AGREEMENT dated18 August 2004
Exhibit 4.2(r)
“CONFIDENTIAL TREATMENT REQUESTED. CONFIDENTIAL PORTIONS OF THIS DOCUMENT HAVE BEEN OMITTED AND HAVE BEEN SEPARATELY FILED WITH THE COMMISSION. CONFIDENTIAL TREATMENT HAS BEEN REQUESTED WITH RESPECT TO THE OMITTED PORTIONS.”
AGREEMENT dated18 August 2004
BETWEEN
AUTOGEN RESEARCH PTY LIMITED(1) ABN 84 074 636 847 of ▇▇▇▇▇▇▇ ▇▇▇▇, ▇▇▇▇▇ ▇▇▇▇▇ ▇▇▇▇▇▇▇▇ ▇▇▇▇ (“Autogen Research”)
AND
INTERNATIONAL DIABETES INSTITUTE ACN 007 342 412 of ▇▇▇ ▇▇▇▇▇▇▇ ▇▇▇▇, ▇▇▇▇▇▇▇▇▇, ▇▇▇▇▇▇▇▇ ▇▇▇▇ (“IDI”)
RECITALS
A. On 14 January 1998 Autogen Research (then named Autogen Pty Ltd) and IDI entered into an agreement entitled Research, Licence and Commercialisation Agreement (“Research Agreement for Human Diabetes / Obesity Discovery”) setting out the terms and conditions for research to be carried out with the participation of the parties.
B. The parties now agree to extend the Research Agreement subject to the terms and conditions of this Agreement.
AGREEMENT
1. EXTENSION OF TERM
With effect on and from 1 July 2004 the parties agree that the term of the Research Agreement is extended until 30 June 2005 (“Extended Term”) (unless the Research Agreement is earlier terminated in accordance with its terms). During the Extended Term the terms and conditions of the Research Agreement will continue to apply except to the extent to which they are inconsistent with anything set out in this Agreement, in which case the provisions of this Agreement will prevail to the extent of the inconsistency.
2. PAYMENT AND RESEARCH PROPOSAL DURING EXTENDED TERM
During the Extended Term:
(a) the budget set out in Schedule 1 to this Agreement will be substituted for any payment program previously applying under the Research Agreement; and
(b) the research plan (including any milestones set out therein) set out in Schedule 2 to this Agreement will be substituted for any research proposal and workplans previously applying under the Research Agreement.
(1) Autogen Research Pty Ltd is a wholly-owned subsidiary of ChemGenex Pharmaceuticals Limited (ABN 79 000 248 304).
1
|
EXECUTED AS AN AGREEMENT |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
SIGNED FOR AND BEHALF of |
|
) |
|
|
|
|
AUTOGEN RESEARCH PTY |
|
) |
|
|
|
|
LIMITED ABN 84 074 636 847 |
|
) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|||
|
Director |
|
|
|
|
|
|
|
|
|
|
|
|
|
SIGNED FOR AND BEHALF of |
|
) |
|
|
|
|
INTERNATIONAL DIABETES |
|
) |
|
|
|
|
INSTITUTE ACN 007 342 412 |
|
) |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
||
|
Secretary |
|
|
|
Director |
|
2
SCHEDULE 1
Payments July 1, 2004 – June 30, 2005 payable by Autogen to IDI as follows:
Funding consists of quarterly payments of $[*], plus GST. The first payment will be made in September 2004. Subsequent payments will be made upon satisfactory completion of the performance reviews required under clause 5 of the Research Agreement, those performance reviews (for the avoidance of doubt) to be completed by the following dates –
• 12 October, 2004
• 12 January, 2005
• 12 April, 2005
IDI is required to forward the reviews to Autogen and following satisfactory performance, IDI will then invoice Autogen for quarterly payment of research in advance.
3
SCHEDULE 2. – INTERNATIONAL DIABETES INSTITUTE GENETICS RESEARCH LABS
CHEMGENEX PHARMACEUTICALS HUMAN GENOMICS PROGRAM RESEARCH PLAN
2004 / 2005
Human Genomics Gene Discovery Program
This section outlines our research plan to identify genes influencing quantitative physiological measures of Type 2 diabetes, obesity and related metabolic disorders using several populations. QTLs identified in Mexican American families from the San Antonio Family Heart Study (SAFHS), Wisconsin families from the MRCOB study and Mauritian families, genome wide scanned by AGT Biosciences, will be investigated.
