Regulierte Intramembranproteolyse des Bri2-Proteins

Regulated intramembrane proteolysis (RIP) is a conserved mechanism for the proteolytic processing of transmembrane proteins, from bacteria to human. In general, RIP is described as a two-step cleavage process. The first proteolytic processing, called “Ectodomain-Shedding”, shortens the ectodomain of the transmembrane protein. The second proteolytic event takes place within the transmembrane domain of the substrate and is accomplished by an intramembrane protease. There are different classes of intramembrane proteases known. Whereas the Site-2 protease belongs to the class of metallo intramembrane proteases, serine intramembrane proteases are represented by rhomboid. According to a conserved GxGD motif in their active site, aspartyl intramembrane proteases are named GxGD intramembrane proteases.

The degradation products of RIP may mediate important signaling functions, e.g. they play an important role in the expression of certain cytokines and cellular division or differentiation processes. Pathologically, RIP plays a major role in Alzheimer’s disease: Aβ-peptides that are critically associated with Alzheimer’s disease and accumulate within the patients’ brain, are products of the processing of the amyloid precursor protein (APP) by γ-secretase, an intramembrane aspartyl protease of the GxGD-family.

γ-Secretase is a target of intense pharmaceutical research, since inhibition of the protease, which would cause reduced production of Aβ-peptides, is a promising approach for Alzheimer’s disease prevention and treatment. Besides γ-secretase, signalpeptide peptidase (SPP) and the signalpeptide peptidase-like proteases (SPPL; SPPL2a,b,c and SPPL3) belong to the family of GxGD-proteases. SPP/SPPL-proteases play an important role in cellular signalling routes. One example is the expression of the cytokine interleukin-12, which is controlled by SPPL2a and/or SPPL2b by processing tumor necrosis factor α (TNFα).

Some inhibitors of γ-secretase that target the active site of the protease cross-react with SPP/SPPL-proteases. The reason for cross-reactivity is most likely a high similarity of the active sites between the GxGD-proteases. The application of unspecific γ-secretase inhibitors for Alzheimer’s disease treatment could cause severe side effects due to their interference of SPP/SPPL-dependent signaling pathways. Therefore, a potential cross reactivity with SPP/SPPL-proteases must be evaluated in the development of novel specific γ-secretase inhibitors.

For this reason, a comparison of the GxGD-proteases in respect to their catalytic properties and their employed mechanism of substrate selection is necessary. While many substrates are known for γ-secretase and several substrates have been described for SPP, only a limited number of SPPL-substrates are published, impeding a detailed comparison. This study describes the identification of the Bri2 protein as a novel substrate for SPPL-proteases.

The processing of Bri2 is an example of RIP. First, Bri2 is processed in its ectodomain by the serine protease Furin and the metallo protease ADAM-10, resulting in the secretion of parts of the Bri2 ectodomain. This proteolytic processing takes place in late cellular secretion compartments. In a further proteolytic event, the membrane retained fragment of Bri2 is cleaved by SPPL2a or by SPPL2b, whereas SPP and SPPL3 are not able to process Bri2. Intramembrane proteolysis of Bri2 by SPPL2a and SPPL2b results in a soluble intracellular domain and a low molecular C-terminal peptide that is secreted into the extracellular space.

While Bri2 is a substrate for SPPL2a and SPPL2b, its homolog Bri3 is not. This finding shows that SPPL2b binds and processes transmembrane proteins very selectively. A detailed comparison of the SPPL2b-dependent processing of Bri2 with chimeric proteins that consist of combinations of Bri2 and Bri3 helped to address several questions in regard to the requirements of SPPL2b-dependent intramembrane proteolysis e. g. which influence intra- or extra-cellular domains of the substrate have on SPPL2b cleavage efficiency. It is shown that the luminal juxtamembrane domain, as well as a short stretch of cytoplasmic amino acids play a decisive role for SPPL2b cleavage efficiency. The data also illustrate that the amino acid sequence within the transmembrane domain of a SPPL2b-substrate affects the proteolytic activity. Strikingly, the results show that SPPL2b selects potential substrates in regard to the size of their ectodomain and that SPPL2b cleavage efficiency – as it is known for γ-secretase – correlates negatively with an increasing size of the substrate ectodomain. In agreement with this finding, ectodomain-shedding of Bri2 by ADAM-10 facilitates intramembrane proteolysis by SPPL2b.

The novel data for the RIP of Bri2 allows a detailed comparison of SPPL2b and γ- secretase. Both proteases are located in late cellular secretion compartments and carry a highly conserved active site. Ectodomain-shedding increases the cleavage efficiency of both SPPL2b and γ-secretase, whereby even the preferred size of the ectodomain of a substrate is highly similar. The substrate selection is very likely to be specific for each protease. The main difference, however, between SPPL2b and γ-secretase seems to be the formation of the γ-secretase complex compared to SPPL2b, that does not need protein co-factors for activity. Therefore, in order to avoid cross-reactivity with SPP/SPPL-proteases, γ-secretase complex formation may serve as the better target for the development of specific γ-secretase inhibitors rather than the active site.