Isolation and partial characterization of the exocellular polysaccharides of Penicillium charlesii

ARCHIVES

OF

Isolation

BIOCHEMISTRY

and

AND

BIOPHYSICS

Partial

Heterogeneity

J. F. PRESTON of

Biochemistry, Received

of the

and

Composition

Exocellular

Polysac-

of High

Molecular

Polysaccharides’

III,2 ERLINDA

LAPIS,

College of Biological St. Paul, Minnesota May

Exocellular

charlesii

of Penicillium

in Size Weight

Department

(1969)

Characterization

charides III.

324-334

134,

14, 1969;

accepted

J. E. GANDER3

AND

Sciences, 55101 July

of Minnesota,

University 29, 1969

Extracellular polysaccharides of the fungus Penicillium charlesii G. Smith, fractionated on DEAE-Sephadex and Sephadex G-50, were further resolved on Sephadex G-100. Specific enzyme assays for glucose, galactose, and mannose and analysis using gas-liquid chromatography revealed that the most highly resolved preparations contained only these three hexoses in significant amounts and were continuously heterogeneous with respect to size and composition. The highest molecular weight portion was shown to be a phosphopolysaccharide. The ratio of anhydrohexose:P decreased as the particle size decreased. Sedimentation-rate studies supported gel-filtration experiments in demonstrating gross differences in size within a single fraction isolated from DEAESephadex. --

Fractionation of extracellular polysaccharides of the fungus Penicillium churl&i on DEAE-Sephadex has been described (1). It has been observed that the major polysaccharide fractions so obtained contain phosphorus which comigrates with large molecular weight carbohydrate fractions on both DEAE Sephadex and Sephadex gels. Since Russel’s viper venom phosphodiesterase and alkaline phosphatase (2) released no phosphate from the polymers while acid hydrolysis did, it was assumed that the phosphorus was attached to the polymers through covalent bonds.

Filtration on Sephadex G-50 of a major fraction (VII-l) isolated from DEAESephadex revealed that much of this fraction was composed of polymers with molecular weights in excess of 10,000. Analysis of mild acid hydrolyzates for free galactose and optical rotation studies indicated that these large molecular weight polymers contained galactose in the furanose form (1). Further studies presented in this paper seek to characterize the high molecular weight polysaccharide fraction with respect to size and carbohydrate composition. MATERIALS

1 This investigation was supported by Grants AM 03713 and AM 09257 from the U. S. Public Health Service and by the University of Minnesota Agricultural Experiment Station, Scientific Journal Series No. 6838, Agricultural Experiment Station, University of Minnesota, St. Paul, Minnesota 55101. 2 Present address: Department of Microbiology, Yale University, New Haven, Connecticut 06510. * Recipient of a Research Career Development Award, U. S. Public Health Service.

AND

METHODS

Growth and cultivation of organism. Stock cultures were maintained as previously described (1). Cultivation of the organism and preparation of the medium for chromatography has also been described for cold medium (1) as well as that labeled with Na2H32P04 (2). Column chromatography. Growth medium was first fractionated in large batches without prior concentration on 61 X 4-cm columns of DEAESephadex A-25, medium porosity. The resin was 324

