- Email: [email protected]
Structural determinants in substrate recognition by proton-amino acid symports in plasma membrane vesicles isolated from sugar beet leaves
ARCHIVES
OF BIOCHEMISTRY
Vol. 294, No. 2, May
AND
BIOPHYSICS
1, pp. 519-526,1992
Structural Determinants in Substrate Recognition by Proton-Amino Acid Symports in Plasma Membrane Vesicles Isolated from Sugar Beet Leaves Zhen-Chang
Li*
and Daniel
R. Bush*st,’
tPhotosynthesis Research Unit, U.S.D.A. Agricultural Research Service and the *Department Plant Biology, University of Illinoi-s, 190 PABL, 1201 W. Gregory Dr., Urbana, Illinois 61801
Received
October
31, 1991, and in revised
form
January
6,1992
Amino acids are actively transported across the plasma membrane of plant cells by proton-coupled symports. Previously, we identified four amino acid symports in isolated plasma membrane vesicles, including two porters for the neutral amino acids. Here we investigated the effect of amino acid analogues on neutral amino acid transport to identify structural features that are important in molecular recognition by Neutral System I (isoleucine) and Neutral System II (alanine and leucine). DIsomers of alanine and isoleucine were not effective transport antagonists of the L-isomers. These data are characteristic of stereospecificity and suggest that the positional relationship between the a-amino and carboxyl groups is an important parameter in substrate recognition. This conclusion was supported by the observation that &alanine and analogues with methylation at the (Ycarbon, at the carboxyl group, or at the a-amino group were not effective transport inhibitors. Specific binding reactions were also implicated in these experiments because substitution of the a-amino group with a space filling methyl or hydroxyl group eliminated transport inhibition. In contrast, analogues with various substitutions at the distal end of the amino acid were potent antagonists. Moreover, the relative activity of several analogues was influenced by the location of sidechain branches and Neutral Systems I and II were resolved based on differential sensitivity to branching at the B-carbon. The kinetics of azaserine and p-nitrophenylalanine inhibition of leucine transport were competitive. We conclude that the binding site for the carboxyl end of the amino acid is a well-defined space that is characterized by compact, asymmetric positional relationships and specific ligand interactions. Although the molecular interactions associated with the distal portion of the amino acid were less 1 To whom correspondence should be addressed at USDA/ARS Department of Plant Biology, University of Illinois, 190 PABL, W. Gregory Dr., Urbana, IL 61801. Fax (217) 244-4419. 0003-9861/92 Copyright 0 All
rights
$3.00
1992 by Academic Press, of reproduction in any form
of
and 1201
restrictive, this component of the enzyme-substrate complex is also important in substrate recognition because the neutral amino acid symports are able to discriminate between specific neutral amino acids and exclude the acidic and basic amino acids. c 1esz ACUIWIC m, IIIC.
Amino acids are actively transported across the plasma membrane of the plant cell by proton-coupled symports (l-3). These transport systems link amino acid accumulation to the protonmotive force generated by a Ptype, H+-pumping ATPase (4, 5). Recently, we used isolated plasma membrane vesicles and imposed proton electrochemical potential differences to study proton amino acid symports found in sugar beet leaf tissue (2). Amino acid transport was carrier-mediated and electrogenie. Moreover, inter-amino acid transport competition suggestedthat at least four amino acid symports, an acidic, a basic, and two neutral, are present in the plant plasma membrane (2,6). We are interested in understanding the biochemistry of these porters because they play a central role in nitrogen distribution in the plant and, since many tissues of the higher plant are heterotrophic, amino acid transport can have a profound influence on growth and development. Previously, we identified two neutral amino acid symports in isolated plasma membrane vesicles based on differences in specific activity, species distribution, and the kinetics of inter-amino acid transport competition (2,6). Perhaps the most striking distinction, however, was the initial observation that isoleucine, valine, and threonine were weak transport antagonists of the other aliphatic neutral amino acids while, in contrast, these amino acids were subject to transport inhibition by all neutral amino acids. Upon inspection, a distinguishing feature of isoleucine, valine, and threonine is that each has a branched methyl or hydroxyl group at their p-carbon. We suggested 519
Inc. reserved.
520
LI
AND
that branching at this position contributes to an unfavorable steric interaction with the active site of one of the symports identified. Thus, we defined Neutral System I as a general porter for all neutral amino acids and Neutral System II as a general porter that specifically excludes isoleucine, valine, and threonine. Resolution of two transport systems based on differential affinities for the neutral amino acids suggests specific molecular interactions must occur between the porters and their respective substrates. In animal cells, a variety of amino acid transport systems have been characterized (7, 8). These systems are distinguished from one another based on differential affinities for specific groups of amino acids and as a function of their dependence upon a cotransported ion. Over the years, detailed studies with amino acid analogues have provided insight into substrate binding to their respective transport proteins (7). For example, some studies have placed Na+ and amino acid binding in close proximity while others have shown that the length, charge distribution, and hydrophobicity of the amino acid side chains are important factors of molecular recognition (9-13). In contrast to animal cells, few molecular characterizations of the amino acid transport systems in plants have been reported. Although the a-amino and the carboxy1 groups have been suggested to be important in substrate recognition (14), little is known about the molecular features that influence amino acid binding to specific symports (3). In the results reported here, we used isolated membrane vesicles and amino acid analogues to investigate structural features of the neutral amino acids that are important for molecular recognition by the neutral amino acid symports. Evidence is presented which shows that the topography of the carboxyl end of the amino acid is very important in molecular recognition and that sidechain branching, length, and polarity also play significant roles. EXPERIMENTAL Plant mute&l.
PROCEDURES
BUSH BSA, pH adjusted to 8.0 with BTP). After differential centrifugation, microsomal pellets were resuspended in 3 ml of 350 mM sorbitol containing 5 mM K-phosphate buffer (pH 8.0) and 0.1 mu DTT. PMVs were purified from the resuspended microsomes by the aqueous twophase partitioning method as described previously (15). Phase purified PMVs were washed by centrifuging in 35 ml of resuspension buffer (330 mM sorbitol, 2 mM Hepes pH adjusted to 8.0 with BTP, 10 mM KC1 and 0.1 mu D’IT) at 50,OOOg for 45 min. The final pellet of plasma membrane vesicles was resuspended in 1 ml of resuspension buffer (4-6 mg protein ml-‘) and used for amino acid uptake experiments.
Transport assays. All assays of amino acid transport activity were carried out at 1O’C. A stable transmembrane pH difference (15) was generated by diluting a small aliquot of PMVs (intravesicular pH 8.0) into 200 pl acidic transport solution (prepared from resuspension buffer by adjusting pH to 6.0 with solid Mes), containing 5 ELM valinomycin and 0.2 to 0.5 gCi of the “‘C-labeled amino acid in the presence or absence of 5 I.IM CCCP. Unlabeled substrate amino acid and amino acid analogs or D-isomers were added to the transport solution to the desired final concentrations. After 2-4 min, while the transport rate was linear (2), the suspended membrane vesicles were collected and washed by a Millipore filtration technique (15). Accumulated radioactivity was measured by scintillation spectrometry. ApH-dependent transport of the radiolabeled amino acid was defined as the difference between the accumulation in the absence and in the presence of CCCP (2,15). Each experiment was repeated at least three times. Protein determination. Proteins were measured by the Markwell method (16) after soluhilization in 0.2% sodium deoxycholate and precipitation in 6.2% trichloroacetic acid (17). BSA Fraction V (Sigma) was used as the protein standard.
RESULTS Inhibition of alanine and isokucine transport by stereoisomers of select amino acids. We examined the effect of different amino acid stereoisomers on isoleucine (Neutral System I) and alanine (Neutral System II) transport to identify macroscopic features of the transported amino acid that may be involved in substrate recognition by the neutral amino acid symports. The results in Table I show
TABLE I StereoisomerEffects on ApH-Dependent Alanine and IsoleucineTransnort
Sugar beet (Beta vulgaris L. cv Great
Western) seeds were geminated in moist vermiculite. When the seedlings were about 1 in. tall, they were transferred to a hydroponic system (two plants per 13 liters), containing one-half strength Hoagland’s solution and 1 mM NaCl. Plants were maintained under a 10-h photoperiod (450 pmol me2 8-i) at 24”C, 70% RH and a 14-h dark period at 18”C, 70% RH. Fully expanded leaves were harvested for experimental analysis.
