SEARCH

SEARCH BY CITATION

Keywords:

  • folate receptor;
  • cardiac neural crest;
  • migration;
  • siRNA;
  • heart defects

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Folate supplementation reduces the incidence of congenital heart defects, but the nature of this protective mechanism remains unclear. Immunolabeling demonstrated that the neural tube and neural crest (NC) cells were rich in the high-affinity folate receptor FOLR1and during the early stages of development FOLR1 was found principally in these cells. Suppression of Folr1 expression in the nascent cardiac NC by site-directed short-interfering RNA (siRNA) altered cardiac NC cell mitosis and subsequent migration patterns leading to abnormal development of the pharyngeal arch arteries (PAA) and outflow tract. qPCR analysis demonstrated that the siRNA treatment significantly reduced Folr1 24 hr after treatment. These treatments also significantly reduced mitosis in the neural tube, but adjacent, nontreated areas were unaffected. In summary, a brief reduction in the expression of Folr1 during a critical stage of NC development had long-term consequences for the development of the PAA and outflow tract. Developmental Dynamics 239:1136–1144, 2010. © 2010 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Congenital heart defects occur at a rate of approximately 1% of live births, typically are considered to be the most common among congenital malformations that require medical intervention, and are the congenital defect most likely to result in the death of a newborn (Hoffman,1995; Yagel et al.,1997; Centers for Disease Control and Prevention,1998; Rosano et al.,2000; Wren et al.,2000; Hoffman and Kaplan,2002). Although the etiology of congenital heart defects remains largely unknown, there is substantial experimental evidence indicating that the cardiac neural crest is a key player in heart development, and the perturbation of the cardiac neural crest results in abnormal heart development in both avian and mammalian models (e.g., Kirby et al.,1983; Hutson and Kirby,2007; Porras and Brown,2008).

The most widely studied and best known role of the cardiac neural crest is its differentiation into mesenchyme-like cells that form the walls of the proximal great vessels and the aorticopulmonary septum (Kirby et al.,1983,1985; Besson et al.,1986). When the cardiac neural crest is perturbed by surgical ablation or interruption of its migration by pharmacologic or genetic alterations, the typical result is abnormal development of the pharyngeal arches that may be related to misalignment of the arterial pole of the heart, and abnormal septation of the truncus arteriosus. These kinds of defects are the most common developmental defects of the cardiovascular system that are observed in the human population, and may include ventricular septal defects, persistent truncus arteriosus, transposition of the great arteries, double outlet right ventricle, tetralogy of Fallot, interrupted aortic arch, and pulmonary atresia (e.g., Kirby et al.,1983; Cavieres and Smith,2000; Nishijima et al.,2000; Boot et al.,2003,2004; Rosenquist et al.,2007; Varadkar et al.,2008).

Conotruncal development appears to be a folic acid-sensitive process. Several population-based studies have demonstrated that periconceptional folate supplementation reduces the occurrence of conotruncal defects by as much as 50% (Shaw et al.,1995; Czeizel,1995,1998; Botto et al.,1996,2000; Burgoon et al.,2002). Furthermore, abnormal folate metabolism that results from the methylenetetrahydrofolate reductase 677C>T mutation produces a significantly increased risk for conotruncal but not other heart defects (van Beynum et al.,2006). We (Rosenquist et al.,1996; Brauer and Rosenquist,2002) and others (Boot et al.,2003,2004) have shown that the hyperhomocysteinemia that accompanies low folate or abnormal folate metabolism impairs the development of the cardiac neural crest, resulting in abnormal development of the conotruncal region of the heart.

Recent evidence reported by members of this research team indicate that abnormal expression of folate receptor genes in the mouse embryo model is associated with abnormal development of the heart and other neural tube- or neural crest-derived structures (Zhu et al.,2007; Taparia et al.,2007). Furthermore, Saitsu et al. (2003) demonstrated that the expression of the folate receptor Folbp1 (also called Folr1, which is homologous with human Folr1 [FRalpha] and avian Folr1) was expressed at high levels in the dorsal neural tube during neural crest formation, indicating its potential role in neural crest development. These data, taken together with the results summarized above, led us to hypothesize that conotruncal heart defects that result from reduced availability of folate or reduced folate transport may occur because of an effect upon the cardiac neural crest and that reduced expression of the folate receptor specific to the cardiac neural crest would have an adverse effect upon its ability to support pharyngeal arch and conotruncal development.

