DECODING THE GENETIC SOFTWARE OF PARKINSONS DISEASE

Transcript

• 00:01Great. So I'm excited to
• 00:03kick off with, talk about
• 00:05the work in my laboratory.
• 00:07And and, with the ambitious
• 00:09goal to decode
• 00:11and simulate
• 00:13the genetic software of Parkinson's
• 00:14disease. We are very grateful
• 00:16to our sponsors. This is
• 00:18really a team effort. Parkinson's
• 00:20disease
• 00:21is a very important,
• 00:23problem. There it's actually now
• 00:25the fastest growing
• 00:27brain disease in the world,
• 00:29outpacing,
• 00:30the growth rates of Alzheimer's
• 00:32disease.
• 00:33There are ten million people
• 00:34with Parkinson's disease in the
• 00:35world today, and this in
• 00:37by some estimates, this number
• 00:39may increase to twenty five
• 00:40millions in in twenty fifty.
• 00:43And the burden for patients
• 00:45and their families and the
• 00:46health care costs are an
• 00:48enormous.
• 00:49And today's
• 00:50medicine
• 00:52does not stop the disease.
• 00:54It is reactive,
• 00:55and it treats patients
• 00:57as if they were all
• 00:58the same.
• 01:01So we want to change
• 01:02this paradigm.
• 01:04And and the big question
• 01:06that we're trying to tackle
• 01:07is, can we develop
• 01:09a predictive,
• 01:12precise,
• 01:13and preventive
• 01:14medicine for Parkinson's disease?
• 01:17And to to make this
• 01:19possible,
• 01:20we are we have set
• 01:21ourselves a very ambitious goal,
• 01:22and that is to make,
• 01:24digital twins of Parkinson's brain
• 01:26cells and and Parkinson's,
• 01:30patients.
• 01:30And so how are we
• 01:31going about this?
• 01:36And under the hood
• 01:38are, ten multimodal,
• 01:40multiomics,
• 01:41technologies.
• 01:43We're doing short and and
• 01:45long single,
• 01:46read sequencing of single cells.
• 01:48We've done about a million.
• 01:50We're we're going moving towards
• 01:52ten million.
• 01:53We do spatial transcriptomics,
• 01:56on the spot level and
• 01:58on the cellular and subcellular
• 02:00level, with the xenon,
• 02:02looking at ATAC seq for
• 02:04open chromatin and whole genome
• 02:06sequences.
• 02:08And we're integrating all of
• 02:09that. And
• 02:11so what are the components
• 02:13to develop,
• 02:14digital twins of of brain
• 02:16cells? Well, number one, we
• 02:17need to know all the
• 02:18brain cells. So far, we
• 02:20have cataloged
• 02:21ninety
• 02:22two,
• 02:23cell types
• 02:24based on the sequencing of
• 02:25a million brain cells.
• 02:32And and we have started
• 02:34to map,
• 02:35the the
• 02:37the brain space, the layers,
• 02:39and and the spatial niches,
• 02:41to which the cell types,
• 02:43can be localized.
• 02:45And,
• 02:46doctor Junjun Dong will delve
• 02:48into this,
• 02:50in much more detail in
• 02:51in in a in a
• 02:52second.
• 02:53But here here is sort
• 02:55of an initial version of
• 02:57the integration of cells and
• 02:59space,
• 03:00done by Jacob Parker, who
• 03:02is somewhere here.
• 03:04They are,
• 03:05showing
• 03:06how these different
• 03:08cell types
• 03:09localize
• 03:11to specific layers in the
• 03:13temporal cortex or to specific
• 03:15niches,
• 03:16in in in the human
• 03:17midbrain.
• 03:20So what can we do
• 03:21with this Atlas? Well,
• 03:23number one, we can,
• 03:25lay out all this multi,
• 03:28omic
• 03:29datasets and integrate them, and
• 03:32we can,
• 03:33see how disease progresses,
• 03:36in this dynamic,
• 03:38view of transcriptional
• 03:40changes.
• 03:41And,
• 03:42again, the next talk will
• 03:44look at the dynamic
• 03:45evolution of the disease across
• 03:48space and time.
• 03:50We can also link this
• 03:52to pathology
• 03:54and to clinical phenotypes
• 03:56with the ultimate goal to
• 03:57use
• 03:58this,
• 04:00molecular
• 04:01signatures
• 04:02to predict and prevent,
• 04:04disease in in patients and
• 04:06prevent this progression.
