This Article Abstract Full Text (PDF) All Versions of this Article: biophysj.106.095208v1 92/7/2570    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Nakagawa, Y. Articles by Takahashi, H. Search for Related Content PubMed PubMed Citation Articles by Nakagawa, Y. Articles by Takahashi, H.

Molecular Analysis of TCR and Peptide/MHC Interaction Using P18-I10-Derived Peptides with a Single D-Amino Acid Substitution

Yohko Nakagawa ^*, Hiroto Kikuchi  and Hidemi Takahashi ^*

^* Department of Microbiology and Immunology and Department of Physics, Nippon Medical School, Tokyo 113-8602, Japan

Correspondence: Address reprint requests to Hidemi Takahashi, Department of Microbiology and Immunology, Nippon Medical School, 1-1-5 Sendagi, Bunkyo-ku, Tokyo 113-8602, Japan. Tel.: 81-3-3822-2131 ext. 5381; Fax: 81-3-3316-1904; E-mail: htkuhkai{at}nms.ac.jp.

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
For the structural analysis of T-cell receptor (TCR) and peptide/MHC interaction, a series of peptides with a single amino acid substitution by a corresponding D-amino acid, having the same weight, size, and charge, within P18-I10 (aa318–327: RGPGRAFVTI), an immunodominant epitope of HIV-1 IIIB envelope glycoprotein, restricted by the H-2D^d class I MHC molecule, has been synthesized. Using those peptides, we have observed that the replacement at positions 324F, 325V, 326T, and 327I with each corresponding D-amino acid induced marked reduction of the potency to sensitize targets for P18-I10-specific murine CD8⁺ cytotoxic T lymphocytes (CTLs), LINE-IIIB, recognition. To analyze further the role of amino acid at position 325, the most critical site for determining epitope specificity, we have developed a CTL line [LINE-IIIB(325D)] and its offspring clones specific for the epitope I-10(325v) having a D-valine (v) at position 325. Taking advantage of two distinct sets of CD8⁺ CTLs restricted by the same D^d, three-dimensional structural analysis on TCR and peptide/MHC complexes by molecular modeling was performed, which indicates that the critical amino acids within the TCRs for interacting with 325V or 325v appear to belong to the complementarity-determining region 1 but not to the complementarity-determining region 3 of Vβ chain.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Immune responses to viral infection include both humoral and cell-mediated effector mechanisms. The major effector cells in cellular immunity are CD8 molcule-expressing cytotoxic T lymphocytes (CTLs) that can recognize and kill virus-infected cells. In general, endogenously synthesized antigens such as virus-derived proteins are fragmented inside of the cells and are presented on the cell in conjunction with class I major histocompatibility complex (MHC) molecules. Such processed epitope peptides associated with the class I MHC molecules can be recognized by CTLs via their specific T cell receptors (TCRs).

The TCRs expressed on the cell surface of T lymphocytes contain similar structural patterns with immunoglobulin-like domains, comprising one variable and one constant, as well as a transmembrane domain and a short cytoplasmic tail. The specificity for T-cell recognition seems to be determined by the variable domains, TCR V and TCR Vβ, within two heterodimeric subsets, TCR and TCRβ. Several recent findings have indicated that the TCR and β heterodimers are oriented to the long axis of the epitope-peptide/MHC complex (1), in which the V domain appears to cover the amino-terminal half of the epitope peptide, whereas Vβ is located over the carboxyl-terminal portion of the epitope (2).

Among those variable V and Vβ domains, three hypervariable complementarity-determining regions (CDRs), termed CDR1, CDR2, and CDR3, seem to directly interact with the peptide/MHC complex. Because the degree of variability is the greatest in the CDR3 loop generally, and it is positioned more closely over the epitope peptide than other CDR1 and CDR2 loops, the antigen specificity has been considered to be associated with the CDR3 but not with CDR1 or CDR2, which were predicted to interact principally with the MHC molecules (3,4). Indeed, according to a recent report on the murine K^b class I MHC molecule-restricted epitope octapeptide (pKB1: KVITFIDL) recognized by KB5-C20 TCR (5), TCR plasticity is primarily restricted to the CDR3 loops of the Vβ domain. Nevertheless, recent crystallographic analyses on various TCR and peptide/MHC interactions have suggested the possibility of direct contact for both CDR1 and CDR3 in the TCR and TCRβ chains with the antigenic peptide/MHC complex (5,6). Therefore, to understand more precise molecular interactions determining T-cell specificity through TCR-mediated peptide/MHC complex recognition, we took advantage of the following known materials to accomplish the analysis.

We have established CD8⁺, H-2D^d class I MHC molecule-restricted murine CTL line, LINE-IIIB, specific for the envelope glycoprotein 160 (gp160) composed of 900 amino acids derived from one of the most commonly used IIIB strains of human immunodeficiency virus type-1 (HIV-1), a causative agent for acquired immunodeficiency syndrome (AIDS) (7). Then, we have identified an immunodominant epitope within the gp160 as a 15-residue peptide, P18IIIB (aa315–329: RIQRGPGRAFVTIGK), for the LINE-IIIB recognition (7) as well as the minimal active 10-residue peptide, P18-I10 (aa318–327: RGPGRAFVTI) within P18IIIB (8). Moreover, although the position of P18IIIB is located in the hypervariable portion (termed V3 domain) of the viral envelope, the site has turned out to be recognized by various isolate-specific CTLs in an isolate-specific manner (9,10), and a number of distinct class I MHC molecules did present the P18IIIB to each specific CTL (11). Furthermore, the P18IIIB in the V3-domain was found to be overlapped with the major determinant sites for neutralizing antibodies against HIV-1 in an isolate-specific manner (12–14) and also to be recognized by CD4-positive helper T lymphocytes specific for HIV-1 in a class II MHC molecule-restricted manner (8,15). In addition, human CTLs did see the P18IIIB when presented by HLA A2 and A3 (16). These findings indicate that the P18IIIB appears to be a highly attractive epitope for the development of peptide-based vaccine against AIDS, and thus, it is important to study the precise interaction between the epitope P18IIIB and their specific TCRs to study the manner of T-cell-mediated immune responses.

