Incorporation of retinoic acid into proteins via retinoyl-CoA

Biochimica et Biophysica Acta, 998 (1989) 69-74 Elsevier

69

BBAPRO 33439

Incorporation of retinoic acid into proteins via retinoyl-CoA B r i t t a R e n s t r o m * a n d H e c t o r F. D e L u c a Department of Biochemistry, Universityof Wisconsin-Madison, College of Agricultural and Life Sciences, Madison, WI (U.S.A. ) (Received 6 December 1988) (Revised manuscript received 8 May 1989)

Key words: Vitamin A; Retinoid; Retinoic acid; Differentiation

The incorporation of tritiated retinoic acid into proteins has been studied in cell-free extracts from rat liver and kidney. Incubation with retinoic acid in the presence of ATP and CoA resuited in a CoA intermediate. This CoA intermediate is a substrate for enzymes that incorporate the labeled retinoic acid moity into proteins ( M r = 14000-60000) as detected with SDS-imlyacrylamide gel electrophoresis. In the microsomal fraction, the label was found in a single protein ( M r 30000). Retinoie acid is linked to the protein via a thioester bond as indicated by neutral hydrolysis with hydroxylamine and the bond's sensitivity to reducing agent. The incorporation of labeled retinoic acid into the protein is very rapid but decreases upon prolonged incubation, suggesting a high turnover of retinoic acid in the protein.

Introduction Retinoic acid is formed irreversibly from retinal and has been identified as a natural metaborite of retinol in rats [1]. Retinoic acid can support the functions of vitamin A in growth [2] and in the control of epithelial cell differentiation [3], but, unlike retinal and retinol, it cannot be stored and is rapidly metabolized [4,5]. Miller and DeLuca [6] studied the metabolism of all-trans-retinoic acid in river microsomes and detected a less polar metaborite that was identified as ethyl retinoate. The formation of ethyl retinoate proved to be a specific enzymatic process and is greatly stimulated by the addition of CoA, suggesting the formation of a retinoic acid CoA intermediate, retinoyl-CoA. In attempts to determine the biological significance of retinoyl-CoA, we have attempted to determine the fate of the retinoic acid moity in retinoyl-CoA. The retinoic acid moity appears to be transferred to tissue proteins. It is well documented that fatty acid via a CoA intermediate can covalently bind to lipid, steroids and

proteins [7-11]. The present report demonstrates that retinoyl-CoA is indeed an intermediate in retinoic acid metaborism, and that it serves as a substrate providing retinoic acid moieties for proteins via a thioester bond. Retinoyl-CoA is also an intermediate for an unknown less polar metaborite, tentatively identified as an ester. Experimental procedures

Materials All-trans-[3H]retinoic acid (20.5 Ci/mmol) was suppried by the Hoffmann-La Roche Company (Nutley, N J). All other chemicals were obtained through commercial suppliers and were of analytical grade. All solvents were HPLC grade.

Animals Weanling male rats were obtained from SpragueDawley (Madison, WI) and were maintained on a purified vitamin A-deficient diet.

Preparation of subcellular fractions * Present address: University of Wisconsin-Madison, Medical Science Center, Cardiology, 1300 University Avenue, Mad/son, WI 53706, U.S.A. Abbreviations: DTT, dithiothreitol; CoA, coenzyme A; SDS, sodium dodecyl sulfate; TCA, trichloroacetic acid; THF, tetrahydrofuran; BHT, butylated hydroxytoluene. Correspondence: H.F. DeLuca, Department of Biochemistry, University of Wisconsin, 420 Henry Mall, Madison, Wisconsin 53706, U.S.A.

All procedures were carried out a 4 ° C. Rats were kiiled 8 weeks after weaning and their rivers and kidneys quickly removed and rinsed in ice-cold 0.1 M Tris-HCl buffer (pH 7.4) containing 0.25 M sucrose. This tissue was finely minced and homogenized in 0.1 M Tris-HCl buffer (pH 7.4) containing 0.25 M sucrose, 5 mM DTT and 10 mM MgCI 2. Homogenates were filtered through three layers of cheesecloth and centrifuged at 400 × g for 10 ~fin. The supernatant was

0167-4838/89/$03.50 © 1989 Elsevier Science Publishers B.V. (Biomedical Division)

70 centrifuged at 10000 × g for 10 rain and the resulting supernatant at 100000 × g for 1 h. Pellets were resuspended in 0.1 M Tris-HCl buffer (pH 7.4) containing 0.25 M sucrose, 5 mM DTT and 10 mM MgCI 2. Microsomal preparations were used immediately or stored at - 7 0 °C no longer than 2 months prior to use. Protein concentration was determined according to the method of Bradford [12].

