SEARCH

SEARCH BY CITATION

Keywords:

  • Cripto;
  • Embryonic stem cells;
  • Tumor;
  • Parkinson's disease;
  • Behavior

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgements
  8. References

Embryonic stem (ES) cells have been suggested as candidate therapeutic tools for cell replacement therapy in neurodegenerative disorders. However, limitations for the use of these cells lie in our restricted knowledge of the molecular mechanisms involved in their specialized differentiation and in the risk of tumor formation. Recent findings suggest that the EGF-CFC protein Cripto is a key player in the signaling pathways controlling neural induction in ES cells. Here we show that in vitro differentiation of Cripto−/− ES cells results in increased dopaminergic differentiation and that, upon transplantation into Parkinsonian rats, they result in behavioral and anatomical recovery with no tumor formation. The use of knockout ES cells that can generate dopamine cells while eliminating tumor risk holds enormous potential for cell replacement therapy in Parkinson's disease.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgements
  8. References

Cell replacement therapies have focused mainly on the use of human fetal mesencephalic tissue transplantation in patients with Parkinson's disease (PD). However, technical difficulties and ethical considerations have prevented the widespread development of this technique. On the contrary, embryonic stem (ES) cells have many characteristics required for an optimal cell source for cell replacement therapy in PD [13]. However, the molecular mechanisms as well as the signaling pathways implicated in neural generation in the context of stem cells have not been extensively evaluated and need to be further characterized. Indeed, both the growth and the differentiation potential of ES cells need to be controlled, and the risk associated with the growth of non-neural tissues needs to be eliminated.

The differentiation of mammalian ES cells in culture has been reported to follow the hierarchical set of signals that regulate embryonic development in the generation of the germ layers and specification of cell types. Moreover, mechanisms operating during neurogenesis in embryonic development are thought to dictate the neuronal population being generated from ES cells [46]. For instance, dilution of ES cell concentration in vitro facilitates neural differentiation [7]. Interestingly, this effect could be mimicked by bone morphogenetic factor inhibitors such as noggin [8] and Cerberus [9], as well as by using ES cells with a targeted mutation of Smad4 gene [7]. According to the in vitro observations, the transplantation of low doses of ES cells in a Parkinson rat model resulted in neuronal dopamine(DA)–containing grafts, supporting the notion that establishment of neural identity from uncommitted mammalian ES cells occurs by default. However, these grafting experiments also showed that lack of engraftment due to insufficient cells or tumor formation accounted for 44% of the animals [10], indicating that the implementation of such strategy in therapeutics remains an obstacle.

Recent data have highlighted a key role of the GPI-anchored EGF-CFC Cripto protein in preventing neural differentiation of ES cells [11]. Cripto is the original member of the EGF-CFC family defined by two conserved adjacent motifs: an epidermal growth factor (EGF)–like domain and a unique cysteine-rich domain, the CFC domain [12]. Both genetic evidence and biochemical approaches demonstrate that Cripto acts as a coreceptor for the transforming growth factor β ligand Nodal [13, 14]. Cripto-dependent Nodal signaling acts through the Activin type I serine/threonine kinase receptor (Alk4) and the type II receptor (ActRII) that, once activated, phosphorylates the downstream transcriptional coactivator Smad2 [1517]. Knockout mice for the cripto gene show early embryonic lethality due to gastrulation defects and are mostly constituted by anterior neuroectoderm [18]. Furthermore, embryoid bodies (EBs) derived from Cripto-deficient (Cr−/−) ES cells selectively lose the ability to form beating cardiac myocytes in vitro and show an enhanced neural differentiation ability, thus suggesting that Cripto could represent a key molecule required for both induction of cardiomyocyte differentiation and repression of neural differentiation [11, 19].

