Is REST required for ESC pluripotency?

Arising from: S. K. Singh, M. N. Kagalwala, J. Parker-Thornburg, H. Adams & S. Majumder Nature 453, 223–227 (2008)10.1038/nature06863 ; Singh et al. reply The DNA-binding protein REST (also called NRSF) is a transcriptional repressor that targets many neuronal genes1,2 and is abundant in human and mouse pluripotent embryonic stem cells (ESCs)3,4,5. In a recent Letter to Nature, Singh et al.6 suggested that REST controls the self-renewal and pluripotency of ESCs, because they found that ESCs in which a single REST allele was disrupted (Fig. 1a, β-geo-stop insertion) had reduced alkaline phosphatase activity and expressed lower levels of several pluripotency-associated genes6. Here we show that partial or complete loss of functional REST protein does not abrogate ESC potential as reflected by marker gene expression. These data are consistent with earlier reports7,8, and argue that REST is not required for maintaining ESC pluripotency. Figure 1 ESCs deficient in REST retain stem cell characteristics. a, Genetic alterations to the Rest locus analysed here (phosphoglycerate kinase 1-neomycin resistance gene (Pgk-Neo) insertion7) and by Singh et al.6 (β-geo-stop insertions in the gene trap ESC lines RRC and YHC). Rectangles represent exons; coding regions are in black. b, REST, Oct4 and lamin B protein levels in wild-type (+/+), homozygous (-/-) and heterozygous (+/-) Rest ESC lysates, and (marked by an asterisk) in wild-type ESCs transfected with siRNA or shRNA constructs targeting Rest (siRest and shRest) or a control sequence (siCtrl and shCtrl). c, Chromatin immunoprecipitation of Rest+/+ (black), Rest+/- (grey) and Rest-/- (white) ESCs using anti-REST (left panel) and anti-Ezh2 (right panel) versus control antibody (immunoglobulin G, IgG). REST binds SG38457 (also known as Fam70b) and Syt4 (which both contain a RE1 motif) but not Math1 (also known as Atoh1; RE1-negative). No significant binding of REST was detected in Rest-/- ESCs. Ezh2-binding at Math1 (ref. 10) confirmed that chromatin fragments were intact. Error bars represent the standard deviation of three experiments. d, Alkaline phosphatase (AP) activity of mutant ESCs (percentage AP-positive colonies ± standard deviation). e, REST, Oct4 and lamin B protein levels in two additional Rest+/- (D05 and F12) and three wild-type ESC lysates. f, Transcript levels in D05 (dark grey), F12 (light grey) and wild-type (black) ESCs are shown (and, for comparison, retinoic-acid-treated embryoid bodies (white)) relative to wild type. Values were normalized to housekeeping genes and error bars show standard deviation of 4–6 experiments. PowerPoint slide