HETEROGENEITY

OF

prepared in the free base form, packed, and eluted with 3-5 vol of 0.01 M K2B407 prior to addition of the sample. About 1 liter of medium was made 0.01 M with respect to KsB40~ by adding an appropriate amount of 0.25 M KzB40~ and fractionated on the DEAE-Sephadex columns using a routine similar to that used for small columns (1), but using larger volumes of eluting buffer. The column was eluted successively with 7.5 liters 0.01 M K2B407 , 4.0 liters of 0.05 M KzB40, , 1.0 liter distilled deionized water, 1.0 liter 0.01 N HCl:0.03 N LiCl, 4-5 liters 0.01 N HC1:0.06 N LiCl, 5-6 liters 0.05 N HCl, and 2 liters 0.5 N HCl. Fraction VII-l, the major carbohydrate fraction, eluted with 0.01 N HC1:0.06 N LiCl, fraction VII-2 with 0.05 N HCl, and fraction VII-3 with 0.5 N HCl. Each fraction was individually neutralized to pH 7 with either concentrated HCl or KOH and concentrated in vucuo at 30-40’ to a small volume for chromatography on Sephadex gels. Fraction VII-l was fractionated on Sephadex G-50 as described (1). Fractions VII-2 and VII-3 were treated in a similar way. The resulting high molecular weight fractions, VII-l G501, VII-2 G501, and VII-3 G501, were concentrated and desalted on Sephadex G-25 as previously described (1). Filtration studies were carried out on columns 2.8 cm in diameter packed to a height of 3 cm with Sephadex G-25 coarse uniform bead grade followed by Sephadex G-100 to a height of 77.5 cm. Both gels were preequilibrated with 0.1 M NaCl prior to packing and samples were eluted with 0.1 M NaCl. The volume of fractions collected varied from one experiment to the next, but the flow rates were generally 0.6 ml per min. Void volumes were determined for each gel column with Blue Dextran (Pharmacia .Fine Chemicals, Inc.). Sephadex G-100 columns were also calibrated for the elution volume of bovine serum albumin and Dextran-40 (Pharmacia Fine Chemicals, Inc.). Sedimentation velocity determination. Sedimentation rates were measured on a Spinco Model E analytical ultracentrifuge in an An-D rotor and a standard cell of 12-mm light path. A concentrated sample was desalted on a Sephadex G-25 column and made 0.2-0.5% with respect to total carbohydrate and 0.1 M with respect to NaCl. Centrifugation was carried out in 0.1 M NaCl at 59,730 rpm at 20”. Photographs were taken at, regular intervals of schlieren patterns at a bar angle of 50”, and migrations of schlieren peaks were determined with a micrometer. Analytical methods. Samples were hydrolyzed in 0.01 N HzS04 for 90 min or in 1.0 N HzSOa for 5 hr in sealed tubes and were neutralized with BaCOa as described previously (1). All enzyme assays were done on samples which had been passed through Dow 5OK+.

POLYSACCHARIDES

325

Hydrolysis in 1.0 N NaOH was carried out as described by Archibald et al. (3). Radioactivity as 32P was estimated with a gasflow, thin-windowed Geiger detector (Nuclear Chicago Model D 47) after plating and drying liquid samples on aluminum planchets 1.25 in. in diameter. Total carbohydrate was determined by the method of Dubois et al. (4). Glucose was used as a standard with corrections for the different extinction coefficients of mannose and galactose when the relative concentrations of these could be determined independently (1). Galactose and glucose were determined with galactose oxidase and Glucostat respectively, purchased from Worthington Biochemical Corporation. Details of the assays are described in the company’s technical bulletin. Mannose was quantitatively measured by reducing with sodium borohydride, converting the mannitol to fructose with mannitol dehydrogenase, and measuring the fructose calorimetrically. Mannitol dehydrogenase was partially purified from Lactobacillus brevis strain A.T.C.C. 367 up to fraction “ammonium sulfate II” as described by Martinez et al. (5). The enzyme preparation was stored with little loss in activity for 4 months at -16” in 2 X 1O-6 M mercaptoethanol and 0.05 M sodium phosphate buffered at pH 6.5. Mannose was converted to mannitol by treating 0.20 ml of sample with 0.10 ml of 0.05 M NaBH4 in an ice bath and incubating the mixture at room temperature for 10 min. Any remaining NaBH4 was destroyed by the addition of 0.05 ml of 0.4 N HCl. Each sample was then neutralized with 0.05 ml of 0.38 N KOH. Mannitol was determined wit’h the mannitol dehydrogenase as described by Martinez et al. (5) with minor modifications. To each neutralized sample was added 0.2 ml of the following mixture: 0.015 M sodium pyruvate, 0.04 M NAD, 0.02 mg per ml lactate dehydrogenase (Crystalline muscle enzyme from C. P. Boehringer and Sons), 0.04 M Tris-HCl buffer, pH 8.5, and 5 units.4 per ml of mannitol dehydrogenase. The samples were allowed to stand at room temperature for 130 min after which time they were assayed for fructose by the method of Roe (6) as modified by Davis and Gander (7). A standard curve relating initial mannose concentration to absorbancy values obtained from the assay of fructose was linear over a range of 0.020.10 rmoles of mannose. Glucose, galactose, and tetraborate had no effect on the quantitative determination of mannose by this method. Qualitative analysis of polysaccharide composition was performed routinely by paper chromatog4 One unit of enzyme al. (5) is that quantity change in optical density

as defined by Martinez et required to produce a of 1.0 per min at 340 rnp.

