Survey
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project
Mathematical modelling assessing the likely effectiveness of anti-HIV gene therapy Associate Professor John M. Murray School of Mathematics and Statistics, & The Kirby Institute, University of New South Wales, Sydney NSW 2052, Australia [email protected] Melbourne Workshop February 2012 1 Outline • A growing HIV epidemic despite antiretroviral therapy. • Gene therapy as a new treatment. • Mathematical modelling of effect of ribozyme gene therapy delivered to hematopoietic stem cells – an ordinary differential equation model. • Different gene therapy classes. Melbourne Workshop February 2012 2 The HIV Pandemic Melbourne Workshop February 2012 3 Number of people living with HIV is increasing • At the end of 2009 there were 33 million people living with HIV in the world. • Each year this amount increases: – In 2009 there were 2.6 million new infections and 1.8 million AIDS deaths. – Over 6 million people in low/middle income countries receives antiretroviral therapy (ART). – However global funding is not increasing. Global HIV/AIDS Response, Progress Report 2011 UNAIDS Melbourne Workshop February 2012 4 The number of people living with HIV in developed countries is increasing • The size and average age of the HIV-infected population in developed countries are both increasing. • The complications associated with both the disease and ART will be compounded by age-related factors. Men and women living with Diagnosed HIV in Australia Melbourne Workshop February 2012Cysique et al., Sexual Health, 2011 5 Treatment of HIV with antiretroviral drugs • Successful in reducing HIV-related morbidity and mortality. • Does not eliminate virus. AIDS deaths in Australia cART becomes available Melbourne Workshop February 2012 Palmer et al., PNAS, 2008 6 Other treatment approaches are needed • ART requires daily use for the rest of a person’s life. • The lengthening of an HIV-infected person’s life with ART has been tremendously beneficial but this also results in longer duration of ART. • More HIV-infected people and longer duration of ART results in increased direct cost – though this is outweighed by the benefits. • Ageing of the HIV-infected population and the resulting need for drugs for age-related issues can lead to complications with concurrent ART. • Some individuals have resistant virus to many of the available drug classes. • A non-drug approach with a single application would be beneficial. Melbourne Workshop February 2012 7 HIV infects CD4+ cells • HIV virions bind to CD4 receptors on the surface of immune cells such as CD4+ T cells, monocytes, macrophages. • Also require attachment with a co-receptor, initially CCR5 (later CXCR4). • After infection, the viral genome HIV RNA, is reverse transcribed into HIV DNA and inserted into the host cell genome, to begin production of new virus. De Clerq, IJAA, 2008 Melbourne Workshop February 2012 8 Gene therapy in other diseases • Gene therapy involves the transfer of genetic material into cells of an individual to treat an underlying illness either through the expression of advantageous genes or the silencing of disadvantageous ones. • Has been used to treat other illnesses such as SCID-X1 (‘bubble babies’). Melbourne Workshop February 2012 9 Anti-HIV gene therapy • Potentially a single infusion of genetransduced cells will provide life-long protection from AIDS. • Avoids toxicity issues associated with ART. • What cells should be targeted? Tutorvista.com – CD4+ T cells or, – Hematopoietic stem cells (HSC, CD34+ cells). Melbourne Workshop February 2012 Rossi, Nat. Biotech, 2007 10 What cells do you transduce with the gene therapeutic – CD4+ T cells or CD34+ hematopoietic stem cells (HSC)? Ledger et al., Cell-delivered gene therapy for HIV, in Recent Translational Research in HIV/AIDS, 2011. Melbourne Workshop February 2012 11 Recent HIV gene therapy trials Cohen, Science, 2007 Melbourne Workshop February 2012 12 Delivery of gene therapy only to CD4+ T cells will miss HIV reservoirs • Collection of CD4+ T cells for transduction with gene therapy occurs through apheresis from peripheral blood. • A large percentage of CD4+ T cells do not traffic to peripheral blood. • Macrophages in tissue are a source of HIV infection that would not be affected by the therapy. Melbourne Workshop February 2012 13 Output of new CD4+ T cells from thymus decreases with age • Introducing gene therapy into stem cells of an adult will result in slow delivery of protected CD4+ T cells. Murray et al., Imm. & Cell Biol., 2003. Melbourne Workshop February 2012 14 Mathematical modelling of HIV • First mathematical modelling of HIV that had a significant impact described changes in virus levels (HIV RNA/ml) after the start of ART (Perelson et al., Science 1996, Wei et al., Nature 1995). • Simple mathematical models following changes in HIV RNA after ART estimated for the first time the turn-over of infected cells I (t1/2 ~2 days) and virions V (t1/2 6 hours). • Strength of