Next Article in Journal
The Potential of Natural Compounds in UV Protection Products
Previous Article in Journal
DNA Methylation in the CYP3A Distal Regulatory Region (DRR) Is Associated with the Expression of CYP3A5 and CYP3A7 in Human Liver Samples
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Molecules
Volume 29
Issue 22
10.3390/molecules29225408
molecules-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Crystallographic Fragment Screening of a Bifunctional Proline Catabolic Enzyme Reveals New Inhibitor Templates for Proline Dehydrogenase and L-Glutamate-γ-semialdehyde Dehydrogenase

by
Kaylen R. Meeks
1,
Alexandra N. Bogner
1,†,
Jay C. Nix
2 and
John J. Tanner
1,3,*
1
Department of Biochemistry, University of Missouri, Columbia, MO 65211, USA
2
Molecular Biology Consortium, Advanced Light Source, Lawrence Berkeley National Laboratory, Berkeley, CA 94720, USA
3
Department of Chemistry, University of Missouri, Columbia, MO 65211, USA
*
Author to whom correspondence should be addressed.
†
Current address: Lilly Biotechnology Center, Eli Lilly and Company, San Diego, CA 92121, USA.
Molecules 2024, 29(22), 5408; https://doi.org/10.3390/molecules29225408
Submission received: 23 October 2024 / Revised: 12 November 2024 / Accepted: 13 November 2024 / Published: 16 November 2024
(This article belongs to the Section Chemical Biology)
Browse Figures
Versions Notes

Abstract

The proline catabolic pathway consisting of proline dehydrogenase (PRODH) and L-glutamate-γ-semialdehyde (GSAL) dehydrogenase (GSALDH) catalyzes the four-electron oxidation of L-proline to L-glutamate. Chemical probes to these enzymes are of interest for their role in cancer and inherited metabolic disease. Here, we report the results of a crystallographic fragment-screening campaign targeting both enzymes. A unique aspect of our approach is the screening of both enzymes simultaneously using crystals of the bifunctional PRODH-GSALDH enzyme, proline utilization A (PutA). A 288-fragment library from Zenobia was screened in crystallo in cocktails of six fragments. Validation X-ray crystallography with individual fragments identified seven crystal hits distributed in the PRODH active site, GSALDH aldehyde substrate-binding site, and GSALDH NAD+ adenine-binding site. The fragment bound in the PRODH active site, 4-methoxybenzyl alcohol, is structurally distinct from all known PRODH inhibitors as it lacks an anionic anchor and stabilizes open conformations of the active site, motivating the study of eighteen analogs. In total, thirteen crystal structures with resolutions ranging from 1.32 Å to 1.80 Å were determined, resolving the poses and interactions of seven fragments from the Zenobia library and five analogs of 4-methoxybenzyl alcohol. These results expand the chemical space of probes targeting proline catabolic enzymes and provide new structural information for further inhibitor development.

Graphical Abstract

1. Introduction

Proline catabolism comprises the reactions catalyzed by proline dehydrogenase (PRODH) and L-glutamate-γ-semialdehyde (GSAL) dehydrogenase (GSALDH) [1]. PRODH is a flavoenzyme that catalyzes the FAD-dependent oxidation of L-proline to Δ1-pyrroline-5-carboxylate (P5C) (Figure 1). The hydrolysis of P5C generates GSAL, which is the substrate for GSALDH. Also known as ALDH4A1, GSALDH is a member of the aldehyde dehydrogenase (ALDH) superfamily and catalyzes the NAD+-dependent oxidation of GSAL to L-glutamate.
The arrangement of proline catabolic genes varies among organisms. In eukaryotes, the genes are distinct and encode separate PRODH and GSALDH enzymes. Both enzymes are localized to the mitochondria, with PRODH located in the inner mitochondrial membrane and GSALDH in the matrix. In contrast, in some bacteria, most notably Gram-negative bacteria, the PRODH and GSALDH genes are combined into the putA gene, which encodes a bifunctional enzyme, proline utilization A (PutA) [1].
PRODH is a target of inhibitor discovery because of its role in the metabolism of certain cancer cells. The role of PRODH in cancer is complicated and depends on cancer type and tumor microenvironment [2,3,4,5,6]. As a p53-induced anti-cancer protein, PRODH induces apoptosis and senescence through ROS signaling in different types of cancers, whereas PRODH promotes malignant phenotypes of certain tumors under stresses such as hypoxia. Breast cancer is an example of the latter. PRODH expression is upregulated in human metastatic breast tissue compared to primary breast tumor tissue [7]. Relevant to inhibitor discovery, the noncovalent PRODH inhibitor S-(-)-tetrahydro-2-furoic acid (THFA) was shown to inhibit the spheroidal growth of breast cancer cells and impair the formation of lung metastases in mouse models of breast cancer [7]. Also, mechanism-based inactivators that covalently modify the FAD of PRODH show activity in cancer cells [8,9].
The inhibition of GSALDH is also of interest. ALDHs are markers of cancer stem cells and may also play functional roles in the self-protection, differentiation, and/or expansion of stem cell populations [10]. Although ALDH1A1 and ALDH3A1 are most often associated with cancer stem cells [11,12], high ALDH4A1 expression has been reported in prostate cancer cells and primary cultures of human prostate cancer [13]. Inhibitors of ALDH4A1 could also potentially be used as pharmacological chaperones [14,15] to stabilize the misfolded variants of ALDH4A1 produced in patients with hyperprolinemia II, a metabolic disorder caused by mutations in the ALDH4A1 gene which can result in neurological problems, including intellectual disability [16,17].
Herein we report the results of a crystallographic fragment-screening campaign of proline catabolic enzymes. A unique aspect of our approach was the use of the bifunctional enzyme PutA as the target, which enabled the simultaneous screening of both PRODH and GSALDH. Crystals of PutA from Sinorhizobium meliloti (SmPutA, UniProt F7X6I3) were soaked with the Zenobia Express-Zen-Core 288™ library in cocktails of six fragments. Validation X-ray crystallography using twenty-four individual compounds (1–24 in Figure 2) revealed seven crystal hits distributed in the PRODH active site, GSAL-binding site, GSALDH NAD+-binding site, and surface sites (7, 9, 10, 11, 13, 19, and 20). The fragments in the GSAL site engage known elements of aldehyde substrate recognition and provide starting points for the development of competitive inhibitors of GSALDH. Fragment 20 in the PRODH active site is unique among PRODH inhibitors and motivated the study of eighteen analogs (25–42), which resulted in five additional crystal structures (25, 39–42). The inhibition of enzyme activity by the crystal hits was also tested. These results expand the chemical space of probes targeting proline catabolic enzymes and provide new structural information for further inhibitor development.

2. Results

2.1. Crystallographic Fragment Screening and Validation

Crystallographic fragment screening was performed by soaking crystals of SmPutA with compounds from the Zenobia Express-Zen-Core 288™ X-ray Mix of Six Fragment Screen. Each crystal was soaked for 48 h in a cocktail of six fragments at 10 mM of each compound (60 mM total fragment concentration). The soaks were performed in duplicate. Sixty-one X-ray diffraction data sets were collected. Data sets were not obtained for two wells (B1, H2) due to the deterioration of the crystals during soaking. The high-resolution limits of the data sets ranged from 1.53 Å to 2.98 Å with a median of 1.75 Å and a mean of 1.80 ± 0.25 Å. Eight data sets had high-resolution limits better than 2 Å. Visual inspection of the electron density maps suggested hits from the following wells of the library: A1, B4, D4, E3, E4, and H1 (the fragment cocktail compositions are provided in a spreadsheet in Supplementary Materials). The clearest hits were compounds 7 and 13, bound to the GSAL-binding site of the GSALDH active site. The other hits were characterized by ambiguous electron density features that required follow-up validation X-ray crystallography to identify the fragments and their poses.
The hits from cocktailed fragment screening were validated by soaking SmPutA crystals with individual compounds (1–24). These experiments resulted in the determination of eight crystal structures, resolving the poses and interactions of seven fragments, bound to the GSAL pocket of the GSALDH active site (7, 13, 20), the NAD+ adenine site in the GSALDH active site (9, 10, 11, 19), the PRODH active site (20), and surface sites (9). The high-resolution limits of the structures, binding locations of the fragments, and PDB deposition IDs are listed in Table 1. The complete X-ray diffraction data processing and refinement statistics are listed in Table S1.

