SYNTHETASES

CROSS-REFERENCES TO RELATED APPLICATIONS [0001] The present application claims priority to U.S. Provisional Appl. No. 62/309,398, filed on September 20, 2020, which application is incorporated herein by reference in its entirety.

STATEMENT AS TO RIGHTS TO INVENTIONS MADE UNDER FEDERALLY SPONSORED RESEARCH AND DEVELOPMENT [0002] This invention was made with government support under Grant Nos. R01GM094523, R01AI130684, U01GM120419, and U01GM125288 awarded by the National Institutes of Health. The government has certain rights in the invention.

BACKGROUND OF THE INVENTION

[0003] Sialic acids such as A- ace tv 1 n e uram i n i c acid (Neu5Ac), A-glycolylneuraminic acid (Neu5Gc), keto-deoxynonulosonic acid (Kdn), and their derivatives play important roles in human physiology and pathology including but not limited to human homeostasis, immunology, allergy, transplantation, microbial and viral infection, host-microbe interaction, and cancer. Their structurally similar bacterial 8 or 9 carbon 2-keto sugars such as legionaminic acid (Leg), pseudaminic acid (Pse), and keto-deoxyoctulosonic acid (Kdo) are also part of bacterial virulence factors. These monosaccharides are added mainly by glycosyltransferase-catalyzed reactions which require activated sugar nucleotide donors such as cytidine-5'-monophospho-/V-acetylneuraminic acid (CMP-sialic acid). The donors themselves are formed by CMP-sugar synthetases such as CMP-sialic acid synthetase (CSS), CMP -legionaminic acid synthetase, CMP-pseudaminic acid synthetase, or CMP-Kdo synthetase etc.

[0004] Neu5Ac2en is an inhibitor designed to mimic the planar, oxocarbenium-like transition state of sialidases. Zanamivir and oseltamivir are modified Neu5Ac2en compounds approved by the FDA for treatment of influenza A virus infection and which act through selective inhibition of influenza sialidase. To date, however, the inhibition of CMP-sialic acid synthetase (CSS) or analogous bacterial sugar activation enzymes by Neu5Ac2en or other 2,3-dehydro-2-deoxy alpha-keto acids has not been reported.

BRIEF SUMMARY OF THE INVENTION

[0005] Provided herein are methods for inhibiting CMP-sugar synthetases. The methods include contacting a CMP-sugar synthetase with an effective amount of a compound selected from a dehydro-sugar, an anhydro-sugar, and salts thereof.

[0006] Also provided herein are methods for treating an infection, such as a bacterial infection or a viral infection. The methods include administering to a subject in need thereof an active agent selected from the group consisting of a dehydro-sugar, an anhydro-sugar, and pharmaceutically acceptable salts thereof, in an amount sufficient to inhibit a CMP-sugar synthetase in the subject, thereby treating the infection.

[0007] Also provided herein are methods for treating cancer. The methods include administering to a subject in need thereof an active agent selected from the group consisting of a dehydro-sugar, an anhydro-sugar, and pharmaceutically acceptable salts thereof, in an amount sufficient to inhibit a CMP-sugar synthetase in the subject, thereby treating the cancer.

BRIEF DESCRIPTION OF THE DRAWINGS [0008] FIG. 1 A shows the structure of ligand-free NmCSS homodimer. Each monomer consists of a globular nucleotide-binding domain and an extended dimerization domain where it interacts with its partner monomer. b-Sheets are labeled with numbers and a-helices with letters, with the exception of the 3io helix located in the dimerization domain of each monomer.

[0009] FIG. IB shows the structure of NmCSS homodimer with ligands CMP and Neu5Ac2en bound in the active site, drawn as sticks at right and left. Citrate binding is drawn in sticks at dimer interface.

[0010] FIG. 2 shows the binding orientation of CTP in the active site, including CTP bound in the A subunit of the dimer with two calcium ions ligated to the phosphate moieties. Calcium ions (A) and (B) work together with the P-loop (residues 12-22) to properly position and stabilize the negative charge on the triphosphate group. Calcium ion A ligates to the a phosphate, Asp211, and four water molecules. Asp209, which was observed to adopt two conformations with one binding to the Ca ²⁺ (A) water ligands, together with Asp211 make up the conserved DXD motif. Calcium ion B ligates to all three phosphates of CTP together with four water molecules. The dashed lines indicated calcium ligations, hydrogen bonds between protein and Ca-CTP, and P-loop stabilization interactions between Argl2-Asp78-Lysl6.

[0011] FIG. 3A shows the Neu5Ac2en inhibition of NmCSS with 1 mM Neu5Ac.

[0012] FIG. 3B shows the inhibition of NmCSS with varied concentrations of different inhibitors.

[0013] FIG. 4 A shows that after CTP and Ca ²⁺ bind to monomer A, the enzyme maintains the open state until arrival of sialic acid.

[0014] FIG. 4B shows that as Neu5Ac2en binds to the nucleotide binding (NB) domain of monomer A (white), interactions involving 5 residues from the dimerization domain of monomer B (brown) and sialic acid (yellow sticks) cause the distance between dimerization and NB domain to decrease by up to 8 A.

[0015] FIG. 4C shows that after formation of product, contacts between dimerization domain and sialic acid are broken, and the active site is re-opened.

[0016] FIG. 4D shows that superposition of nucleotide binding domains of the A monomers from CTP-bound structure (green) onto CMP- Neu5Ac2en structure (yellow). Relative orientation of the dimerization domain rotates closed upon binding Neu5Ac2en (highlighted by arrow). Taken together, FIGS. 4A-4D show the mechanism of active site closing.

[0017] FIG. 5A shows a sialic acid analog (Neu5Ac2en) having numerous contacts with both the nucleotide-binding domain of monomer A (white sticks) and the dimerization domain of monomer B (brown sticks). Three ordered water molecules indicate a solvent pocket which may allow for substrate variation.

[0018] FIG. 5B shows a summary of contacts between NmCSS and Neu5Ac2en. Taken together, FIGS. 5A and 5B show the ligands involved in dimer closing.

[0019] FIG. 6 shows the product-bound open state of NmCSS active site. After product formation, enzyme releases pyrophosphate and Ca ²⁺ (B) and leaves behind Ca ²⁺ (A) and product. Interactions between sialic acid and the nucleotide binding domain are maintained, but those between sialic acid and the dimerization domain of the opposite monomer are broken.

[0020] FIG. 7A shows NmCSS:CMP:Neu5Ac2en with overlay of CTP (sticks) from NmCSS:CTP structure, highlighting residues from the opposite monomer which interact with sialic acid and possibly with CTP.

[0021] FIG. 7B focuses on a small selection of relevant protein-sugar interactions in EcKdsB:CTP: Kdo2en for comparison to Lysl42 and Argl65 of NmCSS. Kdo2en is a non reacting analog of KdsB's natural substrate.

[0022] FIG. 7C focuses on a small selection of relevant protein-sugar interactions in MmCSS:CMP-sialic acid for comparison to Lysl42 and Argl65 of NmCSS. A bridging arginine between the dimerization domain and substrate of the opposite monomer is a common feature.

[0023] FIG. 8A shows a model of the NmCSS active site in the presence of both substrates. The conformation of CTP in the presence of Ca ²⁺ (A) and Ca ²⁺ (B) superimposed into the NmCSS:CMP:Neu5Ac2en active site in which Neu5Ac2en has been replaced with Neu5Ac suggests a clash between Ca ²⁺ (A) and Argl65 (thin sticks). However, a different rotamer of Argl65 (thick sticks) would allow for the observed Ca ²⁺ (A) placement while preserving the negative-charge stabilization of CTP’s g-phosphate by the arginine guanidinium group.

[0024] FIG. 8B shows a proposed mechanism for NmCSS activity i, CTP binds to the nucleotide binding (NB) domain first in addition to two divalent cations ii, next, Neu5Ac enters the active site and associates initially with the NB domain but quickly {Hi) draws together the NB and dimerization domains to form the closed state (all residues involved in Neu5Ac binding are summarized below the structure for simplicity iv, divalent cation M ²⁺ coordinates the C2-OH, allowing a water molecule to deprotonate C2-OH long enough for the hydroxyl O to attack the a-phosphate. v, after product formation, pyrophosphate (PPi) and M ²⁺ (B) are released from the active site and the active site re-opens.

DETAILED DESCRIPTION OF THE INVENTION [0025] The present invention is based in part on the discovery that sugar derivatives such as Neu5Ac2en bind to and inhibit CMP sialic acid synthetases such as Neisseria meningitidis CSS (NmCSS), which was examined during crystallographic studies as described herein. Further investigation revealed an approximate Ki of 7 mM, well below the KM of NmCSS toward sialic acid. NmCSS shares sufficient sequence similarity with human CSS and the family of enzymes which activate bacterial alpha-keto acids, pointing to a common mechanism by this class of enzymes.

[0026] Human CSS is an attractive target for anti-cancer immunotherapies. Some cancer cells overexpress sialic acid on the surface and the interaction of these sialic acids with sialic acid-binding immunoglobulin-type lectins (siglecs) on the surface of immune cells inhibits immune surveillance against cancer. Overexpression of sialyl Lewis x and its sulfated derivatives also affect cancer metastasis. While humans express at least 15 distinct siglecs and at least 20 distinct sialyltransferases, there is only a single human CSS expressed from the CMAS gene. The ability to reduce expression of sialic acid through inhibition of a single gene is attractive. Targeting the inhibitors to cancer is a potentially effect anti-cancer therapeutic strategy.

[0027] Selective inhibition of CMP-Kdo synthetase, CMP-legionaminic acid synthetase, CMP-pseudaminic acid synthetase and other bacterial homologs of CSS is expected to provide broad-spectrum and/or narrow-spectrum antibiotic activity. Kdo in lipopolysaccharides as well as LegA and PseA in capsular polysaccharides are part of virulence factors of pathogenic bacteria. Inhibiting the expression of these glycans is a viable antibiotic strategy. The molecular differences between Neu5Ac and these bacterial sugars provides orthogonality between the enzymes which process them. Compounds such as Kdo2en, PseA2en, and LegA2en may therefore be employed for highly selective inhibition of bacterial CMP-sugar synthetases. Furthermore, these inhibitors very closely mimic the natural sugars from which they are derived, potentially eliminating evolutionary resistance pathways that would otherwise enable CMP-sugar synthetase activity in the target organism without inhibition by the sugar derivative.

I. Inhibition of CMP-Sugar Synthetases

[0028] Provided herein are methods for inhibiting CMP-sugar synthetases. The methods include contacting a CMP-sugar synthetase with an effective amount of a compound selected from a dehydro-sugar, an anhydro-sugar, and salts thereof.

