IbpAB small heat shock proteins are not host factors for bacteriophage φX174 replication

Bacteriophages exploit host proteins for successful infection. Small heat shock proteins are a universally conserved family of stress-induced molecular chaperones that prevent irreversible aggregation of proteins. Two small heat shock proteins, IbpA and IbpB, are a class of holding modulators or "holdases", which bind partially folded proteins and await ATP-driven foldin g chaperones for refolding. Bacteriophage φX174 is a small, icosahedral, and non-tailed virus belonging to the Microviridae . During φX174 infection of Escherichia coli C122, IbpA and IbpB were previously found to be the most highly upregulated host protein s, with expression levels comparable to φX174 proteins. In this work, to understand the role of IbpA and IbpB during φX174 infection, we used a hybrid approach of CRISPR interference and genomic knockouts to disrupt the ibpA and ibpB genes. We show that these two proteins do not appear to be necessary for efficient φX174 replication, and moreover, their absence has no effect on φX174 fecundity.


Introduction
Bacteriophages (phages) are viruses that obligately infect and replicate within a bacterial host.Phages are the most abundant entities within the biosphere, with an estimated global population of ~ 10 31 individual phage particles that outnumber their hosts by orders of magnitude (1)(2)(3).Phage-host interactions represent a vital ecological component of microbial systems with an estimated 10 23 viral infections occurring per second in the global oceans (4).Genomic knockout studies and genome-wide screens have revealed that successful phage infections rely on diverse bacterial host factors, including phage receptors and regulators, lipopolysaccharide (LPS) biosynthesis genes, and DNA replication and transcription machinery components (5)(6)(7)(8)(9).Among all identified host factors, the heat shock proteins (Hsp) are one of the most upregulated host factors during phage infections, at both transcription and translation levels (10)(11)(12).
Hsp are a class of molecular chaperones and are produced by bacteria under challenging conditions such as temperature, oxidative, or osmotic stresses.Their fundamental role is to facilitate the folding of unfolded and partially folded proteins through ATP-driven binding and releasing proteins in their functional states (13,14).Several Hsps have been deemed essential to the development of viable phage virions.Phage λ requires the oriλ-O-P-DnaB nucleoprotein structure to form an initiation complex for DNA replication; the Hsps DnaK, DnaJ, and GrpE (DnaKJE) release P protein from the λ initiation complex, and ClpX degrades O protein, freeing DnaB helicase to unwind the phage DNA (15)(16)(17).The Hsp DnaJ, independently facilitates the excision of prophage KplE1 of Escherichia coli K12 through interactions with the recombination directionality factor TorI (18).DnaJ has also been found to enhance the lysis of host cells by E. coli phage MS2 via the formation of a membrane-associated complex with lysis protein L (19).Additionally, the GroEL/GroES chaperonins assist with phage capsid morphogenesis by facilitating the correct folding of coat proteins (20)(21)(22).
Bacteriophage φX174 is a small, icosahedral, and non-tailed virus belonging to Microviridae (23) and work has shown that mutations in DnaJ/DnaK/GrpE/GroEL/GroES make the E. coli host more sensitive to lysis by E protein (24).In addition to this work, we have recently shown that the small heat shock proteins (sHsps) IbpA and IbpB were the most upregulated host proteins during φX174 infection, with levels matching or exceeding phage protein, which suggests they are important for efficient phage infection (10).
The sHsps are a widespread and diverse family of low-molecular weight Hsps with monomeric mass of 12-43 kDa.Structurally, sHsps share a highly conserved α-crystallin domain (ACD) (~ 90-100 amino acid residues) that is required for the formation of sHsp homodimer, which is the building block of higher oligomers, flanked by variable and disordered N-and C-terminal extensions (25,26).sHsps have been extensively studied for their role as "holdases" to prevent irreversible and uncontrolled aggregation of proteins, in an ATP-independent manner.Some sHsps also exhibit sequestrase activity that facilitates refolding of proteins in near-native state, and cytoprotective aggregase activity through the formation of deposits of misfolded proteins (25).sHsps are conserved across all domains of life.The dissemination of sHsps across all domains of life and the fact that genomes often contain multiple paralogs is indicative of their importance to cells, and has been discussed in detail (25,27).