2.2. Poses and Interactions of Fragments 7, 13, and 20 Bound to the GSAL-Binding Site

Fragments 7 and 13 were bound to the GSAL substrate-binding site of the GSALDH active site. The electron density for both fragments was unambiguous and supported the inclusion of the fragments at high occupancy (Figure 3). The refined occupancies of 7 were 0.83 and 0.86 in chains A and B of the asymmetric unit, respectively. Fragment 13 was modeled as a mixture of the R and S enantiomers, with occupancies of 0.55/0.45 (R/S) in chain A and 0.53/0.47 (R/S) in chain B.
Fragments 7 and 13 engaged well-known recognition elements of ALDH superfamily enzyme active sites (Figure 4A,B) [18,19]. The carboxylates formed hydrogen bonds with the backbone of the anchor loop (1000s loop in SmPutA). The carboxylates also formed an ion pair with Arg843 and a hydrogen bond with Ser845. The heterocycles were flanked by two Phe side chains, which formed the so-called aromatic box of ALDH active sites. All these recognition elements have also been observed in the structures of mouse GSALDH and SmPutA complexed with the product glutamate (PDB IDs 4V9K and 9BBO).
The fragments also exhibited interactions not found in the product complex. The dioxane ring of 7 forms a hydrogen bond with Arg843 (Figure 4A). The amine group of 13 donated a hydrogen bond to Glu674; both enantiomers showed this interaction (Figure 4B).
The fragments also differed from the substrate GSAL in that they did not occupy the oxyanion hole. In the glutamate complex, which is a model for the GSAL Michaelis complex, the side-chain carboxylate formed hydrogen bonds with a conserved asparagine (Asn707) and the backbone amino group of the catalytic cysteine (Cys844) (Figure 4C). These two interactions define the oxyanion hole of ALDH superfamily enzymes. The cyclic structures of 7 and 13 were perpendicular to the side chain of the glutamate product and, thus, could not access the oxyanion hole (Figure 4C).
One instance of fragment 20 was found in the GSALDH site. This feature was observed in only one chain of the two structures determined for this fragment (replicate structure 2 in Table 1 and Table S1). We note that the primary binding site for 20 is the PRODH site (vide infra). Nevertheless, the electron density for 20 in the GSALDH site was convincing and allowed modeling of the fragment at an occupancy of 0.71 in chain B (Figure 5A). Fragment 20 occupied the same space as the dioxane ring of 7 (Figure 5B). The hydroxyl of 20 forms a hydrogen bond to Arg843, while the phenyl ring was packed between one of the aromatic box residues (Phe708) and the aliphatic chain of Glu674 (Figure 5A). The O-methyl group contacts Ile712. Notably, 20 did not interact with the anchor loop and, thus, did not mimic the substrate GSAL, unlike 7 and 13.

2.3. Poses and Interactions of Fragments 9, 10, 11, and 19 Bound to the GSALDH NAD+-Binding Site

Fragments 9, 10, 11, and 19 bound to the GSALDH NAD+-binding site and, more specifically, overlap with the adenine group of NAD+ (Table 1). Fragments 9 and 10 were modeled into this site with high occupancy in both chains of the asymmetric unit (Q = 0.85–1.00). Convincing electron density for 11 was observed only in chain A, and the fragment was modeled with an occupancy of 1.00. The density for 19 was strong in both chains but lacked detailed features to support a single-binding mode. Therefore, this fragment was modeled in two plausible conformations in each chain, with a total occupancy of 0.9 per chain.
Mimicking the adenine of NAD+, fragments 9, 10, 11, and 19 were wedged between two α-helices of the Rossmann dinucleotide-binding domain (Figure 6A). One edge of the fragment made contact with β1 of the Rossman fold, while the other edge was exposed to solvent. The latter was analogous to the N6-N7 edge of adenine in NAD+.
The fragments in the adenine site formed relatively few electrostatic interactions with the enzyme. The carboxylate of 9 formed an ion pair with Lys730 (Figure 6B, right panel) and hydrogen bonds with water molecules. Fragments 10 and 11 formed hydrogen bonds only with water molecules.
As mentioned above, 19 was modeled in two conformations. In chain A, the two conformations differed by a rotation of 180° that swapped the thiourea and fluorine groups (Figure 6C, left). In chain B, the two conformations differed in the position of the thiourea groups (Figure 6C, right). In three of the four poses, the F atom formed hydrogen bonds with water molecules, while, in one of the poses, the F atom formed hydrogen bonds with Lys730.

2.4. Poses and Interactions of Fragment 9 Bound to a Surface Site

Electron density for two copies of 9 was observed on the surface of the protein in a crystal contact region where the NAD+-binding domain of one protein met the PRODH domain of another protein. We note that this interface is not present in the in-solution dimer, which was validated previously by small-angle X-ray scattering [20]. The two surface-bound fragments (9’ (Q = 0.88) and 9” (Q = 0.80) in Figure 6B) were packed between the instance of 9 in the adenine site and the PRODH domain of a symmetry-related molecule. One of the surface fragments participated in the octahedral coordination of a metal ion, which was modeled as Mg2+ since the crystallization solution contained 0.1 M MgCl2 (9’ in Figure 6B). The coordination sphere of the Mg2+ ion consisted of O and N atoms of 9’ plus four water molecules. Although N is less common than O as a Mg2+ ligand [21], searching the PDB with the MESPEUS server reveals many high-resolution examples of histidine side chains coordinating Mg2+ (e.g., 3AYX) [22]. The numerous crystal contacts stabilizing 9’ and 9” suggest that these sites are most likely artifacts of crystallization.

