cysteine mutagenesis

1

J. Biol. Chem. (3/11/02)

Evaluation of Critical Structural Elements of UDP-sugar Substrates and Certain Cysteine

Residues of a Vertebrate Hyaluronan Synthase

Philip E. Pummill and Paul L. DeAngelisψ

Dept. of Biochemistry and Molecular Biology

Oklahoma Center for Medical Glycobiology

Univ. of Oklahoma Health Sciences Center

940 Stanton L. Young Blvd., Oklahoma City, OK 73104

ψ To whom correspondence should be addressed. Phone: (405) 271-2227 Fax: (405) 271-3092;

email: [email protected]

Running Title: Elements of UDP-Sugar Precursors and Cysteines of HAS

Key Terms: hyaluronic acid, hyaluronate, or hyaluronan; synthase; polysaccharide;

glycosyltransferase; sugar-nucleotide; active site

* This work was supported by a National Institutes of Health grant (GM56497) to P.L.D.

Copyright 2002 by The American Society for Biochemistry and Molecular Biology, Inc.

JBC Papers in Press. Published on April 9, 2002 as Manuscript M202456200 by guest on A

pril 12, 2018http://w

ww

.jbc.org/D

ownloaded from

http://www.jbc.org/

2

ABSTRACT

The hyaluronan [HA] synthases catalyze the addition of two different monosaccharides from

UDP-sugar substrates to the linear heteropolysaccharide chain. In order to accomplish this task,

the HA synthases must be able to bind and to transfer from both UDP-sugar substrates. Until

now, it has been impossible to distinguish between these two abilities. We have created a mutant

of xlHAS1, a HA synthase from Xenopus laevis, that allows for the examination of the enzyme’s

ability to bind substrate only. The ability of different compounds to protect the xlHAS1(C337S)

mutant enzyme from loss of activity due to treatment with N-ethylmaleimide, a cysteine

modifying reagent, yields information on the relative affinity of a variety of nucleotides and

nucleotide-sugars. We have observed that the substrate-binding selectivity is more relaxed than

the specificity of catalytic transfer. The only attribute that appears to be absolutely required for

binding is a nucleotide containing two phosphates complexed with magnesium ion. The role of

certain cysteine residues in catalysis was also evaluated. C307 of xlHAS1 may play a role in

catalysis or in maintaining structure. Mutation of C337 raises the UDP-GlcUA Michaelis

constant (Km), suggesting that this residue participates in UDP-GlcUA substrate binding or in

catalytic complex formation.

by guest on April 12, 2018

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

3

INTRODUCTION

HA1 is a glycosaminoglycan composed of alternating repeats of the disaccharide (→4)-β-D-

GlcUA(1→3)-β-D-GlcNAc(1→). This polysaccharide is abundant in vertebrates where it plays

structural, recognition, and signaling roles (1). The enzymes that catalyze the formation of HA,

the HA synthases, are dual-action glycosyltransferases that catalyze the transfer of both GlcUA

and GlcNAc (2, 3). These membrane-associated enzymes utilize UDP-linked sugar precursors.

We have reported previously that xlHAS1 is highly specific for the authentic HA substrates,

UDP-GlcUA and UDP-GlcNAc; the C4 epimers or UDP-glucose will not support HA

biosynthesis (4).

The vertebrate, the streptococcal, and the viral enzymes are comparable in size and have

regions or short sequence elements with considerable similarity (2). A few of these putative

elements, for example the DXD-containing motif, are similar to other glycosyltransferases that

produce various α- or β-linked polysaccharides from UDP-sugars (5-7). However, the exact role

of these motifs in the structure and/or the function of the polypeptide are only recently being

investigated. In view of the close amino acid sequence similarities among many

glycosyltransferases, it is quite likely that these residues are involved in binding common

determinants of UDP-sugars (e.g. uridine ring, phosphate groups) and/or catalyzing the transfer

of sugars residues.

1 The abbreviations used are: HA, hyaluronan, hyaluronate, or hyaluronic acid; HAS, HA synthase; Glc, Glucose;

GlcUA, glucuronic acid; GlcNAc, N-acetylglucosamine; Gal, Galactose; GalUA, galacturonic acid; GalNAc, N-

acetylgalactosamine; Tris, tris(hydroxymethyl)aminomethane; DTT, dithiothreitol; EDTA,

ethylenediaminetetraacetic acid; NEM, N-ethylmaleimide; Km, Michaelis constant; and Ki, inhibition constant.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

4

X-ray crystal structures have been obtained for several different glycosyltransferases from

bacteria, a bacteriophage, and vertebrates that utilize UDP-sugars as well as for a bacterial UDP-

glucose dehydrogenase (8-18). All of these structures show extensive hydrophobic interactions

with the uracil ring and hydrogen bonding with the functional groups of the uracil as well as with

the ribose hydroxyls and phosphates (Table 1). Many of these transferases use the DXD motif to

coordinate the divalent metal cation and interact with the phosphate groups of UDP (5-7, 14-16).

In lieu of a three-dimensional structure or active-site labeling data, the direct measurement or

analysis of the binding of substrates to glycosyltransferases is often quite difficult or even

impossible due to the low relative affinity of nucleotide-sugars for the enzymes. For example,

the Km value of xlHAS1 for UDP-sugars ranges from ~100 µM to almost 1 mM, depending on

experimental conditions (ref. 4; Table 2). Direct binding assays would require any washing steps

removing unbound substrate to be completed in a few seconds time. Equilibrium dialysis

experiments would require long times (which can be problematic for labile UDP-sugars) and

yield relatively weak signals.