In E. coli, IbpA and IbpB are organised under a single operon (ibpAB) within the σ 32 regulon and the independent regulation of ibpB within the σ 54 regulon.These proteins have been found specifically associated with inclusion bodies and intracellular aggregates of overexpressed proteins during both heat shock and recombinant protein expression (28)(29)(30).
Following heat-induced transcription of ibpAB operon, the bicistronic transcript is processed into monocistronic ibpA and ibpB transcript; the translation of ibpA and ibpB is controlled by an RNA thermometer in its 5' untranslated region, forming a secondary structure that blocks the binding of 30S ribosomal subunits at low temperature (31).It has also been found that oligomeric IbpA suppresses translation of its own mRNA transcript as well as that of ibpB (32).Collectively, ibpA and ibpB genes are essential when cells experience prolonged heat stress, > 50°C for 4 hours (33).IbpA and IbpB form oligomeric assemblies with partially folded proteins and deliver the substrates to the ATP-driven DnaKJE-ClpB bi-chaperone network, for refolding (34)(35)(36)(37)(38). IbpA and IbpB are known to act cooperatively, with IbpA stably interacting with aggregating substrates, preventing inter-and intra-molecular hydrophobic interactions by reducing the size of IbpAB-substrate assembly, while IbpB is required for the dissociation of the bound substrate from the assembly for further processing (35,36,39,40).Despite this cooperation and differentiation of function, the loss of either chaperone alone results only in a minimal disruption to cell growth (41), as they are thought to be functionally redundant (42), while DnaKJE-dependent refolding of denatured proteins would still be expected to occur in absence of either IbpA or IbpB, although at a considerably lower rate and efficacy (43,44).Because of this known functional redundancy and genome-wide screens showing individually disrupted ibpA and ibpB genes were non-essential for phage replication (e.g., T4, T7, and λ phages of E. coli) as well as not conferring phage resistance (5-7, 9, 45), we suspected their role in phage replication could be functionally redundant and warranted further investigation.
In our previous work, the IbpA and IbpB were the most highly upregulated of any host protein during bacteriophage φX174 infection of E. coli C122 (10).Thus, we reasoned that these host proteins could either be a defense mechanism deployed by the cell to attempt to sequester the phage proteins in an unfolded state and prevent capsid formation, or that these proteins were an important host factor that enabled large numbers of φX174 capsid proteins to be held in a nearly folded state to prevent the formation of inclusion bodies, and then assembled efficiently into the capsid when needed in the replication cycle.
In this work we determined the role of IbpA/B in the φX174 infection cycle using a combination of CRISPR gene knockouts and CRISPRi knockdowns, and demonstrate that they are collectively non-essential for φX174 replication.Our results suggest that IbpA and IbpB are neither a host phage defense factor nor an essential host factor required for viral replication.

IbpA/B are functionally redundant for φX174 replication but not for heat stress tolerance
To determine the effect of these proteins on bacteriophage φX174 replication we created single knockouts of ibpA and ibpB in the E. coli C122 strain using CRISPR-Cas9 (Table 1).We designed the gene knockouts which resulted in highly truncated IbpA (11% of wildtype) and IbpB (31%) proteins by incorporating pre-mature stop codons into the coding sequence (Fig. 1A).We confirmed these knockouts through whole genome sequencing.As expected from previous work (29,32,42,46), both C122ΔibpA and C122ΔibpB strains displayed no growth defects in rich liquid media at 37°C.However, growth defects at 45°C were apparent (Fig. 1B, C and D).Phages were added to exponential-phase cultures at an MOI of 5 and bacterial cultures were grown for a total of 24 h with OD600 measured every 5 min.Absorbance changes over time were compared to a non-infected control culture.The liquid assay score is the difference in the area under the growth curves of the phage-infected sample and the non-infected control (47).No inhibition of bacterial growth results in a score of zero, whereas complete of absence of growth gives a score of 100.Brackets with numbers above refer to Student's two-tailed t-test p-values.(F) RT-qPCR measurement of ibpA and ibpB transcript targets comparing the expression of the target genes when infected to the uninfected strains to determine the expression induction.To determine what effect the loss of one of the sHsps had on φX174 replication, we infected C122ΔibpA and C122ΔibpB strains along with wild-type C122 and measured changes in OD600 over time.The results showed that infections in the single-gene knockout strains were indistinguishable from wild-type E. coli at MOI 5 (Fig. 1E).A reduced MOI 0.001 infection also resulted in similar results between single-gene ibpA or ibpB knockouts and wild-type (Supplemental Fig. S1), but didn't induce ibpAB strongly, and thus MOI 5 was used going forward.Measuring the disruption effects of the ibpA and ibpB knockout gene sequences on transcript abundance of the ibpAB operon using RT-qPCR showed that the ibpA disruption appears to impair expression of both ibpA and ibpB transcripts when cells are infected with φX174, while the ibpB disruption does not (Fig. 1F).
