JOURNAL OF CELLULAR PHYSIOLOGY 145:347-355 (1990)

Involvement of Hexose Transport in Myogenic Differentiation P.A. KUDO AND T.C.Y.

Lo*

Department of Biochemistry, University of Western Ontario, London, Ontario, Canada N6A 5Cl A high (HAHT) and a low (LAHT) affinity hexose transport system are present in undifferentiated rat L6 myoblasts; however, only the latter can be detected in multinucleated myotubes. This suggests that HAHT is either down-regulated or modified as a result of myogenesis. The present investigation examined the relationship between HAHT and myogenic differentiation. While myogenesis could be inhibited by the potent hexose transport inhibitor phloretin, it was not affected by phlorizin which had no effect on hexose transport. This relationship was further explored using six different HAHT-defective mutants. All six mutants, altered in either the HAHT transport affinity (Type I mutants) or capacity (Type II mutants), were impaired in myogenesis. Since these mutants were selected from both mutagenized and non-mutagenized cells with different reagents, or with different concentrations of the same reagent, the deficiency in myogenesis was likely due to changes in HAHT properties. This notion was confirmed by the observation that growth of Type I mutants in high D-glucose concentrationscould rectify the defect in myogenesis. D-glucose was unlikely to rectify the defect in myogenesis, if this defect was due to a second unrelated mutation that may have arisen during isolation of the mutants. Since both types of mutants were not altered in LAHT, D-glucose should still be taken up into the cells. The fact that the glucose-mediated increase in fusion could not be observed in Type I I mutants (deficientin the HAHT transporter) suggested that myogenesis was dependent on the presence of D-glucose or its metabolites in specific HAHT-accessible compartments. It is tempting to speculate that trans-actingregulators involved in myogenesis may be synthesized from the glucose metabolites in these specialized

HAHT-accessible compartments. Rat myoblast L6 is a clonable continuous line originally isolated from rat embryonic skeletal muscle (Yaffe, 1968). Mononucleated myoblasts will proliferate when there is ample su ply of growth factors (Endo and Nadal-Ginard, 1987). owever, upon the depletion of growth factors, myoblasts will cease DNA synthesis, withdraw from the cell cycle (terminal commitment), and eventually fuse to form multinucleated myotubes (terminal differentiation) (Sanwal, 1979; Pearson, 1981; Wakelam, 1985; Endo and Nadal-Ginard, 1987; Schneider and Olson, 1988; Florini and Magri, 1989). This morpholo ical differentiation is accompanied by biochemical di erentiation. Besides the cessation of DNA synthesis, quite a number of muscle-specific genes, such as creatine phosphokinase, acetylcholine receptor, myosin heavy chain, myosin light chains, or-actin, a- and P-tropomyosin, and troponin T, are expressed (Garfinkel et al., 1982; Endo and NadalGinard, 1987; Florini and Magri, 1989). The finding that over 40 different compounds can inhibit myogenesis at several steps suggest that the differentiation pathway is composed of a sequential cascade of multiple steps leading to terminal differentiation. Thus inactivation of any one of the essential components will block myogenesis. Although the molecular mechanisms and components involved in the differentiation pathway

R

%

0 1990 WILEY-LISS, INC.

have not been clearly defined, the myd and MyoDl genes and the extracellular matrix receptor, integrin, have been implicated in Dlavina important roles in this pathway (Daiis et al., 19871Pihey’et al., 1988; Menko and Boettiger, 1987). A high (HAHT) and a low (LAHT) affinity hexose tramp& system are present in undifferentiated rat myoblast L6 (D’Amore and Lo, 1986a,b,c). D-glucose, 2-deoxy-D-glucose(dGlc),and D-galactose are taken up predominantly by HAHT, whereas 3-O-methylglucose (MeGlc) is taken up primarily by LAHT. Myogenic differentiation was recently found to alter the dGlc, but not the MeGlc, transport affinity (Chen and Lo, 1989). The Km values of dGlc uptake were elevated upon the onset of fusion, and were directly proportional to the extent of fusion. Treatment of cells with 5-bromo2’-deoxyuridine (BrdUrd) abolished not only myogenesis but also the myogenesis-induced change in dGlc transport affinity. Moreover, the latter alteration could not be observed in a myogenesis-impaired mutant, D1; changes in dGlc transport affinity were observed only Received March 16, 1990; accepted August 8, 1990. *To whom reprint requests and correspondence should be addressed.

348

KUDO AND LO

when this mutant was grown under myogenesis-permissive conditions (Chen and Lo, 1989). It was surmised that myogenic differentiation might alter the intrinsic properties or the turnover rate of HAHT. It should be mentioned that the LAHT transport affinity remained unchanged in both undifferentiated myoblasts and in myotubes (Chen and Lo, 1989). The observed down-regulation or modification of HAHT suggests that this energy-dependent function (D’Amoreand Lo, 1986a) is no longer required once the myoblasts have committed to terminal differentiation. The question that arises is whether HAHT is required for myogenesis. It is conceivable that once the cells have committed to biochemical differentiation, the functioning of HAHT is no longer required. If this were the case, then hexose transport inhibitors should abolish myogenesis; mutants defective in HAHT should also be impaired in myogenesis, and increased glucose uptake via HAHT in these mutants might restore their ability to form myotubes. The present investigation was designed to examine these possibilities. Our findings suggest that myogenesis is dependent on the functioning of the HAHT transporter.