Enzyme assay Fractions (1-2.5 mg of protein) were incubated at 37°C with 22 pmol [3H]retinoic acid (22 pmol, 20.5 Ci/mmol), 0.1 M Tris-HCl buffer (pH 7.4), 27 mM MgCI2, 1 mM DTT, 50 mM sucrose, 10 mM ATP and 150/~M CoA in a final volume of 1 ml. For labeling with the acyl-CoA, the fractions were incubated in the same assay mixture except that 10 mM MgCI 2 was used, ATP and CoA were omitted and 40-75 nmol [3H]retinoyl-CoA (6/~Ci//~mol) was used as substrate instead of the acid. Reactions were stopped by freezing the sample immediately in dry ice. Assay mixtures were stored at - 70 o C under argon. Reaction mixtures used for detecting metabolites other than proteins were lyophilized and then extracted with T H F / H 2 0 (2: 1) or CHCI3/MeOH (2 : 1) containing BHT (50/tg/ml) to prevent non-specific oxidation. The aqueous extracts were pooled and brought to dryness with lyophilization. The organic fractions were dried under a stream of N 2. The aqueous and organic soluble materials were dissolved in 200/~1 H20 and 200 /~1 MeOH, respectively, and subjected to HPLC analysis. Samples used for protein analysis were dialyzed (Spectra/Pot membranes Mr cutoff= 12000-14000) against 0.1 M NaCI for 7 h and H20 overnight to eliminate uureacted substrate and low molecular weight molecules [13]. After dialysis, the proteins were precipitated with 5~ TCA or directly lyophilized and stored as a powder at - 7 0 °C before further treatment. The protein mixture was also purified using a Sephadex G-100 column eluted with 50 mM Tris-HCl buffer (pH 7.4) and 0.5~o SDS to isolate the protein fraction free from the substrate. Fractions were counted and absorbance measured at 280 nm. Dialysis was less time consuming and, therefore, preferred to gel filtration.

Electrophoresis Lyophilized protein was added to an electrophoresis sample buffer containing 10~o glycerol, 2.3% SDS, 6.2 mM Tris-HC1 buffer (pH 7.2) and bromophenol blue and set at room temperature for 60 rain. SDS-polyacrylamide slab gel electrophoresis was performed with 0.1~ SDS and 9~ acrylamide. The gels were developed with autoradiography, Coomassie blue or silver staining. Some gels were cut into 3 mm slices,

digested in hydroge~t peroxide and the label incorporation was determined by scintillation counting using 3a70b or Bio-Safe II (Packard, Downers, Grove, IL).

Other methods For hydroxylaafine hydrolysis [10|, the proteins were incubated with 1 M of hydroxylamine hydrochloride at pH 7.0 overnight. Base hydrolysis of retinoyl-CoA was carried out by the addition of 200/tl of 0.1 M KOH in E t O H / H 2 0 (1 : 1) and incubation for 30 min at 37 o C.

Preparation of [ JH]retinoyl-CoA Labeled retinoyl-CoA was synthesized according to a modified method of Kutner et al. [14]. all-trans-[3H]Ret inoic acid (20.5 Ci/mmol) was added to a final concentr~tion of !-5/~M and 200/~1 dioxane along with an equimolar amount of BHT and dried three times with dioxane under N 2 stream. This preparation was treated with 20 /tM N-hydroxysuccinimide in 200 /~1 dioxane and 20 #M N,N'-dicyclohexylcarbodiimide in 200 /~1 dioxane and the reaction mixture was incubated at 37°C for 1 h under N 2. The solvent was evaporated under N 2. A solution of 0.5-1 ml containing 15 /~M CoA in T H F / H 2 0 (2: 1), adjusted slowly to pH 8.0-8.5 with 1~ NaHCO 3, was added to the dried ester, mixed well, flushed with N 2 and incubated with stirring at 37°C overnight. The reaction was terminated by the addition of CHCI3/MeOH (2:1) containing BHT (50 /tg/ml). The unreacted materials were extracted with CHCI3/MeOH (2:1). The purity of the aqueous product was verified by HPLC [14]. An aliquot was hydrolyzed and the product identified by HPLC. Chromatography was performed with a Beckman model 47.3 liquid chromatography system (Beckman Instruments, Arlington Heights, IL) with a Waters U6K injector (Waters Associates) and a Beckman 160 fixed wavelength UV detector. A Spherisorb C8 reverse-phase column (4.6 mm x 25 cm, Phenomenex, Rancho Palos Verdes, CA) and a Zorbax-ODS Cls reverse-phase column (4.6 mm × 25 cm, DuPont, Wilmington, DE) were used to separated the aqueous and organic soluble material, respectively. Results