We reasoned that even with the low transcriptional induction of the ibpA and ibpB transcripts in C122ΔibpA, there could still be functional IbpB protein produced at a level that could mask any beneficial or deleterious effects on φX174 replication (Fig. 1E).
We next attempted to create a double knockout of the ibpAB genes using CRISPR/Cas9 in a similar manner to the single-gene disruptions.We found our initial attempts to create this strain resulted in obtaining transformants with off-target effects that fell into two categories: (1) cryptic recombinants restoring the ibpB gene in screened colonies (Fig. 2A), or (2) apparently successful disruption of both genes but with off-target mutation elsewhere in the genome.

C122ΔibpAB
In a strain where we confirmed that both ibpA and ibpB genes were disrupted (Fig. 2B), we found that it displayed a severe a growth defect at 25°C, 30°C, 37°C, and 45°C (Fig. 2C).Whole genome sequencing of this strain revealed that an area of ~694.5 kb long sequence resulted in approximately twice the read coverage of adjacent areas of the C122ΔibpAB sequence and of the corresponding area of the C122 genome sequence (Fig. 2D).Furthermore, this putatively duplicated region was present outside of the ibpAB gene region target by the CRISPR guide RNAs (Fig. 2D) and contained 606 coding sequences, seven ncRNAs, tmRNA, three rRNAs, and 10 tRNAs, thus making the precise assessment of any epistatic effects unlikely.One coding sequence known to be involved in the same pathway as ibpAB that was present within this genomic region was the clpB gene, the product which in cooperation with DnaK uses energy derived from ATP hydrolysis to dissociate the denatured protein aggregates from sHsp-substrate complex (51).
Complementation of the C122ΔibpAB strain with plasmid-borne ibpAB genes failed to fully restore the growth phenotype (Fig. 2E) and thus we did not investigate this strain further.
Due to the difficulty of obtaining a clean double ibpA and ibpB knockout on the C122 strain genome, and all previously made double ibpAB knockouts were in E. coli strain backgrounds that are not susceptible to φX174 infection (52), we chose to deploy CRISPR interference (CRISPRi) in the single C122ΔibpA and C122ΔibpB knockout backgrounds to create the same effect with an inducible knockdown.CRISPRi employs a catalytically dead Cas9 (dCas9), which forms a complex with a single-guide RNA (sgRNA) (consisting of a gRNA and a structural scaffold) and is directed to the target gene to silence its expression (53,54).
We designed gRNAs that targeted the coding strand of ibpA and ibpB genes because they more efficiently repress transcription than noncoding strand targets (53,54).The target region for the sgRNAs was within the first 100 nucleotides of the CDS and was upstream of the sequence predicted to be the protein active site (55), such that translation of the incomplete mRNA transcript (if any) would result in a non-functional truncated protein.
Using the C122ΔibpA and C122ΔibpB strains as background, we introduced the pFR56::ibpB and pFR56::ibpA plasmids into these strains to create C122ΔibpA(pFR56::ibpB) and C122ΔibpB(pFR56::ibpA) strains where CRISPRi-mediated transcriptional inhibition could be induced within the non-disrupted ibpA or ibpB coding sequence, thus effectively silencing both IbpA and IbpB protein production from both genes.We also introduced into both C122ΔibpA and C122ΔibpB strains a plasmid (pFR56::ns) expressing a control gRNA which does not have a specific target in the E. coli genome and is a control for the effect of expression of an sgRNA in the presence of dCas9 itself.Based on previous work, the nonspecific sgRNA is not expected to have a phenotypic effect (7).Lastly, we transformed pFR56::ibpA, pFR56::ibpB, and pFR56::ns into wild-type C122 strain as a control.
