The disease-linked R336C mutation in cystathionine β-synthase disrupts communication with the PLP cofactor, yet maintains the enzyme's overall structural integrity

The disease-linked R336C mutation in cystathionine bsynthase disrupts communication with the PLP cofactor, yet maintains the enzyme’s overall structural integrity Carolina Conter 1 , Reyes N u~ nez-Franco 1 , Duaa Walid Al-Sadeq 2,3 , Carmen Fern andez-Rodr ıguez 1 , Naroa Goikoetxea-Usandizaga 1,4 , Gheyath K. Nasrallah 3,5 , Michail Nomikos 2 , Maria Luz Martinez-Chantar 1,4 , Alessandra Astegno 6 , Gonzalo Jim enez-Os es 1,7 and Luis Alfonso Mart ınez-Cruz 1 1 Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), Derio, Spain 2 College of Medicine, QU Health, Qatar University, Doha, Qatar 3 Biomedical Research Center, Qatar University, Doha, Qatar 4 Centro de Investigaci on Biom edica en Red de Enfermedades Hep aticas y Digestivas (CIBERehd), Madrid, Spain 5 Department of Biomedical Science, College of Health Sciences, QU Health, Qatar University, Doha, Qatar 6 Department of Biotechnology, University of Verona, Italy 7 Ikerbasque, Basque Foundation for Science, Bilbao, Spain Keywords crystallography; cystathionine b-synthase; homocystinuria; hydrogen sulfide; molecular dynamics; pyridoxal-50-phosphate Correspondence C. Conter, G. Jim enez-Os es and L. A. Mart  ınez-Cruz, Center for Cooperative Research in Biosciences (CIC bioGUNE), Basque Research and Technology Alliance (BRTA), Bizkaia Technology Park, Building 801A, Derio 48160, Spain Tel: +34 944061304 (C.C.); +34 946572537 (G.J.-O.); +34 944061318 (L.A.M.-C.) E-mail: [email protected];gjoses@ cicbiogune.es;[email protected] Carolina Conter and Reyes N u~ nez-Franco contributed equally to this article. (Received 2 December 2024, revised 23 February 2025, accepted 16 April 2025) doi:10.1111/febs.70116 Cystathionine b-synthase (CBS) is a pyridoxal-phosphate (PLP)-dependent enzyme essential for the reverse transsulfuration pathway, where homocysteine and serine combine to form cystathionine, the immediate precursor of cysteine. Mutations in the CBS gene cause homocystinuria, a disorder associated with intellectual disability, multisystem complications, and reduced life expectancy. The CBS p.R336C mutation, linked to severe pyridoxine non-responsiveness, results in reduced enzyme activity, previously attributed to protein instability and weakened substrate and PLP binding. To clarify the effects of the pathological R336C mutation, we performed biochemical, biophysical, and crystallographic analyses, as well as molecular dynamics simulations. Our findings show that the R336C mutation minimally impacts the structural environment around residue 336, does not cause enzyme misfolding, and does not impair the binding of PLP or the allosteric activator S-adenosylmethionine (AdoMet) binding. Instead, the mutation induces subtle reorientations in nearby hydrophobic residues, including F185 and Y381, altering intramolecular contacts that perturb the interaction between asparagine 149 and the O3 oxygen of PLP. This alteration is known to potentially shift the tautomeric equilibrium of the PLP Schiff base from its catalytically active ketoenamine form to the inactive enolimine form, which aligns with the reduced activity of the R336C variant. Additionally, the R336C mutation disrupts intermolecular contacts between the catalytic core and Bateman module, altering the Bateman module’s intrinsic mobility in the enzyme’s basal state and potentially Abbreviations AdoMet, S-adenosylmethionine; CBL, cystathionine b-lyase; CBS, cystathionine b-synthase; CD, circular dichroism; CNNM, cyclin M magnesium transporter; Cth, cystathionine; Cys, cysteine; H 2 S, hydrogen sulfide; HCU, homocystinuria; Hcys, homocysteine; HsCBS, human cystathionine b-synthase; ITC, isothermal titration calorimetry; k cat , turnover number (catalytic rate constant); k cat /K m , catalytic efficiency; K d , dissociation constant; K m , Michaelis constant; LDH, lactate dehydrogenase; MD, molecular dynamics; NADH, nicotinamide adenine dinucleotide (reduced form); PCA, principal component analysis; PDB, Protein Data Bank; PEG, polyethylene glycol; PLP, pyridoxal-phosphate; SEC, size-exclusion chromatography; Ser, serine; SYNROT, synchronized rotational shift; T m , melting temperature; TRANS, transversal displacement; TDS, entropy change; DH, enthalpy change. 4933The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. affecting the cavity opening required for catalysis. Importantly, the R336C variant retains the native enzyme’s ability to assemble into polymeric chains in crystals, preserving its filament formation capacity. Introduction Cystathionine b-synthase (CBS) (EC4.2.1.22) is a crucial cytosolic enzyme that initiates the reverse transsulfuration pathway by catalyzing the b-replacement of serine (Ser) and homocysteine (Hcys) to form cystathionine (Cth), thereby removing toxic homocysteine [1–5]. Beyond this primary role, CBS is also engaged in alternative reactions that produce hydrogen sulfide (H 2 S), a gaseous signaling molecule with relevant biological functions [6,7]. Mutations in the CBS gene result in classical homocystinuria (HCU), (MIM #236200) [8], an autosomal recessive disorder of methionine metabolism characterized by elevated levels of homocysteine in the blood and urine, increased blood methionine, and reduced plasma cystathionine. If untreated, HCU progresses to cardiovascular, skeletal, and visual impairments, along with neurological complications resembling Alzheimer’s disease [2,9–13]. HCU has an estimated global incidence of 1/344 000 [14,15], with varying rates reported across regions, such as 1/900 000 in Japan [16], or 1.6/1 000 000 in Spain [17]. Extensive studies conducted thus far have facilitated the identification of numerous alleles in both homozygous and heterozygous HCU patients, resulting in approximately 200 annotated pathogenic variants [18,19]. The majority of them, approximately 87%, are missense variants that are thought to impact CBS mRNA stability, protein folding, or biochemical function [20–28]. The remaining mutations are nonsense, deletions and insertions [29,30]. In some cases, like p.I278T, the mutations are widely distributed across populations [20,29,30], while in others they are confined to a specific population. For example, the p.G307S mutation accounts for 71% of the mutant alleles in Ireland [31], whereas p.T191M represents between 40% and 75% in the Iberian Peninsula and Colombia [32]. A particularly interesting case is p.R336C, with the highest incidence in Qatar (1/1800 births) [33]. Over the years, several hypotheses have been proposed to explain the pathogenic effect of the R336C mutation. The proposals range from a change in the local environment of the mutated residue that impairs the ability of the protein to bind substrates (Hcys and Ser), pyridoxal 50-phosphate (PLP) cofactor, and the allosteric activator S-adenosylmethionine (AdoMet) [34], to a misfolding of the polypeptide chain that promotes its ubiquitination and degradation by the proteasome [34]. More recent studies conducted on recombinant CBS contradict some of these statements and suggest that mutations do not necessarily alter the protein’s secondary structure, although they may affect other parameters such as its thermal stability [35]. However, without the three-dimensional structure of this variant to confirm or refute these hypotheses, the effect of the mutation remains disputed. Structurally, human CBS (HsCBS) exhibits a modular domain arrangement and a complex regulatory mechanism [36]. Each polypeptide chain consists of three functional segments. The N-terminal domain binds a heme cofactor, which serves both structural and regulatory roles [20,37–39]. The central catalytic domain of the protein contains PLP, the active form of vitamin B6, covalently linked as a Schiff base to the e-amino group of lysine 119 (K119) (internal aldimine). Additionally, the C-terminal domain, referred to as the ‘Bateman module’, consists of a tandem of CBS motifs and functions as an allosteric activator when it binds AdoMet (Fig. 1). This binding counteracts the influence of the C-terminal autoinhibitory domain [40–43]. Understanding the impact of point mutations on enzyme function requires in-depth knowledge of its catalytic activity and activation mechanism. The catalytic activity has been extensively studied in homologous enzymes from various organisms [44–48]. These studies have capitalized on the remarkable evolutionary conservation of the enzyme’s catalytic site. The catalytic reaction occurs in two distinct sequential steps following a ping-pong mechanism and requires the presence of the PLP cofactor within the catalytic cavity, which remains covalently bound to a conserved lysine (K119 in humans). The entry of the first substrate (Ser or cysteine (Cys)) into the cavity leads to the formation of an external aldimine with the PLP. Concurrently, a displacement of several structural elements of the catalytic domain (collectively referred to as the ‘mobile subdomain’ [43]) occurs, facilitating the interaction of the substrate with amino acids (e.g., 4934 The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. Altered PLP tautomerism impairs HsCBS-R336C activity C. Conter et al. 