Increased Insulin-like Growth Factor Binding Protein-1 Phosphorylation in Decidualized Stromal Mesenchymal Cells in Human Intrauterine Growth Restriction Placentas

Selection of Study Subjects

Human placenta samples were collected through our ongoing collection for studies on IUGR.^19–22 Pregnant women were recruited at London Health Sciences Centre (LHSC). Samples were collected under approval by the Western University Research Ethics Board. The infant/maternal clinical data were obtained from their medical records. Birth weight percentiles were calculated using standardized gender-specific growth charts. IUGR was defined as having a birth weight below the 10th percentile, while severe IUGR was a birth weight below the 3rd percentile. Four out the five IUGR samples used were from infants with a birth weight <3rd percentile and with the presence of at least one sign of fetal compromise (such as low ponderal index, abnormal umbilical artery Doppler or oligohydramnios) as an additional criterion to minimize the risk of inclusion of constitutionally small fetuses. Mothers of IUGR pregnancies were otherwise healthy. Control subjects selected were women delivering a term infant with a birth weight for AGA defined as 25th to 75th percentile. Gestational age (GA) was determined by last menstrual period. Healthy controls were selected to match as closely as possible the GA range. We compared IUGR placenta (33–37 weeks gestation) matched with AGA controls (n=5 each). The samples selected for this cohort were matched for inclusion and exclusion criteria briefly as follows: Women with singleton pregnancies were included. Pregnancies with congenital abnormalities, diabetes, thyroid disorders, chronic hypertensive disorders or pre-eclampsia, premature rupture of membranes, placental abruption, fetal congenital and genetic abnormalities, chorioamnionitis, smokers, drug abuse, and malnutrition were excluded, as in our previous studies.^19–22

Tissue Collection

Placentas from IUGR and AGA pregnancies (n=5 each) were randomly selected from both vaginal and/or cesarean section and collected immediately after delivery. Based on review of the pathological reports (undertaken at the LHSC), all of the placentas were negative for signs of infection, and also contained no major infarcts. The tissue samples were collected and full thickness blocks were obtained. All tissues were sectioned serially (consistent thickness of 5 µm; Biotron Integrated Microscopy Laboratory, The University of Western Ontario [UWO]) and mounted on pretreated slides (Superfrost Plus, Fisher Scientific, Fairlawn, NJ). The selection of regions within each block of the placenta was based on the criteria that ensured incorporating all levels/regions from chorionic plate to basal plate. GA-matched AGA and IUGR placenta tissues (n=5 each) were placed side-by-side on a slide and labeled as shown in Fig. 2A. Six representative images from each of the placental tissues (n=5 each) were captured (Fig. 2A, red squares and image represented in Fig. 2B) and used for dual immunofluorescence staining and image analysis as described further.

OPEN IN VIEWER

Figure 2.

Flowchart showing experimental design and methodology for dual immunofluorescent staining and image analysis. Image Pro Premier software-based quantification, and statistical analysis. IUGR and matched AGA placenta (n=5) were stained simultaneously for vimentin, plus total, or phosphorylated (Ser 101, Ser119, or Ser169) IGFBP-1. Representative images (n=6 regions/sample-red outlines as shown) were captured as multichannel images from the BPD of each placenta tissue (n=5), using standardized microscope settings (A and B). A manual ROI was drawn to isolate the decidualized cells within the basal plate (C). Within this ROI on the Vimentin channel (Alexa 660-shown green), vimentin positive mesenchymal cells were counted to create cell outlines (C, D). Cell outlines from (C) were converted into multicell ROIs and applied onto the raw channel image from the matching total or phosphorylated (Ser 101, Ser119, and Ser169) IGFBP-1 image—Alexa 568-shown in (E) grayscale for better viewing. Within the cellular ROIs, high expressing cells only were counted, using IGFBP-1 signal itself to isolate truly decidualized cells (blue outlines, red arrows) from nondecidualized cells—green outlines, yellow arrows (F). A cell-sorted image was created to illustrate the intensity differences in signal between the IUGR and AGA. Matching cells dotted red arrows (between E and G). Data were generated from within the blue outlined, truly decidualized cell population in (E). Scale bar = 100 µM. Abbreviations: IUGR, intrauterine growth restriction; AGA, appropriate for gestational age; IGFBP-1, insulin-like growth factor binding protein-1; BPD, basal plate deciduas; ROI, region of interest; GA, gestational age.