2.5. Poses and Interactions of Fragment 20 Bound to the PRODH Active Site

Fragment 20 was bound to the PRODH active site. Interestingly, the electron density for 20 in this location was not evident in the maps obtained from soaking crystals with fragment cocktails. Instead, diffuse electron density features presumably representing multiple compounds from well H1 were observed in the GSALDH active site. However, upon soaking crystals with 20 alone at the higher concentration of 67 mM, a clear electron density feature was observed in the PRODH active site (Figure 7A).
Two replicate structures of SmPutA complexed with 20 were determined. Replicate structure 1 (1.64 Å resolution) was obtained by soaking a crystal with 67 mM of 20 for 24 h. Upon observing a new PRODH active site conformation, we also co-crystallized SmPutA with 20 mM 20 and then included 67 mM 20 during cryoprotection (replicate structure 2, 1.32 Å resolution); the exposure of the crystal to 67 mM 20 during cryoprotection was brief (a few minutes). The fragment was modeled with occupancies of 0.66 and 0.78 in replicate structure 1 and 0.68 and 0.76 in replicate structure 2.
The poses of fragment 20 in the two structures were similar. Fragment 20 was bound to the L-proline substrate-binding site of the PRODH active site (Figure 7A). One face of the benzene ring stacked parallel against the si face of the FAD isoalloxazine. The other face of 20 made contact with the side chains of conserved lysine and arginine residues but was mostly unobstructed and exposed to a solvent-filled cavity that led to the GSALDH active site. The hydroxyl of 20 occupied the binding site of a conserved water molecule observed in PRODH structures. Apparently, the binding of 20 displaced the conserved water molecule. The hydroxyl of 20 formed hydrogen bonds to the phenol of Tyr373 and the carbonyl of Ala307 (Figure 7A). The methyl group of 20 occupied a hydrophobic pocket near a conserved leucine (Leu449). One edge of 20 made contact with the roof of the active site formed by Asp306 and Tyr473, while the other edge made contact with the O2’ hydroxyl of the FAD ribityl chain.
The FAD of PRODH is known to adopt different conformations in response to ligand binding and reduction [23]. The dominant FAD conformation in the presence of 20 appears to be the one associated with the binding of proline analog inhibitors and, by inference, the substrate L-proline. In this conformation (Michaelis complex conformation), the isoalloxazine, which was oxidized, was planar, and the O2’ hydroxyl of the ribityl chain pointed to the proline-binding site (Figure 7A). The O3’ and O4’ hydroxyls were located below the dimethylbenzene and pyrimidine parts of isoalloxazine, respectively. This conformation was observed at full crystallographic occupancy in replicate structure 1. Two conformations were observed in replicate structure 2. The Michaelis complex conformation was modeled with an occupancy of 0.59/0.72 (chains A/B), while the conformation associated with the unoccupied proline site was modeled at an occupancy of 0.41/0.28. This second FAD conformation differed by a crankshaft rotation of the ribityl chain that moved the O2’ hydroxyl below the pyrimidine ring of the isoalloxazine (Figure 7A, right-hand panel).
Although 20 occupied the proline site and stabilized the Michaelis complex FAD conformation, it did not fully mimic the substrate L-proline. Figure 7B compares the structures of SmPutA complexed with 20 and the proline analog THFA. Major differences between the two structures includes the side-chain rotamer of Leu449 and helix α8. In the THFA complex, Leu449 adopted rotamer #1, which allowed the close packing of the side chain against the aliphatic portion of the tetrahydrofuran ring of THFA. In contrast, in the complex with 20, Leu449 flipped to rotamer #2 to avoid a steric clash with the O-methyl group of the fragment.
The differences in α8 conformations were substantial. This helix is a key structural element of the active site. In the THFA complex, as well as in other proline analog complexes, α8 was highly ordered and contributed three conserved side chains that made contact with the proline analog: Arg488 and Arg489 ion pair with the carboxylate of THFA, while Tyr485 was packed against the aliphatic portion of the tetrahydrofuran ring (Figure 7B). These interactions closed the active site and buried the proline analog. In contrast, the electron density for α8 in the complexes with 20 was weak, implying conformational flexibility. As a result, several side chains and entire residues of α8 were omitted from the final models. Thus, the active site when 20 was bound appeared to be more flexible and open compared to the THFA complex. Thus, the fragment screen revealed a new PRODH active site conformation.

2.6. Inhibition of GSALDH Activity by Hits from Crystallographic Screening

The seven hits from the crystallographic screening of the Zenobia library were tested for the inhibition of GSALDH activity. The fragments were included in the assay at 0.25–5 mM, depending on compound solubility or spectral interference at 340 nm. The most notable inhibition occurred with 7 and 13, whose poses and interactions mimicked the substrate GSAL (Figure 8A). Then, 7 and 13 were further tested in assays performed with varying P5C/GSAL and fragment concentrations at a fixed NAD+ concentration of 0.2 mM. The data were fitted to a global competitive inhibition model, resulting in Ki values of 0.32 mM ± 0.02 mM for 7 and 1.5 mM ± 0.1 mM for 13 (Figure 8B,C).

2.7. Crystal Structures of SmPutA Complexed with Analogs of 20

Compound 20 is chemically different from other PRODH inhibitors, which are proline analogs with an anionic anchor, and the structure of SmPutA complexed with 20 revealed a new, open PRODH active site conformation. We therefore explored this new class of PRODH binder by studying analogs of 20. Eighteen analogs were used in the crystallization experiments (25–42 in Figure 2). Five structures were determined at high-resolution limits of 1.32–1.47 Å, resolving the poses of compounds 25, 39, 40, 41, and 42 in the PRODH active site (see Table S1 for the crystallographic statistics). The electron density for these compounds was strong and covered the entire ligand (Figure 9). For all but compound 25, the ligand was modeled in both chains in the asymmetric unit; compound 25 was modeled only in chain B. The occupancies of 39–42 ranged from 0.67 to 0.88; the occupancy of 25 was 0.68.
Compound 25 was isostructural to 20, except 25 had two alcohol groups, whereas 20 had one alcohol group and a methoxy group. Interestingly, this minor difference affected the interactions of the fragment with the enzyme. Indeed, 25 was bound to the same location as 20, stacking in parallel against the si face of the FAD isoalloxazine (Figure 9A). As with replicate structure 2 of fragment 20, the electron density for the FAD was consistent with two conformations. One of the alcohol groups of 25 bound to the conserved water site, while the other was in the nonpolar pocket. The conformation of the former alcohol group was different from that of 20. While the hydroxyl of 20 was positioned optimally to form hydrogen bonds with Ala307 and Tyr373, the hydroxyl of 25 was rotated by 180° so that it would form hydrogen bonds only with Ala307. In this conformation, the hydroxyl could not form a hydrogen bond with Tyr373, and, furthermore, the adjacent methylene group clashed with the hydroxyl of Tyr373 (C–O distance of 3.0 Å). The other hydroxyl of 25 (the one in the nonpolar pocket) did not form a hydrogen bond with the enzyme. Thus, the replacement of the methoxy of 20 with an alcohol group, paradoxically, diminished hydrogen bonding with the enzyme.
Compounds 39–42 were structurally similar bicyclic 5-6 systems consisting of benzyl alcohol fused to a five-membered ring heterocycle containing either O (39–41) or N (42). As with 20 and 25, the ring system of these compounds stacked against the si face of the FAD isoalloxazine. Unexpectedly, the hydroxyls of 39-41 did not occupy the conserved water site and instead pointed to the nonpolar pocket near Leu449 (Figure 9B–D). The hydroxyl groups of these compounds did not form hydrogen bonds to the enzyme, similar to 25.
The pose of 42 was distinctly different from the other bicyclic 5-6 compounds. Compound 42 was flipped by 180° so its hydroxyl group occupied the conserved water pocket, and the five-membered heterocycle occupied the nonpolar pocket (Figure 9E). The hydroxyl of 42 formed hydrogen bonds with both the carbonyl of Ala307 and the phenol of Tyr373. Thus, the alcohol group of this compound was optimally oriented for hydrogen bonding, similar to 20. We validated the modeling of 42 by refining four alternative “flipped” poses in which the hydroxyl group was in the nonpolar pocket, like 39–41 (Figure S1). The flipped poses had a lower correlation coefficient with the polder map, calculated as CC(1,3), indicating a worse fit to the map [24]. For example, CC(1,3) degrades from 0.92 for the pose in Figure 9E to 0.82–0.86 for the flipped poses. Also, the refined occupancies of the fragment decreased in three of the four alternative refinements (Figure S1). Furthermore, flipping the fragment resulted in strong positive and negative Fo-Fc features on the ligand (Figure S1). These results validated the pose of 42 shown in Figure 9E.

