Background

The Golgi complex (Golgi) plays a central role in sorting and modifying secretory proteins and lipids (cargos) (Glick and Luini, 2011; Klumperman, 2011; Lu and Hong, 2014). In mammalian cells, the Golgi consists of a laterally linked network of Golgi stacks. As the structural unit of the Golgi, a Golgi stack comprises 4–7 flattened and densely packed membrane sacks called cisternae. The stacked cisternae of the Golgi can be further divided into three zones, namely the cis, medial, and trans-Golgi zones. At the trans-side of the Golgi stack, the trans-Golgi network localizes immediately outside the trans-Golgi zone and comprises tubular and vesicular membrane profiles. The secretory cargos synthesized in the ER reach the cis-Golgi and transit the medial before eventually exiting at the trans-side of the Golgi.

The structural unit of the Golgi, the Golgi stack, has an extremely diverse and complex morphology. It has an amorphous and twisted shape and many attached tubular and vesicular membrane profiles. The spatial resolution of the conventional light microscope is ~250 nm, which is insufficient to resolve the 200–400 nm thick Golgi stack. Therefore, under a fluorescence microscope, the mammalian Golgi appears as a perinuclear lump with almost unresolvable structural features. The difficulty in imaging the Golgi is also due to the dense and random orientation of the Golgi stacks. Hence, it is critical to perform multicolor imaging to identify different zones of the same Golgi stack. Although possessing a spatial resolution of 40–100 nm, super-resolution microscopic techniques such as stimulated emission depletion microscopy and single-molecule localization microscopy have difficulty resolving the Golgi's cisternal organizations, since it is very challenging for them to perform multicolor imaging (Schermelleh et al., 2019). In the past, conventional and super-resolution light microscopy have had only limited success in imaging the Golgi cisternae with primarily qualitative data, mostly revealing the relative distributions of no more than three Golgi proteins (Dejgaard et al., 2007; Bottanelli et al., 2016; Tie et al., 2016; Zhang et al., 2020; Hao et al., 2021).

Nocodazole treatment scatters the densely aggregated Golgi as disk-shaped individual structural units with relatively uniform morphological appearances. Extensive data have demonstrated that nocodazole-induced Golgi ministacks (hereafter Golgi ministacks) reliably represent the native Golgi stack (Rogalski et al., 1984; Van De Moortele et al., 1993; Cole et al., 1996; Trucco et al., 2004; Tie et al., 2018). Therefore, Golgi ministacks are a valid Golgi model. We developed a quantitative method called GLIM (Golgi localization by imaging the center of fluorescence mass) (Tie et al., 2016 and 2017). GLIM can achieve ~30 nm spatial resolution along the cis-trans Golgi axis. However, it only provides a numerical value denoting the axial localization but does not generate a pictorial representation of the cisternal distribution of Golgi proteins. By averaging en face views of Golgi ministacks, we later developed an en face averaging method to study the lateral organization of Golgi proteins (Tie et al., 2018). Recently, we have further advanced the averaging approach and introduced the side-averaging method to visualize the cisternal organization of Golgi ministacks directly (Tie et al., 2022). Averaging the Golgi ministack images can effectively enhance the common structural features and suppress the morphological variations or noises among individual Golgi ministacks. Side-averaging provides a systematic and quantitative imaging approach to reveal the axial and lateral cisternal distributions of Golgi proteins. Here, we present a detailed side-averaging protocol to image the intra-Golgi distribution of four Golgi proteins: giantin, GalT-mCherry, GM130, and GFP-OSBP.