326

PRESTON,

LAPIS.

raphy of acid hydrolyzates and reducing sugars detected with alkaline silver nitrate. Details of the conditions used were the same as described previously (1). Gas-liquid chromatography of carbohydrates. The polymeric material was hydrolyzed in 1 N HCl as previously described (1) and the hydrolyzate was evaporated to dryness in l-ml glass-stoppered volumetric flasks several times to remove the HCl. The residue was dissolved in pyridine and held at room temperature for 30 min. The pyridine was removed under reduced pressure and the residue was dissolved in the silylating mixture (pyridine: hexamethyldisilazane : trimethylchlorosilane: 9: 3 : 1, v/v/v) and allowed to react as described by Sweeley et al. (8). After incubation for 1 hr, the solvent was removed and the residue was dissolved in a minimum of hexane. One-microliter samples were injected into a F and M Model 609 gas chromatograph with a 5-ft copper column of 3% SE-52 and a hydrogen-flame ionization detector. The inlet pressure of He and Hz were 10 and 8 psi respectively. The temperature was programed from 80” to 170’ at 2.3” min-1. RESULTS

Fractionation of VII-l on Xephadex G-100. When polysaccharide fraction VII-l obtained from a DEAE-Sephadex A-25 column was fractionated on Sephadex G-50, a major fraction was eluted in the void volume (1). This fraction was concentrated to a volume of 4.0 ml in vacua, desalted on Sephadex G-25, again concentrated and fractionated

100

AND

on a 480-ml column of Sephadex G-100 preequilibrated with 0.1 M NaCl as described in the Methods section. A chromatogram describing the elution of total carbohydrate is given in Fig. 1. The rather diffuse peak of carbohydrate was divided into three fractions referred to as GlOOI, GlOOII, and GlOOIII respectively. A fourth fraction designated as GlOOIV corresponds to a position of elution expected for small molecules such as monosaccharides. The symbols Vo and VssA refer to the void volume and the volume at which bovine serum albumin eluted respectively. Each fraction was concentrated or diluted with water to give a final carbohydrate concentration appropriate for the enzymatic assay of galactose, glucose, and mannose.Portions of each fraction were hydrolyzed in 0.01 N or 1.0 N H.SO+ deionized, and again concentrated or diluted to give appropriate carbohydrate concentrations. The enzymatic analyses for galactose, glucose, and mannose were carried out on unhydrolyzed and hydrolyzed samples as indicated in the Materials and Methods section. The results of these analyses are given in Table I. Fractionation of VII-i G601-32Pon Sephadex G-100. When Penicillium charlesii was grown on a medium supplemented with Na2H32P0 all major polysaccharide fractions obtainzd from chromatography on DEAF-

300

200 EFFLUENT,

GANDER

400

500

ml

FIQ. 1. Chromatography of VII-l G5OI on Sephadex G-100. A 5.0.ml portion of concentrated VII-l G501, which had been desalted on Sephadex G-25, was added to a 481~ml column of Sephadex G-100 preequilibrated with 0.1 M NaCl and eluted with the same. Fractions of about 4 ml were collected, measured accurately for volume, and assayed for total carbohydrate. The arrow from VO indicates the position at which Blue Dextran was eluted; the arrow from Vns~ indicates the position at which bovine serum albumin eluted. Details of column dimensions, packing procedures, and assays are described m the text.

HETEROGENEITY

TABLE ENZYMATIC

Treatment

Unhydrolyzed 0.01 N HzS04 1.0 N HzSO,

o The limits

I

DETERMINATIONS OF GLUCOSE, GALACTOSE, AND SACCHARIDE FRACTIONS AFTER DIFFERENT DEGREES Fraction

GlOOI@ GlOOII GlOOIII GlOOP GlOOII GlOOIII GlOOI” GlOOII GlOOIII to the fractions

327

OF POLYSACCHARIDES

MANNOSE IN VII-l OF HYDROLYSIS

Mannose, pmoles Hexose, pm&s

Galactose, imKJes Hexose, pm&s

0.014 0.011 0.011 0.046 0.073 0.023 0.307 0.407 0.466

0.00 0.00 0.00 0.51 0.29 0.21 0.53 0.46 0.27

POLY-

Glucose, jmoles Hexose, pm&s

0.005 0.008 0.004 0.034 0.045 0.040 0.140 0.176 0.344

listed here are defined in Fig. 2 and are described in the text.