2.8. Inhibition of PRODH Activity by 20 and Its Analogs

Fragment 20 and twelve analogs were tested for the inhibition of PRODH activity. The analogs tested included the five strong crystal hits plus seven others for which a weak electron density suggested the possibility of binding but was insufficient for structure determination. An assay with the known inhibitor THFA was included as a positive control. Each compound was included in the assay at 5 mM, except for 41, which was used at 3 mM due to solubility issues. THFA lowered PRODH activity to 10% relative to having no inhibitor (Figure 10). Fragment 42 produced a notable lowering in activity to 25%.

3. Discussion

Here, we described a crystallographic fragment-screening campaign of the proline catabolic enzymes PRODH and GSALDH. A unique aspect of this study was the simultaneous screening of two enzymes in a single crystal, which was possible using the bifunctional PRODH-GSALDH enzyme, PutA. We identified binders of both PRODH and GSALDH. All the meaningful crystallographic hits targeted an active site, either the proline-binding site of PRODH, the NAD+ adenine site of GSALDH, or the GSAL substrate site of GSALDH (Table 1). One fragment (9) was bound to a remote site; however, this association with the protein appeared to have been enabled by crystal packing interactions.
The two fragments that bound to the GSAL site (7, 13) were notable in that they inhibited enzyme activity. We note that it is common in crystallographic fragment screening to find compounds that bind to the active site and yet do not inhibit enzyme activity, presumably because the affinity is below the detection limit of the enzyme activity assay [25,26]. The competitive inhibition constants of 0.32 mM for 7 and 1.5 mM for 13 were better than those of proline and hydroxyproline stereoisomers, which were in the range of 1.3–30 mM [27]. Possible strategies for developing new GSALDH inhibitors include testing analogs of 7 and 13 and template docking, a virtual ligand screening approach which leverages high-resolution crystal structures of target–fragment complexes to reveal key interactions in the active site which are used to seed docking [28].
Our work provides a proof of concept for using crystallographic fragment screening to develop inhibitors of ALDH superfamily enzymes. The human ALDH superfamily comprises 19 enzymes, which catalyze the NAD+-dependent oxidation of numerous substrates, including small aldehydes such as acetaldehyde, amino acid derivatives, and lipids [29,30,31]. ALDHs play important roles in the detoxification of reactive aldehydes, amino acid metabolism, embryogenesis and development, neurotransmission, oxidative stress, and cancer. Some ALDH isoforms are being studied as potential cancer therapy targets [11,32]. Also, mutations in certain ALDH genes are associated with inherited metabolic diseases. Based on our experience here with GSALDH, fragment screening could be a productive strategy for developing new inhibitors and pharmacological chaperones for other ALDHs.
The discovery of fragment 20 as a PRODH binder is potentially significant in chemical probe development targeting this enzyme. All previously known PRODH inhibitors feature an anionic anchor, which interacts with conserved arginine and lysine residues. Most of these inhibitors have a carboxylate anchor and are close proline analogs, with the best example being THFA [33]. We recently showed that the sulfonate group also serves as an anchor, as demonstrated by the structure of PRODH complexed with 1-hydroxyethane-1-sulfonate [34]. Inhibitors with anionic anchors stabilize a tightly closed conformation of the active site in which the inhibitor is surrounded by several side chains and the FAD isoalloxazine. The congested nature of the closed active site restricts the size of compounds that can be accommodated, and the elaboration of proline analogs as an inhibitor discovery strategy has been challenging. Fragment 20 and its analogs (25, 39–42) are notable because they lack an anionic anchor and are bound to an open form of the active site. The latter feature suggests that it may be possible to identify larger analogs of these compounds with improved affinity.
Interest in 20 as a novel PRODH binder motivated the study of several analogs. Indeed, 20 has limited interactions with the enzyme, which is expected for a fragment [35,36]. Although limited in number, the interactions formed by fragments are thought to be of a “high quality” and provide a basis for fragment-to-lead optimization. The high quality interactions of 20 appear to be the hydrogen bonds with a backbone carbonyl and tyrosine side chain. These interactions are possible because the hydroxyl occupies the conserved water site. Additionally, displacement of the conserved water molecule may be considered another high-quality interaction, which is entropic in origin. Fragment 42 also forms these hydrogen bonds. Surprisingly, 25, 39, 40, and 41 do not form these bonds, owing to a 180° rotation of the alcohol group of 25 and flipping of the other compounds so that the hydroxyl occupies the nonpolar pocket instead of the conserved water site. The notable inhibition of PRODH activity by 42 supports the notion that the aforementioned hydrogen bonds are important for binding. Testing analogs of 42 and template docking based on the SmPutA-42 structure could be promising inhibitor design strategies.
Finally, our experience with fragment 20 demonstrates a potential pitfall of cocktailing fragments. We used a screen in which crystals were soaked with solutions containing six fragments, with each fragment present at 10 mM. In the initial cocktail soak, the electron density for 20 in the PRODH site was not evident. Instead, a diffuse electron density was observed in the GSALDH active site. Surprisingly, during follow-up crystal-soaking studies using compound 20 alone at 67 mM, we observed strong and reproducible electron density for the fragment in the PRODH active site, which led to a novel PRODH inhibitor, compound 42. These results support the assertion of Schiebel et al. that the use of fragment cocktails can lead to artifacts such as reduced occupancy, crystal damage, and interfragment reactions, which diminish the success rate of fragment screening [25].

4. Materials and Methods

4.1. Protein Production

The bifunctional PRODH-GSALDH enzyme, proline utilization A (PutA), from Sinorhizobium meliloti (SmPutA, UniProt F7X6I3) was used in the X-ray crystallography and enzyme activity assays. SmPutA was expressed in Escherichia coli BL21(DE3) cells and purified using immobilized metal affinity chromatography, followed by anion exchange chromatography and size-exclusion chromatography, using procedures similar to those described previously [20,23,27,37]. The N-terminal His tag was removed with TEVP after the affinity chromatography step. The buffer used in size-exclusion chromatography contained 50 mM Tris, 150 mM NaCl, 0.5 mM EDTA, and 0.5 mM TCEP at pH 7.5. Fractions from size-exclusion chromatography were concentrated to 12 mg/mL, distributed into 50 µL aliquots, flash-frozen in liquid nitrogen, and stored at −80°C.

4.2. Crystallization, Crystal Soaking, and Co-Crystallization

Fragment screening was performed by soaking crystals of SmPutA. The crystals were grown using procedures similar to those described previously [20,23,27,37]. Briefly, SmPutA (6 mg/mL) was incubated with 5 mM NAD+ and used in sitting drop vapor diffusion crystallography at 13 °C. The reservoir solution contained 0.1–0.4 M ammonium sulfate, 0.1 M sodium formate, 0.1 M MgCl2, 0.1 M HEPES at pH 7.5, and 16–26% (w/v) PEG 3350. Crystal drops were streaked with SmPutA microcrystals. A minimum of 96 crystals were needed to perform the entire fragment screen in duplicate.
The Zenobia Express-Zen-Core 288™ X-ray Mix of Six Fragment Screen (Zenobia Fragments, a division of Zenobia Therapeutics, San Diego, CA, USA) was purchased in a 48-well tray format. Each well contained the dried-down residue of six fragments. The names and SMILES of the compounds are provided in a spreadsheet in Supplementary Materials. The fragments were resuspended by adding 10 µL of crystal stabilization buffer containing 0.2 M ammonium sulfate, 0.1 M sodium formate, 0.1 M MgCl2, 0.1 M HEPES at pH 7.5, and 20% (w/v) PEG 3350, so that the final fragment concentration was 10 mM per fragment (60 mM total of the six fragments). Two SmPutA crystals were transferred into each well and soaked for 48 h at 13 °C. The soaked crystals were flash-cooled in liquid nitrogen after the addition of 13% (v/v) PEG 200.
To validate the fragments contributing to ligand electron density features in the fragment screen, individual compounds were purchased and used in crystal soaking and co-crystallization. A total of 24 compounds were tested in these follow-up experiments (compounds 1–24 in Figure 2). In addition, eighteen analogs of 20 were also studied (25–42 in Figure 2). The compound concentrations used in soaking and co-crystallization are listed in Table S2. Except for 11 and 42, the compounds were dissolved in 1 M Tris pH 8.5. Compound 11 was dissolved in 100% DMSO and then diluted into the cryo-buffer for crystal soaking. Compound 42 was dissolved in 50% (w/v) PEG 3350 at 200 mM for the crystallization studies to overcome the lack of solubility in Tris buffer.

