J. Biochem. 86, 1049-1054 (1979)

Isolation and Characterization of Mucopolysaccharides from Rat Liver Mitochondria1 Yasunori KOZUTSUMI, Nobuyuki ITOH, Kazuhisa FUJIMOTO, Toshisuke KAWASAKI, and Ikuo YAMASHINA Department of Biological Chemistry, Faculty of Pharmaceutical Science, Kyoto University, Kyoto, Kyoto 606 Received for publication, May 24, 1979

Mucopolysaccharides were isolated from rat liver mitochondria which had been labeled with "S-sulfate. They were prepared from trichloroacetic acid (TCA)-insoluble and -soluble fractions of lipid-free mitochondria. These fractions were digested with pronase exhaustively, and the mucopolysaccharides were recovered in the void volume fractions of gel filtration of the pronase digests on Sephadex G-50, monitored by radioactivity determination. Identification of these mucopolysaccharides was based on electrophoresis on cellulose acetate film using three different media, enzymatic and chemical degradations specific to each type of mucopolysaccharide, using chondroitinases, heparitinase, and nitrous acid. From the TCA-insoluble fraction, chondroitin sulfate A and dermatan sulfate were obtained in a ratio of about 1 : 2, based on ^S-radioactivities, whereas the TCA-soluble fraction yielded chondroitin sulfates A/C, dermatan sulfate, and heparan sulfate in a ratio of about 1 : 3 : 12. The total amount of mitochondrial mucopolysaccharides was about 3 mg/g protein, distributed between the TCA-insoluble and -soluble fractions in a ratio of about 1 : 3.

The occurrence of mucopolysaccharides (or proteoglycans) in cellular organelles has been shown for various cells (1-10). In a previous study to characterize complex carbohydrates in rat liver mitochondria (11, 12), we found that considerable amounts of hexosamine were detected in the void volume fraction of gel filtration of a pronase digest of the whole lipid-free mitochondria, while

most of the glucosamine was in the retarded fractions consisting of glycopeptides. The void volume fraction contained all the galactosamine and a portion of the glucosamine in the pronase digest. We have now undertaken to characterize this fraction and found that most of the hexosamine could be attributed to mucopolysaccharides, identified as chondroitin sulfates, dermatan sulfate, and heparan sulfate.

1

This study was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science and Culture of Japan. Abbreviations: JDi-4S, 2-acetamido-2-deoxy-3-0-{£D-gluco-4-enepyranosyl-uronic acid)-4-C>-sulfo-D-gaIactose; JDi-6S, its 6-sulfate isomer. Vol. 86, No. 4, 1979

1049

EXPERIMENTAL PROCEDURES Preparation of Rat Liver Mitochondria—Male rats, Wistar strain, weighing 120-150 g were used, and mitochondria were isolated from the livers

1050

Y. KOZUTSUMI, N. ITOH, K. FUJIMOTO, T. KAWASAKI, and I. YAMASHINA

essentially according to the method of de Duve et al. (13). Eight rats received 125 ftCi of " S Na,SO 4 (111 mCi/nmol) by intraperitoneal injection 2 h prior to death and their mitochondria were mixed with those from rats without isotope administration. The following procedures were carried out at 0-5°C unless otherwise stated. Isolation of Mitochondrial Mucopolysaccharides —The isolated mitochondria (9.4 g protein from 150 rats which had received 1 mCi of MS-sulfate in total) were suspended in 0.1 M borate buffer, pH 8.6, to give a protein concentration of 1 %, and trichloroacetic acid (TCA) was added to give a final concentration of 10%. After allowing the suspension to stand for 1 h, it was centrifuged at 8,000 rpm for 10 min to yield the TCA-insoluble fraction (IT). Mucopolysaccharides were also recovered from the supernatant (TS). TI was treated with 95% ethanol repeatedly, each time with suspension and centrifugation, to remove TCA. After the pH of the ethanol washings became 5-6, the residue was suspended in 500 ml of chloroform-methanol ( 2 : 1 , v/v) and then the suspension was left standing overnight with stirring. The residue collected by centrifugation was washed with chloroform-methanol and then with acetone. The precipitate was dried in a desiccator. The acetone powder was suspended in 500 ml water, then the suspension was heated at 100°C for 15 min to denature proteins and to kill lysosomal enzymes which might have been present. The heated suspension was digested with pronase (initially 204 mg, then 48 mg after 48 h) in 0.1 M borate buffer, pH 8.0. After a total of 96 h digestion, when the increase in ninhydrin value reached a plateau, the digest was centrifuged, and then the supernatant was concentrated to 75 ml. This was applied to a column (4 x 88 cm) of Sephadex G-25 equilibrated with 0.05 M pyridineacetic acid buffer, pH 5.0. Mucopolysaccharides, as monitored by "S-radioactivity, were excluded from the gel. Their further fractionation is described in " RESULTS." The TS fraction was treated with ether to remove TCA, then the aqueous phase concentrated to 86 ml was applied to a column (4 x 88 cm) of Sephadex G-25. About one-seventh of the total "S-radioactivity was found in the void volume