The situation is further complicated by the extreme difficulty in purifying the vertebrate

HASs in a native state; most binding assessments should be performed on enriched or purified

membrane preparations because many other nucleotide binding proteins exist. Therefore, we

employed an indirect method, utilizing protection from inactivation mediated by a chemical

modification agent, NEM, to assess the relative affinity of a vertebrate HAS for a wide range of

compounds. These molecules have some or most of the structural elements of the authentic

UDP-sugars for HA biosynthesis. Our assumption is that a compound interacting with the

substrate-binding pocket or cleft will block NEM’s access to the site and protect the enzyme

from chemical modification and subsequent inactivation. We found that some structural


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

5

elements of the UDP-sugar substrates, including the two phosphate groups, are critical for

binding to xlHAS1. However, we have also found that some compounds with variations in the

sugar, base, or ribose can bind to xlHAS1 at a putative substrate-binding site but do not support

HA biosynthesis. This observation indicates that the substrate-binding requirements of the

enzyme are more relaxed than the catalytic requirements. Protection experiments suggest that

one or more cysteines might be part of or close to a putative substrate-binding site. We also

found that several cysteines of xlHAS1 were dispensable, but C307 may play a direct or a

structural role.

EXPERIMENTAL PROCEDURES

Production of Recombinant xlHAS1 Wild Type and Cysteine Mutant Enzymes – All reagents

were from Sigma or Fisher unless noted otherwise. The construction and the use of the xlHAS1

expression plasmid for studies in yeast were previously described (4, 19). Basically, the xlHAS1

polypeptide was cloned into the pYES2 vector (Invitrogen) under control of the GAL1 promoter

to form pYES/DG+. Site-directed mutagenesis was performed on pYES/DG+ using the

QuikChange Kit (Stratagene). Seven cysteine codons were altered using pairs of synthetic

oligonucleotides containing either the partially degenerate codon (TYS, where Y = C or T; S = G

or C) or the serine codon (TCT) to obtain a variety of mutants. Plasmids derived from

independent transformants were sequenced to verify the presence of mutations at the various

cysteine codons. The entire open reading frame of each mutant was also verified by sequencing.

The following mutants were generated: C117F, C117L, C117S, C210S, C239S, C298F, C298L,

C298S, C304S, C307S, C337S, C239S/C337S, C304S/C337S, and C307S/C337S. The plasmids

were transformed into Saccharomyces cerevisiae BJ5461 yeast (a pleiotrophic protease deficient


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

6

strain; Yeast Genetic Stock Center, Berkeley) by the lithium acetate/poly(ethyleneglycol)

method (20).

Yeast with recombinant plasmids were routinely grown to a suitable biomass in uracil-

deficient synthetic media with 0.1% glucose and 5% glycerol until OD600 was 0.3. Upon

induction with galactose (1 % final), xlHAS1 wild type or mutant enzyme accumulated in the

plasma membrane fraction. Crude membranes were prepared by disruption with silica/zirconia

beads (0.5 mm) in a MiniBead-Beater-8 (Biospec) and harvested by ultracentrifugation. The

membrane pellet was suspended in 50 mM Tris, pH 7.5, 0.1 mM EDTA, 1 µM E-64, 1 mM

benzamidine, 0.2 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride, and 5 µg/ml pepstatin.

Protein was quantitated by the Coomassie dye-binding assay (Pierce) using a bovine serum

albumin standard (21).

Polysaccharide Synthase Assays and Analyses – The incorporation of sugars into high

molecular-weight HA polysaccharide was monitored using UDP-[14C]GlcUA (~290 mCi/mmol;

NEN Life Sciences Products Inc.) and/or UDP-[3H]GlcNAc (29.2 Ci/mmol; NEN Life Sciences

Products Inc.) precursors as described previously (4, 19). Briefly, crude membranes were

incubated at 30°C in Tris buffer, pH 7.5, with MgCl2 and the UDP-sugar precursors.

Unincorporated, labeled UDP-sugars were separated from the HA product using paper

chromatography. HA at the origin of the paper strip was detected by liquid scintillation counting.

Assays were set so that <5% of the radiolabeled substrate was consumed and the enzyme

concentration was in the linear range. All HAS assays throughout this work were performed in

duplicate and the values were averaged.

The apparent Km values for the substrates were obtained by holding one radiolabeled UDP-

sugar at a constant and saturating concentration while titrating the other UDP-sugar. The


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

7

apparent Ki values (concentration of inhibitor required to reduce activity by 50%) of various

compounds were obtained by determining the HAS activity in the presence of varying

concentrations of the inhibitory compounds.

Chemical Modification and Enzyme Protection – For protection experiments, the enzyme

was first incubated with 200-1000 µM protecting compound on ice for 10 minutes. The enzyme

was then treated with 1 mM NEM at 15°C (5 µl reaction volume). After 15 minutes, the reaction

mixtures were diluted to 50 µl with DTT-containing buffer to quench any residual NEM and then

the residual HAS activity was determined.

For affinity experiments, the enzyme was treated with 1 mM NEM at 15°C in the absence or

presence of increasing concentrations of a protecting compound. The apparent affinity values of

various compounds for xlHAS1 were obtained indirectly by assessing their ability to protect the

enzyme from chemical inactivation. The apparent affinity values equal the concentration of

protecting compound required to yield 50% of the maximum protected activity. All kinetic data

was analyzed by graphing with rectangular hyperbola transformation in Sigma-Plot (Jandel

Scientific).

