Modeling the MHC-I pathway

Modeling the MHC-I pathwayBjörn Peters178576#36291.11.1.110576#357111.1.212352#3781.1.313445#378142152.1162.1.1172.1.22.1.3182.219202.3212.4115#36205#39240#3122315#56216#40508#34623242.52.5.12.625520#420262.727521#4212829483#5482.830174#36316#4031508#3583233342.935507#554362.103373.1383.239576#33340508#384413.342242#25576#336433.444291#288304#9145304#27146314#4547304#2713.4.148493.4.25051304#2713.552287#249533.6543.755564574.14.1.1584.1.24.1.3594.1.4363#2914.1.56061363#2914.24.2.16263496#31264551#329654.2.24.3664.3.1166#3755#21262#3767360#75126#53146#2068104#524.3.269300#1004.3.2.1101#87704.3.2.2714.3.2.379#3688#37323#165724.44.4.14.4.1.14.4.1.2734.4.1.34.4.1.44.4.24.4.2.174586#47975586#45876435#405774.4.2.2784.4.34.4.3.179600#30780814.4.3.2828384854.4.3.3864.4.3.4874.588894.690591929394959669798 References99100101102103104105106 Abbreviations107 Acknowledgments108109 Lebenslauf110 Publikationen111 Erklärung112enTable of contentsHelpModeling the MHC-I pathwayDissertationzur Erlangung des akademischen Grades
doctor rerum naturalium
(Dr. rer. nat.)
im Fach Biophysikeingereicht an der
Mathematisch-Naturwissenschaftlichen Fakultät I
der Humboldt-Universität zu Berlinvon
Diplom-Physiker Björn Peters geboren am 18.5.1973 in Hamburg

Präsident der Humboldt-Universität zu Berlin
Prof. Dr. Jürgen Mlynek

Dekan der Mathematisch-Naturwissenschaftlichen Fakultät I
Prof. Dr. Michael Linscheid Prof. Hermann-Georg Holzhütter Prof. Reinhard Heinrich Prof. Robert Tampe eingereicht am: 24.02.2003Tag der mündlichen Prüfung: 10.07.2003 Summary

A major task of the immune system is to identify cells that have been infected by a virus or that have mutated, and discriminate them from healthy cells. This duty is assigned to cytotoxic T-lymphocytes (CTL), which scan epitopes presented to them on cell surfaces derived from intracellular proteins through the MHC-I antigen processing pathway. The goal of this work is to provide computational methods that allow to predict which epitopes get presented from the large pool of peptide candidates contained in intracellular proteins. This is achieved by examining the selective influence of three major steps in the pathway: peptide generation by the proteasome, peptide transport into the ER by TAP, and binding of peptides to MHC-I molecules.

For peptide binding to MHC-I, a new algorithm is developed that combines a matrix-based method describing the contribution of individual residues to binding with pair coefficients describing pair-wise interactions between positions in a peptide. This approach outperforms several previously published prediction methods, and for the first time quantifies the impact of interactions in a peptide. The distribution of the pair coefficient values shows that interactions are not limited to amino acids in direct contact, but can also play a role over longer distances. Compared to the matrix entries, the pair-coefficients are rather small, explaining why methods completely ignoring interactions can nevertheless make good predictions.

Next, a novel algorithm is developed to predict the TAP affinities of peptides of any length. Longer peptides are important because several MHC-I epitopes are generated by N-terminal trimming of precursor peptides transported into the ER by TAP. As the true in vivo precursors of an epitope are not known, a generalized TAP score is established which averages across the scores of all precursors up to a certain length. The ability of this TAP score to discriminate between epitopes and random peptides shows that the influence of TAP is a consistent, strong pressure on the selection of MHC-I epitopes.

Using predicted TAP transport efficiencies as a filter prior to the prediction of MHC-I binding affinities, it is possible to further improve the already very high classification accuracy achieved using MHC-I affinity predictions alone. Such a 2-step prediction protocol failed when predictions of C-terminal proteasomal cleavages were combined with MHC-I affinity predictions. This disappointing result is thought to be caused by the lack of a sufficiently large set of quantitative and consistent experimental data on proteasomal cleavage rates, which are more difficult to measure and interpret than the affinity assays used to characterize peptide binding to TAP and MHC-I. Therefore, a new protocol for the evaluation of proteasomal digests is developed, which is applied to a series of experiments. This novel protocol addresses two problems: (1) Using mass-balance equations, a method is developed to quantify peptide amounts from MS-signals. (2) By introducing the first kinetic model of the 20S proteasome capable of providing a satisfactory quantitative description of the whole time course of product formation, cleavage probabilities can be extracted reliably from proteasomal in vitro digests.

Prediction Antigen Processing MHC TAP Proteasome Zusammenfassung

Das Immunsystems muss gesunde Zellen von infizierten und Krebszellen unterscheiden können, um letztere selektiv zu bekämpfen. Dies ist die Aufgabe der CTL-Zellen, die dazu auf der Zelloberfläche präsentierte Peptide die aus intrazellulären Proteinen der jeweiligen Zelle stammen untersuchen. Diese präsentierten Peptide (Epitope) werden durch den MHC-I Antigenpräsentationsweg hergestellt. Das Ziel dieser Arbeit ist es Methoden zu entwickeln die Epitope aus der großen Zahl prinzipiell in Proteinen enthaltener Peptide heraussuchen können. Dazu wird die Selektivität dreier wichtiger Komponenten des Präsentationsweges untersucht: Die Herstellung der Peptide durch das Proteasom, der Transport in das ER durch TAP, und das Binden von Peptiden an leere MHC-I Moleküle.

Zur sequenzbasierten Vorhersage der Bindung von Peptiden an MHC-I Moleküle wurde ein neuer Algorithmus entwickelt. Dieser kombiniert eine Matrix, welche die individuellen Beiträge einzelner Reste zur Bindung beschreibt, mit Paarkoeffizienten, die Wechselwirkungen zwischen verschiedenen Positionen im Peptid beschreiben. Dieser Ansatz macht bessere Vorhersagen als bisher publizierte Methoden, und quantifiziert erstmals den Einfluss von Wechselwirkungen innerhalb eines Peptids auf die Bindung. Die Verteilung der Werte der Paarkoeffizienten zeigt, dass sich Wechselwirkungen nicht auf benachbarte Aminosäuren beschränken. Im Vergleich zu den Matrixeinträgen sind die Werte der Paarkoeffizienten klein, was erklärt warum Vorhersagen die Wechselwirkungen komplett vernachlässigen trotzdem gut sein können.

Erstmals wurde ein Algorithmus zur Vorhersage der TAP-Transportseffizienz von Peptiden beliebiger Länge entwickelt. Das ist deshalb wichtig, da viele MHC-I Epitope als N-terminal verlängerte Prekursoren in das ER transportiert werden. Für die Vorhersage der Transportfähigkeit eines potentiellen Epitopes wird deshalb über die Transporteffiziens des Epitopes selbst und seiner Prekursoren gemittelt. Mit Hilfe dieser Definition von Transportfähigkeit wird gezeigt, dass TAP einen starken selektiven Einfluss auf die Auswahl von MHC-I Epitopen hat.

Indem man Peptide die als 'nicht-transportierbar' vorhergesagt werden als mögliche Epitope ausschließt, kann man die ohnehin schon hohe Qualität von MHC-I Bindungsvorhersagen weiter steigern. So eine zweistufige Vorhersage scheitert, wenn man statt des TAP Transports die Vorhersage der Generierbarkeit eines Epitopes durch das Proteasom als Filter verwenden möchte. Dieses schlechte Abschneiden der proteasomalen Schnittvorhersagen wird auf eine mangelhafte experimentelle Datenbasis zurückgeführt, da proteasomale Schnittraten schwieriger zu messen und interpretieren sind als die Affinitätsdaten für TAP und MHC-I. Um die experimentelle Datenbasis in Zukunft verbessern zu können, wurde ein neues experimentelles Protokoll entwickelt und an einer Reihe von Experimenten getestet. Dabei werden zwei Probleme behandelt: (1) Durch die Verwendung von Massenbilanzen werden MS-Signale in quantifizierte Peptidmengen umgerechnet. (2) Durch das erste kinetische Modell des Proteasomes das die Entstehung und den Abbau von Peptiden während eines Verdaus zufrieden stellend beschreiben kann, können aus den Verdaudaten Schnittraten bestimmt werden.

Vorhersage Antigen Prozessierung MHC TAP Proteasom Table of contents

1 Introduction - the MHC-I pathway
- 1.1 Structure and function of the main pathway components
  - 1.1.1 The proteasome generates peptides by degrading proteins
  - 1.1.2 TAP transports peptides into the ER
  - 1.1.3 MHC-I molecules present bound peptides on the cell surface
2 Peptide binding to MHC-I
- 2.1 Overview of existing prediction methods
  - 2.1.1 Matrix prediction methods: BIMAS, SYFPEITHI and PM
  - 2.1.2 The independent binding assumption
  - 2.1.3 General prediction methods: ANN, CART and the additive method
- 2.2 Experimental datasets
- 2.3 Obtaining predictions from published methods
- 2.4 Introducing the stabilized matrix method (SMM)
- 2.5 Evaluating prediction quality
  - 2.5.1 Statistical significance for differences in AUC
- 2.6 Comparison of matrix based predictions: SMM, PM, BIMAS and SYFPEITHI
- 2.7 Comparison of general predictions: ANN, CART and the additive method
- 2.8 Extending SMM with pair coefficients
- 2.9 Distribution of pair coefficient values
- 2.10 Summary
3 Peptide transport by TAP
- 3.1 Published prediction methods of in vitro TAP affinity
- 3.2 Comparison of affinity predictions for 9-mers
- 3.3 Predictions of TAP affinities for longer peptides
- 3.4 Using TAP transport predictions for the identification of epitopes
  - 3.4.1 TAP transport predictions for individual MHC-I alleles
  - 3.4.2 Consequences of the uncertainty as to which N-terminally extended precursors are generated in vivo
- 3.5 Combining TAP transport predictions with predictions of MHC-I affinity
- 3.6 Confidence in the values of the free parameters α and L
- 3.7 Summary
4 Peptide generation by the proteasome
- 4.1 Evaluating published algorithms predicting proteasomal cleavage
  - 4.1.1 NetChop
  - 4.1.2 PaProc
  - 4.1.3 FragPredict
  - 4.1.4 Identifying epitopes using proteasomal cleavage predictions
  - 4.1.5 Combining proteasomal cleavage predictions with predictions of MHC-I affinity
- 4.2 Problems with evaluating experimental proteasome digests
  - 4.2.1 A single snapshot of a digest does not provide reliable cleavage rates
  - 4.2.2 MS-signals do not give quantified peptide amounts
- 4.3 Novel protocol of experimental evaluation
  - 4.3.1 Determining peptide amounts from MS-signals
  - 4.3.2 Kinetic modeling
    - 4.3.2.1 Procession rate
    - 4.3.2.2 Cleavage probability
    - 4.3.2.3 Definition and estimation of rate constants
- 4.4 Application and testing of novel protocol
  - 4.4.1 Experimental setup
    - 4.4.1.1 Peptide synthesis
    - 4.4.1.2 Purification and analysis of 20S proteasome complexes
    - 4.4.1.3 Peptide digestion assays.
    - 4.4.1.4 HPLC-MS analysis
  - 4.4.2 Comparing theoretical and experimentally derived fragment amounts
    - 4.4.2.1 Calibration curves
    - 4.4.2.2 Assessment of peptide amounts from MS signals using the mass balance method
  - 4.4.3 Fitting the kinetic model to the experimental data
    - 4.4.3.1 Comparison of experimental and theoretical time-dependent amount profiles
    - 4.4.3.2 Assessing the variability of model parameters with a jack-knife procedure
    - 4.4.3.3 Checking the equivalence of model computations based on either the mass balance method or experimental calibration curves
    - 4.4.3.4 Adequate fitting of data requires a length-dependent procession rate
- 4.5 Differences between constitutive- and immuno-proteasomal digests
- 4.6 Summary
5 Summary of main results and conclusions
6 Outlook
References
Abbreviations
Acknowledgments
Lebenslauf
Publikationen
Erklärung