fractions with the rest in the retarded fraction (see Fig. 2). The former was further fractionated as described in " RESULTS." Analytical Methods—Neutral sugars were determined by the orcinol-H,SO4 reaction of Hewitt (14), and uronic acid by the carbazole method of Bitter and Muir (15). Sialic acid was determined by the resorcinol method of Jourdian et al. (16). The ninhydrin reaction was carried out according to Yemm and Cocking (17). Electrophoresis—For preparation, cellulose acetate block (0.5 x6 x 17 cm, Cellogel, Chemetron, Italy) and for analysis, cellulose acetate film (Sepraphore III, Gelman, U.S.A.) were used. The technical details are given elsewhere (1, 2). The block and film were stained with 0.5% toluidine blue in 3 % acetic acid for mucopolysaccharides and with Schiff's reagent after periodate oxidation according to Fairbanks et al. (18) for glycopeptides. Nitrous Acid Degradation—Nitrous acid degradation of mucopolysaccharides was carried out according to the method of Kraemer (5). Determination of Radioactivity—Samples dissolved in water were mixed with a toluene-Triton X-100 scintillation mixture (79), and counting was carried out with a Beckman LS-100 scintillation counter. Quenching was monitored by the external standard ratio method. Reagents—Chondroitinases AC and ABC were purchased from Seikagaku Kogyo Co. Ltd., Tokyo. Heparitinase was prepared from a culture of Flavobacterium heparinum according to Linker and Hovingh (20) with some modification. Hyaluronic acid, chondroitin sulfates A and C, dermatan sulfate, and chondroitin were purchased from Seikagaku Kogyo Co. Ltd., Tokyo. Chondroitin sulfate H was a gift from Prof. N. Seno of Ochanomizu University. Heparan sulfate was isolated from pig aorta according to Klemer and Kraska (21). RESULTS Isolation of Mucopolysaccharides from Rat Liver Mitochondria—Mucopolysaccharides from TI, eluted at the void volume fraction from Sephadex G-25 as monitored by "S-radioactivity determination, were collected by lyophilization. The dried material was dissolved in an appropriate amount of 0.05 M pyridine-acetic acid buffer, pH /. Biochem.

MUCOPOLYSACCHARIDES FROM RAT LIVER MITOCHONDRIA 5.0, and the solution applied to a column of Sephadex G-50 equilibrated with the same buffer. The elution pattern is shown in Fig. 1. The mucopolysaccharide fraction (TT-1) was collected by lyophilization. This was treated with 5 % TCA to remove nucleic acids (their presence was revealed by electrophoresis on cellulose acetate film), and the supernatant was lyophilized after removal of TCA by extraction with ether to give purified mucopolysaccharides, MPS-TI. TS was fractionated on a Sephadex G-25 column (Fig. 2), then the void volume fraction was lyophilized. The dried material was digested with pronase, and the digest was further fractionated on a column of Sephadex G-50 as for TL Purified mucopolysaccharides, MPS-TS, were obtained in the void volume fraction. The yields of MPS-TI and MPS-TS were 3.0% and 8.7%, respectively, of the total M Sradioactivity in the isolated mitochondria. Uronic acid determination revealed that specific radioactivities (cpm/uronic acid) were practically the same for these two fractions. These values permitted us to estimate the mucopolysaccharide content of mitochondria, i.e. 466 nmol uronic acid/g protein or about 3 mg mucopolysaccharides/

1051

g protein. Most of the radioactivity was found in the fractions from TS which were retarded in Sephadex G-25. Characterization of these materials of low molecular weight was not carried out. Electrophoretic Examination of the Isolated Mucopolysaccharides—Three different media were used, and the patterns obtained are shown in Fig. 3. On staining with toluidine blue, MPS-TI gave two spots in media B and C, and apparently only one in medium A. It appeared that MPSTI consisted of chondroitin sulfate A and dermatan sulfate, with heterogeneity with respect to sulfate content. MPS-TS gave one elongated spot migrating as fast as dermatan sulfate. On staining with Schiff's reagent after periodate oxidation, MPSTI gave a spot migrating slower than hyaluronic acid (Fig. 3A). This appeared to be glycopeptides (for a similar example, see Ref. 1), although no conclusive identification was carried out. Separation of mucopolysaccharides from glycopeptides was possible by preparative electrophoresis on cellulose acetate block using medium A. Mucopolysaccharides were eluted from the block with water. For further identification of mucopolysaccharides of MPS-TI, the sample purified by the electrophoresis was used.

