SEARCH

SEARCH BY CITATION

Keywords:

  • antigen presentation;
  • epitopes;
  • ligand–receptor interaction;
  • MHCII ;
  • HLA-DM

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

The recognition by CD4+ T cells of peptides bound to class II MHC (MHCII) molecules expressed on the surface of antigen-presenting cells is a key step in the initiation of an adaptive immune response. Presentation of peptides is the outcome of an intracellular selection process occurring in dedicated endosomal compartments involving, among others, an MHCII-like molecule named HLA-DM (DM). The impact of DM on the epitope selection machinery has been known for more than 15 years. However, the mechanism by which DM skews the presented repertoire in favour of kinetically stable complexes has remained elusive. Here, a review of the most recent observations in the field is presented, pointing to the possibility that DM decides the survival of a peptide–MHCII complex (pMHCII) on the basis of its conformational flexibility, which is a function of the ‘tightness’ of interaction between the peptide and the MHCII at a specific region of the binding site.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

Class II MHC (MHCII) molecules are transmembrane heterodimeric proteins expressed on the surface of antigen-presenting cells, and they initiate or propagate immune responses by presenting antigenic peptides to CD4+ T lymphocytes.[1] The MHCII molecules feature a high level of polymorphism, predominantly restricted to the peptide-binding site. This groove-shaped domain is the main structural characteristic of the MHCII and defines its function. Each individual expresses a small number of different MHCII molecules. Hence, each of these must be able to bind a large number of different peptides to ensure an immune response against many possible pathogens.[2]

The MHCII-restricted presentation of peptides to CD4+ T cells can be considered the outcome of an intracellular selection process. MHCII molecules are transported from the endoplasmic reticulum through the Golgi to the MHCII compartment (MIIC) as complexes with the chaperone protein invariant chain (Ii).[3, 4] Ii stabilizes the nascent MHCII and prevents the binding of other endoplasmic reticulum-resident polypeptides. Upon arrival in the MIIC, the Ii molecule is cleaved by proteases, leaving a peptide fragment termed CLIP in the MHCII binding groove. CLIP is then released by the action of HLA-DM (DM) to allow antigenic peptides derived from the fragmentation of engulfed proteins to bind MHCII. The exchange role of DM is not limited to CLIP, as it can promote the exchange of peptides to select for a kinetically stable peptide–MHCII complex (pMHCII) repertoire.[5]

Structural properties of petides bound to class II

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

The MHCII binding site consists of two α helices laterally enclosing a platform formed by eight strands of β sheet. Because the groove is open at both ends, peptides of various lengths can interact with the MHCII as a type II polyproline helix.[6] Hydrophobic side chains of the peptide are sequestered within polymorphic pockets at the extremities of the binding site (‘major anchors’, usually indicated as P1 and P9 pockets, numbered from the N-terminus to the C-terminus). Smaller pockets or shelves generate auxiliary anchoring sites (P4, P6, P7). Depending on the allele, ionic interactions may be involved. The interaction between peptide side chain and the deep pocket at P1 position is often considered a dominant source of binding energy.[7] Finally, a conserved array of hydrogen bonds (H-bonds) is established between MHCII side chains and peptide main chain atoms. In particular, residues α51, α53, α62, α69, α76, β81 and β82 of the MHCII are involved in forming this set of interactions (reviewed in ref. [2]

Multiple conformers of complexed and empty MHCII molecules

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

The conformation of different pMHCII complexes is nearly identical as identified in crystallographic analysis. These usually stable forms of the class II molecule are referred to as closed or ‘compact’.[8] However, there is evidence that MHCII are structurally flexible and can adopt different conformations.[9-12] A ‘floppy’ species with reduced mobility in non-boiled non-reducing (also known as ‘gentle’) SDS–PAGE has been observed in vitro at low pH [8] and as an intermediate in the thermal denaturation and folding pathways for some murine MHCII. The ‘floppy’ species has also been observed in vivo for some MHCII produced in mice lacking Ii, in which the cellular trafficking is altered.[13] Alternative conformational states have been indicated also with respect to peptide loading ability.[14, 15] The ‘peptide-receptive’ form is generated after release of a bound peptide and can rapidly bind a new peptide at endosomal pH (kon ≈ 105 m−1 s−1), whereas in the absence of a peptide this isomer is unstable, inactivating with a half-life of a few minutes into the ‘peptide-averse’ form. The latter isoform does not itself bind peptide but can slowly (t1/2 ≈ 3 hr for the murine I-Ek,[16] t1/2 ≈ 15 hr for the human MHCII allele HLA-DR1 [17]) isomerize into the active molecule. For the ‘averse’ form, the peptide-binding reaction has a complicated kinetic behaviour, which has led to a proposed multistep peptide-binding pathway in which an initial pMHCII undergoes a unimolecular change to generate the stable complex.

