ABSTRACT
The type of capsular polysaccharide (CPS) on the cell surface of Acinetobacter baumannii can determine the specificity of lytic bacteriophage under consideration for therapeutic use. Here, we report the isolation of a phage on an extensively antibiotic resistant ST2 A. baumannii isolate AB5001 that carries the KL3 CPS biosynthesis gene cluster predicting a K3-type CPS. As the phage did not infect isolates carrying KL3 or KL22 and known to produce K3 CPS, the structure of the CPS isolated from A. baumannii AB5001 was determined. AB5001 produced a variant CPS form, K3-v1, that lacks the β-d-GlсpNAc side chain attached to the d-Galp residue in the K3 structure. Inspection of the KL3 sequence in the genomes of AB5001 and other phage-susceptible isolates with a KL3 locus revealed single-base deletions in gtr6, causing loss of the Gtr6 glycosyltransferase that adds the missing d-GlсpNAc side chain to the K3 CPS. Hence, the presence of this sugar profoundly restricts the ability of the phage to digest the CPS. The 41-kb linear double-stranded DNA (dsDNA) phage genome was identical to the genome of a phage isolated on a K37-producing isolate and thus was named APK37.1. APK37.1 also infected isolates carrying KL116. Consistent with this, K3-v1 resembles the K37 and K116 structures. APK37.1 is a Friunavirus belonging to the Autographiviridae family. The phage-encoded tail spike depolymerase DpoAPK37.1 was not closely related to Dpo encoded by other sequenced Friunaviruses, including APK37 and APK116.
IMPORTANCE Lytic bacteriophage have potential for the treatment of otherwise untreatable extensively antibiotic-resistant bacteria. For Acinetobacter baumannii, most phage exhibit specificity for the type of capsular polysaccharide (CPS) produced on the cell surface. However, resistance can arise via mutations in CPS genes that abolish this phage receptor. Here, we show that single-base deletions in a CPS gene result in alteration of the final structure rather than deletion of the capsule layer and hence affect the ability of a newly reported podophage to infect strains producing the K3 CPS.
KEYWORDS: Acinetobacter baumannii, capsular polysaccharide, K locus, KL3, K3-v1, APK37.1
INTRODUCTION
On a global scale, Acinetobacter baumannii has been listed as one of the six leading bacterial pathogens responsible for the majority of deaths associated with antimicrobial resistance (1). In recent years, lytic bacteriophage (phage) that selectively infect and lyse bacteria have been reinvestigated for use in treating otherwise untreatable multidrug-resistant infections and have shown promise as therapeutic alternatives or adjuncts to antibiotic chemotherapy (2–5). In most cases, the initial receptor involved in phage adsorption to the A. baumannii cell surface is the capsular polysaccharide (CPS) layer surrounding the bacterial cell (6–9). However, individual strains produce just one of a large number of CPS types found in the species. Hence, screening for phage activity against diverse collections of A. baumannii strains that differ in the type of CPS produced is needed to build versatile phage libraries or biobanks for a more targeted approach to therapeutics screening.
CPS type is largely determined by genes found at the chromosomal K locus (KL) for CPS biosynthesis in A. baumannii genomes (10), and more than 240 different sets of genes have been identified at this location to date (11–13). However, the CPS structures for only ~75 different KL have been determined. Differences between known structures include differences in the types of carbohydrate constituents and noncarbohydrate decorations present in the repeating oligosaccharide (OS) K-units that make up the CPS as well as the anomeric configurations of the glycosidic linkages between sugars and also between K-units (14–22). As more and more CPS structural data have become available, prediction of CPS structures from whole-genome sequences via detection of specific sets of CPS biosynthesis genes at the K locus has become possible using the bioinformatics typing tool Kaptive (11, 13).
Treatment using alternate phage therapies is most appropriate for infections that fail to respond to clinically available antibiotics, and such strains often carry either KL3 or KL22 CPS biosynthesis gene clusters. In a recent analysis of publicly available A. baumannii genome assemblies, ~20% of sequenced isolates were found to carry KL3 or KL22 (11), which differ only in the presence of a single gene (annotated as pgt1 in Fig. 1A) that has no known role in CPS biosynthesis (Fig. 1A). The differences between these KL therefore do not affect the carbohydrate composition of the CPS produced (23–26), and the shared structure is referred to as the K3 type (11). This CPS type has been associated with increased virulence (27, 28) and also with carbapenem resistance in nosocomial cases of bacteremia and pneumonia and increased rates of admission to intensive care units (29). Hence, there is a need for phage specific for K3.
FIG 1.
(A) Comparison of the A. baumannii KL3, KL22, KL37, and KL116 CPS biosynthesis gene clusters. Figures were drawn using EasyFig (44) with K locus sequences from NCBI accession numbers CP012004.1 (KL3, ATCC 17978), KC526920.1 (KL22, LUH5537), KX712115.1 (KL37, UV_1036), and CP020590 (KL116, 15A34). The scale bar is shown below. Predicted functional categories of gene products are indicated by colors (legend shown below), and gray shading is protein sequence identity determined by tblastx with the scale also shown below. (B) Positions of the deletions or insertions in the gtr6 gene and length of predicted gene products. Length of the gene is shown above, and length of gene product (blue box) is shown. Black arrows indicate the locations of the single-base deletion in AB5001/AB4932/AB4957 and MAR-13-1378 or the ISAba13 insertion sequence in KL22 of HUMC1 (25). Hashed portion in MAR-13-1378 and HUMC1 products indicate extension past the deletion/insertion to the nearest alternate stop codon.
The K3 structure includes a main chain comprised of a →3)-α-d-Galp-(1→6)-β-d-Glсp-(1→3)-β-d-GalpNAc-(1→ trisaccharide, with two different sugars, 2,3-diacetamido-2,3-dideoxy-d-glucuronic acid (d-GlcNAc3NAcA) and a d-GlcpNAc branching from the d-Galp residue at positions C4 and C6, respectively (Fig. 2A) (23–26, 30). Recently, a study showed that the loss of the branching d-GlcpNAc sugar in A. baumannii isolate HUMC1 carrying KL22 via interruption of the gene for the Gtr6 glycosyltransferase by an ISAba13 insertion sequence significantly increased virulence of cells by inhibiting phagocytosis in vivo (25). Although the loss of this sugar appears clinically important, the effect of this structural variation on phage susceptibility has not been explored.
FIG 2.

CPS structures from A. baumannii isolates. (A) K3-type from isolates with KL3 (ATCC 17978 [23] and SMAL [30]) or KL22 (15827 [25], LUH5537 [24], and SK44 [26]). The degree of O-acetylation of d-GlcNAc3NAcA is ~50% in KL3 strains and ~90% in KL22 strains. (B) K3-v1 CPS type and MPS and OS from A. baumannii AB5001. The degree of O-acetylation of d-GlcNAc3NAcA is ~50%. (C) K37 CPS (24, 38). (D) K116 CPS (38). Glycosyltransferase, Itr, and Wzy names are shown next to the linkage each is assigned to.
In this study, we report the properties of a phage isolated on an A. baumannii isolate carrying the KL3 locus and thus predicted to produce a K3 CPS. The phage did not grow on known K3 producers, and the reason for this was traced to an alteration in the CPS structure caused by single-base deletions in the gtr6 gene that alter the final structure.