We designed an experiment to test this hypothesis, using the chicken embryo model. We chose this model because it has proven especially useful for the study of heart development and has been the basis for key experiments that showed the relationship between the neural crest and pharyngeal arch/conotruncal defects (reviewed by Hutson and Kirby,2007). Chicken embryos are uniquely accessible for experimental manipulation during key developmental stages (see http://www.nih.gov/science/ models), permitting the application of a wide variety of useful techniques that cannot be applied to mammalian embryos in utero, including electroporation and direct application of selective reagents (Rosenquist et al.,2007; Lie et al.,2010). It is important that, after these experimental manipulations, the viable and rapidly developing embryo can be studied in vivo.

Uptake of folate is facilitated by at least three different transporters (Salazar and Ratnam,2007). Among these, the high-affinity receptor coded by the gene Folr1 is associated with high mitotic rates, e.g., in cancer cells (Hao et al.,2007) or with the dynamic growth processes during embryonic development (Saitsu et al.,2003; Sato et al.,2008), and appears to be the key to receptor-mediated endocytosis of folate during early embryonic development. Therefore, we designed experiments to assess the role of the avian homolog of Folr1 during neural crest formation. We determined that the folate receptor gene Folr1 was highly expressed in the avian model in the neural crest and neural tube during neurulation, consistent with the results shown by Saitsu et al. (2003) in the mammalian embryo during similar stages of development. Subsequently, we inserted by means of electroporation an short-interfering RNA (siRNA) directly into the neural tube at the site of formation of the premigratory cardiac neural crest. This treatment reduced the expression of Folr1 during a key period of cardiac neural crest development and inhibited mitosis in the neural tube, thereby apparently reducing the formation of migratory cardiac neural crest cells. By applying simultaneous incorporation of the gene for green fluorescent protein (GFP), we were able to follow neural crest migration. A brief but critical interruption in expression of Folr1 was associated with a failure of neural crest cell migration into the pharyngeal arches, resulting in abnormal pharyngeal arch artery development and an increased risk of abnormal heart development.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Observation of whole embryos in vivo following the transfection of the gene for GFP by means of electroporation into the nascent neural tube demonstrated that GFP fluorescence was detected from approximately the mid-mesencephalon region to about somite 4, independent of whether the GFP was delivered in control solutions (Fig. 1A–C) or solutions containing siRNA designed to inhibit expression of the folate receptor Folr1 (Fig. 1D–F). Control solutions included solutions with GFP and no siRNA, or GFP with nonsense siRNA; there were no differences in results between these two controls.

