In vivo characterization of the bacterial intramembrane-cleaving protease RseP using the heme binding tag-based assay iCliPSpy

Kupke, Thomas; Götz, Rabea M.; Richter, Florian M.; Beck, Rainer; Lolicato, Fabio; Nickel, Walter; Hopf, Carsten; Brügger, Britta

doi:10.1038/s42003-023-04654-z

Download PDF

Article
Open access
Published: 18 March 2023

In vivo characterization of the bacterial intramembrane-cleaving protease RseP using the heme binding tag-based assay iCliPSpy

Communications Biology volume 6, Article number: 287 (2023) Cite this article

1797 Accesses
6 Altmetric
Metrics details

Subjects

Abstract

Regulated intramembrane proteolysis (RIP) describes the protease-dependent cleavage of transmembrane proteins within the hydrophobic core of cellular membranes. Intramembrane-cleaving proteases (I-CliPs) that catalyze these reactions are found in all kingdoms of life and are involved in a wide range of cellular processes, including signaling and protein homeostasis. I-CLiPs are multispanning membrane proteins and represent challenging targets in structural and enzyme biology. Here we introduce iCLiPSpy, a simple assay to study I-CLiPs in vivo. To allow easy detection of enzyme activity, we developed a heme-binding reporter based on TNFα that changes color after I-CLiP-mediated proteolysis. Co-expression of the protease and reporter in Escherichia coli (E. coli) results in white or green colonies, depending on the activity of the protease. As a proof of concept, we use this assay to study the bacterial intramembrane-cleaving zinc metalloprotease RseP in vivo. iCLiPSpy expands the methodological repertoire for identifying residues important for substrate binding or activity of I-CLiPs and can in principle be adapted to a screening assay for the identification of inhibitors or activators of I-CLiPs, which is of great interest for proteases being explored as biomedical targets.

Mechanism-based traps enable protease and hydrolase substrate discovery

Article Open access 16 February 2022

Ten catalytic snapshots of rhomboid intramembrane proteolysis from gate opening to peptide release

Article 30 September 2019

A closed translocation channel in the substrate-free AAA+ ClpXP protease diminishes rogue degradation

Article Open access 10 November 2023

Introduction

I-CLiPs cleave transmembrane proteins within the lipid bilayer (reviewed in¹). This type of proteolytic activity is conserved throughout all domains of life, and contributes to a wide range of cellular processes, including protein maturation and degradation, and cellular signaling. I-CLiPs can be divided into different groups depending on the type of catalysis. Although these proteases generally have low substrate specificity, some exclusively cleave type I transmembrane proteins, such as mammalian gamma-secretase², which is associated with Alzheimer’s disease, while other proteases target type II transmembrane proteins. Examples of the latter are members of the SPPL2 (signal peptide peptidase-like 2) family, whose best-known substrate is tumor necrosis factor alpha (TNFα), a key regulator of the inflammatory response³. I-CLiPs are involved in highly coordinated RIP signaling events. In general, signal transduction by RIP begins with cleavage of the extracellular domain of transmembrane proteins in a process known as ectodomain shedding, and then the remaining membrane-bound fragment is processed by I-CLiPs^4,5,6. A well-characterized I-CLiP in E. coli is the inner membrane zinc metalloprotease RseP^7,8,9 (formerly called YaeL or EcfE). RseP activates the cell envelope stress response in E. coli by cleaving the anti-σ^E type II transmembrane protein RseA. This signaling cascade is initiated by the membrane-bound serine protease DegS^10,11,12, which acts as a sheddase and removes the periplasmic domain of RseA. The remaining N-terminal fragment 1-148 of RseA is then cleaved by RseP between residues 108 and 109¹³.

RseP is a transmembrane protein with four transmembrane helices and two tandemly arranged periplasmic PDZ domains that serve as size exclusion filters for the substrate^{8,9,14,15,16,17}. The Zn²⁺ ion required for activity is bound by the side chains of His22, His26 and Asp402^7,8,18. An important structural element involved in substrate recognition and cleavage is the membrane-embedded β-hairpin-like structure formed by residues Ile61 to Glu75 [membrane-reentrant β-loop, β-MRE]¹⁹. In addition, the GFG motif of RseP and a membrane-bound amphiphilic helix (residues Pro323 to Thr350) contribute to substrate discrimination and binding^20,21. The recently deposited crystal structure of RseP with the inhibitor batimastat provides mechanistic insights into intramembrane proteolysis and suggests a gating mechanism to regulate substrate entry²².

RseP also cleaves signal peptides generated by leader peptidase-mediated processing of presecretory proteins^9,23,24. This low substrate specificity prompted us to investigate whether reporter substrates could be used to study RseP in vitro and in vivo. Ideally, such a reporter would change its spectroscopic characteristics upon proteolytic processing and also contain a tag that would allow affinity purification and detection.

In a recent study, we used an E. coli expression model to identify TNFα, its proteolytical processing products and other substrates of SPPL2a/b proteases as heme (= heme B) binding proteins²⁵. In vivo, TNFα is cleaved by the sheddase TACE, releasing an N-terminal fragment (NTF) and a soluble cytokine²⁶. The NTF is further processed by the intramembrane-cleaving proteases SPPL2a/b, resulting in two intracellular domains (ICDs) that differ in length [TNFα-(1-39) and TNFα-(1-34)²⁷]. We have described that heme binding depends on both a cytosolic cysteine residue proximal to the membrane (Cys30) and oligomerization of the transmembrane domain (TMD). Proteolytic processing of the N-terminal fragment of TNFα alters the nature of ferric heme binding, switching from stable, hexa-coordinated ligation by two Cys30 side chains in dimeric TNFα-(1-39) (so-called bis-thiolate ligation) to low affinity, penta-coordinated heme bound by only one Cys30 side chain in the shorter, monomeric ICD TNFα-(1-34)²⁵ (Fig. 1a). Deletion of five of the nine TMD residues still present in TNFα-(1-39) (residues Leu35 to Leu39) removes the dimerization motif S³⁴XXS³⁷ and thus prevents heme promoted dimerization of TNFα-(1-34) (Fig. 1a). Bis-thiolate ligated ferric heme shows a split Soret band with absorbance maxima at about 370 nm and 450 nm and an additional absorbance maximum at about 550 nm, so that TNFα-(1-39) with bound heme is a green colored peptide. In contrast, TNFα-(1-34) has only one Soret band at about 370 nm and appears uncolored due to its low heme binding affinity²⁵ (Fig. 1a).

**Fig. 1: The principle of the heme binding tag-based iCliPSpy assay.**

In the present study, we exploit the different spectroscopic properties of substrate TNFα-(1-39) and product TNFα-(1-34) to investigate intramembrane processing catalyzed by RseP in vivo. We show that E. coli cells co-producing an MBP-TNFα-(1-39) fusion protein (MBP, maltose binding protein) optimized for iCliPSpy and an active RseP protease are uncolored/white, while cells expressing the inactive RseP enzyme together with the reporter are green. To confirm the processing of TNFα-(1-39) catalyzed by RseP, we used mass spectrometry. By determining the intact mass and additionally the sequence of tryptic fragments of the processed substrate proteins, we show that RseP trims tripeptides and cleaves L/F-S peptide bonds within the substrate used. To test whether this assay for I-CliPs allows the identification of residues within RseP that are important for activity and/or substrate binding, we generated mutant rseP genes by error-prone PCR²⁸. Using this approach, we identified Tyr69 as a critical residue for RseP activity and/or substrate binding. Molecular dynamics (MD) simulations of the AlphaFold^29,30 structure of RseP (with bound Zn²⁺ ion) in a lipid bilayer with a lipid composition mimicking E. coli membranes show that residue Tyr69 is located in the membrane/cytosol environment near the active site and may play a role in positioning the substrate peptide through hydrophobic interactions. These results are in agreement with the recently published crystal structure of RseP and the model proposed for substrate binding, in which the so-called edge strand of RseP (residues Gly67 to Val70 including Tyr69 and part of the β-MRE) forms an antiparallel β-sheet with the substrate TMD²². Our data show that iCliPSpy is a simple and sensitive assay to identify and study I-CLiP activities in vivo.

