Pregnant human peripheral leukocyte migration during several late pregnancy clinical conditions: a cross-sectional observational study

Pregnant women were recruited from the Royal Alexandra Hospital in Edmonton, Alberta with informed, written consent prior to inclusion in the study. Specific inclusion criteria were used to place women into six study groups for blood sampling: sTL (spontaneous normal labour delivered vaginally at term, 37–42 wk, n = 24), TNL (elective caesarean section at term without labour, n = 16), PTL (preterm in labour, 22–36 wk, n = 10), PTNL (normal women sampled preterm but not in labour, 22–36 wk, n = 16), TPTL (threatened preterm labour as identified by uterine contractions or cervical dilation, 22–36 wk, n = 11) and pPROM (preterm with premature rupture of membranes but without contractions or dilated cervix, 22–36 wk, n = 8). Patients who had disease or medication, multiple pregnancy, fetal congenital disease, and clinical signs of infection (body temperature over 38 °C) were excluded from this study.

The extracted chemoattractant was used in the migration assay. Fetal membranes were collected from fifteen women who underwent term spontaneous labour without complications. Exclusion criteria were the same as for leukocyte collection. The preparation of the chemoattractant was as published with minor modifications [9, 10, 14, 15]. From each of the fifteen sets of fetal membranes a 4cm square piece was cut. Each of these fifteen square pieces were then washed with 1× phosphate buffered saline (PBS) for 3 times to remove blood and debris, and homogenized with 4mL of Dulbecco’s Modified Eagle’s Medium (DMEM) (Thermo Fischer Scientific Inc., Ottawa, ON, Canada). Peptidase inhibitors were omitted as tests including them demonstrated they had no effect upon the chemoattraction potential of the final preparation. Following centrifugation at 3,000 × g for 30 min at 4 °C and then 20,000 × g for 2 h at 4 °C (Thermo Scientific™ Sorvall™ ST 16R, Thermo Fischer Scientific Inc., Ottawa, ON, Canada),the supernatants from each piece were collected and pooled together. Protein concentrations (BCA method) were adjusted to 4μg/μL with DMEM. Pooled chemoattractant extracts were aliquoted and stored at−80 °C. For each experiment a vial(s) was placed on ice to thaw. All experiments in this study used vials prepared and frozen from the same batch of chemoattractant and were performed within one year of the original preparation. There was no change in the activity of chemoattractant in that time and the chemoattractant performed similarly to batches prepared at other times (data not shown).

Leukocytes were prepared as published with minor modifications [9, 10]. Peripheral blood samples were collected by venipuncture upon recruitment into the study and granting of consent using a standardized protocol for each subject in each of the groups. Leukocytes present in peripheral maternal blood samples drawn into 6mL heparinized tubes were isolated and used in the LMA. Five mL of the anticoagulated blood were combined with 1mL HetaSep (Stemcell, Vancouver, BC, Canada) to remove erythrocytes through sedimentation. Samples were placed in a humidified incubator at 37 °C for 10 min to allow sedimentation of erythrocytes. Approximately 3mL of the top, leukocyte-rich plasma layer were collected and washed with four-fold of 1× PBS. Leukocytes were sedimented using gentle centrifugation (120 × g for 10 min at 20 °C without the brake). The supernatant was discarded and leukocytes resuspended in 4mL Hyclone™ Roswell Park Memorial Institute 1640 medium (RPMI) (Thermo Fischer Scientific Inc., Ottawa, ON, Canada) containing 2.0mM L-glutamine. A Bright Line™ hemacytometer (Sigma-Aldrich, St. Louis, MO, USA) was used to count leukocytes. The number of dead leukocytes were recorded using Trypan blue method and the suspension mixture was only used with a viability rate >95%. The leukocyte suspension was diluted using RPMI to a final concentration of 1x10⁵ cells/50μL and used in the LMA within an hour of isolation.

The procedure used was published [9, 10, 14, 15] with recent modifications to improve the assay performance. Modified Boyden chemotaxis chambers (AP48; Neuro Probe, Gaithersburg, MD, USA) were used in the assay. Twenty-five μL of the chemoattractant extract (100μg total protein) or DMEM as negative control were placed in the lower chamber to create a slightly positive meniscus. A polycarbonate membrane with 3μm pores (Neuro Probe, Gaithersburg, MD, USA) was next placed over the lower chamber followed by a rubber gasket and then the upper chamber. Previously we used a filter with 5μm pores, but we found it allowed too many leukocytes through in the control (blank) tubes resulting in high background counts. We thoroughly tested the system performance with the smaller pores and consequent low blanks (ca. 50–100 cells) and found that the number of cells that migrated was directly dependent upon the amount of chemoattractant placed in the lower chamber and was directly proportional to the number of cells placed in the upper chamber (Fig. 1).

Fifty μL of suspension containing 100,000 leukocytes was pipetted into each well of the upper chamber. The chamber was incubated in a humidified atmosphere at 37 °C containing 5% CO₂ for 90 min. Following incubation, the upper chamber liquid was removed and the lower chamber samples (into which leukocytes migrated) were individually collected in BD Falcon™ 5mL polystyrene round bottom tubes (Thermo Fischer Scientific Inc., Ottawa, ON, Canada) and combined with 300μL OptiLyse (Beckman Coulter, Mississauga, ON, Canada) to lyse red blood cells and fixed leukocytes. For the subsets analysis, antibodies were added to the tube before adding OptiLyse. Samples were incubated in the dark for 15 min and 1mL 1× PBS was added to each tube. Following centrifugation at 1,000 × g for 10 min at 4 °C, the supernatant was discarded and 475μL 1× PBS with 1% formalin was added to fix leukocytes. Lower chamber samples were then kept covered in a 4 °C cold room until quantification using flow cytometry (BD FACSCanto™ II, BD Biosciences, San Jose, CA, USA).

