A Mouse Model for Fetal Maternal Stem Cell Transfer during Ischemic Cardiac Injury

Abstract

Fetal cells enter the maternal circulation during pregnancies and can persist in blood and tissues for decades, creating a state of physiologic microchimerism. Microchimerism refers to acquisition of cells from another individual and can be due to bidirectional cell traffic between mother and fetus during pregnancy. Peripartum cardiomyopathy, a rare cardiac disorder associated with high mortality rates has the highest recovery rate amongst all etiologies of heart failure although the reason is unknown. Collectively, these observations led us to hypothesize that fetal cells enter the maternal circulation and may be recruited to the sites of myocardial disease or injury. The ability to genetically modify mice makes them an ideal system for studying the phenomenon of microchimerism in cardiac disease. Described here is a mouse model for ischemic cardiac injury during pregnancy designed to study microchimerism. Wild‐type virgin female mice mated with eGFP male mice underwent ligation of the left anterior descending artery to induce a myocardial infarction at gestation day 12. We demonstrate the selective homing of eGFP cells to the site of cardiac injury without such homing to noninjured tissues suggesting the presence of precise signals sensed by fetal cells enabling them to target diseased myocardium specifically. Clin Trans Sci 2012; Volume 5: 321–328

Introduction

Microchimerism is a term first presented in 1977 by Liegoise et al. ¹ who reported on a steady state low level proliferation of allogeneic bone marrow cells in the murine species. They later demonstrated this phenomenon in maternal tissue during and long after pregnancy. ² Pregnancy may therefore result in the physiologic transfer of a stem cell population that is fetal in nature. Since the inception of this idea decades ago the potential consequences of fetal maternal stem cell transfer in humans has gained appreciable notice.

The presence and persistence of fetal stem cells has been implicated in contributing to some autoimmune diseases in the maternal host. ³ Questions have been raised with regards to the possibility that an immune reaction between fetal and maternal cells could result in maternal disease due to the long term presence of fetal cells in the semiallogeneic maternal body. In addition, most initial reports that appeared on this topic implicated microchimerism as the inciting factor in autoimmune diseases. “Bad microchimerism” was first hypothesized in the rheumatology research community by Nelson and colleagues; they hypothesized that the persistence of fetal stem cells in maternal tissue after pregnancy led to some autoimmune diseases which may in fact be alloimmune diseases with clinical and pathological characteristics similar to graft‐versus‐host disease. ³ The group considered as evidence the observation that some autoimmune diseases do occur at a higher frequency in women than men in the presence of age‐specific incidence patterns amongst other observations. ³

Systemic sclerosis is one of the more extensively studied diseases associated with microchimerism. Some reports demonstrated the presence of significant levels of fetal cells in skin cells from women with systemic sclerosis ⁴ , ⁵ and in the peripheral blood of women with scleroderma as compared to controls. ⁶ , ⁷ Ichikawa et al. demonstrated that although the presence of fetal cells was not specific for systemic sclerosis, there were more fetal cells in systemic sclerosis patients than in controls. ⁸ Autopsy specimen studies from multiple tissues from women affected with systemic sclerosis showed that male cells of putative fetal origin were most frequently observed in the spleen. ⁹ Taken together, these data implicated fetal cell microchimerism as the cause of these autoimmune diseases.

Reports emerging shortly after conflicted with these data. Studies on Sjogren’s syndrome ¹⁰ and primary biliary cirrhosis ¹¹ reported the absence of fetal cell microchimerism in both of these autoimmune disorders. Furthermore, studies on hepatitis, thyroid disease, and cervical cancer all reported the presence of fetal cell microchimerism albeit nonautoimmune in nature. ¹² , ¹³ , ¹⁴ Gannage’s study on the potential role of fetal microchimerism in autoimmune disease led to the conclusion that fetal microchimerism is unlikely to be a risk factor for the development of connective tissue disease. ¹⁵ The researchers found that the proportion of microchimerism between systemic sclerosis, other connective tissue diseases and controls were similar. ¹⁵ The emergence of these conflicting data led to an alternative hypothesis.