[0029] In some embodiments, the compound is a dehydro-sugar or a salt thereof. In some embodiments, the dehydro-sugar is a 2-deoxy-2,3-dehydro-nonulosonic acid or a 2-deoxy- 2,3-dehydro-octulosonic acid. [0030] In some embodiments, the 2-deoxy-2,3-dehydro-nonulosonic acid is a 2-deoxy -2,3- dehydro-sialic acid or a salt thereof. In some embodiments, the 2-deoxy-2,3-dehydro- nonulosonic acid is a compound according to Formula I: or a salt thereof, wherein:

R ¹ , R ² , R ³ , R ⁴ , and R ⁶ are independently selected from the group consisting of -NHC(0)R ^a , -Ns, -NH ₂ , -NHR ^a , -OC(0)R ^a , -OH, -F, and hydrogen;

R ⁵ is selected from the group consisting of -NHAc, -NHGc, -NHC(0)R ^a , -N3, -NH2, -OC(0)R ^a , -OH, -F, and hydrogen; each R ^a is independently selected from the group consisting of optionally substituted Ci-12 alkyl, optionally substituted Ci-12 haloalkyl, optionally substituted C3-10 cycloalkyl, optionally substituted G,-in aryl, and optionally substituted C7-22 arylalkyl;

Ac is -C(0)CH3; and Gc is -C(0)CH ₂ OH.

[0031] In some embodiments, at least one of R ¹ , R ² , R ³ , and R ⁴ is other than -OH when R ⁵ is

-NHAc or -NHGc. In some embodiments, at least one of R ¹ , R ² , R ³ , and R ⁴ is other than - OH when R ⁵ is -NHAc, -NHGc, or -OH. [0032] In some embodiments, at least one of R ¹ , R ² , R ³ , and R ⁴ is other than -OH when R ⁶ is hydrogen and R ⁵ is -NHAc or -NHGc. In some embodiments, at least one of R ¹ , R ² , R ³ , and R ⁴ is other than -OH when R ⁶ is hydrogen and R ⁵ is -NHAc, -NHGc, or -OH.

[0033] In some embodiments, R ⁶ is hydrogen. In some embodiments, R ⁶ is -F.

[0034] In some embodiments, the 2-deoxy -2, 3-dehydro-sialic acid is selected from:

and salts thereof.

[0035] In some embodiments, the dehydro-sugar is a 2-deoxy-2,3-dehydro-octulosonic acid or a salt thereof. In some embodiments, the 2-deoxy-2,3-dehydro-octulosonic acid is a compound according to Formula II: or a salt thereof, wherein:

R ¹¹ , R ¹² , R ¹⁴ , and R ¹⁵ are independently selected from the group consisting of -NHC(0)R ^a , -Ns, -NH ₂ , -NHR ^a , -0C(0)R ^a , -OH, -F, and hydrogen;

R ¹³ is selected from the group consisting of -NHAc, -NHGc, -NHC(0)R ^a , -N3, - NFh.

-0C(0)R ^a , -OH, -F, and hydrogen; each R ^a is independently selected from the group consisting of optionally substituted Ci-12 alkyl, optionally substituted Ci-12 haloalkyl, optionally substituted C3-10 cycloalkyl, optionally substituted Ce-io aryl, and optionally substituted C7-22 arylalkyl;

Ac is -C(0)CH3; and Gc is -C(0)CH ₂ 0H.

[0036] In some embodiments, R ¹⁵ is hydrogen. In some embodiments, R ¹⁵ is -F. [0037] In some embodiments, the 2-deoxy-2,3-dehydro-octulosonic acid is: or a salt thereof.

[0038] In some embodiments, the dehydro-sugar is a 2-deoxy-2,3-difluoro-nonulosonic acid or a salt thereof. In some embodiments, the 2-deoxy-2,3-difluoro-nonulosonic acid is a compound according to Formula III: or a salt thereof, wherein:

R ²¹ , R ²² , R ²³ , and R ²⁵ are independently selected from the group consisting of -NHC(0)R ^a , -Ns, -NH ₂ , -NHR ^a , -OC(0)R ^a , -OH, -F, and hydrogen; R ²⁶ is hydrogen and R ²⁷ is -F, or R ²⁶ is -F and R ²⁷ is hydrogen;

R ²⁸ is -F; each R ^a is independently selected from the group consisting of optionally substituted Ci-12 alkyl, optionally substituted Ci-12 haloalkyl, optionally substituted C3-10 cycloalkyl, optionally substituted C6-10 aryl, and optionally substituted C7-22 arylalkyl;

Ac is -C(0)CH3; and Gc is -C(0)CH ₂ OH.

[0039] In some embodiments, the 2-deoxy-2,3-difluoro-nonulosonic acid is a compound according to Formula Ilia: (Ilia), a compound according to Formula Illb: (Illb), or a salt thereof.

[0040] In some embodiments, R ²⁸ is -F. In some embodiments, R ²⁶ is -F and R ²⁷ is hydrogen. In some embodiments, R ²⁶ is hydrogen and R ²⁷ is -F. [0041] In some embodiments, the 2-deoxy-2,3-difluoro-nonulosonic acid is selected from the group consisting of: and salts thereof. [0042] In some embodiments, the compound is an anhydro-sugar or a salt thereof. In some embodiments, the anhydro-sugar is a 2,7-anhydro-nonulosonic acid. In some embodiments, the 2,7-anhydro-nonulosonic acid is a compound according to Formula IV: or a salt thereof, wherein

R ³¹ , R ³² , R ³⁴ , R ³⁶ , and R ³⁷ are independently selected from the group consisting of

-NHC(0)R ^a , -Ns, -NH ₂ , -NHR ^a , -0C(0)R ^a , -OH, -F, and hydrogen; R ³⁵ is selected from the group consisting of-NHAc, -NHGc, -NHC(0)R ^a , -N3,

-NH2, -OC(0)R ^a , -OH, -F, and hydrogen; R ^a is selected from the group consisting of optionally substituted Ci-12 alkyl, optionally substituted Ci-12 haloalkyl, optionally substituted C3-10 cycloalkyl, optionally substituted Ce-io aryl, and optionally substituted C7-22 arylalkyl;

Ac is -C(0)CH3; and Gc is -C(0)CH ₂ OH.

[0043] In some embodiments, at least one of R ³¹ , R ³² , and R ³⁴ is other than -OH when R ³⁵ is

-NHAc or -NHGc. In some embodiments, at least one of R ³¹ , R ³² , and R ³⁴ is other than -OH when R ³⁵ is -NHAc, -NHGc, or -OH.

[0044] In some embodiments, at least one of R ³¹ , R ³² , and R ³⁴ is other than -OH when R ³⁶ and R ³⁷ are hydrogen and R ³⁵ is -NHAc or -NHGc. In some embodiments, at least one of R ³¹ , R ³² , and R ³⁴ is other than -OH when R ³⁶ and R ³⁷ are hydrogen and R ³⁵ is -NHAc, - NHGc, or -OH.

[0045] In some embodiments, R ³⁶ and R ³⁷ are hydrogen. In some embodiments, R ³⁶ is hydrogen and R ³⁷ is -F. In some embodiments, R ³⁶ is -F and R ³⁷ is hydrogen.

[0046] The inhibitors used in the methods of the invention can be prepared by any suitable method including, for example, those described in WO 2018/098342 and WO 2018/201058, as well as those described by Li and Chen, etal. ( J Org. Chem. 2019, 84, 6697-6708), which references are incorporated herein by reference in their entirety.

[0047] CMP-sugar synthetases, including CMP-sialic acid synthetases and CMP-Kdo synthetases are described, for example, by Munster-Kiihnel et al. ( Glycobiology , 2004,

14(10): 43R-51R; Topics in Current Chemistry-SialoGlyco Chemistry and Biology I, 2013, 366:139-167). The functional active unit of CMP-sialic acid synthetases such as N. meningitidis CMP-sialic acid synthetase is a homodimer. During catalysis the nucleotide resides in a hydrophobic pocket formed by amino acid residues of monomer A, while the sugar is bound by amino acid residues of a substrate loop in monomer B. Dimerization domains extend out of central b-sheets and intertwine to form the dimer. CMP-Kdo synthetases such as those found in certain E. coli strains are also active dimers but contain a flatter dimerization interface without extended domains that entwine and lock the two monomers together. [0048] CMP-sugar synthetases include, but are not limited to, CMP-sialic acid synthetases (CSS), e.g., CSS’s from N. meningitidis (UniProt Accession No. P0A0Z8); H. sapiens (UniProt Accession No. Q8NFW8); D. melanogaster (UniProt Accession No. Q8IQV0); M. musculus (UniProt Accession No. Q99KK2); Rainbow trout, (). mykiss (UniProt Accession No. Q90WG6); H. influenzae (UniProt Accession No. A0A0D0HR91); S. agalactiae (EMBL Accession No. U19899); H. ducreyi (GenBank Accession No. U54496); E. coli (SwissProt Accession No. P13266); C. coli (SwissProt Accession No. Q45982); C. jejuni (EMBL Accession No. AAK91728); and L. pneumophila, (EMBL Accession No. AJ007311).

[0049] In some embodiments, the CSS is a CSS from N. meningitidis, H. sapiens, D. melanogaster, M. musculus, Rainbow trout, (). mykiss, H. influenzae, S. agalactiae, H. ducreyi, E. coli, C. coli, C. jejuni, or L. pneumophila. In some embodiments, the CSS contains an amino acid sequence having at least about 70% sequence identity to a reference sequence such as those set forth under UniProt Accession No. P0A0Z8, UniProt Accession No. Q8NFW8, UniProt Accession No. Q8IQV0, UniProt Accession No. Q99KK2, UniProt Accession No. Q90WG6, UniProt Accession No. A0A0D0HR91, EMBL Accession No. U19899, GenBank Accession No. U54496, SwissProt Accession No. P13266, SwissProt Accession No. Q45982, EMBL Accession No. AAK91728, or EMBL Accession No. AJ007311. The CSS may contain an amino acid sequence having, for example, at least about 70%, at least about 71%, at least about 72%, at least about 73%, at least about 74%, at least about 75%, at least about 76%, at least about 77%, at least about 78%, at least about 79%, at least about 80%, at least about 81%, at least about 82%, at least about 83%, at least about 84%, at least about 85%, at least about 86%, at least about 87%, at least about 88%, at least about 89%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, or at least about 99%, sequence identity to the reference sequence.