4.3. X-Ray Crystal Structure Determination

Shutterless X-ray diffraction data were collected at the NECAT beamlines 24-ID-C and 24-ID E of the Advanced Photon Source and beamlines 4.2.2 and 8.2.1 of the Advanced Light Source. The data were processed with automated processing pipelines, which use XDS for indexing, integration, and scaling [38], POINTLESS for space group determination [39], AIMLESS for merging [40], and CTRUNCATE for the conversion of intensities to amplitudes and the generation of a test set of reflections (2% or 5%) for Rfree calculations [41]. The space group is P21 with a dimer in the asymmetric unit.
The initial phases were calculated by Fourier synthesis in Phenix [42] using a model derived from PDB entry 5KF6 or 6X9A. Phenix was used for crystallographic refinement. The B-factor model consisted of one TLS group per protein chain and isotropic B-factors for non-hydrogen atoms. Coot was used for iterative model building [43]. Electron density features representing fragments were identified by manual inspection of the 2Fo-Fc and Fo-Fc maps in both active sites, as well as by analysis with the blob picker and difference map peak tools in Coot. The data processing and refinement statistics are listed in Table S1.

4.4. Enzyme Inhibition Assays

Enzyme activity assays were performed in 96-well plates in a BioTek Epoch 2 microplate spectrophotometer (Agilent, Santa Clara, CA, USA) at 23 °C. The total assay volume was 200 µL. The PRODH activity assay measured the dihydroquinazolinium compound formed by the reaction of the product P5C with ortho-aminobenzaldehyde (o-AB) by absorbance at 443 nm (ε443 = 2.59 mM−1cm−1). The electron acceptor menadione was included in the assay to allow for the catalytic cycling of PRODH by oxidizing the reduced FAD. The assays contained final concentrations of 4 mM o-AB and 0.1 mM menadione in a buffer of 20 mM MOPS pH 7.5 and 10 mM MgCl2. For the analysis of inhibition, 25 mM L-proline and the inhibitor were pipetted into the 96-well plate and then a master mix including enzyme, menadione, and o-AB in the assay buffer was added to the plate by a multichannel pipette to initiate the reaction. The final concentration of enzyme in the assay was 200 nM. The initial rates were estimated from a linear regression of the first 30 min of the progress curve using Origin 2020 version 9.7.0.188 software.
The GSALDH activity of SmPutA in the presence of inhibitors was measured by monitoring NADH production at 340 nm with L-P5C as the substrate. The assay buffer contained 100 mM sodium phosphate at pH 7 and 1 mM EDTA. The SmPutA concentration in each assay was 230 nM. The initial rates were estimated from the linear regression of the first 5 min of the progress curve. To estimate the inhibition constants, L-P5C was the variable substrate, and the initial rate as a function of both the variable substrate and the inhibitor concentrations was fitted globally to the standard model of competitive inhibition using Origin 2020 version 9.7.0.188 software.

Supplementary Materials

The following supporting information can be downloaded at https://www.mdpi.com/article/10.3390/molecules29225408/s1, Table S1: X-ray diffraction data collection and refinement statistics; Table S2: Concentrations of inhibitors used in soaking, co-crystallization, and cryoprotection; Figure S1: Refinement of four flipped poses of 42; and Spreadsheet with the names and SMILES of Zenobia fragments.

Author Contributions

K.R.M.: conceptualization, methodology, investigation, writing—original draft, writing—review and editing, and visualization. A.N.B.: conceptualization, methodology, and investigation. J.C.N.: investigation. J.J.T.: conceptualization, writing—original draft, writing—review and editing, validation, supervision, project administration, and funding acquisition. All authors have read and agreed to the published version of the manuscript.

Funding

The research reported in this publication was supported by the National Institute of General Medical Sciences of the National Institutes of Health under award number R01GM132640. K.R.M. was supported in part by a Wayne L. Ryan Fellowship through the Ryan Foundation.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

The coordinates and structure factor amplitudes for the crystal structures were deposited in the Protein Data Bank with the following accession codes: 7, 9DL2; 9, 9DL3; 10, 9DL4; 11, 9DL5; 13, 9DL6; 19, 9DL7; 20 (replicate 1), 9DL8; 20 (replicate 2), 9DL9; 25, 9E0A; 39, 9E0B; 40, 9E0C; 41, 9E0D; and 42, 9E0E.

Acknowledgments

We thank Jonathan Schuerman for their help with X-ray diffraction data collection and processing at NECAT. We thank Oseeyi Daudu from the Becker lab at the University of Nebraska for supplying D,L-P5C. This work is based upon research conducted at the Northeastern Collaborative Access Team beamlines, which are funded by the National Institute of General Medical Sciences from the National Institutes of Health (P30 GM124165). The Eiger 16M detector on 24-ID-E is funded by an NIH-ORIP HEI grant (S10OD021527). This research used resources of the Advanced Photon Source, a U.S. Department of Energy (DOE) Office of Science User Facility operated for the DOE Office of Science by Argonne National Laboratory under contract no. DE-AC02-06CH11357. Beamlines 4.2.2 and 8.2.1 of the Advanced Light Source, a DOE Office of Science User Facility under contract no. DE-AC02-05CH11231, are supported in part by the ALS-ENABLE program funded by the National Institutes of Health, National Institute of General Medical Sciences, grant P30 GM124169-01.

Conflicts of Interest

Author Alexandra N. Bogner was employed by the company Eli Lilly and Company. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