0.25 b.% 16.0-

&X-JO EFFLUENT,

ml

FIG. 2. Chromatography of VII-l G501-3”P on Sephadex G-100. A 5.0-ml was eluted on a 481-ml column of Sephadex under identical conditions to gram given in Fig. 1. Open circles connected by solid lines represent values circles connected by broken lines represent values for phosphorus, measured

Sephadex A-25 were labeled. Further fractionation of fraction VII-l on Sephadex G-50 showed most of the 32Peluted in the void volume along with the major portion of VII-l (2). When this polymer fraction was concentrated, desalted, and again concentrated and further fractionated on a 480-ml column of Sephadex G-100 as was used for the studies described above, the elution diagram shown in Fig. 2 was obtained. It is evident that different portions of the chromatogram show markedly different total

cocentrated desalted sample those used for the chromatofor total carbohydrate; solid as @P.

carbohydrate to =P ratios. Most apparent is the higher carbohydrate: 32P ratio of the highest molecular weight fraction (analogous to VII-l G50I GlOOI of Fig. 1) compared to the smaller molecular weight fractions (GlOOII and GlOOIII). It is also apparent that GlOOII has the lowest carbohydrate: 32Pratio.

Chromatography of VII-l

G5OI GlOOI 32P

072 Sephadex G-100. The high molecular weight fraction, GlOOI, represented by the effluent obtained between 150 and 170 ml

328

PRESTON,

LAPIS,

AND GANDER

0 0

100 EFFLUENT,

ml

FIG. 3. Chromatography of VII-l G50I G1001-32P on Sephadex G-100. A 5.0.ml sample of concentrated and desalted VII-1 G501 G1001-32P was chromatographed on the same 481-n-J column of Sephadex G-100 used for the chromatogram in Fig. 2. Open circles joined by solid lines represent values for total carbohydrate; closed circles joined by broken lines represent values for P determined as @P; open triangles joined by solid lines represent values for the ratio of anhydrohexose residues to P.

shown in Fig. 2 was concentrated, desalted, and again concentrated and chromatographed on the same column from which it was initially obtained. The reults of total carbohydrate and 32Panalysis are presented in Fig. 3. The polysaccharide fraction eluted in a volume identical to that in which it was initially obtained, indicating that little or no degradation occurred during a second gel filtration. The top of the figure shows a plot of the carbohydrate-to-phosphorus ratio as a function of elution volume and demonstrates that the ratio constantly changed with the position of elution. Fractionation of partial acid and basehydrolyzates of VII-lG501 GlOOII on Sephadez G-25. Portions of fraction VII-l G501 GlOOII were hydrolyzed in either 0.01 N H$Od at 100” for 90 min or in 1.0 N NaOH at 100” for 3 hr and were neutralized and concentrated as described in the Methods section. When concentrates of these hydrolyzates were fractionated on Sephadex G-25, the results shown in Fig. 4 were obtained. It is evident that hydrolysis in mild acid releasedabout 30 % of the total carbohydrate (Pig. 4a), most of which eluted in a position

expected for hexoses.Paper chromatographic and GLC analyses (not shown) of a complete acid hydrolyzate of the limit polymer resulting from 0.01 N acid hydrolysis which eluted in the void volume of Sephadex G-25 revealed that it contained mannose and glucose but no galactose. In contrast, hydrolysis in 1.0 N NaOH produced three fractions upon Sephadex G-25 filtration (Fig. 4b). Fraction VII-l G501 GlOOII 1.0 N NaOH G251 (tubes 21-32 in Fig. 4b) represents the “limit” fraction after base hydrolysis; it comprised about 90% of the total carbohydrate and contained mannose and galactose as the only hexoses detected after hydrolysis in 1.0 N acid followed by paper chromatography and GLC analyses (not shown). Fraction VII-l G501 GlOOII 1.0 N NaOH G2511 (tubes 3642 in Fig. 4b) comprised about 10% of the total carbohydrate and released glucose as the only significant hexose component upon complete acid hydrolysis. Less than 2% was eluted from the G25 column in the position of free hexoses. Sedimentation velocity studies on VII-l G50I GlOOII, VII-l G5OI GiOOIII, VII-2