thumbnail image

Figure 1. Migration of green fluorescent protein (GFP) labeled cells. A–F: Whole embryos were photographed under epi-illumination to demonstrate cells emitting GFP fluorescence. To maximize contrast, software was used to convert the GFP green emission to pink, and the lighter green autofluorescence of the surrounding embryonic tissues to blue. Embryos were treated with a control solution (A,C, GFP-only; B, GFP and a negative control short-interfering RNA [siRNA]), or with GFP and an siRNA for Folr1 silencing (D,E,F), electroporated into the dorsal neural tube at axial level somites 2–4, at age 29–33 hr (7–9 somites, neural fold stage, Hamburger and Hamilton [HH] stages 9–10 [Hamburger and Hamilton,1951]). “Days” in the figure legends below refers to the time elapsed after this treatment. Magnification varies from figure-to-figure, however, in all cases, the bar equals 0.5 mm. A,C,D,F: Because of the exponential growth of the embryos during the period of observation, the embryos in C and F are, on the average, 4.54 times larger than those in A and D. The photographic enlargements are adjusted accordingly to permit illustration of the whole head and thorax, consequently, the resolution and granular detail is reduced from left to right in these figures. A,D: Day 1 after treatment. A: Day 1 after treatment (age 55–59 hr, 29–32 somites, HH stage 17), embryos given a control solution showed a bolus of labeled cells in the pharyngeal arches (P) that was continuous with the labeled area in the neural tube, obvious at the dorsum of the embryo. D: Few GFP-positive cells were found in the pharyngeal arches of embryos of the same age that had been treated with siRNA to silence expression of Folr1 (P). H, heart. B,E: Day 2 after treatment. B: Two days after treatment (age 79–83 hr, 43–44 somites, HH stage 21), the bolus of GFP-labeled cells migrated past the lateral mid-line into pharyngeal arches 3, 4, and 6 in embryos given a control solution (P). E: There were substantially fewer GFP-positive sites in the pharyngeal arches in embryos that had received GFP+ targeted siRNA (P). H = heart. C,F: Day 3 after treatment. C,F: Three days after treatment (age 103–107 hr, genesis of somites complete, HH stage 26), migratory cells were near the heart (H) in embryos given a control solution (C, P) but in the GFP+siRNA group, there were few detectible GFP-labeled cells in the pharyngeal arches (F, P).

Download figure to PowerPoint

Migration of the cardiac neural crest away from the neural tube was reduced when the expression of Folr1 was knocked-down with siRNA. In embryos treated with control solutions, by 1 day after treatment (total incubation approximately 57 hr, approximately Hamburger and Hamilton [HH] stage 17) a bolus of labeled cells that was continuous with the labeling of the neural tube was detected in the dorsal-most pharyngeal arch artery segments (Fig. 1A), whereas few labeled cells were found to have migrated out of the neural tube in embryos treated with inhibitive siRNA (Fig. 1D).

By 2 days after treatment (total incubation approximately 81 hr, approximately HH stage 21), the bolus of labeled cells had advanced past the lateral mid-line into pharyngeal arches 3, 4, and 6 and remained continuous with the set of labeled cells in the neural tube (Fig. 1B). As had been the case 24 hr earlier, there was little evidence of migration into the pharyngeal arches in embryos that had received both GFP and functional siRNA (Fig. 1E).

At 3 days after treatment (total incubation approximately 105 hr, approximately HH stage 26) the migratory bolus of labeled cells was near the heart in the embryos treated with control solutions (Fig. 1C). In the functional siRNA-treated group on the other hand, there were few detectible GFP-labeled cells in the pharyngeal arches (Fig. 1F).

The embryos shown in Figure 1 were selected as typical of the surviving embryos that were analyzed with in vivo fluorescence microscopy during each time-point: Figure 1A represents a total of 27 embryos; Figure 1D, 18 embryos; Figure 1B, 21 embryos; Figure 1E, 19 embryos; Figure 1C, 15 embryos; and Figure 1F, 14 embryos.

Sections of control embryos from HH stage 17 embryos, approximately 24 hr after injection of GFP or GFP/nonsense siRNA into the lumen of the neural tube, were treated with antibodies against GFP and the neural crest marker 20B4. GFP was found in the dorsal neural tube, and in cells that had migrated out of the neural tube (Fig. 2A,B); an overlay process reveals substantial overlap between 20B4-positive cells and GFP-positive cells (Fig. 2C), supporting the evidence shown in the whole-mount in vitro immunofluorescence preparations in Figure 1A–C, that is, that the GFP-labeled cells in the pharyngeal arches were neural crest cells.

thumbnail image

Figure 2. Green fluorescent protein (GFP) and the neural crest marker 20B4. A: Twenty-four hours after insertion of GFP or GFP/mis-sense siRNA by electroporation, immunolabeling of transverse sections of Hamburger and Hamilton (HH) stage 17 embryos demonstrated the presence of GFP in the dorsal neural tube (N) and in the cells that had migrated from the neural tube. Isolated areas containing GFP in the epidermal layer (arrows) is a common preparation artifact. B–D: A section taken in sequence with that shown in A, was immunolabeled with an antibody marker for neural crest cells. Some of the nascent neurons in the neural tube (N) also express the marker. There is substantial overlap with the cells positive for 20B4 and GFP, as shown in the orange–yellow areas of the overlay preparation in C. Original magnification, ×20; images combine epifluorescence and phase/contrast.

