LanCL proteins are not Involved in Lanthionine Synthesis in Mammals

LanC-like (LanCL) proteins are mammalian homologs of bacterial LanC enzymes, which catalyze the addition of the thiol of Cys to dehydrated Ser residues during the biosynthesis of lanthipeptides, a class of natural products formed by post-translational modification of precursor peptides. The functions of LanCL proteins are currently unclear. A recent proposal suggested that LanCL1 catalyzes the addition of the Cys of glutathione to protein- or peptide-bound dehydroalanine (Dha) to form lanthionine, analogous to the reaction catalyzed by LanC in bacteria. Lanthionine has been detected in human brain as the downstream metabolite lanthionine ketimine (LK), which has been shown to have neuroprotective effects. In this study, we tested the proposal that LanCL1 is involved in lanthionine biosynthesis by constructing LanCL1 knock-out mice and measuring LK concentrations in their brains using a mass spectrometric detection method developed for this purpose. To investigate whether other LanCL proteins (LanCL2/3) may confer a compensatory effect, triple knock-out (TKO) mice were also generated and tested. Very similar concentrations of LK (0.5–2.5 nmol/g tissue) were found in LanCL1 knock-out, TKO and wild type (WT) mouse brains, suggesting that LanCL proteins are not involved in lanthionine biosynthesis.

Scientific RepoRts | 7:40980 | DOI: 10.1038/srep40980 with LK in the mammalian brain 24 , and more recently, a cell-permeable ester derivative of LK (LKE) was shown to have anti-inflammatory and anti-apoptotic function in NSC-34 motor neuron-like cells 24 . A neuritogenesis effect of LKE was demonstrated in primary chick dorsal root ganglia culture, and LKE treatment delayed the onset of clinical paralysis and increased the survival of SOD1 G93A mutant mice 24 . Moreover, in a mouse model of Alzheimer's disease, administration of LKE also substantially mitigated cognitive decline 25 , and LKE was found to protect neuronal cells in a mouse model of cerebral ischemia 26 . These in vivo results suggest neuroprotective and neuritogenic activity of LK.
Linking the relative abundance of LanCL1 in brain, the presence of LK in brain, and the observed interaction of LK and LanCL1, Hensley et al. proposed that lanthionine or LK could be an allosteric regulator of LanCL1 or the product of a LanCL1-catalyzed reaction 24 . Since LanCL1 binds to the Cys of glutathione 3 , and because bacterial LanC catalyzes addition of Cys to dehydrated Ser residues in peptides, it was suggested that glutathione could be one of the substrates of LanCL1 during lanthionine formation ( Fig. 1) 24 . Indeed, precedent in Nature for amino acid crosslinking in a protein followed by proteolysis to generate a bioactive small molecule is found in the biosynthesis of thyroid hormone 27 .
In this study, we tested the hypothesis that LanCL1 may be involved in LK biosynthesis by creating a LanCL1 knockout (KO) mouse line and quantifying LK in the brain. Concentrations of about 0.5-2.5 nmol/g tissue were detected in both WT and LanCL1 KO mouse brains, suggesting LanCL1 may not play a major role in lanthionine synthesis in vivo. To rule out a potential compensatory effect from other LanCL proteins, triple knock-out (TKO) mice were also generated. The LK levels in WT and TKO mice were similar, confirming that despite the chemically appealing hypothesis, LanCL proteins are likely not involved in LK synthesis.

Results
Generation of Lancl1, Lancl2 and Lancl3 knock-out mice. Zinc finger nuclease (ZFN) mRNA designed to target the third exon of the LanCL1 gene was injected into FVB embryos at the pronuclear stage followed by transfer into pseudo pregnant female FVB mice, resulting in seventy three founder mice. Genotyping of tail DNAs by PCR-based Surveyor assay revealed that eight founder mice harbored mutations at the expected ZFN cleavage site. The PCR products were subcloned and sequenced to characterize the mutations. Notably, each PCR product contained a mixture of wild-type and mutant sequences, suggesting that all founder mice were mosaic. All eight genomic DNAs were confirmed to contain deletions of 1-nucleotide to 48-nucleotides at the expected ZFN site. We chose a founder mouse with a 19-nucleotide deletion (Fig. 2a) to breed with wild-type FVB mice for establishing a germline-transmitted mutant mouse line and further characterization. Heterozygous and homozygous mutant mice were identified by PCR genotyping (Supplementary Fig. S1a), and a Mendelian ratio was observed for both genotypes. The 19-nucleotide deletion in exon 3 would result in a frame shift in the remaining coding sequence, and the resulting premature stop codon in the mRNA would lead to nonsense-mediated mRNA decay and hence knockout of the gene product. Tissues were harvested from young adult mice, homogenized, and subjected to western blot analysis. Complete loss of the LanCL1 protein was confirmed in the brain, heart, and liver of homozygous mutant mice ( Supplementary Fig. S1e). From here on, we refer to the homozygous mutant mice as Lancl1− /− .  24 . LanCL1 may catalyze the conjugation of GSH to protein-bound Dha, followed by peptidase-mediated digestion of the reaction product (proteolysis sites indicated by curled lines) to form lanthionine. Transamination to form lanthionine ketimine has been documented previously. An enamine tautomer can also be formed (not shown).
