High-resolution Microbiome Analysis Reveals Exclusionary Klebsiella Species Competition in Preterm Infants at Risk for Necrotizing Enterocolitis

doi:10.21203/rs.3.rs-2468771/v1

Download PDF

Article

High-resolution Microbiome Analysis Reveals Exclusionary Klebsiella Species Competition in Preterm Infants at Risk for Necrotizing Enterocolitis

https://doi.org/10.21203/rs.3.rs-2468771/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 16 May, 2023

Read the published version in Scientific Reports →

You are reading this latest preprint version

Intestinal colonization with Klebsiella has been linked to necrotizing enterocolitis (NEC), but methods of analysis usually failed to discriminate Klebsiella species or strains. A novel ~2,500-base amplicon (StrainID) that spans the 16S and 23S rRNA genes was used to generate amplicon sequence variant (ASV) fingerprints for Klebsiella oxytoca and Klebsiella pneumoniae species complexes (KoSC and KpSC, respectively) and co-occurring fecal bacterial strains from 10 preterm infants with NEC and 20 matched controls. Complementary approaches were used to identify cytotoxin-producing isolates of KoSC. Klebsiellaspecies colonized most preterm infants, were more prevalent in NEC subjects versus controls, and replaced Escherichiain NEC subjects. Single KoSC or KpSC ASV fingerprinted strains dominated the gut microbiota, suggesting exclusionary Klebsiella competition for luminal resources. Enterococcus faecaliswas co-dominant with KoSC but present infrequently with KpSC. Cytotoxin-producing KoSC members were identified in most NEC subjects and were less frequent in controls. Few Klebsiellastrains were shared between subjects. We conclude that inter-species Klebsiella competition, within an environment of KoSC and E. faecalis cooperation, appears to be an important factor for the development of NEC. Preterm infants seem to acquire Klebsiella primarily through routes other than patient-to-patient transmission.

Biological sciences/Microbiology/Bacteria

Biological sciences/Microbiology/Clinical microbiology

Biological sciences/Microbiology/Communities

Health sciences/Medical research/Paediatric research

Health sciences/Medical research/Translational research

Health sciences/Gastroenterology/Gastrointestinal diseases

Health sciences/Pathogenesis/Infection

Necrotizing enterocolitis (NEC) is the most common and lethal gastrointestinal emergency of premature infants¹. The premature intestine is uniquely hyper-responsive to pathogen-associated molecular patterns (PAMPS)², which likely explains the destructive inflammatory response observed in this disorder. Along these lines, it is generally accepted that NEC pathogenesis involves activation of Toll-like receptor 4 (TLR4) signaling cascades by lipopolysaccharides (LPS) from the outer membranes of gut colonizing Gram-negative bacteria^3,4. Whether colonization with particular Gram-negative bacteria is a feature of the dysbiotic microbial profiles that incite this runaway pro-inflammatory response remains an open question.

Gram-negative bacteria belonging to the genus Klebsiella have been linked to NEC in several studies, but methods of analysis often failed to discriminate Klebsiella species^5–7. Most NEC microbiome studies have used 16S rRNA sequencing, which surveys complex microbial communities but typically lacks the depth to classify bacteria at the species or even genus levels. Members of the genus Klebsiella are highly diverse and include the Klebsiella oxytoca and Klebsiella pneumoniae species complexes (KoSC and KpSC, respectively) along with several more genetically distant species^8,9. We previously employed a combination of 16S rRNA sequencing, selective culture systems, polymerase chain reaction and mass spectrometry to identify cytotoxin-producing K. oxytoca in the fecal microbiota of premature infants with NEC¹⁰. Toxigenic strains of KoSC produce the enterotoxins tilivalline and tilimycin and are the causative agent of antibiotic-associated hemorrhagic colitis (AAHC) in older children and adults^11–13. Alternate approaches, such as genome-resolved metagenomics, have similarly identified Klebsiella as dominating the fecal microbiome of NEC subjects, while suggesting that strains of K. pneumoniae were more often associated with the disease¹⁴. Improved resolution of intestinal microbiota would help to clarify the potential role of different Klebsiella species and strains in the preterm gut.

Advances in sequencing platforms have greatly improved the accuracy of long-read, high-throughput approaches for surveying microbial communities¹⁵. We recently described a novel ~ 2,500 base rRNA (StrainID) amplicon spanning the bacterial 16S and 23S rRNA genes that maps to a new, custom 16S-23S rRNA database to achieve species and strain level resolution of the gut microbiome in preterm infants¹⁶. The StrainID amplicon contains sufficient sequence variability such that a single Klebsiella strain can produce up to 8 different amplicon sequence variants (ASVs) which, when combined, create a unique fingerprint. Differentiation of closely related Klebsiella, E. coli and Enterobacter with this method enabled ASV fingerprinting to track specific strains in the hospital environment¹⁶. In the present study, we used this novel deep-sequencing approach to obtain evidence that colonization of the premature infant gut with either KpSC or KoSC strains appears to be a common precursor to the onset of NEC.