References

  1. Tanner, J.J. Structural Biology of Proline Catabolic Enzymes. Antioxid. Redox Signal. 2019, 30, 650–673. [Google Scholar] [CrossRef] [PubMed]
  2. Phang, J.M. Proline Metabolism in Cell Regulation and Cancer Biology: Recent Advances and Hypotheses. Antioxid. Redox Signal. 2019, 30, 635–649. [Google Scholar] [CrossRef] [PubMed]
  3. Phang, J.M.; Liu, W. Proline metabolism and cancer. Front. Biosci. 2012, 17, 1835–1845. [Google Scholar] [CrossRef] [PubMed]
  4. Phang, J.M.; Liu, W.; Hancock, C.; Christian, K.J. The proline regulatory axis and cancer. Front. Oncol. 2012, 2, 60. [Google Scholar] [CrossRef]
  5. Burke, L.; Guterman, I.; Palacios Gallego, R.; Britton, R.G.; Burschowsky, D.; Tufarelli, C.; Rufini, A. The Janus-like role of proline metabolism in cancer. Cell Death Discov. 2020, 6, 104. [Google Scholar] [CrossRef]
  6. Xu, X.; Zhang, G.; Chen, Y.; Xu, W.; Liu, Y.; Ji, G.; Xu, H. Can proline dehydrogenase-a key enzyme involved in proline metabolism-be a novel target for cancer therapy? Front. Oncol. 2023, 13, 1254439. [Google Scholar] [CrossRef]
  7. Elia, I.; Broekaert, D.; Christen, S.; Boon, R.; Radaelli, E.; Orth, M.F.; Verfaillie, C.; Grunewald, T.G.P.; Fendt, S.M. Proline metabolism supports metastasis formation and could be inhibited to selectively target metastasizing cancer cells. Nat. Commun. 2017, 8, 15267. [Google Scholar] [CrossRef]
  8. Scott, G.K.; Mahoney, S.; Scott, M.; Loureiro, A.; Lopez-Ramirez, A.; Tanner, J.J.; Ellerby, L.M.; Benz, C.C. N-Propargylglycine: A unique suicide inhibitor of proline dehydrogenase with anticancer activity and brain-enhancing mitohormesis properties. Amino Acids 2021, 53, 1927–1939. [Google Scholar] [CrossRef]
  9. Scott, G.K.; Yau, C.; Becker, B.C.; Khateeb, S.; Mahoney, S.; Jensen, M.B.; Hann, B.; Cowen, B.J.; Pegan, S.D.; Benz, C.C. Targeting Mitochondrial Proline Dehydrogenase with a Suicide Inhibitor to Exploit Synthetic Lethal Interactions with p53 Upregulation and Glutaminase Inhibition. Mol. Cancer Ther. 2019, 18, 1374–1385. [Google Scholar] [CrossRef]
  10. Muzio, G.; Maggiora, M.; Paiuzzi, E.; Oraldi, M.; Canuto, R.A. Aldehyde dehydrogenases and cell proliferation. Free Radic. Biol. Med. 2012, 52, 735–746. [Google Scholar] [CrossRef]
  11. Muralikrishnan, V.; Hurley, T.D.; Nephew, K.P. Targeting Aldehyde Dehydrogenases to Eliminate Cancer Stem Cells in Gynecologic Malignancies. Cancers 2020, 12, 961. [Google Scholar] [CrossRef] [PubMed]
  12. Bu, J.; Zhang, Y.; Wu, S.; Li, H.; Sun, L.; Liu, Y.; Zhu, X.; Qiao, X.; Ma, Q.; Liu, C.; et al. KK-LC-1 as a therapeutic target to eliminate ALDH(+) stem cells in triple negative breast cancer. Nat. Commun. 2023, 14, 2602. [Google Scholar] [CrossRef] [PubMed]
  13. van den Hoogen, C.; van der Horst, G.; Cheung, H.; Buijs, J.T.; Lippitt, J.M.; Guzman-Ramirez, N.; Hamdy, F.C.; Eaton, C.L.; Thalmann, G.N.; Cecchini, M.G.; et al. High aldehyde dehydrogenase activity identifies tumor-initiating and metastasis-initiating cells in human prostate cancer. Cancer Res. 2010, 70, 5163–5173. [Google Scholar] [CrossRef] [PubMed]
  14. Convertino, M.; Das, J.; Dokholyan, N.V. Pharmacological Chaperones: Design and Development of New Therapeutic Strategies for the Treatment of Conformational Diseases. ACS Chem. Biol. 2016, 11, 1471–1489. [Google Scholar] [CrossRef] [PubMed]
  15. McCorvie, T.J.; Yue, W.W. Chapter 13—Structure-guided discovery of pharmacological chaperones targeting protein conformational and misfolding diseases. In Protein Homeostasis Diseases; Academic Press: Cambridge, MA, USA, 2020; pp. 281–308. [Google Scholar]
  16. Geraghty, M.T.; Vaughn, D.; Nicholson, A.J.; Lin, W.W.; Jimenez-Sanchez, G.; Obie, C.; Flynn, M.P.; Valle, D.; Hu, C.A. Mutations in the Delta1-pyrroline 5-carboxylate dehydrogenase gene cause type II hyperprolinemia. Hum. Mol. Genet. 1998, 7, 1411–1415. [Google Scholar] [CrossRef]
  17. Phang, J.M.; Hu, C.A.; Valle, D. Disorders of proline and hydroxyproline metabolism. In Metabolic and Molecular Basis of Inherited Disease; Scriver, C.R., Beaudet, A.L., Sly, W.S., Valle, D., Eds.; McGraw Hill: New York, NY, USA, 2001; pp. 1821–1838. [Google Scholar]
  18. Riveros-Rosas, H.; Gonzalez-Segura, L.; Julian-Sanchez, A.; Diaz-Sanchez, A.G.; Munoz-Clares, R.A. Structural determinants of substrate specificity in aldehyde dehydrogenases. Chem. Biol. Interact. 2013, 202, 51–61. [Google Scholar] [CrossRef]
  19. Pemberton, T.A.; Tanner, J.J. Structural basis of substrate selectivity of Delta(1)-pyrroline-5-carboxylate dehydrogenase (ALDH4A1): Semialdehyde chain length. Arch. Biochem. Biophys. 2013, 538, 34–40. [Google Scholar] [CrossRef]
  20. Luo, M.; Gamage, T.T.; Arentson, B.W.; Schlasner, K.N.; Becker, D.F.; Tanner, J.J. Structures of Proline Utilization A (PutA) Reveal the Fold and Functions of the Aldehyde Dehydrogenase Superfamily Domain of Unknown Function. J. Biol. Chem. 2016, 291, 24065–24075. [Google Scholar] [CrossRef]
  21. Harding, M.M. The architecture of metal coordination groups in proteins. Acta Cryst. 2004, D60 Pt 5, 849–859. [Google Scholar] [CrossRef]
  22. Lin, G.Y.; Su, Y.C.; Huang, Y.L.; Hsin, K.Y. MESPEUS: A database of metal coordination groups in proteins. Nucleic Acids Res 2024, 52, D483–D493. [Google Scholar] [CrossRef]
  23. Campbell, A.C.; Becker, D.F.; Gates, K.S.; Tanner, J.J. Covalent Modification of the Flavin in Proline Dehydrogenase by Thiazolidine-2-Carboxylate. ACS Chem. Biol. 2020, 15, 936–944. [Google Scholar] [CrossRef] [PubMed]
  24. Liebschner, D.; Afonine, P.V.; Moriarty, N.W.; Poon, B.K.; Sobolev, O.V.; Terwilliger, T.C.; Adams, P.D. Polder maps: Improving OMIT maps by excluding bulk solvent. Acta Crystallogr. D Struct. Biol. 2017, 73 Pt 2, 148–157. [Google Scholar] [CrossRef]
  25. Schiebel, J.; Radeva, N.; Krimmer, S.G.; Wang, X.; Stieler, M.; Ehrmann, F.R.; Fu, K.; Metz, A.; Huschmann, F.U.; Weiss, M.S.; et al. Six Biophysical Screening Methods Miss a Large Proportion of Crystallographically Discovered Fragment Hits: A Case Study. ACS Chem. Biol. 2016, 11, 1693–1701. [Google Scholar] [CrossRef] [PubMed]
  26. Meeks, K.R.; Ji, J.; Protopopov, M.V.; Tarkhanova, O.O.; Moroz, Y.S.; Tanner, J.J. Novel Fragment Inhibitors of PYCR1 from Docking-Guided X-ray Crystallography. J. Chem. Inf. Model 2024, 64, 1704–1718. [Google Scholar] [CrossRef]
  27. Campbell, A.C.; Bogner, A.N.; Mao, Y.; Becker, D.F.; Tanner, J.J. Structural analysis of prolines and hydroxyprolines binding to the l-glutamate-gamma-semialdehyde dehydrogenase active site of bifunctional proline utilization A. Arch. Biochem. Biophys. 2021, 698, 108727. [Google Scholar] [CrossRef]
  28. Muller, J.; Klein, R.; Tarkhanova, O.; Gryniukova, A.; Borysko, P.; Merkl, S.; Ruf, M.; Neumann, A.; Gastreich, M.; Moroz, Y.S.; et al. Magnet for the Needle in Haystack: “Crystal Structure First” Fragment Hits Unlock Active Chemical Matter Using Targeted Exploration of Vast Chemical Spaces. J. Med. Chem. 2022, 65, 15663–15678. [Google Scholar] [CrossRef]
  29. Vasiliou, V.; Thompson, D.C.; Smith, C.; Fujita, M.; Chen, Y. Aldehyde dehydrogenases: From eye crystallins to metabolic disease and cancer stem cells. Chem. Biol. Interact. 2013, 202, 2–10. [Google Scholar] [CrossRef]
  30. Vasiliou, V.; Nebert, D.W. Analysis and update of the human aldehyde dehydrogenase (ALDH) gene family. Hum. Genomics 2005, 2, 138–143. [Google Scholar] [CrossRef]
  31. Sophos, N.A.; Vasiliou, V. Aldehyde dehydrogenase gene superfamily: The 2002 update. Chem. Biol.Interact. 2003, 143–144, 5–22. [Google Scholar] [CrossRef]
  32. Lavudi, K.; Nuguri, S.M.; Pandey, P.; Kokkanti, R.R.; Wang, Q.E. ALDH and cancer stem cells: Pathways, challenges, and future directions in targeted therapy. Life Sci. 2024, 356, 123033. [Google Scholar] [CrossRef]
  33. Bogner, A.N.; Tanner, J.J. Structure-affinity relationships of reversible proline analog inhibitors targeting proline dehydrogenase. Org. Biomol. Chem. 2022, 20, 895–905. [Google Scholar] [CrossRef] [PubMed]
  34. Meeks, K.R.; Bogner, A.N.; Tanner, J.J. Screening a knowledge-based library of low molecular weight compounds against the proline biosynthetic enzyme 1-pyrroline-5-carboxylate 1 (PYCR1). Protein Sci. 2024, 33, e5072. [Google Scholar] [CrossRef] [PubMed]
  35. Scott, D.E.; Coyne, A.G.; Hudson, S.A.; Abell, C. Fragment-based approaches in drug discovery and chemical biology. Biochemistry 2012, 51, 4990–5003. [Google Scholar] [CrossRef]
  36. Rees, D.C.; Congreve, M.; Murray, C.W.; Carr, R. Fragment-based lead discovery. Nat. Rev. Drug Discov. 2004, 3, 660–672. [Google Scholar] [CrossRef]
  37. Campbell, A.C.; Prater, A.R.; Bogner, A.N.; Quinn, T.P.; Gates, K.S.; Becker, D.F.; Tanner, J.J. Photoinduced Covalent Irreversible Inactivation of Proline Dehydrogenase by S-Heterocycles. ACS Chem. Biol. 2021, 16, 2268–2279. [Google Scholar] [CrossRef]
  38. Kabsch, W. XDS. Acta Crystallogr. D Biol. Crystallogr. 2010, 66 Pt 2, 125–132. [Google Scholar] [CrossRef]
  39. Evans, P. Scaling and assessment of data quality. Acta Crystallogr. D Biol. Crystallogr. 2006, 62 Pt 1, 72–82. [Google Scholar] [CrossRef]
  40. Evans, P.R.; Murshudov, G.N. How good are my data and what is the resolution? Acta Crystallogr. D Biol. Crystallogr. 2013, 69 Pt 7, 1204–1214. [Google Scholar] [CrossRef]
  41. Evans, P.R. An introduction to data reduction: Space-group determination, scaling and intensity statistics. Acta Crystallogr. D Biol. Crystallogr. 2011, 67 Pt 4, 282–292. [Google Scholar] [CrossRef]