We first confirmed the efficiency of the CRISPRi by RT-qPCR of φX174 infected cells.
The expression of dCas9 was induced by DAPG (50 µM) at early exponential phase and we allowed 40 min for dCas9 to be produced before phage infection.We did not expect ibpA or ibpB transcripts to be expressed prior to φX174 infection because they are within a stress response regulon (10, 12, 56, 57) and we did not subject the cells to temperature-shock.At 60 min post-infection, we found upregulation of both ibpA and ibpB transcripts in the absence of dCas9 expression, whereas the amount of ibpA transcripts were less than 2-fold above non-infected samples, and ibpB transcripts were unchanged when dCas9 was expressed (Fig. 3B).Together, these observations suggest that the gRNAs targeting the ibpA and ibpB transcripts were able to direct dCas9 binding and prevent the majority of transcription from these loci.In particular, the C122ΔibpA(pFR56::ibpB) strain should represent a true knockout of functional IbpA and IbpB proteins, since the ibpB transcript is fully blocked by CRISPRi, and the slightly induced ibpA transcript contains a disrupted coding sequencing, presumably resulting in a non-functional IbpA protein.
We next measured φX174 lysis efficiency in the absence of the IbpAB proteins by comparing liquid lysis scores of each strain undergoing ibpA or ibpB CRISPRi with the same strain expressing a nonspecific sgRNA.We found no difference between φX174 replication efficiencies in strains with wild-type ibpAB expression and those with a combination of genomic disruption and CRISPRi knockdown of the ibpAB genes (Fig. 3C).Using a lower MOI of 0.001 also showed no significant effect of ibpAB disruption on φX174 replication (Supplemental Fig. S2).
Because some infection deficiencies can be obscured in bulk culture measurements (58), we next wanted to measure the efficiency of each infection event through the efficiency of centre of infection assay.We first grew E. coli strains, induced CRISPRi, and infected cells as before, except in this case we washed cells following infection to remove unadsorbed phage, followed by co-plating with wild-type C122 as an indicator strain.Thus, any plaques observed from this method show the result of an initial successful infection event, but not the cumulative effect of multiple rounds of infection within cells undergoing CRISPRi.Similar to the liquid lysis assay (Fig. 3C), we did not observe significant differences in φX174 infection efficiency when infecting ibpAB knockdown strains (Fig. 3D).
CRISPRi-mediated ibpAB knockdown strains were grown in phage-LB at 37°C/240 rpm to early exponential phase when dCas9 was induced by DAPG (50 µM).At 40 min after induction, the bacteria were infected with wild type φX174 at an MOI of 5. Cells were harvested at 60 min postinfection, and RNA was extracted.RT-qPCR against ibpA and ibpB targets was performed in biological and technical triplicate.(C) φX174 liquid culture virulence against CRISPRi-mediated ibpAB knockdown strains.The ibpAB sgRNAs are named for their target and ns (non-specific) being an sgRNA that lacks a specific target.(D) φX174 efficiency of centre of infection against CRISPRimediated knockdown.Student's two-tailed t-test p values reported.

Discussion
In this work, we examined the role of two sHsps, IbpA and IbpB, during φX174 infection of E. coli C122, using a combination of single-gene genomic knockout and CRISPRi-mediated transcription inhibition.We found that the virulence and plaque-forming ability of φX174 was unaffected by the loss of the ibpA and ibpB genes, and thus they are not host factors for this virus.This result is unexpected given the two sHsps have been found amongst the most enriched host proteins during phage infections (10,12,56,57), and leaves us with the question of why they are highly expressed during φX174 infection?
The IbpA and IbpB proteins seem to be important for stabilising denatured protein aggregates.They have been found massively upregulated when there is intracellular protein aggregation as a result of stresses, including elevated temperature, recombinant protein overexpression, and oxidative damage (32,33,(59)(60)(61).The sHsps are widespread and diverse family of molecular chaperones that are found across all domains of life.Although they are not in the bacterial minimal gene set (62) nor in the E. coli essential gene set (7), mutations in genes encoding sHsps have been linked to several human diseases, including Charcot-Marie-tooth disease, cataracts, cardiomyopathy, and motor neuropathy (63)(64)(65).