METHODS AND MATERIALS Cell lines and culture media

Whole cell transport studies Transport studies were carried out with six-well Falcon plates (D’Amore et al., 1986a). As indicated in our previous studies, the pro erties of the high affinity hexose transport system ( AHT) were examined by using a low concentration ran e (0.05-1 mM) of dGlc, whereas those of the low affinity hexose transport system (LAHT) were examined using MeGlc (0.1-1 mM) as the transport substrate. Medium was aspirated and each well was washed with 10 ml of DhosDhatebuffered saline (PBS; 137 mM-NaC1/ 2.7 mM-KhI 8.1 mM-Na2HP04/1.5 mM-KH2POdJ0.9 mM-CaC12,2H20/ 0.5 mM-MgC1,). Nine hundred microliters of the UDtake buffe; (FBS containin 1 mg/ml bovine serum albumin) were added to eac well. Transport studies were carried out at 23°C and were initiated by adding 100 pl of radioactive substrate to the desired final concentration. At appropriate times, uptake was terminated by washing the cells rapidly (less than 15 sec) twice with 10 ml of ice-cold PBS. In the case of MeGlc uptake, washes were always with cold PBS containing 1 mM mercuric chloride. Regardless of the transport substrate used, 1 min uptake assays were performed samples were taken at 15,30,45, and 60 sec after thc addition of radioactive substrate. Cells were solubilized with 1 ml of 0.1% Triton X-100; 0.8-ml aliquots were counted in 10 ml of scintillation fluid. All transport studies were carried out in duplicate. Background counts were determined by adding excess D-glucose (10 mM), and were subtracted from the data. Cells in two wells on each plate were detached with 0.1% trypsin, and the average cell number er well was determined using a Coulter counter. Un er the above-mentioned conditions, the uptake of dGlc and MeGlc are linear with time, and over 95% of the dGlc taken up is in the phosphorylated form (D’Amore and Lo, 1986a).

2

!i

Yaffe’s L6 rat skeletal myoblast (Yaffe, 1968) was maintained in Alpha medium (Flow Laboratories) supplemented with 10% (vh) horse serum (Flow Laboratories) and gentamicin (50 pglml; Gibco) as previously described (D’Amore and Lo, 1986a). Transfers were routinely made every 3 days (before fusion); 0.1% trypsin was used to detach cells from the plates. Unless indicated otherwise, cells were routinely plated at a density of 1.5 x lo5 cells/well in six-well (35 X 15 mm) Falcon plates. Determination of fusion index and nuclei density Cells were plated in six-well Falcon plates in Alpha Mutagenesis and isolation of mutants medium containing different concentrations of D-gluWith the exception of mutant D23 (Chen and Lo, cose, phloretin, phlorizin or BrdUrd. On different days 1988), the properties of the hexose transport mutants after subculturing, cells were treated with 1mM ZnS04 described in this manuscript have not been previously in 20% DMSO to swell the nuclei, and then fixed in reported. They were isolated from both mutagenized 2.5% glutaraldehyde in PBS (Chen and Lo, 1989). Cells and non-mutagenized rat myoblast L6 cells, by meth- were then stained with 6% Giemsa. The extent of cell ods described earlier (D’Amoreet al., 1986b).Mutagen- fusion was determined by calculating the ratio of the esis was carried out by treating actively growing L6 number of nuclei in myotubes to the total number of cells for 18 h with 250 pgiml ethyl methanesulfonate. nuclei in each field (Morris and Cole, 1972). Only Cells were allowed to recover in fresh medium for 3 structures containing at least three nuclei were considdays and then seeded at lo4 cells per 100 mm plate. ered as myotubes. Usually each field contained at least Both mutagenized and non-muta enized cells were 50 nuclei. Ten fields were counted for each set of selected by growth in fructose megum supplemented determinations. Calculation of the density of nuclei with 2-deoxy-D-glucose, or phloretin at appropriate was determined by dividing the average number of concentrations. Fructose medium was Alpha medium nuclei per field by the total area of the field. prepared without glucose; however, it was suppleEnzyme assays mented with 0.1% fructose and dialysed horse serum. After the seventh day, colonies were picked by placing For all enzyme assays, cells were grown in 150 mm cylindrical sterile cloning rings around them, placing tissue culture plates for 2 days to subconfluency. Media trypsin into the well, and after incubation for 5 min, were aspirated and cells were washed twice with citrate drawing up the cells with a Pasteur pipet. After clon- saline. After detaching the cells by incubation in citrate ing, the hexose transport and myogenic properties of saline, cells were harvested by centrifugation and the mutants were determined. Mutants were then kept suspended in homogenization buffer-5 mM 4 4 2 - h ~ at -80°C until further investigation. droxyethy1)-1-piperazineethanesulfonic acid (Hepes),

B

349

HEXOSE TRANSPORT AND MYOGENESIS

pH 7.5,l mM phenylmethylsulfonyl fluoride (PMSF)and homogenized by forcing through a 27-gauge 0.5inch needle ten times. The homogenate was then spun down in a Beckman Airfuge for 10 min at 96,600g. The supernatant obtained was kept at -20°C until use. Hexokinase activity was determined by a coupled enzyme assay using glucose-6-phosphate dehydrogenase and glucose (D’Amore et al., 198613). Phosphofructokinase was assayed by measuring the disappearance of NADPH in a coupled enzyme assa as described by Schneider et al. (1978). Glucose-6-p osphate dehydrogenase was measured by monitoring the formation of NADPH at 340 nm (Lohr and Waller, 1974). Glucose phosphate isomerase activity was assayed by the method of King (1974). The discrepancies in specific activities between the present and previous determinations (D’Amore et al., 198613) were due to differences in growth conditions and sample preparations; Sam ies were centrifuged at 96,600g for 10 min, instea of 2,OOOg for 5 min. Materials 2-Deox~-D-[1,2-~Hlglucose (50 Ciimmol) and 3-0[methyl- HI-methyl-D-glucose (50 Cilmmol) were purchased from ICN Biochemicals Inc. Phloretin, phlorizin and 5-bromo-2’-deoxy-uridine(BrdUrd) were purchased from Sigma Chemical Co. All other chemicals were obtained from commercial sources and were of the highest available purity.