HETEROGENEITY

OF

“0

0

“s 1

0

lo

20

30 FRACTION

40

50

60

;I3 x

70

NUMBER

FIG. 4. a. Chromatography of VII-l G501 GlOOII 0.01 N HzSOI hydrolyzate on Sephadex G-25. An 8.3-ml portion of a concentrated, neutralized 0.01 N H&O* hydrolyzate of fraction VII-l G501 GlOOII was fractionated on a 750~ml column (3 X 120 cm) of Sephadex G-50 preequilibrated with 0.1 M NaCl. Fractions of 12 ml were collected and assayed for total carbohydrate. The arrow from VO defines the void volume determined with Blue Dextran; the arrow from Vo defines the elution volume of free glucose. b. Chromatography of VII-l G501 GM011 1.0 N NaOH hydrolyzate on Sephadex G-25. A 7-ml portion of a concentrated, neutralized 1.0 N NaOH hydrolyzate of fraction VII-l G501 GM011 was fractionated on a 750-ml column (3 X 120 cm) of Sephadex-650 preequilibrated with 0.1 M NaCl. Fractions of 12 ml were collected and assayed for total carbohydrate. VO and Vo are defined in the same way they were in Fig. 4a above.

GbOI,

and VII-S G&U. Fractions correto GlOOI, GlOOII, and GlOOIII, as defined in Fig. 1, were accumulated upon fractionating several liters of media. Sufficient amounts of VII-l, GlOOII, and GlOOIII, along with VII-2 G501 and VII-3 G501, were thus obt,ained to carry out sedimentation velocity studies as described in the methods section. At the same time, mild sponding

329

POLYSACCHARIDES

acid and base hydrolyzates of VII-l G501 GlOOII were analyzed. Sedimentation plots of each of the fractions are shown in Fig. 5. Typical schlieren patterns are shown in Fig. 6. Photographs showing the rate of migration of acid- or alkali-treated fraction VII-l G501, VII-2 G501, and VII-3 G501 compared to fraction VII-l G501 after the beginning of the run. Pictures were taken at five times for each fraction to permit calculations of sedimentation rates. Figure 5 shows that Fractions VII-l, VII-2, and VII-3 are all relatively high molecular weight polymers migrating at a uniform rate in the force field of the ultracentrifuge. Figure 6 shows reasonable weight average molecular weight homogeneity. Dextran 40 and Dextran 80 fractions obtained from Pharmacia Fine Chemicals, Inc. were analyzed on the ultracentrifuge for comparison with the carbohydrate fractions collected. Again pictures were taken at different times during the run to permit calculation of sedimentation rates. These rates are presented in Table II as S values along with those determined for the various fractions and partial hydrolyzates. Data sheets provided by the manufacturer indicate that Dextran 40 has a AT5 of 39,800 and a i%IIn6of 25,600 while Dextran 80 has a %I, of 72,000 and a M, of 40,500. gi‘, values were determined by light scattering while ft, values were determined by end-group analysis. Gas-liquid chromatographic analyses. Qualitative gas-liquid chromatography (GLC) was conducted on a SE-52 column temperature programed between 80 and 170”. Samples of fractions VII-lG50-I, VII-2G50-I, and VII-3G50-I were hydrolyzed in 1 N HzS04 and silylated as described in the Methods section. Figure 7a, 7b, and 7c show the elution patterns obtained. In fraction VII-l (Fig. 7a) major substances were eluted at 5, 9, 27.5, 28.5, 29.5, 31, and 35.5 min. The substance eluting at 5 min had a retention time similar to glycerol. The substances eluting with retention times of 27.5 to 35.5 min had retention times identical to the various isomers of mannose, 6 Weight 6 Number

average average

molecular molecular

weight. weight.

330

PRESTON,

LAPIS,

AND

GANDER

18300

i I.8200 I z E 1.8100-

I8000

0

5

IO dt x 10-10

15

FIG. 5. Schlieren plots of fractions VII-l (u), VII-2 (0), VII-3 (A), VII-l 3 hr at 100” (A), and VII-l treated with 0.01 N HzSOc at 100” for 90 min (0).

galactose and glucose. Small quantities of other substances had retention times of 9 and approximately 20 min. The elution pattern of fractions VII-2 and VII-3, shown in Fig. 7b and 7c, respectively, also contained the same hexoses but in much smaller quantity relative to the quantity of nonhexose material. In addition, material was eluted with a retention time of 13.5 min, a retention time on this column similar to threitol. These fractions also contained large quantities of material with a retention time of glycerol and only slightly smaller quantity of material with a retention time of 2.5 min. Gas-liquid chromatography (not shown) was also carried out on the polymeric material obtained after treatment with 0.01 N HzS04 or 1.0 N NaOH. All of the galactose was removed by the acid treatment and only glucose was removed by treatment with alkali.

