organization of functional groups of liver bilitranslocase

Biochimica et Biophysica Acta, 1041 (1990) 195-200 195 Elsevier

BBAPRO 33776

Organization of functional groups of fiver bilitranslocase

Sabina Passamonti and Gian Luigi Sottocasa Dipartimento di Biochimica Biofisica e Chimica delle Macromolecole, Unioersitd degli Studi di Trieste, Trieste (Italy)

(Received 25 June 1990)

Key words: Bilitranslocase; Bilirubin transport; (Liver)

Bilitranslocase transport activity can be described as consisting of three functional fractions, which depend on two distinct classes of sulfhydryl groups, on the one hand, and on the guanido groups of arginine residues, on the other. Each fraction accounts for approx. 50% transport activity. The pattern of transport activity inhibition resulting from step-wise derivatization of these functional groups indicates that, in general, derivatization of arginine residues prevents that of one class of sulfhydryl groups and vice versa, indicating their close location in the protein. Nevertheless, under appropriate conditions, derivatization of both functional groups can be achieved; however, the inhibitory effect produced is not additive. Hence, these two fractions overlap functionally and are likely to belong to a common functional domain of the protein. On the contrary, the other class of sulfhydryl groups can be derivatized, regardless of the state of the arginine residues.

Introduction

Bilitranslocase is a plasma membrane cartier protein confined to the sinusoidal pole of the liver cell [1-4]. Its function is to mediate transport of bilirubin [5] and functional analogues, such as sulfobromophthalein (BSP) [4,6,7] and Thymol blue [8], from plasma into the cell.

Bilitranslocase activity is measured in vitro as the rate of valinomycin/potassium-induced BSP uptake by plasma membrane vesicles [4].

Using specific functional group reagents, this parameter has been shown to be determined by two functional groups, the sulfhydryl groups of cysteine residues [9,10] and the guanido groups of arginine residues [11].

On the basis of their reactivity with the reagent 5,5'-dithiobis(2-nitrobenzoate) (DTNB) [9], the sulfhydryl groups have been distinguished in two classes, one DTNB-sensitive and the other DTNB-insensitive. Each of the two accounts for approx. 50% of transport activity. Interestingly, arginine residues account also for approx. 50% of the activity [11].

Abbreviations: DTNB, 5,5'-dithiobis(2-nitrobenzoate); 2-ME, 2-mercaptoethanol; NEM, N-ethylmaleimide; pHMB, p-hydroxymercury- benzoate; BSP, sulfobromophthalein.

Correspondence: Gian Luigi Sottocasa, Dipartimento di Biochimica Biofisica e Chimica delle Macromolecole, Universit/~ degli Studi di Trieste, via A. Valerio 32, 1-34127 Trieste, Italy.

The fact that BSP transport activity in plasma membrane vesicles can be described in terms of three functional fractions, i.e., one dependent on DTNB-sensitive sulfhydryl groups, another on DTNB-insensitive ones, and the third on arginine residues, poses the problem of how these fractions overlap.

The two classes of sulfhydryl groups controlling the whole of BSP transport activity in plasma membrane vesicles, though apparently unrelated, have indeed been shown to belong to the same cartier system [9].

Needless to say that there is no doubt concerning the link between the arginine residues and bilitranslocase transport activity [11].

Thus, it should be particularly interesting to assess the inhibition of bilitranslocase transport activity resulting from the concomitant modification of arginine residues and either one or both classes of sulfhydryl groups. This would answer the general question whether the functional groups reactivities can be affected mutually.

In previous work [9,10], this approach provided val- uable information regarding the function of native bilitranslocase.

In this work, the same approach was expected to enlarge the knowledge about this protein. Data in this work show that arginine residues and DTNB-sensitive sulfhydryl groups are quite close in the protein. From the functional point of view, their respective fractions of activity have been found to overlap completely, suggesting the possibility that they belong to a common domain of the protein. The DTNB-insensitive class of sulfhydryl groups accounts for the residual functional

0167-4838/90/$03.50 © 1990 Elsevier Science Publishers B.V. (Biomedical Division)

196

fraction of activity, that can be suppressed by selective modification.