  42. Afonine, P.V.; Grosse-Kunstleve, R.W.; Echols, N.; Headd, J.J.; Moriarty, N.W.; Mustyakimov, M.; Terwilliger, T.C.; Urzhumtsev, A.; Zwart, P.H.; Adams, P.D. Towards automated crystallographic structure refinement with phenix.refine. Acta Crystallogr. D Biol. Crystallogr. 2012, 68 Pt 4, 352–367. [Google Scholar] [CrossRef]
  43. Emsley, P.; Lohkamp, B.; Scott, W.G.; Cowtan, K. Features and development of Coot. Acta Crystallogr. D Biol. Crystallogr. 2010, 66 Pt 4, 486–501. [Google Scholar] [CrossRef]
Molecules 29 05408 g001
Figure 1. Reactions of proline catabolism.
Figure 1. Reactions of proline catabolism.
Molecules 29 05408 g001
Molecules 29 05408 g002
Figure 2. Compounds used in the validation and follow-up X-ray crystallography experiments. The boxes indicate compounds for which a strong electron density was observed and crystal structures were determined and deposited in the PDB.
Figure 2. Compounds used in the validation and follow-up X-ray crystallography experiments. The boxes indicate compounds for which a strong electron density was observed and crystal structures were determined and deposited in the PDB.
Molecules 29 05408 g002
Molecules 29 05408 g003
Figure 3. The locations of validated hit compounds in the tertiary structure of SmPutA. The mesh represents polder omit maps (4σ). The PRODH half of SmPutA is shown on the left in beige, and the GSALDH half is shown on the right in white. FAD is shown as yellow sticks. NAD is shown as gray sticks.
Figure 3. The locations of validated hit compounds in the tertiary structure of SmPutA. The mesh represents polder omit maps (4σ). The PRODH half of SmPutA is shown on the left in beige, and the GSALDH half is shown on the right in white. FAD is shown as yellow sticks. NAD is shown as gray sticks.
Molecules 29 05408 g003
Molecules 29 05408 g004
Figure 4. Poses and interactions of 7 and 13: (A) fragment 7; (B) fragment 13; and (C) comparison of the binding modes of 7 (fragment, protein, and NAD+ in yellow, PDB ID 9DL2) and 13 (fragment, protein, and NAD+ in orange, PDB ID 9DL6) with the product L-glutamate (glutamate and protein in blue, PDB ID 9BBO).
Figure 4. Poses and interactions of 7 and 13: (A) fragment 7; (B) fragment 13; and (C) comparison of the binding modes of 7 (fragment, protein, and NAD+ in yellow, PDB ID 9DL2) and 13 (fragment, protein, and NAD+ in orange, PDB ID 9DL6) with the product L-glutamate (glutamate and protein in blue, PDB ID 9BBO).
Molecules 29 05408 g004
Molecules 29 05408 g005
Figure 5. Fragment 20 in the GSALDH active site: (A) polder omit map (4σ) and interactions; and (B) comparison of the poses of 20 (fragment, protein, and NAD+ in blue, PDB ID 9DL9) and 7 (fragment, protein, and NAD+ yellow, PDB ID 9DL2).
Figure 5. Fragment 20 in the GSALDH active site: (A) polder omit map (4σ) and interactions; and (B) comparison of the poses of 20 (fragment, protein, and NAD+ in blue, PDB ID 9DL9) and 7 (fragment, protein, and NAD+ yellow, PDB ID 9DL2).
Molecules 29 05408 g005
Molecules 29 05408 g006
Figure 6. Fragments 9, 10, 11, and 19 bound to the NAD+-binding site of the GSALDH active site: (A) overlap of all four fragments with the adenine of NAD+, where the NAD+ is from PDB entry 5KF6; (B) locations and polder omit maps (4σ) for fragment 9 bound to the adenine site (denoted by 9) and two surface sites formed by crystal contacts (denoted 9’ and 9”), where the protein in the asymmetric unit is white, and the symmetry-related chain is light blue; (C) the four conformations modeled for 19 in chain A (left) and chain B (right). The polder omit maps are shown (4σ).
Figure 6. Fragments 9, 10, 11, and 19 bound to the NAD+-binding site of the GSALDH active site: (A) overlap of all four fragments with the adenine of NAD+, where the NAD+ is from PDB entry 5KF6; (B) locations and polder omit maps (4σ) for fragment 9 bound to the adenine site (denoted by 9) and two surface sites formed by crystal contacts (denoted 9’ and 9”), where the protein in the asymmetric unit is white, and the symmetry-related chain is light blue; (C) the four conformations modeled for 19 in chain A (left) and chain B (right). The polder omit maps are shown (4σ).
Molecules 29 05408 g006
Molecules 29 05408 g007
Figure 7. Fragment 20 bound to the PRODH active site: (A) electron density (polder omit, 4σ) and interactions for 20, with replicate structures 1 and 2 shown in the left and right panels, respectively, and the FAD shown in yellow; and (B) comparison of the PRODH active sites complexed with 20 and the proline analog THFA (shown in green). Replicate structures 1 (white) and 2 (black) are shown in the left and right panels, respectively.
Figure 7. Fragment 20 bound to the PRODH active site: (A) electron density (polder omit, 4σ) and interactions for 20, with replicate structures 1 and 2 shown in the left and right panels, respectively, and the FAD shown in yellow; and (B) comparison of the PRODH active sites complexed with 20 and the proline analog THFA (shown in green). Replicate structures 1 (white) and 2 (black) are shown in the left and right panels, respectively.
Molecules 29 05408 g007
Molecules 29 05408 g008
Figure 8. Inhibition of GSALDH activity of SmPutA by fragments: (A) Relative activity measurements. The bars represent the initial rate with L-P5C at 200 µM, NAD+ at 50 µM, and the fragment at 5 mM. The error bars represent the standard deviation from three technical replicates. The data are normalized to the rate of SmPutA in the absence of an inhibitor. * Due to low compound solubility, 11 was screened at 1 mM and 19 was screened at 3.8 mM; due to the absorbance of 10 at 340 nm, this fragment was tested at 0.25 mM. (B,C) Inhibition of GSALDH activity of SmPutA by 7 and 13 with L-P5C as the variable substrate and NAD+ fixed at 0.2 mM. The data were fitted to a global competitive inhibition model. The Ki for 7 is 0.32 mM ± 0.02 mM. The Ki for 13 is 1.5 mM ± 0.1 mM.
Figure 8. Inhibition of GSALDH activity of SmPutA by fragments: (A) Relative activity measurements. The bars represent the initial rate with L-P5C at 200 µM, NAD+ at 50 µM, and the fragment at 5 mM. The error bars represent the standard deviation from three technical replicates. The data are normalized to the rate of SmPutA in the absence of an inhibitor. * Due to low compound solubility, 11 was screened at 1 mM and 19 was screened at 3.8 mM; due to the absorbance of 10 at 340 nm, this fragment was tested at 0.25 mM. (B,C) Inhibition of GSALDH activity of SmPutA by 7 and 13 with L-P5C as the variable substrate and NAD+ fixed at 0.2 mM. The data were fitted to a global competitive inhibition model. The Ki for 7 is 0.32 mM ± 0.02 mM. The Ki for 13 is 1.5 mM ± 0.1 mM.
Molecules 29 05408 g008
Molecules 29 05408 g009
Figure 9. Electron density (polder omit, 4σ) and interactions for analogs of 20 bound to the PRODH active site: (A) fragment 25; (B) fragment 39; (C) fragment 40; (D) fragment 41; and (E) fragment 42. The FAD is colored in yellow.
Figure 9. Electron density (polder omit, 4σ) and interactions for analogs of 20 bound to the PRODH active site: (A) fragment 25; (B) fragment 39; (C) fragment 40; (D) fragment 41; and (E) fragment 42. The FAD is colored in yellow.
Molecules 29 05408 g009
Molecules 29 05408 g010
Figure 10. Inhibition of PRODH activity by 20 and its analogs. The bars represent the initial rate with L-proline at 25 mM, menadione at 0.1 mM, o-AB at 4 mM, and the fragment at 5 mM. The error bars represent the standard deviation from three technical replicates. The data are normalized to the rate of SmPutA in the absence of an inhibitor. + represents activity in the presence of the positive control, THFA (5 mM). * Due to low compound solubility, 41 was screened at 3 mM. Also, because of aqueous solubility limitations, the assays for 38 and 42 included 0.5% DMSO.
Figure 10. Inhibition of PRODH activity by 20 and its analogs. The bars represent the initial rate with L-proline at 25 mM, menadione at 0.1 mM, o-AB at 4 mM, and the fragment at 5 mM. The error bars represent the standard deviation from three technical replicates. The data are normalized to the rate of SmPutA in the absence of an inhibitor. + represents activity in the presence of the positive control, THFA (5 mM). * Due to low compound solubility, 41 was screened at 3 mM. Also, because of aqueous solubility limitations, the assays for 38 and 42 included 0.5% DMSO.
Molecules 29 05408 g010
Table 1. Resolution of structures and binding locations of fragments.
Table 1. Resolution of structures and binding locations of fragments.
FragmentResolution (Å)Binding LocationPDB ID
71.55GSAL site9DL2
91.77NAD+ adenine site, surface9DL3
101.80NAD+ adenine site9DL4
111.77NAD+ adenine site9DL5
131.42GSAL site9DL6
191.72NAD+ adenine site9DL7
20 (replicate 1)1.64PRODH active site9DL8
20 (replicate 2)1.32PRODH active site, GSAL site9DL9
251.39PRODH active site9E0A
391.47PRODH active site9E0B
401.32PRODH active site9E0C
411.33PRODH active site9E0D
421.37PRODH active site9E0E
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Meeks, K.R.; Bogner, A.N.; Nix, J.C.; Tanner, J.J. Crystallographic Fragment Screening of a Bifunctional Proline Catabolic Enzyme Reveals New Inhibitor Templates for Proline Dehydrogenase and L-Glutamate-γ-semialdehyde Dehydrogenase. Molecules 2024, 29, 5408. https://doi.org/10.3390/molecules29225408