Sample Preparation

A stratified random sampling procedure was used. The decidua region was identified as the superficial layer on the maternal facing side of the placenta, to ensure regions were inclusive of decidual (basal plate decidua [BPD]) tissue for IHC as confirmed based on morphology through H&E staining. The sections were treated according to the standard immunofluorescent technique for detection of IGFBP-1 and vimentin.

Antibodies

The following primary antibodies were used: IGFBP-1 mAb 6303 (Catalog #0027301; Medix Biochemica, Kauniainen, Finland), mAb vimentin (Catalog #M7020; Dako, Burlington, Ontario, Canada). Pre-immune serum used as negative controls were Dako Mouse Immunoglobin and Dako Rabbit Immunoglobin Fraction (Normal) each. We used our prevalidated custom made IGFBP-1 polyclonal antibodies targeting pSer101, pSer119, and pSer169^13,23 (generated at YenZyme Antibodies LLC, San Francisco, CA). Secondary antibodies were antirabbit Alexa 568 (Catalog #A11036) and antimouse Alexa 660 (Catalog #A21055; ThermoFisher, Waltham, MA).

Immunohistochemistry

The tissue sections were incubated for 10 min in 3% H₂O₂ to quench endogenous peroxidase. After a 2 × 5 min PBS wash, endogenous Biotin was blocked using an Avidin-Biotin blocking kit. Briefly, avidin solution was applied for 10 min, rinsed in PBS, followed by biotin solution for 10 min, to prevent reagent avidin molecules from adhering to the endogenous biotin known to be present in placental tissues. After a brief rinse with PBS, tissues were blocked using Dako Background Sniper (Biocare Medical, Concord, CA) for 10 min. They were rinsed with PBS and IGFBP-1 (mAb 6303) as primary antibody (1:10,000; 100 µL) was applied for overnight incubation at 4C, using pre-immune serum as a negative control. All incubations were performed in a humidified chamber. Next, Vectastain Elite kit biotinylated secondary antibody matching the primary host (Vector Laboratories, Burlington, ON, Canada; 1:200) was applied for 45 min at room temperature followed by Vectastain Elite Avidin-biotin peroxidase complex for a further 45 min. Next, 3,3′-Diaminobenzidine (DAB) staining was performed for 2 min per slide with Cardassian DAB (Biocare Medical, Concord, CA). For counterstaining, tissues sections were stained with methyl green (Vector Laboratories, Burlington, ON, Canada) preheated to 60C for 5 min and the slides were dehydrated, cleared in xylene and mounted with Permount (Fisher Scientific, Burlington, ON, Canada).

Dual Immunofluorescence

After deparaffinization and following blocking and a PBS wash, the slides were incubated with a mixed 120 µL solution of primary antibodies; vimentin plus IGFBP-1 antibody or vimentin plus individual phospho-site-specific IGFBP-1 antibodies (pSer101, pSer119, and pSer169) for 1 hr. The dilutions for the antibodies were Vimentin (1:25), rabbit polyclonal IGFBP-1 (1:10,000), and each of the phospho-site-specific IGFBP-1 antibodies (pSer101, pSer119, and pSer169; 1:1000). Pre-immune serum was used as a negative control. Combinational secondary antibodies (100 µL), Alexa 568 (polyclonal antirabbit) and Alexa 660 (monoclonal antimouse; 1:400) were added and incubated at room temperature for 1 hr. The tissues were then counter-stained with 100 µL of 4′,6-Diamidino-2-Phenylindole, Dihydrochloride (DAPI; Life Technologies, Burlington, ON, Canada) for 2 min with a 1:200 dilution, rinsed in PBS, mounted with Prolong Gold Mounting Media (Life Technologies, Burlington, ON, Canada) and sealed with nail polish.

