Hepatic microsomal composition studies in the Gunn rat

Biockimrea et Biopkysica Acta, 750 (1983) 419-423 Elsevier Biomedical Press

419

BBA Report BBA 50036

HEPATIC MICROSOMAL GLENN

R. GOURLEY,

COMPOSITION

WILLIAM

MOGILEVSKY

STUDIES IN THE GUNN RAT

and GERARD

B. ODELL

U~~#e~sit,v of Wisconsin, School of Medicine, ~epu~tment of Pedi~t~~~s,Madison, WI 53792 (U.S.A.) (Received

July 19th, 1982)

The homozygous jaundiced (3) Gunn rat exhibits hepatic microsomal enzyme activities which vary from markedly decreased to normal when compared with the non-jaundiced (JJ) Gunn rat. In order to determine if an alteration in microsomal lipid might be related to these observations, cholesterol, phospholipid, fatty acid and fluorescence polarization determinations were carried out in Gunn rats of both genotypes. Significant differences in microsomal palmitic, stearic and arachidonic acid composition were present, but these were not striking. Fluorescence polarization data best fit a two-phase linear model for both genotypes with no significant differences in breakpoint temperatures. In J rats, the anisotrophy parameter ((r’/r) 1) -’ was significantly greater than that seen in JJ rats at both 25 and 37”C, indicating a decreased membrane fluidity in the juandiced animals. Alterations in enzyme microenvironment due to subtle changes in lipid composition may be related to the different enzyme activities observed in Gunn rats.

The microenvironment of membrane-bound enzymes is clearly important in determining enzyme activity [ 11.The function of enzymes embedded in the endoplasmic reticulum, like enzymes of many other biological membranes, is intimately related to the microenvironment provided by membrane lipids [2]. The jaundiced (2) Gunn rat lacks liver UDPglucuronyltransferase activity (EC. 2.4.1.17) for bilirubin, yet is able to produce bilirubin monoand diglucuronides when bilirubin dimethyl diester is used as substrate 131.In vitro gfucuronidation activities for other aglycones vary from markedly reduced to normal levels, when compared to the non-jaundiced (JJ) Gunn rat [4-61. A second unrelated enzyme, aminopyrine demethylase, a hepatic cytochrome P-450 axed-function oxidase (EC 1.14.14.1), is located in the same endoplasmic reticulum membrane as glucuronyl transferase and segregates in the same deficient fashion [7]. Glucuronyl transferase requires phospholipids to Abbreviation:

DPH,

1,6-diphenyl-i,3,5-hexatriene.

00052760/83/0000-0000/$03.00

0 1983 Elsevier Biomedical

Press

exhibit maximum activity [8]. A similar dependence on lipids is seen in the microsomal cytochrome P-450 enzymes [9]. It has been suggested that alterations in the microen~ronment of the microsomal matrix may be related to the enzyme activities demonstrated in Gunn rats [3]. The concomitant 45% decrease in phospholipids and 82% decrease in activity of rotenone-insensitive NADPH-cytochrome c reductase in microsomes prepared from Morris 7777 hepatoma cells, as compared with normal liver, exemplifies this possibility [lo]. With this in mind, the lipid composition of jaundiced and non-jaundiced Gunn rat liver microsomes was studied by direct measurement and fluorescence polarization techniques. Male Gunn rats were maintained on a standard high-fat diet and water ad libitum. Bilirubin glucuronyltransferase activity was determined using the assay of Strebel and Ode11 [ 1l] in fresh liver microsomes prepared in 0.15 M KCl. Microsomal protein was determined as previously described [7].