The identified unimolecular kinetic step has been interpreted as an indication of the structural rearrangement the MHC and the peptide experience during complex formation.[16-18] A series of biophysical studies provided evidence in support of the hypothesis that peptide binding induces structural rearrangements in the MHCII.[15, 19, 20] Peptide-free DR1 appears to have a larger hydrodynamic radius than the peptide-occupied form (29 Å versus 35 Å) and also a decreased helicity, as measured by circular dichroism. These modifications would be accompanied by partial folding/unfolding of the β1 helix residue 58–69, which is the epitope of an antibody specific for the human MHCII devoid of peptide.[21] Some of the conformational modifications observed in this region have been correlated with binding and release of short peptides that would be able to fill only the P1 pocket and extend only for a few residues. These results have been interpreted as the evidence that P1 pocket occupancy would be able to trigger a global conformational change within the protein, which propagates from the peptide-binding site to the opposite end of the β subunit. However, complete conversion to the compact, stable form would be possible only with contributions from both side chain and main chain interactions.[20]

Molecular dynamics simulations have also identified regions that may be involved in the peptide-binding-induced modification.[22, 23] These studies have confirmed that the β58–70 amino acidic sequence is such a region, and it may exist in an equilibrium of conformational states. Residues α51–54 appear to constitute a very flexible region as well. These amino acids are part of an extended strand close to the P1 pocket, and they undergo a dramatic rearrangement during peptide binding or release. Indeed, upon simulated removal of the peptide, the α50–59 region of DR would fill the N-terminal end of the peptide-binding site occupying, in part, the area where the antigenic peptide is usually found. A sharp kink would form at Gly α58, allowing the region α50–59 to fold into the binding site, taking the place of the bound peptide in the P1 to P4 region.

DM mechanism: an open question

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

Despite its discovery 15 years ago, the mechanism of DM action has remained poorly understood. Initially, DM was identified through the study of mutant B-cell lines that expressed only CLIP/MHCII complexes on their surface. Genetic mapping studies localized the defect to the class II region, and subsequent work showed that transfection of functional DM genes could correct the antigen presentation defect.[24, 25] As DM was so structurally similar to MHCII, the mechanism by which it would promote CLIP release and antigenic peptide loading was not immediately obvious.[26] Structural and biochemical evidence suggested that DM does not function by binding to peptide. However, using purified DM and DR molecules, many groups were able to show that DM is able to catalyse the release of CLIP from the antigen-binding groove, while at the same time promoting the binding of antigenic peptides.[27, 28] Further observations suggested that DM could act in an enzyme-like fashion, and that the rate of peptide exchange was directly proportional to the intrinsic dissociation rate of the pMHCII complex.[29] These results led to the hypothesis that DM functions as a general purpose peptide exchange catalyst.[30] However, experiments examining the activity of DM during peptide loading in vivo suggested that DM also has the ability to act as an MHCII-specific chaperone by stabilizing empty MHCII under low pH conditions.[31-33] In contrast to the expected 1 : 1 ratio, quantitative immunoblot analysis demonstrates a 5 : 1 molar ratio of MHCII to DM, which is more consistent with a catalytic role for DM than simply chaperone-like.[34]

