THE OF CHEMISTRY Vol. 266. 36, 25, pp. 24276-24286,1991 ... · 24276 . Catabolism of Chitin Oligosaccharides 24277 menclature of the hydrolyses. Insofar as we are aware, all published

0 1991 by The American Society for Biochemistry and Molecular Biology, Inc THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 266. No. 36, Issue of December 25, pp. 24276-24286,1991

Printed in U.S.A.

Chitin Utilization by Marine Bacteria DEGRADATION AND CATABOLISM OF CHITIN OLIGOSACCHARIDES BY VIBRIO FURNISSZP

(Received for publication, October 18,1990)

Chemotaxis of the marine bacterium Vibrio furnissii to chitin oligosaccharides has been described (Bassler, B. L., Gibbons, P. J., Yu, C., and Roseman, S. (1991) J. Biol. Chem. 266, 24268-24275). Some steps in catabolism of the oligosaccharides are reported here. GlcNAc, (GlcNAc)z, and (GlcNAc), are very rapidly consumed by intact cells, about 320 nmol of GlcNAc equivalents/min/mg of protein. (G1cNAc)r is utilized somewhat more slowly. During these processes, there is virtually no release of hydrolysis products by the cells. The oligosaccharides enter the periplasmic space (via specific porins?) and are hydrolyzed by a unique membrane-bound endoenzyme (chitodextrinase) and an exoenzyme (N-acetyl-B-glucosaminidase; 8-Glc- NAcidase). The genes encoding these enzymes have been cloned and expressed in Etwherichia coli. The chitodextrinase cleaves soluble oligomers, but not chitin, to the di- and trisaccharides, while the periplasmic 8-GlcNAcidase hydrolyzes the GlcNAc termini from the oligomers. The end products in the periplasm, GlcNAc and (G1cNAc)z (possibly (GlcNAc)3) are catabolized as follows. (a) Disaccharide pathway, A (Glc NAc)z permease is apparently expressed by Vibrio furnissii. Translocated (G1cNAc)z is rapidly hydrolyzed by a soluble, cytosolic 8-GlcNAcidase, and the GlcNAc is phosphorylated by an ATP-dependent, constitutive kinase to GlcNAc-6-P. ( b ) Monosaccharide pathway, Periplasmic GlcNAc is taken up by Enzyme IINng of the phosphoeno1pyruvate:glycose phosphotransferase system, yielding GlcNAc-6-P, the common intermediate for both pathways. Finally, GlcNAc-6-P + Ac- + GlcNHz-6-P + Fru-6-P + NH3.

(G~CNAC)~ is probably the “truen inducer of the chitin degradative enzymes described in this report and, de- pending on its concentration in the growth medium, differentially induces the periplasmic and cytosolic 8- GlcNAcidases. The disaccharide pathway appears to be the most important when the cells are confronted with low concentrations of the oligomers (e.g. in chemotaxis swarm plates). The relative activities of the induced enzymes suggest that the rate-limiting steps in oligosaccharide catabolism are the glycosidase activities in the periplasm.

* This work was supported by Contract N0001485-K-0072 from the Office of Naval Research. This paper is Contribution No. 1478 from the McCollum-Pratt Institute of The Johns Hopkins University. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Supported by National Institutes of Health Training Grant 51‘32 GM07321 and by W. R. Grace and Company. Present address: The Agouron Institute, 505 Coast Blvd. South, La Jolla, CA 92037.

The accompanying papers (1, 2) and a subsequent report‘ show that a marine bacterium, Vibrio furnissii, can adhere to a chitin analogue, immobilized N-acetylglucosamine (Glc NAc)’ via a unique lectin, can “sense” chitin oligosaccharides by a complex chemotactic system and can also sense and catabolize PTS sugars. The utilization of chitin oligosaccharides by V. furnissii is the subject of this report.

The process of chitin degradation has been actively investigated for many decades (4). In the 1930s’, the importance of chitin turnover in the survival of marine ecosystems and the key role marine bacteria play in this process were recognized by Zobell and Rittenberg (5). Nonetheless, the genetic and biochemical mechanisms involved in chitin degradation by marine microorganisms have not been fully elucidated. Teleo- logical reasoning predicts that chitinolytic bacteria should possess specific mechanisms for acquiring and utilizing soluble, extracellular chitin oligosaccharides formed by the partial hydrolysis of the insoluble polymer. Otherwise, diffusion and water currents would deprive the bacteria of these potential nutrients prior to their ultimate conversion to GlcNAc. We report here that V. furnissii does indeed “capture” chitin oligosaccharides and subsequently degrades them by a complex series of reactions.

The first studies on chitinolytic enzymes appear to be those of Zechmeister and Toth, who chromatographed extracts of almond emulsin (6) and Helix pornatia (7) on bauxite columns. In each case they separated two enzymes, an endochitinase and an exo-N-acetyl-P-glucosaminidase. The latter catalyzed only the hydrolysis of small oligosaccharides, yielding GlcNAc, while the former was termed a “polysac- charidase.” Subsequently, chitinases and exo-N-acetyl-@-glucosaminidases have been isolated from a wide variety of species (8-13), including bacteria, yeast, fungi, and higher organisms.

It is important to clarify one aspect of the literature relating to both the mechanism of chitin solubilization and the no-

’ C. Yu, B. L. Bassler, J. A. Stock, and S. Roseman, submitted for publication.

* The abbreviations used are: GlcNAc or Nag, N-aCetyl-D-glUCOSa- mine; ASW, artificial sea water; DMF, N,N-dimethylformamide; MUF, 4-methylumbelliferone; PEP, phosphoenolpyruvate; PTS, phosphoeno1pyruvate:glycose phosphotransferase system; kb, kilo- base(s); Hepes, 4-(2-hydroxyethyl)-l-piperazineethanesulfonic acid; HPLC, high performance liquid chromatography; DTT, dithiothrei- tol. Oligosaccharides derived from chitin are @,l+ 4-linked oligomers of GlcNAc and are designated (GlcNAc)?, for N,N’-diacetylchito- biose; (GICNAC)~, (GICNAC)~, (GlcNAc)s, and (GlcNAc)e, for the corresponding chitotriose, -tetraose, -pentaose, and -hexaose derivatives, respectively. Glycosides of N-acetylglucosamine and the oligosaccharides are pyranosides and are designated by their aglycon and ano- meric configurations as follows: methyl a-GlcNAc, methyl P-GlcNAc, MUF-P-GlcNAc, MUF-P-(GlcNAc)2, phenyl a-GkNAc, phenyl B- GlcNAc; PN-GlcNAc, p-nitrophenyl 0-GlcNAc; PN-(GlcNAc)y, p- nitrophenyl P-~-N,N”diacetylchitobioside. GlcNAcidase signifies N - acetyl-P-glucosaminidase, i x . the exoenzyme.

24276

Catabolism of Chitin Oligosaccharides 24277

menclature of the hydrolyses. Insofar as we are aware, all published reports indicate that only two types of enzymes are involved in converting chitin to GlcNAc, chitinases which act on the insoluble particle, and exo-N-acetyl-P-glucosaminidases which cleave at the nonreducing termini of the soluble products formed by the chitinase (13). The designation “chitodextrinase” has been used (14, 15), but the properties of these preparations appear to be identical with those of chitinases, in that they act on particulate and colloidal chitin. We suggest that the term chitodextrinase be reserved for endoen- zymes that catalyze the hydrolysis of the intermediate, soluble oligosaccharides. That is, a chitodextrinase is unable to solu- bilize chitin, but can catalyze the hydrolysis of high to low molecular weight soluble chitin oligosaccharides. One of the V. jurnissii enzymes described in this report exhibits precisely these properties and is therefore a chitodextrinase. Thus, in at least this marine Vibrio, three classes of enzymes are required for the complete degradation of chitin to GlcNAc, chitinases, chitodextrinases, and exo-N-acetyl-0-glucosaminidases.

In addition to studies on chitinases, the structural genes encoding some N-acetyl-P-glucosaminidases (p-Glc- NAcidases) have been cloned (16-24) and a few have been sequenced (25-28). Current work has focused on the homol- ogies between the P-GlcNAcidase genes from different species t o assess potential evolutionary relationships, especially to the human gene since a defect in the P-GlcNAcidase results in Tay-Sachs disease. The studies most relevant to our work, however, are those involving the genus Vibrio, the most ubiq- uitous of the marine bacteria (29). Vibrios apparently play a crucial role in converting chitin to a utilizable form in the marine environment. Soto-Gil and Zyskind (16) and Janna- tipour et al. (17) cloned the 0-GlcNAcidase gene from V. harveyi into Escherichia coli and found that the enzyme was transported to the outer membrane of the E. coli host following cleavage of a signal sequence. The amino acid sequence deduced from the cloned gene is similar to the a-chain of human 0-GlcNAcidase (26). In V. harveyi, P-GlcNAcidase activity is induced by ( G ~ C N A C ) ~ in the growth medium (16). A P-GlcNAcidase has also been cloned into E. coli from V. vulnificus by Wortman et al. (18), who suggest that this enzyme catalyzes the complete degradation of chitin to GlcNAc.

As shown below, both GlcNAc and its oligosaccharides are consumed a t surprisingly high rates by V. jurnissii. These compounds are converted to intracellular fructose-6-P by a complicated sequence of enzymatic reactions, partially characterized in this report. The scheme presented in Fig. 1 summarizes our concepts. At least three glycosidases convert chitin oligosaccharides to products that can be translocated across the V . jurnissii cytoplasmic membrane. Colloidal chitin is first solubilized by an “extracellular” chitinase, the soluble oligosaccharide products penetrate the outer membrane and are degraded in the periplasmic space to GlcNAc, (GlcNAc),, and (GlcNAc)3 by two hydrolases (one chitodextrinase and one P-GlcNAcidase) which appear to be bound to the cytoplasmic membrane.3 GlcNAc is transported and phosphorylated via the PTS, yielding cytoplasmic GlcNAc-6-P. I t appears that (G1cNAc)n (and perhaps (G~CNAC)~) formed in the periplasmic space is transported by a specific permease and then hydrolyzed in the cytoplasm by a second P-GlcNAcidase. Cytoplasmic GlcNAc is subsequently phosphcrylated by an ATP-dependent kinase (30 and the present report) yielding

’’ The chitodextrinase and 0-GlcNAcidase are membrane-bound, but are not found in preparations of the outer membrane (unpublished experiments; data kindly provided by Nemat Keyhani).