Results

Substrate and product of the reaction catalyzed by RseP

An effective production strategy for the heme-binding peptide TNFα-(1-39) as potential substrate of RseP and other I-CLiPs, is its expression as MBP fusion protein in E. coli²⁵. To ensure the correct topology [considering that full-length TNFα is a type II transmembrane protein: N-terminus in (cytoplasm) and C-terminus out (periplasm/lumen of the ER)], MBP was fused to the N-terminus of TNFα-(1-39) without the signal peptide (residues -24 to -1, Fig. 1b). TNFα-(1-39) contains only the N-terminal nine amino acid residues L³¹FLSLFSFL³⁹ of the TNFα TMD and is therefore found in both the membrane and cytosolic fractions^25,27. Instead of the MBP construct described in our recent study²⁵, we used a mutant MBP protein (MBP_mut) containing additional substitutions, including I317V, which further increases the affinity for amylose³¹ (Fig. 1b). To increase the amount of heme bound by TNFα, we used the variant peptide TNFα-(1-39)-L31P, which shows a stronger interaction with heme compared to the wildtype peptide²⁵. In addition, MBP_mut and TNFα-(1-39)-L31P are linked in MBP_mut-TNFα-(1-39)-L31P by a shortened linker peptide (Fig. 1b), which should increase the proteolytic stability of the fusion protein. We expressed the putative product of RseP-mediated intramembrane proteolysis, TNFα-(1-34), in the same manner [MBP_mut-TNFα-(1-34)-L31P]. The MBP_mut-TNFα-(1-39)-L31P, MBP_mut-TNFα-(1-34)-L31P and His-MBP-TNFα-(1-39)-L31P genes (Fig. 1b) were cloned into the pETDuet-1 vector (Supplementary Table 1 and Supplementary Note 1) and the recombinant plasmids were transformed into E. coli T7 Express cells. For in vivo experiments, expression of MBP fusions was performed without IPTG induction on agar plates (Fig. 2a). A general disadvantage of the T7 expression system is the already existing low protein expression without IPTG induction, also called leaky expression. However, leaky expression can be an advantage, especially in the expression of membrane proteins, as the low-level expression prevents the formation of inclusion bodies^32,33 and also favors heme-binding of TNFα-(1-39)-L31P, as the heme concentration in cells is extremely low³⁴.

**Fig. 2: Characterization of substrate and product of the iCliPSpy assay.**

After several days at room temperature cells expressing MBP_mut-TNFα-(1-39)-L31P were ocher green, while cells expressing His-MBP-TNFα-(1-39)-L31P were only faintly colored and cells expressing MBP_mut-TNFα-(1-34)-L31P or the empty vector control pETDuet-1 were white (Fig. 2a, b). SDS-PAGE analysis of whole cells scraped from agar plates confirmed the production of the different MBP-TNFα fusion proteins and showed that reduced expression of His-MBP-TNFα-(1-39)-L31P is responsible for the less intense color of these cells (Fig. 2c). Therefore, we selected the MBP_mut-TNFα-(1-39)-L31P, and MBP_mut-TNFα-(1-34)-L31P constructs for further experiments. The membrane-associated MBP fusion proteins were then solubilized with the non-ionic detergent n-dodecyl-β-D-maltopyranoside (DDM) and purified by amylose affinity chromatography in the presence of DDM. UV/Vis spectroscopy of the purified fusion proteins confirmed the ligation of heme by bis-thiolate in MBP_mut-TNFα-(1-39)-L31P and by thiolate in MBP_mut-TNFα-(1-34)-L31P (Fig. 2d-f). MBP_mut-TNFα-(1-34)-L31P with thiolate ligated heme shows an absorbance maximum at about 392 nm, as the addition of DDM results in a bathochromic shift of the absorbance maximum by ∼20 nm (Fig. 2f). Cells expressing MBP_mut-TNFα-(1-34)-L31P are white because the heme content is very low. Heme reconstitution of MBP_mut-TNFα-(1-34)-L31P, however, resulted in a red-brown color of the fusion protein (Fig. 2d, f). Taken together, our results show that recombinantly expressed reporter substrate and mimicked product of I-CliP-mediated proteolysis in E. coli differ in their spectroscopic properties, resulting in visible changes in the coloring of bacterial colonies. Next, we tested whether our reporter MBP_mut-TNFα-(1-39)-L31P also serves as a substrate for bacterial I-CLiPs.

Visualization of the in vivo activity of RseP

A well-characterized bacterial I-CLiP is E. coli RseP, which we used to develop the I-CLiP assay iCliPSpy. A previously published in vivo assay for monitoring RseP activity^9,23 uses E. coli rseP rseA double-disrupted cells carrying plasmids expressing RseP-Myc⁸ and an HA-MBP-tagged, truncated version of RseA, RseA140, as a substrate for RseP. RseA140 cleavage by RseP was monitored by immunoblot analysis^9,23 and this assay has been successfully used to study the structure, function and substrate binding of RseP by site-directed mutagenesis^19,20,21,35. Since the in vivo activity of RseP is not directly visible when looking at colonies of E. coli cells over-expressing RseP-Myc and HA-MBP-RseA140, we sought to develop an assay that allows simple detection of I-CLiP activity based on the color of the colonies.

To test whether MBP_mut-TNFα-(1-39)-L31P can be cleaved by RseP, RseP-Myc⁸ wild type (wt, Supplementary Fig. 1a) or the catalytically inactive variant RseP-Myc H22F were expressed together with our I-CliP reporter. Since RseP is essential for bacterial growth⁷, we performed iCliPSpy in presence of the chromosomal copy of rseP. In principle, the RseP-dependence of growth can be circumvented by down-regulating the porins OmpA and OmpC³⁶. Our data, however, show that the amount of chromosomally encoded RseP is not sufficient to degrade overproduced MBP_mut-TNFα-(1-39)-L31P to any appreciable extent (Fig. 3a), so that additional genetic modification was not required.

**Fig. 3: Visualization of in vivo intramembrane proteolysis catalyzed by RseP-Myc.**

Co-expression of the I-CliP reporter with RseP-Myc wt resulted in white cells (Fig. 3a), indicating that the overexpressed RseP had cleaved TNFα-(1-39)-L31P. The H22F substitution in RseP, which leads to a loss of zinc binding, inhibited RseP activity, so that co-expression of MBP_mut-TNFα-(1-39)-L31P with rseP-Myc H22F resulted in green E. coli colonies (Fig. 3a). SDS-PAGE analysis of whole cells scraped from the agar plate confirmed comparable expression levels of the MBP_mut-TNFα fusions and showed that the RseP cleavage product of MBP_mut-TNFα-(1-39)-L31P has a similar apparent molecular weight as MBP_mut-TNFα-(1-34)-L31P, indicating that only a few amino acid residues were cleaved off by RseP (Fig. 3b). Expression of rseP-Myc wt and rseP-Myc H22F genes was confirmed by western blotting (Fig. 3c). Higher expression levels of rseP-Myc H22F compared to rseP-Myc wt were less pronounced in case of co-expression with MBP_mut-TNFα-(1-39)-L31P (compare the first two lanes with lanes five and six of the immunoblot shown in Fig. 3c). As expected and despite this higher expression, proteolytic cleavage of MBP_mut-TNFα-(1-39)-L31P by RseP-Myc H22F was not observed. In vivo processing of MBP_mut-TNFα-(1-39)-L31P was also confirmed for untagged RseP (Supplementary Figs. 1a-c and 2a, b). However, when a 10x HisTag (His₁₀) was fused directly without a linker peptide to the carboxy terminus of RseP, the activity of RseP was decreased (Supplementary Fig. 1a-c), suggesting that the interaction of the periplasmic His₁₀ Tag with the PDZ domains (Supplementary Fig. 1b) could affect folding, stability or activity of RseP.

In general, our results suggest that green-colored colonies appear when sufficient MBP_mut-TNFα-(1-39)-L31P is present. White colonies, however, are observed when MBP_mut-TNFα-(1-39)-L31P has been cleaved or when the reporter is expressed at very low concentrations. Therefore, an expression rate of the I-CLiP reporter that is independent of the sequence of the co-expressed rseP gene (Fig. 3b and Supplementary Fig. 2b) is an essential requirement for this assay.

Processing changes heme binding properties of the substrate

Our data suggest that intramembrane proteolysis catalyzed by RseP affects the dimerization motif S³⁴LFS³⁷ of TNFα-(1-39) and consequently alters the heme binding mode of the TNFα peptide, consistent with the color change. To test whether MBP_mut-TNFα-(1-39)-L31P indeed serves as a substrate for RseP cleavage, membrane-associated proteins solubilized with DDM from IPTG-induced E. coli cells co-expressing MBP_mut-TNFα-(1-39)-L31P with either rseP wt or rseP H22F were enriched via amylose affinity chromatography (Fig. 4a), analyzed by SDS-PAGE (Fig. 4b) and the heme coordination mode was determined by UV/Vis spectroscopy (Fig. 4c). The ratio of active and correctly folded protease to substrate may differ after IPTG-induced overexpression compared to leaky expression in the in vivo assay. This would lead to different proportions of fully processed substrate. Indeed, we observed that both amylose affinity columns were green, but the color was more intense in the case of co-expression with rseP H22F (Fig. 4a). MBP_mut-TNFα-L31P proteins purified from rseP wt over-expressing cells showed a slightly lower apparent molecular weight than MBP_mut-TNFα-L31P proteins purified from rseP H22F over-expressing cells (Fig. 4b). The UV/Vis spectrum of MBP_mut-TNFα-L31P purified from rseP H22F over-expressing cells showed absorbance maxima at 371 nm and 451 nm, demonstrating bis-thiolate ligation and confirming that RseP H22F does not cleave MBP_mut-TNFα-(1-39)-L31P. For MBP_mut-TNFα-L31P purified from rseP wt over-expressing cells, however, the absorbance at 451 nm was significantly reduced, the absorbance maximum at 371 nm was shifted to 373 nm and the absorbance at 392 nm was increased [Fig. 4c, see also Fig. 2f, which shows a UV/Vis absorbance spectrum of the putative RseP cleavage product MBP_mut-TNFα-(1-34)-L31P, which would be expected if proteolysis was complete and the dimerization motif disrupted]. As expected, unprocessed and RseP-processed MBP_mut-TNFα-(1-39)-L31P fusion proteins were also present in the cytosolic fraction (Supplementary Fig. 3a-e; compare also Kupke et al.²⁵).