Using a series of peptides with an alanine (A) substitution at each position, we observed that amino acids at positions 322R and position 324F were critical for D^d binding and that position 325V within P18-I10 was essential for interacting with TCRs (17). Also, C-terminus 327I appears to be critical for D^d binding to form the D^d-binding motif (8,18). In addition, we found the HIV-MN isolate-specific CTLs also saw the corresponding minimal active site, MNT10 (aa318–327: IGPGRAFYTT) in association with the same D^d molecules, and replacement of just a single residue, 325V with 325Y, within the P18-I10 or vice versa within the MNT10 was sufficient to reciprocally interchange the specificities for these two non-cross-reactive sets of CTLs (9,10). Thus, a single side chain at position 325 can play a critical role in determining the epitope specificity within both P18-I10 and MNT10 presented by the same class I MHC molecule D^d for CD8⁺ CTL TCR recognition mediated by the CDRs.

It has been reported that the charge of the amino acid might also affect interaction between TCRs and peptide/MHC complexes (19). Indeed, when negatively charged glutamic acid (E) at position 436 within HIV-1-envelope-derived helper T-cell epitope T1 (aa428–443; KQIINMWQEVGKAMYA) (20) was substituted with either uncharged alanine (A) or size-conservative, uncharged glutamine (Q), stimulatory capacity of the substitute peptides for T1-specific T hybridomas was significantly enhanced, although charge-conservative aspartic acid (D) substitution did not show any enhancement (21). Moreover, substitution at position 6Q in the immunodominant CTL epitope for vesicular stomatitis virus (RGYVYQGL, VSV8) presented by K^d class I MHC molecules to a negatively charged residue such as 6E or 6D induced a change at position 93S of TCR CDR3 to a positively charged residue 93R or 93K (22). Therefore, particularly to reduce the influence of charges of each amino acid as well as the size and molecular weight on TCR-mediated recognition, a series of P18-I10-derived peptides with a single amino acid substitution by D-amino acid at each corresponding site have been synthesized. Using those D-amino acid–substituted peptides, we found apparent reduction of specific cytotoxic activity in the 325V-specific LINE-IIIB cells by the replacement of 325V with D-type valine at position 325, represented as I10(325v). Then, we attempted to establish CTL lines specific for the epitope I-10(325v) by immunization with dendritic cells pulsed with the peptide, I10(325v) (23). We have successfully generated both a 325(v)-specific CTL line and clones that did not cross-react with the original P18-I10.

Taking advantage of two distinct sets of CD8⁺ CTL clones specific for either P18-I10 bearing L-type valine or I10(325v) having D-type valine at position 325 presented by the same MHC molecules D^d, we attempted to study the three-dimensional (3D) structural analysis on TCRs and peptide/MHC complexes. Here, using molecular modeling analysis, we would like to show that the critical amino acids for interacting with P18-I10 in determining epitope specificity appear to be the peptide DMSHET within CDR1 of Vβ7, whereas those for interacting with I10(325v) appear to be TNSHNY within CDR1 of Vβ8.3.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Mice
Female BALB/c (H-2^d) mice were purchased from Charles-River Japan Inc. (Tokyo, Japan). The mice were 6 to 10 weeks of age and were maintained in a specific-pathogen-free environment. All experiments were performed according to the guidelines of the NIH Guide for the Care and Use of Laboratory Animals.

Synthetic peptides
Peptides were synthesized and purified as described previously (24). Table 1 summarizes the peptides used in this study; L-amino acids are represented as capital letters, and D-amino acids as lower-case letters.

Monoclonal antibodies
We used fluorescein isothiocyanate-conjugated antimouse CD3 (145-2C11), βTCR (H57-597), CD4 (RM4-5), and CD8 (53-6.7) monoclonal antibodies (PharMingen, San Diego, CA) to determine the cell surface molecules of the established CTL lines and clones.

CTL lines and clones
The P18-I10-specific CTL line, LINE-IIIB, was generated as described previously (7). Based on a previously reported procedure (23), a CTL line specific for I10(325v) was generated from spleen cells of BALB/c mice immunized with I10(325v)-pulsed splenic dendritic cells. Briefly, immune spleen cells were restimulated in vitro with mitomycin C-treated I10(325v)-pulsed syngeneic BALB/c.3T3 fibroblasts in 24-well culture plates containing 1.5 ml of complete T-cell medium composed of RPMI 1640 medium supplemented with 2 mM L-glutamine, 50 µM 2-ME, 100 U/ml penicillin, 100 µg/ml streptomycin, 10% heat-inactivated FCS, and 10% Rat T-STIM (Collaborative Biomedical Products, Bedford, MA). To establish CTL lines, the generated CTLs were maintained by biweekly stimulation with the mitomycin C-treated I10(325v)-pulsed Neo and were termed LINE-IIIB(325D) cells. CTL clones were established from bulk CTL lines using a limiting dilution technique in 96-well U-bottomed microplates, as described previously (15).

CTL assay
The cytolytic activity of the CTL lines and clones was measured, as previously described (27), using a standard 5-h ⁵¹Cr-release assay with various ⁵¹Cr-labeled targets, as indicated in the figure legends.

Flow cytometric analysis
Flow cytometric analysis was performed to determine the surface molecule expression of the established CTL lines and clones using a FACScan analyzer (Becton Dickinson Immunochemical Systems, Mountain View, CA). We harvested 5 x 10⁵ cells, washed them twice with serum-free RPMI 1640, and then pelleted them. Fluorescein isothiocyanate-conjugated monoclonal antibodies were added to pellets, and they were then incubated for 30 min at 4°C. Then, the cells were washed three times and resuspended with phosphate-buffered saline (PBS) containing 0.1% bovine serum albumin and 0.1% sodium azide. Dead cells were gated out by forward and side scatter based on propidium iodide uptake. Ten thousand events were acquired for each sample and analyzed using Cell Quest software (Becton Dickinson, Franklin Lakes, NJ).

mRNA extraction, reverse transcription, and PCR amplification
Poly(A) tail-bearing mRNA was isolated from CTL clones using the Fast Track mRNA Isolation Kit (Invitrogen, Carlsbad, CA) according to the manufacturer's instructions. Reverse transcription and PCR amplification of purified mRNA were performed with a GeneAmp RNA PCR Kit (PE Biosystems, Foster City, CA). For the synthesis of T-cell receptor Vβ-specific cDNA, Vβ7-specific primer (ACATCCCTAAAGGATACAGGG) and Vβ8-specific primer (ATATCCCTGATGGGTACAAGG) were used in conjunction with Cβ primer (CCGATGGGAGCACACGAACCCTTAAG). For the synthesis of T-cell receptor V-specific cDNA, V2-specific primer (AGCAATTCTGAACTGCAGTTA), V3-specific primer (CAGCCCGATGCTCGCGTCACT), and V16-specific primer (ATGGACTGTGTGTATGAAAC) were used in conjunction with C primer (ACTGGACCACAGCCTCAGCGTC).