[0050] CMP-sugar synthases also include CMP-Kdo synthetases, e.g., CMP-Kdo synthetases from E. coli (enzyme EckspU; UniProt Accession No. P42216); E. coli K12 (enzyme EcKdsB; UniProt Accession No. P04951); A. aeolicus (UniProt Accession No. 066914); Z. mays (EMBL Accession No. AJ242474); H. influenzae (SwissProt Accession No. P44490); N. meningitidis (EMBL Accession No. AAF41093); and C. trachomatis (SwissProt Accession No. U15192). [0051] In some embodiments, the CMP-Kdo synthetase is a CMP-Kdo synthetase from E. coli, (e.g., E. coli K 12). A. aeolicus, Z. mays, H. influenzae, N. meningitidis, or C. trachomatis . In some embodiments, the CMP-Kdo synthetase contains an amino acid sequence having at least about 70% sequence identity to a reference sequence selected such as those set forth under UniProt Accession No. P42216, UniProt Accession No. P04951, UniProt Accession No. 066914, EMBL Accession No. AJ242474, SwissProt Accession No. P44490, EMBL Accession No. AAF41093, or SwissProt Accession No. U15192. The CMP- Kdo synthetase may contain an amino acid sequence having, for example, at least about 70%, at least about 71%, at least about 72%, at least about 73%, at least about 74%, at least about 75%, at least about 76%, at least about 77%, at least about 78%, at least about 79%, at least about 80%, at least about 81%, at least about 82%, at least about 83%, at least about 84%, at least about 85%, at least about 86%, at least about 87%, at least about 88%, at least about 89%, at least about 90%, at least about 91%, at least about 92%, at least about 93%, at least about 94%, at least about 95%, at least about 96%, at least about 97%, at least about 98%, or at least about 99%, sequence identity to the reference sequence.

[0052] For sequence comparison, typically one sequence acts as a reference sequence, to which test sequences are compared. When using a sequence comparison algorithm, test and reference sequences are entered into a computer, subsequence coordinates are designated, if necessary, and sequence algorithm program parameters are designated. Default program parameters can be used, or alternative parameters can be designated. The sequence comparison algorithm then calculates the percent sequence identities for the test sequences relative to the reference sequence, based on the program parameters. For sequence comparison of nucleic acids and proteins, for example, BLAST and BLAST 2.0 algorithms can be used, which are described in Altschul et al. (1990) J. Mol. Biol. 215: 403-410 and Altschul et al. (1977) Nucleic Acids Res. 25: 3389-3402, respectively. Software for performing BLAST analyses is publicly available at the National Center for Biotechnology Information website. The BLAST algorithms provide a statistical analysis of the similarity between two sequences (see, e.g., Karlin and Altschul, Proc. Nat Ί. Acad. Sci. USA, 90: 5873-5787 (1993)). One measure of similarity provided by the BLAST algorithm is the smallest sum probability (P(N)), which provides an indication of the probability by which a match between two nucleotide or amino acid sequences would occur by chance.

[0053] In some embodiments, the CMP-sugar synthetase is a CMP-sialic acid synthetase (also referred to as a cytidine-5'-monophospho-/V-acetylneuraminic acid synthetase in the case of Neu5Ac), e.g., a microbial CMP-sialic acid synthetase or human CMP-sialic acid synthetase. In some embodiments, the microbial CMP-sialic acid synthetase is N. meningitidis CMP-sialic acid synthetase (UniProt Accession No. P0A0Z8). In some embodiments, the CMP-sugar synthetase is a CMP-Kdn synthetase (e.g., Zebrafish CMP-Kdn synthetase, UniProt Accession No. H9BFW7), a CMP-legionaminic acid synthetase (e.g., CMP-legionaminic acid synthetase from C. concisus strain 13826, UniProt Accession No. A7ZFS0), or a CMP-pseudaminic acid synthetase (e.g., CMP-pseudaminic acid synthetase from C. jejuni subsp. jejuni serotype 0:2; UniProt Accession No. Q0P8U6). In some embodiments, the CMP-sugar synthetase is a CMP-Kdo synthetase (e.g., enzyme EckspU; UniProt Accession No. P42216).

[0054] Inhibiting the CMP-sugar synthetase generally includes contacting the CMP-sugar synthetase with an amount of the inhibitor (a 2-deoxy-2,3-dehydro-nonulosonic acid, a 2- deoxy-2,3-dehydro-octulosonic acid, or a salt thereol) that is sufficient to reduce the activity of the CMP-sugar synthetase as compared to the CMP-sugar synthetase activity in the absence of the inhibitor. For example, contacting the synthetase with the inhibitor can result in synthetase inhibition levels ranging from about 1% to about 99% (i.e.. the activity of the inhibited synthetase ranges from 99% to 1% of the activity of the synthetase in the absence of the compound). The level of synthetase inhibition can range from about 1% to about 10%, or from about 10% to about 20%, or from about 20% to about 30%, or from about 30% to about 40%, or from about 40% to about 50%, or from about 50% to about 60%, or from about 60% to about 70%, or from about 70% to about 80%, or from about 80% to about 90%, or from about 90% to about 99%. The level of synthetase inhibition can range from about 5% to about 95%, or from about 10% to about 90%, or from about 20% to about 80%, or from about 30% to about 70%, or from about 40% to about 60%. In some embodiments, contacting the synthetase with the inhibitor will result in essentially complete (i.e., 100%) inhibition of the synthetase.

II. Methods of Treatment

[0055] Also provided herein are methods for treating an infection, such as a bacterial infection or a viral infection. The methods include administering to a subject in need thereof an active agent selected from the group consisting of a dehydro-sugar, an anhydro-sugar, and pharmaceutically acceptable salts thereof, in an amount sufficient to inhibit a CMP-sugar synthetase in the subject, thereby treating the infection. In some embodiments, the infection is caused by a microbe that expresses one or more CMP-sugar synthetases but does not express a sialidase (e.g., a microbe such as E. coli EHV2 or other E. coli strain that does not express a sialidase, H. influenzae, or N. meningitidis).

[0056] Also provided herein are methods for treating cancer. The methods include administering to a subject in need thereof an active agent selected from the group consisting of a dehydro-sugar, an anhydro-sugar, and pharmaceutically acceptable salts thereof, in an amount sufficient to inhibit a CMP-sugar synthetase in the subject, thereby treating the cancer. In some embodiments, the cancer is a lymphoma (e.g., diffuse large B-cell lymphoma, medium to large B-cell lymphoma, or follicular lymphoma) or a leukemia (e.g., B-precursor leukemia).

[0057] The methods of the invention may be further be employed for the treatment of diseases and conditions mediated by siglecs (sialic-acid binding immunoglobulin-like lectins) and/or selectins. Examples of siglec-mediated conditions include, but are not limited to, autoimmune conditions (such as systemic lupus erythematosus and rheumatoid arthritis), graft-versus-host disease, infections (such as infections caused by HIV-1, PRRSV, C. jejuni, and Group B Streptococcus), sepsis, eosinophilia, cancer (including, but not limited to, lymphoma and leukemia such as acute myeloid leukemia), neuroinflammation, neurodegeneration, Alzheimer’s disease, pre-eclampsia, asthma, chronic lung inflammation, COPD, acute lung injury, allergies, osteoporosis, and tissue damage. Examples of selected- mediated conditions include, but are not limited to, atherosclerosis, ischemia-reperfusion injury, arterial thrombosis, and deep vein thrombosis, and sickle cell anemia. Siglec- and selectin-mediated diseases are described, for example, by Macauley et al. {Nat Rev Immunol, 2014, 14: 653-666) and McEver {Cardiovascular Research, 2015, 107(3): 331-339).

[0058] CMP-sugar synthetase inhibitors as described herein can be administered at any suitable dose in the methods for disease treatment. In general, a CMP-sugar synthetase inhibitor is administered at a dose ranging from about 0.1 milligrams to about 1000 milligrams per kilogram of a subject’s body weight {i.e., about 0.1-1000 mg/kg). The dose of CMP-sugar synthetase inhibitor can be, for example, about 0.1-1000 mg/kg, or about 1-500 mg/kg, or about 25-250 mg/kg, or about 50-100 mg/kg. The dose of CMP-sugar synthetase inhibitor can be about 1, 2, 3, 4, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, 60, 65, 70, 75, 80,

85, 90, 95, 100, 150, 200, 250, 300, 350, 400, 450, 500, 550, 600, 650, 700, 750, 800, 850, 900, 950 or 1000 mg/kg. The dosages can be varied depending upon the requirements of the patient, the severity of the disorder being treated, and the particular formulation being administered. The dose administered to a patient should be sufficient to result in a beneficial therapeutic response in the patient. The size of the dose will also be determined by the existence, nature, and extent of any adverse side-effects that accompany the administration of the drug in a particular patient. Determination of the proper dosage for a particular situation is within the skill of the typical practitioner. The total dosage can be divided and administered in portions over a period of time suitable to treat to the condition or disorder.

[0059] CMP-sugar synthetase inhibitors can be administered for periods of time which will vary depending upon the nature of the particular disorder, its severity, and the overall condition of the subject to whom the CMP-sugar synthetase inhibitor is administered. Administration can be conducted, for example, hourly, every 2 hours, three hours, four hours, six hours, eight hours, or twice daily including every 12 hours, or any intervening interval thereof. Administration can be conducted once daily, or once every 36 hours or 48 hours, or once every month or several months. Following treatment, a subject can be monitored for changes in his or her condition and for alleviation of the symptoms of the disorder. The dosage of the CMP-sugar synthetase inhibitor can either be increased in the event the subject does not respond significantly to a particular dosage level, or the dose can be decreased if an alleviation of the symptoms of the disorder is observed, or if the disorder has been remedied, or if unacceptable side effects are seen with a particular dosage.

[0060] A therapeutically effective amount of a CMP-sugar synthetase inhibitor can be administered to the subject in a treatment regimen comprising intervals of at least 1 hour, or 6 hours, or 12 hours, or 24 hours, or 36 hours, or 48 hours between dosages. Administration can be conducted at intervals of at least 72, 96, 120, 144, 168, 192, 216, or 240 hours (i.e., 3, 4, 5, 6, 7, 8, 9, or 10 days). In certain embodiments, administration of one or more CMP- sugar synthetase inhibitors is conducted in periods ranging from several months to several years.

[0061] Also provided are pharmaceutical compositions for the administration of CMP- sugar synthetase inhibitors, including 2-deoxy-2,3-dehydro-sialic acids and 2,7-anhydro- sialic acids as described herein. The pharmaceutical compositions can be prepared by any of the methods well known in the art of pharmacy and drug delivery. In general, methods of preparing the compositions include the step of bringing the active ingredient into association with a carrier containing one or more accessory ingredients. The pharmaceutical compositions are typically prepared by uniformly and intimately bringing the active ingredient into association with a liquid carrier or a finely divided solid carrier or both, and then, if necessary, shaping the product into the desired formulation. The compositions can be conveniently prepared and/or packaged in unit dosage form.