Similar procedures were performed to generate Lancl2− /− and Lancl3− /− mice. The resulting LanCL2 KO mouse line bears a 2-nt deletion at the ZFN target site in exon four and the LanCL3 KO mouse line has a 37-nt deletion in exon one (Fig. 2a). The homozygous mutant mice were identified by genotyping ( Supplementary Fig. S1b,c), and complete loss of the target proteins was confirmed by western blotting (Fig. 2b-d Generation and validation of LanCL triple knock-out (KO) mice. Double KO mouse lines were generated by crossing single KO mice. After three rounds of breeding, all three double KO mouse lines were obtained. We then crossed double KO mouse lines to generate LanCL1/2/3 triple KO mice. The complete knock-out of all three Lancl genes was verified by genotyping ( Supplementary Fig. S1d). Brain tissue was collected, homogenized and subjected to western blotting. The absence of all three LanCL proteins confirmed the successful generation of the TKO mouse line (Fig. 2e). Again, no gross abnormality was observed in the TKO mice.

PITC detection of LK in WT and LanCL1 KO mouse brains.
Cavallini and colleagues reported a method for detection and quantification of LK, which involves reaction with phenylisothiocyanate (PITC) and subsequent analysis of the absorbance of the product (PTH-LK) at 380 nm 28,29 . Using this method, LK has been shown to be present in human and bovine brain at a concentration of 0.5-1 nmol/g tissue 30,31 , as well as in human urine samples with a relative concentration of 3-140 μ g/g creatinine 28 . We initially used this method of LK detection. First, a synthetic standard was derivatized with PITC and the product analyzed by HPLC with detection at 380 nm to determine the elution time (Fig. 3a). Then WT mouse brain homogenate was treated with the derivatization reagents, and the resulting sample was analyzed under the same conditions (Fig. 3b). Standard synthetic LK was subsequently spiked into the mouse brain homogenates before sample preparation. The peak with a retention time of 25.8 min increased in intensity, indicating that it corresponds to LK (Fig. 3c). As a negative control, brain sample that was not treated with the derivatization reagents showed no corresponding peak. To further characterize the material giving rise to the peak, the fraction eluting at 25 min was subjected to liquid chromatography (LC) coupled to tandem mass spectrometry (MS/MS) analysis using multiple reaction monitoring (MRM), which confirmed that the peak corresponded to derivatized LK ( Supplementary Fig. S2d). Using the optimized PITC method, brain LK levels in thirteen WT mice were determined as shown in Fig. 3d. We observed rather small HPLC absorbance values at 380 nm, indicating small amounts of LK in WT mouse brains.

LC/MS/MS detection of lanthionine metabolites in WT and LanCL1 KO mice.
Because the detected amounts of PTH-LK in WT mouse brain approached the detection limit of HPLC-UV (Fig. 3b), we investigated mass spectrometry based detection without derivatization. To detect LK with higher sensitivity and specificity, a triple quadrupole mass spectrometer coupled with HPLC was employed using MRM looking at specific mass transitions in tandem electrospray ionization MS. Unlike the PITC method, which only detected LK, MS enabled the detection of all related metabolites. To obtain a more comprehensive assessment of the in vivo lanthionine level, specific methods were developed to directly detect lanthionine and LK. The possible sodium adduct of LK (LK-Na) was also examined ( Supplementary Fig. S2), and a specific fragmentation transition of 190.1 Da → 144.1 Da was chosen to detect and quantify LK. Figure 4a shows a typical MRM chromatogram of WT mouse brain, illustrating that the signal-to-noise of LK detection in WT mouse brain using this method was high. Free lanthionine and LK-Na were also observed in mouse brain sample but with much lower signal intensity compared to standards ( Supplementary Fig. S2b,c), suggesting they are a minor contribution to lanthionine species. Because of the higher levels of LK and much better resolution, we focused on quantifying the LK levels in mouse brain.