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License S147 in humans) involved in the progression of the external aldimine toward an aminoacrylate intermediate. This intermediate reacts in the second step with the second substrate (Hcys) to ultimately generate the products (Cth and H 2 O (or H 2 S if the first substrate was Cys)). In contrast to the high evolutionary conservation exhibited by CBS across its various catalytic reactions, the regulation of the enzyme differs significantly among organisms. The best-studied case is the human enzyme, and involves a conformational change from a basal state—where the Bateman module compresses Fig. 1. 3D-Structure of HsCBS. (Top) Schematic representation of the HsCBS domain organization. The heme-binding site is shown in red, the catalytic core (resembling the PLP-fold type II family) is highlighted in yellow, and the Bateman module is depicted in blue. (Middle) HsCBS monomer in its basal (left) and activated (right) AdoMet-bound states, represented with PYMOL (The PyMOL Molecular Graphics System, version 2.2.3, Schr€ odinger, LLC). (Bottom) Dimeric HsCBS. The location of the different mutations is represented with spheres. S2 indicates the location of the AdoMet-binding cavity in the Bateman module. The catalytic core, the interdomain linker and the Bateman module, are colored in yellow, green and blue, respectively. The PLP, HEME and AdoMet molecules are shown in sticks. 4935The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. C. Conter et al. Altered PLP tautomerism impairs HsCBS-R336C activity 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License the mobile subdomain of the catalytic core, blocking access to the catalytic cavity [49]—to an activated state, in which the occlusive effect exerted by the Bateman module is released upon binding AdoMet [50]. The binding of AdoMet to the Bateman module occurs at the designated S2 site, disrupting some of the preexisting interactions among the amino acids in the beta sheets of the two CBS domains. The stabilization of the AdoMet molecule in the S2 cavity induces a relative rotation of the CBS motifs, which sterically affects the a-helices of the CBS2 motif, situated just above the mobile subdomain of the catalytic core. This steric hindrance is subsequently relieved through a displacement of the Bateman module, which removes the occlusive effect that blocks access to the catalytic cavity. The precise pathway by which the Bateman module reaches its final activated state remains unknown. However, it ultimately leads to the association of this module with the corresponding region of the complementary subunit that forms each dimer (Fig. 1). The interaction between complementary Bateman modules is facilitated by the adequate length of the interdomain linker that connects them to their respective catalytic cores resulting in the formation of a disc-shaped assembly known as the ‘CBS module’. This assembly remains stable as long as AdoMet is bound to the corresponding S2 cavities. The CBS module is structurally positioned above the mobile subdomains of the complementary catalytic cores, interacting with them solely through the interdomain linker that connects them to the catalytic core (Fig. 1). In this context, the pressure that obstructed substrate access and kept the enzyme in a basal activity state is relieved, enabling the enzyme to attain its ‘activated state’ [49,50]. Recent findings from cryo-electron microscopy have revealed the filamentous nature of the human CBS enzyme, along with the drastic structural changes that occur in the CBS polymer during the transition from the basal conformation to the activated state upon AdoMet binding at the S2 sites of each dimeric species [51]. This study presents the first crystal structure of the R336C variant of HsCBS in its basal conformation, a critical mutation contributing significantly to the high incidence of classical homocystinuria in Qatar. The structural data is compared with previously obtained data from the native protein [43,49] and its variant D444N [49]. The analysis is further supported by comprehensive kinetic, biochemical, biophysical, and computational studies, offering deeper insights into the impact of this mutation on the intrinsic mobility of the Bateman module and on the tautomeric equilibrium of the PLP cofactor. Results In 2013, we discovered that shortening the flexible loop (L512–529) connecting the last two b-strands of the CBS2 domain in human CBS did not affect the catalytic activity of the resulting protein construct (HsCBS-D516-525) or its response to the allosteric effector, AdoMet, compared to the full-length native enzyme (FL-HsCBS) [50]. Instead, this artificial modification avoided the formation of insoluble aggregates and reduced the oligomeric complexity of the enzyme, leading to the predominant formation of stable dimers. More importantly, this approach was crucial for achieving the crystallization of HsCBS in its two conformational states: ‘basal’ [50] and ‘AdoMet-bound activated’ [49]. Using a similar strategy, we have generated the recombinant construct HsCBS-D516-525R336C which harbors the R336C mutation. As expected, we found that constructs lacking amino acid residues 516–525 (HsCBS-D516-525-WT and HsCBS-D516-525-R336C) predominantly exist as stable dimers (Fig. S1B) within the concentration range typically used in purification protocols (less than 5mgmL 1 ). Structural, kinetic, and biophysical properties of the HsCBS-R336C variant To investigate the effects of the R336C substitution in HsCBS, we expressed, purified, and biochemically characterized the HsCBS-D516-525-R336C variant. In parallel, we purified HsCBS-D516-525-WT to be used as a reference control. The purity of both proteins was confirmed by SDS/PAGE analysis, which showed a single band at approximately 60 kDa (Fig. S1A), corresponding to the expected molecular weight of the monomer. We compared HsCBS-D516-525-WT and HsCBS-D516525-R336C at 25 °C, the temperature typically chosen for biochemical and biophysical analyses [35,52,53], and at 37 °C, the physiological temperature, to assess the impact of the mutation under these conditions. The purified HsCBS-D516-525-R336C protein displayed a red color, indicating proper heme binding. The absorption spectrum of the HsCBS-D516-525-R336C (Fig. 2A) showed the characteristic features of Fe(III)- heme coordination. In particular, the spectrum displayed a sharp Soret band at 428 nm, a dband at 364 nm, and a broad a/babsorption envelope with a maximum at 553 nm, confirming the correct heme incorporation. These spectral features closely resembled those of the wild-type enzyme (Fig. 2A), suggesting that the R336C mutation does not disrupt heme binding. Comparison of the absorption spectra at 25 °C 4936 The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. Altered PLP tautomerism impairs HsCBS-R336C activity C. Conter et al. 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License and 37 °C (Fig. 2A) revealed no significant differences for either WT or the R336C variant. Further structural characterization by far-UV circular dichroism (CD) spectroscopy (Fig. 2B) revealed a high a-helical content in HsCBS-D516-525-R336C, comparable to that of the wild-type enzyme. The spectra at 25 °C and 37 °C showed no significant changes in secondary structure for both proteins, suggesting that the R336C mutation does not alter the structural integrity of the enzyme under either condition. Thermal unfolding monitored by CD at 222 nm (Fig. 2C) showed melting temperatures (T m ) of 66.3 0.3 °C for the WT and 63.0 0.2 °C for the mutant. Although the mutation results in a measurable reduction in thermal stability, the relatively small difference of 3 °CinT m underscores that the overall structural integrity of the enzyme is largely preserved. However, it is important to note that these values are still quite far from the physiological temperature. To assess structural stability over time, the proteins were incubated at 37 °C and monitored by CD at 222 nm over 60 min (Fig. 2D). Both HsCBS-D516-525-WT and HsCBS-D516-525R336C showed stable signals, indicating no significant unfolding or aggregation during the incubation period. Next, AdoMet binding