Ti-1

\ f Hi / i

0.75- 03

1

l

1.0-

0.5-

n

20

30 40 Tub* Nunbw

i i

faso-•

t

50

Q2

I

i

| 0 . 2 5 -•

0.1

60

Fig. 1. Gel filtration of the mucopolysaccharide fraction prepared from the TCA-insoluble (TT) fraction of rat liver mitochondria. The preparation procedure of the sample is described in the text. The sample dissolved in 20 ml of 0.05 M pyridine-acetic acid buffer, pH 5.0, was applied to a column (2.4 x 85 cm) of Sephadex G-50 equilibrated with the same buffer. Gel filtration was carried out at a rate of 15 ml per h, and 5 ml was collected in each tube. Aliquols were monitored for "S-radioactivity (—•—), sialic acid (—O—), and hexose (--O--). Vol. 86, No. 4, 1979

Nurrfctr

Fig. 2. Gel filtration of the TCA-soluble (TS) fraction of rat liver mitochondria. The preparation procedure of the sample is described in the text. The sample (86 ml) was applied to a column of Sephadex G-25 (4 x 88 cm) equilibrated with 0.05 M pyridine-acetic acid buffer, pH 5.0. The flow-rate was 90 ml per h, and 15 ml was collected in each tube. Aliquots were monitored for "S-radioactivity (—•—), sialic acid (—O—), and hexose (--O--).

1052

Y. KOZUTSUMI, N. ITOH, K. FUJIMOTO, T. KAWASAKI, and I. YAMASHINA (A)

ChS-A

O

ChS-6 .MPS-TS

o

MPS-TI HS HA MPSTI (B)

ChS-C ChS-A MPS-TS MPSTI ChS-B HS

(C)

o

o o

0 o

o Dtst*nc«(cm)

ChS-H

o

Fig. 4. Chromatography of chondroitinases AC and ABC digests of MPS-TI. (A) A sample corresponding to 707 cpm **S was digested with chondroitinase AC in MPS-T1 o0 50 fi\ of 0.1 M Tris-HCl buffer, pH 7.0, containing 0.1 CH unit of the enzyme at 37°C for 2 h. The digest was 1 2 cm fractionated by gel filtration on a Sephadex G-25 ? Origin column. The radioactivities were detected in the void Fig. 3. Electrophoresis of mucopolysaccharides from volume fraction and the fractions corresponding to rat liver mitochondria. MPS-TT and MPS-TS are the disaccharides.. An aliquot of the disaccharide fractions mucopolysaccharides prepared from the TI and TS was applied to a paper and developed using the followfractions, respectively. Electrophoresis was carried out ing solvent, n-butyric acid: 0.5 M aq. ammonia (5 : 3, as follows: (A) 1 M acetic acid-pyridine buffer, pH 3.5, v/v). The paper was cut into segments, 1 cm wide, c 0.5 mA/cm, 30 min, (B) 0.3 M calcium acetate, 1 mA/cm, and eluted with 0.1 M H O at 50 C for 30 min. Radio3 h, and (C) 0.1 M HC1, 2 mA/cm, 40 min. Strips were activities in the eluates were determined. (B) The void stained with toluidine blue except for the one sample of volume fraction from the AC digest was evaporated to MPS-TI marked with an asterisk. This sample was dryness, and the residue was digested with chondroitinase stained with Schiff's reagent after periodate oxidation. ABC in 50 fi\ of 0.1 M Tris-HCl buffer, pH 7.0, conChS-A, chondroitin sulfate A; ChS-B, dermatan sulfate; taining 0.1 unit of the enzyme at 37°C for 2 h. The ChS-C, chondroitin sulfate C; ChS-H, chondroitin sul- digest was treated as in (A). fate H; HS, heparan sulfate; HA, hyaluronic acid; CH, chondroitin. of about 1 : 2 and the dermatan sulfate is partially Sequential Enzymatic and Chemical Treatments oversulfated. This is consistent with the change of the Isolated Mucopolysaccharides—About one in electrophoretic pattern of MPS-TI after the third of MPS-TI, as determined by "S-radioactivity, chondroitinase digestion: The faster moving part was degraded by chondroitinase AC to produce of the spot in Fig. 3(A), the faster moving spot disaccharides. The resistant fraction, which had in Fig. 3(B), and the slower moving spot in Fig. been recovered by gel filtration on Sephadex G-25 3 ( Q disappeared after the chondroitinase AC from the chondroitinase digest, was degraded by digestion, and the remaining spots disappeared chondroitinase ABC. The digests were examined after the chondroitinase ABC digestion. by paper chromatography and the results are MPS-TS was similarly digested with the shown in Fig. 4. The chondroitinase AC digest chondroitinases, then treated with NaNO,, and the contained only JDi-4S, but the chondroitinase results are shown in Fig. 5. Based on the radioABC digest showed two spots, one of which was activities in the pooled fractions, as indicated by identified as ADi-4S derived from normally sulfated bars, the composition of MPS-TS was estimated dermatan sulfate. The other one might be a to be 6% or less of chondroitin sulfate A plus C, disaccharide with an extra sulfate group. Thus, it 20% of dermatan sulfate, and 74% or more of may be concluded that MPS-TI consists of chon- N-sulfated mucopolysaccharides. This heterodroitin sulfate A and dcnnatan sulfate in a ratio geneous nature may be consistent with the electroChS-B

