SYNOPSIS: In this proposal, the Principal Investigator (PI) plans to utilize different human embryonic stem cells (hESCs) in a common model of injury, in this case, a spinal cord injury. Over the years, multiple different stem cell lines have been generated, but they were never reliably and accurately compared side-by-side along different lineages with regard to cell fate after differentiation and with functional outcome. This application seeks to determine the effects that different methods of generating neural stem cells in vitro will have on the ability of these cells to survive transplantation into the injured spinal cord, integrate within the damaged host tissue and mediate any functional recovery.
IMPACT & SIGNIFICANCE: There have been no systematic studies of the relative survival, engraftment, fate, and efficacy of different multipotent neural stem cells versus the lineage-restricted types after transplantation in an injury model. Therefore, a systematic approach such as this is really quite critical if one hopes to use stem cells as therapeutic tools (which is no different than optimizing a new chemical based on structure activity relationships for venture use in small molecule drug delivery). Thus, in general, this approach is timely.
Overall, the use of multiple stem cell lines is not particularly original scientifically and none of the experiments in this proposal are using novel or original methodologies to derive stem cells or to engraft them. However, that is not really the point to this proposal. The main aim is to systematically evaluate different stem cell lines and their derivation in a common neural injury model and, although not original, it is an important scientific approach; therefore originality carries little weight in the decision regarding the significance/merit of this research plan.
This is one of the first attempts to actually characterize whether or not the derivation of different stem cell populations leads to equally effective outcomes after injury and transplantation. In that sense, this is a novel and innovative set of experiments. Assuming that the different methods for deriving these cells does lead to significant differences in gene expression (which is a big assumption) whether the expression of a particular set of genes is responsible for the functional outcome once transplanted will be based strictly on correlations. Additionally, because a cell line expresses certain genes at a particular level in vitro does not necessarily mean they will express those same genes (versus others) when placed into a more complex environment such as the injured spinal cord. The applicant will be able to determine what effects this “niche” will have on cell differentiation, but these will be more descriptive than mechanistic studies since the niche will remain relatively undefined. Thus while a great deal of data will be generated, the impact that the in vitro data will have on the field will likely be limited. Although we may still not know why, we can hope that the in vivo studies will demonstrate that a particular cell line when derived in a specific way responds to the environment of the injured spinal cord so as to maximize cell-directed functional recovery.
The strength of this application is that it targets an important problem which is the question of whether generation of hESCs by different means affects their utility in the generation of specific cell types and repair of tissue injury.
QUALITY OF THE RESEARCH PLAN: This application is focused on the generation of human neural stem cells (hNSCs) from hESCs by 3 different methods to determine optimal protocols for survival, integration, and generation of desired cell fates in the host niche. Aim 1 tests the hypothesis that non-tumorigenic hNSCs can be derived from hESCs. Aim 2 tests the hypothesis that the methods by which hNSCs are generated is critical to the phenotypic characteristics of these cells in responding to the injured central nervous system (CNS) niche, including migration and cell fate potential after transplantation into the injured CNS, with the presence of serum during generation biasing cells towards an astrocytic fate. The rationale for such experiments is that promotion of recovery will be related to cell fate predisposition. Aim 3 tests the hypothesis that in the absence of FACS sorting or other specific derivation methods, the hNSCs generated will be biased towards production of endothelial cells. Aim 4 tests the hypothesis that hNSCs derived in appropriate ways will exhibit greater responsiveness to the niche presented by the injured CNS than hESC-derived progenitor-restricted cell populations, leading to improvements in functional recovery after in vivo transplantation.
One reviewer considers this is a well-organized research plan that includes some appropriate and convincing preliminary data. Also, the PI has prior experience and good collaborators are included, which all suggest that the goals are achievable. Collaborators include Peter Donovan who is highly experienced in stem cell derivation and Dr. Lesley Lock, who is experiencing in handling hESC.
The first Aim uses three previously described methods to derive hNSCs from 2 hESC lines, giving a total of 6 hNSC lines, followed by assessment of self-renewal capacity, FISH screening for chromosomal aberrations, analysis of ability to generate astrocytes, neurons or oligodendrocytes, and analysis of gene expression by microarray. No hypotheses are tested beyond different derivation protocols yielding different outcomes. Neuronal characterization is described by a reviewer as superficial. Interpretation of microarray analysis is not at all clear to the reviewer as to the questions asked or how they will be answered.
The second Aim takes the cell lines generated in Aim 1, transplants them into the contusion-injured spinal cord and performs BBB analysis, locomotor recovery, assays of allodynia, and histology. This aim will utilize methodologies and analyses with which the investigators are familiar. Importantly, the investigators include experiments to evaluate the fate of cells using a variety of immune markers for neurons, glia and will also include quantitative methods, for example stereology. The importance of the transplanted cells will be assayed by injection of diphtheria toxin to selectively ablate human cells (which are more sensitive). No alternative hypotheses are presented. Analysis of inflammation by protein and cDNA is discussed, but rationale and interpretation are not provided.