QTL Selection
Three or more QTLs from the SAFHS, MRCOB or Mauritius studies will be analysed simultaneously to increase the number of potential disease genes under investigation. This approach should provide choice in candidate gene selection and allow for certain QTLs to be excluded in the event that high probability candidates are identified that have already been covered by patents or that do not represent good drug targets. Selected QTLs have been fine mapped to an average of 2 cM resolution to narrow the target region to the minimum achievable interval. Current QTLs under investigation include:
|
• |
|
[*] |
|
SAFHS |
|
[*] genes covered |
|
• |
|
[*] |
|
SAFHS |
|
[*] genes covered |
|
• |
|
[*] |
|
Mauritius |
|
[*] genes covered |
Future QTLs may include:
|
• |
|
[*] |
|
SAFHS |
Positional Candidates and Druggable Gene Ranking Strategy
The ranking strategy has been altered to include a stage of the SNPScan where all putative druggable genes will be evaluated with equal priority as positional candidates ranked on functional annotation. This measure has been introduced in response to a general concern that lack of speed in identifying disease genes in this program will result in an increasingly limited intellectual property space for protection of newly identified disease genes. Additionally, since chemically tractable or druggable targets are of prime interest to AGT Biosciences for development of therapies, it is critical that genes that fall directly into this class and are located under a linkage peak are investigated first.
Five criteria shall be used in the ranking scheme relying on functional annotations derived from published literature and ontological grouping based on primary amino acid sequence analysis; membership of families known to be chemically tractable by small molecule drugs, or amenable to antibody or protein type therapeutics; association of SNP markers across the region to the phenotype; gene expression level differences in mammalian tissues discordant for the phenotype; and location of the candidate gene relative to the centre of the linkage peak (human and baboon tissue). While these measures have been devised based on the latest theory and methods available in the field, due to the complex nature of the disease’s underlying genetic architecture and mechanism of pathogenesis, this strategy cannot
4
guarantee that the actual disease gene will be among the top ranked candidates.
Ranking Criteria Summary
1. Functional positional candidates as determined by online database information mining
2. Chemical tractability as determined by membership of families targeted by therapeutics
3. SNPScan for association of variants to disease – gene centric SNP selection
4. Gene expression levels by microarray in tissue
5. Location relative to centre of linkage peak
SNPScan
The SNPScan shall proceed in stages, the first stage shall focus on the top ranked candidate genes based on their known or inferred function and differential expression profiles in diseased vs. normal tissues (if available). The second stage shall include those genes likely to be druggable. Depending upon the number of genes and qualifying SNPs, these stages may be combined or spread over additional stages to facilitate their genotyping in the laboratory. The third through to nth stage of the scan shall examine SNPs in all genes starting in the centre of the linkage peak and progressively working outward. Each QTL shall be divided up into 5 or more stages depending upon the size of the interval. At each iterative stage, the latest release of the online SNP databases will be reviewed for newly submitted SNPs appearing in genes that were analysed in earlier stages. Qualifying SNPs will be incorporated into the next stage.
SNP Selection
The gene centric approach of SNP selection for each gene is as follows: SNPs will be selected from public databases for each gene, and shall include all exonic SNPs and SNPs in conserved flanking regions identified by comparison to other mammalian genomes. In particular, SNPs will be included if they fall within the putative promoter region (usually within 1-2 kb of the transcriptional start site), and up to 1 kb extending into intron 1 based on GGG trinucleotide and CpG dinucleotide content. SNPs falling within the 3’ UTR and flanking region will also be included if conserved. SNPs lying within conserved intronic and intergenic regions will also be included as these may comprise alternate exons (not yet annotated), enhancer regions, distant exons or complete genes not predicted by current software algorithms. However, these latter SNPs in the current database are sufficiently numerous to require sub-sampling. Therefore, we propose to initially select 1-2/SNPs per conserved region (size up to 10kb), and limit these selections to those SNPs with evidence of being polymorphic (i.e. “-cluster” or “–2-hit” on database). In the event the QTL is fully scanned without detecting a disease related gene, we may initiate the typing of additional intergenic and intronic SNPs.
Where a SNP is found associated or trending toward association (p<0.1), we shall type this within the entire cohort. If the association remains or strengthens, we shall proceed with typing additional SNPs within the gene (intronic) and those that may have appeared on the database since the initial selection.
5
Genotyping Cohort
The selected SNPs shall be typed in the founders of the family collection in which the QTL was identified to determine wether the SNP is polymorphic in the cohort, and whether it is associated with a disease related trait. At the conclusion of typing for each stage and prior to commencement of genotyping on the next stage, the data shall be analysed by the AGT Biosciences Centre for Statistical Genomics. If a positive association is identified between a SNP and a trait, the association shall be verified by typing the SNP in the whole family cohort and the association signal again measured. An increase in the significance of the association suggests that the genetic variation is influencing the trait or is in LD with a variant that is influencing the trait. If this is observed a conditional linkage analysis on these variants will be performed to determine its contribution to the QTL signal.
If no significant trait association among any of the SNPs typed within a SNPScan stage is observed, the next progressive stage of the SNPScan can commence. Since, as mentioned above, the GenBank and dbSNP databases are constantly being updated, the region will likely require re-annotation. If this is the case, a review of all previously SNPscanned regions shall be carried out so as to include genes and SNPs that appear on the new database release. Additionally a review of the tools and available mammalian genomic sequence shall be made to improve our detection of evolutionarily conserved regions.