Flow cytometry was used to quantify the number of leukocytes that migrated from the upper chamber down into the lower chamber through the polycarbonate membrane. Each sample had 25μl of CountBright™ Absolute Counting Beads (Life Technologies, Burlington, ON, Canada) added prior to analysis using flow cytometry. The flow cytometer was set to analyze the samples for 30 sec and data were collected using BD FACSDiva software (BD Biosciences, San Jose, CA, USA). Leukocytes were gated based on their forward and side scatter. The total number of leukocytes migrated was calculated using the number of beads added to the tube multiplied by the ratio of leukocytes to beads in the sample aliquot taken by the flow cytometer. The number of leukocytes that migrated in the blank, or negative control, wells containing DMEM was subtracted from all samples in the final calculations.

Leukocyte subtypes that migrated were also analyzed using flow cytometry in a few test cases. Multitest 6-color TBNK Reagent and V450 Mouse Anti-Human CD14 (BD Bioscience, CA, USA) were used for staining cells. Granulocytes were considered as CD45⁺ CD3⁻ CD19⁻ CD16/56⁻ CD14⁻cells.

Data were analyzed and graphs were drawn with GraphPad Prism Version 5.0 software (GraphPad Prism Software, CA, USA). Data were tested for normal distribution and then significance testing was performed with an unpaired t test for two-group comparison. One-way ANOVA followed by Tukey’s multiple comparison post hoc test was used to detect statistically significant differences between the study groups of over two. P < 0.05 was accepted as significant.

This is the first report demonstrating that maternal human peripheral leukocytes increase their migration toward a term fetal membrane chemotactic signal when in labour at term or preterm or in the early stages of labour such as pPROM. The exciting aspect of this migration assay is that it has the potential to differentiate those subjects who will not deliver for several weeks from those subjects experiencing symptoms who will deliver within a week.

The percentages of various leukocytes that migrated in the Boyden chamber to the chemotactic signal vary from their distribution when collected and placed into the upper chamber; more neutrophils appear to be more sensitive in labouring women. This suggests that the migrated leukocytes, especially neutrophils, developed greater responsiveness to the chemotactic substance in human fetal membranes than other cells. This would be consistent with reports that demonstrated higher percentages of neutrophils than were in peripheral blood in late gestation or that invaded gestational tissues at term labour [1‐3, 16, 17]. However those studies also report increasing invasion of monocytes, suggesting that migration and invasion may be differentially regulated. Thus it is important to note that this assay assesses just one aspect of the more complex process of leukocyte invasion of the uterus prior to delivery.

The increasing responsiveness of the leukocytes to the chemoattractant with labour onset suggests that either there are more leukocyte receptors to the chemoattractant, better coupling of receptors with intracellular signal transduction mechanisms and/or an upregulation of the migration apparatus in the cells. In future investigation we intend to determine whether studying migration of neutrophils only will improve the parameters of the study. At present, though, it is clear that total leukocyte migration is an effective outcome measure for assessing proximity to delivery.

We did not assess the chemical nature of the chemoattractant present in the crude homogenates of human fetal membranes. Data (not shown) indicates it is also present in uterine tissues from mice and rats and these rodent chemoattractants are able to attract human leukocytes. We are in the process of identifying it chemically with expert collaborators. Preliminary analysis suggests it is not one of the common chemokines and isolating it has proven difficult. We will continue our analysis using a variety of biochemical isolation techniques. Considerable more work is required before it will be identified.

The receiver-operating characteristic curve utilized all the data available (Fig. 3) to determine preliminary parameters of the assay. While preliminary, these data are highly encouraging in that the positive predictive value, negative predictive value, sensitivity and specificity are very high. They suggest that with further improvement in terms of simplifying the technical aspects so it can become a point of care test, the assay described here may form the basis for a future clinical diagnostic test.

We demonstrated an interesting and potentially clinically applicable characteristic of circulating leukocytes in late gestation women and those in various clinical situations. Further refinement and improvement will both reveal basic biological mechanisms of late gestation in normal and inflammatory situations plus possibly lead to an informative diagnostic test to predict preterm birth risk. One limitation of this study is that it is only cross-sectional with separate subjects in each determination. Future work comprising longitudinal samples from a single subject during pregnancy, at term or preterm, threatened preterm or ruptured membranes would be very informative.

Leukocyte migration has the potential with further development, refinement and simplification to become a diagnostic test that would be valuable for identifying asymptomatic women who may be at risk for preterm birth. In addition, many symptomatic women present with signs of preterm labor but only 10–20% will deliver preterm. It is important from a treatment and also a health care costing perspective to know which of these women to treat or admit to hospital, and which to send home. Existing tests, the fetal fibronectin (fFN) test and the insulin-like growth factor binding protein-1 (IGFBP-1) test, have high negative prediction values, but low positive predictive values [17‐19]. They indicate who will not deliver, but they do not indicate who will deliver within one to two weeks. A test with a high positive predictive value, along with a reasonable negative predictive value, would be a significant advance in the management of pregnant women.