The Cysteine Protease Legumain Is Upregulated by Vitamin D and Is a Regulator of Vitamin D Metabolism in Mice

Legumain is a lysosomal cysteine protease that has been implicated in an increasing amount of physiological and pathophysiological processes. However, the upstream mechanisms regulating the expression and function of legumain are not well understood. Here, we provide in vitro and in vivo data showing that vitamin D3 (VD3) enhances legumain expression and function. In turn, legumain alters VD3 bioavailability, possibly through proteolytic cleavage of vitamin D binding protein (VDBP). Active VD3 (1,25(OH)2D3) increased legumain expression, activity, and secretion in osteogenic cultures of human bone marrow stromal cells. Upregulation of legumain was also observed in vivo, evidenced by increased legumain mRNA in the liver and spleen, as well as increased legumain activity in kidneys from wild-type mice treated with 25(OH)D3 (50 µg/kg, subcutaneously) for 8 days compared to a control. In addition, the serum level of legumain was also increased. We further showed that active legumain cleaved purified VDBP (55 kDa) in vitro, forming a 45 kDa fragment. In vivo, no VDBP cleavage was found in kidneys or liver from legumain-deficient mice (Lgmn−/−), whereas VDBP was cleaved in wild-type control mice (Lgmn+/+). Finally, legumain deficiency resulted in increased plasma levels of 25(OH)D3 and total VD3 and altered expression of key renal enzymes involved in VD3 metabolism (CYP24A1 and CYP27B1). In conclusion, a regulatory interplay between VD3 and legumain is suggested.


Introduction
Legumain is a cysteine endopeptidase with strict specificity for hydrolysis of peptide bonds C-terminally of asparagine residues [1], hence the synonym asparaginyl endopeptidase (AEP).Although mainly lysosomal, legumain has also been shown to be secreted and detectable in plasma.In addition, legumain is postulated to have autocrine/paracrine functions (reviewed in [2]).Legumain is highly expressed in the kidneys, liver, and spleen [3] and is described as being involved in the pathogenesis of several disorders, such as cardiovascular diseases (reviewed in [2]).We have previously shown that legumain expression is altered in the bone microenvironment of patients with osteoporosis [4].An increasing Mini Kit was purchased from QIAGEN, Hilden, Germany.GentleMACS™ M Tubes were purchased from Miltenyi Biotec, Bergisch Gladbach, Germany.EconoSpin ® spin columns were purchased from Epoch Life Science, Missouri City, TX, USA.Primers for mouse legumain, VDBP, CYP27B1, CYP24A1, RPLP0, GAPDH, and β-actin were purchased from Ebersberg, Germany.Bovine mature active legumain (36 kDa) was acquired as previously described [23].

Identification of Putative Vitamin D-Responsive Elements in the LGMN Promoter Region
The nucleotide sequence cut-off of the human legumain (LGMN) gene promoter (accession no.: NM_005606) was set to 1485 base pairs downstream and 15 base pairs upstream of the transcription start site.The nucleotide sequence was retrieved using the Sequence Retrieval Tool in the Eukaryotic Promoter Database (https://epd.epfl.ch(accessed on 20 September 2022)).Putative vitamin D-responsive elements (VDRE) were identified using the PROMO database [24,25] with the maximum matrix dissimilarity rate set to 15.

Harvesting of Cell-conditioned Media and Lysates
Cell-conditioned media were collected and centrifuged at 800 rpm for 5 min at 4 • C, and the supernatants were frozen at −20 • C. Adherent cells were washed with PBS before harvesting in legumain lysis buffer (100 mM sodium citrate, 1 mM disodium-EDTA, 1% n-octyl-β-D-glucopyranoside, pH 5.8) for quantitation of legumain activity or Buffer RLT Plus for mRNA isolation.Cell lysates harvested in legumain lysis buffer were frozen (−70 • C) and thawed (30 • C) three times before centrifugation at 10,000× g for 5 min, and the supernatants were frozen at −20 • C or directly analyzed.Total protein concentrations in cell lysates were determined by measuring absorbance at 595 nm in a microplate reader (Wallac Victor ® 3™ or Wallac Victor ® Nivo™, Perkin Elmer, Boston, MA, USA) according to Bradford [27] and the manufacturer.Bovine serum albumin (0-400 µg/mL) was used to generate a standard curve for the calculation of total protein concentrations.All measurements were performed in triplicates.