to HsCBS-D516-525-WT and HsCBS-D516-525-R336C was assessed by isothermal titration calorimetry (ITC) at both 25 °C and 37 °C (Fig. S2, Table S1). At 25 °C, both proteins showed similar AdoMet binding stoichiometry (N), with values of 0.48 0.05 for HsCBS-D516-525-WT and 0.45 0.01 for HsCBS-D516-525-R336C. The dissociation constants (K d ) were also comparable, with 429 31 nMfor HsCBS-D516-525-WT and 454 60 nMfor HsCBS-D516-525-R336C, reflecting strong AdoMet affinity. The thermograms obtained are comparable to those previously published for the HsCBS-D516-525-WT, confirming the consistency of our data with previous studies [52]. The enthalpy change (DH) was similar for both variants (8.8 0.3 kcalmol 1 for HsCBS-D516-525-WT and 8.7 0.1 kcalmol 1 for HsCBS-D516-525-R336C), while the entropic contribution (TDS) remained minimal (0.2 kcalmol 1 for HsCBS-D516-525-WT and 0.1 kcalmol 1 for HsCBS-D516-525-R336C), suggesting that the binding is primarily driven by enthalpic forces. At 37 °C, a temperature closer to physiological conditions, the AdoMet binding stoichiometry remained comparable between variants (0.44 0.07 for HsCBS-D516-525-WT and 0.48 0.05 for HsCBS-D516-525-R336C), though a slight increase in K d was observed (505 52 nMfor WT and 550 60 nMfor R336C), suggesting a modest reduction in affinity at higher temperatures. The DHvalues were lower at 37 °C(4.8 0.1 kcalmol 1 for WT and 5.9 0.1 kcalmol 1 for R336C), with a corresponding increase in the entropic contribution (TDS of 4.1 kcalmol 1 for WT and 2.9 kcalmol 1 for R336C). These data indicate that while enthalpy remains the main driving force, the role of entropy becomes more pronounced at physiological temperatures. Overall, the data confirmed that the HsCBS-D516-525-R336C variant was able to bind AdoMet with similar affinity to the wild-type enzyme across different temperatures, suggesting that the mutation does not significantly affect the AdoMet binding site. Subsequently, we verified if HsCBS-D516-525R336C was catalytically responsive to AdoMet activation in both canonical and alternative reactions at 37 °C. Kinetic analyses of the canonical CBS reaction (L-serine +L-homocysteine ?L-cystathionine +H 2 O) were performed to assess the functional effects of the R336C mutation in the presence and absence of AdoMet (Fig. 2G, Table 1). We also performed the same analyses for the HsCBS-D516-525-WT (Fig. 2E, Table 1). Notably, the k cat value for the HsCBS-D516525-R336C variant in the absence of AdoMet was lower (0.70 s 1 ) compared to that of the wild-type enzyme (2.2 s 1 ). Although the presence of AdoMet increased the activity of both the wild-type and mutant enzymes, the R336C variant remained less efficient, with a k cat of 1.50 s 1 compared to 4.9 s 1 for the wild-type. The K m values for the mutant were comparable to those of the wild-type enzyme, indicating that substrate binding affinity was not significantly affected by the mutation. However, the catalytic efficiency (k cat /K m ) was lower for the R336C variant. In the absence of AdoMet, the catalytic efficiency was reduced 2.6-fold for the R336C variant (0.12 mM 1 s 1 versus 0.31 mM 1 s 1 for the wild type), and in the presence of AdoMet, it was reduced 3.9-fold (0.23 mM 1 s 1 versus 0.90 mM 1 s 1 for the wild type). In the alternative reaction producing H 2 S(Lcysteine +L-homocysteine ?L-cystathionine +H 2 S), the HsCBS-D516-525-R336C variant also exhibited reduced catalytic parameters (Fig. 2H, Table 1) compared to HsCBS-D516-525-WT (Fig. 2F, Table 1). The k cat for H 2 S production by the HsCBS-D516-525R336C mutant without AdoMet was 0.60 s 1 , lower than the 1.4 s 1 observed for the wild-type enzyme. The mutant also remained less active in the presence of AdoMet, with a k cat of 1.20 s 1 compared to 3.3 s 1 for the wild type. The catalytic efficiency of the HsCBS-D516-525-R336C variant for H 2 S production was severely impaired. In the absence of AdoMet, the k cat /K m was reduced 3-fold (0.076 mM 1 s 1 versus 0.23 mM 1 s 1 for the wild type), and in the presence 4937The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. C. Conter et al. Altered PLP tautomerism impairs HsCBS-R336C activity 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License of AdoMet, the reduction was about 4.2-fold (0.15 mM 1 s 1 versus 0.63 mM 1 s 1 for the wild type). Overall, the data suggest that the R336C mutation preserves the structural integrity and cofactor binding but reduces the catalytic efficiency of the enzyme in both canonical and alternative reactions. 4938 The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. Altered PLP tautomerism impairs HsCBS-R336C activity C. Conter et al. 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Overall structure of HsCBS-D516-525-R336C The R336C mutation barely alters the overall fold of human CBS To investigate the factors contributing to the drastic reduction in enzyme activity in the R336C variant and encouraged by the biophysical data indicating a structurally stable protein, we proceed to crystallize the HsCBS-D516-525-R336C construct both in the absence and presence of two of its substrates, Ser or Cys, as well as its allosteric activator, AdoMet. The decision to use HsCBS-D516-525-R336C was based on our laboratory’s extensive experience on the structural analysis of CBS enzymes from various species [45,48,49], which has historically posed a significant challenge. As expected, the crystallization of HsCBS-D516-525-R336C was equally difficult, and despite significant experimental efforts, only a few tiny crystals grown in the absence of substrates and AdoMet could be obtained. These crystals formed clusters of small, flat, red-colored needles that diffracted Xrays to an average resolution of approximately 3.6  A. The final electron density maps were of excellent quality, enabling accurate model refinement with robust statistical support (Table S2). Intriguingly, the resolved HsCBS-D516-525-R336C structure revealed an overall folding pattern that, far from exhibiting disordered or unstructured regions, showed a remarkably high similarity to the native enzyme (r.m.s.d. =0.18) (Fig. 3). Structurally, the catalytic core belongs to the b-family of PLP-dependent enzymes. Like the native protein, it is composed of 13 a-helices and two b-sheets consisting of four (b4–b7) and six (b2, b3, and b8–b11) strands, respectively (Fig. 3A). Loops L145–148, L171–174, and L191–202 are prominently positioned at the entrance of the catalytic site, between the core and the Bateman module of the complementary monomer. These loops, along with strands b4, b5, and b6—which precede them in the polypeptide chain— and helix a7, which follows the L191–202 loop, are all clearly discernible. The catalytic core and the Bateman module are linked by a long segment (residues 382–411, in green in Fig. 3A) that includes two ahelices, a15 and a16. The Bateman module includes a pair of interleaved CBS motifs (CBS1, 412–471; CBS2, 477–551) that show an aabba and an ababba fold, respectively, as formerly described in the native enzyme [50]. The two CBS motifs engage through their twoor three-stranded b-sheets, with the long edges of this bilayer interface giving rise to the two main symmetry-related cavities of the Bateman module (S1 and S2). The more exposed S2 cavity corresponds to the AdoMet-binding site. Additionally, each short Nterminal helix (a17 or a20) is an essential component of the other CBS motif, engaging in antiparallel packing with its C-terminal b-strand (b16 or b13) and the a-helix (a22 or a19). This configuration leads to a nested structure formed by both CBS motifs, exhibiting pseudo-C2 symmetry. Similarly to the native enzyme [50],HsCBS-D516525-R336C forms basket-shaped symmetrical dimers in the absence of AdoMet (Fig. 3E), with the catalytic core of each subunit interacting with both the catalytic core and the Bateman regulatory module of the opposing subunit. The positioning of the catalytic core and regulatory domain remains unaltered with respect to the native enzyme (Fig. 3F) and places the Bateman module of the first subunit just above the entrance to the catalytic site of the second subunit, which restricts substrate access to this cavity, explaining why the basal conformation is poorly active either in the native enzyme or in the R336C variant. As formerly described in the native HsCBS enzyme [50], the interaction between the Bateman module and the core in Fig. 2. Spectroscopic and kinetic properties of HsCBS-D516-525-WT and HsCBS-D516-525-R336C. (A) UV–visible absorption spectrum of 15 lMpurified HsCBS-D516-525-WT (blue, 25 °C; black, 37 °C) and HsCBS-D516-525-R336C (green, 25 °C; red, 37 °C) recorded in 20 mM sodium phosphate buffer pH 7.5. (B) Far-UV-CD spectra of 0.2 mgmL 1 HsCBS-D516-525-WT (blue, 25 °C; black, 37 °C) and HsCBS-D516525-R336C (green, 25 °C; red, 37 °C) in 20 mMsodium phosphate buffer pH 7.5. (C) Thermal denaturation of 0.2 mgmL 1 HsCBS-D516525-WT (black) and HsCBS-D516-525-R336C (red) monitored by circular dichroism signal at 222 nm in 20 mMsodium phosphate buffer pH 7.5. (D) Time-dependent stability of 0.2 mgmL 1 HsCBS-D516-525-WT (black) and HsCBS-D516-525-R336C (red) monitored at 222 nm over 60 min at 37 °C. (E, G) Steady-state enzyme kinetics for the canonical reaction (L-Ser +L-Hcys ?L-Cth +H 2 O) using the coupled CBL-LDH assay at pH 8.6. (E) HsCBS-D516-525-WT and (G) HsCBS-D516-525-R336C in the absence (black) and presence (green) of AdoMet. (F, H) Steady-state enzyme kinetics for the H 2 S-producing reaction (L-Cys +L-Hcys ?L-Cth +H 2 S) using the lead acetate assay. (F) HsCBS-D516525-WT and (H) HsCBS-D516-525-R336C in the absence (black) and presence (blue) of AdoMet. Data in (A–D) are the average of three experiments from three different independent purifications. Data shown in (E–H) were obtained from three independent experiments using three different purification batches. Error bars represent the Standard Error of the Mean (SEM). Data were fitted using the Michaelis–Menten equation with GRAPHPAD PRISM software, version 10.2.3 (GraphPad Software, San Diego, CA, USA). The kinetic parameters derived from the fit are summarized in Table 1. 4939The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. C. Conter et al. Altered PLP tautomerism impairs HsCBS-R336C activity 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License the HsCBS-D516-525-R336C variant exhibits greater contribution of secondary structural elements from the CBS2 motif compared to CBS1. This interaction involves helices a21 and a22 from CBS2, along with helix a19 and the loop L482–486, which links helix a20 to strand b14 in CBS1. Additionally, various elements from the core domain, including helix a7, strand b6, and loops L171–174 and L192–202, contribute to the formation of the interface between the regulatory domain and the catalytic core. The R336C mutation does not impair PLP and AdoMet binding A priori, given its position midway through helix a12, with the side chain exposed to the solvent and distant from the catalytic site (Fig. 3A,C), the substitution of a cysteine in place of arginine at position 336 should not impact the catalytic capacity of the CBS enzyme. However, this initial expectation contrasts with our kinetic assays (Fig. 2D,E, Table 1) despite maintaining an unperturbed overall three-dimensional fold and the immediate neighborhood surrounding residue 336 (Fig. 3C,D and Fig. S3). Based on these findings that rule out total or partial denaturation of the mutant protein, our analysis focused on other structural factors that could potentially impair the enzyme’s activation mechanisms and catalytic process. First, we investigated whether the R336C mutation compromises or somehow distorts the S2 AdoMet-binding site within the Bateman module. Again, the crystal structure showed that mutation R336C does not alter the orientation of the CBS motifs or the residues involved in AdoMet interaction. Moreover, the residues at the interface between the Bateman module and the catalytic core remain unchanged. These findings are consistent with the biophysical characterization, confirming that the R336C variant retains its ability to bind AdoMet (Fig. S2 and Fig. 2D,E). Then, we assessed whether the R336C mutation affects the binding of the PLP cofactor within the catalytic cavity, as patients carrying the R336C mutation are not responsive to treatment with pyridoxine [35]. Addressing this question is not feasible through spectroscopic measurements, as the absorption signal from the PLP (~410 nm) is obscured by that of the heme group (~428 nm). Instead, we used our crystallographic data. As illustrated in Fig. 3B, the presence of the cofactor within the catalytic cavity was unequivocally confirmed by Polder omit maps showing distinct residual electron density at the potential PLP cofactor location during the refinement of the final model. The crystal structure of the R336C variant also indicated that the orientation of the cofactor mimics that found in the native enzyme, enabling it to maintain the interactions typically established with key residues within the catalytic cavity (Fig. 3C,D and Fig. S3). Of note, the heme group was also present in our structure, displaying an interaction pattern with the protein consistent with that of the native enzyme. The R336C mutation alters the intrinsic dynamics of the Bateman module Given the challenges in explaining the effect of the R336C mutation solely through the crystallographic data, which represent an averaged snapshot of a protein’s spatial and temporal structure, captured in a local energy minimum under specific buffered conditions, we investigated the dynamic behavior of HsCBS-D516-525-R336C and compared it with HsCBS-D516-525-WT (native enzyme) and a highly active variant, HsCBS-D444N, previously studied [50], using microsecond molecular dynamics (ls-MD) simulations. To uncover the most significant conformational transitions in each system, we applied principal Table 1. Steady-state enzyme kinetics of HsCBS-D516-525-WT from compared to HsCBS-D516-525-R336C for both the canonical and alternative H 2 S-forming reactions at 37 °C. Data are mean SEM. HsCBS-D516-525-WT HsCBS-D516-525-R336C AdoMet +AdoMet AdoMet +AdoMet L-Ser +L-Hcys ?L-Cth +H 2 O k cat (s 1 ) 2.2 0.1 4.9 0.1 0.70 0.02 1.50 0.03 K m L-Ser (mM) 7.1 0.5 5.6 0.5 5.6 0.5 6.4 0.5 k cat /K m (mM 1 s 1 ) 0.31 0.03 0.90 0.09 0.12 0.01 0.23 0.02 L-Cys +L-Hcys ?L-Cth +H 2 S k cat (s 1 ) 1.4 0.1 3.3 0.1 0.60 0.02 1.20 0.04 K m L-Cys (mM) 6.1 1.0 5.2 0.8 7.8 0.9 7.6 0.8 k cat /K m (mM 1 s 1 ) 0.23 0.05 0.63 0.11 0.076 0.011 0.15 0.02 4940 The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. Altered PLP tautomerism impairs HsCBS-R336C activity C. Conter et al. 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License component analysis (PCA) as a method for dimensionality reduction. This analysis, which aligns with previous findings from our laboratory [48], enabled the deconvolution of each protein’s MD trajectory and highlighted two main distinct types of structural changes. The first one primarily affects the intrinsic Fig. 3. Crystal structure of HsCBS-D516-525-R336C. (A) Ribbons representation of the protein monomer in the basal conformation. The core, the interdomain linker and the Bateman module are colored in cyan, green and orange, respectively. PLP and heme are shown in sticks; (B) Polder omit map confirming clear residual electron density for the PLP cofactor bound to residue K119 (in sticks) within the catalytic cavity (map represented at 3r). Similar density is observed at the equivalent position in the complementary subunit of the dimer; (C, D) The vicinity of the amino acid residues C336 and R336 in both the mutant and the native enzyme crystals shows minimal changes; (E) HsCBS-D516-525R336C dimers found in the crystals. The two complementary subunits are colored in blue and yellow, respectively. (F) Overlay of HsCBS-D516525-R336C and native HsCBS-D516-525 crystal structures. All panels were represented with UCSF CHIMERA (version 1.13.1) [80]. 4941The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. C. Conter et al. Altered PLP tautomerism impairs HsCBS-R336C activity 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License that the R336C mutation altered catalytic activity but did not impact AdoMet binding or activation. The K d remains in the nanomolar range, and k cat increases 2–3 fold in the presence of the activator, as previously reported [50]. Our crystallographic data align with biochemical and biophysical findings, showing that the enzyme’s fold remains unaltered and its active site intact, including PLP’s proper anchoring to the catalytic K119 residue (Fig. 1). This finding contrasts with previous hypotheses attributing the pyridoxine insensitivity observed in R336C patients to defects in PLP binding, structural instability, partial unfolding, or the inability to bind AdoMet. Although this structural integrity might seem inconsistent with the severe patient phenotype, it emphasizes the need to examine potential dynamic changes that could subtly disrupt enzyme function. Our molecular dynamics (MD) simulations provide further insight into the effects of the R336C mutation, revealing that it alters the mobility of the Bateman module, as well as interactions within the catalytic domain that determine the tautomeric equilibrium of the Shiff base. Specifically, the mutation disrupts key electrostatic and hydrophobic contacts that, while not destabilizing the Bateman structure, affect its flexibility and positioning. Additionally, the R336C mutation, although distant from the catalytic center, appears to disrupt interactions between the conserved N149 and the PLP phenol group, potentially affecting the cofactor’s tautomeric state. Specifically, the mutation might favor the non-catalytic enolimine form of PLP over the active ketoenamine form required for catalysis. Prior studies have associated CBS inactivation with tautomeric shifts induced by CO binding to PLP [39], suggesting that the R336C mutation similarly stabilizes an inactive PLP state, thus hindering catalytic function. This insight into tautomeric equilibrium offers a potential mechanism for pyridoxine insensitivity in R336C patients and highlights an avenue for developing drugs that stabilize the active ketoenamine form in R336C and possibly other CBS variants. A notable feature of the R336C mutant observed in our crystallographic analysis is its tendency to form filamentous structures. This fact, which we first described after resolving the crystal structure of the basal conformation of the human enzyme [50], gains new significance following the electron microscopy studies recently conducted by McCorvie et al.