Matched pairs used for analysis. During the study period, fecal samples from 10 preterm infants who developed NEC were available for analysis; samples from 20 matched controls were also analyzed. There were no significant differences in clinical characteristics between the two groups (Table 1). Adequate DNAs were extracted from 143 (44 NEC and 99 control) time points.

Table 1

Baseline characteristics of the subjects.
Characteristic	NEC (N = 10)	Control (N = 20)	P value
Birthweight (grams)	817 (778–1125)	820 (720–1092)	0.681
Gestational age at birth (weeks)	25.6 (24.7–27.8)	25.6 (24.9–28.5)	0.940
Sex			0.333
Girls	1 (10)	5 (25)
Boys	9 (90)	15 (75)
Race			0.310
Black	3 (30)	8 (40)
White	3 (30)	9 (45)
Other	4 (40)	3 (15)
Apgar scores
1 min	5 (1.5–6.8)	3.5 (2.0-5.5)	0.701
5 min	7 (7.0–8.0)	8 (6.0–8.0)	0.813
Number of singleton births	10 (100)	18 (90)	N/A
Number of births by caesarean delivery	7 (70)	14 (70)	1.000
Age of necrotizing enterocolitis (days)	27 (19.3–34.0)	N/A	N/A
Percentage of days on human milk	70 (64.0–95.0)	91 (73.8–97.0)	0.737
Days of antibiotics prior to NEC or sample	4 (2.3–7.5)	5 (3.0–8.0)	0.558
Age first stool analyzed (days)	12.5 (8.3–15.5)	10 (7.0-14.3)	0.610
Number of stools analyzed	5 (3.0-5.8)	5 (3.0–6.0)	0.621
Data are median (interquartile range) or n (%); Abbreviations: NEC, necrotizing enterocolitis; N/A, not applicable.

Klebsiella species colonized most preterm infants, were more prevalent in NEC subjects versus controls, and replaced Escherichia in NEC subjects. StrainID amplicon sequencing returned a total of 3,462,451 reads with a mean of 24,213 ± 1443 reads per sample. Klebsiella was the predominant genus identified for both cases and controls, accounting for a combined 30.6% of total reads from the two groups (Fig. 1A). In individual samples, the mean number of Klebsiella reads was 10,181 ± 1777 and 6,100 ± 1007, respectively, in the NEC and control groups (Fig. 1B; P = 0.03). When normalizing to total reads, Klebsiella accounted for 36.7 ± 5.6% of reads in the NEC group and 27.7 ± 3.7% of reads in the control group (Fig. 1C; P = 0.055). For individual subjects, Klebsiella was present at a relative abundance of > 7% in 8 of 10 NEC cases and 10 of 20 controls (P = 0.11), and at a relative abundance of > 2% in 9 of 10 NEC cases and 12 of 20 controls (P = 0.09). Additionally, the relative abundance of Escherichia was 9.5 ± 2.4% in the control group and 0.04 ± 0.02% in the NEC group (Fig. 1D; P = 0.005). There were no significant differences for other major taxa including Enterococcus, Enterobacter, Clostridia, Veillonella, Staphylococcus and Streptococcus between the groups (Fig. 1A).

Differentiation of Klebsiella species revealed exclusionary competition between members of KoSC and KpSC. We next sought to use StrainID fingerprinting to differentiate Klebsiella species in our samples. All subjects that harbored Klebsiella at a relative abundance of > 2% (9 NEC and 12 controls) were used in the analysis. By plotting the composite relative abundance of reads assigned to Klebsiella species for each subject, we observed that 4 of 9 NEC cases were heavily dominated by KoSC, while the remaining five were heavily dominated by KpSC (Fig. 2A). A similar pattern of KoSC or KpSC dominance was found in the controls (Fig. 2B). KoSC and KpSC were present together in appreciable amounts in only 4 of 21 subjects (Cases 1 and 3, Controls 2 and 5).

To further clarify the relationship between KoSC and KpSC in the 4 subjects with appreciable amounts of both, we determined their relative abundances in successive fecal samples over time and in relation to clinical parameters (Fig. 3). Case 1 demonstrated a pattern of KoSC dominance throughout successive weeks despite detection of low levels of KpSC at later time points; no antibiotics were administered during the time points prior to the development of NEC. Enteral feeds were fortified human milk and the diet was supplemented with carbohydrate/fat powder and liquid protein 8 and 4 days, respectively, before the development of NEC. Case 3 displayed the reverse pattern, with KpSC dominance despite detection of low levels of KoSC in later weeks; no antibiotics were administered during the time points prior to the development of NEC. Enteral feeds were fortified human milk and the diet was supplemented with carbohydrate/fat plus amino acid powder 8 days before the development of NEC. Controls 2 and 5, on the other hand, demonstrated dramatic shifts between KoSC and KpSC over successive weeks. Control 2 received treatment with Meropenem and Rifampin for a non-intestinal infection during the interval period but no dietary changes. Control 5 was receiving feeds of fortified human milk and the diet was supplemented with carbohydrate/fat powder 5 days before an increase in KoSC, and with formula 4 days before the disappearance in KpSC; no antibiotics were provided during the time points.

StrainID amplicon sequencing demonstrated that very few Klebsiella strains were shared between subjects. To differentiate the Klebsiella strains involved, we analyzed the StrainID amplicons using DADA2 to infer distinct ASV groups that correspond to individual strains¹⁶. ASV fingerprint patterns representing distinct KoSC or KpSC strains were present in all 9 NEC cases and 12 controls (Fig. 4A and B). Most strains among the NEC cases and controls were unique, indicating that transmission of dominant strains within the subject populations was not occurring. Case 9 and Control 17 shared two distinct patterns, indicating they were likely colonized with the same KoSC strains at different time points (Fig. 4A), while case 10 and Control 19 shared patterns indicating they were likely colonized with the same KpSC strain (Fig. 4B).

Cytotoxin-producing KoSC members were more prevalent in NEC subjects versus controls. Members of KoSC can harbor a biosynthetic gene cluster responsible for generating the enterotoxins tilimycin and tilivalline¹², which may facilitate mucosal damage in NEC¹⁰. To determine if cytotoxin-producing KoSC was more prevalent in NEC cases versus controls, and in KoSC- versus KpSC-dominated microbiomes, we analyzed stool samples using PCR and a selective culture system as described in the Methods. Toxin-positive KoSC members were identified and cultured from 8 of 10 NEC cases and 5 of 20 controls (P = 0.004) (Table 2; and Supplementary Tables S1, S2 and Figures S1, S2). In one control subject (Control 2), KoSC was not detected by PCR or culture, despite KoSC reads being present in the StrainID analysis; this subject received broad-spectrum antibiotics during specimen collection. In all remaining subjects, KoSC isolates were recovered when StrainID identified appreciable KoSC reads. Toxin-positive KoSC strains were detected in several KpSC-dominated microbiomes of NEC subjects (Cases 3, 5 and 10) and in one NEC subject (Case 8) with a low (< 2%) overall abundance of Klebsiella. Toxin-positive KoSC was identified in one control (Control 20) with a KpSC-dominated microbiome; this infant was treated for suspected (Bell’s stage 1), but not confirmed, NEC. No KoSC strains were identified in controls with a low (< 2%) overall abundance of Klebsiella. Collectively, these data demonstrate the greater prevalence of cytotoxin-producing KoSC in NEC cases compared to controls; KoSC was either the dominant Klebsiella species or secondary to KpSC in the majority of NEC cases.

Table 2

KoSC isolates recovered from cases and controls.
Subject	Taxonomic assignment	Toxin +/-
Case 1	K. grimontii^a,b	+
Case 2	K. michiganesis^a	+
Case 3	K. michiganesis^a	+
Case 4	K. grimontii^a K. michiganesis^a	+ -
Case 5	K. grimontii^a K. michiganesis^a	+ -
Case 8	K. oxytoca^b	+
Case 9	Unclassified KoSC^b*	+
Case 10	K. grimontii^b	+
Control 5	K. michiganesis^a	+
Control 10	K. michiganesis^a	+
Control 14	K. grimontii^b	+
Control 17	K. oxytoca^b	+
Control 20	Unclassified KoSC^b*	+
Toxin +/- denotes the presence or absence of npsA/B. ^aTaxonomic assignment by average nucleotide identity (ANI)/OrthoANI⁴⁰. ^bTaxonomic assignment by Cosic et al. PCR scheme⁴¹. Maps closest to K. grimontii.* Abbreviations: KoSC, Klebsiella oxytoca species complex; K, Klebsiella.

Enterococcus faecalis was co-dominant with KoSC but present infrequently with KpSC.

To ascertain if other bacteria were associated with KoSC or KpSC, the prevalence of other microbes was determined by StrainID. The analysis revealed E. faecalis to be co-dominant with KoSC; E. faecalis accounted for 11% and 1% of the reads, respectively, in KoSC- and KpSC-dominated microbiomes (P = 0.02, Fig. 5). No significant associations were found between KoSC or KpSC and the abundance of Clostridia, Staphylococcus, Veillonella, Enterobacter or Escherichia spp.

A substantial body of evidence points to LPS-induced TLR4 signaling cascades as causal to the profound intestinal inflammatory response observed in infants with NEC³. While gut dysbiosis with Gram-negative bacteria is believed to be a key antecedent event^4,17, there is uncertainty as to whether gut colonization with particular bacterial species predisposes to the development of NEC. In this regard, it is unlikely that all colonizing Gram-negative bacteria share the same pathogenic potential to incite intestinal damage¹⁸. Members of the genus Klebsiella have been implicated in previous studies^5–7; however, specific species identification is generally not possible using phenotypic characteristics or short segments of the 16S rRNA gene (e.g., V4)⁹. Here, we utilized a novel long-read amplicon spanning the 16S and 23S rRNA genes to determine the relative abundance of KoSC and KpSC across NICU subjects at risk for developing NEC. We found that Klebsiella was ubiquitous, colonizing the majority of preterm infants; consistent with previous reports^{5–7, 10,14}, Klebsiella was more prevalent in NEC subjects compared to non-NEC controls. We further reveal a pattern of colonization in which the gut microbiota was heavily dominated by either KoSC or KpSC, suggesting competition between the two for luminal resources. The data provide new insights into the microbial community structure in the preterm gut and how members of the genus Klebsiella may contribute to this catastrophic outcome of prematurity.

Klebsiella spp. are opportunistic pathogens that are frequently isolated from hospital environments and a leading cause of nosocomial infections^8,9. The healthy human microbiome also serves as a potential reservoir for infection; gut colonization with Klebsiella spp. is a well-established precursor for extraintestinal infection in adults^19,20. While Klebsiella spread among adults is often multi-strain^19,20, single strain outbreaks also occur²¹. Our analysis of Klebsiella ASV fingerprint patterns identified specific strains in the fecal microbiota of preterm infants and demonstrated that very few KoSC or KpSC strains were shared between subjects. The absence of dominant strains argues against there being a common environmental reservoir (e.g., sink, bedding). We did not sample parents in our cohort, and, given the broad distribution of strains, it is possible that parental transmission or alternate reservoirs were involved.

Production of the cytotoxins tilimycin and tilivalline by KoSC can cause AAHC in older children and adults^11–13 and has been linked to NEC in premature infants¹⁰. Our analysis detected cytotoxin-producing KoSC in the majority of NEC subjects, supporting a causal role for this bacterium in the disease. Surprisingly, several toxin-producing isolates were recovered from NEC subjects whose gut microbiomes were dominated by KpSC. The pattern of either KoSC or KpSC dominance suggests exclusionary Klebsiella species competition, and the context in which competitive stress could increase the expression of virulence factors²² warrants further investigation. It is possible that intestinal injury represents collateral damage as the result of competitive mechanisms unleashed during Klebsiella inter-species warfare. In the preterm gut, these processes may be further exacerbated by activation of TLR-4 signaling pathways in response to Klebsiella LPS ^2,3,23,24. Thus, the culmination in the development of NEC could involve competitive microbial communities interacting with bacterial signaling receptors (e.g., TLR4) on the premature intestine.

Members of KoSC utilize a broader array of sugars than KpSC which enables them to outcompete KpSC in murine models of gut colonization²⁵, while providing the fermentative energy to support cytotoxin production²⁶. In Case 3 and Control 5, an increase in KoSC abundance occurred following an increase in dietary carbohydrates. The development of NEC also occurred in Cases 1 and 3 after an increase in dietary carbohydrates. Preterm infants are at risk for carbohydrate malabsorption because they have under-developed intestinal brush border enzymes²⁷. The availability of undigested carbohydrates may serve as the impetus for KoSC and KpSC to engage each other in conflict over valuable resources.

Provision of broad-spectrum antibiotics enhances colonization with KpSC²⁸. Along these lines, the dramatic shift in KoSC to KpSC observed in Control 2 corresponded with antibiotic treatment for a non-intestinal infection. Many members of KpSC also exploit the type VI secretion system (T6SS) which can facilitate intra- and inter-species killing depending on environmental cues²⁹. Deciphering the contextual factors in the gut in association with diversity of genes regulating sugar utilization, cytotoxin production, anti-microbial resistance and T6SS would assist in establishing predictive patterns of KoSC versus KpSC dominance.

E. coli was found to be significantly higher in the non-NEC controls, which could indicate that this microbe filled a niche that was occupied by Klebsiella in the cases. Enhanced sugar utilization by members of KoSC is known to facilitate colonization resistance against E. coli ³⁰; thus, an alternate explanation is that E. coli was replaced by Klebsiella. Several reports have linked colonization of Enterococcus with KpSC^31,32. In the current study, we found that E. faecalis was predominantly associated with KoSC, rather than KpSC. Recently, a broad scale genomic analysis differentiated KoSC from KpSC by the presence of genes involved in a type II system to secrete pullulanase, a debranching enzyme that breaks down complex sugars³³. Enterococci do not express glycosidases that degrade mucosal polysaccharides, and their carbohydrate utilization is limited to less complex sugars³⁴. Therefore, it is possible that E. faecalis co-colonizes with KoSC via cooperative carbohydrate metabolism³⁵, whereby E. faecalis takes advantage of pullulanase converting complex polysaccharides into small fermentable sugars. Cytotoxin production by KoSC also may be contributory as tilimycin was recently reported to enhance Enterococcus growth and restrict E. coli colonization³⁶.

In summary, StrainID amplicon sequencing and ASV fingerprinting provides improved resolution to differentiate KoSC and KpSC in fecal samples and suggests that the two species are in direct competition in the preterm gut. Untangling the contextual factors, genetic diversity of specific strains and competitive mechanisms that can result in intestinal injury are important to understand how these microbes contribute to this devastating disease of prematurity.

Study population. Subjects were cared for at two affiliated NICUs in Hartford and Farmington, CT. Infants with a gestational age of less than 32 weeks were enrolled from March 2017 to October 2019; those with known congenital malformations of the intestine or not expected to survive beyond the first week were excluded. Study subjects underwent routine care and informed written consent was obtained from a parent on behalf of their infant. Cases were infants whose clinical courses and radiographic findings were consistent with Bell’s stage 2 or 3 NEC³⁷. Two control subjects were matched to each NEC case. The study was approved by the Institutional Review Board of Connecticut Children’s Medical Center and all research was performed in accordance with relevant guidelines and regulations.

Collection of samples and clinical data. Fecal samples were collected on an approximate weekly basis using sterile disposable spatulas during diaper changes, placed into sterile containers, and immediately frozen at − 80°C until processing. Samples collected prior to and up to 2–3 weeks after the diagnosis of NEC were included in the analysis. Samples from control infants were time-matched by the closest chronological age corresponding to case samples. Clinical data were obtained from enrolled infants including demographics, gestational age, mode of delivery, day of life (DOL) of NEC, DOL of sample acquisition, exposure to antibiotics, and diet.

StrainID amplicon sequencing. Fecal samples were processed and analyzed as previously described¹⁶. Briefly, fecal DNA was purified, PCR amplified, and pooled for sequencing using the Complete StrainID Kit (StrainID set A [barcodes 1 to 96]; Intus Biosciences, Farmington, CT) according to the manufacturer’s instructions. Amplicon libraries were created using the SMRTbell express template prep kit 2.0 (catalog number 100-938-900; PacBio). The library was sequenced on a Sequel IIe system (Pacific Biosciences) at the University of Delaware, Delaware Biotechnology Institute Sequencing and Genotyping Center, Newark, DE. The selected reads from each sample were primer trimmed and filtered to reads within the length range of 1,900 to 3,000 bp. The trimmed and filtered reads were analyzed manually via a histogram to identify peaks of read lengths that are likely to represent unique amplicons. The corresponding read length ranges (i.e., the 2,400- to 2,405-bp range from each sample) were passed to DADA2³⁸ and pooled for ASV inference¹⁶. A sequence table of ASV abundance per sample was produced as part of the DADA2 output, and a heat map was generated in R using the sequence table. SBanalyzer 2.4 (Intus Biosciences) was used to map ccs reads to the Athena database and assign taxonomic identification¹⁶.

Identification of cytotoxin-producing KoSC in fecal samples. DNA from fecal samples and clinical isolates was extracted using the DNeasy Power Soil Kit and DNeasy Blood and Tissue Kit (Qiagen, Germantown, MD), respectively. Fecal samples were screened by PCR for the KoSC-specific gene pehX and genes encoding enzymes in the cytotoxin biosynthetic pathway (npsA and npsB)³⁹. For culturing KoSC, a loopful of fecal material was inoculated into 5 ml of Luria-Bertani (Lennox) broth (LB) and grown for 24 h with shaking at 225 rpm at 37°C. Serial dilutions of fecal cultures were plated on hydroxybenzoic agar as previously described¹⁰. Individual KoSC colonies were confirmed as toxin-positive or -negative by PCR (pehX, npsA/B)¹⁰. For taxonomic assignments of the isolates, libraries were made using NexteraXT and sequenced on MiSeq v2 500 cycle (Illumina Inc., San Diego, CA) at UConn Microbial Analysis, Resources, and Services, Storrs, CT. The sequenced genomes were compared across various KoSC phylogroups and an average nucleotide identity (ANI)/OrthoANI value of 98% or greater indicated a matching phylogroup^26,40. Isolates that were not analyzed by (ANI)/OrthoANI were classified using a PCR-typing strategy to detect the following genes: npsA, bla_OXY−1, bla_OXY−2, bla_OXY−4/6, orfABC, leupAB, and orfA’⁴¹. Isolates were further characterized by their ability to metabolize D-melezitose⁴². The primer sequences used were as follows: npsA 5’-GCGCTGTTATGGTTCCCGT-3’ and 5’-CCGGGCACGCTTGTTACATC-3’; npsB 5’-TGCAGGGTACGCTAAATATTTTAGCT-3’ and 5’-ACCCACTTACTTTGCGTATAACCAAT-3’; pehX 5’-GATACGGAGTATGCCTTTACGGTG-3’ and 5’-TAGCCTTTATCAAGCGGATACTGG-3’; all other primer sequences for the PCR-typing strategy are as published⁴¹.

Statistics. The analyses were carried out using GraphPad Prism version 9.3.0 (GraphPad Software, San Diego, CA). Mann-Whitney U test was used for quantitative data and Chi-square test for categorical data. A minimum threshold of 400 reads per sample was used to reduce outliers in the microbiome analysis.

Acknowledgments: The authors thank Justin Radolf for his guidance and critical evaluation of the manuscript, Juan Salazar and James Moore for their encouragement and support. We are indebted to the NICU staff at Connecticut Children’s Medical Center for their assistance with collecting stool samples. We also thank all the patients and families that participated in this study.

Author Contributions: A.P.M. designed the study, supervised data collection and experiments, analyzed and interpreted the data and wrote the manuscript. S.C., K.U. and M.J.C. contributed to the study design, performed experiments, analyzed and interpreted the data and wrote the manuscript. K.R. and N.L. performed experimental work. S.L. recruited subjects, collected clinical data and assisted with sample analysis. Y.Z. was involved with the analysis and interpretation of microbial sequencing data. E.J., D.G. and M.D.D. helped with the microbiome analysis and drafting the manuscript. All authors reviewed the manuscript.

Data Availability: StrainID amplicon data and the whole-genome assemblies for KoSC isolates are deposited under Bioproject accession number PRJNA908822.

Additional Information: This work was supported by funds from the Connecticut Children’s Department of Research (to A.P.M. and M.J.C.) and the Connecticut Children’s Stevenson Fund for Microbiome Research (to A.P.M.). Competing interest statement: M.D.D. is a founder and shareholder of Intus Biosciences. D.G. and E.J. are employees and shareholders of Intus Biosciences. A.P.M. serves on the Advisory Board and is a shareholder of Intus Biosciences. Intus Biosciences was previously named Shoreline Biome. The remaining authors declare no potential conflicts of interest.

Neu, J. & Walker, W. A. Necrotizing enterocolitis. N. Engl. J. Med 364, 255–264, doi:10.1056/NEJMra1005408 [doi] (2011).
Nanthakumar, N. et al. The mechanism of excessive intestinal inflammation in necrotizing enterocolitis: an immature innate immune response. PLoS. One 6, e17776, doi:10.1371/journal.pone.0017776 [doi] (2011).
Hackam, D. J. & Sodhi, C. P. Bench to bedside - new insights into the pathogenesis of necrotizing enterocolitis. Nat Rev Gastroenterol Hepatol 19, 468–479, doi:10.1038/s41575-022-00594-x (2022).
Warner, B. B. et al. Gut bacteria dysbiosis and necrotising enterocolitis in very low birthweight infants: a prospective case-control study. Lancet, doi:S0140-6736(16)00081-7 [pii];10.1016/S0140-6736(16)00081 – 7 [doi] (2016).
Frantz, I. D., III, L'heureux, P., Engel, R. R. & Hunt, C. E. Necrotizing enterocolitis. J. Pediatr 86, 259–263, doi:S0022-3476(75)80485-9 [pii] (1975).
Sim, K. et al. Dysbiosis anticipating necrotizing enterocolitis in very premature infants. Clin. Infect. Dis 60, 389–397, doi:ciu822 [pii];10.1093/cid/ciu822 [doi] (2015).
Torrazza, R. M. et al. Intestinal microbial ecology and environmental factors affecting necrotizing enterocolitis. PLoS. One 8, e83304, doi:10.1371/journal.pone.0083304 [doi];PONE-D-13-28041 [pii] (2013).
Wyres, K. L., Lam, M. M. C. & Holt, K. E. Population genomics of Klebsiella pneumoniae. Nat Rev Microbiol 18, 344–359, doi:10.1038/s41579-019-0315-1 (2020).
Yang, J. et al. Klebsiella oxytoca Complex: Update on Taxonomy, Antimicrobial Resistance, and Virulence. Clin Microbiol Rev 35, e0000621, doi:10.1128/CMR.00006-21 (2022).
Paveglio, S. et al. Cytotoxin-producing Klebsiella oxytoca in the preterm gut and its association with necrotizing enterocolitis. Emerg. Microbes. Infect 9, 1321–1329, doi:10.1080/22221751.2020.1773743 [doi] (2020).
Schneditz, G. et al. Enterotoxicity of a nonribosomal peptide causes antibiotic-associated colitis. Proc. Natl. Acad. Sci. U. S. A 111, 13181–13186, doi:1403274111 [pii];10.1073/pnas.1403274111 [doi] (2014).
Unterhauser, K. et al. Klebsiella oxytoca enterotoxins tilimycin and tilivalline have distinct host DNA-damaging and microtubule-stabilizing activities. Proc. Natl. Acad. Sci. U. S. A 116, 3774–3783, doi:1819154116 [pii];10.1073/pnas.1819154116 [doi] (2019).
Hogenauer, C. et al. Klebsiella oxytoca as a causative organism of antibiotic-associated hemorrhagic colitis. N. Engl. J. Med 355, 2418–2426, doi:355/23/2418 [pii];10.1056/NEJMoa054765 [doi] (2006).
Olm, M. R. et al. Necrotizing enterocolitis is preceded by increased gut bacterial replication, Klebsiella, and fimbriae-encoding bacteria. Sci Adv 5, eaax5727, doi:10.1126/sciadv.aax5727 (2019).
Tedersoo, L., Albertsen, M., Anslan, S. & Callahan, B. Perspectives and Benefits of High-Throughput Long-Read Sequencing in Microbial Ecology. Appl Environ Microbiol 87, e0062621, doi:10.1128/AEM.00626-21 (2021).
Graf, J. et al. High-Resolution Differentiation of Enteric Bacteria in Premature Infant Fecal Microbiomes Using a Novel rRNA Amplicon. mBio 12, doi:10.1128/mBio.03656-20 (2021).
Mai, V. et al. Fecal microbiota in premature infants prior to necrotizing enterocolitis. PLoS. One 6, e20647, doi:10.1371/journal.pone.0020647 [doi];PONE-D-11-04694 [pii] (2011).
Casadevall, A. The Pathogenic Potential of a Microbe. mSphere 2, doi:10.1128/mSphere.00015-17 (2017).
Gorrie, C. L. et al. Gastrointestinal Carriage Is a Major Reservoir of Klebsiella pneumoniae Infection in Intensive Care Patients. Clin Infect Dis 65, 208–215, doi:10.1093/cid/cix270 (2017).
Martin, R. M. et al. Molecular Epidemiology of Colonizing and Infecting Isolates of Klebsiella pneumoniae. mSphere 1, doi:10.1128/mSphere.00261-16 (2016).
Snitkin, E. S. et al. Tracking a hospital outbreak of carbapenem-resistant Klebsiella pneumoniae with whole-genome sequencing. Sci Transl Med 4, 148ra116, doi:10.1126/scitranslmed.3004129 (2012).
Cornforth, D. M. & Foster, K. R. Competition sensing: the social side of bacterial stress responses. Nat Rev Microbiol 11, 285–293, doi:10.1038/nrmicro2977 (2013).
Leaphart, C. L. et al. A critical role for TLR4 in the pathogenesis of necrotizing enterocolitis by modulating intestinal injury and repair. J. Immunol 179, 4808–4820, doi:179/7/4808 [pii] (2007).
Egan, C. E. et al. Toll-like receptor 4-mediated lymphocyte influx induces neonatal necrotizing enterocolitis. J. Clin. Invest 126, 495–508, doi:83356 [pii];10.1172/JCI83356 [doi] (2016).
Osbelt, L. et al. Klebsiella oxytoca causes colonization resistance against multidrug-resistant K. pneumoniae in the gut via cooperative carbohydrate competition. Cell Host Microbe 29, 1663–1679 e1667, doi:10.1016/j.chom.2021.09.003 (2021).