Individual knockout studies have shown the ibpA and ibpB genes to be non-essential in E. coli K-12 strains.Independent single deletions of ibpA or ibpB using TraDIS (41) and CRISPR interference pools (6, 66) also show these genes are individually nonessential.
In contrast to the redundant function of ibpA and ibpB in K-12 strains, we see in this work that disruption of either ibpA or ibpB causes severe growth defects at 45°C in the E. coli C122 strain (Fig. 1B-D).Although not an identical stress to heat shock, we have previously shown that dnaK and clpB genes are upregulated in C122 undergoing φX174 infection, with dnaK upregulated 2.5-fold and clpB upregulated 2.2-fold (10).Our inability to completely disrupt both ibpA and ibpB simultaneously in C122 without compensatory mutations (Fig. 2) may suggest that the ibpAB genes display synthetic sickness or lethality characteristics in this strain that are not present in K-12 strains.Other work by Rousset et al. (7) fits with this hypothesis, as they have shown that there are significant differences in gene essentiality between E. coli strains.
Another possibility is that cryptic recombination may have occurred in some previous reports and this was not detected because confirmation of mutants was done by PCR product sizing or similar low-resolution methods.With the advent of whole-genome sequencing, previously invisible confounding events such as the sequence duplication we observed (Fig. 2) may partially explain the differences we see to past work.
Despite the sHsps being highly upregulated during many different phage infections (10,12,56,57) we have shown in this work that they do not appear to play a role in the Microviridae phage φX174 replication cycle.Previously, Young, et al (1989) found that lysis sensitivity increased in E. coli when plasmid encoded φX174 E protein and heat-shock genes dnaK, dnaJ, groEL, and grpE were present (24).Additionally, other folding chaperones were shown to be important for successful phage replication in phage PRD1, HK97, and T4, which all rely on the GroES/EL chaperones (20,21,67).In other phage infections, the IbpA/B genes were upregulated to some of the highest levels of any transcript or protein in P. aeruginosa and E. coli during PRR1 phage infection (57), along with other σ 32 regulated genes ( 56), but they did not investigate the essentiality of these two sHSPs for phage infection.
Although more typically associated with cellular stress due to protein overexpression and increased temperatures (25), the sHsps appear to have another major role in maintaining membrane stability, which has been demonstrated in both prokaryotes and eukaryotes.In cyanobacterium Synechocystis sp.PCC 6803, HspA (or Hsp17) was found specifically associated with the thylakoid membrane under heat stress and functioned to reduce fluidity back to more normal levels (68)(69)(70).Under salt stress, increased expression of GroEL/GroES was observed in a Synechocystis mutant that lacked hspA; GroEL/GroES are also known to assist the folding of membrane-associated proteins and stabilize the lipid membrane structure (71,72).Similarly, the Lo18 sHsp from the lactic acid bacterium Oenococcus oeni was found associated with the cell membrane because of increased membrane disorder, rather than cellular protein aggregation (73).In vitro studies have also shown that Lo18 interacts with liposomes and increases the molecular order of the lipid bilayer (74,75).In Lactobacillus plantarum, deletion of hsp 18.55 caused increased membrane fluidity when exposed to ethanol shock, whereas L. plantarum overexpressing Hsp18.55 had significant reduction in membrane fluidity compared to the wild-type strain (76).Functional characterisation of Hsp20 from the thermoacidophilic crenarchaeon Sulfolobus acidocaldarius has revealed that sHsps stabilise archaeal and bacterial membrane possibly through hydrophobic interactions with membrane lipids (archaeosomes and liposomes) (77).Furthermore, HSP17 (a sHsp from Caenorhabditis elegans), when heterologously expressed in E. coli, was found to be partially localised to periplasmic space and associated with inner membrane to maintain cell envelope integrity when E. coli were grown at lethal temperature (50°C) (78).The IbpA and IbpB proteins were previously found associated with the outer membrane in E. coli (30,33) and are known to form a complex with lipoprotein NlpI and localise to the outer membrane during cell division (59).
