Cold Mechanical Isolation of Placental Macrophages as a Method to Limit Procedure-Induced Activation of Macrophages

Isolation of macrophages is a strategy that is widely used to study their phenotype and functionalities in vitro. At present, no consensus has been reached on the optimal isolation method. Several isolation procedures have been proposed to isolate macrophages from the placenta. These protocols differ in terms of phenotype of isolated macrophages, their origin (e.g., fetal or maternal), tissue sampling technique, the trimester during which the tissue is processed, digestion and dissociation procedures, and FACS gating strategies. Single-cell suspensions are usually obtained through enzymatic digestion that requires elevated temperatures to maximize enzymatic efficacy, or through mechanical dissociation that is typically performed at 4°C. To our knowledge, most protocols concerning isolation of placental cells thus far have relied on the use of enzymatic digestion (1–4).

During the past decade, new technological developments have rapidly increased the scientific possibilities to study and understand cellular heterogeneity. For approaches such as single-cell analysis, considerable efforts have been made to develop new isolation protocols that optimize cell yield, purity, and viability. Enzymatic digestion with proteolytic enzymes has become a well-established isolation method as there is common consensus that this improves digestion abilities and thus results in optimal cell yield and viability. To approximate the transcriptional profiles of cells in vivo it is of great importance to minimize the potential influence of the isolation procedures on the transcriptional and phenotypic states of cells ex vivo (5). Total tissue needs to be dissociated to obtain single-cell suspensions for downstream analysis (Fig. 1), which is inevitably associated with exposure to nonphysiological conditions that can cause cell stress, injury, excitotoxicity, and death. Although this occurs in both warm enzymatic digestion and cold mechanical dissociation, it has been proposed that maintaining cells at 4°C leads to metabolic inactivation, thereby enhancing the preservation of transcriptomic and proteomic characteristics (6).

In brain studies, it has been well established that warm enzymatic digestion procedures can induce artificial transcriptional perturbations, resulting in aberrant ex vivo gene expression signatures of brain cells (5–9). These transcriptional changes were most outspoken in microglia, the tissue-resident macrophages of the CNS, and other brain myeloid cells (5). More specifically, enzymatic digestion at 37°C induced deregulations in gene and protein expression, indicative of microglial immune activation at both the transcriptomic and translational levels (6). The genes that were upregulated after enzymatic digestion were mainly immediate early genes (IEGs), immune-signaling genes, and stress-induced genes (i.e., heat shock proteins [HSPs]) (5). In flow cytometry studies, heat- and enzyme-induced alterations in the percentage of isolated live cells and macrophage surface markers have been described as well. It has been suggested that proteolytic cleavage of cell surface receptors can occur with enzymatic digestion, which can cause misinterpretation of flow cytometry data. Moreover, changes in cytokine and chemokine levels were confirmed to be induced by warm enzymatic digestion, as these were significantly different from the in vivo state (5). Besides microglia, immune cell subsets such as T and B lymphocytes, NK cells, and macrophages isolated from the mouse kidney have shown alterations in their activation status after enzymatic digestion at high temperatures, as this resulted in higher expression of IEGs and HSPs compared with dissociation on ice with a cold-active protease (10). Although these studies were not able to dissect which of the changes were due to the thermal shock and which were caused by enzymatic digestion, it is unequivocal that the observed alterations likely introduce unwanted biological bias.

To understand the function of macrophages in pregnancy and its complications, it is important to use a method for placental macrophage isolation that limits changes to the phenotypical and functional characteristics. In addition, minimizing methodological bias avoids biological misinterpretation and contributes to enhanced reproducibility. We therefore developed a protocol that eliminates the enzymatic digestion step, which makes incubation at 37°C unnecessary. By using flow cytometry, FACS, and quantitative real-time PCR, we examined whether our method, based on cold mechanical isolation, resulted in sufficient immune cell yield and viability while limiting immune cell activation.

Term placentas (n = 10) were collected following scheduled primary cesarean sections at the University Medical Center Groningen (Groningen, the Netherlands). Placentas were included from healthy pregnancies without maternal or fetal disease, with maternal age between 18 and 40 y old, and gestational age at birth 36–42 wk of age. Tissue was collected anonymously and used according to the Code of Conduct for Responsible Use following the guideline from the Federation of Medical Scientific Associations with approval of the Medical Ethics Review Committee.