Ledala, N. et al. Bacterial Indole as a Multifunctional Regulator of Klebsiella oxytoca Complex Enterotoxicity. mBio, e0375221, doi:10.1128/mbio.03752-21 (2022).
Kien, C. L. Colonic fermentation of carbohydrate in the premature infant: possible relevance to necrotizing enterocolitis. J. Pediatr 117, S52-S58, doi:S0022-3476(05)81131-X [pii];10.1016/s0022-3476(05)81131-x [doi] (1990).
Jiao, Y. et al. Risk factors for carbapenem-resistant Klebsiella pneumoniae infection/colonization and predictors of mortality: a retrospective study. Pathog Glob Health 109, 68–74, doi:10.1179/2047773215Y.0000000004 (2015).
Storey, D. et al. Klebsiella pneumoniae type VI secretion system-mediated microbial competition is PhoPQ controlled and reactive oxygen species dependent. PLoS Pathog 16, e1007969, doi:10.1371/journal.ppat.1007969 (2020).
Oliveira, R. A. et al. Klebsiella michiganensis transmission enhances resistance to Enterobacteriaceae gut invasion by nutrition competition. Nat Microbiol 5, 630–641, doi:10.1038/s41564-019-0658-4 (2020).
Caballero, S. et al. Distinct but Spatially Overlapping Intestinal Niches for Vancomycin-Resistant Enterococcus faecium and Carbapenem-Resistant Klebsiella pneumoniae. PLoS Pathog 11, e1005132, doi:10.1371/journal.ppat.1005132 (2015).
Collingwood, A. et al. Epidemiological and Microbiome Associations Between Klebsiella pneumoniae and Vancomycin-Resistant Enterococcus Colonization in Intensive Care Unit Patients. Open Forum Infect Dis 7, ofaa012, doi:10.1093/ofid/ofaa012 (2020).
Long, H., Hu, Y., Feng, Y. & Zong, Z. Genome Analysis of Klebsiella oxytoca Complex for Antimicrobial Resistance and Virulence Genes. Antimicrob Agents Chemother 66, e0218321, doi:10.1128/aac.02183-21 (2022).
Pultz, N. J., Hoskins, L. C. & Donskey, C. J. Vancomycin-resistant Enterococci may obtain nutritional support by scavenging carbohydrate fragments generated during mucin degradation by the anaerobic microbiota of the colon. Microb Drug Resist 12, 63–67, doi:10.1089/mdr.2006.12.63 (2006).
Engevik, M. A. et al. Mucin-Degrading Microbes Release Monosaccharides That Chemoattract Clostridioides difficile and Facilitate Colonization of the Human Intestinal Mucus Layer. ACS Infect Dis 7, 1126–1142, doi:10.1021/acsinfecdis.0c00634 (2021).
Kienesberger, S. et al. Enterotoxin tilimycin from gut-resident Klebsiella promotes mutational evolution and antibiotic resistance in mice. Nat Microbiol 7, 1834–1848, doi:10.1038/s41564-022-01260-3 (2022).
Bell, M. J. et al. Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Ann. Surg 187, 1–7 (1978).
Callahan, B. J. et al. High-throughput amplicon sequencing of the full-length 16S rRNA gene with single-nucleotide resolution. Nucleic Acids Res 47, e103, doi:5527971 [pii];10.1093/nar/gkz569 [doi] (2019).
Greimel, T. M. et al. Toxin-Producing Klebsiella oxytoca in Healthy Infants: Commensal or Pathobiont? J Pediatr Gastroenterol Nutr 74, e1-e7, doi:10.1097/MPG.0000000000003299 (2022).
Lee, I., Ouk, K. Y., Park, S. C. & Chun, J. 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] (2016).
Cosic, A. et al. Variation in Accessory Genes Within the Klebsiella oxytoca Species Complex Delineates Monophyletic Members and Simplifies Coherent Genotyping. Front Microbiol 12, 692453, doi:10.3389/fmicb.2021.692453 (2021).
Merla, C. et al. Description of Klebsiella spallanzanii sp. nov. and of Klebsiella pasteurii sp. nov. Front Microbiol 10, 2360, doi:10.3389/fmicb.2019.02360 [doi] (2019).

Competing interest reported. M.D.D. is a founder and shareholder of Intus Biosciences. D.G. and E.J. are employees and shareholders of Intus Biosciences. A.P.M. serves on the Advisory Board and is a shareholder of Intus Biosciences. Intus Biosciences was previously named Shoreline Biome. The remaining authors declare no potential conflicts of interest.

SupplementaryTablesFiguresRevFN.pdf

Download PDF

Journal Publication

published 16 May, 2023

Read the published version in Scientific Reports →

Editorial decision: Major revision
01 Mar, 2023
Reviews received at journal
28 Feb, 2023
Reviews received at journal
17 Feb, 2023
Reviews received at journal
15 Feb, 2023
Reviewers agreed at journal
03 Feb, 2023
Reviewers agreed at journal
01 Feb, 2023
Reviewers invited by journal
01 Feb, 2023
Editor assigned by journal
01 Feb, 2023
Editor invited by journal
17 Jan, 2023
Submission checks completed at journal
17 Jan, 2023
First submitted to journal
11 Jan, 2023

You are reading this latest preprint version

High-resolution Microbiome Analysis Reveals Exclusionary Klebsiella Species Competition in Preterm Infants at Risk for Necrotizing Enterocolitis

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Results

Discussion

Methods

Declarations

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1