I 20 treated with 1 N KOH for

Analysis of complete hydrolyzates of fraction VII-l GlOO-I, -11, and -111 showed that the sum of the individual hexoses equaled the total hexosesdetermined as total carbohydrate, precluding the existence of significant amounts of other hexoses or pentoses. That these represent the major aldosespresent is shown by the GLC chromatographic analysis of VII-I G50-I ‘(Fig. 7a). The specificity of the mannitol dehydrogenase (4), as well as the sensitivity of the overall mannose assay, should make it useful in the routine quantitative assay of mannose in the presence of other sugars. A similar assay has already proved useful in the qualitative analysis of GDP mannuronic acid (9). The quantitative determinations of mannose, glucose, and galactose in hydrolysates of polysaccharides fractionated on Sephadex G-100 demonstrates basic differences in composition of various fractions tested and suggests a continuous heteroDISCUSSION geneity throughout these fractions. The The enzymatic assay of mannose, glucose, higher molecular weight fraction GUI01 is and galactose in hydrolyzates containing all richest in galactose and poorest in glucose three sugars permitted the unequivocal de- while the opposite is true for GlOOIII (note termination of the relative amounts of each 1.0 N acid hydrolyzates in Table I). Manin heteropolysaccharides containing all three. nose, to a lesser extent than glucose, is

HETEROGENEITY

OF

POLYSACCHARIDES

FIG. 6. Sedimentation velocity studies of fractions VII-l G501 GlOOII, VII-2 G501, and VII-3 G501. Solutio sns 0.5°]0 with respect to anhydrohexose except as noted and 0.1 M NaCl were centrifuge1 d on a Spinco Model E. Pictures of schlieren patterns appearing about 40 min after reaching fldl speed are given above. Details of procedure are given in the text,. a. Schlieren patt,ern for VII-l G501 GlOOII in thl e leading cell and VII-l G501 GlOOII after treatment with 0.01 N H#OI 90 min, 100”~ followed by chror natography IIn Sephadex G-50 (as shown in Fig. 5a). b. Schlieren pattern for VII-l G501 GlOOII after treatat 100” for 3 hr is shown in trailing peak as compared to the same material p rior to ment w ith 1 N NaOH pattern for VII-2 G501. The concentration was O.ZfyO with respect to pc jlysachydrol; qsis. c. Schlieren charide :. d. Pchlieren pattern for VII-3 G501.

332

PRESTON,

LAPIS,

AND

GAIL’DER TABLE

SEDIMENTATION

AND VII-l G501 THOSE OF DEXTRAK

II

OF GlOOIII

RATES

VII-l

G501

COMPARED 40 AND DEXTRAN

GlOOII WITH

80

Carbohydrate -

Fraction

I

I hIan1Ilose -:-

VII-l

--I.

c

I

M Gal M Gal Gal Glu

GlU

G501 GlOOII 0.01 N H,SO, G25I 1 N NaOH G25I VII-l G501 GlOOIII VII-2 G5OI VII-3 G5OI Dextran 40 Dextran 80

2.1 3.7

39,sOc I 72,00C I

provided See text

by for

-

-

a From Chemicals,

8

I6 RETENTION

24 TIME (min)

32

FIG. 7. Linear temperature programed separation of trimethysilyl derivatives of carbohydrates from P. charlesii polysaccharides. The column used was a 3y0 SE-52, programed from 80 to 170” at 2.3 mia1. Figure 7a, 7b, and 7c represents the silyl derivatives in VII-l GM-I, VII-2 G50-I, and VII-3 G50-I, respectively.

present in a higher concentration the smaller the size of the polysaccharide fraction studied. The relevance of these relationships to biologically functional polymer structure must await more detailed structural analysis now underway. A further testimony to the constantly changing structure of any of the polysaccharide fractions studied is presented in Fig. 2. The continuously changing anhydrohexose: P ratio within fraction VII-l G501 demonstrates a continuous heterogeneity within the most discrete carbohydrate fraction defined in Fig. 1. It is evident from Table I and the GLC analysis that most of the galactose, as high as 96% in GlOOI, is labile to 0.01 N acid, indicating at least this much may be present in the furanose form. The resistance of the

data Inc.

