Predicting of molecular targets for immunotherapy against tumourleukaemic cells

1

• Predicting of molecular targets for
  immunotherapy against tumour/leukaemic cells
• 2) Immunoinformatics within a distributed
  computational environment

Mark Halling-Brown
Supervised by David Moss
2
Overview of Work
Prediction of Minor Histocompatability Antigens
SiPep algorithm and SNPBinder database Halling-Bro
wn et al. 2005. Sipep A System For The
Prediction Of Tissue-specific Minor
Histocompatability Antigens, Submitted to Journal
of Immunology Analysis of human SNPs (Work
Ongoing paper to be written)
Distributed Computational Pipelines
Antigenic Peptide Prediction Pipeline using ICENI
II Halling-Brown et al. 2005.Development of
Antigenic Peptide Prediction Pipeline using ICENI
II, to be submitted to Bioinformatics Development
of Workflow Resource Manager (Work Ongoing -
paper to be written)
3
Prediction of Minor Histocompatability Antigens
4
Treatment of Leukaemia
Leukaemia Patient

• Tissue Typing
• Blood Type A positive
• HLA A_0201, A3, B51

• mHAg, HA-1 HA-1H

V L H D D L L E A
HLA-A0201
5
Treatment of Leukaemia
Leukaemia Patient

• Tissue Typing
• Blood Type A positive
• HLA A_0201, A3, B51
• mHAg, HA-1 HA-1H

Blood Type A positive HLA A_0201, A3, B51 mHAg,
HA-1 HA-1H
6
Treatment of Leukaemia
Leukaemia Patient
Bone Marrow Donor Register
Blood Type A Negative HLA A_0201, A1, B51 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A1, B51 mHAg,
A2 Positive
Blood Type A Negative HLA A_0201, A3, B31 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A3, B31 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Positive
Blood Type A Negative HLA A_0201, A3, B31 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A3, B51 mHAg,
HA-1 HA-1H
Blood Type A positive HLA A_0201, A1, B51 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A1, B51 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Positive
Matching Blood Type MHC Alleles.
7
Treatment of Leukaemia
Leukaemia Patient
Bone Marrow Donor Register
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Negative
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Negative
Blood Type A positive HLA A_0201, A3, B51 mHAg,
HA-1 HA-1H
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Positive
8
Treatment of Leukaemia
Leukaemia Patient
Bone Marrow Donor Register
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Positive
Blood Type A positive HLA A_0201, A3, B51 mHAg,
A2 Negative
Blood Type A positive HLA A_0201, A3, B51 mHAg,
HA-1 HA-1H
Matching mHAg (HA-1). Looking for HA-1R.
9
Treatment of Leukaemia
Leukaemia Patient
Bone Marrow Donor
Chemotherapy
Radiotherapy
Blood Type A positive HLA A_0201, A3, B51 mHAg,
HA-1 HA-1H
Blood Type A positive HLA A_0201, A3, B51 mHAg,
HA-1 HA-1R
10
Treatment of Leukaemia
Leukaemia Patient
Bonor Marrow Donor
Bone Marrow Transplant
Blood Type A positive HLA A_0201, A3, B51 mHAg,
HA-1 HA-1H
Blood Type A positive HLA A_0201, A3, B51 mHAg,
HA-1 HA-1R
11
Donor Lymphocyte Infusion (DLI)
Bone Marrow Donor
Cured Patient
Leukaemia Relapse Patient
T-cells
Blood Type A positive HLA A_0201, A3, B51 mHAg,
HA-1 HA-1H
Blood Type A positive HLA A_0201, A3, B51 mHAg,
HA-1 HA-1R
12
Minor Histocompatability Antigens
HA-1H
HA-1R
V L H D D L L E A
V L R D D L L E A
HLA-A0201
HLA-A0201
mHAgs are immunogenic peptides from polymorphic
proteins
13
mHAgs Leukaemia Relapse Treatment
If we select a donor/patient pair that differ at
a specific mHAg Donor Lymphocyte Infusion using
Donor cytotoxic T cells specific to that mHAg.
Leukaemia Relapse Patient
Bone Marrow Donor
mHAg specific T-cells
V L H D D L L E A
V L R D D L L E A
14
Predicting mHAgs
348474
Peptide
Expression
15256
Protein
Proteasomal Cleavage
SAAP
12893538
Binding Predictions
45094
SNP
Binding Predictions
Proteasomal Cleavage
SNP
SAAP
Protein
Peptide
Expression
VLRDDLLEA
-1.46
G/A
R/H
NP_036424
5.28
Thyroid gland
VLHDDLLEA
0.768
Halling-Brown et al, 2005. SiPep A system for
the prediction of Minor Histocompatability
Antigens. Submitted to Journal of Immunology
15
Building a Query
Find all SNP-containing peptides
WHERE
All five scoring methods agree that the mutation
affects the binding to HLA-A0201
AND
The SNP has frequency data AND the contig
frequency lies between 0.4 and 0.8
AND
The source protein is expressed in Blood or
Lymph or Spleen
AND
The proteasome cleavage prediction score is
greater than 5.0
Search
16
Searching
Find all SNP-containing peptides
348,474
WHERE
All five scoring methods agree that the mutation
affects the binding to HLA-A0201
796
AND
The SNP has frequency data AND the contig
frequency lies between 0.4 and 0.8
69
AND
The source protein is expressed in Blood or
Lymph or Spleen
18
AND
The proteasome cleavage prediction score is
greater than 5.0
9
17
Screenshot of SiPep
18
Screenshot of SiPep
19
Distributed Computational Pipelines
20
Antigenic Peptide Prediction Pipeline
Web Client
OpenLaszlo Server
ICENI II Grid Middleware
Birkbeck Application Server
Imperial Application Server
FindConsensus
BIMAS
NHLAPred
SYFPEITHI
CombiPred
Immunoproteasome
Halling-Brown et al, 2005. Implementing an
antigenic peptide prediction pipeline with ICENI
II. Submitted to Bioinformatics
21
Antigenic Peptide Prediction Pipeline
22
Workflow Management Software
23
Workflow Management Software
24
Antigenic Peptide Prediction Pipeline
25
Results from mHAg predictions
26
mHAg Discovery Process
Predicted mHAgs