G * GlcNAc

FIG. 1. Degradation of chitin oligosaccharides by V. furnissii. Oligosaccharides, (GlcNAc),, enter the periplasmic space ( I ) . This step may require specific porins. The oligomers are either hydrolyzed by the periplasmic P-GlcNAcidase (2) or the chitodextrinase ( 3 ) . Hydrolysis by the periplasmic 0-GlcNAcidase gives GlcNAc and (GlcNAc), - ,; the latter is acted on again by these enzymes. The chitodextrinase yields (GlcNAc)3 and (GICNAC)~ as final products. (G1cNAc)s is the most effective substrate for the periplasmic p- GlcNAcidase and is hydrolyzed to (GlcNAc), and GlcNAc. Monosac- charide pathway, the monomer formed in the periplasmic space is transported by Enzyme IINag ( 4 ) and enters the cytosol as GlcNAc-6- P. Disaccharide pathway, (GICNAC)~ formed by the chitodextrinase and the periplasmic 0-GlcNAcidase is transported into the cytosol by the (GlcNAc):! permease (5) and hydrolyzed to 2 GlcNAc by the cytoplasmic 0-GlcNAcidase ( 6 ) . Cytosolic GlcNAc is then phosphorylated to GlcNAc-6-P by an ATP-dependent kinase (7). Thus, both pathways give a common intermediate, GlcNAc-6-P, which is deacetylated ( 8 ) and deaminated (9) to the catabolites Fru-6-P, acetate. and ammonia.

GlcNAc-6-P. Finally, GlcNAc-6-P generated by either of the above pathways is deacetylated and deaminated (31, 32) producing fructose-6-P.

Preliminary reports for some of these steps have been presented (33-35).

MATERIALS AND METHODS4

RESULTS

Utilization of GlcNAc

V. jurnissii consumes both GlcNAc and GlcNAc oligosaccharides with remarkable efficiency, as shown in Fig. 2. V . jurnissii was grown in medium containing either lactate or GIcNAc as the sole carbon source, washed, and resuspended in medium containing 0.6 mM GlcNAc at a final cell density equivalent to 25 yg of total protein per ml. Cells pregrown on lactate (uninduced) utilized GlcNAc after a 15- to 20-min lag period, whereas cells grown on the sugar (induced) immediately began consuming GlcNAc, utilizing 80% of the monosaccharide in 1 h. Results similar to the latter were also obtained when V. jurnissii was grown in lactate medium containing either 0.6 mM GlcNAc or (GlcNAc),.

Under these conditions, V. furnissii consumed GlcNAc at a rate of 320 nmol/min/mg of protein at 25 “C, which is comparable to the optimal Glc consumption rate reported for E. coli, 300 nmol/min/mg of protein a t 37 “C (46). Additionally, the disappearance of 48 nmol of GlcNAc in 1 h by 2.5 pg of cell protein calculates to the utilization of 2 times more GlcNAc than total cell mass (dry weight) per h.

E. coli can transport GlcNAc by at least two different PTS membrane proteins, IINag and IIM”” (47). In contrast, V. furnissii apparently only transports GlcNAc via IINaK, step 4 in Fig. 1. Thus, the mutant, AP801, defective in IINap, neither ferments nor grows on GlcNAc (see “Materials and Meth-

‘ Portions of this paper (including “Materials and Methods,” Foot- notes 5 and 6, and Fig. 7) are presented in miniprint at the end of this paper. Miniprint is easily read with the aid of a standard magnifying glass. Full size photocopies are included in the microfilm edition of the Journal that is available from Waverly Press.

24278 Catabolism of Chitin Oligosaccharides

0 20 40 60 80

Time ( m i d FIG. 2. N-Acetylglucosamine utilization by V. furnissii. V.

furnissii was grown in the synthetic media described below. The cultures were harvested a t midexponential phase, washed three times with buffered 50% ASW by centrifugation at 5000 rpm, diluted to 5 X 10’ cells/ml in buffered 50% ASW containing 0.6 mM GlcNAc. The suspensions were shaken at room temperature, 200 pl aliquots were centrifuged (12,000 rpm, 25 “C) at the indicated times, and 100 /*I of the supernatants were analyzed for GlcNAc by the Morgan-Elson method. 0, V. furnissii grown on 55 mM lactate (uninduced); A, cells pregrown on 55 mM GlcNAc (induced); 0, cells grown on 55 mM lactate plus 0.6 mM GlcNAc; +, cells grown on 55 mM lactate containing 0.6 mM (G1cNAc)Z. Each aliquot (100 /*I) contained 2.5 pg of total cell protein.

ods”). However, this mutant grows on (G~CNAC)~ as the sole source of carbon. This key result suggests a separate trans- porter for the disaccharide in V. furnissii (step 5 in Fig. 1) and a potentially alternate pathway of catabolism. In V. furnissii, cytosolic GlcNAc (formed by the hydrolysis of (GlcNAc),) is phosphorylated by an ATP-dependent kinase (steps 6 and 7), then deacetylated to GlcNH2-6-P, and deaminated to fructose-6-P, steps 8 and 9 in Fig. 1, respectively. The latter two enzymatic reactions are also responsible for converting GlcNAc-6-P transported via the PTS, to Fru-6-P. Table I gives the activities and inducibility of the deacetylase and deaminase and shows that the GlcNAc kinase is constitutive.

Utilization of Oligosaccharides

Intact V. furnissii cells also utilized chitin oligosaccharides at a remarkable rate. Three examples are shown in Fig. 3. The surprising result, and the one to be emphasized, is that the cells consumed these oligosaccharides with little to no release of the expected intermediate hydrolysis products. The first two panels (Fig. 3, A and B) show the disappearance of 0.6 mM (GlcNAc)z and (GlcNAc)3, respectively. From 80-90% of each substrate was consumed within the 2-h time course of the experiment. There was virtually no release of the monomer in the case of (GlcNAcL, and almost no mono- or disaccharide was released when (GlcNAc)s was the substrate. (GlcNAc), was utilized more slowly (Fig. 3C); about 50% disappeared within the time course of this experiment, and the tetrasaccharide was completely consumed within 4 h (data not shown). Again, virtually no degradative products appeared in the medium. Thus, V. furnissii is capable of consuming GlcNAc oligosaccharides rapidly and completely, such that all of the carbon (and presumably the amino nitrogen) in these compounds is made available to the cells. Most of the remaining experiments described in this report are concerned with the mechanisms underlying this phenomenon.

Permeabilization of the cytosolic membrane with toluene

TABLE I Activities of GlcNAc catabolic enzymes in extracts of V. furnissii V. furnissii was grown in buffered 50% ASW supplemented with

NH3, Pi, and the carbon source or combination of carbon sources indicated in the table. The cells were harvested a t midexponential phase, washed, and fractionated as described under “Materials and Methods.” Membranes were assayed for the PTS Enzyme IINag activity and soluble extract for the deacetylase, deaminase, and GlcNAc kinase activities. All assays were performed under conditions where the activity was proportional to the amount of subcellular fraction added, and the rate was constant with the time of incubation. Activi- ties are expressed as nanomoles of product formed per min per mg of protein in the crude extract at 25 “C.

Activity

IINaB Deacety- Deamin- GlcNAc-

nmoljminjrng protein

Strain Carbon source lase ase kinase

7225 Peptone-YE” NDb <5 6 ND 0.5% Lactate 7 10 10 200 0.5% GlcNAc 190 120 1074 ND 0.2% Lactate + 165 135 1025 180

0.5% Lactate + 17 <5 190 ND 0.2% GlcNAc

0.6 mM GlcNAc

0.1% (GlcNAc), 156 53 980 ND AP801 0.2% Lactate + <1 <5 <5 188

0.2% GlcNAc 0.1% (GIcNAc)z‘ <I 30 482 ND

a YE is yeast extract.

‘Growth rates (generation times in hours) of strains 7225 and AP801 were: lactate minimal medium, strain 7225, 1.5, and AP801, 2.0; GlcNAc minimal medium, strain 7225,1.5, and AP801, no growth; (GlcNAc)* minimal medium, strain 7225, 1.4, and AP801, 2.5.

ND, not determined.

A. (GIcNAc), B. (GlcNAcl, c. (GlcNAcir ‘20- m m

Time (min)

FIG. 3. Degradation of chitin oligosaccharides by intact and toluene-permeabilized V. furnissii cells. Cells were grown to midexponential phase in lactate, 50% ASW medium containing 0.6 mM (GlcNAc)p as inducer, washed, and suspended to a density of 5 X

abilized with toluene as described under “Materials and Methods.” 10’ cells/ml in toluenizing buffer. Half of the suspension was perme-

The following oligosaccharides a t 0.6 mM concentrations were added to 3-5-ml aliquots of the suspensions containing intact or permeabilized cells: (GlcNAc)*, A and D; (GIcNAc)~, B and E or (GlcNAc).+, C and F. Intact cells, pawls A-C; permeabilized cells, panels D-F. The tubes were shaken (200 rpm, 25 “ C ) , 200-pl aliquots were removed at the indicated times and centrifuged, and 5 p1 of the supernatants were injected into a Dionex column to analyze simultaneously for substrates (-) and products (0- - -0). The results are plotted as “GlcNAc-equivalents,” which gives the concentrations of products as percent of the theoretical yield.

releases low molecular weight compounds from the cell (such as ATP) and stops the complete catabolism of the products of chitin degradation. Following this treatment, the final product of chitin oligosaccharide breakdown was the mono-


saccharide, GlcNAc. In all three experiments with (GlcNAc),, n = 2-4,100% of the substrate was recovered as the monomer (Fig. 3, D, E, and F ) , and the degradations were accomplished much more rapidly than in the intact cells. That is, (GlcNAc), was completely converted to GlcNAc in less than 5 min. (GlcNAc)3 initially yielded both the monomer and the dimer, and the dimer was subsequently degraded to monomer. The tetramer was still utilized slowly, initially yielding all the possible products, mono-, di-, and trisaccharide, but by the end of the experiment the major product was GlcNAc. These results indicate that hydrolysis of the tri- and tetrasaccharide (and presumably of higher oligosaccharides) in the periplasm, not transport, is the rate-limiting step in “capture.” Transport does appear to be the rate-limiting step for (GlcNAc),.

