Multicellular organisms are capable of generating a wide range of different cell types from the same genome through differentiation, during which distinct gene expression patterns are gradually specified along a branchwork of lineages. The highly orchestrated and robust developmental process requires tight regulation of cell potency, which limits developmental competence towards given lineages to specific stem cell populations. It remains elusive whether cell potency is actively acquired in lineage intermediates, by the activation of pioneer transcription factors, or progressively and irreversibly restricted during development. Importantly, the molecular basis of cell potency is unclear. We speculate that gene potency, which refers to the ability of a gene to be activated in response to proper transcriptional activators or developmental cues, is the foundation of cell potency. In order to measure gene potency, we took advantage of cell fusion assay, wherein genome of specific cell type was exposed to regulators from a disparate lineage. The potency of a silent gene could be evaluated by whether it could be activated post cell fusion. Transcriptome profiling of hybrid cells led to discovery of two group of silent genes: occluded genes and activatable genes. Despite both being silent before cell fusion, activatable and occluded genes showed distinct transcriptional potential in response to transcriptional activators from fusion partners. While activatable genes are readily activated after cell fusion, occluded genes remain silent even in the fusion cells where the same genes of the fusion partner are actively expressed. Fusion of the same mouse SPRET fibroblast cell line with a serious of mouse cell lines of various differentiation potential indicates that restriction of cell potency is accompanied by decreasing of activatable genes and increasing of occluded genes. Importantly, while naïve pluripotent embryonic stem cells (ESCs) are able to reprogram occluded genes into a potent activatable or active status, de-occlusion ability is disabled in primed pluripotent stem cells (EpiSCs), which are derived from epiblast corresponding to developmental stage prior to lineage differentiation. The shutoff of de-occlusion capacity before onset of differentiation renders gene occlusion irreversible in every downstream lineage, suggesting that developmental competence is progressively restricted along developmental process. Consistently, neural differentiation of fibroblast x EpiSC fusion cells suggests that occluded fibroblast genes, but not potent EpiSC genes, cannot be activated during the lineage differentiation process, arguing against the theory of active acquirement of developmental potential in lineage intermediates. Collectively, these finding indicates the fundamental role of irreversible gene occlusion in progressive restriction of cell potency in developmental process. Mechanismly, we found that chromatinization with recombinant histones without post-translational modifications commits a minigene to occlusion after electroporated into somatic cells, indicating that gene occlusion is the default choice of unprotected chromatin. In contrast, activatable status requires protection by binding of placeholders to the target genes. Deletion of Sox2 or Olig2 impaired target gene potency, switching activatable or expressed genes into occluded status. Based on the observations, we propose a placeholder model in regulation of gene potency during lineage differentiation process. Before differentiation, stem cells possess placeholders that bind to and prevent fully chromatinization of regulatory regions of genes to be activated in downstream lineages, assuring their activatable status. In response to differentiation signal, stem cell placeholders disappear and different groups of activatable genes are activated in different downstream lineages. Genes of alternative lineages loss protection from placeholders and are spontaneously occluded by chromatin so that they will never be activated even if their transcription activators reappear in the future.