solute cotransporter family

BIOCHIMICA ET B1OPHYSICA ACTA

ELSEVIER

Biochimica et Biophysica Acta, 1365 (1998) 60-64

BB3

Topology and function of the Na ÷/proline transporter of Escherichia coli, a member of the Na ÷/solute cotransporter family Heinrich Jung* Universitiit Osnabriick, Fachbereich Biologie/Chemie, Abteilung Mikrobiologie, D-49069 Osnabriick, Germany Received 27 January 1998; received in revised form 4 March 1998; accepted 4 March 1998

Abstract The Na ÷/proline transporter of Escherichia coli (PutP) is a member of the Na +/solute cotransporter family (SCF) and catalyzes the uptake of proline by a Na + dependent transport mechanism. Hydropathy profile analysis suggests that the protein consists of 12 transmembrane domains (TMs) that traverse the membrane in zigzag fashion connected by hydrophilic loops. However, analysis of a series of putP-phoA (PutP-alkaline phosphatase) and putP-lacZ (PutP-[3galactosidase) fusions and site-directed labeling of the transporter indicate a 13-helix motif with the N-terminus on the outside and the C-terminus facing the cytoplasm. The findings are discussed with respect to a common topological motif for all members of the SCF. Furthermore, amino acid substitution analysis indicates that the N-terminal part of PutP is important for ion binding. Thus, Asp55 (putative TM II) is essential for transport and proposed to interact directly with Na +. The functional importance of TM II is further confirmed by the observation that replacement of Arg40, Ser50, Ala53, or Ser57 alters transport kinetics dramatically. © 1998 Elsevier Science B.V.

Keywords: Secondary transport; Protein topology; Site-directed mutagenesis; Protein labeling; Gene fusion

1. Introduction

Accumulation of a variety of solutes against a concentration gradient is driven by transmembrane electrochemical H + or Na ÷ gradients [1]. Functional analysis and sequence comparison indicate that the membrane proteins catalyzing these transport reactions fall into different families [2,3]. The Na÷/ solute cotransporter family (SCF) encompasses more than 35 homologous proteins from Archaea, Bacteria, Yeast, Insects, and Mammals [4]. Certain members of this family have been implicated in human diseases (i.e., glucose/galactose malabsorption, thyroid dis*Fax: +49-541-9692870; E-maih [email protected]

orders) [5,6]. Kinetic analyses have resulted in the proposal of an ordered binding mechanism for Na +coupled solute transport [7,8]. However, information on protein structure at the atomic level is not available yet. Finally, the molecular mechanism by which these proteins use energy stored in electrochemical ions gradients for substrate accumulation remains enigmatic. In order to gain new insights into this mechanism our attention is focused on the Na +/proline transporter of Escherichia coli (PutP), a member of the SCF. The transporter is an integral protein of the cytoplasmic membrane that utilizes free energy released from downhill transport of Na + to drive accumulation of proline [8-10]. The putP gene encoding the Na÷/proline transporter has been cloned and se-

0005-2728/98/$19.00 © 1998 Elsevier Science B.V. All rights reserved. PII: S0005-2728 (98)00044-9

61

H. Jung I Biochimica et Biophysica Acta 1365 (1998) 60-64

served within the members of the SCF, with a less polar amino acid (i.e., Cys) stabilized PutP in an apparently Na + independent high affinity conformation for proline (Quick and Jung, manuscript in preparation). This phenomenon was accompanied by only marginal proline uptake rates (about 1% of wild-type Vmax). The apparent independence of proline binding from the Na + concentration could at least partially be attributed to an enhanced Na ÷ affinity of Asp187-->Cys-PutP. In addition, reaction of PutP containing a single Cys at position 187 with N-ethylmaleimide (NEM) was inhibited by Na + but not by proline. Based on these results it was suggested that Asp187 is located close to the pathway of the coupling ion through the membrane and may, i.e., be important for the release of Na ÷ on the cytoplasmic side of the membrane. A reduced Na ÷ dependence of proline binding was also observed upon replacement of Arg257 in PutP with Cys [16]. However, the effect of the substitution of Arg257 on proline uptake was much less dramatic as in case of Asp187. In the human Na÷/glucose transporter (SGLT1), like PutP a member of the SCF, replacement of Asp176 by Ala altered the kinetics of charge transfer while Asn at this position had no effect on charge movement. It was suggested that removal of the polar side chain caused a change in the rate constants for a

quenced [11,12], and the gene product has been solubilized from the membrane, purified, reconstituted into proteoliposomes, and shown to be solely responsible for Na +/proline transport [13,14]. Here I describe results of recent studies focused on the membrane topology of PutP and the identification of amino acids that are of functional importance.