Quantification of LK levels in mouse brain with isotopically-labeled LK. The LC/MS/MS method provided more straightforward detection of LK and related metabolites than the PTH-LK method and is less likely to be affected by other metabolites because no derivatization is needed. To further improve quantification, an isotopically labeled synthetic internal standard was employed. The synthetic isotope-labeled LK was prepared from uniformly 13 (Fig. 4). Mouse brain to which isotopically-labeled LK was not added served as a negative control; no signal was observed for the 194.1 Da → 148.0 Da transition indicating that endogenous metabolites do not coincidentally fragment to give the same transition. Next, we spiked in known concentrations of isotope-labeled LK and measured the concentrations using the indicated transition with synthetic standards to generate a standard curve. Endogenous LK concentrations were calculated by comparing peak areas with isotope labeled LK of known concentration added to the brain samples. In order to achieve a higher intensity of peak signal, two or three mouse brains were pooled. The same method was applied to TKO mouse brains. We detected LK concentrations in brains of four, two and one month old mice and the results are summarized in Fig. 5. Compared with bovine brain, mouse brain has a slightly higher concentration of LK at 0.5-2.5 nmol/g tissue. More importantly, no statistically significant difference was observed between WT, LanCL1 KO, and TKO mouse brains of all ages examined.

Discussion
The biogenesis of lanthionine in mammals is elusive. Lanthionine has long been thought to be a by-product from the transsulfuration pathway. The pyridoxal phosphate (PLP)-dependent enzyme cystathionine β -synthase (Cβ S) catalyzes the first step in the transsulfuration pathway, condensing homocysteine with serine or cysteine to form cystathionine, a bis-amino acid detected in the same organs as lanthionine 28,31 . Cystathionine γ -lyase (CSE) catalyzes the second step in the pathway, the α ,γ -elimination of cystathionine to give cysteine, α -ketobutyrate and ammonia ( Supplementary Fig. S3a). It was reported that Cβ S and CSE can also catalyze condensation of cysteine with either serine or another cysteine to form lanthionine ( Supplementary Fig. S3b). This activity was observed in vitro as a minor reaction 32,33 , and evidence of Cβ S or CSE being the biosynthetic enzymes for lanthionine in vivo is missing. The discovery of mammalian homologs of bacterial lanthionine forming enzymes led Hensley and colleagues to propose that they may be involved in lanthionine formation in mammalian cells 24 . LanCL1 was a particularly attractive target for lanthionine formation in brain because it binds LK, binds glutathione, is ubiquitously present in brain, and appears to perform an antioxidative, neuroprotective function in brain, similar to the activities displayed by LKE.
A seemingly missing piece in this hypothesis is the mechanism by which the Dha would be formed. The bacterial lanthipeptide enzymes that dehydrate Ser to Dha do not have mammalian sequence homologs, but alternative routes of generating Dha have been reported 34 . Therefore, Hensley and coworkers proposed that LanCL1 may catalyze the addition of the Cys in glutathione to a protein-or peptide-bound Dha, followed by trimming of the product by peptidases to form lanthionine (Fig. 1). In support of the model, the adduct of GSH to dehydroalanine (glutathionyl lanthionine, gLan) was synthesized and shown to have biological effects similar to that of LK in ALS mouse models 24 . Intriguingly, a very recent crystal structure of one of the bacterial enzymes that dehydrates Ser shows that structurally they strongly resemble mammalian lipid kinase-like proteins 35 , suggesting that perhaps enzymatic formation of Dha in mammals is indeed feasible. We therefore set out to test the involvement of LanCL1 in lanthionine formation.
LK has been previously detected in human and bovine brain by PITC derivatization and subsequent HPLC analysis relying on the absorbance of the PTH-LK derivative 22,23 . Here we report the detection of LK in mouse brain using both PITC derivatization and a new LC/MS/MS method. A peak corresponding to PTH-LK was detected in WT and LanCL1 KO mouse brain with low intensity that approached our detection limit. Therefore, an LC/MS/MS method was developed in which LK was directly detected without a derivatization step. Use of isotope labeled LK as internal standard, which is expected to undergo the same potential losses during sample preparations, was employed to further reduce quantification error. The concentration of LK in mouse brain determined by the LC/MS/MS method was 0.5-2.5 nmol/g tissue, which is somewhat higher than that found in bovine brain by the PITC method. The LK levels in four, two and one month old mouse brains showed no significant differences, indicating the production of LK is not strongly age dependent.