Immunochemical Detection of Polypeptides – The xlHAS1 and mutant proteins were

quantitated by Western blot analysis for assessment of the relative specific HAS activity. After

SDS-PAGE separation, the proteins in the gel were transferred to nitrocellulose by semi-dry

transfer. The blot was blocked with BSA and incubated with the primary reagent composed of

serum (1:1,000) from rabbits immunized with a fusion protein containing 1-166 residues of

xlHAS1 (gift of I. Dawid, ref. 22). Protein A-alkaline phosphatase detection with 5-bromo-4-

chloro-3-indolyl phosphate and nitroblue tetrazolium was used to visualize the immunoreactive

bands.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

8

RESULTS

Mutant Enzyme Expression and HAS Activity – The importance of some of the various

cysteines in xlHAS1 that are conserved among many Class I HASs (bacterial, viral, and

vertebrate) was assessed by site-directed mutagenesis. The membrane preparations containing

the various enzymes were tested for enzyme expression and HAS activity. The results are shown

in Table 2. The cysteine to serine mutation was typically found to be the least altering to protein

expression and activity in comparison to substitution with leucine or phenylalanine. The HAS

specific activity varied slightly among cysteine to serine mutants except for enzymes containing

the C307S mutation. xlHAS1(C307S) and xlHAS1(C307S/C337S) were expressed at levels

similar to that of wild type, but xlHAS1(C307S) retained less than 10% of wild type activity

while xlHAS1(C307S/C337S) had no detectable HAS activity. The Km values were similar for

all mutants except the series with the C337S mutation; UDP-GlcUA binds with lower affinity to

these mutants as assessed by higher Km values (Table 2).

Loss of HAS Activity Due to NEM Modification – In 1979, it was reported that a cysteine

modifying reagent, p-chloromercuribenzoate, inhibited the release of HA by streptococcal HAS

(23). We found that this reagent inactivated xlHAS1-catalyzed polymerization of HA (data not

shown). We tested the effect of NEM, a more selective cysteine modifying reagent, on the HAS

activity of the wild type and mutant enzymes. Membranes were incubated with varying

concentrations of NEM and the residual HAS activity was determined. xlHAS1 wild type and all

cysteine to serine mutants were inactivated by low levels of NEM, retaining <10% of their HAS

activity after treatment with 200 µM NEM (Fig. 1 and data not shown). xlHAS1 mutants

containing the C337S mutation were slightly more resistant to NEM-mediated loss of HAS


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

9

activity, with IC50 values (the concentrations of NEM that reduced the HAS activity by 50%) of

~75 µM NEM compared to ~35 µM NEM for all other mutants and wild-type (Fig. 1 and data

not shown). Results similar to wild type were obtained for xlHAS1(C307S) when more total

protein was used to accommodate the lower activity of this mutant (data not shown).

Protection from Loss of HAS Activity Due to NEM Modification – To determine if the

authentic HA substrate UDP-sugars, UDP-GlcUA and UDP-GlcNAc, could protect the wild-type

and mutant enzymes from loss of HAS activity due to NEM modification, membranes were

incubated with substrates before NEM treatment. If a substrate binds to the active site, then

NEM added later will be excluded from the site and the site’s modification rate will be

decreased. The HAS activity was then determined and compared to the activity of a parallel

aliquot of enzyme not treated with NEM. When preincubated with substrates, only

xlHAS1(C337S) could be protected from loss of HAS activity due to NEM. The modification of

C337 probably inactivates the enzyme by an indirect mechanism because the C337S mutant still

retains HAS activity. This mutant enzyme allows analysis of substrate binding characteristics

because an “irrelevant” (i.e. non-protectable) inactivation pathway has been eliminated. The

UDP-sugar protection effect also required Mg++ because no protection was observed when 10

mM EDTA chelator was added instead of Mg++ (data not shown).

Certain other structurally related compounds also protected xlHAS1(C337S) to various

extents from NEM-mediated loss of HAS activity (Table 3). The protecting compounds include

several UDP-sugars, thymine-containing nucleotides, and several other nucleotide triphosphates.

No nucleotide monophosphate protected the enzyme from NEM-mediated loss of HAS activity

(Table 3 and data not shown). The monosaccharides GlcUA and GlcNAc provided little

protection (Table 3).


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

10

The ability of some molecules to protect xlHAS1(C337S) from NEM-mediated loss of HAS

activity allowed for the investigation of the enzyme's apparent affinities for the various

compounds. Apparent affinity values were determined by titrating the protectant and measuring

the residual HAS activity after NEM treatment. UDP-GlcNAc yielded the highest maximum

protection, protecting ~60% of the HAS activity observed in the control not treated with NEM

(Fig. 2A). UDP-GalNAc, the C4 epimer, protected less than 20% of the control activity,

therefore, no apparent affinity value was obtained for this compound. The apparent affinity

values for all of the other compounds tested were about 10-4 M, except for UDP and UTP (Fig.

2B). Although UDP and UTP provided higher maximum protection than many of the other

compounds, xlHAS1(C337S) displayed approximately 10-fold lower apparent affinity (~10-3 M)

for UDP and UTP. When UDP or UTP were included together with UDP-GlcNAc, an apparent

affinity value similar to that of UDP-GlcNAc alone (~10-4 M) was obtained.

Inhibition of HAS Activity – The protection from NEM-mediated inactivation observed with

the added compounds is presumably due to the compounds binding to the enzyme and shielding

one or more cysteine residues from NEM modification. This cysteine(s) is hypothesized to be

either part of or near a substrate-binding site. To test this assumption of active site occupancy,

the compounds were tested for their ability to inhibit HAS activity (Table 4). As shown in

Figure 3, many of the compounds that protect xlHAS1(C337S) from NEM-mediated loss of HAS

activity (solid bars) also inhibit wild-type xlHAS1 (open bars). Similar inhibition was observed

with xlHAS1(C337S) (data not shown). Apparent Ki values were determined at two different

substrate concentrations and are listed in Table 4. The Ki values were higher when tested under

the higher authentic substrate concentration conditions. Km values of xlHAS1 were determined

for both substrates in the presence of UMP, UDP, and UDP-Glucose (Table 5). Although UMP


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

11

had little effect on the Km values, both UDP and UDP-Glucose raised the values considerably.