Analysis of V. furnissii utilization of higher oligosaccharides is more complicated. Presumably, additional enzymes, not yet characterized, are required for the dissimilation of larger oligomers. This topic is discussed in the preceding report (2).

Induction of P-GLcNAcidases Growth of V. furnissii in the presence of chitin oligosaccha-

rides induced the synthesis of P-GlcNAcidase activity. The kinetics of induction of the ( i n uiuo) P-GlcNAcidase activity, as measured by monitoring the rate of formation of p-nitrophenol from PN-GlcNAc, are shown in Fig. 4. V. furnissii was grown in lactate, 50% ASW medium or in this medium containing 0.6 mM (GlcNAc),, n = 2-4. P-GlcNAcidase activity was barely detectable in cells grown on lactate alone, whereas (GlcNAc), induced P-GlcNAcidase activity 100- to 200-fold. Significant activity could be detected after 30 min, while full induction usually occurred at 2 h (the generation time is 1.5 h). The higher oligosaccharides, (GlcNAc)3 and (GlcNAc),, also induced P-GlcNAcidase activity, but to a lesser extent than (GlcNAc)*. Presumably, the higher oligosaccharides must first be hydrolyzed to the “true” inducer, (GlcNAc),, before induction can occur.

A variety of GlcNAc glycosides were analyzed for their

- - I

0 60 120 180 240

Time of Induction ( m i d

FIG. 4. Kinetics of induction of 8-GlcNAcidase by GlcNAc oligosaccharides. The level of 0-GlcNAcidase activity in V. furnissii grown in medium containing 0.6 mM (GlcNAc)P, (GlcNAc)a, or (GlcNAc), was measured by the PN-GlcNAc assay. Cultures of V. furnissii were grown to midexponential phase, and one of the GlcNAc oligosaccharides was added at the times indicated prior to harvesting the cells. All the cultures were harvested simultaneously and washed three times with buffered 50% ASW, and the activity of (3-Glc- NAcidase was determined in crude extracts. +, V. furnissii grown in the presence of (GlcNAc)*; A, (GlcNAc),; and 0, (GlcNAc),. The data presented in this figure represent the minimal induction by each of the oligosaccharides. Generally, induction by 0.6 mM (GlcNAc)? in lactate, 50% ASW medium was 100-200-fold over the control.

Cytosol ic

\ Intact cells

”

01 I 10 1 0 0 I O 0 0

(GlcNAc), Concentration (pM)

FIG. 5. Effect of the concentration of inducer, (GlcNAc)z, on the subcellular distribution of V. furnissii 8-GlcNAcidase activity. V. furnissii was grown in lactate, 50% ASW medium containing either no (GlcNAc)* or (GlcNAc), at the indicated concentrations. At midexponential phase, the cultures were harvested and washed. P-GlcNAcidase activity was measured by the continuous PN- GlcNAc assay using intact cells (+) and crude extracts (A) (total activity). The difference between the levels of activity in the extracts and the intact cells is designated cytosolic activity (0).

ability to induce P-GlcNAcidase activity in V. furnissii. Each oligosaccharide analogue was tested at 0.6 mM concentration in lactate, 50% ASW medium. Both MUF-@-(GlcNAc),, and MUF-P-(GlcNAc)3 gave full induction of the 8-GlcNAcidase activities. Fluorescent MUF was observed in these cultures, indicating that V. furnissii had hydrolyzed these analogues to the known inducers, (GlcNAc)* and (GlcNAc)3. Of the remaining glycosides tested, only MUF-P-GlcNAc showed slight inducing activity, and we cannot explain this result. The following compounds were inactive as inducers: GlcNAc, phenyl P-GlcNAc, phenyl a-GlcNAc, p-nitrophenyl P- GlcNAc, methyl a-GlcNAc, methyl P-GlcNAc, and MUF- thio-P-GlcNAc.

Subcellular Location of P-GlcNAcidase Actiuities

The first evidence suggesting the existence of two distinct 0-GlcNAcidase enzymes is presented in Fig. 5. V. furnissii was grown in lactate, 50% ASW medium containing either no inducer or (GlcNAc), at concentrations varying from 0.19 PM to 2.4 mM. Each culture was grown to midexponential phase, harvested, and intact cells and cell extracts were assayed for P-GlcNAcidase activity. The intact cell assays measured the activity located on the cell surface and/or in the periplasmic space, while the crude extract yielded the total P-GlcNAcidase activity. The difference between these two activities repre- sented the cytosolic activity. Fig. 5 shows that at low concentrations of inducer (less than 24 pM), the 6-GlcNAcidase activity was located almost exclusively in the cytosol. In contrast, at higher (GlcNAc), inducer concentrations, two pools of P-GlcNAcidase activity were expressed with about 30% of the total activity observed with intact cells.

Separation of V. furnissii membranes from the soluble fraction confirmed the above results. Two cultures of V. furnissii were grown in lactate, 50% ASW medium, containing either low (4.8 pM) or high (0.6 mM) concentrations of (GlcNAc)*. The cultures were harvested and washed, and p- GlcNAcidase activity was measured in intact cells, crude extracts, soluble, and membrane fractions. The combined activity from the supernatant and membrane fractions equalled that found in each crude extract. In experiments employing low levels of inducer, 100% of the observed p-


GlcNAcidase activity was located in the soluble f r a ~ t i o n . ~ Quite different results were obtained with high levels of (GlcNAc),, such as 0.6 mM. Under these conditions, the P- GlcNAcidase activity was approximately equally distributed between soluble and membrane-associated species (soluble = 45%, membrane = 58%).

Thus, the two sets of experiments indicate that V. furnissii contains more than a single species of p-GlcNAcidase. One P- GlcNAcidase is located in the cytosol and induced by low concentrations of (GlcNAc),, while the second is expressed by intact cells and remains associated with the membrane. The sec. nd enzyme is therefore located on the cell surface or in the periplasmic space and is only induced by the presence of high levels of (GlcNAc), (greater than 24 p ~ ) . The combined p-GlcNAcidase activity remained cell-associated under all conditions examined.

Specificities of the Periplasmic and Cytosolic P-GlcNAcidases

The experiments described above employed the synthetic analogue PN-GlcNAc for measuring the P-GlcNAcidases. The specificities of the two 6-GlcNAcidases were determined using chitin oligosaccharides. Not only did the two enzymes have separate subcellular locations and dissimilar induction pat- terns, but they also exhibited distinct substrate specificities.

V. furnissii was grown in lactate, 50% ASW medium containing 0.6 mM (GlcNAc):! to induce both the cytosolic and periplasmic P-GlcNAcidases. Substrate specificities of the respective P-GlcNAcidases were determined by incubating the soluble and particulate fractions with various concentrations of the oligosaccharides (GlcNAc),, n = 2-6, and the mixtures were assayed for GlcNAc appearance by the Morgan-Elson method. The results of the experiments using 5 mM concentrations of oligosaccharides are shown in Fig. 6, A and B. The periplasmic p-GlcNAcidase (Fig. 6A) hydrolyzed GlcNAc from each of the oligosaccharides a t approximately the same rate when the substrate was the dimer, trimer, tetramer, or pentamer. Cleavage of the hexamer occurred after a 10-min lag period. We cannot explain the lag; possibly, the membranes changed their permeability properties during the early part of the incubation. As shown below, the hexasaccharide is a substrate for the periplasmic enzyme when it is expressed in E. coli.

The cytosolic P-GlcNAcidase exhibited a dramatically different substrate specificity from the periplasmic form of the enzyme (Fig. 6 B ) . The soluble enzyme rapidly hydrolyzed the disaccharide (GIcNAc)~ and slowly acted on the trisaccharide (GlcNAc)a. The larger oligomers (GIcNAc)~, (GlcNAch, and (GlcNAc)6 were virtually inactive as substrates for the cytosolic P-GlcNAcidase.

Finally, the data taken together provide strong evidence for two 6-GlcNAcidases. The soluble, cytosolic enzyme is induced by low concentrations of (GlcNAc)z and hydrolyzes predomi- nately the disaccharide. The membrane-bound activity is expressed by intact cells, is induced by high concentrations of (GlcNAc)z, and hydrolyzes all of the oligomers, although not at equal rates. The potential role that each p-GlcNAcidase plays in “capturing” chitin oligosaccharide hydrolysis products is addressed under “Discussion.”

Although essentially no activity was detected in membranes prepared from cells induced with low levels of (GlcNAc)l, some 0- GlcNAcidase activity, from 15-25% of the total, was detected in the corresponding intact cells. As shown in Fig. 5, however, these values were close to background, and their validity is open to question. Alternatively, PN-GlcNAc may be transported by the (putative) (GlcNAc)? permease and cleaved in the cytosol which would also account for this result.