Selection of Placental Regions and Image Standardization and Acquisition

AxioImager Z1 Epifluorescent Microscope and Zen Pro software (Carl Zeiss Canada Ltd, Toronto, ON, Canada) was utilized to capture images using a 20 × objective. The fluorescent lamp was preheated for 15 min prior to imaging to stabilize lamp intensity. For image uniformity, camera settings were then set to avoid saturation and autofluorescent background, and consistently applied at the same settings for all imaging.

Representative images (n=6) from each of the dually stained placenta tissues from AGA and IUGR group (n=5 each; represented in Fig. 2A), stained with vimentin plus one of total IGFBP-1 or IGFBP-1 phosphorylation at Ser101, Ser119 and Ser169 (Fig. 2A and B) were acquired within the BPD regions (Fig. 2A, red squares, n=6 each, and image represented in Fig. 2B). This allowed coverage of large areas for each placenta, which subsequently provided an adequately large number of cells/image to represent per cell data to quantitate our results.

Quantitative Image Pro Software Analysis

Data Presentation and Statistical Analysis

Statistics were performed using GraphPad Prism 5 (Graph Pad Software Inc., CA). Dual immunofluorescent images captured from six different regions of each placental tissue section from AGA and IUGR were used to quantify the mean intensity for each placenta. We then calculated the average of the mean intensities across the samples (Fig. 2A) and compared the intensities of total and phosphorylated IGFBP-1 (pSer101/pSer119/pSer169), respectively, in AGA and IUGR placental tissues (n=5 each). The mean intensity of the control (AGA) was assigned an arbitrary value of 100. All mean values for IUGR were expressed relative to this mean. To compare means, unpaired Student’s t-test was performed, and the results were expressed as mean ± SEM with n=5 samples. Significance was accepted at p<0.05.

Clinical Characteristics

The clinical characteristics of the AGA and IUGR pregnancies included in this study are as follows: The mean maternal age in the AGA group was 36 ± 4.58 compared with 31.80 ± 2.78 years in the IUGR group, with no significant differences. Mean GAs were similar, with 34.8 ± 0.97 and 35.6 ± 0.68 weeks in AGA and IUGR, respectively. Cesarean section and vaginal deliveries were included in both groups. The mean placental weight was 449 ± 66.87 and 386 ± 36.01 grams in the AGA and IUGR groups, respectively. The mean birth weight for IUGR infants (1812 ± 194 grams) was significantly lower than AGA (2323 ± 28 grams) and so were the birth weight percentiles, which were 28.75 and 2.80 for AGA and IUGR, respectively (p<0.05). The other clinical data did not show significant differences between study groups. All IUGR newborns (n=5) were confirmed to be growth restricted with birth weight <3rd percentile except for one at 7th percentile.

Relatively Higher Abundance of IGFBP-1 in Decidual Cells of Placenta

We first aimed to immunohistochemically (Biotin-Avidin staining) localize decidual cells in the placenta tissue sections (Fig. 1). The decidua basalis or the basal plate, which is the maternal part of the placenta, forms an important contact between maternal and fetal tissue where the decidual marker IGFBP-1 is localized. ¹⁵ We located the basal plate region in tissue sections based on overall morphology and vimentin positive staining. Maternal decidual cells of the basal plate region, which were identified by immunoreactivity to vimentin (not shown), expressed IGFBP-1 in high abundance (Fig. 1B–D, Arrow-1). Based on the signal intensity using equal exposure criteria, as expected, IGFBP-1 was expressed in most decidual cells of the basal plate but in variable abundance (Fig. 1B and D). The maternal side of villus parenchyma includes a thin basal plate corresponding to the maternofetal junction made up of trophoblasts, stromal, and fibroblast-like cells. Most of the staining for IGFBP-1 was observed in stromal-like decidual cells (Fig. 1B and D, Arrow-1), however, unexpectedly, there was also staining for IGFBP-1 observed in extravillous trophoblast cells (Fig. 2A–D, Arrow-2) possibly due to secreted IGFBP-1. No significant IGFBP-1 staining was observed in the villous trophoblasts or in negative controls (without any primary antibody or with pre-immune serum, not shown). IGFBP-1 was also largely absent in the cells of fetal origin, such as the fetal capillary endothelium, therefore, we concluded that the expression of IGFBP-1 was characteristically limited to the basal plate region of the decidua. Utilizing GA-matched AGA and IUGR placental samples, the relative expression of IGFBP-1 in decidual cells, shown in Fig. 2, indicates relatively more abundant IGFBP-1 expression in IUGR (Fig. 1B and D) compared with AGA placenta (Fig. 1A and C).