Legumain-Deficient Mice
Legumain-deficient (Lgmn −/− ) mice were produced using CRISPR/Cas9 gene targeting in C57BL/6J mouse embryos following established molecular and animal husbandry techniques [28].A single guide RNA (sgRNA) was designed to target within exon 1 of Lgmn (target with protospacer-associated motif underlined GGATGGAGGCAAG-CACTGGGTGG) and co-injected with polyadenylated Cas9 mRNA into C57BL/6J zygotes.Microinjected embryos were cultured overnight and introduced into pseudo-pregnant foster mothers.Pups were screened by PCR and Sanger sequencing of ear-punch DNA and a founder mouse was identified that carried a 10 bp frame-shift deletion in exon 1.The targeted allele was maintained by breeding on a C57BL/6J background.

Treatment of Mice with 25(OH)D 3 and Tissue Harvesting
Twelve-week-old female legumain wild-type (Lgmn +/+ ) and legumain-deficient (Lgmn −/− ) mice were bred and housed under standard conditions (21 • C, 55% relative humidity) on a 12 h light/dark cycle.The mice were injected subcutaneously (s.c.) on day 0, 2, 4, and 7 with 50 µg/kg 25(OH)D 3 (Vicotrat ® ) in 5% DMSO and 95% saline (n = 7) or an equal volume of vehicle (5% DMSO and 95% saline control, n = 7)).After the final injection, the mice were fasted overnight and anesthetized before blood was collected by retro-orbital bleeding and plasma was obtained after centrifugation and frozen at −80 • C. Subsequently, the mice were euthanized, and kidneys, liver, and spleen were collected, snap-frozen in liquid nitrogen, and stored at −80 • C. Mice experiments were carried out in accordance with permissions issued by the Danish Animal Experiments Inspectorate (2022-15-0201-01225). Tissue samples were homogenized in gentleMACS™ M Tubes (Miltenyi Biotec) using a gentleMACS™ Octo Dissociator (Miltenyi Biotec) in either TRI Reagent ® (Sigma) or lysis buffer for subsequent mRNA isolation or protein analysis, respectively.

Quantitative PCR
Total RNA was extracted from cell lysates harvested in Buffer RLT Pluss using an RNeasy ® Plus Kit according to the manufacturer's protocol or from tissue homogenates by chloroform phase separation and subsequent EconoSpin column purification (Epoch Life Science).RNA was quantified using Nanodrop™ (Thermo Scientific, Waltham, MA, USA) and stored at −80 • C until analysis.Complementary DNA (cDNA) was synthesized from 2 µg mRNA using the High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific) and ProFlex™ 3 × 32-well thermal cycler (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA, USA) and stored at −20 • C until analysis.Primers (Supplementary Table S1) were designed by the Primer Express software version 1 (Applied Biosystems, Thermo Fisher Scientific).Gene expressions were examined by real-time quantitative PCR (qPCR) using Power SYBR™ Green PCR Master Mix and the Applied Biosystems StepOnePlus™ Instrument with the accompanying software StepOne™ Version 2.3 (Applied Biosystems, Thermo Fisher Scientific).Gene expression was normalised against the geometric means of the CT values of housekeeping controls (RPLP0, GAPDH, β-actin, 18s) [30].
All vitamin D metabolite assays met the requirements specified by vitamin D external quality assessment (DEQAS) scheme (http://www.deqas.org/;accessed on 30 January 2023).The 25OHD 3 and 25OHD 2 assays showed <6% accuracy bias against the Center for Disease Control and Prevention (CDC) reference measurement (RMP) target values on the DEQAS scheme.

Statistical Analysis
The data are represented as mean ± SEM.Student t-test, Kruskal-Wallis, Mann-Whitney, simple linear regression, and one-way or two-way ANOVA were performed when appropriate.Statistical significance was considered at p < 0.05.All calculations were performed with GraphPad Prism (Version 9.0; GraphPad Software, Inc., San Diego, CA, USA).