Tme (mid FIG. 6. Determination of V. furnissii membrane-bound

(periplasmic) and cytosolic @-GlcNAcidase substrate specificities. V furnissii was grown in lactate, 50% ASW medium containing 0.6 mM (GlcNAc), to induce both the cytosolic and periplasrnic forms of the P-GlcNAcidase. The culture was harvested at midexponential phase and washed three times with buffered 50% ASW at 0 “C, the cells were ruptured by passage through a French Press, and the membranes were separated from the soluble fraction by ultracentri- fugation (see “Materials and Methods”). For both fractions (soluble and particulate), protein equivalent to 1.5 X lo8 cells was incubated at 5 mM concentrations with each of the chitin oligosaccharides, (GlcNAc),, n = 2-6, in buffered 50% ASW at 25 “C. GlcNAc was measured by the Morgan-Elson method. A, membrane-bound S- GlcNAcidase; B, cytosolic enzyme.

V. furnissii Chitodextrinase: Induction and Subcellular Distribution

When V. furnissii is grown on colloidal chitin agar plates, clear zones are observed around the colonies, implying that the cells produce a chitinase; this hydrolase may be secreted or, alternatively, may be released as the result of cell lysis. Additionally, V. furnissii produces at least one chitodextrinase. Evidence for the existence of the latter was first obtained by hydrolysis of PN-(GlcNAc),, which is not cleaved by either of the P-GlcNAcidases (Table 11). To determine whether hydrolysis of PN-(GlcNAc), was or was not catalyzed by the secreted/cell surface chitinase, a mutant was isolated which could not grow on colloidal chitin (no clearing occurred around the colonies on colloidal chitin plates). However, the mutant retained the ability to hydrolyze both PN-GlcNAc and PN- (GlcNAc),. Since the p-GlcNAcidases cannot hydrolyze PN- (GlcNAc),, cleavage of this analogue by extracts of the mutant suggested that wild type V. furnissii expresses both a chitinase and a chitodextrinase.

In experiments analogous to those presented for the p- GlcNAcidases, induction of chitodextrinase activity was tested by including different concentrations of (GlcNAch in the growth medium. The activity on the cell surface or in the periplasmic space was assayed with intact V. furnissii. How- ever, attempts to measure the “total” chitodextrinase activity failed. The addition of toluene to permeabilize the cytosolic membrane or French Press treatment of the cells resulted in as much as 100% loss of the activity compared to that obtained with intact cells. We are unable to explain the loss of activity at this time. When chitodextrinase activity was detected in the crude extract, it was found exclusively in the washed particulate fraction. Furthermore, no chitodextrinase activity (assayed with PN-(GlcNAc),) was detected in the growth medium. The chitodextrinase activity observed in intact V. furnissiiwas induced by (GlcNAc):!. Maximal induction, about 35-fold, occurred with 0.6 mM (GICNAC)~ in the lactate, 50% ASW medium, similar to the induction results obtained with


TABLE I1 Kinetic data for the p-GlcNAcidases and chitodextrinase

A summary of kinetic data ( Vmax and K,,, values) is presented in the table. V . furnissii and the two E. coli transformants HB101:pBB20 and HB101:pBB22 were grown in minimal media in the absence and presence of the inducer, 0.6 mM (GlcNAc)2. The V. furnissii p- GlcNAcidase (exo-) Vmax was measured in crude extracts using PN- GlcNAc. The calculated value includes both the soluble and membrane forms. The chitodextrinase (endo-) Vmax was measured in intact-induced V. furnissii cells. The true V,,, of this enzyme in V. furnissii may be larger (see text). The K,,, of the chitodextrinase could not be determined in V. furnissii. The VmaX values for the E. coli transformants were measured in crude extracts. No change in either the V,,, or K,,, values was observed after growth of the E. coli transformants in the presence of (GlcNAc)n. PN-GlcNAc and PN- (GlcNAc)z were the substrates for measuring 0-GlcNAcidase (exo-) and chitodextrinase (endo-) activities, respectively. The assays were performed in 50 mM sodium phosphate buffer, pH 7.4, and room temperature. Production of p-nitrophenol was continuously moni- tored at 400 nm.

VmaX K, Cell type Uninduced Induced Induced

Exon Endo’ Exo Endo Exo Endo pmollrninlpg protein P M

V. furnissii 2.0 1.0 333 35 307 ND‘ E. coli 912 - 833 -

E. coli - 19.4 - 25 - 66 HB101:pBB20

HB101:~BB22

286 -

~ a Exo refers the 8-GlcNAcidase activity. Endo refers to the chitodextrinase activity.

The line under “endo” signifies undetectable rates of hydrolysis with PN-(GlcNAc)z, i.e. ~ 0 . 1 pmol/min/mg, while the corresponding lines under “exo” signify <2 pmol/min/mg with PN-GlcNAc.

the P-GlcNAcidases. Again we stress, that, in these studies, only a portion of the enzymatic activity may have been detected in the assay. The substrate specificity of the chitodextrinase was determined using the cloned gene product and is discussed below.

‘ ND, not determined.

Kinetic Studies with p-Nitrophenyl Glycosides Apparent kinetic constants for the 0-GlcNAcidases and the

chitodextrinase were determined and are summarized in Table 11. In these experiments, extracts of the E. coli transformants were used to study the cloned 0-GlcNAcidase and the chitodextrinase, extracts of V. furnissii to study the combined 0- GlcNAcidases, and intact V. furnissii cells for the chitodextrinase. It is evident that the gene products expressed in the two E. coli transformants are distinct. The gene product in pBB2O (P-GlcNAcidase) hydrolyzes only PN-GlcNAc, while the enzyme cloned into pBB22 (chitodextrinase) is specific for the corresponding trisaccharide analogue PN-(GlcNAc)2. Additionally, the two fragments of V. furnissii DNA showed no homology by Southern blotting.

The P-GlcNAcidases and chitodextrinase were inducible in V. furnissii (Table 11). In contrast, the E. coli strains, HB101:pBB20 and HB101:pBB22, constitutively expressed high levels of the (3-GlcNAcidase and chitodextrinase activities, respectively, and (GlcNAc)* had no inducing effect on these activities. The same results were obtained with E. coli HB101:pBB25, which also expresses the P-GlcNAcidase. The constitutive level of the /3-GlcNAcidase in the E. coli transformant was 3 times greater than the maximally induced levels of the combined P-GlcNAcidases in V. furnissii. The chitodextrinase was expressed at about the same level in the transformant as in induced V. furnissii. In the only case in which a comparison could be made (the 8-GlcNAcidases),

roughly the same K,,, value was obtained for the p-Glc- NAcidase expressed by the transformant and by V. furnissii, which is actually a reflection of the mixture of two Glc- NAcidases (Fig. 7).

Both cloned enzymes were completely soluble in cell-free extracts of the transformants. Furthermore, no activity was detected with intact E. coli cells, but full activity was observed following permeabilization of the cells with toluene. These results indicate that neither enzyme was transported out of the cytosol in the E. coli transformants, as they are in V. furnissii.

Specificity Studies with the Cloned Enzymes

/3-GlcNAcidase-To determine which of the p-Glc- NAcidases had been cloned into E. coli HB101:pBB20, extracts of this transformant were incubated with each of the chitin oligosaccharides (GlcNAc),, n = 2-6 for various time periods, the mixtures were boiled to stop the reaction, dena- tured protein was removed by centrifugation, and the supernatant fluids were analyzed for released GlcNAc by the Mor- gan-Elson method. The results are shown in Fig. 8, and it is evident that the cloned enzyme exhibits the same substrate specificity (qualitatively) as the membrane-associated, i.e. periplasmic, enzyme of V. furnissii. Each of the oliogsacchar- ides was hydrolyzed by the cloned enzyme, the trisaccharide was the preferred substrate, while the least effective substrate, the disaccharide, was utilized at about 20% of this rate. The cloned enzyme (Fig. 8) appears to give different kinetics than the particulate enzyme from V. furnissii (Fig. 6). The membrane-bound enzyme gave rates which were approximately constant with the time of incubation, whereas the rates with the cloned enzyme were not, particularly with the disaccharide. However, the cloned enzyme was assayed with 0.5 mM substrates (Fig. 8), whereas the V. furnissii membrane-bound enzyme was assayed a t 5 mM concentrations (Fig. 6). An extensive kinetic analysis will be required to fully explore the apparent differences. It is possible, for example, that the reaction product, GlcNAc, is a potent inhibitor of the enzyme a t low disaccharide concentrations, and, if true, this would have considerable physiological implications.

0 3 6 9 I2

Time (rnin) FIG. 8. Substrate specificity of the 8-GlcNAcidase cloned

into E. coli HB101:pBBZO. A crude extract of E. coli HB101:pBB20 containing the cloned V. furnissii 0-GlcNAcidase was incubated with each of the chitin oligosaccharides (GlcNAc),, n = 2- 6 at 0.5 mM. The GlcNAc released from each oligosaccharide was measured by the Morgan-Elson method. The symbols for the substrates are: 0, A, A, + and 0 designate GlcNAc release from (GlcNAc)z, (GlcNAc)3, (GlcNAc),, (GlcNAcIs, and (GlcNAc)e, respectively.


Chitodextrinase-Substrate specificity determinations with the cloned chitodextrinase employed the TosoHaas Amide 80 HPLC column system (see “Materials and Methods”). The E. coli HB101:pBB22 extract was used at protein concentrations ranging from 20 pg to 1 mg of protein per incubation (650 pl total volume), and the oligosaccharide concentrations ranged from 0.2 mM to 2 mM. Fig. 9 shows typical data from one such experiment, using 1 mM pentasaccharide as the substrate. After 1 min, (GlcNAc)z and (GlcNAc)3 were observed, and continued incubation resulted in increased quantities of these products, but neither GlcNAc nor the tetrasaccharide was observed at any time during the incubation. Integration of the areas beneath each peak demonstrated that 1 nmol each of (GlcNAc)2 and (GlcNAc)3 was produced for each nanomole of (GlcNAc):, hydrolyzed. The use of more enzyme resulted in the complete degradation of the pentasaccharide in 20 min.