i

day-6 cultures, less than 15% and 1%fusion were observed in cells grown in the presence of 50 and 100 pM phloretin, respectively; whereas about 83% fusion was observed with control cells (Fig. 1A). On the contrary, around 70% fusion was observed in myoblasts grown in 50 or 100 pM phlorizin. This clearly demonstrated that phloretin, but not phlorizin, was a potent inhibitor of myogenesis. The inability to fuse could not be attributed to the cell number before the onset of fusion (Fi . 1B).The observed slower rates of growth in 50 pM ph oretin could be due to complete inhibition of LAHT and HAHT by a concentration of phloretin 6 to 16 times higher than the respective Ki values for transport (D’Amore and Lo, 1986a). The above observation suggests that HAHT may conceivably be involved in myogenesis. This serves as the basis for further investigation.

Y

r;

RESULTS Effect of hexose transport inhibitors on myogenesis We have recently demonstrated that the properties of the rat myoblast high affinity hexose transport system (HAHT)were altered during myogenesis (Chen and Lo, 1989). In the present study, the relationship between HAHT and myogenesis was further examined using hexose transport inhibitors. Cytochalasin B (CB), BrdUrd, and phloretin have been demonstrated to be potent inhibitors for hexose transport and myogenesis (D’Amore and Lo, 1986a; Chen and Lo, 1988, 1989; Endo and Nadal-Ginard, 1987; Delain and Wahrmann, 1975 and Fig. 1A). Unfortunately, CB and BrdUrd can also act on other proteins with similar affinity (Chen and Lo, 1988; Sanwal, 1979; Pearson, 1981; Ray et al., 1987). Thus they cannot be used to examine the relationshi between hexose transport and myogenesis. Both p loretin and its analogue, phlorizin, have been used as specific affinity labels in the identification and isolation of hexose transporters (Fannin et al., 1981; Gibbs et al., 1988). Phloretin was 30-233 times more effective in inhibiting hexose transport than other cellular functions. Its inhibition constant for the rat myoblast HAHT was 3 pM,whereas those for succinate oxidation, anion transport, neutrophil activation, and thyroid hormone uptake were 700,210,100, and 88 pM, respectively (D’Amore and Lo, 1986a; De Jonge et al., 1983; Snow et al., 1978; Shefcyk et al., 1983; Movius et al., 1989).Phloretin was also six times more effective than phlorizin in inhibiting HAHT (DAmore and Lo, 1986a). Fusion indices of myoblasts grown in the presence of phloretin or phlorizin were determined (Fig. 1A). In

i

loo

i

A

uo

0 VI

GO -

3

Ls

DAYS AFTER SUBCULTURING

11:

w

g

400

L,

z 0 ‘

2

3

4 5 6 DAYS AFTER SUBCULTURING

7

Fig. 1. Effects of hexose transport inhibitors on myogenic differentiation. Rat L6 myoblasts were grown in the presence of hexose transport inhibitors. Fusion indices and the nuclei densities were determined as described in the text. A shows the fusion indices on different days after subculturing. B shows the density of the nuclei at different stages of growth of the control and phloretin treated cells. denotes untreated L6 cells. v A denote cells grown in the presence of 50 and 100 pM phlorizin. A , V denote cells grown in the presence of 50 and 100 pM phloretin. o denotes cells grown in the presence of 7.5 pM BrdUrd. ~

350

KUDO AND LO

Types of hexose transport mutants The involvement of HAHT in myogenesis was further examined using HAHT-defective mutants. In order to ensure that only the effect of HAHT was examined, a number of HAHT-defective mutants were used in this study, as it would be highly unlikely that an identical second mutation (one that alters myogenesis) could occur in all these mutants. Both mutagenized and non-mutagenized L6 myoblasts were used for the selection of mutants; they were selected for the ability to grow in a variety of toxic hexose transport substrates. The hexose transport and myogenic properties of these mutants were monitored routinely. They were found to remain unchanged with subculturing. At least two types of HAHT-defective mutants were detected. Type I mutants were altered in the Km values for dGlc transport, without changin their transport capacities (Fig. 2A). Mutants D2 and 21 were selected on the basis of their ability to grow in 0.1 mM and 0.25 mM dGlc, respectively. These two mutants were isolated from ethyl methanesulfonate (EMS)-mutagenized L6 cells (D'Amore et al., 198613). The Km values for dGlc u take by mutants D2 and D21 were 1.1mM and 1.0 m , respectively, whereas the corresponding value for the parental L6 cells was 0.6 mM. Mutant P2 was a spontaneous mutant selected for its ability to grow in 250 pM phloretin. The Km value for this mutant was 1.1 mM. These three mutants were not altered in their LAHT kinetic properties (Fig. 2B). The Km and Vmax values for MeGlc uptake were around 3.5 mM and 357 pmoles/105 celldmin, respectively. This also served to indicate that these mutants were not altered in the general membrane property. These mutants exhibited the normal level of hexokinase, glucose-6-phosphate dehydro enase, glucose phosphate isomerase, and phospho ructokinase (Table 1).Thus the defect in dGlc uptake was likely due to a specific alteration of HAHT. Type I1 mutants were altered in the Vmax, but not the Km, values for dGlc uptake (Fig. 3A). Mutants D9 and D18 were selected for their ability to grow in 0.05 mM and 0.25 mM dGlc, respectively. The former was a sDontaneous mutant. whereas the latter was selected

I.

