The thalamus (dorsal thalamus) is a diencephalic structure flanked dorsally by the pretectum and ventrally by the prethalamus (Nakagawa, 2019). It develops from progenitor pools into a complex nuclear array and establishes extensive connections with cortical, subcortical, and cerebellar regions (Jones, 1991; Guillery and Sherman, 2002). Accumulating evidence indicates that the thalamus serves not merely as a passive relay station but also plays pivotal roles in sensory processing and functions as a hub for cognitive-behavioral processes, including conscious awareness, memory, and executive control (Štillová et al., 2015; Wolff and Vann, 2019; Crosson, 2021).

During diencephalic development, ventricular zone (VZ) progenitors are organized into three domains along the anteroposterior (AP) axis: p1 (pretectum), p2 (epithalamus and thalamus), and p3 (prethalamus) (Chatterjee and Li, 2012). Within the p2 domain, thalamic VZ progenitors partition along the AP axis into rostral (rTh) and caudal (cTh) thalamic domains. This early patterning establishes the thalamic nuclear blueprint, with core architecture largely determined by birth (Nakagawa, 2019). Key protein-coding genes, including Sonic hedgehog (Shh), transcription factor 7 like 2 (Tcf7l2), and gastrulation brain homeobox 2 (Gbx2), have been identified as crucial regulators of thalamic fate determination, regionalization, and nuclear organization (Mallika et al., 2015; Lipiec and Bem, 2020; Govek et al., 2022). Many studies demonstrate that precisely timed expression of the transcription factor GBX2 is essential for maintaining thalamic neuronal identity and orchestrating nuclear organization in the early embryo (Chen et al., 2009; Mallika et al., 2015). Despite these advances, molecular mechanisms governing mammalian thalamic organization remain incompletely characterized.

Long noncoding RNAs (lncRNAs), defined as >200 nucleotides (nt) with minimal protein-coding potential, exhibit pronounced enrichment in the brain (Sarropoulos et al., 2019). Two decades ago, Bejerano et al. reported approximately 500 noncoding ultra-conserved regions (UCRs) exhibiting perfect sequence conservation across mammalian genomes. These UCRs predominantly overlap RNA-processing genes or localize proximal to genes regulating transcription and development (Bejerano et al., 2004). Sandelin et al. further demonstrated that the UCRs frequently flank genes encoding developmental transcription factors (Sandelin et al., 2004). Notably, some lncRNAs with high evolutionary conservation exhibit disproportionate enrichment in the mammalian brain compared to peripheral organs (Aprea and Calegari, 2015; Sarropoulos et al., 2019). It is established that the mammalian brain is characterized by complicated and fine-grained spatiotemporal regulation. Disruptions in embryonic processes may cause neurodevelopmental disorders (NDDs), such as intellectual disability (ID) and autism spectrum disorder (ASD) (Wegiel et al., 2010; Homan et al., 2014; Li et al., 2019). Meanwhile, population studies reveal aberrant expression and/or mutations in multiple lncRNAs among patients with NDDs (Zarrei et al., 2019). Nevertheless, in vivo functional validation of lncRNAs, especially their roles in thalamic development, remains critically underexplored.

In this study, we indicated that colorectal neoplasia differentially expressed (CRNDE), a conserved noncoding RNA enriched during early central nervous system (CNS) development, is stage-specifically highly expressed in mouse embryonic thalamic progenitors of the ventricular zone. Using a Crnde knockout mouse model, we revealed that Crnde deletion disrupted thalamic neural progenitor differentiation during embryogenesis and caused reduced thalamic size at birth. Our research uncovers the vital role of long noncoding RNA in regulating thalamic formation and offers direct insights into the potential etiology of neurodevelopmental diseases.