One possible explanation for our results that show ibpAB is dispensable for φX174 replication, is that IbpA/B may be upregulated in response to both the large increase in protein production due to phage infection as well as changes to the membrane integrity during infection.A potentially similar response is seen with the Phage Shock Protein (Psp) complex which repairs cell envelope integrity, and is upregulated in both φX174 and PRD1 infections of E. coli (10,56,79,80).Phage φX174 expresses a single lysis protein E during infection and its mechanism of action is thought to be through inhibiting a key enzyme (MraY) in cell wall peptidoglycan synthesis (81).We speculate that the progressive disruption to peptidoglycan structure and cell envelope integrity results in a signal that causes the upregulation of the ibpAB genes.The produced proteins may provide some transient protection to cell membrane integrity prior to lysis, but they are ultimately overwhelmed by lysis protein production and burst of phage progeny.
For routine φX174 phage amplification, a single plaque was added to mid-exponential phase culture of bacteria in phage-LB (LB supplemented with 2 mM CaCl2), and phage were purified as previously described (82).Phage titers were measured by the double-layer overlay method (83).Growth media, buffers, and reagents were sterilised by autoclaving and/or filtering through 0.2 µm membrane prior to use in the experiments.
The bacterial strains, phage, and plasmids used in this study are presented in Table 1.Synthetic DNA was acquired from Integrated DNA Technologies (Coralville, IA, USA), primers from Azenta Life Sciences (South Plainfield, NJ, USA), and they are described here (Supplemental Table S1).

CRISPR/Cas9 mediated gene knockouts
Custom sgRNA plasmids, knockouts, and plasmid curing were performed as described (84,85), except for transformation.Transformations were performed with chemically competent cells using the Mix and Go! E. coli Transformation and Buffer Set (Zymo Research, USA).Primers, sgRNA sequences, and donor DNA are described here (Supplemental Table S1).Confirmation of incorporated targeting sgRNA sequences and knockouts were performed through PCR amplification of targeted regions, followed by PCR product purification using the GenElute TM PCR Clean-Up Kit (Sigma Aldrich, U.S.A).All sequencing throughout was performed using the Macrogen sequencing service (Macrogen, South Korea).
Whole genome sequencing of C122, C122ΔibpA, C122ΔibpB, and C122ΔibpAB strains was performed by preparing by Illumina sequencing libraries from gDNA samples using Nextera DNA Flex Library Preparation Kit.Libraries were sequenced on an Illumina MiSeq instrument (Sequencing @ UTS).Read processing and mapping was performed using Geneious Prime software.

Assembly of plasmid for trans expression of ibpAB genes
The native ibpAB operon (NZ_LT906474 coordinates: 16,177-17,263) was assembly into linearized pULTRA-CNF plasmid using NEBuilder Assembly (New England Biolabs), resulting in pULTRA::ibpAB.A version of pULTRA-CNF lacking the inserted gene (pULTRA::empty) was also created (Table 1).Sequence-confirmed plasmids were transformation into C122ΔIbpAB and wild-type C122 using the Mix and Go! E. coli Transformation and Buffer Set (Zymo Research, U.S.A).Transformants were selected and grown in phage-LB under a selective marker spectinomycin (50 μg/mL).

CRISPRi plasmid construction
The dCas9-sgRNA plasmid expression system used for CRISPRi, pFR56, harbors a constitutively expressed sgRNA and dcas9 under the control of a DAPG-inducible PhlF promoter (7).We predicted all possible gRNAs targeting ibpA or ibpB using the CRISPR Guide RNA Design tool (https://crispr-browser.pasteur.cloud/guide-rna-design)(86).The gRNA sequence with the highest score on the non-template (coding) strand was selected, because binding of gRNA to non-template strand has been demonstrated more efficient in repressing transcription than binding to the template strand (6,53).The selected gRNAs targeting ibpA or ibpB were cloned on pFR56 using a pair of divergent primers (87) (Supplemental Table S1), and the assembled gRNA plasmids were transformed into chemically competent NEB Turbo E. coli cells (New England Biolabs, C2984H).Transformants were selected on LB-Cm agar plates and PCR amplification of targeted regions was used to confirm correct plasmid assembly, followed by Sanger sequencing (Macrogen, South Korea).