Placental samples were processed within 30 min after primary cesarean section. Three biopsies of 0.5 × 0.5 cm were taken from placenta parenchyma and were stored in RNAlater at the start of the experiment, after 45 min on ice (cold mechanic protocol) and after 45 in a 37°C water bath (warm enzymatic protocol). For the isolation of macrophages, biopsies of parenchyma with decidua were taken from the middle and from each quadrant of the placenta and stored in cold PBS. Biopsies for the warm enzymatic and cold mechanical protocol were taken from the same placenta to correct for biological variances between placentas. Decidual and villous tissues were manually separated on ice using scissors, and the decidua was scraped clean from villi using a scalpel. For the warm enzymatic and cold mechanical protocol, equal amounts of decidual and villous tissues were used.

Table I describes the reagents that were used. All reagents were free of endotoxins as stated in the product specification and certificate of analysis provided by the manufacturing company. Separated decidual and villous tissues were minced and washed twice with calcium- and magnesium-free PBS by centrifugation (1000 rpm [=205 G], acceleration [acc]: 9, brake: 9, 4°C). Tissue was then mechanically dissociated using a gentleMACS either with cold HBSS with phenol red containing 0.6% glucose and 15 mM HEPES or with Accutase (warm protocol). The gentleMACS was located in a room at 4°C. For villous tissue, the preprogrammed program m-heart 02.01 was used, and for basalis, program m-heart 02.01 was followed by m-heart 02.02. For the warm enzymatic protocol, decidual and villous tissue was put into C-tubes (Miltenyi Biotec) with Accutase (1:2) in a 37°C shaking water bath for 45 min. To obtain a single-cell suspension, tissue was filtered over a 300-μm sieve followed by a 106-μm sieve, and the filtrate was collected for both procedures. For the rest of the protocol, the suspensions for the warm enzymatic and cold mechanical protocol were both further processed on ice. Cells were pelleted at 300 × g for 10 min (acc: 9, brake: 9, Hettich Rotina 420R, 4°C), resuspended in HBSS with phenol red containing 0.6% glucose and 15 mM HEPES, and loaded onto a Ficoll gradient. After 45 min of centrifugation at 800 × g (acc: 4, brake: 0, 4°C), a clear cell layer at the Ficoll/HBSS interface was observed, with a distinct separation of the RBCs in the tip of the tube. Cells from the Ficoll/HBSS interface were pipetted into a new tube, washed with HBSS with phenol red containing 0.6% glucose and 15 mM HEPES, and pelleted by centrifugation at 300 × g for 10 min (acc: 9, brake: 9, 4°C). See Fig. 2 for a schematic outline of the procedure.

To recover macrophages and determine purity and viability, we performed flow cytometry and FACS. Cell pellets were washed and resuspended in HBSS without phenol red containing 0.6% glucose, 15 mM HEPES, and 1 mM EDTA and incubated on ice with FcR blocking reagent (1:10) for 10 min, after which cells were stained with monoclonal anti-CD45, CD14, CD80, CD86, CD163, and CD206 Abs (5 μl per sample) (see Table II) in HBSS without phenol red containing 0.6% glucose, 15 mM HEPES, and 1 mM EDTA for 30 min on ice in the dark. To check for viability, 1 μl of propidium iodide was added to the flow cytometry samples and 1 μl of DAPI to the FACS samples. The gating strategy can be found in Fig. 3. For flow cytometry (NovoCyte Quanteon, Agilent Technologies), stained cells were analyzed using NovoExpress flow cytometry software. With FACS (Sony SH800S cell sorter) CD45⁺CD14⁺DAPI⁻ cells were collected for downstream analysis.

