A Second Uniquely Human Mutation Affecting Sialic Acid Biology

Siglecs are immunoglobulin superfamily member lectins that selectively recognize different types and linkages of sialic acids, which are major components of cell surface and secreted glycoconjugates. We report here a human Siglec-like molecule (Siglec-L1) that lacks a conserved arginine residue known to be essential for optimal sialic acid recognition by previously known Siglecs. Loss of the arginine from an ancestral molecule was caused by a single nucleotide substitution that occurred after the common ancestor of humans with the great apes but before the origin of modern humans. The chimpanzee Siglec-L1 ortholog remains fully functional and preferentially recognizes N-glycolylneuraminic acid, which is a common sialic acid in great apes and other mammals. Reintroducing the ancestral arginine into the human molecule regenerates the same properties. Thus, the single base pair mutation that replaced the arginine on human Siglec-L1 is likely to be evolutionarily related to the previously reported loss of N-glycolylneuraminic acid expression in the human lineage. Siglec-L1 and its chimpanzee Siglec ortholog also have a different expression pattern from previously reported Siglecs because they are found on the lumenal edge of epithelial cell surfaces. Notably, the human genome contains several Siglec-like pseudogenes that have independent mutations that would have replaced the arginine residue required for optimal sialic acid recognition. Thus, additional changes in the biology of sialic acids may have taken place during human evolution. Siglecs are immunoglobulin superfamily member lectins that selectively recognize different types and linkages of sialic acids, which are major components of cell surface and secreted glycoconjugates. We report here a human Siglec-like molecule (Siglec-L1) that lacks a conserved arginine residue known to be essential for optimal sialic acid recognition by previously known Siglecs. Loss of the arginine from an ancestral molecule was caused by a single nucleotide substitution that occurred after the common ancestor of humans with the great apes but before the origin of modern humans. The chimpanzee Siglec-L1 ortholog remains fully functional and preferentially recognizes N-glycolylneuraminic acid, which is a common sialic acid in great apes and other mammals. Reintroducing the ancestral arginine into the human molecule regenerates the same properties. Thus, the single base pair mutation that replaced the arginine on human Siglec-L1 is likely to be evolutionarily related to the previously reported loss of N-glycolylneuraminic acid expression in the human lineage. Siglec-L1 and its chimpanzee Siglec ortholog also have a different expression pattern from previously reported Siglecs because they are found on the lumenal edge of epithelial cell surfaces. Notably, the human genome contains several Siglec-like pseudogenes that have independent mutations that would have replaced the arginine residue required for optimal sialic acid recognition. Thus, additional changes in the biology of sialic acids may have taken place during human evolution. N-glycolylneuraminic acid N-acetylneuraminic acid polymerase chain reaction Siglec-like molecule 1 untranslated region N-acetyllactosamine The great apes are the closest evolutionary relatives of humans, with the chimpanzee/bonobo clade likely having shared a last common ancestor with humans about 6–7 million years ago (1Darwin C. The Descent of Man, and Selection in Relation to Sex. D. Appleton and Co., New York1871Crossref Google Scholar, 2King M.C. Wilson A.C. Science. 1975; 188: 107-116Crossref PubMed Scopus (1901) Google Scholar, 3Sibley C.G. Comstock J.A. Ahlquist J.E. J. Mol. Evol. 1990; 30: 202-236Crossref PubMed Scopus (102) Google Scholar, 4Ruvolo M. Mol. Biol. Evol. 1997; 14: 248-265Crossref PubMed Scopus (259) Google Scholar, 5Goodman M. Am. J. Hum. Genet. 