[51],in which linear filaments, structurally related but differently arranged, are also described. There, and using full-length CBS constructs, the authors showed that the L512–529 loop region is essential for enzyme polymerization. Our crystallized constructs, lacking residues 516–525, exhibit flattened, linear-shaped polymers that show altered interdimer interactions and a reduction in helicity, yet retain the b-sheet interactions between the Bateman modules of consecutive dimers. Structurally, these C222 1 filaments are closely related to the filaments observed by cryo-EM by a counterclockwise rotation between consecutive dimers (Movie S7). As pointed out in 2013, our HsCBS-D516-525 construct also forms alternative filaments that polymerize via a-helix interactions between the Bateman modules of subsequent CBS dimers (Fig. 6B, Movie S5). We propose that this type of filament, only observed in crystals so far, represents an alternative filamentous stable assembly of the basal state of the enzyme that is formed without the contribution of the L512–529 loop. The observation of such filamentous structures in different CBS variants—including disease-causing mutants such as R336C and D444N—suggests that pathogenic mutations might not only impair catalytic activity but also modify the range of polymeric assemblies CBS can adopt in different tissues according to expression levels or, complementarily, alter the enzyme activity by modifying the polymerization pattern. Supporting this hypothesis is the observation that minor structural variations, such as the removal of specific residues within a protein loop like L512–529, can induce a 90-degree shift in the relative orientation of consecutive dimers within the filament, disrupting the helicity characteristic of the native enzyme. Our new data suggest that not all filament types are equally affected by amino acid alterations or substitutions. The functional significance of different types of CBS filaments remains uncertain, but understanding their effects on enzyme stability and activity could help guide therapeutic strategies. If polymerization enhances CBS functionality, inhibiting filament formation could be counterproductive for homocystinuria patients. Conversely, filament inhibitors could benefit conditions with CBS overexpression, such as Down syndrome, where abnormally increased expression (and polymerization) of the enzyme may contribute to pathology. In summary, our study shed light on the complex effects of the R336C mutation in CBS, revealing that it disrupts catalytic efficiency without compromising structural integrity. Our findings suggest that the mutation introduces dynamic changes, particularly in the Bateman module’s mobility, that affect the active site and substrate accessibility rather than causing significant structural changes. Insights into the potential shift in the tautomeric states of the PLP cofactor, driven by conformational changes in surrounding 4948 The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. Altered PLP tautomerism impairs HsCBS-R336C activity C. Conter et al. 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License residues, suggest a mechanistic basis for pyridoxine insensitivity in R336C patients, highlighting a potential therapeutic target to stabilize the active ketoenamine form. Furthermore, the identification of filamentous structures in R336C adds a compelling dimension to our understanding of CBS functionality, although the precise effects on the stability and activity of the enzyme remain incompletely understood. Future research aimed at elucidating the functional significance of CBS filaments in the context of pathogenic mutations, along with targeted approaches that integrate both the structural dynamics of CBS and the regulatory role of its polymeric forms, could significantly enhance treatment strategies for homocystinuria and other CBS-related disorders. Materials and methods Materials Pyridoxal 5’-phosphate (PLP), L-homocysteine, L-serine, L-cysteine, lactic dehydrogenase, isopropyl-b-D-thiogalactoside (IPTG), nicotinamide adenine dinucleotide (reduced form) (NADH), lead acetate, sodium chloride, HEPES, S-adenosylmethionine (AdoMet), and imidazole were all purchased from Merck Millipore (Burlington, MA, USA). Tacsimate and polyethylene glycol (PEG) 3350 were purchased from Hampton Research Corporation (Aliso Viejo, CA, USA). All the chemicals were of the highest grade available. Protein production and spectroscopic measurements The HsCBS-D516-525-R336C mutant was generated via site-specific mutagenesis on the pET28a-HsCBS-D516-525WT plasmid, using the QuikChangeII site-directed mutagenesis kit (Agilent Technologies Inc, Santa Clara, CA, USA), following the manufacturer’s protocol. The primers used to introduce the mutation were as follows: forward 50-CAGTCATATGCCTTCTGAGACCCCCCA-30and reverse 50-GATACTCGAGTCACTTCTGGTCCCGCTC30. After sequence validation, the pET28a-HsCBS-D516525-WT and pET28a-HsCBS-D516-525-R336C plasmids were then transformed into Escherichia coli Rosetta (DE3) expression cells (Novagen, Merck KGaA, Darmstadt, Germany). The expression and purification of both proteins followed established procedures for the wild-type protein [28]. Absorption spectra were acquired at 25 °C and 37 °C using a Jasco V630 UV–visible spectrophotometer in 20 mMsodium phosphate buffer at pH 7.5. CD experiments were recorded using a JASCO J-815 spectropolarimeter equipped with a Peltier temperature controller. FarUV CD spectra of protein samples (0.2 mgmL 1 )in 20 mMsodium phosphate pH 7.5 at 25 °C and 37 °C were measured between 190 and 250 nm at 25 °C, using a scanning speed of 50 nmmin 1 and a 1-mm quartz cuvette. A minimum of three scans were averaged and corrected against the corresponding buffer blank [53,63]. Thermal unfolding profiles of 0.2 mgmL 1 HsCBS-D516-525-WT and HsCBS-D516-525-R336C were recorded by measuring the CD signal at 222 nm in a temperature range between 20 °C and 100 °C (scan rate 1.5 °Cmin 1 )in20mM sodium phosphate buffer pH 7.5 using a quartz cuvette with 1-mm pathlength [63,64]. The loss of secondary structure at 37 °C over time was measured by continuously monitoring the signal at 222 nm at a protein concentration of 0.2 mgmL 1 in 20 mMsodium phosphate buffer at pH 7.5 using a 1-mm path length cuvette [65]. Oligomeric state determination The oligomeric state of HsCBS-D516-525-WT and HsCBS-D516-525-R336C was assessed BY size-exclusion chromatography on a Superdex 200 Increase 10/300 GL column (Cytiva, Marlborough, MA, USA) in 50 mMHepes pH 7.5, 150 mMNaCl, and 0.1 mMDTT. A high molecular weight gel filtration calibration kit was used to generate the calibration curve, following the protocol described in [66]. Enzyme activity assays The activity of CBS variants was evaluated using a continuous coupled assay involving cystathionine b-lyase (CBL from E. coli) and lactate dehydrogenase (LDH), as described in [66]. Assays were conducted at 37 °C and pH 8.6 using a Jasco V630 spectrophotometer, with a final reaction volume of 200 lL. The reaction mixture included 20 lMPLP, 0.2 mMNADH, 1.8 lMLDH, 1.5 lMCBL, 0.1–7lMCBS variants, and 2 mML-Hcys. The reaction was initiated by adding 1–40 mML-Ser, and NADH oxidation was monitored at 340 nm. For H 2 S production, measurements were taken by tracking the formation of lead sulfide at 390 nm (e 390 =5500 M 1 cm 1 ) as described in [63],at37°C. The assay was carried out with 1–40 mMLCys and 2 mML-Hcys. Where specified, both assays were performed in the presence of 0.5 mMAdoMet. Isothermal titration calorimetry (ITC) ITC experiments were performed using a TA Instruments Nano ITC. Protein samples at concentrations of 20–30 lM (monomer) were titrated with 250–300 lMAdoMet in 35–44 injections of 0.8 lL each, with 120-s intervals between injections at 25 °Cand37°C. A control titration of AdoMet into buffer without protein was conducted, showing no significant heat changes in the reference cell. Data were analyzed using the TA Instruments software, and a one-site binding model was applied. The reported values represent the mean SEM 4949The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. C. Conter et al. Altered PLP tautomerism impairs HsCBS-R336C activity 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License from at least three independent titrations, each performed with separate protein preparations. Protein crystallization Protein crystallization experiments were performed using the vapor-diffusion method in a sitting drop technique, following a protocol adapted from [48]. Crystallization droplets were set up using an automated MOSQUITO nanodispenser (SPT Labtech Ltd., Melbourn, UK) in 96-well MRC plates (Molecular Dimensions Ltd., Sheffield, UK), mixing 200 nL of protein solution with 200 nL of precipitant solution, and incubated at 293 K. Optimal crystals of HsCBS-D516-525R336C were obtained in hanging drops containing 16% (wt/vol) PEG 3350 and 3% Tacsimate at pH 8.0 with a protein concentration of 16 mgmL 1 . Crystals typically formed within 6 days. The drops were composed of 1 lL of protein solution and 1 lL of precipitant. Crystals were harvested using nylon loops and cryopreserved in a solution containing 16% (wt/vol) PEG 3350, 3% Tacsimate pH 8.0, and 21% (vol/vol) glycerol as a cryoprotectant. All X-ray datasets were collected at Synchrotron beamlines XALOC (ALBA, Barcelona, Spain) and were processed as reported in [48].Thestructures were determined by molecular replacement methods using PHASER [67],fromthePHENIX SUITE [68], with the structure of the HsCBS (PDB ID code 4LOD)[50], as the initial search model. Refinement was done with PHENIX.REFINE [69]. The model was built with COOT [70]. Figures were prepared with PYMOL (The PyMOL Molecular Graphics System, Version 2.2.3, Schr€ odinger, LLC, New York, NY, USA) and UCSF CHIMERA (version 1.13.1). The crystal characteristics and refinement statistics are summarized in Table S1. Molecular modeling and dynamics The crystallographic structure of HsCBS-D516-525-R336C and the wild-type homodimer (PDB ID 4LOD for HsCBS) [50] were used as templates. In both chains of the homodimer, K119 residues were covalently attached to the PLP cofactors, forming internal aldimines. Parameters for these modified residues were derived with the Amber antechamber module, using the gaff2 force field and partial charges adjusted to fit the electrostatic potential generated via HF/631G(d) using the RESP method [71]. Charges were computed based on the Merz-Singh-Kollman approach using Gaussian 16 [72]. Heme groups were modeled following all-atom parameters from Giammona’s work (D. A. Giammona, Ph.D. thesis, University of California, Davis, 1984). Molecular dynamics (MD) simulations for each homodimer were performed using the Amber 22 suite [73], with the ff14SB [74] and gaff2 [75] force fields. The initial structures were neutralized with Na + or Cl  ions and placed in a cubic TIP3P water box [76], maintaining a 10  A buffer between the solute and the box edge. A two-stage geometry optimization was employed: the first stage minimized the positions of solvent molecules and ions, while the second stage involved unrestrained minimization of all atoms in the simulation cell. Systems were then heated from 0 to 300 K (25 °C) and from 0 to 310 K (37 °C) under a constant pressure of 1 atm and periodic boundary conditions, with harmonic restraints of 10 kcalmol 1 applied to the solute using the Andersen temperature coupling method. The time step during heating was set to 1 fs to allow any potential inhomogeneities to equilibrate. Water molecules were handled using the SHAKE algorithm [77] to fix the angle between hydrogen atoms throughout the simulations. Long-range electrostatic interactions were modeled using the particle mesh Ewald method [78], with an 8  A cutoff for Lennard-Jones interactions. Each system underwent equilibration for 2 ns with a 2 fs time step at 300 K and 310 K and constant volume. The production phase consisted of a 2000 ns NVT trajectory at 300 K and 310 K with a 2 fs time step, controlled by the Andersen thermostat. Principal component analysis (PCA) was performed using the Bio3d program [79]. Representative snapshots of the production phase were generated via the cpptraj module in Amber and visualized in PYMOL. Acknowledgements This research was supported by Spanish Ministerio de Ciencia e Innovaci on (MICINN), Grants No PID2019109055RB-I00 and PID2022-141748OB-I00, to LAM-C; PID2021-125946OB-I00 to GJ-O; Italian Ministry of University and Research (MUR)-PRIN 2022 grant No. 20224BYR59 to AA; MICINN: PID2020-117116RB-I00 integrado en el Plan Estatal de Investigaci on Cient ıfica y T ecnica e Innovaci on, cofinanciado con Fondos FEDER (for MLM-C); Ministerio de Ciencia e Innovaci on, Programa Retos-Colaboraci on RTC2019-007125-1 (for MLM-C); Instituto de Salud Carlos III, Proyectos Investigaci on en Salud DTS20/00138 (for MLM-C); La Caixa Scientific Foundation (HR17-00601) (for MLMC); La Caixa Consortium (for MLM-C); Ayudas Fundaci on Cient ıfica AECC para proyectos coordinados (IGTP-AECC_2022-042) (for MLM-C); Transferencia tecnol ogica 2022 (6/12/TT/2022/00001) (for MLM-C); Desarrollo Tecnol ogico en Salud (DTS20/00138) (for MLM-C); Ayudas a proyectos de investigaci on y desarrollo en salud (2023333041) (for MLM-C); Caixa Impulse Innovation 2023 (CI23-20155) (for MLM-C) to MLM-C; GSRA grant (GSRA6-1-0413-19013) to DWAS. We also thank MINECO for the Severo Ochoa Excellence Accreditation (CEX2021-001136-S). Conflict of interest LAM-C provides ad hoc consulting to Travere Therapeutics, which clinically develops Pegtibatinase as an enzyme replacement therapy for classical 4950 The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. Altered PLP tautomerism impairs HsCBS-R336C activity C. Conter et al. 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License homocystinuria. The rest of the authors declare no conflict of interest. Author contributions CC conceived the biophysical, biochemical, and structural study, performed the experiments, analyzed the data, and wrote the manuscript. RN-F performed the MD study and analyzed the data. DWA-S, CF-R, NG-U, GKN, and MN analyzed the data and contributed to the discussion of the results. MLM-C contributed to the discussion of the results and supported funding acquisition. AA contributed to the discussion of the results and supported funding acquisition. GJ-O conceived the MD study, contributed to the discussion of the results, and supported funding acquisition. LAM-C conceived the overall study, provided financial support, analyzed the data, and wrote the manuscript. All authors read and approved the final manuscript. Data availability statement All data generated or analyzed during this study are included in this published article and its Supporting Information files. The crystal structure of HsCBS-D516525-R336C (PDB code 9HIF) is accessible through the Protein Data Bank (https://www.rcsb.org). References 1 Yadav PK & Banerjee R (2012) Detection of reaction intermediates during human cystathionine b-synthasemonitored turnover and H2S production. J Biol Chem 287, 43464–43471. 2 Mudd SH, Finkelstein JD, Irreverre F & Laster L (1964) Homocystinuria: an enzymatic defect. Science 143, 1443–1445. 3 Finkelstein JD, Mudd SH, Irreverre F & Laster L (1964) Homocystinuria due to cystathionine synthetase deficiency: the mode of inheritance. Science 146, 785–787. 4 Gerritsen T, Vaughn JG & Waisman HA (1962) The identification of homocystine in the urine. Biochem Biophys Res Commun 9, 493–496. 5 Carson NAJ & Neill DW (1962) Metabolic abnormalities detected in a survey of mentally backward individuals in Northern Ireland. Arch Dis Child 37, 505– 513. 6 Singh S & Banerjee R (2011) PLP-dependent H2S biogenesis. Biochim Biophys Acta Proteins Proteomics 1814, 1518–1527. 7 Kabil O & Banerjee R (2010) Redox biochemistry of hydrogen sulfide. J Biol Chem 285, 21903–21907. 8 Hamosh A (2004) Online mendelian inheritance in man (OMIM), a knowledgebase of human genes and genetic disorders. Nucleic Acids Res 33, D514–D517. 9 Castro R, Rivera I, Blom HJ, Jakobs C & De Almeida IT (2006) Homocysteine metabolism, hyperhomocysteinaemia and vascular disease: an overview. J Inherit Metab Dis 29,3–20. 10 Makris M (1998) Hyperhomocysteinemia is a risk factor for venous and arterial thrombosis. Br J Haematol 101,18–20. 