Identification of the GalNAc kinase amino acid sequence.

A new kinase that forms GalNAc-1-P was purified from pig kidney cytosol and identified on gels by labeling with N3-[32P]ATP (Pastuszak, I., Drake, R., and Elbein, A. D. (1996) J. Biol. Chem. 271, in press). A 50-kDa labeled protein was eluted, digested with trypsin, and the sequences of four peptides representing 49 amino acids showed 90% identity to sequence of human galactokinase reported to be on chromosome 15. To resolve this dilemma, activities and substrate specificities of galactokinase and GalNAc kinase from human and pig kidney, as well as of galactokinase from the yeast clone transfected with the cDNA from presumptive human galactokinase, were compared. The purified galactokinases phosphorylated galactose, but not GalNAc, whereas GalNAc kinase also phosphorylated galactose when this sugar was present at millimolar concentrations. Extracts of gal 1(-) yeast clone, transfected with presumptive human galactokinase cDNA, had very low galactokinase activity even when yeast were grown on galactose, but good activity with GalNAc. On the other hand, the wild type yeast phosphorylated galactose, but not GalNAc. These data indicate that the sequence reported for galactokinase on chromosome 15 is that of GalNAc kinase, which can phosphorylate galactose when this sugar is present at millimolar concentrations. This transfection thus allows the yeast mutant to grow slowly on galactose-containing media.

Galactose, N-acetylgalactosamine, and galacturonic acid are all common components of complex carbohydrates in eucaryotic cells, and they all have the same configuration of hydroxyl groups at carbons 2 through 5, as well as similar pathways of biosynthesis (1). Thus, the major pathway for synthesis and activation of galactose is via the conversion of UDP-glucose to UDP-galactose by a 4-epimerase (2). However, another pathway for the formation of UDP-galactose also exists in some animal tissues such as liver and kidney. This pathway involves the phosphorylation of galactose in the one position by a specific galactokinase (3) and then transfer of galactose-1-P to UDP-glucose to give UDP-galactose and glucose-1-P (4).
For synthesis of UDP-GalNAc and UDP-galacturonic acid, the literature indicates that these two sugars are produced from the corresponding glucose derivatives, i.e. UDP-GlcNAc and UDP-glucuronic acid, by the action of specific UDP-sugar 4-epimerases (5). In fact, the UDP-gal 4-epimerase was purified to homogeneity, and the homogenous protein was shown to catalyze the epimerization of UDP-GlcNAc to UDP-GalNAc at the same rate as the epimerization of UDP-glucose to UDPgalactose (5). In mung bean seedlings, a UDP-galacturonic acid pyrophosphorylase was identified (6), suggesting the presence of a series of reactions for the activation of galacturonic acid that involves another kinase and the above-mentioned pyrophosphorylase.
Recent studies from our laboratory have identified and purified a GalNAc kinase from pig liver that phosphorylates GalNAc in the one position to form GalNAc-␣-1-P (7). In addition, we previously purified a UDP-GlcNAc pyrophosphorylase from pig kidney and pig liver and found that this enzyme had strong activity with GalNAc-1-P and UTP for the synthesis of UDP-GalNAc (8). Thus, animal cells may also possess another mechanism for activation of GalNAc, and this pathway probably represents a salvage mechanism to rescue GalNAc arising from the degradation of complex carbohydrates. The purified GalNAc kinase was subjected to SDS-gel electrophoresis and a 50-kDa band that had strong GalNAc kinase activity, and that was specifically labeled with N 3 -[ 32 P]ATP, was subjected to peptide sequencing. Four peptides from this protein, containing a total of 49 amino acids, showed 90% homology to the sequence reported from the human galactokinase gene located on chromosome 15 (9). The cDNA for this galactokinase was isolated from human HepG2 cells by expression cloning using a yeast mutant that lacked galactokinase and could not grow on galactose. Transfection with this cDNA enabled this clone to grow on galactose (9).
In order to resolve the dilema between these two sequences, we have reexamined the specificity of the human and pig kidney GalNAc kinase, as well as the galactokinase isolated from pig and human kidney. We have also examined these enzymatic activities in the parent yeast and in the yeast clone transfected with the human expression library. This report demonstrates that the previously reported sequence is that of GalNAc kinase and shows that this enzyme can phosphorylate galactose when this sugar is present at millimolar concentrations.