In the attempt to reconcile DM's catalytic activity on the dissociation of the bound peptide with the one facilitating loading of peptide into the MHCII groove, many groups began to investigate the mechanism by which DM molecules interact with MHCII. Unfortunately the crystal structures of DM or the murine H2-M [35] did not reveal any obvious structural features that might explain peptide exchange activity for either molecule. Clearly, an association of DM to DR appeared to be required, as DM/MHCII complexes could be immunoprecipitated from solubilized cells under low pH conditions.[36] Indeed, the altered conformations of both MHCII and DM induced by low pH may favour binding.[37] To date, any attempt to co-crystallize MHCII/DM complexes has failed, but it now appears likely that the lateral face of the MHCII molecule near the N-terminus of the bound peptide is the site of interaction (Fig. 1).[38-40]

image

Figure 1. Modelled structure of DM/peptide–MHC class II (pMHCII) complex. Top (a), front (b) and side (c) view of the DM/CLIP/DR3 complex generated through molecular docking simulation. DM α-chain is shown in light green, β-chain in dark green. DR3 α-chain is shown in blue, β-chain in pale pink. The peptide is shown as stick in yellow. Coordinates are from ref. 38. The model was visualized with PyMol.[69]

Download figure to PowerPoint

The structural studies of the DM/MHCII interaction have not been sufficient to outline a conclusive mechanism of DM activity. Several works have been published in which the focus was on determining the characteristics that make a pMHCII complex susceptible to DM-mediated peptide release. The initial hypothesis postulated that the intrinsic dissociation rate of the complex was directly related to its susceptibility to DM-mediated exchange, and the factor by which the DM-catalysed rate constant for peptide release exceeded the rate constant of the uncatalysed reaction was indicated as j factor.[29]

The observation that the j factor was constant for complexes with different off-rates suggested that DM promotes peptide release by destabilizing sequence-independent interactions, such as the H-bond network. Indeed, several works have indicated the H-bond network as a viable target of DM activity, possibly promoting or stabilizing a form of the MHCII in which one or more of the H-bonds from the peptide main chain to the MHCII are broken.[41, 42] In particular, it was proposed that DM specifically targets the H-bond formed by the conserved histidine at position β81 in MHCII molecules.[43] HLA-DR1 molecules with an asparagine substitution at this position were reported to form highly unstable peptide complexes, and peptide dissociation was not further enhanced by DM, possibly because DM cannot disrupt a hydrogen bond that does not exist in the mutant molecule.

Nevertheless, not all the observations can be explained by postulating a disruptive activity of DM on one or multiple H-bonds. In particular, the evidence that the destabilization of single H-bonds has a cooperative effect on peptide stability [44, 45] is hard to reconcile with the sequence-independent j factor. Moreover, different reports have shown that complexes unable to form the H-bond at position β81,[46-48] as well as any other conserved H-bonds,[46] are still susceptible to DM-mediated peptide release.

A model of DM activity that is becoming increasingly accepted postulates that DM would recognize a specific and flexible conformation of class II, rather than a kinetically unstable pMHCII. The first evidence in support of this model was gained through the analysis of a mutant DR1, DR1βG86Y.[49] This mutant remains permanently in a receptive form when empty, most likely because the tyrosine substituting the wild-type glycine fills the P1 pocket and prevents the flexible N-terminal region from collapsing. DR1βG86Y forms only short-lived complexes with the peptide but features low affinity for DM. As the conformations of the mutant DR1 and wild-type (wt)DR1 bound to low-affinity peptides feature different levels of rigidity, and DM was shown to interact preferentially with the latter, it was proposed that the flexibility present in the wtDR1 loosely bound to a low-affinity peptide was determinant for DM/pDR1 interaction.

Structural features of DM labile pMHCII complex

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

If conformational traits of the pMHCII complex are crucial for the interaction with DM, the next step towards a comprehensive model of DM activity is defining the structure of the DM-labile conformer. Our inability to resolve the crystal structure of the DM/pMHCII triad suggests a great structural flexibility of the pMHCII complex targeted by DM. However, two reports have provided important insights into the conformational aspects that render a pMHCII complex amenable to DM-mediated peptide exchange. The first was based on the analysis of αF54-substituted DR1 molecules.[50] These mutants were shown to be more susceptible to DM-mediated peptide release than wtDR1 bound to a high-affinity peptide, they featured increased affinity for DM, and increased peptide vibration, especially in the H-bonding network at the N-terminal site of the complex. The crystal structure of the mutant MHCII identified peculiar structural features at this site of the pMHCII dyad, in particular a reorientation of the α45–50 region and changes in the flanking extended strand regions (α39–44 and α51–54). Importantly, the aforementioned molecular dynamics studies have predicted that the wtDR1 may also assume a conformation that resembles the one shown by the αF54C mutant. Moreover, many of the residues in the MHCII postulated to be important for DM interaction are located within this region and undergo rearrangement during the modifications indicated by the analysis of the mutant MHCII.