Tables

Table 1: SMM scoring matrix for binding to HLA-A0201
Table 2: Comparison of prediction quality
Table 3: Pair coefficient values
Table 4: TAP consensus matrix
Table 5: Individual alleles
Table 6: Combined TAP and MHC-I predictions
Table 7: Amount dependent signal deviations monitored in the calibration curves
Table 8: Estimated model parameters for the pp89-25mer digests
Table 9: Estimated model parameters for the LLO-27mer digests

Images

Figure 1: Schematic overview of the MHC-I antigen processing and presentation pathway
Figure 2: Structure of the 20S yeast proteasome published by (Groll, et al., 1997).
Figure 3: Epitope (blue) in the binding groove of the MHC-I α-chain (orange). Structure published by (Khan, et al., 2000).
Figure 4: MHC-I bound epitope is scanned by T-cell receptor. Structure published in (Garboczi, et al., 1996).
Figure 5: Cross validated distance as a function of λ
Figure 6: ROC curve for matrix based prediction methods on the Blind-set.
Figure 7: ROC curves for general prediction methods on the HLA-A2 Blind-set.
Figure 8: Classification tree for peptides binding to HLA-A0201
Figure 9: Using cross-validation to determine the optimal λ'
Figure 10: Distribution of pair coefficients.
Figure 11: Comparison of predicted and measured in vitro TAP affinity values of 9-mer peptides
Figure 12: The Cross validated distance of measured and predicted TAP affinities is plotted as a function of λ
Figure 13: Comparison of predicted and measured in vitro TAP affinity values for peptides longer than 9 amino acids.
Figure 14: ROC curves for the HLA-X dataset.
Figure 15: Prediction quality for the HLA-X dataset as a function of the maximal precursor length
Figure 16: Prediction quality for the H2-X dataset as a function of the maximal precursor lengths
Figure 17: Prediction quality on a simulated dataset
Figure 18: ROC curves for the combined TAP and MHC-I prediction on the HLA-A0201 dataset
Figure 19: ROC curves for proteasomal cleavage predictions
Figure 20: ROC curves for proteasomal cleavage + MHC-I binding predictions
Figure 21: Different proteasome species with identical cleavage preference can produce large differences in individual fragment amounts
Figure 22: Re-processing of peptides makes the relative amounts of fragments associated with each cleavage site time dependent
Figure 23: Possible partitions of a peptide containing 2 cleavage sites
Figure 24: Fragments of the pp89-25-mer
Figure 25: Fragments of the LLO-27mer
Figure 26: Calibration curve for the 11mer peptide 5-DMYPHFMPTNL-15 contained in the pp89-25mer.
Figure 27: Comparison of experimentally and theoretically determined signal conversion coefficients for fragments derived from the pp89-25mer
Figure 28: Time courses of peptide amounts for the pp89-25mer digests
Figure 29: Variability of model parameters for the pp89-25mer digests assessed by a jack-knife procedure: Experimental peptide amounts determined by the mass balance method
Figure 30: Variability of model parameters for the LLO-27mer digests assessed by a jack-knife procedure: Experimental peptide amounts determined by the mass balance method
Figure 31: Variability of model parameters for the pp89-25mer digests assessed by a jack-knife procedure: Experimental peptide amounts determined by using calibration curves