this modelling was related to dI kVT I dt dV NI cV dt – Experimental data – The “right” (simplest) model to incorporate data and biology Melbourne Workshop February 2012 15 Mathematical modelling of HIV gene therapy • • • HIV researchers at Johnson & Johnson Research Australia had developed an anti-HIV gene therapy and wanted to estimate its impact. The therapy, a ribozyme that bound to and cleaved part of HIV RNA, was to be delivered to HSC. How many HSC need to be transduced with the gene therapy for an observable effect in viral load after 1 year? – HIV RNA assays are accurate to about 0.5 log10 so this was the minimum change needed. • • Delivery of gene therapy to HSC and observations and changes estimated from effects on viral production from CD4+ T cells in peripheral blood. Needed modelling of the downstream effects of gene therapy in HSC to impact throughout body. Melbourne Workshop February 2012 HIV sequence bound and cleaved by ribozyme Hammerhead ribozyme Amado et al., Human Gene Therapy 2004 16 Naive and Memory T cells Naive T cells Memory T cells Antigen Activated effector cells against antigen • Each T cell can target only a single antigen – Clonal Selection Theory. So need lots of different T cell clones. • Cells exported from the thymus are naive – they have not yet seen the antigen they are specific for. • When a T cell comes in contact with its antigen it becomes activated and proliferates to produce many copies of itself. Most of these die once the antigen has been cleared. • At some stage in this process some of the activated cells become memory cells. If the antigen re-occurs then there are now more memory cells specific for that antigen (compared to the original number of naive cells) and these are faster to respond and more efficient at clearing antigen. • The purpose of vaccination is to produce memory cells that will eliminate the infection before it has a chance to takeWorkshop off. February 2012 Melbourne 17 Naive and Memory T cells • If we want to get our gene therapy into all/most T cells then also need them to develop into memory cells. So dependent on general antigenic stimulation of naive cells via activated cells. • Productive HIV infection preferentially occurs in activated cells – HIV usually infects through the CCR5 co-receptor, so activated and memory cells. – Integration of HIV DNA into the host cell genome requires some energy which is more likely in activated cells. • Memory cells can have long lifespans. Memory to previous antigen exposure can exceed 50 years. • Describing how “protected” T cells expand in the periphery requires description of the dynamical processes underlying generation of memory cells. • HIV infection is then overlaid on the normal system. Melbourne Workshop February 2012 18 • Mathematical modelling of gene therapy delivered to hematopoietic stem cells (HSC) Need to incorporate the homeostatic processes regulating T cells to properly model the effects of gene therapy in HSC that produce protected T cells. • Deliver gene to HSC in bone marrow. • These produce T cell precursors that mature in thymus. • New T cells exported from thymus are naive to antigen (N). • Memory T cells (M) have arisen from previous antigen contact. • Activated T cells (A) arise from antigen stimulation. These cells are more susceptible for productive infection by virus (V) to give rise to infected cellsMelbourne (I). Workshop February 2012 19 Change in cell populations and virus thymus Production and homeostasis of CD4+ T cells activation death – n(V)N – nN reversion + mnM dN/dt = s(age) dA/dt = (n(V)N+m(V)M)(1 – A/A) – aA – – aA – k(V)A dM/dt = aA + 2k(V)A – m(V)M – mnM – mM dI/dt = k(V)A – II dV/dt = NI – cV Melbourne Workshop February 2012 Standard HIV infection model 20 Infection predominantly occurs in tissue and includes long-lived and latently infected cells • Infected macrophages are a source of infection. Their longer lifespans than productively infected CD4+ T cells (half-life of ~1 day) enables continued transmission. • As well as productively infected CD4+ T cells, HIV establishes a reservoir of latently infected CD4+ T cells (HIV integrated but not producing virus), that with reactivation keeps infection going. • Need to model the effect of gene therapy in HSC leading to protected macrophages and impact on replenishment of latent pool. Melbourne Workshop February 2012 21 Model of macrophage infection and transmission M infected macrophages Ld CD4+ T cells with defective unintegrated HIV DNA Lu with competent unintegrated HIV DNA Li with integrated HIV DNA P productively infected CD4+ T cells V free virus Melbourne Workshop February 2012 22 Model of macrophage infection and transmission I kvV (1 )u Lu i Li I V NI cV Ld k d M Ld Lu ku M ( u ) Lu L L ( ) L i u u i i i Standard HIV model with source from latent infection Generation of latent infection and long-lived infected cells M k m M m M Melbourne Workshop February 2012 23 The hard part of mathematical modelling - parameter values • Approximately 