DNA sequence analysis
PCR products were separated by electrophoresis on a 1% of agarose gel and sliced bands were purified by EASYTRAP glass powder (Takara Bio., Siga, Japan). We then analyzed the purified DNA by a direct-sequence technique, using the ABI PRISM Big Dye Terminator Cycle Sequencing FS Ready Reaction Kit (PE Biosystems), and we analyzed the sequences on the ABI PRISM 377XL DNA Sequencing System (PE Biosystems).

Molecular modeling
Because the actual 3D TCR structures for both Vβ7 and Vβ8.3 were unknown, we carried out comparative and homology modeling to elucidate the spatial relation between TCRβ (Vβ7 or Vβ8.3) and P18-I10 or I10(325v). Comparative modeling predicts the 3D structure of a given protein sequence (target) primarily on the basis of its alignment to one or more proteins of a known 3D structure (templates). It is usually difficult to accurately determine the 3D structure from a protein sequence by theoretical procedures. However, because the 3D TCRβ structure consists of a sandwich comprising a four-stranded antiparallel β-sheet and a three-stranded antiparallel β-sheet that are linked by a disulfide bond, and because the core portion was tightly bound by a hydrogen bond, the 3D structure of the TCRβ, especially the core portion, could be reliably predicted.

Following the above procedure, the so-called threading or 3D template-matching method (28) could be implemented to select templates. For this procedure, we used LIBRA (29) (http://www.ddbj.nig.ac.jp/search/libra_i-e.html) software in which compatible structures of a target sequence are sought from the structural library chosen from the Protein Data Bank (PDB), and the target sequence and 3D profile are aligned by simple dynamic programming. According to the alignment, sequence remounts on the structure and its fitness are evaluated by the pseudoenergy potential. The scores are then sorted from the best-matched templates and shown along with their alignments. Based on the obtained alignments between the template and the target Vβ sequence, a 3D model is calculated by the MODELLER software (30–33), by which five 3D models were obtained from five templates used.

To examine the interaction between the calculated TCR and P18-I10, TCR/peptide/class I MHC complex (PDB code 1kj2) was used for two reasons. First, the 3D structure of the MHC part of 1kj2 is almost identical to that of MHC used in our experiment, which is also resolved (code 1bii) and shows a H-2D^d class I MHC molecule presenting the HIV-derived peptide P18-I10 (RGPGRAFVTI) (34). The amino acid sequence identity between 1kj2 and 1bii is 90%. Second, 1kj2 is a TCR/peptide/MHC complex (5), whereas 1bii is only a peptide/MHC complex. In our 3D models, Vβ7 or Vβ8.3 was fitted to the TCRβ position in 1kj2 and was drawn by Mol Feat software (FiatLux, Tokyo, Japan).

Quantum chemical calculation
To study the effect of electric charge on the TCR recognition response, we calculated both the electronic state of P18-I10 and any changes therein based on a single amino acid substitution (325V with 325T) within the P18-I10 using the PM5 molecular orbital method (MOPAC2002) (35). In this calculation, hydrogen atoms were first added to the PDB 3D structures, and then the net charges of all atoms in them were obtained, with the PDB 3D structure being maintained.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
Effect of a single amino acid substitution at position 325 within P18-I10 on LINE-IIIB recognition
First, to see the effect of a single amino acid substitution at position 325 where epitope specificity for LINE-IIIB recognition is determined, a series of P18-I10-derived peptides shown in Table 1 have been synthesized. Similar to our previous findings (10), we confirmed that LINE-IIIB cross-reacted with an aliphatic residue, such as I, L, or A at position 325, in addition to reacting with the original residue V. Moreover, LINE-IIIB cross-reacted with positively charged residues such as R and H but not with the negatively charged E or the uncharged T, S, or P at position 325 (Fig. 1 A). Here, we should focus on the case of T, because the 3D structure of the side chain of T is very similar to that of V. When only one carbon atom of the side chain of V is exchanged for one oxygen atom (Fig. 1 B; red), the amino acid becomes T, except for the hydrogen atoms. According to the quantum chemical calculation using Hamiltonian PM5, the net charge of the carbon atom in V was about –0.3e, whereas that of the oxygen atom in T is about –0.4e; e represents the elementary charge. This may be the reason why the recognition of V at position 325 by LINE-IIIB was dramatically changed by substitution with T. These results indicate that the charges of the amino acids within the epitope may affect the interaction between TCRs and peptide/MHC complexes.

View larger version (17K): [in this window] [in a new window]   FIGURE 1  Effect of a single amino acid substitution at position 325 within P18-I10 on LINE-IIIB recognition. (A) Ten thousand of P18-I10-specific CTL line (LINE-IIIB cells) were added to 5000 51Cr-labeled BALB/c.3T3 fibroblast target cells in the presence of 3 µM of the substituted peptides at position 325 within P18IIIB, as shown in Table 1. Standard errors of the means of triplicate cultures were <5% of the mean in each case. Results are representative of three independent experiments. (B) Ball-and-stick model of peptide P18-I10. The amino acid side chain containing the valine (V) at position 325 within P18-I10 and threonine (T) are shown in the circle. Carbon, oxygen, nitrogen, and hydrogen are shown in gray, red, blue, and white, respectively.

View larger version (11K): [in this window] [in a new window]   FIGURE 2  Effect of a single amino acid substitution with D-amino acid within an epitope peptide, P18-I10, on LINE-IIIB recognition. Ten thousand LINE-IIIB cells were added to 5000 51Cr-labeled BALB/c.3T3 fibroblast target cells in the presence of 3 µM of the substituted peptides with a single amino acid substitution by D-amino acid at each corresponding site (represented by lower-case letters in Table 1). Standard errors of the means of triplicate cultures were <5% of the mean in each case. Each experiment was performed at least three times.