Sequence (Agilent, USA).The assay was performed in biological triplicates.The OD600 values as a function of time were plotted to generate the growth curve.The growth rate and the maximum population size at stationary phase were calculated using Growthcurver R package (88).

In vitro bacterial killing assay
Bacterial growth and OD600 monitoring were described as above.When OD600 reached ~0.15, DAPG was added to the cultures to a final concentration of 50 µM to induce the expression of dCas9.Bacterial strains were allowed to grow for a further 40 minutes before infection with φX174 phage at MOI=5 and 0.001.In the control experiment, 1 µL carrier (70% (v/v) ethanol) was added as mock induction, and 2 µL phage-LB was added as mock phage infection.The assay was performed in biological triplicates.The virulence of φX174 against E. coli C122 strains was assessed by the liquid assay score as previously described, which is equal to the area between the control curve and the phage treatment curve; a score of zero indicate no inhibition of bacterial growth, whereas complete absence of bacterial growth of phage-infected culture results in a score of 100 (47).The area under each curve was calculated using Growthcurver R package (88).

Efficiency of centre of infection (ECOI) assay
Centre of infection assays were conducted as previously described (89).Overnight bacterial cultures were diluted 100-fold in 10 mL of phage-LB or phage-LB-Cm, and were grown at 37°C/240 rpm.The OD600 was monitored in a spectrophotometer (UV-1280, Shimadzu, Japan).DAPG was added to a final concentration of 50 µM when OD600 reached ~0.3, and bacteria continued to grow for a further 40 min.Subsequently, 2 mL of bacterial culture was centrifuged and resuspended in 500 µL of phage-LB.To achieve an MOI of 0.001 or less, 60 pfu phage lysates were added.Phage were allowed to adsorb for 5 min at 37°C/900 rpm in a thermomixer (Eppendorf).The cells were placed on ice for 5 min and then washed twice with 500 µL of phage-LB to remove free phage.The cells with adsorbed phage were co-plated with 200 µL of wild type E. coli C122 overnight culture following the standard soft agar overlay method (83), and incubated overnight at 37°C.Resulting plaques were enumerated and the efficiency of centre of infection (ECOI) was calculated.The experiment was performed in four biological replicates.

Statistical analysis
All statistical analyses were conducted with R program language (version 4.2.1)(https://cran.r-project.org/bin/macosx/big-sur-arm64/base/R-4.2.1-arm64.pkg).Student's two-tailed t-test were performed to assess the differences in bacterial growth rate, liquid assay scores, ECOI of φX174, and fold changes of ibpA/B expression levels.

FIG 1
FIG 1 Single ibpA and ibpB gene disruptions result in growth inhibition at elevated temperatures but no difference in φX174 replication efficiency.(A) ibpAB gene knockout scheme.Wild-type ibpA is 355 bp long while knockout introduces premature stop codon resulting in coding sequence of only 39 bp (11%).Wild-type ibpB is 233 bp long while knockout introduces premature stop codon resulting in coding sequence of only 77 bp (31%).Grey bars indicate area of each gene removed after repair and faded gene symbol indicates wild-type extent of gene.(B-D) Growth rate comparison between 37°C and 45°C in ibpA and ibpB knockout strains.(E) Comparison of φX174 virulence against wild type C122, C122ΔibpA, and C122ΔibpB bacterial strains.The genetic background of the E. coli host denoted by + (wild-type gene present) or -(gene disrupted).Phages were added to exponential-phase cultures at an MOI of 5 and bacterial cultures were grown for a total of 24 h with OD600 measured every 5 min.Absorbance changes over time were compared to a non-infected control culture.The liquid assay score is the difference in the area under the growth curves of the phage-infected sample and the non-infected control(47).No inhibition of bacterial growth results in a score of zero, whereas complete of absence of growth gives a score of 100.Brackets with numbers above refer to Student's two-tailed t-test p-values.(F) RT-qPCR measurement of ibpA and ibpB transcript targets comparing the expression of the target genes when infected to the uninfected strains to determine the expression induction.

