Entomopathogenic nematodes are roundworms that, together with their bacterial symbionts, parasitize and kill insects, utilizing the bacteria as a food source (Mucci et al., 2022) . These nematodes are both economically and scientifically important since they can be used as biocontrol agents against different insect pests and are used as a model to better understand the molecular basis of animal-microbe symbioses (Koppenhöfer et al., 2020; Eleftherianos and Heryanto, 2020; Heppert et al., 2022). Over the past two decades, bacterial genes and pathways involved in both insect virulence and nematode mutualism have been identified, with the majority of efforts focusing on a few Xenorhabdus-Steinernema species pairs, particularly Xenorhabdus nematophila and Steinernema carpocapsae (Herbert and Goodrich-Blair, 2007; Hasan et al., 2019; Stock, 2019; Cao and Goodrich-Blair, 2020; Lefoulon et al., 2022) . However despite many efforts, reverse genetics and transgenesis in the nematode Steinernema carpocapsae have not succeeded. A recently isolated strain of Steinernema hermaphroditum is an attractive candidate for nematode genetic engineering because, unlike most Steinernematidae including S. carpocapsae which are dioecious, S. hermaphroditum is hermaphroditic with rare males (Bhat et al., 2019) . For this reason, S. hermaphroditum and their symbiotic bacterial partner Xenorhabdus griffiniae are being explored as a new genetic model system for studying symbiosis (Cao et al., 2022) . Towards this goal, we sought to adapt genetic tools to investigate the biology of X. griffiniae . Since Xenorhabdus strain variability can influence host interactions (Murfin et al., 2015a, Murfin et al., 2015b, Ciezki et al., 2015; Mollah et al., 2020) and genetic tractability (Cuív et al., 2015), we focused on two strains of X. griffiniae , HGB2507 ( S.h. isolate 5) and HGB2511 ( S.h. isolate 9), isolated concurrently from the same population of the S. hermaphroditum nematode partner (Cao et al., 2022) . Here we describe optimization of conjugation and transposon mutagenesis in these X. griffiniae strains.

To determine if X. griffiniae strains are amenable to conjugation and transposon mutagenesis, we used a conjugatable oriR6K suicide plasmid, pKV124, carrying a mini-Tn 10 transposon and a promoter-less lacZ gene (Visick and Skoufos, 2001) ( Fig. 1A ). In the standardized conjugation protocol for X. nematophila, ampicillin (or carbenicillin can be used interchangeably) is utilized for selecting Xenorhabdus bacteria over E. coli donor cells , because X. nematophila is naturally resistant to ampicillin (Gotz et al., 1981) . To apply this protocol, we first tested the carbenicillin resistance among ten isolates of X. griffiniae (Table 1). Our data show that X. griffiniae exhibited variability in their resistance against carbenicillin among different isolates, suggesting strain-to-strain variation (or phenotypic variation) could occur among these isolates. To facilitate the selection of X. griffiniae from E. coli donor cells without the use of negative antibiotic selection, we adapted the existing conjugation protocols available for X. nematophila to deliver this plasmid from diaminopimelic acid (DAP)-dependent E. coli pir -encoding donor cells into X. griffiniae recipient cells. After donor and recipient mating, X. griffiniae exconjugants were selected by virtue of their resistance to chloramphenicol and their ability to grow in the absence of DAP ( Fig. 1B ). Because at least some Xenorhabdus can efficiently kill E. coli (Muangpat et al., 2020; Hillman, 2020), we varied the ratio of E. coli donor to Xenorhabdus recipient in our conjugations. We tested three different recipient to donor ratios; 1:1, 1:3, and 1:10 and found that a 10-fold higher concentration of E. coli donor relative to X. griffiniae recipient, yielded the highest numbers of exconjugants ( Fig. 1D ). Exconjugants from a total of five plates from a single conjugation were picked and streaked multiple times to ensure clonal purity (see Methods). For this pilot optimization effort, a total of 192 individual exconjugants were picked from selection plates and arrayed in 2 x 96 well plates with LB medium and chloramphenicol, grown overnight at 30°C, and frozen at -80˚C with 20% glycerol for ease of phenotypic screening and subsequent recovery of mutants of interest.

The success of transposition was assessed by estimating frequencies of two phenotypes associated with genome disruption by the Tn 10 transposon: auxotrophy caused by disruption of biosynthetic loci and expression of lacZ caused by insertion downstream of active promoters. Each exconjugant was retrieved from the 96-well plates and streaked in a grid format onto either LB pyruvate (LBP) or Glucose Minimal Medium (GMM) with 5-Bromo-4-chloro-3-indolyl β-D-galactopyranoside (X-gal), a chromogenic substrate for β-galactosidase. Auxotrophs were identified as those that grew on LBP but not on GMM. Note that some carryover of nutrients can occur when streaking from LBP to GMM, which may allow some growth of auxotrophs. Mutants with lacZ insertions downstream of an active promoter were identified as those with blue color on either substrate ( Fig. 1C ).