11 Seshadri S, Beiser A, Selhub J, Jacques PF, Rosenberg IH, D’Agostino RB, Wilson PWF & Wolf PA (2002) Plasma homocysteine as a risk factor for dementia and Alzheimer’s disease. N Engl J Med 346, 476–483. 12 Welch GN & Loscalzo J (1998) Homocysteine and atherothrombosis. N Engl J Med 338, 1042–1050. 13 Folin M, Baiguera S, Gallucci M, Conconi MT, Di Liddo R, Zanardo A & Parnigotto PP (2005) A crosssectional study of homocysteine-, NO-levels, and CTfindings in Alzheimer dementia, vascular dementia and controls. Biogerontology 6, 255–260. 14 De Franchis R, Sperandeo MP, Sebastio G, Andria G & The Italian Collaborative Study Group on Homocystinuria (1998) Clinical aspects of cystathionine b-synthase deficiency: how wide is the spectrum? Eur J Pediatr 157, S67–S70. 15 Naughten ER, Yap S & Mayne PD (1998) Newborn screening for homocystinuria: Irish and world experience. Eur J Pediatr 157, S84–S87. 16 Tada K, Tateda H, Arashima S, Sakai K, Kitagawa T, Aoki K, Suwa S, Kawamura M, Oura T, Takesada M et al. (1984) Follow-up study of a nation-wide neonatal metabolic screening program in Japan: a collaborative study group of neonatal screening for inborn errors of metabolism in Japan. Eur J Pediatr 142, 204–207. 17 Garc ıa-Jim enez MC, Baldellou A, Garc ıa-Silva MT, Dalmau-Serra J, Garc ıa-Cazorla A, G omez-L opez L, Pedr on Giner C, Alonso Luengo O, Pe~ na Quintana L, Luz Couce M et al. (2012) Estudio epidemiol ogico de las enfermedades metab olicas con homocistinuria en Espa~ na. An Pediatr (Barc) 76, 133–139. 18 Landrum MJ, Lee JM, Benson M, Brown G, Chao C, Chitipiralla S, Gu B, Hart J, Hoffman D, Hoover J et al. (2016) ClinVar: public archive of interpretations of clinically relevant variants. Nucleic Acids Res 44, D862–D868. 19 Stenson PD, Ball EV, Mort M, Phillips AD, Shiel JA, Thomas NST, Abeysinghe S, Krawczak M & Cooper DN (2003) Human gene mutation database (HGMD): 2003 update: HGMD 2003 UPDATE. Hum Mutat 21, 577–581. 20 Jano s ık M, Oliveriusov a J, Jano s ıkov a B, Sokolov aJ, Kraus E, Kraus JP & Ko zich V (2001) Impaired heme binding and aggregation of mutant cystathionine b4951The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. C. Conter et al. Altered PLP tautomerism impairs HsCBS-R336C activity 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License synthase subunits in homocystinuria. Am J Hum Genet 68, 1506–1513. 21 Ko zich V, Sokolov a J, Klatovsk a V, Krijt J, Jano s ık M, Jel ınek K & Kraus JP (2010) Cystathionine bsynthase mutations: effect of mutation topology on folding and activity. Hum Mutat 31, 809–819. 22 Hn ızda A, Majtan T, Liu L, Pey AL, Carpenter JF, Kod ı cek M, Ko zich V & Kraus JP (2012) Conformational properties of nine purified cystathionine b-synthase mutants. Biochemistry 51, 4755–4763. 23 Hn ızda A, Jurga V, Rakov aK&Ko  zich V (2012) Cystathionine beta-synthase mutants exhibit changes in protein unfolding: conformational analysis of misfolded variants in crude cell extracts. J Inherit Metab Dis 35, 469–477. 24 Kopeck a J, Krijt J, Rakov aK&Ko  zich V (2011) Restoring assembly and activity of cystathionine bsynthase mutants by ligands and chemical chaperones. J Inherit Metab Dis 34,39–48. 25 Melenovsk a P, Kopeck a J, Krijt J, Hn ızda A, Rakov a K, Jano s ık M, Wilcken B & Ko zich V (2015) Chaperone therapy for homocystinuria: the rescue of CBS mutations by heme arginate. J Inherit Metab Dis 38, 287–294. 26 Katsushima F, Oliveriusova J, Sakamoto O, Ohura T, Kondo Y, Iinuma K, Kraus E, Stouracova R & Kraus JP (2006) Expression study of mutant cystathionine bsynthase found in Japanese patients with homocystinuria. Mol Genet Metab 87, 323–328. 27 Gaustadnes M, Wilcken B, Oliveriusova J, McGill J, Fletcher J, Kraus JP & Wilcken DE (2002) The molecular basis of cystathionine?-synthase deficiency in Australian patients: genotype-phenotype correlations and response to treatment. Hum Mutat 20, 117–126. 28 Al-Sadeq DW, Conter C, Thanassoulas A, Al-Dewik N, Safieh-Garabedian B, Mart ınez-Cruz LA, Nasrallah GK, Astegno A & Nomikos M (2024) Biochemical and structural impact of two novel missense mutations in cystathionine b-synthase gene associated with homocystinuria. Biochem J 481, 569–585. 29 Kraus JP, Janos ık M, Ko?ich V, Mandell R, Shih V, Sperandeo MP, Sebastio G, De Franchis R, Andria G, Kluijtmans LAJ et al. (1999) Cystathionine?-synthase mutations in homocystinuria. Hum Mutat 13, 362–375. 30 Moat SJ, Bao L, Fowler B, Bonham JR, Walter JH & Kraus JP (2004) The molecular basis of cystathionine ß-synthase (CBS) deficiency in UK and US patients with homocystinuria. Hum Mutat 23, 206. 31 Gallagher PM, Ward P, Tan S, Naughten E, Kraus JP, Sellar GC, McConnell DJ, Graham I & Whitehead AS (1995) High frequency (71%) of cystathionine bsynthase mutation G307S in Irish homocystinuria patients. Hum Mutat 6, 177–180. 32 Urreizti R, Asteggiano C, Cozar M, Frank N, Vilaseca MA, Grinberg D & Balcells S (2006) Functional assays testing pathogenicity of 14 cystathionine-beta synthase mutations. Hum Mutat 27, 211. 33 Zschocke J, Kebbewar M, Gan-Schreier H, Fischer C, Fang-Hoffmann J, Wilrich J, Abdoh G, Ben-Omran T, Shahbek N, Lindner M et al. (2009) Molecular neonatal screening for homocystinuria in the Qatari population. Hum Mutat 30, 1021–1022. 34 Gupta S, Gallego-Villar L, Wang L, Lee H, Nasrallah G, Al-Dewik N, H€ aberle J, Th€ ony B, Blom HJ, BenOmran T et al. (2019) Analysis of the Qatari R336C cystathionine b-synthase protein in mice. J Inherit Metab Dis 42, 831–838. 35 Al-Sadeq DW, Thanassoulas A, Islam Z, Kolatkar P, Al-Dewik N, Safieh-Garabedian B, Nasrallah GK & Nomikos M (2022) Pyridoxine non-responsive p.R336C mutation alters the molecular properties of cystathionine beta-synthase leading to severe homocystinuria phenotype. Biochim Biophys Acta Gen Subj 1866, 130148. 36 Singh S, Madzelan P & Banerjee R (2007) Properties of an unusual heme cofactor in PLP-dependent cystathionine b-synthase. Nat Prod Rep 24, 631–639. 37 Taoka S, Lepore BW, Kabil € O, Ojha S, Ringe D & Banerjee R (2002) Human cystathionine b-synthase is a heme sensor protein. Evidence that the redox sensor is heme and not the vicinal cysteines in the CXXC motif seen in the Crystal structure of the truncated enzyme. Biochemistry 41, 10454–10461. 38 Majtan T, Singh LR, Wang L, Kruger WD & Kraus JP (2008) Active cystathionine b-synthase can Be expressed in heme-free Systems in the Presence of metalsubstituted porphyrins or a chemical chaperone. J Biol Chem 283, 34588–34595. 39 Weeks CL, Singh S, Madzelan P, Banerjee R & Spiro TG (2009) Heme regulation of human cystathionine bsynthase activity: insights from fluorescence and Raman spectroscopy. J Am Chem Soc 131, 12809–12816. 40 Majtan T, Pey AL, Gimenez-Mascarell P, Mart ınezCruz LA, Szabo C, Ko zich V & Kraus JP (2017) Potential pharmacological chaperones for cystathionine Beta-synthase-deficient homocystinuria. In Targeting Trafficking in Drug Development (Ulloa-Aguirre A & Tao Y-X, eds), pp. 345–383. Springer International Publishing, Cham. 41 Bateman A (1997) The structure of a domain common to archaebacteria and the homocystinuria disease protein. Trends Biochem Sci 22,12–13. 42 Baykov AA, Tuominen HK & Lahti R (2011) The CBS domain: a protein module with an emerging prominent role in regulation. ACS Chem Biol 6, 1156–1163. 43 Ere~ no-Orbea J, Oyenarte I & Mart ınez-Cruz LA (2013) CBS domains: ligand binding sites and conformational variability. Arch Biochem Biophys 540,70–81. 44 Koutmos M, Kabil O, Smith JL & Banerjee R (2010) Structural basis for substrate activation and regulation 4952 The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. Altered PLP tautomerism impairs HsCBS-R336C activity C. Conter et al. 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License by cystathionine beta-synthase (CBS) domains in cystathionine b-synthase. Proc Natl Acad Sci U S A 107, 20958–20963. 45 Gim enez-Mascarell P, Majtan T, Oyenarte I, Ere~ no-Orbea J, Majtan J, Klaudiny J, Kraus JP & Mart ınez-Cruz LA (2018) Crystal structure of cystathionine b-synthase from honeybee Apis mellifera. J Struct Biol 202,82–93. 46 Devi S, Abdul Rehman SA, Tarique KF & Gourinath S (2017) Structural characterization and functional analysis of cystathionine b-synthase: an enzyme involved in the reverse transsulfuration pathway of Bacillus anthracis.FEBS J 284, 3862–3880. 47 Tu Y, Kreinbring CA, Hill M, Liu C, Petsko GA, McCune CD, Berkowitz DB, Liu D & Ringe D (2018) Crystal structures of cystathionine b-synthase from Saccharomyces cerevisiae: one enzymatic step at a time. Biochemistry 57, 3134–3145. 48 Fern andez-Rodr ıguez C, Oyenarte I, Conter C, Gonz alez-Recio I, N u~ nez-Franco R, Gil-Pitarch C, Quintana I, Jim enez-Os es G, Dominici P, MartinezChantar ML et al. (2021) Structural insight into the unique conformation of cystathionine b-synthase from toxoplasma gondii. Comput Struct Biotechnol J 19, 3542–3555. 