Download figure to PowerPoint

The 3-day survival rate for embryos that were treated by electroporation was approximately 50% regardless of whether the treatment was GFP-only, or GFP + siRNA. We considered this an excellent result that permitted us to follow a significant number of embryos in ovo for analysis of neural crest cell migration during the early stages of pharyngeal arch artery development and aorticopulmonary septation. However, there were no survivors in the GFP/siRNA group after 7 days, and no survivors in the GFP group after 8 days. This steep mortality curve appears to be a typical effect of electroporation upon the long-term survival of avian embryos. The 1-day survival rate of approximately 90% in the current study is consistent with our previous results using the same electroporation method (Lie et al.,2010), and with the results of other, similar, studies (Momose et al.,1999); the 3-day survival and subsequent steeper mortality rate is also consistent with our previous results (Lie et al.,2010).

Although there was essentially no difference in survival among the GFP-only, GFP + nonsense siRNA, and GFP + siRNA groups, there were obvious differences in the developmental and migratory behavior of neural crest cells, as described in our observations of whole embryos. In sections of HH stage 26 embryos, 3 days after treatment, cells positive for the neural crest cell marker 20B4 were abundant in the pharyngeal arch arteries of embryos that were treated with GFP-only (Fig. 3A,B) but were rare or absent in the poorly developed pharyngeal arch arteries in the embryos given GFP + siRNA (Fig. 3C,D). The obvious failure of neural crest cells to migrate toward the heart resulted in a significant difference in the rate of occurrence of abnormal developmental events between the two groups (Fig. 4). Pharyngeal arch or heart abnormalities that were detectible either grossly or histologically, and that are defined as neural crest-related defects, included the following: complete absence of one or more pharyngeal arch arteries 3, 4, or 6; no detectible development of the vessel wall in the pharyngeal arch arteries by stage 24–25 (e.g., Fig. 3D); and complete absence of the initiation of aorticopulmonary septation/ectomesodermal formation in the outflow tract indicating persistence of the primitive truncus arteriosus. In one case, there was a complete absence of cushion formation, however this is not typically associated with a neural crest migration problem. Other abnormal developmental events with a neural crest/neural tube connection included failure of neural tube closure and craniofacial defects including absence of midline closure and absence of one eye. Embryos that were abnormal showed an average of two different defects.

thumbnail image

Figure 3. Neural crest cell marker in the branchial arch arteries. A,B: Transverse sections taken through the areas of the third, fourth, or sixth branchial arch arteries in Hamburger and Hamilton [HH] stage 26 embryos treated with control solutions demonstrated that cells positive for green fluorescent protein (GFP; A) as well as the neural crest cell marker 20B4 (B) were detected in the walls of the pharyngeal arch arteries (P), where the nascent media indicated robust, normal artery development. C,D: On the other hand GFP-labeled (C) and 20B4-labeled (D) cells were rare or absent in the poorly developed pharyngeal arch arteries in the embryos given GFP along with targeted short-interfering RNA (siRNA (P). In both control and GFP/siRNA samples, cells positive for both GFP and 20B4 also were found in the neural tube (N). Original magnification, ×20; images combine epifluorescence and phase/contrast. Phase/contrast is more prominent in C and D to enable visualization of the artery walls in the absence of contrasting immunofluorescence. The areas shown in B,D were ipsilateral with A,C, but are reversed photographically for ease of comparison.