Thus, it remains to be determined whether increased neural induction in Cr−/− ES cells may contribute to the generation of specific neuronal types such as dopaminergic neurons that could be used in therapeutic applications. Of significant concern regarding ES cell replacement therapy is the risk of teratoma formation. Worth noting, Cripto is also overexpressed in a wide range of epithelial cancers, including breast, pancreatic, ovarian, and colon carcinomas, and, more recently, antibody blockade of Cripto has been shown to suppress tumor cell growth [14, 2022]. We therefore hypothesize that inhibition of Cripto signaling in Cr−/− ES cells, in addition to directing ES cells to increased neural commitment, may block teratoma formation, thus enhancing the therapeutic potential of ES cells.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgements
  8. References

ES In Vitro Differentiation

Both Cr+/+ and Cr−/− ES cells [11, 19] were cultivated as EBs, as previously described [23, 24]. EBs were plated on gelatin-coated plates and allowed to differentiate in medium alone or supplemented with sonic hedgehog (Shh) (500 ng/ml, R&D Systems Inc., Minneapolis, http://www.rndsystems.com) and/or fibroblast growth factor-8 (FGF-8) (20 ng/ml, R&D Systems). Differentiated EBs were fixed in 4% paraformaldehyde and stained with the following primary antibodies: mouse βIII-tubulin (Tuj1) (1:1,000, Promega, Madison, WI, http://www.promega.com) and rabbit anti-tyrosine hydroxylase (anti-TH) (1:150, Pel-Freez, Rogers, AZ, http://www.pelfreez-bio.com/home.html). Appropriate cyanin-2 (Cy-2) and Rhodamine-labeled secondary antibodies (Jackson Immuno Research, West Grove, PA, http://www.jacksonimmuno.com) and Hoescht counterstaining were used for visualization. The proportion of neurons (labeled for using the immature neuronal marker βIII-tubulin [Tuj1-ir]) within the colonies was noted, and the number of dopaminergic neurons per square micrometer of EB (identified using a marker for the precursor enzyme in DA synthesis, TH [TH-ir]) was counted. Differentiation experiments were performed in triplicate.

RNA Preparation and Reverse Transcription–Polymerase Chain Reaction

Total RNA was extracted using TRIzol reagent (Gibco Invitrogen, Carlsbad, CA, http://www.invitrogen.com). Reverse transcription–polymerase chain reaction (RT-PCR) was performed as previously described [25]. HPRT was used as internal control.

6-OHDA Lesioning and Rotational Behavior

Eighteen adult male Sprague-Dawley rats (250–350 g; from B&K Universal AB, Sollentuna, Sweden) were housed and treated according to the guidelines of the European Community and local ethics committee. Animals, anesthetized with sodium pentobarbital (60 mg/kg i.p.), received unilateral stereotaxic injections of 6-hydroxydopamine (6-OHDA) (42 μg) into the substantia nigra pars compacta (SNpc) as previously described [26]. Lesioned animals were selected for transplantation based on their response to amphetamine-induced rotational behavior. Motor asymmetry was quantified 6 days after lesioning and then weekly after grafting was performed. Circling behavior was measured for 3 minutes at 15, 30, 45, and 60 minutes after injection. Animals making greater than five rotations per minute (lesion greater than 80%) were selected for grafting.

Transplantation of ES Cells

ES cells were prepared as previously described [10]. The day before grafting, animals began immunosuppression (15 mg/kg daily, cyclosporine-A, Novartis, Basel, Switzerland, http://www.novartis.com) and then received cyclosporin daily (10 mg/kg) for the duration of the study. One week after lesioning, selected animals were again anesthetized and stereotaxically injected with 1 μl of cell suspension (1,000 cell/μl) at two sites (from bregma: anterior 1 mm, lateral 3.0 mm, ventral 4.5 and 5.0 mm, incisor bar 0 mm) according to Paxinos and Watson [27]. Cells were injected slowly through a 22-gauge, 10-μl Hamilton syringe, and the needle was left in place for 2 minutes after injection. Five rats received control Cr+/+ ES cells, five rats received Cr−/− ES cells, and eight rats were sham injected (culture medium).

Immunohistochemistry and Cell Counts

Seven weeks after grafting, rats were killed by an overdose of sodium pentobarbital (100 mg/kg−1 i.p.) and perfused with 400 ml of 0.1 M phosphate-buffered saline (PBS) (pH 7.4; 37 °C) followed by 125 ml of warm 4% paraformaldehyde (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) in 0.1 M PBS and 0.2% picric acid (4 °C; pH 7.4) and then 125 ml of chilled 4% paraformaldehyde. Brains were removed, postfixed for 1 hour, and then left overnight at 4 °C in 20% sucrose and PBS solution. The following day, 20-μm-thick coronal sections were cut serially through the striatum (12 series) and 50 μm through the SNpc (4 series) and mounted directly onto superfrost slides. SNpc lesion size was confirmed in all animals using a fractionator sampling design to determine the number of (0.25%) cresyl violet–stained neurons in the SNpc, as previously described [26].

Striatal sections were stained with mouse-specific surface antigen antibody, M2 (1:100, Hybridoma Bank, Iowa City, Iowa), to identify graft tissue followed by Cy-2 secondary. TH immunohistochemistry was performed as described above or using visualization diaminobenzidine. Throughout this study, analysis of variance with Tukey's post hoc tests was used with statistical differences set at the level of p ≤ .05 unless stated otherwise.

Results and Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgements
  8. References

Cripto−/− ES Cell Differentiation Increases DA Cells by Increasing Neuronal Precursors

To establish whether Cr−/− cells have increased potential to give rise to dopaminergic cells, we first examined their differentiation potential in vitro in the presence of Shh and FGF-8, two factors important for the generation of ventral midbrain DA neurons. As previously reported [11], most (>70%) of Cr−/− ES cell–derived EBs showed regions of high Tuj1 labeling, which were also substantially greater than Cr+/+ EBs (Figs. 1C, 1G). Treatment of cultures with Shh and/or FGF-8 dramatically increased the proportion of EBs expressing Tuj1 and increased the proportion of the EBs showing immunoreactivity, particularly in Cr−/− cultures treated with Shh/FGF-8 (Figs. 1C–J). Examination of TH-ir cells within the cultures revealed that in the absence of cripto, cells more frequently differentiate into TH-ir cells (Figs. 1A, 1K). Indeed, Cr−/− showed a 24-fold increase in the number of TH cells compared with Cr+/+ ES cell cultures. In addition, the number of TH-ir cells could be additionally increased (up to 102%) by the addition of Shh or Shh plus FGF-8 (Figs. 1A, 1K–N).

thumbnail image

Figure Figure 1.. In vitro differentiation of Cr−/− ES cells increases the number of dopamine cells by increasing neuronal precursors. (A): Number of TH-ir cells per square micrometer of EB revealed that Cr−/− ES cell cultures differentiate more preferentially into TH-ir cells than Cr+/+ and that these effects could be increased by addition of Shh and/or FGF-8 (mean ± SD; p < .0001). (B): Polymerase chain reaction confirms upregulation of TH in differentiated Cr−/− EBs and after Shh treatment. Other markers involved in DAergic development and maturation were also seen to be upregulated in Cr−/− cultures. (C–J): Tuj1/Hoescht staining in 7-day differentiated Cr−/− and Cr+/+ EB cultures following control conditions and treated with Shh and/orFGF-8. (K–N): TH-ir staining in Cr−/− cultures following control conditions and treated with Shh and/or FGF-8; note substantial increase in TH-ir cells after Shh/FGF-8 treatment compared with control conditions. Abbreviations: Cr, Cripto; DAT, dopamine transporter; EB, embryoid body; ES, embryonic stem; FGF-8, fibroblast growth factor-8; Shh, sonic hedgehog; TH-ir, tyrosine hydroxylase immunoreactive.

Download figure to PowerPoint

RT-PCR confirmed the histochemical observations. mRNA levels for TH were greatly increased in Cr−/− cultures, with even greater levels seen in those cells treated with Shh. Furthermore, DA transporter (DAT) levels were increased in Cr−/− cells, with DAT being a marker turned on late in DA cell differentiation. Morphologically, the cells also showed a typical bipolar mature shape (Fig. 1), indicating that not only were these cultures rich in TH and DAT but that in fact the cells develop a mature phenotype. PCR also revealed increased mRNA levels of Wnt1 both in untreated and Shh-treated cultures. We recently showed that Wnt1 increases TH cell number by regulating Nurr1-ir precursor proliferation [28]. These results suggest that suppression of Cripto not only increases the neuronal pool for differentiation into various neural subtypes but also has effects on Wnt1 expression levels, which may result in proliferation and differentiation of DA precursors into neurons. Finally, c-ret, the tyrosine kinase receptor for glial cell–derived neurotrophic factor that regulates DA neuron survival and differentiation, was also expressed at higher levels in Cr−/− ES cells. Thus, our results show that Cr−/− ES cells not only have increased capacity to generate neurons but can also generate increased numbers of TH-ir neurons.

Cripto−/− ES Cell Grafts Do Not Give Rise to Teratoma Formation in Parkinsonian Rats

One of the greatest tribulations that exists for ES cell replacement therapy is the development of teratomas from undifferentiated ES cells present within the grafted population. In this regard, Cripto is known to be overexpressed in several tumors [14, 20, 22]. We therefore wished to examine whether Cr−/− ES cell grafts formed teratomas in vivo with the hypothesis that suppression of the Cripto protein may result in reduced tumor formation. We noted that all Cr+/+ ES cell–grafted animals developed teratomas. Some tumors were small and surrounded with dense TH areas (Fig. 2A), whereas other tumors were extensive (Figs. 2B, 2C), invading forebrain, midbrain, and hindbrain structures. Surprisingly, none of the animals receiving Cr−/− ES cell grafts showed any tumor-like formations by 7 weeks after grafting (Fig. 2D).

thumbnail image

Figure Figure 2.. Cr+/+ ES cell grafts result in teratoma formation in Parkinsonian rats. (A): A small tumor seen within the striatum of a Cr+/+ ES cell–grafted rat, surrounded by TH-ir dense region. (B): Large teratoma from a Cr+/+ ES cell–grafted rat. (C): One of many TH-ir dense regions seen within the tumor formation shown in (B). (D): Example of a Cr−/− ES cell–grafted striatum showing no tumor formation and extensive TH innervation. Abbreviations: Cr, Cripto; ES, embryonic stem; TH-ir, tyrosine hydroxylase immunoreactive.

Download figure to PowerPoint

Several rodent studies have shown tumor formation after ES cell transplantation [10, 25]. The use of different ES differentiation protocols [25, 2932] and cell sorting [33] have reduced the risk of tumor formation after cell replacement therapy in Parkinsonian rodents; however, the use of ES cells for therapy is presently still unfeasible. Finding a gene that regulates proliferation and being able to manipulate cells before or after implantation with a “switch” to prevent tumor formation has become a key quest in the development of ES cell replacement therapies. The results shown here illustrate that gene knockout technology could be used to suppress teratoma formation and to enhance ES cell capacity to differentiate into neurons.

Cripto−/− ES Cell Grafts Result in Anatomical and Behavioral Improvements in Parkinsonian Rats

Given the ability for Cripto-deficient ES cells to give rise to increased number of TH cells in vitro, we wished to assess the ability of Cr−/− ES cell transplants to restore function in Parkinsonian rats.

We created an animal model of PD by injecting the neurotoxin 6-OHDA into the right SNpc of rats to produce a complete lesion. The number of SNpc neurons was counted in untreated and 6-OHDA–treated rats to determine lesion size. In total, untreated Sprague-Dawley rats had 12,161 ± 780 SNpc cells. All lesioned animals showed between 87% and 100% loss of SNpc cells, with no significant difference seen between the lesioned groups (Fig. 3A), confirming our behavioral selection of animals for grafting.

thumbnail image

Figure Figure 3.. Cr−/− ES cell grafts result in anatomical and behavioral improvements in Parkinsonian rats. (A): Total number of SNpc cells in untreated and 6-hydroxydopamine–lesioned animals, confirming that extensive lesions were created in all grafted animals, sham, Cr+/+, and Cr−/−. (B): Number of TH-ir cells seen within the striatum of sham, Cr+/+, and Cr−/−-grafted animals. (C–E): Photomicrographs of grafts from sham-operated, Cr+/+-grafted, and Cr−/−-grafted rats. Grafted tissue could be detected by M2-ir (green), a mouse-specific antibody, and TH-ir cells (red) could be seen throughout Cr+/+ and Cr−/− grafts. (F, H): High-power pictures of grafted cells stained with TH-ir, illustrating typical bipolar mature dopamine morphology with cells also making extensive connections. (G, H): An example of a smaller, more isolated graft seen in a Cr−/−animal,TH-ir.(I):Amphetamine-induced rotational behavior in sham, Cr+/+, and Cr−/− ES cell–grafted animals; note the behavioral improvements seen in ES cell–grafted animals compared with sham-operated animals. Data are mean ± SD. Abbreviations: Cr, Cripto; ES, embryonic stem; SNpc, substantia nigra pars compacta; TH-ir, tyrosine hydroxylase immunoreactive.

