Sara Terzoli

Apical radial glial cells (aRGs) are highly polarized epithelial cells that drive human cortical development through tightly regulated lineage progression. A distinctive feature of aRGs is their non-canonical Golgi organization, raising the possibility that specialized intracellular trafficking and secretory processes contribute to progenitor behaviour. However, how Golgi architecture contributes to human corticogenesis remains poorly understood. Here, we investigate this question in the context of COG5 deficiency, a rare congenital disorder of glycosylation associated with neurodevelopmental defects, using patient-derived fibroblasts and brain organoids, N-glycomics, glycolipidomics, and single-cell transcriptomics. We show that COG5 deficiency disrupts Golgi organization and impairs Golgi-to-ER retrograde trafficking. At the neurodevelopmental level, COG5-deficient organoids display altered ventricular organization and reduced basal progenitor abundance. Single-cell transcriptomic analysis reveals rewired developmental trajectories characterized by premature differentiation from apical progenitors with activation of neuronal pathways. Despite the absence of major abnormalities in plasma glycoproteins glycosylation, glycolipidomic analysis identified marked defects in glycosphingolipid maturation within patient-derived brain organoids, indicating selective impairment of neural lipid glycosylation. Transcriptomic and biochemical analyses further revealed dysregulation of developmental pathways, including reduced Notch signalling and altered localization of both cleaved Notch1 and the Golgi-associated protein ACBD3. Consistent with impaired progenitor maintenance, COG5-deficient organoids displayed altered cleavage plane orientation in apical progenitors. Together, our findings suggest a link between Golgi organization, intracellular trafficking, membrane glycosylation, and lineage progression in human neural progenitors, identifying Golgi architecture as an important regulator of apical radial glial cell biology and cortical development.

Supplementary Table 3. List of significant genes (FDR-adjusted p-value < 0.05, and |Log2-FoldChange| > 0.25, detected in not less than 10% of cells) identified from the comparison Liver vs Colon within the inflammatory or immunoregulatory TAMs. In green are indicated the genes up-regulated in the liver, while in orange are indicated the colon-associated ones. Table related to Fig. 5A.

Supplementary Figure 5. Differences between inflammatory and immunoregulatory TAMs in colon and liver

Supplementary Figure 2. Transcriptomic characterization of healthy liver and Kupffer cells

Supplementary Figure 4. Pseudotime trajectory analysis shows macrophage blood derivation

Supplementary Table S2: List of genes used to compute the gene module scores related to the antigen-presentation (n genes = 24), lipid-associated (n genes = 32), angiogenic (n genes = 31), and cytotoxic-inflammatory (n genes = 21) macrophage scores. Table related to Fig. 2E, 3E-G, supplementary figure 2F, and supplementary figure 3E.

Supplementary Figure 1. scRNA-sequencing analysis of early phases synchronous CRLM patients

Supplementary Table 4. List of genes used to compute the gene module scores related to the cell cycle proliferation score (n genes = 57). Table related to Fig. 5D.

Supplementary Table 1. List of significant genes (FDR-adjusted p-value < 0.05, and |Log2-FoldChange| > 0.25, detected in not less than 10% of cells) identified from in each tissue-specific cluster within pathologic colon, liver, and peripheral blood. Table related to Fig. 2D, 3D, and supplementary figure 3B.