[0062] The pharmaceutical compositions can be in the form of sterile injectable aqueous or oleaginous solutions and suspensions. Sterile injectable preparations can be formulated using non-toxic parenterally-acceptable vehicles including water, Ringer’s solution, and isotonic sodium chloride solution, and acceptable solvents such as 1,3-butane diol. In addition, sterile, fixed oils are conventionally employed as a solvent or suspending medium. For this purpose, any bland fixed oil can be employed including synthetic mono- or diglycerides. In addition, fatty acids such as oleic acid find use in the preparation of injectables.

[0063] Aqueous suspensions contain the active materials in admixture with excipients including, but not limited to: suspending agents such as sodium carboxymethylcellulose, methylcellulose, oleagino-propylmethylcellulose, sodium alginate, polyvinyl-pyrrolidone, gum tragacanth and gum acacia; dispersing or wetting agents such as lecithin, polyoxyethylene stearate, and polyethylene sorbitan monooleate; and preservatives such as ethyl, «-propyl, and / hydroxy benzoate.

[0064] Oily suspensions can be formulated by suspending the active ingredient in a vegetable oil, for example, arachis oil, olive oil, sesame oil or coconut oil, or in a mineral oil such as liquid paraffin. The oily suspensions can contain a thickening agent, for example beeswax, hard paraffin or cetyl alcohol. These compositions can be preserved by the addition of an anti-oxidant such as ascorbic acid.

[0065] Dispersible powders and granules (suitable for preparation of an aqueous suspension by the addition of water) can contain the active ingredient in admixture with a dispersing agent, wetting agent, suspending agent, or combinations thereof. Additional excipients can also be present.

[0066] The pharmaceutical compositions can also be in the form of oil-in-water emulsions. The oily phase can be a vegetable oil, for example olive oil or arachis oil, or a mineral oil, for example liquid paraffin or mixtures of these. Suitable emulsifying agents can be naturally- occurring gums, such as gum acacia or gum tragacanth; naturally-occurring phospholipids, such as soy lecithin; esters or partial esters derived from fatty acids and hexitol anhydrides, such as sorbitan monooleate; and condensation products of said partial esters with ethylene oxide, such as polyoxyethylene sorbitan monooleate.

[0067] Pharmaceutical compositions containing the CMP-sugar synthetase inhibitors described herein can also be in a form suitable for oral use. Suitable compositions for oral administration include, but are not limited to, tablets, troches, lozenges, aqueous or oily suspensions, dispersible powders or granules, emulsions, hard or soft capsules, syrups, elixirs, solutions, buccal patches, oral gels, chewing gums, chewable tablets, effervescent powders, and effervescent tablets. Compositions for oral administration can be formulated according to any method known to those of skill in the art. Such compositions can contain one or more agents selected from sweetening agents, flavoring agents, coloring agents, antioxidants, and preserving agents in order to provide pharmaceutically elegant and palatable preparations.

[0068] Tablets generally contain the active ingredient in admixture with non-toxic pharmaceutically acceptable excipients, including: inert diluents, such as cellulose, silicon dioxide, aluminum oxide, calcium carbonate, sodium carbonate, glucose, mannitol, sorbitol, lactose, calcium phosphate, and sodium phosphate; granulating and disintegrating agents, such as com starch and alginic acid; binding agents, such as polyvinylpyrrolidone (PVP), cellulose, polyethylene glycol (PEG), starch, gelatin, and acacia; and lubricating agents such as magnesium stearate, stearic acid, and talc. The tablets can be uncoated or coated, enterically or otherwise, by known techniques to delay disintegration and absorption in the gastrointestinal tract and thereby provide a sustained action over a longer period. For example, a time delay material such as glyceryl monostearate or glyceryl distearate can be employed. Tablets can also be coated with a semi-permeable membrane and optional polymeric osmogents according to known techniques to form osmotic pump compositions for controlled release.

[0069] Compositions for oral administration can be formulated as hard gelatin capsules wherein the active ingredient is mixed with an inert solid diluent (such as calcium carbonate, calcium phosphate, or kaolin), or as soft gelatin capsules wherein the active ingredient is mixed with water or an oil medium (such as peanut oil, liquid paraffin, or olive oil).

[0070] The CMP-sugar synthetase inhibitors described herein can also be administered topically as a solution, ointment, cream, gel, suspension, mouth washes, eye-drops, and the like. Still further, transdermal delivery of the CMP-sugar synthetase inhibitors can be accomplished by means of iontophoretic patches and the like. The compound can also be administered in the form of suppositories for rectal administration of the drug. These compositions can be prepared by mixing the drug with a suitable non-irritating excipient which is solid at ordinary temperatures but liquid at the rectal temperature and will therefore melt in the rectum to release the drug. Such materials include cocoa butter and polyethylene glycols.

[0071] In some embodiments, an CMP-sugar synthetase inhibitor described herein is administered via intraperitoneal injection. In some embodiments, the CMP-sugar synthetase inhibitor is administered orally. In some embodiments, the CMP-sugar synthetase inhibitor is administered intravenously.

[0072] CMP-sugar synthetase inhibitors can be administered alone or in combination with one or more additional therapeutically active agents including, but not limited to, neuraminidase inhibitors (e.g., oseltamivir, zanamivir, peramivir, and the like), viral replication inhibitors (e.g., rimantadine and the like), macrolide antibiotics (e.g., erythromycin and the like), cephalosporin antibiotics (e.g., cefuroxime, ceflacor and the like), and fluoroquinolone antibiotics (e.g., ciprofloxacin, moxifloxacin, and the like). CMP-sugar synthetase inhibitors can also be administered in combination with cytotoxic/anti-cancer agents including, but not limited to, angiogenesis inhibitors (e.g., bevacizumab, ranibizumab, and the like), anthracy dines (e.g., doxorubicin, daunorubicin, and the like), platins (e.g., cisplatin, oxaliplatin, carboplatin, and the like), antimetabolites (e.g., 5-fluorouracil, methotrexate, and the like), kinase inhibitors (e.g., erlotinib, gefitinib, and the like), nucleoside analogs (e.g., gemcitabine, cytarabine, and the like), and taxanes (e.g., paclitaxel, docetaxel, and the like). The CMP-sugar synthetase inhibitors may also be administered in conjunction with one or more siglec-targeted therapies, such as a Siglec-2(CD22)-specific or Siglec 3(CD33)-specific antibody conjugated to calicheamicin (e.g., gemtuzumab ozogamicin and inotuzumab ozogamicin) or a bacterial protein toxin (such as moxetumomab pasudotox). The CMP-sugar synthetase inhibitors may also be administered in conjunction with one or more selectin-targeted therapies (e.g., rivipansel, bimosiamose, efomycine M, or the like).

III. Examples

Example 1. Materials and Methods

[0073] Expression and purification of His-tagged NmCSS. Neisseria meningitidis

CMP-sialic acid synthetase (NmCSS) was expressed and purified as described previously [Yu, et al. (2004) Bioorg Med Chem 12, 6427-6435], with the following exceptions: cells were lysed using a microfluidizer instead of lysozyme and DNasel, and lysate was clarified in a Beckman Avanti J-20 centrifuge using the J-20 rotor at 14,000rpm for 40 minutes before loading on the Ni ²⁺ -NTA column. NmCSS can routinely be purified to >95% purity at yields of up to 175 mg/L of cell culture.

[0074] Crystallization of CSS complexes. Crystals of CSS:CMP-sialic acid complex and substrate-free CSS were grown at room temperature by sitting-drop vapor-diffusion. A volume of 1 pL of protein solution containing 10 mg/mL of CSS, CTP (8 mM), and Neu5Ac (8 mM) was mixed with 1 pL of reservoir (0.16 M calcium acetate, 0.08 M sodium cacodylate, pH 6.5, 14.4% PEG 8000, and 20% glycerol). After a month of crystal growth, a single crystal was flash-cooled in liquid nitrogen. The resultant structure revealed that the product, CMP-Neu5Ac, occupied the active site along with a calcium ion in an octahedral coordination geometry. The presence of the product was surprising since it was previously thought that CSS could only use Mg ²⁺ to catalyze the formation of the product [Blacklow, et al. (1962) J Biol Chem 237, 3520-3526] Although it is possible that trace Mg ²⁺ could have been present in the crystallization buffer, the longer ligand distances between 2.3-2.6 A of the metal ion suggest that it was calcium instead of magnesium, which has typical ligand distances of ~2.2 A [Andreini, et al. (2013) Nucleic Acids Res 41, D312-319; Zheng, et al. (2017) Acta Crystallogr D Struct Biol 73, 223-233] Presumably when given a month to react during crystal growth, CSS catalyzed the reaction with the non-preferred cation, Ca ²⁺ . Eight- month-old crystals from this same drop were soaked in MgCh (250 mM) for thirty minutes to try to replace Ca ²⁺ with the preferred Mg ²⁺ before flash-cooling in liquid nitrogen. However, rather than yielding the desired Mg ²⁺ :CMP-Neu5Ac structure, diffraction data showed no electron density for any ligand in the active site of these crystals. The crystal structure determined under these conditions is presented here as the ligand-free structure.

[0075] Crystals of NmCSS:CTP were grown at room temperature by sitting-drop vapor- diffusion. A volume of 1 pL of protein mixture containing NmCSS (15 mg/mL), CTP (1 mM), and Neu5Ac (1 mM) was mixed with 1 pL of reservoir (0.1 M imidazole, pH 8.0, 6% PEG 8000, and 0.2 M calcium acetate) and allowed to equilibrate. Twenty-four hours later, rectangular crystals were observed. Assuming that catalysis in the presence of Ca ²⁺ would be significantly slower than with Mg ²⁺ , shortly after crystals were observed, a single crystal was selected and moved into immersion oil prior to flash-cooling in liquid nitrogen. It was anticipated that when crystals flash-cooled shortly after nucleation (~24 hours), the presence of Ca ²⁺ rather than Mg ²⁺ might inhibit or slow the enzyme catalysis enough during crystallization to trap a CTP plus Neu5Ac in a “Michalis-like” complex. However, only electron density for CTP was seen in the active site.

[0076] Crystals of NmCSS:CMP:Neu5Ac2en were grown at room temperature using hanging-drop vapor-diffusion. A volume of 200 nL of a protein solution containing NmCSS (15 mg/mL), CTP (1 mM), Neu5Ac2en (10 mM), and MgCk (100 mM) was mixed with 200 nL of reservoir (0.1 M sodium citrate/citric acid, pH 5.5, and 20% PEG 3000). After three days of equilibration, thick-needle-shaped crystals were observed. A single crystal was selected and transferred to a cryogenic solution containing 30% ethylene glycol before being flash-cooled in liquid nitrogen. This structure resulted in electron density that clearly defines CMP and Neu5Ac2en, suggesting that the presence of Mg ²⁺ allowed for the hydrolysis of CTP to CMP, and Neu5Ac2en bound in the active site with CMP.