All 192 exconjugants tested grew on LBP, but 9 (4.7%) failed to grow on GMM. This is consistent with frequencies of auxotrophic mutants observed in other bacteria such as Bacillus subtilis (Youngman et al., 1983) , although higher than the 1-2% observed for Salmonella enterica and Escherichia coli ( Fig. 1C ) (Kleckner et al., 1975; Nichols et al., 2011) . Based on these previous reports, we would expect ~2% (approximately 4/192) of insertions within genes to cause an auxotrophy. The higher than expected frequency of auxotrophs among the X. griffiniae HGB2511 exconjugants tested here may be due to the presence of sibling clonal colonies, derived from replication of mutants prior to plating. Sibling clonal colonies can be reduced by decreasing the amount of time that cells are permitted to grow after conjugation and before selection, and by using independent conjugations to build mutant libraries. It is also possible that if we allowed the exconjugants to grow longer on the GMM, some of the 9 mutants we identified would eventually grow.

During growth on X-gal, 12/192 (6.2%) and 83/192 (43%) exconjugants produced blue color indicative of β-galactosidase activity on LBP and GMM, respectively ( Fig. 1C ). This indicates that the Tn 10 can insert in an orientation that places the lacZ gene downstream of active promoters. The successful identification of exconjugants that displayed GMM-induced β-galactosidase activity (relative to LBP) demonstrates the feasibility of screening an X. griffiniae library for conditionally-active promoters using pKV124-mediated transposition. Additionally, the locations of two of the Tn 10 insertions were mapped using arbitrary PCR with pKV124-miniTn 10 transposon-specific primers (see Methods, Mapping). When comparing the Tn 10 flanking sequences to the Xenorhabdus BMMCB genome available on the Magnifying Genomes platform (MaGe, Vallenet et al., 2013) we found that one of the Tn 10 insertions was within the coding region of a gene with sequence similarity to type III restriction enzymes (LDNM01_v1_400040). The other insertion was in an intergenic region upstream of a gene with homology to a glycosyltransferase family 4 protein within a predicted LPS biosynthesis locus (LDNM01_v1_1170034). Each of the Tn 10 insertion candidates tested returned an unambiguous sequencing result, suggesting that they may be single insertions. However, this is not definitive, and there is still the possibility that we have multiple transposon insertions within the genomes of the library mutants we isolated. Future follow up experiments, such as Southern blotting or whole genome sequencing of representative mutants, will be necessary to ensure that in general we have one transposon insertion per isolate. These results establish an effective conjugation and mutagenesis protocol for X. griffiniae symbionts of S. hermaphroditum nematodes .

Strains and plasmids

X. griffiniae recipient strains were isolated from S. hermaphroditum (India) (Bhat et al., 2019) as described in Cao et al., 2022. Two isolates (HGB2507 and HGB2511) were tested for conjugation and transposition. The E. coli donor strain (HGB1333) used for this study is pKV124-miniTn 10 transformed into BW29427 (also known as WM3064) (Visick and Skoufos, 2001, and Dehio and Meyer, 1997) . BW29427 requires diaminopimelic acid (DAP) to grow in LB (Dehio and Meyer, 1997) and carries the pir gene, encoding the protein π necessary for replication of oriR6K plasmids, such as pKV124-miniTn 10 (Visick and Skoufos, 2001) . The DAP requirement facilitates counter-selection against the E. coli donor by removal of DAP from the medium.

Conjugation

Conjugation conditions were determined and optimized for 2 different isolates of X. griffiniae with E. coli carrying the pKV124-miniTn 10 . A more detailed protocol is provided below. For optimization, each of the two X. griffiniae isolates were mixed with the E. coli donor strain at different ratios of X. griffiniae to E. coli (1:1, 1:3, 1:10). The mixtures were then spotted on to LBP + DAP plates and grown for 3 days, after which they were plated on LBP + 15µg/ml chloramphenicol for selection. The colonies were confirmed to be X. griffiniae through their distinctive rust color compared to cream-colored E. coli. Individual exconjugants (192) of X. griffiniae HGB2511 were streaked for isolation on LB with 15µg/ml chloramphenicol and inoculated into a 96 well plate with dark LB and 15µg/ml chloramphenicol, and frozen at -80˚C.

Phenotype screening

LBP solid media were made by dissolving yeast extract (5g), tryptone (10g), NaCl (5g), pyruvate (1g), and agar (20g), in 1 Liter of ddH ₂ O and autoclaving prior to pouring plates (~25ml each). GMM solid media were made by combining a salts solution consisting of: nicotinic acid (100mg), KH ₂ PO ₄ (3g), K ₂ HPO ₄ (7g), (NH ₄ ) ₂ SO ₄ (2g), pH to 7.0 using NaOH and H ₂ O added to 500ml; and an agar solution consisting of: 15g Agar, 1g sodium pyruvate, and 450ml H ₂ O. After autoclaving (30min) and cooling until just warm to the touch, the salts and agar solution were combined a with 10ml of a 0.2µm filter-sterilized solution of SL4 salts, 9.1g glucose, 0.05g MgCl ₂ ᐧ(H ₂ O) ₆ , (this solution was mixed, heated to dissolve, and filter-sterilized prior to addition to the salt and agar solution mixture). The GMM plates were then poured at a volume of ~25ml per plate. Plates were top-treated with 50μl of a 10mg/ml X-gal stock solution, resulting in a final concentration of 40μg/ml. Bacteria were streaked onto plates in a grid pattern and were incubated for 2 days at 30˚C and then examined. By visualizing the plates, pigments that expressed any shade of blue, dark or light blue, were labeled as blue clones expressing lacZ. For the pigments with no blue shading, which included ones that were yellow, brown, and cream colored, they were labeled as non- lacZ expressing clones. Any streak that showed any growth was counted as a grown clone, while those who expressed no growth (not a single colony) were counted as clones that did not grow ( Fig. 1E ).

Transposon Insertion Mapping

For two exconjugate clones, the location of mini-Tn 10 transposon insertions in the X. griffiniae genome were mapped using arbitrarily-primed polymerase chain reaction (PCR) following the protocol from Saavedra et al., 2017. Briefly, the exconjugates were inoculated from glycerol stocks into 5ml dark LB + 15μg/ml chloramphenicol and grown at 30˚C overnight. Genomic DNA was extracted using the PureLink™ Genomic DNA Mini Kit (Invitrogen) per the manufacturer instructions. In the nested PCR reactions, 40ng of genomic DNA was used in the first round of PCR with primers pkv124_F_1: 5’-GCATCTGCCAGTTTGAGG; ARB6: 5’-GGCCACGCGTCGACTAGTACNNNNNNNNNNACGCC, an annealing temperature of 30˚C and extension time of 1 minute; and 20ng of DNA was used in the second round of PCR with primers (pkv124_F_2: 5’-GCGATTAAGTTGGGTAACGCCAG ; ARB2: 5’-GGCCACGCGTCGACTAGTAC) an annealing temperature of 60˚C and extension time of 1:20 minute. Ten microliters of the products from both rounds of PCR reaction were visualized on a 1.5% agarose gel and then the remaining product (40μl) was purified using the DNA Clean & Concentrator kit (Zymo Research). The products of the second reaction were then sequenced using Sanger Sequencing, and X. griffiniae genomic DNA sequences were identified by sequence similarity search using the BLASTN algorithm (Altschul et al., 1990) against the Xenorhabdus BMMCB genome LDNM01.1 on the MAGE Microscope platform (Vallenet et al., 2013) .

Conjugation Protocol:

Step 1. Streak donor and recipient strains onto plates from glycerol stocks stored at -80˚C and grow for 2 days at room temperature or 1 day at 30˚C. This step is necessary because strains, especially Xenorhabdus , do not always grow well in liquid media when inoculated straight from freezer stocks.

Step 2. Inoculate a heavy streak of bacteria (through a dense streak or multiple colonies) from the plates into 5ml dark LB in 2 tubes, one containing E. coli and one containing Xenorhabdus . Supplement E. coli medium with 30μg/ml chloramphenicol, 0.3mM DAP, and 150μg/ml streptomycin. Grow Xenorhabdus overnight at 30˚C and E. coli overnight at 37˚C.

Step 3. Subculture each strain separately into fresh dark LB media without antibiotics. Subculturing is typically done at 1:10 (500μl in 5ml), but can be done at a different dilution, e.g., 1:5 (1000μl in 5ml), to ensure timely growth of strains for use in Step 4. Supplement E. coli medium with 0.3mM DAP. Monitor growth using a spectrophotometer at Abs ₆₀₀ starting around 2 hours post inoculation.

Step 4. When an optical density (OD ₆₀₀ ) of ~0.8 is reached, mix donor and recipient together in an eppendorf tube at the desired ratio (e.g., 1:1, 1:3, 1:10) to a total volume of 1200μl. Spin the cell suspension in a benchtop microfuge at maximum speed at room temperature in order to pellet the cells. Resuspend in 25μl dark LB and spot the entire volume onto an LBP plate with 0.3mM DAP. (Do not add antibiotics at this step.) Incubate plates for 2 days at room temperature or overnight at 30˚C (If no colonies grow, allow plates to grow an additional day at 30˚C).

Step 5. Using a sterile stick or loop, scrape up the bacterial growth and resuspend well in 1ml dark LB in an eppendorf tube and mix by pipetting and vortexing until no clumps of bacteria remain. Dilute 1:10 by taking 100 μl of cells and transferring to an eppendorf tube with 900μl dark LB. Plate 100μl of both the original resuspension and the 1:10 dilution onto separate on LBP supplemented with antibiotic, in this case 15μg/ml chloramphenicol to select for Xenorhabdus with miniTn 10 transposon insertions, and spread across the surface of the plate using a sterile spreader or sterile glass beads. Leave out the DAP at this stage to select against growth of the E. coli donor.

Step 6. Incubate plates for 2 days at room temperature or overnight at 30˚C, and then select individual colonies for further validation.

Description: Table 1: Carbenicillin resistance test among X. griffiniae isolates. Resource Type: Image. DOI: 10.22002/whdjy-csp86