The “good microchimerism” hypothesis suggests that persistent fetal cells are found within clinically affected tissue as a response mechanism to maternal injury and therefore may provide a rejuvenating source of fetal progenitor cells attempting to participate in maternal tissue repair.

Animal models have been widely utilized to study the mechanistic basis for heart failure and to develop strategies aimed at mitigating the effects of heart disease. Given the ease of generating specific genetic models, mice are particularly useful for fetal maternal stem cell transfer and microchimerism studies. Here we describe a mouse model with ischemic cardiac injury during pregnancy in order to examine the phenomenon of fetal maternal stem cell transfer. This model was designed to test whether cells of fetal origin home to sites of injury and participate in repair.

Materials and Methods

Animals

Wild‐type (WT) female virgin mice (B6CBA F1/J) and enhanced green fluorescent protein (eGFP) transgenic male mice (C57BL/6tg(ACTbeGFP)10sb/J) were purchased from Jackson Laboratories (Bar Harbor, ME, USA). All mice used were between the ages of 3–6 months. All animal care was in compliance with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health, as well as institutional guidelines at Mount Sinai’s School of Medicine.

Anterolateral myocardial infarction (MI)

Room temperature was controlled at 25°C. Mice were anesthetized with 8–100 mg/kg ketamine + 5 mg/kg xylazine intraperitoneally, and then fixed in the supine position by tying the legs and the upper jaw. After the left chest is shaved and disinfected with 70% ethanol, the skin is delicately dissected by a lateral 1.5 cm cut along the left side of the sternum. A 4–0 USP polypropylene suture is passed along the edge of the incision before the skin is dissected. The subcutaneous tissues are detached along the inferior fringe of the left pectoralis major muscles, and the left pectoralis major muscles are then refracted. The 4th intercostal space is exposed and delicately dissected 1 cm with the aid of microforceps. Self‐retaining microretractors are then used to separate the 3rd and 4th ribs enough to get adequate exposure of the operating region, but the ribs are kept intact. The heart is squeezed out of the chest by pressing the thorax lightly. The heart is held with the thumb and forefinger of the left hand, and its apex is pointed to the left side of the head. A 5–0 USP polypropylene suture is passed from the left fringe of the pulmonary infundibulum to the lower right of the left auricle, a distance of about 2–3 mm. The LAD and the great cardiac vein are ligated together. In the sham control group, the 5–0 USP polypropylene suture is passed but not ligated. The heart is replaced in the thorax immediately after the LAD ligation or sham procedure. The blood and air in the thorax is squeezed out using the forefinger. The thorax is closed with the suture that had been prepared before the thoracotomy.

DNA extraction

Total DNA was prepared from tissues using the Dneasy mini kit according to manufacturer’s instructions (Qiagen, Valencia, CA, USA).

Quantitative PCR (q‐PCR)

Quantitative PCR reactions were performed with iQ (SYBR Green Supermix) on the iQ5 Real‐Time PCR Detection System (Bio‐Rad, Hercules, CA, USA). The PCR protocol consisted of one cycle at 95°C (10 minutes) followed by 40 cycles of 95°C (15 seconds) and 60°C (1 minute). Fold changes in gene expression were determined using the comparative CT method (Ct method) ¹⁶ with normalization to the endogenous control Apolipoprotein B (ApoB). Primers used for RT‐qPCR experiments are as follows:

• 1GFP‐forward 5′‐CATCGAGCTGAAGGGCATC‐3′,
• 2GFP‐reverse 5′‐TGTTGTGGCGGATCTTGAAG‐3′,
• 3ApoB‐forward 5′‐AAGGCTCATTTTCAACAATTCC‐3′,
• 4ApoB‐reverse 5′‐GGACACAGACAGACCAGAAC‐3′.