Materials and Methods

Solutions of the reagents used were: 0.44 M phenylglyoxal (Sigma, St. Louis, MO, U.S.A.), freshly dissolved in ethanol/water (1:1, v/v); 0.6 M methylglyoxal (Sigma) in water; 25 mM DTNB (Sigma) in 0.1 M Hepes (pH 7.4); in the experiment shown in Fig. 4 (panel A, inset) 45 mM DTNB was dissolved in pyri- dine; 3.9 mM pHMB (Sigma) in 1 mM NaOH; 33 mM NEM (Sigma) dissolved in water; 4.0 mM cupric sulfate (Carlo Erba, Milano, Italy) in 0.1 M sodium-potassium tartrate (pH 7.5); 1.1 M arginine (Merck AG, Darm- stadt, F.R.G.) in water; 0.4 M 2-mercaptoethanol (BDH, Poole, U.K.), dissolved in water; BSP (Serva, Heidel- berg, F.R.G.) in water. All other reagents are commer- cially available products.

Vesicles were prepared, stored and utilized as described in Ref. 8.

Inhibition of electrogenic BSP transport by functional group reagents was obtained by step-wise reaction of the latter with vesicles. The inactivation reactions were started by the addition of 0.17 vol. reagent to a tube containing 0.83 vol. vesicles (10.2 + 1.04 mg protein/ml), already equilibrated at 36 ° C. The second addition was of 0.03 vol. of the second reagent. It was checked that the solvents of the reagent solutions, in the comparatively highly buffered suspension, did not in- fluence per se the results.

Reactivations were obtained by adding 2-ME and/or arginine in a minimal, negligible volume. Additions of volumes less than 0.003 ml were made with a calibrated microsyringe.

Reactions were stopped by diluting 25 /xl samples, withdrawn from the pre-incubation mixture, in the transport assay medium.

Measurements of BSP electrogenic transport activity were carried out by the spectrophotometric technique described in Ref. 4 as applied in Ref. 8. The test was started by adding 25/~1 (= 0.255 mg protein) of pre-incubated vesicles to 1.975 ml transport assay medium, composed of 0.1 M potassium phosphate buffer (pH 8.1) and 15-25 /~M BSP. Under these conditions, the degree of inactivation by the various reagents was found to be always the same. The valinomycin-induced uptake phase was started by adding 2 /lg of the ionophore dissolved in 2/~1 of methanol [8], at room temperature. The wavelength pair was 580-514.4 nm. As in previous work [4,8-11], the activity of bilitranslocase has been measured by the initial rate of BSP uptake induced by valinomycin, in the presence of potassium. All the data presented in this paper are expressed as percent change of this parameter.

Q. )

I(00- CO

co

~- 50-

O_ U? m

A . . . . . . . . q00-~

50- PhG

B

DINB

I I I I I I

5 115 25 ' 5 1'5 2'5 rain rain

Fig. 1. Time-course of inhibition of BSP transport activity by DTNB (A) and by phenylglyoxal (B) and effect of the addition of either phenylglyoxal (A) or DTNB (B). Experimental conditions of pre-incubation: 40 mM Hepes (pH 7.4), 60 mM NaC1, 10.2_+1.04 mg protein/ml (o . . . . . . o). (A) 2.1 mM DTNB (o); at 10 min, addition of 20 mM phenylglyoxal (PhG) (e). (B) 20 mM phenylglyoxal (PhG) (o); at 15 rain (arrow), addition of 2.1 mM DTNB (e); T = 36°C. At the time intervals indicated, 25/t l samples were withdrawn from the preincubation mixture, diluted in 2.0 ml transport assay medium and tested for BSP uptake rate, as described under Materials and

Methods.

Protein determination was performed by the Bio-Rad protein assay, taking y-globulin (standard I) as the standard.

Results

Previous work has shown that the maximal inhibitory effect of either DTNB (a sulfhydryl group reagent) [9,10] or phenylglyoxal (an arginine residue reagent) [11] on bilitranslocase activity in rat liver plasma membrane vesicles is no greater than 50%. The minimal effective concentrations are 1.5 and 20 mM, respectively.