RESULTS
A. baumannii AB5001 is an extensively antibiotic-resistant GC2 isolate carrying KL3.
A. baumannii isolate AB5001 was recovered from a blood infection of a US military patient in 2008 (31). Using Kaptive (11), the available draft genome sequence (NCBI accession number LREN00000000.1) (32) was found to include the KL3 locus and hence predicted to produce a K3-type CPS. The genome belongs to ST2 in the A. baumannii Institut Pasteur multilocus sequence typing (MLST) scheme and ST452 in the Oxford MLST scheme (Table 1), identifying it as a strain belonging to the major globally disseminated clonal lineage, Global Clone 2 (GC2), which includes pan-resistant strains for which bacteriophage therapy is warranted. AB5001 is recorded as resistant to most agents currently used for therapy, including imipenem, broad-spectrum cephalosporins, aminoglycosides, amikacin and gentamicin, sulfonamides, tetracycline, and fluoroquinolones (see Table S1 in reference 31). Consistent with this phenotype, from the draft genome sequence, it was found that AB5001 carries the oxa23 gene for resistance to carbapenems and blaPER-1, which confers resistance to ceftazidime, cefotaxime, and aztreonam. AB5001 also carries the aphA1 (kanamycin and neomycin resistance) gene and the aphA6 (amikacin resistance) and aacC1 (gentamicin resistance) genes. The sul2 sulfonamide resistance, tet(B) tetracycline resistance, and strAB streptomycin resistance genes are also present.
TABLE 1.
A. baumannii strains tested in this study
| Strain name | Country | Yr | Source | STiph | STOXi | K locus | Resistance determinants | NCBI accession no. |
|---|---|---|---|---|---|---|---|---|
| ATCC 17978a | France | 1951 | Meninges | ST437 | ST112 | KL3 | sul2 | CP012004.1 |
| ATCC 19606 | USA | <1948 | Urine | ST52 | ST931 | KL3 | sul2 | CP045110.1 |
| AB5711 (MRSN1310) | USA | 2009 | Blood/sepsis | ST2 | ST452 | KL3 | aacC1, aphA6, blaPER-1, oxa23 (x3), strA (x2), strB, sul2 | AHAJ00000000.1 |
| AB5001 (MRSN954) | USA | 2008 | Blood/sepsis | ST2 | ST452 | KL3 | aacC1, aphA1, aphA6, blaPER-1, blaTEM-116, oxa23, strA, strB, sul2 | LREN00000000.1 |
| AB4932 (MRSN949) | USA | 2008 | Sputum | ST2 | ST452 | KL3 | aacC1, aphA6, blaPER-1, oxa23, strB, sul2 | LREK00000000.1 |
| AB4957 (MRSN951) | USA | 2008 | Sacral/osteomyelitis | ST2 | ST452 | KL3 | aacC1, aphA6, blaPER-1, oxa23, strB, sul2 | LREL00000000.1 |
| MAR 13-1378 | Russia | 2013 | Respiratory tract | ST78 | Newb | KL3 | NDc | JAPQKD000000000 |
| LUH5537e | Netherlands | Unknown | Unknown | ST2 | ST1839 | KL22 | aacC1, aadA1, aphA1, blaTEM-1D, catA1, qacE, sul1 | KC526920.1 d |
| RES-1687 | Russia | 2003 | Skin and soft tissue | NDc | NDc | KL22 | aadA24, aadB, aphA6, sul2 | JAPQKA000000000 |
| REV-3116 | Russia | 2007 | Blood | ST2 | ST281 | KL22 | aacC1, aadA1, aphA1, blaTEM-1D, qacE, sul1 | JAPQKC000000000 |
| NIPH 146f | Czech Republic | 1993 | Wound | ST25 | ST276 | KL37 | APOU00000000.1 | |
| KZ-1101 | Kazakhstan | 2016 | Skin/soft tissue | ST132 | ST2213 | KL37 | JAPYKX000000000.1 | |
| MAR-303g | Russia | 2011 | Inpatient | ST12 | ST953 | KL116 | aadB, aphA6, blaOXA-120 | JANSJS000000000 |
| TP1 | USA | 2016 | Peritoneal drain | ST570 | ST1578 | KL116 | aac(6′)-Ib-cr, aacA4, aadA1, aphA1, armA, blaTEM-1D, catB8, mph(E), msr(E), oxa23, qacE, sul1 | NZ_CP056784.1 |
| TP3 | USA | 2016 | Pancreatic drain | ST570 | ST1578 | KL116 | aacA4, aadA1, aphA1 (x2), aphA6, armA, blaNDM-1, blaTEM-1D, catB8, mph(E), msr(E), oxa23, qacE, sul1 | NZ_CP060013.1 |
The structure of CPS produced by this strain was determined previously (23, 25). CPS includes the d-GlcNAc branch.
Single locus variant of ST1961 and ST2331.
ND, not detected.
Only K locus sequence is available.
The structure of CPS produced by this strain was determined previously (24). CPS includes the d-GlcNAc branch.
The structure of CPS produced by this strain was determined previously (38).
For STip: sequence type in Institut Pasteur scheme.
For STOX: sequence type in Oxford scheme.
Isolation of a new A. baumannii bacteriophage.
Samples of sewage water collected in 2018 from the Moscow region in Russia were used to isolate phage that could infect the extensively antibiotic-resistant A. baumannii isolate AB5001 carrying KL3. A phage producing relatively large clear plaques (about 3 to 4 mm in diameter) surrounded by visible halos (Fig. 3A) was isolated on a lawn of AB5001. The phage was recovered, and the nucleotide sequence of the genome was determined. The phage genome sequence was identical to that of a phage independently isolated from the same sewage sample on a lawn of A. baumannii isolate KZ-1101 that carries the KL37 CPS biosynthesis gene cluster (Fig. 1A). Hence, AB5001 and KZ-1101 were infected by one and the same phage, which, following the nomenclature used by Popova and colleagues (9), was assigned the name APK37.1.
FIG 3.
Characterization of APK37.1. (A) Plaques with opaque haloes formed by phage APK37.1 on a lawn of A. baumannii AB5001. A magnified image of a plaque is shown below, with the bar indicating halo diameter. (B) Pairwise nucleotide sequence comparison of APK37.1 and Fri1 genomes. Shading shows DNA sequence identity, and the scale is shown below. Open reading frames are colored by functional category with the key shown below. The figure was constructed by Easyfig (44). (C) Proteomic phylogeny of related Friunavirus genomes. Circles on node tips indicate phage that have demonstrated specificity for strains that carry the KL37 or KL116 CPS biosynthesis gene clusters. Red denotes phage APK37.1.
Phage APK37.1.
The APK37.1 genome was annotated and submitted to GenBank under accession number MZ967493.1. APK37.1 has a linear double-stranded DNA (dsDNA) genome of 40,966 bp with terminal repeats of 399 bp and an overall G+C content of 39.2%, consistent with other phage specific to A. baumannii (9). A total of 56 predicted open reading frames (ORFs) all oriented in the same direction were found (Fig. 3B), with one ORF, named gp35, predicted to encode a DNA-dependent RNA polymerase (GenPept accession number UAW07714.1), which is characteristic of members of the phage family Autographiviridae (33). Genome comparison of APK37.1 through calculation of average nucleotide identity (ANI) revealed an ANI of 92.76% with the prototype podophage Fri1 (GenBank accession number NC_028848.1), which belongs to Friunavirus of the Beijerinckvirinae subfamily within the family Autographiviridae (34). A pairwise sequence comparison of APK37.1 and Fri1 revealed significant sequence homology (Fig. 3B), particularly in the “late genome” region that includes genes for common podophage structural proteins (gp39-gp48 in APK37.1). A proteomic tree calculated from the nucleotide sequences of related A. baumannii podophage in the Friunavirus genus (Fig. 3C) revealed that APK37.1 groups among members of this family but in a different clade to Fri1. Taken together, the data indicated that APK37.1 belongs to the Friunavirus genus in the Autographiviridae family.
Phage APK37.1 cannot lyse other K3-producing strains.