In sharp contrast to (GlcNAc),, neithr the di- nor trisaccharide was hydrolyzed after 20 min of incubation. These experiments were performed using 2 mM concentrations of the oligosaccharides and large quantities of the extract (1 mg of protein).

The tetrasaccharide, which is the smallest oligosaccharide substrate for the chitodextrinase, disappeared with the con- comitant appearance of only one product, (GlcNAc)z. In all experiments with the tetrasaccharide (various enzyme and (GlcNAc), concentrations), the reaction rate changed dramatically during the incubation; an initial rapid cleavage was followed by a decrease in the rate, and in no case was the substrate completely utilized.

When (G~CNAC)~ was the substrate, results similar to those observed for (GlcNAc)5 were obtained. The hexamer was completely hydrolyzed within 20 min with sufficient enzyme in the mixture, yielding both (GlcNAc)z and (G1cNAc)B. Since these products are not substrates of the enzyme, this result implies that (GlcNAc), was either cleaved once to 2 molecules of (GlcNAc)a or cleaved twice yielding 3 molecules of (GlcNAc)*. The disaccharide can only be produced from (G~CNAC)~ with the simultaneous formation of (GlcNAc),. However, tetrasaccharide production was never observed in

R Incubation Time= 0 min

20 min

(GlcNAcI, n= I 2 3 4 5 6 Elution 01 t t t t t t

FIG. 9. Chromatograms illustrating the kinetics of hydrolysis of (GlcNAc)5 by the cloned chitodextrinase. E. coli HB101:pBB22 was grown to midexponential phase, harvested, washed, and a crude extract was prepared as described in the text. Replicate incubation mixtures were prepared containing (in final volumes of 650 pl): 1 mM (GlcNAc)6, 1.0 mg of protein, and 50 mM sodium phosphate buffer, pH 7.4. Reactions were initiated by adding the extract, incubated at 25 “C, and analyzed as described under “Materials and Methods.” The figure shows representative chromatograms for the zero time, 1- and 20-min time incubations.

these experiments. One explanation for these results is that the enzyme acts processively on (GICNAC)~, producing (GlcNAc), and enzyme-bound (GlcNAc),.

Our preliminary kinetic experiments did not permit estimates of apparent K, values because the hydrolysis rates changed with time of incubation, and initial rates were particularly difficult to measure at low substrate concentrations. Initial rates of hydrolysis at 1 mM substrate concentrations could be estimated, however, and were (pmol/min/pg of protein): (GlcNAc),, 10; (GlcNAch 7; (GlcNAcL, 4; (GlcNAcL and (GlcNAc)*, 0.

Finally, it should be emphasized that the chitodextrinase showed a consistent pattern of hydrolysis with regard to glycosidic linkage. In each case, the first and last glycosidic bonds in the oligosaccharide were completely resistant to hydrolysis by the enzyme. Thus, (GlcNAc), could only give the disaccharide, (GlcNAc)s only the di- and trisaccharides, and (GlcNAc),6 could yield di-, tri-, and tetrasaccharides, but actually yields only the dimer and trimer for the reasons discussed above. The requirement for internal glycosidic bonds for hydrolysis explains why the di- and trisaccharide must be the end products of action of this enzyme.

DISCUSSION

V. furnissii grows on agar surfaces with colloidal chitin as the sole source of both carbon and nitrogen. Clear zones are evident around large colonies, implying the secretion of a soluble chitinase, or, alternatively, a membrane-bound form of chitinase which is solubilized during cell lysis. The chitinase extracted from the agar rapidly solubilizes particulate chitin, whereas the chitodextrinase described in this report does not. The extracellular chitinase is not expressed by cells grown in liquid culture and is not induced by (GlcNAcL in the growth medium. These results are consistent with those of Zobell and Rittenberg (5), who demonstrated that efficient solubilization of chitin occurred only after marine bacteria had been allowed to settle onto strips of chitin. The V. furnissii chitinase is under investigation.’

Soluble oligosaccharides liberated by the extracellular chitinase elicit immediate physiological responses in V. furnissii by inducing the expression of a number of proteins. Some of the latter are soluble and others are membrane-associated; they include cytoplasmic and periplasmic proteins, enzymes, transporters, chemoreceptors, and possibly porins and periplasmic solute binding proteins. The primary function of the induced proteins is to provide V. furnissii with the apparatus necessary to find, capture, and degrade the chitin oligosaccharides. The preceding paper (2) reports the chemotactic systems for finding the oligosaccharides and here we describe some steps in their capture and degradation. This discussion is aimed at interpreting the results of oligosaccharide utilization by intact V. furnissii (Fig. 3, A-C) and in the membrane- permeabilized cells (Fig. 3, D-F). Since some of the important degradative reactions occur within the periplasm, it is perti- nent to note that the periplasmic space occupies from 20-40% of the total cell volume in E. coli and Salmonella typhimurium (48), organisms that are closely related to V. furnissii.

Induced V. furnissii cells consume GlcNAc at 25 “C at approximately the optimal rate reported for glucose consumption by E. coli at 37 “C (300 nmol/min/mg of protein (46)). Furthermore, only 15 min are required for induction of the following proteins: Enzyme I (and presumably HPr) and Enzyme IIN”g of the PTS (31, which together are required for chemotaxis to, and uptake of, GlcNAc by these cells (49, 50);

C. Rowe, unpublished studies.


the catabolic enzymes, GlcNAc-6-P deacetylase and GlcNHz- 6-P deaminase (Table I). The cytosolic, ATP-dependent, GlcNAc kinase is constitutive.

In addition to GlcNAc, V. furnissii is remarkably efficient at capturing virtually all of the carbon (and presumably the nitrogen) contained in the oligosaccharides (Fig. 3). The di- and trisaccharides were almost completely consumed in 2 h, i.e. about half the rate of GlcNAc on a molar basis, and at an equivalent rate on a weight basis, while the tetrasaccharide was utilized somewhat more slowly. Therefore, intact cells catabolize 2 times more oligosaccharide than total cell mass (dry weight) in 1 h, with little to no release of the expected intermediate hydrolysis products. The remaining experiments were concerned with defining some of the steps in these processes.

When (GlcNAc)?, (GlcNAc),, or (GlcNAc), is added to the growth medium, in addition to the GlcNAc degradative enzymes mentioned above, V. furnissii expresses the following: a periplasmic chitodextrinase, a periplasmic P-GlcNAcidase, and a cytoplasmic P-GlcNAcidase. There is barely detectable activity of the hydrolases in uninduced V. furnissii, and induction requires (GlcNAc),. The tri- and tetrasaccharide also act as inducers, but since all active inducers are convertible to the disaccharide, we suggest that it is the true inducer. (GlcNAc)z also acts as an inducer (probably indirectly) of the enzymes required for GlcNAc catabolism. V. furnissii utilizes GlcNAc at an identical rate (Fig. 2) following induction by either the mono- or the disaccharide.

One of the most interesting results reported here is the differential induction of the periplasmic and cytosolic P- GlcNAcidases (Fig. 5). At low inducer concentrations (less than 24 PM), only the cytoplasmic form of the P-GlcNAcidase is induced, while at higher (GlcNAc), concentrations, both enzymes are induced. Thus, under conditions where limiting quantities of (GlcNAc), are available to the cells, only the disaccharide pathway is operating (Fig. l ) , but, when there is a greater supply of (GlcNAc),, both the monosaccharide and disaccharide pathways are utilized. Thus, V. furnissii is capable of efficiently utilizing oligosaccharides under a variety of physiological conditions.

It is of interest to compare induction of the hydrolases in V. furnissii with induction of the chemoreceptors (discussed in the preceding paper (2)). Both groups of proteins are expressed by including (GlcNAc), in the growth medium, and, since maximal induction is obtained under the same conditions (0.6 mM (GlcNAc), in lactate medium), both systems appear to be coordinately regulated. However, when V. furnissii is grown on (GlcNAc), as the sole carbon source, the chemotactic response is repressed, while the hydrolases remain maximally induced. Therefore, regulation of the genes encoding the glycosidases probably has some features in common with regulation of those encoding the chemotaxis apparatus, but there must be additional specific control of the chemotaxis system.

The upper limit of oligosaccharide size that can be hydrolyzed by the chitodextrinase has not been defined, but we know that it hydrolyzes (GlcNAc),, n = 4-6, while it is completely inactive with the di- and trisaccharides, which are therefore the final products of hydrolysis of the higher oligosaccharides. For reasons offered under “Results,” it seems that the chitodextrinase is a processive enzyme and does not release intermediate hydrolysis products into the medium. A complete kinetic characterization of this interesting, and apparently novel enzyme, awaits purification to homogeneity?

The periplasmic P-GlcNAcidase is an exoenzyme, which

N. Keyhani, unpublished studies.

hydrolyzes the nonreducing GlcNAc termini of all of the oligosaccharides tested, although at different rates (Figs. 6A and 8). (GlcNAc), is the most and (GlcNAc), the least effective substrate for the periplasmic P-GlcNAcidase. In sharp contrast, the cytoplasmic P-GlcNAcidase rapidly hydrolyzes the disaccharide, slowly cleaves the trisaccharide, and shows little to no activity with the higher oligomers (Fig. 6B).

The genes encoding the chitodextrinase and the periplasmic p-GlcNAcidase have been cloned into and expressed in E. coli, and the enzymes were particularly useful for determining their respective substrate specificities. The cloned enzymes remain entirely in the cytoplasm of the transformants, which may indicate that the signal sequence used for translocating periplasmic proteins in V. furnissii is not recognized by E. coli. The cloned hydrolases are constitutively expressed by the E. coli transformants at levels as great as, or exceeding, the maximally induced levels in V. furnissii. Furthermore, the transformants are not induced by (GlcNAc)?. Nevertheless, the V. furnissii DNA fragments cloned into E. coli may contain the regulatory regions necessary for control of expression in V. furnissii, but the transformants may lack a repres- sor(s) which regulates expression in the Vibrio. Intact E. coli transformants were unable to utilize the disaccharide, despite their high levels of P-GlcNAcidase (and the ATP-dependent GlcNAc/Glc specific kinase (31)). Presumably, the transformants lack the disaccharide permease (discussed below), so that (GlcNAc), cannot be transported into the cell to regulate gene expression or to be catabolized.

