Structure of a Low Molecular Weight Form of Glycogem ... · Structure of a Low Molecular Weight...

T~IE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 247, No. 5, Issue of March 10, pp. 13GlF1367, 1972

Printed in U.S.A.

Structure of a Low Molecular Weight Form of Glycogem

Isolated from the Liver in a Case of Glycogen

Storage Disease*

(Received for publication, July 16, 1971)

RONALD D. EDSTROM

From the Departmtxt of Biochemistry, The Medical School, I/‘niaersity of Minnesota, Miw~eapolis, Hinnesota 55455

SUMMARY

An incompletely branched, low molecular weight form of glycogen has been isolated from liver tissue of a patient with glycogen storage disease. Digestion with pullulanase and @-amylase, oxidation by periodate, Smith degradation, spec- tral studies of the iodine complexes, osmotic pressure measurements, and end group analysis gave results which indicate a molecular weight range of 8,000 to 15,000 with an average chain length of 17 to 18 glucose units per branch point. The substance was found in heart and kidney tissues in addition to liver. In an attempt to evaluate the role of branching enzyme in the disease, it was found that the branching enzyme was not stable in normal frozen liver and disappeared with a half-life of about 5 weeks. Clinical, chemical, and pathological factors in this case were similar to those previously reported in the case of glycogen storage disease type IV.

An unusual form of glycogen has been found in a liver, taken at autopsy, of a patient who had exhibited storage of polysaccharide in several tissues. Osmotic pressure measurements and reducing end group analysis of the polysaccharide indicated an average molecular weight of 8,000 to 15,000. Normal liver glycogen has been reported to have a molecular weight range of 2 to 30 x lo6 (2). The snnlogous plant polysacchnride, amylopectin, has been reported to have a molecular weight of between 10 aud 100 x 106. Chemical determination of molecular weights of glyc*ogelI have not generally been made because the very high molecvhlr weight yields very few reducing terminals and also due to the destruction of the reducing terminal sugar in the usual methods of glycogen extraction by digestion of the tissue with strong base (3). Another important structural charact.eristic of

* Tl~ese studies were supported in part by an American Cancer Society Institutional Research Grant, a University of Minnesota Gradlmte School Grant-in-Aid and United States Public Tlealth Service Grant AM-10127. A portion of this work has bean pub- lished in preliminary form (1).

glycogen is the high degree of branching in the molecule. A typical glycogen sample contains between 10 and 15 glucose units for each cu-(1 ,6) branch point while the number for amylopectin is 20 to 30. The material studied in this work has 17 to 18 glucose units in each chain.

Only two glycogen storage diseases have been previously described in which glycogen with an abnormal structure was isolated from the tissues. Limit dextrinosis (type III), a disease in which the debranching enzyme activities (amylo-l , 6-glucosidase and oligo-1 ,4 + 1,4-glucantransferase) are missing, results in an accumulation of a glycogen having abnormally short outer chains with a higher than usual number of o(-( 1, A)-linked glucosyl units (4). Amylopectinosis (type IV) has been characterized by Brown and Brown (5) as a condition in which the branching enzyme is missing. This enzyme, (Y-1,4-glucan: LY-1,4-glucan B-glycosyltransferase (EC 2.4.1.18), forms branches by transfer of seg- ments of the 01.( 1,4)-linked glucan to position 6 of glucose units within the chain. When branching enzyme is absent, the resultant storage product is a form of glycogen which exhibits many of the properties of amylopectin. These include a shift toward higher wave lengths in the spectrum of the iodine complex and an increased extent of degradation by /3-amylase and phosphorylase; both of these effects are caused by the polymer having fewer a-( 1,6) branch points than normal glycogen. Recently Mercier and Whelan (6) investigated the storage 1)olysarcharide from a case of type IV glycogenosis and concluded that at least some of the branches contained as few as 5 glucose units.

In a prelimimlry report (I), we showed t,hat the polyglucan isolated from the diseased liver had properties consistent with the unusually low molecular weights indicated above and a degree of branching intermediate between glycogen and amylopectin. In t,he present paper, we have confirmed the molecular weight by an independent method and established the average chain length by direct measurement. In addition enzyme studies are presented which were carried out in an attempt to establish the factors which might be responsible for deposition of this material in the liver and other t’issues. The clinical and pathological factors which were found in the present case to be similar t#o those first described by Anderson (7) for amylopectinosis are in a separat,e publication (8). Possible relationships between the material isolated in the present case and the material stored in amylopectinosis (type IV) are discussed.

1360

by guest on August 19, 2018

http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Issue of Alarch 10, 1972 Ii. Il. Eclstronz 1361

ISXPERIMEKTAL PROCEDURE

il,Inferials-I’ancreatic RNase and pancreatic DNase were purchased from Worthington Biochemical Corp., Freehold, N. J., sweet potato /3-amylase from Sigma Chemical Co., St. Louis, MO. Pronase, rabbit liver glycogen, and amylopectin were obtained from Calbiochem, Los Angeles. Pullulanase was a gift of Dr. Thomas Nelson, 13aylor University. Tritum-labeled sodium borohydride (New England Nuclear Corp., Boston) (300 mCi per mmole) was dissolved in 1 N sodium hydroxide for storage and diluted just prior to use. DEhE-cellulose was obtained from Distillation Products, Rochester, N. Y.