Download figure to PowerPoint

Striatal examination revealed that grafts had survived in 80% of animals, shown by the presence of cells specifically stained with antibodies against mouse glia (M2). As expected, no M2-ir or TH-ir cells were observed within the striatum of sham-operated animals, with only a few remaining TH-ir fibers present (Fig. 3C). Total counts of TH-ir cells in ES cell–grafted animals showed no significant difference, with 9,918 ± 2,968 TH-ir cells found in Cr+/+ ES cell grafts and 8,467 ± 1,056 in Cr−/− grafts (Figs. 3B, 3D, 3E). In some instances, TH-ir cells were seen to be dispersed throughout the striatum (Figs. 3D, 3E), whereas other grafts showed pockets of TH-ir cells (Figs. 3G, 3H). These results showed that the grafts survived and had the ability to differentiate into TH-ir cells, as previously reported [25].

We finally wished to establish the functionality of these TH cells by assessing behavioral improvements in the ES cell–grafted Parkinsonian animals. Behavioral assessment of the animals revealed a nonsignificant improvement (17%) in sham-operated animals (Fig. 3I), which may be explained by modest compensatory mechanisms such as collateral sprouting and hypertrophy of remaining fibers. Conversely, both Cr+/+ and Cr−/− ES cell–grafted animals showed significant increases in behavior by 6 weeks (63% and 67%, respectively), and unlike Cr+/+-treated rats, Cr−/−-treated rats continued to improve, showing a 74% increase by 7 weeks. (Fig. 3I). It is worth mentioning that Cr+/+ cells, despite generating similar numbers of TH-ir cells as Cr−/− grafts, did not result in the same degree of functional recovery. The slow differentiation of Cr+/+ cells in vitro correlated well with the slow behavioral recovery of Cr+/+-grafted animals during the first 4 weeks compared with Cr−/− grafted animals. By weeks 5 to 6, the behavior of Cr+/+ was similar to Cr−/− grafted animals, suggesting that a significant number of dopamine-producing cells was generated. However, by week 7, the overgrowth of the tumors in these Cr+/+-grafted rats resulted in reduced behavioral recovery. Interestingly, those animals having the larger tumors showed the greatest deterioration in behavioral improvement, reflecting the damage caused by inflammation and compression of adjacent structures. We thus conclude that the beneficial effects of the TH+ cells derived from the tumors are compromised by the growth of the tumor. This was clearly not the case for Cr−/− cells, which showed behavioral recovery in the absence of tumor formation. Overall, our results demonstrate that the striatal TH-ir cells observed within the Cr−/− grafts are functional, able to restore rotational behavior in Parkinsonian rats, and, more important, do not lead to the formation of teratomas.

Conclusions

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgements
  8. References

Our data suggest that Cripto is an important regulator of ES cell fate. In this study we show that suppression of cripto allows cells to spontaneously differentiate to a neural fate, increasing the cellular pool for dopaminergic differentiation. Furthermore, our data revealed that Cr−/− cells, when grafted at low density, can generate TH cells and restore behavior in animal models of PD. More important, we show that suppression of cripto reduces tumor formation, providing important new insights into the use of ES cell replacement therapy for PD. Thus, our findings open the door to the future development of knockout ES cells as a tool in regenerative medicine and to the development of drugs to inhibit Cripto-receptor interaction as an alternative means to direct cell fate in ES cells.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgements
  8. References

This study was supported by the Swedish Foundation for Strategic Research, Swedish Royal Academy of Sciences, Knut and Alice Wallenberg Foundation, European Commission (Euro Stem Cell Program), Juvenile Diabetes Research Foundation, Swedish MRC, and Karolinska Institutet (to E.A.), and the Associazione Italiana Ricerca sul Cancro (AIRC) (to G.M.), and Ministero Istruzione Universita' e Ricerca (MIUR-progetto FIRB) (to M.G.P.). Clare L. Parish is a Human Frontiers Science Program Long-Term Fellow at Karolinska Institute. We thank Anna Aliperti for proofreading the manuscript. C.L.P. and S.P. contributed equally to this study as co-first authors; E.A. and G.M. contributed equally to this study as co-last authors.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results and Discussion
  6. Conclusions
  7. Acknowledgements
  8. References