1,25(OH) 2 D 3 Regulates Legumain Expression in Pre-Osteoblastic Cells
Given the role of VD 3 in regulating the expression of several bone-related factors, we first aimed to investigate if VD 3 could regulate the expression of the legumain encoding gene (LGMN).Analysis of the human LGMN gene promoter region using in silico analysis by the PROMO database revealed the presence of four potential vitamin D-responsive elements (VDRE) at the following nucleotide positions relative to the transcription start site: Cells 2024, 13, 36 6 of 16 nucleotide −638 (dissimilarity (ds) = 4.62%), −536 (ds = 8.08%), −474 (ds = 8.93%), and −402 (ds = 6.93%) (Figure 1A).This suggested a possible regulation of LGMN expression by VD 3 .To test this hypothesis in a cell-based model, the effect of VD 3 on legumain mRNA expression was investigated in osteogenic hBMSC cultures in the presence or absence of 1,25(OH) 2 D 3 (10, 50 or 100 nM).We found a dose-dependent increase in legumain mRNA expression, reaching significance at 100 nM 1,25(OH) 2 D 3 (Figure 1B).

1,25(OH)2D3 Regulates Legumain Expression in Pre-Osteoblastic Cells
Given the role of VD3 in regulating the expression of several bone-related factors, we first aimed to investigate if VD3 could regulate the expression of the legumain encoding gene (LGMN).Analysis of the human LGMN gene promoter region using in silico analysis by the PROMO database revealed the presence of four potential vitamin D-responsive elements (VDRE) at the following nucleotide positions relative to the transcription start site: nucleotide −638 (dissimilarity (ds) = 4.62%), −536 (ds = 8.08%), −474 (ds = 8.93%), and −402 (ds = 6.93%) (Figure 1A).This suggested a possible regulation of LGMN expression by VD3.To test this hypothesis in a cell-based model, the effect of VD3 on legumain mRNA expression was investigated in osteogenic hBMSC cultures in the presence or absence of 1,25(OH)2D3 (10, 50 or 100 nM).We found a dose-dependent increase in legumain mRNA expression, reaching significance at 100 nM 1,25(OH)2D3 (Figure 1B).To further investigate the effect of VD 3 on legumain expression and proteolytic activity, osteogenic hBMSC were cultured with or without 1,25(OH) 2 D 3 (10, 50, or 100 nM) or 25(OH)D 3 (100, 250, 500, or 1000 nM) for 7 days.Immunoblot analysis showed a dosedependent tendency of increased levels of 36 kDa mature legumain in the presence of 1,25(OH) 2 D 3 , reaching significance at 100 nM 1,25(OH) 2 D 3 (Figure 1C,D).However, the expression was not significantly affected by 25(OH)D 3 .The effect of VD 3 on legumain function was investigated by quantifying the proteolytic activity of legumain in the lysates.Increased legumain activity was observed in cells treated with 50 or 100 nM 1,25(OH) 2 D 3 and with 1000 nM 25(OH)D 3 (Figure 1E).As legumain can also be secreted and mediate autocrine/paracrine functions, we investigated whether VD 3 could alter legumain secretion.ELISA measurements of legumain in the conditioned media showed increased legumain secretion by pre-osteoblastic cells in the presence of 50 nM 1,25(OH) 2 D 3 (Figure 1F).

25(OH)D 3 Administration Increases Legumain Expression and Activity In Vivo
To investigate whether VD 3 also regulated the levels of legumain in vivo and whether legumain expression is important for vitamin D metabolism through VDBP processing (see below), high dose 25(OH)D 3 (50 µg/kg) or vehicle was subcutaneously (sc) administrated to wild-type (Lgmn +/+ ) and legumain-deficient (Lgmn −/− ) C57BL6/J mice for 8 days.Legumain deficiency in the kidneys, liver and spleen of Lgmn −/− mice was verified by immunoblotting and qPCR (Figure S1).In the wild-type mice, qRT-PCR analysis showed increased expression of legumain mRNA in the liver and spleen from the 25(OH)D 3treated compared to control mice (Figure 2A).In addition, immunoblot analysis showed a tendency towards increased level of 36 kDa mature legumain in the kidneys, liver, and spleen of 25(OH)D 3 -treated mice, although not statistically significant (Figure 2B,C).No prolegumain (56 kDa) was observed in these organs.Furthermore, an increased level of legumain proteolytic activity was detected in the kidneys of the 25(OH)D 3 -treated mice compared to control mice (Figure 2D).Importantly, ELISA measurement of legumain in the plasma revealed increased circulating legumain levels in mice treated with 25(OH)D 3 versus control (Figure 2E).Plasma levels of VD 3 metabolites in 25(OH)D 3 and vehicletreated mice were also measured and showed a positive correlation between the level of 1,25(OH) 2 D 3 and circulating legumain (Figure 2F).