EXPERIMENTAL PROCEDURES
Materials-[1-3 H]N-acetylgalactosamine (GalNAc, 10 -25 Ci/mmol) and [6-3 H]galactose (20 Ci/mmol) were purchased from American Radiolabeled Chemicals, Inc. Unlabeled sugars and various adsorbents used for purification of the enzymes were from Sigma. Coomassie Blue, protein assay reagent, sodium dodecyl sulfate, acrylamide, and hydroxyapatite were from Bio-Rad. All other chemicals were from reliable chemical sources and were of the best grade available.
Kinase Assays-GalNAc kinase activity was assayed in incubation mixtures of 100 l containing the following components: 200 M [ 3 H]GalNAc (30,000 cpm), 5 mM ATP, 5 mM MgCl 2 , 5 mM NaF, 100 mM Tris-HCl buffer, pH 8.5, and various amounts of the enzyme preparations to be examined. Following an incubation for the appropriate time (usually 5 min), the reaction was stopped by heating at 100°C for 1 min, and the incubation mixture was applied to a column of DE52, contained in a Pasteur pipette (about 1.5 ml of resin). The column was washed with at least 5 column volumes of 10 mM (NH 4 )HCO 3 to remove unbound material and then GalNAc-1-P was eluted with 500 mM (NH 4 )HCO 3 . Aliquots of the wash and eluate were removed and assayed for their radioactive content by scintillation counting. In some experi-ments, the effect of increasing substrate concentration on the rate of phosphorylation was determined.
Galactokinase was assayed essentially as described above for the GalNAc kinase except that these incubations contained [ 3 H]galactose (30,000 cpm) as the substrate rather than GalNAc. Incubations usually contained 200 M galactose, 5 mM ATP, 5 mM MgCl 2 , 5 mM NaF, 100 mM Tris buffer, pH 8.5, and various amounts of the enzymes to be examined. Assays for phosphorylation were done by ion exchange chromatography as described above. The effect of galactose concentration on the rate of phosphorylation was also examined with various enzyme preparations.
Growth and Assay of Yeast Clones-Strains of Saccharomyces cerevesiae were grown in liquid culture in a medium containing 2% peptone, 1% yeast extract, and 2% of either glucose or galactose. Flasks were incubated on a rotary shaker at 30°C for 24 -48 h. Cultures of yeast were maintained on the same media containing 2% bactoagar in screw cap tubes. Yeast were harvested by centrifugation, washed with buffer, and ruptured by homogenization in the presence of glass beads. Cell debris and membranes were removed by high speed centrifugation and the soluble or cytosolic fraction was assayed for the GalNAc kinase and galactokinase activities.
Digestion and Sequencing of the GalNAc Kinase-Proteinase digestions of the GalNAc kinase and peptide sequencing were done by the Harvard Microchemistry Facility, Cambridge, MA. Purified proteins were separated by SDS-polyacrylamide gel electrophoresis and electrotransferred to polyvinylidiene difluoride membrane. Individual protein bands were excised and submitted to in situ digestion with trypsin (10). The resulting peptide mixture was separated by microbore high performance liquid chromatography using a Zorbax C18 (1 ϫ 150 mm) reverse-phase column on a Hewlett-Packard 1090 HPLC 1 /1040 diode array detector. Optimum fractions from the chromatogram were chosen based on differential UV absorbance at 205, 277, and 292 nm, peak symmetry, and resolution. Peaks were further screened for length and homogeneity by matrix-assisted laser desorption time-of-flight mass spectrometry (MALDI-MS) on a Finnigan Lasermat 2000 (Hemel, United Kingdom), and selected fractions were submitted to automated Edman degradation on an Applied Biosystems 494A or 477A (Foster City, CA). Details of strategies for the selection of peptide fractions, and their microsequencing have been described previously (11). Alternatively, tryptic peptide sequences were determined by microcapillary HPLC/electrospray ionization/tandem mass spectrometry on a Finnigan TSQ7000 triple quadrupole mass spectrometer (San Jose, CA) as described previously (12).

RESULTS AND DISCUSSION
The GalNAc kinase that catalyzes the phosphorylation of GalNAc from ATP to form GalNAc-␣-1-P was purified about 1275-fold from cytosolic extracts of pig kidney. At the final stage of purification, the enzyme fraction was chromatographed on a column of DEAE Cibacron blue, and fractions were assayed for GalNAc kinase activity and were also incubated with the photoaffinity label, N 3 -[ 32 P]ATP, and run on SDS gels to identify ATP binding proteins. The affinity labeling with the specific substrate was done on each column fraction in order to definitively identify the GalNAc kinase band for sequencing. A 50-kDa band showed maximum labeling in the same column fraction that contained maximum GalNAc kinase activity, whereas other fractions with low, or no, GalNAc kinase activity had little of the 50-kDa band, but more of a 66-kDa band that also labeled with the ATP probe (7). Thus it seems most likely that the 50-kDa band is the GalNAc kinase.
This 50-kDa band was electroeluted onto polyvinylidiene difluoride membranes and subjected to trypsin digestion and peptide sequencing by the Harvard Microchemistry Facility. Four peptides, containing 17, 9, 5 and 18 amino acids, were FIG. 1. Effect of galactose concentration on the activity of the purified pig kidney GalNAc kinase. Purified enzyme was incubated with various amounts of radioactive galactose as indicated in the presence of ATP, and the amount of radioactivity binding to DE52 was determined as a measure of phosphorylation of the substrate. For comparison, the effect of concentration of radioactive GalNAc on the enzymatic activity is shown. E, galactose concentration; q, GalNAc concentration.