Analytical Methods-Sialic acid was measured after hydrolysis in 0.1 N H2S04 for 30 min at 100” by t8he method of Aminoff (9). Hexose was determined by the anthrone method (10). For specific determination of the glucose content, the polysaccharides were hydrolyzed in 2 N HCl for 2 hours at 100”. After neutral- ization of the acid with 2 N NaOH, the glucose was determined with Glucostat (Worthington Biochemical Corp., Freehold, N. J.). Hesosamine was estimated by the method of Elson and Morgan (11) as modified by Roseman and Daff ner (12). After sapon- fication with 1 N KOH at 100” for 30 min, acidification with HCl, and extraction with chloroform, fatty acids were estimated by the method of Duncombe (13). Estimation of 6-deoxyhexose content was carried out by the method of Dische and Shettles (14). Iironic acids were determined by the carbazole method (15). Total phosphate analysis was performed by the method of Ames and Dubin (16). Protein was estimated by the method of Lowry et al. (17). Amino acid content was determined after hydrolysis under nitrogen gas in 6 N HCl at 115” for 21 hours by a Spinco amino acid analyzer. Reducing sugar was assayed by the alkaline ferricyanide procedure (18) and by the method of Conrad et al. (19). In this latter method, the reducing end groups are converted to glucitol by reduction with NaB3H4. The amount of tritium in the glucitol is proportional to the reducing group content. Radioactivity was determined by liquid scintillation counting.

Paper chromatography was used for qualitative analysis of the polysaccharide after hydrolysis. Descending chromatography was used with Whatman No. 1 paper and the following solvent systems: Solvent System A, butane-1-ol-acetic acid-water (5 : 1: 2); Solvent System B, propane-1-ol-water (8 : 2) ; Solvent System C, butane-1-ol-ethanol-Hz0 (10 : 1: 2). Paper electrophoresis was carried out on Whatman 3MM paper with a Savant Flat-plate apparatus. Buffers were: Buffer I, 1 M formic acid; Buffer II, 0.05 M sodium germanate, pH 10 (20); and Buffer III, 0.05 M

triethylammonium acetate, pH 4.5. Sugars were detected by the periodate-benzidine dip of Gordon et ul. (21) or the alkaline silver stain of Trevelyan (22).

Periodate oxidation of the polysaccharides was carried out in the dark. Samples of the polysaccharides (30 mg) were dissolved in 22 ml of water by heating to 100” and after cooling to O”, were treated with 3 ml of 1 M sodium periodate. Samples (3 ml) were withdrawn at various times and 0.2 ml of ethylene glycol was added to destroy the residual periodate. After 30 min in the dark, the formic acid was titrated with 0.01 N NaOI-I to a brom- cresol purple end point (pH 5.2 to 6.8).

Iodine spectra were determined on solutions containing 0.1 mg of polysaccharide (solubilized by heating) in 1 ml of 0.2% KI -t 0.02$& 12 against a blank containing no polysaccharide.

Assays for branching enzyme (ol-1,4-glucan: oc-1,4-glucnn 6-

glycosyltransferase) were performed by the method of Krismau (23). Enzyme was prepared from liver samples by the method of Brown and Brown (5). The decrease in the absorbance of the 1,KI complex with amylopectin at 520 nm is taken as the measure of branching activity after correction for oc-amylase activity.

Molecular weight determinations by osmotic flow measurement, were carried out by the method of Johnson.’ A Technicon autoanalyzer continuous flow dialysis unit was used with the standard Technicon cellophane membrane. Volume changes in the two compartments were measured while pumping the solvent (water) and the solution through the membraneseparated chan- nels at identical rates with a standard Technicon metering pump. With the intake and outlet lines connected to calibrated reser- voirs, to allow continuous flow, it was possible to measure changes in volume of 5 ~1 in a total sample volume of 8 ml.

Isolation of Glycogen-The usual methods of glycogen isolation by digestion of the tissues with strong base could not be used due to the destruction of the reducing end group of the polymers by the alkali (30% KOH) (3). Two alternate procedures were developed to permit isolation of minimally degraded glycogen. The two methods summarized below have been described in detail in an earlier publication (1). Portions of the frozen liver (10 g) were homogenized in cold 0.9% NaCI. The homogenate was immediately added to 9 volumes2 of acetone and the precip- itate washed with acetone and ether, then dried in uacuo at 50” (yield was 2.1 g). Method A depends on removal of protein and nucleic acids by enzymatic means followed by fractionation into the cold water-soluble (Preparation I) and the hot water-soluble (Preparation II) fraction. Starting with 1 g of acetone powder, the average yield of Preparation I was 23 mg (range, 21 to 24) (0.48% wet tissue) and for Preparation II the average was 55 mg (range 50 to 66) (1.2% wet tissue).

Method B utilized the direct extraction of the acetone powder by hot water without the use of enzymes. The procedure yielded an average of 92 mg of Preparation III (1.9% w-et tissue) per g of powder with a range of 63 to 122.

A special preparation of rabbit liver glycogen was made with the procedure for preparation of polymer I, Method I above. After killing by exsanguination, the liver was removed from a rabbit, and immediately frozen in Dry Ice. From 10 g of frozen liver were obtained 2 g of acetone powder which yielded 142 mg of glycogen after three reprecipitations with 50% ethanol (glucose content 95 %).