Initial attempts to detect the in vitro formation of retinoyl-CoA utilized homogenized tissue and microsomes from rat kidney incubated with [3H]retinoic acid (ATP, MgCIz, CoA) and analyzed for product by HPLC. A product comigrating with synthetic retinoyl-CoA was detected (Fig. 1A). The time course indicated that maximum production of retinoyl-CoA occurred after only 5 min incubation and then decreased (Table I), an indication that the product was utilized in other reactions. No retinoyl-CoA was detected with boiled tissue preparations (Fig. 1B) or in the presence of other denaturing

5.0 ~-

4.5 -

rl

At

71

lOO

4.0 ~,

90

"D

80

°

.o

i

,,.

' " i=0

1.8

I

.e

.o II .. "-...

lOO

2.0

~0

~~

S

90

"o

1.6

80

1.4

70

=

1.2

eo

~

",

-r

..,I

c)

1.0

so

5"

2.0

40

(~

x

0.8

4O

~

1.5

30

0 -1-

E .~ o.e

3o

,1,

1.0

20

I I I

t~

0.4

20

&

0.5

10

0.2

10

2.5

L

0

5

50

I

I

I

I

I

I

I

I

I

10

15

20

25

30

35

40

45

50

Fraction

I

Y

55

60

,i

0

S

,

10

=

i

15

20

Number

.

,

25

30

.

•

35

40

•

45

,

50

,

55

,

60

0

,,,-

65

70

Fraction Number

Fig. 1. HPLC profile from fresh rat kidney microsomes (A) or heated kidney microsomes (B) incubated with [3H]retinoic acid. Sp[,~:nsorb C 8 column (4.6 m m x 2.5 cm, Phenomenex) was eluted with 10 mM NH4OAc in methanol/water gradient system at a flow rate of 1.0 ml/min with UV monitoring at 340 nm. The arrows indicate the elution position of retinoyl-CoA (Fraction 43) and all-trans-retinoic acid (Fraction 50).

agents such as SDS buffer for electrophoresis, demonstrating it to be an enzymatic process (data not shown). The in vitro product co-chromatographed with an

TABLE I Production of [~H]retinoyl-CoA in microsomes from rat kidney Incubation conditions (min)

dpm of [ 3H]retinoylCoA per mg protein

dpm of [3H]retinoylCoA per mg protein as ~ of total dpm incubated

1 3 5 10

886 1200 6221 n.d.

0.24 0.46 1.24 n.d.

1

2

3

,4

5 10 s

n.d., not detectable.

I

2.0

100

1.8

90

1.6

80

1.4

70

1.2

60

¢-

o 4-1

(,1

bP

1.0

tD

0 ,X

E ix "r o)

"I -r 0

i.

u.

.g (1) o (I)

%

5O 40

0.8 0.6 0.4

t3o 20

0.2

10

0

~, -!

O

5

10

15

Fraction

20

25

30

¢D

:C 0 i I I

v

35

Number

Fig. 2. HPLC profile of hydrolyzed [3H]retinoyl-CoA isolated from rat kidney microsomes. Zorbax-ODS Cls column (4.6 ram×25 cm, Dupont) was eluted with 10 mM NH4OAc in methanol/water gradient system at a flow rate of 1.0 ml/min with UV monitoring at 340 nm. The arrows indicate the elution position of retinoyl-CoA (Fraction 17) and all-trans-retinoic acid (Fraction 20).

Fig. 3. Retinoic acid labeling of proteins in fractions from rat kidney. The fractions were incubated with [3H]retinoyl-CoA for 5 rain and analyzed by SDS-polyacrylamide gel electrophoresis. Lane 1, 400 x g supernatant; lane 2, 10000× g supernatant; lane 3, 10000 × g pellet; lane 4, 100000x g supernatant; 5, 100000x g pellet. The positions of 14C molecular weight markers are indicated.

