THEJOURNAL or BIOLOQICAL CHEMISTRY
Vol. 235, No.6,May 1960
Printed in U.S.A.
Glucosamine
V. ENZYMATIC SYNTHESIS
SUDHAMOY GHOSH,~ HAROLD J. BLUMENTHAL,~
From the Rackham Arthritis Research Unit and the Departments of Biological Chemistry and Bacteriology, The Unive&ty
of Michigan, Ann Arbor, Michigan
(Received for publication, November 27, 1959)
Studies in viva on the biosynthesisof n-glucosamine,with the
useof specificallylabeledglucosein the rat (1,2) and in group A
streptococci (3-5) showedthat the glucosecarbon chain waspre-
servedduring its conversionto the hexosamine. Leloir and Car-
dini (6, 7) noted that extracts obtained from Neurosporacrassa
converted glutamineand hexosephosphateto glutamic acid and
a product tentatively characterized asglucosamine6-phosphate.
Owing to the presenceof phosphoglucoisomerase in the extract,
it wasnot known whether glucose6-phosphateor fructose6-phos-
phate was the acceptor of the amido group of glutamine. A
preliminary report (8) indicated that such extracts could be
fractionated to remove isomeraseand that glucosamine6-phos-
phate synthesisoccurredwith fructose 6-phosphatebut not with
glucose6-phosphate. Satisfactory stoichiometry studies were
not possiblewith thesepreparationsdue to their lability and to
the considerableactivity of contaminating glutaminase. Subse-
quent work in this laboratory was performed with extracts ob-
tained from bacteria and from rat liver (9).
Studieson the rat liver enzyme were first reported by Pogell
and Gryder (lo), who obtained activity with both hexosephos-
phates. On the basisof relative activity studies and stability
experiments,they concludedthat glucose6-phosphateand not
fructose 6-phosphatewas the active substrate in the reaction
with glutamine.
The present paper gives the purification proceduresand prop-
erties of the enzymesfrom N. crassa, Escherichiacoli, and rat
liver. As previously noted (9) each of the preparations was
found to be specificfor fructose 6-phosphate. Leloir (6, 7) sug-
gestedthat the enzyme be namedtransamidasewhen its hexose
phosphatespecificity was unknown. In view of the presentre-
sults, we suggestthat the enzyme be called L-glutamine-n-fruc-
tose 6-phosphatetransamidase.
EXPERIMENTAL
Matirials-The hexosamines,their phosphateesters,and their
N-acetyl derivatives were preparedas previously described(II-
* The Rackham Arthritis Research Unit is supported by a grant
from the Horace H. Rackham School of Graduate Studies of The
University of Michigan. This investigation was supported in
Dart bv a arant from the National Institutes of Health. and a arant
from the Michigan Chapter, Arthritis and Rheumatism Fo&da-
tion.
t Postdoctoral Fellow, Helen Hay Whitney Foundation.
$ Present address, Department of Bacteriology, The University
of Michigan, Ann Arbor, Michigan.
5 Present address, Department of Biochemistry, Duke Univer-
sity, Durham, North Carolina.
Metabolism
OF GLUCOSAMINE 6-PHOSPHATE*
EUGENE DAVIDSON,~ AND SAUL ROSEMAN
13). The followingwerecommercialproducts: glucose-6-P,man-
nose-6-P, fructose-6-P, glutamine, glutamic acid, glucosamine
HCl, Tris, EDTA,’ TPN, protamine sulfate, polymyxin sulfate,
Veronal. For someof the morecritical studies,fructose-6-Pwas
purified by ion exchangechromatography. Azaserineand DON
Downloaded from http://www.jbc.org/ by guest on August 27, 2017
were gifts from Dr. H. M. Crooks, Jr., Parke Davis and Com-
pany. Potato phosphatasewas prepared by an unpublished
method of Dr. A. Kornberg. Calcium phosphategel was pre-
pared by the method of Singerand Kearney (14), and contained
32 mg of solidsper ml.
A buffer mixture, hereafter called GPV, was routinely used
during the purification of the E. coli and N. crassaenzymes. The
mixture contained the following in pmolesper ml: 150, potas-
sium phosphate buffer, pH 7.0; 60, n-glutamine; 10, EDTA.
Methods-Glucosamine-6-Pwasoccasionallyestimatedby the
acetylacetone method of Blix (15), but routinely by the pro-
cedure given in detail below. Fructose-6-Pwas determinedby
the Roe method (16); protein wasdeterminednephelometrically
(17).
Mixtures of hexosephosphate,particularly fructose-6-P, glu-
tamine yield false positive colors in the acetylacetone procedure,
especiallyin the presenceof protein, salts,and others. Further-
more, sincefructose-6-Pand glutamineyield morespuriouscolor
than glucose-6-Pand glutamine, it is difficult to comparetrans-
amidase activity in the presence of the two hexose phosphates
when phosphoglucoisomerase is present;heat-inactivated enzyme
or “zero time” controls are not adequateblanks. For this rea-
son, a procedurewasdevised for the estimation of glucosamine-
6-P by a modification of the method of Morgan and Elson (18,
19). Samplescontaining glucosamine-6-P(0.02 to 0.20 pmole)
in 0.80 ml of water were treated with 0.10 ml of saturated Na-
HCO, solution and 0.10 ml of ice-cold, 5%, aqueousacetic an-
hydride solution, in that order. After vigorous shaking, the
tubesweremaintainedat roomtemperaturefor 3 minutes,heated
in a boiling water bath for 3 minutes to destroy excessacetic
anhydride, and cooledto room temperature. The sampleswere
treated with 0.20 ml of 0.80 M sodium borate solution (adjusted
to pH 9.0), mixed, and heated in a boiling water bath for 3 min-
utes. After cooling, the sampleswere treated with 6.0 ml of
Ehrlich’s reagent (1.Og of p-dimethylaminobenzaldehydeadded
to 1.25ml of 10 N HCl and then diluted to 100 ml with glacial
acetic acid). Finally, the samples were incubated for 20 min-
utes at 37”, cooled, the optical densitiesmeasuredin a Beck-
1 The abbreviations used are: EDTA, disodium ethylenedia-
minetetraacetate; DON, 6-diazo-5-oxo-n-norleucine; azaserine, O-
diaxoacetyl-n-serine.
1265
1266 Glucosamine Metabolism. V Vol. 235, No. 5

man spectrophotometer at 585 rnp, and compared with glucosa- months, but after this time the extractable activity declined
mine standards treated in the same manner. With this modi- rapidly.
fication of the Morgan-Elson method, glucosamine-6-P yields N. crassa was grown at temperatures ranging from 25 to 28”
approximately 85% of the optical density obtained with an equi- in Erlenmeyer flasks on a rotary shaker. Sucrose-minimal me-
molar solution of glucosamine. dium (20) was used in these studies under conditions previously
Glutamine and glutamic acid were determined by a ninhydrin described. After growth for 18 to 20 hours, the culture was
method after paper electrophoresis. Known volumes of sam- filtered through cotton gauze, the mycelia washed with water,
ples were applied to l-inch strips of Whatman 3 MM paper and lyophilized, and stored at -18” in a vacuum in a desiccator.
subjected to electrophoresis in the Spinco model R paper elec- About 4 g of dried mycelia were obtained from 1 liter of the me-
trophoresis apparatus with 0.01 M citrate, 0.02 M phosphate dium.
buffer, pH 6.0, at 450 volts for 1 hour. After being dried at Puri$cation of Enzyme from E. co&-Unless otherwise indi-
110” for 7 minutes, the paper strips were dipped in a solution cated, all operations were conducted between 0 and 4” and cen-
containing 0.5% ninhydrin in acetone. After 8 hours at room trifugations for 15 minutes at 20,000 X g.
temperature, the color densities were determined in a Spinco Step 1. The frozen E. co& cells (50 g) were ground in a chilled
Analytrol Densitometer. The densities were proportional to mortar with alumina (50 g; A-301, Aluminum Company of
concentrations of glutamic acid and glutamine between 0.025 America) and the mixture was extracted with 125 ml of 0.15 M
pmole to 0.20 pmole. Under these conditions, glutamine, glu- KC1 solution containing 1 mg per ml of reduced glutathione.
tamic acid, glucosamine, and glucosamine-6-P are readily separa- After centrifugation, the supernatant solution was collected, the
ble. residue further extracted with another 75 ml of KCl-glutathione

Downloaded from http://www.jbc.org/ by guest on August 27, 2017

Enzyme Assay-The assay mixture routinely used for the en- solution, and the two supernatant fluids combined (crude ex-
zyme obtained from rat liver contained the following in pmoles tract).
per ml: 10, fructose-6-P; 12, n-glutamine; 50, potassium phos- Step 2. The crude extract, 180 ml, was treated with 75 ml
phate buffer, pH 7.7; 1, EDTA; and protein to be assayed. Af- of 2% polymyxin sulfate solution, which had been adjusted to
ter incubation at 37” for 30 minutes, the reaction was stopped pH 7.0 with dilute ammonium hydroxide. After being stirred
by heating at 100” for 2 minutes, and an aliquot of the super- for 10 minutes, the mixture was centrifuged, and the residue dis-
natant fluid was used for determination of glucosamine-6-P. carded.
Assay by the acetylacetone procedure yielded large and variable Step 3. DEAE-cellulose (Type 20; Brown and Company) was
blanks, whereas assay by the acetic anhydride method gave equilibrated with a solution containing 0.01 M potassium phos-
blanks which showed less than 10% of the absorbancies obtained phate buffer, pH 7.0, and 0.02 M KCl. The cellulose suspension
with active protein fractions. (160 ml containing 73 mg dry weight of cellulose per ml) was cen-
The fractions obtained from E. coli were assayed with the fol- trifuged, and the residue was stirred gently for 5 to 10 minutes
lowing incubation mixtures &moles per ml): 20, fructose-6-P; with 240 ml of the polymyxin supernatant solution. Most of the
15, n-glutamine; 37.5, potassium phosphate buffer, pH 7.0; 2.5, enzyme was absorbed by the cellulose; the supernatant fluid was
EDTA; and protein. The remainder of the procedure was con- discarded after centrifugation. The cellulose residue was then
ducted as outlined above. washed and extracted by stirring with 200 ml portions of 0.008
Assay mixtures for the N. crassa enzyme contained the follow- M potassium phosphate buffer, pH 7.0, containing KC1 of increas-
ing in pmoles per ml: 20, fructose-6-P; 30, glutamine; 75, potas- ing molarities; extractions were conducted for 5 minutes followed
sium phosphate buffer, pH 6.6; 5, EDTA; and protein. Incuba- by centrifugation. The KC1 concentrations were: 0.10 M, 0.15
tions and assays were performed as outlined above, except the M, 0.20 M, and 0.50 M. The enzyme activity was found in the
period of incubation which was 60 minutes to increase the yield latter two fractions, which will be called cellulose eluates I and
of product. II.
In each case, 0.05 to 0.4 unit of enzyme was added per ml of Step 4. Cellulose eluate I (180 ml) was treated with 14.4 ml
incubation mixture. A unit of enzyme is defined as the quantity of 2% protamine sulfate solution adjusted to pH 6.5. After
that produced 1 pmole of glucosamine-6-P per 30 minutes under stirring for 5 to 10 minutes, the precipitated protein was col-
the conditions of assay described above. Routine controls in the lected by centrifugation, washed once with 18 ml of a solution
above assays included a complete, but nonincubated mixture, a containing 0.015 M KC1 and 0.008 M potassium phosphate buffer,
mixture without substrate, and a mixture without enzyme. The pH 7.0, and the enzyme then extracted from the residue with
product of enzyme action, glucosamine&P, yields less color on a 19.8 ml of a solution containing 18 ml of 0.15 M pyrophosphate
molar basis than glucosamine by the method described above buffer, pH 6.7, plus 1.8 ml of GPV. The supernatant solution,
(12, 13); in routine fractionation procedures no correction was containing the enzyme, was collected by centrifugation (pro-
made for this difference. tamine eluate I).
Growth of Microorganisms-E. coli B was grown either in a The supernatant liquid (194 ml) obtained after the addition of
g-gallon fermenter or in 5-gallon carboys containing media of the protamine to cellulose eluate I, was again treated with protamine
following composition in g per liter: Na2HP04, 6.0; KHZPOI, sulfate solution (67 ml of a 2 y0 solution), and the precipitate col-
3.0; NaCI, 5.0; NH&l, 2.0; MgS04, 0.10; glucose, 8.0. The glu- lected by centrifugation. The precipitate was washed with 20
cose and MgS04 were sterilized separately and added to the salt ml of a solution containing 0.008 M potassium phosphate buffer,
mixture. After growth at 3334” for 18 hours, with vigorous pH 7.0, and 0.015 M KCl, and finally extracted with a mixture of
aeration, the bacteria were harvested in a Sharples centrifuge, 5 ml of 0.15 M pyrophosphate buffer, pH 6.7, and 0.5 ml of GPV
washed with 0.10 M KCl, and stored in the frozen state until solution. After centrifugation, the supernatant solution was col-
ready for use. The cells could be stored for periods up to 2 lected (protamine eluate II).
May 1960 S. Ghosh, H. J. Blumenthal, E. Davidson, and S. Roseman 1267

Protamine precipitation of the enzyme followed by extraction formed between 0 and 4”. Centrifugations were conducted at
removes phosphohexoisomerase. The pyrophosphate buffer is 20,000 x g for 5 minutes with the exception of the calcium phos-
effective as an eluting agent, since it combines with protamine to phate gel steps which were performed for 3 minutes.
form an insoluble precipitate. The crude extract was prepared by treating 0.7 g of the lyoph-
Pur$cation of Enzyme from Rat Liver-All operations were ilized mycelia with 20 ml of GPV buffer (see above) and about
conducted between 0 and 4”. Liver (30 g) obtained from ex- 7 g of glass beads (described above) in a jacketed semimicro
sanguinated adult rats was rapidly chilled by dropping into an Waring Blendor. The fluid circulating around the Blendor was
ice cold solution (90 ml) containing 0.125 M KCl, 0.004 M EDTA, maintained at -2” and the homogenization was performed for
2 mg per ml mercaptoethanol, adjusted to pH 7.2. The liver 3 minutes. Centrifugation yielded a slightly turbid, supernatant
was removed from the solution and ground thoroughly in a fluid, which was light orange in color (crude extract).
chilled mortar with approximately 11 g of glass beads (Super- The crude extract (12 ml) was treated with 4.6 ml of a 2%
brite pavement marking beads, 115 regular; Minnesota Mining solution of protamine sulfate. After occasional stirring for 10
and Manufacturing Company). The above mixture was slowly minutes, centrifugation yielded a clear tan supernatant fluid
added to the mortar and, after being ground, was centrifuged at (protamine supernate).
18,000 X g for 1 hour. The supernatant liquid Fraction I (82 Calcium phosphate gel (2.2 ml) was slowly added with stirring
ml) was treated with 2oJ, polymyxin sulfate solution (18 ml; ad- to 15 ml of the supernatant solution, and the mixture was stirred
justed to pH 7.0 with ammonia) and t,he precipitate formed on for an additional 10 minutes before centrifugation. The gel was
stirring was removed by centrifuging at 20,000 x g for 10 min- then washed for 2 minutes with 10 ml of dilute GPV buffer (1
utes. The supernatant solution (98 ml), was designated Frac- part GPV, 3 parts water). After centrifugation, the supernatant

Downloaded from http://www.jbc.org/ by guest on August 27, 2017

tion II. fluid, which contained most of the phosphohexoisomerase, was
DEAE-cellulose (7.5 g dry weight) was equilibrated with a discarded. The gel was again washed by suspension in 10 ml
solution containing 0.02 M KC1 and 0.01 M potassium phosphate of dilute GPV buffer (1 part GPV, 1 part water), and then cen-
buffer, pH 7.0. The mixture was filtered through cotton gauze, trifuged. Finally, the enzyme was eluted by suspension of the
and the cellulose was added to 95 ml of Fraction II. After gel in 15 ml of stock GPV buffer, followed by stirring for 10
stirring the mixture gently for 10 minutes, the cellulose was sepa- minutes before centrifugation. The supernatant fluid was
rated from the liquid (Fraction III) by filtering through cotton treated as described below or was lyophilized to yield a white
gauze, and was then washed twice with 200 ml of solution con- fluffy powder which could be stored at -18” in a vacuum. For
taining 0.10 M KCl, and 0.010 M potassium phosphate buffer, pH further fractionation, the powder was dissolved in a volume of
7.0. The washings were discarded. water equal to that removed by lyophilization; this fraction will
The cellulose was suspended in 50 ml of solution containing be called calcium phosphate gel I.
0.20 M KCl, 25 mg of mercaptoethanol, and 0.010 M potassium Ammonium sulfate solution, saturated at 2’, and adjusted to
phosphate buffer, pH 7.2, and the mixture stirred gently for 5 pH 6.3 with aqueous ammonia was added to the gel eluate ob-
minutes to elute the enzyme. After removing the cellulose by tained in the preceding step (volume for volume), and after
filtration through gauze, it was once again extracted with the stirring occasionally for 15 minutes, the mixture was centrifuged
KCl-mercaptoethanol-potassium phosphate solution. The elu- for 15 minutes. The precipitate was dissolved in 6 ml of dilute
ates were combined and centrifuged to remove a small amount GPV buffer (1 part GPV, 9 parts water).
of suspended cellulose (Fraction IV, approximately 100 ml.). The enzyme solution (5.0 ml) was treated with 0.7 ml of cal-
Fraction IV was then dialyzed for 2 hours with stirring against cium phosphate gel, and the suspension centrifuged after stand-
5 liters of solution containing 0.005 M potassium phosphate, 0.0005 ing for 10 minutes. The gel was washed once with 10 ml of dilute
M EDTA, and 0.50 mg of mercaptoethanol per ml, pH adjusted GPV buffer (1 part GPV, 4 parts water), and the enzyme was
to 7.2. A small precipitate, formed during dialysis, was removed eluted from the gel by resuspension in 5 ml of stock GPV buffer.
by centrifugation. The yield of supernatant liquid (Fraction V) The final supernatant fluid was colorless, and contained the most
was approximately 100 ml. highly purified N. crassa enzyme fraction (calcium phosphate gel
Calcium phosphate gel (0.3 mg on dry weight basis per mg II).
protein in Fraction V) was added to Fraction V, stirred gently A summary of the data on the purification of the various en-
for 5 minutes, and centrifuged. The supernatant liquid (Frac- zyme preparations is presented in Tables I, II, and III.
tion VI) was removed, and the gel washed twice with 5 ml of Storage and Stability of Enzyme Preparations-Purified E. coli
0.10 M phosphate buffer, pH 7.0, and finally eluted twice with enzyme preparations were relatively stable in GPV buffer at 0”
2 ml of a mixture containing 0.30 M potassium phosphate buffer, to 2” for periods up to 3 or 4 weeks. Freezing and thawing in-
pH 7.2, 0.005 M EDTA, and 5 mg per ml of mercaptoethanol activated the enzyme; various techniques were tested to stabilize
(Fraction VII). the preparation without success.
Fraction VII was dialyzed for 3 hours against 2 liters of 0.005 The rat liver enzyme preparations were very unstable, there-
M potassium phosphate buffer, 0.0005 M EDTA, 0.5 mg per ml of fore requiring that studies with these preparations be conducted
mercaptoethanol, pH adjusted to 7.2. After dialysis, a small on the day they were made. When Fraction VII was stored in
precipitate was removed by centrifugation, and the supernatant 0.30 M phosphate buffer, pH 7.2, 0.005 M EDTA, and 5 mg per
liquid was retained (Fraction VIII). ml mercaptoethanol, it lost half of its activity in 24 hours and all
Owing to the instability of this enzyme, the entire procedure the activity in about 48 hours either at 0” or at -18’. Neither
outlined above was conducted as rapidly as possible, generally glutamine, glucose-6-P, fructose-6-P, nor a variety of other sub-
taking approximately 9 hours. stances appeared to stabilize the enzyme preparations.
Pur$ication of N. crassa Extract-All operations were per- The N. crassa enzyme preparations were even less stable than
1268 Glucosamine Metabolism. V Vol. 235, No. 5

TABLE I 2 to 3 hours against solutions containing EDTA, mercapto-

Purijcation of enzyme from Escherichia coli ethanol, phosphate buffer, and so forth. The E. coli enzyme ex-
hibited variable stability to dialysis, usually retaining full ac-
Specific activity with*
tivity for 3 to 4 hours. These losses in activity on dialysis were
Fraction Total activity
Fru6c$se- Glug$se- not reversed by the addition of boiled extracts or possible co-
factors.
units mits/mg unit/mg
E$ects of Enzyme Concentration, Time, pH, and Substrate Con-
centration on Reaction Rates--In these studies, although adequate
Crude extract. . .. 195 0.13 0.13
Polymyxin supernatant 189 0.46 0.46 data were obtained with the rat liver and E. coli enzymes, the
Cellulose eluate I. . 63 1.3 0.35 experimental results obtained with the N. crossa enzyme were
Cellulose eluate II.. .. 88 1.3 not as satisfactory, probably owing to the instability of the latter
Protamine eluate I. . . . 29 4.5 0.00 material. This point is demonstrated in Fig. 1, which indicates
Protamine eluate II.. 16 7.3 0.00 that the rate of glucosamine formation with the N. craSsa en-
zyme, was linear for only 30 minutes when the temperature was
* Fractions which were inactive with fructose-6-P showed no
maintained at 30”. As shown in Fig. 2, the purified E. coli and
activity with glucose-6-P.
rat liver enzymes catalyzed the formation of glucosamine-6-P
linearly with time. Owing to the nonlinearity of the curve ob-
TABLE II tained with the N. crassa enzyme, the kinetics are used only to
PurZfication of_ enzume
” from rat liver approximate the K, for fructose-6-P.
-

Downloaded from http://www.jbc.org/ by guest on August 27, 2017

Specific activity with* The reaction velocity as a function of rat liver and E. coli en-
Fraction Total activity - zyme concentrations is shown in Fig. 3.
The effect of pH on rat liver enzyme is shown in Fig. 4; the
Frugc-tpose-
G1%-- MazYse- reaction proceeds at a maximum rate in phosphate, Veronal, and
units units/mg unit/mg unit/mg Tris-maleate buffers at pH 7.7,7.7, and 7.9, respectively. Little
I 20.2 0.023 0.023 0.018 activity of the rat liver enzyme was observed in Tris-maleate or
II 20.2 0.032 0.032 0.019 Verona1 buffers unless EDTA was added to the incubation mix-
IV 11.4 0.44 0.00 0.040 tures. The pH optimum for the E. coli enzyme in Tris-maleate
VII 5.4 4.7 0.00 0.00 buffer is 7.9 (Fig. 5). The pH optimum of the Neurospora en-
VIII 1.5 2.1 zyme in phosphate buffer was found to be 6.7.
The effects of substrate concentrations on the reaction veloci-
* Fractions which were inactive with fructose-6-P showed no
activity either with glucose-6-P or mannose-6-P. ties are shown in Figs. 6 and 7; here again, adequate data could
only be obtained with the E. coli and rat liver enzymes. The
K, values were: 1.6 X 10” M for glutamine and 3.8 X lo-’ M
TABLE III
for fructose-6-P with the rat liver enzyme; and 6.5 X 1O-4 M for
Purijication of enzyme from Neurospora crassa glutamine and 2.0 X 10e3 M for fructose-6-P with the E. coli
Specific activity with* enzyme. Studies with the N. crassa enzyme yielded an approxi-
Fraction Total activity mate K, of 7.5 X 1O-4 M for fructose-6-P. Glutaminase ac-
Fructose-6-P Glucose-6-P tivity in the N. crassa enzyme prohibited substrate studies with
units units/mg unit/rig glutamine.
Crude extract ................ 12.5 0.16 0.22 Substrate Xpeci$city-Crude extracts obtained from E. coli, rat
Protamine supernatant ....... 10.1 0.45 0.47 liver or N. crassa all showed transamidase activity in the pres-
Gel eluate I .................. 9.0 1.85 0.44 ence of either glucose-6-P or fructose-6-P. The rat liver extract
Gel eluate II ................. 2.4 2.05 0.00 was also tested with mannose-6-P since Dr. Alfred J. Bollet
(University of Virginia) informed us that mannose-6-P would
* Fractions which were inactive with fructose-6-P showed no serve as substrate for these extracts. However, as the enzymes
activity with glucose-6-P. were purified free from phosphohexoisomerases, only fructose-6-P
acted as the hexose substrate, and only glutamine as the nitrogen
the rat liver preparations. Both the crude extracts and the pro- donor. Specificity studies were conducted with the purified en-
tamine supernatant solutions lost approximately 50% of their zymes from rat liver and E. coli. The following substances were
activities in 12 hours either at 2” or at - 18’. The more purified inactive when substituted for fructose-6-P in the usual assay
preparations lost all activity at either temperature. Attempts system: fructose, glucose-6-P, mannose-6-P (tested only with the
to stabilize the activities in solution with a variety of substances rat liver enzyme). The following could not substitute for L-glu-
were unsuccessful. The most stable preparation was the lyoph- tamine as nitrogen donors: NH&l, asparagine, asparagine +
ilized material obtained after calcium phosphate gel I step, which ATP, NH&l + glutamic acid + ATP, glutamic acid.
could be stored for about 1 week at - 18” in a vacuum with rela- Reversibility Studies-Experiments were performed to de-
tively little loss in specific activity. termine whether the reaction was measurably reversible, i.e. the
All the preparations were irreversibly inactivated when ex- conversion of glucosamine-6-P and glutamic acid to fructose-6-P
posed briefly to a solution of pH 5.5, or below. and glutamine. In all experiments, the purified enzyme prep-
Neither the N. crassa nor rat liver enzymes were stable to di- arations were used. Initially, the reversibility of the reaction
alysis, losing from 70 to 75% of their activities when dialyzed for was studied by measuring fructose-6-P formation directly; no
May 1960 8. Ghosh, H. J. Blumenthal, E. Davidson, and X. Roseman 1269

Downloaded from http://www.jbc.org/ by guest on August 27, 2017

30 60 90 120 150 180
TIME (MINUTES)
Fm 1. Glucosamine formation as a function of time by N. crassa enzyme. The incubation mixture contained per ml: 20
pmoles of fructose-6-P, 30 pmoles of glutamine, 75 pmoles of phosphate buffer, pH 6.6, 5 pmoles of EDTA, 60 rg of protein from gel
eluate I; incubated at 30”. The products, glucosamine (due to phosphatase activity) and glucosamined-P, are presented as “glu-
cosamine.”

FIG. 3. Effect of protein concentration on glucosamine-6-P

TIME (MINUTES) (Gm-6-P) formation. Rat. liver enzyme (A) ; E. coli enzyme (0).
The rat liver incubation mixtures contained per ml: 16.6 pmoles
FIG. 2. Glucosamine-6-P (Gm-6-P) formeti’on as a function of
The E. coli incubation of fructose-6-P, 12 /Imoles of glutamine, 3.33 &moles of EDTA, 100
time. Rat liver enzyme (A) ; E. coli (0). pmoles of phosphate buffer, pH 7.5, protein from Fraction VIII.
mixtures contained ner ml: 40 Dmoles of fructose-6-P, 48 moles of The E. coli incubation mixtures contained per ml: 40 &moles of
glutamine, 37.5 pmoles of phosphate buffer, pH 7.0, 2.5 pmoles of fructose-6-P, 48 Bmoles of glutamine, 37.5 rmoles of phosphate
EDTA, 42 pg of protein from protamine eluate II fraction. The buffer, pH 7.0,2.5pmoles of EDTA, protein from protamine eluate
rat liver incubation mixtures contained per ml : 25 pmoles of fruc- II. Incubated at 37” for 1 hour with rat liver enzyme and 30 min-
tose-6-P, 30 pmoles of glutamine, 50 pmoles of phosphate buffer, utes with E. coli enzyme.
pH 7.6,2.5pmoles of EDTA, 55.5 pg of protein from Fraction VIII.
Incubated at 37”.
following in 3 ml: 10 pmoles of glucosamine-6-P, 10 pmoles of
synthesis of hexose phosphate was noted. Attempts were then glutamic acid, 2 pmoles of TPN, 1.6 units of yeast phosphohexo-
made to “pull” the reaction towards the left by adding phospho- isomerase,2 1.3 units of glucose-6-P dehydrogenase, and 0.4 unit
hexoisomerase, glucose-6-P dehydrogenase, and TPN to the re- 2 Purified yeast phosphoglucoisomerase was a gift from Dr.
action mixtures. Typical incubation mixtures contained the Armand J. Guarino, Department of Biological Chemistry, The
1270 Glucosamine Metabolism. V Vol. 235, No. 5

0.5 I I I I I

TRIS-MALEATE
g OA- .
PHOSPHATE l

9- 0.3-
co
k
; 0.2- * t
5 VERONAL
0
$ O.l-

I I I 4 I
6.8 22 7.6 8.0 8.4 8
PH
FIG. 4. Activity of rat liver enzyme as a function of pH.
incubation mixtures contained per ml: 10 pmoles of fructose-6-P,
10 rmoles of glutamine, 2 pmoles of EDTA, 30 pg of protein from

Downloaded from http://www.jbc.org/ by guest on August 27, 2017

Fraction VIII of rat liver enzyme, and 50 pmoles buffer. The re-
action mixtures were incubated at 37” for 1 hour. Glucosamine- I I I I I I

6-P = Gm-6-P. 7.0 l 7.4 7.8 8.2 8.6 9.0

PH

of E. coli enzyme, the mixture being adjusted to pH 7.8. The FIG. 5. Activity of E. coli enzyme as a function of pH. One
milliliter of incubation mixture contained: 20 moles of fructose-
rat liver enzyme incubation mixture contained: 40 pmoles of 6-P, 6 pmoles of glutamine, 5 pmoles of EDTA, 160 pg of protein
potassium phosphate buffer, pH 7.6,2 pmoles of EDTA, 3 pmoles f rom protamine eluate II, and 50 pmoles of Tris-maleate buffer.
of glutamic acid, 3 pmoles of glucosamine-6-P, 1.2 units of phos- Incubated at 37” for 30 minutes. Gm-6-P = glucosamine-6-P.

FIG. 6. Effect of fructose-6-P concentration on glucosamine-6-P contained: 48 pmoles of L-glutamine, 37.5 pmoles of phosphate
formation. Rat liver enzyme (A), E. coli enzyme (0). Rat liver buffer, pH 7.0,2.5 pmoles of EDTA, 28 pg of protein from prota-
incubation mixtures contained per ml: 6 pmoles of L-glutamine, mine eluate II. Incubated at 37’ for30 minutes. K,,, values were
50 pmoles of phosphate buffer, pH 7.5, 2.5 pmoles of EDTA, 35 determined by the method of Lineweaver and Burk (21). F-6-
rg of protein from Fraction VIII. E. coli incubation mixtures P = fructose-6-P.

phohexoisomerase, 30 units of glucose-6-P dehydrogenase, 0.5
pmole of TPN, and 0.18 unit of rat liver transamidase. No
University of Michigan. The specific activity of the enzyme was
40 to 50 units per mg of protein; 1 unit converted 1 pmole of fruc-
tose-6-P to glucose-6-P per minute at 25” under optimum condi-
tions.
TPNH formation was detected even after prolonged incubation,
although the addition of 0.01 pmole of fructose-6-P to either
mixture resulted in an immediate reduction of a corresponding
amount of TPN. Similar results were obtained with the N.
crassa enzyme.
Xtoichiometry-The stoichiometric conversions of fructose-6-P
May 1960 S. Ghosh, H. J. Blumenthal, E. Davidson, and X. Roseman 1271

and glutamine to glucosamine-6-P and glutamic acid by the puri- TABLE IV

fied enzymes from E. coli and rat liver are shown in Table IV. Stoichiometry studies
Studies with the most highly purified N. crassa enzyme were
Reaction catalyzed by enzyme from
unsatisfactory, since this preparation still contained active phos-
phatase and glutaminase. In a typical experiment, at the end Rat liver
Compound I
of a 30-minute incubation, 1.1 pmoles of fructose-6-P had disap- E.colii
peared whereas 0.80 pmole total hexosamine (glucosamine and ExpePnt
Experiment
11t
glucosamine-6-P) was formed; however, approximately 20 times
more glutamic acid was formed than fructose-6-P utilized. pWWlt?/?d p?MltX/ntl /moles/ml
Characterization of Hexosamine Phosphate-The enzymatic Fructose-6-P. . -0.89 -4.6 -30
product obtained with the E. coli fraction was characterized as Glutamine................. -0.87 -5.9 -28
follows. The incubation mixture was deproteinized with per- Glucosamine-6-P. f0.77 +4.6 +32
chloric acid, and the excess perchloric acid removed by the addi-
tion of solid KHCOa in an ice bath, and the filtrate applied to a
Glutamic acid..

* Incubation mixture
-I
contained
f0.94 +5.1

per ml: 3.54 pmoles of fructose-

$32

Dowex 50, hydrogen column. The hexosamine phosphate was 6-P, 4.91 pmoles of glutamine, and 0.86 ml of Fraction VIII (dia-
eluted with water (10) and the solution evaporated under reduced lyzed 3 hours) containing, in this case, 190 rg of protein. Incu-
pressure almost to dryness. Finally, the hexosamine phosphate bated at 37” for 2 hours.
was crystallized from the syrup by the careful addition of acetone. t Incubation mixture contained per ml: 5.8 pmoles of fructose-
After one recrystallization from aqueous acetone, the material 6-P, 7.7 pmoles of glutamine, and 0.88 ml of Fraction VIII (dia-
exhibited the same powder x-ray diffraction pattern as authentic lyzed 1 hour) containing 195 fig of protein. Incubated at 37” for

Downloaded from http://www.jbc.org/ by guest on August 27, 2017

glucosamine-6-P (11). 3 hours.
The smaller quantities of glucosamine-6-P available from the $ Incubation mixture contained per ml: 54 pmoles of fructose-
6-P, 40 pmoles of glutamine, 5 mg of azaserine, and 0.33 ml of
purified rat liver preparation discouraged attempts at crystalliza-
protamine eluate II containing 8.5 mg of protein. The specific
tion, and the product was therefore characterized as follows.
activity of the enzyme had declined to 2.8 on storage; azaserine
After deproteinization of the incubation mixture, the hexosamine was added to inhibit glutaminase activity. The mixture was in-
phosphate was isolated by ion exchange chromatography as de- cubated for 1 hour at 37”, and in contrast to the other experiments
scribed above. An aliquot was then treated with glucosamine- described in the table, where the samples were heated for 2 min-
6-P deaminase (22) which yielded fructose-6-P as expected. A utes to stop the reactions, in this case, the reaction was stopped
part of the hexosamine phosphate was treated with potato phos- by adding perchloric acid to 0.5 M concentration. After centrifu-
phatase and the resultant hexosamine isolated by ion exchange gation, excess perchloric acid was removed with potassium bicar-
chromatography according to the technique of Garde11 (23). bonate.
The hexosamine peak coincided exactly with that of a glucos-
amine indicated that it was N-acetylglucosamine and not N-ace-
amine standard. The isolated hexosamine hydrochloride was
tylmannosamine or N-acetylgalactosamine (24). When the free
N-acetylated (12, 13); chromatography of the N-acetylhexos-
hexosamine was treated with ninhydrin (25), it yielded a product
I I I which showed the same RF value as n-arabinose. These results
5
indicated that the product of action of the purified rat liver en-
zyme was glucosamine-6-P.
Inhibition of Transamidase Reactionby DON-The following
substances were tested as inhibitors of the rat liver and E. coli
enzymes: DON, azaserine, y-glutamylhydrazide, and DL-z-amino-
4-(methylsulfinyl)-butyric acid. DON was by far the most po-
tent inhibitor of the enzymes as shown in Fig. 8. In addition
to its well known effect as an inhibitor of purine synthesis, there-
fore, DON may prevent cell growth by inhibiting the formation
of glucosamine.
Distribution of Transamidasein Fungi-Many fungi contain
chitin in their cell walls (26). The following strains of organisms
were found to contain transamidase activity when grown on a
minimal medium (20) and extracted as described above for N.
I I t 1 1 I .
crassa:N. crassa Em-5297a,3 Glowzerella cingulataBOT, Helmin-
(0) 2 4 6 8 IO 12
I thospmium sativum BOT, Neurospora tetrasperma BOT, Penicil-
iA) I lium species BACT, Dactylium dendroides QM 508, Aspergillus
2[6L”TAM:NE] x1043M 5 ’
parasiticus QM 883 and QM 884, Penicillium notatum BACT,
FIG. 7. Effect of L-glutamine concentration on glucosamine-6-P Aspergillus flavus QM 870. Negative results were obtained with
(Gm-6-P) formation. Rat liver enzyme (A); E. coli enzyme (a).
Rat liver incubation mixtures contained per ml: 5 #moles of fruc- 3 The abbreviations after the names of the organisms indicate
tose-6-P, 50 pmoles of phosphate buffer, pH 7.5, 2.5 rmoles of the sources from which the cultures were obtained: Em-5297a, G.
EDTA, 35 rg of protein from Fraction VIII. E. coli incubation W. Beadle, California Institute of Technology; BOT and BACT,
mixtures contained per ml: 40 pmoles of fructose-6-P, 37.5 rmoles Departments of Botany and Bacteriology, The University of
of phosphate buffer, pH 7.0, 2.5 pmoles of EDTA, 28 fig from pro- Michigan, respectively; QM, Quartermaster Culture Collection;
tamine eluate II. Incubated at 37” for 30 minutes. Km values NRRL, Northern Regional Research Laboratories, ATCC, Amer-
were determined by the method of Lineweaver and Burk (21). ican Type Culture Collection.
1272 Glucosamine Metabolism. V Vol. 235, No. 5

I I 1 I I Propertiesof E. coli Glutaminase-Apparently, there have been

few studies on bacterial glutaminases. The E. coli extracts ex-
hibited powerful glutaminase activity, and some characteristics
of this enzyme were determined. In these experiments, the
polymyxin supernatant fraction was used. The effects of en-
zyme concentration and tie on the hydrolysis of glutamine are
shown in Figs. 9 and 10. The pH optimum for the enzyme ac-
tivity was 4.9.
DON is a powerful inhibitor of the E. coli glutaminase. For
example, about 1.1 pmoles per ml of DON inhibited the activity
approximately 50% in an incubation mixture containing 67
pmoles per ml of glutamine. Azaserine was much less effective
as an inhibitor of the glutaminase, since 92 pmoles of azaserine
per ml yielded a 12’j!& inhibition of the enzyme under the same
conditions.

DISCUSSION

(A) I 3 4 5 The results obtained with purified enzymes obtained from

Doff (,Ug. PER ml.) typical bacterial, fungal, and mammalian cells indicate that they

Downloaded from http://www.jbc.org/ by guest on August 27, 2017

FIG. 8. Inhibition of rat liver and E. coli enzymes by DON. catalyze the reaction: Fructose-6-P + glutamine -+ glucosamine-
Rat liver enzyme (A); E. coli enzyme (0). Rat liver incubation 6-P + glutamic acid. A suitable name for the enzyme is n-glu-
mixtures contained per ml: 10 pmoles of fructose-6-P, 6 pmoles of tamine-n-fructose 6-phosphate transamidase.
glutamine, 30 pmoles of phosphate buffer, pH 7.5, 2.5 pmoles of These data apparently conflict with the previous report of
EDTA, 44 pg of protein from Fraction VIII. E. coli incubation Pogell and Gryder (lo), who concluded that glucose-6-P is the
mixtures contained per ml: 40 pmoles of fructose-6-P, 12 pmoles of
glutamine, 37.5 pmoles of phosphate buffer, pH 7.0, 2.5 pmoles of hexose phosphate substrate for the rat liver enzyme. Since both
EDTA, and 200 pg of protein from protamine eluate II. Incu- hexose phosphates acted as substrates for their preparations, it
bated at 37” for 1 hour with rat liver enzyme and 30 minutes with is suggested that their preparations were contaminated with
E. coli enzyme. phosphohexoisomerase. In the present studies, all of the puri-
fied enzymes exhibited marked specificity for fructose-6-P and

I51 glutamine. Furthermore, the activities of the enzymes from the

three sources were followed with fructose-6-P and glucose-6-P
(also mannose-6-P in the case of the rat liver enzyme) during
2
purification; no fraction was active with glucose-6-P while inactive
E with fructose-6-P. These results, therefore, clearly indicate an
E IO- absolute requirement for fructose-6-P by the transamidase.
a
Extensive studies have been conducted to determine whether
more than one enzyme is involved in the biosynthesis of glucos-
amine-6-P from fructose-6-P. Since there was no evidence for

PROTEIN (,t/g. PER ml.)

FIG. 9. Hydrolysis of glutamine by E. coli enzyme as a function
of protein concentration. One milliliter of incubation mixture
contained: 67pmoles of glutamine, 38.5pmoles of acetate buffer pH
4.9. Incubated at 37” for 10 minutes.

extracts of Aspergillus niger BACT and AspergillusoryzaeATCC

7252. There was no correlation between the chitm content (20)
of the organism and the enzymatic activity of the extracts. N.
craSsa Em-5297a, the organism used in the studies described
IO 20 40 50 60
above, gave the most active extracts. Finally, the crude extracts
Tk (MINUTES)
obtained from the Penicillium species were active with fructose-
FIG. 10. Activity of E. coli glutaminase as a function of time.
6-P, but exhibited little activity with glucose-6-P, thus confirm-
Incubation mixtures contained per ml: 67 pmoles of glutamine,
ing the results of the enzyme fractionation described above for 38.5 pmoles of acetate buffer, pH 4.9, 160 pg of protein from poly-
the N. crassa extracts. myxin supernatant fraction of E. coli enzymes. Incubated at 37”.
May 1960 S. Ghosh, H. J. Blumenthal, E. Davidson, and S. Roseman 1273

separation of activities in the fractionation of the various en- REFERENCES

zymes, the available data suggest that the reaction is catalyzed 1. BECKER, C. E., AND DAY, H. G., J. Biol. Chem., 201,795 (1953).
by a single protein. In addition, studies on possible cofactor re- 2. RIEDER, S. V., AND BUCHANAN, J. M., J. Biol. Chem., 232,951
quirements involving dialysis and treatment with ion exchange (1958).
3. ROSEMAN, S., MOSES, F. E., LUDOWIEG, J., AND DORFMAN,
resin, were negative, suggesting that the enzyme does not require A., J. Biol. Chem., 203, 213 (1953).
cofactors. 4. ROSEMAN, S., LUDOWIEG, J., MOSES, F. E., AND DORFMAN, A.,
The transamidase reaction appears to be unique since the en- J. Biol. Chem., 206, 665 (1954).
ergy required is derived solely from the cleavage of the amide 5. TOPPER, Y. J., AND LIPTON, M. M., J. Biol. Chem., 203, 135
(1953).
bond of glutamine. In the usual biochemical reaction involving
6. LELOIR, L. F., AND CARDINI, C. E., Biochim. et Biophys. Acta,
the transfer of the amide nitrogen of glutamine to an acceptor 12, 15 (1953).
other than water (such as in purine biosynthesis (27) and DPN 7. LELOIR, L. F., in D. E. GREEN (Editor), Currents in biochemi-
formation (28)) an additional source of energy like ATP is re- cal research 1966, Interscience Publishers, New York, , 1956,,
quired. Finally, the transamidase reaction is not measurably p. 585.
8. BLUMENTHAL, H. J., HOROWITZ, S. T., HEMERLINE, A., AND
reversible, even in the presence of enzymes which would utilize ROSEMAN. S.. Bacteriological woceedinas. Societv of Ameri-
any fructose-6-P which might be formed. can Bacteriologists, Baltimore, 1955, p.“l37. ”
Previous experiments (22) demonstrated that glucosamine-6-P 9. ROSEMAN, S., DAVIDSON, E. A., BLUMENTHAL, H. J., AND
can also be formed from fructose-6-P and ammonia in the pres- DOCKRILL, M., Bacteriological proceedings, Society of Ameri-
can Bacteriologists, Baltimore, 1958, p. 107.
ence of glucosamine-6-P deaminase. Although the equilibrium 10. POGELL, B. M., AND GRYDER, R. M., J. Biol. Chem., 228, 701
of the deaminase reaction is far towards the formation of fruc- (1957).

Downloaded from http://www.jbc.org/ by guest on August 27, 2017

tose-6-P, this equilibrium can be shifted towards glucosamine-6-P 11. ROSEMAN, S., AND LUDOWIEG, J., J. Am. Chem. SOL, 76, 301
synthesis by coupling the deaminase with glucosamine-6-P acety- (1954).
12. DISTLER, J. J., MERRICK, J. M., AND ROSEMAN, S., J. Biol.
lase and acetyl-CoA. Thus glucosamine-6-P can be formed by
Chem., 230, 497 (1958).
two types of systems, the transamidases described here and by 13. SPIVAK, C. T., AND ROSEMAN, S., J. Am. Chem. SOL, 81, 2403
reversing the deaminase reaction. In both cases fructose-6-P (1959).
serves as the hexose phosphate acceptor for the nitrogen, which 14. SINGER, T. P., AND KEARNEY, E. B., Arch. Biochem., 29, 190
in one case is derived from glutamine and in the other from am- (1950).
15. BLIX, G., Acta Chem. Stand., 2, 467 (1948).
monia. The relative importance of these two pathways in the 16. ROE, J. H., AND PAPADOPOULOS, N. M., J. Biol. Chem., 210,
synthesis in viva of glucosamine-6-P remains to be determined. 703 (1954).
The experiments of Lowther and Rogers (29) are of interest in 17. HAWK, P. B., OSER, B. L., AND SUMMERSON, W. H., Practical
this connection. Fresh, intact cells of Streptococcus hemolyticus physiological chemistry, 12th edition, Blakiston Company,
Philadelphia, 1947.
synthesized hyaluronate with either ammonia or glutamine as 18. ROSEMAN, S., AND DAFFNER, I., Anal. Chem., 28, 1743 (1956).
the sole nitrogen source, whereas aged cells utilized only gluta- 19. REISSIG, J. L., STROMINGER, J. L., AND LELOIR, L. F., J. Biol.
mine. Chem.; 21’7,959 (1955).
SUMmRY 20. BLUMENTHAL. H. J.. AND ROSEMAN. I S.. , J. Bacterial.. , 74. 222
(1957). ’ ’
Studies with purified enzymes obtained from Escherichiacoli, 21. LINEWEAVER, H., AND BURK, D., J. Am. Chem. SOL, 66, 658
rat liver, and Neurospora crassa indicate that these enzymes cata- (1934).
22. COMB, D. G., AND ROSEMAN, S., J. Biol. Chem., 232,807 (1958).
lyze the conversion of n-fructose 6-phosphate and n-glutamine to 23. GARD~LL, S.; Acta Chem. Scan&, 7, 207 (1953).
glucosamine 6-phosphate and glutamic acid. The preparations 24. COMB. D. G.. AND ROSEMAN. , S.. , J. Am. Chem. Sot.. , 80. , 497
are relatively labile, specific for the two indicated substrates, (1958). ’
strongly inhibited by 6-diazo-5-oxo-L-norleucine, and show no 25. STOFFYN, P. J., AND JEANLOZ, R. W., Arch. Biochem., 62, 373
(1954).
cofactor requirements. The reaction is apparently catalyzed by
26. RICHARDS, A. G., The integument of arthropods, University of
one enzyme and is not measurably reversible. Minnesota Press, Minneapolis, 1951.
An active, soluble glutaminase was obtained from E. co&i. 27. WOLSTENHOLME, G. E. W., AND O’CONNOR, C. M. (Editors),
Some of its properties are recorded; the enzyme is strongly in- Ciba Foundation symposium on the chemistry and bioloou of
hibited by 6-diazo-5-oxo-L-norleucine. purines, J. and A.Churchill, Ltd., London,-1957. “” -
28. PREISS. J.. AND HANDLER. , P.. , J. Am. Chem. SOL. I 79. , 4246
(1957). ’
Acknowledgment-We are most grateful to Mrs. Carol Lusty 29. LOWTHER, D. A., AND ROGERS, H. J., Biochem. J., 62, 304
and Mrs. Mary Do&ill for their expert assistance. (1956).
 Glucosamine Metabolism: V. ENZYMATIC SYNTHESIS OF
GLUCOSAMINE 6-PHOSPHATE
Sudhamoy Ghosh, Harold J. Blumenthal, Eugene Davidson and Saul Roseman
J. Biol. Chem. 1960, 235:1265-1273.

Access the most updated version of this article at

http://www.jbc.org/content/235/5/1265.citation

Alerts:
• When this article is cited
• When a correction for this article is posted

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

Downloaded from http://www.jbc.org/ by guest on August 27, 2017

This article cites 0 references, 0 of which can be accessed free at
http://www.jbc.org/content/235/5/1265.citation.full.html#ref-list-1