**Fig. 4: Characterization of the RseP cleavage product in vitro.**

Solubilized membrane-associated MBP_mut-TNFα-L31P proteins purified from E. coli co-overexpressing either inactive rseP H22F or active rseP wt were then separated by gel filtration in presence of DDM into heme-binding (monomeric or dimeric) MBP_mut-TNFα-L31P and heme-free monomeric MBP_mut-TNFα-L31P fusion proteins (Fig. 4d). Heme-free monomers are always present as heme is not stoichiometrically bound by TNFα-ICDs²⁵. The higher content of heme-free monomers of MBP_mut-TNFα-L31P in the case of RseP wt compared to RseP H22F is probably related to a lower heme binding affinity of the processed fusion protein (compare UV/Vis spectra in Fig. 4c), because hydrophobic residues within the heme-binding motif SRRCPFLSLFSFL of TNFα may also contribute to heme binding and because bis-thiolate ligation stabilizes heme binding²⁵. The heme-binding MBP_mut-TNFα-L31P fusion protein obtained in the case of RseP H22F eluted at 15.40 ml, whereas the protein observed for RseP wt eluted at 16.05 ml (corresponding to a lower apparent molecular weight) from the size exclusion column (Fig. 4d). Since MBP_mut-TNFα-(1-39)-L31P contains only a single cysteine residue [Cys30 of TNFα-(1-39)-L31P,] bis-thiolate ligation of heme requires dimerization²⁵ (Fig. 1a). Therefore, the eluted proteins were again characterized by UV/Vis spectroscopy to distinguish between monomeric and dimeric heme-binding proteins (Fig. 4e). The UV/Vis spectra of the heme-binding proteins purified by gel filtration confirm that dimeric MBP_mut-TNFα-(1-39)-L31P is present when MBP_mut-TNFα-(1-39)-L31P was co-expressed with rseP H22F but not with rseP wt. In the latter case, the absorbance maximum at about 392 nm (measured in the presence of DDM) shows that processed MBP_mut-TNFα-L31P coordinates heme via a single cysteine residue and is monomeric. It appears that RseP-processed (monomeric) heme binding MBP_mut-TNFα-L31P is not uniform, as the observed elution volume is as mentioned either 16.05 ml (when measuring the absorbance at 280 nm) or 15.85 ml (absorbance at 371 nm; Fig. 4d). We assume that at least two different processing products with different heme binding affinities (and thus different heme content) are present. At higher protein concentrations, a processing product that still contains part of the dimerization motif may dimerize to a small extent and therefore also shows an absorbance maximum at about 450 nm (and 373 nm; compare spectrum “+ RseP wt” of Fig. 4e with the corresponding spectrum of Fig. 4c).

In summary, gel filtration shows that cleavage by RseP results in MBP_mut-TNFα fusion proteins that differ in molecular weight and heme binding properties from the substrate MBP_mut-TNFα-(1-39)-L31P.

RseP processing sites in the MBP_mut-TNFα substrate

The UV/Vis absorbance spectrum of the MBP_mut-TNFα-(1-39)-L31P fusion protein purified from cells co-overexpressing rseP wt shows that the heme-coordinating residue Cys30 is still present in the proteolytic cleavage product, but that the bis-thiolate ligation of heme has been largely lost. Consequently, cleavage occurred within TNFα TMD residues 31-39 of the reporter protein, disrupting the dimerization motif (compare Fig. 1). To determine the exact RseP processing site(s) within the C-terminal TNFα TMD residues P³¹FLSLFSFL³⁹ by mass spectrometry, MBP_mut-TNFα-(1-39)-L31P was again co-expressed with either rseP wt, empty vector control pCOLADuet-1, rseP-Myc wt or rseP-Myc H22F. The MBP-TNFα fusion proteins produced were purified from the membrane fraction by amylose affinity chromatography and then analyzed by SDS-PAGE (Fig. 5a, b) and UV/Vis spectroscopy (Fig. 5c) to confirm the experimental data shown in Fig. 4 and to ensure RseP catalyzed processing within TNFα-(1-39)-L31P prior to mass spectrometric analysis. As previously observed, the MBP_mut-TNFα fusion proteins that still bind heme via bis-thiolate ligation have a slightly higher molecular weight than the fusion proteins that bind heme via mono-thiolate ligation (Fig. 5a, b), but the resolution of SDS-PAGE is too low to separate differently truncated proteins. Two complementary mass spectrometric analyses were then performed to determine the proteolytic cleavage sites: (i) analysis of the intact mass of the MBP-TNFα fusion proteins (Fig. 5d, e and Supplementary Fig. 4a-d). and (ii) sequence analysis and quantification of their tryptic peptides (Supplementary Fig. 4e). For the first approach, MBP-TNFα fusion proteins obtained from the four different genetic backgrounds were separated by reversed-phase (RP)-liquid chromatography (LC) and the intact mass information was obtained by coupled electrospray ionization mass spectrometry (ESI-MS) and compared with the calculated average molecular weights of C-terminally truncated MBP_mut-TNFα-(1-39)-L31P fusion proteins (Fig. 5d, e). The detected major degradation product MBP_mut- … CP³¹ [lacking the terminal eight amino acid residues of TNFα-ICD-(1-39)] is not an RseP processing product, as it is also present in the absence of over-expressed active protease (Fig. 5e). Residues F³²LSLFSFL³⁹ could be cleaved off by unspecific proteolytic processing (independent of RseP), as to our knowledge there is no E. coli K12 endopeptidase that specifically cleaves Pro-X peptide bonds. We assume that MBP_mut- … CP³¹ is detectable, despite the absence of TNFα TMD residues, because the membrane fraction has not been washed and therefore also contains soluble proteins (but also the hydrophobic heme cofactor bound to the CP motif might target this fusion protein to the membrane).

**Fig. 5: Determination of RseP cleavage sites.**

The full-length MBP_mut-TNFα-(1-39)-L31P fusion protein characterized by the C-terminal sequence … C³⁰PFLSLFSFL³⁹, however, was identified both by its intact mass and by sequence analysis of tryptic peptides (see Supplementary Note 2), for “Empty vector control” and “RseP-Myc H22F” (inactive RseP protein) purifications (Fig. 5e, and Supplementary Fig. 4a, b, e). This full-length MBP_mut-TNFα-(1-39)-L31P protein is able to bind heme by bis-thiolate ligation, resulting in the characteristic absorbance maxima of 370 nm, 450 nm and 550 nm^25,37,38.

However, when MBP_mut-TNFα-(1-39)-L31P was co-expressed with rseP wt or rseP-Myc wt, MBP_mut-TNFα-(1-39)-L31P ( = MBP_mut- … CPFLSLFSFL³⁹) was cleaved between residues F36 and S37 as well as between L33 and S34 of TNFα-(1-39)-L31P in both cases, resulting in the major processing products MBP_mut- … CPFLSLF³⁶ and MBP_mut- … CPFL³³ for both wt proteases (Fig. 5e, and Supplementary Fig. 4c-e). Upon co-expression with rseP wt, we found small amounts of three additional cleavage products, MBP_mut- … CPF³², MBP_mut- … CPFLS³⁴ (Fig. 5e, and Supplementary Fig. 4c) and MBP_mut- … CPFLSLFS³⁷ (Supplementary Table 2). We do not know whether these truncated MBP_mut-TNFα-(1-39)-L31P proteins are also RseP processing products or whether the two major RseP processing products are further degraded by other E. coli proteases. MBP_mut- … CPFL³³ can no longer bind heme by bis-thiolate ligation, since the loss of the C-terminal pentapeptide L³⁵FSFL³⁹ already leads to mono-thiolate ligated heme²⁵ (Fig. 2). MBP_mut- … CPFLSLF³⁶, which lacks the second serine residue Ser37 important for (heme promoted) dimerization²⁵, could still dimerize to a small extent and bind heme by bis-thiolate ligation (compare absorbance of the processing products of RseP at 450 nm, Fig. 4c).