From both the placental biopsies and FACS-sorted macrophages, RNA was extracted using a RNeasy Micro kit (Qiagen) according to the manufacturer’s instructions. Total RNA concentrations were measured with NanoDrop (ND1000). cDNA was synthesized using RevertAid reverse transcriptase and RiboLock RNase inhibitor (all Fermentas, St. Leon-Rot, Germany). The PCR reaction contained iQ SYBR Green supermix (Bio-Rad) and was performed on a Thermo Fisher Scientific QuantStudio 7 flex. Primers were designed with National Center for Biotechnology Information Primer-BLAST software and ordered from Biolegio (Nijmegen, the Netherlands). Results were explored using QuantStudio real-time PCR software. Primers that were used can be found in Table III.

Statistical analyses were performed using IBM SPSS Statistics (version 26) and GraphPad Prism (version 9.4.1). As biopsies for the warm enzymatic and cold mechanical protocol were taken from the same placenta, the samples were considered matched and all statistical tests were performed pairwise. For comparison of normally distributed gene expression levels between 4 and 37°C, a paired samples t test was performed. For comparison of cell counts and median marker expression with flow cytometry, a Wilcoxon signed-rank test was performed because data were not normally distributed. All boxplots show median + interquartile range (IQR), and whiskers indicate the 5th to 95th percentiles. Values of p < 0.05 were considered significant.

To determine whether cell populations differed between the cold mechanical and warm enzymatic digestion protocols, we compared single-cell suspensions obtained by both procedures (Figs. 1–3, Tables I–III). The absolute cell count of total cells was significantly higher with the warm enzymatic procedure compared with the cold mechanical protocol for both chorionic villi and decidua basalis (p = 0.013 and p = 0.007, respectively) (Fig. 4). The significant differences observed are mainly driven by the proportion of granulocytes as determined by their properties on the forward/side scatter plot. When comparing cell populations, we isolated significantly more granulocytes within the warm enzymatic procedure for both villous and decidual tissues (Fig. 5A). In contrast, a significantly higher proportion of leukocytes and macrophages was obtained from the villi with the cold mechanical procedure; for the basalis, there were no significant differences (Fig. 5B, 5C).

The percentage of singlets was similar in both protocols. Similarly, the percentage of viable cells did not significantly differ for villous tissue. For basalis, a significant higher percentage of viable cells was isolated with the warm enzymatic procedure compared with the cold mechanical procedure (p = 0.028). Our cold mechanical protocol yielded a single-cell suspension with ∼93% viability in villous tissue as determined by live/dead staining with flow cytometry. For basalis, the mean viability was ∼87% (Fig. 6).

Next, we studied phenotypical differences between macrophages isolated with the warm enzymatic and cold mechanical protocol. Macrophages were defined based on the expression of both CD45 (common leukocyte Ag) and CD14 (myelomonocyte lineage marker). In villous tissue, we observed a slightly higher median expression of CD45 on cells isolated with the cold mechanical protocol compared with the warm enzymatic protocol. In the cells from the basalis no differences were observed (Fig. 7A, 7B). There were no significant differences in expression of CD14 (Fig. 7C, 7D).

Frequently used surface markers for the proinflammatory response of macrophages are CD80 and CD86, surface receptors that regulate T cell activation (11). In our study, no significant differences were observed in the expression of CD80 and CD86 between the cold mechanical or warm enzymatic protocol. On the other side of the spectrum are CD163 and CD206 that are expressed on macrophages with a more anti-inflammatory character (12). We observed a significantly increased expression of CD163 in the CD163^high and CD163^low population in both villous and decidual macrophages in the cold mechanical procedure compared with the warm enzymatic procedure (Fig. 8). There were no differences observed in the expression of the macrophage mannose receptor (CD206) in both protocols. In summary, our data show that cold mechanical dissociation yields macrophages that express higher levels of CD163 compared with warm enzymatic digestion, likely indicative of a more anti-inflammatory type.

To determine whether the two isolation protocols resulted in a different activation status of decidual and villous macrophages, we measured gene expression of cytokines, IEGs, and HSPs expressed by isolated macrophages. In both decidual and villous macrophages isolated with warm enzymatic digestion, there was a significantly higher expression of TNFα (p = 0.013), IL1β (p = 0.005), and IL10 (p = 0.005) (Fig. 9A, 9B, 9D). For TGFβ1 no significant differences were observed (Fig. 9C) (13). In villous macrophages, the expression of FOS, an IEG, was significantly higher in the cold mechanical protocol (p = 0.047). This was not observed in macrophages isolated from decidual tissue (Fig. 9E). There was no difference in the expression of JUN. Another group of factors that are known to be induced by hyperthermia are HSPs (14). Expression of HSP27 remained stable regardless of the isolation procedure. For HSP70, there was a significant induction in the warm enzymatic procedure for both villous and decidual macrophages (p = 0.012 and p = 0.009, respectively) (Fig. 9G, 9H). To conclude, the warm enzymatic protocol resulted in more macrophage activation compared with the cold mechanical protocol, shown by upregulation of proinflammatory gene expression.