Smith Degradation-Samples of polysnccharide (10 mg) were dissolved in 5 ml of water at 100”. After cooling to room tern- perature, 5 ml of 0.1 M NaI04 were added. After 18 hours at room temperature, 200 ~1 of ethylene glycol was added and the samples were dialyzed. The oxidized polymer was treated with 1 ml of 1 M NaBaH (specific activity 5 mCi per mmole) for 16 hours at room temperature followed by 3 ml of 1 x HCl. After evaporation, & uucuo, the dry residues were dissolved in 2 ml of 2 N HCl and heated at 100” for 1 hour. The boric acid was removed by distillation as methyl borate and the entire sample was electrophoretically treated in Buffer III for 1 hour at 2500 volts. The radioactive material which remained at the origin was eluted and chromatographed on 3-mm paper in Solvent System C for 24 hours. The strips were scanned for radioactivity and the

1 Personal communication from Dr. J. Johnson, Department of Physiology, University of Minnesota.

2 An error iu the preliminary report indicated 19 volumes of acetone (1).


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

1362

2.3

Low Molecular Weight Glycogen in Glycogen Storage Disease Vol. 247, No. 5

I .o

.9

.8

Wavelength ( n m )

FIG. 1. Absorption spectra of iodine polysaccharide complexes: Amylopectin, 0 ; rabbit liver glycogen, l ; Polymer I, q ; II, v; III, n .

areas in the peaks corresponding to the glycerol and erythritol standards were determined by triangulation.

RESULTS

Structural Studies

Qualitative and Quantitative Analysis of Polysaccharides I, II, and

III

Composition-Quantitative analysis of the three preparations indicated that they were homopolymers containing only glucose in appreciable amounts (95 to 100%). Since only negligible amounts (<I %) of phosphate, fatty acids, uranic acid, 6-de- oxyhexoses, sialic acid, or hexosamines were found in preparations I and II, contamination by nucleic acids, mucopolysaccharides, glycolipids, or glycoproteins was minimal. The presence of nucleic acids in Preparation III was indicated by an absorption at 260 nm equivalent to 2% nucleic acid based on an assumed average molar extinction coefficient of lo4 M-’ cm-l. The only other significant, detectable contaminating substance was a variable but small amount of protein (2 to 5%). Repeated fractionation by the procedures described for isolation of the glycogen reduced the values to less than 1%. Quantitative analysis indicated a complete spectrum of amino acids in this protein contaminant. Paper chromatography of acid hydrolysates (2 N HCl, loo”, 2 hours) of Preparations I and II showed only one reducing sugar spot corresponding to glucose in Solvent Systems A and B. In Preparation III a weak spot corresponding to ribose was also found (presumably from nucleic acid).

TABLE I

Degradation of polysaccharides by P-amylase

Polysaccharide

I ............................. II. .............................. III. ............................. Amylopectin .................... Glycogen .......................

Maximal extent of conversion to maltose

No previous treatmenr

After pullulanase digestionb

% %

G5

50 58 84

52 99 45 45

a Reaction mixtures contained: polysaccharide, 0.25 mg; sodium acetate buffer, pH 4.6, 0.10 M; and fi-amylase, 500 Mg. Total volume of 1 ml. Portions, 50~1, were analyzed for reducing sugar at intervals up to 90 min. The extent of degradation was maximal at 15 min.

a Reaction mixtures contained: polysaccharide, 2 mg; sodium

acetate, pH 5, 0.03 M; and pullulanase, 250 pg. Total volume = 1 ml. After 25 hours at 37”, there was no further release of reducing

groups. The samples were heated to 100” for 2 min, cooled to 37”, and 50 pg of p-amylase added. Portions, 10 ~1, were analyzed for reducing sugar at intervals up to 18 hours. No increase in reducing value occurred after 2 hours.

Iodine Spectra-In Fig. 1 are shown the spectra of the iodine complexes of the three polysaccharides compared with those of amylopectin and rabbit liver glycogen. The data are presented in the form of logarithm of absorbance as a function of wave length to allow comparison of the shape and position of the peak without regard to the extinction coefficients. The longer wave length of maximal absorption for amylopectin is due to the greater average chain length per branch of 20 to 30 compared to 10 to 15 glucose units for glycogen. Preparations I, II, and III have absorption maxima of 525, 510, and 518 nm, respectively, compared to 480 and 530 nm for glycogen and amylopectin. These spectra are suggestive of a greater average number of glucose units per branch for the three fractions than is found in normal glycogen. The rather low absorbance of the iodine complex with Preparation II may be due to the same factors which cause its tendency toward aggregation and low solubility. This insolubility phenomenon is reminiscent of the retrogradation of starch solutions which results in a decrease in ability to form blue complexes with iodine (24).

/3-Amylase Digestion-Polyglucans such as glgcogen and amylopectin which contain a-(1,4) linkages are digested by fi-amylase from the nonreducing ends by successive removal of maltose units. Since the enzyme is unable to pass the a-(1,6) branch points, the extent of degradation is an indication of the amount of glucose in the external chains. In Table I, the maximal ex- tents of degradation of the three fractions by /3-amylase are displayed as compared to the hydrolysis of glycogen and amylopectin. The greater degree of hydrolysis of the three preparations than is found for rabbit liver glycogen is consistent with an average outer chain glucose content somewhat greater than is normal for glycogen. The finding that the samples are more susceptible to /3-amylolysis than was amylopectin was somewhat surprising. The lower wave length of maximal absorption for the Iz complexes of Preparations I, II, and III compared to that of amylopectin indicated a slightly shorter average chain length


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Issue of March 10, 1972 R. D. Edstrom 1363

for the three preparations compared to amylopectin. A possible reconciliation between these two sets of data could be made if an unusually large amount of the glucose was present in the outer chains, i.e. those in which none of the glucose units are bound in the six positions. A second set of data in Table I shows the extent of /3-amylolysis after treatment with the a-(1,6)-glucanos- lyase, pullulanase. This enzyme can completely debranch amylopectin but has little effect on the more highly branched glycogen molecule (6).