I

0.10

>

-

\

0.05

0.00

8

8

f

. B

>

TYPE I MUTANTS

-

0 08

0.06

\ 3

0.04 0.02

n on -_ -2

/

0

4

2

6

8

10

12

1 / s

Fig. 2. Hexose transport properties of Type I mutants. Transport studies were carried out as described in the text. The amount of hexose taken up a t 23°C in 15, 30, 45, and 60 sec was used to calculate the initial rate of uptake. The initial rate of uptake (V) is expressed as nmoles of hexose taken up per min per 1 x lo5 cells. S refers to the concentration of hexose (in mmoles per litre) used. An equal amount of counts was added to each hexose concentration. A and B are double-reciprocal plots of dGlc and McGlc uptake by the Type I mutants, respectively. 0, L6; A , D2;A , D21; 0, P2.

TABLE 1. Glycolytic enzyme activities in type I and type I1 HAHT- transuort mutants' Enzyme activity (nmoles/min/mg protein) Type I mutants Enzyme Hexokinase Glucose-6-P dehydrogenase Glucose phosphate isomerase Phosphofructokinase

Parental L6

D2

D21

P2

D9

235.2 15.6 (100%) 68.7 4.5 (100%) 1,205 f 4 (10%;) 254.6 f 35.8 (100%)

252.7 f 20.1 (107%) 70.5 5.1

213.8 f 16.4 (91%) 54.4 4.4 (79%) 1,5183z 4 (126%) 209.9 f 16.1 (82%)

205.4 f 15.1 (87%) 59.9 3.9 (87%) 1,386 f 29 (115%) 211.5 f 17.6 (83%)

258.8 25.9 (109%) 68.8 f 5.7 ( 100%) 1,212 4 (101%) 244.7 k 9.1 (96%)

* *

*

(103%) 1,094 k 62 (91%) 251.0 15.5 (99%)

*

+

*

+

*

Type I1 mutants D18

D23

224.1 f 10.5 (95%) 58.4 f 4.5 (85%) 1,218 f 15 (101%) 218.8 f 19.6

225.2 10.5 (96%) 69.1 f 5.1 (101%) 1,418 f 1 (117%) 201.1 19.7

(86%)

(79%)

+

*

'Cells were grown in normal Alpha medium in 150mm plates for 2 days. Cells were harvested and homogenized as described in the text. Homogenates were centrifuged for 10 min a t 96,600gand enzyme activities were determined in the supernatant.

351

HEXOSE TRANSPORT AND MYOGENESIS

from EMS-muta enized L6 cells. The Vmax values for mutants D9 an!i D18 were 109 and 125 pmoles/105 cells/min, respectively, whereas that for the parental L6 cells was around 250 pmoles/105celldmin (Fig. 3A). These mutants were not altered in their LAHT kinetic properties (Fig. 3B). The well-characterized mutant D23 also belongs to this category. This spontaneous mutant was isolated for its ability to grow in 0.05 mM dGlc. Characterization by transport kinetics, CB binding, photolabelling, and immunoblotting studies (DAmore et al., 1986b; Chen et al., 1988;Chen and Lo, 1988, 1990) revealed that this mutant contained only residual levels of the HAHT transporter; the level of the LAHT transporter remained normal. These Type I1 mutants were similar to the Tme I mutants in that they were not altered in the gliklytic enzymes examined (Table 1). Both Type I and Type I1 HAHT- mutants were defective in myogenesis Having established the nature of the HAHT-defective mutants, their ability to undergo myogenic differentiation was then examined. All the Type I mutants examined were impaired in myogenesis. Significant

0 10

1

1

I

I

/

GO

K

20

0 0

1

2 3 4 5 DAYS AFTER SUBCULTURING

G

7

T Y P E I MUTANTS

T Y P E I1 LIUTANTS /'

O

L

1

.

2

.

"

3

4

5

L

6

A

7

DAYS AFTER S U B C U L T U R I N G

Fig. 4. Ability of Type I mutants to undergo myogenic differentiation. Fusion indices and nuclei densities were determined as described in the text. A shows the fusion indices of various mutants on different days after subculturing. B shows the nuclei densities ofthese cultures. 0 , L6; a ,D2; A , D21; 0, P2.

reduction in the degree of fusion was observed with day 4 cultures; only 3 to 7% fusion was observed with the Type I mutants, whereas around 24% fusion was observed with the parental L6 cells (Fig. 4A). About 10% fusion was observed with day 5 cultures of all three Type I mutants; this was substantially lower than the 40% fusion observed with their parental L6 cells (Fig. 4A). This inability of the mutants to fuse was not due T Y P E I1 MUTANTS o'lo to differences in growth rates or initial cell densities (Fig. 4B). As expected, the number of nuclei ceased to increase upon the onset of fusion. Since these mutants were selected by growth in different toxic reagents, and from mutagenized and non-mutagenized cells, their 0.04 inability to fuse was likely to be associated with the alteration in HAHT. All three clones of Type 11 mutants examined were 0.02 also impaired in myogenesis. However, they differed from the Type I mutants in that hardly any rnyotubes " " ' 0 00 L could be observed in day 4 and day 5 cultures (Fig. 5A). - 2 / 0 4 G 0 10 !2 About 9%, 17%,and 0% fusion were observed with day 1 / s 6 cultures of mutants D9, D18, and D23, respectively. Fig. 3. Hexose transport properties of Type I1 mutants. Transport Thus the reduction in transport capacity seemed to studies were carried out essentially as described in Figure 3. A and B are double-reciprocalplots of dGlc and McGlc uptake, respectively. 0 , have a more dramatic effect on myogenesis. Again, this L6; 0 , D9;A , D18. inability to fuse could not be attributed to changes in '

IB-

-

I

'A''

T _ _

"

352

KUDO AND LO

growth rates or density of the culture (Fig. 5B). Since the HAHT transporter could hardly be detected in mutant D23 (Chen and Lo, 19881, this transporter might conceivably be essential for myogenesis.