17 year half-life for total thymic output with age. • death rates = log(2)/(half-life) • half-lives: naïve cells 10 years (McLean & Michie, PNAS, 1995) memory cells 10 years activated cells 8 days infected cells 2 days free virus 6 hours (Perelson et al., Science, 1996) • reversion of 5% of activated to memory: a = 0.05a • clonal expansion: on average 7 to 8 rounds of cell division, =200 • homeostatic level of activated cells: A=60 • virions produced per infected cell per day: N=320 • half-life of unintegrated HIV DNA 2.3 days • other parameters chosen to duplicate HIV RNA and DNA levels in 5 untreated seroconverters, 4 with AZT, 9 with ART Melbourne Workshop February 2012 24 Effects of OZ1 gene therapy • Modelled the effects of delivering the anti-HIV ribozyme, termed OZ1, targeting and cleaving a conserved region of HIV-1. • CD34+ stem cells were assumed transduced with OZ1 to P%. • Assumed this delivered same percentage of pre-thymocytes with this gene. • Effects of these genes in CD4+ T cells and macrophages based on in vitro experiments. Melbourne Workshop February 2012 25 OZ1 assumptions from in vitro analyses • OZ1+CD4+ T cells 1/10th as likely to be infected. • Infected OZ1+CD4+ T cells 1/20th as productive. • OZ1+ macrophages not infected. • Sensitivity to these and other assumptions performed. • Simulations in Matlab. Melbourne Workshop February 2012 26 Simulated HIV progression without gene therapy Uninfected HIV infected Total CD4+ T cells Memory cells Naive cells Murray et al., Journal of Gene Medicine, 2009. Melbourne Workshop February 2012 27 Estimated effects of gene therapy from simulations Log10 HIV RNA/ml reduction: After 6 months After 1 year After 2 years Percentage OZ1+ HSC Melbourne Workshop February 2012 • Estimated transduction of approximately 20% of HSC to achieve observable decrease in viral levels at 1 year. • HIV RNA assays are accurate to about 0.5 log10 so this was the minimum change needed. 28 Phase II trial of OZ1 Nature Medicine, 2009 • A randomized, doubleblinded trial of OZ1 was conducted. • Delivered in addition to ART and evaluated during structured treatment interruptions. • It was shown to be safe with some effect. Melbourne Workshop February 2012 29 Current HIV gene therapy • • • • • • • The OZ1 trials was important as it established that cell-delivered gene transfer is safe and biologically active in the setting of HIV. Most focus around HIV gene therapy now is on reduction of CCR5 expression in cells. CCR5 is a molecule on the surface of immune cells involved in normal immune signalling. It is also used as a coreceptor by HIV to infect a cell. Some individuals have a genetic mutation (‘Delta 32’), that results in lower or no CCR5 on their cells. They are less likely to be infected, and if they are then they progress more slowly. The “Berlin Patient” who had a bone marrow transplant from an uninfected individual with the Delta 32 mutation, is the only person to have cleared HIV infection (Hutter et al., NEJM 2009). The hope is an anti-CCR5 gene therapy will perform similarly without the need to find a matching bone marrow donor. Melbourne Workshop February 2012 Photobucket 30 Advantages of anti-CCR5 gene therapy • Gene therapy against CCR5 expression is targeting a cellular process. Normal cellular mutation is very low so the likelihood of an individual’s cells changing back to CCR5+ is negligible. • Targeting a viral process, as is done in ART, does not have this advantage as HIV mutates very quickly. • Additionally gene therapy against CCR5 is a Class I therapy, as it stops infection, rather that reducing viral production once infected. Melbourne Workshop February 2012 Ledger et al., 2011 31 Mathematical modelling of different gene therapy classes • Lund et al., (Bulletin of Math. Biol. 1997), von Laer et al., Applegate et al (Retrovirology 2010) have mathematically demonstrated the relative advantages of the different gene therapy classes. • Class III genes provide no selection advantage. • Evolutionary pressure from HIV infection and cell death allow Class 1 containing cells to multiply and protect. Von Laer et al., Journal of Theoretical Biology 2006 Melbourne Workshop February 2012 32 Von Laer et al., Journal of Gene Medicine 2006 Summary • HIV gene therapy is an expanding field with great promise. • Mathematical modelling has an important role in the development of gene therapy. • Estimation of the effects of this therapy require analysis of a complicated immune-virus dynamical system. • Biomedical researchers appreciate the part mathematicians can play. However they rarely communicate in the language of mathematics. • For there to be an effective interaction, mathematicians must make themselves familiar with the biomedical background, and speak to the researchers in their own language. Melbourne Workshop February 2012 33 Acknowledgements • Johnson & Johnson Research. • Calimmune. • Dr G. Symonds, JJR/Calimmune. • Australian Research Council Linkage Grant Melbourne Workshop February 2012 34