[0077] In an effort to obtain a true pre-catalytic structure, experiments for co-crystallizing NmCSS with the non-hydrolyzable CTP analog CYtidine-5 -|(a.P)-methyleno| triphosphate (CMPCPP), Neu5Ac, and Mg ²⁺ were also attempted. Although crystals diffracted well, no electron density for the analog was observed in the active site.

[0078] Data collection, processing, and structure refinement. Data for the substrate-free crystal were collected on beamline 24-ID-C at the Advanced Photon Source (APS) at Argonne National Laboratory. Data for all protein: substrate complexes were collected on beamline 7-1 at the Stanford Synchrotron Radiation Lightsource (SSRL). All X-ray diffraction data intensities were processed with XDS [Kabsch, et al. (2010) Acta Crystallogr D Biol Crystallogr 66, 125-132] Scaling for the ligand-free enzyme-only and NmCSS:CMP- sialic acid complex was done using AIMLESS (CCP4) [Evans, et al. (2013) Acta Crystallogr D Biol Crystallogr 69, 1204-1214], and scaling for NmCSS:CTP and NmCSS:CMP- Neu5Ac2en was done using the program XSCALE [Kabsch, supra]. The previously published NmCSS structure (PDBID: 1EYR) [Mosimann, et al, (2001) JBiol Chem 276, 8190-8196] was used as a model for molecular replacement in CCP4 to solve the phases for all data sets. Data collection and refinement statistics are summarized in Table 1.

[0079] The molecular replacement solution for the NmCSS :CMP:Neu5Ac2en structure was initially difficult to interpret because of poor electron density. To resolve this issue, the NmCSS molecular replacement model monomer was broken up into two separate domains. Domain A was made up of residues 1-138 and 170-225 (nucleotide binding domain), and domain B of residues 139-169 (dimerization domain). The two subunits were used as ensembles in PHASER [McCoy, et al. (2007) J Appl Crystallogr 40, 658-674] for an initial round of molecular replacement. The resultant molecular replacement solution had a much improved, interpretable electron density map and revealed more of the closure between the two domains. All models were refined with REFMAC5 (CCP4) [Murshudov, et al. (2011) Acta Crystallogr D Biol Crystallogr 67, 355-367] and PHENIX [Afonine, et al. (2012) Acta Crystallogr D Biol Crystallogr 68, 352-367]

[0080] Citrate activation assay. A 20-pL reaction mixture containing NmCSS (20 ng), MgCh (10 mM), Tris-HCl (100 mM, pH 8.5), CTP (1 mM), Neu5Ac (1 mM), and sodium citrate (0, 1, 5, 10, or 20 mM) was incubated at 37°C in duplicate for 10 minutes. The reaction was stopped with 20 pL of pre-cooled ethanol and analyzed by capillary electrophoresis according to a previously-reported protocol [Li, et al. (2012) Appl Microbiol Biotechnol 93, 2411-2423] The experiment was repeated on a separate day and all percent conversion values were averaged for each concentration of citrate.

[0081] NmCSS mutagenesis and activity assays. NmCSS mutants E162A, E162Q, and R165A were prepared by Q5 site-directed mutagenesis, expressed in 50 mL cultures, and purified by Ni ²⁺ -NTA chromatography. For assays, a 40-pL reaction mixture containing NmCSS (4 ng) or its mutant (50 ng of E162A or E162Q, or 5 pg of R165A), MgCh (10 mM), Tris-HCl (100 mM, pH 8.5), CTP (3 mM), Neu5Ac (1 mM) was incubated at 37 °C in duplicate for 30 minutes. The reaction was stopped with 40 pL of pre-cooled methanol and quantified with an Infinity 1290-11 HPLC equipped with a UV-Vis detector (Agilent Technologies, CA) and a CarboPac PA100 anion exchange column (Dionex), using a gradient from 0% to 90% 2 M ammonium acetate, pH 6.0, over 10 minutes against water.

[0082] Neu5Ac2en and calcium chloride inhibitions assays. For Neu5Ac2en and calcium chloride inhibition assays, reactions were carried out at 37 °C for 30 minutes in 40 pL reaction mixtures containing CTP (3 mM), Neu5Ac (2 mM), NmCSS (8 ng), MgCh (10 mM), and Tris-HCl buffer (100 mM, pH 8.5) with or without Neu5Ac2en (0, 0.01, 0.02, 0.05, 0.1, 0.2, 0.5, 1, 2, 5, 10 mM) or CaCh (0, 5, 10, 20, 50, or 100 mM). The reactions were quenched and analyzed similarly to that described for the mutant assays. Data were analyzed using Grafit 5.0. Example 2. Catalytic cycle of Neisseria meningitidis CMP-sialic acid synthetase illustrated by high-resolution protein crystallography.

Overall structure

[0083] The ligand-free structure crystallizes with one monomer in the asymmetric unit with the crystallographic 2-fold generating the biologically functional dimer. The ligand-free structure presented here has an improved resolution (1.75 A) compared to the formerly published NmCSS ligand-free structure (2.00 A) (PDBID: 1EZI) [Mosimann, supra]. The two structures superimpose with a 0.70 A RMSD for all 216 modelled equivalent alpha- carbons. NmCSS exists as a homodimer where each monomer is composed of a globular nucleotide-binding domain and an extended dimerization domain (FIG. 1A). The dimerization domain comprises a ~35 residue insert corresponding to residues 136-171 and “domain-swaps” with the dimerization domain of the crystallographic related monomer to generate a tightly entwined complex. Improved resolution and more favorable crystal contacts allowed for the resolution of the structure of the loop between aD and aE (residues 71-80), which was disordered in the previously published ligand-free structure (PDBID: 1EZI) [Mosimann, supra]. The carbon backbone of the P-loop (residues 12-22) was visible in the electron density, but the side-chain density for Lysl6 was poor and was modeled as an alanine stub. The average B-value for this region remained high at 68.7 A ² , compared to an average B-value of 28.5 A ² for the entire structure.

Structure of NmCSS Calcium-CTP complex

[0084] The NmCSS calcium-CTP complex crystallized with a dimer in the asymmetric unit, and the structure was determined to 1.8 A resolution. Many residues involved in nucleotide binding were elucidated in the NmCSS :CDP structure published previously (PDBID: 1EYR) [Mosimann, supra]. However, the NmCSS:CDP structure revealed two conformations of the diphosphate moiety of CDP, resulting in ambiguity of understanding how Mg:CTP binds and the relative orientation and interactions between substrates Mg:CTP and Neu5Ac. The structure reported here clearly shows CTP with two Ca ²⁺ ions binding in the active site of monomer A with the conformation of the triphosphate moiety pointing in a similar direction to Conformation I in the previously published CDP structure (FIG. 2). The CTP-bound structure also reveals that monomer B binds a CTP molecule without calcium ligands, resulting in a slightly different conformation for the triphosphate moiety. This difference highlights the effect of divalent cations in the positioning of CTP. However, the electron density for the b and g phosphates is weak suggesting partial CTP hydrolysis in the B subunit. Nevertheless, for the A subunit, two calcium ions were observed to bind to the CTP triphosphate (FIG. 2). The triphosphate group wraps around Ca ²⁺ (B), allowing the b- phosphate to form a hydrogen bond with Seri 5 and the b and g phosphate to interact with Argl2 of the P-loop. This triphosphate conformation binding to Ca ²⁺ (B) exposes the backside of the a-phosphate to solvent or a potential sialic acid that would bind in the active site. However, in the absence of Ca ²⁺ in the B subunit, the triphosphate group adopts a staggered conformation with no interaction between the b-phosphate and Serl5 or Argl2. As a result, the side of the a-phosphate, which should be exposed, is pointed back toward the ribose moiety and is less accessible to a nucleophilic attack. Unless otherwise noted, all further discussion of the NmCSS:CTP structure refers strictly to the CTP:Ca ²⁺ :Ca ²⁺ complex, focusing on the A subunit.

[0085] Binding of CTP:Ca ²⁺ :Ca ²⁺ stabilizes the P-loop, as evidenced by its lower average B-value (18.9 A ² ) and clear electron density as compared to that of the ligand-free structure, including defining the conformation of Lysl6. The highly conserved, positively-charged P- loop residues Argl2 and Lys21 ion-pair with the negatively charged triphosphate group of CTP (FIG. 2). Argl2 and Lys21 were found to be conserved in a protein sequence alignment analysis of Neisseria meningitidis (Nm), human (Hs), Drosophila melanogaster (Dm), mouse (Mm), rainbow trout (Om), zebra fish (Dr), Streptococcus agalactiae (Sa), Haemophilus influenzae (Hi), two E. coli CMP-Kdo synthetases (EckpsU and EcKdsB), and Aquifex aeolicus CMP-Kdo synthetase (AaKdsB) was analyzed. Conserved residues that interact with CTP in were identified, as well as conserved residues involved in sialic binding, P-loop residues, and the metal-binding motif DXD. Electrostatic interactions are present between Lys21 and the a-phosphate group as well as between Argl2 and the b- and g-phosphates.

Seri 5 hydrogen-bonds with the b-phosphate, and the backbone amino groups of both Lysl6 and Glyl7 fix the proper positioning of the g-phosphate through hydrogen-bonding interactions (FIG. 2). The amide group of Asn22 in the P-loop makes two hydrogen bonds to the 2’- and 3’-OH of the ribose sugar.

[0086] The CTP:Ca ²⁺ :Ca ²⁺ binding pocket P-loop is also stabilized by an interaction between conserved Lysl6 and conserved Asp78, which in turn interacts with Argl2 (FIG. 2). This interaction helps to position the loop (residues 71-80) that binds to and selects for cytidine base of the nucleotide, where Arg71 donates two hydrogen bonds to 02 and N3 of cytidine, and the main chain carbonyl of Ala80 accepts a hydrogen bond from N4 (FIG. 2). [0087] The two calcium ions in the NmCSS:CTP structure — denoted Ca ²⁺ (A) and Ca ²⁺ (B) — each play different but important roles in the chemistry of binding and catalysis. Ca ²⁺ (B) interacts more directly with the triphosphate group, binding to all three phosphates in pentagonal bipyramidal primary coordination sphere. Meanwhile, Ca ²⁺ (A) coordinates only the a-phosphate directly. Conserved Asp209 and Asp211, of the DXD sequence, position Ca ²⁺ (A) through both direct coordination (Asp211) and indirect coordination (Asp209) via Ca ²⁺ (A)-bound waters. (FIG. 2) This finding supports previous activity studies of the poorly- functioning D209A and D211 A NmCSS mutants [Horsfall, et al, (2010) FEBS J 277 , 2779- 2790] These studies showed that, while the D209A and D211A mutations are both detrimental to enzyme function, increased concentrations of Mg ²⁺ increases the reaction rate of D211 A two times more than it increases the reaction rate for D209A. This is consistent with the structure that demonstrates Asp211 directly coordinates the divalent metal ion, but Asp209 interacts with the metal ion water ligands. Sequence conservation also suggests a less direct role for Asp209, because it is a glutamate in some species, including Drosophila.