MPS-TS

J. Biochem.

1053

MUCOPOLYSACCHARIDES FROM RAT LIVER MITOCHONDRIA

charides further, this fraction was treated with heparitinase from Flavobacterium heparinum.

JO (B)

JO

40

*CH

SO UH

1 »

1

u To

20

M

to

Sub-

stantial degradation occurred. However, the digest contained degradation products of a wide range of molecular size, in contrast to the more extensive digestion of heparan sulfate from pig aorta with the heparitinase. This suggests that these heparan sulfates may differ in the internal structure.

SO

DISCUSSION

10 40 ELutlon votunw(ml)

Fig. 5. Sequential enzymatic and chemical treatments of MPS-TS. A sample corresponding to 6,003 cpm "S was digested with chondroitinase AC in 330 /il of 0.1 M Tris-HCl buffer, pH 7.0, containing 0.8 unit of the enzyme at 37°C for 2 h. The digest was fractionated on a column of Sephadex G-50 (1.3 x 46 cm) equilibrated with 0.05 M pyridine-acetic acid buffer, pH 5.0. The flow-rate was 18 ml per h, and 2.2 ml was collected in each tube. The two fractions obtained, AC-I and AC-IT, were separately collected and evaporated to dryness. The residue from AC-n was subjected to paper chromatography as in Fig. 4, and that from AC-I was digested with chondroitinase ABC in 330 p\ of 0.1 M Tris-HCl buffer, pH 7.0, containing 0.8 unit of the enzyme at 37°C for 2 h. The digest was fractionated as for the AC digest. The undigested fraction (ABC-I) was evaporated to dryness, and the residue was treated with 1.8% NaNO, in 1 ml of 10% acetic acid at room temperature for 80min. The reaction mixture was fractionated by gel filtration as for the AC digest. The ABC-II fraction was evaporated to dryness, and the residue was examined as for AC-II. (A), Untreated sample; (B), after digestion with chondroitinase AC; (Q, after digestion with chondroitinase ABC of AC-I; (D), after nitrous acid treatment of ABC-I.

phoretic pattern shown in Fig. 3. The dermatan sulfate seemed to have an atypical structure since its chondroitinase ABC digest gave about 70% of JDi-4S and 30% of an unidentified spot with an Rf value lower than that of JDi-6S on paper chromatography. To characterize the iV-sulfated mucopolysacVol. 86, No. 4, 1979

It is now accepted that mucopolysaccharides occur in plasma membranes of various types of cells {1-8). Their occurrence in intracellular organelles has also been shown recently (9, 10). We discovered the occurrence of dermatan sulfate in isolated rat liver mitochondria in 1974 (22). To characterize all the mitochondrial mucopolysaccharides occurring in rather minute amounts, mucopolysaccharides were labeled with "S-sulfate and extensively fractionated. It was found important to kill lysosomal enzymes before carrying out pronase digestion to solubilize mitochondrial mucopolysaccharides since various glycosidases and sulfatases of probable lysosomal origin often degrade the mucopolysaccharides substantially. The purity of the isolated mitochondria was very high, and the amounts of contaminating microsomes and lysosomes were estimated to be at most 2 % each of the total protein, based on the activities of marker enzymes (glucose-6-phosphatase for microsomes and acid phosphatase for lysosomes). Thus, the mitochondrial mucopolysaccharides are practically free from mucopolysaccharides from these organelles since the mucopolysaccharide contents of microsomes and lysosomes are about the same as that of mitochondria (unpublished data obtained in our laboratory). When successfully solubilized, mitochondrial mucopolysaccharides were recovered in both the TCA-insoluble (TI) and -soluble (TS) fractions in a ratio on weight basis of about 1:3. It is unlikely that this separation is due to incomplete precipitation of proteoglycans since the mucopolysaccharide patterns of TI and TS differed considerably. In view of the low protein content of proteoglycans from various sources (usually about 10%, see Ref. 4 inter alia), they should not