Before discussing the alternate pathways for oligosaccharide utilization, we shall consider the putative disaccharide permease. The evidence for such a permease is summarized as follows. (a ) ( GlcNAc)2, but not GlcNAc, induces expression of the glycosidases and the oligosaccharide chemoreceptors. ( b ) GlcNAc is transported by the PTS via IINag (3) and is phosphorylated by toluene-permeabilized V. furnissii cells when supplemented with phosphoenolpyruvate. No phosphorylation of (G1cNAc)n was observed under the same conditions, and the presence of (GlcNAc), did not inhibit GlcNAc phosphorylation. ( c ) When the PTS translocates and phos- phorylates a disaccharide, the hydrolase that splits the transport product is invariably a phosphoglycosidase, i.e., it specif- ically cleaves the disaccharide phosphate (47). In V. furnissii, the cytosolic enzyme is a P-GlcNAcidase. ( d ) The mutant defective in IINag, V. furnissii AP801, which cannot grow on nor utilize GlcNAc, grows at normal rates on (GlcNAc), and is inducible for both the hydrolases and oligosaccharide chemoreceptor proteins (Table I and Ref. 2). The permease remains to be characterized and its specificity defined, but it may translocate (GlcNAc), in addition to (GlcNAc),. We suggest this possibility because the trisaccharide is an end product of digestion by the periplasmic chitodextrinase and is a substrate, albeit less effective than the disaccharide, for the cytosolic P-GlcNAcidase.

We propose that a chitin oligosaccharide enters the periplasmic space, possibly via an oligosaccharide porin, and then binds to a specific solute binding protein.’O In the periplasmic space, the oligosaccharide is hydrolyzed to GlcNAc by the p- GlcNAcidase and to variable quantities of (GlcNAc), and (GlcNAc), by the chitodextrinase. (GlcNAc), is also hydrolyzed by the periplasmic P-GlcNAcidase to the monomer and

lo Specific porins and a solute binding protein which acts as a chemoreceptor for maltose are induced by maltose and maltodextrins in E. coli (3). N. Keyhani has shown that induction of V. jumissii with (GlcNAc), results in the expression of a major protein in the outer membrane. The outer membrane preparations show no hydrolase activities.


dimer. The monomer is transported by the PTS and enters the cytosol as GlcNAc-6-P, while the disaccharide (and possibly some trisaccharide) is translocated by the permease and hydrolyzed to GlcNAc by the cytoplasmic P-GlcNAcidase. The cytoplasmic GlcNAc formed in this step is phosphorylated by the ATP-dependent kinase. Thus, both transport systems (the PTS and the (GlcNAc)a permease), ultimately yield GlcNAc-6-P. The latter is deacetylated and deaminated to acetate, ammonia, and fructose-6-P.

As mentioned, our aim has been to quantitatively explain how V. frrnissii utilizes oligosaccharides without releasing the interme..iate hydrolysis products (Fig. 3). Full understanding of oligosaccharide capture by V. furnissii awaits rigorous kinetic characterization of the different enzymes and transporters. These studies would enable us to predict the rate of utilization of each oligosaccharide and the function of each protein in converting an oligomer to fructose-6-P.

We do not yet possess the data to make such an analysis, but our first crude attempts to explain the observed results are summarized as follows. The relative rates of the known periplasmic membrane, cytoplasmic enzymes, and transporters in intact V. furnissii cells that were maximally induced for the glycosidases with 0.6 mM (GlcNAc)z in lactate medium are as follows (nmol/min/lO* cells): chitodextrinase = 0.7-1.3 for (GlcNAc),, n = 4,5,6; periplasmic p-GlcNAcidase = 0.3- 0.6 for (GlcNAc),, n = 2-6; (GlcNAc)z-permease = 0.7; GlcNAc-permease (lINag) = 1.5; cytosolic P-GlcNAcidase = 12 for (GlcNAc),; cytosolic ATP-dependent GlcNAc kinase = 0.9; GlcNAc deacetylase = 1.3; and deaminase = 10.” The methods and assumptions used to obtain these estimates are given in the footnote; the least reliable values are those for the chitodextrinase. We emphasize that these measurements and/or estimates apply only to initial rates at 0.6 mM substrate concentrations.

The key points and predictions that arise from these data are the following. (i) In general, the activities of the cytoplasmic degradative enzymes are equal to or greater than the activities of the transporters, which in turn are more active than the periplasmic p-GlcNAcidase. Thus, the cleavage products formed in the periplasmic space are transported and catabolized before they can diffuse into the extracellular space. (ii) In “the monosaccharide pathway” of Fig. 1, GlcNAc uptake via the PTS proceeds a t about the same rate as the deacetylase, which implies that only low levels of GlcNAc-6- P would accumulate inside the cells at the steady state; GlcNAc-6-P could be the inducer of the nag operon. Con- versely, it appears that GlcNH2-6-P does not accumulate in these cells. The levels of deaminase are so high as to suggest immediate and complete conversion of GlcNH2-6-P to fructose-6-P and NH3. (iii) Under less optimal inducingconditions (Fig. 5 ) , the “disaccharide pathway” is used exclusively for (GlcNAc)* catabolism because the periplasmic P-GlcNAcidase is not induced. Once in the cytoplasm, (G1cNAc)s is hydro- ~ _ _ _ _

’’ The relative rates of hydrolysis by the chitin degradative enzymes or transport by the permeases was determined as follows: Chitodex- trinase, the rates of hydrolysis of the oligosaccharides by extracts of E. coli HB101:pBB22 were converted to the comparable activities in V. furnksii by normalizing the rates in the two cell types with PN- (GICNAC)~. These values are considered the least reliable. Periplasmic 0-GlcNAcidase, from hydrolysis of the oligosaccharides by V. furnissii membran9 (Fig. 6A). (GlcNAc), permease, from the rate of (GICNAC)~ disappearance with intact V. furnissii (Fig. 3A). (G1cNAc)- permease, from the rate of GlcNAc utilization by intact V. fumissii (Fig. 2). Cytosolic P-GlcNAcidase, from the hydrolysis of chitin oligosaccharides using the soluble fraction prepared from V. furnissii (Fig. 6B). The ATP-dependent GlcNAc kinase, deacetylase, and deaminase activities were also measured in the soluble fraction from V. furnissii (Table I).

lyzed by the most active enzyme in either pathway, the cytosolic p-GlcNAcidase, and the resulting GlcNAc is phosphorylated by the constitutive ATP-dependent kinase. Since the kinase appears to be a t least as active as the permease, no accumulation of free internal GlcNAc is predicted. (iv) If only the induced periplasmic p-GlcNAcidase hydrolyzed the tri-, tetra-, penta-, and hexasaccharides, then the above consider- ations apply, i.e. hydrolysis in the periplasm is the rate- limiting step in oligosaccharide utilization, and the products of hydrolysis are transported and catabolized as rapidly as they are formed. A key, as yet undetermined factor, is of course the contribution of the periplasmic chitodextrinase in these processes, and this will be especially true for the higher oligosaccharides.

In this report we have primarily considered catabolism of the di-, tri-, and tetrasaccharide, but have not yet adequately investigated utilization of the penta- and hexasaccharides, nor of higher oligosaccharides. We know that the penta- and hexasaccharides are excellent carbon sources for cell growth and chemotaxis on swarm plates. Preliminary experiments show that growth on the penta- or hexasaccharide induces hydrolases and/or porins in addition to the series of proteins described in this report (2).

Other questions remain unanswered. What are the sizes of the chitodextrins produced by the extracellular chitinase? Are specific porins induced to permit penetration of the larger oligosaccharides into the periplasmic space where they can be acted on by the chitodextrinase? If so, what is the exclusion limit of these porins? Are periplasmic solute binding proteins induced to salvage low levels of the oligosaccharides? Do the genes encoding these proteins and enzymes exist in operons, and, if so, how are they regulated?

Finally, the results presented here must be viewed as no more than a starting point in understanding the sophisticated mechanisms underlying chitin oligosaccharide utilization by V. furnissii.

REFERENCES 1. Yu, C., Lee, A., Bassler, B. L., and Roseman, S. (1991) J. Bid.

Chem. 266,24260-24267 2. Bassler, B. L., Gibbons, P. J., Yu, C., and Roseman, S. (1991) J.

Biol. Chem. 266, 24268-24275 3. Macnab, R. M. (1987) in Escherichia coli and Salmonella lym-

phimurium: Cellular and Molecular Biology (Neidhardt, F. C., Ingraham, J. L., Low, K. B., Magasanik, B., Schaechter, M., and Umbarger, H. E., eds) Vol. 1, pp. 732-759, American Society for Microbiology, Washington, D. C.