The second model has been suggested by analysing DM interaction with peptide/HLA-DR2 variants, indicating that DM specifically binds DR molecules in which the N-terminal site of the complex is emptied.[51] Indeed, this study clearly showed that DM did not interact with DR molecules loaded with a covalently bound peptide, whereas deletion of the first three N-terminal residues of the linked peptide (and the relative H-bond network) was associated with strong DM binding. Therefore it appears that the weakening of the cluster of interactions between the peptide and the binding groove at the N-terminal precedes DM binding.

Hence, DM would play a critical role in the decision-making process as to whether a complex will be selected for presentation based on the conformational flexibility of the N-terminal side, inclusive of the P1 pocket and surrounding H-bond network, associated with the binding state of the peptide in this region.

Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

Considering the magnitude of structural modifications that both the peptide and the MHCII binding groove undergo during interaction, the question of DM-mediated peptide exchange has been approached in terms of DM effect on the folding–unfolding of the complex.[47, 52] From a methodological standpoint, measurements of folding and conformational rearrangements can be performed via analysis of cooperative effects.[53, 54] In the absence of DM, peptide binding to and release from MHCII were shown to be cooperative.[44, 55, 56] When the same analysis was performed in the presence of DM, no cooperativity could be observed in the release of the pre-bound peptide.[52] This evidence was interpreted as an indication that DM promotes a dramatic disruption of the interactions between MHCII and the peptide, so that the typical coordinate unfolding of the intrinsic release is not present. Interestingly, measuring cooperativity for the exchange peptide revealed that the latter needs to fold into the groove more efficiently than the pre-bound to displace it, and DM increases the energetic threshold that the exchange peptide has to overcome to displace the pre-bound. Importantly, through different biophysical approaches, that report also showed that DM requires an exchange peptide (of proper affinity) at equimolar or greater concentrations than the preformed complex to promote the maximal extent of exchange the system would realize based on the relative binding affinities of the two peptides. Hence, the exchange peptide appears to play the important role of ‘cofactor’ in DM-mediated release of the pre-bound peptide.

However, one aspect of DM-mediated peptide dissociation observed in the latter work was particularly intriguing. A small, though measurable, release of peptide was detected even in the absence of any exchange peptide. A follow-up article recently published has provided a possible explanation for this phenomenon.[57] At the beginning of a peptide release assay in the presence of DM and in the absence of an exchange peptide, pMHCII complexes would assume two conformers, which can be identified through ‘gentle’ SDS–PAGE, one of which is DM-labile (L form). The DM-stable conformer (S form) does not release peptide in the presence of DM, until an exchange peptide is added. Probably the most interesting observation was that the incubation of isolated S conformer with an equimolar amount of exchange peptide in the absence of DM results in the formation of a conformer with an electrophoretic mobility similar to that of L, which in turn is DM labile. This evidence sheds light on DM's requirement for an exchange peptide to promote the release of the pre-bound ligand.

Taken together, the most recent observations of DM-mediated peptide release indicate that the pMHCII complex needs to assume a specific conformation (αF54C mutants, DR2 mutants and the L form mentioned in the latter report) to interact with DM. The generation of this conformer is, to a certain extent, a function of the affinity of the bound peptide. However, it appears that the presence of exchange peptides, rather than a characteristic intrinsic to the complex, is critical in promoting the formation of complexes with increased affinity for DM. In the endosomal milieu a similar mechanism would provide a chance for any of the available peptides to attempt to fold the MHCII. In a contrasting scenario, the first peptide that can complex with an MHCII in a form with low affinity for DM would freeze the epitope selection machinery, limiting the breadth of the presented antigenic repertoire.