View larger version (24K): [in this window] [in a new window]   FIGURE 3  Comparison of specificity of the newly established CTL lines and clones against two peptides, peptide P18-I10 and I10(325v). We examined the cytolytic activity of the following distinct CTL lines and clones using a 5-h 51Cr-release assay. To test the peptide specificity, effector cells and 51Cr-labeled BALB/c.3T3 fibroblast targets (E/T ratio was 10:1) were incubated in the presence of various concentrations of either P18-I10 or I10(325v). (A) LINE-IIIB(325D) specific for I10(325v) was used for effector cells. (B) LINE-IIIB specific for P18-I10 was used for effector cells. (C) The CTL clones, IIE11(D) and IIA4(D), derived from LINE-IIIB(325D) were used as effector cells. (D) The CTL clones, IIH7(L) and IB9(L), from LINE-IIIB were used as effector cells. Standard errors of the means of triplicate cultures were <5% of the mean in each case. Each experiment was performed at least three times.

Specificity and characterization of the established CTL clones
We further examined the fine specificities of the CTL clones using a series of substituted peptides, each with a single amino acid substitution of either the L- or D-type having an aliphatic or aromatic structure at position 325 in P18-I10 (Table 1).

Among the D-specific clones, IIE11(D) did not cross-react with I10(325I) at all and was strictly specific for the D-type substitution such as I10(325i) and I10(325a) as well as for the original I10(325v) (Fig. 4 A), whereas IIA4(D) showed a weak cross-reactivity for both types of isoleucine, I10(325I) and I10(325i), as well as for the original I10(325v) (Fig. 4 B). In addition, both of the L-type-specific CTL clones, IIH7(L) and IB9(L), cross-reacted with I10(325I) as well as with the original P18-I10 (Fig. 4, C and D). These findings indicate that the CTL clones can specifically distinguish the optical isomers at position 325.

View larger version (23K): [in this window] [in a new window]   FIGURE 4  Effects of a single amino acid substitution at position 325 within P18-I10 on target sensitization for CTL recognition. We measured the cytolytic activity of the CTL lines and clones using a 5-h 51Cr-release assay. For the analysis of fine peptide specificity, effector cells and 51Cr-labeled BALB/c.3T3 fibroblast targets were mixed with 10 µM of each substituted peptide. The following four clones were used as effectors (E/T ratio was 20:1): (A) clone IIE11(D), (B) clone IIA4(D), (C) clone IIH7(L), and (D) clone IB9(L). Standard errors of the means of triplicate cultures were <5% of the mean in each case. Each experiment was performed at least three times.

TCR-sequences of the established clones
Taken together, the two groups of CTL clones with high specificity to the substituted P18-I10-derived peptides, having either the L-type of valine (V) or D-type of valine (v) at position 325, expressed both CD8 and β TCRs, and were restricted by the same class I MHC molecule, D^d. Therefore, taking advantage of the unique combinations of CTL clones, we attempted to perform a precise analysis of the interaction between the TCRs of those clones and the amino acid at position 325 within epitope to determine the specificity by comparing their TCR sequences.

To investigate the actual amino acid sequences of the TCRs in the CTL clones, we extracted their mRNA and analyzed the nucleotide sequences of the - and β-chain transcripts after a cDNA synthesis and PCR amplifications. The V usage of the I10(325v)-specific clones, IIE11(D) and IIA4(D), were both V16 (36), whereas that of the P18-I10-specific clones, IIH7(L) and IB9(L), were V3 (37) and V2 (BLASTN Accession U88296), respectively (Fig. 5). It should be noted that all of the four distinct clones used the same uncommon J gene segment, 18BBM142, determined from a murine alloreactive T-cell hybridoma specific for I-A^bm12 (38). In contrast, the Vβ usage of the 325(v)-specific clones, IIE11(D) and IIA4(D), was Vβ8.3 (39) with the Jβ 2.2 segment (40) bearing the same CDR3, "DWGGS," whereas the Vβ usage of the 325(L)-specific clones, IIH7 and IB9, was Vβ7 with the Jβ1.1 segment (41) encoding a distinct sequence, "LGYT" and "LGVT," respectively, in the CDR3 flap (Fig. 5). These results strongly indicate that the TCRβ chains of those clones may be responsible for the specificity of the epitope.

View larger version (20K): [in this window] [in a new window]   FIGURE 5  Nucleotide and amino acid sequences that form the V(D)J region of the TCR and -β chain from the established CTL clones. We analyzed the amino acid sequences of TCRs in two groups of CTL clones with high specificity to substituted P18-I10-derived peptides, having either the D-type of valine (v) (IIE11(D) and IIA4(D)) or the L-type of valine (V) at position 325 (IIH7(L) and IB9(L)) restricted by the same class I MHC molecules, Dd.

View larger version (37K): [in this window] [in a new window]   FIGURE 6  (A) Compatible templates with excellent Standardized Scores (SD value) for both Vβ7 and Vβ8.3 selected from the LIBRA software. Bold abbreviations are defined as follows: Rk, rank position; StrC, structural code of PDB (the last character is a subunit name); Lsr, length of the structural template; Lal, length of the aligned region; Rsc, raw score of the structural template; SD, standardized score; Rs/N, raw score (Rsc) normalized by the alignment length (Lal); ID%, sequence identity. (B) Sequence alignment of TCR Vβ7, Vβ8.3, and each of the five template proteins (1a6wH, 1dn0D, 1tcrA, 1ao7D, 1kb5A for Vβ7 and 1hxmB, 1etzB, 1fo0A, 1h5bA, 1d1fL for Vβ8.3) obtained from the LIBRA software. In this figure, the amino acid sequences of CDR1 and CDR3 from TCR Vβ7 or Vβ8.3 are drawn in red and green, respectively. In 10 template proteins, the corresponding portion of the CDR1 and CDR3 regions are also drawn in red and green, respectively. Conserved cysteine residues upstream of CDR1 and CDR3 are drawn in magenta. Two or three amino acids between the conserved cysteine residues and CDR1 are drawn in blue (see Discussion).