Smith Degradation-In treatment of glycogen or amylopectin by this procedure, erythritol would result from all glucose units except those at the nonreducing ends of the chains, where glycerol would be found. The ratio of erythritol to glycerol is a direct measurement of the number of glucose units per branch. The results for two independent runs were: rabbit liver glycogen, 13.4, 12.4; amylopectin, 23.1,25.6; polysaccharide III, 17.6, 18.7.

Period& Oxidation-The time course of formic acid production by periodic acid oxidation of polysaccharides I and II is shown in Fig. 2. Oxidation of amylopectin and rabbit liver glycogen are shown for comparison. The final yields of formic acid (mole of HCOOH per mole glucose) were polysaccharide I, 0.098; polysaccharide II, 0.079; rabbit liver glycogen, 0.071; amylopectin, 0.037. In polymers of a-(1,4)-linked glucose, 1 mole of formic acid is produced by periodate oxidation of each nonreducing terminus while as many as 2 moles may be produced at the reducing end. In highly branched, high molecular weight glucans such as glycogen or amylopectin, the ratio of reducing ends to nonreducing ends is very low. In such cases, the potential con- tribution of the 2 moles of formic acid from the 4-linked reducing glucose units may be ignored and the production of formic acid can be considered to be proportional to the number of branches in the molecule. For low molecular weight polymers, the release of formic acid by oxidation at the reducing terminal is not trivial. Under identical conditions to those used in Fig. 2, maltose yielded 2.2 HCOOH per mole. The theoretical value of 3 HCOOH per mole was probably not reached due to the stability of the formate ester produced at C-l (25). Thus, the relatively large amount sf formic acid produced from Polysaccharides I and II would indicate either a degree of branching higher than that found for normal glycogen or a very low molecular weight. If the molecular weights of the preparations were high enough to make the reducing end insignificant, the average chain length (moles of glucose per branch) would be the reciprocal of the quantity of formic acid produced per mole of glucose. The values in the present instance would be: polysaccharide I, 10.2; polysaccharide II, 12.3; rabbit liver glycogen, 13.9; amylopectin, 25.6. Inter- pretation of the data in this manner is not consistent with the data obtained from the /I-amylase digestion, Smith degradation, and the iodine spectra determination. This leaves the possibility of a low molecular weight as the cause of these unusual results.

20 40

Time (hours)

FIG. 2. Formic acid production by periodate oxidation of polysaccharides I (0) and II (A), rabbit liver glycogen (A), and amylopectin (0).

where L;is a filtration coefficient; u is the reflection coefficient (when u = 1 the membrane is impermeable to solute, when (r = 0, solvent and solute are equally permeable) ; R is the gas constant (8.31 x 10’ dyne cm/deg mole); T is the absolute temperature and AC, is the difference in molar osmotic concentration across the membrane.

When the pressure is equal on both sides, AP = 0 and

J, = -L,uRTAC, (‘4

the negative sign indicating the direction of solvent flow from C,i to CaZ where AC, = C,z - C,r. Where the measurements are made under circumstances when there is no solute on Side 1, AC, = c,z, so

J, = -L,aRTC,z

The rate of flow (Q) from Side 1 to Side 2 is

Q = -A.J, = L,AuRTC,z (4)

where A is the area of the membrane. By rearranging Equation 4

C,n = Q/L,AuRT (5)

L, is evaluated from Equation 1 by measurement of J, when AC, = 0 at a known Ap and was found to be 2.0 X lo-l1 cm*/ dyne sec.

Molecular Weight Determination-Attempts at estimation of molecular weight by sedimentation in the ultracentrifuge or passage through calibrated gel filtration columns were not successful due to the tendency for aggregation and precipitation of the polysaccharides.

A was found to be 35.4 cm2. The reflection coefficient, u, was estimated to be very close to 1 on the basis of negligible amounts of glucose containing material being found on the solvent side (Side 1) of the membrane at the end of the run (membrane permeable to solvent). Accurate determination of u for inulin (molecular weight 4600) yielded a value of 0.8 to 0.9 for the same system. It seems reasonable that the assumption of u = 1.0 could involve an overestimate of molecular weight by not more than 20%. In any case, an underestimate would not be possible.

For a solution containing 2.6 X lOma g per ema at 37”, Q was Measurement of molecular weight by determination of osmotic

flow is based on the permeability equations of Kedem and Kat- chalsky (26).

determined to be 5.2 x 10e6 cm3 per sec. This value yielded a molar concentration (from Equation 4) of

Flux t,hrough a membrane (Jv) is c, = 2.9 x 10-T ms

J, = L,(Ap - uRTA.C.) (1) which gives a molecular weight of 9100.

60 80

(3)


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

1364 Loll: JIolecular Weight Glycogen ill Glycogeu &wage Disease

I Polysaccharide molecular weight Method of analysis

Alkaline ferricyanidc. Reduction by NaB3H4.

Osmotic pressure.

i II

14,700 I 13,700

III RLG”

9,600

12,800 1.4 x 107

(10,900)* j c i >30,000

n Rabbit liver glycogcn, isolated by the method used for polymer I.

*This value for polymer III by the borotritide reduction is based on electrophoretic purification of [3H]glucitol after acid

hydrolysis. c Ijue to the insolubility of polysaccharides II and III, it was

not, possible to obtain osmotic pressure measurements.

