Intramembrane cleaving proteases (ICliPs) are required for reverse signaling and membrane protein degradation. One class of these proteases is represented by the GxGD-type aspartyl proteases. GxGD describes a signature sequence, which discriminates these proteases from conventional aspartyl-proteases (Fig 1). Members of the class of the GxGD-type aspartyl proteases are the Alzheimer's disease related presenilins (PS), the signal peptide peptidase (SPP) and their homologues (SPPLs), as well as the bacterial type IV prepilin peptidases (TFPP). TNFa and Bri2 have been identified as the first substrates that undergo cleavage within their transmembrane domain by SPPL2b, one member of the SPP/SPPL family. An initial substrate shedding is mandatory to allow subsequent intramembrane cleavage by SPPL2b, but structural elements within the substrates transmembrane and juxtamembrane domains are additionally required to define a substrate.
The research project associated to the DZNE aims to understand the homologies and differences between the PS-family and the SPP/SPPL-family and test potential therapeutic agents against Alzheimer’s disease on cross reactivity with SPP/SPPL-proteases.
We recently identified the first physiological substrates of SPPL3 which are implicated in modification of N- and O-linked glycans as well as in glycosaminoglycan biosynthesis. Verification of substrates in cell culture and in tissues of SPPL3 knockout mice confirmed SPPL3 as one of the major proteases responsible for releasing the luminal domain of various glycosyltransferases and glycosidases into the extracellular space (Fig.2). Since the catalytic center of glycan-modifying enzymes is localized within their ectodomain, shedding of these enzymes reduces their catalytic activity in the Golgi. Consequently, expression levels of SPPL3 affect the glycosylation status of proteins in the secretory pathway. Increased SPPL3-expression results in hypoglycosylation of many, if not all, secretory and membrane proteins, while reduced levels of SPPL3 produce hyperglycosylated proteins (Fig. 2). In a physiological context, this enables cells to rapidly change the glycan pattern of many proteins within the secretory pathway by adapting the expression level of one protease.
In various collaborative projects, we were able to also extend the substrate portfolio of other SPP/SPPL family members by candidate approaches and analysed substrate determinants for certain substrate/enzyme combinations. For instance, the invariant chain of CD74 and the transferrin receptor-1 were identified as SPPL2a and SPPL2b substrates, respectively and Neuregulin 1 Type III turned out to be processed by g-secretase, SPPL2a and SPPL2b.
In the future one major focus will be the identification of SPPL2b substrates. SPPL2b is highly expressed in brain, but its physiological function remains elusive. Furthermore, it is known that glycosylation patterns change during the development of neurodegenerative diseases like Alzheimer’s. Thus, we will now evaluate whether SPPL3 is implicated in these changes that may also influence the generation of Ab-peptides.