There was little effect on Vmax with any inhibitor. Overall, these alterations in kinetics are the

hallmark of competitive inhibition.

DISCUSSION

During the final preparation of this manuscript, it was reported that NEM inactivated HASs

from Group A and C Streptococcus (24, 25). These enzymes contain six and four cysteine

residues, respectively. The xlHAS1 polypeptide has 19 cysteines, therefore it is not surprising

that this enzyme is sensitive to treatment with NEM. When xlHAS1 is treated with biotin-

maleimide, the Western blot band shifts to ~5 kDa larger (data not shown), indicating that there

are ~10 free, readily available cysteines. xlHAS1 enzymes containing the C337S mutation were

slightly more resistant than wild type enzyme to loss of HAS activity due to NEM, indicating

that this residue is responsible in part for the NEM-mediated loss of activity. There are

obviously other cysteine residues that are modified by NEM since none of the cysteine to serine

mutants or double mutants tested thus far were completely resistant to NEM-mediated loss of

HAS activity.

The mutation of C337 to serine caused a large decrease in the enzyme's apparent affinity for

UDP-GlcUA (but not UDP-GlcNAc, Table 2), suggesting that C337 is somehow involved in the

binding of UDP-GlcUA or in GlcUA transfer. This result indicates that C337 is probably close

to the substrate-binding or catalytic site. The C337 residue, however, is probably not buried in a

substrate pocket or cleft because UDP-sugars could not protect the wild-type enzyme from

NEM-mediated inactivation.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

12

In a recent study on a rat glucosylceramide synthase, it was found that C207 was the primary

residue involved in the inactivation by NEM (26). It has also been recently found that the C226S

mutation in the HAS from equisimilis (24) and the C225S mutation in the HAS from pyogenes

(25) caused a reduction of about 90% and 50%, respectively, in the HAS activity. Interestingly,

based on sequence alignments, these residues roughly correspond to C307 in xlHAS1, which is

conserved in all known Class I HASs. As shown in Table 1, the C307S mutant lost almost all

HAS activity. A double mutant with C307S/C337S lost all measurable HAS activity. This

finding suggests that C307 plays a role in substrate binding, catalysis, or enzyme folding; the

latter explanation may not be as likely due to the mutant xlHAS1 enzyme's proper membrane

localization and good expression level.

Even though NEM inactivates the streptococcal HASs, it has been determined that no

cysteines are required for enzyme activity in these HASs (24, 25). Although it was speculated

that one or more of these cysteines are located in or near the active sites, the localization of the

cysteines to a substrate-binding site was not demonstrated by substrate protection from

inactivation. We show here that some of the 19 cysteines in xlHAS1 might be involved in

substrate binding or catalysis, based on kinetic and protection data. This involvement might not

be direct, but it is clear that loss or modification of some of these cysteines has significant effects

on the vertebrate enzyme’s ability to function as a HAS.

The fact that xlHAS1 is very specific for the substrates utilized in the HAS reaction could be

due to specificity at either the binding step or the catalytic step; our results in this report indicate

that the latter is most likely. Table 3 and Figure 3 show that many different compounds are able

to bind to the enzyme and protect it from NEM-mediated loss of HAS activity. The protection

observed with these non-substrate compounds was usually lower than that observed with the


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

13

authentic UDP-sugar substrates (Table 3). One of the most important structural elements appears

to be the pyrophosphate moiety because none of the nucleotide monophosphates tested were able

to protect the enzyme. It appears that xlHAS1 is able to bind purine-containing nucleotides (eg.

ATP, GDP) as well as many different pyrimidine-containing nucleotide-sugars (eg. UDP-

glucose). This finding suggests two potential hypotheses: (i) the enzyme can accommodate these

different shapes or (ii) the protection seen with these compounds is primarily due to interactions

with the phosphate groups. Neither simple phosphate ion nor pyrophosphate ion, however, were

able to protect xlHAS1(C337S) from NEM-mediated loss of HAS activity at a concentration of 1

mM (data not shown). This lack of effectiveness could be due to (a) the charge state differences

among these various phosphate ions, (b) the smaller size of these compounds (i.e. less sterically

hindered of access to pocket or cleft) allowing modification of cysteines that are normally

protected by other larger pyrophosphate-containing compounds, and/or (c) a requirement for

important interactions with the nucleotide base.

Mg++ was required for the protection phenomenon, thus it appears that the HAS enzyme

binds a UDP-sugar/metal complex. As mentioned earlier, the DXD-containing motif has been

implicated in coordination of the divalent metal cation and interaction with the phosphate groups.

All known Class I HASs contain a DXD-containing motif as described by Wiggins and Munro

(27) as well as an XDD motif similar to that seen in the putative UDP-sugar transferase SpsA of

Bacillus subtilis (7,8). When the first aspartate in DXD or the second aspartate in XDD was

mutated to glutamate in mouse HAS1, there was a 99% or greater loss of HAS activity (28).

These aspartate residues are probably involved in interactions with Mg++ and/or the phosphate

groups.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

14

Recently, the crystal structure for SpsA with dTDP has been obtained (8). This structure

shows that SpsA is able to accommodate the methyl group at position 5 of the pyrimidine ring.