View larger version (21K): [in this window] [in a new window]   FIGURE 7  Comparison between the predicted 3D structure for 1dn0D by the MODELLER software and that of the same protein registered in PDB. The 3D structure of the protein (PDB cord 1dn0D) was predicted by the MODELLER, a molecular modeling software, using a protein (e.g., PDB cord 1a6wH) as a template. The predicted 3D structure for 1dn0D is drawn in blue, and the PDB 3D structure is drawn in red. Their backbone representations in the molecular modeling are drawn from rotated two distinct views (A and B).

View larger version (40K): [in this window] [in a new window]   FIGURE 8  3D structures representing the interaction between TCR Vβ7 (Vβ8.3) and P18-I10/H-2Dd Complex or the interaction between TCR Vβ7 (Vβ8.3) and I10(325v)/ H-2Dd complex. (A–H) The ternary complex of TCR Vβ7 or Vβ8.3 onto either P18-I10 or I10(325v) bounded to the H-2Dd class I MHC molecules were illustrated by computer-based molecular modeling (see Materials and Methods). Horizontal gray backbone represents the H-2Dd class I MHC molecule, and the yellow ball format indicates the Dd-bounded epitope peptide P18-I10. Vertical overlapping white backbones indicate either TCR Vβ7 or Vβ8.3. In these figures, Vβ7 and Vβ8.3 made from five distinct template proteins were overlaid in one figure. The CDR1 and CDR3 loops from both TCR Vβ chains are drawn in cyan and red, respectively. The L-type of valine (V) and the D-type of valine (v) at position 325, where epitope specificity appears to be determined, are shown in bright green and bright red, respectively. The 3D-structures of the (A) Vβ7 /H-2Dd/P18-I10, (B) enlarged figure of A, (C) Vβ7 /H-2Dd/I10(325v), (D) enlarged figure of C, (E) Vβ8.3 /H-2Dd/I10(325v), (F) enlarged figure of E, (G) Vβ8.3/H-2Dd/P18-I10, and (H) enlarged figure of G are shown in eight independent panels.

To substantiate the interaction between valine at position 325 and TCR-CDR1, the distance between terminal atoms in the side chain of amino acids within the epitope peptide and atoms in the main chain of the nearest portion within the TCR was calculated. As shown in Table 2, first to confirm the reliability of our molecular modeling, the distance between the two terminal atoms (OD1 and OD2) in the side chain of the seventh amino acid, aspaltic acid (D), within epitope peptide pKB1(aa: KVITFIDL) and the nearest portion in the TCR from OD1 or OD2 was determined using an already reported TCR/peptide/MHC complex, 1kj2 (PDBcode) obtained from x-ray crystallographic analysis (5). Based on the above observation, the distance between the two terminal atoms (CG1 and CG2) in the side chain of 325V or 325v and their surrounding atoms in the main chain of the obtained TCRs, Vβ7 and Vβ8.3, calculated (Tables 3 and 4) and compared with that of 1kj2. The results indicate that the distance between 325V and CDR1 in Vβ7 is similar to the distance in the case of 1kj2, whereas the distance between 325v and CDR1 in Vβ7 is too small, which may induce conformational interference between them. In contrast, the distance between 325v and CDR1 in Vβ8.3 is similar to 1kj2, but the distance between the 325V and the CDR1 in Vβ8.3 appeared to be too far from that of 1kj2. These results again agreed with our experimental results that Vβ8.3 recognized the epitope I10(325v) having D-valine at position 325 via CDR1 but did not recognize the P18-I10 containing L-valine at that position.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
In our previous study, we found that the amino acids at positions 322R and 324F were critical for D^d binding, and those at position 325V were essential for interacting with TCRs (8,17). Also, the C-terminus 327I appears to be a key amino acid for D^d binding to form the D^d-binding motif (8,18). In this study, we demonstrated that the substitution of a positively charged 322R by 322r did not result in measurable changes in target sensitization, although in our previous study the substitution with uncharged alanine (A) completely eliminated the capacity to sensitize the targets (17), indicating that a positive electric charge must be critical for D^d binding at position 322, and that a reduction of charge in the amino acid might diminish the epitope potency for T-cell activation. As shown in Fig. 1 B, the side chains of V and T have the same 3D structure, except for the hydrogen atoms, if one terminal carbon atom of the side chain of V is exchanged for one oxygen atom. According to the quantum chemical calculation, the net charge of the carbon or the oxygen is –0.3e or –0.4e, respectively. In addition, three hydrogen atoms, which have positive charges, are added to the carbon, whereas one hydrogen atom is added to the oxygen. Thus, T generates a greater negative net charge than V. The influence of this charge difference was evident when LINE-IIIB recognized the peptide. It is likely that T had a greater repulsive force against the oxygen atoms in the CDR1 loop compared with V. This repulsive force may account for the significant reduction in the recognition by LINE-IIIB after substitution with T. These findings indicate that the charge of an amino acid at a specific position within an epitope can affect its binding capacity to MHC molecules and/or the interaction with a TCR.

Although the method we applied here for modeling the TCR/peptide/class I MHC complex was not directly based on crystallographic analysis, our computer-based molecular modeling was still accurate. Indeed, as demonstrated in Fig. 7, the predicted 3D structure for 1dn0D by MODELLER was similar to that determined experimentally, and the core regions appeared to be nearly identical. This reflects the fact that the 3D structures of the core regions of the TCR consist of a four-stranded antiparallel β-sheet and a three-stranded anti-parallel β-sheet linked by a disulfide bond. Although the predicted 3D structure of CDR3 seems slightly different from the experimental 3D structure of CDR3 because of the large loop, the predicted 3D structure of CDR1 appears almost the same as the experimental 3D structure. These results suggest that the TCR domain predicted by the present molecular modeling methods can become useful and reliable tools for the structural analysis of TCRs.

In these sorts of structural analyses, it should be acknowledged that a substitution of D-valine (v) for L-valine (V) would result in a conformational interference between the class I MHC molecule and the D-valine (v), which itself could decrease the recognition response by LINE-IIIB. However, as shown in Fig. 9, no conformational interference occurred in this study. Thus, I10(325v) would also be associated with the class I MHC molecule so that TCRs may recognize the epitope. Therefore, our findings suggest that the change in the recognition response by LINE-IIIB after the D-valine (v) substitution reflects the interaction with the TCR.