LK was detected in both WT and LanCL1 KO mouse brains by both methods at similar concentrations, suggesting that LanCL1 is not required for LK biogenesis. Furthermore, the similar concentrations of LK in TKO mouse brains ruled out a compensatory effect and more importantly, suggested that the three LanCL proteins are not involved in LK synthesis in vivo. Whether LK found in mammals is a by-product of Cβ S or CSE or whether there exists another biosynthetic route is currently not clear. We can also not rule out that LK is derived from commensal bacteria, many of which produce lanthipeptides. We note that our results do not exclude the possibility of other types of functional correlation between LanCL1 and LK, such as the aforementioned allosteric effector model. A systematic screening for potential LanCL1 co-substrates is needed to define its function in neuroprotection.

Methods
Antibodies and Chemicals. Anti-LanCL1 antibody was purchased from Bethyl Laboratories (Montgomery, TX). Anti-LanCL2 antibody was generated by Proteintech Group (Chicago, IL) using full length recombinant LanCL2 as the antigen as previously described 15  PITC derivatization. Mouse brain sample preparation and PITC derivatization were performed essentially as described previously with minor modifications 28 . Freshly harvested mouse brain (about 0.5 g of either WT or KO) was homogenized four times with an OMNI TH homogenizer in 5 mL of 30% acetonitrile/water using 30 s pulse and pause cycles. The homogenate was deproteinated by adjusting the acetonitrile/water ratio to 2:1 and centrifuged at 23,700 × g for 10 min. The supernatant was concentrated under N 2 flow to 1 mL before PITC derivatization. One mL of concentrated brain homogenate was mixed with 90 μ L of PITC and 3 mL of coupling buffer (acetonitrile:pyridine: triethylamine:H 2 O = 10:5:2:3) in a 5 mL reaction vial and stirred for 30 min at 20 °C. The solution was then dried at 40 °C using a rotary evaporator and re-dissolved in 1 mL of 10 mM potassium acetate buffer (pH = 8.0). Standard LK (100 ng) was dissolved in 30% acetonitrile/water, followed by addition of 1 mL of coupling buffer and 30 μ L of PITC. Reaction conditions and work-up were identical to those described for the brain sample.
Mouse husbandry and microinjection. All animal experiments in this study followed protocols approved by the Animal Care and Use Committee at the University of Illinois at Urbana-Champaign. FVB mice were maintained on a 12 h/12 h light/dark cycle with access to water and food. Microinjections of ZFN mRNAs (Sigma) were performed in the Transgenic Mouse Facility in the Roy J. Carver Biotechnology Center at the University of Illinois at Urbana-Champaign. LanCL1, LanCL2, and LanCL3 ZFN mRNAs were diluted to 2.5-4 ng/μ L with 10 mM Tris buffer, 0.1 mM EDTA (pH 7.5) and injected in FVB embryos at the pronuclear stage before transferring to pseudo pregnant females.
Genomic DNA isolation from mouse tail tips. For PCR amplification followed by Surveyor mismatch endonuclease assay 37 , mouse tail tips (2-5 mm) were dissolved by incubating with 600 μ L of extraction buffer (20 mM Tris pH 7.5-8.0, 50 mM EDTA, 100 mM NaCl, 0.5% SDS and 500 μ g/mL proteinase K) for 2-3 h at 55 °C. Then 240 μ L of high salt solution (4.21 M NaCl, 0.63 M KCl, 10 mM Tris pH 8.0) was added to the dissolved tail and the sample was incubated for 30 min at 4 °C to precipitate proteins. The precipitated proteins were removed by centrifugation at 16,100 × g for 10 min at 4 °C and the supernatant was transferred to a new tube. Genomic DNA was precipitated by adding 2x volume of ethanol followed by centrifugation at 16,100 × g for 10 min at 4 °C. The precipitated DNA pellet was washed once with 80% ethanol and dissolved in 200-400 μ L of TE buffer (1 mM Tris pH 8.0, 0.1 mM EDTA) after a brief air dry. For PCR amplification followed by agarose gel examination, genomic DNA was extracted from mouse tail tips using KAPA Express Extract Kits following the manufacturer's instructions.