To assess the effects of temperature on total tissue, we took three biopsies from every placenta. The first biopsy served as a control, the second biopsy was stored on ice for 45 min and the third biopsy was placed in a warm water bath at 37°C for 45 min. Similar to the macrophage populations, we measured gene expression of cytokines, IEGs, and HSPs. There were no significant differences in gene-expression between the biopsy stored on ice and the biopsy stored at 37°C as determined by quantitative real-time PCR (Supplemental Fig. 1).

We demonstrate that warm enzymatic digestion results in higher gene expression of proinflammatory cytokines, IEGs, and HSPs in isolated placental macrophages compared with cold mechanical dissociation. Based on these findings, we propose a protocol based on cold, mechanical dissociation to isolate macrophages from the placenta, both from the decidua and chorionic villi. With this protocol, we were able to preserve yield and viability of the macrophage population while limiting immune cell activation in comparison with warm, enzymatic isolation.

A reason to use enzymatic digestion is to preserve yield and viability of the isolated macrophages (1, 3). As expected, the absolute count of isolated cells was higher using enzymatic digestion at warm temperatures. These higher cell numbers were mainly driven by granulocytes, as the number of isolated granulocytes was significantly higher with the warm enzymatic procedure. Indeed, the relative proportion of villous leukocytes and macrophages was higher in the cold mechanical procedure for villous tissue and did not differ significantly for decidual tissue. Differences between decidual and villous macrophages are possibly due to the tissue structure, as we experienced that decidual tissue was more difficult to homogenize. Altogether, our data suggest that different immune cell populations are more resistant to different isolation techniques. Most importantly, we show that cold mechanical isolation does not compromise the yield of placental macrophages.

An important reason to opt for cold mechanical isolation is the activation of immune cells that is caused by warm enzymatic digestion. Albeit both cold mechanical dissociation and warm enzymatic digestion will result in some transcriptomic alterations in comparison with their in situ profile, it has been shown that a “heat shock” as experienced by cells in enzymatic digestion procedures can induce elaborate transcriptional aberrations within a short window of time (15). Furthermore, the expression of cell surface markers may be altered because of mechanical stress and enzymatic trimming, respectively. Cell surface marker expression might even be further modified after enzymatic digestion when cells are stored at physiological temperatures, as this keeps cells in their metabolically active state, leading to endocytosis and degradation of surface proteins (6, 15).

For downstream analysis undeviating surface marker expression is desirable. In our study, only mild effects were observed on protein level. This was particularly pronounced for CD163, where CD163 expression was higher in macrophages isolated at 4°C compared with macrophages isolated at 37°C with enzymatic digestion. Although we studied cell surface markers that are most commonly used to phenotypically define macrophage subsets with flow cytometry, we note that there could be effects on cell surface proteins that belong to other macrophage subsets.

Larger effects were observed for the gene expression levels of cytokines, IEGs, and HSPs, which were all increased with warm, enzymatic digestion. We hypothesize that the effects of warm temperatures and the use of digestive enzymes are not large enough to cause major differences in cell surface markers and hence observe phenotypical changes.

However, we do show that warm enzymatic digestion is significant enough to induce more rapid inflammatory changes in terms of cytokine, IEG, and HSP expression differences. Macrophages are the body’s first line of defense and are thus fast responders to inflammatory signals. Compared to cold mechanical isolation, we observed an increased expression of TNFα, IL1β, and IL10 in both decidual and villous macrophages using the warm enzymatic protocol. Moreover, we found HSP70, a protein that responds quickly to inflammation, to be increased. In macrophages, HSP70 is a known for its importance in the mediation of thermal effects on their function (16). Interestingly, another commonly used IEG, FOS, was found to be increased in villous macrophages in the cold mechanical procedure. In pregnancy, FOS is a transcription factor involved in the control of trophoblast invasion in the placenta (17). Why FOS expression was increased in the cold mechanical procedure remains to be unresolved and further research is needed.