Small amounts of the major processing product of RseP, MBP_mut- … SRRCPFLSLF³⁶, were also identified for the “Empty vector control” but not for “RseP-Myc H22F” by LC-MS/MS analysis (Supplementary Table 2). It is likely that minor cleavage of the substrate protein by chromosomally encoded RseP (and other E. coli proteases) occurs, which can be detected by MS/MS after tryptic digest of purified MBP fusion proteins. This processing of MBP_mut-TNFα-(1-39)-L31P was not detected in intact mass analysis (shown in Fig. 5e and Supplementary Fig. 4a).

In summary, mass spectrometric analysis confirms that RseP cleaves the TNFα substrate within the TMD residues 31-39. The different heme binding modes of substrate and processing product(s) determined by UV/Vis spectroscopy are demonstrably related to C-terminal truncation of the substrate. The mass spectrometric analysis suggests a processive cleavage mechanism of RseP and also gives insights into the specificity of RseP.

Combining error-prone mutagenesis of RseP with iCliPSpy

To identify residues in RseP that are critical for protease activity and/or substrate binding, we combined the in vivo assay iCliPSpy with error-prone mutagenesis of the rseP gene (Fig. 6a, b). We selected seven green E. coli clones co-producing an inactivated RseP enzyme and the substrate MBP_mut-TNFα-(1-39)-L31P for sequencing of the rseP gene. In six out of seven clones, we found premature stop codons and frameshift mutations in addition to several amino acid exchanges (Fig. 6c), and therefore excluded them from further analysis. In contrast, the mutant designated RseP G1 contained eleven non-synonymous mutations, but neither premature stop codons nor frameshift mutations. We mapped the eleven amino acid exchanges onto the AlphaFold wt structure of RseP and found that the Y69H substitution is located within the edge of the β-MRE at a position close to the active-site within the substrate binding groove (Fig. 6d). The edge structure was shown to be important for both protease activity and substrate binding^19,20. For example, the Y69P substitution in RseP significantly impairs the cleavage of the model substrate HA-MBP-RseA148¹⁹. However, the introduction of a Pro residue can significantly alter the structure of the β-MRE^19,20. To further investigate the role of Tyr69 in protease function without altering the overall structure of the β-MRE, we tested whether the single substitution Y69H, in which Tyr69 is exchanged for a more hydrophilic histidine residue, reduces RseP activity in vivo. Assuming that the color of colonies on agar plates reflects the proteolytic activity of RseP, we found a decreased in vivo activity of RseP Y69H compared to RseP wt (Fig. 6e), which was more pronounced for RseP G1 and the active-site variant RseP H22F (Fig. 6e). Next, we focused on a more detailed characterization of residues Tyr69 and Tyr428 by MD simulations and site-directed mutagenesis. Residue Tyr428 of TMD 4 was included in our studies because it is localized on the opposite side of Tyr69 at the putative substrate binding groove, and Tyr69 and Tyr428 together may work in gating substrates²² into the binding groove (Fig. 6d).

**Fig. 6: Error-prone mutagenesis of *rseP* in combination with the iCliPSpy assay.**

MD simulations of membrane-embedded RseP

To further investigate the role of Tyr69 in the function of RseP, we performed MD simulations of RseP based on the structure predicted by AlphaFold^29,30 (with a Zn²⁺ ion manually added to the active site), as the pdb file 7W6X for the crystal structure of full-length RseP²² was not publicly available. To mimic the native environment, RseP was embedded in an artificial membrane resembling the lipid composition of E. coli membranes (see Supplementary Table 3 and Supplementary Note 3) and MD simulations (Fig. 6f and Supplementary Fig. 5a-e) were performed on a time scale of 2 microseconds. The structures obtained showed only minor changes in the RseP active site-containing membrane core compared to the AlphaFold prediction used for the simulations (and compared to the RseP crystal structure). However, the MD simulations indicated that the periplasmic tandem PDZ domain is flexible (Supplementary Fig. 5e). The MD simulations of membrane-embedded RseP predict that Tyr69 and Tyr428 are localized at the cytosol/membrane interface adjacent to the active site of RseP as part of the substrate binding groove (Fig. 6f and Supplementary Fig. 5). The physiological importance of a lipid molecule placed by MD simulation next to/in the substrate binding groove, however, is less clear. In the case of the MD simulation shown in Fig. 6f and Supplementary Fig. 5 this lipid was phosphatidylethanolamine (PE), in other simulations we found cardiolipin, suggesting that overall hydrophobicity is the major determinant for insertion into the substrate binding pocket of RseP.

In our in vivo assay (Fig. 6g), the Y428H substitution had no significant impact on the proteolytic activity of RseP. Furthermore, the double substitution Y69H/Y482H was indistinguishable from the single substitution Y69H (Fig. 6g), suggesting that Tyr428 neither plays a role in binding the substrate MBP_mut-TNFα-(1-39)-L31P nor is important for catalytic activity. To support this result, MBP fusion proteins purified from E. coli cells co-overexpressing MBP_mut-TNFα-(1-39)-L31P with wt or mutant rseP genes were analyzed by SDS-PAGE (Supplementary Fig. 6a) and UV/Vis spectroscopy (Supplementary Fig. 6b). When MBP_mut-TNFα-(1-39)-L31P was co-expressed with rseP Y69H or with rseP Y69H/Y428H, heme remained bound to the reporter via bis-thiolate ligation (Supplementary Fig. 6b), confirming that intramembrane proteolysis by RseP is impaired when Tyr69 is exchanged for His. However, RseP Y428H behaved like the wt enzyme, processing the C-terminal residues of the reporter substrate (compare Fig. 5) and shifting heme binding towards a monomeric, thiolate-bound form.

Discussion

Recently, we observed that the N-terminal domain of TNFα comprising residues 1-39, TNFα-(1-39), dimerizes and binds heme via two cysteine residues, whereas the proteolytically processed peptide TNFα-(1-34) binds heme via one cysteine residue, as the dimerization motif S³⁴XXS³⁷ within the TMD residues 31-39 of TNFα is disrupted²⁵. In the present study, we exploit these spectroscopic properties of TNFα to develop a simple chromophore-based in vivo assay to monitor I-CliP activity. We chose E. coli RseP as a benchmark enzyme, which is well characterized by both in vitro and in vivo assays²³. Our in vivo studies show that co-expression of TNFα-(1-39)-L31P with active rseP wt results in white E. coli colonies, while co-expression with inactive RseP H22F results in green colonies. After purification of the MBP-tagged reporter protein we confirmed by UV/Vis spectroscopy and mass spectrometry that TNFα-(1-39)-L31P is indeed cleaved by RseP. Mass spectrometric analysis showed that RseP engages in the trimming of MBP_mut-TNFα-(1-39)-L31P at L/F-S peptide bonds and releases the tripeptides S³⁷FL³⁹ and S³⁴LF³⁶, whereas the native substrate RseA is cleaved by RseP between Ala108 and Cys109¹³. However, the low substrate specificity of RseP and other I-CLiPs is well known^23,35. Interestingly, tripeptide trimming is also observed for the processing of amyloid β-peptide 49 by γ-secretase, and in this case the trimming mechanism has recently been studied in detail using a novel peptide Gaussian accelerated molecular dynamics method³⁹. In the future, MD simulations may also help to determine the mechanism of RseP more precisely and provide better insight into the mechanism of I-CLiPs in general.

The crystal structures of the tandem PDZ domain of RseP¹⁶ and of the transmembrane core domain of an S2P metalloprotease from Methanocaldococcus jannaschii⁴⁰ led to an AlphaFold structure prediction^29,30 with very high confidence for RseP. We used this prediction for mapping mutations (Fig. 6d) and for MD simulations of RseP (Fig. 6f and Supplementary Fig. 5), because the pdb file for the crystal structure of RseP²² was only available as we were working on the revised version of the manuscript but not in the beginning of our studies. Later, it also turned out that the AlphaFold structure prediction for the membrane core of RseP is very similar to the published crystal structure of RseP (Supplementary Fig. 5f), so that we did not repeat the MD simulations using the crystal structural data. Thus, our approach led to an overall correct picture of the structure of membrane-embedded RseP without knowledge of the crystal structure. Although the physiological importance of a lipid molecule placed next to/into the substrate binding groove by MD is unclear, this may be an important discovery as the lipid has to be exchanged for the substrate protein during catalysis. Future work will show whether this bound lipid is part of a lipid/substrate exchange during RseP-mediate cleavage of proteins.

Using the iCliPSpy in vivo assay, we identified the inactive RseP G1 mutant with eleven non-synonymous mutations, including the substitution Y69H, by error-prone mutagenesis. We focused on characterizing the single mutant RseP Y69H, which has not been previously described. Although the AlphaFold structures of RseP wt and RseP Y69H are nearly identical except for the Y69H exchange, we found decreased proteolytic cleavage of the substrate TNFα-(1-39) for RseP Y69H. The finding of decreased activity of RseP Y69H is consistent with the importance of the edge strand (residues Gly67 to Val70) for substrate binding by RseP, as it had been previously shown that proline substitutions of the edge residues Gly67, Gly68, Tyr69, and Val70 perturb the β-strand conformation of the β-MRE and abolish the proteolytic activity of RseP^19,20,22. It is tempting to speculate that Tyr69 engages in a hydrophobic contact to a side chain of the unwound substrate peptide within the substrate binding groove and that the Y69H exchange reduces the affinity for the substrate. The crystal structure of RseP now shows that the bound inhibitor batimastat is in close contact to the side chains of Phe44 (which is part of the GFG motif²⁰) and Tyr69. How the other mutations within the G1 mutant contribute to the impairment of RseP activity is less clear.

When the photoreactive hydrophobic amino acid, p-benzoyl-L-phenylalanine (pBPA) was incorporated at position 69, 71 or 74 of RseP (e.g. RseP Y69pBPA), RseP retained its proteolytic activity and the substrate RseA was photo-crosslinked to RseP, leading to a model in which the β-MRE binds to substrate TMDs and stabilizes their extended conformation^19,20,22. In the recently proposed mechanism for substrate accommodation and cleavage in RseP, it is assumed that RseP is in equilibrium between a gate-open and gate-closed conformation, that the substrate is bound by the edge residues in the gate-open conformation, and that in a final step the transmembrane segment of the bound substrate is extended by the edge strand to promote proteolytic cleavage²².

Although the obtained results were largely confirmative, our studies convincingly demonstrate the power of the combination of error-prone mutagenesis with the iCliPSpy assay for the characterization of RseP and possibly other I-CliPs at the molecular level.

It is not known what fraction of the substrate protein MBP_mut-TNFα-(1-39)-L31P forms a dimeric complex with heme and what proportion is monomeric, since we neither know the heme-binding constant nor the equilibrium constant for (heme-promoted) dimerization of TNFα-(1-39). Also, the dimerization mechanism itself is not elucidated, but we favor a model in which heme is first bound by monomeric TNFα-(1-39) and then this protein-heme complex binds a second TNFα-(1-39) monomer assisted by the SXXS dimerization motif. We assume (considering the sizes of the peptide binding groove and the active-site) that monomeric TNFα-(1-39) (without ligated heme) and not the dimeric (heme binding) one is finally cleaved by RseP. This does not exclude that initially present dimeric TNFα-(1-39) is bound by RseP and that beta-strand addition dissociates the dimer before cleavage. RseP-catalyzed cleavage of the dimerization motif within monomeric TNFα-(1-39) will result in decreased amounts of peptide TNFα-(1-39) available for (heme promoted) dimerization and increased amounts of cleaved peptides binding heme by only one Cys residue.

In summary, we present iCLiPSpy, a chromophore-based assay for in vivo analysis of intramembrane proteolysis. In addition, combining iCliPSpy with error-prone mutagenesis can help in targeted protein design, for example to increase protease activity. In principle, iCLiPSpy can be extended to other I-CLiPs (that cleave type II transmembrane proteins) as well as to other protein substrates of the THOMAS (Thiolate-Heme Oligomeric type II trans-Membrane proteins with heme binding mode adjusted by SPPL2a/b) protein family²⁵. For example, CD74-NTF also binds heme via bis-thiolate ligation²⁵ and can be used to expand the repertoire of substrates to study I-CLiPs with different substrate specificities. The investigation of other I-CLiPs, such as SPPL2a/b proteases, using iCLiPSpy depends on their expression in E. coli and moreover, in vivo detection of heme binding substrate proteins such as CD74-NTF is only possible at high expression levels (changing the color of cells). Further research and development is needed to transfer this approach to an eukaryotic expression system such as yeast. In combination with the over-expression of I-CliP reporters such as TNFα-(1-39)-L31P or CD74-NTF, green yeast colonies would turn into colorless colonies due to the activity of I-CLiPs. Such a system would be adaptable for high-throughput screening approaches aimed at discovering inhibitors or activators of I-CLiPs, which is of great interest for proteases being explored as biomedical targets.

In pathogenic bacteria, site-2 proteases play an important role in sensing host signals and regulating the expression of virulence genes during infection⁴¹. In Vibrio cholerae, for example the I-CLiP YaeL (a homolog of RseP) is involved in the inactivation of the membrane localized transcription factor TcpP via a two-step RIP pathway. TcpP activates transcription of toxT, which encodes the direct activator of toxin and pilus genes⁴². In Streptococcus pneumoniae, RseP cleaves the competence protein ComM to control its quantity. ComM protects S. pneumoniae from CbpD-mediated lysis during competence^43,44. The iCLiPSpy assay can provide a simple yet powerful tool for the identification/detection and characterization of these and other I-CLiPs from pathogenic bacteria and opens up the possibility of searching for small molecule modulators of protease activity in an in vivo model system.

Methods

Expression plasmids

MBP_mut-TNFα-(1-39)-L31P, MBP_mut-TNFα-(1-34)-L31P, His-MBP-TNFα-(1-39)-L31P and E. coli rseP wt, rseP-Myc wt and mutant genes, respectively, optimized for expression in E. coli were synthesized by Invitrogen GeneArt (Thermo Fisher Scientific). MBP_mut contains substitutions increasing the affinity for amylose (A312V, I317V), reducing the surface entropy (D82A, K83A, E172A, N173A, and K239A) and increasing the crystal-packing interactions between neighboring MBP molecules (A215H/K219H)³¹. MBP only contains the substitution A312V (numbering of amino acid residues refers to the sequence of periplasmic MBP lacking the signal peptide). To minimize differences in expression rates of rseP wt, rseP H22F, rseP Y69H, rseP Y69H/Y428H, rseP Y428H and rseP-Myc wt, rseP-Myc H22F genes due to different codon usage, the sequences only differ in the exchanged codons. Cloning into NcoI/XhoI sites of expression vectors pETDuet-1 (ColE1 replicon, ampicillin resistance) and pCOLADuet-1 (COLA replicon, kanamycin resistance; compatible with pETDuet-1), respectively, was performed using standard molecular biology techniques. The DNA sequence of all cloned genes was confirmed using appropriate primers. DNA sequences of all constructed plasmids (listed in Supplementary Table 1) are given in Supplementary Note 1, so that the exact amino acid sequences of all proteins used in this study are also available. Due to the cloning procedure the second residue (Leu; codon CTG) of all RseP variants is exchanged for Val (codon GTG). RseP L2V and RseP L2V H22F (and so on) are referred to as RseP wt and RseP H22F, respectively, in this paper. Transformation and expression were performed with NEB (New England Biolabs) 5α and T7 Express E. coli cells, respectively.

In vivo cleavage of MBP_mut-TNFα-(1-39)-L31P by RseP

T7 Express E. coli cells co-transformed with pTK1130 [= pETDuet-1-MBP_mut-TNFα-(1-39)-L31P] and pCOLADuet-1 encoded rseP genes were streaked from bacterial glycerol stocks onto selection plates [LB agar substituted with 2 g glucose l^-1, ampicillin (100 µg ml^-1) and kanamycin (50 µg ml^-1), but containing no δ-aminolevulinic acid hydrochloride⁴⁵], incubated overnight at room temperature or 37 °C. Photos of the plates were taken after incubation for several additional days at room temperature. Expression of MBP_mut-TNFα-(1-39)-L31P and rseP occurred because the T7 expression system is leaky.

Purification of membrane-associated MBP fusion proteins

As recently described, we used the MBP tag for amylose affinity purification of the MBP-TNFα fusions in combination with n-dodecyl-β-D-maltopyranoside (DDM, Avanti) as detergent for membrane protein solubilization and subsequent chromatography²⁵. E. coli T7 Express E. coli cells transformed with the appropriate MBP-TNFα encoding pETDuet-1 plasmids (see Supplementary Table 1) were grown in LB medium substituted with 2 g glucose l^-1 and ampicillin (100 µg ml^-1) at 37 °C. Co-expression experiments of pTK1130 [= pETDuet-1-MBP_mut-TNFα-(1-39)-L31P] and pCOLADuet-1 encoded rseP genes, were also performed in E. coli T7 Express cells. Cells were grown in this case in LB medium substituted with 2 g glucose l^-1, ampicillin (100 µg ml^-1) and kanamycin (50 µg ml^-1). For both kinds of experiments, over-expression was induced with 50 µM isopropyl 1-thio-β-D-galactopyranoside (IPTG, Sigma-Aldrich) at an A₆₀₀ of about 0.5, and cultures were then further shaken at 37 °C for about 2 h. Cells from a one liter culture were harvested, resuspended in 30 ml TN buffer (50 mM Tris/HCl pH 7.4 at room temperature and 150 mM NaCl) containing 1 mM phenylmethylsulfonyl fluoride (PMSF, Sigma-Aldrich), frozen in liquid nitrogen and stored at -80 °C. For the experiments shown in Figs. 2d-f, 4, and Supplementary Fig. 3, 0.6 mM δ-aminolevulinic acid hydrochloride was added before induction with IPTG⁴⁵. However, δ-aminolevulinic acid hydrochloride was not used in all other experiments. Thawed cells were again substituted with 1 mM PMSF and then disrupted by sonication. The membrane pellet obtained after 30 to 45 min centrifugation at 40,000x g and 4 °C was resuspended in 15 ml TN buffer containing 2% of the detergent DDM. Then, membrane proteins were solubilized at 4 °C for about 30 to 45 min in a 50 ml tube using a tube roller mixer. The supernatant obtained after centrifugation for 30 min at 40,000x g and 4 °C was subjected to amylose (NEB) affinity chromatography at room temperature. Amylose affinity columns were equilibrated and washed with TN buffer containing 0.2% DDM. MBP fusion proteins were eluted with 10 mM maltose (Sigma-Aldrich) in equilibration buffer and UV/Vis spectra of the eluted protein solutions were recorded at room temperature in a 1 cm Hellma quartz absorption cuvette using the NanoDrop 2000c spectrophotometer from Thermo Fisher Scientific with elution buffer as blank²⁵.

Gel filtration

For gel filtration experiments, eluted MBP fusion proteins were concentrated using Vivaspin 500 columns at 4 °C and then stored at 4 °C. Gel filtration was carried out as described²⁵ at 4 °C using an ÄKTA pure system from GE Healthcare and a Superose 6 Increase 10/300 GL column for high-resolution size exclusion chromatography. 100 µl of protein aliquots were subjected to the column equilibrated in TN buffer containing 0.2% DDM at a flow rate of 0.4 ml min^-1. The elution was followed by absorbance at 280 nm and simultaneously for the detection of heme at additional wavelengths (such as 371 nm and 451 nm), and 0.4 ml fractions were collected. The void volume of the used column was determined with Blue dextran 2000 (GE Healthcare) to be 8.5 ml.

Mass spectrometric analysis of RseP processing products

Intact masses of MBP-TNFα fusion proteins were determined using reversed-phase (RP) ultra high-performance liquid chromatography (UHPLC) coupled to electrospray ionization mass spectrometry (RP-ESI-MS). Briefly, RP separation was performed on a 1290 UHPLC (Agilent Technologies) equipped with a PLRP-S column (1.0 ×50 mm, 5 µm, 1000 Å; Agilent Technologies) at 80 °C and 0.35 ml/min. Solvent A was 0.1% formic acid (FA) in water. Solvent B was 0.1% FA in acetonitrile. Total run time was 17 min and the gradient was as follows. 0-0.5 min: 20% B, 0.5-8.5 min: linear increase from 20% to 50% B, 8.5-9.0 min: linear increase from 50% to 90% B, 9.0-11.0 min: 90% B, 11.0-11.5 min: linear decrease from 90% to 20% B, and 11.5-17 min: re-equilibration at 20% B. MS analysis was performed (min 4 to 13) in positive polarity scan mode on an Impact II ESI-QqTOF mass spectrometer (Bruker Daltonics), in a mass range from m/z 600 to 7000. Following ESI source parameters were used: end plate offset 500 V, capillary voltage 4500 V, nebulizer gas 3.0 bar, dry gas 12.0 l/min, dry temperature 300 °C. High mass calibration was performed using cesium fluoroheptanoate in enhanced quadratic calibration mode. Mass spectra were processed in DataAnalysis software version 4.4 (Bruker Daltonics). In brief, the protein is getting detected with different charge states in parallel, that are then deconvoluted to obtain the molecular weight of the protein. Masses were determined using maximum entropy charge deconvolution in a range of 10,000 – 50,000 Da. To confirm RseP cleavage specificity in a complementary analysis, tryptic peptides of MBP-TNFα fusion proteins were sequenced by LC-MS/MS (for a detailed setup of this peptide analysis see Supplementary Note 2). Since no extracted ion chromatograms (XIC) of the intact mass analysis could be obtained due to the nature of the analysis, quantification was performed by LC-MS/MS on the peptide level (compare Supplementary Table 2).

Error-prone PCR

Error-prone PCR²⁸ was performed using the JBS (Jena Bioscience) Error-prone kit with plasmid pTK1121 (= pCola-Duet-1-rseP wt) as template for the rseP gene and using the forward and reverse PCR primers TK154 (5´-AACTTTAATAAGGAGATATACCATGG-3´) and TK155 (5´-GGTTTCTTTACCAGACTCGAGTTATTA-3´) (NcoI and XhoI restriction sites are underlined). The rate of mutagenesis of this method is in the range of 0.6-2.0% and the mutations are randomly distributed throughout the amplified PCR product²⁸. The PCR product was purified using the Qiagen PCR purification kit, double digested with NcoI and XhoI and then isolated by agarose gel electrophoresis using the Qiagen gel extraction kit. The NcoI/XhoI rseP (error prone) PCR fragment was then ligated with NcoI/XhoI double-digested pCOLA-Duet-1 and the ligation mixture was transformed into E. coli NEB5α cells. An aliquot of the transformation mixture was used to inoculate 50 ml LB + glucose medium. Plasmids of cells grown overnight at 37 °C were then purified using the Nucleobond PC100 Midi kit. The purified plasmid pCola-Duet-1-rseP (error prone) was subsequently co-transformed with pTK1130 [= pETDuet-1-MBP_mut-TNFα-(1-39)-L31P] into NEB T7 Express E. coli cells. The diluted transformation mixture was spread onto selection plates containing ampicillin and kanamycin, incubated overnight at 37 °C and then for two days at room temperature. White and green clones were then transferred to a fresh selection plate, incubated overnight at 37 °C and then for one additional day at room temperature. Plasmid DNA [pETDuet-1-MBP_mut-TNFα-(1-39)-L31P and pCola-Duet-1-rseP (error prone)] of green clones was purified and the sequence of the mutated rseP genes was determined using the pCOLA-Duet-1 specific primers TK160 (5´-CTCCTGCATTAGGAAATTAATACG-3´) and TK161 (5´-GTGCTTCTCAAATGCCTGAGG-3´) and in addition by using several different rseP specific primers.

SDS-PAGE and immunoblotting

As described²⁵, proteins were separated using Tricine-SDS-polyacrylamide (10%) gel electrophoresis under reducing (1x sample buffer containing 10 mM dithiothreitol, DTT) conditions⁴⁶. Molecular weights of standard proteins (all blue prestained protein standard, Bio-Rad) are indicated on the side of the gel pictures. For documentation, destained gels were scanned using an Epson Perfection V850 Pro scanner with 1200 d.p.i. resolution. Proteins were electrophoretically transferred to PVDF membrane (Immobilon-FL, Merck Millipore) after SDS-PAGE by semi-dry blotting. Myc-tagged proteins were detected with a monoclonal anti-Myc Tag antibody (Sigma-Aldrich SAB1305535, clone 9E10, diluted 1:500) as first antibody, an Alexa 680 goat-anti-mouse antibody (Life Technologies, Thermo Fischer Scientific, 0.8 µg ml^-1) as secondary antibody and then using the Odyssey Imaging System (Li-Cor Biosciences).

AlphaFold and MD simulations

We used the AlphaFold^29,30 predicted structure of E. coli K12 RseP (with a Val residue in position 2) for our MD simulations. Into the predicated structure we manually placed a Zn²⁺ ion coordinated by the side chains of residues H22, H26 and D402. The modified structure was then embedded in silico into a described E. coli model membrane system containing about 75% cardiolipin and about 25% PE with multiple types of acyl chains⁴⁷ using the CHARMM GUI membrane builder interface^48,49. Six MD simulations for membrane-embedded RseP wt, were performed each for 2 µs using GROMACS⁵⁰ software and CHARMM36m forcefield⁵¹ (for a detailed setup of the methodology compare Supplementary Table 3 and Supplementary Note 3). To choose the statistically most representative structures for illustration, the obtained trajectories were analyzed by clustering using the GROMOS method⁵² applying an RMSD cutoff of 0.15 nm. All pdb files were visualized with UCSF Chimera⁵³.