Comparative CT method (▵▵CT method)

Briefly, the threshold cycle number (C _T) was obtained as the first cycle at which a statistically significant increase in fluorescence signal was detected. Data were normalized by subtracting the C _T values of ApoB from the C _T values of eGFP. Each reaction was done in triplicate and the C _T values were averaged. The ▵▵C _T was calculated as the difference of the normalized C _T values (▵C _T) of the treated and control samples: ▵▵C _T=▵C _{T treated}–▵C _{T control}. ▵▵C _T was converted to fold change by the following formula: fold change = 2^−▵▵CT. ¹⁶ The fold differences in gene expression are represented as the mean ± standard deviation (SD). A minimum of three samples was evaluated for each group at each time point (n= 8 for experimental group at 1 week, n= 5 at 2 weeks; n= 3 for shams at 1 and 2 weeks; n= 4 for noninfarcted control at 1 week, n= 5 at 2 weeks). The fold‐differences calculated using the ▵▵C _T method are usually expressed as a range, which is a result of incorporating the error of the ▵▵C _T value into the fold difference calculation. The error bars represent the top and bottom range of the fold‐difference. p‐values were determined by a two‐tailed paired Student’s t‐test from the ▵C _T values.

Absolute quantitation method

Q‐PCR was performed utilizing genomic DNA extracted from whole hearts. A sensitivity test ¹⁷ , ¹⁸ was performed by mixing serial dilutions of DNA from GFP transgenic mouse hearts with each of three quantities of DNA from virgin female WT mouse hearts (0, 10,000, and 100,000 pg) and real‐time PCR for amplification of GFP was performed. 1 GFP cell amongst 100,000 cells of WT background can be detected. GFP is present as two copies per cell in the transgenic mouse we are utilizing. ¹⁹

Immunofluorescence

Maternal heart ventricular 4‐μm‐thick sections were fixed for 20 minutes and then blocked with 10% donkey serum (Jackson Immunoresearch, West Grove, PA, USA) for 1 hour at room temperature (RT). Each section was incubated with the primary antibody for 1 hour at RT, followed by a secondary antibody for an additional 1 hour at RT and counterstained with DAPI. Finally the sections were incubated for 5 minutes with Sudan Black (0.7% in 70% EtOH) and cover‐slipped with mounting media (DAKO, Carpinteria, CA, USA). Slides were imaged using a Zeiss LSM‐510 Meta confocal microscope (Carl Zeiss, Munich, Germany). The primary antibodies rabbit anti‐GFP (ABCAM #AB6556, Cambridge, MA, USA) and mouse anti‐alpha actinin (Santa Cruz #15335, Santa Cruz, CA), were used. The secondary antibodies used (Alexa Fluor‐488 and Alexa Fluor‐568) were purchased from Molecular Probes (Invitrogen, Carlsbad, CA, USA).

Spectral Scanning

Spectral scanning was performed using a Leica Microsystems (Leica, Mannheim, Germany) TCS SP5 confocal microscope. Images were collected using the lambda scan mode from 545–705 nm with a 10 nm bandwidth per image. The 543 nm HeNe laser was used for excitation and images were collected at 512 × 512 pixels using the 63x/1.4NA HCX PL APO oil lens. Regions of interest (ROIs) were selected around both sample and control cells. The mean intensity versus wavelength for each respective ROI was then plotted on a graph and compared to the Alexa Fluor 568 spectral profile.

Isolation of maternal cardiac cells

Chest wall was opened to expose heart then perfused with 10 mL PBS, using a 23‐gauge needle. Entire heart was dissected out (atria and ventricle) and extraneous tissue removed. Small amounts of serum‐free medium (DMEM, Cellgro, Manassas, VA, USA) were added to prevent the heart from drying out. Hearts from 3–4 adult mice were minced and placed in serum‐free medium. Tissue was digested with pronase at 1 mg/mL (Calbiochem, Gibbstown, NJ, USA) in a spinning incubator for 1 hour at 37°C. Cells were spun at 3000 rpm for 5 minutes and supernatant was removed. 5mL of warm (37°C) complete medium (DMEM supplemented with 10% fetal bovine serum [Cellgro, Manassas, VA, USA]) was added to the tube. (NOTE: No glycine was necessary to inactivate the pronase, as the serum in the medium does this). Skeletal/cardiac muscle was triturated in the medium. During trituration, small aliquots of tendon‐free solution were transferred to an empty 50 mL tube. Above procedure was repeated by adding 5 mL aliquots of medium to the tube every few triturations until a final tendon‐free volume of 35–45 mL was achieved. Solution was filtered through a 70‐micron mesh filter to remove small pieces of tendon. Filtered solution was spun at 3000 rpm for 5 minutes. Pellet was resuspended in 3 mL of medium then 21 mL of red blood cell (RBC) lysis buffer (Ebiosciences, San Diego, CA, USA) was added. After inverting a few times, filtered solution was spun at 3000 rpm for 5 minutes. Supernatant was removed and the pellet was resuspended in 1 mL 1× PBS with antibiotics. Cells were counted.