4. Johnstone, J. (1908) Conditions of Life in the Sea, Cambridge University Press, Cambridge

5. Zobell, C. E., and Rittenberg, S. C. (1937) J. Bacteriol. 35, 275- 287

6. Zechmeister, L., and Toth, G. (1939) Enzyrnologia 7, 165-169 7. Zechmeister, L., Toth, G., and Vajda, E. (1939) Enzyrnologia 7,

8. Jeuniaux, C. (1959) Arch. Intern. Physiol. Biochem. 67,115-116 9. Trumbly, R. J., Robbins, P. W., Belfort, M., Ziegler, F. D., Maley,

10. Kinoshita, K., Taniguchi, N., Makita, A., Narita, M., and Oikawa,

11. Pedraza-Reyes, M., and Lopez-Romero, E. (1989) J. Gen. Micro-

12. Hara, S., Yamamura, Y., Fujii, Y., Mega, T., and Ikenaka, T.

13. Colowick, S. P., and Kaplan, N. 0. (1988) Methods Enzymol.

14. Tracey, M. (1955) Biochern. J. 61,579-586 15. Fischer, E., and Stein, E. (1960) in The Enzymes (Boyer, P. D.,

Lardy, H., and Myrback, K., eds) Vol. 4, p. 301, Academic Press, New York

16. Soto-Gil, R. W., and Zyskind, J. W. (1984) in Chitin, Chitosun

170-175

F., and Trimble, R. B. (1985) J. Biol. Chern. 260, 5683-5690

K. (1988) J. Biochem. (Tokyo) 104,827-831

biol. 135, 211-218

(1989) J. Biochem. (Tokyo) 105, 484-489

161, 403-524


and Related Enzymes (Zakikas, J. P., ed) pp. 209-223, Academic 31. Roseman, S. (1957) J. Biol. Chem. 226, 115-124 Press, New York 32. Comb, D. G., and Roseman, S. (1958) J. Biol. Chern. 232, 807-

J . W. (1987) J. Bacteriol. 169, 3785-3791 33. Bassler, B. L., Gibbons, P. J., and Roseman, S. (1989) Biochem.

Appl. Enuiron. Microbiol. 52, 142-145 34. Yu, C., Lee, A,, and Roseman, S. (1987) Biochem. Biophys. Res.

W. (1988) J. Bacteriol. 170, 2984-2988 35. Bassler, B. L., Bouma, C., and Roseman, S. (1988) J. Cell Biol.

17. Jannatipour, M., Soto-Gil, R. W., Childers, L. C., and Zyskind, 827

18. Wortman, A. T., Somerville, C. C., and Colwell, R. R. (1986) Biophys. Res. Comrnun. 161,1172-1176

19. Joshi, S., Kozlowski, M., Selvaraj, G., Iyer, V. N., and Davies, R. Cornrnun. 149,86-92

20. Kuranda, M. J., and Robbins, P. W. (1987) Proc. Natl. Acad. Sci. 107,535A

21. Robbins, P. W., Albright, C., and Benfield, B. (1988) J. Biol. (1951) J . Biol. Chern. 193, 265-275

22. Broglie, K. E., Gaynor, J. J., and Broglie, R. M. (1986) proc. Natl. 38. Reissig, J., Strominger, J., and Leloir, L. (1955) J. Bid. Chem 217,959-966

23, Fuchs, R, L., ~ ~ p h ~ ~ ~ ~ ~ , s, A., and Drahos, D, J. (1986) ~ ~ ~ l , 39. Maniatis, T., Frits+ E. F., and Sambrook, J. (1982) Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory

24. Horwitz, M., Reid, J., and Ogrydziak, D. (1984) in Chitin, Chito- Press, Cold Spring Harbor, NY

U. S. A. 84,2585-2589 36. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J.

Chem. 263,443-447 37. Morgan, W., and Elson, L. (1934) Biochem. J . 28,988

Acad. Sci. U. S. A. 83, 6820-6824

Enuiron. Microbiol. 51, 504-509

san, and Related Enzymes (Zakikas, J. P., ed) pp. 191-208, 40' Marmur, J. (lg61) J . Mol. Bioi. 3 9 208-218 Academic Press, New York

O'Dowd, B. F., Tropak, M., Willard, H. F., Anderson, M.-J., 43. Southern, E. (1975) J , Mol. Bioi, 98, 503-517

41. Mandel, M., and Higa, A. (1970) J. Mol. Biol. 53, 159-162 42. Pulley-Blank, D., and Michalak, M. (1983) Mol. Biol. Rep. 9 , 25. Korneluk, R. G., Mahuran, D. J., Neote, K., Klavins, M. H., 191-195

J' A'' and R' A. (1986) J' 2619 44. O'Brien, M., and Colwell, R. R. (1987) Appl. Enuiron. Microbiol.

26. Soto-Gil, R. W.9 and Zyskind, J. W. (1989) J . Bid. Chern. 264, 45. Anthony, R,, and spector, L. (1971) J , ~ i ~ l , chern, 247, 2120- 8407-8413 53,1718-1720

14778-14783 27. Robbins, P. W., Trimble, R. B., Wirth, D. F., Hering, C., M a h , 46. Schultz, K. L,, and ~ i ~ ~ , R, s, (1964) ~ ~ ~ h , Mikriobiol, 48, 1-20

2125

F., Maleyy G. F-, Dasr R., Gibson* *. w.* Roya1* N.z and 47. Meadow, N. D., Fox, D. K., and Roseman, S. (1990) Annu. Reu. Biemann, K. (1984) J. Biol. Chern. 259,7577-7583 Biochem. 59, 497-542

28. Kameiy K.* YamamUra, y. , Haray s.9 and Ikenaka, T. (1989) J. 48. Stock, J., Rauch, B., and Roseman, S. (1977) J . Biol. Chem. 252, Biochem. (Tokyo) 105,979-985 7850-7861

29. Colwek R. R. (1984) Vibrios in the Enuironment, John Wiley 49. Adler, J., Hazelbauer, G. L., and Dahl, M. M. (1973) J . Bacteriol. and Sons, Inc., New York 115,824-847

30. Asensio, C., and Ruiz-Amil, M. (1966) Methods Enzymol. 9, 421- 50. Lengler, J., Auburger, A.-M., and Percher, A. (1981) Mol. & Gen. 425 Genet. 183,163-170

Materials- The following SUbstlnCeB were purchased from the indicated sources: chitin OligOSaCCharides from Seikagaku America Inc.

IRIF, DMF. PN-GlcNAC, PN-(GlcNAc)a, methyl a- and U-ClCNAc. and L-amino (St Petersburg. FL); GlcNAc, (GICNAC)I, (GlcNAc),, (G1CNACl.. chitin,

acids, Signa Chemical CO. (St. Louie. Y O ) ; the glycosides WF-GICNAc and H ~ ~ F - ( G ~ E N A E ) ~ , from Calbiochem (Id Jolla, CAI; reagants Used far preparation Of bacterial media were Of the highest purity available from DifEO LabDratOriea (Detroit, MI1 and J. T. Baker (Phlllipsburq.

used in DNA cloning experiments. the vector pBR322, restriction N.J.); Hepe8 from Reeearch Organics Inc. (Cleveland, OH); reagents

enzymes and buffers, bacterial alkaline phosphatase, DNA ligase, and ultra-pure Agarose, from Bethasda Research IdbOratories ICaithPrSbYrg, Mol: radioisotomes from the DY POnt-New Ensland Nuclear CO. Iwilminq-

little to no &ior in this reaction.

designated APBol, vas isolated as follows: The parental strain, Y .

guanidine as demoribod i l l , and the cmlls screened for GIcNAc nom- furnissii 1225, was nutagenired with N-methyl-N~-nitr-O-N"-nitrosa-

fermentoTS on M E plates containing 1: NaCl and 0 . 5 8 GlCNAc. NO"- fermenting colonies were streaked on MBINaCl platen containing 0.5t glucose or mannose. Positive fermentors Of both glucose and mannose were picked and purified by repetitive single colony ieolation. The mutant, designated AP801. did not spontaneously revert to positive GICNAc fermentation, and some Of its growth Charactaristics and biochemical analyses a1111 given in Table I. Crude eXtraCtI) of the mutant contained normal levels ai the s ~ l u b l e PTS proteins (Enzyme I, HPr, and I l l r L r ) . and these extracts phosphorylated GlCNAc when supplemented with membranes f m m the parental strain. The only detectable defect in strain APE01 was in the FTS membrane protein. 11"". which catalyzes the transport and phosphorylation Of GlcNAc. Tu0 other

deaminssa. Yere induced bv 1GlCNAC). but not bv ClcNAc lToblr, 11. and proteins encoded by genes in the same operon, the deacetylos= and

Isolation and Characterization of a 11% Rutant---A mutant,

~~

thus the mutant could grow on the disaccharide.

nii 1225 was grown at 25-C (1) in minimal-lactate medium containing nener.cl~-buffertld 50a artificial sea water (desiqnated lactate-50:

Growth or Bacteria, and Cloning of v . fumissii Genes- Y. .€uu&z

. . ..

Ask msdiunl, or. when specified. in a complex marine medium containing Peptone and Yeast Extract. Inocula yere grown overnight to stationary phase in the complex medium, diluted 1:25 into the lactate-5Ol ASW medium, grown to mid-exponential phase and harvested as described below. When specified, 55 nM GlCNAc or 5 Oligosaccharide was substituted for the lactate. 8-GlcNAcidase activity wae induced in p.

. .

mcI flnal Concentration) in the lactate growth nadium. LvrnLaaii by including the (GICNAEI. oliqoners, n = 2, 1. or 4 (at 0 . 6

(19) containing 208 glycerol and grown overnight at 37'C in LB with E . HBIOI wae *torad at -18T in Luria-Pertmi medium (LB)

overnight cultire at a 1:20 dilution and the Culture groii-ti mid: "

HBlOl transformants were g r o w With 15 ug per m1 ampicillin in LB- exponsntioil phase aS measured by the absorbance at 590 nm. E . rpli

medium was prepared ae described (39) and contalned 0.2: glycerol as broth. and 10 ug per m l ampicillin on LB-agar plates. 19 minimal

the carbon source.