Statistics and reproducibility

No statistical analyses were performed. Wt and mutant proteins were always purified in parallel using the same purification procedure including the same buffers and detergent solutions (and using cells grown under the same conditions). The precise numbers of purifications are indicated in the figure legends.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All underlying data of this study are available in the main article and Supplementary Information. The source data behind Figs. 2f, 4c-e, 5c are provided as Supplementary Data 1 and the DNA sequences of all constructed plasmids are included in Supplementary Note 1. The uncropped version of Fig. 2c is included as Supplementary Fig. 7. The mass spectrometry proteomics data have been deposited with the ProteomeXchange Consortium via the PRIDE⁵⁴ partner repository with the dataset identifier PXD038785 and 10.6019/PXD038785. All other data are available from the corresponding authors upon reasonable request.

Code availability

No customed or in-house software was used in the analysis of the MD simulations. AlphaFold2 model and output files as well as MD simulations parameter files are publicly available at Zenodo: https://doi.org/10.5281/zenodo.7612846⁵⁵.

References

Kühnle, N., Dederer, V. & Lemberg, M. K. Intramembrane proteolysis at a glance: from signalling to protein degradation. J. Cell. Sci. 132, jcs217745 (2019).
Article PubMed Google Scholar
Wolfe, M. S. Unraveling the complexity of γ-secretase. Semin Cell Dev. Biol. 105, 3–11 (2020).
Article CAS PubMed PubMed Central Google Scholar
Bradley, J. R. TNF-mediated inflammatory disease. J. Pathol. 214, 149–160 (2008).
Article CAS PubMed Google Scholar
Brown, M. S., Ye, J., Rawson, R. B. & Goldstein, J. L. Regulated intramembrane proteolysis: a control mechanism conserved from bacteria to humans. Cell 100, 391–398 (2000).
Article CAS PubMed Google Scholar
Lichtenthaler, S. F., Lemberg, M. K. & Fluhrer, R. Proteolytic ectodomain shedding of membrane proteins in mammals-hardware, concepts, and recent developments. EMBO J. 37, e99456 (2018).
Article PubMed PubMed Central Google Scholar
McCarthy, A. J., Coleman-Vaughan, C. & McCarthy, J. V. Regulated intramembrane proteolysis: emergent role in cell signalling pathways. Biochem. Soc. Trans. 45, 1185–1202 (2017).
Article CAS PubMed Google Scholar
Dartigalongue, C., Loferer, H. & Raina, S. EcfE, a new essential inner membrane protease: its role in the regulation of heat shock response in Escherichia coli. EMBO J. 20, 5908–5918 (2001).
Article CAS PubMed PubMed Central Google Scholar
Kanehara, K., Akiyama, Y. & Ito, K. Characterization of the yaeL gene product and its S2P-protease motifs in Escherichia coli. Gene 281, 71–79 (2001).
Article CAS PubMed Google Scholar
Akiyama, Y., Kanehara, K. & Ito, K. RseP (YaeL), an Escherichia coli RIP protease, cleaves transmembrane sequences. EMBO J. 23, 4434–4442 (2004).
Article CAS PubMed PubMed Central Google Scholar
Alba, B. M., Leeds, J. A., Onufryk, C., Lu, C. Z. & Gross, C. A. DegS and YaeL participate sequentially in the cleavage of RseA to activate the σ^E-dependent extracytoplasmic stress response. Genes Dev. 16, 2156–2168 (2002).
Article CAS PubMed PubMed Central Google Scholar
Alba, B. M., Zhong, H. J., Pelayo, J. C. & Gross, C. A. degS (hhoB) is an essential Escherichia coli gene whose indispensable function is to provide σ^E activity. Mol. Microbiol. 40, 1323–1333 (2001).
Article CAS PubMed Google Scholar
Kanehara, K., Ito, K. & Akiyama, Y. YaeL (EcfE) activates the σ^E pathway of stress response through a site-2 cleavage of anti-σ^E, RseA. Genes Dev. 16, 2147–2155 (2002).
Article CAS PubMed PubMed Central Google Scholar
Li, X. et al. Cleavage of RseA by RseP requires a carboxyl-terminal hydrophobic amino acid following DegS cleavage. Proc. Natl Acad. Sci. USA 106, 14837–14842 (2009).
Article CAS PubMed PubMed Central Google Scholar
Hizukuri, Y. & Akiyama, Y. PDZ domains of RseP are not essential for sequential cleavage of RseA or stress-induced σ^E activation in vivo. Mol. Microbiol. 86, 1232–1245 (2012).
Article CAS PubMed Google Scholar
Hizukuri, Y. et al. A structure-based model of substrate discrimination by a noncanonical PDZ tandem in the intramembrane-cleaving protease RseP. Structure 22, 326–336 (2014).
Article CAS PubMed Google Scholar
Inaba, K. et al. A pair of circularly permutated PDZ domains control RseP, the S2P family intramembrane protease of Escherichia coli. J. Biol. Chem. 283, 35042–35052 (2008).
Article CAS PubMed PubMed Central Google Scholar
Kanehara, K., Ito, K. & Akiyama, Y. YaeL proteolysis of RseA is controlled by the PDZ domain of YaeL and a Gln-rich region of RseA. EMBO J. 22, 6389–6398 (2003).
Article CAS PubMed PubMed Central Google Scholar
Koide, K., Maegawa, S., Ito, K. & Akiyama, Y. Environment of the active site region of RseP, an Escherichia coli regulated intramembrane proteolysis protease, assessed by site-directed cysteine alkylation. J. Biol. Chem. 282, 4553–4560 (2007).
Article CAS PubMed Google Scholar
Akiyama, K. et al. Roles of the membrane-reentrant beta-hairpin-like loop of RseP protease in selective substrate cleavage. Elife 4, e08928 (2015).
Article PubMed PubMed Central Google Scholar
Akiyama, K., Hizukuri, Y. & Akiyama, Y. Involvement of a conserved GFG motif region in substrate binding by RseP, an Escherichia coli S2P protease. Mol. Microbiol 104, 737–751 (2017).
Article CAS PubMed Google Scholar
Miyake, T., Hizukuri, Y. & Akiyama, Y. Involvement of a Membrane-Bound Amphiphilic Helix in Substrate Discrimination and Binding by an Escherichia coli S2P Peptidase RseP. Front Microbiol 11, 607381 (2020).
Article PubMed PubMed Central Google Scholar
Imaizumi, Y. et al. Mechanistic insights into intramembrane proteolysis by E. coli site-2 protease homolog RseP. Sci. Adv. 8, eabp9011 (2022).
Article CAS PubMed PubMed Central Google Scholar
Hizukuri, Y., Akiyama, K. & Akiyama, Y. Biochemical Characterization of Function and Structure of RseP, an Escherichia coli S2P Protease. Methods Enzymol. 584, 1–33 (2017).
Article CAS PubMed Google Scholar
Saito, A. et al. Post-liberation cleavage of signal peptides is catalyzed by the site-2 protease (S2P) in bacteria. Proc. Natl Acad. Sci. USA 108, 13740–13745 (2011).
Article CAS PubMed PubMed Central Google Scholar
Kupke, T., Klare, J. P. & Brügger, B. Heme binding of transmembrane signaling proteins undergoing regulated intramembrane proteolysis. Commun. Biol. 3, 73 (2020).
Article CAS PubMed PubMed Central Google Scholar
Horiuchi, T., Mitoma, H., Harashima, S., Tsukamoto, H. & Shimoda, T. Transmembrane TNF-α: structure, function and interaction with anti-TNF agents. Rheumatol. (Oxf.) 49, 1215–1228 (2010).
Article CAS Google Scholar
Fluhrer, R. et al. A gamma-secretase-like intramembrane cleavage of TNFα by the GxGD aspartyl protease SPPL2b. Nat. Cell Biol. 8, 894–896 (2006).
Article CAS PubMed Google Scholar
Cadwell, R. C. & Joyce, G. F. Randomization of genes by PCR mutagenesis. PCR Methods Appl 2, 28–33 (1992).
Article CAS PubMed Google Scholar
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Article CAS PubMed PubMed Central Google Scholar
Varadi, M. et al. AlphaFold Protein Structure Database: massively expanding the structural coverage of protein-sequence space with high-accuracy models. Nucleic Acids Res 50, D439–D444 (2022).
Article CAS PubMed Google Scholar
Bokhove, M. et al. Easy mammalian expression and crystallography of maltose-binding protein-fused human proteins. J. Struct. Biol. 194, 1–7 (2016).
Article CAS PubMed PubMed Central Google Scholar
Briand, L. et al. A self-inducible heterologous protein expression system in Escherichia coli. Sci. Rep. 6, 33037 (2016).
Article CAS PubMed PubMed Central Google Scholar
Zhang, Z. et al. High-level production of membrane proteins in E. coli BL21(DE3) by omitting the inducer IPTG. Micro. Cell Fact. 14, 142 (2015).
Article Google Scholar
Hanna, D. A. et al. Heme dynamics and trafficking factors revealed by genetically encoded fluorescent heme sensors. Proc. Natl Acad. Sci. USA 113, 7539–7544 (2016).
Article CAS PubMed PubMed Central Google Scholar
Koide, K., Ito, K. & Akiyama, Y. Substrate recognition and binding by RseP, an Escherichia coli intramembrane protease. J. Biol. Chem. 283, 9562–9570 (2008).
Article CAS PubMed Google Scholar
Douchin, V., Bohn, C. & Bouloc, P. Down-regulation of porins by a small RNA bypasses the essentiality of the regulated intramembrane proteolysis protease RseP in Escherichia coli. J. Biol. Chem. 281, 12253–12259 (2006).
Article CAS PubMed Google Scholar
Barr, I. et al. DiGeorge critical region 8 (DGCR8) is a double-cysteine-ligated heme protein. J. Biol. Chem. 286, 16716–16725 (2011).
Article CAS PubMed PubMed Central Google Scholar
Faller, M., Matsunaga, M., Yin, S., Loo, J. A. & Guo, F. Heme is involved in microRNA processing. Nat. Struct. Mol. Biol. 14, 23–29 (2007).
Article CAS PubMed Google Scholar
Bhattarai, A. et al. Mechanism of tripeptide trimming of amyloid β-peptide 49 by β-Secretase. J. Am. Chem. Soc. 144, 6215–6226 (2022).
Article CAS PubMed Google Scholar
Feng, L. et al. Structure of a site-2 protease family intramembrane metalloprotease. Science 318, 1608–1612 (2007).
Article CAS PubMed Google Scholar
Schneider, J. S. & Glickman, M. S. Function of site-2 proteases in bacteria and bacterial pathogens. Biochim Biophys. Acta 1828, 2808–2814 (2013).
Article CAS PubMed PubMed Central Google Scholar
Teoh, W. P., Matson, J. S. & DiRita, V. J. Regulated intramembrane proteolysis of the virulence activator TcpP in Vibrio cholerae is initiated by the tail-specific protease (Tsp). Mol. Microbiol. 97, 822–831 (2015).
Article CAS PubMed Google Scholar
Marquart, M. E. Pathogenicity and virulence of Streptococcus pneumoniae: Cutting to the chase on proteases. Virulence 12, 766–787 (2021).
Article CAS PubMed PubMed Central Google Scholar
Straume, D., Stamsas, G. A., Salehian, Z. & Havarstein, L. S. Overexpression of the fratricide immunity protein ComM leads to growth inhibition and morphological abnormalities in Streptococcus pneumoniae. Microbiol. (Read.) 163, 9–21 (2017).
Article CAS Google Scholar
Verderber, E. et al. Role of the hemA gene product and delta-aminolevulinic acid in regulation of Escherichia coli heme synthesis. J. Bacteriol. 179, 4583–4590 (1997).
Article CAS PubMed PubMed Central Google Scholar
Schägger, H. & Jagow, G. Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368–379 (1987).
Article PubMed Google Scholar
Wu, E. L. et al. CHARMM-GUI Membrane Builder toward realistic biological membrane simulations. J. Comput Chem. 35, 1997–2004 (2014).
Article CAS PubMed PubMed Central Google Scholar
Jo, S., Kim, T. & Im, W. Automated builder and database of protein/membrane complexes for molecular dynamics simulations. PLoS One 2, e880 (2007).
Article PubMed PubMed Central Google Scholar
Jo, S., Lim, J. B., Klauda, J. B. & Im, W. CHARMM-GUI Membrane Builder for mixed bilayers and its application to yeast membranes. Biophys. J. 97, 50–58 (2009).
Article CAS PubMed PubMed Central Google Scholar
Lindahl, E., Hess, B. & van der Spoel, D. GROMACS 3.0: A package for molecular simulation and trajectory analysis. J. Mol. Mod. 7, 306–317 (2001).
Article CAS Google Scholar
Huang, J. et al. CHARMM36m: an improved force field for folded and intrinsically disordered proteins. Nat. Methods 14, 71–73 (2017).
Article CAS PubMed Google Scholar
Daura, X. et al. Peptide Folding: When simulation meets Experiment. Angew. Chem. Int. Ed. 38, 236–240 (1999).
Article CAS Google Scholar
Pettersen, E. F. et al. UCSF Chimera-a visualization system for exploratory research and analysis. J. Comput Chem. 25, 1605–1612 (2004).
Article CAS PubMed Google Scholar
Perez-Riverol, Y. et al. The PRIDE database resources in 2022: a hub for mass spectrometry-based proteomics evidences. Nucleic Acids Res. 50, D543–D552 (2022).
Article CAS PubMed Google Scholar
Lolicato, F. & Beck, R. AlphaFold2 model and Molecular dynamics simulations files for “In vivo characterization of the bacterial intramembrane-cleaving protease RseP using the heme binding tag-based assay iCliPSpy”. https://doi.org/10.5281/zenodo.7612846. (2022).
Zhang, L. & Guarente, L. Heme binds to a short sequence that serves a regulatory function in diverse proteins. EMBO J. 14, 313–320 (1995).
Article CAS PubMed PubMed Central Google Scholar

