Role of Biofilm Formation in Ureaplasma Antibiotic Susceptibility and Development of Bronchopulmonary Dysplasia in Preterm Neonates

Abstract

BACKGROUND

Ureaplasma respiratory tract colonization is a risk factor for bronchopulmonary dysplasia (BPD) in preterm infants, but whether Ureaplasma isolates from colonized infants can form biofilms is unknown. We hypothesized that Ureaplasma isolates vary in capacity to form biofilms that contribute to their antibiotic resistance and ability to evade host immune responses. Study objectives were to 1) determine the ability of Ureaplasma isolates from preterm neonates to form biofilms in vitro; 2) compare the susceptibility of the sessile and planktonic organisms to azithromycin and erythromycin; and 3) determine the relationship of biofilm-forming capacity in Urea-plasma isolates and the risk for BPD.

METHODS

Forty-three clinical isolates from preterm neonates and five ATCC strains were characterized for their capacity to form biofilms in vitro and antibiotic susceptibility was performed on each isolate pre- and post-biofilm formation.

RESULTS

Forty-one (95%) clinical and 4 of 5 (80%) ATCC isolates formed biofilms. All isolates were more susceptible to azithromycin (Minimum Inhibitory Concentration, MIC₅₀ 2 μg/mL) than erythromycin (MIC₅₀ 4 μg/mL), and biofilm formation did not significantly affect antibiotic susceptibility for the 2 tested antibiotics. The MIC₅₀ and minimum biofilm inhibitory concentrations (MBIC₅₀) for U. urealyticum clinical isolates for azithromycin were higher than for MIC₅₀ and MBIC₅₀ for U. parvum isolates. There were no differences in MIC or MBICs among isolates from BPD infants and non-BPD infants.

CONCLUSIONS

Capacity to form biofilms is common among Ureaplasma spp. isolates, but biofilm-formation did not impact MICs for azithromycin or erythromycin.

INTRODUCTION

The genital mycoplasmas Ureaplasma parvum (serovars 1, 3, 6, and 14) and U. urealyticum (serovars 2, 4, 5, 7-13) are common commensals in the female genitourinary tract, but colonization with these organisms during pregnancy has been increasingly associated with adverse outcomes such as infertility, stillbirth, intrauterine infection, and preterm birth.^{1, 2} Vertical transmission may occur in utero or at the time of birth. The organisms can be detected in amniotic fluid,³ placental tissues,⁴ cord blood,^5-7 neonatal respiratory⁸ and gastric secretions,⁹ and cerebral spinal fluid.⁶ Detection of ureaplasmal organisms in preterm infants has been associated with an increased incidence of bronchopulmonary dysplasia (BPD),^{10, 11} necrotizing enterocolitis (NEC),¹² meningitis,¹³ and abnormal cranial ultrasounds.² Although the organisms are susceptible to macrolide antibiotics in vitro ,¹⁴ trials of erythromycin and azithromyicn in women and infants have failed to demonstrate efficacy for eradication in intrauterine and neonatal lung compartments, prevention of maternal-fetal vertical transmission in humans^{15, 16} and experimental pregnancy models,¹⁷ or prevention of BPD in infected infants.^18-24 The mechanisms by which these organisms evade host defenses have not been elucidated.

Biofilms are sessile bacteria attached to a substratum encased in complex structures with an extracellular matrix of polysaccharide, lipid, DNA, and protein.²⁵ Biofilm formation increases microbial resistance to host defenses and antibiotics. Since Mycoplasma species have a limited genome and lack cell walls, it was unknown until recently that these species had the capacity to form biofilms.^25-29 García-Castillo et al.³⁰ reported that seven of nine Ureaplasma clinical isolates from the urine of healthy men or those with urethritis or chronic prostatitis formed biofilms in vitro. Whether Ureaplasma isolates from colonized infants can form biofilms is unknown. We hypothesized that Ureaplasma isolates vary in capacity to form biofilms that contribute to their antibiotic resistance and ability to evade host immune responses. This study 1) determined the ability of Ureaplasma spp. isolates from preterm neonates to form biofilms in vitro; 2) compared the susceptibility of the sessile and planktonic organisms to azithromycin and erythromycin; and 3) determined the relationship of biofilm-forming capacity in Urea-plasma spp. isolates and the risk for BPD.

MATERIALS AND METHODS

Ureaplasma isolates

Ureaplasma isolates were selected from banked isolates derived from respiratory specimens from preterm neonates admitted to the University of Maryland Medical Center (UMMC) Neonatal Intensive Care Unit that had previously been characterized by real-time PCR for species and in some cases, serotype classification.⁸ All isolates were within 2 passages from the original patient-derived culture. Five laboratory strains representing U. parvum serotypes 1 (27813), 3 (700970), 6 (27818), and 14 (33697) and U. urealyticum serotype 8 (27618) were obtained from the American Tissue Culture Collection (ATCC, Rockville, MD). All isolates stocks were frozen at −80°C prior to experiments.

Preparation of Antibiotic Stock Solutions

Erythromycin and azithromycin (Sigma, St Louis, MO) were used in both conventional antibiotic susceptibility testing and in biofilm susceptibility testing. Stock solutions of 1.6 mg/ml antibiotic in ethanol were prepared and stored at −20 °C until ready for use. A 1:100 dilution (16 ug/mL) of this stock solution in 10B broth medium was prepared for serial dilution in microtiter plates for conventional antibiotic susceptibility testing and biofilm susceptibility testing.

Conventional Minimum Inhibitory Concentration (MIC) Susceptibility Testing

Conventional MIC susceptibility testing with erythromycin and azithromycin were performed for each isolate in quadruplicate by the broth microdilution method as previously described.³⁰ Cultures of each clinical and laboratory isolate were grown at 37°C for 15-18 hours overnight. One hundred microliter of overnight culture diluted 1:100 in 10B broth (pH 6.0) for final concentration 1 × 10⁴ color changing units/ml was added to each well of a 96-well microtiter plate (Nunc, Thermo Fisher Scientific, Waltham, MA), prepared with serial dilutions of erythromycin and azithromycin. The plates were incubated at 37°C and observed until there was a color change indicating a pH change as a result of Ureaplasma growth in the wells without antibiotics. The MIC was defined as the lowest concentration preventing bacterial growth at the time that growth was observed in the positive control wells.

Biofilm Susceptibility Assay

Biofilm susceptibility assays were performed in duplicate as previously described with modifications³⁰. Briefly, 100 μL of a 1:100 dilution of an overnight culture was added to 100 μL of 10B medium (pH 6.0) in each well of a 96-well microtiter plate. For bacterial biofilm formation, pegs of a modified lid (Nunc Immuno™ TSP system, Thermo Fisher Scientific, Waltham, MA) were immersed into the wells and the plates incubated at 37°C until a color change was observed. The peg lids were washed to remove planktonic organisms by dipping the lids ten times into each of two 96-well plates containing PBS (−Ca²⁺, −Mg²⁺). Immediately following the second wash, the lids were placed onto another 96-well plate containing serial dilutions of both erythromycin and azithromycin, each in quadruplicate. To transfer the biofilm forming organisms to the antibiotic-containing medium, the plates were centrifuged at 2,000 x g for ten minutes at 4°C. The peg lids were discarded and replaced with a sterile standard lid. The plates were incubated at 37°C and observed until color change in the positive control wells. The MBIC was defined as the lowest concentration preventing bacterial growth at the time that growth was observed in the positive control wells. Similarly, biofilm formation was determined by growth in the positive control wells of the antibiotic plate.

Clinical outcomes

The clinical characteristics of the cohort from which the banked isolates were obtained have been previously described.⁸ The respiratory specimens were obtained after parental consent was obtained for protocols approved by the University of Maryland School of Medicine Institutional Review Board. Bronchopulmonary dysplasia severity was defined according to NIH criteria.³¹ Infants receiving supplemental oxygen at 28 d were assessed at 36 wk postmenstrual age or discharge whichever came first, and were classified as mild BPD if breathing room air; moderate BPD, if receiving supplemental oxygen <30%; or severe BPD, if receiving positive pressure support and/or supplemental oxygen ≥30% at the assessment time point.

Statistical analysis

Data were analyzed by Chi square and Wilcoxon signed rank test. A p value <0.05 was considered significant.

RESULTS

The banked clinical isolates selected for biofilm-forming capacity included 26 U. parvum and 17 U. urealyticum isolates (see Table, Supplemental Digital Content 1, http://links.lww.com/INF/B387 which summarizes MICs and MBICs for all isolates tested). Forty-one (95%) clinical and 4 of 5 (80%) ATCC isolates tested formed biofilms under in vitro condtions. The three non-biofilm forming isolates included ATCC 27618, a U. urealyticum serovar 8 strain, and two U. parvum serovar 6 isolates. All isolates were more susceptible to azithromycin (Minimum Inhibitory Concentration, MIC₅₀ 2 μg/mL) than erythromycin (MIC₅₀ 4 μg/mL), and biofilm formation did not significantly affect antibiotic susceptibility for the two tested antibiotics. There was a trend towards lower conventional MICs for erythromycin (p=0.07) and statisically significant lower MICs for azithromycin (p=0.006) for the ATCC strains compared to the MICs for these antibiotics for the clinical isolates (Fig. 1). No tested isolate was resistant to azithromycin. Seven of 48 isolates (14.5%) were resistant to erythromycin (MIC₅₀ ≥ 16 μg/ml) as planktonic cells (see Table, Supplemental Digital Content 1, http://links.lww.com/INF/B387). Of these, 2 were also resistant as biofilm-forming cells. Two additional isolates were sensitive to erythromycin as planktonic cells, but were resistant after biofilm formation.

Figure 1.

Comparison of antibiotic susceptibility of planktonic and sessile cultures of ATCC and clinical isolates. Conventional MIC susceptibility testing with erythromycin and azithromycin were performed for each ATCC (N=5) and clinical (N=43) isolate in quadruplicate by the broth microdilution method and the MBIC determined following biofilm-formation as previously described ³⁰. Data are presented as box plots with the bottom, top, and line through the middle of the box corresponding to the 25th percentile, 75th percentile, and 50th percentile (median), respectively. The whiskers extend from the 10th percentile to the 90th percentile.

Erythro, erythromycin, AZI, azithromycin

*p=0.006 vs AZI MIC₅₀ ATCC strains by Wilcoxon sign rank test

The MIC₅₀ and minimum biofilm inhibitory concentrations (MBIC₅₀) for U. urealyticum clinical isolates for azithromycin were higher than for MIC₅₀ and MBIC₅₀ for U. parvum isolates (Fig. 2). Both U. parvum planktonic and biofilm-forming cells were more susceptible to azithromycin than to erythromycin. The clinical outcomes of the patients from whom the isolates were selected included 14 no BPD, 10 mild BPD, 15 moderate-severe BPD, and 4 deaths.There were no differences in MIC₅₀ or MBIC₅₀ among isolates from BPD/death infants and non-BPD infants (Fig. 3).

Figure 2.

Comparison of antibiotic susceptibility of planktonic and sessile cultures of U. parvum and U. urealyticum isolates. Conventional MIC susceptibility testing with erythromycin and azithromycin were performed for each U. parvum (N=26) and U. urealyticum (N=17) isolate in quadruplicate by the broth microdilution method and the MBIC determined following biofilm-formation as previously described ³⁰. MIC₅₀ and MBIC₅₀ for erythromycin (Erythro) and azithromycin (AZI). Data are presented as box plots with the bottom, top, and line through the middle of the box corresponding to the 25th percentile, 75th percentile, and 50th percentile (median), respectively. The whiskers extend from the 10th percentile to the 90th percentile.

*p<0.01 compared to U. parvum isolate erythromycin MIC₅₀ and MBIC₅₀ by Wilcoxon sign rank test

†p<0.001 compared to azithromycin MIC₅₀ and MBIC₅₀ U. urealyticum isolates by Wilcoxon sign rank test

Figure 3.

Comparison of clinical isolates MIC₅₀ and MBIC₅₀ by clinical outcomes. There were no significant differences in antibiotic susceptibility among biofilm-forming isolates from infants with and without BPD/death. Data are presented as box plots with the bottom, top, and line through the middle of the box corresponding to the 25th percentile, 75th percentile, and 50th percentile (median), respectively. The whiskers extend from the 10th percentile to the 90th percentile.

DISCUSSION

Considerable effort has focused on determining whether the 10 distinct U. urealyticum and 4 U. parvum serotypes differ in pathogenic potential. Ureaplasma parvum serovars are the most common in clinical specimens, but some earlier antibody-based studies suggested that specific U. urealyticum serovars may have greater pathogenic potential.^32-34 Utilizing recently developed serovar specific primers/probes for all 14 serovars,³⁵ we previously demonstrated that serovars 3 and 6, alone and in combination, accounted for 96% of U. parvum isolates from colonized preterm infants in our institution.⁸ In our population, the U. urealyticum isolates were commonly a mixture of 2 or more serovars. Serovar 11 was the most common U. urealyticum serovar, detected as a single serovar or in combination with other serovars in 56% of U. urealyticum-positive samples. There were no statistical differences among the U. parvum and U. urealyticum species in BPD severity, but the sample size was inadequate to detect a difference at the serovar level. We proposed that there might be virulence genes that are common across multiple serovars that could be targeted to prevent or ameliorate the adverse effects of Ureaplasma infection in the immature lung. Using the genotyped, banked isolates from our previous study,⁸ the current study was conducted to determine whether biofilm-forming capacity is shared among the Ureaplasma serovars.

The present study confirms that most clinical Ureaplasma isolates derived from preterm infants and U. parvum laboratory strains form biofilms in vitro. We tested isolates representing both species and multiple isolates from the same serovar. Although both non-biofilm forming clinical isolates were serovar 6 strains, other serovar 6 isolates formed biofilms in vitro, suggesting that biofilm-forming capacity is not serovar-specific. Four of five ATCC isolates formed biofilms, indicating that the biofilm-forming capacity is likely stable after multiple passages.

Our study extends the observations of in vitro biofilm-forming capacity of clinical Ureaplasma isolates derived from healthy men and those with urethritis and chronic prostatitis.³⁰ In the previous study, two U. parvum serovar 3, five U. urealyticum serovar 7 and 2 of four U. urealyticum serovar 13 isolates formed biofilms in vitro. Scanning electron microscopy demonstrated that biofilm-forming bacteria adhere to the prostate ductal epithelium in vivo, contributing to organism persistence, chronic inflammation, and antibiotic treatment failures.³⁶ Recently, Romero and co-workers reported a case of M. hominis in vivo biofilm in the intrauterine cavity detected as “amniotic fluid sludge” by transvaginal ultrasound and the complex structure confirmed by scanning electron microscopy.³⁷ Mycoplasma pulmonis intranasally inoculated in mice formed biofilms on the tracheal epithelium in vivo.²⁵ These studies suggest that ureaplasmas colonizing the preterm infant respiratory epithelium may form biofilms that protect the organisms from host defenses and antibiotic treatment. To confirm the biologic relevance of the current in vitro studies, biofilm-formation of clinical Ureaplasma isolates should be tested in in vivo experimental models.

Since erythromycin and azithromycin are the most common antibiotics used clinically in preterm infants to treat Ureaplasma infections, we tested whether biofilm formation by Ureaplasma isolates altered susceptibility to these antibiotics. In agreement with previous studies of in vitro susceptibilities to macrolides, the Ureaplasma isolates in the current study were more susceptible to azithromycin than erythromycin.^38-40 In contrast to the decreased susceptibility to erythromycin of sessile cells compared to planktonic cells of Ureaplasma isolates from adult males,³⁰ there was no detectable difference between erythromycin or azithromycin MIC₅₀ and MBIC₅₀ for the Ureaplasma isolates from preterm infants. Since the activity of macrolides in vitro is affected by pH of the medium,^{41, 42} as well as inoculum size,⁴¹ the higher MICs observed in our study may be due, in part, to the acidic pH (6.0) of the 10B broth and variable inoculum size of the transferred sessile cells from the peg lids.

Despite in vitro susceptibility of Ureaplasma to erythromycin and azithromycin, ^{14, 43} trials of erythromycin or azithromycin therapy in the first few weeks of life in Ureaplasma colonized preterm infants have failed to demonstrate efficacy to prevent BPD,^{18, 19, 24} or eradicate respiratory tract colonization.²¹ The failure to prevent BPD in these studies may have been due to the small sample size of each study, or to the initiation of antibiotic therapy too late to prevent the lung inflammation and injury that contribute to the pathogenesis of BPD. Recently, Walls et al.⁴⁴ demonstrated that azithromycin, but not erythromycin prophylaxis, improved outcomes and reduced inflammation in a murine neonatal Ureaplasma infection model. This suggests that azithromycin may be effective if administered immediately after birth. An initial single dose pharmacokinetic study of 10 mg/kg azithromycin in infants 24-28 weeks gestation suggested that this dose was well-tolerated, but likely insufficient to maintain azithromycin concentrations above the MIC₅₀ for Ureaplasma.⁴⁵ In the current study, none of the isolates as planktonic or sessile cells were resistant to azithromycin, but 10-15% were resistant to erythromycin. Additional clinical pharmacokinetic/pharmacodynamic studies are needed to determine the optimal dosing regimen for azithromycin to eradicate Ureaplasma from the respiratory tract and potentially prevent BPD. Biofilm formation may require higher doses and longer course of therapy for effective eradication.