Rectification of the myogenesis defect by growth in D-glucose If the defect is myogenesis is indeed due to reduced levels of D-glucose or its metabolites in HAHT-accessible metabolic compartments, then growth of these HAHT-defective mutants in higher D-glucose concentrations may be able to restore their ability to form myotubes. The effect of D-glucose on fusion by day 6 cultures of Type I and Type I1 mutants was therefore examined (Fig. 6). D-glucose, up to a concentration of 150 mM, had no significant effect on the fusion index and nuclei density of the parental L6 cells. In fact, a slight drop in fusion index was observed with 150 mM D-glucose. On the other hand, the fusion index of the Type I mutant, D2, was dramatically elevated when grown in higher D-glucose concentrations; about 24%, 48%, and 65% fusion were observed a t 25,100, and 150 mM D-glucose, respectively; the nuclei density was not significantly altered at these concentrations. Although not indicated here, both mutants D21 and P2 exhibited a similar increase in fusion index when grown in high D-glucose concentrations. Since these mutants were altered only in their HAHT transport affinity, D-glucose should still be taken up by HAHT, albeit at a lower efficiency. Thus this study suggested that the deficiency in myogenesis was probably due to the reduced level of D-glucose or its metabolites in HAHT-accessible compartments. In other words, HAHT is essential in maintaining the supply of D-glucose or its metabolites which may, directly or indirectly, be required for myogenesis. A different picture was observed with the Type I1 mutants. Increasing D-glucose concentrations did not significantly alter the fusion index. About 12% and 18%fusion were observed with mutant D9 at 25 and 150 mM D-glucose, respectively (Fig. 6). Increasing the D-glucose concentration also did not increase the degree of fusion by mutant D23 (Fig. 6). Since these mutants were not altered in their LAHT, this suggested that D-glucose taken up via LAHT had no effect on myogenesis. Thus it may be inferred that the functioning of HAHT, and not LAHT, is required for myogenesis, probably in delivering D-glucose to the appropriate metabolic compartment.

DISCUSSION Myogenic differentiation is generally thought to be composed of three sequential steps: biochemical differentiation, terminal commitment, and terminal differentiation (Endo and Nadal-Ginard, 1987). Biochemical differentiation is characterized by the induction of a number of unlinked muscle-specific genes, and the down-regulation of genes associated with proliferation (Schneider and Olson, 1988). The accumulation of muscle-specific proteins during myogenesis is thought to involve transcriptional, post-transcriptional, and translational controls (Nguyen et al., 1983; Emerson et al., 1986; Breitbart et al., 1987; Endo and NadalGinard, 1987; Ewton et al., 1988; Kaufman and Foster,

$,O 80

I

N

2

ni

n

A

_I

w

rn

5 DAYS AFTER SUBCULTURING

1

0

0

82

0

3

6

4

7

T Y P E 11 MUTANTS

1200

0 2

1

3

4

5

6

7

DAYS AFTER SUBCULTURING

Fig. 5. Ability of Type I1 mutants to undergo myogenic differentiation. Fusion indices and nuclei densities were determined as described in the text. A shows the fusion indices of various mutants on different days after subculturing. B shows the nuclei densities of these cultures. 0 , L6; 0, D9; A , D18; v , D23.

120

-E-T X I

-

100 -

80-

W II P

2: 0 ; 2.L

60-

4020

+-

L------.

~

0

.

F 50 100 150 D CLUCOSh CONCENTRAIION IN G R O W I I I MLDlUM (mM) - ,

-

-

Fig. 6. Effect of D-glucose on the ability of Type I and Type I1 mutants to undergo myogenic differentiation. In this study, cells were grown in the presence of different concentrations of D-glucose. The fusion indices (closed symbols) and nuclei densities (open symbols) of day 6 cultures were determined as described in the text. 0,0,L6; A , A , D2; .,a, D9; v , v, D23.

353

HEXOSE TRANSPORT AND MYOGENESIS

1988).Stretches of highly conserved sequences, such as the CArG motif, are present in upstream regions of genes encoding for cardiac and skeletal actin, &-cardiac myosin heavy chain, cardiac and skeletal myosin light chain-2, and cardiac troponin T (Bergsma et al., 1986; Minty and Kedes, 1986).The interaction between these cis-sequences and specific trans-acting factors is thought to play an important role in myogenesis (Blau et al., 1985; Pinney et al., 1988; Schneider and Olson, 1988). As yet, no muscle-specific transcription factors have been isolated, nor is it known whether a common factor can co-regulate multiple muscle-specific genes. However, it is conceivable that mutants altered in the synthesis in these trans-acting factors should also be defective in myogenesis. We have recently demonstrated that the high affinity hexose transport system (HAHT) was down-regulated upon the formation of myotubes (Chen and Lo, 1989). This suggested that the functioning of HAHT might be associated with myo enesis. It is conceivable that HAHT may be required or myogenesis, and it is no longer required once it has completed its mission. If this were the case, then inactivation of HAHT may also abolish myogenesis. The observations that cytochalasin B and 5-bromo-2-deoxyuridineare potent inhibitors for these two processes may serve as circumstantial evidence for this notion (Chen and Lo, 1989; Endo and Nadal-Ginard, 1987; Delain and Wahrmann, 1975). Unfortunately these two reagents are also known to act on other target proteins with similar affinit (Chen and Lo, 1988;Ray et al., 1987).More definitive rliochemical evidence can be obtained through studies with phloretin. Although phloretin can also affect other cellular functions, its inhibition constant for hexose transport is 30-233 times lower than those for other functions (D’Amore and Lo, 1986a; De Jonge et al., 1983; Snow et al., 1978; Shefcyk et al., 1983; Movius et al., 1989). Furthermore, it is six times more effective in inhibiting the rat myoblast HAHT than its analogue, phlorizin (D’Amore and Lo, 1986a). Thus phloretin may be regarded as a specific hexose transport inhibitor. As shown in Figure lA, less than 15%and 1%fusion were observed in day 6 cultures grown in the presence of 50 and 100 FM phloretin, respectively, whereas about 83% fusion was observed with the untreated cells. Similar concentrations of phlorizin did not have much effect on fusion. Thus, this finding su gests that hexose transport may conceivably be invo ved in myogenesis. More definitive proof of this hypothesis comes from studies with hexose transport mutants. In order to ensure that only the effect of HAHT was examined, six different independently isolated mutants from both mutagenized and non-mutagenized rat L6 myoblasts were used. These mutants were selected by growth in different types and different concentrations of toxic hexose transport inhibitors. Routine monitoring of the hexose transport and myogenic properties of these mutants indicated that these properties remained unchanged with subculturing. Two types of HAHT-defective mutants have been identified. Type I mutants were altered in the HAHT transport affinity, with no change in transport capacity (Fig. 2). In other words, they retained the ability to take up D-glucose, albeit with much reduced efficiency. Type I1 mutants were reduced