Download figure to PowerPoint

thumbnail image

Figure 4. Embryo phenotype after treatment. The process of electroporation per se (green fluorescent protein [GFP]-only group, GFP) resulted in a significant increase in major structural defects that were ascertainable by gross or histologic observation of embryos surviving at least through Hamburger and Hamilton [HH] stage 28, consisting mainly failure of neural tube closure. The same frequency of abnormalities occurred with GFP and a negative control short-interfering RNA (siRNA−). In both cases, the occurrence of abnormal artery or heart development was insignificant. When embryos were exposed to a Folr1-suppressing siRNA, there was a significant increase in the rate of occurrence of structural abnormalities in the pharyngeal arches or heart. The most common structural abnormalities at this stage were complete bilateral absence of one or more pharyngeal arch arteries 3, 4, or 6; or no detectable development of the vessel media in one or more of the pharyngeal arch arteries by stage 24–25. *P ≥ 0.001 compared with untreated; **P ≥ 0.001 compared with GFP-only.

Download figure to PowerPoint

The knock-down of Folr1 expression was confirmed by quantitative polymerase chain reaction (qPCR). Twenty-four hours after GFP + siRNA treatment, expression of Folr1 detected by qPCR was reduced significantly compared with embryos that had undergone treatment with GFP-only (Fig. 5). Subsequently, Folr1 gene expression rebounded to levels that were higher than baseline, 48–72 hr. By 96 hr after treatment, there were no significant differences between the groups of embryos.

thumbnail image

Figure 5. Quantitative polymerase chain reaction (qPCR) analysis for Folr1 RNA. Summary of the qPCR data illustrating the change (Ln(RQ) in the expression of the folate receptor in embryos following electroporation with green fluorescent protein-positive (GFP+) short-interfering RNA (siRNA) to Folr1 compared with embryos electroporated with GFP-alone. For these experiments embryos were electroporated with either GFP or GFP+ FRsiRNA and then the expression of Folr1 was determined at various time points using qPCR. Differences among groups were determined by the comparative Ct method, where 2−[ΔΔ]Ct = relative quantitation (RQ). Expression of Folr1 declined significantly in the 24 hr following electroporation of siRNA; Folr1 expression rebounded significantly and returned to the untreated baseline in approximately 96 hr. *P ≥ 0.001 compared with GFP-alone.

Download figure to PowerPoint

Results cited above showed (1) that the electroporation process produced a result specific to the dorsal neural tube/neural crest; and (2) that the inclusion of a Folr1-suppressing siRNA was associated with reduced mitosis of neural crest cells resulting in fewer cells migrating, and significantly reduced expression of Folr1. To confirm that these results were consistent with the true distribution of Folr1 at the time of neural crest formation/migration, we used the public domain morphometric analysis system NIH Image to test the distribution of anti-FOLR1 antibody in the HH stage 17 embryo. The results showed that 72% of the labeling for FOLR-positive cells were in the neural tube and 18% was in the neural crest, supporting further the probability that the qPCR-detected decline in Folr1 reflected an siRNA knockdown effect specific to these regions; and confirming earlier reports on the spatial distribution of the folate receptor (Saitsu et al.,2003; Sato et al.,2008).

One of the predictable effects of their inability to sequester sufficient folate is a reduction of proliferation in the neural tube and neural crest cell precursors. We applied a Ki67 proliferation index (method based upon Huuhtanen et al.,1999) to transverse sections through the neural tube at the site of siRNA injection in HH stage 21 embryos that were harvested 48 hr after treatment with GFP-only, or with GFP and siRNA to silence expression of Folr1. There was no difference between the two groups in the proliferation index of cells in the somites: 55% of cells were labeled in the GFP-only group vs. 53% of cells in the GFP + siRNA group, P > 0.10. On the other hand, there was a significant reduction in the proliferation index of the cells in the neural tube that had been exposed to GFP + siRNA: 72% of cells were labeled in the GFP-only group vs. 31% of cells in the GFP + siRNA group, P < 0.001. The GFP-only or control group was not significantly different from the GFP + siRNA group 3 days after treatment using the same indexing technique (results given as text only). This result suggests that a principal effect of decreased expression of Folr1 was a significant reduction in mitotic activity of cells in the neural tube, at the site of siRNA electroporation.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The results of this experiment indicate that the duration of the siRNA-induced knock-down of Folr1 was relatively brief, on the order of 24–48 hr; yet there was a longer-term impact that resulted in an obvious failure of an adequate number of neural crest cells migrating into the pharyngeal arches, therefore, the pharyngeal arch artery walls had few or no precursors of neural crest-derived smooth muscle cells, leading ultimately to abnormal, or at the very least significantly delayed, development of the cardiac outflow tract. This long-term impact following a treatment of short duration is consistent with earlier studies in which a 24-hr exposure to high levels of exogenous homocysteine also resulted in a decrease of cardiac neural crest cells in the pharyngeal arches, and abnormal heart development (Rosenquist et al.,1996; Brauer and Rosenquist,2002; Tierney et al.,2004). It appears that a 24-hr interval in the midst of the short but critical period of induction and specification of the cardiac neural crest may be sufficient for disruption of the delicate, complex interplay of the neural tube with the epidermis on the one hand, or with migratory routes embedded within the mesenchyme on the other. Indeed, the induction, specification, migration, and differentiation of neural crest cells are extraordinarily complex processes (Santiago and Erickson,2002; Sauka-Spengler and Bronner-Fraser,2008; Thiery and Sleeman,2006), and it may be assumed that the disruption of Folr1 expression results in an alternative cascade of events that leads to abnormal development of structures reliant upon the neural crest.

However, the initiating event for the developmental effects found in the current study may be as simple as a transitory disruption in the proliferation of the neural crest cell precursors during a key developmental period, as suggested by the altered Ki67 mitotic-index in the neural tube. A high rate of cell division is characteristic of cells with a robust expression of folate receptors, and interference with the facilitated transport of folate by blocking its receptor is a key to the diminished proliferation of such cells (e.g., ovarian or breast cancer cells, see Lee et al.,2006). Furthermore, as discussed previously by a laboratory collaborating with our group (Tierney et al.,2004), progenitors arising in the folate receptor–enriched dorsal neural tube must undergo a cell cycle-dependent epithelial–mesenchymal transition to progress to migratory neural crest cells, and inhibiting the G1/S progression, as folate deficiency does (Brown et al.,2006; Tang et al.,2008), blocks neural crest cell formation (see also Burstyn-Cohen and Kalcheim,2002). Thus, it may be postulated that impaired expression of the folate receptor for 24–48 hr slows neural crest precursor cell proliferation sufficiently to inhibit the formation of the migratory neural crest during a stage that is critical to the successful initiation of a cascade of events that that are keys to neural crest, pharyngeal arch artery and, ultimately, outflow tract development.

Results from another of the laboratories collaborating in this study showed multiple developmental abnormalities related to the massive die-off of neural tube and neural crest cells in Folbp1−/− (Folr1−/−) knockout mouse (Piedrahita et al.,1999; Tang et al.,2004; Zhu et al.,2007), damage that was far more extensive than is the case for the current study. These contrasting results demonstrate a difference between the global loss of Folbp1 and its reduced expression in a highly focused spatiotemporal manner. However, in both cases, developmental delays or abnormalities were observed in areas of the cardiovascular system where cardiac neural crest formation, migration, and differentiation are critical.

The procedure used in the present study limited the knock-down effect to a specific time and place where neural tube closure and neural crest formation were the dominant developmental activities. The short-term negative impact of this procedure upon total Folr1 expression supports the finding of Saitsu et al. (2003) and Sato et al. (2008), where the site of neural tube closure/neural crest formation also is a principal site of Folr1 expression in the early (2–10 somite) embryo. The apparent rebound effect of the Folr1 expression that occurred between 48 and 96 hr subsequent its initial knockdown is consistent with the magnitude of Folr1 up-regulation that can occur within 8 hr of the onset of a folate deficiency in several cell lines or in kidney cells (Hsueh and Dolnick,1993; Said et al.,2000; Zhu et al.,2001; Sadasivan et al.,2002; Antony et al.,2004). Additionally, the expression of the folate receptor can be induced during folate insufficiency in some tissues where it is not normally expressed (Xiao et al.,2005). The increases in expression of this gene are thought to be due to both an increase in transcription rate and an increase in its RNA stability (Sadasivan, et al.,2002). Therefore, it is possible that there is a compensatory overexpression of Folr1 expression consequent to siRNA knockdown. In addition, expression of Folr1 may increase in tissues other than the neural tube as development progresses, and some of these tissues were included in the qPCR analyses.

In summary, the results of this experiment indicate that a relatively brief interruption in expression of Folr1 at a key time and place in the early embryo can impair the formation, and consequently the migration, of cardiac neural crest cells into the pharyngeal arches, thereby increasing the risk of abnormal heart development, supporting the hypothesis that normal development of the cardiac neural crest is exquisitely dependent upon the ability of nascent neural crest cells to sequester folate.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Animals

Chicken embryos from pathogen-free fertile eggs (SPAFAS, Charles River, North Franklin, CT) were incubated at 37.5°C/57% humidity, and were treated, observed or harvested at each of the following stages of development: age 29–33 hr (7–9 somites, neural fold stage, HH stages 9–10 [Hamilton and Hamburger, 1951]); age 55–59 hr (29–32 somites, HH stage 17); age 79–83 hr (43–44 somites, HH stage 21); and age 103–107 hr (all somites, HH stage 26).

Electroporation

These procedures followed the methods as outlined by Krull (2004) and previously used in our laboratory (Lie et al.,2010). Briefly, embryos at seven to nine somites (seven to nine somites, neural fold stage, HH stages 9–10) were visualized by the injection of a small amount of ink under the embryo. Then electrodes were inserted through a small hole in the vitelline membrane and were situated on either side of the embryo anterior to somite 1. Electroporation solutions were injected using the CUY21 Electroporator (NEPA Gene Company, Japan) with an attached Picospritzer III (Parker Hannifin Corporation; Fairfield, NJ). The electroporation solution contained either a Pcax-EGFP plasmid alone (4,790 bp eukaryotic expression vector containing an EGFP insert, a generous gift from Dr. C. Krull at the University of Michigan); a Pcax-EGFP plasmid with two folate receptor siRNAs, each 500 μg/ml; or the plasmid with two negative control siRNAs.

Positive siRNAs were targeted to the chicken folate receptor gene mRNA (sequence NM_204834 Gallus Gallus folate receptor 1 (adult) AKA FOLR1 (orthologous with human FRα).

Target 1.

AAG GAC TAC GTC ACA TGC AAA-offset 427 had a sense r(GGA CUA CGU CAC AUG CAA A)dTdT and an antisense r(UUU GCA UGU GAC GUA GUC C)dTdT sequence.

Target 2.

CTG GTC CAA CTC CTA CAA ATA offset 564 had a sense r(GGU CCA ACU CCU ACA AAU A)dTdT and an antisense r(UAU UUG UAG GAG UUG GAC C)dAdG sequences.

Negative control siRNAs were targeted to the sequence ATT TCT CCG AAC GTG TCA CGT.

Negative control 1.

Sense, UUC UCC GAA CGU GUC ACG UdT dT; Antisense, ACG UGA CAC GUU CGG AGA AdT dT.

Negative control 2.

Sense, UUC UCC GAA CGU GUC ACG UdT dT; Antisense, ACG UGA CAC GUU CGG AGA AdT dT.

After the control or folate receptor siRNA solutions were injected between the third and fourth somite, a transfecting charge was applied (0.5–1 ohms, 4 pulses at 20 volts for 50 ms, 50 ms between pulses). Embryos that had visible heartbeats were harvested at 24, 48, 72, or 84 hr after electroporation and were analyzed grossly and histologically; developmental defects were recorded and catalogued according to type (Rosenquist et al.,1996; Andaloro et al.,1998).

RNA isolation.

RNA was isolated from harvested embryos using the RNAqueous kit (Ambion, Austin, TX) following the manufacturer's protocol. In summary, selected regions were placed in a guanidinium salt solution diluted with ethanol and passed through a glass filter cartridge to bind RNA. RNA was recovered by a low ionic strength solution. The purity and concentration of the isolated RNA was determined by its ultraviolet-absorbance at 260/280 and 230 nm.

Real-time qPCR.

The protocol for these procedures has been published previously (Rosenquist et al.,2007). Real-time (R-T) qPCR was performed using Applied Biosystems (ABI, Foster City, CA) 7900HT fast real-time PCR system. Primers and probes were designed using Primer Express software (ABI) to span an exon–exon boundary that would result in an amplified fragment of approximately 200–500 bp. The qPCR was performed on cDNA, from the previously performed R-T reaction, using Taqman and the primer/probe sets (labeled with FAM dye and an MGB). The primer-probe sets for these studies were as follows: 18S-rRNA Fwd: tcc cct ccc gtt act tgg at; Rev: gcg ctc gtc ggc atg ta; Probe: act gtg gta att cta gag cta; Folate Receptor Fwd: tgc tcc ctg gaa gga caa tg; Rev: act ggt ccc tgt ggg ctt ct; Probe: cac ggc caa cac c. From these reactions, the gene expression plot was generated and analyzed using the SDS software (ABI).

Immunohistochemistry and Morphometry

Alternate embryos that were not used for qPCR or for whole-mount preparations were harvested intact and sections were prepared for immunohistochemistry according to a method developed for early chicken embryos (Rosenquist et al.,1996). Sections were incubated 16 hr at 4°C with primary antibody, then rinsed and incubated for 1 hr at room temperature in secondary antibody (5 μg/ml). Slides were rinsed in PBST and water and coverslipped with Vectashield Hardset (Vector Labs, Burlingame, CA). A series of slides incubated in the absence of either a primary or secondary antibody served as background controls. Concentrations and sources of the primary antibody that were used in these experiments were as follows: folate receptor, 1:100 (FR N20; Santa Cruz Biotechnology, Santa Cruz, CA); GFP, 1:400 (Invitrogen); neural crest cell marker, 20B4 supernatant 1:1 (Developmental Studies Hybridoma Bank, Iowa City, IA), Ki67 mouse IgG 1:50, mitosis detection (BD, Franklin Lakes, NJ). The secondary antibodies were either Alexa Fluor 488 or Alexa Fluor 555 (Invitrogen, Carlsbad, CA) at 5 μg/ml.

For determining the proportion of FOLR1 contained within the early embryonic neural tube at the site of electroporation, the number of red pixels per anatomic area of a given histologic region was analyzed in photographic images using Scion Image for Windows, the PC version of the public domain image processing program NIH Image.

Analysis of Embryo Phenotypes

Embryonic phenotypes and the rate of occurrence of major structural defects were analyzed from gross specimens and from reconstructed histological sections, according to our published method (Rosenquist et al.,1990,1996,1999; Latacha and Rosenquist,2005).

Proliferation Index

A Ki67 proliferation index was applied to the neural tubes of GFP-only embryos (n = 6) vs. GFP+siRNA embryos (n = 6) at 2 days after treatment (total incubation approximately 81 hr, approximately HH stage 21) (method based upon Huuhtanen et al.,1999, an analytical method used for comparisons of tumors vs. normal tissue). For each embryo sampled, 75 random cells were counted in (1) neural tube and (2) somites from each of six sections.

Data Analysis

Data were entered into an electronic database using Excel 2007 (Microsoft) and normalized to the number of deaths or defects per 10 embryos. Groups were compared using a Student's t-test modified for small groups with differing n (Rosenquist et al.,1999):

  • equation image

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

The cooperation and assistance of Dr. Catherine E. Krull is gratefully acknowledged. We also thank Janice A. Taylor and James R. Talaska of the Confocal Laser Scanning Microscope Core Facility at the University of Nebraska Medical Center for providing assistance with confocal microscopy and the Nebraska Research Initiative and the Eppley Cancer Center for their support of the Core Facility. T.H.R. and P.I. were funded by the National Heart, Lung and Blood Institute; NIH.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES