Leptotrichia amnionii sp. nov., a Novel Bacterium Isolated from the Amniotic Fluid of a Woman after Intrauterine Fetal Demise

A novel bacterium was isolated and characterized from the amnioticfluid of a woman who experienced intrauterine fetal demise inthe second trimester of pregnancy. The bacterium was a slow-growing,gram-negative anaerobic coccobacillus belonging to the genusLeptotrichia. Unlike Leptotrichia sanguinegens, the isolatedid not grow in chopped-meat glucose broth or on sheep bloodagar upon subculturing. The isolate was characterized by sequencingand analyzing its 16S rRNA gene. The 1,493-bp 16S ribosomalDNA sequence had only 96% homology with L. sanguinegens. Severalphylogenetic analyses indicated that L. amnionii is a distinctspecies and most closely related to L. sanguiegens.

INTRODUCTION

Top

Introduction

Molecular-based diagnostic and identification methods for fastidiousor uncultivable bacteria have resulted in the recognition ofmany new pathogenic microorganisms (3, 16). One of the mostsuccessful methods is PCR amplification and sequencing of thebacterial 16S rRNA gene. This method has been successfully appliedto environmental as well as clinical samples (6, 15). The largerRNA sequence databases at GenBank and at Ribosomal DatabaseProject II allow for a quick comparison of 16S ribosomal DNA(rDNA) sequences and accurate identification of bacteria thatare difficult to identify on the basis of phenotypic propertiesalone (10). The use of this method has greatly expanded thelist of indigenous microbial flora of humans and has helpedin recognizing the numerous opportunistic pathogens that causeinfections related to severe physiological stress and immunosuppressiondue to chemotherapy.

Leptotrichia species are slow-growing, gram-negative anaerobicflora of the oral cavity and genital tract (5). Colonizationby Leptotrichia species has been reported in over 40% of childrenless than a year old (19). Leptotrichia buccalis, which is consideredindigenous oral flora, has been associated with endocarditisin patients with Down's syndrome (2) and bacteremia in neutropenicchildren and adults (14, 22). They seem to colonize permucosalimplants of edentulous patients (12) and, not surprisingly,are often considered contaminants if isolated from clinicalspecimens.

Leptotrichia sanguinegens has recently been proposed as an agentof postpartum and neonatal bacteremia (4). It has not been identifiedfrom a healthy individual. We describe an isolate that is relatedto the species L. sanguinegens, but is different in its genotypicproperties and nutritional requirements (4). For this isolate,we proposed the name "L. amnionii sp. nov." (from "amnion,"the extraembryonic membrane enveloping the embryo in utero andcontaining the amniotic fluid), to signify its source of isolation.

CASE REPORT

Top

Case report

A 27-year-old previously healthy, multiparous female in thesecond trimester of pregnancy presented to the emergency roomwith severe headache, neck and back pain, and a temperatureof 102°F. The abdominal examination demonstrated no guardingor rebound. A purulent vaginal discharge was noted, but thephysical examination was otherwise normal. Fetal heart toneswere present. The patient was hospitalized. Initial laboratoryvalues demonstrated a leukocyte count of 7,800 with the differentialshowing 1 metamyelocyte, 10 band forms, 85 segmented neutrophils,and 3 lymphocytes. The hemoglobin level was 11.5 g/dl, and theC-reactive protein level was 11.4 mg/dl. A urinalysis was unremarkable,showing no evidence of infection. A wet preparation of the vaginaldischarge demonstrated no abnormal organisms or evidence ofsignificant vaginal infection. A PCR test for Chlamydia trachomatiswas negative. Cerebral spinal fluid evaluation was normal. Thesymptoms gradually resolved, and the patient was discharged.Six days later, the patient was seen in the outpatient clinic.No fetal heart tones could be heard, and an ultrasound confirmedan intrauterine fetal demise. The patient was admitted for uterineevacuation by labor induction. Prior to induction, an amniocentesiswas performed. The amniotic fluid was turbid and brown in colorand had a distinct foul smell. A gram stain of the amnioticfluid demonstrated gram-negative coccobacilli. A slow-growing,gram-negative anaerobic coccobacillus was recovered. Scant growthof Bacteroides fragilis and Propionibacterium acnes was observedin cultures of the placenta. The mother was given amoxicillin-clavulanicacid and had an uneventful recovery.

MATERIALS AND METHODS

Top

Materials and Methods

Microbiology. The amniotic fluid and the placenta tissue specimens were culturedon blood and chocolate agar under both aerobic and anaerobicconditions at 37°C. The template DNA for 16S rDNA PCR wasprepared from a few colonies that were isolated on the prereducedblood agar incubated anaerobically. The DNA was extracted witha Qiagen DNA extraction kit (Qiagen, Inc., Valencia, Calif.).Broad-range prokaryotic PCR primers (23) and nested sequencingprimers (17, 23) were used to amplify and sequence the 16S generRNA. The methodology has been described previously (17).

Phylogenetic analysis. The rDNA sequence of the L. amnionii sp. nov. was aligned witha database of archaeal, bacterial, and eucaryal SSU rRNA sequences(ca. 10,000 sequences in total) by using the ARB software package(18). Both BLAST analysis and the parsimony insertion tool ofARB tentatively placed the L. amnionii sequence within the bacterialdivision of Fusobacteria. Consequently, a subset of the ARBalignment, which included the Leptotrichia species of the divisionFusobacteria (including the species of the genus Leptotrichia),as well as members of other, outlying bacterial divisions, wasselected for phylogenetic analysis. Both full-length data setsand sequence alignments minimized by the use of the Lane mask(8) were analyzed. The sequences of Methanococcus jannaschiiand Sulfolobus acidocaldarius were selected as out-groups forphylogenetic analysis. The dendrogram presented in Fig. 2 wasconstructed by evolutionary distance analysis (neighbor joiningwith Olsen correction) with the ARB package (18). The robustnessof this tree was assessed by bootstrap resampling (>100 replicates)of evolutionary distance trees by using weighted least-squaresmean analysis with Kimura two-parameter or maximum-likelihoodcorrection of evolutionary distances (PAUP* version 4.0b2) (20).Parsimony and maximum-likelihood analyses (ARB or PAUP*) providedresults that were substantially similar to those of the evolutionarydistance algorithm.

View larger version (56K):
[in this window]
[in a new window]
FIG. 2. Evolutionary distance dendrogram of selected leptotrichial and fusobacterial 16S rRNA sequences, including that of the Leptotrichia-like sp. isolate. Two archaeal species, Methanococcus jannaschii (M59126) and Sulfolobus acidocaldarius (D14876), were chosen as out-groups for phylogenetic analysis. Sequences are identified by species name and GenBank accession number. Branch points supported by >90% bootstrap values are indicated by solid circles. Open circles represent branch points with bootstrap values in the range 75 to 89%. Branch points without circles were not resolved (bootstrap values in the range <75%) as specific groups by this analysis. The bar at the bottom indicates the number of nucleotide changes per site.

Nucleotide sequence accession number. The 16S rRNA sequence of the L. amnionii sp. nov. was depositedin GenBank and given accession no. AY078425.

RESULTS AND DISCUSSION

Top

Results and Discussion

Numerous gram-negative coccobacilli were observed in the amnioticfluid along with numerous neutrophils. Anaerobic culture ofthe amniotic fluid on blood and chocolate agar resulted in verysmall gray colonies, <1 mm in diameter, following 72 h ofincubation. Gram stain of the colonies revealed gram-negativecoccobacilli, including some filamentous forms (Fig. 1). Therewas no growth on blood agar incubated under aerobic conditions,nor was there anaerobic growth on kanamycin and vancomycin orMueller-Hinton agars upon subculturing. Viral cultures werenegative. Since the bacterium resembled L. sanguinegens, itwas inoculated into chopped meat glucose (CMG) broth and incubatedunder aerobic conditions (4). This medium did not sustain growth,as evidenced by the lack of turbidity of the medium. The organismwas extremely fastidious and did not survive beyond the thirdsubculture. There was insufficient growth to perform biochemicalor fatty acid analysis.

View larger version (78K):
[in this window]
[in a new window]
FIG. 1. Gram stain of amniotic fluid colonies demonstrating gram-negative pleomorphic bacilli.

The isolate was identified and characterized by PCR amplificationof the 16S rRNA gene by using broad-range eubacterial primersFD1 and RD1 (17, 23). The PCR product was directly sequencedas described previously (17). A 1,493-nucleotide consensus sequencewas created and edited with DNAsis software (Hitachi Corporation)and compared with the sequences deposited in the GenBank database.The submitted sequence had only 96% homology to L. sanguinegens(GenBank accession no. L37789).

Phylogenetic analysis. Based on its unique source of isolation, inability to grow onknown special media, such as CMG broth, and unique 16S rDNAsequence, it was evident that this bacterium was related to,but different from, L. sanguinegens (4). The phylogenetic relationshipof L. amnionii with other species of the bacterial divisionFusobacteria, including species of the genus Leptotrichia, wasinferred by evolutionary distance, parsimony, and maximum-likelihoodanalyses. Figure 2 shows a representative evolutionary distancedendrogram. Parsimony and maximum-likelihood analyses gave qualitativelysimilar results. Bootstrap resampling of data provided strongsupport for a specific association of L. amnionii with othermembers of the genus Leptotrichia. L. sanguinegens was identifiedas the closest neighbor of L. amnionii (bootstrap values of99 and 100% for distance and parsimony analyses, respectively).

Clinical significance. Three species of Leptotrichia, L. buccalis, L. trevisanii, andL. sanguinegens (also called Sneathia sanguinegens) (1), havebeen associated with human infections (Table 1). Fifty-ninepercent of the patients were immunosuppressed due to malignancy.Four cases of L. sanguinegens bacteremia were associated withpregnancy, and two neonates were infected (4).

View this table:
[in this window]
[in a new window]
TABLE 1. Comparison of clinical features of Leptotrichia species isolated until 2001

L. buccalis has been well characterized and is part of the normaloral flora. It has been isolated from babies <1 year of age,and 40% of babies seem to be carriers (19). L. trevisanii isa recently identified bacterium that was recovered from a patientwith myeloid leukemia. L. sanguinegens has been proposed asan agent of postpartum and neonatal bacteremia. At this time,we suspect that L. amnionii is indigenous to the urogenitaltract and is an opportunist in the appropriate clinical situations.Like other Leptotrichia-related clinical cases, this bacteriumwas isolated from a clinical condition that is physiologicallysomewhat analogous to having the same stress and immunosuppressionas an underlying malignancy. Hanff et al. described the presenceof a strong odor from two neonatal cases of infection possiblydue to L. sanguinegens (4), as was detected from the amnioticfluid in our case. It appears that L. amnionii is not a blood-lovingmicrobe like L. sanguinegens, because it failed to grow on bloodagar. Additional clinical isolates will help establish its trueecological niche and pathogenic potential.

Description of Leptotrichia amnionii sp. nov. The name "L. amnionii" (am' .on... L. gen. n., amnionii) is derived from the word"amnion." The organism is characterized by pleomorphic coccobacillus,long, nonmotile, fusiform cells. Some cells are joined end toend in a filamentous form. L. amnionii grows anaerobically onblood agar after 3 days of incubation and is closely relatedto L. sanguinegens based on its 16S rDNA sequences.

ACKNOWLEDGMENTS

We thank Karen Park, Carol Murray, and Teresa Aspeslet for technicalassistance, Alice Stargardt for help in preparing the manuscript,and Norman Pace for helpful discussions.

This work was supported in part by grants from the MarshfieldMedical Research Foundation.

FOOTNOTES

* Corresponding author. Mailing address: Clinical Research Center, Marshfield Medical Research and Education Foundation, 1000 North Oak Avenue, Marshfield, WI 54449. Phone: (715) 389-5363. Fax: (715) 389-3808. E-mail: shuklas{at}mmrf.mfldclin.edu.

REFERENCES

Top