Legumain Cleaves VDBP In Vitro and In Vivo
VDBP has previously been reported as a legumain substrate [22]; thus, we aimed to investigate the possible role of legumain in the regulation of VD 3 metabolism.First, we examined VDBP processing by legumain using incubation of purified VDBP from human plasma with or without purified active bovine legumain, followed by immunoblot analysis.Cleavage of full-length VDBP (55 kDa) by active legumain generated a VDBP cleavage product of approximately 45 kDa, which was not observed in the absence of legumain (Figure 3A).In addition, purified VDBP was incubated with or without lysate from legumain over-expressing HEK293 (M38L) cells [26], and a similar cleavage product (~45 kDa) was detected (Figure S2).
To further investigate the role of legumain in VDBP processing in vivo, the abovementioned wild-type (Lgmn +/+ ) and legumain-deficient (Lgmn −/− ) mice were treated (sc) with 25(OH)D 3 or vehicle for 8 days.Immunoblot analysis of VDBP in homogenates from the liver and kidney of Lgmn −/− mice did not show the generation of the 45 kDa VDBP cleavage product compared to the wild-type control (Lgmn +/+ ) mice (Figure 3B-D).Interestingly, significantly decreased expression of full-length VDBP was detected in the liver from Lgmn −/− compared to control mice, as observed by immunoblotting (Figure 3B,E), whereas legumain deficiency did not alter the levels of full-length VDBP in the kidneys (Figure 3B,F).No effect of 25(OH)D 3 treatment on VDBP levels or its processing was observed in kidneys or liver from either Lgmn +/+ or Lgmn −/− mice (Figure 3B-F).The level of VDBP in plasma was analyzed using ELISA and showed decreased circulating VDBP levels in Lgmn −/− compared to Lgmn +/+ mice (Figure 3G).In addition, qRT-PCR analy-sis revealed a significantly decreased level of VDBP mRNA expression in the liver from Lgmn −/− mice (Figure 3H).We also observed a tendency towards decreased levels of VDBP in plasma and VDBP mRNA expression in the liver from 25(OH)D 3 -treated wild-type mice (Figure 3G and Figure 3H, respectively).

Legumain Cleaves VDBP In Vitro and In Vivo
VDBP has previously been reported as a legumain substrate [22]; thus, we aimed to investigate the possible role of legumain in the regulation of VD3 metabolism.First, we examined VDBP processing by legumain using incubation of purified VDBP from human plasma with or without purified active bovine legumain, followed by immunoblo analysis.Cleavage of full-length VDBP (55 kDa) by active legumain generated a VDBP cleavage product of approximately 45 kDa, which was not observed in the absence of legumain (Figure 3A).In addition, purified VDBP was incubated with or without lysate from legumain over-expressing HEK293 (M38L) cells [26], and a similar cleavage product (~45 kDa) was detected (Figure S2).