Although the substrate specificity or transfer activity of SpsA has yet to be determined, it is

obvious that this protein has the ability to bind both dTDP and UDP (7,8). Our findings indicate

that xlHAS1 can not only accommodate a methyl group at position 5 of the pyrimidine ring, but

also a bromine or iodine atom at position 5, or a thiol at position 2. This suggests that there are

no intimate or essential interactions between xlHAS1 and positions 2 and 5 of the pyrimidine

ring. Interestingly, of the enzymes listed in Table 1, only half show an interaction with the

carbonyl at position 2 of the uridine in the crystal structure. xlHAS1 may interact with the

nucleotide base primarily by the hydrophobic effect due to its relatively promiscuous binding of

nucleotides, as assessed by the protection data.

When the two structures mentioned above for SpsA with different nucleotides are compared,

a shift in the contacts with the ribose ring can be seen to accommodate the loss of the 2' hydroxyl

(7,8). Since xlHAS1 can not only bind dTDP but also ddTTP, similar shifts are probably made

to accommodate the loss of either the 2' hydroxyl or both the 2' and 3' hydroxyls. The proteins

shown in Table 1 interact with the ribose hydroxyls but the requirements and importance in

catalysis or binding are not known. In the case of xlHAS1, contacts with the ribose hydroxyls

may not be essential.

Many of the compounds that provide substantial protection from NEM-mediated loss of HAS

activity also appear to be competitive inhibitors of xlHAS1. The Ki values of these inhibitors are

higher in the presence of increased substrate concentration (Table 4). UDP and UDP-Glucose

alter the Km for the two HAS substrates but do not significantly affect the Vmax (Table 5); these

results are indications of competitive inhibition.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

15

The apparent affinity obtained from Km studies is a measure of both the enzyme's ability to

bind the substrate as well as its ability to catalyze the addition of the sugar to the growing HA

chain. The apparent affinity values for various nucleotides (obtained from Figure 2) are

measurements of the ability of the enzyme to bind a substrate analog without concern for

catalytic ability. The protection from NEM observed with all compounds is probably due to

protection of one or more cysteines at a putative substrate-binding site. The UDP-GlcNAc

binding site is the most likely candidate because the most protection was observed with this

substrate. However, at this stage, it is impossible to determine which particular cysteines are

being protected and whether the cysteines participate in catalysis in the putative substrate-

binding site.

In Figure 3, dTDP, UDP, and thioUDP show higher inhibition of xlHAS1 in comparison to

their protection ability. xlHAS1(C337S) was found to have a much lower apparent affinity for

UDP and UTP than for the UDP-sugars (Fig. 2B). This result could be explained in several

ways, including (i) relative steric hindrance, (ii) multisite inhibition, and/or (iii) allosteric

regulation. As described below, we believe the first two effects may be responsible, in part, for

the observed disparity.

First, it is possible that the lack of sugar moieties on these nucleotides exposes one or more

cysteines in a substrate-binding pocket, which are normally protected by UDP-sugars, to

modification by NEM. The different levels of maximal protection observed in Figure 2 suggest

that there may be variations in the number of cysteines protected by the different compounds.

Specifically, it appears that at least one cysteine is protected by the GlcNAc portion of the UDP-

GlcNAc molecule. This cysteine is not efficiently protected by any of the other compounds,

leading to lower maximal protection values observed for all other compounds (Fig. 2).


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

16

Interestingly, UDP-GalNAc gave the lowest maximal protection, indicating that the position of

the C4 hydroxyl is probably critical for efficient or high affinity UDP-GlcNAc binding.

A second potential explanation for higher inhibition than protection is that nucleotides

without sugar moieties may be able to bind to both putative substrate-binding sites, thus

competing with both substrates simultaneously. This competition would explain the difference

in UDP-GlcNAc Km values observed with different concentration of UDP-GlcUA with xlHAS1

(Table 2), streptococcal HASs (29), and mouse HASs (30). The protection ability of the UDP-

like compounds, however, might be due to binding at only one of these sites (probably the UDP-

GlcNAc site), thus yielding a lower value in our NEM modification experiments.

A third possible explanation for the observation of greater inhibition compared to protection

is allosteric regulation; a site distinct from the catalytic sites would modulate polymerization

upon binding UDP. HA is extruded out of the cell once it is produced, therefore, it would be

difficult for the cell to determine the amount of HA produced directly. However, UDP, the

byproduct of the HAS reaction, may serve as a measure of synthesis rate or extent. Thus, the

local UDP concentration level near the synthase might serve as an internal indicator of the

amount of HA produced. However, the kinetics finding that UDP and UTP act as competitive

inhibitors does not support the solely allosteric control hypothesis. The sensitivity of the

vertebrate HASs to UDP may be an adequate potential negative feedback loop to control

synthesis levels by a competitive mechanism.

No three-dimensional structure for any HAS is available, thus the direct contacts between

substrate and enzyme are not known. Our work is the first assessment of the critical elements of

UDP-sugars required for binding to any HAS. A flexible xlHAS1 binding pocket probably

interacts with the hydrophobic nucleotide base and a metal-complexed pyrophosphate group.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

17

The identification of these critical elements of the substrate may allow for the future design of

HAS inhibitors to curtail HA polymer production in certain disease states. Based on sequence

similarities, it is probable that the streptococcal and vertebrate HASs interact with the same

elements. Also, we have made tentative assignments of the roles of two conserved cysteines

found in all vertebrate HASs. Further work on the details of the catalytic mechanism should

illuminate the nature of sugar transfer specificity.

Acknowledgements – We thank Dr. Ann Achyuthan for technical assistance in creating several

of the cysteine mutants and performing some chemical modification trials. We also thank Tasha

Arnett, Wei Jing, and Carissa White for aid in performing the many synthase activity assays and

for comments on the manuscript.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

18

REFERENCES

1. Laurent, T.C., and Fraser, J. R. E. (1992) FASEB J. 6, 2397-2404

2. Spicer, A. P., and McDonald, J. A. (1998) J. Biol. Chem. 273, 1923-1932

3. Weigel, P. H., Hascall, V. C., and Tammi M. (1997) J. Biol. Chem. 272, 13997-14000

4. Pummill, P. E., Achyuthan, A. M., and DeAngelis, P. L. (1998) J. Biol. Chem. 273, 4976-

4981

5. Breton, C. and Imberty, A. (1999) Curr. Opin. Struct. Biol. 9, 563-571

6. Ünligil, U. M. and Rini, J. M. (2000) Curr. Opin. Struct. Biol. 10, 510-517

7. Tarbouriech, N., Charnock, S. J., and Davies, G. J. (2001) J. Mol. Biol. 314, 655-661

8. Charnock, S. J. and Davies, G. J. (1999) Biochemistry 38, 6380-6385

9. Vrielink, A., Rüger, W., Driessen, H. P. C. and Freemont, P. S. (1994) EMBO J. 15, 3413-

3422

10. Moréra, S., Imberty, A., Aschke-Sannenborn, U., Rüger, W., and Freemont, P. S. (1999) J.

Mol. Biol. 292, 717-730

11. Moréra, S., Larivière, L., Kurzeck, J., Aschke-Sannenborn, U., Freemont, P. S., Janin, J., and

Rüger, W. (2001) J. Mol. Biol. 311, 569-577

12. Gastinel, L. N., Cambillau, C., and Bourne, Y. (1999) EMBO J. 18, 3546-3557

13. Ramakrishnan, B. and Qasba, P. (2001) J. Mol. Biol. 310, 205-218

14. Gastinel, L. N., Bignon, C., Misra, A. K., Hindsgaul, O., Shaper, J. H., and Joziasse, D. H.

(2001) EMBO J. 20, 638-649

15. Persson, K., Ly, H. D., Dieckelmann, M., Wakarchuk, W. W., Withers, S. G., and Strynadka,

N. C. J. (2001) Nature Struct. Biol. 8, 166-175


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

19

16. Pedersen, L. C., Tsuchida, K., Kitagawa, H., Sugahara, K., Darden, T. A., and Negishi, M.

(2000) J. Biol. Chem. 275, 34580-34585

17. Ünligil, U. M., Zhou, S., Yuwaraj, S., Sarkar, M., Schachter, H., and Rini, J. M. (2000)

EMBO J. 19, 5269-5280

18. Campbell, R. E., Mosimann, S. C., van de Rijn, I., Tanner, M. E., and Strynadke, N. C. J.

(2000) Biochmistry 39, 7012-7023

19. DeAngelis, P. L., and Achyuthan, A. M. (1996) J. Biol. Chem. 271, 23657-23660

20. Gietz R. D., Schiestl R. H., Willems A. R., Woods R. A. (1995) Yeast 11, 355-360

21. Bradford, M. M. (1976) Anal. Biochem. 72, 248-254

22. Rosa, F., Sargent, T. D., Rebbert, M. L., Michaels, G. S., Jamrich, M., Grunz, H., Jonas, E.,

Winkles, J. A., and Dawid, I. B. (1988) Dev. Biol. 129, 114-123

23. Sugahara, K., Schwartz, N. B., and Dorfman, A. (1979) J. Biol. Chem. 254, 6252-6261

24. Kumari, K., Tlapak-Simmons, V. L., Baggenstoss, B. A., and Weigel, P. H. (2001) J. Biol.

Chem. epub ahead of print, M110638200

25. Heldermon, C. D., Tlapak-Simmons, V. L., Baggenstoss, B. A., and Weigel, P. H. (2001)

Glycobiology 11, 1017-1024

26. Marks, D. L., Dominguez, M., Wu, K., and Pagano, R. E. (2001) J. Biol. Chem. 276, 26492-

26498

27. Wiggins, C. A. R., and Munro, S. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 7945-7950

28. Yoshida, M., Itano, N., Yamada, Y., and Kimata, K. (2000) J. Biol. Chem. 275, 497-506

29. Tlapak-Simmons, V. L., Baggenstoss, B. A., Kumari, K., Heldermon, C., and Weigel P. H.

(1999) J. Biol. Chem. 274, 4246-4253


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

20

30. Itano, N., Sawai, T., Yoshida, M., Lenas, P., Yamada, Y., Imagawa, M., Shinomura, T.,

Hamaguchi, M., Yoshida, Y., Ohnuki, Y., Miyauchi, S., Spicer, A. P., McDonald, J. A., and

Kimata, K. (1999) J. Biol. Chem. 274, 25085-25092


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

21

Table 1

Polypeptide-Substrate Contacts in Enzymes Utilizing UDP-sugars

The residues reported to make distinct interactions with the nucleotide portion of the substrate

based on X-ray crystallography and the nature of the protein/substrate interaction are listed for

eight different enzymes that utilize UDP-sugars. The enzymes are: 1, Bacillus subtilis SpsA

(7,8); 2, bacteriophage T4 β-glucosyltransferase (9-11); 3, bovine β1,4-galactosyltransferase T1

(12,13); 4, bovine α1,3-galactosyltransferase (14); 5, Neisseria meningitidis LgtC

galactosyltransferase (15); 6, human β1,3-glucuronyltransferase I (16); 7, rabbit N-

acetylglucosaminyltransferase I (17); and 8, Streptococcus pyogenes UDP-glucose

dehydrogenase (18).


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

22

Enzyme #

Interaction 1 2 3 4 5 6 7 8

uridine

stacking

Y11 F213

G214

I238

V243

R191

F226

Y139

W195

I198

Y11 Y84 I187 V215

uridine O2 V136 D8 H190 S253

uridine N3 D39 I238 R189 V136 D8 D113 D144 N251

uridine O4 R71 I238 R189 N10 N251

ribose 2'

OH

D98 E272 P187 F134 A6 P82

D195

D212 D402

ribose 3'

OH

T9 E272 D252

V253

R202

V226

I104 D195 D212 G257

M319

α-PO4 K13 S189

R191

R269

Q187

K250

R156 R117

V321

Y249

K320

β-PO4 R191

R195

Y265

R269

K279

W314

H78

G247

R310 S322 E145

K320

metal-PO4 D99 E163 D254

M344

H347

D225

D227

D103

D105

H244

D194

D196

D213


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

23

Table 2

Enzyme Expression, HAS Activity, and Substrate Affinity

Membrane preparations were tested for enzyme expression, HAS activity, and substrate affinity

(apparent Km) as described in Experimental Procedures. Protein expression and HAS activity

were categorized as: >50% of wild type (+++), 10-50% of wild type (++), 0.1-10% of wild type

(+), or <0.1% of wild type (–). Constant precursor concentrations were 1.2 mM UDP-GlcUA for

the UDP-GlcNAc Km determination and 2.4 mM UDP-GlcNAc for the UDP-GlcUA Km

determination, unless noted otherwise by a footnote. All data points were obtained in duplicate.

Standard deviations are given for experiments performed at least three times.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

24

xlHAS1

Enzyme

Protein

Expression

HAS

Activity

UDP-GlcNAc

Km (µµµµM)

UDP-GlcUA

Km (µµµµM)

Wild Type +++ +++ 400 ± 100, 260a, 700b 190 ± 40, 110c

C117F – – n.d. n.d.

C117L +++ ++ n.d. n.d.

C117S +++ +++ 340 120

C210S +++ ++ 400 120

C239S +++ +++ 320 110

C298F + + n.d. n.d.

C298L + + n.d. n.d.

C298S +++ +++ 430 160

C304S +++ +++ 470 180

C307S +++ + n.d. n.d.

C337S +++ +++ 400 ± 100, 880b 700 ± 170

C239S/C337S ++ +++ 1000 930

C304S/C337S +++ ++ 1100 890

C307S/C337S ++ – n.d. n.d.

n.d. - not determined

Alternative concentrations of substrate were used to assess these particular Km values: a, 0.6 mM

UDP-GlcUA; b, 2.4 mM UDP-GlcUA; c, 1.2 mM UDP-GlcNAc.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

25

Table 3

Nucleotide Protection of xlHAS1 from NEM Inactivation

Four independent sets of protection experiments are shown. Membranes containing xlHAS1

(C337S) were incubated with 200 µM (Runs 1-3) or 1 mM (Run 4) of the indicated substrate or

analog compound for 10 minutes on ice followed by treatment with 1 mM NEM for 10 minutes

at 15°C. The membranes were then diluted 10-fold into assay buffer (containing 10 mM DTT to

quench the remaining NEM) and assayed as described in Experimental Procedures. The values

(averaged duplicates) indicate the percentage of HAS activity remaining compared to an

identical incubation performed without NEM (i.e. higher percentage means better protection).

Without protection under these conditions, ~80-90% of the enzyme is inactivated by the NEM.

The best protectant in all cases is UDP-GlcNAc. At least slight protection is observed for all

pyrophosphate-containing nucleotides.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

26

Protection (%)

Compound Run # 1 2 3 4

none 8 19 13 16

ATP 26 29 38

CTP 20 33 17 40

GTP 29 26 29

ITP 27 31 13 35

dTMP 13 12

dTDP 73 42 38

dTTP 69 40 39

ddTTP 65 46 23 51

UMP 12 11 15

UDP 47 43 25 41

UTP 60 44 25 46

5-BromoUTP 63 42 37

5-IodoUTP 57 40 34

2-ThioUDP 63 54 40

UDP-Mannose 59 22 57

UDP-Xylose 16 35

UDP-Galactose 30 52

UDP-GalUA 49 38 25 38

UDP-GalNAc 34 34 15 27

UDP-Glucose 69 41 29 53

UDP-GlcUA 55 42 23 39

UDP-GlcNAc 77* 68* 44* 69*

GlcUA 15 13

GlcNAc 19 14

*compound providing most protection in a given run.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

27

Table 4

Inhibition Constants for Uridine Nucleotides

Apparent Ki values were determined for each compound by titrating each compound (0-4.5 mM)

into the standard synthase assay; the Ki value equals the inhibitor concentration required to give

50% inhibition. The results from two independent experiments are shown. All compounds

completely inhibited HAS activity at elevated concentrations. Values were determined at low

substrate concentration (0.3 mM UDP-GlcUA and 0.6 mM UDP-GlcNAc) or at high substrate

concentration (1.2 mM UDP-GlcUA and 2.4 mM UDP-GlcNAc). Ki values increase with higher

substrate concentrations, indicating a competitive mode of inhibition.

low substrate high substrate

Ki (µM) Ki (µM)

Compound Exp. 1 Exp. 2 Exp. 1 Exp. 2

UDP 150 130 550 600

UDP-Glucose 290 230 1100 1200

UTP 600 450 1000 ≥1500


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

28

Table 5

Effect of Various Nucleotides on Apparent Substrate Affinity

Apparent Km and Vmax values were determined for xlHAS1 in the presence of 200 µM (Exp. 1) or

300 µM (Exp. 2) of the indicated compound. The constant UDP-sugar concentrations were 1.2

mM UDP-GlcUA for the UDP-GlcNAc Km determination or 2.4 mM UDP-GlcNAc for the

UDP-GlcUA Km determination. Vmax values are in units of pmoles of sugar transferred per

minute.

UDP-GlcNAc

Km (µµµµM)

UDP-GlcUA

Km (µµµµM)

UDP-GlcNAc

Vmax

UDP-GlcUA

Vmax

Compound Exp. 1 Exp. 2 Exp. 1 Exp. 2 Exp. 1 Exp. 2 Exp. 1 Exp. 2

none 250 270 130 170 160 180 110 90

UMP – 310 – 250 – 200 – 150

UDP 720 ≥1500 200 460 150 210 100 100

UDP-Glucose 480 630 300 690 130 180 110 130


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

29

FIGURE LEGENDS

Fig. 1. NEM Effect on HAS Activity. Membranes containing wild-type xlHAS1 or assorted

mutant enzymes (60 µg total protein) were incubated with varying concentrations of NEM for 15

minutes at 15°C. The membranes were then diluted 10-fold into assay buffer and assayed as

described in Experimental Procedures. Results for representative mutants are shown. Legend:

wild type (solid circle), C239S (solid diamond), C304S (solid triangle), C337S (solid square),

C239S/C337S (open diamond), C304S/C337S (open triangle). The mutants containing the

C337S mutation are more resistant to NEM-mediated inactivation relative to the enzymes with

C337.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

30

Fig. 2. Protecting Compound Affinities. Apparent affinity values were determined for each

compound by titrating in the compound up to 3 mM prior to 1 mM NEM treatment at 15°C for

10 minutes. The membranes were then diluted 10-fold into assay buffer (containing 10 mM

DTT to quench the remaining NEM) and assayed for HAS activity with 0.6 mM UDP-GlcUA

and 1.2 mM UDP-GlcNAc. Results are given as percentage form of [(activity of NEM-treated,

compound-protected enzyme) / (activity of untreated control with same amount of compound)] –

[(activity of NEM-treated, unprotected enzyme) / (activity of untreated control with no

compound)]. Averages of duplicate results from up to four independent experiments are shown

along with rectangular hyperbola curves. Panel A, UDP-Glucose (solid triangle, solid line),

UDP-Galactose (open triangle, dashed line), UDP-GlcUA (solid square, solid line), UDP-

GalUA (open square, dashed line), UDP-GlcNAc (solid diamond, solid line), and UDP-GalNAc

(open diamond, dashed line). Panel B, UDP (solid circle), UTP (solid square), UDP-GlcNAc

(solid diamond, solid line), equimolar UDP + UDP-GlcNAc (open circle, dashed line), and

equimolar UTP + UDP-GlcNAc (open square, dashed line). Although the rectangular hyperbola

curves for UDP + UTP gave a maximum protection of ~40% and ~70% of control, respectively,

the apparent affinities were about 10-fold lower than that obtained for the other compounds.


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

31

Fig. 3. Comparison of Protection from NEM Inactivation Versus Inhibition of HAS

Activity. For protection experiments, percentages of control activity for Run 4 from Table 3 are

represented graphically with solid bars. For inhibition experiments, membranes containing

xlHAS1 were assayed with or without 1 mM of the indicated compound in addition to 0.6 mM

UDP-GlcUA and 1.2 mM UDP-GlcNAc. Open bars indicate the percent inhibition of HAS

activity compared to the parallel reaction without added compound. UDP-GlcNAc is best

protectant, but nucleotides with at least two phosphates protect a moderate amount.

(*, authentic substrate is not applicable for inhibition study)


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

32

Fig. 1

0

0.2

0.4

0.6

0.8

1

1.2

0 50 100 150 200

[NEM] µµµµM

Fra

ctio

n o

f C

ontr

ol A

ctiv

ity


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

33

Fig. 2

A

B

[Protectant] mM

0.0 0.5 1.0 1.5 3.0

% P

rote

ctio

n

0

10

20

30

40

50

60

70

[Protectant] mM

0.0 0.5 1.0 1.5 3.0

% P

rote

ctio

n

0

10

20

30

40

50

60

70


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

34

Fig. 3

0

20

40

60

80

100

none

AT

P

CT

P

GT

P

ITP

dTM

P

dTD

P

dTT

P

ddT

TP

UM

P

UD

P

UT

P

5-B

rom

oUT

P

5-Io

doU

TP

2-T

hioU

DP

UD

P-M

anno

se

UD

P-X

ylos

e

UD

P-G

alac

tose

UD

P-G

alU

A

UD

P-G

alN

Ac

UD

P-G

luco

se

UD

P-G

lcU

A

UD

P-G

lcN

Ac

% C

ontr

ol A

ctiv

ity (

Pro

tect

ion)

% I

nhib

ition

of

HA

S A

ctiv

ity

* *


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Philip E. Pummill and Paul L. DeAngelisresidues of a vertebrate hyaluronan synthase

Evaluation of critical structural elements of UDP-sugar substrates and certain cysteine

published online April 9, 2002J. Biol. Chem.

10.1074/jbc.M202456200Access the most updated version of this article at doi:

Alerts:

When a correction for this article is posted•

When this article is cited•

to choose from all of JBC's e-mail alertsClick here


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/lookup/doi/10.1074/jbc.M202456200

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;M202456200v1&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/early/2002/04/09/jbc.M202456200.citation

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=early/2002/04/09/jbc&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/early/2002/04/09/jbc.M202456200.citation

http://www.jbc.org/cgi/alerts/etoc

http://www.jbc.org/

cysteine mutagenesis

Documents