The experiments shown in Fig. 1 are aimed at esti- mating the level of inhibition of BSP transport activity in plasma membrane vesicles obtained by two-step reaction with DTNB and phenylglyoxal.

Fig. 1A shows that addition of 2.1 mM DTNB to plasma membrane vesicles inhibits BSP transport activity by approximately 50%, a value that cannot be lowered by subsequent treatment with 20 mM phenylglyoxal. The same picture is obtained when BSP transport activity is inhibited first by phenylglyoxal and then by DTNB (Fig. 1B). In both cases, it has been confirmed that the additions of the second reagent to vesicles pre-incubated in the absence of the first reagent did display its typical inhibitory effect (not shown).

These results indicate that the effect of the two reagents is not additive.

Two possibilities are open: (a) the second reagent cannot react because its target is no longer available for derivatization (no matter why: steric hindrance, conformational changes etc.); (b) the second reagent has reacted with the functional group, but this derivatization has no functional relevance any more.

To discriminate between the two alternatives, the experiment reported in Fig. 2 has been carried out. It

197

priM

O_ 1 phG 2-ME 03 []

I I I I I I I I 5 15 25 35 min

Fig. 2. Time-course of inhibition of BSP transport activity by pHMB. Effect of the additions of phenylglyoxal, 2-ME and arginine. Experi- mental conditions of pre-incubation: 0.3 mM pHMB (o); at 1 min, addition of 20 mM phenylglyoxal (PhG) (o); at 16 min addition of 1.0 mM 2-ME (A); at 18 min and 25 rain, additions of 30 mM arginine

(Arg) (11). Other details as in Fig. 1.

shows that treatment of vesicles with 0.3 mM p-hy- droxymercurybenzoate (pHMB) brings about 50% inhibition of BSP transport activity, as expected. It has already been reported [9,10] that, under these conditions, DTNB-sensitive sulfhydryl groups are derivatized. Similarly to what described in Fig. 1A, subsequent treatment with phen'ylglyoxal does not alter the level of inhibition. 2-ME promotes the rapid, full recovery of transport activity, which, however, is only transient. A slow, progressive decay of activity ensues soon after. Arginine addition, causing phenylglyoxal dissociation from arginine residues in proteins [11,12], either prevents activity from decaying, when added at the peak of recovery, or allows it to recover, when added at a time of plain depression.

The fact that transport activity can be restored by addition of 2-ME shows two facts, i.e.: (a) inhibition is caused only by pHMB, since 2-ME cannot displace phenylglyoxal from arginine residues [11], and (b) the arginine residues have not reacted with phenylglyoxal, otherwise 2-ME would have promoted reduction of sulfhydryl groups without promoting recovery of activity, because of persistent arginine derivatization.

The slow decay of activity seen after 2 rain-treatment with 2-ME proceeds with a time-course consistent with that of phenylglyoxal inhibition of BSP transport activity (compare with Fig. 1B). The effect of arginine con- firms that inhibition is due to the reaction of phenylglyoxal with the arginine residues. This starts only when pHMB has been removed from DTNB-sensitive sulfhydryl groups.

This experiment shows that the first alternative holds true, namely that the second reagent has no access to arginine residues.

In order to establish whether the same happens when the order of the additions is reversed, the experiment in Fig. 3 has been run, where phenylglyoxal is the first

reagent used. BSP transport activity is shown to fall to 50% thereupon. N EM is then added, at the concentra- tion of 1.2 mM, which has been shown to react only with the DTNB-sensitive class of sul fhydryl groups [9,10]. Following this treatment, no further inhibition ensues, similarly to what can be seen in Fig. 1, panel B. After that, 2-ME is added in order to remove unreacted NEM. It might be recalled that 2-ME does not cleave the thioetheric bond of any possible NEM-derivative of sulfhydryl groups [13]. Subsequent arginine addition causes BSP transport activity to rise again.

This result shows that only arginine residues were derivatized (and subsequently set free by arginine), while DTNB-sensitive sulfhydryl groups were prevented from reacting with NEM.