72 10

9

03 a) .o "~

8

x

4

m

7

a

z

2 1 0

0

5

10

15

20

25

30

35

40

45

50

Gel Slice

Fig. 5. Retinoic acid labeling of proteins in microsomes from rat kidney analyzed by SDS-acrylamide gel electrophoresis. Gels were sliced into 3 mm slices and counted. Arrow indicates position of 14C molecular weightmarker Mr = 30000.

Fig. 4. Retinoic acid labeling of proteins in microsomes from rat kidney and liverat various incubationtimes. Lane 1, 5 min kidney; 2, 60 min kidney; 3, 5 min kidney boiled enzyme; 4, 1 rain liver. The positions of 14C molecularweightmarkersare indicated.

authentic sample of retinoyl-CoA on HPLC (Fig. 1A). The metabolite was isolated following reverse-phase HPLC separation and subjected to base hydrolysis. The hydrolyzed product was applied to HPLC. The label comigrated with all-trans-refinoic acid (Fig. 2). [3H]Retinoyl.CoA was investigated further. Homogenized tissue and subcellular fractions from both kidney and fiver were incubated with [3H]retinoyl-CoA. The formation of labeled proteins was analyzed by Sephadex gel filtration, TCA precipitation and SDS-polyacrylamide gel electrophoresis. These data showed all fractions contained some labeled proteins (Fig. 3) and the highest ratio between labeled protein per mg protein incubated was found in the microsomes (data not shown). Gel electrophoresis showed several labeled proteins in the soluble fraction ranging from M r 14000-60000 (Fig. 3) but the microsomal fractions of both fiver and kidney showed major labeled protein in the M r 30 000 region (Fig. 4). The dpm determination of gel slices indicated that kidney tissue contained two major peaks of radioactivity and 3 or 4 small or insignificant peaks. The major peak at Fraction 43 is free retinoic acid, and the peak at Fraction 27 is a protein of approx. M r 30000 (Fig. 5). In fiver, 4 major peaks of radioactivity appeared, 2 of which are about M r 30000.

One did not enter the gel and the other is free retinoic acid (Fig. 6). The incorporation of retinoic acid from retinoyl Co-A is prevented by heating the tissue preparations to 100 ° C for 10 min or by SDS denaturation, demonstrating it to be an enzymatic process. The labeKag was very rapid but time-dependent (Tables II and III). Maximum incorporation was reached after 5 min in kidney tissue and 0-1 rain in fiver tissue with a rapid decrease upon prolonged incubation. A metabolite less polar than the substrate was detected in fiver microsomes. This product was purified by 10 9 (1) (.1

8

o ~

u)

7

w

o)

t3

6 5

x

4

E

3

o" "lo "I-

2 1 0 0

.5

10

15

20

25

30

35

40

45

50

55

Gel S l i c e

Fig. 6. Retinoic acid labeling of proteins in microsomesfrom rat fiver analyzed by SDS-acrylamidegel electrophoresis.Gel was sliced into 3 mm slices and counted. Arrow indicates position of 14C molecular weightmarker Mr = 30000.

73 TABLE II

Incorporation of [3H]retinoic acid into microsomal protein from rat kidney incubated with [3H]retinoyl-CoA Incubation conditions (min)

dpm in TCA precipitate per mg protein

dpm in TCA precipitate per mg protein as • of total dpm incubated

5 tO 60

67930 1~361 12150

3.3 2.3 0.2

TABLE 11I

Incorporation of [~H]re'tinoic acid in microsomalprotein from rat liver incubated with [ ~H]retinoyl-CoA Incubation conditions (min)

dpm in TCA precipitate per mg protein

dpm in TCA precipitate per mg protein as ~ of total dpm incubated

1 3 5 30

4508 2776 1361 773

1.41 0.69 0.41 0.13

HPLC and was found to undergo base hydrolysis. The hydrolyzed product comigrated on I~PLC with alltrans-retinoic acid, indicating an ester function but no further investigation was done.