AMA Style

Meeks KR, Bogner AN, Nix JC, Tanner JJ. Crystallographic Fragment Screening of a Bifunctional Proline Catabolic Enzyme Reveals New Inhibitor Templates for Proline Dehydrogenase and L-Glutamate-γ-semialdehyde Dehydrogenase. Molecules. 2024; 29(22):5408. https://doi.org/10.3390/molecules29225408

Chicago/Turabian Style

Meeks, Kaylen R., Alexandra N. Bogner, Jay C. Nix, and John J. Tanner. 2024. "Crystallographic Fragment Screening of a Bifunctional Proline Catabolic Enzyme Reveals New Inhibitor Templates for Proline Dehydrogenase and L-Glutamate-γ-semialdehyde Dehydrogenase" Molecules 29, no. 22: 5408. https://doi.org/10.3390/molecules29225408

APA Style

Meeks, K. R., Bogner, A. N., Nix, J. C., & Tanner, J. J. (2024). Crystallographic Fragment Screening of a Bifunctional Proline Catabolic Enzyme Reveals New Inhibitor Templates for Proline Dehydrogenase and L-Glutamate-γ-semialdehyde Dehydrogenase. Molecules, 29(22), 5408. https://doi.org/10.3390/molecules29225408

Article Metrics

Back to TopTop