The data of Figs. 2 and 3 allow to conclude that derivatization of the DTNB-sensitive sulfhydryl groups bars that of the arginine residues and vice versa. The basis of this occurrence could, however, be mere mutual steric hindrance by the relatively large reagents used. The experiment of modification of DTNB-sensitive sulfhydryl groups prior to that of arginine residues has therefore been repeated, using the small cupric ion, instead of pHMB, as a sulfhydryl group reagent [9]. Fig. 4A shows that 0.4 mM cupric sulfate cuts down transport activity by 50%. This is to be ascribed to the modification of DTNB-sensitive sulfhydryls, since further addition of DTNB does not deepen the inhibition level, as shown by the experiment in the inset. Subse- quent treatment with phenylglyoxal is ineffective and 2-ME reestablishes transport activity. In the experiment shown in Fig. 4B, the arginine residue reagent methylglyoxal (which has been shown to behave as phenylglyoxal in BSP transport inhibition [11]) substitutes for the larger phenylglyoxal. Nevertheless, methylglyoxal treatment does not affect the half-maximal inhibition of

t00 -~ / $C0 i PhG

m 4~ o_ 50- -, / \

m0313- I~EM 2-ME

/ I I I I 15 25 rain

Fig. 3. Time-course of inhibition of BSP transport activity by phenylglyoxal (PhG). Effects of NEM, 2-ME and arginine. Experimental conditions of pre-incubation: 20 mM phenylglyoxal (©); at 15 min, addition of 1.2 mM NEM (O); at 17 min, addition of 1.2 mM 2-ME (*); at 18 min, addition of 20 mM arginine (11). Other details as

in Fig. 1.

198

100~--~ ++o - - - - -o~ - - -~ . . .o ':"~l'~cu "+ 100~1~ . . . . . .,o_ ~ ~ _o.. ~ _ __c

5o- ~ ~ o 50-

° ~ L i i r I i i

5 15 25 35 5 15 25 35 min min

lOO . . . . . . . . . o

50- (3. MeG 2 - M E 03

c rn ae

i 5 15 25

rnin

Fig. 4. Time-course of inhibition of BSP transport activity either by cupric sulfate (A and B) or pHMB (C). Effect of either phenylglyoxal (A) or methylglyoxal (B) and (C). Experimental conditions of pre-incubation: (A) 0.4 mM cupric sulfate (o); at 1 rain, addition of 20 mM phenylglyoxal (O); at 15 rain, addition of 1.3 mM 2-ME (A); inset: 0.4 ram cupric sulfate (o); at 1 rain, addition of 1.5 raM DTNB (O). (B) 0.4 mM cupric sulfate (o); at 1 min, addition of 20 mM methylglyoxal (MEG) (O); at 15 rain, addition of 1.3 mM 2-ME (A); at 19 rain, addition of 34.8 mM arginine (I). (C) 0.3 mM pHMB (o); at 1 rain, addition of 20 raM rnethylglyoxal (O); at 15 rain, addition of 1.4

raM 2-ME (A). Other details as in Fig. 1.

transport activity reached by prior reaction with 0.4 mM cupric sulfate. Interestingly, though, in this case, 2-ME does not spur the recurrent restoration of transport activity, which can be induced only after addition of free arginine. These results show that the second modification, though achievable under these conditions, brings about no additional functional modification. This picture cannot be reproduced using pHMB instead of cupric sulfate, as shown in Fig. 4C, indicating that methylglyoxal can no longer have access to arginine residues when the DTNB-sensitive sulfhydryl groups are modified by a reagent larger than cupric ion.

In order to investigate the relationship between arginine residues and DTNB-insensitive sulfhydryl groups, experiments meant at testing the possibility to obtain full inhibition of BSP transport activity through derivatization of both functional groups have been performed.