IGFBP-1 Expression in Decidual Mesenchymal Cells Is Relatively Higher in IUGR Compared with AGA

Next, to quantify IGFBP-1 expression only in a subpopulation of truly decidualized mesenchymal cells of the placenta, we utilized a dual immunofluorescence approach combined with Image Pro software analysis. IGFBP-1 expression (merged yellow, Fig. 3A; Vimentin, pseudo-green/Alexa 660, Fig. 3B; and IGFBP-1, pseudo-red/Alexa 568, Fig. 3C) was assessed in a subpopulation of truly decidualized mesenchymal cells (Fig. 3A–L). Based on Image Pro software analysis as described in Materials and Methods and in flowchart (Fig. 2 and Supplemental Fig. 1), IGFBP-1 channel (Alexa 568, Fig. 3D and H and grayscale, Fig. 3I and J) shows the expression of IGFBP-1 within the outlines created in the vimentin channel (Alexa 660). The underlying IGFBP-1 signals cropped from the image (Fig. 3I and J) and recreated are shown in Fig. 3K and L in IUGR and AGA samples, respectively. By clearly isolating the signals within the vimentin positive cells, exclusive of overall background signals, overall cell-sorted images show enhanced IGFBP-1 signal intensity in the IUGR (Fig. 3L) versus AGA (Fig. 3K) likely to be in decidua. Accounting for IGFBP-1 only in the subpopulation of decidualized mesenchymal cells when they were removed from surrounding cells/tissue, we minimized the potential for false positive results. Furthermore, exclusion of primary antibody in the negative controls showed that the signal differences between the IUGR and AGA were highly specific to IGFBP-1 and were not due to autofluorescence or nonspecific binding of secondary antibodies (Supplemental Fig. 2).

OPEN IN VIEWER

Figure 3.

Relative dual immunofluorescence detection of total IGFBP-1 levels in AGA and IUGR placental tissues. Representative dual immunofluorescence IHC images of merged channels with IGFBP-1 (pseudo-red), vimentin (pseudo-green), and nuclei/DAPI (blue; A and E). Vimentin (pseudo-green, Alexa 660; B and F), IGFBP-1 (pseudo-red, Alexa 568; C and G). Cell outlines of vimentin positive cells applied to IGFBP-1 channel (Alexa 568; D and H). Alexa 568 image (D and H) also shown in grayscale for better viewing (I and J). Utilizing set thresholds, IGFBP-1 expression was then quantified within cell outlines (from I and J). Cell outlines were converted into multicell ROIs and applied onto the raw channel. A cell-sorted image was created to illustrate the intensity differences in signal between the IUGR and AGA (K and L). A few high IGFBP-1 expressing cells shown by red arrows, representative of the subpopulation of truly decidualized cells. IUGR pregnancies were noted to have higher expression of IGFBP-1 within the decidua of placental tissues compared with GA-matched AGA placenta (n=5 each). Scale bar = 50 µm. Abbreviations: IGFBP-1, insulin-like growth factor binding protein-1; AGA, appropriate for gestational age; IUGR, intrauterine growth restriction; IHC, immunohistochemistry; DAPI, 4′,6-Diamidino-2-Phenylindole, Dihydrochloride; ROI, region of interest; GA, gestational age.

IGFBP-1 Phosphorylation at Ser169, Ser119 and Ser101 in Decidual Mesenchymal Cells Is Markedly Enhanced in IUGR

Similarly, using the same methodology as used with Image Pro based software analysis for total IGFBP-1 (Figs. 2 and 3), we next evaluated IGFBP-1 phosphorylation (Alexa 568) at Ser169 (Fig. 4A–H), Ser 119 (Fig. 5A–I) and Ser101 (Fig. 6A–I) within the vimentin positive (Alexa 660) subpopulation of decidualized mesenchymal cells in IUGR and AGA placenta. Phosphorylated IGFBP-1 channel (Alexa 568; pseudo-red, Figs. 4–6, D and H) and the corresponding grayscale images (Figs. 4 –6, I and J) indicate IGFBP-1 phosphorylation within the outlines created in the vimentin channel (Alexa 660). The underlying IGFBP-1 phosphorylation (Ser 169, Ser119, and Ser101) signals cropped from the respective images and recreated are shown in Figs. 4 –6 (K and L). The data with overall cell-sorted images provide a better visual prospective on changes in IGFBP-1 phosphorylation at all three serine sites in IUGR compared with AGA relative to pseudo colored images. The expression was likely to be mainly in truly decidualized mesenchymal cells.

OPEN IN VIEWER

Figure 4.

Relative dual immunofluorescence detection of phosphorylated IGFBP-1 (pSer169) in AGA and IUGR placenta tissues. Representative dual immunofluorescence IHC images of merged channels of phosphorylated IGFBP-1 (pSer169; pseudo-red), vimentin (pseudo-green), and nuclei/DAPI (blue; A and E). Vimentin (pseudo-green, Alexa 660; B and F), IGFBP-1 (pseudo-red, Alexa 568; C and G). Cell outlines of vimentin positive cells applied to IGFBP-1 channel (D and H). Alexa 568 image (D and H) also shown in grayscale for better viewing (I and J). Utilizing set thresholds, IGFBP-1 (pSer169) expression was then quantified within cell outlines (from I and J). Cell outlines were converted into multicell ROIs and applied onto the raw channel. A cell-sorted image was created to illustrate the intensity differences in signal between the IUGR and AGA (K and L). A few high phosphorylated IGFBP-1 (pSer169) expressing cells shown by red arrows (K and L). IUGR placental tissues exhibited a relatively higher IGFBP-1 phosphorylation (pSer169) compared with GA-matched AGA samples (n=5 each). Scale bar = 50 µm. Abbreviations: IGFBP-1, insulin-like growth factor binding protein-1; AGA, appropriate for gestational age; IUGR, intrauterine growth restriction; IHC, immunohistochemistry; DAPI, 4′,6-Diamidino-2-Phenylindole, Dihydrochloride; ROI, region of interest; GA, gestational age.

OPEN IN VIEWER

Figure 5.

Relative dual immunofluorescence detection and quantitative analysis of phosphorylated IGFBP-1 (pSer119) in AGA and IUGR placenta tissues. Representative dual immunofluorescence IHC images of merged channels of phosphorylated IGFBP-1 (pSer119; pseudo-red), vimentin (pseudo-green), and nuclei/DAPI (blue; A and E). Vimentin (pseudo-green, Alexa 660; B and F), IGFBP-1 (pseudo-red, Alexa 568; C and G). Cell outlines of vimentin positive cells applied to IGFBP-1 channel (D and H). Alexa 568 image (D and H) also shown in grayscale for better viewing (I and J). Utilizing set thresholds, IGFBP-1 phosphorylation (pSer119) was then quantified within cell outlines (from I and J). Cell outlines were converted into multicell ROIs and applied onto the raw channel. A cell-sorted image was created to illustrate the intensity differences in signal between the IUGR and AGA (K and L). A few high expressing phosphorylated IGFBP-1 (pSer119) in IUGR placental tissues exhibited a relatively higher IGFBP-1 phosphorylation (pSer169) compared with GA-matched AGA samples (n=5 each). Scale bar = 50 µm. Abbreviations: IGFBP-1, insulin-like growth factor binding protein-1; AGA, appropriate for gestational age; IUGR, intrauterine growth restriction; IHC, immunohistochemistry; DAPI, 4′,6-Diamidino-2-Phenylindole, Dihydrochloride; ROI, region of interest; GA, gestational age.

OPEN IN VIEWER

Figure 6.

Relative dual immunofluorescence detection and quantitative analysis of phosphorylated IGFBP-1 (pSer101) in AGA and IUGR placenta tissues. Representative dual immunofluorescence IHC images of merged channels of phosphorylated IGFBP-1 (pSer101; pseudo-red), vimentin (pseudo-green), and nuclei/DAPI (blue; A and E). Vimentin (pseudo-green, Alexa 660; B and F), IGFBP-1 (pseudo-red, Alexa 568; C and G). Cell outlines of vimentin positive cells applied to IGFBP-1 channel (D and H). Alexa 568 image (D and H) also shown in grayscale for better viewing (I and J). Utilizing set thresholds, IGFBP-1 phosphorylation (pSer101) was then quantified within cell outlines (from I and J). Cell outlines were converted into multicell ROIs and applied onto the raw channel. A cell-sorted image was created to illustrate the intensity differences in signal between the IUGR and AGA (K and L). A few high expressing phosphorylated IGFBP-1 (pSer101) cells shown by red arrows. IUGR placental tissues exhibited a relatively higher IGFBP-1 phosphorylation (pSer169) compared with GA-matched AGA samples (n=5 each). Scale bar = 50 µm. Abbreviations: IGFBP-1, insulin-like growth factor binding protein-1; AGA, appropriate for gestational age; IUGR, intrauterine growth restriction; IHC, immunohistochemistry; DAPI, 4′,6-Diamidino-2-Phenylindole, Dihydrochloride; ROI, region of interest; GA, gestational age.

Relative Quantitation of Decidual IGFBP-1 and IGFBP-1 Phosphorylation at Ser 169, 119, and 101 in IUGR

We then quantified the relative changes in IGFBP-1 expression and phosphorylation at the three different serine sites in decidual mesenchymal cells of IUGR and AGA placenta. Utilizing set thresholds, levels of total IGFBP-1 expression and IGFBP-1 phosphorylation at Ser169, Ser119, and Ser101 were quantified within the outlines (Figs. 3 –6, I and J) and are depicted in Fig. 2F. Based on second binary threshold within the mesenchymal cell outlines, we measured the subpopulation of decidualized mesenchymal cells to collect data only from highly IGFBP-1 expressing cells as described earlier (Fig. 2F).

The average of intensity measurements from six representative images (n=6) per placental tissue from AGA and IUGR group (n=5 each) were used for statistical analysis. The intensity comparison of total IGFBP-1 within the vimentin positive decidual cell population between AGA and IUGR placenta (n=5) showed relatively higher expression of total IGFBP-1 in IUGR decidua (+85%; p=0.0001; Fig. 7A) compared with GA-matched AGA decidua. Similarly, our data show relative IGFBP-1 phosphorylation in IUGR was markedly greater at Ser169 (+100%; p=0.0021) compared with AGA (Fig. 7B). Phosphorylation was also increased at Ser119 (+60%; p=0.0064; Fig. 7C) followed by Ser101 (+40%) compared with AGA (Fig. 7D). By measuring IGFBP-1 phosphoisoforms only in the specific decidual cell types, particularly in the heterogeneous cell population of the basal plate region, we have provided novel information to show that IGFBP-1 phosphorylation at Ser169 (Fig. 7B) comparative to the other two phosphorylation sites was relatively greater in IUGR versus AGA placenta.

OPEN IN VIEWER

Figure 7.

Quantitative analysis of the relative expression of decidual total IGFBP-1 and phosphorylated IGFBP-1 in AGA and IUGR placental tissues. Utilizing set thresholds, IGFBP-1 expression and phosphorylation was quantified within the white outlines shown in grayscale in Figs. 3 to 6, panels I and J in AGA and IUGR subjects. IUGR group was noted to have higher expression of total IGFBP-1 (A); phosphorylated IGFBP-1, pSer169 (B); phosphorylated IGFBP-1, pSer119 (C) and phosphorylated IGFBP-1, pSer101 (D) within the decidua of placental tissues compared with GA-matched AGA placenta (n=5 each). Values are given as means ± SEM. p≤0.05 versus control as per unpaired Student’s t-test. Abbreviations: IGFBP-1, insulin-like growth factor binding protein-1; AGA, appropriate for gestational age; IUGR, intrauterine growth restriction; GA, gestational age.