First published online January 25, 2007
Stem Cells
Vol. 25 No.
4
April 2007, pp.
1070
-1088
doi:10.1634/stemcells.2006-0397; www.StemCells.com
© 2007 AlphaMed Press
Report on the Workshop "New Technologies in Stem Cell Research," Society for Pediatric Research, San Francisco, California, April 29, 2006
 |
ABSTRACT
|
|---|
INTRODUCTION: This is a meeting report on the workshop "New Technologies in Stem Cell Research," which was presented to pediatric residents, fellows, and faculty at the Society for Pediatric Research meeting in San Francisco, California, on April 29, 2006. Four speakers presented an overview of selected topics related to the current status of methods used to study stem cells. The topics presented at the workshop focused on RNA interference, mesenchymal stem cells, expression analysis, and gene therapy. In the first report, Drs. Jerry Cheng and Kathleen Sakamoto summarize the application of RNA interference in stem cells. Second, Dr. Edwin Horwitz describes basic approaches to the isolation and purification of mesenchymal stem cells. Third, Drs. Stanislav Karsten, Lorelei Shoemaker, and Harley I. Kornblum discuss methods in expression analysis of stem cells. Fourth, Dr. Punam Malik reports on the use of gene therapy for hemoglobinopathies using autologous stem cells.
Disclosure of potential conflicts of interest is found at the end of this article.
Key Words. Stem cells • RNA interference • Expression profiles • Neural stem cells • Mesenchymal stem cells • Purification Gene therapy • Hematopoietic stem cells
|
FOOTNOTES
|
|---|
Correspondence: Kathleen M. Sakamoto, M.D., Ph.D., Division of Hematology-Oncology, Mattel Children's Hospital at UCLA, 10833 Le Conte Avenue, Los Angeles, California, 90095-1752. Telephone: 310-794-7007; Fax: 310-206-8089; e-mail: kms{at}ucla.edu
Received June 28, 2006;
accepted for publication January 5, 2007.
First published online in STEM CELLS EXPRESS January 25, 2007.
RNA Interference and Stem Cells
Jerry C. Chenga,
Kathleen M. Sakamotoa,b,c
aDivision of Hematology/Oncology, Department of Pediatrics, Gwynne Hazen Cherry Memorial Laboratories and Mattel Children's Hospital, Jonsson Comprehensive Cancer Center and
bDepartment of Pathology and Laboratory Medicine, David Geffen School of Medicine at University of California Los Angeles, Los Angeles, California, USA;
cDivision of Biology, California Institute of Technology, Pasadena, California, USA
|
ABSTRACT
|
|---|
RNA interference (RNAi) is a powerful tool with which to study gene function, especially in stem cells. Small interfering RNAs (siRNAs) can effectively be introduced either with a vehicle or through viral vectors to transiently or stably inhibit the expression of a particular gene target. Much is known about the optimization of siRNAs and method of delivery in mammalian cells. In this review, we discuss design considerations for siRNAs, methods of delivery, optimization of siRNAs, applications to study genes in stem cells, therapeutic applications, and remaining hurdles. With recent advances in RNAi, it is likely that application of this technology will increase in the future.
Disclosure of potential conflicts of interest is found at the end of this article.
Key Words. RNA interference • Stem cells • Lentivirus
|
INTRODUCTION
|
|---|
RNA interference (RNAi) describes the inhibition of gene expression by double-stranded RNAs (dsRNAs) developed in the mid-1990s [1]. Guo and Kemphues discovered that sense RNA was as effective as antisense RNA for suppressing gene expression in nematode worms (Caenorhabditis elegans) [2]. This was followed by the introduction of dsRNA into worms. When single-stranded antisense RNA and double-stranded RNA were introduced into worms, it was found that dsRNA was more effective than either strand individually in downregulating genes [1].
RNAi is a multistep process that involves the generation of small interfering RNAs (siRNAs) in vivo through the activity of the RNase III endonuclease Dicer. The resulting 21- to 23-nucleotide (nt) siRNAs mediate degradation of their complementary RNA [3]. It is now thought that RNAi induces gene silencing through various mechanisms. One is by sequence-specific targeted gene silencing. The second is through translational repression (microRNAs). Finally, it has been reported that RNAi maintains silenced regions of chromosomes [3].
|
BASIC MECHANISMS OF RNAI
|
|---|
Long dsRNAs are the precursors of the siRNAs that trigger the RNAi effect. When dsRNAs enter cells, they are cleaved by an RNase III-like enzyme known as Dicer into siRNAs (Fig. 1). These 21–23-nt siRNAs form part of a siRNA– protein complex known as RNA-induced silencing complex (RISC), which contains helicase activity that unwinds the two strands of RNA molecules, allowing the antisense strand to bind the targeted RNA [4–7]. RISC also has endonuclease activity that hydrolyzes the target RNA at the site where it binds the antisense strands. Formation of RISC is critical for mRNA degradation. Therefore, the RISC complex mediates the sequence-specific degradation of the target RNAs that contain homologous sequences to the siRNA.