FIG. 2. Effect of concentration of galactose or GalNAc on the enzymatic activity of the pig (or human) kidney galactokinase.
The partially purified galactokinase was incubated with various amounts of the radioactive substrates, galactose or GalNAc, plus ATP, and the amount of radioactivity binding to DE52 was determined as a measure of phosphorylation of the sugars. E, galactose concentration; q, GlcNAc concentration.

Pig GalNAcK
Gln-Ser-Leu-Phe-Ala-Thr-Lys-Pro-Gly-Gly-Gly-Ala-Leu-Val-Phe-Leu-Glu-Ala Identification of the GalNAc Kinase Amino Acid Sequence 23654 sequenced. As shown in Table I, these four peptides showed 90% homology to the amino acid sequence of human galactokinase deduced from the cDNA of a HepG2 expression library (9). This cDNA expression library was introduced into the yeast strain YM20, which contains a deletion of the gene encoding galactokinase (GAL1). Transformants were plated on media containing galactose as the carbon source and one transformant in 150,000 was identified. The cDNA involved in "curing" this mutant encodes for a 458-amino acid polypeptide of 50,386 daltons.
Since the sequence of the purified pig kidney GalNAc kinase was identical to that of the reported human galactokinase, we reexamined the specificity of our GalNAc kinase to determine whether it had any activity on galactose. Fig. 1 shows that the purified enzyme can phosphorylate galactose when this sugar is present at millimolar concentrations. A rough estimation of the K m for galactose based on this experiment is about 4 mM. The data in Fig. 1 also demonstrate that this enzyme was much more active toward GalNAc and showed approximately the same activity with 50 -100 M GalNAc as with 10 mM galactose. Similar results were observed with the partially purified GalNAc kinase from human kidney.
We also purified the galactokinase from pig and human kidney and examined the specificity of this enzyme as shown in Fig. 2. The GalNAc kinase and the galactokinase were previously shown to be well separated from each other by chromatography on phenyl-Sepharose (7). Fig. 2 shows that the kidney galactokinase had good phosphorylating activity toward galactose with an approximate K m of 0.5 mM, but it did not show any detectable activity toward GalNAc, even when this sugar was present at 10 mM concentrations. Thus the purified 50-kDa protein from pig kidney appears to be a true GalNAc kinase, but it does have the ability to phosphorylate galactose when this sugar is present in high concentrations. Since the transformants are selected based on their ability to grow on plates containing 2% galactose (about 10 mM), it seems likely that introduction of the GalNAc kinase into these cells would allow them to grow, albeit slowly, on galactose.
In order to determine if the yeast clone had been rescued by introduction of the GalNAc kinase gene, the wild type yeast and the transformant were grown on peptone-yeast extract medium containing either galactose or glucose as the sugar source. The cells were then ruptured, and the cytosolic fraction was assayed for the presence of galactokinase or GalNAc ki-  Identification of the GalNAc Kinase Amino Acid Sequence 23655 nase. Fig. 3 shows that the wild type S. cerevesiae, having the normal galactokinase gene, has strong phosphorylating activity for galactose when the cells were grown on galactose (panel B), but very low galactokinase activity when cells were grown on glucose (panel A). More importantly, the wild type cells had no ability to phosphorylate GalNAc, regardless of whether they were grown on glucose or galactose (panels A and B).
On the other hand, the yeast clone obtained by transformation with the HepG2 cDNA library showed quite different kinase activity as seen in Fig. 4. In this case, the cell free extracts had very low activity for phosphorylating galactose, regardless of whether cells were grown on glucose or galactose. However, these cells were able to phosphorylate GalNAc, and this activity was present regardless of whether cells were grown on glucose or galactose. In fact, the activity was significantly better when cells were grown on glucose, probably because these cells grow much better on glucose. The data with the various yeast clones are summarized in Table II, which shows the reasonably high GalNAc activity in the transformed yeast and the complete absence of this activity in wild type.
An earlier report had indicated that the human galactokinase resided on chromosome 17, but the subunit molecular weight was in conflict. One study reported a molecular mass of 38 kDa (13), while another reported it as 55 kDa (14). A second "galactokinase" was identified in the HepG2 expression library by complementing a yeast strain with a galactokinase deficiency that allowed this yeast to grow on galactose. This gene encoded a protein of 458 amino acids (50,386 kDa) and was localized to chromosome 15 (9). The data provided in this paper demonstrate that the sequence for this second gene is really that of the human (and pig) GalNAc kinase and that this enzyme is able to utilize galactose as a substrate and phosphorylate it to form galactose-1-P. Thus introduction of the gene for GalNAc kinase into the galactokinase negative yeast will apparently "cure" this organism and allow it to grow, although poorly, on galactose. Thus, the sequence of the GalNAc kinase and its location on chromosome 15 have now been established.