To confirm the specificity of APK37.1 for K3 CPS, the phage was applied to a small collection of 10 other isolates (Table 1) with whole-genome sequences that were known or were predicted to include either KL3 (n = 6) or KL22 (n = 4) sequences at the K locus. These strains included ATCC 17978 (KL3) and LUH5537 (KL22), for which the chemical structure of the CPS had been determined previously and shown to be the K3 type (23–25). However, APK37.1 did not infect these two strains. In addition to AB5001, phage APK37.1 could infect only three other isolates that carried KL3 (Table 2). Hence, overall, three KL3 isolates (43%) and all three KL22 isolates (100%) were found to be resistant to APK37.1 infection. The difference in APK37.1 infectivity suggested that AB5001 might produce a different or modified CPS structure, and, as the chemical structures of CPS produced by the isolates susceptible to APK37.1 infection have not been previously elucidated, AB5001 was chosen for further chemical analyses by NMR.
TABLE 2.
Susceptibility of A. baumannii isolates with KL3 or KL22 to APK37.1 infection
| Strain name | K locus | Susceptibility to APK37.1 | gtr6 frameshift | Frameshift details | First base position |
|---|---|---|---|---|---|
| ATCC 17978 | KL3 | − | − | − | − |
| ATCC 19606 | KL3 | − | − | − | − |
| AB5711 | KL3 | − | − | − | − |
| LUH5537 | KL22 | − | − | − | − |
| RES-1687 | KL22 | − | − | − | − |
| REV-3116 | KL22 | − | − | − | − |
| AB5001 | KL3 | + | + | −A in string of 7 × A | 81 |
| AB4932 | KL3 | + | + | −A in string of 7 × A | 81 |
| AB4957 | KL3 | + | + | −A in string of 7 × A | 81 |
| MAR 13-1378 | KL3 | + | + | −A in string of 9 × A | 440 |
Structural elucidation of the AB5001 CPS.
A high-molecular-mass CPS was isolated from cells of A. baumannii AB5001 by the phenol-water procedure (35). Sugar analysis of the CPS showed the presence of Glc, Gal, and GalN in a ratio of 1.0 to 0.5 to 1.0, respectively.
The CPS was studied by nuclear magnetic resonance (NMR) spectroscopy, including one-dimensional 1H and 13C NMR experiments and two-dimensional 1H,1H correlation spectroscopy (COSY), total correlation spectroscopy (TOCSY), rotating-frame nuclear Overhauser effect spectroscopy (ROESY) (Fig. S1 in the supplemental material), 1H,13C heteronuclear single quantum coherence (HSQC) (Fig. S2), and heteronuclear multiple-bond correlation (HMBC) experiments. Five sugar spin systems, including those for α-Gal (unit A), β-Glc (unit B), β-GalNAc (unit C), β-GlcNAc3NAcA, and β-GlcNAc3NAcA-4Ac (unit D′) were identified (Table 3), all monosaccharides being in the pyranose form.
TABLE 3.
1H and 13C NMR chemical shifts (δ, ppm) of the CPS of A. baumannii AB5001 and MPS derived from O-deacetylation of the CPS
| Sugar residue | C-1 H-1a |
C-2 H-2 |
C-3 H-3 |
C-4 H-4 |
C-5 H-5 |
C-6 H-6 (6a,6b) |
|---|---|---|---|---|---|---|
| CPSb | ||||||
| →3)-α-d-Galp-(1→ A |
99.7 | 66.8 | 81.6 | 77.0 | 71.4 | 61.7 |
| 4.94 | 3.73 | 3.92 | 4.40 | 3.64 | 3.65, 3.79 | |
| →6)-β-d-Glcp-(1→ B |
106.0 | 74.4 | 77.1 | 70.3 | 77.6 | 66.6 |
| 4.56 | 3.32 | 3.48 | 3.60 | 3.59 | 3.73, 4.01 | |
| →3)-β-d-GalpNAc-(1→ C |
105.0 | 52.9 | 82.0 | 71.4 | 76.0 | 62.5 |
| 4.65 | 4.14 | 3.87 | 3.93 | 3.71 | 3.82, 3.82 | |
| β-d-GlcpNAc3NacA-(1→ D |
102.6 | 54.8 | 56.2 | 71.4 | 75.0 | 174.1 |
| 5.08 | 3.84 | 4.03 | 3.64 | 4.03 | ||
| β-d-GlcpNAc3Nac4AcA-(1→ D′ |
102.4 | 54.4 | 54.2 | 72.1 | 77.1 | 174.1 |
| 5.08 | 3.96 | 4.23 | 4.96 | 3.90 | ||
| MPSc | ||||||
| →3)-α-d-Galp-(1→ A |
100.0 | 69.1 | 81.8 | 77.4 | 71.8 | 62.1 |
| 4.93 | 3.73 | 3.92 | 3.40 | 3.93 | 3.65, 3.77 | |
| →6)-β-d-Glcp-(1→ B |
106.2 | 74.7 | 77.5 | 70.7 | 75.9 | 67.1 |
| 4.54 | 3.31 | 3.48 | 3.59 | 3.59 | 3.75, 4.00 | |
| →3)-β-d-GalpNAc-(1→ C |
105.3 | 53.2 | 82.2 | 69.5 | 76.3 | 62.8 |
| 4.66 | 4.13 | 3.88 | 4.16 | 3.71 | 3.82, 3.82 | |
| β-d-GlcpNAcA-(1→ D |
102.9 | 55.1 | 56.5 | 71.6 | 77.1 | 175.2 |
| 5.10 | 3.85 | 4.03 | 3.67 | 3.97 | ||
| OSd →3)-α-d-Galp-(1→ A |
100.1 | 68.8 | 81.7 | 77.1 | 71.8 | 61.8 |
| 4.88 | 3.71 | 3.92 | 4.39 | 3.93 | 3.65, 3.77 | |
| →3)-β-d-GalpNAc-(1→ C |
105.4 | 54.2 | 72.6 | 69.1 | 76.3 | 62.6 |
| 4.57 | 4.00 | 3.72 | 3.96 | 3.69 | 3.83 | |
| β-d-GlcpNAcA-(1→ D |
102.6 | 54.9 | 56.0 | 71.6 | 77.2 | 175.9 |
| 5.08 | 3.82 | 4.04 | 3.62 | 3.88 | ||
| →1)-Gro B′ |
70.1 | 71.4 | 63.9 | |||
| 3.56, 3.73 | 3.91 | 3.62, 3.66 | ||||
H NMR chemical shifts are italicized.
Chemical shifts for the N-acetyl group are δH 1.90 to 2.07, δC 22.2 to 23.8 (Me), and 175.8 to 176.0 (CO).
Chemical shifts for the N-acetyl group are δH 2.00 to 2.01, δC 23.4 to 23.7 (Me), and 176.1 to 176.3 (CO).
Chemical shifts for the N-acetyl group are δH 1.99 to 2.01, δC 23.7 to 24.1 (Me), and 176.0 to 176.1 (CO).
In the 1H NMR spectrum, there were five signals for anomeric protons at δ 4.56 to 5.08, five signals of H-2 and H-3 of the NAc-linked carbons at δ 3.87 to 4.23 and H-4 of unit D′ at an OAc-linked carbon at δ 4.96. The 13C NMR spectrum showed five signals for anomeric carbons at δ 99.7 to 106.0, signals of N-linked C-2 and C-3 of units C, D, and D′ at δ 52.9 to 54.8, and a signal of a CO2H group (C-6) of units D and D′ at δ 174.1.
The 1Н and 13С NMR spectra of the CPS were assigned using two-dimensional 1Н,1Н COSY, 1Н,1Н TOCSY, and 1Н,13C HSQC (Fig. S2) experiments. The positions of substitution and the sequence of the monosaccharide residues were established by 1Н,1Н ROESY (Fig. S1) and 1Н,13C HMBC experiments, which showed correlations between atoms of the neighboring monosaccharide residues.