420

Lipids were extracted according to the method of Bligh and Dyer [ 121. Fatty acids were determined in an aliquot of this whole microsomal extract, using a Hewlett-Packard 5830A gas chromatograph after derivatization to their methyl esters [13]. Total phosphorus was determined using the method of Ames and Dubin [14]. Phospholipids were separated by the thin-layer chromatography method of Skipski et al. [15] into phosphatidylethanolamine, phosphatidylinositol plus phosphatidylserine and phosphatidylcholine. These phospholipids (Sigma Chemical Co., St. Louis) were used as standards. Phospholipid spots were developed by exposure to iodine vapor and quantitated after scraping by eluting into chloroform/methanol/H,0 (1 : 2 : 0.8) and measuring phosphorus as stated above. Fatty acid composition of the three phospholipid groups was determined as stated above after derivatization to their methyl esters. Cholesterol was measured in the microsomal suspension using the method of Zlatkis and Zak [16]. For fluorescence polarization determinations, the lipid-soluble fluorescent probe 1,6-diphenyl1,3,5-hexatriene (DPH, Aldrich Chemical Co., Inc., Milwaukee WI) was used as a 2 mM stock solution in tetrahydrofuran. In a typical experiment, microsomes (containing 50 pg protein/ml) in 0.1 M Tris-HCl buffer, pH 7.6, were incubated lo-15 min with DPH, 4. 10-j mM. The polarization of fluorescence, P, was measured using an Elscint model MV-la microviscosimeter (Elscint Inc., Hackensack, NJ). This instrument displays P = (I,, - I,)/(Z,, + I,), where I,, and I, are the fluorescence intensities oriented, respectively, parallel and perpendicular to the direction of polarization of the exciting light. Interpretation of this data utilizes the Perrin equation, which can be written: r”/r = 1 + 37/p, where Y” is the maximal limiting anisotropy (0.362 for DPH [17]), r is the anisotropy, 7 is the excited state fluorophore lifetime and p is the rotational relaxation time. Hence, at constant 7, the parameter ((r’/r) - I)-’ is directly proportional to the rotational relaxation time and provides a quantitative index of the resistance of the environment to rotational motion [ 181. The fluorescence polarization was expressed as the anisotropy parameter ((r’/r) - l)-‘, where r = (Z,, I, )/( I,, + 2 I, ). Fluorescence polarization was de-

termined from 5 to 40°C with readings obtained approximately every 2°C using fresh microsomes. Temperatures of the microsomal suspension were measured directly with a digital thermometer (Digitec model 5810), which reads to O.l”C. Since there is some debate about the shape of the resulting function when ln((r’/r)1))’ is plotted vs. reciprocal absolute temperature, four models were compared in their ability to fit the data obtained. These models were linear, two-phase linear, three-phase linear and quadratic. Linear and quadratic regressions determined these respective models. The best two- and three-phase linear models were determined using a computer program based on the theory developed by Hudson [19]. This program has been filed with the Biomedical Computing Technology Information Center, Vanderbilt Medical Center, Nashville, TN. In order to test the adequacy of each model, residual plots were made. F-tests were performed to decide if a two-phase was superior to a linear model and if a three-phase was superior to a two-phase model. Breakpoints and anisotropy parameters were compared in jj and JJ Gunn rats using the two-phase model. Mean values were compared between JJ and jj rats using Student’s t-test. The composition of liver microsomes from adult male JJ and jj Gunn rats with respect to total cholesterol, total phospholipid, protein and bilirubin glucuronyltransferase is shown in Table I. Also included is information regarding age and weight of these rats, since microsomal composition changes with development [20]. There is no significant difference in weight between these two groups, although the meanjj age is about 1 month greater. This reflects the fact that jaundiced Gunn rats sometimes do not grow as fast as the JJ group. There is no significant difference in protein, cholesterol, total phospholipid or cholesterol/ phospholipid molar ratio between the two groups. Since the cholesterol method employed in this measures total cholesterol (i.e., free study cholesterol and cholesterol esters), this result is higher than that determined by investigators using methods which measure only free cholesterol [20]. Our total phospholipid results are in agreement with similar measurements made in hepatic microsomes from other rats [lo].

421

TABLE

I

COMPOSITION

OF LIVER

MICROSOMES

IN VARIOUS

Results are expressed as mean f S.D. P values greater 10 animals in each group.

ADULT

MALE GUNN

than 0.05 are indicated

RATS

as n.s. (not significant).

Determinations

Significance

Rats

Age (days) Weight (g) Protein (mg/ml) Bilirubin glucuronyltransferase (nmol bilirubin conjugated/ 30 min per mg N)

Total phospholipid (pmol phosphorus/mg Cholesterol/phosphoIipid ratio

JJ

JY

135 + 18 434 k58 8.49 + 1.45

164 421 9.39

68.9

Total cholesterol (nmoI/mg protein) protein)

f 19.3

0

II

FATTY

ACID

* 35 *39 f 1.17

< 0.025 n.s. n.s.

+0

i 0.001

0.110+

0.015

O.llOk

0.014

ns.