f

f

in the HAHT transport capacity, with no change in transport affinity (Fig. 3). One of the Type I1 mutants examined, mutant D23 was previously demonstrated to contain only residual amounts of the HAHT transporter (DAmore et al., 1986b; Chen et al., 1988; Chen and Lo, 1988,1990).Thus only an insignificant amount of D-glucose should be taken up via HAHT in this mutant. It was important to note that the LAHT activities remained unaltered in both Type I and Type I1 mutants (Figs. 2 and 3). An examination of the fusion properties of these mutants revealed that both Type I and Type I1 mutants were impaired in myogenesis (Figs. 4,5). The fusion indices of day 5 cultures of Type 1 and Type I1 mutants were around 10% and 1%, respectively, whereas that for their parental L6 cells were around 40% (Figs. 4A, 5A). The inability of the mutants to fuse was not due to differences in growth rates or the initial cell densities (Figs. 4B, 5B). Since all six independently isolated HAHT-defective mutants were impaired in myogenesis, it might be concluded that the inability to form myotubes were probably due to defects in HAHT. It should be mentioned that the inability to fuse is not an inherent property of any mutant isolated from rat L6 myoblasts. For example, both the concanavalin A-resistant mutant, D1, and the wheat germ agglutinin-resistant mutant, WGAR’, were isolated by procedures similar to those used in the present study; these two mutants were able to form myotubes (Chen and Lo, 1989; Zeuner et al., 1989; Gilfix and Sanwal, 1982,1984).Myoblasts from human muscle biopsies have also been isolated and cloned by similar procedures; the cloned myoblasts were found to retain their ability to differentiate (Mesmer and Lo, 1989). Further indication of the involvement of HAHT in myogenesis comes from studies using higher D-glucose concentrations (Fig. 6). If myogenesis is dependent on the amount of D-glucose internalized via HAHT, then increasing this internal D-glucose level may restore the mutants’ ability t o form myotubes. When grown in 150 mM D-glucose, the fusion index of Type I mutants increased from 24% to 65% in day 6 cultures, whereas about 83% fusion was observed with L6 cells grown in 150 mM D-glucose (Fig. 6). This notion was further supported by comparing the fusion indices of Type I and Type I1 mutants; the relatively higher fusion indices of Type I mutants may be due to the ability of these cells to take up more D-glucose via HAHT (Figs. 4 and 5). Glucose would not be expected to rectify the defect in myogenesis, if this defect was due to a second unrelated mutation in the mutants. While this study shows that HAHT- mutants are defective in myogenesis, it should be mentioned not all myogenesis-impaired mutants are im aired in HAHT. For example, mutant DUS4 was unab e to differentiate even though it was not impaired in HAHT (D’Amore and Lo, 1988). Thus it may be surmised from these observations that the myogenic pathway is composed of a sequential cascade of multiple steps, one of which is dependent on the functioning of the HAHT transporter. The next obvious question is whether the lack of fusion is due to a general reduction of the internalized D-glucose, or the absence of a functional HAHT transporter. Studies with phloretin and the HAHT-defective

f

354

KUDO AND LO

mutants (Figs. 1 and 6) suggested that a certain minimal level of D-glucose or its metabolites was required for myogenesis. Since both Type I and Type I1 mutants can still take up D-glucose via LAHT (Figs. 3, 4), the defect in myogenesis is not likely due to a reduction of D-glucose or its metabolites in the overall intracellular space, but rather in specific intracellular metabolic compartments. This notion was confirmed by growth of Type I1 mutants in higher glucose concentrations (Fig. 6). Mutant D23 has been shown to contain only residual amounts of the HAHT transporter (Chen and Lo, 1988; Chen et al., 1988); thus most of the glucose should be taken up via LAHT. Since this mutant could not differentiate when grown in 150 mM D-glucose, it might be surmised that D-glucose taken up via LAHT could not be used for myogenesis. In other words, only the D-glucose taken up via HAHT could be used for myogenesis. This suggests a specific role of the HAHT transporter in delivering D-glucose to specific metabolic compartments. The above observation suggests that the D-glucose taken up via HAHT may be confined, at least initially, in certain metabolic compartments, which are not readily accessible to the D-glucose internalized via LAHT. Metabolic corn artments can be in the form of physical permeability arriers, interaction with subcelM a r structures or with multienzyme complexes, or compartmentation by diffusion (Sies, 1982; Kalant et al., 1987). It is interesting to note that two immunologically distinct glucose transporters have recently been demonstrated to reside in different metabolic compartments in rat adipocytes (Zorzano et al., 1989). Furthermore, a biphasic pattern of glucose utilization has recently been observed in endothelial cells; thus suggesting compartmentation of internalized glucose (Lee et al., 1989). In view of the findin that myogenesis is dependent on the functioning of H HT, it is tempting to speculate that trans-acting regulators involved in myogenesis may be synthesized in the specialized HAHT-accessible compartments, using D-glucose or its metabolites as the substrate. A reduction in the amount of D-glucose internalized by HAHT in these compartments will therefore affect the synthesis of the trans-acting regulators, thus altering the ability to undergo myogenic differentiation. This may explain why hexose transport inhibitors can block myogenesis, and why HAHTdefective mutants are unable to fuse. Attempts are being made to identify the trans-acting regulators by comparing D-glucose-labelled components in mutant D23 and its parental L6 cells. ACKNOWLEDGMENTS This investigation was supported by operating grants from the Medical Research Council of Canada to T.C.Y.L.