View larger version (66K):
[in this window]
[in a new window]
|
Figure 1. siRNA pathways that target mRNA for degradation. Abbreviations: dsRNA, double-stranded RNA; RISC, RNA-induced silencing complex; shRNA, short hairpin RNA; siRNA, small interfering RNA.
|
|
|
WHAT IS A DESIRABLE TARGET FOR RNAI?
|
|---|
Desirable targets of RNAi include genes that are amplified or overexpressed in cells leading to a specific phenotype. Additional targets include aberrant proteins that are encoded by dominant mutant alleles. An example is oncogenes that produce transformation in mammalian cells. However, genes that are abundantly expressed or have a prolonged half-life may not be efficiently inhibited. Similarly, genes that are redundant may not be effectively downregulated.
The advantages of RNAi are that the targeted degradation is very specific and can result in variable levels of downregulation such that gene dosage effects can be studied. This technology is much easier, quicker, and less expensive than generating knockout mice. RNAi can also be used to inhibit expression of multiple genes at the same time [8–10].
|
DESIGN OF SIRNA
|
|---|
The use of siRNAs has become a common method of downregulating gene expression to screen gene function in many cell types, including stem cells. Although long dsRNAs (>30 nt) are effective in suppressing gene expression in plants, Drosophila, and C. elegans, long dsRNAs are cleaved by Dicer to form siRNAs when introduced into mammalian cells, and these siRNAs lead to mRNA degradation. However, in mammalian cells, long dsRNAs activate the interferon response pathway, leading to nonspecific mRNA degradation. The dsRNA-dependent protein kinase (PKR) is activated, resulting in nonspecific translational inhibition [11, 12]. Therefore, the usefulness of dsRNA in mammalian cells is limited.
In general, 21–23-nt siRNAs are too short to activate the nonspecific dsRNA response pathway, but they are effective in inhibiting the expression of specific targets. There are several limitations of using this technology in mammalian cells. In fungi, plants, and worms, siRNAs can be replicated in vivo. In mammalian cells, siRNAs do not prime the synthesis of dsRNA to form additional siRNAs, which may explain why this technology is less effective [9]. Nevertheless, there are several examples in which siRNAs are effective in a variety of mammalian cell types, including stem and progenitor cells [1, 13].
Optimization of siRNAs in mammalian cells is dependent on several factors. One is the accessibility of the target sequence to the desired mRNA substrate. Previous reports have suggested that selecting a target sequence 100–200 nts away from the translational initiation sequence AUG of the gene is desirable [1]. However, successful inhibition of gene expression has also been reported for siRNAs targeting various sequences, including the 3' untranslated region [14]. Targeting of the 3' untranslated region is also useful if rescue experiments are to be performed. There is no reliable way to predict or identify the ideal sequence for siRNA. Several reports have suggested that sequences that form the stems of the hairpin siRNAs, the loop size, and the sequences at the base of the loop might also affect siRNA-induced gene inhibition. Other determinants include thermodynamic stability; siRNA with lower thermodynamic stability for base pairing at the 5' end of antisense (guide) strand and in the middle of the siRNA were more effective at RNAi than those that had stronger base pairings in these regions due to affects on uptake of guide strand into RISC and enhancing RISC binding to target mRNA.