1999; 64: 31-39Abstract Full Text Full Text PDF PubMed Scopus (191) Google Scholar). Human genomic DNA sequences differ on average by only 1–2% from those of these great apes (3Sibley C.G. Comstock J.A. Ahlquist J.E. J. Mol. Evol. 1990; 30: 202-236Crossref PubMed Scopus (102) Google Scholar, 4Ruvolo M. Mol. Biol. Evol. 1997; 14: 248-265Crossref PubMed Scopus (259) Google Scholar, 5Goodman M. Am. J. Hum. Genet. 1999; 64: 31-39Abstract Full Text Full Text PDF PubMed Scopus (191) Google Scholar). Thus, the altered expression of relatively few gene products may underlie some of the obvious morphological and functional differences between the species. We and others recently discovered an inactivating mutation in the human gene encoding CMP- N-acetylneuraminic acid hydroxylase (6Chou H.H. Takematsu H. Diaz S. Iber J. Nickerson E. Wright K.L. Muchmore E.A. Nelson D.L. Warren S.T. Varki A. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 11751-11756Crossref PubMed Scopus (446) Google Scholar, 7Irie A. Koyama S. Kozutsumi Y. Kawasaki T. Suzuki A. J. Biol. Chem. 1998; 273: 15866-15871Abstract Full Text Full Text PDF PubMed Scopus (299) Google Scholar), an enzyme that is functional in the great apes (6Chou H.H. Takematsu H. Diaz S. Iber J. Nickerson E. Wright K.L. Muchmore E.A. Nelson D.L. Warren S.T. Varki A. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 11751-11756Crossref PubMed Scopus (446) Google Scholar). This explains the human-specific loss of a major sialic acid,N-glycolylneuraminic acid (Neu5Gc)1 (8Muchmore E.A. Diaz S. Varki A. Am. J. Phys. Anthropol. 1998; 107: 187-198Crossref PubMed Scopus (146) Google Scholar). The sialic acids are a family of 9-carbon sugars that are abundantly expressed on cell surfaces and secreted glycoconjugates of animals of the deuterostome lineage (9Varki A. Glycobiology. 1992; 2: 25-40Crossref PubMed Scopus (485) Google Scholar, 10Schauer R. Sialic Acids: Chemistry, Metabolism and Function, Cell Biology Monographs. 10. Springer-Verlag, New York1982Google Scholar, 11Kelm S. Schauer R. Int. Rev. Cytol. 1997; 175: 137-240Crossref PubMed Google Scholar). Located mostly at the outer end of glycan chains on glycoproteins and glycolipids, sialic acids mediate a variety of recognition events involving pathogenic microbes and toxins, as well as endogenous animal lectins (12Varki A. FASEB J. 1997; 11: 248-255Crossref PubMed Scopus (490) Google Scholar, 13Karlsson K.A. Mol. Microbiol. 1998; 29: 1-11Crossref PubMed Scopus (147) Google Scholar). Siglecs are the largest family of such endogenous sialic acid-recognizing lectins defined to date (14Crocker P.R. Clark E.A. Filbin M. Gordon S. Jones Y. Kehrl J.H. Kelm S. Le Douarin N. Powell L. Roder J. Schnaar R.L. Sgroi D.C. Stamenkovic K. Schauer R. Schachner M. Van den Berg T.K. Van der Merwe P.A. Watt S.M. Varki A. Glycobiology. 1998; 8: VCrossref PubMed Google Scholar, 15Crocker P.R. Varki A. Trends Immunol. 2001; 22: 337-342Abstract Full Text Full Text PDF PubMed Scopus (339) Google Scholar). All 10 reported human Siglecs are type I membrane proteins, consisting of an amino-terminal Ig V-set domain, variable numbers of Ig C2-set domains, a single-pass transmembrane domain, and a cytoplasmic tail typically containing tyrosine-based signaling motifs. Sialic acid recognition is mediated by the first Ig V-set domain (15Crocker P.R. Varki A. Trends Immunol. 2001; 22: 337-342Abstract Full Text Full Text PDF PubMed Scopus (339) Google Scholar, 16Law C.L. Aruffo A. Chandran K.A. Doty R.T. Clark E.A. J. Immunol. 1995; 155: 3368-3376PubMed Google Scholar, 17Nath D. Van der Merwe P.A. Kelm S. Bradfield P. Crocker P.R. J. Biol. Chem. 1995; 270: 26184-26191Abstract Full Text Full Text PDF PubMed Scopus (97) Google Scholar, 18Van der Merwe P.A. Crocker P.R. Vinson M. Barclay A.N. Schauer R. Kelm S. J. Biol. Chem. 1996; 271: 9273-9280Abstract Full Text Full Text PDF PubMed Scopus (140) Google Scholar, 19Vinson M. Van der Merwe P.A. Kelm S. May A. Jones E.Y. Crocker P.R. J. Biol. Chem. 1996; 271: 9267-9272Abstract Full Text Full Text PDF PubMed Scopus (107) Google Scholar), and certain amino acid residues invariant to this domain are known to be involved in interactions with the sialic acid ligand (20May