TABLE III Paper chromatography and electrophoresis of tritium-labeled

h?J&wl?/SiS p1.OchCt

Substance ! Rg~ueitol ( Electrophoresis” Chromatography

Mannitol.. 1.07 1.06

Galactitol. 1.13 1.03

Sylitol I (23

Gl~~cosamiuitol. 0.88

3H-product. 1.00 ~ 1.00

ct System II. * Solvent A.

TIHLI.: IV Summary of structural data

Structural feature I

Moles glucose/reducing group. HCOOH to glucose. HCOOH to reducing group. Number of chains

(HCOOH/reducing group) - 1.2. Average chain length. .

51 0.098 5.0

3.8 13.4

II

88

0.079

A.9

5.7 15.4

For the enzymntically purified rabbit liver glycogen, an estimate of molecular weight in excess of 30,000 was found.

Chemical analysis of end groups was also performed to estimate the molecular weights of the polymers. Table II summarizes the molecular weight data determined by the alkaline ferricyanide and tritiated sodium borohydride methods. Because of the high blank values always found in the borotritide method, the analysis on polysaccharide III was repeated with the following modi- fication. The radioactive product was hydrolyzed in 2 N HCl for 1 hour at 100”. The dried samples were subjected to electrophoresis in Buffer System I. Individual strips were scanned for radioactivity and the area of the radioactive peak corresponding to glucitol was determined. The data obtained for the standard curve from this experiment was nonlinear due to the nonlinear response of the chromatogram-scanner system over the three decade range of the standard curve. It was found that a least

squares fit with the power series y = aXb where y was the area (cm2) and X the amount (micromoles) provided a good fit with a correlation coefficient of 0.9959 with a = 329.9 and b = 0.679.

The value for glucitol is reported in parentheses in Table II for polymer III. To confirm the identity of the radioactive material as glucitol, a portion was chromatographed in Solvent System X for 72 hours and the remainder subjected to electrophoresis in sodium germanate buffer for 90 min at 50 volts per cm. These results are shown in Table III.

X-ray D@~~ction Studies-The relative insolubility of am\-lose and amylopectin as compared to glycogen has been ascribed to the formation of semicrystalline regions in the molecules where parallel chains of CL-(~) 4).linked glucose units can form hydrogen- bonded structures unimpeded by c+(l) 6) branches. Evidence for this crystallinity has been demonstrated by x-ray diffraction studies (27). Diffraction patterns for polysacrharides I and II showed definite evidence of crystallinity in the two isolated polymers as compared to normal glycogen, although not as great as amylopectin.

Su~~mary of Structural Data-With the data from the molecular weight determination and periodate oxidations, the calcu- lations displayed in Table IV were performed. The values for the number of moles of glucose per reducing group are based on the average for the two methods of molecular weight determination. Multiplication of these numbers by the amount of formic acid produced in periodate oxidation gives the number of

moles of formate produced per reducing group. Subtraction oE the 1.2 moles of formic acid potentially produced by the oxidation of the first 2 carbon atoms of the reducing terminal glucose (in maltose) gives the average number of chains per molecule. The average chain length is derived by division of the number of moles of glucose per reducing group by the number of chains per reducing group. The values of 13.4 and 15.4 for the average chain lengths are somewhat lower than indicated by the Smith degradation. The uncertainty of the number of moles of formic acid produced from the reducing end makes this a less satisfactory method than the Smith degradation for determining arerage chain length.

Enzyme Studies

The storage of an uur~suxl polysaccharide such as this must be the result of deletion or aberration of enzyme activity. The following studies were undertaken to see if such a defect could be substantiated.

Branching Bnzy?,ze-&says for branching enzyme activity in the frozen diseased liver sample showed negligible amount,s to be present after 18 months storage at -21”. Because the activity was not measllred at the time of death, it was of interest to see if branching activity in the tissue specimen would survive 18 months under those conditions of storage. Portions of liver were removed at autopsy from individuals aged 0 to 5 years who had died of accidental causes or at birth. The tissue samples were frozen immediately in Dry Ice and stored at -21”. Samples were taken at weekly intervals and the branching enzyme was assayed. Values are expressed as changes in absorbance at 520 nm per min per mg of protein. The resultant activities were analyzed by least squares and found to fit the equation y = aebt where y is the amount of activity at time t, b is the first order rate constant, and a is the extrapolated value of activity at zero time. Fig. 3 shows the rate of loss of branching activity in the tissue during storage. The average value for all five samples is plotted


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

ISSUC of March 10, 1972 R. D. Edstrom 1365

2 4 6 s IO I2 14 16

Weeks Post mortem

FIG. 3. Loss of liver branching enzyme activity during storage of human liver tissue at -21”.

TIBLE T T

Aclivif~ of rut liver branching enzyme on polysaccharide polymer III

AA per min mg protein Substrate /-

520 nm I 460 nm

Glycogen ............. +0.010 ! +O.lGO Amylopectin ........ -0.252 -0.100

Polymer III ....... ~ -0.380 -0.248

against time after death. The value of the first order rate constant was found to be 0.134 per week giving a half-life of 5 weeks for the loss of branching activity on storage of human liver at -21”.