0.444+

0.038

0.460 i

0.042

n.s.

0.248 f

0.035

0.239 + 0.027

ns.

COMPOSITION

OF LIVER MICROSOMES

An analysis of the major phospholipids in the liver microsomes of JJ and jj Gunn rats shows no significant difference in these three phospholipid groups. The fatty acid compositions of the three major phospholipid groups in liver microsomes from JJ andjj Gunn rats was also determined and reveals no striking differences. Both the relative amounts of the three phospholipid groups and the fatty acid composition of these groups are in good agreement with similar measurements in male

FROM

VARIOUS

ADULT

MALE GUNN

Results are expressed as mean* S.D. P values greater than 0.05 are indicated as n.s. (not significant). nine animals in each group. Individual fatty acids are shown as percentage of total fatty acids. Fatty

acid

Common

Palmitic Stearic Oleic Linoleic Eicosatrienoic Arachidonic Docosahexenoic Animal Animal

age (days) weight (g)

C number

: double bonds

16:O 18:O 18:l 18:2 20:3(n-6) 20:4 (n -6) 22:6 (n-3)

RATS

Determinations

Rats name

(P)

molar

An analysis of the fatty acid composition of JJ and jj Gunn rat liver microsomes is shown in Table II. Thejj group contains significantly greater amounts of stearic and arachidonic acid and significantly less palmitic acid. These differences in palmitic and arachidonic acid composition remain significant (Student’s t-test: P < 0.025), even if the individual group size is decreased to six animals each, thereby removing any significant difference in age or weight.

TABLE

were made on

were made on

Significance

JJ 17.43* 22.52 $10.13+ 17.165 1.75 + 24.38 k 6.63& 134 432

+I9 k61

0.7 1.01 0.99 1.77 0.85 2.02 1.01

15.70* 23.34 f 9.67* 16.57 + 1.58i 26.43k 6.71 + 167 423

1.13 0.74 0.71 1.09 0.52 1.14 0.78

+37 &41

< 0.025 < 0.05 n.s. n.s. ns. -z 0.01 n.s. < 0.025 n.s.

(P)

422

TABLE

III

FLUORESCENCE

POLARIZATION

DATA

FROM

LIVER

MICROSOMES

OF VARIOUS

ADULT

MALE GUNN

RATS

Results are expressed as mean kS.D. using the two-phase linear model. P values greater than 0.05 are indicated as n.s. (not significant). Determinations were made on nine animals in each group. DPH was used as the fluorescent probe, as described in the methods section. Rats

Significance

JJ Breakpoint ((r”/r)-

temperature 1))’

At 25’C At 37°C

(“C)

23.4

(P)

JY k 3.1

0.623 + 0.03 1 0.43 1 * 0.028

Wistar rats [21]. Hence, these data are not presented. The Gunn strain of rat was originally developed from the Wistar rat. Fluorescence polarization when plotted logarithmically as a function of reciprocal absolute temperature may be linear or may exhibit discontinuities or breakpoints. The curvilinearity of such plots has also been noted [22,23]. In order to assess our data, we used a computer to fit a linear, two-phase linear, three-phase linear and quadratic model to the data. Residual plots, used to assess model adequacy, demonstrated non-random patterns in the linear model. The other three models had random residual plots. An F-test to compare linear and two-phase linear models showed that the two-phase model significantly improved the closeness of fit in 17/18 data sets. Using the F-test to compare the two- and three-phase linear models revealed that the three-phase linear model significantly improved closeness of fit in only 3/18 rats. We are unable to compare closeness of fit between the two-phase linear and quadratic models. However, we find the residual sum of squares (RSS) was lower for the two-phase linear model in 16/18 rats. Mean square (RSS/(number of observations _ number of parameters)) was lower for the twophase model in 7/9 JJ rats and 7/9jj rats. Hence, it appears that a two-phase linear model (i.e., one breakpoint) most closely fits these data. Table III shows fluorescent polarization data in JJ and jj Gunn rats, using the two-phase linear model. There is no significant difference in breakpoint temperature between these two groups. However, the anisotropy parameter (( r’/r) - 1)) ’