GlcNAcidaae and chitodextrinase genes Were generoia g1fts of n!: Carolyn Boyma. Briefly. genomic DNA was extracted from Y. &d 111. These fragnents Were ligated into the plasmid pBR322 by the method Of MarmUr ( 4 0 1 and digested to completion with h HI or

praviausly digested Wlth Um HI or Kind 111, and bacterial alkaline

E. Epli HBlol transformed by the standard procedure (41). Plasmids phosphatase. The library was prepared by amplifying tha plasnida in

Were prepared as described (421. DNA homology between the cloned p.

assesbed by the method of Southern (41). genes enccding the E-GlcNAcidaSa and chitodextrinases Was

The DNA libraries used to obtain the cloned V. fvrnlssii

E. d i transformants were acrsened for B-GlcNlcld-.a m d rh(*n- dextrinaseaotivities e.= follows ( i 4 1 . After Overnight growth on LE- ampicillin plates, transformants were transferred to freah LB-smpicil- lm plates ln grids and grown overnight at 31.C. A replica Of each grid was transferred to Whatman NO 1 filter paper. and the papers Sprayed with a Solution Of WF-GlcNAc (when screening for 8-GIENACida- Se activity) or MUF-(GlCNAC)r (when screening for chitodextrinase actlvity). Each solution was prepared by dissolvinq 9.5 mg of the respective HUT-glycoside in one nl Of N.N-di.ethylfo~..ide. and diluting to 50 nl With 0.1 L1 Tria. pH 7.4. After spraying, the papers were incubated for 15 min at 3l'C. and wain mraved With a solution

"""

of saturated sodium bicarbonate (to enhance fliorissence) . The paper-. were immediately viewed under long wavelength ultra-violet light. Transformants harboring plasmids containing E-GlcNAcidass or Chlto- dextrinase ICtlYitieB were fluorescent. These FDlonieS were DiCksd from the original LB-ampicillin grids, purified by single colbny isolation and screened again for the respective enzymatic activity. A total of 6000 transformants were screened; 3 of tha clones exprasaed E-ClcNAcidase activity and harbored an identlsal plasmid (designated pBB20) containing a 12.5 kb fraqment Of Y. fvrniaaFi DNA. Tan trans.. formants exprsssed chitodextrinase activity and harborsd a plasmid

of the 12.5 kb fragnant of Y. (called pBB22) containing a 5.5 kb fragnant of W DNA. Digeation

ligation into pBRl22 resulted in a plasmid lpBB25) containing a 4.5 kb frDigDlent Of Y. U!XdS%U DNA which expressed full E-GlcNAcidase "+'.,4*"

" DNA in pBB20 with U I and

and E . G Q U cell8 Yere harveeted at mid-exponential growth pha.. . In all of the experiments described below, both th. Y.

Cell Culture8 yere chilled to 0-4°C. centrifuged for 5 mln at 6000 x g. and unless otherwise specified, the cell pellets Yere washed three times with the following buffaro at 0-4'c: y. fvEniaaii with "toluan- rpll with toheniring buffer containlnq 0.5\ NaC1 in place Of 1.58. iEing" buffer (50 nM Hspss,Cl.. pH 1.4, 10 mcI MqC1,. 1.5% NOICI); E .

Hashed pellets were resuspended to the indicated call densities in the above buffers, and the suepaneions Used immediately.

" - -. - -, .


the baseline

Assays for GlcNAcidase and Chitodextrinase- Tyo types of Sub-

oligosaccharides. and synthetic p-nitrophenyl glycosides of GlCNAC and strates were used for assaying the chitinolytic enzymes, chitin

respectively). In each case, preliminary experiments were pereorned (GlcNAc), (for lneasuring 6-GlcNAcidase and chitodextrinase activity

to determine optimum pH, ionic Strength and other parameters. Both cloned enzymes exhibited a pH optimum between 7.4 and 7.8 and neither Of the enzymes displayed a divalent Cation requirement. A 1 1 data

where the activity Vas proportional to the quantity of protein added reported below with cell free extracts were obtained under conditions

to the mixture. and the rate was constant with time of incubation unless otherwise indicated. Controls included complete mixtures incubated for zero nin. mixtures lacking protein, mixtures lacking substrate, or mixtures containing heat-inactivated enzymes. Activities were measured With intact cells, permeabilized cells, crude extroicte, and sub"celIu1ar fractions.

was routinely measured by determining the rate of formation pf e; nitrophenol from PN-GlCNAc. Intact or permeabilized y. cells were Suspended in tolueniring buffer at a density Of 5 x IO' cells per 1111. and 100 wl were incubated at 25'C With 100 111 Of 660 a PN-GICNAC in 0.1 I Tris, pH 7.4. Incubations were conducted from 5 to

mixture to quench the reaction. Cells were removed by centrifugation 30 nin at 25'C after which 3 nl of 1 B Tris. pH 11. were added to each

The quantity of p-nitrophenol formed pel min vas calculated using an and the absorbancies of the supernatant fluids determined at 400 nm.

extinction coefficient of 18.300 (I x cm)". This assay was also Used with the E. transformants except that the toluenizing buffer contained less NaC1 (as indicated above).

(a ) 8YdlOly.iS Of Syntl).tic *Ub#trat.m. 8-GlcNAcidase activity

'This method was developed by Dr. Mark Hardy, to whom the authors are deeply indebted. Dr. Hardy also analyzed the incubation mixtures containing these oligosaccharides. ' The structural reasons for this apparent anomaly will be consid-

ered elsewhere (Y. C. Lee, submitted for publication).

Chitodextrinass activity was measured exactly an described for the 6-GlcNAcidase except that the analogue, PN-(GlFNAC),, was used as substrate. In general. longer incubation periods with this

assay. substrate were necessary (30 to 60 lain) than in the 8-GlcNAcidase

strates, PN-GlCNAC and PN-(GlCNAC),, Were quantitated spectrophoto- metrically by continuously nonitorinq the rates of formation of p- nitrophenol at 400 nn. Cuvettes contained 600 "1 of 110 a PN-GlcNAC

PN-(GlCNM). in 0.1 B Tris,CI~ buffer pH 7.4; suitable aliquot.

tored at 25'C. At H 7.4, the extinction coefficient Of p-nitrophenol lusually 5 p l l Of cell extracts were added. and the rmactions m n i -

is 16,200 (E x c m P . Althouqh the synthetic PN-(GlcNAc)* substrata vas Cleaved by the

Ehitodextrinase, the hydrolysis rate vas very slow (both with y. cells and extracts. and With extract. of the E. d trans-

studying the in - chitodextrinase activity since the presence Of formant). Nevertheless. PN-(GlCNAE)z Was -he substrate of Choice for

the 0-GlcNAcidares in p . fvraiaaii calls and extracts precluded the use Of Che chitin oligosaccharides for determinations of chico- dextrinase activity. The substrate specificity of the chitcdextrinase

GQU HB101:pBB22 transformant, which did not contain 8-G1cNAcida.a was determined with the chitin oligosaccharides and extracts of the E.

When possible, initial rates Of hydrolysis of the synthetic sub-

. . .

activity. (bl I I Y d r O l Y s i s of chitin olIaosacch.rid.. by th. E-OloYloid.....

asaayad for 8-GlsNAcidasie activity using quantities of each fraction Both ~ y i o s ~ l i s ind membrana fractions prepared f;om Y.'lurni..ii were

equivalent to 1.5 x 10' cells. Assay mixtures contained either 0.5 &I or 3.0 mp1 Of one Of the chitin oligosacchoiridee in SO &I sodium phosphate buffer, pH 7.4. After 5, 10, and 15 min at 25.C. aliquot. (200 "1) were removed from the tubas. hasted for 2 min at 10O'C. centrifuged for 5 min at 12,000 x g. and the resultinq supematsnt~

Controls contained oligosaccharides without protein. Thtl specificity analyzed for the appearance Of GlCNAE by the Korgan-Elson preaedure.

of the cloned 8-GlcNAsidaee was determined similarly. Cell extract. Of E. GQU HBlOl:pBB20 ( 1 5 yg protein) Were added to chitin oliqosac-

pl of 50 sodium phosphate buffer. pH 7.4, incubated at 25.C tram 0 charide BOlUtions ranqing in concentration from 0.2 &I to 10 &I in 250

to 120 min (the time Of incubation depended on the concentration of oligosaccharide used), and analyzed as described above.

The roeaificitv Of +.he chitcdextrinase vas determined With the cloned (c) nydrolysim of oli.pmaccmrid.. by ta. 01on.d oaitod.xtrin....

gene b;oduct~u;ing extracts Of E. GQ.U H0101:pBB22. The chitin oligo-

phosphate buffer, pH 7.4, at concentrations of 0.2, O.k, 1.0 and 2.0 saccharides, (GICNAE)n, n = 2 to 6 , were dissolved in 50 &I sodium

mH in volumes of 650 ul. Prior to addition of the Eel1 extract. a 100

and~from 0 to 8 h for mixtures containing.ths pmta- and hexasaccharide. The aliquot6 were immediately centrifuged at 12,000 X 9 for 5 rain to remove the bacteria, and the supernatant fluids frozen for later analysis by one of the two HPLC procedures described above.

membranes. GlcNAa-6-phosphate deacetylaaa and C1CNHl-6-phoQphlte deaminase activities were determined as described (3, 31. 32).

using acetate kinase (45). The activity Of the ATP-dependent GlcNAc Deacetylase activity was measured by determining acetate formation

kinase was measured as follows. Reaction mixtures (50 p1 volume), contained 10 mp1 ATP, KgCI,, and KF, DTT at 0.5 &I, Tris-C1 (50 &I, pH 8 . 0 ) . (LLC]-GICNAC ( l W , 250 dpnlnmole) and various a.OUnt(l Of Y.

measured as described ( 3 ) . NO ATP-dependent kinase activity Was detected in the membranes.

Assays Cor GlcNAc Catabolic En~ynes---II"'~ activity in washed

soluble extract. Formation of GlfNAc-6-phODphate was

THE OF CHEMISTRY Vol. 266. 36, 25, pp. 24276-24286,1991 ... · 24276 . Catabolism of Chitin Oligosaccharides 24277 menclature of the hydrolyses. Insofar as we are aware, all published

Documents