A.P. Robinson R.C. Vinson M. Crocker P.R. Jones E.Y. Mol. Cell. 1998; 1: 719-728Abstract Full Text Full Text PDF PubMed Scopus (239) Google Scholar). In particular, all Siglec V-set domains have a conserved arginine residue that forms a salt bridge with the carboxylate group of sialic acids. Experimental mutation of this residue markedly diminishes binding in all Siglecs studied to date (18Van der Merwe P.A. Crocker P.R. Vinson M. Barclay A.N. Schauer R. Kelm S. J. Biol. Chem. 1996; 271: 9273-9280Abstract Full Text Full Text PDF PubMed Scopus (140) Google Scholar,19Vinson M. Van der Merwe P.A. Kelm S. May A. Jones E.Y. Crocker P.R. J. Biol. Chem. 1996; 271: 9267-9272Abstract Full Text Full Text PDF PubMed Scopus (107) Google Scholar, 21Tang S. Shen Y.J. DeBellard M.E. Mukhopadhyay G. Salzer J.L. Crocker P.R. Filbin M.T. J. Cell Biol. 1997; 138: 1355-1366Crossref PubMed Scopus (129) Google Scholar, 22Angata T. Varki A. Glycobiology. 2000; 10: 431-438Crossref PubMed Scopus (99) Google Scholar, 23Angata T. Varki A. J. Biol. Chem. 2000; 275: 22127-22135Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar). Here we report a human molecule that has many features of Siglecs but lacks robust sialic acid recognition because of a specific mutation of the “essential” arginine residue, which remains conserved in its great ape orthologs. We also consider potential connections to the previously described human-specific mutation involving sialic acid biology and search the human genome for Siglec-like pseudogenes, several of which turn out to have independent mutations replacing the “essential” arginine residue. After this work was completed, another group independently reported the cloning of the same molecule, which they called S2V, a “putative” Siglec (24Yu Z. Lai C.-M. Maoui M. Banville D. Shen S.-H. J. Biol. Chem. 2001; 276: 23816-23842Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar). Our study shows that the “essential” arginine residue, which was mutated to cysteine specifically in the human lineage, is required for easily detectable sialic acid-dependent recognition typical of previously reported Siglecs. By comparison, we demonstrate robust sialic acid-dependent recognition by the fully functional chimpanzee ortholog and by the human molecule when the arginine is restored. Thus, we do not consider the native human molecule abona fide Siglec; instead, we call it Siglec-like molecule 1 (Siglec-L1). A human expressed sequence tag clone (GenBankTM accession number AI132995) encoding transmembrane and cytosolic domains of Siglec-L1 was identified by BLAST search (25Altschul S.F. Madden T.L. Schaffer A.A. Zhang J. Zhang Z. Miller W. Lipman D.J. Nucleic Acids Res. 1997; 25: 3389-3402Crossref PubMed Scopus (59768) Google Scholar) of the GenBankTM data base. A cDNA for the Siglec-L1 full-length coding sequence was obtained by 5′-rapid amplification of cDNA ends from human fetal liver Marathon-Ready cDNA library (CLONTECH) using primers 5′R-SP1 (5′-GGGAGAGTTTGGTCATCAGGCACG-3′) and AP1 (included in the kit), followed by a nested PCR using primers 5′R-SP2 (5′-CTGATGTCCTGTTGCACAGCATGG-3′) and AP2 (included in the kit). The 3′-untranslated region (UTR) was obtained likewise, using 3′R-SP1 (5′-GTGGGCGTGGGGGATACAGGCATGG-3′) and AP1 for the first PCR, followed by a nested PCR using 3′R-SP2 (5′-CCAGTATGCATCCCTCAGCTTCCAC-3′) and AP2. The DNA fragments were cloned into pCRII-TOPO (Invitrogen), and the sequences were analyzed. Based on these sequences, new primers were designed: 5′Expr (5′-CCTTCTAGAGCCACCATGCTACTGCTGCTGCTACTGCTGCC-3′; theXbaI site is underlined), 3′-UTR (5′-CTGAAGCTTCTAAAGGAATCAGTCTCGGG-3′; theHindIII site is underlined), and 3′Chi 3D (5′-ATCTCCTTGGAAGACAGTCATGGTC-3′). The PCR product with 5′Expr and 3′-UTR (containing full coding sequence) was digested withXbaI and HindIII and ligated into appropriate sites of pcDNA3.1 (Invitrogen) for functional analysis. The PCR product with 5′Expr and 3′Chi 3D was digested with XbaI and ligated into the XbaI-EcoRV sites of EK-Fc-pEDdC vector for recombinant fusion protein production. Mammalian cell transfection with this construct gives a fusion protein of Siglec-L1 (first three Ig-like domains) and human IgG Fc tail (22Angata T. Varki A. Glycobiology. 