With these insights, a ‘compare-exchange’ model of DM mechanism has been suggested [52] (Fig. 2), in which the presence of exchange peptide generates a structural rearrangement of the pMHCII complex possibly by colliding into the α54F or other regions of the MHCII molecule that can trigger morphological modifications. The conformational changes may promote a weakening of the H-bond network at the N-terminal of the complex and, depending on the distributed binding energy of the complex, promote an initial DM-independent release of the peptide, leaving the P1 pocket emptied. Once devoid of peptide, the N-terminal side of the complex would feature an increased structural fluctuation, favouring the number of microstates in which the α45–50 region is reoriented of about 20° and features a partial unwinding from a tight 310 helix toward a more canonical α-helical pitch.[50] This rearrangement is accompanied by a modification of the shape and volume of the P1 pocket. The rearranged complex would feature a high affinity for DM and would be susceptible to DM activity. The binding of DM might trigger a dramatic destabilization of the remaining interactions between the MHCII and the loosely tethered pre-bound peptide. At this point a metastable intermediate is reached, with DM bound to an MHCII interacting with two peptides. The exchange peptide would now have a chance to be tested by the MHCII based on the ability to form a structurally tight complex. However, the geometry of the intermediate allows the pre-bound peptide to rebind if the exchange peptide does not succeed in forming a closed complex. DM would be released from the complex once this assumes a collapsed conformer with the cluster of interactions between the peptide and the MHCII at the N-terminal stabilized.

image

Figure 2. A model for exchange peptide-induced peptide–MHC class II (pMHCII) complex conformational rearrangement as the basis for DM peptide exchange activity. MHCII molecules stably complexed with peptides in a tight conformation are not susceptible to DM interaction (1). The addition of an exchange peptide promotes a conformational rearrangement of the pMHCII complex leading to a partial release of the pre-bound peptide at the N-terminal of the complex. The rearranged structure features an increased flexibility, with the result that residues important for the interaction with DM have a greater probability than in the stable complex to be available for DM binding (2). The two-peptide/MHCII intermediate has a great affinity for DM and is susceptible to DM activity. The pre-bound peptide is further destabilized and remains loosely tethered to MHCII (3). The exchange peptide can now attempt to form a complex (4) and fold itself and the binding groove in a tight conformer (5). The geometry of the intermediate allows the pre-bound peptide to rebind in case the exchange peptide does not succeed (6). To the extent that either peptide forms a tight complex with MHCII, the residues important for DM binding are re-engaged, directly or indirectly, in the interaction with the peptide. As a consequence, DM is released, and the complex can undergo a new cycle of ‘compare-exchange’ or be shifted to the plasma membrane depending on availability of peptides and pH.

Download figure to PowerPoint

The above model of DM-mediated peptide exchange is consistent with what has been proposed on the basis of molecular dynamics simulation analysis,[23] suggesting a dual role for DM during peptide binding. First, because of the stabilization of the groove in an ‘exchangeable’ form, DM shifts the control of peptide binding from kinetic to thermodynamic. Second, because of the competition by DM for binding to the P1 pocket ‘neighbourhood’, the effective free energy threshold for peptide binding is increased. Hence, only peptides with a sufficient affinity for binding can compete for the P1 pocket, which in turn also results in DM dissociation.

A critical aspect of the ‘compare-exchange’ model is the existence of an MHCII/two-peptide intermediate. Such an intermediate was also proposed for the exchange reaction in the absence of DM. In particular, the two-peptide/one MHC complex has been adopted to explain observations from several groups indicating an accelerated release of a pre-bound peptide either at the cell surface or in vitro in the presence of free peptide.[12, 58-60] Initially it was thought that the effect of accelerated dissociation was specific because only I-Ed binding peptides were able to accelerate the dissociation of the hen egg lysozyme 107-116/I-Ed complex either on the surface of cells or in purified forms in solution, and high-affinity I-Ed binders did not affect the half-life of purified ovalbumin 323–339/I-Ad complexes.[60] There is evidence that peptides that may not feature a high affinity for a given allele can promote release of a peptide bound to that allele.[58] The replacement reaction accelerated by a second peptide was indicated as push-off, and was experimentally observed in gels first,[59] and subsequently in solution.[12]