Exhaustive genetic variation analysis of ranked candidate genes
Once ranked the candidate disease gene(s) will be evaluated by exhaustive variant analysis by resequencing of up to 50 individuals from the selected family cohort to identify all genetic variants segregating in the population. To maximise the chances of identifying functional variants within the candidate gene, we shall resequence all exons together with conserved regions from intronic, 5’ and 3’ flanking regions as determined by alignment with other mammalian species (eg mouse). Those variants shall then be genotyped in the entire cohort and tested for whether they contribute to the original linkage signal and thus influence the variation in the quantitative trait.
Bioinformatics
The GeneSniffer script will continue to be used to rank functional and druggable positional candidate genes. Further development for the script includes:
• Inclusion of rank parameter based on proximity to peak of linkage signal
• Development of a statistical assessment of the significance of the genes total “hit score” by evaluating scores obtained under the null hypothesis.
Summary
1 QTL Selection.
2 Fine Mapping of QTL.
3 Positional candidate gene selection.
4 SNPScan of public SNPs using association, conditional linkage and localised linkage disequilibrium.
6
5 Microarray based gene expression levels of genes in disease vs. normal tissue (Human, Baboon, ISR).
6 Location of gene relative to linkage peak.
7 Determine patent position on top ranked candidates.
8 Hypothesis development
9 Proceed with full variant analysis of top ranked genes based on SNPScan and expression data combined with commercial objectives.
10 Identify all genetic variation in gene by resequencing.
11 Type genetic variants in extended family pedigrees.
12 Test whether genetic variation accounts for linkage signal – identification of disease related gene. Continue through steps 7-12 on progressively lower ranked genes until disease influencing gene identified.
7
Israeli Sand Rat Candidate Genes
We shall continue to evaluate genes discovered by differential expression in the Israeli sand rat polygenic animal model of type 2 diabetes and obesity. Dependent upon functional information available for these genes (literature based and AGT generated), genetic variants will be genotyped in selected cohorts where QTLs are found for relevant phenotypes. Variation within the gene will then be assessed for association with relevant available quantitative traits by the AGT Biosciences Centre for Statistical Genomics.
SNPs will be selected as above according to our gene centric approach and genotyped first in the founders of the cohort to determine whether polymorphic. If they are polymorphic within the cohort, they shall be typed in the entire cohort and statistical association analysed. Genes currently undergoing statistical analysis include:
|
• |
|
[*] |
|
[*] |
|
MRCOB & SAFHS |
|
• |
|
[*] |
|
[*] |
|
MRCOB |
|
• |
|
[*] |
|
[*] |
|
Mauritius |
|
• |
|
[*] |
|
[*] |
|
MRCOB |
Genes that may be analysed in the future
|
• |
|
[*] |
|
[*] |
|
To be determined |
SELS was found to be strongly associated with several circulating markers of inflammation in the MRCOB cohort. We shall further investigate this association in the SAFHS cohort in addition to phenotyping of other relevant markers in this cohort.
Metabolic Syndrome QTL 3q27 – (NIH and NHMRC funding)
Funding was obtained during the 2003/2004 financial year to investigate the positional candidate genes [*] and [*] under the [*] linkage peak identified by ▇▇▇▇▇ ▇▇▇▇▇▇▇▇ in the MRCOB cohort. These genes will be re-sequenced in 50 individuals selected from families in this cohort contributing most strongly to the linkage signal. Identified variants will then be genotyped in the whole cohort and association and conditional linkage analysis undertaken to determine the level of contribution of the gene to the linkage signal. Additionally, gene expression level differences will be analysed using a BAC chip and cDNA obtained from muscle biopsies of selected MRCOB study participants.
A recent re-run of the GeneSniffer script revealed another strong functional positional candidate (majority homolog score) in the [*] region [*] (guanine nucleotide binding protein (G protein), beta polypeptide 4). This gene will be considered for re-sequencing and genotype analysis.
8
|
Milestones |
|
Completion Date |
|
|
|
|
|
|
|
SNPScan [*] |
|
|
|
|
Complete typing SNPs in all genes |
|
March 2005 |
|
|
Commence re-sequencing on selected genes |
|
June 2005 |
|
|
SNPScan [*] |
|
|
|
|
Complete typing SNPs in potentially druggable genes |
|
December 2004 |
|
|
Complete 1-2 additional SNPScan stages |
|
June 2005 |
|
|
SNPScan [*] |
|
|
|
|
Complete typing SNPs in potentially druggable genes |
|
March 2005 |
|
|
Complete one additional SNPScan stage |
|
June 2005 |
|
|
Israeli Sand Rat Candidate Genes |
|
|
|
|
[*] complete re-sequencing |
|
October 2004 |
|
|
[*] complete genotyping and phenotyping |
|
October 2004 |
|
|
[*] complete statistical analysis |
|
November 2004 |
|
|
[*] complete statistical analysis |
|
June 2005 |
|
|
Metabolic Syndrome [*] |
|
|
|
|
[*] complete re-sequencing |
|
December 2004 |
|
9