[0088] Divalent cations are strictly required for catalytic activity [Mizanur, et al. (2008) Appl Microbiol Biotechnol 80, 757-765; Horsfall, supra]. While the majority of CSS enzymes require Mg ²⁺ or Mn ²⁺ , most are either inhibited or display greatly diminished activity with Ca ²⁺ [Mizanur (2008), supra]. It was also confirmed that Ca ²⁺ inhibits NmCSS with an IC50 value of 8.8 ± 0.4 mM (data not shown). However, given that Ca ²⁺ coordination is very similar to Mg ²⁺ coordination geometries, this CTP:Ca ²⁺ :Ca ²⁺ complex structure reported here is an ideal Michaelis complex analog of the first substrate binding in this ordered Bi-Bi catalytic mechanism. Furthermore, the CTP-Ca ²⁺ (B) substrate complex here is very similar to the CTP-Mg ²⁺ complex observed in E. coli CMP-Kdo synthetase, where the Mg ²⁺ ion is coordinated by all three phosphates of CTP [Heyes, et al. (2009) J Biol Chem 284, 35514-35523] Unfortunately, this CMP-Kdo synthetase structure did not allow for observation of the second Mg ²⁺ ion binding site but it was speculated that it would bind a in similar location to the Ca ²⁺ (A) ion reported here. However, a Mg ²⁺ ion was observed in another CMP-Kdo synthetase structure in a similar location to the Ca ²⁺ (A) ion [Jelakovic (2001) J Mol Biol 312, 143-155]

Binding of sialic acid induces active site closing

[0089] Conformational changes are seen in the active site of E. coli CMP-Kdo-forming enzyme upon binding of an analog of 2-keto-3-deoxymanno-octulonic acid (Kdo) [Heyes; Jelakovic; supra], leading to the hypothesis that a similar closed conformation may occur in NmCSS. To successfully capture this state, 2-deoxy-2,3-dehydro-/V-acetylneuraminic acid (Neu5Ac2en) was utilized. Neu5Ac2en is a dehydrated sialic acid analog that lacks the attacking C2-OH group responsible for formation of CMP-sialic acid and having instead a double bond between C2 and C3. In the absence of this hydroxyl group, Neu5Ac2en can bind in the active site pocket with CTP, but traps the enzyme in an intermediate state that parallels the enzyme structural state before catalysis occurs. Inhibition studies confirmed that Neu5Ac2en was an inhibitor with an IC50 value of 0.13 ± 0.01 mM under the assay conditions used (FIG. 3).

[0090] Neu5Ac2en was successfully trapped in the NmCSS active site; however, the structure revealed that during crystallization, the enzyme hydrolyzed CTP to CMP, and CMP was observed bound in the active site with bound Neu5Ac2en (FIG. IB). Nevertheless, the crystal structure of this ternary complex revealed the enzyme in a more closed state, relative to ligand-free and Ca:CTP bound structures (FIG. IB and FIG. 4).

[0091] The closed state of NmCSS:CMP:Neu5Ac2en is brought about by rotation about pivot points located at the termini of b-strands 5 and 8 at the base of the dimerization domain. The flexible nature of the residues at this location at the boundary of the nucleotide-binding and dimerization domain, including conserved Serl32, Alal33, and Glyl76, enables active site closure without enacting large conformational change in the individual domains. The nucleotide-binding domain (residues 1-134 and 175-224) of the CTP-bound and the CMP:Neu5Ac2en-bound conformations superimpose with an RMSD of only 0.430 A for all modeled equivalent alpha carbons. At the same time, the dimerization domain (residues 135— 174) of the CTP-bound and the CMP:Neu5Ac2en-bound conformations superimpose with an RMSD of only 0.658 A for all modeled equivalent alpha carbons. While the domains themselves remain rigid, the angle between them changes significantly, moving NmCSS into the closed conformation resulting in the dimerization domain of one subunit closing in on the active site of the other subunit in the dimer. In the CMP:Neu5Ac2en-bound structure, the dimerization domain has rotated so that the b5/b6 loop and 310 helix cover most of the exposed surface of the occupied active site, closing off the active site from bulk solvent (FIG. 4).

[0092] Active site closing is initiated through interactions between the sialic acid analogue and the dimerization domain of the other monomer B (FIG. 4B and FIGS. 5 A and 5B). Lysl42 of monomer B, located at the end of the dimerization domain’s 3io helix, and the backbone nitrogen of Argl65, reach into the active site of monomer A to interact with Neu5Ac2en’s carboxylate group. Lysl42 and the Leul61 backbone carbonyl each form a hydrogen bond with the C7-OH. The Glul62 backbone carbonyl oxygen hydrogen-bonds with C4-OH. After initial closing, which involved interactions between protein and sialic acid, the closed conformation is stabilized by several inter-monomer interactions. Glnl66(B) of the 3 io helix bridges the gap between monomers through hydrogen bonding with the backbone carbonyl of Asp78(A) (FIG. 4B). Hisl38(B) and Argl73(A) interact via t- t- stacking and Hisl38(B) also forms a salt bridge with Asp209(A).

[0093] In addition to the interactions with the dimerization domain of the opposite monomer, sialic acid also interacts with several highly or partially conserved residues in the nucleotide binding domain. The methyl group in the A-acetyl ofNeu5Ac2en lies in a hydrophobic pocket made up of Leul02, Tyrl79, and Phel92. Additionally, The Ne2 of GlnKM hydrogen-bonds with 08 of Neu5Ac2en. The Ser82 backbone nitrogen hydrogen- bonds to 04 of Neu5Ac2en while its Og hydrogen-bonds with the A-acetyl nitrogen (FIG. 5 A and 5B).

[0094] In spite of having Mg ²⁺ present in crystallization solution, no cations are observed in the active site of the NmCSS:CMP:Neu5Ac2en structure. This is most likely due to the absence of the cytidine nucleotide’s b- and g- phosphates, which were hydrolyzed during the course of the crystallography experiment.

Study of enzyme activation by citrate

[0095] Unaccountable electron density was observed at the dimer interface in the NmCSS:CMP:Neu5Ac2en structure. Given the crystals were grown in the presence of citrate (100 mM), a citrate molecule was modeled in which fits the electron density acceptably. This citrate molecule bridges the gap between the two dimerization domains and also forms a conduit between the one dimerization domain and the nucleotide binding domain of opposite monomer. Citrate's two terminal carboxylate groups reach across the space between the two monomers at their b5/b6 loop, hydrogen-bonding with the backbone nitrogens of the Glul37 in each. The citrate also connects two side chains from opposite monomers, the central carboxylate interacting with the Ndΐ of Hisl38 and the one of the terminal carboxylates interacting with the Ne2 of His204. [0096] Considering the importance of the b5/b6 loop in enzyme closing and the role of His 138 in stabilizing the NmCSS closed conformation, it was hypothesized that citrate may affect CSS activity. Indeed, after adding citrate to the reaction buffer, the production of CMP- sialic acid was increased moderately (data not shown). Further work will be necessary to determine the precise role of citrate in the regulation of CSS activity.

Structure with CMP-Neu5Ac product bound

[0097] The crystal structure of NmCSS with CMP-Neu5Ac product bound revealed a structure in the open state conformation similar to the ligand-free and Ca-CTP bound state of NmCSS. This suggests that after binding CTP and Neu5Ac, product formation drives the re opening of the active site, possibly to release products. Sialic acid interactions with the nucleotide-binding domain are maintained, but as a new bond is formed between the sialic acid and the a-phosphate, the sugar group moves deeper into the active site pocket of the nucleotide binding domain breaking interactions between the sugar and dimerization domain, and NmCSS transitions back into an open state. At this point, Ca ²⁺ (B) leaves the active site with the pyrophosphate, but Ca ²⁺ (A) remains, coordinating the Neu5Ac carboxylate group, an oxygen in the phosphate moiety of CMP-Neu5Ac, and four ordered water molecules in an octahedral geometry (FIG. 6). The ordered waters are also held in place with the help of the metal-binding DXD motif of Asp209 and Asp211. The DXD-containing loop is nudged open a small amount so Asp211 no longer coordinates Ca ²⁺ (A) directly as observed in the Ca-CTP bound structure.

Comparison with other structures

[0098] With less than 30% sequence identity to NmCSS, the crystal structures of E. coli CMP-Kdo synthetases KdsB [Heyes, supra ] and KpsU [Jelakovic, supra ] as well as mouse CSS (MmCSS) [Krapp, S., et al. (2003) J Mol Biol 334, 625-637] have parallels as well as differences with the structures presented here. Although several residues are highly conserved, the process and means of closed state formation varies from species to species.

[0099] While the E. coli CMP-Kdo synthetases are active dimers, they do not possess dimerization domains that entwine and lock the two monomers together as observed in NmCSS. This flatter dimerization interface only buries 2673 A ² of surface area of both monomers in the closed state compared to NmCSS closed state, which buries 4460 A ² . However, the E. coli structures also display a conformational change that closes the active site pocket in the presence of two substrates, but not after product formation, reminiscent of what was observed for NmCSS.

[0100] While the NmCSS :CTP: sialic acid structure could not be captured, the KdsB:CTP: 2-keto-2,3-deoxymanno-octulonic acid (Kdo2en) structure indicates that the guanidinium group of Argl64 (Argl65 in NmCSS) is involved in CTP g-phosphate interaction (FIGS. 7A and 7B). Superposition of the NmCSS:CTP and NmCSS :CMP:Neu5Ac2en structures suggests a similar function for Argl65 in NmCSS (FIG. 7A). Because the E. coli CMP-Kdo synthetases form different dimerization interactions and lack the dimerization domain, no residues of one monomer contact the sugar bound in the active site of the partner monomer. However, Glu210 of the nucleotide binding domain of KdsB hydrogen bonds with C4-OH of the Kdo analog through its carboxylate group and backbone nitrogen (FIG. 7B). Structurally, KdsB Glu210 is in the same position as Glul62 of NmCSS; however, in KdsB, Glu210 is from the same monomer as the active site in which Kdo2en is situated, but in NmCSS Glul62 is from the dimerization domain of the opposite monomer. KdsB Argl57, which ion- pairs with Kdo2en carboxylate group in a position similar to that of CSS Lysl42, is also from the same monomer.