PCR for Surveyor assays. The  Founder identification using Surveyor mismatch endonuclease assay. Genomic DNA was extracted from mouse tail tips cut from three week old pups and a 400 bp fragment surrounding the ZFN cutting site was amplified by primer sets ZFN_F and ZFN_R. The PCR product (≥ 50 ng/μ L) was denatured and hybridized using the following program: 95 °C for 10 min; 95 °C to 85 °C, − 2 °C/s; 85 °C to 25 °C, − 0.1 °C/s; 4 °C, indefinitely. The rehybridized PCR product (about 15 μ L) was incubated with 1 μ L of enhancer S and 1 μ L of Nuclease S (Integrated DNA Technologies) for 1 h at 42 °C. The cleaved products were resolved by 3% NuSieve DNA agarose gel electrophoresis. RNA and protein extraction from mouse tissues. Mouse brain, heart and liver were dissected and snap frozen immediately in liquid nitrogen. Frozen tissues were homogenized by grinding with a pestle and mortal. For protein extraction, the homogenized powder was lysed in tissue extraction buffer (200 nM Tris HCl pH 7.4, 150 mM NaCl, 1 mM EGTA, 1 mM EDTA, 2.5 mM sodium pyrophosphate, 1 mM β -glycerol phosphate, 1 mM sodium orthovanadate, 2% Triton X-100 and a protease inhibitor cocktail) and rotated for 30 min at 4 °C. After centrifugation at 16,100 × g for 10 min, the supernatant was transferred to a new tube and boiled with 1x volume of 2x Laemmili sample buffer for SDS-PAGE and Western blotting. RNA was extracted from the homogenized powder using the RNeasy Kit (QIAGEN) following the manufacturer's instructions.

LC/MS/MS of mouse brain samples.
Freshly harvested mouse brain (about 0.5 g of either WT or KO) was homogenized four times with an OMNI TH homogenizer in 5 mL of 30% acetonitrile/water using 30 s pulse and pause cycles. Homogenate was separated into two aliquots. To one aliquot, isotopically labeled LK standard (100 ng) was added. The other aliquot was processed without addition of labeled LK. The two aliquots were deproteinated by addition of acetonitrile to adjust the acetonitrile/water ratio to 2:1 and centrifuged at 23,700 × g for 10 min. The supernatant was collected and dried under N 2 flow. Dried sample was resuspended in 400 μ L of 30% acetonitrile/water followed by centrifugation at 16,100 × g for 5 min. The supernatants were then collected for LC/MS/MS. Samples were analyzed with a 5500 QTRAP LC/MS/MS system (AB Sciex, Foster City, CA) in the Metabolomics Laboratory of the Roy J. Carver Biotechnology Center, University of Illinois at Urbana-Champaign. The 1200 series HPLC system (Agilent Technologies, Santa Clara, CA) includes a degasser, an autosampler, and a binary pump. The LC separation was performed on an Agilent Zorbax SB-Aq column (4.6 × 50 mm, 3.5 μ m), with a gradient from 100% A (0.1% formic acid in water) to 99% B (0.1% formic acid in acetonitrile) in 6 min at a flow rate of 0.45 mL/min. The autosampler was set at 5 °C. The injection volume was 5 μ L. Positive and negative ion mass spectra were acquired under electrospray ionization (ESI) with the ion spray voltage at 5500 V and − 4500 V, respectively. The source temperature was 500 °C. The curtain gas, ion source gas 1, and ion source gas 2 were 33 psi, 65 psi, and 50 psi, respectively. Multiple reaction monitoring (MRM) was used to measure related metabolites.
HPLC. Analytical reversed-phase high-performance liquid chromatography (RP-HPLC) was performed on an Agilent 1260 Infinity System with a Hypersil Gold C18 column (250 mm × 4.6 mm, particle size 5 μ ). The program ran from 98% buffer A (0.05 M ammonium acetate, pH 6.5) to 60% buffer B (acetonitrile: H 2 O = 7:3) in 30 min and then to 100% buffer B in 5 min at a flow rate of 1 mL/min. All HPLC solvents were filtered with a Millipore filtration system equipped with a 0.22 μ m PVDF membrane filter prior to use. Statistical analysis. All data analysis was conducted using GraphPad Prism software V6.0 (GraphPad, San Diego, CA, USA). Comparisons of WT, LanCL1KO and TKO LK levels were made using one-way analysis of variance (ANOVA), assuming Gaussian distribution. Multiple comparisons were also conducted in conjunction with ANOVA by comparing the mean of one group with other two groups using Turkey's test.