Stability of the retinoic acid protein linkage The stability of [3H]retinoic acid-bound protein was investigated by treatment with reducing agents and hydroxylamine at neutral pH. Addition of mercaptoethanol in the electrophoresis sample buffer caused cleavage of the bond and loss of labeled protein. Treatment with neutral hydroxylamine also reduced the amount of radioactivity associated with the protein.

not is unknown and requires further work on the identification o f the p r o t e i n s r e t i n o y l a t e d a n d h o w this m i g h t affect their function. R e t i n o y l - C o A does n o t a p p e a r to be i n v o l v e d in m e t a b o l i c d e g r a d a t i o n o f r e t i n o i c acid, since n o lipid solubl¢~ p r o d u c t s were d e t e c t e d f r o m [ 3 H ] r e t i n o y i - C o A . F u r t h e r ~ o x i d a t i o n d o e s n o t a p p e a r a feasible r o u t e of d e g r a d a t i o n for this h i g h l y c o n j u g a t e d i s o p r e n o i d

material. There are several proteins in liver and kidney that become retinoylated with retinoyl-CoA. Tile soluble fraction contained several proteins labeled with retinoic acid. However, the highest specific labeling took place on a M r 30000 protein found in the microsomal fraction. Since pure microsomes were not prepared, a conclusion that this protein is found in the membrane that constitutes microsoraes cannot be made. The nature of the protein is not known and is under investigation. Retinoyl-CoA formed by tissue in vitro was identified by co-chromatography with synthetic retinoyl-CoA and by alkaline hydroly~,is to yield all-trans-retinoic acid, leaving no ,doubt that it is formed. Furthermore, CoA, ATP and r¢tinoic acid were required for its formation. The retinoyl moiety on proteins was found stable to denaturing conditions of SDS-PAGE, TCA precipitation and gel filtration. The retinoyl bond to protein can be cleaved with reducing agents and neutral hydroxylamine strongly suggesting a thioester linkage [7-11]. These results show the participation of retinoic acid in a new reaction of undetermined significance. Acknowledgements This work was supported by a program project grant No. DK-14881 from the National Institutes of Health and by the Harry Steenbock Research Fund of the Wisconsin Alumni Research Foundation.

Discussion References

The mechanism of action of retinoic acid in inducing differentiation of epithelial cells remains unknown, although the possible existence of a retinoic acid receptor has been demonstrated by molecular biology techniques [15,16], suggesting a mechanism similar to that of the steroid hormones. Until this mechanism is firmly established, other mechanisms must be considered. In this report we have demonstrated that kidney tissue is capable of producing retinoyl-CoA from CoA and retinoic acid. This product is formed rapidly and is rapidly turned over. Retinoyl-CoA is a substrate for the retinoylation of certain proteins. The reaction is similar to the previously reported fatty acylation of proteins [7-11]. The retinoylated protein(s) appear to lose the retinoyl group rapidly, suggesting a high turnover rate. Whether retinoylation of proteins is of functional significance or

1 Emerick, R.J., Zile, M. and DeLuca, H.F. (1967) Biochem. J. 102, 606-611. 2 Zile, M. and DeLuc~- H.F. (1968) Biochem. J. 94, 302-308. 3 Dowling, J.E. and Wa~d, G. (1960) Proc. Natl. Acad. Sci. USA 46, 587-608. 4 Roberts, A.B. and DeLuca, H.F. (1967) Bioche,~,.J. 10:?,600-605. 5 Zachman, R.D., Dunagin, P.E. and Olson, J.A. (1966) J. Lipid Res. 7, 3-9. 6 Miller, D.A. and DeLuca, H.F. (1985) Proc. Na~l. Acad. Sci. USA 82, 6419-6422. 7 Lichtenstein, A.H. and Brecher, P. (1980) ~ Biol. Chem. 255, 9098-9104. 8 Robinson, M., Blank, M.L. and Snyder, F. (198~) J. Biol. Chem. 260, 7889-7895. 9 Lakin-Thomas, P.L. and Brody, S. (1986) J. Biol. Chem. 261, 4785-4788. 10 Olson, E.N., Towler, D.A. and Glaser, L. (1985) J. Biol. Chem. 260, 3784-3790.

74 11 Wall, L.A., Byers, D.M. and Meighen, E.A. (1984) J. Bacteriol. 720-724. 12 Bradford, M.M. (1974) Anal. Biochem. 72, 248-254. 13 Allen, G. (1981) in Laboratory Techniques in Biochemistry and Molecular Biology. Sequencing of Proteins and Peptides, p. 23, Elsevier Science Publishers, New York.

14 Kutner, A., Renstrom, B., Schnocs, H.K. and DeLuca, H.F. (1986) Proc. Natl. Acad. Sci. USA 83, 6781-6784. 15 Petkovich, M., Brand, N.J., Krust, A. and Chambon, P. (1987) Nature 330, 444-450. 16 Giguere, V., Yang, N., Segui, P. and Evans, R.M. (1988) Nature 331, 91-94.