In particular, the action of a push-off peptide, dynorphin A (dynA-[1-13]) was examined on the dissociation kinetics of the PCC-(89–104)/I-Ek complex. Kinetic analysis, fluorescence resonance energy transfer (FRET), and 19F NMR analysis determined the molecular mechanism of push-off. The results indicated that the first step of push-off is indeed the formation of a two-peptide/one-MHC complex in solution. Although estimates of the relative proportion of the two-peptide/MHCII complex were low in those studies, (1·0–0·1%), these complexes were preferentially associated with the ‘open’ conformer of the pMHCII complex during PAGE analysis.[12]

Evidence in support of a two-peptide/MHCII transition state in the presence of DM is provided by FRET experiments in which peptide-to-peptide energy transfer was detected only in reactions containing a preformed complex and an exchange peptide.[52] Further support for this model is provided by kinetic stability of pMHCII complexes in the presence of DM and the absence of an exchange peptide.[52, 57, 47] In consideration of the correlation between two-peptide intermediates and ‘open’ conformers, the observed DM-associated increase in inter-peptide FRET has been interpreted as evidence that DM recognizes the ‘open’ MHCII resulting from the interaction with the two peptides.

An important step in defining the two-peptide/MHCII intermediate and refining the exchange mechanism in general will be mapping the location where the exchange peptide interacts with the pre-bound peptide/MHCII complex. Exchange peptides with different chemistry need to be recognized, so one possibility is that the competitor peptide interacts with a distinct (presumably less polymorphic) site present across MHCII alleles. Analysing the ‘peptide exchangeability’ of MHCII molecules carrying ad hoc mutations in the absence or presence of DM might be an approach to address these questions.

Interestingly, the possibility that the two-peptide/MHCII intermediate and the push-off mechanism occur both in the absence of DM at neutral pH and in the presence of DM at acid pH broadens the possibilities for loading MHCII molecules efficiently under different conditions. Consequently, the question arises as to whether a similar breadth of binding conditions also takes place in vivo and whether it might regulate alternative loading or recycling pathways of class II MHC molecules.

Effect of MHC-II polymorhism on DM activity

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

The extensive polymorphism characterizing MHCII molecules affects the stabilities of class II heterodimers and plays a role in determining the extent to which DM exerts its function. In vitro experiments have shown allele-dependent association of DM with empty class II.[32] Studies performed in transfected cells have identified the allele-specific requirement of DM during class II-restricted antigen presentation, however different groups reached contradictory conclusions.[61-64] It is likely that the complementation assays adopted in those works to investigate DM activity could be affected by additional experimental variables, such as abnormal expression levels and functional contributions by recipient cell lines, impairing our ability to evaluate the significance of these observations. To rectify these technique-related inconsistencies, mutant mice were generated expressing known ratios of different MHC class II alleles and Ii chain via homologous recombination in embryonic stem cells. Experiments conducted in these animals showed clear evidence for distinctive isotype-specific modes of peptide capture and dependence on DM.[65, 66]

These studies led to an investigation of the possibility that human MHCII molecules also feature a diversified DM and/or Ii requirement for appropriate trafficking and antigen presentation. Presentation by HLA-DR depends on both Ii and DM, and DM activity on several DR alleles (DRB1*0101, 0201, 0301, 0401, 0402, 0404) has been demonstrated. The availability of crystal structures for several DR molecules in complex with relevant epitopes and the relative facility to purify large amounts of these proteins in a stable form have led to a focus of the analysis of DM on the interaction with these specific alleles. A significant deviation from this trend is constituted by a recent report showing that DR, DQ and DP differ markedly in their requirements for Ii and DM, despite having 70% amino acid sequence similarity. For instance, it seems that Ii is sufficient for DQ to attain a stable SDS conformation in the absence of DM, and SDS-stable DQ5 dimers can be identified through dimer-specific antibodies recognizing the stable conformation. These observations are consistent with studies conducted on DQ alleles, suggesting that DM-independent antigen presentation by these MHCII may constitute a risk factor for autoimmune disease.[67]

Therefore it appears that DM can interact and function on a variety of MHCII alleles; however, the actual requirement of DM for efficient antigen presentation may be isotype-specific. We are not fully aware of the reasons as to how and why the effect of DM on epitope selection differs on an allele basis. If DM recognizes a flexible conformation of the pMHCII complex and promotes a destabilization of the interactions near the P1 pocket, it is plausible that DM-independent alleles may feature an increased rigidity related to a specific pocket structure that renders such alleles a low-affinity (or overall ‘insensitive’) target of DM activity. Structural analysis and in vivo studies of these different isotypes will contribute to increase our understanding of the different paths of epitope selection and it will indicate whether we need alternative mechanisms to explain the outcome of DM interaction with different MHCII alleles. Moreover, a deeper analysis of the molecular properties of DP and DQ conformation and stability and their looser DM requirement for proper trafficking may offer an explanation as to why some autoimmune diseases are linked to these alleles.

pH-dependence of DM-mediated peptide exchange

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

An interesting aspect of the interaction between DM and MHCII that has not received enough attention is the dependence on protonation of DM function. It has been evident since the initial studies that the ability of DM to promote peptide exchange has an optimum at pH 4·5–5·5 and it is dramatically weakened at pH 7. Through time-resolved fluorescence anisotropy and fluorescence binding studies with 8-anilinonaphthalene-1-sulphonate, conformational rearrangements of DM and HLA-DR3 promoted by pH changes were probed. With this approach it was shown that both molecules increased their degree of non-polarity upon protonation, and that the interaction between DM and DR limited the exposure of these pH-sensitive non-polar areas to solvent. The observation that acid pH facilitates the uncovering of interior non-polar patches that are usually shielded at neutral pH was interpreted as the driving mechanism behind the enhanced DM–DR association in lysosomal conditions and the increased catalytic action of DM on class II peptide loading.[37] Subsequently, acquisition of CD and fluorescence spectra confirmed that DM exists in spectroscopically distinguishable, rapidly interconvertible states at pH 7 and pH 5.[68]

In consideration of the structural modifications consequent to changes in protonation, a more thorough analysis of the effect of pH on peptide binding and DM activity should be pursued. As suggested in past reports, a deeper understanding of the role played by pH and its modifications within the MIIC would point to possible mechanisms of regulation of the epitope selection process. For instance, one could speculate that depending on the availability of exchange peptides and the pH present in the endosomal milieu, DM would be able to operate as a peptide editor. As the endosomal pH moves toward neutral values, DM-assisted exchange machinery becomes less efficient until it stalls. The final compact complex can be shifted to the plasma membrane for presentation. Because exchange appears to be a function of peptide KD, the probability of finding a high-affinity peptide in a compact conformer is the greatest. However, to the extent that a low-affinity peptide generates a DM resistant conformer in the proximity of neutral pH, this mechanism also allows such ligands to be exposed for T-cell recognition.

Conclusions

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

The work of several laboratories has advanced our understanding of the mechanisms by which DM affects peptide exchange and skews epitope selection. However, resolving the structure of the DM/pMHCII complex at atomic resolution remains a crucial step toward the definition of the rules governing DM function. The ability to link pMHCII binding energetics, complex conformation and DM function will be reached only through structural studies, providing critical insights to define DM activity.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References

I wish to specially thank Dr Jack Gorski for his remarkable mentorship and for his inspiring creative thinking. Funding for the research described here was from National Institutes of Health grant R01AI63016 to Dr Gorski. This work was supported by National Institute of General Medical Sciences of the National Institutes of Health under Award Number P20GM103395 and by the Pfizer-sponsored Aspire Award Number WS1907326. The content is solely the responsibility of the author and does not necessarily represent the official views of the National Institutes of Health or Pfizer. The author has no financial conflicts of interest.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Structural properties of petides bound to class II
  5. Multiple conformers of complexed and empty MHCII molecules
  6. DM mechanism: an open question
  7. Structural features of DM labile pMHCII complex
  8. Generation of the DM-labile conformer and a ‘compare-exchange’ model of DM-mediated peptide exchange
  9. Effect of MHC-II polymorhism on DM activity
  10. pH-dependence of DM-mediated peptide exchange
  11. Conclusions
  12. Acknowledgements
  13. References