Activity of Branching li:nzyme on Polymer III-An alternative reason for the low degree of branching in these polysaccharides might be their unsuitability as substrates for branching enzyme. A l)repsrat,ion of branching enzyme was made from rat liver and the standard branching enzyme assay was carried out with this preparation with rabbit liver glycogen, amylopectin, and Preparation III. Table V contains the results of this experiment. It can be seen that the enzyme caused significant branching of amylopectin and Preparation III without appreciable effect on the normal glycogen. The absence of a-amylase was shown by

TABLE VI

rl’issr~e distribution of polysawharide polymers I and I1

TiSSlE Amount

m&“/g 1issue

Heart I. Heart II. Kidney I. . Kidney II. Liver I. Liver II. i

1.4 26

1.5

7.5 14.6

Iz wave length maximum

510 515

505

525 510

% 78

100

50 95-100

99

the failure of the enzyme preparation to cause a decreased absorbance in the iodine complex of rabbit liver glycogen at 460 nm. Samples were removed at the beginning and after 24 min of in- cubation of branching enzyme with polymer III and their 1~ complex spectra were taken. The wave length of maximum absorbance shifted from the region of 515 nm to the 490 to 495 area during Ohat 24-min period indicating that branching did occur.

Distribution of Polysaccharides in Tissues

Samples of heart and kidney tissue were treated in the same way as described for the liver polysaccharide isolation by enzy matic digestion. In Table VI are displayed the amounts of polysaccharide isolated, the glucose content, and the wave length of maximum absorption for the iodine complexes. Skeletal mus- cle was not available for analysis. No cold water insoluble polysaccharide could be isolated from kidney tissues. In the case of polymer I from heart and polymer II from kidney, the low yield precluded extensive purification.

DISCUSSION

The present work described the characterization of an abnormal glycogen which differs in several respects from the usual form.

The results of structural studies on the isolated material were all consistent with a low molecular weight, branched polgglucan. The various preparations were found to have molecular weights in the range of 8,000 to 15,000 with an average chain length of 17 to 18 glucose units per branch point. The spectra of the iodine complexes indicated a chain length of between that usually found for glycogen (10 to 15) and the 20 to 30 normally associated with amylopectin. The unusually large amount of formic acid produced during periodate oxidation was attributed to the formic acid arising by oxidation of carbon atoms 1 and 2 of the glucose at the reducing end of the carbohydrate chain. Digestion of the polysaccharide by p-amylase produced significantly more maltose than would be expected for a high molecular weight polymer having an average out’er chain length between glycogen and amylopectin. However, as the molecular weight of such a polysaccharide is drastically reduced, much more of the glucose is contained in outer, unbranched chains. This concept is based on

the findings of Gunja-Smith et al. (28) that nearly all of the glucose chains in glycogen have lengths close to the average so that there are few, if any, very long, multiply branched chains as in the usual description of the model of Meyer and Bernfeld (29). Using models based on the proposal of Gunja-Smith et al. for polymers having the properties demonstrated for Preparations I and II (four and six chains of lengths between 13.4 and 15.4 glucose units), an average of 61 and 637, of the glucose was


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

1366 Low Molecular Weight Glycogen iu Glycogen Storage Disease Vol. “47, so. 5

found to be external to all 1-6 branches.3 Thus, it is then possible for the degree of degradation by an exo-enzyme such as fl-amylase to increase markedly without an increase in the average number of glucose units per chain.

Determination of molecular weights of polysaccharides by analysis of reducing end groups has not always been successful in the case of high molecular weight polymers. By using the reducing agent sodium borohydride labeled with tritium and then purifying and quantifying the labeled glucitol, it is possible to avoid the overestimations due to polysaccharide degradation and the presence of contaminating substances. Physical measurement of molecular weight of preparation by an osmotic flow method gave a value which was in agreement with the two chemical methods.

The usual method for the isolation of glycogen from tissues is that of Somogyi (30). Digestion with 30% KOH destroys the proteins, nucleic acids, and saponifies the triglycerides leaving the glycogen more or less intact. According to Orrell et al. (31), the degradation occurring by this method may reach an extent sufficient to cause a 50.fold loss in molecular weight. Recently, Beuding et al. have studied the sedimentation characteristics of glycogens from human storage disease types I, II, III, and VI (32). With a cold water extraction method, they isolated native glycogens from a variety of sources having sedimentation coefficients in the range of 179 to 997. Such values would corre- spond to molecular weights of several millions of daltons. In the case under present discussion, the glycogen was not extract- able by cold water. Orrell et al. have also stated that glycogen extracted with hot water is somewhat degraded with respect to particle weight (31). Greenwood and Manners using hot water extraction of rabbit liver glycogen found molecular weights in the range of 2 to 6 x lo6 daltons (33). Therefore, a reduction to the level of a molecular weight of 10,000 from over lo6 could not be achieved by a simple heating at 100” in neutral solution. The similar results found with both methods indicates that there was no significant degradation of the polysaccharides by the enzyme treatments in Method A. Additionally, rabbit liver glycogen isolated by Method A retained a high molecular weight. The range in yields from the combined preparations (Prepara- tions I and II) of Method A was 71 to 90 mg per g of acetone powder, compared to a range of 63 to 122 mg per g of acetone powder for Method B (Preparation III). The variation in yields may have been due to the difficulty of keeping the slightly soluble material in solution during purification.

Since analyses were done on autopsy material, post-mortem changes must be considered. The following arguments can be made against such a possibility. The liver was removed at autopsy within 2 hours of death and placed in a -70” freezer. After freezing, the tissues were transferred to a -21” freezer where they have remained. In electron micrographs of sections taken at biopsy, the polysaccharide material had the same mor- phological appearance as in sections removed from the frozen organ (8). In earlier, preliminary studies done on liver biopsy samples, it was found that little glucose-containing material could be extracted from homogenates with cold water (5 mg per g of tissue), while hot 0.1 N H2S04 did release substantial amounts of glucose-containing materials (13 mg per g of tissue). Un-

3 Ten models for each were calculated with a table of random numbers for assigning the position of the branch points. The results were an average for the ten. The ranges were 57 to 74% for Preparation I and 45 to 72% for Preparation II.

fortunately, by the time the storage product was identified as glycogen, biopsy material was no longer available.

The clinical and pathological findings in this case are similar to those found for the six previously reported cases of glycogeu storage disease type IV (5, 7, 34-35). In all of those cases the storage product has been described as being similar to amylo- pect,in based on some or all of the following criteria: (a) an iodine spectrum having an absorbance maximum in the 510. to 535.nm region, (b) relative insolubility in cold water, and (c) an unusually high extent of digestion by phosphorylase (50 to 58%) or fl- amylase (57 to 66%). In one of those cases periodate oxidation was performed and the production of formic acid measured (35). The results indicated an average chain length of 20. The glycogen was isolated in that case by digestion of the tissue with KOH which converts the reducing end group to a saccharinic acid (3). Such an end group would not yield 2 moles of formic acid upon treatment with periodate as was the case with material under present investigation. Thus, their finding that periodate osida- tion yielded an amount of formic acid consistent with an average chain length of about 20 should not be interpreted to exclude the possibility of a low molecular weight polymer. The only infor- mation concerning the size of the glycogen molecules isolated from the tissues of a case of IV glycogenosis is a brief statement by Cori that unpublished ultracentrifugation studies showed a molecular weight smaller than that of normal liver glycogen (38).

Mercier and Whelan (6) analyzed the glycogen isolated from the liver in one of the previously described cases of type IV disease (36). They took advantage of the enzyme pullulanase which cleaves the a-( 1,6) linkages in amylopectin to the extent of 95% but has no action on the same linkages in normal human liver glycogen. Treatment of the resultant products with p- amylase causes complete degradation of amylopectin and 47y0 degradation for glycogen. They found type IV glycogen to be degraded to the extent of 80% by this method. The 84y0 degradation found for Preparation III in this study is strikingly similar.

An enzymatic deficiency has been shown in type IV glycogenosis by Brown and Brown (5). They were unable to detect branching enzyme in either the liver or leukocytes. More recently, Legum and Nitowsky have shown that both parents of a child4 having died of type IV glycogenosis had approximately 5Oa/, of the level of branching enzyme in their leukocytes as compared to normal controls (39). Because of the many similar- ities between the present case and the authentic type IV glycogenosis, the frozen, diseased liver was tested for branching enzyme activity. Although none could be detected, subsequent studies on the rate of disappearance of branching activity from normal liver under identical conditions of storage showed that it could not be detected after 18 months even if it had been present originally. The polysaccharide under present study contains fewer branches than normal glycogen in addition to having a much smaller molecular weight. There does not appear to be any basis for postulating a causal relationship between these two factors. With purified branching enzyme, it was possible to show increased branching of the isolated polysaccharide as shown by changes in absorption spectrum with iodine. Thus, the polymer would appear to be a suitable substrate for the enzyme.

Because little is known about either the regulation of polymer size in glycogen biosynthesis and degradation, or the de novo

1 Case reported by Sidbury et al. (34).


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Issue of March 10, 1972 R. D. Edstrom 1367

initiation of glycogen synthesis, one may only speculate on the causes of accumulation of such a product. There are at least three possible causes for the formation of such material: (a) a partial degradation of a polysaccharide having an unusually low number of branch points, (b) an inability of the synthesizing system to extend and branch low molecular weight chains, or (c) excessive initiation of glycogen molecule synthesis. The first of these would require the presence of a depolymerizing activity which would act only on high molecular weight substrates having an unusually low number of branch points. There is no such enzyme known to be present in liver tissue. A plausible argu- ment for the second possibility can be found in the work of Gold- emberg who has shown that glycogen synthetase (a-glucan- UDP :n-glucose glucosyltransferase) (EC 2.4.1.11) works better with larger branched molecules than with oligosaccharides or only slightly branched acceptors such as amylose or amylopectin (40). Such a selective affinity together with the relative insolubility of the smaller slightly branched polymers once formed might be sufficient to account for their accumulation. The third possibility cannot be discussed since little is known concerning the initiation of glycogen synthesis.

Acknowledgments-I would like to thank Doctors Joseph Lar- ner and William Krivit for their thoughtful criticisms and sug- gestions, Mrs. Janice Yager and Miss Patricia Bowe for their

outstanding technical assistance, and Dr. John Johnson and

Mrs. M. A. Simonds for their assistance in the osmotic flow measurements.

1. 2.

3.

4.

5.

6.

7. 8.

9.

REFERENCES

EDSTROM, R. D. (1970) Arch. Biochem. Biophys. 137, 293 FRENCH, D. (1964) in W. J. WHELAN AND M. P. CAMERON

(Editors), Control ofglycogen metabolism, pp. 7-24, Churchill, London

STETTEN, M. R., AND KBTZEN, H. M. (1961) J. Amer. Chem. sot. 83, 2912

ILLINGWORTH, B., CORI, G. T., AND CORI, C. F. (1956) J. Biol. Chem. 218, 123

BROWN, B. I., AND BROWN, D. H. (1966) Proc. Nut. Acad. Sci. U. S. A. 56, 725

MERCIER, C., AND WHEL.IN, W. J. (1970) Eur. J. Biochem. 16, 579

ANDERSON, D. H. (1956) Lab. Invest. 6, 11 KRIVIT, W., SHARP, H. L., Lx);:, J. C., LARNER, J., BND ED-

STROM, R. I)., Amer. J. Med. in press. AMINOFF, D. (1961) Biochem. J. 81, 384

10.