View larger version (27K): [in this window] [in a new window]   FIGURE 9  (A and B) 3D structures of P18-I10/H-2Dd complex and I10(325v)/H-2Dd complex. Gray ribbon and ball format represent the H-2Dd class I MHC molecule, and the yellow ball format indicates the Dd-bounded epitope peptide, P18-I10 or I10(325v). In the H-2Dd class I MHC molecule, only the positions for interaction with 325V are drawn in the ball format, for clarity. The L-type of valine (V) and the D-type of valine (v) at position 325 are shown in green and red, respectively. Radius of ball format indicates van der Waals radius.

Although the epitope specificity created by TCRs has been reported to be dependent mainly on the amino acid sequences in the CDR3 regions for both the TCR and β chains (43,44), crystallographic analyses on various TCR and peptide/MHC interactions have suggested that both the CDR1 and CDR3 in the TCR and TCRβ chains might contact with the antigenic peptide/MHC complex (1,2,5,6), particularly with the carboxyl-terminal portion of the peptides (2). Moreover, recent reports have shown that the CDR1 in Vβ10 participated in class I MHC molecule-mediated T-cell recognition (45,46). In these reports, significant alteration in the capacity to bind class I MHC molecules and in the ability to respond to the peptide/MHC complex was demonstrated when a single amino acid substitution was introduced into the CDR1 of TCR Vβ10 by site-directed mutagenesis. Indeed, we have shown here that the CDR1 in the TCRβ chains appeared to interact directly with the key amino acid for determining epitope specificity. Thus, if the most critical amino acid for determining the epitope specificity is located near the C-terminal portion of a peptide such as P18-I10, not only the CDR3 but also the CDR1 in the TCRβ chain may be involved in determining antigen specificity.

It is also important to consider the role of the primary structures in TCR recognition. As demonstrated in Fig. 6 B, there are two cysteine (C) residues positioned upstream of CDR1 (magenta) and CDR3 (magenta) of both TCR Vβ7 and Vβ8.3. These residues bind each other via a disulfide bond and must be a basic structure of TCRs because they are conserved in most of the murine TCR sequences (47) . The two or three amino acids (blue) between the conserved cysteine (C) residues and the CDR1 would be key amino acids in forming the 3D structure of CDR1, although they do not interact directly with the epitope peptide. If only these amino acids are exchanged for other amino acids, the 3D structure of CDR1 affecting the recognition of an amino acid at position 325 would change. Indeed, these amino acids are highly variable for both TCR Vβ7 and Vβ8.3 (Fig. 6 B) and for various other murine TCR sequences (47), whereas the regions just after CDR1 are mostly conserved. These amino acids are likely to participate in the peptide recognition by creating small changes in the 3D structure of CDR1. In this regard, both CDR1 and the two or three amino acids between the cysteine (C) and CDR1 may play a important role in recognizing position 325 within P18-I10 or I10(325v).

	ACKNOWLEDGEMENTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
We thank Dr. Megumi Takahashi for her assistance with the DNA sequence analysis of the TCRs.

This work was supported in part by grants from the Ministry of Education, Science, Sport, and Culture, from the Ministry of Health and Labor and Welfare, Japan; from the Japanese Health Sciences Foundation; and from and by the Promotion and Mutual Aid Corporation for Private Schools of Japan.

Submitted on August 16, 2006; accepted for publication December 13, 2006.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION ACKNOWLEDGEMENTS REFERENCES

 
1. Garboczi, D. N., P. Ghosh, U. Utz, Q. R. Fan, W. E. Biddison, and D. C. Wiley. 1996. Structure of the complex between human T-cell receptor, viral peptide and HLA-A2. Nature. 384:134–141.[CrossRef][Medline]

2. Garcia, K. C., M. Degano, R. L. Stanfield, A. Brunmark, M. R. Jackson, P. A. Peterson, L. Teyton, and I. A. Wilson. 1996. An alphabeta T cell receptor structure at 2.5 A and its orientation in the TCR-MHC complex. Science. 274:209–219.[Abstract/Free Full Text]

3. Garboczi, D. N., and W. E. Biddison. 1999. Shapes of MHC restriction. Immunity. 10:1–7.[Free Full Text]

4. Hennecke, J., and D. C. Wiley. 2001. T cell receptor-MHC interactions up close. Cell. 104:1–4.[CrossRef][Medline]

5. Reiser, J. B., C. Gregoire, C. Darnault, T. Mosser, A. Guimezanes, A. M. Schmitt-Verhulst, J. C. Fontecilla-Camps, G. Mazza, B. Malissen, and D. Housset. 2002. A T cell receptor CDR3beta loop undergoes conformational changes of unprecedented magnitude upon binding to a peptide/MHC class I complex. Immunity. 16:345–354.[CrossRef][Medline]

6. Garcia, K. C., M. Degano, L. R. Pease, M. Huang, P. A. Peterson, L. Teyton, and I. A. Wilson. 1998. Structural basis of plasticity in T cell receptor recognition of a self peptide-MHC antigen. Science. 279:1166–1172.[Abstract/Free Full Text]

7. Takahashi, H., J. Cohen, A. Hosmalin, K. B. Cease, R. Houghten, J. L. Cornette, C. DeLisi, B. Moss, R. N. Germain, and J. A. Berzofsky. 1988. An immunodominant epitope of the human immunodeficiency virus envelope glycoprotein gp160 recognized by class I major histocompatibility complex molecule-restricted murine cytotoxic T lymphocytes. Proc. Natl. Acad. Sci. USA. 85:3105–3109.[Abstract/Free Full Text]

8. Takeshita, T., H. Takahashi, S. Kozlowski, J. D. Ahlers, C. D. Pendleton, R. L. Moore, Y. Nakagawa, K. Yokomuro, B. S. Fox, and D. H. Margulies. 1995. Molecular analysis of the same HIV peptide functionally binding to both a class I and a class II MHC molecule. J. Immunol. 154:1973–1986.[Abstract]

9. Takahashi, H., S. Merli, S. D. Putney, R. Houghten, B. Moss, R. N. Germain, and J. A. Berzofsky. 1989. A single amino acid interchange yields reciprocal CTL specificities for HIV-1 gp160. Science. 246:118–121.[Abstract/Free Full Text]

10. Takahashi, H., Y. Nakagawa, C. D. Pendleton, R. A. Houghten, K. Yokomuro, R. N. Germain, and J. A. Berzofsky. 1992. Induction of broadly cross-reactive cytotoxic T cells recognizing an HIV-1 envelope determinant. Science. 255:333–336.[Abstract/Free Full Text]

11. Shirai, M., C. D. Pendleton, and J. A. Berzofsky. 1992. Broad recognition of cytotoxic T cell epitopes from the HIV-1 envelope protein with multiple class I histocompatibility molecules. J. Immunol. 148:1657–1667.[Abstract]

12. Palker, T. J., M. E. Clark, A. J. Langlois, T. J. Matthews, K. J. Weinhold, R. R. Randall, D. P. Bolognesi, and B. F. Haynes. 1988. Type-specific neutralization of the human immunodeficiency virus with antibodies to env-encoded synthetic peptides. Proc. Natl. Acad. Sci. USA. 85:1932–1936.[Abstract/Free Full Text]

13. Rusche, J. R., K. Javaherian, C. McDanal, J. Petro, D. L. Lynn, R. Grimaila, A. Langlois, R. C. Gallo, L. O. Arthur, P. J. Fischinger, D. P. Bolognesi, S. D. Putney, and T. J. Matthews 1988. Antibodies that inhibit fusion of human immunodeficiency virus-infected cells bind a 24-amino acid sequence of the viral envelope, gp120. Proc. Natl. Acad. Sci. USA. 85:3198–3202.[Abstract/Free Full Text]

14. Goudsmit, J., C. Debouck, R. H. Meloen, L. Smit, M. Bakker, D. M. Asher, A. V. Wolff, C. J. Gibbs, Jr., and D. C. Gajdusek. 1988. Human immunodeficiency virus type 1 neutralization epitope with conserved architecture elicits early type-specific antibodies in experimentally infected chimpanzees. Proc. Natl. Acad. Sci. USA. 85:4478–4482.[Abstract/Free Full Text]

15. Takahashi, H., R. N. Germain, B. Moss, and J. A. Berzofsky. 1990. An immunodominant class I-restricted cytotoxic T lymphocyte determinant of human immunodeficiency virus type 1 induces CD4 class II-restricted help for itself. J. Exp. Med. 171:571–576.[Abstract/Free Full Text]

16. Clerici, M., D. R. Lucey, R. A. Zajac, R. N. Boswell, H. M. Gebel, H. Takahashi, J. A. Berzofsky, and G. M. Shearer. 1991. Detection of cytotoxic T lymphocytes specific for synthetic peptides of gp160 in HIV-seropositive individuals. J. Immunol. 146:2214–2219.[Abstract]

17. Takahashi, H., R. Houghten, S. D. Putney, D. H. Margulies, B. Moss, R. N. Germain, and J. A. Berzofsky. 1989. Structural requirements for class I MHC molecule-mediated antigen presentation and cytotoxic T cell recognition of an immunodominant determinant of the human immunodeficiency virus envelope protein. J. Exp. Med. 170:2023–2035.[Abstract/Free Full Text]

18. Corr, M., L. F. Boyd, E. A. Padlan, and D. H. Margulies. 1993. H-2Dd exploits a four residue peptide binding motif. J. Exp. Med. 178:1877–1892.[Abstract/Free Full Text]

19. Jorgensen, J. L., U. Esser, B. Fazekas de St Groth, P. A. Reay, and M. M. Davis. 1992. Mapping T-cell receptor-peptide contacts by variant peptide immunization of single-chain transgenics. Nature. 355:224–230.[CrossRef][Medline]

20. Cease, K. B., H. Margalit, J. L. Cornette, S. D. Putney, W. G. Robey, C. Ouyang, H. Z. Streicher, P. J. Fischinger, R. C. Gallo, C. DeLisi, and J. A. Berzofsky. 1987. Helper T-cell antigenic site identification in the acquired immunodeficiency syndrome virus gp120 envelope protein and induction of immunity in mice to the native protein using a 16-residue synthetic peptide. Proc. Natl. Acad. Sci. USA. 84:4249–4253.[Abstract/Free Full Text]

21. Boehncke, W. H., T. Takeshita, C. D. Pendleton, R. A. Houghten, S. Sadegh-Nasseri, L. Racioppi, J. A. Berzofsky, and R. N. Germain. 1993. The importance of dominant negative effects of amino acid side chain substitution in peptide-MHC molecule interactions and T cell recognition. J. Immunol. 150:331–341.[Abstract]

22. Zhang, W., S. Honda, F. Wang, T. P. DiLorenzo, A. M. Kalergis, D. A. Ostrov, and S. G. Nathenson. 2001. Immunobiological analysis of TCR single-chain transgenic mice reveals new possibilities for interaction between CDR3alpha and an antigenic peptide bound to MHC class I. J. Immunol. 167:4396–4404.[Abstract/Free Full Text]

23. Takahashi, H., Y. Nakagawa, K. Yokomuro, and J. A. Berzofsky. 1993. Induction of CD8⁺ cytotoxic T lymphocytes by immunization with syngeneic irradiated HIV-1 envelope derived peptide-pulsed dendritic cells. Int. Immunol. 5:849–857.[Abstract/Free Full Text]

24. Takahashi, H., Y. Nakagawa, G. R. Leggatt, Y. Ishida, T. Saito, K. Yokomuro, and J. A. Berzofsky. 1996. Inactivation of human immunodeficiency virus (HIV)-1 envelope-specific CD8⁺ cytotoxic T lymphocytes by free antigenic peptide: a self-veto mechanism? J. Exp. Med. 183:879–889.[Abstract/Free Full Text]

25. Margulies, D. H., G. A. Evans, K. Ozato, R. D. Camerini-Otero, K. Tanaka, E. Appella, and J. G. Seidman. 1983. Expression of H-2Dd and H-2Ld mouse major histocompatibility antigen genes in L cells after DNA-mediated gene transfer. J. Immunol. 130:463–470.[Abstract]

26. Abastado, J.P., C. Jaulin, M.P. Schutze, P. Langlade_Demoyen, F. Plata, K. Ozato, and P. Kourilsky. 1987. Fine mapping of epitopes by intradomain K^d/D^d recombinants. J. Exp. Med. 166:327–340.[Abstract/Free Full Text]

27. Takahashi, M., E. Osono, Y. Nakagawa, J. Wang, J. A. Berzofsky, D. H. Margulies, and H. Takahashi. 2002. Rapid induction of apoptosis in CD8⁺ HIV-1 envelope-specific murine CTLs by short exposure to antigenic peptide. J. Immunol. 169:6588–6593.[Abstract/Free Full Text]

28. Bowie, J. U., N. D. Clarke, C. O. Pabo, and R. T. Sauer. 1990. Identification of protein folds: matching hydrophobicity patterns of sequence sets with solvent accessibility patterns of known structures. Proteins. 7:257–264.[CrossRef][Medline]

29. Ota, M., and K. Nishikawa. 1997. Assessment of pseudo-energy potentials by the best-five test: a new use of the three-dimensional profiles of proteins. Protein Eng. 10:339–351.[Abstract/Free Full Text]

30. Sali, A., and T. L. Blundell. 1993. Comparative protein modelling by satisfaction of spatial restraints. J. Mol. Biol. 234:779–815.[CrossRef][Medline]

31. Fiser, A., R. K. Do, and A. Sali. 2000. Modeling of loops in protein structures. Protein Sci. 9:1753–1773.[Abstract]

32. Baker, D., and A. Sali. 2001. Protein structure prediction and structural genomics. Science. 294:93–96.[Abstract/Free Full Text]

33. Marti-Renom, M. A., M. S. Madhusudhan, A. Fiser, B. Rost, and A. Sali. 2002. Reliability of assessment of protein structure prediction methods. Structure. 10:435–440.[Medline]

34. Achour, A., K. Persson, R. A. Harris, J. Sundback, C. L. Sentman, Y. Lindqvist, G. Schneider, and K. Karre. 1998. The crystal structure of H-2Dd MHC class I complexed with the HIV-1-derived peptide P18–I10 at 2.4 A resolution: implications for T cell and NK cell recognition. Immunity. 9:199–208.[CrossRef][Medline]

35. Stewart, J. J. P. 1996. Application of localized molecular orbitales to the solution of semiempirical self-consistent field equations. Int. J. Quant. Chem. 58:133–146.[CrossRef]

36. Sherman, D. H., P. S. Hochman, R. Dick, R. Tizard, K. L. Ramachandran, R. A. Flavell, and B. T. Huber. 1987. Molecular analysis of antigen recognition by insulin-specific T-cell hybridomas from B6 wild-type and bm12 mutant mice. Mol. Cell. Biol. 7:1865–1872.[Abstract/Free Full Text]

37. Tan, K. N., B. M. Datlof, J. A. Gilmore, A. C. Kronman, J. H. Lee, A. M. Maxam, and A. Rao. 1988. The T cell receptor V alpha 3 gene segment is associated with reactivity to p-azobenzenearsonate. Cell. 54:247–261.[CrossRef][Medline]

38. Bill, J., J. Yague, V. B. Appel, J. White, G. Horn, H. A. Erlich, and E. Palmer. 1989. Molecular genetic analysis of 178 I-Abm12-reactive T cells. J. Exp. Med. 169:115–133.[Abstract/Free Full Text]

39. Cerasoli, D. M., M. P. Riley, F. F. Shih, and A. J. Caton. 1995. Genetic basis for T cell recognition of a major histocompatibility complex class II-restricted neo-self peptide. J. Exp. Med. 182:1327–1336.[Abstract/Free Full Text]

40. Horwitz, M. S., Y. Yanagi, and M. B. Oldstone. 1994. T-cell receptors from virus-specific cytotoxic T lymphocytes recognizing a single immunodominant nine-amino-acid viral epitope show marked diversity. J. Virol. 68:352–357.[Abstract/Free Full Text]

41. Plaksin, D., K. Polakova, P. McPhie, and D. H. Margulies. 1997. A three-domain T cell receptor is biologically active and specifically stains cell surface MHC/peptide complexes. J. Immunol. 158:2218–2227.[Abstract]

42. Bowie, J. U., R. Luthy, and D. Eisenberg. 1991. A method to identify protein sequences that fold into a known three-dimensional structure. Science. 253:164–170.[Abstract/Free Full Text]

43. Kaye, J., and S. M. Hedrick. 1988. Analysis of specificity for antigen, Mls, and allogenic MHC by transfer of T-cell receptor alpha- and beta-chain genes. Nature. 336:580–583.[CrossRef][Medline]

44. Lai, M. Z., Y. J. Jang, L. K. Chen, and M. L. Gefter. 1990. Restricted V-(D)-J junctional regions in the T cell response to lambda-repressor. Identification of residues critical for antigen recognition. J. Immunol. 144:4851–4856.[Abstract]

45. Bellio, M., Y. C. Lone, O. de la Calle-Martin, B. Malissen, J. P. Abastado, and P. Kourilsky. 1994. The V beta complementarity determining region 1 of a major histocompatibility complex (MHC) class I-restricted T cell receptor is involved in the recognition of peptide/MHC I and superantigen/MHC II complex. J. Exp. Med. 179:1087–1097.[Abstract/Free Full Text]

46. Lone, Y. C., M. Bellio, A. Prochnicka-Chalufour, D. M. Ojcius, N. Boissel, T. H. Ottenhoff, R. D. Klausner, J. P. Abastado, and P. Kourilsky. 1994. Role of the CDR1 region of the TCR beta chain in the binding to purified MHC-peptide complex. Int. Immunol. 6:1561–1565.[Abstract/Free Full Text]

47. Clark, S. P., B. Arden, D. Kabelitz, and T. W. Mak. 1995. Comparison of human and mouse T-cell receptor variable gene segment subfamilies. Immunogenetics. 42:531–540.[Medline]

This Article Abstract Full Text (PDF) All Versions of this Article: biophysj.106.095208v1 92/7/2570    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager