| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1,2
* Department of Biological Chemistry, and
Department of Immunology, the Weizmann Institute of Science, Rehovot, Israel
3Correspondence: Department of Biological Chemistry, the Weizmann Institute of Science, Rehovot, 76100 Israel. E-mail: yechiel.shai{at}weizmann.ac.il
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-helical transmembrane domain (TMD) of the TCR chain, and the ß-sheet 5–13 region of the 16 N-terminal aa of FP (FP1–16). Deciphering the molecular mechanism of the immunosuppressive activity of FP provides a new potential target to overcome the immunosuppressant activity of HIV, and in addition a tool for down-regulating immune mediated inflammation.—Bloch, I., Quintana, F. J., Gerger, D., Cohen, T., Cohen, I. R., and Shai, Y. T-Cell inactivation and immunosuppressive activity induced by HIV gp41 via novel interacting motif.
Key Words: membrane proteins • protein-membrane interaction • peptide-membrane interaction • transmembrane domain • recognition within membranes
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2
3)
. Several regions in gp41 contribute to the fusion process; however, the 33 residues at the N terminus (the fusion peptide, FP1–33) are of outstanding importance. Within FP1–33, two different domains can be identified based on their biochemical characteristics and their role in membrane fusion. The N terminus of FP1–33 contains a 16 aa-long hydrophobic stretch (FP1–16), which inserts itself into the target membrane and is followed by a C-terminus hydrophilic region that lies parallel to the cell surface (4
5
6)
. Thus, during the process of membrane fusion, FP1–16 anchors gp41 to the target membrane, and conformational changes in the three-dimensional structure of gp41 bring the membranes together and eventually complete the fusion process.
In addition to its fusogenic activity, FP1–33 harbors immunosuppressive activities (7
, 8)
. We have recently reported that FP can interact with the T cell receptor (TCR) and interfere with T cell activation (8)
. This activity of FP might function to suppress HIV-specific immunity and facilitate the spread of HIV to uninfected cells (8)
. Moreover, this immunosuppressive activity of FP might be exploited as an immunotherapeutic tool. Indeed, we have shown that FP can reduce the clinical signs of the experimental autoimmune T cell-mediated disease adjuvant arthritis (AA) (8)
.
The contribution of the different regions of FP to membrane fusion is well characterized; however, their role in the immunosuppressive activity of FP is still unknown. In this work, we identified the TCR/FP1–33 interaction at the molecular level, providing a 3D model of the TCR/FP complex structure. A clear understanding of the molecular interactions mediating the immunosuppressive activity of FP will facilitate the evaluation of its contribution to HIV pathology and its exploitation as an immunotherapeutic tool.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Further details can be found in the supplementary materials.
Induction and assessment of adjuvant arthritis (AA)
AA was induced in groups of 6 rats by injecting 50 µl of Mt suspended in IFA (0.5 mg/ml) at the base of the tail, as previously reported (8)
. Further details can be found in the supplementary materials.
Fluorescent labeling of peptides
Resin-bound peptides were treated with 4-chloro-7-nitrobenz-2-oxa- 1,3-diazole fluoride [nitro benzoxadiazole (NBD)-F] or 5-carboxytetramethylrhodamine, succinimidyl ester (Rho-SUC), respectively. The NBD-F and Rho-SUC fluorescent probes were purchased from Molecular Probes (Junction City, OR, USA). The reaction with NBD-F took place in DMF, and the reaction with Rhodamine in DMF containing 2% diisopropylethylamine as described previously (10)
. The fluorescent probes were used in an excess of two equivalents, leading to the formation of resin-bound N-terminal NBD or Rhodamine peptides. After 1 h, the resins were washed thoroughly with DMF and then with methylene chloride. The resin was dried under nitrogen flow and then cleaved for 3 h with TFA 95%, H2O 2.5%, and triethylsilane 2.5%. The labeled peptides were cleaved as described above and purified on a RP-HPLC C4 column as described before (8)
. Unless stated otherwise, stock solutions of concentrated peptides were maintained in DMSO to avoid aggregation of the peptides prior to use. The final concentration of DMSO in each experiment (5% v/v) had no effect on the system.
Colocalization of peptides with membrane molecules
A2b cells that had been activated for 72 h with PPD and APC were fixed with 4% paraformaldehyde for 15 min on ice and washed with PBS. In control experiments resting A2b cells (8–10 d after a previous activation by APC incubated with antigen) were used. A2b cells activated with PPD were also investigated without fixation with 4% paraformaldehyde. The cells were then treated with 2% BSA in PBS at room temperature to block unspecific binding. After 30 min the cells were divided into aliquots containing 50,000 cells per 100 µl and either TCR-FITC or CD4-FITC was added (1:100) for 2 h. The Rho-labeled FP fragments were added during the last 5 min of incubation at a final concentration of 0.5–1µM. The cells were then washed with PBS and deposited onto a glass slide. The labeled cell samples were observed under a fluorescence confocal microscope. FITC excitation was set at 488 nm, with the laser set at 20% power to minimize bleaching of the fluorophore. Fluorescence was recorded from 505–525 nm. Rhodamine excitation was set at 543 nm, with the laser set at 5% power. Fluorescence data were collected from 560 nm and up.
Coimmunoprecipitation of fluorescence-labeled peptides with TCR molecules
Activated A2b T cells (2x106) were cultured for 1 h at 37°C in the presence of FP fragments (25 µg/ml), or medium alone, and lysed for 15 min on ice in 0.1 ml lysis buffer (11)
. Insoluble material was removed by centrifugation at 10,000 g for 10 min at 4°C. The lysate was then incubated overnight with Protein A-plus Agarose beads (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) bound to antibodies to the rat TCR, CD28, actin, HSP60 or MHC class I (data shown only for TCR and MHC class I). The antibodies reactive against the rat TCR (clone R73) or HSP60 (clone LK1) were purified from the respective hybridomas at our lab; antibodies to rat CD28, actin and the rat MHC class I were purchased from Serotec (Oxford, UK). After an overnight incubation at 4°C, the beads were washed with lysis buffer and boiled for 10 min, and the protein supernatant run in a 16% SDS-PAGE. The presence of coimmunoprecipitated peptide was detected by the Typhoon 9400 variable mode imager.
Fluorescence resonance energy transfer (FRET) measurements
NBD-labeled peptides served as donors in the FRET measurements to the acceptor Rho-labeled peptides. Prior to the measurements lipid mixture of PC:SM:Cho (1:1:1 w/w/w ratio) was prepared in CHCL3:MeOH (2:1 v/v ratio), this lipid mixture was then dissolved in PBS buffer filtered by miniextruder reaching filter pore size of 0.1 µm to obtain large unilamellar vesicles (LUVs). The NBD-Labeled crude protein (CP) was introduced into the LUV and fluorescence spectra were read (Aminco Bowman series 2 luminescence spectrometer) in the presence or absence of Rho-labeled FP fragments. In FRET competition experiments, after the addition of both donor and acceptor peptides to the system, an unlabeled competing peptide was added and the fluorescence intensity was measured before and after this addition. In all fluorescence experiments involving FP fragments, we used peptide/lipid ratio of 
8 x 10–5 to ensure that FP concentration is kept below the fusogenic concentration according to the former reported value of 5 x 10–3 (12)
. In this peptide/lipid ratio, the peptide fragments reside inside the membrane. Energy transfer efficiency (E) was determined experimentally from the ratio between the fluorescence intensity of the donor molecules in the presence (Ida) and in the absence (Id) of acceptor molecules at the wavelength of the maximal donor’s emission. The percentage of energy-transfer efficiency E (%) is given by:
 |
Statistical significance
The InStat 2.01 program (Graph Pad Software, San Diego, CA, USA) was used for statistical analysis. The Student’s t test and the Mann-Whitney U test were conducted to assay significant differences between the different experimental groups. Further details on the experimental methods including the entire simulation procedure are given in the Supplemental Data.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Since FP1–16 is known to insert itself into the T cell membrane to trigger membrane fusion, we investigated whether FP1–16 could also interact with the TCR and interfere with T cell activation. We, therefore, synthesized a set of peptides spanning the whole hydrophobic N terminus (FP1–16) or parts of it (FP1–8 and FP9–16) and studied their ability to interfere with antigen-triggered T cell proliferation (Table 1
). Primed LNC were taken from Mt immunized rats 26 d after immunization, when they show a strong proliferative response to the immunodominant mycobacterial antigens PPD and Mt176–90 (9
, 13)
. FP1–16 and FP1–33 displayed similar inhibitory effects when tested at a concentration of 25 µg/ml (Fig. 1
). However, at lower concentrations, FP1–16 was less active than was FP1–33 (Fig. 1)
. No inhibition of T cell proliferation was observed when the T cells where incubated with FP1–8 or FP9–16. Thus, the inhibitory effects of FP1–33 on T cell activation in vitro can be recapitulated to some degree by its N-terminal hydrophobic region, contained in FP1–16.
|
|
Inhibition of T cell activation in vivo
Adjuvant arthritis (AA) is a T cell-mediated experimental autoimmune disease inducible in Lewis rats by immunization with Mt in oil. We have previously shown that FP1–33 can interfere with T cell activation in vivo and ameliorate the course of AA (8)
. To analyze the effect that FP1–16 might have on T cell activation in vivo, we studied its effects on the progression of AA. FP1–16 administration at the time of AA induction led to a significant decrease in the signs of AA, both in terms of clinical AA score (Fig. 2
A) and leg swelling (Fig. 2B
). Note that the in vivo immunomodulatory activity of FP1–16 was equal to that produced by a similar amount of FP1–33. No amelioration was seen when we used a PBS control group (Fig. 2A and B
) or a control peptide (data not shown and (8)
.
|
The activity of the T cells that mediate AA can be followed in vivo by studying the DTH response to Mt (14)
. The administration of FP1–16 led to a decrease of 50% in the Mt-specific DTH response, similar to that observed in FP1–33-treated AA rats (Fig. 2C
). Treatment with the shorter peptides or PBS did not affect the DTH response.
The T cells that mediate AA are characterized by the secretion of IFN
on activation with PPD or mt176–90 ex vivo. Accordingly, treatments that ameliorate AA, such as administration of FP1–33, are usually associated with a decreased INF- secretion by LNC on stimulation with mycobacterial antigens. The administration of the FP1–16 peptide led to a significant (P<0.01) decrease in the secretion of IFN by the LNC of AA rats on stimulation with Mt antigens, comparable with that seen in FP1–33-treated rats (Fig. 2D
). All in all, these results indicate that FP1–16 inhibits T cell activation both in vitro and in vivo.
Colocalization of FP1–16 with the TCR in the membrane of activated T cells
To better understand the molecular interactions mediating the inhibition of T cell activation by FP1–16, we studied the distribution of rhodamine-conjugated FP1–16, FP1–8 and FP9–16 in the membrane of activated rat A2b T cells. Figure 3
shows that FP1–16 colocalized with the TCR, while FP1–8 and FP9–16 showed largely a cytoplasmatic distribution. Thus, FP1–16 reproduced the pattern of colocalization with the TCR observed for FP1–33 (8)
.
|
FP1–16 physically interacts with the TCR
We have shown that FP1–33 can be immunoprecipitated with anti-TCR antibodies, suggesting the formation of a TCR/FP1–33 complex in the membrane of activated T cells. To map the region in FP1–33 that mediates this interaction, we tested the ability of Rho-labeled FP1–16, FP1–8 and FP9–16 to coimmunoprecipitate with TCR molecules, or with MHC-I as a control. Rho-labeled FP1–33 and FP1–16 could be coprecipitated with TCR molecules, although the interaction seemed to be weaker for FP1–16; no coimmunoprecipitation was seen when we used the control peptide Rho-F9L or a control antibody (Ab) to MHC-I (Fig. 4
A). We compared the affinity of the TCR/FP1–33 and the TCR/FP1–16 interactions by performing a series of coprecipitation experiments using different concentrations of Rho-labeled FP1–16 or FP1–33. Figure 4B
shows that, although both FP1–16 and FP1–33 can be coprecipitated with the TCR receptor, the interaction mediated by FP1–33 seems to be of higher affinity: lower concentrations of FP1–33 were needed to detect a significant coprecipitation with TCR molecules (Fig. 4B
). Thus, the 16 N terminus aa of FP1–33, FP1–16, can account for much of the TCR/FP1–33 interaction.
|
Interaction between FP1–16 and TCR TMD within the TCR complex
The coimmunoprecipitation results showed in Fig. 4
suggested a physical interaction between FP1–16 and the TCR, but did not identify the target sequence recognized within the TCR complex. However, the hydrophobic nature of FP1–16 suggested that it interacts with components of the TCR located within the membrane, such as the transmembrane domain (TMD) of the TCR. To study the potential interaction of the TCR TMD with FP1–16, we competed the coimmunoprecipitation of FP1–16 and TCR molecules with a peptide representing the TMD of the TCR chain (TCR) (termed CP) or a mutated CP peptide (2G CP) (Table 1)
. 2G CP is a mutant peptide derived from CP, where the two charged aa known to mediate the interaction with CD3 subunits (15
, 16)
have been substituted by glycines. Competition with unlabeled CP (in a 4x molar excess) led to a significant decrease in the amount of Rho-labeled FP1–33 coimmunoprecipitated (Fig. 5
A). However, no decrease was observed when the coimmunoprecipitation reaction included an excess of the control peptide 2G CP. These results suggest that FP1–16 targets the TCR TMD in an interaction that involves the two positive charges of the TCR TMD.
|
We also studied the interaction between FP1–33 and the TCR TMD by following fluorescence resonance energy transfer (FRET) between NBD-labeled CP and Rho-labeled FP1–33 synthetic peptides in lipid vesicles composed of PC:SM:Cho (1:1:1 w/w/w ratio). When a 4-fold molar excess of unlabeled CP was added to lipid membranes containing NBD-labeled CP and Rho-labeled FP1–33, a significant decrease (P<0.01) in the FRET signal was detected (Fig. 5B
). No change was detected on the addition of 2G CP (Fig. 5B
). These results suggest a direct interaction between the TCR TMD and FP within the T cell membrane.
Molecular dynamics simulation: equilibration properties
To confirm and visualize our experimental results, we conducted a 30 ns molecular dynamics simulation with an explicit atom description. We used the sequence of the FP1–16 with the sequence of the TCR peptide extended by the three additional C-terminal aa AGF. A model of the complex between FP1–16 and TCR CP was created and positioned in a fully hydrated DMPC bilayer (Fig. 6
A). We used DMPC bilayers in the MD simulation because of the following criteria: i) DMPC bilayers have been intensively studied both experimentally and theoretically, which allows us to validate the correctness of the simulation parameters by comparing measurable entities (e.g., electronic density, lipid order parameter SCD, bilayer thickness, etc.) of the membrane to those of previous studies; ii) the use of DMPC bilayers in the simulation allows us to maintain the most biologically relevant bilayer phase (L phase) in the biologically relevant temperature (310°K); iii) in the simulation conducted in this study we used an atomistic detailed membrane model rather than an implicit model of membrane. The use of DMPC lipids preserves the global parameters required for describing a model membrane, and furthermore allows careful examination of the specific lipid-peptide interactions; iv) the DMPC membrane model does capture several important natural membrane properties as it is zwitterionic and it creates a hydrophobic layer. The evolution of the total energy, density, and box volume of the system during the 30 ns simulation was primarily used for determining whether the system had reached equilibrium (simulation profiles are presented in Supplemental Fig. 1A-C
). A decrease in the simulation box volume was observed during the first 300 ps restrained run and in the first 5 ns of the full unrestrained run, which enabled the system to reach a pressure of 1 bar. The simulation box volume decreased from 254.06 nm3 to the final equilibrium volume of 227.68 nm3. Based on these parameters, equilibrium was achieved after 12 ns. Thus, we considered the last 15 ns of the simulation to be in equilibrium. Statistical and structural analysis was made on the basis of data taken from the equilibrated phase only.