[0101] The dimerization domain of mouse CSS, which is similar to NmCSS, does interact with the sugar in the opposite monomer’s active site through two arginine residues. Similar to Argl65 of NmCSS, the Arg202 backbone amino group of MmCSS binds to the Neu5Ac carboxylate group. Interestingly, in MmCSS the arginine guanidinium group ion-pairs with the Neu5Ac carboxylate group as well, giving the Arg202 side chain a role in closed-state stabilization. The R202A mutation has been shown to be detrimental but not fatal to enzyme activity, while the NmCSS R165A mutation results in a complete loss of activity [Munster, et al. (2002) J Biol Chem 277, 19688-19696] The ability of MmCSS to function without Arg202 may be due to additional closed-state stabilization from MmCSS Argl99, which hydrogen bonds with both the Neu5Ac C7 hydroxyl and the A-acetyl carbonyl group (FIG. 7C). The position of Argl99 allows it to also form three new interactions with Glu211 and Gln229 of the partner monomer (FIG. 7C). Due to this extra closed-state stabilization, MmCSS remains closed even after product formation. NmCSS Argl65, however, may play a different role in directing NmCSS enzyme-substrate conformation, as discussed below. Modeling of CSS:CTP:Ca ²⁺ :Ca ²⁺ :Neu5Ac

[0102] Argl65 is strictly conserved in both CMP-sialic acid and CMP-Kdo synthetases, yet there are no explicit interactions between Argl65 side chain atoms and Neu5Ac2en or CMP in the NmCSS:CMP:Neu5Ac2en structure. A superposition of CTP from the NmCSS:CTP structure onto the NmCSS:CMP:Neu5Ac2en structure indicates that Argl65(B) could form electrostatic interactions with the g-phosphate of CTP (FIG. 7A), which would corroborate with the function of its equivalent residue in KdsB. The side chain electron density for Argl65 in NmCSS:CMP:Neu5Ac2en is weak and the position shown in FIG. 7A would not be possible in the presence of Ca ²⁺ (A). However, since the structure is missing the complete triphosphate moiety, it cannot be regarded as a true intermediate state of the NmCSS catalytic cycle. An alternative Argl65 rotamer has been modeled which maintains the ion pair between Argl65 and the g-phosphate, but does not clash with Ca ²⁺ (A) (FIG. 8A).

[0103] Since Argl65 is a long, positively charged residue capable of reaching across the gap between the dimerization and nucleotide binding domains, the highly conserved nature of Argl65 may be due to the need for proper positioning of the backbone of residue 165.

Perhaps the anchoring of Argl65 with the g-phosphate is necessary in order to allow its backbone amino group to interact with the Neu5Ac carboxylate group.

Activity of NmCSS mutants

[0104] The four crystals structures presented above led to the hypothesis that the functional role of two key residues in helping to stabilize the closing of the active site and are therefore important in the catalytic mechanism proposed below. Specifically, it appears that Glul62 and Argl65 may play an important role by reaching across the dimer interface to possibly interact with substrates in the other monomer’s active site. Glul62 and Argl65 interact with the sialic acid and the triphosphate moiety from CTP, respectively, in the other subunit’s active site (FIG. 4B). Glul62 also stabilizes the active-site closure by hydrogen bonding with the main chain nitrogen of Phel92 in the other monomer.

[0105] Indeed, the E162A and E162Q mutants displayed six-fold lower activity than the wild-type NmCSS, while the R165A mutant was approximately six hundred-fold less active than the wild-type (Table 2). These results suggest that the active-site closer, stabilized by cross-monomer interactions with residues Glul62 and Argl65, is necessary for efficient catalysis. DISCUSSION

Basis for substrate tolerance

[0106] In the chemoenzymatic synthesis of complex carbohydrates, NmCSS has proven itself to be a very useful enzyme, activating several forms of sialic acids as diverse as 2-keto- 3-deoxy-D-manno-octulosonic acid (Kdo) and /V-azidoacetylneuraminic acid (Neu5AcN3). The structures presented provide a better understanding of the basis of this tolerance. First off, the flexible nature of the Lysl42 side chain would likely allow for substitutions at 07 as well as small changes in the position of the Neu5Ac carboxylate group. Additionally, analysis of the active site pocket after closing suggests that a solvent pocket near C7, C8, and C9 provides space for bulky modifications to 09 (FIG. 4A). It could also provide space for sialic acid to shift deeper into the active site pocket in order to allow A-acetyl modification as well. In fact, these are the modifications that are tolerated by the enzyme [Yu, supra]. Modification of the C8 hydroxyl group, however, results in a drastic decrease in activity [Li, supra]. In light of the structures presented here, this is likely due to the importance of the Glnl 04-08 interaction in the initial docking of sialic acid.

[0107] Although few crystal structures of NmCSS have been available previously, several groups have successfully created mutants of NmCSS with increased or modified substrate specificity. The S81R mutant of NmCSS, for example, has been shown to better tolerate 08- methyl substitutions [Yi, et al. (2013 ) Advanced Synthesis & Catalysis 355, 3597-3612] Model-guided mutagenesis has also created mutants that have more general tolerance for N- acyl-modified sialic acids [Yi, supra]. Despite these success stories however, there is still a need for mutants that can accept substrates with more than one modification such as Neu5Gc80Me or Kdo80Me. Based on the NmCSS:CMP:Neu5Ac2en and NmCSS:CMP- Neu5Ac structures presented here, shortening or mutating the loop between b-sheet 8 and 9 may allow more room for C8 modifications as well as provide more space for small shifts of sialic acid due to multiple modifications..

Role of Mg ²⁺ in catalysis

[0108] Previous studies have suggested that only Mg ²⁺ can be used by NmCSS to catalyze nucleotide transfer to sialic acid [Blacklow, supra]. Since product was formed over a month's time in the crystallization drop with Ca ²⁺ rather than Mg ²⁺ , clearly the identity of the cation is not strictly limited to Mg ²⁺ , but catalysis using Ca ²⁺ may be much slower than with Mg ²⁺ . In fact, calcium has been reported to substitute for Mg ²⁺ in other CSS enzymes, although usually with less catalytic efficiency [Schauer, et al. (1980), Hoppe Seylers Z Physiol Chem 361, 641- 648; Bravo, et al. (2001) Biochem J 358, 585-598; Rodriguez-Aparicio, et al. (1992) J Biol Chem 267, 9257-9263] FIG. 2 demonstrates the important role of divalent cations in proper positioning of the triphosphate groups. Divalent metal M ²⁺ (B) coordinates all three phosphate groups and orients the scissile bond between the a-phosphate and b-phosphate to be along the same axis as the attack by Neu5Ac on the a-phosphate. M ²⁺ (B) also increases the leaving group potential of the pyrophosphate moiety by stabilizing its negative charge. Even with proper orientation and the presence of a good leaving group, however, the C2-OH of Neu5Ac must first be deprotonated before it can attack the a-phosphate, and there is no basic residue within close enough proximity to play this role. M ²⁺ (A) likely solves this problem by decreasing the /?Ka of the hydroxyl group by ligating directly to the M ²⁺ (A) ion. Without wishing to be bound by any particular theory, it is believed that both the anomeric C2-OH and the Cl-carboxylate group of Neu5Ac directly ligate to M ²⁺ (A), which positions the anomeric C2 hydroxyl group in prime position for a nucleophilic attack on the a-phosphate. This model is corroborated by the NmCSS:CMP-Neu5Ac structure that has the Neu5Ac carboxylate group ligated to Ca ²⁺ (A) after catalysis. Hydroxyl activation via Mg ²⁺ coordination has also been proposed for the catalysis of KdsB [Heyes, supra], which uses the metal to activate 2-keto-3-deoxymanno-octulonic acid (Kdo). Based on structural analysis and EPR data [Heyes, supra, Schmidt, et al. (2011) PLoS One 6, e23231], the locations of the Mg ²⁺ in both EcKdsB and AaKdsB have been elucidated. Overlay of AaKdsB with the NmCSS:CTP:Ca:Ca structure shows that the metal positions in these structures are similar, supporting a similar mechanism as well.

Proposed catalytic cycle for NmCSS

[0109] Careful analysis of each structure presented here allows for the hypothesis of an enzyme conformation for each step in the catalytic cycle of NmCSS catalysis (FIG. 8B). At the start of the catalytic cycle, CTP binds in the active site pocket along with M ²⁺ (A) and M ²⁺ (B), which properly position the phosphate groups for catalysis and allow space for Neu5Ac to enter the active site pocket. Argl2 moves to interact with the b- and g-phosphates. With CTP in this position, Neu5Ac loosely binds in the available space in the active site of monomer A, with its carboxylate group and the anomeric carbon hydroxyl group in the coordination shell of M ²⁺ (A). The binding of Neu5Ac initiates active site closure as the interactions between sialic acid and Lysl42, Leul61, Glul62, and Argl65 pull sialic acid, CTP, and the nucleotide binding domain toward the dimerization domain (FIG. 4). Once initiated, the closed state is stabilized by Argl65 ion pairing with the g-phosphate and several inter-monomer interactions. As the dimerization domain comes near to Neu5Ac, coordination to M ²⁺ (A) allows Neu5Ac-C2-OH to be deprotonated, with the proton possibly being shuttled through water ligands to Asp209. The restricted movement by the presence of the dimerization domain forces the deprotonated oxygen toward CTP's a-phosphate, which has an increased electrophilic nature due to the two M ²⁺ ions that coordinate the triphosphate group and the Argl65 to g-phosphate ion pair. As the bond between the a-phosphate and b- phosphate breaks and the new bond between Neu5 Ac-02 and a-phosphate forms, the a- phosphate passes through a trigonal bipyramidal transition state, and pyrophosphate and its associated M ²⁺ (B) leave the active site. Once product has been formed, the sugar is more closely associated with the nucleotide-binding domain and can no longer maintain its hydrogen bonding and electrostatic interactions with the dimerization domain, so NmCSS:CMP-Neu5Ac reverts back to the open state. The product diffuses from the active site and the cycle can begin again.

[0110] Although NmCSS has been used in chemoenzymatic synthesis of sialic acid- containing structures for over a decade, the causes for its broad substrate tolerance and the role of divalent cations in catalysis has finally been shown through the four structures presented here. This work describes how sugar interaction with both monomers of the NmCSS homodimer are responsible for active site closure and the role of divalent cations in the transfer of CMP to sialic acid, laying the groundwork for future structure-guided enzyme engineering.

Example 3. Inhibition of NmCSS by dehydro-sugars and anhydro-sugars.

[0111] Reaction conditions for Neu5Ac2en IC50 studies were CTP (3 mM), Neu5Ac (1 mM, 2 mM, or 5 mM), NmCSS (8 ng), MgCk (10 mM), and Tris-HCl (100 mM, pH 8.5) in 40 pL reactions. Neu5Ac2en concentrations were 0, 10 mM, 20 mM, 50 mM, 100 mM, 200 mM, 500 mM, 1 mM, 2 mM, 5 mM, and 10 mM. Results with 1 mM Neu5Ac are shown in FIG. 3 A.