Fluorescence‐activated cell sorting (FACS)

Cells were sorted utilizing a MoFlo high‐speed cell sorter (Dako Cytomation, Carpinteria, CA, USA). Both eGFP+ (cells of fetal origin) and eGFP– (cells of fetal and maternal origin) populations were collected. Data analysis was performed using FlowJo Software (Tree Star, Ashland, OR, USA).

Results

Mouse model

WT virgin female mice, age 3–6 months, were crossed with heterozygous eGFP transgenic male mice as previously described by Bianchi et al. ²⁰ , ²¹ The female mice underwent ligation of the left anterior descending (LAD) artery in order to induce an anterolateral myocardial infarction (MI) at gestation day 12 ( Figure 1 ). Consistent with our previous work, this results in approximately 50% left ventricular infarction and an overall survival rate of 70%. ²² Congruent with Mendelian inheritance approximately 50% of the embryos express eGFP providing a reliable method of tracking the fate of the fetal cells. Gestation day 12 was chosen as we had observed that induction of myocardial injury preceding gestation day 12 resulted in increased rates of resorbtion of embryos likely due to the hypoxic insult. Furthermore, other groups have shown that fetal cells were first observed in most organs around gestation day 11. ¹⁷ MI after gestation day 12 resulted in increased rates of abortion of embryos due to the hemodynamic consequences of the volume overload state in late pregnancy. Maternal cardiac tissue was then analyzed for eGFP at various time points before and after delivery ( Figure 1 ).

Figure 1.

Schematic of experimental mouse model. Virgin wildtype females were mated with eGFP transgenic males and underwent LAD ligation at E12. Mice were sacrificed and analyzed at multiple time points afterwards.

Quantifying eGFP in the maternal heart and circulation

Comparative method (Pfaffl method)

EGFP expression was initially quantified in injured maternal hearts, sham‐operated maternal hearts and noninjury control maternal hearts. Postpartum females were sacrificed at 1 or 2 weeks post‐MI. Total DNA was extracted from each total heart and eGFP expression was analyzed using the Pfaffl ¹⁶ mathematical model for relative quantification by q‐PCR ( Figure 2A and B ). This measure was taken to ensure correct data analysis as the Pfaffl method accounts for differences in amplification efficiencies of the target and reference genes ¹⁶ whereas, the traditional ▵▵CT method ²³ assumes that both the target and the reference genes are amplified with efficiencies near 100% and within 5% of each other. In our case, the amplification efficiencies of the target and reference genes were similar (∼2) and within the acceptable range (1.9–2.1) for q‐PCR therefore either method yields similar results. In fact, the ▵▵CT method is a special case of the Pfaffl method where the amplification efficiencies of the target and reference genes equal 2 ( Figure 2A ). Infarcted hearts harvested at 1‐week post‐MI contained 120 times more eGFP than controls (p= 0.0003) and 20 times more eGFP than shams (p= 0.0027). Infarcted hearts harvested at 2 weeks post‐MI contained 12 times more eGFP than controls (p= 0.0001) and 8 times more eGFP than shams (p= 0.0001; Figure 2B ).

Figure 2.

Comparative quantitation of eGFP in maternal hearts. (A) Pfaffl method flow chart for relative quantification. (B) Quantitative PCR results of relative eGFP expression utilizing Pfaffl method.

Absolute method

Absolute quantification utilizing q‐PCR determined the eGFP cell numbers in control, sham‐operated, and MI hearts. We generated standard curves for both eGFP and the internal control ApoB by plotting Ct values for different quantities of known amounts of DNA from the eGFP transgenic mice versus the DNA quantity in nanograms. Equations fitting these curves and the mouse genome conversion factor ¹⁷ for the strain of mouse utilized were used to extrapolate the DNA quantities in our experimental samples. The results show that 1.3% and 1.7% of the total heart at 1 and 2 weeks after injury, respectively was composed of eGFP+ cells ( Figure 3 ).