|
A root mean square deviation (RMSD) profile showed that the peptide structure reached stability after 4.5 ns of the unrestrained run. Supplemental Fig. 1D shows the averaged RMSD profile of the complex and the separate RMSD for both TCR CP and FP1–16. These results indicate that the CP is much more stable and in fact keeps its helical structure throughout the simulation while the FP looses its initial conformation rapidly and adopts a ß-strand. This latter conformation remained stable for the rest of the simulation.
Molecular dynamics simulation: general bilayer properties
To estimate the validity of a simulation, one should compare the obtained system parameters with that of the relevant experimental system. Such a comparison can be made with experiments done on fluid, fully hydrated lipid bilayers, which are the most biologically relevant (L phase) to such systems. The methods used for detection of bilayer parameters vary from volumetric measurements to optical methods such as diffuse X-ray scattering, liquid crystallographic X-ray and Neutron diffraction measurements (17
, 18)
. The density profile of the simulation box along the bilayer surface normal (the z-axis of the simulation box) shows typical bilayer properties with no water penetration into the lipid environment while the peptide complex remains inside the bilayer (Supplemental Fig. 2A). The average area per lipid headgroup is 62Å (2)
; this is in agreement with the determined experimental value (19
20
21)
. Bilayer thickness (Dpp), which is the average distance between lipid headgroups at each monolayer, is 35Å (obtained from the electron density profiles of the bilayer) and is close to the experimental value of 36Å (17)
for a pure hydrated DMPC bilayer. The microscopic structure of lipid molecules and the ordering of acyl chains are characterized through the deuterium order parameter SCD. The deuterium order parameter measures the orientational order of the lipid tails. It can be computed for each carbon atom in the lipid acyl chains by averaging over all equivalent atoms in all DMPC. Although slightly lower, as expected by the perturbation of the peptides used, SCD values, which are shown in Supplemental Fig. 2B, are comparable with recent published results from a DMPC bilayer simulation (22)
.
The ß-sheet folded FP1–16 interacts with the -helical TMD of the TCR
We found that the potential interface between the two peptides in the complex is stable throughout the production run. The structural data show that the largest contribution to complex stability was made by residues GALFLGFLG, in the 5–13 region of FP1–16, which interacts with the TCR TMD (Fig. 6A and B
). In the TCR TMD, the Arg and Lys residues were found in the interpeptide interface contributing the most to the stability of the complex (Fig. 6B
). Interestingly, these observations are in agreement with our experimental results, showing the need of the Arg and Lys aa to compete with the coimmunoprecipitation of FP1–33 and the TCR (Fig. 5)
. As it seems from the equilibrium phase of the complex, the two charged residues interact with the polar atoms of the FP backbone to stabilize the interaction, mainly through hydrogen bonds. However, as both Arg and Lys have long and bulky side chains, they are in close contact to the FP residues. Specifically, Arg of the TCR TMD interacts with the two N terminus Leu of the FP, and Lys of the TCR TMD interacts with the C-terminus Leu and Ala of the FP. In addition, during the equilibrium phase of the simulation (last 15 ns) the FP N terminus Phe shifts from one rotameric state to another. Only in the second rotameric state, the TCR TMD Phe interacts with the FP Phe. This shift occurs at 25 ns and remains stable until the end of the simulation. Therefore, we can conclude that it does not play a crucial role in the initial formation of the complex. However, this interaction contributes to the complex stability by expanding the contact surface between the two peptides. Energetically, there is no significant change in the overall energy profile of the complex interaction energy, probably due to compensation by other elements within the complex.
Remarkably, analysis of the FP1–16 C dihedral angles demonstrated a shift from the initial -helical structure into a ß-strand, as shown in the Ramachandran plot in Fig. 6C
. The TCR chain kept its helical structure throughout the entire simulation (Fig. 6C
). Thus, the molecular dynamics simulation suggests a direct interaction between residues 6 to 12 of the ß-sheet folded FP and the -helical TMD of the TCR -chain.
FP5–13 inhibits T cell activation
Our molecular dynamics simulation suggested that the 5–13 region in FP mediated the interaction between FP and the TCR TMD (Fig. 6)
. To confirm this finding experimentally, we synthesized a peptide corresponding to the 5–13 region of FP1–16 (GALFLGFLG, named FP5–13) and studied its ability to interfere with T cell activation. We have described previously that FP inhibits the activation of the A2b T cell line by its cognate peptide Mt176–90 (8)
; this inhibition results from the activity of FP on A2b T cells, and not on the APC presenting Mt176–90 (8)
. To study the effect of FP5–13 on T cell activation, A2b cells were incubated with increasing concentrations of FP5–13 and stimulated with the Mt176–90 and irradiated APC. Figure 7
shows that addition of FP5–13 led to a dose-dependant inhibition of T cell proliferation; this effect was statistically significant (P<0.005) when compared to the effect of the control peptide p277. Hence, the interaction between FP and the TCR TMD is mediated by the 5–13 region of FP1–16.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
TMD is targeted by FP1–16, which interferes with the antigen-triggered activation of T cells in vivo and in vitro. The molecular dynamic simulations pinpointed the interaction to the 5–13 region of FP1–16. Indeed, we found that FP5–13 interferes with the antigen-triggered activation of T cells in vitro.
The insertion of FP into the membrane of the target cells anchors the molecular machinery that promotes HIV membrane fusion (23
24
25)
. Recent data indicate that this event takes place at the interface of the APC/T cell interaction, leading to the invasion of the T cell via the immune synapse (26
27
28)
. The immune synapse is an organized array of receptors and signaling molecules responsible for T cell activation on recognition of the agonist T cell epitope displayed on the surface of the APC. Thus, the localization of the immunosuppressive motif in FP1–16 suggests that the APC/T cell physical interaction is exploited by the virus to simultaneously invade the T cell and interfere with T cell activation. Taken together with the fact that HIV preferentially infects HIV specific T cells (29
30
31
32)
, our observations might explain the unsolved puzzle of decreased HIV-specific T cell immunity amid general T cell hyperactivation observed in HIV-infected patients in the early stages of the infection (33
, 34)
. Other immunosuppressive activities described for HIV might collaborate in the takeover of the immune response that finally leads to the collapse of the host’s immune system (25
, 35)
.
Obviously, FP plays a central role in viral replication; the virus cannot replicate unless it infects a cell. Viral proteins are under the constant pressure of the host’s immune system, which they escape by the generation of mutant variants. However some viral peptides, like FP, cannot accommodate a large degree of variability without compromising their functionality. Indeed, FP is one of the most conserved sequences within the viral genome, and HIV strains harboring a mutant FP show a decreased ability to infect target cells and replicate in vivo (36)
. In this context, the colocalization of the fusogenic and the immunosuppressor activities in FP might diminish the immunogenicity of FP, protecting it from the immune system’s pressure. This hypothesis is supported by the increased immunogenicity found in FP mutants that harbor a decreased immunosuppressive activity (8)
.
In this work we identified the cellular target of FP experimentally. Similarly to what we previously described for FP1–33 (8)
, FP1–16 colocalizes with the TCR in the T cell membrane (Fig. 3)
. Indeed, FRET studies suggested that FP1–16 implants within 50 Å of the TCR (data not shown). Moreover, FP1–16 could be coprecipitated with the TCR from the T cell membrane (Fig. 4)
. Thus, the signals required for guiding the colocalization of FP1–16 to the membrane and the interaction with the TCR are included in the first 16 N-terminal aa contained in FP1–16. Using two different experimental systems, a competitive immunoprecipitation assay and FRET (Fig. 5A,B
), we could conclude that FP directly interacts with the TMD of the TCR -chain. This interaction is mediated by the two positively charged aa in the TCR TMD (arginine and lysine); their removal led to a significant decrease in the interaction with FP. The participation of the arginine and lysine residues of the TCR TMD in the interaction with the FP is further supported by the results of the molecular dynamics simulation (Fig. 6)
and by the inhibition exerted by FP5–13 (Fig. 7)
.
Analysis of the experimental data and the computer simulation highlighted two outstanding features of the TCR/FP interaction: 1) it is a specific interaction between a ß-sheet (the FP) and an -helix (the CP); 2) FP interacts with polar aa in the target TCR TMD, although FP itself is not charged. Although unusual, such interactions might be favored by the special features of intermolecular interactions within the membrane milieu (37
, 38)
.
The FP manifests structural diversity; it has been shown to adopt both -helical and ß-sheet conformations (39
40
41
42
43
44
45
46
47
48
49)
. However, a 13C FTIR structural model of FP within the 70 N-terminal aa of gp41 shows that FP is likely to adopt a ß-sheet conformation in the context of the ectodomain (50)
. The results of the molecular dynamics simulation showed an unusual conformational transition of the FP1–16 peptide from an -helix into a ß-strand during the simulation trajectory. This conformational change allowed the creation of hydrogen bonds between the TCR polar residues and the FP backbone’s carbonyl atoms that further stabilized the FP/TCR complex inside the membrane. Thus, our model suggests that polar residues on the TCR TMD contribute to the stabilization of transmembrane complexes by interacting with the backbone of FP1–16. This observation is in agreement with the results depicted in Fig. 5
: the FP/TCR interaction is diminished when the two polar aa in the TCR TMD are substituted with glycines.
Nevertheless, our model does not completely satisfy the energetic requirements of the hydrogen donor/acceptor atoms of the backbone between an -helix and a single ß-strand inside the membrane. Note, however, that the gp41 protein forms trimers (51)
on the viral membrane and that FP itself oligomerizes in solution (44
, 50)
; this would allow the FP to adopt a ß-sheet conformation and still satisfy the energy requirement for hydrogen bonds that would stabilize its interaction with the -helix TCR TMD. The formation of FP trimers could not be accounted for in our model because of the need to reduce the complexity of the system while focusing primarily on the FP/TCR interaction. To our knowledge, this molecular dynamics model is the first prediction of such a complex transmembrane structure.
The targeting of the TCR/CD3 interaction by FP is reminiscent of the use of the CP peptide as an immunomodulator to inhibit T cell activation (52)
and interfere with the progression of autoimmune arthritis and diabetes (52
, 53)
. Notably the interaction of CP with the TCR/CD3 complex is also dictated by some unexpected structural constraints (53)
. Our results suggest that HIV might have naturally evolved a strategy to interfer with the TCR/CD3 interaction as a way to thwart a specific immune response to gp41. Overall, deciphering the molecular mechanism of the immunosuppressive activity of FP provides a new potential target to overcome the immunosuppressant activity of HIV, and in addition a tool for down-regulating immune-mediated disorders.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
2 Current address: Center for Neurologic Diseases, Harvard Medical School, Boston, MA, USA.
Received for publication August 9, 2006. Accepted for publication September 22, 2006.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Related Articles
This article has been cited by other articles:
|
|
 |
|
  A. B. Sigalov Interaction between HIV gp41 fusion peptide and T cell receptor: putting the puzzle pieces back together FASEB J, June 1, 2007; 21(8): 1633 - 1634. [Full Text] [PDF]
|
|
|
|
|
|
F. J. Quintana, D. Gerber, I. Bloch, I. R. Cohen, and Y. Shai Response to "Interaction between HIV gp41 fusion peptide and T cell receptor: putting the puzzle pieces back together" FASEB J, June 1, 2007; 21(8): 1635 - 1635. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||
