Materials and Methods
Animals
Animal work was approved by The University of Western Australia’s Animal Ethics Committee. Thirty date-mated ewes ( Ovis aries) carrying singleton pregnancies received a single intraamniotic injection of 10 7 color change units (CCU) of UP at 55 days’ gestation. Fetal viability was confirmed in 28 of 30 animals by ultrasound imaging at 116 days gestation. Animals were assigned randomly to receive either: (1) a single maternal intramuscular and saline solution injection (control; n = 8); (2) a single intraamniotic erythromycin injection (3.2 mg/kg estimated fetal weight) and repeated maternal intramuscular erythromycin injections (500 mg; once every 8 hours for 3 days; n = 10), or (3) a single maternal 500 mg intramuscular injection and 3 intraamniotic erythromycin injections once every 48 hours (as per group 2; n = 10). Predominantly intramuscular (group 2) and predominantly intraamniotic (group 3) regimens were designed to deliver a supra–minimal inhibitory concentration (MIC) dose of erythromycin to maternal and fetal tissues, taking into account the poor transplacental passage of maternally administered erythromycin to the amniotic fluid (AF) and of intraamniotic administered erythromycin to the maternal circulation. Animals were killed at 125 days gestation. Tissues for UP quantification and inflammatory analysis were snap-frozen. Fetal lung (right upper lobe) was perfusion fixed in 10% neutral buffered formalin for 24 hours before paraffin embedding. Chorioamnion was immersed in 10% neutral buffered formalin for 24 hours before paraffin embedding.
Culture
UP was cultured with the use of 10B broth (Melbourne University Media Preparation Unit, Melbourne, Australia). Quantification and erythromycin resistance was assessed as described by Beeton et al.
Nucleic acid extraction
DNA was extracted directly from a 250 μL sample of cord blood plasma with the Siemens Sample Preparation Kit (version 1.0; Siemens, Munich, Germany) on a Kingfisher Duo extraction platform (Thermo Fisher Scientific Inc, Waltham, MA) as per manufacturer’s instructions. All extracts were eluted in a final volume of 100 μL of elution buffer (Siemens).
Total RNA was extracted from fetal tissues (lung right lower lobe, internal groin skin, liver and chorioamnion) with TRIzol (Life Technologies, Carlsbad, CA) and treated with Turbo-DNase (Life Technologies) as previously reported.
UP nucleic acid detection
UP DNA in cord blood plasma was detected with a real-time polymerase chain reaction (PCR) assay targeting the urease gene of UP as described by Yi et al, on a thermocycler (ViiA7; Life Technologies). Reaction mixtures were (final volume 20 μL): 1x Taqman FAST Advanced Master Mix, 0.9 μmol/L primers UU1613F and UU1524R, 0.25 μmol/L probe UU-parvo (FAM) 5 μL template DNA, and nuclease-free water (all Life Technologies). PCR cycling conditions were initial denaturation/Taq activation at 95°C for 20 seconds, 40 quantification cycles of 95°C for 1 second, and 60°C for 20 seconds (data acquisition phase).
UP RNA in fetal tissues was detected as described previously. Reaction mixtures were (final volume 20 μL) 10 μL EXPRESS SuperScript qPCR SuperMix Universal (Life Technologies), 2 μL EXPRESS SuperScript Mix for One-Step qPCR (Life Technologies), 50 nmol/L ROX reference dye (Life Technologies), 0.9 μmol/L primers UU1613F and UU1524R, 0.25 μmol/L probe UU-parvo (FAM), 125 ng total RNA template, and nuclease-free water. Reaction cycling conditions were 15 minutes reverse transcription at 50°C and a denaturation/polymerase activation at 95°C for 20 seconds, 40 cycles of 95°C for 3 seconds, and 60°C for 30 seconds (data acquisition phase). UP RNA levels were quantitated with ViiA7 real-time PCR system software (version 1.2.1; all Life Technologies) with a standard curve ( R 2 = 0.993) that consisted of UP control RNA at 35 ng, 17.5 ng, 8.75 ng, and 4.375 ng per 20 μL reaction.
Inflammation
Ovine-specific hydrolysis probes and PCR primers were used to perform quantitative PCR reactions. Reactions were performed with an EXPRESS One-Step SuperScript qRT-PCR Kit (all Life Technologies) that contained 400 ng RNA template in a final volume of 20 μL in accordance with manufacturer’s instructions. Reaction conditions were as described for UP RNA analysis. Quantification cycle values were normalized to 18S recombinant RNA and expressed as fold changes relative to control values. Reaction efficiencies were within limits that were proposed in the minimum information for publication of quantitative real-time PCR experiments guidelines.
Histologic findings
Five-micrometer paraffin sections of fetal lung and chorioamnion were stained with Meyer’s hematoxylin and eosin.
Qualitative scoring of airspace inflammatory cell infiltration was performed by a single investigator who was blinded to treatment groups. Six fields (total magnification, ×200) were scored for each animal. Airspace infiltration and consolidation was graded in the following manner: 0 = no inflammatory cells in airspace; 1 = airspace inflammatory cells, no consolidation; 2 = airspace inflammatory cells + limited microconsolidation (1-4 per field) foci; 3 = airspace inflammatory cells + numerous (>4 per field) but predominantly discrete microconsolidation foci; and 4 = airspace inflammatory cells + confluent airspace consolidation.
Qualitative scoring of histologic chorioamnionitis was performed by a single investigator who was blinded to treatment groups with a modified Redline staging system as described previously. Briefly, the tissue plane and severity of inflammatory cell infiltration was graded in the following manner: 0 = no inflammatory cells identifiable/no chorioamnionitis; 1 = stage 1, grade 1 chorioamnionitis; 2 = stage 1, grade 2 chorioamnionitis; 3 = stage 2, grade 2 chorioamnionitis; and 4 = stage 3, grade 2 chorioamnionitis.
Statistics
All values represent mean ± standard deviation (SD). Analyses were performed with IBM SPSS Statistics for Windows (version 20.0; IBM Corporation, Armonk, NY). Data were assessed for normality with Shapiro-Wilk tests. For normally distributed data, mean differences were tested for significance with one-way analysis of variance, with a probability value of .05 accepted as significant. Multiple post-hoc comparisons were performed with Tukey’s test. Between groups differences in nonparametric data were tested for significance with Kruskal-Wallis one-way analysis of variance, with a probability value of .05 accepted as significant. Multiple post-hoc comparisons were performed with Rank-Sum tests, with a probability value corrected for the number of multiple comparisons.
Results
All animals randomized to treatment survived to the conclusion of their respective protocols. There were no significant differences in birthweight, arterial cord blood pH, pO 2 , pCO 2 , or AF pH between treatment groups ( Table 1 ).
| Group | Delivery weight, kg | Arterial CB | Amniotic fluid pH | Ureaplasma parvum serovar 3 | ||||
|---|---|---|---|---|---|---|---|---|
| pH | pO 2 | pCO 2 | Positive arterial CB culture | Positive amniotic fluid culture | Erythromycin-resistant amniotic fluid | |||
| Intraamniotic U parvum serovar 3 control (n = 8) | 2.6 ± 0.4 | 7.2 ± 0.1 | 14.5 ± 6.0 | 74.3 ± 16.1 | 7.1 ± 0.1 | 0 (0/8) | 100% (8/8) | 0 (0/8) |
| Intraamniotic + repeated intramuscular erythromycin (n = 10) | 2.9 ± 0.3 | 7.1 ± 0.2 | 17.0 ± 7.1 | 66.2 ± 24.2 | 7.0 ± 0.2 | 10% (1/10) | 60% (6/10) | 33% (2/6) |
| Intramuscular + repeated intraamniotic erythromycin (n = 10) | 3.0 ± 0.3 | 7.2 ± 0.2 | 16.2 ± 4.9 | 67.4 ± 22.8 | 6.9 ± 0.2 | 0 (0/10) | 30% (3/10) | 33% (1/3) |
UP quantification
Animals that were treated with single maternal intramuscular and repeated intraamniotic erythromycin injections had significantly ( P = .003) reduced viable UP in the AF, compared with animals that were treated with saline solution (1.1 × 10 4 ± 3.3 × 10 4 vs 1.6 × 10 5 ± 3.4 × 10 5 CCU, respectively). UP-positive AF from the 3 animals from the single maternal intramuscular injection and repeated intraamniotic erythromycin injection group contained 10 5 , 10 3 , and 10 2 CCU UP, respectively. Single intraamniotic and repeated maternal intramuscular erythromycin treatment did not significantly reduce viable AF UP compared with control animals that were treated with saline solution ( Figure 1 ). Repeated intraamniotic erythromycin administration resulted in an absence of culturable AF UP in 7 of 10 animals and UP PCR–negative chorioamnion and skin in 5 of 10 animals. However, neither single maternal intramuscular and repeated intraamniotic treatment nor single intraamniotic and repeated maternal intramuscular erythromycin treatment completely cleared UP from the amniotic environment.
Of those animals that were culture positive for AF UP, none of the isolates from the saline solution control group, 2 isolates (33%) from the single intraamniotic and repeated maternal intramuscular group, and 1 isolate (33%) from the single maternal intramuscular and repeated intraamniotic treatment groups were erythromycin resistant in vitro (MIC >32 μg/mL at 72 hours). Bacteremia was uncommon in all groups, irrespective of treatment ( Table 1 ). The one animal that tested culture-positive for cord blood plasma UP also tested positive for cord blood plasma UP RNA. The erythromycin MIC for the AF isolate from this single intraamniotic and repeated maternal intramuscular erythromycin-treated animal was >64 μg/mL after 72 h in vitro culture.
UP RNA was detected in the fetal lung, skin, and chorioamnion of all saline solution–treated control animals ( Table 2 ). All erythromycin-treated animals were positive for UP RNA in the fetal lung. All erythromycin- and saline solution–treated animals were negative for UP RNA in the fetal liver. Erythromycin treatment (irrespective of delivery regimen) was associated with a significant reduction in UP RNA ( P < .05) in the fetal skin and lung compared with saline solution–treated control animals. Animals that were treated with single maternal intramuscular injections and repeated intraamniotic erythromycin injections had significantly reduced UP RNA in the chorioamnion, compared with saline solution–treated control animals ( Figure 2 ).
| Group | Positive U parvum serovar 3 polymerase chain reaction, % | ||||
|---|---|---|---|---|---|
| Fetal lung | Fetal skin | Fetal liver | Arterial | Chorioamnion | |
| Intraamniotic U parvum serovar 3 control (n = 8) | 100 (8/8) | 100 (8/8) | 0 (0/8) | 25 (2/8) | 100 (8/8) |
| Intraamniotic + repeated intramuscular erythromycin (n = 10) | 100 (10/10) | 50 (5/10) | 0 (0/10) | 20 (2/10) | 37.5 (3/8) a |
| Intramuscular + repeated intraamniotic erythromycin (n = 10) | 100 (9/9) a | 50 (5/10) | 0 (0/10) | 0 (0/10) | 50 (5/10) |
a U parvum serovar 3 polymerase chain reaction could not be performed on all group animals.
Fetal inflammation
Animals that were treated with single maternal intramuscular and repeated intraamniotic erythromycin injections had significantly reduced ( P < .05) levels of fetal lung and skin interleukin-8, chorioamnion interleukin-1β and -6, and tumor necrosis factor-α messenger RNA (mRNA), relative to saline solution–treated control animals ( Figure 3 ). Compared with saline solution–treated control animals, there was a significant ( P < .05) reduction in chorioamnion tumor necrosis factor-α, but no significant difference in fetal lung or skin cytokine or chemokine mRNA levels in animals that were treated with single intraamniotic and repeated maternal intramuscular erythromycin injections. There was no significant difference in liver cytokine or chemokine mRNA levels between groups. Levels of liver serum amyloid A mRNA were reduced in erythromycin-treated animals (irrespective of delivery regimen), compared with saline solution–treated control animals ( Figure 4 ).
Histologic findings
There was no significant difference in qualitative airspace inflammation scoring between groups ( Figure 5 ). Animals from erythromycin treatment and saline solution control groups had, on average, airspace inflammatory cells and between 1 and 4 micro-consolidation foci in each of the 6 right upper lobe fields that were examined for each animal.