• Peptides synthesised and refolded with MHC class
  I
• Correctly refolded MHC class I is purified by
  gel filtration using an HPLC Superdex column
• Elution is monitored by spectrophotometer

ANRI
27
Predictions sent to ANRI
28
Predictions sent to ANRI
29
S(L/P)ANVSSCL bound with A2
I Aggregate II Protein/Peptide III Peptide
A2/(Control)
A2/SLANVSSCL
Optical density (mAU)
A2/SPANVSSCL
Time
30
S(L/P)ANVSSCL bound with A2
A2/(Control)
A2/SLANVSSCL
Optical density (mAU)
A2/SPANVSSCL
Time
31
Predictions sent to ANRI
32
Typing for Presence of SNPs
As a first step primers have been designed for
PCR typing of SNPs in normal subjects.
54ºc
72ºc
H20
Sample 1
PLLDLAAYD Outer primers
Sample 2
H20
SMDPLKLFD Outer primers
Sample 1
Sample 2
33
Acknowledgments
Anthony Nolan Research Institute Dr Paul
Traver Ruby Quartey-Papafio
London E-science Centre John Darlington Jeremy
Cohen William Lee Vesso Novo
BBSRC
34
Extra Slides
35
Other peptides bound with A3
NKDFFLSRL
PLLDLAAYD
QSLYSLTGL
SMDPLKLFD
36
Stem Cell Transplantation
Predominantly used as part of the treatment for
certain types of malignant diseases, mainly
leukaemia, lymphoma or myeloma that involve the
bone marrow (BM).
PBSC Peripheral Blood Stem cells
37
(No Transcript)
38
Transplantation Antigens

• Major
• HLA MHC alloantigens induce a harmful immune
  response at antibody and T lymphocyte immune
  levels
• ABO blood groups - ABO incompatibility is
  important only in antibody-mediated injury of the
  graft.
• Minor Minor Histocompatability Antigens (mHAgs)
  induce a harmful immune response at T
  lymphocyte immune level

39
Single Nucleotide Polymorphisms

• A Single Nucleotide Polymorphism or SNP
  ("snip") is a single base mutation in DNA. AgtT,
  AgtC, AgtG, TgtG..
• 90 of all human DNA polymorphisms (Collins et
  al, 1998)