FIG 2
FIG 2 Difficulty obtaining a double ibpAB knockout without off-target effects.(A) Cryptic recombination between ibpA 5'-end and ibpB gene 3'-remnants creates ibpB gene with identical amino acid sequence to IbpB protein.(B) Apparent double ibpAB disruption where a premature stop codons were introduced into ibpA and ibpB genes resulting in coding sequences of only 39 bp (11%) and 77 bp (31%) of the wild-type genes, respectively.Grey bars indicate area of each gene removed after repair and faded gene symbol indicates wild-type extent of gene.(C) Apparent doubleknockout ibpAB strain growth characteristics at 25° -45°C compared to isogenic wild-type C122 strain.(D) Deep sequencing coverage of C122ΔIbpAB strain (blue) compared to wild-type C122 (red) reveals putative 694.5 kb sequence duplication outside ibpAB region cut by CRISPR/Cas9.Black bar indicates region of increased read coverage with approximate genomic coordinates of boundaries.(E) Complementation of C122ΔIbpAB strain with plasmid-borne ibpAB genes shows only partial restoration of growth defect at 30°C and 37°C.Plasmid pULTRA::ibpAB denotes pULTRA harboring native ibpA/B operon while pULTRA::empty denotes empty pULTRA plasmid backbone.Variance of one standard deviation (n = 3) in measurements shown as faded region of same colour around growth curves.
F D L S P L M R Q W I G IbpB M R N F D L S P L M R Q W I G Biological triplicates of E. coli C122 were grown in 1.45 mL of phage-LB or phage-LB-Cm in a 96-well deep-well plate at 37°C in an Infors MT multitron pro shaker at 240 rpm rotating orbitally at a 25 mm diameter.Expression of dCas9 was induced at early exponential phase by DAPG to a final concentration of 50 µM.The bacteria were further grown for 40 min before infection with wild type φX174 at an MOI of 5.At 60 min after infection, samples were pelleted by centrifugation at 5,000 RCF for 10 min at 4°C and resuspended in 250 µL ice-cold 1x phosphate-buffered saline (PBS) at pH 7.4, RNAprotect bacterial reagent (Qiagen, 76506) was added and the cells treated according the manufacturer instructions.Following protection, RNA was isolated using the RNeasy Mini Kit (Qiagen, 74106) with the optional DNase step according to the manufacturer instructions.The quality and quantity of RNA were determined using a NanoDrop spectrophotometer (NanoDrop One, Thermo Fisher Scientific, Waltham, MA, USA).All RNAs were reverse transcribed into cDNA using the High Capacity cDNA Reverse Transcription Kit (ThermoFisher Scientific, 4368814).RT-qPCR was performed with the LightCycler 480 SYBR Green I Master (Roche, 04707516001) in a LightCycler 480 II (Roche, USA) following manufacturer's instructions in 10 µL volumes with an annealing temperature of 64°C.RT-qPCR was performed in technical triplicate and biological triplicate.Relative gene expression was computed using the comparative CT method (90) after normalization to the cysG housekeeping gene with the infected samples as the treated samples and the uninfected the untreated.

Table 1 Description of Plasmids, Strains, and Phage Used in this Study
(50) strain harbouring genomic ibpB knockout and pFR56 plasmid targeting ibpA This study Bacteriophage Bacteriophage φX174 Phage of E. coli C synthesized to be identical to Sanger GenBank sequence NC_001422.1(50) pFR56::ns Plasmid used for cloning gRNA.Contains chloramphenicol resistance marker, a control non-specific single gRNA (sgRNA) under a constitutive promoter, and dcas9 under the control of a 2,4-diacetylphloroglucinol (DAPG)-inducible PhlF promoter GenBank accession no.MT412099 (A gift from David Bikard Lab) pFR56::ibpA gRNA plasmid targeting ibpA gene.Contains chloramphenicol resistance marker, sgRNA under a constitutive promoter, and dcas9 under the control of a DAPG-inducible PhlF promoter This study pFR56::ibpB gRNA plasmid targeting ibpB gene.Contains chloramphenicol resistance marker, sgRNA under a constitutive promoter, and dcas9 under the control of a DAPG-inducible PhlF promoter This study Strain Escherichia coli C122 Wild-type host strain of bacteriophage ΦX174.
-verified plasmids were transformed into chemically competent C122, C122ΔibpA, and C122ΔibpB cells prepared by the Mix and Go! E. coli Transformation and Buffer Set (Zymo Research, USA).