2. Amino acids of functional importance Spontaneous and site-directed mutagenesis have been utilized in an attempt to identify amino acid residues in PutP involved in ligand binding and transport. Out of five acidic residues in the N-terminal domain analyzed Asp55 proved to be essential for transport while Glu75, Asp33, and Asp34 were shown to be dispensable for function [15] (cp. Figs. 1 and 3 for location of residues). Kinetic analysis of active transport catalyzed by Asp55--->Glu-PutP revealed a 50-fold decrease of the apparent affinity of the protein for Na ÷ ions compared to the wild-type transporter. On the other hand, only a relatively small alteration of the apparent affinity for proline was observed. These results suggested that Asp55 is located at or close to a binding site of the coupling ion [15]. Replacement of Asp187, a residue that is con-

SE rv~

PERIPLASM

L2

L4

I(I~(~GI~I ~

TFG S E rv~ F Y L E

L6 KQKS

lV

L

S

GAV

IEN

A NQNA ~g~/

pA H D

V D

I ~ 1 r~-I r---=~ ~

F

L/0 EN P R N V

/.12 F

q4tX, ~;[l a,

1~_

~

! M(NHO N L1 wp RLRGAV

%

~G~ ~

S KDE Fit

V~EyNNNAL yFT LPD L3

DIs7 L5

pQ IH L

HV HSI

T

LYK L7

L9

E K QS HA

N ~ '~ R R p MT AA¢ LIlF.,AFRKQM

YHSA

I,COurl)

L13

CY'I~PLASM Fig. 1. Secondary structure model of the Na ÷/proline transporter based on hydropathy analysis of the amino acid sequence [11]. Transmembrane domains

are representedas rectanglesand numberedwith Roman numerals. Arabic numeralscorrespondto hydrophiliccytoplasmicand periplasmic loops. Functionalimportantaminoacid residuesare highlighted.Positionsused in the proteinlabelingstudyare encircled.

62

H. Jung / Biochimica et Biophysica Acta 1365 (1998) 60-64

partial step in the reaction cycle, i.e., the conformational changes of the unloaded transporter [7]. Besides acidic residues, Ser57 in PutP was found to be critical for transport [17]. Characterization of PutP containing Ala, Cys, Gly, or Thr in place of Ser57 revealed a drop of the apparent affinity for Na ÷ and proline up to two orders of magnitude with little effect on Vmax. It was proposed that the Ser residue is located at or close to the binding site(s) for Na ÷ and/or proline. The functional importance of this region in PutP was further confirmed by the observation that replacement of Arg40, Ser50, or Ala53 dramatically altered transport kinetics (Quick, Leifker and Jung, unpublished information). The results discussed above are consistent with the idea that cation binding is determined by interactions with the N-terminal part of the transporters. This suggestion is in agreement with the finding that a cluster of mutations that alter the Li ÷ sensitivity of Salmonella typhimurium was found in the 5' region of putP [18]. Also in the melibiose transporter of E. coli (MelB), a transport protein which is not homologous to PutP, amino acid residues in the N-terminal domain are implicated in Na ÷ binding [19]. The C-terminal part of members of the SCF is proposed to be involved in recognition of the organic substrate. Thus, sugar binding to SGLT1 was shown to be determined by the C-terminal half of the protein [20]. Furthermore, reaction of Cys344 in PutP with NEM was blocked by proline while Na ÷ was less efficient (Jung and Tebbe, unpublished information). Taking also into account that substitution of Cys344 by Ser yielded a reduced proline binding activity [21] the results suggest a location of the Cys at or near a proline binding site. However, the concomitant effect of the substitution of Ser57 in PutP on the protein affinity for Na ÷ and proline indicates a fight linkage between both ligand binding sites.

3. Topology Based on hydropathy profile analysis a secondary structure model was proposed according to which PutP consists of a short hydrophilic N-terminal region, 12 transmembrane domains (TMs) in et-helical conformation that traverse the membrane in zigzag fashion connected by hydrophilic loops, and a

hydrophilic C-terminal tail [ 11] (Fig. 1). Experimental evidence for the cytoplasmic location of the Cterminus came from immunological studies [22]. The 12-helix motif was predicted as a common structural feature of all members of the SCF [3]. We have tested this hypothesis applying a gene fusion approach and site-directed labeling techniques (Leifker, Ruebenhagen, Tholema, Quick, Tebbe and Jung, manuscript in preparation). Analysis of a series of putP-phoA and putP-lacZ fusions yielded a reciprocal activity pattern of the reporter proteins alkaline phosphatase (PhoA) and 13-galactosidase (LacZ) that was in agreement with the topology of TMs III to XII of the 12-helix model. Interestingly however, analysis of PutP-PhoA hybrid proteins with junction points in putative cytoplasmic loop 3 revealed a sharp increase in alkaline phosphatase activity upon shifting the junction point from Arg96 (last amino acid of the PutP fragment) to Trp90. Assuming that about half of a TM is needed to translocate alkaline phosphatase across the membrane [23] Trp90 is placed into TM II leaving Arg96 at the cytoplasmic border of the putative helix (Fig. 1). This modification shifts putative TM II by eight amino acids towards the C-terminus creating a large periplasmic loop of rather amphipathic character. Placement of PutP-PhoA junction points into this loop did not yield conclusive results. More information on PutP topology was gained by monitoring the accessibility of Cys residues placed close to the N- or C- terminus to thiol reagents. Cys at the position of Ile3, Thr5, or Ser502 was accessible to the membrane permeant thiol reagent 3-(Nmaleimidylpropionyl)biocytin (BM) in intact E. coli cells (Fig. 2). Furthermore, Cys at position 502 did not react with the membrane impermeant probe 4acetamido - 4' - maleimidylstilbene - 2,2' - disulfonate (SM) in intact cells but became accessible to this reagent after cell disruption. These results confirm a cytoplasmic location of the C-terminus as has already been indicated by immunological studies [22] and our gene fusion analysis. In contrast, the thiol groups at positions 3 and 5 were highly accessible to SM in intact cells (Fig. 2), and cell disruption had no influence on the labeling yield. This observation contradicts the 12-helix motif and indicates a periplasmic location of the N-terminus. In addition, the results point to an uneven number of TMs.

H. Jung / Biochimica et Biophysica Acta 1365 (1998) 60-64

PutP

Preincubation with Stilbenedisulfonate

I3C

+

T5C

+

PERIPLASM

$502C +

L3

L1 (Nil,) M

V A G

A

Streptavidin Blot

63

I

FLS .GI

S E

Protein Detection Fig. 2. Location of the N-terminus in the Na÷/proline transporter of E. coli. Cells of E. coli WG170 containing PutP with a single Cys at a given position were incubated with 200 IxM 3-(N-maleimidylpropionyl)biocytin (BM) for 30 rain at room temperature. Where indicated, cells were preincubated with 1 mM 4-acetamido-4'-maleimidylstilbene-2,2'-disulfonate for 15 min. After labeling, the protein was solubilized, purified by Ni-NTA chromatography and subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis. Reaction with BM was detected with Streptavindin HRP.

I%,,I •

R

~

T L K ! N Y FDD

L2

In a search for residues that might form an additional TM we replaced 19 amino acids in putative loop 2 and the adjacent TM II individually with Cys (Fig. 1). However, the thiol groups at all positions analyzed showed only a low reactivity towards BM or did not react at all. This result might be explained by a hydrophobic environment [24] or a buried location of the Cys residues. This conclusion is in agreement with the fact that an antibody raised against an amino acid sequence in putative loop 2 did not react with the protein [7]. Thus, the experimental results are consistent with the idea that amino acids of putative loop 2 form an additional TM. Further support for this idea comes from a site-directed spinlabeling study. This method analyzes relative collision frequencies between freely diffusing paramagnetic probe molecules of different solubility in the compartments of the system and an immobilized nitroxide spin-label [25]. The results obtained so far confirm a shift of amino acids (i.e., Arg40, Phe45, Ser50) from former loop 2 into the membrane bilayer (Steinhoff and Jung, unpublished information). Based on these studies a new secondary structure model of PutP is proposed according to which the protein consists of 13 TMs with the N-terminus on the outside and the C-terminus facing the cytoplasm. An additional TM is formed by amino acid residues of former loop 2 placing the functionally important residues Asp55 and Ser57 into TM II (Fig. 3). The new model is in agreement with the recently

N

R R

V H ET Y NA

NN L T L4 LPD

CYTOPLASM Fig. 3. New topological arrangement of the N-terminal part of the Na +/proline transporter. The model is based on the analysis of a series of putP-phoA and putP-lacZ fusions and site-directed labeling of the transporter.

proposed topological arrangement of TMs I to XIII of SGLT1 which is based on an N-glycosylation study [26]. The results support the idea of a common topological motif for members of the SCF according to which, i.e., the bacterial transporters for proline (PutP) and pantothenate (PanF) and the mammalian Na÷/I transporter are composed of 13 TMs [4]. Transporters with a C-terminal extension [i.e., the human SGLT1 and myoinisitol transporter (SMIT1)] are proposed to have an additional 14 th TM [4,26]. In this context it is interesting to note that a melB-phoA fusion analysis revealed a 12-helix motif for MelB, a member of the Na÷/galactoside family (SGF) [27]. The differences in primary structure and topology of members of the SCF and SGF suggest that these families evolved independently of each other.

Acknowledgements

This work was financially supported by the Deutsche Forschungsgemeinschaft (SFB 171-C 19).

64

H. Jung I Biochimica et Biophysica Acta 1365 (1998) 60-64

References [1] H.R. Kaback, K. Jung, H. Jung, J. Wu, G.G. Priv6, K. Zen, in: A.M. Tartakoff, R.E. Dalbey (Eds.), Advances in Cell and Molecular Biology of Membranes and Organelles, JAI Press, Greenwich, Vol. 4, 1995, 129-144. [2] M.D. Marger, M.H. Saier, Trends Biol. Sci. 18 (1993) 13-20.

[3] J. Reizer, A. Reizer, M.H. Saier Jr., Biochim. Biophys. Acta 1197 (1994) 133-166. [4] E. Turk, E.M. Wright, J. Membr. Biol. 159 (1997) 1-20. [5] M.G. Martin, E. Turk, M.P. Lostao, C. Kemer, E.M. Wright, Nat. Genet. 12 (1996) 216-220. [6] G. Dai, O. Levy, N. Carrasco, Nature 379 (1996) 458-460. [7] E.M. Wright, D.D.F. Loo, M. Panayotova-Heiermann, M.P. Lostao, B.A. Hirayama, B. Mackenzie, K. Boorer, G. Zampighi, J. Exp. Biol. 196 (1994) 197-212. [8] I. Yamato, Y. Anraku, in: E.P. Bakker (Ed.), Alkali Cation Transport Systems in Procaryotes, CRC Press, Boca Raton, FL, 1993, pp. 53-76. [9] L.M.D. Stewart, I.R. Booth, FEMS Microbiol. Lett. 19 (1983) 161-164. [10] C.-C. Chen, T. Tsuchiya, Y. Yamane, J.M. Wood, T.H. Wilson, J. Membr. Biol. 84 (1985) 157-164. [11] J.M. Wood, D. Zadwomy, Can. J. Biochem. 58 (1980) 787-796. [12] T. Nakao, I. Yamato, Y. Anraku, Mol. Gen. Genet. 208 (1987) 70-75. [13] K. Hanada, I. Yamato, Y. Anraku, J. Biol. Chem. 263 (1988) 7181-7185.

[14] C.-C. Chen, T.H. Wilson, J. Biol. Chem. 261 (1986) 25992604. [15] M. Quick, H. Jung, Biochemistry 36 (1997) 4631-4636. [16] M. Ohsawa, T. Mogi, H. Yamamoto, I. Yamato, Y. Anraku, J. Biol. Chem. 170 (1988) 5185-5191. [17] M. Quick, S. Tebbe, H. Jung, Eur. J. Biochem. 239 (1996) 732-736. [18] R.S. Myers, S.R. Maloy, Mol. Microbiol. 2 (1988) 749755. [19] M.L. Zani, T. Pourcher, G. Leblanc, J. Biol. Chem. 268 (1993) 3216-3221. [20] M. Panayotova-Heiermann, D.D.F. Loo, C.-T. Kong, J.E. Lever, E.M. Wright, J. Biol. Chem. 271 (1996) 1002910034. [21] K. Hanada, T. Yoshida, I. Yamato, Y. Anraku, Biochim. Biophys. Acta 1105 (1992)61-66. [22] Y. Komeiji, K. Hanada, I. Yamato, Y. Anraku, FEBS Lett. 256 (1989) 135-138. [23] J. Calamia, C. Manoil, Proc. Natl. Acad. Sci. USA 87 (1990) 4937-4941. [24] T. Kimura, M. Suzuki, T. Sawai, A. Yamaguchi, Biochemistry 35 (1996) 15896-15899. [25] W.L. Hubbell, C. Altenbach, in: S.H. White (Ed.), Membrane Protein Structure--Experimental Approaches, Oxford University Press, Oxford, 1994, pp. 224-248. [26] E. Turk, C.J. Kemer, M.E Lostao, E.M. Wright, J. Biol. Chem. 271 (1996) 1925-1934. [27] T. Pourcher, E. Bibi, H.R. Kaback, G. Leblanc, Biochemistry 35 (1996) 4161-4168.