Figure 3.

Absolute quantitation of eGFP in maternal hearts. This was done using standard curves for both eGFP and internal control Apolipoprotein B by plotting Ct values for different quantities of known amounts of DNA from eGFP transgenic mice versus the DNA quantity in nanograms. The mouse genome conversion factor was also utilized to calculate the DNA quantities in our experimental samples.

Immunofluorescence method

In a separate group of infarcted and control mice, immunofluorescence analysis with confocal microscopy was utilized to detect eGFP+ cells in ventricular tissue sections of maternal hearts at various time points subsequent to myocardial injury ( Figures 4A–C ). EGFP+ cells were noted in infarct zones and peri‐infarct zones of infarcted maternal hearts at 1, 2, 3, and 4 weeks post‐MI ( Figure 4A ). Negligible numbers of eGFP cells were noted in noninfarct zones of the infarcted maternal hearts ( Figure 4A ). These results were verified by obtaining spectral profiles of sample cells (eGFP+) and control cells (eGFP–; Figure 4B ). Further analysis of ventricular sections at 3 weeks post injury indicated that 3% of nuclei within the infarct and border zones belonged to eGFP positive cells ( Figure 4C ). In comparison, only 0.2% of nuclei in noninjured zones of the same ventricular sections belonged to eGFP cells ( Figure 4C ). Impressively, 50% of the eGFP cells within the infarct area expressed the cardiac marker alpha‐actinin ( Figure 4C ) although no organized sarcomeres were detected in the vast majority of these cells, suggestive of an immature cardiac phenotype. ²⁴

Figure 4.

(A) Immunofluorescence sections of maternal ventricular myocardium at various time points depicting the green fluorescence of eGFP+ fetal cells. (B) Spectral profiles of eGFP+ cells versus endogenous maternal cardiac tissue as controls. (C) Absolute cell quantitation by immunofluorescence and computing an index‐ratio of eGFP+ nuclei to total nuclei; approximately 50% of eGFP+ nuclei belonged to cells that coexpressed alpha‐actinin. (continued on next page)

Flow cytometry method

Analysis of cell suspensions by flow cytometry show that fetal cells selectively home to the injured maternal heart and not to uninjured organs. The cell suspensions were generated from various organs and tissues harvested from mice subjected to cardiac injury. To establish appropriate flow cytometry gates, corresponding cell populations obtained from age‐matched pregnant WT female mice mated with WT males were used. Approximately 1.1% of the cells in the injured heart were eGFP+ before delivery and this number rose significantly to ∼6.3% immediately after delivery ( Figure 5 ). In blood, ∼1.3% of cells were eGFP+ before delivery and this number rose to ∼3.6% after delivery, although this increase was not statistically significant ( Figure 5 ).

Figure 5.

EGFP positive cells analyzed by flow cytometry before/after delivery. In both injured hearts and blood, the numbers of eGFP+ cells increase after delivery although this is a significant increase in the injured heart but not in blood.

Discussion

The results presented here demonstrate the selective homing of fetal cells to injured maternal myocardium with no such homing to uninjured tissues in our murine model of ischemic cardiac injury during pregnancy. These results are comparable to previous studies conducted in similar murine models, whereby wild‐type female mice were mated with GFP expressing males. In these prior models fetal microchimerism has been detected in brain, liver, kidney, blood vessel, and skin injuries. ²⁵ , ²⁶ , ²⁷ , ²⁸

Nassar et al. showed that fetal cells responding to wounds generated on normal and bleomycin‐induced fibrotic skins of parous or pregnant WT females were mainly progenitor cells. The majority of fetal cells expressed leukocyte markers during early phases of wound healing and endothelial markers in later phases of healing. ²⁵ A study examining the homing of fetal cells to the brain in pregnant WT female mice ²⁷ illustrated that there were more fetal cells in the maternal brain 4 weeks after delivery than on the day of delivery. Furthermore, there were a higher number of fetal cells within the injured regions after an excitotoxic lesion to the brain with the cells adopting markers suggestive of perivascular macrophage, neuron, astrocyte, and oligodendrocyte like cells. ²⁷ Fetal endothelial progenitor cells acquired by pregnant WT female mice participated in maternal angiogenesis at blood vessel inflammation sites where inflammation was previously induced by contact hypersensitivity using Oxazolone. ²⁶ Wang et al. used a similar model to investigate microchimerism in injured livers and kidneys of pregnant female WT mice. They noted that GFP positive cells expressing albumin, a marker for hepatocytes, could be detected in the injured liver. They also detected GFP positive cells in the kidneys within the tubular basement membrane, indicating formation of tubular cells. ²⁸ Taken together, these studies further strengthen our hypothesis that fetal cells home to injured maternal tissues at a higher frequency in comparison to uninjured tissue and points to the presence of precise signals sensed by cells of fetal origin that may enable them to target diseased tissues specifically.

Bianchi and colleagues have shown that male microchimeric cells bearing epithelial, leukocyte or hepatocyte markers could be detected in a variety of diseased maternal tissue specimens suggesting the presence of fetal cells that may have multilineage capacity. ²⁰ The tissues in this study were from women with no history of organ transplantation or blood transfusions with the exception of one thereby suggesting that the likely source of the male microchimeric cells were male fetuses carried by the women in previous pregnancies.

Anecdotal reports of the association of fetal cell microchimerism with the maternal response to injury in humans have been demonstrated. For example, one group analyzed a liver biopsy specimen from a woman with hepatitis C. ¹³ The woman had stopped treatment and nevertheless did well clinically with subsequent disease ablation. Using dual‐color Fluorescence In‐Situ Hybridization (FISH) studies utilizing X and Y chromosome probes, thousands of male cells contained within her liver biopsy were detected. These cells were morphologically indistinguishable from cells in the surrounding liver tissue. The only possible source of stem cells was a pregnancy she had terminated approximately two decades prior as she had never received a blood transfusion and neither was she a twin. ¹³ Similar results were demonstrated in cervical cancer tissue of affected women. Furthermore, researchers were able to show that the fetal cells located within the cancerous tissue were diploid thereby excluding the possibility that they were persistent spermatocytes. ¹²

The study of microchimerism using male DNA as a marker in female tissue ¹² , ¹³ , ²⁹ has its limitations. For example lack of diseased and healthy tissue and insufficient pregnancy histories are major stumbling blocks. Furthermore, fetal cell acquisition by females who have given birth to daughters cannot be detected using similar methods making the described murine model invaluable. The use of homozygous GFP expressing male mice results in all cells of fetal origin inheriting the transgene and therefore the detection of higher numbers of fetal cells. ³⁰ We however used heterozygous GFP‐expressing males in our model as it has been shown that cardiomyopathies can occur with expression of GFP in the heart in a dose‐dependent manner. ³¹ , ³² The use of heterozygous eGFP expressing males was a measure to further minimize the possible deleterious effects of high levels of expression of foreign proteins such as GFP on cardiac function.

Cell surface markers expressed on the fetal cell types involved in fetal maternal cell trafficking during pregnancy in the uninjured state have been previously studied by Bianchi and colleagues. ³⁰ Their results suggest that a heterogeneous population of progenitor and differentiated cells were found in different maternal organs such as lung, liver, spleen, blood, bone marrow, kidney, heart, thymus, and brain ³⁰ albeit at different relative proportions. Of note, the study was limited to antigens expressed on hematopoietic cells with only Endoglin/CD105 being found in the heart at a significant median frequency. ³⁰ Furthermore, analysis was conducted in an uninjured state. In our model, Cdx2 expressing cells were found at a relatively high frequency (38%) in the injured maternal heart. ³³ Our results suggest that injury may result in recruitment of cell types that are different from those recruited in the uninjured state.

EGFP positive fetal cells identified by immunofluorescence were detected as individual cells or as clusters within the infarct and border zones of the injured myocardium ( Figure 4A ). The presence of eGFP positive cells in clusters within the injured myocardium may be an indication that the fetal cells are proliferating after homing to the injury sites and are thus stem or progenitor in nature. We have previously shown that when isolated from the injured heart and cultured in vitro, these fetal cells exhibit clonal expansion. ³³ An alternative explanation may be that these cells converge at sites permeated with signals that act as “attractants” of fetal cells to injured myocardium.