Discussion
In the present study, VD 3 was identified as an inducer of legumain expression and proteolytic activity in pre-osteoblasts and mouse tissues.In addition, the cleavage of VDBP by legumain was, for the first time, demonstrated in vivo.Interestingly, legumain deficiency resulted in transcriptional downregulation of hepatic VDBP synthesis, resulting in reduced levels of circulating VDBP.Furthermore, legumain deficiency also altered VD 3 metabolism due to changes in the renal expression of key metabolic enzymes (CYP27B1 and CYP24A1), resulting in altered basal levels of VD 3 metabolites, as well as in response to 25(OH)D 3 treatment.
Initially, in silico studies indicated the presence of vitamin D-responsive elements (VDRE) in the promoter region of the legumain encoding gene (LGMN).Therefore, we hypothesized that VD 3 could be a regulator of legumain expression.Our cell-based studies using osteogenic cultures of human BMSC showed increased mRNA expression, proteolytic activity, and secretion of legumain by pre-osteoblasts in the presence of 1,25(OH) 2 D 3 .Although the promoter of the LGMN gene contains VDRE, it is most likely that the enhancing effect of VD 3 on legumain expression is mediated through an indirect mechanism, as a direct transcriptional regulation of legumain expression by VD 3 would likely result in a more pronounced effect.
We also observed increased legumain activity in the presence of the VD 3 metabolite 25(OH)D 3 .This was likely due to the conversion of 25(OH)D 3 to 1,25(OH) 2 D 3 by the pre-osteoblasts, as CYP27B1 is expressed and functional in these cells [33][34][35].In addition, administration of 25(OH)D 3 to wild-type mice increased the expression and activity of legumain in various tissues and, importantly, increased the circulating levels of legumain in the plasma.These data provide strong evidence that VD 3 is an upstream regulator of legumain expression.It has previously been shown that there is minimal overlap in genes regulated by VD 3 between different cell types or species [36].Interestingly, we observed that legumain was regulated in a similar manner in human pre-osteoblastic cells and mice.However, whether the functional and physiological consequences of VD 3 -induced production of legumain are conserved in mice and humans is currently not known.
Identification of VD 3 as an upstream regulator of legumain expression provides new insights into the role of VD 3 in modulating cellular processes beyond its well-known roles in, i.e., regulation of calcium homeostasis.The ability of VD 3 to regulate legumain expression suggests a possible involvement of VD 3 in legumain-mediated physiological and pathological processes.In this regard, and since legumain has an inhibitory role in osteoblast maturation [4], it is possible that legumain plays a role in the inhibition of osteoblast differentiation and reduction of bone mass associated with a high dose of VD 3 administration [37][38][39].
The present study provides evidence that VDBP is processed by legumain both in vitro and in vivo, corroborating a previous study presenting VDBP as a legumain substrate [22].We observed no VDBP processing in mouse kidneys or liver upon legumain deficiency, which could possibly lead to an increased level of VDBP in the circulation.However, interestingly, significantly lower plasma levels of VDBP were observed in Lgmn −/− compared to Lgmn +/+ mice.Renal dysfunction manifested as decreased glomerular filtration rate, increased plasma creatinine, and fibrosis, and premature senescence has been demonstrated in legumain-deficient mice [40,41].Whether the decrease in circulating VDBP levels in Lgmn −/− mice is caused or exacerbated by proteinuria is not known.However, the present data show a significant decrease in mRNA and protein expressions of VDBP in the liver upon legumain deficiency, which suggests a negative regulatory feedback loop that ensures decreased hepatic production of VDBP to counteract the systemic lack of VDBP processing by legumain.In addition, the observed increase in total VD 3 metabolite concentration in conjunction with decreased VDBP levels upon legumain deficiency indicates that proteinuria is not the cause of the reduced plasma VDBP level as the absolute majority of VD 3 metabolites are bound to VDBP and would be excreted along with the carrier protein [42,43].In a normal state, the plasma VDBP level is in a substantial surplus with regard to the VD 3 metabolite levels, and the binding capacity of VDBP far exceeds the level of available VD 3 metabolites [44][45][46].In addition, VD 3 metabolites are also bound to albumin, although to a lesser extent.Therefore, the observed increase in total VD 3 metabolite concentration seen in Lgmn −/− mice is not a contradiction to the decrease in plasma VDBP.
The mice used for in vivo experiments were kept on a regular diet (chow) with sufficient amounts of dietary VD 3 .Therefore, in order to provoke detectable changes in the levels of circulating VD 3 metabolites, high doses of parenteral 25(OH)D 3 were administered.However, the total exposure was within the range of what has previously been used in comparable experiments [47,48], and the detected levels of VD 3 metabolites were well below what has been considered toxic [48].Results in the present study showed increased plasma levels of total VD 3 and 25(OH)D 3 in Lgmn −/− compared to Lgmn +/+ mice, which could be explained by reduced tissue distribution of VD 3 or reduced clearance due to decreased levels of VDBP upon legumain deficiency.In addition, our data indicated a tendency towards increased plasma levels of 1,25(OH) 2 D 3 in Lgmn −/− mice, which could reflect increased total VD 3 and 25(OH)D 3 plasma levels upon legumain deficiency.However, the lack of major changes in the plasma levels of 1,25(OH) 2 D 3 upon legumain deficiency indicates the presence of legumain-independent mechanisms that could play a role in the release of VD 3 from VDBP in the kidneys, which is required for hydroxylation to the active 1,25(OH) 2 D 3 .
We observed a significantly decreased level of CYP24A1 mRNA expression in kidneys from Lgmn −/− mice.However, it is intriguing that the plasma level of 24,25(OH) 2 D 3 did not decrease in these mice.CYP27B1 is the key enzyme involved in the hydroxylation of 25(OH)D 3 and the production of active 1,25(OH) 2 D 3 .Expression of CYP27B1 mRNA in kidneys of Lgmn −/− mice was significantly decreased upon 25(OH)D 3 administration.Taking into account the increased plasma levels of total VD 3 and 25(OH)D 3 in Lgmn −/− mice, together with significantly decreased expression of the VD 3 catabolizing enzyme CYP24A1 upon legumain deficiency, decreased renal expression of CYP27B1 mRNA in Lgmn −/− mice could be a feedback mechanism to avoid high levels of 1,25(OH) 2 D 3 production and its associated side effects such as hypercalcemia [53].This is in line with a previous study indicating decreased renal expression of CYP27B1 in mice that are unable to catabolize VD 3 due to CYP24A1 deficiency [53].