Preparation of plasma mernbmne vesicles.
Plasma membrane vesicles (PMVs) were prepared at 4’C as described by Bush (15). Briefly, 40 g of leaves was homogenized in 350 ml of grinding solution (250 mM sorbitol, 50 mM Hepes: 10 mM KCl, 3 mM EGTA, 3 mM D’IT, and 1% of
‘Abbreviations used: BSA, bovine serum albumin; BTP, 1,3bis[tris(hydroxymethyl)methylamino]propane; CCCP, carbonyl cyanide m-chlorophenylhydrazone; DTT, dithiothreitol; EGTA, ethyleneglycol bii(/3aminoethyl ether)NJJ’-tetraacetic acid; Hepes, 4-(2-hydroxyethyl) l-piperazineethanesulfonic acid; Mes, 2-(N-morpholino)ethanesulfonic acid, PMV, plasma membrane vesicle; T-P-L, tabtoxinine-@-lactam
Transport
substrate
Competitor
L-Alanine
L-Isoleucine
Control L-Alanine D-Alanine L-Isoleucine D-Isoleucine L-Serine D-Serine
100% 101+ 94* 107 + 762 103 +
100% 50+ 91+ 100* 58rt 87 zk
11 7 9 5 7
6 5 9 7 10
Note. Transport solutions contained 0.1 mM [“Clalanine or [“C]isoleucine and 0.5 mhf of the competing amino acid. ApH-dependent transport activity was defined as the difference between accumulation in the absence and that in the presence of 5 pM CCCP. Results are expressed as percentage of control + SD (n = 4). 100% = 1.5 nmol alanine or 0.09 nmol isoleucine/mg protein/min.
SUBSTRATE
RECOGNITION
BY PLANT TABLE
ApH-Dependent
Alanine
and Isoleucine
Transport
PROTON-AMINO
521
ACID SYMPORTS
II
in the Presence
of Amino
Acid Analogs
Containing
Distinct
R-Groups
Substrate Amino acid analogs Control 0 II CH,-C-0-CH,-CH-COOH
0-Acetylserine
0 II N=N=CH-C-0-CHZ--CH-COOH
Azaserine
Alanine
Isoleucine
100%
100%
46 2 1
46 2 2
29 rt 1
39 * 5
44 f 2
59 + 3
58 2 8
26k 1
53 2 2
58 + 6
40 2 2
16 + 2
32 + 6
24 2 4
61e4
65 2 5
I NHz
(t
0-Benzylserine
/
0-Methylserine
-
\
CH2-0-CH2-CH-COOH
I NH2 CH,-0-CH,-CH-COOH
I NH2 p-Nitrophenylalanine
(0.1 mM)*
NO2
CH,-CH-COOH
NH2 CH3-CH2-CH2-CH-COOH
Norvaline
I NH2
Norleucine
CH3-CH2--CH2-CH,-CH-COOH NH2
Methionine
sulfoxide
CH,-S-CH,-CH,-CH-COOH d
I NH2
Note. Transport solutions included 0.1 mM [i4C!]alanine or [i4C]isoleucine in the absence (control) or presence of 0.5 mM amino acid analogue*. ApH-dependent transport activity was expressed as percentage of the control 2 SD (n = 3).
that L-serine and L-alanine inhibit alanine and isoleucine transport while L-isoleucine was not effective against alanine. In contrast, D-isomers of alanine, isoleucine, and serine had no effect on alanine transport while isoleucine transport was decreased by approximately 10%. These results suggest both Neutral Systems I and II are specific for L-amino acids, although Neutral System I may exhibit some recognition of the D-isomers. Inhibition of alanine and isoleucine transport by various amino acid anulogues. The effect of several amino acid analogues on alanine and isoleucine transport was analyzed to determine what portion of the amino acid molecule might be important in molecular recognition. Three major classes of analogues were examined, including (i) hydrophobic or hydrophilic R group additions, (ii) replacement of the a-carbon amine with a spacefilling group or moving it to the P-carbon, and (iii) methylation at the a-carbon, at the a-amino group, or at the carboxyl group.
The results in Table II report the effect of various R group configurations on alanine and isoleucine transport. In every case, these analogues were effective transport inhibitors. This was true for both hydrophobic and hydrophylic additions, as well as for large ringed structures. These data suggest the binding sites of Neutral Systems I and II are not very restrictive with respect to the distal portion of the amino acid. It should be noted, however, that the presence of charged groups in this region, such as found in the acidic and basic amino acids, significantly reduces transport antagonism (2). The second class of amino acid analogues examined was chosen to explore the positional significance of the a-amino group in substrate recognition (Table III). Analogues lacking the a-amino group were not active transport antagonists. This observation supports the notion that the amino group plays an important role in substrate recognition. In addition, substitution of the amine with
522
LI
ApH-Dependent
Alanine and Isoleucine Transport
AND
BUSH
TABLE
III
in the Presence of Amino Acid Analogs Containing Substituted Carboxyl and Amino Groups Substrate
Amino
acid analogue
Alanine
Control
Isoleucine
100%
&Alanine
NH2 -
N-Methyl-k&nine
CHx -
CHr -
COOH
100%
104 + 10
CH3--CH-COOH
83+
7
91+
6
70 + 10
lOO+
8
84+
9
101 f
3
101+
7
91*
5
1Olk
7
110+
7
118 + 10
113 f
7
92+
6
108+
7
97+
6
102 *
2
106 f
3
I NH I CHs CH3
2-Aminoisobutyrate
(AIB)
CH3 -
DL-Isoserine
C - COOH I NHs
NHr-CH,-CH-COOH I OH
Dl&Aminoisobutyrate
NH*
-CHr
-CH-COOH
I CH3
Glycolic
acid
HO-CHr-COOH
Sarcosine
CHs -NH-CHx
-COOH CHs
2-Methylaminoisobutyrate
CHx -
C -
COOH
I NH
I CH, 0
II L-Serine
methylester
HO-CH,-CH-C-0-CHx
I NHz Note.
Experimental
conditions
and expression
of transport
activity
a space filling methyl or hydroxyl group, or placing the amine within the linear side chain, did not restore activity. This suggests the charge status of the amino group, or at least its ability to form specific ligand interactions, is also a critical factor in molecular recognition. We also examined amino acid analogues which shifted the amine to the @-carbon. &Aminoisobutyrate and isoserine showed little or no inhibition of either alanine or isoleucine transport (Table III). In contrast, @-alanine had no influence on alanine transport, yet it decreased isoleucine transport by 20%. These data provide further evidence in support of the conclusion that the a-amino group is
were the same as in Table
II.
important in molecular recognition and, moreover, they highlight the significance of the positional relationship of the amino group with respect to the rest of the molecule. The last class of analogues examined included those with specific substitutions at the o-amino, a-carbon, and carboxyl positions (Table III). Methylation of the a-carbon or a-amino group practically eliminated transport antagonism. For example, 2-aminosiobutyric acid (AIB) and N-methylalanine inhibited isoleucine transport to a limited extent while they had no significant impact on alanine transport. Likewise, methylation of a carboxyl oxygen (serine methylester) eliminated transport inhi-
SUBSTRATE
llleucine
g
E
6.0
-
RECOGNITION
BY
PLANT
(PM)
B
llleucine
(PM)
FIG. 1.
Double reciprocal plot of the kinetics of ApH-dependent leutine transport in the presence of (A) azaserine and (B) p-nitrophenylanine at 0 pM (O), 100 pM (0), and 200 pM (M).
bition. Taken together, the impact of the analogues listed in Table III on alanine and isoleucine transport emphasizes the importance of the a-amino and carboxyl end of the molecule in substrate recognition. Competitive inhibition of leucine transport by azaserine and p-nitrophenylalanine. We examined the kinetics of leucine transport inhibition by azaserine and p-nitrophenylalanine to test the assumption that these analogues are competing for binding to the active site of Neutral System II. Double reciprocal plots of azaserine and pnitrophenylalanine inhibition of leucine transport were diagnostic of competitive inhibition (Fig. 1) and Ki values, estimated from Dixon plots, were approximately 75 PM for both analogues. These results are consistent with the notion that these amino acid analogues are competing for binding at the active site of the porter. Effect of amino acid length on alunine transport. In animal cells, Oxender and Christensen (9) observed increased transport antagonism against amino acid Transport Systems A and L activity as the length of the competing amino acid’s apolar R group increased. To determine if parallel molecular interactions occur for Neutral System II, we tested several amino acids and analogues, at a fixed concentration, for their ability to inhibit alanine transport. Glycine, which has no carbon in the R group side chain, inhibited alanine transport by 30% while norleucine, which contains 4 carbons in the R group, in-
PROTON-AMINO
ACID
523
SYMPORTS
hibited transport by approximately 70% (Fig. 2). When the side chain contained one carbon (alanine), transport competition was approaching maximum. These observations implicate the length of the R group as a contributing factor in substrate binding to the active site. Effect of branching patterns on alanine and isolewine transport. We reported previously that amino acid branching at the P-carbon appears to be an important structural property which differentiates the neutral amino acids with respect to transport through Neutral Systems I and II (2, 6). In the current investigation, a number of amino acids and amino acid analogues representing a variety of branching patterns were tested for their relative abilities to inhibit isoleucine (Neutral System I) and alanine (Neutral System II) transport. As the branching position moved away from the carboxyl carbon, transport inhibition increased from zero to 40% for both porters (Fig. 3). It is noteworthy, however, that different patterns of transport antagonism were observed. Alanine transport was progressively inhibited as the branch point moved away from the carboxyl end of the analogues. In contrast, isoleucine transport decreased rapidly with maximal inhibition evident with branching at the ,&carbon. These results are consistent with our previous suggestion that amino acid binding to Neutral System II is limited by a comparatively restricted molecular space associated with the B-carbon. DISCUSSION
Two neutral amino acid transport systems are present in plasma membrane vesicles isolated from sugar beet leaf tissue (6). These proton-coupled symports were initially resolved because of their differential affinities for isoleutine, valine, and threonine. We suggestedthat the presence of a methyl or hydroxyl group at the P-carbon of these amino acids blocks accessto the active site of Neutral Sys-
0 % 2 I= 2 $
60-
40-
u z
l
+Serine
20 .-0” z =E
0,
I 0.0
1.0
Number
FIG. 2.
1 2.0
of Atoms
I 3.0
I 4.0
in the Sidechain
Inhibition of ApH-dependent alanine transport as a function of antagonist sidechain length. Transport was assayed after 3 min with 0.1 mM [“Clalanine and 0.5 mM antagonist. Specific antagonists were glycine (0 atoms), alanine (1 atom), serine (2 atoms), norvaline (3 atoms), and norleucine (4 atoms).
524
LI
=::
AND
50
2 F lQ = B 2 a
40
30
20
z .-5
10
E .c +
0
ae
-I 0
1
Number
2
3
of Atoms Away Carboxyl Carbon
4
from
5
the
FIG. 3. Effect of antagonist sidechain position on ApH-dependent alanine (0) and isoleucine (0) transport. Transport assays were as described in Fig. 2. Amino acids representing specific branching positions were 2-aminoisobutyrate, isoleucine (or valine in the Ile transport), leutine, and methionine sulfoxide, respectively.
tern II and, based on this hypothesis, we became interested in identifying structural elements of the neutral amino acids that are important in molecular recognition. Stereospecificity is a characteristic of enzyme-mediated reactions which provides macroscopic information about the three-dimensional requirements for substrate orientation within the enzyme active site. Since Neutral Systems I and II discriminate against D-iSOmerS of their respective transport substrates (Table I), we propose that the a-amino and carboxyl groups of the neutral amino acids are associated with specific ligands in the active sites of these porters and that these ligands must be asymmetrically positioned. Furthermore, stereospecificity requires at least three unique binding sites for the groups surrounding the chiral carbon (18). Therefore, the position of the R group must also play an important role in molecular recognition. Stereospecificity in amino acid transport was also reported in cultured tobacco cells (19, 20) and corn scutellum (21). However, the degree of sensitivity to stereoisomers of the neutral amino acids may vary from one species to another. For example, Jung and Luttge (22) and Felle (14) used electrophysiological evidence to conclude that D- and L-amino acids are transported by the same carrier. Stereospecificity suggests the positional relationship between the a-amino and the carboxyl groups is an important parameter in molecular recognition. This conclusion is consistent with the results generated with the pamino acid analogues. /3-Alanine was not an active antagonist of amino acid transport through Neutral System II and it was only marginally effective against Neutral System I (Table III). Likewise, isoserine and /3-aminoisobutyrate were not active against either porter (Table III). These results suggest a critical distance between the (Yamino and the carboxyl groups must be maintained for
BUSH
competitive binding to the active site. Taken together with the three-dimensional restrictions imposed by stereospecificity, the region of the active site associated with the carboxyl end of the neutral amino acids appears to be a very well-defined space which plays an important role in substrate recognition. Similar conclusions have been reached for a number of amino acid transport systems in animal cells (7, 13) and in an aquatic liverwort (14). Analogues with methyl additions to either the a-amino or the carboxyl groups were not effective transport inhibitors, although Neutral System I exhibited some sensitivity to N-methylalanine. These data support the notion that the a-amino and carboxyl groups are important participants in substrate recognition and provide further insight into the molecular interactions associated with the enzyme-substrate complex. Methylation of the (Yamino or carboxyl groups interferes with several reactions of potential significance. For example, methylation may simply produce a steric interaction which restricts access to the active site. This kind of hindrance would be the result of a relatively confined binding cavity which does not accommodate extraneous atoms. This idea is consistent with the positional restrictions placed on the system as a result of stereospecificity. Additionally, several forces are involved in substrate binding to enzyme active sites, including van der Waals interactions, H-bonds, hydrophobic effects, and electrostatic bonds. Presumably, methylation of the a-amino or carboxyl groups also eliminates electrostatic linkages and H-bonds which may be important in substrate recognition and binding. This conclusion, which assumes the amino acids are zwitterionic while bound, is supported by the observation that substitution of the a-amino group with a space filling methyl or hydroxyl group produces a completely inactive analogue (Table III). We conclude from these data that specific positional and chemical requirements must be satisfied for successful binding to occur. Previously we argued that global features of the neutral amino acids cannot account for differences in substrate recognition by Neutral Systems I and II (6). This conclusion was supported in the current investigation. Amino acid analogues with a variety of R groups were effective antagonists of amino acid flux through both transport systems (Table II). Since these R groups included hydrophobic and hydrophylic constituents, as well as large ringed structures, it appears that molecular interactions with the distal portion of the neutral amino acids is relatively unrestricted. The exception to this generalization, i.e., when there is a branch point at the P-carbon, is the distinguishing molecular feature of those amino acids excluded from Neutral System II. It should be noted, however, that there were subtle differences between Neutral Systems I and II in their relative sensitivity to the analogues in Table II (e.g., norvaline and o-methylserine) and, perhaps more significantly, these symports were comparatively insensitive to transport antagonism by the
SUBSTRATE
RECOGNITION
BY
PLANT
acidic and basic amino acids which place charged groups in the distal portion of the molecule (2). Thus, although molecular interactions with this region of the amino acid are relatively unconstrained, there are restrictions which effectively limit amino acid binding to the neutral symports as a function of the distal region of the molecule. In animal cells the number of atoms (including carbon, sulfur, and oxygen) in the linear side chain is an important factor in substrate binding. For example, amino acids and analogues with four or five carbons in the side chain have the highest affinity for Transport Systems ASC and Ly+ (11-13). Similarly, amino acid analogues against System A and System L were more effective as the length of the side chain increased to seven atoms (9). In the present investigation, the sensitivity of Neutral System II to analogues of increasing length was similar to that observed for System A and System L. Transport antagonism increased from zero carbon (glycine) to four carbons (norleucine) in the side chain (Fig. 2). If oxygen atoms in the side chain are considered roughly equivalent in length to carbon, serine could be equated with a two-carbon side chain analogue. Note, however, that its activity as a transport antagonist was lower than that of alanine (Fig. 2). This observation suggests polarity can also play a role in molecular recognition in this region of the amino acid. We suggested in our earlier studies that tolerance to branching at the P-carbon is an important property differentiating Neutral System I from Neutral System II. We noted that branching at this position places a dense electron cloud in close proximity to the a-amino group and concluded that this results in unfavorable steric interactions with the binding cavity of Neutral System II. The experiments reported here extend this hypothesis. Methylation at the a-amino, a-carbon, or P-carbon position increases the electron cloud distribution at the carboxy1 end of N-methylalanine, 2-aminosiobutyrate, and valine. These amino acids exhibited little or no inhibition of alanine transport yet isoleucine transport was inhibited from 16 to 40% (Table 3 and Fig. 3, (2)). These results support the idea that the binding cavity associated with the (Y- and P-carbons is, comparatively, more restricted in Neutral System II than in Neutral System I. When the side chain is branched at the P-carbon, antagonist activity against Neutral System I approaches maximum. In contrast, activity against Neutral System II is very low and only increases as the branch point moves out from the pcarbon (Fig. 3). An underlying assumption of the experiments reported here is that the transport antagonists were competing for substrate binding at the active sites of the neutral amino acid symports. The kinetics of Neutral System II transport inhibition by azaserine and p-nitrophenylalanine support this conclusion (Fig. 1). Additionally, our previous studies of inter-amino acid transport competition with isolated vesicles did not reveal evidence of allosteric regulation of Neutral System I or II (6). Therefore, even
PROTON-AMINO
ACID
SYMPORTS
525
though the kinetics of transport inhibition were not examined for every analogue, we believe the data presented here are the results of specific molecular interactions associated with substrate recognition. However, since these porters are secondary active transport systems that are dependent on a stable transmembrane proton electrochemical potential, it is possible that indirect effects on this driving force could interfere with amino acid flux. We examined this possibility in earlier experiments by determining the effect of simultaneous sucrose transport on amino acid flux. Sucrose transport is also mediated by a proton-coupled symport in these membrane preparations (15) and, thus, if competition for the protonmotive force is a reasonable explanation for our observations, we should see sucrose-dependent changes in the transport rate. Such changes were not observed and, therefore, we do not attribute changes in the transport rates to indirect competition for the driving force. Taken together, we conclude that these amino acid analogues are indeed competing with the native substrate for binding to the porter active sites. The structure-activity relationships reported here provide an experimental basis for correlating specific molecular interactions with substrate binding to the neutral amino acid symports. Moreover, these data provide a conceptual framework for constructing a working model of the binding sites of these porters. Three general features of this model stand out. First, that region of the protein which binds the carboxyl end of the amino acid is characterized by a well-defined cavity which is structurally limited, stereospecific, and where successful binding is driven by specific ligand interactions. This site does not tolerate extraneous atoms or space filling substitutions which eliminate or reduce required binding reactions. The second feature of this model is the relatively unrestricted binding site for the distal portion of the amino acid. This section of the porter has few physical barriers which interfere with substrate binding, although it is important to recall that this region is sensitive to polar interactions and binding is weak when charged groups are present. The third feature of this model, which separates Neutral Systems I and II, focuses on the transition zone between the carboxyl and the distal regions of the amino acids. Here, Neutral System II is characterized by steric interactions which block effective binding of amino acids containing substitutions on the P-carbon. In contrast, amino acid binding to Neutral System I is not as structurally confined in this region. Taken together, a model is emerging which describes a tight fitting binding pocket for the carboxyl end of the amino acid, fewer structural limitations around the P-carbon, and a wide open cavity which binds the distal region. Perhaps the carboxyl end of the molecule is buried within the porter with the distal portion of the amino acid inside a large channel leading to the carboxyl binding pocket.
526
LI
A
OH
NH,
I
I
AND
ing the knowledge obtained in this study to identify potential photoaffinity probes, such as azaserine and p-nitrophenylalanine, to help identify these important transport systems in the plant.
-CH,--CH,-C-COOH H2P-i:
,B-“\\,
BUSH
I
ti
REFERENCES 1. Bush, D. R., and Lang&on-Unkefer, 487-490.
B
0
2. Li, Z-C., 3. Reinhold, 45-83.
II OH-P-CH,-NH-CH,-COOH I OH FIG.
4.
Structure
of tabtoxinine-p-lactam
(A) and glyphosate
(B).
This model is useful because it can be used to predict whether exotic biological compounds are transported through the neutral amino acid symports. For example, tabtoxinine+lactam (T-@-L) is a toxic a-amino acid analogue produced by several closely related pathovars of Pseudomonas syringae. This compound is an irreversible inhibitor of glutamine synthetase and this activity is directly involved in eliciting the chlorotic symptoms these pathogens induce in their hosts (23). Recently, T-P-L transport into plant cells was shown to be mediated by an amino acid transport system (1). That result is consistent with the conclusions drawn here because the structural features of T-P-L (Fig. 4) suggest it can be transported through either Neutral System I or II. This conclusion is supported by the observation that T-0-L transport into cultured cells was inhibited by alanine (24). In a contrasting example, glyphosate is a widely used herbicide whose transport into the cell has been suggested to be mediated by an amino acid porter (25). However, it does not contain the necessary structural elements required for recognition by either the neutral amino acid symporters or (by inference) the acidic or basic porters (Fig. 4). Indeed, recent experiments examining glyphosate transport into isolated plasma membrane vesicles have failed to demonstrate symport mediated flux (Riechers and Bush, unpublished data). The results reported here provide important clues about the molecular interactions involved in substrate recognition by the plant neutral amino acid symports. Future advances will require information about the membrane proteins mediating these reactions. Therefore, we are us-
P. J. (1988)
Plant
Phyaiol.
88,
and Bush, D. R. (1990) Plant PhysioL 94,268-277. L., and Kaplan, A. (1984) Annu. Reu. Plant Physiol.
35,
4. Sze, H. (1985) Annu. Reu. Plant Physiol. 36,175208. 5. Serrano, R. (1989) Annu Rev. Plant Physiol. Pkmt 61-94.
Mol.
BioL 40,
6. Li, Z-C., and Bush, D. R. (1991) Plant Physiol. g&1338-1344. 7. Christensen, H. N. (1989) in Methods in Enzymology (Fleischer, S., and Fleischer, B., Eds.), Vol. 173, pp. 576-616, Academic Press, San Diego. M. E., and Kilberg, M. S. (1999) J. Biol. Chem. 265, 8. Mailliard, 14,321-14,326. 9. Oxender, D. L., and Christensen, H. N. (1963) J. Biol. Chem. 238, 3686-3699. 10. Thomas, E. L., and Christensen, H. N. (1970) Biochim. Biophys. Acta 40,277-283. E. L., and Christensen, H. N. (1971) J. Biol. Chem. 246, 11. Thomas, 1682-1688. H. N., Thomas, E. L., and Handiogten, M. E. (1969) 12. Christensen, Biochim. Biophys. Acta 193, 228-230. 13. Christensen, H. N. (1979) Advance 14. Felle, H. (1981) Planta 152,505-512.
Enzymol.
49,41-101.
15. Bush, D. R. (1989) Plant Physial. 89,1318-1323. M. K., Haas, S. M., Tolbert, N. E., and Bieber, L. L. 16. Markwell, (1981) in Methods in Enzymology (Lowenstein, J. M., Ed.), Vol. 72, pp. 296-303, Academic Press, San Diego. A., and Weinstein, D. (1976) Anal. Biochem. 70, 24117. Bensadoun, 250. Structure and Function, 2nd ed., pp. 224, 18. Fersht, A. (1984) Enzyme Freeman, New York. 19. Harrington, 378. 20. Blackman, 261-266.
H. M., and Henke, M.
S., and McDaniel,
R. R. (1981)
Plant
C. N. (1980)
PhysioL Plant
67,373-
Physiol.
66,
21. Steward, C. R. (1971) Plant Physiol. 47,157-161. 22. Jung, K. D., and Luttge, U. (1980) Planta 150,230-235. R. D. (1971) in Morophological and Biochemical Events in 23. Durbin, Plant Parasite Interactions (Akai, S., Ouchi, S., Eds.), pp. 369-385, Mochizuke Publishing Co., Japan. P. J. (1987) Plant PhysioL 85, 24. Bush, D. R., and Langston-Unkefer, 845-849. E. D., Widholm, J. M., and Slife, F. W. (1983) Plant PhysioL 25. Nafziger, 71,623-626.
OF BIOCHEMISTRY
Vol. 294, No. 2, May
AND
BIOPHYSICS
1, pp. 519-526,1992
Structural Determinants in Substrate Recognition by Proton-Amino Acid Symports in Plasma Membrane Vesicles Isolated from Sugar Beet Leaves Zhen-Chang
Li*
and Daniel
R. Bush*st,’
tPhotosynthesis Research Unit, U.S.D.A. Agricultural Research Service and the *Department Plant Biology, University of Illinoi-s, 190 PABL, 1201 W. Gregory Dr., Urbana, Illinois 61801
Received
October
31, 1991, and in revised
form
January
6,1992
Amino acids are actively transported across the plasma membrane of plant cells by proton-coupled symports. Previously, we identified four amino acid symports in isolated plasma membrane vesicles, including two porters for the neutral amino acids. Here we investigated the effect of amino acid analogues on neutral amino acid transport to identify structural features that are important in molecular recognition by Neutral System I (isoleucine) and Neutral System II (alanine and leucine). DIsomers of alanine and isoleucine were not effective transport antagonists of the L-isomers. These data are characteristic of stereospecificity and suggest that the positional relationship between the a-amino and carboxyl groups is an important parameter in substrate recognition. This conclusion was supported by the observation that &alanine and analogues with methylation at the (Ycarbon, at the carboxyl group, or at the a-amino group were not effective transport inhibitors. Specific binding reactions were also implicated in these experiments because substitution of the a-amino group with a space filling methyl or hydroxyl group eliminated transport inhibition. In contrast, analogues with various substitutions at the distal end of the amino acid were potent antagonists. Moreover, the relative activity of several analogues was influenced by the location of sidechain branches and Neutral Systems I and II were resolved based on differential sensitivity to branching at the B-carbon. The kinetics of azaserine and p-nitrophenylalanine inhibition of leucine transport were competitive. We conclude that the binding site for the carboxyl end of the amino acid is a well-defined space that is characterized by compact, asymmetric positional relationships and specific ligand interactions. Although the molecular interactions associated with the distal portion of the amino acid were less 1 To whom correspondence should be addressed at USDA/ARS Department of Plant Biology, University of Illinois, 190 PABL, W. Gregory Dr., Urbana, IL 61801. Fax (217) 244-4419. 0003-9861/92 Copyright 0 All
rights
$3.00
1992 by Academic Press, of reproduction in any form
of
and 1201
restrictive, this component of the enzyme-substrate complex is also important in substrate recognition because the neutral amino acid symports are able to discriminate between specific neutral amino acids and exclude the acidic and basic amino acids. c 1esz ACUIWIC m, IIIC.
Amino acids are actively transported across the plasma membrane of the plant cell by proton-coupled symports (l-3). These transport systems link amino acid accumulation to the protonmotive force generated by a Ptype, H+-pumping ATPase (4, 5). Recently, we used isolated plasma membrane vesicles and imposed proton electrochemical potential differences to study proton amino acid symports found in sugar beet leaf tissue (2). Amino acid transport was carrier-mediated and electrogenie. Moreover, inter-amino acid transport competition suggestedthat at least four amino acid symports, an acidic, a basic, and two neutral, are present in the plant plasma membrane (2,6). We are interested in understanding the biochemistry of these porters because they play a central role in nitrogen distribution in the plant and, since many tissues of the higher plant are heterotrophic, amino acid transport can have a profound influence on growth and development. Previously, we identified two neutral amino acid symports in isolated plasma membrane vesicles based on differences in specific activity, species distribution, and the kinetics of inter-amino acid transport competition (2,6). Perhaps the most striking distinction, however, was the initial observation that isoleucine, valine, and threonine were weak transport antagonists of the other aliphatic neutral amino acids while, in contrast, these amino acids were subject to transport inhibition by all neutral amino acids. Upon inspection, a distinguishing feature of isoleucine, valine, and threonine is that each has a branched methyl or hydroxyl group at their p-carbon. We suggested 519
Inc. reserved.
520
LI
AND
that branching at this position contributes to an unfavorable steric interaction with the active site of one of the symports identified. Thus, we defined Neutral System I as a general porter for all neutral amino acids and Neutral System II as a general porter that specifically excludes isoleucine, valine, and threonine. Resolution of two transport systems based on differential affinities for the neutral amino acids suggests specific molecular interactions must occur between the porters and their respective substrates. In animal cells, a variety of amino acid transport systems have been characterized (7, 8). These systems are distinguished from one another based on differential affinities for specific groups of amino acids and as a function of their dependence upon a cotransported ion. Over the years, detailed studies with amino acid analogues have provided insight into substrate binding to their respective transport proteins (7). For example, some studies have placed Na+ and amino acid binding in close proximity while others have shown that the length, charge distribution, and hydrophobicity of the amino acid side chains are important factors of molecular recognition (9-13). In contrast to animal cells, few molecular characterizations of the amino acid transport systems in plants have been reported. Although the a-amino and the carboxy1 groups have been suggested to be important in substrate recognition (14), little is known about the molecular features that influence amino acid binding to specific symports (3). In the results reported here, we used isolated membrane vesicles and amino acid analogues to investigate structural features of the neutral amino acids that are important for molecular recognition by the neutral amino acid symports. Evidence is presented which shows that the topography of the carboxyl end of the amino acid is very important in molecular recognition and that sidechain branching, length, and polarity also play significant roles. EXPERIMENTAL Plant mute&l.
PROCEDURES
BUSH BSA, pH adjusted to 8.0 with BTP). After differential centrifugation, microsomal pellets were resuspended in 3 ml of 350 mM sorbitol containing 5 mM K-phosphate buffer (pH 8.0) and 0.1 mu DTT. PMVs were purified from the resuspended microsomes by the aqueous twophase partitioning method as described previously (15). Phase purified PMVs were washed by centrifuging in 35 ml of resuspension buffer (330 mM sorbitol, 2 mM Hepes pH adjusted to 8.0 with BTP, 10 mM KC1 and 0.1 mu D’IT) at 50,OOOg for 45 min. The final pellet of plasma membrane vesicles was resuspended in 1 ml of resuspension buffer (4-6 mg protein ml-‘) and used for amino acid uptake experiments.
Transport assays. All assays of amino acid transport activity were carried out at 1O’C. A stable transmembrane pH difference (15) was generated by diluting a small aliquot of PMVs (intravesicular pH 8.0) into 200 pl acidic transport solution (prepared from resuspension buffer by adjusting pH to 6.0 with solid Mes), containing 5 ELM valinomycin and 0.2 to 0.5 gCi of the “‘C-labeled amino acid in the presence or absence of 5 I.IM CCCP. Unlabeled substrate amino acid and amino acid analogs or D-isomers were added to the transport solution to the desired final concentrations. After 2-4 min, while the transport rate was linear (2), the suspended membrane vesicles were collected and washed by a Millipore filtration technique (15). Accumulated radioactivity was measured by scintillation spectrometry. ApH-dependent transport of the radiolabeled amino acid was defined as the difference between the accumulation in the absence and in the presence of CCCP (2,15). Each experiment was repeated at least three times. Protein determination. Proteins were measured by the Markwell method (16) after soluhilization in 0.2% sodium deoxycholate and precipitation in 6.2% trichloroacetic acid (17). BSA Fraction V (Sigma) was used as the protein standard.
RESULTS Inhibition of alanine and isokucine transport by stereoisomers of select amino acids. We examined the effect of different amino acid stereoisomers on isoleucine (Neutral System I) and alanine (Neutral System II) transport to identify macroscopic features of the transported amino acid that may be involved in substrate recognition by the neutral amino acid symports. The results in Table I show
TABLE I StereoisomerEffects on ApH-Dependent Alanine and IsoleucineTransnort
Sugar beet (Beta vulgaris L. cv Great
Western) seeds were geminated in moist vermiculite. When the seedlings were about 1 in. tall, they were transferred to a hydroponic system (two plants per 13 liters), containing one-half strength Hoagland’s solution and 1 mM NaCl. Plants were maintained under a 10-h photoperiod (450 pmol me2 8-i) at 24”C, 70% RH and a 14-h dark period at 18”C, 70% RH. Fully expanded leaves were harvested for experimental analysis.
Preparation of plasma mernbmne vesicles.
Plasma membrane vesicles (PMVs) were prepared at 4’C as described by Bush (15). Briefly, 40 g of leaves was homogenized in 350 ml of grinding solution (250 mM sorbitol, 50 mM Hepes: 10 mM KCl, 3 mM EGTA, 3 mM D’IT, and 1% of
‘Abbreviations used: BSA, bovine serum albumin; BTP, 1,3bis[tris(hydroxymethyl)methylamino]propane; CCCP, carbonyl cyanide m-chlorophenylhydrazone; DTT, dithiothreitol; EGTA, ethyleneglycol bii(/3aminoethyl ether)NJJ’-tetraacetic acid; Hepes, 4-(2-hydroxyethyl) l-piperazineethanesulfonic acid; Mes, 2-(N-morpholino)ethanesulfonic acid, PMV, plasma membrane vesicle; T-P-L, tabtoxinine-@-lactam
Transport
substrate
Competitor
L-Alanine
L-Isoleucine
Control L-Alanine D-Alanine L-Isoleucine D-Isoleucine L-Serine D-Serine
100% 101+ 94* 107 + 762 103 +
100% 50+ 91+ 100* 58rt 87 zk
11 7 9 5 7
6 5 9 7 10
Note. Transport solutions contained 0.1 mM [“Clalanine or [“C]isoleucine and 0.5 mhf of the competing amino acid. ApH-dependent transport activity was defined as the difference between accumulation in the absence and that in the presence of 5 pM CCCP. Results are expressed as percentage of control + SD (n = 4). 100% = 1.5 nmol alanine or 0.09 nmol isoleucine/mg protein/min.
SUBSTRATE
RECOGNITION
BY PLANT TABLE
ApH-Dependent
Alanine
and Isoleucine
Transport
PROTON-AMINO
521
ACID SYMPORTS
II
in the Presence
of Amino
Acid Analogs
Containing
Distinct
R-Groups
Substrate Amino acid analogs Control 0 II CH,-C-0-CH,-CH-COOH
0-Acetylserine
0 II N=N=CH-C-0-CHZ--CH-COOH
Azaserine
Alanine
Isoleucine
100%
100%
46 2 1
46 2 2
29 rt 1
39 * 5
44 f 2
59 + 3
58 2 8
26k 1
53 2 2
58 + 6
40 2 2
16 + 2
32 + 6
24 2 4
61e4
65 2 5
I NHz
(t
0-Benzylserine
/
0-Methylserine
-
\
CH2-0-CH2-CH-COOH
I NH2 CH,-0-CH,-CH-COOH
I NH2 p-Nitrophenylalanine
(0.1 mM)*
NO2
CH,-CH-COOH
NH2 CH3-CH2-CH2-CH-COOH
Norvaline
I NH2
Norleucine
CH3-CH2--CH2-CH,-CH-COOH NH2
Methionine
sulfoxide
CH,-S-CH,-CH,-CH-COOH d
I NH2
Note. Transport solutions included 0.1 mM [i4C!]alanine or [i4C]isoleucine in the absence (control) or presence of 0.5 mM amino acid analogue*. ApH-dependent transport activity was expressed as percentage of the control 2 SD (n = 3).
that L-serine and L-alanine inhibit alanine and isoleucine transport while L-isoleucine was not effective against alanine. In contrast, D-isomers of alanine, isoleucine, and serine had no effect on alanine transport while isoleucine transport was decreased by approximately 10%. These results suggest both Neutral Systems I and II are specific for L-amino acids, although Neutral System I may exhibit some recognition of the D-isomers. Inhibition of alanine and isoleucine transport by various amino acid anulogues. The effect of several amino acid analogues on alanine and isoleucine transport was analyzed to determine what portion of the amino acid molecule might be important in molecular recognition. Three major classes of analogues were examined, including (i) hydrophobic or hydrophilic R group additions, (ii) replacement of the a-carbon amine with a spacefilling group or moving it to the P-carbon, and (iii) methylation at the a-carbon, at the a-amino group, or at the carboxyl group.
The results in Table II report the effect of various R group configurations on alanine and isoleucine transport. In every case, these analogues were effective transport inhibitors. This was true for both hydrophobic and hydrophylic additions, as well as for large ringed structures. These data suggest the binding sites of Neutral Systems I and II are not very restrictive with respect to the distal portion of the amino acid. It should be noted, however, that the presence of charged groups in this region, such as found in the acidic and basic amino acids, significantly reduces transport antagonism (2). The second class of amino acid analogues examined was chosen to explore the positional significance of the a-amino group in substrate recognition (Table III). Analogues lacking the a-amino group were not active transport antagonists. This observation supports the notion that the amino group plays an important role in substrate recognition. In addition, substitution of the amine with
522
LI
ApH-Dependent
Alanine and Isoleucine Transport
AND
BUSH
TABLE
III
in the Presence of Amino Acid Analogs Containing Substituted Carboxyl and Amino Groups Substrate
Amino
acid analogue
Alanine
Control
Isoleucine
100%
&Alanine
NH2 -
N-Methyl-k&nine
CHx -
CHr -
COOH
100%
104 + 10
CH3--CH-COOH
83+
7
91+
6
70 + 10
lOO+
8
84+
9
101 f
3
101+
7
91*
5
1Olk
7
110+
7
118 + 10
113 f
7
92+
6
108+
7
97+
6
102 *
2
106 f
3
I NH I CHs CH3
2-Aminoisobutyrate
(AIB)
CH3 -
DL-Isoserine
C - COOH I NHs
NHr-CH,-CH-COOH I OH
Dl&Aminoisobutyrate
NH*
-CHr
-CH-COOH
I CH3
Glycolic
acid
HO-CHr-COOH
Sarcosine
CHs -NH-CHx
-COOH CHs
2-Methylaminoisobutyrate
CHx -
C -
COOH
I NH
I CH, 0
II L-Serine
methylester
HO-CH,-CH-C-0-CHx
I NHz Note.
Experimental
conditions
and expression
of transport
activity
a space filling methyl or hydroxyl group, or placing the amine within the linear side chain, did not restore activity. This suggests the charge status of the amino group, or at least its ability to form specific ligand interactions, is also a critical factor in molecular recognition. We also examined amino acid analogues which shifted the amine to the @-carbon. &Aminoisobutyrate and isoserine showed little or no inhibition of either alanine or isoleucine transport (Table III). In contrast, @-alanine had no influence on alanine transport, yet it decreased isoleucine transport by 20%. These data provide further evidence in support of the conclusion that the a-amino group is
were the same as in Table
II.
important in molecular recognition and, moreover, they highlight the significance of the positional relationship of the amino group with respect to the rest of the molecule. The last class of analogues examined included those with specific substitutions at the o-amino, a-carbon, and carboxyl positions (Table III). Methylation of the a-carbon or a-amino group practically eliminated transport antagonism. For example, 2-aminosiobutyric acid (AIB) and N-methylalanine inhibited isoleucine transport to a limited extent while they had no significant impact on alanine transport. Likewise, methylation of a carboxyl oxygen (serine methylester) eliminated transport inhi-
SUBSTRATE
llleucine
g
E
6.0
-
RECOGNITION
BY
PLANT
(PM)
B
llleucine
(PM)
FIG. 1.
Double reciprocal plot of the kinetics of ApH-dependent leutine transport in the presence of (A) azaserine and (B) p-nitrophenylanine at 0 pM (O), 100 pM (0), and 200 pM (M).
bition. Taken together, the impact of the analogues listed in Table III on alanine and isoleucine transport emphasizes the importance of the a-amino and carboxyl end of the molecule in substrate recognition. Competitive inhibition of leucine transport by azaserine and p-nitrophenylalanine. We examined the kinetics of leucine transport inhibition by azaserine and p-nitrophenylalanine to test the assumption that these analogues are competing for binding to the active site of Neutral System II. Double reciprocal plots of azaserine and pnitrophenylalanine inhibition of leucine transport were diagnostic of competitive inhibition (Fig. 1) and Ki values, estimated from Dixon plots, were approximately 75 PM for both analogues. These results are consistent with the notion that these amino acid analogues are competing for binding at the active site of the porter. Effect of amino acid length on alunine transport. In animal cells, Oxender and Christensen (9) observed increased transport antagonism against amino acid Transport Systems A and L activity as the length of the competing amino acid’s apolar R group increased. To determine if parallel molecular interactions occur for Neutral System II, we tested several amino acids and analogues, at a fixed concentration, for their ability to inhibit alanine transport. Glycine, which has no carbon in the R group side chain, inhibited alanine transport by 30% while norleucine, which contains 4 carbons in the R group, in-
PROTON-AMINO
ACID
523
SYMPORTS
hibited transport by approximately 70% (Fig. 2). When the side chain contained one carbon (alanine), transport competition was approaching maximum. These observations implicate the length of the R group as a contributing factor in substrate binding to the active site. Effect of branching patterns on alanine and isolewine transport. We reported previously that amino acid branching at the P-carbon appears to be an important structural property which differentiates the neutral amino acids with respect to transport through Neutral Systems I and II (2, 6). In the current investigation, a number of amino acids and amino acid analogues representing a variety of branching patterns were tested for their relative abilities to inhibit isoleucine (Neutral System I) and alanine (Neutral System II) transport. As the branching position moved away from the carboxyl carbon, transport inhibition increased from zero to 40% for both porters (Fig. 3). It is noteworthy, however, that different patterns of transport antagonism were observed. Alanine transport was progressively inhibited as the branch point moved away from the carboxyl end of the analogues. In contrast, isoleucine transport decreased rapidly with maximal inhibition evident with branching at the ,&carbon. These results are consistent with our previous suggestion that amino acid binding to Neutral System II is limited by a comparatively restricted molecular space associated with the B-carbon. DISCUSSION
Two neutral amino acid transport systems are present in plasma membrane vesicles isolated from sugar beet leaf tissue (6). These proton-coupled symports were initially resolved because of their differential affinities for isoleutine, valine, and threonine. We suggestedthat the presence of a methyl or hydroxyl group at the P-carbon of these amino acids blocks accessto the active site of Neutral Sys-
0 % 2 I= 2 $
60-
40-
u z
l
+Serine
20 .-0” z =E
0,
I 0.0
1.0
Number
FIG. 2.
1 2.0
of Atoms
I 3.0
I 4.0
in the Sidechain
Inhibition of ApH-dependent alanine transport as a function of antagonist sidechain length. Transport was assayed after 3 min with 0.1 mM [“Clalanine and 0.5 mM antagonist. Specific antagonists were glycine (0 atoms), alanine (1 atom), serine (2 atoms), norvaline (3 atoms), and norleucine (4 atoms).
524
LI
=::
AND
50
2 F lQ = B 2 a
40
30
20
z .-5
10
E .c +
0
ae
-I 0
1
Number
2
3
of Atoms Away Carboxyl Carbon
4
from
5
the
FIG. 3. Effect of antagonist sidechain position on ApH-dependent alanine (0) and isoleucine (0) transport. Transport assays were as described in Fig. 2. Amino acids representing specific branching positions were 2-aminoisobutyrate, isoleucine (or valine in the Ile transport), leutine, and methionine sulfoxide, respectively.
tern II and, based on this hypothesis, we became interested in identifying structural elements of the neutral amino acids that are important in molecular recognition. Stereospecificity is a characteristic of enzyme-mediated reactions which provides macroscopic information about the three-dimensional requirements for substrate orientation within the enzyme active site. Since Neutral Systems I and II discriminate against D-iSOmerS of their respective transport substrates (Table I), we propose that the a-amino and carboxyl groups of the neutral amino acids are associated with specific ligands in the active sites of these porters and that these ligands must be asymmetrically positioned. Furthermore, stereospecificity requires at least three unique binding sites for the groups surrounding the chiral carbon (18). Therefore, the position of the R group must also play an important role in molecular recognition. Stereospecificity in amino acid transport was also reported in cultured tobacco cells (19, 20) and corn scutellum (21). However, the degree of sensitivity to stereoisomers of the neutral amino acids may vary from one species to another. For example, Jung and Luttge (22) and Felle (14) used electrophysiological evidence to conclude that D- and L-amino acids are transported by the same carrier. Stereospecificity suggests the positional relationship between the a-amino and the carboxyl groups is an important parameter in molecular recognition. This conclusion is consistent with the results generated with the pamino acid analogues. /3-Alanine was not an active antagonist of amino acid transport through Neutral System II and it was only marginally effective against Neutral System I (Table III). Likewise, isoserine and /3-aminoisobutyrate were not active against either porter (Table III). These results suggest a critical distance between the (Yamino and the carboxyl groups must be maintained for
BUSH
competitive binding to the active site. Taken together with the three-dimensional restrictions imposed by stereospecificity, the region of the active site associated with the carboxyl end of the neutral amino acids appears to be a very well-defined space which plays an important role in substrate recognition. Similar conclusions have been reached for a number of amino acid transport systems in animal cells (7, 13) and in an aquatic liverwort (14). Analogues with methyl additions to either the a-amino or the carboxyl groups were not effective transport inhibitors, although Neutral System I exhibited some sensitivity to N-methylalanine. These data support the notion that the a-amino and carboxyl groups are important participants in substrate recognition and provide further insight into the molecular interactions associated with the enzyme-substrate complex. Methylation of the (Yamino or carboxyl groups interferes with several reactions of potential significance. For example, methylation may simply produce a steric interaction which restricts access to the active site. This kind of hindrance would be the result of a relatively confined binding cavity which does not accommodate extraneous atoms. This idea is consistent with the positional restrictions placed on the system as a result of stereospecificity. Additionally, several forces are involved in substrate binding to enzyme active sites, including van der Waals interactions, H-bonds, hydrophobic effects, and electrostatic bonds. Presumably, methylation of the a-amino or carboxyl groups also eliminates electrostatic linkages and H-bonds which may be important in substrate recognition and binding. This conclusion, which assumes the amino acids are zwitterionic while bound, is supported by the observation that substitution of the a-amino group with a space filling methyl or hydroxyl group produces a completely inactive analogue (Table III). We conclude from these data that specific positional and chemical requirements must be satisfied for successful binding to occur. Previously we argued that global features of the neutral amino acids cannot account for differences in substrate recognition by Neutral Systems I and II (6). This conclusion was supported in the current investigation. Amino acid analogues with a variety of R groups were effective antagonists of amino acid flux through both transport systems (Table II). Since these R groups included hydrophobic and hydrophylic constituents, as well as large ringed structures, it appears that molecular interactions with the distal portion of the neutral amino acids is relatively unrestricted. The exception to this generalization, i.e., when there is a branch point at the P-carbon, is the distinguishing molecular feature of those amino acids excluded from Neutral System II. It should be noted, however, that there were subtle differences between Neutral Systems I and II in their relative sensitivity to the analogues in Table II (e.g., norvaline and o-methylserine) and, perhaps more significantly, these symports were comparatively insensitive to transport antagonism by the
SUBSTRATE
RECOGNITION
BY
PLANT
acidic and basic amino acids which place charged groups in the distal portion of the molecule (2). Thus, although molecular interactions with this region of the amino acid are relatively unconstrained, there are restrictions which effectively limit amino acid binding to the neutral symports as a function of the distal region of the molecule. In animal cells the number of atoms (including carbon, sulfur, and oxygen) in the linear side chain is an important factor in substrate binding. For example, amino acids and analogues with four or five carbons in the side chain have the highest affinity for Transport Systems ASC and Ly+ (11-13). Similarly, amino acid analogues against System A and System L were more effective as the length of the side chain increased to seven atoms (9). In the present investigation, the sensitivity of Neutral System II to analogues of increasing length was similar to that observed for System A and System L. Transport antagonism increased from zero carbon (glycine) to four carbons (norleucine) in the side chain (Fig. 2). If oxygen atoms in the side chain are considered roughly equivalent in length to carbon, serine could be equated with a two-carbon side chain analogue. Note, however, that its activity as a transport antagonist was lower than that of alanine (Fig. 2). This observation suggests polarity can also play a role in molecular recognition in this region of the amino acid. We suggested in our earlier studies that tolerance to branching at the P-carbon is an important property differentiating Neutral System I from Neutral System II. We noted that branching at this position places a dense electron cloud in close proximity to the a-amino group and concluded that this results in unfavorable steric interactions with the binding cavity of Neutral System II. The experiments reported here extend this hypothesis. Methylation at the a-amino, a-carbon, or P-carbon position increases the electron cloud distribution at the carboxy1 end of N-methylalanine, 2-aminosiobutyrate, and valine. These amino acids exhibited little or no inhibition of alanine transport yet isoleucine transport was inhibited from 16 to 40% (Table 3 and Fig. 3, (2)). These results support the idea that the binding cavity associated with the (Y- and P-carbons is, comparatively, more restricted in Neutral System II than in Neutral System I. When the side chain is branched at the P-carbon, antagonist activity against Neutral System I approaches maximum. In contrast, activity against Neutral System II is very low and only increases as the branch point moves out from the pcarbon (Fig. 3). An underlying assumption of the experiments reported here is that the transport antagonists were competing for substrate binding at the active sites of the neutral amino acid symports. The kinetics of Neutral System II transport inhibition by azaserine and p-nitrophenylalanine support this conclusion (Fig. 1). Additionally, our previous studies of inter-amino acid transport competition with isolated vesicles did not reveal evidence of allosteric regulation of Neutral System I or II (6). Therefore, even
PROTON-AMINO
ACID
SYMPORTS
525
though the kinetics of transport inhibition were not examined for every analogue, we believe the data presented here are the results of specific molecular interactions associated with substrate recognition. However, since these porters are secondary active transport systems that are dependent on a stable transmembrane proton electrochemical potential, it is possible that indirect effects on this driving force could interfere with amino acid flux. We examined this possibility in earlier experiments by determining the effect of simultaneous sucrose transport on amino acid flux. Sucrose transport is also mediated by a proton-coupled symport in these membrane preparations (15) and, thus, if competition for the protonmotive force is a reasonable explanation for our observations, we should see sucrose-dependent changes in the transport rate. Such changes were not observed and, therefore, we do not attribute changes in the transport rates to indirect competition for the driving force. Taken together, we conclude that these amino acid analogues are indeed competing with the native substrate for binding to the porter active sites. The structure-activity relationships reported here provide an experimental basis for correlating specific molecular interactions with substrate binding to the neutral amino acid symports. Moreover, these data provide a conceptual framework for constructing a working model of the binding sites of these porters. Three general features of this model stand out. First, that region of the protein which binds the carboxyl end of the amino acid is characterized by a well-defined cavity which is structurally limited, stereospecific, and where successful binding is driven by specific ligand interactions. This site does not tolerate extraneous atoms or space filling substitutions which eliminate or reduce required binding reactions. The second feature of this model is the relatively unrestricted binding site for the distal portion of the amino acid. This section of the porter has few physical barriers which interfere with substrate binding, although it is important to recall that this region is sensitive to polar interactions and binding is weak when charged groups are present. The third feature of this model, which separates Neutral Systems I and II, focuses on the transition zone between the carboxyl and the distal regions of the amino acids. Here, Neutral System II is characterized by steric interactions which block effective binding of amino acids containing substitutions on the P-carbon. In contrast, amino acid binding to Neutral System I is not as structurally confined in this region. Taken together, a model is emerging which describes a tight fitting binding pocket for the carboxyl end of the amino acid, fewer structural limitations around the P-carbon, and a wide open cavity which binds the distal region. Perhaps the carboxyl end of the molecule is buried within the porter with the distal portion of the amino acid inside a large channel leading to the carboxyl binding pocket.
526
LI
A
OH
NH,
I
I
AND
ing the knowledge obtained in this study to identify potential photoaffinity probes, such as azaserine and p-nitrophenylalanine, to help identify these important transport systems in the plant.
-CH,--CH,-C-COOH H2P-i:
,B-“\\,
BUSH
I
ti
REFERENCES 1. Bush, D. R., and Lang&on-Unkefer, 487-490.
B
0
2. Li, Z-C., 3. Reinhold, 45-83.
II OH-P-CH,-NH-CH,-COOH I OH FIG.
4.
Structure
of tabtoxinine-p-lactam
(A) and glyphosate
(B).
This model is useful because it can be used to predict whether exotic biological compounds are transported through the neutral amino acid symports. For example, tabtoxinine+lactam (T-@-L) is a toxic a-amino acid analogue produced by several closely related pathovars of Pseudomonas syringae. This compound is an irreversible inhibitor of glutamine synthetase and this activity is directly involved in eliciting the chlorotic symptoms these pathogens induce in their hosts (23). Recently, T-P-L transport into plant cells was shown to be mediated by an amino acid transport system (1). That result is consistent with the conclusions drawn here because the structural features of T-P-L (Fig. 4) suggest it can be transported through either Neutral System I or II. This conclusion is supported by the observation that T-0-L transport into cultured cells was inhibited by alanine (24). In a contrasting example, glyphosate is a widely used herbicide whose transport into the cell has been suggested to be mediated by an amino acid porter (25). However, it does not contain the necessary structural elements required for recognition by either the neutral amino acid symporters or (by inference) the acidic or basic porters (Fig. 4). Indeed, recent experiments examining glyphosate transport into isolated plasma membrane vesicles have failed to demonstrate symport mediated flux (Riechers and Bush, unpublished data). The results reported here provide important clues about the molecular interactions involved in substrate recognition by the plant neutral amino acid symports. Future advances will require information about the membrane proteins mediating these reactions. Therefore, we are us-
P. J. (1988)
Plant
Phyaiol.
88,
and Bush, D. R. (1990) Plant PhysioL 94,268-277. L., and Kaplan, A. (1984) Annu. Reu. Plant Physiol.
35,
4. Sze, H. (1985) Annu. Reu. Plant Physiol. 36,175208. 5. Serrano, R. (1989) Annu Rev. Plant Physiol. Pkmt 61-94.
Mol.
BioL 40,
6. Li, Z-C., and Bush, D. R. (1991) Plant Physiol. g&1338-1344. 7. Christensen, H. N. (1989) in Methods in Enzymology (Fleischer, S., and Fleischer, B., Eds.), Vol. 173, pp. 576-616, Academic Press, San Diego. M. E., and Kilberg, M. S. (1999) J. Biol. Chem. 265, 8. Mailliard, 14,321-14,326. 9. Oxender, D. L., and Christensen, H. N. (1963) J. Biol. Chem. 238, 3686-3699. 10. Thomas, E. L., and Christensen, H. N. (1970) Biochim. Biophys. Acta 40,277-283. E. L., and Christensen, H. N. (1971) J. Biol. Chem. 246, 11. Thomas, 1682-1688. H. N., Thomas, E. L., and Handiogten, M. E. (1969) 12. Christensen, Biochim. Biophys. Acta 193, 228-230. 13. Christensen, H. N. (1979) Advance 14. Felle, H. (1981) Planta 152,505-512.
Enzymol.
49,41-101.
15. Bush, D. R. (1989) Plant Physial. 89,1318-1323. M. K., Haas, S. M., Tolbert, N. E., and Bieber, L. L. 16. Markwell, (1981) in Methods in Enzymology (Lowenstein, J. M., Ed.), Vol. 72, pp. 296-303, Academic Press, San Diego. A., and Weinstein, D. (1976) Anal. Biochem. 70, 24117. Bensadoun, 250. Structure and Function, 2nd ed., pp. 224, 18. Fersht, A. (1984) Enzyme Freeman, New York. 19. Harrington, 378. 20. Blackman, 261-266.
H. M., and Henke, M.
S., and McDaniel,
R. R. (1981)
Plant
C. N. (1980)
PhysioL Plant
67,373-
Physiol.
66,
21. Steward, C. R. (1971) Plant Physiol. 47,157-161. 22. Jung, K. D., and Luttge, U. (1980) Planta 150,230-235. R. D. (1971) in Morophological and Biochemical Events in 23. Durbin, Plant Parasite Interactions (Akai, S., Ouchi, S., Eds.), pp. 369-385, Mochizuke Publishing Co., Japan. P. J. (1987) Plant PhysioL 85, 24. Bush, D. R., and Langston-Unkefer, 845-849. E. D., Widholm, J. M., and Slife, F. W. (1983) Plant PhysioL 25. Nafziger, 71,623-626.