-

Pharmacia details.

Fine

unhydrolyzed fractions to galactose oxidase rules out any possibility other than galactopyranose residues involving the sixth carbon in linkages or substitution, or galactofuranose residues. The ease with which mild acid treatment releases free galactose strongly favors the possibility of galactofuranose residues. Optical rotation studies before and after hydrolysis on the VII-l G501 fraction have already supported this possibility (1). It is not possible to estimate accurately the molecular weights of the various fractions on the basis of evidence presented here. While GlOOI eluted in a position between the void volume and the position at which bovine serum albumin was eluted, the fact that a spherical protein of a given molecular weight might be retarded to a greater extent than a less compact polysaccharide having the same molecular weight presents the use of a protein standard to estimate the size of polysaccharides. On the other hand, polysaccharides may bind weakly with the carbohydrate matrix of the Sephadex gel itself and give erroneously low values for molecular weights when compared to protein standards. When Dextran 40 was used as a standard, a diffuse peak was observed which

HETEROGENEITY

OF

was unsymmetrical and showed extensive trailing. It was therefore deemed an unsuitable standard. Based upon the position of the elution of the various fractions from Sephadex it can only be pointed out that all were excluded from Sephadex G-50; and therefore assuming these fractions had properties similar to dextrans with respect to filtration on Sephadex gels, all had molecular weights in excess of 10,000. By the same rationale, none were excluded by Sephadex G-100, and therefore all were of molecular weights less than 100,000. Sedimentation studies by themselves suggest each fraction defined by the elution diagram in Fig. 1 is relatively homogeneous in size although determination of carbohydrate content and carbohydrate: P ratios indicates the polymers are heterogeneous in composition. Comparison with Dextran standards indicates that the li& of GlOOII is about 40,000 while the M, of GlOOIII is considerably less. Since all polysaccharide fractions were eluted in the void volume of Sephadex G-50 and therefore had molecular weights in excess of 10,000, it is probable that M, of GlOOIII is about 20,000. Fraction GlOOI would be expected to have a M, between 50,000 and 100,000. The maximum ratio of hexose:P recorded in Fig. 1 is 180 anhydrohexose units per unit of phosphorus. If all molecular species eluted in GlOOI were phosphopolysaccharides, they must have had a minimal molecular weight of at least 30,000. While the possibility exists that fraction GlOOI contained some phosphopolysaccharides and some neutral polysaccharides, the fact that all fractions studied were retained on DEAE-Sephadex and even the smallest contained phosphorus mitigates against this possibility. It should, however, be noted that if galactocarolose and mannocarolose as characterized by Haworth et nl. (10) (11) were derived from fraction VII-l G501 as noted earlier (l), they must in fact be part of molecules having molecular weights of the order of 50,000. It is probable that the phosphorus present is an integral part of the heteropolysaccharide fraction containing glucose, mannose, and galactofuranose. The sedimentation rates and gel elution

POLYSACCHARIDES

333

patterns (not shown) for fractions VII-2 G501 and VII-3 G501 indicate these fractions contain high molecular weight polysaccharides in the molecular weight range of 20,000 to 50,000. The GLC analysis of VII-2 and VII-3 fractions when compared to VII-l showed the same basic carbohydrate composition. However, fractions VII-2 and VII-3 contained relatively larger fractions of nonhexose components. Material with the elution time of glycerol appeared as a major constituent in all three polymer fractions. However, glycerol has not been positively identified as a component of the polymer. It has been noted that the “limit” polymer obtained by hydrolysis in 0.01 N H&S04 contained no galactose detectable by paper chromatography analysis after hydrolysis in 1.0 N acid. A complete acid hydrolyzate of the intact polymer fraction VII-l G501 GlOOII (Table I) showed it contained 46% galactose, all of which may be assumed to occur in the furanose form based on previous studies of VII-l G501 (1) and GLC analysis and the fact that no galactose was detected in the intact polymer with galactose oxidase. If no other major structural alterations occurred other than the release of free galactose and galactofuranose-containing oligosaccharides, it would be expected that 46% of the average molecule contained within polysaccharide fraction VII-l G501 GlOOII is represented by a section composed exclusively of galactofuranose residues while the remaining 54% is represented by a polymer containing mannose and glucose. This conclusion is supported by the results of the sedimentation velocity studies presented in Table II. This table shows a 48 % decrease occurred in s value upon 0.01 N HzS04 hydrolysis, which agrees well with the fact that the intact polymer contained 46% galactose. In addition, this treatment also released about one-third of the phosphorus. Hydrolysis of VII-l G501 GlOOII in 1.0 N NaOH resulted in the formation of a “limit” polymer devoid of glucose with an s value 11% lower than the intact polymer (Table II). Since composition analysis showed that the intact polymer contained