[0112] Reaction conditions for FIG. 3B, which shows the results obtained using various inhibitors, were: inhibitor (10 mM, 100 mM, or 1 mM), CTP (1.5 mM), Neu5Ac (2 mM), NmCSS (100 ng), MgCh (10 mM), and Tris-HCl (100 mM, pH 8.5) in 40 pL reactions at 37 °C for 30 minutes. [0113] The reactions were stopped with an equal volume (40 pL) of pre-cooled methanol and quantified with an Infinity 1290-11 HPLC equipped with a UV-Vis detector (Agilent Technologies, CA) and a CarboPac PA100 anion exchange column (Dionex). using a gradient from 0% to 90% 2 M ammonium acetate, pH 6.0, over 10 minutes against water.

IV. Exemplary Embodiments

[0114] Exemplary embodiments provided in accordance with the presently disclosed subject matter include, but are not limited to, the claims and the following embodiments:

1. A method of inhibiting a CMP-sugar synthetase, the method comprising contacting the CMP-sugar synthetase with an effective amount of a compound selected from the group consisting of a dehydro-sugar, an anhydro-sugar, and salts thereof.

2. The method of embodiment 1, wherein the CMP-sugar synthetase is a CMP-sialic acid synthetase.

3. The method of embodiment 2, wherein the CMP-sialic acid synthetase is a microbial CMP-sialic acid synthetase or human CMP-sialic acid synthetase.

4. The method of embodiment 3, wherein the microbial CMP-sialic acid synthetase is N. meningitidis CMP-sialic acid synthetase.

7. The method of embodiment 1, wherein the CMP-sialic acid synthetase is a microbial CMP-sialic acid synthetase and the compound is administered to a subject having a microbial infection in an amount sufficient to treat the infection.

5. The method of embodiment 1, wherein the CMP-sugar synthetase is a CMP-Kdn synthetase, a CMP-legionaminic acid synthetase, a CMP-pseudaminic acid synthetase.

6. The method of embodiment 1, wherein the CMP-sugar synthetase is a CMP-Kdo synthetase.

8. The method of embodiment 1, wherein the compound is administered to a subject having cancer in an amount sufficient to treat the cancer.

9. The method of any one of embodiments 1-8, wherein the compound is a dehydro-sugar or a salt thereof. 10. The method of any one of embodiments 1-9, wherein the dehydro sugar is a2-deoxy-2,3-dehydro-nonulosonic acid or a2-deoxy-2,3-dehydro-octulosonic acid.

11. The method of embodiment 10, wherein the dehydro-sugar is a 2- deoxy-2,3-dehydro-nonulosonic acid.

12. The method of embodiment 11, wherein the 2-deoxy-2,3-dehydro- nonulosonic acid is a compound according to Formula I: (I), or a salt thereof, wherein:

R ¹ , R ² , R ³ , R ⁴ , and R ⁶ are independently selected from the group consisting of -NHC(0)R ^a , -Ns, -NH ₂ , -NHR ^a , -OC(0)R ^a , -OH, -F, and hydrogen;

R ⁵ is selected from the group consisting of -NHAc, -NHGc, -NHC(0)R ^a , -N3, -NH2, -OC(0)R ^a , -OH, -F, and hydrogen; each R ^a is independently selected from the group consisting of optionally substituted Ci-12 alkyl, optionally substituted Ci-12 haloalkyl, optionally substituted C3-10 cycloalkyl, optionally substituted Ce-io aryl, and optionally substituted C7-22 arylalkyl;

Ac is -C(0)CH3; and Gc is -C(0)CH ₂ OH.

13. The method of embodiment 12, wherein at least one of R ¹ , R ² , R ³ , and R ⁴ is other than -OH when R ⁵ is -NHAc, -NHGc, or -OH.

14. The method of embodiment 12, wherein at least one of R ¹ , R ² , R ³ , and R ⁴ is other than -OH when R ⁶ is hydrogen and R ⁵ is -NHAc, -NHGc, or -OH.

15. The method of embodiment 12, wherein R ⁶ is hydrogen.

16. The method of embodiment 12, wherein R ⁶ is -F.

17. The method of embodiment 12, wherein the 2-deoxy-2,3-dehydro- nonulosonic acid is selected from the group consisting of:

and salts thereof. 18. The method of any one of embodiments 1-9, wherein the dehydro sugar is a 2-deoxy-2,3-dehydro-octulosonic acid or a salt thereof.

19. The method of embodiment 18, wherein the 2-deoxy-2,3-dehydro- octulosonic acid is a compound according to Formula II: or a salt thereof, wherein:

R ¹¹ , R ¹² , R ¹⁴ , and R ¹⁵ are independently selected from the group consisting of -NHC(0)R ^a , -Ns, -NH ₂ , -NHR ^a , -0C(0)R ^a , -OH, -F, and hydrogen;

R ¹³ is selected from the group consisting of -NHAc, -NHGc, -NHC(0)R ^a , -N3, - NH ₂ ,

-0C(0)R ^a , -OH, -F, and hydrogen; each R ^a is independently selected from the group consisting of optionally substituted Ci-i ₂ alkyl, optionally substituted Ci-i ₂ haloalkyl, optionally substituted C3-10 cycloalkyl, optionally substituted Ce-io aryl, and optionally substituted C7-22 arylalkyl;

Ac is -C(0)CH3; and Gc is -C(0)CH ₂ 0H. 20. The method of embodiment 19, wherein R ¹⁵ is hydrogen.

21. The method of embodiment 19, wherein R ¹⁵ is -F.

22. The method of embodiment 19, wherein the 2-deoxy-2,3-dehydro- octulosonic acid is: or a salt thereof.

23. The method of any one of embodiments 1-9, wherein the dehydro sugar is a 2-deoxy-2,3-difluoro-nonulosonic acid or a salt thereof.

24. The method of embodiment 23, wherein the 2-deoxy-2,3-difluoro- nonulosonic acid is a compound according to Formula III: or a salt thereof, wherein:

R ²¹ , R ²² , R ²³ , and R ²⁵ are independently selected from the group consisting of -NHC(0)R ^a , -Ns, -NH ₂ , -NHR ^a , -0C(0)R ^a , -OH, -F, and hydrogen;

R ²⁶ is hydrogen and R ²⁷ is -F, or R ²⁶ is -F and R ²⁷ is hydrogen;

R ²⁸ is -F;

R ²⁴ is selected from the group consisting of-NHAc, -NHGc, -NHC(0)R ^a , -N3, - NH ₂ ,

-0C(0)R ^a , -OH, -F, and hydrogen; each R ^a is independently selected from the group consisting of optionally substituted Ci-i ₂ alkyl, optionally substituted Ci-i ₂ haloalkyl, optionally substituted C3-10 cycloalkyl, optionally substituted Ce-io aryl, and optionally substituted Ci-22 arylalkyl;

Ac is -C(0)CH3; and Gc is -C(0)CH ₂ 0H. 25. The method of embodiment 24, the 2-deoxy-2,3-difluoro-nonulosonic acid is a compound according to Formula Ilia: (Ilia), a compound according to Formula Illb: (Illb), or a salt thereof.

26. The method of embodiment 24 or embodiment 25, wherein R ²⁶ is -F and R ²⁷ is hydrogen.

27. The method of embodiment 24 or embodiment 25, wherein R ²⁶ is hydrogen and R ²⁷ is -F.

28. The method of embodiment 24, wherein the 2-deoxy-2,3-difluoro- nonulosonic acid is selected from the group consisting of: and salts thereof.

29. The method of any one of embodiments 1-8, wherein the compound is an anhydro-sugar or a salt thereof.

30. The method of any one of embodiments 1-8 and embodiment 29, wherein the anhydro-sugar is a 2,7-anhydro-nonulosonic acid. 31. The method of embodiment 30, wherein the 2,7-anhydro-nonulosonic acid is a compound according to Formula IV: or a salt thereof, wherein

R ³¹ , R ³² , R ³⁴ , R ³⁶ , and R ³⁷ are independently selected from the group consisting of -NHC(0)R ^a , -Ns, -NH ₂ , -NHR ^a , -0C(0)R ^a , -OH, -F, and hydrogen;

R ³⁵ is selected from the group consisting of-NHAc, -NHGc, -NHC(0)R ^a , -N3, -NH2, -0C(0)R ^a , -OH, -F, and hydrogen;

R ^a is selected from the group consisting of optionally substituted Ci-12 alkyl, optionally substituted Ci-12 haloalkyl, optionally substituted C3-10 cycloalkyl, optionally substituted Ce-io aryl, and optionally substituted C7-22 arylalkyl; Ac is -C(0)CH3; and Gc is -C(0)CH ₂ 0H.

32. The method of embodiment 31, wherein at least one of R ³¹ , R ³² , and R ³⁴ is other than -OH when R ³⁵ is -NHAc, -NHGc, or -OH.

33. The method of embodiment 31, wherein at least one of R ³¹ , R ³² , and

R ³⁴ is other than -OH when R ³⁶ and R ³⁷ are hydrogen and R ³⁵ is -NHAc, -NHGc, or -OH.

34. The method of embodiment 31, wherein R ³⁶ and R ³⁷ are hydrogen.

35. The method of embodiment 31, wherein R ³⁶ is hydrogen and R ³⁷ is -F.

36. The method of embodiment 31, wherein R ³⁶ is -F and R ³⁷ is hydrogen.

37. A method for treating an infection, the method comprising administering to a subject in need thereof an active agent selected from the group consisting of a dehydro-sugar, an anhydro-sugar, and pharmaceutically acceptable salts thereof, in an amount sufficient to inhibit a CMP-sugar synthetase in the subject, thereby treating the infection. 38. The method of embodiment 37, wherein the active agent is a compound as set forth in any one of embodiments 9-36.

39. A method for treating cancer, the method comprising administering to a subject in need thereof an active agent selected from the group consisting of a dehydro- sugar, an anhydro-sugar, and pharmaceutically acceptable salts thereof, in an amount sufficient to inhibit a CMP-sugar synthetase in the subject, thereby treating the cancer.

40. The method of embodiment 39, wherein the active agent is a compound as set forth in any one of embodiments 9-36. [0115] Although the foregoing has been described in some detail by way of illustration and example for purposes of clarity of understanding, one of skill in the art will appreciate that certain changes and modifications may be practiced within the scope of the appended claims. In addition, each reference provided herein is incorporated by reference in its entirety to the same extent as if each reference was individually incorporated by reference. Where a conflict exists between the instant application and a reference provided herein, the instant application shall dominate.