The sequence of siRNAs should be carefully designed. The number of nucleotides should be between 19 and 23. The GC content should be between 30% and 50%. The preferred format is AAN19TT. Sequence specificity to at least two nucleotides should be confirmed by Blast comparison of the National Center for Biotechnology Information GenBank database. Finally, one should query against the single nucleotide polymorphism database [10].
|
OPTIMIZATION OF SIRNA
|
|---|
To ensure that the gene of interest is effectively downregulated by the siRNA, it is now recommended that at least three different siRNA sequences per target be designed [15, 16]. More robust knockdown of genes has been reported using this approach of creating "multiplicity" controls. Inhibition of expression has been reported for up to 5–10 days when using "pools" of siRNAs in transfected cells.
siRNA concentrations must also be optimized. In general, concentrations of siRNAs greater than 100 nM are considered to be toxic. Various amounts of siRNAs should be tested for each specific cell type. This should be considered when one is using multiple siRNA sequences. Multiple cell lines should also be tested to validate response and downregulation. Finally, a nucleotide Blast search should be performed to determine whether the siRNA sequence would target another gene. In terms of controls, scrambled or mutated sequences (http://www.sirnawizard.com) and unrelated genes (e.g., luciferase) are commonly used. To validate successful downregulation of the target gene, it is recommended that a Western blot analysis be performed to assess protein levels and Northern blot analysis or reverse transcription-polymerase chain reaction (RT-PCR) to measure RNA levels. Demonstration of lower mRNA levels is critical to rule out a microRNA effect and translational inhibition of gene expression. To control for off target effects, one can measure interferon response genes, including OAS1, OAS2, and INFB1, by RT-PCR [1].
|
DELIVERY OF SIRNA TO CELLS
|
|---|
In mammalian cells, efficiency of siRNA to cells transiently depends on the vehicle or mode of delivery and the cell types. Approaches to introduce siRNAs into cells include a lipid-based vehicle (e.g., Lipofectamine) or a non-lipid-based approach (e.g., calcium phosphate or electroporation). The disadvantages of this approach are that the siRNAs are nonrenewable and are only effective as long as they are bath-applied to cells. An alternative strategy has been to deliver siRNAs through a DNA vector-mediated RNAi approach.
Because of the transient nature of gene silencing produced by oligonucleotide siRNAs and their high costs of chemical synthesis, alternative approaches to introduce siRNAs in plasmid vectors have been developed. A variety of expression vectors are now available. Expression is driven by either the U6 (small nuclear RNA) or H1 RNA polymerase III promoters to drive expression of sequence-specific short hairpin RNAs (shRNAs) in mammalian cells [2]. These systems are based on the expression of siRNAs either as two separate strands or as a single shRNA. It is thought that the shRNAs are processed by Dicer to active siRNAs in vivo [17–19].
For stable expression in stem cells, successful delivery has been demonstrated with viral vectors. Various recombinant viral vectors have been developed to deliver shRNAs in mammalian cells [10, 20]. Lentiviral vectors are especially effective. The reasons for this are that lentiviruses have broader tropism and receptor-independent delivery, that they have the ability to integrate into the genome for stable gene silencing, and that lentiviral transduction and expression of shRNAs do not require cell division for integration into the genome [21]. Lentiviral transduction has been successfully performed in cell lines, mouse hematopoietic stem cells (HSCs), and embryonic stem (ES) cells [22–24].
Adenoviral vectors have also been reported to be useful for delivering siRNAs to target cells. This vector system has been used to downregulate genes in liver. However, this vector system has limited utility in stem cells, since low transduction rates have been found in ES cells and HSCs. This is most likely due to the fact that the receptor for adenovirus is not highly expressed in stem cells [25]. Similarly, adenoviral-associated vectors have been successfully used to deliver RNAi to nonstem cells [1].
If the stable transfection or transduction of siRNAs results in toxic effects to cells, an alternative approach is to use the inducible expression of shRNAs. The tetracycline/doxycycline regulated form of U6 or H1 promoter has been successfully used. If there is leakiness, other inducible systems, such as an ecdysone-inducible system, are more tightly regulated with less background. A newer approach has been a CRE-lox-inducible system [26]. Most recently, a doxycycline-inducible vector that contains a KRAB domain from one-third of zinc finger domains was used in cell lines, mouse ES cells, epithelial breast cancer cells, rat brains, CD34+ cells, and transgenic mice [27].
|
APPLICATION OF RNAI IN STEM CELLS
|
|---|
There is now emerging evidence that RNAi can be used to study gene function and for therapeutic application. ES cells are pluripotent stem cells that are derived from the inner cell mass of the 3.5-day-old mouse blastocyst [1, 28]. These cells are desirable models to study the regulation of development and cell lineage commitment and differentiation, since ES cells can give rise to all three germ layers. This system is a powerful tool with which to study development.
Interestingly, long dsRNA has been used in ES cells, but only when undifferentiated. The reason for this is unknown. In differentiated ES cells, siRNAs have been found to be effective in inhibiting genes, such as PU1 and c-EBPa [1]. A variety of other genes have been downregulated in ES cells, such as Shp-2 and Oct-4. Synthetic shRNAs recently have been shown to be efficiently transfected transiently with Lipofectamine [29]. More commonly, viral vector systems have been used to transduce genes of interest for stable expression of shRNAs.
