The Golgi matrix is a collection of proteins involved in the structure and function of the
Golgi apparatus.[1][2][3] The matrix was first isolated in 1994 as an amorphous collection of 12 proteins that remained associated together in the presence of
detergent (which removed Golgi membranes) and 150
mMNaCl (which removed weakly associated proteins).[4] Treatment with a
protease enzyme removed the matrix, which confirmed the importance of proteins for the matrix structure.[4] Modern
freeze etch[5]electron microscopy (EM) clearly shows a mesh connecting Golgi
cisternae and associated
vesicles.[6][7] Further support for the existence of a matrix comes from EM images showing that ribosomes are excluded from regions between and near Golgi cisternae.[8][9][10][11][12][13]
The Golgin
GMAP210 has functional regions at both ends.The ALPS of GMAP210 binds to curved, but not flat, lipid layers
Structure and function
GRASP domain alignment of
GRASP55 and the GRASP homologue of Cryptococcus neoformansMicroinjection of antibodies to GRASP65 prevents normal Golgi stack formation.
The first individual protein component of the matrix was identified in 1995 as
Golgin A2 (then called GM130).[14] Since then, many other golgin family proteins have been found to be in the Golgi matrix[2] and are associated with the Golgi membranes in a variety of ways.[15][1] For example, GMAP210 (Golgi
Microtubule Associated Protein 210) has an
ALPS (Amphipathic Lipid-Packing Sensor) motif in the N-termal 38 amino acids and an
ARF1-binding domain called GRAB (Grip-Related Arf-Binding) at the C-terminus.[16] Thus, the GRAB-domain can bind indirectly to Golgi cisternae and its ALPS motif can tether vesicles.[17]
Golgins have
coiled-coildomains and are thus predicted to have elongated structures[2] up to 200 nm in length.[18] Most are
peripheral membrane proteins attached at one end to Golgi membranes.[2] They have flexible regions between the coiled-coil domains, which make them ideal candidates for mediating the dynamic vesicle docking to Golgi cisternae and dynamic structure of the Golgi itself.[2]
Golgi reassembly-stacking proteins are an evolutionarily conserved family of proteins in the Golgi matrix.[2]GRASP65 and
GRASP55 are the 2 human GRASPs. These proteins were named from their requirement for accurate Golgi reassembly during an in vitro assay,[2] but they have also been shown to function in vivo, as shown in the accompanying figure.[19] GRASPs associate with lipid bilayers because they are
myristoylated and their
myristic acid residue intercalates into the lipid layer.[7] Their trans oligomerization is controlled by phosphorylation[6] and is thought to explain the fragmentation of the Golgi as required during mitosis.[7]