SNP
ATAGCTCATGCATGCATGC
A
TCATCGATCGACATCAGTC
OR
C
ATAGCTCATGCATGCATGC
TCATCGATCGACATCAGTC
40
Known mHAgs
M.Bleakley S. R. Riddell. 2004. Molecules and
mechanisms of the graft-versus-leukaemia effect.
Nature reviews. 4.371-380
41
Allogeneic Immune Responses

• Graft Rejection (Host versus graft Response
  (HvGR))
• Graft-versus-host disease (GvHD)

42
Graft Rejection

• Rejection can be mediated by antibodies and/or
  lymphocytes

43
Acute Rejection - Direct
44
Acute Rejection - Indirect
45
Allogeneic Immune Responses

• Graft Rejection (Host versus graft Response
  (HvGR))
• Graft-versus-host disease (GvHD)

46
Graft-versus-host disease

• Donor derived mature T cells may be activated to
  reject host antigens and mount a widespread
  attack against the tissues of the host.
• Many organs can be involved, especially the
  skin, liver, eyes, mouth, lungs, GI tract, and
  neuromuscular system.
• 30 to 60 incidence of GvHD
• Ultimately, 20 to 40 of the
• patients die of complications
• associated with GvHD

47
GvHD Solutions

• Donor marrow treated with antibodies to markers
  on mature T cells (anti-CD3, anti-CD4, and CD8)
• Cord blood from the placentas of newborns can
  replace bone marrow as a source of haematopoietic
  cell transplants
• Eliminating all mature T cells before
  transplanting the marrow increases the likelihood
  of recurrent leukaemia
• Graft versus Leukaemia response

48
GvL/GvHD
Non-self
Non-self
Donor T-cell
Donor T-cell
Host mHAg
Host mHAg
MHC
MHC
Leukaemia Cell
Non-Leukaemia Cell
Graft-Versus-Leukaemia
Graft-Versus-Host Disease
49
Frequency of GvHD relapse
IncreasingRelapse
50
Separating GvL from GvHD

Possible targets for GvL effect
Several human minor histocompatibility antigens,
such as HA-1,HA-2, HB-1 and BCL2A1, are only
expressed by haematopoietic cells and are under
investigated as potential targets for a GvL effect
51
Overview
SNPs
dbSNP
SNPs in Proteins
Generate peptides
SNPs containing peptides
MHC binding Predictions
Peptides which affect binding
Other filters
Peptides which are possible mHAgs
52
Database SiPep SNPs in Peptides
Halling-Brown M, Moss D. The SiPep Database
Prediction of Tissue-Specific Minor
Histocompatability Antigens. In preparation
53
Database SiPep SNPs in Peptides
54
Gathered Data

• What information is stored?
• 13,877 Proteins
• 45094 Nonsynonymous SNPs
• 37 Class I MHC alleles
• 20 Amino acids

• Where is the information from?
• NCBI refseq, Unigene, swissprot, Stanford SOURCE
  and NCBI dbSNP
• Using HTTPrequests, SRS and downloaded
  databases

55
Generated Data Peptides
For each mutation on a protein, window around the
mutation to generate all possible peptides which
contain that mutation. Currently 348474 peptides
in the database
...ACBEERGYALEDILAGERAFGSTOUTFAWATERM...
YALEDILAGER
ALEDILAGERA
...ACBEERGYALEDILAGERAFGSTOUTFAWATERM...
LEDILAGERAF
...ACBEERGYALEDILAGERAFGSTOUTFAWATERM...
EDILAGERAFG
...ACBEERGYALEDILAGERAFGSTOUTFAWATERM...
DILAGERAFGS
...ACBEERGYALEDILAGERAFGSTOUTFAWATERM...
. . . . . . . . . . . . . . . . . . .
. . . . . .
SNP
56
Computed Data Binding
Each peptide has its binding score calculated
against available MHC alleles using a variety of
scoring methods.
Scoring Methods
Peptide/Mutated
Consensus
SYFPEITHI
CombiPred
Binder
DILAGERAF
BIMAS
nHLApred
Non-binder
DILAGEYAF
MHCPred
Simulated Annealing
CombiPred Raheel Shaban
Simulated Annealing Barry Smith
57
Computed Data Proteasome Cleavage Prediction

• Proteasome Cleavage
• Immunoproteasome 2 matrices, ProPred Altuvia
  et al
• Constitutive proteasome 1 matrices, ProPred