49 Ere~ no-Orbea J, Majtan T, Oyenarte I, Kraus JP & Mart ınez-Cruz LA (2014) Structural insight into the molecular mechanism of allosteric activation of human cystathionine b-synthase by S-adenosylmethionine. Proc Natl Acad Sci U S A 111, E3845–E3852. 50 Ere~ no-Orbea J, Majtan T, Oyenarte I, Kraus JP & Mart ınez-Cruz LA (2013) Structural basis of regulation and oligomerization of human cystathionine b-synthase, the central enzyme of transsulfuration. Proc Natl Acad Sci U S A 110, E3790–E3799. 51 McCorvie TJ, Adamoski D, Machado RAC, Tang J, Bailey HJ, Ferreira DSM, Strain-Damerell C, Basl eA, Ambrosio ALB, Dias SMG et al. (2024) Architecture and regulation of filamentous human cystathionine beta-synthase. Nat Commun 15, 2931. 52 Pey AL, Mart ınez-Cruz LA, Kraus JP & Majtan T (2016) Oligomeric status of human cystathionine betasynthase modulates AdoMet binding. FEBS Lett 590, 4461–4471. 53 Conter C, Fruncillo S, Favretto F, Fern andez-Rodr ıguez C, Dominici P, Mart ınez-Cruz LA & Astegno A (2022) Insights into domain organization and regulatory mechanism of cystathionine Beta-synthase from Toxoplasma gondii.Int J Mol Sci 23, 8169. 54 Gim enez-Mascarell P, Oyenarte I, Hardy S, Breiderhoff T, Stuiver M, Kostantin E, Diercks T, Pey AL, Ere~ noOrbea J, Mart ınez-Chantar ML et al. (2017) Structural basis of the oncogenic interaction of phosphatase PRL1 with the magnesium transporter CNNM2. J Biol Chem 292, 786–801. 55 Gulerez I, Funato Y, Wu H, Yang M, Kozlov G, Miki H & Gehring K (2016) Phosphocysteine in the PRLCNNM pathway mediates magnesium homeostasis. EMBO Rep 17, 1890–1900. 56 G omez Garc ıa I, Oyenarte I & Mart ınez-Cruz LA (2011) Purification, crystallization and preliminary crystallographic analysis of the CBS pair of the human metal transporter CNNM4. Acta Crystallogr Sect F 67, 349–353. 57 Buey RM, Fern andez-Justel D, Marcos-Alcalde  I, Winter G, G omez-Puertas P, De Pereda JM & Luis Revuelta J (2017) A nucleotide-controlled conformational switch modulates the activity of eukaryotic IMP dehydrogenases. Sci Rep 7, 2648. 58 Gupta S, Lee H-O, Wang L & Kruger WD (2023) Examination of two different proteasome inhibitors in reactivating mutant human cystathionine b-synthase in mice. PLoS One 18, e0286550. 59 Singh LR, Gupta S, Honig NH, Kraus JP & Kruger WD (2010) Activation of mutant enzyme function in vivo by proteasome inhibitors and treatments that induce Hsp70. PLoS Genet 6, e1000807. 60 Lee H, Salami CO, Sondhi D, Kaminsky SM, Crystal RG & Kruger WD (2021) Long-term functional correction of cystathionine b-synthase deficiency in mice by adeno-associated viral gene therapy. J Inherit Metab Dis 44, 1382–1392. 61 Gupta S, Kelow S, Wang L, Andrake MD, Dunbrack RL & Kruger WD (2018) Mouse modeling and structural analysis of the p.G307S mutation in human cystathionine b-synthase (CBS) reveal effects on CBS activity but not stability. J Biol Chem 293, 13921– 13931. 62 Roman JV, Mascarenhas R, Ceric K, Ballou DP & Banerjee R (2023) Disease-causing cystathionine bsynthase linker mutations impair allosteric regulation. J Biol Chem 299, 105449. 63 Conter C, Fruncillo S, Fern andez-Rodr ıguez C, Mart ınez-Cruz LA, Dominici P & Astegno A (2020) Cystathionine b-synthase is involved in cysteine biosynthesis and H2S generation in Toxoplasma gondii. Sci Rep 10, 14657. 64 Pedretti M, Fern andez-Rodr ıguez C, Conter C, Oyenarte I, Favretto F, di Matteo A, Dominici P, Petrosino M, Martinez-Chantar ML, Majtan T et al. (2024) Catalytic specificity and crystal structure of cystathionine c-lyase from Pseudomonas aeruginosa.Sci Rep 14, 9364. 65 Conter C, Oppici E, Dindo M, Rossi L, Magnani M & Cellini B (2019) Biochemical properties and oxalatedegrading activity of oxalate decarboxylase from bacillus subtilis at neutral pH. IUBMB Life 71, 917– 927. 66 Conter C, Favretto F, Dominici P, Martinez-Cruz LA & Astegno A (2023) Key substrate recognition residues 4953The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. C. Conter et al. Altered PLP tautomerism impairs HsCBS-R336C activity 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License in the active site of cystathionine beta-synthase from Toxoplasma gondii.Proteins 91, 1383–1393. 67 McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn MD, Storoni LC & Read RJ (2007) Phaser crystallographic software. J Appl Cryst 40, 658–674. 68 Adams PD, Afonine PV, Bunk oczi G, Chen VB, Echols N, Headd JJ, Hung L-W, Jain S, Kapral GJ, Grosse Kunstleve RW et al. (2011) The phenix software for automated determination of macromolecular structures. Methods 55,94–106. 69 Afonine PV, Grosse-Kunstleve RW, Echols N, Headd JJ, Moriarty NW, Mustyakimov M, Terwilliger TC, Urzhumtsev A, Zwart PH & Adams PD (2012) Towards automated crystallographic structure refinement with phenix.Refine.Acta Crystallogr D 68, 352–367. 70 Emsley P, Lohkamp B, Scott WG & Cowtan K (2010) Features and development of coot.Acta Crystallogr D 66, 486–501. 71 Bayly CI, Cieplak P, Cornell W & Kollman PA (1993) A well-behaved electrostatic potential based method using charge restraints for deriving atomic charges: the RESP model. J Phys Chem 97, 10269–10280. 72 Frisch MJ, Trucks GW, Schlegel HB, Scuseria GE, Robb MA, Cheeseman JR, Scalmani G, Barone V, Petersson GA, Nakatsuji H et al. (2016) Gaussian 16, Revision C.01. Gaussian, Inc., Wallingford, CT. 73 Case DA, Aktulga HM, Belfon K, Cerutti DS, Cisneros GA, Cruzeiro VWD, Forouzesh N, Giese TJ, G€ otz AW, Gohlke H et al. (2023) AmberTools. J Chem Inf Model 63, 6183–6191. 74 Maier JA, Martinez C, Kasavajhala K, Wickstrom L, Hauser KE & Simmerling C (2015) ff14SB: improving the accuracy of protein side chain and backbone parameters from ff99SB. J Chem Theory Comput 11, 3696–3713. 75 Wang J, Wolf RM, Caldwell JW, Kollman PA & Case DA (2004) Development and testing of a general amber force field. J Comput Chem 25, 1157–1174. 76 Jorgensen WL, Chandrasekhar J, Madura JD, Impey RW & Klein ML (1983) Comparison of simple potential functions for simulating liquid water. J Chem Phys 79, 926–935. 77 Miyamoto S & Kollman PA (1992) Settle: an analytical version of the SHAKE and RATTLE algorithm for rigid water models. J Comput Chem 13, 952–962. 78 Darden T, York D & Pedersen L (1993) Particle mesh Ewald: An N log(N) method for Ewald sums in large systems. J Chem Phys 98, 10089–10092. 79 Grant BJ, Rodrigues APC, ElSawy KM, McCammon JA & Caves LSD (2006) Bio3d: an R package for the comparative analysis of protein structures. Bioinformatics 22, 2695–2696. 80 Pettersen EF, Goddard TD, Huang CC, Couch GS, Greenblatt DM, Meng EC & Ferrin TE (2004) UCSF chimera—a visualization system for exploratory research and analysis. J Comput Chem 25, 1605–1612. Supporting information Additional supporting information may be found online in the Supporting Information section at the end of the article. Fig. S1. Purity and quaternary structure of purified recombinant HsCBS-D516-525-WT and HsCBS-D516525-R336C. Fig. S2. Representative isothermal titration calorimetry (ITC) experiments monitoring the binding of AdoMet to HsCBS-D516-525-WT and HsCBS-D516-525-R336C at 25 °C and 37 °C. Fig. S3. PLP-neighborhood in HsCBS and HsCBSR336C. Fig. S4. HsCBS filaments. Table S1. Thermodynamic parameters for AdoMet binding determined by ITC. Table S2. Data collection and refinement statistics. Movie S1. Transversal displacement (TRANS) revealed by PCA analysis of the MD simulation on HsCBSD516-525-R336C. Movie S2. Synchronized Rotational motion (SYNROT) revealed by PCA analysis of the MD simulation on HsCBS-D516-525-R336C. Movie S3. MD analysis of the N149-PLP interaction in HsCBS-D516-525-R336C (basal conformation) at 300 K and 310 K. Movie S4. Three-dimensional structure of the HsCBSD516-525 filaments found in C2221 crystals (PDB ID 4L3V). Movie S5. Three-dimensional structure of the HsCBSD516-525 filaments found in I222 crystals (PDB ID 4L0D). Movie S6. Three-dimensional structure of full-length HsCBS filaments (basal conformation) found by Cryoelectron Microscopy (PDB ID 8S5I). Movie S7. Morphing showing three different views of the beta-sheet mediated counterclockwise rotation of the Bateman modules required to convert the filaments found in the C2221-crystals into the filaments found by cryo-EM. 4954 The FEBS Journal 292 (2025) 4933–4954 ª2025 Federation of European Biochemical Societies. Altered PLP tautomerism impairs HsCBS-R336C activity C. Conter et al. 17424658, 2025, 18, Downloaded from https://febs.onlinelibrary.wiley.com/doi/10.1111/febs.70116 by Spanish Cochrane National Provision (Ministerio de Sanidad), Wiley Online Library on [09/10/2025]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License