!i

s

LITERATURE CITED Bergsma, D.J., Grichnick, J.M., Gossett, L.M.A., and Schwartz, R.J. (1986) Delimitation and characterization of cis-actin DNA sequences required for the regulated expression and transcriptional control of skeletal a-actin ene. Mol. Cell. Biol., 6:2462-2475. Blau, H.M., Pavlath, G.K., faideman, E.C., Chiu, C.P., Silberstein, L., Webster, S.G., Miller, S.C., and Webster, C. (1985) Plasticity of the differentiated state. Science, 230:75%766.

Breitbart, R.E., Andreadis, A., and Nadal-Ginard, B. (1987) Alternative splicing: a ubiquitous mechanism for the generation of multiple protein isoforms from single genes. Annu. Rev. Biochem., 56:467495. Chen, S.R., Kay, M.M.B., and Lo, T.C.Y. (1988) Identification of the rat myoblast L6 hexose transporter(s) with specific antibodies. 4th International Congress of Cell Biology p. 231. Chen, S.R., and Lo, T.C.Y. (1988) Cytochalasin B as a probe for the two hexose transport systems in rat L6 myoblasts. Bioehem. J., 251: 63-72. Chen, S.R., and Lo,T.C.Y. (1989) Regulation of hexose transport in rat myoblasts during growth and differentiation. J. Cell. Physiol., 138:338348. Chen, S.R., and Lo, T.C.Y. (1990) Cytochalasin B photoaffinity labeling of hexose transporters in undifferentiated rat L6 myoblasts. Biochem. Int., 20:747-759. DAmore, T., Cheung, M.O., Duronio, V., and Lo, T.C.Y. (1986a) Stimulation of hexose transport in L6 rat myoblasts by antibody and by glucose starvation. Biochem. J., 238:831-836. DAmore, T., Duronio, V., Cheung, M.O., and Lo, T.C.Y. (198613) Isolation and characterization of hexose transport mutants in L6 rat myoblasts. J . Cell. Physiol., 126:26-36. DAmore, T., and Lo, T.C.Y. (1986a) Hexose transport in L6 rat myoblasts. I. Rate-limiting step, kinetic properties, and evidence for two systems. J . Cell. Physiol., 127:95-105. DAmore, T., and Lo, T.C.Y. (1986b) Hexose transport in L6 rat myoblasts. 11. The effects of sulfnydryl reagents. J. Cell. Physiol., 127: 106-1 13. D’Amore, T., and Lo, T.C.Y. (1986~)Approaches used to examine the mechanism and regulation of hexose transport in rat myoblasts. Biochem. Cell. Biol., 64:1081-1091. DAmore, T., and Lo, T.C.Y. (1988) Properties of hexose transport regulatory mutants isolated from L6 rat myoblasts. Biochem. J. 250:59-64. Davis, R.L., Weintraub, H., and Lassar, A.B. (1987) Expression of a single transfected cDNA converts fibroblasts to myoblasts. Cell, 51:987-1000. De Jonge, P.C., Wieringa, T., van putten, J.P.M., Krans, H.M.J., and van Dam, K. (1983) Phloretin-an uncoupler and an inhibitor of mitochondria1 oxidative phosphorylation. Biochim. Biophys. Acta. 722:219-225. Delain, D., and Wahrmann, J.P. (1975)Is fusion a trigger for myoblast differentiation? Exp. Cell. Res. 93:495498. Emerson, C., Fischman, D., Nadal-Ginard, B., and Siddiqui, M.A.Q., eds. (1986) “Molecular Biology of Muscle Development.” Alan R. Liss, Inc., New York. Endo, T., and Nadal-Ginard, B. (1987) Three types of muscle-specific gene expression in fusion-blocked rat skeletal cells: translational control in EGTA-treated cells. Cell, 49:515-526. Ewton, D.Z., Spizz, G., Olson, E.N., and Florini, J.R. (1988) Decrease in transforming growth factor-binding and action during differentiation in muscle cells. J. Biol. Chem., 263:40294032. Fannin, F.F., Evans, J.O., Gibbs, E.M., and Diedrich, D.F. (1981) Phloretinyl-3’-benzylazide:a high affinity probe for the sugar transporter in human erythrocytes. I. Hexose transport inhibition and photolabeling of mutarotase. Biochim. Biophys. Acta. 649:189201. Florini, J.R., and Magri, K.A. (1989). Effects of growth factors on myogenic differentiation. Am. J. Physiol. 256:C701-C711. Garfinkel, L.G., Periasamy, M., and Nadal-Ginard, B. (1982) Cloning, identification, and characterization of a-actin, myosin light chains 1,2, and 3, a-tropomysin, and troponin C and T. J. Biol. Chem. 257:11078-11086. Gibbs, A.F., Chapman, D., and Baldwin, S.A., (1988) Proteolytic dissection as a probe of conformational changes in the human erythrocyte glucose transport protein. Biochem. J., 256:421-427. Gilfix, B.M., and Sanwal, B.D. (1982) Lectin-resistant mutants of myoblasts. In: Muscle Development (M.L. Pearson and S. Epstein, eds. Cold Spring Harbor Laboratory, New York, pp. 329-336. Gilfix, B.M., and Sanwal, B.D. (1984). Relationship between cell surface asparagine-linked glycoproteins and myoblast differentiation. Analysis of wheat germ agglutinin-resistant mutants. Can. J. Biochem. Cell Biol., 62:60-71. Kalant, N., Parniak, M., and Lemieux, M. (1987) Compartmentation of glucose-6-phosphate in hepatocytes. Biochem. J., 248:927-931. Kaufman, S.J., and Foster, R.F. (1988)Replicating myoblasts express a muscle-specific phenotype. Proc. Natl. Acad. Sei. USA, 85:960f% 9610. King, J. (1974) Glucose phosphate isomerase. In: Methods of Enzy-