DTNB-insensitive sulfhydryl groups can be selec- tively derivatized with NEM, provided that the other class has already been modified by another reagent [9]. This is achieved using 0.3 mM pHMB (as in Fig. 2) prior to NEM addition. After NEM has reacted, 2-ME will set free the DTNB-sensitive class of sulfhydryl groups [9]. The experiment in Fig. 5 shows that selective modification of DTNB-insensitive sulfhydryl groups with NEM depresses activity to 50%. Subsequent addition of phenylglyoxal is shown to cause activity to fall again, indicating that modification of DTNB-insensitive

L l O 0 - t , ~ - o - - o - - - - < > < > - - - - < > - - - - o !

co NEM PhG

5 0 - o_ ,J ' U? m

i i I i ' x ' \ I I

5 1~5 25 35

min Fig. 5. Time-course of inhibition of BSP transport activity by pHMB. Effect of NEM, 2-ME, pheny!glyoxal and arginine. Experimental conditions of pre-incubation: 0.3 mM pHMB (<3); at 1 min, addition of 3.0 mM NEM (O); at 3 rain, addition of 0.5 mM 2-ME (A); at 6 min, addition of 20 ram phenylglyoxal (D); at 21 rain, addition of 40

mM arginine (I). Other details as in Fig. 1.

sulfhydryl groups does not prevent the arginine residues from reacting with phenylglyoxal. The involvement of arginine residues is confirmed by the fact that final arginine addition causes transport activity to recover.

Fig. 6 shows an experiment where the possibility to obtain full inhibition of BSP transport activity by two- step reaction of arginine residues and the DTNB-sensi. tive class of sulfhydryl groups is attempted. This experiment is the same as that shown in Fig. 5, except for the order of the reagents addition. 20 mM phenylglyoxal causes 50% inhibition of transport activity. Addition of 3.0 mM NEM brings down transport activity to 20%. 3.0 mM 2-ME is added, in order to get rid of free NEM. Arginine is added, in order to reconstitute the arginine residues. As shown, this raises transport activity to 50$. The final addition of 1.2 mM NEM brings down transport activity again.

0..

03 CO

100..4

5 0 -

""" PhG

15 25 315

min Fig. 6. Time-course of inhibition of BSP transport activity by phenylglyoxal. Effect of NEM, 2-ME, arginine and NEM again. Experimen- tal conditions of pre-incubation: 20 mM phenylglyoxal (o) , at 15 min, addition of 3.0 raM NEM (e); at 17 rain, addition of 3.0 mM 2-ME (A); at 19 rain, addition of 20 mM axginine ((3); at 33 rain,

addition of 1.2 raM NEM (I). Other details as in Fig. 1.

The fact that 3.0 mM NEM depresses transport activity even after phenylglyoxal inhibition shows that DTNB-insensitive sulfhydryl groups have been derivatized, since the DTNB-sensitive ones are protected by phenylglyoxal derivatization of arginine residues (see Fig. 3). Moreover, addition of 1.2 mM NEM after arginine-induced recovery of a fraction of activity suc- ceeds in bringing down transport activity. This proves again that there are free sulfhydryl groups available for modification, belonging to the DTNB-sensitive class. The other class, in fact, has undergone early modification by 3.0 mM NEM. Data are consistent with the view that modification of both DTNB-insensitive sulfhydryl groups and arginine residues can coexist.

Discussion

Previous work has shown that the rate of valinomycin/potassium-induced BSP transport in plasma membrane vesicles is the specific measure of bilitranslocase activity [3,4]. This parameter has later been shown to be affected by specific reagents of either sulfhydryl groups and arginine residues [9-11]. The fact that bilirubin (at nanomolar concentrations), BSP and thymol blue (at micromolar concentrations) can protect both classes of sulfhydryl groups and arginine residues from selective modification [10-11] indicated that: (a) the reagents used target bilitranslocase; and (b) that bilitranslocase transport activity is not affected by chemical modification of other plasma membrane proteins.

The mutually excluding modification of either DTNB-sensitive sulfhydryl groups and arginine residues responsible for fractions of bilitranslocase transport activity suggests their close proximity. It is reasonable to postulate that they not only belong to the same carrier system, but, even, that they lie very close in a portion of the protein involved in transport function. If one favors only the simplest explanation, namely that plain steric hindrance is responsible of the effects seen in the experiments shown in Fig. 4, it can be guessed that the distance between the DTNB-sensitive sulfhydryl groups and the arginine residues is no greater than a benzene ring, since this substituent can be encumbrant enough to protect the adjacent functional group from reagent's attack.

On the contrary, neither DTNB-insensitive sulfhydryl groups nor arginine residues hinder each others' reaction with selective reagents, suggesting that they are set somewhat further apart.

Fig. 4B shows a quite fortunate case, where derivatization of both DTNB-sensitive sulfhydryl groups by cupric ion and arginine residues by methylglyoxal con- cur. The interesting point here is, however, that the coexisting modifications bring about no additive inhibitory effect on transport activity. Thus, the first modification involved a functional perturbation of the protein,

199

perhaps of conformational nature, impairing transport activity by 50%, but subsequent chemical modification of an adjacent residue is ineffective. This could even be an indication that DTNB-sensitive sulfhydryl groups and arginine residues are not just close to each other, but may reside in the same functional domain of the protein. The chemical modification of this domain, either with reagents specific for sulfhydryl groups or for arginine residues, introduces a functional modification of the protein, so that its performance is fixed to half-maximal efficiency. The residual activity is dependent on the DTNB-insensitive class of sulfhydryl groups only, and can be suppressed by their chemical modification.

It can therefore be concluded that two of the three functional fractions, each of which accounts for 50% transport activity, overlap. DTNB-sensitive sulfhydryl groups and arginine residues share the same functional fraction, which, in turn, is expected to be accounted for by a definite domain of the protein.

The fact that modification of DTNB-sensitive sulfhydryl groups by either the small cupric ion or the large DTNB or pHMB ends in the loss of same fraction of bilitranslocase activity suggests that the size of these reagents is not the crucial factor of inhibition. Rather, the latter seems to be due to a functional modification of bilitranslocase, perhaps triggered by breaking the interaction of DTNB-sensitive sulfhydryl groups with some other site in the protein.

Further work focused on the structural arrangement of this protein will ascertain the validity of these predic- tions.

Acknowledgements

Supported by grants from Ministero della Pubblica Istruzione and Fondo per lo Studio e la Ricerca Scienti- fica delle Malattie del Fegato. Thanks are due to Mr. B. Gazzin for graphical work, and to Prof. L. Ernster for stimulating discussion.

References

1 Tiribelli, C., Lunazzi, G.C., Luciani, M., Panfili, E., Gazzin, B., Liut, G.F., Sandri, G. and Sottocasa, G.L. (1978) Biochim. Bio- phys. Acta 532, 105-112.

2 Lunazzi, G.C., Tiribelli, C., Gazzin, B. and Sottocasa, G.L. (1982) Biochim. Biophys. Acta 685, 117-122.

3 Miccio, M., Baldini, G., Basso, V., Gazzin, B., Lunazzi, G.C., Tiribelli, C. and Sottocasa, G.L. (1989) Biochim. Biophys. Acta 981, 115-120.

4 Baldini, G., Passamonti, S., Lunazzi, G.C., Tiribelli, C. and Sot- tocasa, G.L. (1986) Biochim. Biophys. Acta 856, 1-10.

5 Sottocasa, G.L., Tiribelli, C., Luciani, M., Lunazzi, G.C. and Gazzin, B. (1979) in Function and Molecular Aspects of Biomem- brane transport (Quagliariello, E. et al., eds.), pp. 451-458, Else- vier/North Holland, Amsterdam.

200

6 Sottocasa, G.L., Baldini, G., Sandri, G., Lunazzi, G.C. and Tiri- belli, C. (1982) Biochim. Biophys. Acta 685, 123-128.

7 Persico, M. and Sottocasa, G.L. (1987) Biochim. Biophys. Acta 930, 129-134.

8 Passamonti, S. and Sottocasa, G.L. (1988) Biochim. Biophys. Acta 943, 119-125.



11 Passamonti, S., Battiston, L. and Sottocasa, G.L. (1990) Biochim. Biophys. Acta 1025, 122-126.

12 Takahashi, K. (1968) J. Biol. Chem. 243, 6171-6179. 13 Riordan, J.F. and Vallee, B.L. (1972) Methods Enzymol. 45,

449-456.

organization of functional groups of liver bilitranslocase

Documents