HSCs are a self-renewing population of cells in the bone marrow that gives rise to all differentiated hematopoietic cells [1]. A number of genes have been targeted using RNAi in HSCs. Growth factor receptor genes, clusters of differentiation, chemokines, oncogenes (bcr-abl), tumor suppressors, human immunodeficiency virus genes, globin genes, and RPS19 expression have all been successfully targeted. In most cases, retroviral or lentiviral vector systems were used. Electroporation has been used successfully to introduce dsRNA in HSCs [13]. Lipofectamine has also been reported to effectively transfect oligonucleotide siRNAs into hematopoietic progenitor cells [30]. HSCs that are transduced with shRNAs can then be studied in vitro using methylcellulose colony assays or in vivo in bone marrow transplantation experiments.
|
NEURAL STEM CELLS AND MSCS
|
|---|
Neural stem cells (NSCs) have also been transduced with shRNAs to downregulate genes. Examples of genes inhibited in NSCs by RNAi are MELK, PPAR
and B27.a genes [31–33]. MSCs have been studied using both viral and nonviral methods. Genes inhibited using viral vectors were ß-catenin, Msx2, and mecdin [2, 34]. Nonviral liposomal methods to introduce siRNAs into MSCs have been used to inhibit epidermal growth factor receptor and connective tissue growth factor [35, 36]. Recently, a transfection microarray approach was generated in which siRNAs were applied onto slides that are coated with poly-L-lysine and fibronectin. MSCs were then placed on top of the poly-L-lysine and siRNA sandwich. Fluorescent microscopy was used to then visualize and quantify the degree of downregulation [37, 38]. A similar approach was used with HeLa cells placed on slides treated with siRNAs, in which cells were then followed in real time using time-lapse fluorescent microscopy as a high-throughput method to screen for genes involved in chromosomal segregation [39].
|
SHRNA LIBRARIES
|
|---|
One of the technological advances in the RNAi field has been the development of shRNA libraries to screen for genes that regulate a specific pathway or biological function. Many of the libraries rely on lentiviral vector-based expression. Libraries have been used to identify deubiquitinating enzymes [40], sensitivity to small molecule inhibitors, novel cancer genes, and previously unidentified components of signaling pathways. A recent report from the Broad/Massachusetts Institute of Technology group (The RNAi Consortium) used an shRNA library with 72,600 clones targeting 10,500 human and 5,300 mouse genes [41]. It is anticipated that the numbers of genes targeted could be as high as 15,000 human or mouse genes. Viruses expressing shRNAs can be transiently or stably transduced into mammalian cells [41]. Genes that are involved in a particular cellular process will be identified through identification of the shRNA clones that block the function of the gene. An inducible shRNA library has also been used recently to identify genes that regulate proliferation or survival of diffuse large B cell lymphoma cells to seek novel targets for therapy [42].
|
THERAPEUTIC APPLICATIONS OF RNAI
|
|---|
The field of RNAi is advancing at a rapid pace. The application of RNAi as gene therapy is now being realized. In mice, delivery of siRNA to downregulate Fas by hydrodynamic tail injection resulted in protection from fulminant hepatitis [43]. A recent report by Samakoglu et al. has demonstrated that sickle globin gene can be downregulated in CD34+ cells using a lentiviral shRNA, with a concomitant increase in -globin expression in erythroid-specific manner [44]. Another advance has been the successful RNAi-mediated gene silencing in nonhuman primates. The first report of systemic delivery of APOB siRNA in nonrodent species was recently reported [45]. APOB is a component of low-density lipoprotein (LDL) and regulates the storage and metabolism of cholesterol. A liposomal formulation of APO-B siRNAs was intravenously administered into cynomolgus monkeys with effective inhibition of APOB levels after 48 hours and 11 days. Plasma levels demonstrated that not only LDL and cholesterol levels were lower than controls, but high-density lipoprotein levels were not affected. Although previous success was shown with hydrodynamic tail injection of oligonucleotide siRNAs in rodents, this was the first report of siRNAs successfully targeting a gene in nonrodent models.
|
REMAINING CHALLENGES
|
|---|
Although the field of RNAi has progressed rapidly, there are several hurdles that remain before this technology can be fully applied in human
Top
Abstract
Abstract