Our results show that warm enzymatic digestion is associated with a stronger inflammatory response in placental macrophages compared with cold mechanical dissociation. Consequently, the transcriptional profile of macrophages isolated with warm enzymatic digestion could mistakenly be interpreted as “proinflammatory,” whereas in fact it could be artifactual due the isolation method.

It is noteworthy that all methodologies to obtain a single-cell suspension from total tissue inevitably cause transcriptional changes to some extent, as cells are undergoing tissue dissociation and cell separation steps to get to the population of interest (Fig. 1). During this procedure, cells will be exposed to nonphysiological conditions that can cause cell stress, injury, excitotoxicity, and death. Based on previous studies and our results, it is likely that cold, mechanical isolation induces less activation of placental macrophage populations compared with warm enzymatic digestion due to metabolic inactivation. Our findings are strengthened by the fact that we compared both protocols on tissue biopsies from the same placenta, thereby eliminating the factor of interplacental biological differences that undeniably influence macrophage characteristics. Although we did not compare the isolated macrophages to in vivo macrophage populations, our results are in line with other studies showing that isolated immune cell populations using warm enzymatic digestion substantially differ from their in vivo state (5, 6, 10). This strongly suggests that the macrophage population that we obtained through cold mechanical isolation more closely resembles their true nature.

To further analyze the effects of temperature, we compared gene expression of cytokines, IEGs, and HSPs between placental biopsies taken immediately after delivery and biopsies that were kept on ice or incubated in 37°C to determine the effect of temperature on placental tissue. The transcriptional changes we found in isolated macrophages were not observed in the placental tissue biopsies incubated at 4 and 37°C. A reasonable explanation for this might be that the percentage of cells that respond to thermal changes in the biopsy is not high enough to display an effect in total tissue. Biopsies are comprised of many other cell types besides macrophages, such as fibroblasts, endothelial cells, and trophoblasts, that are not expected to express the inflammatory genes assessed in the current study (Fig. 1). Another explanation could be that the response of macrophages is relatively faster compared with other placental immune cell subsets, resulting in only small changes in cytokines, IEGs, and HSPs gene expression in the total placental tissue. In addition, we did not use enzymatic or mechanical dissociation for the placental biopsy tissues, and only temperature differed. Our results are in line with the study of Marsh et al. (5) who showed that enzymatic dissociation of brain tissue only resulted in a stress response in microglia and not in other brain-associated cell types.

Recently, much research has been conducted on the role of macrophages in pregnancy complications. An imbalance in proinflammatory and anti-inflammatory macrophages has been found to be involved in pregnancy complications such as recurrent spontaneous pregnancy loss, spontaneous preterm birth, preeclampsia, fetal growth restriction, and stillbirth. Numerous reviews have outlined how increased secretion of proinflammatory cytokines from macrophages in the placenta impact gestational processes such as implantation, trophoblast migration and invasion, and angiogenesis (1, 18–22). It is therefore of crucial importance to develop protocols that can obtain macrophage populations that closely resemble their in vivo state to avoid erroneous interpretation of study results. Moreover, comparing results between studies is less reliable when different enzymes or temperatures are used because this can lead to differences in the degree of macrophage activation. This is even more important when moving toward the development of immune-based therapies for pregnancy complications in the future.

In conclusion, we developed a protocol for the isolation of macrophages based on mechanical dissociation at low temperatures and showed that this induces less activation of decidual and villous macrophages compared with warm enzymatic digestion. We believe that this method leads to more reliable results in downstream analysis and therefore significantly improves future studies on the role of placental macrophages in healthy and complicated pregnancies.

The authors have no financial conflicts of interest.

We extend gratitude to the operators of the University Medical Center Groningen flow cytometric unit, H.J. Teunis and T. Bijma for technical support, and to E. Wesseling for contributions to the quantitative real-time PCR analysis.