SYNOPSIS: Dr. Chow proposes to define integration sites in human embryonic stem cells that have been transduced with a lentiviral, a gammaretroviral or an ASLV vector. The methodology he proposes to utilize is well established and has been used in his own laboratory to study retroviral integration. Integration sites will be mapped in 3 sets of cell samples: 1) embryonic stem cells 48 hours after transduction; 2) embryonic stem cells differentiated into hematopoietic progenitors after 10 days in culture; and 3) T-lymphocytes derived from transduced human embryonic stem cells that develop in immunodeficient mice. In addition to mapping sites, the investigator proposes to evaluate the impact of particular integration positions on both transgene expression and the potential of human embryonic stem cells to undergo hematopoietic differentiation both in vitro and in vivo.
SIGNIFICANCE AND INNOVATION: The novelty of the scientific approach is low since the investigator will use established methodology to obtain data that has been obtained in other cell types. As such, the proposed experiments show no true innovation. Several previous studies have already established the bias of retroviral integration for highly transcribed genes. For example, the studies to map the integration site preferences for lentivirus and gammaretrovirus based vectors have been done in a variety of cell types, including CD34 hematopoietic stem cells (Wagner, W., et al., Retroviral integration sites correlate with expressed genes in hematopoietic stem cells. Stem Cells, 2005. 23(8): p. 1050-8). While the knowledge gained from these studies has been valuable, at this point, the knowledge is also generalizable. There are subtle differences in integration site preferences between lenti and gammaretroviruses (slight bias towards the transcription start site for gamma), but there have not been any observed cell-specific differences. Therefore, while mapping the integration site preferences in ES cells would be a nice study to do, reviewers do not consider it one of high scientific significance in that the prediction would be that the results would be analogous to all other systems studied to date - that is with approximately 2-fold bias towards transcriptionally active regions. Limited new knowledge is likely to arise from these studies; nonetheless, the data collected would provide further information regarding the distribution of integration sites of the various vector types in human embryonic stem cells, and the potential to identify common integration sites in hESCs would be of interest. Not explicitly stated, but potentially informative, is the comparison of distribution sites in these 3 cell populations for it may give insights into factors that result in clonal dominance.
STRENGTHS: This established PI has proposed a comprehensive analysis of the distribution of integration sites in undifferentiated human embryonic stem cells, stem cells differentiated into hematopoietic progenitors and mature T-lymphocytes. The investigator is familiar with the methodology and well-qualified to do the study as he has published his own adaptation to the methods commonly used to map integration sites of retroviruses, and has used this successfully to map HIV-1 integration sites (note that this paper is in press, and the applicants have not provided any data showing the performance of this method over pre-existing methods, therefore, it is a bit difficult to measure the utility of the new high throughput method the investigator has developed to map integration sites).
WEAKNESSES: The applicant fails to address several key aspects of viral integration. First, the PI fails to acknowledge that integration often occurs in a gene rich area so that it is difficult to deduce any potential functional consequences of a particular integration site because multiple genes may be affected by the integration. Second, the investigator does not address the potential for particular integration sites that result in gene activation of cell proliferation and/or survival in culture. Clonal dominance may emerge very quickly even in a culture of only a few days in-vitro and certainly upon differentiation in-vivo. The potential of clonal dominance during in-vitro differentiation and subsequent in-vivo T-lymphocytopoiesis is not discussed, and clonal dominance may cause difficulty in establishing the significance of particular integrations with respect to cell differentiation. Other reviewers felt that in the absence of looking at clonal populations (either derived in-vitro or in-vivo by serial transplantation and looking for dominant clones that emerge), it is not clear how their analyses of insertion sites will allow them to make any correlations with phenotypic alterations.
In terms of their proposed studies to identify correlations of biological function with integration sites, they only plan to do this in the hematopoietic lineages, and these studies have also been done extensively by a number of groups (one of the better examples of this type of study: Kustikova, O., et al., Clonal dominance of hematopoietic stem cells triggered by retroviral gene marking. Science, 2005. 308(5725): p. 1171-4.). It is of note that the authors appear to mis-interpret a reference (#16) to suggest that there are cell-specific differences in integration site preferences - in fact this reference shows differences between the vectors analyzed, but not between the different cell types analyzed with the same vector.
Reviewers also had several concerns regarding the experimental design:
1) The method of integration site mapping as described in section C1.1, is based on the use of a 6-bp cutter to identify LTR-containing fragments. Because the method is still PCR-based, and the frequency of 6-bp cutters is significantly lower than 4-bp cutters (used in analogous technologies), the length of the genomic fragment would be significantly longer - presumably 1-several kilobases, as opposed to 100-several hundred base pairs, as in other commonly used methods (I-PCR, LM-PCR, LAM-PCR). It is unclear whether this modification could result in decreased sensitivity of detection of certain integrants, since it would likely bias detection of some integration sites based solely on proximity of the restriction enzyme site vs. frequency of the integrant (for example).
2) The investigator plans to use a lentivirus vector with a GALV pseudotype - the GALV envelope cannot pseudotype lentiviruses. The investigator does not indicate the need to use a version of the GALV envelope with a C-terminal truncation - without this, he will not be able to produce lentiviral vector pseudotypes (Christodoulopoulos, I. and P.M. Cannon, Sequences in the cytoplasmic tail of the gibbon ape leukemia virus envelope protein that prevent its incorporation into lentivirus vectors. J Virol, 2001. 75(9): p. 4129-38.)
3) How many integration sites/experiment would be analyzed? Some statistical analysis indicating the minimum number of integrants that need to be analyzed should be included to provide meaningful data.
DISCUSSION: There was no discussion further following the reviewers' comments.