1054

Y. KOZUTSUMI, N. ITOH, K. FUJIMOTO, T. KAWASAKI, and I. YAMASHINA

be TCA precipitable. For this reason, solubility of proteoglycans in TCA seems to reflect the localization of the proteoglycans within mitochondria. Thus, it is suspected that MPS-TI represents the mucopolysaccharides bound tightly to the membranes of mitochondria whereas MPSTS are those bound loosely to the membranes or those occurring in the intermembrane spaces. In total, mitochondrial mucopolysaccharides are composed of 13% or less of chondroitin sulfates A plus C, about 31 % of dermatan sulfate and 56 % or more of heparan sulfate, based on the radioactivities. Dermatan sulfate and heparan sulfate seem to have atypical structures involving oversulfation, as judged from the electrophoretic patterns of the intact mucopolysaccharides and the chromatographic patterns of the enzymatic digests. This mucopolysaccharide pattern is somewhat different from that reported by Dietrich et al. (10). They claim that dermatan sulfate constitutes 70% and heparan sulfate 30% of the total mitochondrial mucopolysaccharides. This difference is probably due to the difference in procedures used to isolate mucopolysaccharides. It is interesting to compare our data with those of Suzuki et al. (23) for mucopolysaccharides in whole rat liver and a shift in their content in carbon tetrachloride-damaged liver. They found two dermatan sulfates with different electrophoretic mobilities, i.e. different sulfate contents and their liver contents increased with the experimental injury. Our mitochondrial dermatan sulfate had the same electrophoretic mobility as Suzuki's high sulfated dermatan sulfate (a gift from Dr. S. Suzuki of Nagoya University). This is consistent with a finding that mitochondria are altered in some way after administration of carbon tetrachloride (24). At least a part of the dermatan sulfate appears to occur in the outer membranes of mitochondria. In some work performed recently in our laboratory (unpublished data), it was shown that chondroitinase ABC treatment of rat liver mitochondria enhanced the tryptic sensitivity of the rotenoneinsensitive NADH-cytochrome c reductase whereas chondroitinase AC did not.

REFERENCES 1. Funakoshi, I., Nakada, H., & Yamashina, I. (1974) /. Biochem. 76, 319-333 2. Nakada, H., Funakoshi, I., & Yamashina, I. (1975) / . Biochem. 78, 863-872 3. Akasaki, M., Kawasaki, T., & Yamashina, I. (1975) FEBS Lett. 59, 100-104 4. Mutoh, S., Funakoshi, I., & Yamashina, I. (1976) /. Biochem. 80, 903-912 5. Kraemer, P. (1971) Biochemistry 10, 1437-1445 6. Yamamoto, K. & Terayama, H. (1973) Cancer Res. 33, 2257-2264 7. Buonassisi, V. & Root, M. (1975) Biochim. Biophys. Ada 385, 1-10 8. Underhill, C.B. & Keller, J.M. (1975) Biochem. Biophys. Res. Commun. 63, 448-454 9. Bhavanandan, V.P. & Davidson, E.A. (1975) Proc. Natl. Acad. Sci. U.S. 72, 2032-2036 10. Dietrich, C.P., Sampaio, L.O., & Toldes, O.M.S. (1976) Biochem. Biophys. Res. Commun. 71, 1-10 11. Itoh, N., Kawasaki, T., & Yamashina, I. (1974) / . Biochem. 76, 459-^66 12. Itoh, N., Kawasaki, T., & Yamashina, I. (1974) FEBS Lett. 47, 225-228 13. de Duve, C , Pressman, B.C., Gianetto, R., Wittiaux, R., & Appelmans, F. (1955) Biochem. J. 60, 604-617 14. Hewitt, L.F. (1937) Biochem. J. 31, 360-366 15. Bitter, T. & Muir, H. (1962) Anal. Biochem. 4, 330-334 16. Jourdian, G.W., Dean, L., & Roseman, S. (1971) / . Biol. Chem. 246, 43