26.6

+4.1

0.658 k 0.029 0.474 5 0.024

Il.%

< 0.025 < 0.005

in the jj rat is significantly higher than that in the JJ rat at both 25 and 37°C. This indicates that the microsomal membrane from jj rats is significantly less fluid than that from JJ rats. Although the fluorophore we used, DPH, does bind to proteins [24], it is generally accepted that in subcellular membrane preparations, DPH binds to the hydrophobic lipid-rich interior of the membrane, giving a general picture of the state of the bilayer [25]. Fatty acid composition has been shown to significantly affect fluorescence anisotropy determinations with DPH [26]. Hence, it is possible that the observed differences in fatty acid composition between JJ and jj Gunn rats were responsible for the different anisotropy parameters. In explaining the enzyme deficiencies of the Gunn rat (and presumably of the Crigler-Najjar syndrome, of which the Gunn rat is a model) [27], several hypotheses have been suggested. These data indicate that the microsomal membrane composition of jj and JJ Gunn rats is different in fatty acid composition and membrane fluidity. It is possible that this difference may be secondary to prolonged hyperbilirubinemia. It is also possible that these measurements are part of a microsomal matrix abnormality which is itself the primary defect in the Gunn rat. We thank Dr. Milton B. Yatvin, University of Wisconsin, Department of Human Oncology, for use of his microviscosimeter and John Vorpahl and James Freimuth for technical assistance. This research was supported by NIH grants 5 F32HD05873 and AM 21668.

423

References 1 Katchalski, E., Silman, I. and Goldman, R. (1971) Adv. Enzymol. 34, 445-536 2 DePierre, J.W. and Ernster, L. (1977) Annu. Rev. B&hem. 46, 201-262 3 Odell, G.B., Cukier, J.O. and Gourley, G.R. (1981) Hepatology 1, 307-3 15 4 Arias, I.M. (1961) Biochem. Biophys. Res. Commun. 6, 81-84 5 Javitt, N.B. (1966) Am. J. Physiol. 211, 424-428 6 Puuka, R., Tanner, P. and Hanninen, 0. (1973) B&hem. Genet. 9, 343-349 7 Gourley, G.R., Mogilevsky, W. and Odell, G.B. (1982) Biochem. Pharmacol. 31, 1792-1795 8 Berry, C.S., Allistone, J. and Hallinan, T. (1978) Biochem. Biophys. Acta 507, 198-206 9 Wade, A.E. and Norred, W.P. (1976) Fed. Proc. 35, 2475-2479 10 Hostetler, K.Y., Zenner, B.D. and Morris, H.P. (1976) B&him. Biophys. Acta 441, 231-238 11 Strebel, L. and Odell, G.B. (1971) Pediatr. Res. 5, 548-559 12 Bligh, E.G. and Dyer, W.J. (1959) Can. J. B&hem. Physiol. 37, 911-917 13 Morrison, W.R. and Smith, L.M. (1964) J. Lipid Res. 5, 600-608

14 Ames, 15 16 17 18 19 20 21 22 23 24 25 26

B.N. and Dubin,

D.T. (1960) J. Biol. Chem.

235,

769-775 Skipski, V.P., Peterson, R.F. and Barclay, M. (1964) Biothem. J. 90, 364-378 Zlatksi, A. and Zak, B. (1969) Anal. Biochem. 29, 143-148 Shinitzky, M. and Barenholz, Y. (1974) J. Biol. Chem. 249, 2652-2657 Schachter, D. and Shinitzky, M. (1977) J. Clin. Invest. 59, 536-548 Hudson, D.J. (1966) J. Am. Stat. Assn. 61, 1097-l 129 Kapitulnik, J., Tshershedsky, M. and Barenholz, Y. (1979) Science 206, 843-844 Davison, S.C. and Wills, E.D. (1974) Biochem. J. 140, 46 l-468 Londesborough, J. (1980) Eur. J. Biochem. 105, 21 I-215 Brasitus, T.A., Tall, A.R. and Schachter, D. (1980) Biochemistry 19, 125661261 Mely-Goubert, B. and Freedman, M.H. (1980) Biochim. Biophys. Acta 601, 315-327 Wharton, S.A., DeMartinez, S.S. and Green, C. (1980) Biochem. J. 191, 785-790 Schroeder, F. and Goh, E.H. (1980) Chem. Phys. Lipids 25, 207-224

27 Cornelius, 369-372

C.E. and Arias,

I.M. (1972) Am. J. Pathol.

69,