Addition Matrices score is calculated by summing
the scores of each position. E.g
GSPCSCTEPQGSPAP Score C(6) S(5) C(4) T(3)
E(2) P(1) Q(1) G(2) S(3) P(4)
A(5) P(6) 
58
Computed Data Expression
A UniGene cluster contains sequences that
represent a unique gene, as well as
related information such as the tissue types in
which the gene has been expressed and
map location.
E.g. Gene with clones from skeletal muscle (9
unique clones) and cardiac muscle (1 unique
clone). Frequency of 9 (Skeletal muscle
clones in UniGene Cluster X) UniGene Cluster
0.000600 in skeletal muscle
15000 (Total skeletal muscle clones in
UniGene) Frequency of 1
(Cardiac muscle clones in UniGene Cluster
X) UniGene Cluster 0.000100 in
cardiac muscle 10000 (Total cardiac
muscle clones in UniGene) Summation of Cluster
Expression Frequencies, by Tissue 0.000600
0.000100 0.000700 Normalized expression of
Cluster X in skeletal muscle 0.000600 /
0.000700 85.71 Normalized expression of
Cluster X in cardiac muscle 0.000100 / 0.000700
14.29
59
Example Predicted mHAgs
60
Transplants
Source dependant
61
Side effects of Immunosuppression

• They are serious
• Infection is a frequent side effect of
  immunosuppression Controlled by the appropriate
  antibiotic, antiviral drug, etc.
• Cancer Due to decreased immunosurveillance
• 5 or more of transplant recipients will develop
  cancer within a few years of receiving their
  allograft.
• Particularly skin cancers and lymphomas

62
Identification Techniques
High-performance liquid chromatography (HPLC) and
mass spectrometry Peptides are eluted from class
I molecules and fractionated by microcapillary
HPLC to identify fractions that contain the
epitope recognized by cytotoxic T lymphocytes
(CTLs). These fractions are analysed by tandem
mass spectrometry to identify the single peptide
sequence.Databases are then searched to identify
a candidate gene that contains a coding sequence
corresponding to the identified peptide. cDNA
expression cloning A cDNA library is prepared
from RNA from antigen-positive cells. These cDNAs
are cloned into an expression vector and
co-transfected with a plasmid encoding the human
leukocyte antigen (HLA)-restricting allele into
COS cells. The transfected cells are then
screened by co-culture with T cells that are
specific for minor histocompatibility antigen to
identify those that stimulate T-cell cytokine
production.Positive cells are subcloned to
identify those that expresses the cDNA encoding
the antigenic epitope. The antigenic epitope is
then localized by transfecting truncated
deletions of the gene into COS cells or by
prediction algorithms for HLApeptide
binding. Genetic-linkage analysis Genetic-linkage
analysis uses cell lines from large pedigrees,
such as the Centre dEtude Polymorphism Humain
reference families, that have been mapped for
polymorphic genetic markers. Cell lines are
transfected with cDNA encoding the relevant class
I HLA-restricting allele and evaluated for
recognition by minor-histocompatibility-antigen-sp
ecific CTLs. Pairwise linkage analysis is used to
identify the chromosomal region that regulates
the expression of the minor Histocompatibility
antigen. The draft sequence of the human genome
has improved the use of this method for
identifying candidate genes. Polymorphic-peptide
screening Polymorphic sequences of candidate
proteins are screened using prediction algorithms
for HLApeptide binding. Candidate peptides are
then pulsed onto antigen-presenting cells to
assess T-cell recognition.
63
Example results
299248 624718 758946 337600 42143 312456 108292 15
0008 150015 315092 44722
64
Overview

• What are Minor Histocompatability Antigens
  (mHAgs)?
• Treatment of Leukaemia?
• mHAgs in Donor Lymphocyte Infusion
• mHAg Discovery Process
• Results
• Other Results Antigenic Peptide Prediction
  Pipeline

65
Predicting mHAgs
Source Ensemble, ncbi
Peptide
Expression
Source ProPred Altuvia et al
Source Stanford SOURCE
Protein
Proteasomal Cleavage
SAAP
Binding Predictions
Source ensembl
SNP
Source NHLAPred, MHCPred, SYFPEITHI, BIMAS,
CombiPred
SNP
VLRDDLLEA
-1.46
G/A
R/H
NP_036424
5.28
Thyroid gland
VLHDDLLEA
0.768
Halling-Brown et al, 2005. SiPep A system for
the prediction of Minor Histocompatability
Antigens. Submitted to Journal of Immunology