2000; 10: 431-438Crossref PubMed Scopus (99) Google Scholar, 23Angata T. Varki A. J. Biol. Chem. 2000; 275: 22127-22135Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar, 26Patel N. Brinkman-Van der Linden E.C.M. Altmann S.W. Gish K. Balasubramanian S. Timans J.C. Peterson D. Bell M.P. Bazan J.F. Varki A. Kastelein R.A. J. Biol. Chem. 1999; 274: 22729-22738Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar). The chimpanzee Siglec-L1 ortholog cDNA was obtained by reverse transcription-PCR from spleen total RNA (kindly provided by Dr. Elaine Muchmore, San Diego VA Hospital, San Diego, CA) using primers 5′-UTR (5′-TCAGCAACCCTGGCACCTCCAACCC-3′) and 5′R-SP2, followed by nested PCR using primers c5′Expr (5′-CCCTCTAGAGCCACCATGCTACTGCTGCTGCTACTGCTACTGCTGCC-3′; theXbaI site is underlined) and c3′-UTR (5′-CTGAAGCTTCTAAAGGAATCAGCCTCGGG-3′; theHindIII site is underlined). The DNA fragments were digested with XbaI and HindIII and ligated into appropriate sites of pBluescript II and the sequences were analyzed. A full-length cDNA was to pcDNA3.1 for functional analysis. The mutation was into human using the (Invitrogen) with primers and to DNA from and was from cell of (kindly provided by Dr. DNA from human was obtained from the of provided by Dr. Warren DNA fragments containing the first of Siglec-L1 were using primers 5′-UTR and 3′Chi and the first were using the and for some human for and and for A with N-acetyllactosamine was from was a from was as described previously J. J. Biol. Chem. 1996; 271: Full Text Full Text PDF PubMed Scopus Google Scholar) from of and of using acid (kindly provided by Dr. of San The reaction product was with and by and containing of was with of and of J. E. U. J.C. J. Biol. Chem. Full Text PDF PubMed Google Scholar). The reaction products were to in and substitution with sialic acids was as by S. M. Y. K. H. M. PubMed Scopus Google Scholar). expression with protein in the were by full-length cDNA into The were to using The were in in after at for 1 and with in The were with 1 of at for 1 and with 1 of at for The were in and to using the were and for binding human was and for of T. Varki A. J. Biol. Chem. 2000; 275: 22127-22135Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar). immunoglobulin that with human IgG Fc closest of the and were with of followed by with on protein The as described was specific for as by an using and as not were provided by on were and at with as described followed by and with of The were and using were using a with a for of the was by of as described to with human and using the for acid sequences of known Siglec V-set domains were to sequences in the human genome data base in GenBankTM using the DNA sequences were for sites using a This and DNA sequence with known Siglec were to the of in the of the pseudogenes were by PCR amplification and of genomic DNA pseudogenes on the were in of the closest to the was The only the was This is as by the Human DNA that we not as a gene was The of some pseudogenes was with the of the of the human genome data base J.C. 2001; PubMed Scopus Google Scholar). DNA sequences of the and of V-set domains were by and using for amino acid residues and the arginine residue involved in sialic acid and three conserved in known Siglecs (20May A.P. Robinson R.C. Vinson M. Crocker P.R. Jones E.Y. Mol. Cell. 1998; 1: 719-728Abstract Full Text Full Text PDF PubMed Scopus (239) Google sequences for the were because some pseudogenes a and the sequences were for using were using the N. M. Mol. Biol. Evol. Google Scholar) with a on was from the because its coding sequence was by an of of and were by the cDNA and and genomic DNA sequences of these in the GenBankTM data base. The sequence of was by cDNA closest with the genomic DNA region containing The genomic DNA region in the data base to be it was from some fragments in sequences of and were identified by using A sequence in the GenBankTM expressed sequence tag data base (GenBankTM accession number AI132995) to the coding for of known Siglecs. A coding region was obtained 5′-rapid amplification of cDNA The cDNA a new Siglec-like molecule with amino acids a type I membrane protein with Ig-like domains, a single transmembrane domain, and a cytosolic tail with tyrosine-based signaling motifs. All these features are typical of the of the Siglecs (15Crocker P.R. Varki A. Trends Immunol. 