The synthetic genes for the three different variants of murine procathepsin B and the variant of procathepsin L with an N-terminal His6-tag were ordered cloned into a pET28a( +) vector (BioCat GmbH, Heidelberg, Germany), where the genes were inserted via seamless cloning using AACTTTAAGAAGGAGATATACC at the 5´site and the XhoI restriction site at the 3´site, respectively. The propeptide sequences were retained, while the signal peptide sequences were excluded in the design of the synthetic genes. The sequences were codon-optimized for expression in E. coli. All genes contained an N- or C-terminal His6-tag or a His6-tag after the sequence for the N-terminal propeptide for purification via affinity chromatography together with an additional cleavage site for the TEV-protease (ENLYFQS) to have the possibility to cleave off the His6-tag. The procathepsin L construct CTSL_C-6xHis was generated due to the deletion of the N-terminal His6-tag and insertion of a C-terminal His6-tag using site-directed mutagenesis. The N-terminal His6-tag was deleted using the forward primer 5′-ACCCCGAAATTCGATCAG-3′ and the reverse primer 5′-CATGGTATATCTCCTTCTTAAAG-3′ and the C-terminal His6-tag of the pET28a( +) vector was inserted using the forward primer 5′-acttccagagcCTCGAGCACCACCACCA-3′ and the reverse primer 5′-acagattttcATTAACAACCGGATAACTTGCTGC-3′. The amino acid and nucleotide sequences of the variants of procathepsin B and L are listed in the Supplementary Information.

To investigate the function of the C-terminal propeptide of CTSB, the construct CTSB_N-6xHis was used for site-directed mutagenesis where the C-terminal propeptide was deleted. For this variant, the forward primer 5′-GCGCACCGATtagTATTGGGGTC-3′ and the reverse primer 5′-GGAATGCCTGCCACAATTTC-3′ were used.

For gene expression, the pET28a( +) vector containing the desired gene sequence was introduced into E. coli SHuffle® T7 Express cells (New England Biolabs, Frankfurt am Main, Germany) by the heat-shock method. For each gene expression, a single colony of freshly transformed E. coli SHuffle® T7 Express cells with the desired gene sequence was picked and used to inoculate 4 mL lysogeny broth (LB) medium supplemented with 50 µg mL−1 kanamycin. Subsequently, the cultures were grown overnight at 30 °C at 180 rpm. These starter cultures were used to inoculate the 50 mL main cultures in which the cells were grown. For the main cultures either LB, terrific broth (TB), LB autoinduction medium (LB-AIM), or TB autoinduction medium (TB-AIM) were used. When an optical density (OD600) of approximately 0.8–1.0 was reached, the gene expressions of the LB or TB cultures were induced with a final concentration of 0.4 mM isopropyl-β-D-thiogalactopyranoside (IPTG) and were shaken for 26 h at 16 °C at 160 rpm. The LB-AIM and TB-AIM cultures were also shifted to 16 °C for 26 h at 160 rpm when an optical density (OD600) of approximately 0.8–1.0 was reached. The main cultures were harvested by centrifugation for 20 min at 4500 × g at 4 °C and washed once with sodium phosphate buffer (50 mM, pH 6.0). The cell pellets were stored at − 20 °C until further use. All constructs were expressed three times.

The harvested bacteria pellets were resuspended in 4 mL equilibration buffer (50 mM sodium phosphate, 300 mM NaCl, 10 mM imidazole, pH 8.0) for each gram of cell pellet. Afterward, the cells were disrupted using ultrasonication with 50% cycle and 30% power on ice. The sonication procedure consisted of 4 min sonication followed by a 2 min break and another 4 min of sonication. Subsequently, the samples were centrifuged at 10,000 × g for 30 min at 4 °C for separation of the cell debris from the supernatant and the lysates were purified.

Columns of 3 mL Ni-imino diacetate (IDA, Carl Roth, Karlsruhe, Germany) for gravity flow chromatography for the purification of His-tagged proteins were prepared. The columns were washed three times with one column volume of cold MilliQ water followed by five column volume washing steps with equilibration buffer (50 mM sodium phosphate, 300 mM NaCl, 10 mM imidazole, pH 8.0). Afterwards, the clarified lysates containing the desired proteins were transferred onto the Ni-IDA columns, incubated for 30 min on ice, washed 10 times with washing buffer (50 mM sodium phosphate, 300 mM NaCl, 20 mM imidazole, pH 8.0), and eluted in 2 mL fractions with elution buffer (50 mM sodium phosphate, 300 mM NaCl, 250 mM imidazole, pH 8.0). All five elution fractions were pooled.

Prior to protein concentration measurements, the protein samples were rebuffered to remove imidazole in the elution buffer during protein purification. For this, the samples were transferred into centricons with a membrane cutoff of 10 kDa and centrifuged for 20 min at 4500 × g at 4 °C and the protein solutions were filled up to 5 mL with CTSB/CTSL measuring buffer (100 mM sodium acetate, 5 mM calcium chloride, pH 5.5). The two steps of centrifugation and filling up with CTSB/CTSL measuring buffer were repeated two more times.

The protein concentrations were measured at 280 nm via NanoDrop 1000 (Thermo Scientific, Wilmington, DE, USA), and the protein yields were calculated based on the extinction coefficients which were determined using the Expasy tool ProtParam (https://web.expasy.org/protparam/, Table S1).

The recombinant procathepsins were autocatalytically activated. For this, subsequently after rebuffering, the recombinant procathepsins B and procathepsin L were incubated with 10 mM dithiothreitol (DTT) at 37 °C until fully activated protein could be verified by SDS-PAGE analysis. Precipitated protein was removed by filtration after the activation process.

The activities of both cathepsins were determined in measuring buffer (100 mM sodium acetate, 5 mM calcium chloride, pH 5.5) using chromogenic substrates. The activity of cathepsin B was measured using the substrate Z-Arg-Arg-AMC (Bachem, Bubendorf, Switzerland), and the activity of cathepsin L was measured using the substrate Z-Phe-Arg-AMC (Bachem). In both cases, the fluorescence was measured at an extinction wavelength of 360 nm and an emission wavelength of 470 nm. KM values were determined by measuring initial rates with varying substrate concentrations. For CTSB, the final substrate concentrations of 0, 0.05, 0.1, 0.2, 0.5, 1, 2, and 4 mM of Z-Arg-Arg-AMC were used. For CTSL, the final substrate concentrations of 0, 5, 10, 25, 50, 75, 100, and 150 µM were used. Ninety microliters of the substrate in measuring buffer with the addition of 10 mM DTT were prepared in a reaction plate, and the reaction was started by the addition of 10 µL of a 1 µg mL−1 CTSB or CTSL solution. The KM values for CTSB were calculated with a nonlinear regression for the fit function for the Michaelis–Menten kinetics and for CTSL with a nonlinear regression using the substrate-inhibition fit in GraphPad Prism 8.4.3 (GraphPad Software, USA) software.