11.

12. 13. 14.

15. 16. 17.

18. 19.

20.

21.

22.

23. 24.

25. 26.

27. 28.

29.

30. 31.

32.

33.

34.

35.

36.

37.

38. 39. 40.

SEIFTER, S., DAYTON, S., SOVIC, B., AND MUNTWYLER, X. (1950) Arch. Biochem. Biophys. 25, 191

ELSON, L. A., AND MORGBN, W. T. J. (1933) Biochem. J. 27, 1824

ROSEMAN, S., SND DSFFNER, I. (1956) Anal. Chem. 28, 1743 DUNCOMLIE, W. G. (1963) Biochem. J. 88, 7 DISCHE, Z., AND SHETTLES, L. B. (1951) J. Biol. Chem. 192,

579 DISCHE, Z. (1947) J. Biol. Chcm. 167, 189 AMES, B. N., AND DUBIN. D. T. (1960) J. Biol. Chem. 236, 769 LOWRY, 0. H., ROSEBROUGH, N. J., FARR, A. L., ASD II LN-

DALL, R. J. (1951) J. Biol. Chem. 193, 265 PORK, J. T., AND JOHNSON, M. J. (1949) J. Biol. Chem. 181, 149 CONRBD, H. E., BAMBURG, J. R., EPLEY, J. D., AND KINDT,

T. J. (1966) Biochemistry 6, 2808 LINDBERG, B., BND SWAN, B. (1960) Acta Chem. Stand. 14,

1043 GORDON, H. T., THORNBURG, W.. BND WERUM, L. N. (1956)

Anal. Chem. 28, 849 TREVELYAN, W. E., PROCTER, D. P., AND HIIRRISOX, J. S.

(1950) Nature 166, 444 KRISMAN, C. R. (1962) Biockim. Biophys. Acta 66, 307 COLLISON, R. (1968) in J. A. R~DLEY (Editor), Starch and its

derivatives, pp. 194-202, Chapman and Hall, London NEUM~~LLER, G., AND V.ISSEM, E. (1953) Ark. Kemi 6, 235 KEDEM, O., .~ND KATCHALSKY, A. (1958) Biochim. Biowhus.

Acta 27, 229 . ”

BEAR, R. S., AND CORI, C. F. (1941) J. Biol. Chem. 140, 111 GUNJA-SMITH, Z., Mlzns~aLL, J. J., MERCIER, C.. SMITH. E. E..

AND WHELAN, k. J. (1970).Fed. Eur. Biochem: Sot. L>tt. 121 101

MEYER, K. H., AND BERNFELD, P. (1940) Helv. Chim. Acta 23, 875

SOMOGYI, M. (1934) J. Biol. Chem. 104,245 ORRELL, S. A., BUEDING, E., AND REISSIG, M. (1964) in W. J.

WHELAN AND M. P. CAMERON (Editors),’ Con&o1 oi glycogen metabolism. nn. 29-52. Churchill. London __

BUEDING, E., SIDBURY, ‘J., AND O~RELL, S. A. (1970) Biochem. Med. 3, 355

GREENWOOD, C. T., END MIINNERS, D. J. (1957) Proc. Chem. Sot. 26

SIDBURY, J. B., JR., MASON, J., BURNS, W. B., JR., .4ND RUEBNER, B. H. (1962) BUZZ. Johns Hopkins Hosp. 111, 157

HOLEMAN, L. W. J., VAN DER HAAR, J. A., AND DE Vaa~, G. A. M. (1966) Lab. Invest. 16, 357

FERNANDES, J., AND HUIJING, F. (1968) Arch. Dis. Childhood, 43, 347

LEVIN, B.. BURGESS, E. A., .IND MORTIMER, P. E. (1968) Arch. Dis. Childhood, 43, 548

CORI, G. T. (1954) Harvey Lect. 48, 145 LEGUM, C. P., AND NITO~SKY, H. J. (1969) J. Pediat. 74, 84 GOLDEMBERG, S. H. (1962) Biochim. Biophys. Actu 66, 357


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/

Ronald D. Edstroma Case of Glycogen Storage Disease

Structure of a Low Molecular Weight Form of Glycogen Isolated from the Liver in

1972, 247:1360-1367.J. Biol. Chem.

http://www.jbc.org/content/247/5/1360Access the most updated version of this article at

Alerts:

When a correction for this article is posted•

When this article is cited•

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

http://www.jbc.org/content/247/5/1360.full.html#ref-list-1

This article cites 0 references, 0 of which can be accessed free at


http://ww

w.jbc.org/

Dow

nloaded from

http://www.jbc.org/content/247/5/1360

http://www.jbc.org/cgi/alerts?alertType=citedby&addAlert=cited_by&cited_by_criteria_resid=jbc;247/5/1360&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/247/5/1360

http://www.jbc.org/cgi/alerts?alertType=correction&addAlert=correction&correction_criteria_value=247/5/1360&saveAlert=no&return-type=article&return_url=http://www.jbc.org/content/247/5/1360

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

http://www.jbc.org/content/247/5/1360.full.html#ref-list-1

http://www.jbc.org/

Structure of a Low Molecular Weight Form of Glycogem ... · Structure of a Low Molecular Weight...

Documents

Transcript of Structure of a Low Molecular Weight Form of Glycogem ... · Structure of a Low Molecular Weight...