2001; 22: 337-342Abstract Full Text Full Text PDF PubMed Scopus (339) Google Scholar, 23Angata T. Varki A. J. Biol. Chem. 2000; 275: 22127-22135Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar). previously known Siglecs that have only amino-terminal V-set domain, the new molecule has V-set domains that are to other and to those of other Siglecs. of these V-set domains the “essential” arginine residue known to be for robust sialic acid recognition by previously reported Siglecs. A for sialic acid recognition by Siglecs is the of human the molecule P.R. S. A. A.C. Gordon S. G. Kelm S. Bradfield P. J. PubMed Scopus Google Scholar). from the of the “essential” arginine the discovered Siglec-like molecule to as in Siglecs we have reported previously (22Angata T. Varki A. Glycobiology. 2000; 10: 431-438Crossref PubMed Scopus (99) Google Scholar, 23Angata T. Varki A. J. Biol. Chem. 2000; 275: 22127-22135Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar). A recombinant protein the three amino-terminal Ig-like domains of the molecule also to sialic acid binding not in a T. Varki A. J. Biol. Chem. 2000; 275: 22127-22135Abstract Full Text Full Text PDF PubMed Scopus (138) Google Scholar, 26Patel N. Brinkman-Van der Linden E.C.M. Altmann S.W. Gish K. Balasubramanian S. Timans J.C. Peterson D. Bell M.P. Bazan J.F. Varki A. Kastelein R.A. J. Biol. Chem. 1999; 274: 22729-22738Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar). as a Siglec sialic acid recognition (14Crocker P.R. Clark E.A. Filbin M. Gordon S. Jones Y. Kehrl J.H. Kelm S. Le Douarin N. Powell L. Roder J. Schnaar R.L. Sgroi D.C. Stamenkovic K. Schauer R. Schachner M. Van den Berg T.K. Van der Merwe P.A. Watt S.M. Varki A. Glycobiology. 1998; 8: VCrossref PubMed Google Scholar), we have this protein Siglec-like (Siglec-L1). the sequence of the first V-set domain of Siglec-L1 to the domains of and we that the of the “essential” arginine residue in humans, a from the ancestral this we studied Siglec-L1 in the great DNA fragments the first of the Siglec-L1 of great apes were from genomic DNA of the and and the sequences were analyzed. All great ape have the “essential” arginine residue in the first Ig-like domain by a Thus, a single nucleotide substitution to at the first residue of the replaced the ancestral with a residue in humans. the other all great apes as well as humans the “essential” arginine residue in the Ig-like domain 1 that the great apes are fully functional we cloned a full-length cDNA encoding the chimpanzee ortholog (GenBankTM accession number expressed in by this molecule robust sialic acid-dependent A reverse mutation the ancestral arginine residue was in the human Siglec-L1 cDNA and studied in the same as described in the human molecule also robust sialic acid-dependent Thus, the single nucleotide substitution that the to a was for the sialic acid binding of the human is an with another study (24Yu Z. Lai C.-M. Maoui M. Banville D. Shen S.-H. J. Biol. Chem. 2001; 276: 23816-23842Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar) that sialic acid recognition by the native human This be by the that those with only a few in an that to have a of about In we robust sialic acid-dependent recognition by the chimpanzee ortholog as well as the human molecule, with of to the Thus, we are that the native human molecule not have robust sialic acid binding typical of other Siglecs. is that the binding reported in the other study some recognition of other of the sialic acid molecule that in the of the arginine residue. is that all humans are related R.L. Science. 2001; PubMed Scopus Google Scholar, P. C. U. D. P.A. C. G. Proc. Natl. Acad. Sci. U. S. A. 1999; PubMed Scopus Google Scholar). when the nucleotide in the first Ig-like domain occurred to the common origin of modern humans, we studied human genomic DNA from and and three All studied the same mutation the to a that this substitution occurred before the common origin of modern humans. We also that some human have a mutation that have occurred after the common origin of modern humans. This is This the human-specific mutation that the biology of sialic the first in the loss of expression (6Chou H.H. Takematsu H. Diaz S. Iber J. Nickerson E. Wright K.L. Muchmore E.A. Nelson D.L. Warren S.T. Varki A. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 11751-11756Crossref PubMed Scopus (446) Google Scholar). the events are we studied the of the human Siglec-L1 and the chimpanzee ortholog for to recognize and of sialic acids in the were to study with human the human reverse the chimpanzee in the human binding of in this In the human molecule and the chimpanzee ortholog sialic acids. is on a we also that binding of the molecule to sialic acid is markedly A specific Siglec-L1 the lumenal edge of epithelial in such as the and and in the A expression pattern was in chimpanzee are only of Siglec expression the in the M. U. 2000; 29: PubMed Scopus Google Scholar), and in the N. Brinkman-Van der Linden E.C.M. Altmann S.W. Gish K. Balasubramanian S. Timans J.C. Peterson D. Bell M.P. Bazan J.F. Varki A. Kastelein R.A. J. Biol. Chem. 1999; 274: 22729-22738Abstract Full Text Full Text PDF PubMed Scopus (144) Google Scholar). This is the first of a Siglec Siglec-like molecule that is expressed in epithelial all human which may be by the of mutations in some human The amino acid of the V-set domain of known Siglecs to search for other Siglecs and Siglec-like pseudogenes in the human genome sequence 2001; PubMed Scopus Google Scholar). of the Siglec are a region of that they by and from an ancestral We found that this region also contains many Siglec-like The of the V-set domains of these and pseudogenes is in In this many pseudogenes are found with different functional Siglec that they were independently from functional Siglec and the of evolution. of these pseudogenes have mutations at the that would have replaced the “essential” arginine residue. of these mutations to have events The genome sequence shows for Siglec and Siglec-like pseudogenes in the of and of the has an these data that the during and several of the pseudogenes may have for sialic acid binding by mutations the “essential” arginine residue. A likely for the of mutations is that of functional Siglec by some other mutations was followed by mutations to 2001; PubMed Scopus Google and the to which the arginine are the of these mutations of the to be to be by the of the of well as sequence data from the of other such as those of the and are to the of events that to the human The genomic of human Siglec-L1 was from the human genome sequence and with those of its in is a of with the genomic of the that and Siglec-L1 are by a relatively gene the Ig-like domain of Siglec-L1 is to an in the gene in the in of the C2-set domain in The other human and and the gene to this called Maoui M. Banville D. Shen J. 2001; PubMed Scopus Google Scholar), do not this The of three in all of these and the of a DNA between and of Siglec-L1 the of the that the ancestral of Siglec-L1 and may have in region of the ancestral in with the DNA Genet. 1996; PubMed Scopus Google Scholar). This V-set have replaced the the arginine in to to in Siglec-L1 this to in We have described the human-specific and mutation that the biology of sialic acids, the first of which to the loss of expression of the sialic acid the that the chimpanzee Siglec-L1 and the human Siglec-L1 a for binding it likely that the events are evolutionarily which is that the mutation in human Siglec-L1 occurred for the loss of The other is that the loss of in the loss of of this for it to in the it is likely that a gene would have this specific functional the arginine residue required for optimal sialic acid