334

PRESTON,

LAPIS,

17.6 % glucose (Table I), it is evident that s values (uncorrected for concentration dependence) before and after base hydrolysis are not directly proportional to molecular weight. Hydrolysis in 1.0 N NaOH resulted in the release of glucose and glucose-containing oligosaccharides with no destruction of the “limit” polymer comprised of mannose and galactose residues. In contrast to 0.01 N acid hydrolysis, treatment with 1.0 N NaOH did not release phosphorus. Consistent with present data is the general structure given below. The mannan, presumably containing (l-+2). Glucan-Mannan-Galactan and (l-+6) linkages as found in mannocarolose (11) exists as part of a central core since it is found in the “limit” polymer after both partial acid and base hydrolysis. The galactan, comprised of galactofuranose, presumably joined by /3 (1 + 5) linkages as found in galactocarolose (12) may be joined to the mannan core either as a single large section or as two or more smaller sections. Galactose cannot be joined to glucose residues since all of the glucose but none of the galactose is released from the polymer upon 1 N NaOH hydrolysis. As in the case of galactose, glucose may be present as one or more sections joined to the mannan core, but no glucose residues can be joined to galactose residues since mild acid hydrolysis, which released all of the galactose from the polymer, released none of the glucose. At present, little is known of the glucosidic linkages. Treatment of VII-l G.501 GlOOII with HI04 followed by NaBH4 reduction and complete acid hydrolysis did not release any free sugars (13), precluding the existence of (1 + 3) linkages. More detailed structural studies are current.ly underway to determine the nature of those linkages and to establish that the glycosidic linkages in the mannan and galactan sections correspond to those found in mannocarolose and galactocarolose, respectively.

AND

GANDER

The position of attachment of phosphorus is at present unknown. However, the relative acid lability of about one-third of the phosphorus and the relative acid stability of another third surely suggests that phosphorus is attached at more than one position in the polymer. Further studies are under way to assign a precise structural role to the phosphate groups. The limited GLC data suggest the heteropolysaccharide may contain additional components. Fractions VII-2 and VII-3 appear to contain a relatively small percentage of hexose and relatively large percentage of other material, glycerol possibly being one of the major substances. The polymers are currently being investigated to determine if they contain any amino acids or lipid-like constituents. REFERENCES 1. PRESTON, J. F., AND GANDER, J. E., Arch. Biochem. Biophys. 134, 594 (1968). 2. PRESTON, J. F., LAPIS, E., GANDER, J. E., AND WESTERHOUSE, S., Arch. Biochem. Biophys. 134, 316 (1969). 3. ARCHIBALD, A. R., BADDILEY, J., AND BuCHANAN, J. G., Biochem. J. 81, 124 (1951). 4. DUBOIS, M., GILLES, K. A., HAMILTON, J. K., REBERS, P. A., AND SMITH, F., Anal. Chem. 28, 350 (1956). 5. MARTINEZ, G., BARKER, H. A., AND HORECKER, B. L., J. Biol. Chem. 238, 1598 (1963). 6. ROE, H. H., J. Biol. Chem. 107, 15 (1934). 7. DAVIS, J. S., AND GANDER, J. E., Anal. Biothem. 19, 72 (1967). 8. SWEELEY, C. C., BENTLEY, R., MAKITA, M., AND WELLS, W. W., J. Am. Chem. Sot. 83, 2497 (1963). 9. LIN, T., AND HASSID, W. Z., J. Biol. Chem. 241, 3283 (1966). 10. SOHACHMAN, H. K., “Ultracentrifugation in Biochemistry,” p. 216. Academic Press, New York (1959). 11. HOUGH, L., AND PERRY, M. B., J. Chem. Sot. (London) 1962, 2801. 12. HAWORTH, W. N., RAISTRICK, H., AND STACEY, M., Biochem. J. 31, 640 (1937). 13. GANDER, J. E., AND LAPIS, E., unpublished results.