Conclusions
Overall, the present work revealed the role of VD 3 as an upstream enhancer of legumain expression both in vitro and in vivo and that legumain plays a role in the regulation of VD 3 metabolism.This suggests a potential feedback loop where legumain activity can modulate the bioavailability of VD 3 and its metabolites and possibly its downstream physiological processes (Figure 5).These findings provide insight into the intricate relationship between VD 3 and legumain and can possibly open new avenues for research and investigation of novel therapeutic opportunities in various diseases in which VD 3 and legumain play crucial roles.
regulation of VD3 metabolism.This suggests a potential feedback loop where legumain activity can modulate the bioavailability of VD3 and its metabolites and possibly its downstream physiological processes (Figure 5).These findings provide insight into the intricate relationship between VD3 and legumain and can possibly open new avenues for research and investigation of novel therapeutic opportunities in various diseases in which VD3 and legumain play crucial roles.
As 24,25(OH) 2 D 3 is generated by CYP24A1-mediated hydroxylation of 25(OH)D 3 , the lack of change in the plasma levels of 24,25(OH) 2 D 3 in Lgmn −/− mice could be due to the increased plasma levels of 25(OH)D 3 upon legumain deficiency, together with CYP24A1-mediated hydroxylation of 25(OH)D 3 in extra-renal vitamin D-targeted tissues.This notion is supported by studies indicating that extra-renal CYP enzymes are involved in the regulation of VD 3 metabolism [49-51] and the increase in hepatic CYP24A1 mRNA expression in 25(OH)D 3 -treated Lgmn −/− mice.It has recently been shown that extra-renal CYP24A1 ameliorates severe hypercalcemia in mice with kidney-specific CYP24A1 ablation [52].

Figure 5 .
Figure 5. Graphical representation of the suggested interplay between vitamin D and legumain.Left panel: Vitamin D (VD3) promotes legumain expression and activity through transcriptional upregulation of the legumain gene (LGMN).The free fraction of circulating VD3 metabolites diffuse through plasma membranes.25-hydroxyvitamin D (25(OH)D3) is hydroxylated by 1α-hydroxylase (CYP27B1), forming the active metabolite 1α,25-dihydroxyvitamin D (1,25(OH)2D3).1,25(OH)2D3 binds to the nuclear vitamin D receptor (VDR) and promotes transcription of legumain (LGMN).Synthesized prolegumain is either sorted and activated in the endolysosomal system or released to the extracellular environment.Right panel: In the proximal tubular epithelium, 25(OH)D3 bound to vitamin D binding protein (VDBP) is internalized from the tubular lumen through a megalin/cubilin-mediated process.The vitamin D metabolite is released, enabling subsequent hydroxylation by 1α-hydroxylase (CYP27B1) or 24-hydroxylase (CYP24A1), and VDBP is cleaved by legumain in the endolysosomal system.VDBP cleavage by legumain might be important in