HEXOSE TRANSPORT AND MYOGENESIS matic Analvsis. H.U. Beremever. ed. Academic Press. New York. Vol. 2, pp. i113-1117. Lee,T.S., Saltsman, K.A., Ohashi, H. andKing, G.L. (1989)Activation of Drotein kinase C bv elevation of glucose concentration: ProDosal for' a. ~~~. mechanism develoameG of ~ diabetic vascular comdica. ~ ~ in . the . ~ ~ ~ - ~ ~ tions. Proc. Natl. Acad. Sci.cSA, 86~5141-5145. h h r , G.W., and Waller, H.D. (1974) Glucose-6-phosphate dehydrogenase. In: Methods of Enzvmatic Analvsis. H.U. Bermever, ed. Academic Press, New Yo&, Vol. 2, pp. 636-643. Menko, A.S., and Boettiger, D. (1987) Occupation of the extracellular matrix receptor, integrin, is a control point for myogenic differentiation. Cell, 5151-57. Mesmer, O.T., and Lo, T.C.Y. (1989) Hexose transport in human myoblasts. Biochem. J. 262:15-24. Minty, A., and Kedes, L. (1986) Upstream regions of the human cardiac actin gene that modulate its transcription in muscle cells: presence of an evolutionarily conserved repeated motif. Mol. Cell. Biol. 6~2125-2136. Morris, G.E., and Cole, R.J. (1972) Cell fusion and differentiation in cultured chick muscle cells. Exp. Cell Res. 75:191-199. Movius, E.G., Phyillaier, M.M., and Robbins, J. (1989) Phloretin inhibits cellular uptake and nuclear receptor binding of triiodothyronine in human Hep G2 Hepatocarcinoma cells. Endocrinology, 124:1988-1997. Nguyen, H.T., Medford, R.M., and Nadal-Ginard, B. (1983) Reversibility of muscle differentiation in the absence of commitment: analysis of a myogenic cell line temperature-sensitive for commitment. Cell, 34:281-293. Pearson, M.L. (1981) The Molecular Genetics of Development. T. Leighton and W.F. Loomis, eds. Academic Press, Orlando, Florida, pp.361-418. Pinney, D.F., Pearson-White, S.H., Konieczny, S.F., Lantham, K.E., and Emerson, C.P. (1988) Myogenic lineage determination: Evidence for a regulatory gene pathway. Cell, 53:781-793. Ray, J.H., Louie, E., and German, J. (1987) Different mutations are responsible for the elevated sister-chromatid exchange frequencies U

I

I

~~~

I

"

.

355

characteristic of Bloom's syndrome and hamster EM9 cells. Proc. Natl. Acad. Sci. U.S.A., 84:2368-2371. Sanwal, B.D. (1979) Myoblast differentiation. Trends Biochem. Sci., 4:155-157. Schneider, J.A., Diamond, I., and Rozengurt, E. (1978) Glycolysis in quiescent cultures of 3T3 cells: 11. Addition of serum, epidermal growth factor, and insulin increases the activity of phosphohctokinase in a protein synthesis-dependent manner. J. Biol. Chem., 253:872-877. Schneider, M.D., and Olson, E.N. (1988) Control of myogenic differentiation b cellular oncogenes. Mol. Neurobiol., 21-39, Shefcyk, J.,dolski, T.F.P., Volpi, M., Naccache, P.H., and Shaafi, R.I. (1983)Phloretin is a potent inhibitor of rabbit neutrophil activation by chemotactic factors. Biochim. Biophys. Acta 728:97-102. Sies. H. (1982) On metabolic commrtrnentation: Introductorv remarks. In: Metabolic Compartm&tation (H. Sies, ed.) Academic Press, Orlando, Florida, pp. 1-6. Snow. J.W.. Brandts. J.F.. and Low. P.S. (1978) The effects of anion transport' inhibitor's on structural'transitions in erythrocyte membranes. Biochim. Biophys. Acta, 512~579491. Wakelam, M.J.O. (1985) The fusion of myoblasts. Biochem. J.. 228: 1-12. Yaffe, D. (1968) Retention of differentiation potentialities during prolonged cultivation of myogenic cells. Proc. Natl. Acad. Sci. U.S.A., 61:477-483. Zeuner, E., Nandan, D., Brickendon, A., Cates, G.A., Ball, E., and Sanwal, B.D. (1989) Glycoprotein glycans may not be necessary for the differentiation of skeletal myoblasts. Exp. Cell Res., 180: 178-188. Zorzano, A,, Wilkinson, W., Kotliar, N., Thoidis, G., Wadzinkski, B.E., Ruoho, A.E., and Pilch, P.F. (1989) Insulin-regulated glucose uptake in rat adipocytes is mediated by two transporter isoforms present in at least two vesicle populations. J . Biol. Chem., 264:12358-12363.