Download references

Acknowledgements

This work was funded by the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) – Project Number 112927078-TRR83, and by the Ministry of Education and Research (BMBF) – project Drugs4Future (13FH8I05IA) within the “FH-Impuls” framework. CSC-IT Centre for Science Ltd. (Espoo, Finland) is acknowledged for excellent computational resources. The authors gratefully acknowledge the data storage service SDS@hd supported by the Ministry of Science, Research, and the Arts Baden-Württemberg (MWK) and the DFG through grant INST 35/1314-1 FUGG and INST 35/1503-1 FUGG. For the publication fee we acknowledge financial support by DFG within the funding programme „Open Access Publikationskosten“ as well as by Heidelberg University.

Funding

Open Access funding enabled and organized by Projekt DEAL.

Author information

Authors and Affiliations

Heidelberg University Biochemistry Center (BZH), Im Neuenheimer Feld 328, 69120, Heidelberg, Germany
Thomas Kupke, Rainer Beck, Fabio Lolicato, Walter Nickel & Britta Brügger
Center for Mass Spectrometry and Optical Spectroscopy (CeMOS), Mannheim University of Applied Sciences, Paul-Wittsack-Str. 10, 68163, Mannheim, Germany
Rabea M. Götz, Florian M. Richter & Carsten Hopf
Department of Physics, University of Helsinki, Helsinki, Finland
Fabio Lolicato

Authors

Thomas Kupke
View author publications
You can also search for this author in PubMed Google Scholar
Rabea M. Götz
View author publications
You can also search for this author in PubMed Google Scholar
Florian M. Richter
View author publications
You can also search for this author in PubMed Google Scholar
Rainer Beck
View author publications
You can also search for this author in PubMed Google Scholar
Fabio Lolicato
View author publications
You can also search for this author in PubMed Google Scholar
Walter Nickel
View author publications
You can also search for this author in PubMed Google Scholar
Carsten Hopf
View author publications
You can also search for this author in PubMed Google Scholar
Britta Brügger
View author publications
You can also search for this author in PubMed Google Scholar

Contributions

T.K. formed the concept for the in vivo intramembrane-cleaving protease assay, designed the overall research in coordination with B.B. and performed all experiments apart from analytical mass spectrometry which was done by R.G., F.R. and C.H. R.B. and F.L. did the MD simulations. T.K. prepared Figs. 1 to 6 and Supplementary Figs. 1 to 7 (apart from Supplementary Fig. 5e which was prepared by RB). Funding acquisition: B.B., C.H. and W.N. T.K. and B.B. wrote the manuscript. Review and editing: R.G., F.R., F.L., R.B., C.H. and W.N.

Corresponding authors

Correspondence to Thomas Kupke or Britta Brügger.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Communications Biology thanks the anonymous reviewers for their contribution to the peer review of this work. Primary Handling Editors: Ingrid Span and Gene Chong. Peer reviewer reports are available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Peer Review File

Supplementary Information

Description of Additional Supplementary Files

Supplementary Data 1

Reporting Summary

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Kupke, T., Götz, R.M., Richter, F.M. et al. In vivo characterization of the bacterial intramembrane-cleaving protease RseP using the heme binding tag-based assay iCliPSpy. Commun Biol 6, 287 (2023). https://doi.org/10.1038/s42003-023-04654-z

Download citation

Received: 07 June 2022
Accepted: 02 March 2023
Published: 18 March 2023
DOI: https://doi.org/10.1038/s42003-023-04654-z

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.