In the third Aim, the investigators will utilize multipotent human neural stem cells derived from human fetal neurospheres, not selected by FACs sorting, but by using culture conditions that include growth factors (e.g. EGF/FGF-2). The PI argues that these cells will respond to the injury “niche” and will be more likely to differentiate into endothelial cells or cells that are “necessary” for the injury niche. Reviewers find this aim the most difficult to follow and convoluted, suggesting that the author should be much more straightforward in describing experiments rather than use somewhat abstract descriptions. As best one reviewer can tell, the PI hypothesizes that the most undifferentiated cells in this engraftment will respond better, or at least differently, than more committed progenitors. This is not a new idea (often espoused by the Snyder lab). Multiple published experiments to date clearly show that when more undifferentiated cells are used in injury engraftment, the ultimate outcome is mostly glia, some poorly differentiated small neurons, and many undifferentiated cells remain. This aim will use skid mice and injection of stem cells via the bloodstream. This aim is certainly the shakiest because of the lack of preliminary data on the use of this mouse model and behavior analysis in terms of spinal cord injury and the ability to fully quantify results. A reviewer points out that there is no analysis of the possibility that mixed populations can be separated, and no experiments to look at specific effects of endothelial cells (if they are generated).
This is a very comprehensive research plan that will involve a significant amount of effort for all of the proposed in vitro and in vivo studies. The studies rely on previous published methods within the expertise of the applicant or the collaborators. It is likely that a great deal of data will be generated. The most meaningful data will be generated from the in vivo studies while the in vitro data will be descriptive at best. All of the aims depend on being able to generate the proposed cell lines, which is likely given the resources at hand. A major concern is whether the gene profiling experiments will reveal any differences between the differentially derived cell lines. If there are no significant differences, the rest of the application, including the in vivo studies, will not be worth pursuing. Additional preliminary data demonstrating differences in these cell lines depending on the method used to derive them would be extremely helpful and provide much stronger support for proceeding with the animal studies.
STRENGTHS: Strengths identified by reviewers include: the comparison of multiple different stem cell lines in an effort to characterize the potential of different cell lines for cell-based therapeutics; the multiple methods of delivery in a single disease model; the plans to test the contribution of the transplanted NSCs by selective ablation.
Preliminary data show that the proposed methods are within the applicant’s area of expertise; the applicant can grow hESCs, conduct FISH analysis, and can conduct FACS-based cell isolation. The applicant has previously shown benefits of hNSCs in SCI (using the BBB scale and the ladder beam). The applicant also has shown the ability to use DTX to eliminate human cells in vivo and to analyze cells transplanted to the spinal cord.
The proposal includes a thorough analysis of the in vivo data, including anatomical and functional assays, which is considered a strength.
WEAKNESSES: A number of weaknesses of this proposal were identified by the reviewers.
One reviewer finds the overall experimental design difficult to follow and the proposal not well-written. S/he pointed out that comparison of multiple genetic background species can be difficult (including various skid mice) Also, it may be difficult to quantify the actual delivery of cells from direct injection protocols versus peripheral blood delivery (e.g. how will the PI normalize the “therapy dose” by these different approaches?). Overall, this reviewer finds that this is an interesting and certainly valuable approach for comparing and using different stem cell populations, however, adequate controls are not in place to truly make these pharmacokinetic comparisons.
The investigator has some experience with delivery of cells (although the list of papers on engraftment is short). Nevertheless, apparently this research is well-matched to the investigator’s expertise. This reviewer believes the PI is capable of carrying out the proposed studies. However, there is no doubt that this is an ambitious plan for the four year timeline, given the very large effort required to complete good histological, quantifiable, and behavioral analyses.
One reviewer states there are no preliminary data supporting the specific intentions of any aim of this application, just the ability to carry out the appropriate techniques. The work is not hypothesis-driven, and no alternative strategies are provided.
The technical nature of Aim 1 is viewed as a weakness. Because the PI is likely able to generate the proposed cell lines, this aim should be completed and the results included as preliminary data. There is also concern that the proposed experiments in Aim 1 will not demonstrate significant differences in the differentially-derived cell lines. If there are no significant differences, the rationale for the remaining experiments is weakened. No definition of what constitutes a significant difference in gene expression between cell lines is provided within the proposal.
Additional weaknesses include the limited rationale for assuming that in vitro profiling will identify a particular phenotype directly responsible for in vivo effects and for the need for the experiments outlined in Aim 2B. What will be the effects of blood serum proteins that may be present after spinal cord injury? The description of interactions between the transplanted cells and the “niche” of the injured spinal cord, particularly in Aims 3 and 4B were relatively vague. What are the components of the niche and how might they affect NSC differentiation? There also is limited rationale for Aim 4. The hypothesis that multipotent NSCs will have a greater response when transplanted into the injured spinal cord than lineage-restricted cells seems obvious to one reviewer. The applicant’s own work has demonstrated that these multipotent cells can differentiate into a variety of cell types after injury. The issue is how to control them to differentiate into the desired types of cells, something that is not addressed in this application. Alternatively, while not as versatile as multipotent NSCs, restricted cells are preferable in situations where they are required to do only one thing, e.g., OPCs after SCI to re-myelinate demyelinated axons.
The reviewers agree that the generally descriptive, rather than mechanistic, nature of the proposed experiments is a weakness.
Changes to be considered by the PI for future applications should new funding opportunities arise, are to provide preliminary data relevant to the proposal and to make the work more hypothesis-driven. Also, one reviewer suggests that the PI match the research to the requested budget more realistically, as the present budget seems quite excessive for the work proposed.
DISCUSSION: While aspects of this proposal were praised, including the focus on a single disease model and the underlying idea of the proposal, reviewers generally agreed the proposal had substantive weaknesses in the plan, was not sufficiently hypothesis-driven, and insufficient preliminary data was provided to support the plan.