In the 1Н,1Н TOCSY spectrum, there were correlations H-1/H-2–H-4 for the sugars having the galacto configuration (units A and C) and H-1/H-2–H-6 for the sugars having the gluco configurations (units B, D, and D′). Units D and D′ were identified as hexuronic acid derivatives based on the correlation between CO2H (C-6) and H-5 at δC/δH 174.1/4.03 and 174.1/3.90, respectively, in the 1H,13С HMBC spectrum.
Low-field positions at δ 81.6, 82.0, and 66.6 of the signals for C-3 of units A and C and C-6 of unit B, respectively, showed that the CPS is branched, with three monosaccharide residues (A to C) in the main chain and unit D attached as a side chain.
O-Deacetylation was performed to confirm the structure of the CPS shown in Fig. 2B. The modified polysaccharide (MPS) obtained was subjected to Smith degradation, which cleaved the α-d-Galp-(1→6)-β-d-Glсp linkage to give an oligosaccharide (OS), namely, a trisaccharide glycoside with glycerol (Gro) as aglycon. The structures of the MPS and OS (Fig. 2B) were established by NMR spectroscopy as described above for the CPS (for the 1H,13C HSQC spectra, see Fig. S3 and S4; for the assigned 1H and 13C chemical shifts, see Table 3), and the CPS structure was confirmed.
Therefore, the AB5001 CPS includes a trisaccharide backbone of →3)-α-d-Galp-(1→6)-β-d-Glсp-(1→3)-β-d-GalpNAc-(1→ and a 4-O-acetylated d-GlcpNAc3NAcA residue branching from position 4 of the d-Galp in the main chain. The configuration is very similar to that of the K3-type CPS. However, the second side chain, d-GlсpNAc, attached to position 6 of the d-Galp residue is missing (Fig. 2B). Hence, the variant CPS was designated the name K3-v1.
CPS biosynthesis gene cluster in the genome sequence of AB5001.
To determine the reason for this altered phenotype, the sequence at the K locus was reexamined. The CPS biosynthesis gene cluster in the AB5001 whole-genome sequence shares 100% coverage and 98.15% nucleotide sequence identity with KL3 from ATCC 17978 reference sequence (GenBank accession number CP018664.1), which produces the K3 CPS with the d-GlсpNAc side branch (23). All 20 of the expected coding sequences for KL3 were present. However, a length discrepancy of −1 bp was detected between the sequences using the Kaptive typing tool. A pairwise sequence alignment of the KL3 gene clusters from AB5001 and ATCC 17978 identified a single missing “A” residue in the gtr6 gene in a run of seven A residues beginning 81 bases downstream from the start codon (Fig. S5). Translation from the next available start codon would therefore produce a protein of 356 amino acids (aa) (Fig. 1B), as opposed to the 387-aa wild-type protein in ATCC 17978. The Gtr6 glycosyltransferase is known to be responsible for linking the d-GlcpNAc branched sugar (25), and this explains the absence of this sugar in the AB5001 K3-v1 CPS.
Frameshifts in gtr6 determine susceptibility to infection by phage APK37.1.
Inspection of the gtr6 sequence in the collection of KL3 and KL22 isolates used to examine the specificity of APK37.1 revealed that all six resistant isolates carried an intact copy of the gtr6 gene (Table 2). For two other isolates, AB4932 and AB4957, that were susceptible to APK37.1 infection, the same single-base deletion at position 81 was detected in the genome. However, a different single base deletion was detected in the gtr6 sequence from the A. baumannii isolate MAR 13-1378. This deletion is in a string of nine “A” residues beginning at position 440 from the gtr6 start codon (Fig. S5). This frameshift leads to a truncated product with only the first 149 aa the same as the complete Gtr6 sequence (Fig. 1B). Hence, it appears that the single-base deletion near base 81 or base 440 has abrogated the function of Gtr6, leading to loss of the side chain.
Specificity of APK37.1 for other A. baumannii CPS types.
Like other Acinetobacter podophages, APK37.1 demonstrates specificity for the type of CPS produced by A. baumannii strains infected. This is known to be directed by a depolymerase enzyme surrounding the phage tail that recognizes specific CPS structures and hydrolyses a bond in the polysaccharide (6). Visible halos in APK37.1 plaques on an AB5001 lawn (Fig. 3A) suggest depolymerase activity (34, 36), and, consistent with this, a pectate lyase fold (Pfam PF12708) characteristic of tail spike/structural depolymerases was identified in the gp49 protein sequence (black ORF in Fig. 3B; GenPept accession number UAW07728.1) encoded by the APK37.1 genome.
Further inspection of the predicted product of gp49 (here referred to as DpoAPK37.1) revealed that the amino acid sequence shares 52 to 53% amino acid sequence identity (>97% sequence coverage) with the depolymerases encoded by APK37 (gp48; GenBank accession number AZU99445) and AbTP3phi1 (gp48; GenBank accession number UNI74976.1) (Fig. 4A). APK37 and AbTP3phi1 are two other A. baumannii podophage previously shown to infect the K37 CPS-producing A. baumannii isolate NIPH 146 (9) and K116 CPS-producing A. baumannii isolate TP1 (37), respectively, and their encoded depolymerases are 96.2% identical (Fig. 4A). Considering the relationship between depolymerases and the fact that APK37.1 had also been independently isolated on the K37-producing isolate KZ-1101 (see above), the infectivity of APK37.1 on a broader set of strains carrying either KL37 or KL116 was also tested (Table 1).
FIG 4.

Specificity of APK37.1 for more than one CPS type. (A) Amino acid sequence percentage identity matrix of depolymerases encoded by APK37.1, APK37, AbTP3phi1, and APK116 (GenPept accession numbers UAW07728.1, AZU99445.1, UNI74976.1, and QHS01530.1, respectively). (B) APK37.1 spots on lawns of K3-v1, K37, and K116 strains. (C) Amino acid sequence alignment of depolymerases encoded by APK37.1, APK37, and AbTP3phi1. Numbers indicate amino acid ranges. Alignment is colored by identity using the CLUSTAL colormap visualized in MView.
APK37.1 was found to infect all five tested strains that carry either KL37 or KL116 (see Fig. 4B for APK37.1 infection of representative K37- and K116-producing isolates), indicating broad host specificity for different CPS types. Inspection of the phylogeny (Fig. 3A) revealed that APK37.1 belongs to a subclade that is separate from APK37, AbTP3phi1, and one further phage, APK116, that is also known to infect K116 CPS-producing A. baumannii strains (9). However, DpoAPK37.1 shares only 35.4% amino acid identity to the depolymerase from APK116 (Fig. 4A) and is therefore more closely related to the depolymerases from APK37 and AbTP3phi1. Similar to other capsular depolymerases encoded by members of the Friunavirus genus (6, 9), much of the homology between DpoAPK37.1 and depolymerases from APK37 and APK116 is at the N terminus in the first 140 amino acids (Fig. 4C), although APK37.1 and APK37 do share further homology extending up to amino acid position 250. However, the C-terminal sequence of all three depolymerases is divergent. As the C terminus is responsible for receptor recognition, diversity in this sequence may reflect differences in the range of specificity for different CPS types. However, this was not further investigated.
DISCUSSION
In this study, we describe APK37.1, a novel A. baumannii bacteriophage of the Friunavirus genus, and show that it can infect only a subset of isolates that carry the KL3 or KL22 CPS biosynthesis loci. The reason for this was traced to the loss of the d-GlсpNAc side branch in the K3-type structure arising from a single-base deletion in the gtr6 gene. While small changes in the nucleotide sequences of KL have previously been shown to alter phage susceptibility in some isolates (7, 37), in many cases, these changes have, or likely have, involved loss of the CPS, rendering cells resistant to further infection due to removal of the primary phage receptor. To the best of our knowledge, this study is the first report of a single-base deletion in an A. baumannii CPS biosynthesis locus leading to a demonstrated change in the composition of the structure rather than complete deletion of the CPS layer on the cell surface.
Before this study, loss of the d-GlcpNAc side branch in the K3 CPS had been described for the extensively resistant A. baumannii clinical isolate HUMC1 (25), in which an ISAba13 sequence was found to interrupt the gtr6 gene. The same phenotype is produced by two different and independent −1 deletions in gtr6 that we report here. Both of these −1 deletions occur in strings of “A” residues, which may suggest that these regions are more prone to strand slippage during replication and may be occurring at low levels in a population. Hence, these frameshifts are likely maintained in subsequent generations under appropriate selective pressure and without selection may be reversible. Phenotypic reversion complicates treatment using phage; therefore, further work will be needed to establish whether these mutations are phase variable.
Assessment of the gtr6– phenotype in constructed isogenic mutants of KL3- and KL22-carrying strains in a previous study revealed that the absence of the d-GlcpNAc side branch results in inhibition of phagocytosis, leading to increased lethality in vivo (25). Mutations in the gtr6 gene (via both frameshifts and insertion sequence [IS] disruption) therefore lead to changes in not only susceptibility to phage but also resistance to the host immune response. Determination of whether mutations that inactivate Gtr6 are globally disseminated and/or are associated with the most resistant isolates belonging to major antibiotic-resistant clonal lineages is therefore important. However, while the disruption of gtr6 significantly increases the virulence of A. baumannii strains, APK37.1 could be an appropriate choice for therapy or decontamination of clinical environments where outbreaks of highly resistant and hypervirulent gtr6– strains are occurring.
Detection of small differences in KL sequences is likely to prove important if other phage isolated for use in phage therapy are also sensitive to the presence of this CPS side chain. However, while Kaptive is an effective tool to detect and type A. baumannii KL sequences, the genes reported as present are based on the detection of ≥90% of the expected translated amino acid sequence (using default parameters). Hence, as the −1 deletion in the AB5001 gtr6 occurs at the beginning of the gene (Fig. 1B), the translated product meets the minimum 90% coverage threshold, and Kaptive reports the gene as present. Likewise, for HUMC1, the insertion of ISAba13 occurs close to the 3′ end of gtr6 (Fig. 1B), and gtr6 is again recorded as present given that the translated product (to an alternate stop codon) meets the minimum coverage threshold. Manual inspection of the gtr6 gene is therefore currently required to identify all disruptions that may abolish Gtr6 function, and rapid automated detection of Gtr6 loss will warrant more sensitive methods.
Interestingly, other APK37.1-sensitive isolates produce K37 (Fig. 2C) or K116 (Fig. 2D) CPSs that do not include a side branch residue attached to position 6 of the terminal d-Galp residue in the main chain (38). The arrangements of the main chain residues are also closely related to the K3-v1 structure as well as the glycosidic linkages between the oligosaccharide K-units that are formed by the Wzy polymerase. While specific sites of CPS cleavage for DpoAPK37.1 were not determined in this study, it is clear that the absence of the d-GlсpNAc side branch is necessary for susceptibility to infection by phage APK37.1, and the purified depolymerase digests K3-v1 CPS but not K3 (O. Timoshina and M. Shneider, unpublished observations). The depolymerase may hydrolyze a common structural feature shared by K3-v1, K37, and K116. The ability of APK37.1 to infect strains producing three different CPS types allows for broader application, although further work will be needed to explore phage specificity, host interactions, and therapeutic safety.
MATERIALS AND METHODS
Bacterial strains and cultivation.
A. baumannii isolates used in this study (Table 1) were obtained from Mikhail Edelstein (Institute of Antimicrobial Chemotherapy, Smolensk State Medical University, Smolensk, Russia), Bin Liu (TEDA Institute of Biological Sciences and Biotechnology, Nankai University, TEDA, Tianjin, People’s Republic of China), Alexandr Nemec (Laboratory of Bacterial Genetics, National Institute of Public Health, Prague, Czech Republic), Raffaele Zarrilli (Department of Molecular Medicine and Medical Biotechnology, University of Naples Federico II, Naples, Italy), Dean Scholl (AvidBiotics Corp., South San Francisco, CA, USA) and Ry Young (Center for Phage Technology, Texas A&M AgriLife Research and Texas A&M University, College Station, TX, USA). Bacteria were routinely cultivated in 2TY medium at 37°C.
Bioinformatic analysis of bacterial genomes.
Draft genome sequences of A. baumannii strains (sequenced in this study or downloaded from accession numbers listed in Table 1) were compiled for local analyses. For A. baumannii isolates for which a genome assembly was not already available, genomic material was extracted and sequenced on a MiSeq platform using a Nextera DNA library preparation kit (Illumina, San Diego, CA). Resulting short-read data were assembled into contigs using SPAdes v 3.10 (39). Multi locus sequence typing (MLST) was performed by submitting genome assemblies to the PubMLST database available at https://pubmlst.org/organisms/acinetobacter-baumannii. Resistance determinants were detected using ResFinder v. 4.1 (https://github.com/cadms/resfinder) (40). Sequences of CPS biosynthesis gene clusters were identified in genome assemblies using command-line Kaptive v 2.0.4 (41) with the most recent iteration of the A. baumannii KL reference database that includes 241 KL (11). Sequences of the gtr6 gene from each genome were extracted, and sequence differences were identified by a pairwise sequence alignment to gtr6 from ATCC 17978 (GenBank accession number CP012004.1) using CLUSTAL Omega (https://www.ebi.ac.uk/Tools/msa/clustalo/).
Phage isolation, propagation, and sequencing.
A sewage sample collected in 2018 from the Moscow region, Russia, was used to isolate phage APK37.1 on bacterial lawns of A. baumannii isolates AB5001 and KZ-1101 using methods described previously (42). Phage propagation and purification were executed according to the protocols described earlier (9). Genomic material from APK37.1 was isolated from a purified high-titer stock and sequenced as described previously (9). Briefly, ORFs in the sequenced genome of APK37.1 were predicted using GeneMarkS 4.3, manually checked for fidelity, and annotated according to the conventional gp nomenclature system and submitted to GenBank under accession number MZ967493.1. The positions and lengths of terminal repeats were predicted by searching a region of double read depth in comparison with average read depth across the whole genome.
Bioinformatic analysis of phage genomes.
Average nucleotide identity (ANI) for comparison of phage genomes was computed using the OrthoANIu tool (43) and USEARCH over BLAST with default settings. Alignment of APK37.1 and Fri1 was visualized using EasyFig software (44). Nucleotide sequences (FASTA) of phage genomes (listed in Table S1 in the supplemental material) were used to construct a proteomic tree using ViPTree (45). The resulting phylogeny was visualized and annotated using the iTOL web interface (https://itol.embl.de/). Percent identity matrices and multiple pairwise alignments of depolymerase sequences were constructed using CLUSTAL Omega with MView (https://www.ebi.ac.uk/Tools/msa/mview/).
Isolation of the CPSs from A. baumannii isolate AB5001.
Cells of the APK37.1-susceptible A. baumannii isolate AB5001 were harvested by centrifugation (10,000 × g, 20 min), washed with phosphate-buffered saline, suspended in aqueous 70% acetone, precipitated, and dried. Dried cell mass (2.88 g) was treated with 45% aqueous solution of phenol (68°C, 1 h) (35), and the extract was dialyzed without layer separation and freed from insoluble contaminations by centrifugation. The resultant solution was concentrated and treated with cold aqueous 50% CCl3CO2H at 0°C for 1 h. After centrifugation, the supernatant was dialyzed against distilled water. A sample of the native CPS (300 mg) was hydrolyzed with 2% CH3CO2H (100°C, 2 h). Fractionation of the products was performed by gel permeation chromatography on a column Sephadex G-50 Superfine (53 × 3.5 cm; Amersham Biosciences, Sweden) and gave purified CPS (41.7 mg). Elution was performed with 0.1% acetic acid (HOAc) and monitored using a UV detector (Uvicord, Sweden) at 206 nm.
Chemical analyses.
A CPS sample (1 mg) of AB5001 was hydrolyzed with 2 M CF3CO2H (120°C, 2 h), reduced with NaBH4 in 1 M NH4OH (05 mL, 10 mg/mL, 20°C, 1 h), and acetylated with a mixture of pyridine and acetic anhydride (Ac2O) at a ratio of 1:1 (120°C, 2 h). Monosaccharides were analyzed by gas-liquid chromatography (GLC) of the alditol acetates on a Maestro (Agilent 7820) chromatograph (Interlab, Russia) equipped with an HP-5 column (0.32 mm · 30 m) using a temperature program of 160°C (1 min) to 290°C at 7°C min−1.
O-Deacetylation of CPS.
A sample of the CPS from strain AB5001 (46 mg) was heated with aqueous 12% NH4OH (2 mL) at 38°C for 12 h. Fractionation of the products was performed by gel-permeation chromatography on a column Sephadex G-50 Superfine (53 × 3.5 cm; Amersham Biosciences, Sweden) using 0.1% HOAc for elution and was monitored with a UV detector (Uvicord, Sweden) at 206 nm to give an MPS sample (46 mg).
Smith degradation.
A sample of the CPS from strain AB5001 (20 mg) was oxidized with aqueous 0.05 M NaIO4 (2.7 mL) at 20°C for 40 h in the dark and reduced with NaBH4 (80 mg) at 20°C for 16 h. The excess NaBH4 was destroyed with concentrated HOAc, the solution was evaporated, methanol was added to the residues (3 × 1 mL) and evaporated, and the residue was dissolved in 0.3 mL of water and applied to a column (108 × 1.2 cm) of TSK HW-40. A modified polysaccharide was eluted with aqueous 0.1% HOAc and hydrolyzed with 2% CH3CO2H (100°C, 2 h). Fractionation of the products by gel-permeation chromatography on a column (108 × 1.2 cm) of TSK HW-40 in water gave an oligosaccharide (10 mg).
NMR spectroscopy.
Samples were deuterium exchanged by freeze-drying from 99.9% D2O and then examined as solutions in 99.95% D2O. NMR spectra were recorded on a Bruker Avance II 600 MHz spectrometer (Germany) at 60°C. Sodium 3-trimethylsilylpropanoate-2,2,3,3-d4 (δH 0, δC −1.6) was used as an internal reference for calibration. Two-dimensional NMR spectra were obtained using standard Bruker software, and the Bruker TopSpin 2.1 program was used to acquire and process the NMR data. A 60-ms MLEV-17 spin-lock time and a 150-ms mixing time were used in 1H,1H TOCSY and ROESY experiments, respectively. A 60-ms delay was used for evolution of long-range couplings to optimize 1H,13C HMBC experiments for the coupling constant JH,C of 8 Hz. The following correlations were observed in the 1Н,1Н ROESY spectrum of the CPS: H-1 of unit B with H-3 of unit C, H-1 of unit C with H-3 of unit A, H-1 of unit A with H-6 of unit D, and H-1 of units D and D′ with H-4 of unit A.
Data availability.
Whole-genome sequence data from this study are deposited in NCBI under accession numbers MZ967493.1, JAPYKX000000000.1, JAPQKD000000000.1, JAPQKA000000000.1, JAPQKC000000000.1, and JANSJS000000000.1.
ACKNOWLEDGMENTS
We thank Dean Scholl (AvidBiotics Corp.) and Eugene A Sheck (Institute of Antimicrobial Chemotherapy [IAC], Smolensk State Medical University [SSMU]) for providing strains. We also thank Sarah Cahill (Queensland University of Technology, Australia) for bioinformatics assistance. Monosaccharide analysis and CPS structure determination were supported by the Russian Science Foundation (project 19-14-00273). Genetics was supported by an Australian Research Council (ARC) DECRA fellowship (DE180101563) to J.J.K. and an NHMRC Investigator grant (GNT1194978) to R.M.H.
Footnotes
Supplemental material is available online only.
Contributor Information
Johanna J. Kenyon, Email: johanna.kenyon@qut.edu.au.
Daria Van Tyne, University of Pittsburgh School of Medicine.
REFERENCES
- 1.Antimicrobial Resistance Collaborators. 2022. Global burden of bacterial antimicrobial resistance in 2019: a systematic analysis. Lancet 399:629–655. doi: 10.1016/S0140-6736(21)02724-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Schooley R, Biswas B, Gill J, Hernandez-Morales A, Lancaster J, Lessor L, Barr JJ, Reed SL, Rohwer F, Benler S, Segall AM, Taplitz R, Smith DM, Kerr K, Kumaraswamy M, Nizet V, Lin L, McCauley MD, Strathdee SA, Benson CA, Pope RK, Leroux BM, Picel AC, Mateczun AJ, Cilwa KE, Regeimbal JM, Estrella LA, Wolfe DM, Henry MS, Quinones J, Salka S, Bishop-Lilly KA, Young R, Hamilton T. 2017. Development and use of personalized bacteriophage-based therapeutic cocktails to treat a patient with a disseminated resistant Acinetobacter baumannii infection. Antimicrob Agents Chemother 61:e00954-17. doi: 10.1128/AAC.00954-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Wu N, Dai J, Guo M, Li J, Zhou X, Li F, Gao Y, Qu H, Lu H, Jin J, Li T, Shi L, Wu Q, Tan R, Zhu M, Yang L, Ling Y, Xing S, Zhang J, Yao B, Le S, Gu J, Qin J, Li J, Cheng M, Tan D, Li L, Zhang Y, Zhu Z, Cai J, Song Z, Guo X, Chen L-K, Zhu T. 2021. Pre-optimized phage therapy on secondary Acinetobacter baumannii infection in four critical COVID-19 patients. Emerg Microbes Infect 10:612–618. doi: 10.1080/22221751.2021.1902754. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Rao S, Betancourt-Garcia M, Kare-Opaneye YO, Swierczewski BE, Bennett JW, Horne BA, Fackler J, Hernandez LPS, Brownstein MJ. 2021. Critically ill patient with multidrug-resistant Acinetobacter baumannii respiratory infection successfully treated with intravenous and nebulized bacteriophage therapy. Antimicrob Agents Chemother 66:e00824-21. doi: 10.1128/AAC.00824-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Blasco L, Bleriot I, González de Aledo M, Fernández-García L, Pacios O, Oliveira H, López M, Ortiz-Cartagena C, Fernández-Cuenca F, Pascual Á, Martínez-Martínez L, Pachón J, Azeredo J, Tomás M. 2022. Development of an anti-Acinetobacter baumannii biofilm phage cocktail: genomic adaptation to the host. Antimicrob Agents Chemother 66:e01923-21. doi: 10.1128/AAC.01923-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Oliveira H, Costa A, Konstantinides N, Ferreira A, Akturk E, Sillankorva S, Nemec A, Shneider M, Dötsch A, Azeredo J. 2017. Ability of phages to infect Acinetobacter calcoaceticus‐Acinetobacter baumannii complex species through acquisition of different pectate lyase depolymerase domains. Environ Microbiol 19:5060–5077. doi: 10.1111/1462-2920.15620. [DOI] [PubMed] [Google Scholar]
- 7.Gordillo Altamirano F, Forsyth JH, Patwa R, Kostoulias X, Trim M, Subedi D, Archer SK, Morris FC, Oliveira C, Kielty L, Korneev D, O'Bryan MK, Lithgow TJ, Peleg AY, Barr JJ. 2021. Bacteriophage-resistant Acinetobacter baumannii are resensitized to antimicrobials. Nat Microbiol 6:157–161. doi: 10.1038/s41564-020-00830-7. [DOI] [PubMed] [Google Scholar]
- 8.Knirel YA, Shneider MM, Popova AV, Kasimova AA, Senchenkova SN, Shashkov AS, Chizhov AO. 2020. Mechanisms of Acinetobacter baumannii capsular polysaccharide cleavage by phage depolymerases. Biochemistry (Mosc) 85:567–574. doi: 10.1134/S0006297920050053. [DOI] [PubMed] [Google Scholar]
- 9.Popova AV, Shneider MM, Arbatsky NP, Kasimova AA, Senchenkova SN, Shashkov AS, Dmitrenok AS, Chizhov AO, Mikhailova YV, Shagin DA, Sokolova OS, Timoshina OY, Kozlov RS, Miroshnikov KA, Knirel YA. 2021. Specific interaction of novel Friunavirus phages encoding tailspike depolymerases with corresponding Acinetobacter baumannii capsular types. J Virol 95:e01714-20. doi: 10.1128/JVI.01714-20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kenyon JJ, Hall RM. 2013. Variation in the complex carbohydrate biosynthesis loci of Acinetobacter baumannii genomes. PLoS One 8:e62160. doi: 10.1371/journal.pone.0062160. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Cahill SM, Hall RM, Kenyon JJ. 2022. An update to the database for Acinetobacter baumannii capsular polysaccharide locus typing extends the extensive and diverse repertoire of genes found at and outside the K locus. Microb Genom 8:e000878. doi: 10.1099/mgen.0.000878. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kenyon JJ, Hall RM. 2021. Updated analysis of the surface carbohydrate gene clusters in a diverse panel of Acinetobacter baumannii isolates. Antimicrob Agents Chemother 66:e01807-21. doi: 10.1128/AAC.01807-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wyres KL, Cahill SM, Holt KE, Hall RM, Kenyon JJ. 2020. Identification of Acinetobacter baumannii loci for capsular polysaccharide (KL) and lipooligosaccharide outer core (OCL) synthesis in genome assemblies using curated reference databases compatible with Kaptive. Microb Genom 6:e000339. doi: 10.1099/mgen.0.000339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Arbatsky NP, Kasimova AA, Shashkov AS, Shneider MM, Popova AV, Shagin DA, Shelenkov AA, Mikhailova YV, Yanushevich YG, Hall RM, Knirel YA, Kenyon JJ. 2022. Involvement of a phage-encoded Wzy protein in the polymerization of K127 units to form the capsular polysaccharide of Acinetobacter baumannii isolate 36–1454. Microbiol Spectr 27:e0150321. doi: 10.1128/spectrum.01503-21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Shashkov AS, Kenyon JJ, Senchenkova SN, Shneider MM, Popova AV, Arbatsky NP, Miroshnikov KA, Volozhantsev NV, Hall RM, Knirel YA. 2016. Acinetobacter baumannii K27 and K44 capsular polysaccharides have the same K unit but different structures due to the presence of distinct wzy genes in otherwise closely related K gene clusters. Glycobiology 26:501–508. doi: 10.1093/glycob/cwv168. [DOI] [PubMed] [Google Scholar]
- 16.Kenyon JJ, Marzaioli AM, Hall RM, De Castro C. 2014. Structure of the K2 capsule associated with the KL2 gene cluster of Acinetobacter baumannii. Glycobiology 24:554–563. doi: 10.1093/glycob/cwu024. [DOI] [PubMed] [Google Scholar]
- 17.Kasimova AA, Arbatsky NP, Timoshina OY, Shneider MM, Shashkov AS, Chizhov AO, Popova AV, Hall RM, Kenyon JJ, Knirel YA. 2021. The K26 capsular polysaccharide from Acinetobacter baumannii KZ-1098: structure and cleavage by a specific phage depolymerase. Int J Biol Macromol 191:182–191. doi: 10.1016/j.ijbiomac.2021.09.073. [DOI] [PubMed] [Google Scholar]
- 18.Arbatsky NP, Shneider MM, Dmitrenok AS, Popova AV, Shagin DA, Shelenkov AA, Mikhailova YV, Edelstein MV, Knirel YA. 2018. Structure and gene cluster of the K125 capsular polysaccharide from Acinetobacter baumannii MAR13-1452. Int J Biol Macromol 117:1195–1199. doi: 10.1016/j.ijbiomac.2018.06.029. [DOI] [PubMed] [Google Scholar]
- 19.Senchenkova SN, Kenyon JJ, Jia T, Popova AV, Shneider MM, Kasimova AA, Shashkov AS, Liu B, Hall RM, Knirel YA. 2019. The K90 capsular polysaccharide produced by Acinetobacter baumannii LUH5553 contains di-N-acetylpseudaminic acid and is structurally related to the K7 polysaccharide from A. baumannii LUH5533. Carbohydr Res 479:1–5. doi: 10.1016/j.carres.2019.04.008. [DOI] [PubMed] [Google Scholar]
- 20.Kenyon JJ, Kasimova AA, Shashkov AS, Hall RM, Knirel YA. 2018. Acinetobacter baumannii isolate BAL_212 from Vietnam produces the K57 capsular polysaccharide containing a rarely occurring amino sugar N-acetylviosamine. Microbiology (Reading) 164:217–220. doi: 10.1099/mic.0.000598. [DOI] [PubMed] [Google Scholar]
- 21.Kenyon JJ, Marzaioli AM, Hall RM, De Castro C. 2015. Structure of the K6 capsular polysaccharide from Acinetobacter baumannii isolate RBH4. Carbohydr Res 409:30–35. doi: 10.1016/j.carres.2015.03.016. [DOI] [PubMed] [Google Scholar]
- 22.Kenyon JJ, Arbatsky NP, Shneider MM, Popova AV, Dmitrenok AS, Kasimova AA, Shashkov AS, Hall RM, Knirel YA. 2019. The K46 and K5 capsular polysaccharides produced by Acinetobacter baumannii NIPH 329 and SDF have related structures and the side-chain non-ulosonic acids are 4-O-acetylated by phage-encoded O-acetyltransferases. PLoS One 14:e0218461. doi: 10.1371/journal.pone.0218461. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lees-Miller RG, Iwashkiw JA, Scott NE, Seper A, Vinogradov E, Schild S, Feldman MF. 2013. A common pathway for O-linked protein-glycosylation and synthesis of capsule in Acinetobacter baumannii. Mol Microbiol 89:816–830. doi: 10.1111/mmi.12300. [DOI] [PubMed] [Google Scholar]
- 24.Arbatsky NP, Shneider MM, Kenyon JJ, Shashkov AS, Popova AV, Miroshnikov KA, Volozhantsev NV, Knirel YA. 2015. Structure of the neutral capsular polysaccharide of Acinetobacter baumannii NIPH146 that carries the KL37 capsule gene cluster. Carbohydr Res 413:12–15. doi: 10.1016/j.carres.2015.05.003. [DOI] [PubMed] [Google Scholar]
- 25.Talyansky Y, Nielsen TB, Yan J, Carlino-Macdonald U, Di Venanzio G, Chakravorty S, Ulhaq A, Feldman MF, Russo TA, Vinogradov E, Luna B, Wright MS, Adams MD, Spellberg B. 2021. Capsule carbohydrate structure determines virulence in Acinetobacter baumannii. PLoS Pathog 17:e1009291. doi: 10.1371/journal.ppat.1009291. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yang F-L, Lou T-C, Kuo S-C, Wu W-L, Chern J, Lee Y-T, Chen S-T, Zou W, Lin N-T, Wu S-H. 2017. A medically relevant capsular polysaccharide in Acinetobacter baumannii is a potential vaccine candidate. Vaccine 35:1440–1447. doi: 10.1016/j.vaccine.2017.01.060. [DOI] [PubMed] [Google Scholar]
- 27.Hua M, Liu J, Du P, Liu X, Li M, Wang H, Chen C, Xu X, Jiang Y, Wang Y, Zeng H, Li A. 2021. The novel outer membrane protein from OprD/Occ family is associated with hypervirulence of carbapenem resistant Acinetobacter baumannii ST2/KL22. Virulence 12:1–11. doi: 10.1080/21505594.2020.1856560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Loraine J, Heinz E, Soontarach R, Blackwell GA, Stabler RA, Voravuthikunchai SP, Srimanote P, Kiratisin P, Thomson NR, Taylor PW. 2020. Genomic and phenotypic analyses of Acinetobacter baumannii isolates from three tertiary care hospitals in Thailand. Front Microbiol 11:548. doi: 10.3389/fmicb.2020.00548. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Hsieh Y-C, Wang S-H, Chen Y-Y, Lin T-L, Shie S-S, Huang C-T, Lee C-H, Chen Y-C, Quyen TLT, Pan Y-J. 2020. Association of capsular types with carbapenem resistance, disease severity, and mortality in Acinetobacter baumannii. Emerg Microbes Infect 9:2094–2104. doi: 10.1080/22221751.2020.1822757. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Fregolino E, Gargiulo V, Lanzetta R, Parrilli M, Holst O, De Castro C. 2011. Identification and structural determination of the capsular polysaccharides from two Acinetobacter baumannii clinical isolates, MG1 and SMAL. Carbohydr Res 346:973–977. doi: 10.1016/j.carres.2011.03.024. [DOI] [PubMed] [Google Scholar]
- 31.Jacobs AC, Thompson MG, Black CC, Kessler JL, Clark LP, McQueary CN, Gancz HY, Corey BW, Moon JK, Si Y, Owen MT, Hallock JD, Kwak YI, Summers A, Li CZ, Rasko DA, Penwell WF, Honnold CL, Wise MC, Waterman PE, Lesho EP, Stewart RL, Actis RA, Palys TJ, Craft DW, Zurawski DV. 2014. AB5075, a highly virulent isolate of Acinetobacter baumannii, as a model strain for the evaluation of pathogenesis and antimicrobial treatments. mBio 5:e01076-14. doi: 10.1128/mBio.01076-14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Arivett BA, Ream DC, Fiester SE, Kidane D, Actis LA. 2016. Draft genome sequences of Acinetobacter baumannii isolates from wounded military personnel. Genome Announc 4:e00773-16. doi: 10.1128/genomeA.00773-16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Oliveira H, Domingues R, Evans B, Sutton JM, Adriaenssens EM, Turner DT. 2022. Genomic diversity of bacteriophages infecting the genus Acinetobacter. Viruses 14:181. doi: 10.3390/v14020181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Popova AV, Lavysh DG, Klimuk EI, Edelstein MV, Bogun AG, Shneider MM, Goncharov AE, Leonov SV, Severinov KV. 2017. Novel Fri1-like viruses infecting Acinetobacter baumannii—vB_AbaP_AS11 and vB_AbaP_AS12—Characterization, comparative genomic analysis, and host-recognition strategy. Viruses 9:188. doi: 10.3390/v9070188. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Westphal O, Jann K. 1965. Bacterial lipopolysaccharides: extraction with phenol-water and further applications of the procedure, p 83–91. In Whistler R (ed), Methods in carbohydrate chemistry. Academic Press, New York, NY. [Google Scholar]
- 36.Pires DP, Oliveira H, Melo LDR, Sillankorva S, Azeredo J. 2016. Bacteriophage-encoded depolymerases: their diversity and biotechnological applications. Appl Microbiol Biotechnol 100:2141–2151. doi: 10.1007/s00253-015-7247-0. [DOI] [PubMed] [Google Scholar]
- 37.Liu M, Hernandez-Morales A, Clark J, Le T, Biswas B, Bishop-Lilly KA, Henry M, Quinones J, Voegtly LJ, Cer RZ, Hamilton T, Schooley RT, Salka S, Young R, Gill JJ. 2022. Comparative genomics of Acinetobacter baumannii and therapeutic bacteriophages from a patient undergoing phage therapy. Nat Comm 13:3776. doi: 10.1038/s41467-022-31455-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Shashkov AS, Cahill SM, Arbatsky NP, Westacott AC, Kasimova AA, Shneider MM, Popova AV, Shagin DA, Shelenkov AA, Mikhailova YV, Yanushevich YG, Edelstein MV, Kenyon JJ, Knirel YA. 2019. Acinetobacter baumannii K116 capsular polysaccharide structure is a hybrid of the K14 and revised K37 structures. Carbohydr Res 484:107774. doi: 10.1016/j.carres.2019.107774. [DOI] [PubMed] [Google Scholar]
- 39.Bankevich A, Nurk S, Antipov D, Gurevich A, Dvorkin M, Kulikov A, Lesin VM, Nikolenko SI, Pham S, Prjibelski AD, Pyshkin AV, Sirotkin AV, Vyahhi N, Tesler G, Alekseyev MA, Pevzner PA. 2012. SPAdes: a new genome assembly algorithm and its applications to single-cell sequencing. J Comput Biol 19:455–477. doi: 10.1089/cmb.2012.0021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Bortolaia V, Kaas RS, Ruppe E, Roberts MC, Schwarz S, Cattoir V, Philippon A, Allesoe RL, Rebelo AR, Florensa AF, Fagelhauer L, Chakraborty T, Neumann B, Werner G, Bender JK, Stingl K, Nguyen M, Coppens J, Xavier BB, Malhotra-Kumar S, Westh H, Pinholt M, Anjum MF, Duggett NA, Kempf I, Nykäsenoja S, Olkkola S, Wieczorek K, Amaro A, Clemente L, Mossong J, Losch S, Ragimbeau C, Lund O, Aarestrup FM. 2020. ResFinder 4.0 for predictions of phenotypes from genotypes. J Antimicrob Chemother 75:3491–3500. doi: 10.1093/jac/dkaa345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Lam MMC, Wick RR, Judd LM, Holt KE, Wyres KL. 2022. Kaptive 2.0: updated capsule and LPS locus typing for the Klebsiella pneumoniae species complex. Microb Genom 8:e000800. doi: 10.1099/mgen.0.000800. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Popova AV, Zhilenkov EL, Myakinina VP, Krasilnikova VM, Volozhantsev NV. 2012. Isolation and characterization of wide host range lytic bacteriophage AP22 infecting Acinetobacter baumannii. FEMS Microbiol Lett 332:40–46. doi: 10.1111/j.1574-6968.2012.02573.x. [DOI] [PubMed] [Google Scholar]
- 43.Lee I, Kim YO, Park S, Chun J. 2016. OrthoANI: an improved algorithm and software for calculating average nucleotide identity. Int J Syst Evol Microbiol 66:1100–1103. doi: 10.1099/ijsem.0.000760. [DOI] [PubMed] [Google Scholar]
- 44.Sullivan MJ, Petty NK, Beatson SA. 2011. Easyfig: a genome comparison visualizer. Bioinformatics 27:1009–1010. doi: 10.1093/bioinformatics/btr039. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Nishimura Y, Yoshida T, Kuronishi M, Uehara H, Ogata H, Goto S. 2017. ViPTree: the viral proteomic tree server. Bioinformatics 33:2379–2380. doi: 10.1093/bioinformatics/btx157. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental material. Download spectrum.03631-22-s0001.pdf, PDF file, 0.8 MB (826.8KB, pdf)
Data Availability Statement
Whole-genome sequence data from this study are deposited in NCBI under accession numbers MZ967493.1, JAPYKX000000000.1, JAPQKD000000000.1, JAPQKA000000000.1, JAPQKC000000000.1, and JANSJS000000000.1.


