br Experimental procedures br Results

Experimental procedures
Results
Discussion Histone clipping was reported for the first time even before these proteins received their current nomenclature (Johns, 1964; Phillips and Johns, 1959). Surprisingly however, these truncation events are still greatly understudied. Yet, from a practical point of view, techniques such as chromatin-immunoprecipitation run the risk of not detecting the substantial influence that histone clipping might have on the their outcome. All techniques employing specific Epigenetics Compound Library to the H3 N-terminus in general or to modifications thereof in particular, will fail in case of H3 proteolytic cleavage. Also mass spectrometry-based (quantitative) comparisons of e.g. H3K4 and H3K27 methylations devaluate when dealing with samples where clipped histones are present. Histone clipping has been reported in several distinct biological systems (Allis and Wiggins, 1984; Allis et al., 1980; Falk et al., 1990; Gonzalo, 2010; Mahendra and Kanungo, 2000; Mandal et al., 2012, 2013; Tesar and Marquardt, 1990), entailing different biological settings and thus potential roles of this PTM. However, only recently it was admitted into the epigenetic landscape with the discovery of transient histone H3 clipping in differentiating mESC (Duncan et al., 2008) and sporulating yeast (Santos-Rosa et al., 2009). The evolutionarily conserved sequence of histone proteins has been suggested to underlie the surprising occurrence of A21 clipping in both mouse and yeast, and we thus set out to verify whether the H3 clipping that accompanies mESC differentiation also takes place during hESC differentiation. Duncan and colleagues monitored cH3 levels under several differentiation protocols in mESC, namely spontaneous monolayer differentiation after withdrawal of LIF, induced differentiation with RA and embryonic body formation. Depending on the protocol applied, a different cH3 pattern was visualized. Where RA induction leads to an undulating pattern of upcoming and decreasing H3 clipping centered around days 2 and 3 of differentiation, embryonic body formation displayed a faster migrating H3 band which peaked between day 8 and 12 but did not disappear completely after 14days. Here, we report that indeed histone H3 N-terminal clipping occurs in differentiating hESC and that its temporal appearance is equally influenced by the culture conditions. In feeder-free conditions, the two hESC lines tested (UGENT2 and WA01 Oct4-eGFP) show a continuous cH3 pattern after RA induced differentiation, resembling the results of EB formation in mESC. In contrast, when switching to culturing hESC on a feeder layer of MEF, both these cell lines obtain a pattern of upcoming cleavage appearing on day 1, reaching a maximum intensity at day 4 and fading again at the last day of the experiment, similar to the temporal pattern described for RA induced mESC differentiation. Together, this suggests that within both human and mouse, the specific control over histone H3 cleavage during differentiation is profoundly influenced by experimental culture conditions applied. Although the cleavage event in mESC and hESC is appreciably similar, caution should be taken when functionally comparing these events, not in the least because hESC resemble more to mouse epiblast stem cells than regular mouse embryonic stem cells. The latter are considered to represent a more homogeneous and naïve pluripotent state compared to hESC, which are designated to be in a heterogeneous and primed pluripotent condition and thus probably reflect a more developed state (Gafni et al., 2013; Nichols and Smith, 2009; O\'Leary et al., 2012). In contrast to Duncan et al., we could not find any clear correlation between the progression of differentiation and histone H3 clipping. The pluripotency status was monitored in this report by following Oct4 levels, i.e. specifically with an Oct4-eGFP reporter cell line. When monitoring three different differentiation methodologies, H3 cleavage seemed to be unaffected, while the Oct4 status was found to be influenced in the predicted way: a decrease in expression following (induced) differentiation. Considering the open and highly dynamic state of pluripotent hESC chromatin, this might not be such a surprise (Tollervey and Lunyak, 2012). In line with this, on a Sypro Ruby stained SDS PAGE gel we also observed some reduction in H4 band intensity at some time points (data not shown). Whether this indeed is a clipping event remains to be determined. Of note, when Duncan et al. inhibited cathepsin L, H3 clipping was abrogated while Oct4 expression still decreased, adding yet another argument against a direct link between Oct4 expression levels and histone clipping.