islands (GIs) on LYSZa2 and LYSZa5 genomes were predicted by IslandViewer4 239 (Table S2&S3 ). In total, 35 GIs on LYSZa2 and 36 GIs on LYSZa5 were predicted by at least 240 one prediction method. Genomes of LYSZa2 and LYSZa5 were compared with genomes of five 241 other P. aeruginosa strains including PAO1 reference strain and virulence strains including 242 PA14, LESB58, SCV20265 and VFRPA04 ( Figure 1A ). Most of the GIs predicted are specific to 243 LYSZa2 and LYSZa5 genomes ( Figure 1A ). Genes in these GIs are involved in transcriptional 244 regulation, DNA restriction-modification, DNA repair, toxin-antitoxin and secretion systems, 245 showing that these GIs are essential for P. aeruginosa survival and virulence during coinfection 246 with SARS-CoV-2. 247 We then traced the origin of these isolates by constructing phylogenetic tree using genomes of 248 LYSZa2 and LYSZa5 with 22 other clinical or environmental P. aeruginosa genomes selected 249 from NCBI/Pseudomonas genome database (Table S4 ) and one SARS-CoV-2 coinfecting strain 250 published by our group recently, P. aeruginosa LYSZa7 [33] . As seen from the phylogenetic tree 251 ( Figure 1B ), LYSZa2 and LYSZa5 are closed related without evolutionary distance between 252 them, and are in the same phylogenetic cluster with PAO1 reference strain and the hypervirulent 253 isolate LESB58 from CF patient. Interestingly, great evolutionary distance was observed 254 between these two isolates with P. aeruginosa LYSZa7, indicating the distinct evolutionary traits 255 evolved between different P. aeruginosa strains in COVID-19 patients for adaptation.

Single nucleotide polymorphism (SNP) and other genome modifying events were assessed 257 between the ancestry isolates and the progeny isolates respectively using PAO1 as reference 258 (Table S5) . 90 genomic modifying events including SNPs, insertion, deletion and replacement 259 (Table S5) were identified in LYSZa3 with a dN/dS ratio of 0.875, indicating a negative selection 260 during evolution, probably due to the short evolving time between these two isolates. 93 of such 261 genomic modifying events (Table S5) were identified in LYSZa6 comparing to LYSZa5 with a 262 dN/dS ratio of 1.114 indicating a positive selection. Common mutations were found between 263 LYSZa3 and LYSZa6 on genes related to Type VI secretion system and iron transport (Table 1) . 264 Such observation indicated that P. aeruginosa undergoes both adaptive evolution and convergent 265 evolution in COVID-19 patients to survive and modulate its virulence during coinfection. We 266 then performed RNA sequencing to assess the changes at transcriptional level during evolution. (Table S6 ). In LYSZa6 as comparing to LYSZa5, 242 genes were 276 differentially expressed, among which 114 were upregulated and others were downregulated 277 (Table S7) . These differentially expressed genes are illustrated by heatmaps ( Figure 2A&B are alginic acid biosynthetic process and protein secretion by the type VI secretion system 292 (T6SS). 12 genes assigned to alginic acid biosynthetic process (algD, algX, algA, algE, algF, 293 algL, alg44, algJ, algK, alg8, algI, algG) were all upregulated in LYSZa3 for 16.89 to 1091.52 294 folds ( Table 2 ). The expression of same genes also increased significantly in LYSZa6 for 4.45 to 295 138.79 folds (Table 3) . Eight genes assigned to protein secretion by T6SS, PA1657 (hsiB2), 296 PA1658 (hsiC2), PA1659 (hsiF2), PA1660(hsiG2), PA1661(hsiH2), PA1662(clpV2), 297 PA1663(sfa2), PA1666(lip2), were all significantly downregulated in LYSZa3 for 4.46 to 6.47 298 folds (Table 2 ). In LYSZa6, beside the same eight genes mentioned above, 7 other genes 299 assigned to protein secretion by T6SS were also downregulated, including PA1656 (hsiA2), 300 PA1665 (fha2), PA1667 (hsiJ2), PA1668 (dotU2), PA1669 (icmF2), stk1 and stp1, for 4.46 to 301 17.11 folds (Table 3) . Besides these genes assigned to T6SS by GO enrichment analysis, the 302 expression of several other genes involved in T6SS, hcpB, lip3 (PA2364), and dotU3 (PA2362), 303 were also decreased in LYSZa3 (Table 2 ). While the expression of 15 other genes involved in

PA0094(eagT6), PA3484(tse3), PA5266(vgrG6), hcpA and hcpB (Table 3) . P. aeruginosa carries 307 three types of T6SS, HSI-I, HSI-II and HSI-III respectively. As illustrated in Figure 4 , the 308 downregulated genes in both progeny isolates, especially those in LYSZa6, are mostly involved 309 in HSI-II gene cluster, with several others scattered on another two gene clusters. Higher no. of isolates produce excessive alginate and exhibit more mucoid phenotype which matches with the 317 DEG results ( Figure 5A ). As alginate plays an important role in biofilm architecture and 318 development[34], we thus quantified biofilm formation of the isolates by crystal violet staining.

As seen from Figure 5B , aligned with gene expression, an increase in the biofilm formation were 320 observed in LYSZa3 comparing to those of LYSZa2. Similarly, more biofilm formation was 321 observed in LYSZa6 as compared to LYSZa5. RNA-seq analysis showed the adaptive changes in 322 genes responsible for alginate biosynthesis and T6SS protein secretion in the progeny isolates.

In addition, 74 common genes in total were found to be differentially regulated in both progeny 324 isolates, LYSZa3 and LYSZa6 (Table S8) To test the reduction in the competitiveness of the progeny isolates with bacterial neighbors due to suppression of T6SS, we examined the capability of the isolates in killing E. coli cells through 341 bacterial competition assay. The isolates were mixed and cultured with E. coli/placZ in 1:1 ratio 342 respectively. The mixtures were diluted to 10 -3 while triplicate colonies of each dilution were 343 then cultured on plate containing X-gal to test the killing efficiency ( Figure 6A&B ). E. Weakened macrophage cytotoxicity of P. aeruginosa after T6SS suppression 356 We then performed macrophage killing assay to assess the decrease in cytotoxicity of the 357 progeny isolates to eukaryotic cells upon T6SS suppression. Cultures of the isolates were added 358 to infect macrophages individually. Relative lactate dehydrogenase (LDH) release was measured 359 to determine the death of macrophages. More macrophages are killed, higher LDH will be 360 released. As illustrated in Figure 7 , LDH released by macrophages infected by LYSZa3 was 361 much lower than that of LYSZa2. Similar results were observed from LYSZa5 and LYSZa6 362 where macrophages infected by LYSZa5 released a significant higher level of LDH comparing to 363 LYSZa6. LYSZa3 and LYSZa6 possess weaker cytotoxicity comparing to their ancestral isolates. 364 Such results suggested that these P. aeruginosa isolates produced attenuated virulence towards 365 their eukaryotic hosts due to the downregulation of T6SS genes after evolution. However, the increased occurrence of illness and death due to bacterial coinfection and its 371 complications indicated that there is a much underestimated and neglected influence of bacterial 372 coinfection on the disease progression in COVID-19 patients. Here in this study, we isolated the 373 ancestry and the progeny isolate pairs of hospital-acquired P. aeruginosa from two critically ill 374 COVID-19 patients to investigate its adaptive and convergent evolution during coinfection with 375 SARS-CoV-2 virus. We found that P. aeruginosa upregulates its alginate biosynthesis and 376 downregulates its T6SS to increase its fitness in the niche and reduce its virulence to escape from 377 host clearance for longer colonization. To the best of our knowledge, this is the first study 378 describing the evolution of P. aeruginosa during coinfection with SARS-CoV-2 virus in COVID-379 19 patients.

The isolates collected from the two patients belongs to the same sequence type. In addition, the 381 ancestry isolates appeared after more than 10 days of hospitalization, they are therefore hospital-382 acquired P. aeruginosa strain. Genomic sequencing analysis revealed that the genomes of the 383 isolates carry specific GIs for DNA repair and protein secretion systems, and ARGs for multiple 384 drug classes. Common genomic modification events identified between the progeny isolates and 385 the ancestral isolates indicated the genomic changes in genes related to T6SS, tla3 and tli5b3. 386 However, no change was observed in these two genes at transcriptional level most probably due 387 to the short evolving time and the complexity of gene regulation cascades. Although the 388 expression of these two genes were not affected, we indeed observed lots of significant changes 389 at transcriptional level.

Through differential gene expression analysis, we found that genes involved in alginate 391 biosynthesis were upregulated greatly while genes involved in T6SS protein secretion system 392 were significantly downregulated. Alginate is an essential EPS component converting non- 

Overproduction of alginate and increased biofilm formation were observed in the progeny strains 400 in this study, inferring the potential mechanism adopted by P. aeruginosa in COVID-19 patients 401 to interfere with host defense systems. Thick alginate layer is also probably a key contributing Table S7 . ( o r f X ) P A 0 0 7 8 ( t s s L 1 ) P A 0 0 7 9 ( t s s K 1 ) P A 0 0 8 0 ( t s s J 1 ) P A 0 0 8 1 ( f h a 1 ) P A 0 0 8 2 ( t s s A 1 ) P A 0 0 8 3 ( t s s B 1 ) P A 0 0 8 4 ( t s s C 1 ) P A 0 0 8 5 ( h c p 1 ) P A 0 0 8 6 ( t a g J 1 ) P A 0 0 8 7 ( t s s E 1 ) P A 0 0 8 8 ( t s s F 1 ) P A 0 0 7 5 ( p p p A ) P A 0 0 7 7 ( i c m F 1 ) P A 0 0 7 0 ( t a g Q 1 ) P A 0 0 7 1 ( t a g R 1 ) P A 0 0 7 2 ( t a g S 1 ) P A 0 0 7 3 ( t a g T 1 ) P A 0 0 7 4 ( p p k A ) P A 0 0 7 6 ( t a g F 1 ) P A 0 0 8 9 ( t s s G 1 ) P A 0 0 9 0 ( c l p V 1 ) P A 0 0 9 1 ( v g r G 1 ) P A 0 0 9 2 ( t s i 6 ) P A 0 0 9 3 ( t s e 6 ) P A 0 0 9 4 ( e a g T 6 )

HSI-I HSI-II HSI-III 

SARS-CoV-2-triggered neutrophil extracellular traps mediate COVID-19 pathology

COVID-19: consider cytokine storm syndromes and immunosuppression

Respiratory Viral Infection-Induced Microbiome Alterations and Secondary Bacterial