The trend of detecting higher numbers of fetal cells around the time of delivery suggests that delivery seems to cause the numbers of fetal cells entering the maternal circulation to rise. This corresponds with the significant increase in fetal cells homing to the injured hearts noted in our model at delivery and is in conjunction with other studies. ¹⁷ , ³⁴ Fujiki et al. showed that fetal cell microchimerism increases with advancing gestation and was maximal near term. ³⁵ Analysis of the fetal–maternal interface in placenta by Vernochet et al. showed that the number of fetal cells rose regularly and significantly as gestation progressed and was maximal between gestation days 17 and 19. ³⁴ However, in our model fetal cells were still detectable in injured maternal tissue up to 4 weeks after delivery suggesting that after homing these cells engraft within injured maternal tissues.

We present in this study a mouse model of ischemic cardiac injury during pregnancy and demonstrate using a variety of methods (q‐PCR, immunofluorescence, flow cytometry), the ability of fetal cells to preferentially home to injured maternal tissue. A comparison of all three of these methods show a correlated trend towards detection of higher numbers of eGFP+ fetal cells in injured maternal tissues relative to uninjured tissues. However we must point out that limitations with each detection method exist and as such must be taken into account when analyzing different types of maternal tissues for degrees of microchimerism.

EGFP molecules can be lost if cell integrity is disrupted by freezing, sectioning, or permeablization and as such should not be visualized in fresh or frozen tissue sections without prior fixation. ³⁶ , ³⁷ Furthermore, the fluorescence of eGFP is dependent on its conformation ³⁸ , ³⁹ requiring fixation methods that can preserve its conformational integrity. In our hands, fixation of cardiac maternal tissue with a formaldehyde‐based fixative prior to tissue sectioning was imperative to circumvent leaching of soluble eGFP. Cardiomyocytes are highly autofluorescent due to the presence of lipofuscins. Autoflourescence is further excerbated after fixation with a formaldehyde based fixative ⁴⁰ due to fluorescent by‐products generated from aldehydes reacting with amines and proteins. This makes it difficult to distinguish real eGFP signal from background signal even with the use of an anti‐eGFP antibody to detect cells of fetal origin. Subsequent quenching of cardiomyocyte autofluorescence was facilitated by use of Sudan Black ⁴¹ a lysochrome diazo dye. In addition, spectral profiles of ventricular tissue sections were obtained to ensure that the native autofluorescence of cardiomyocytes was not affecting fluorescence images. Indeed the mean intensities of the sample regions were significantly higher than mean intensities of the control regions.

The above mentioned measures cannot be applied when analyzing cells from injured maternal tissues for presence of eGFP+ cells by flow cytometry and as such it is possible that fetal cell numbers maybe underreported (in the case where eGFP+ cells have diminished green fluorescence) or over‐reported (in the case where false positives are detected with the use of an anti‐eGFP antibody). Q‐PCR therefore seems to be the best method of detection in cardiac tissues. Nevertheless, optimization of tissue treatment conditions for different tissue types is still necessary.

Conclusions

In a recent publication we demonstrated that fetal cells not only home to injured maternal myocardium but also differentiate to functional beating cardiomyocytes. ³³ This manuscript describes in detail the mouse model utilized in that study. In summary eGFP+ fetal cells traverse the placenta in response to injury and engraft within the injured maternal myocardium ( Figure 6 ). A better understanding of the mechanism(s) involved in the phenomenon of fetal maternal stem cell transfer and microchimerism may offer new treatment perspectives with regards to not only cardiac injury but also injuries in other tissues/organs.

Figure 6.

Fetal cells traverse the placenta and home to sites of myocardial injury. This model may be useful for assessing the fetal response to injured tissues in organs other than the heart and perhaps lead to the discovery of novel stem/progenitor cells.

Conflict of interest

None.

Sources of funding

This work was supported by a National Institutes of Health (NIH) grant (R01‐HL 088255) and an American Heart Association (AHA) grant.