• 04:09So but one of the
• 04:11very cool applications
• 04:14of this prototype
• 04:16digital twins is
• 04:18that it
• 04:20allows
• 04:21to infer genome function,
• 04:24in particular brain cells.
• 04:27And and and so this
• 04:29allows
• 04:30to
• 04:31look at the genome sequence,
• 04:33input the genome sequence,
• 04:35and have as a readout
• 04:37a prediction
• 04:38of RNA changes in specific
• 04:41brain cells.
• 04:43And,
• 04:45to do this, what are
• 04:46the components that we need?
• 04:48Well, we need to we
• 04:49need to
• 04:51have,
• 04:52all the DNA variants.
• 04:55We want we need to
• 04:56wire them to the RNA
• 04:58changes
• 05:00in specific brain cells,
• 05:03combine convergent,
• 05:05pathways to identify processes, and
• 05:08thereby, delineate
• 05:09the gene regulatory networks from
• 05:12DNA variants
• 05:13to, cellular processes.
• 05:16And if we are successful
• 05:18with this, it will give
• 05:20us targets
• 05:22to,
• 05:23prevent that patient's
• 05:25disease progresses,
• 05:28to slow movements,
• 05:30motor Parkinson's, and cognitive
• 05:32decline, and instead
• 05:34turn patients,
• 05:37into slow progressors
• 05:38that are able
• 05:40to enjoy an awesome quality
• 05:42of life and play golf,
• 05:44for fifteen years and, enjoy
• 05:46time with the grandchildren.
• 05:49So
• 05:50what do we know about
• 05:51the noncode
• 05:53about the, DNA variants? What
• 05:55can we input in into
• 05:56our prototype?
• 06:00They're in principle
• 06:02two
• 06:03two types of,
• 06:06two two ways the genome
• 06:08of functions functions. One is
• 06:10sequence modulation,
• 06:13where
• 06:14mutations
• 06:16change the protein coding sequence.
• 06:18And that's the case in
• 06:20Mendelian
• 06:20forms of the disease that
• 06:22comprise about three percent of
• 06:24all Parkinson's patients.
• 06:26And,
• 06:27Shri Ganachandra
• 06:28and Pietro de Camille will
• 06:30take a deep dive on
• 06:32some of these,
• 06:33familial
• 06:34genes.
• 06:36However,
• 06:37most Parkinson's patients
• 06:40don't have a mutation that
• 06:42changes protein sequence.
• 06:43Instead,
• 06:44there are seven thousand fifty
• 06:46seven
• 06:47non coding DNA variants,
• 06:49linked to Parkinson's disease.
• 06:53They account for up to
• 06:54thirty six percent of the
• 06:55genetic heritability
• 06:57of the disease.
• 06:59But the key question
• 07:01is, how do these function?
• 07:03And, what we have previously
• 07:05seen
• 07:06is that these non coding
• 07:07variants are highly enriched in,
• 07:10cis regulatory
• 07:11regions,
• 07:12of the genome and enhances
• 07:14and promoters. And so we
• 07:16therefore hypothesize
• 07:18that
• 07:19really the key,
• 07:22function
• 07:23genome function perturbed in Parkinson's
• 07:25disease
• 07:26might be modulation
• 07:28of RNA quantity
• 07:30based on cis regulatory,
• 07:32effects of this noncoding variant.
• 07:35How how can we treat
• 07:36this with precision drugs? Obviously,
• 07:38a lot more to do,
• 07:40but we do have some
• 07:41exciting,
• 07:42initial results.
• 07:43And most of all, we
• 07:45have what I think is
• 07:46really a cool way
• 07:48to try our best to,
• 07:50bring new drugs to patients
• 07:52as fast as possible,
• 07:54and that is,
• 07:55machine learning, big data powered,
• 07:58drug repurposing to teach new
• 08:00tricks to old drugs.
• 08:02This work is,
• 08:04performed in collaboration with the
• 08:05University of Burdon.
• 08:07The TronTrees has been a
• 08:09partner with us, first at
• 08:11Harvard and now here at
• 08:12the Stephen and Denise Adams
• 08:14Center.
• 08:15The way this works is,
• 08:18in Norway, we have,
• 08:21access on well curated databases
• 08:24for four point five million
• 08:25Norwegians over fifteen years with
• 08:28fifteen years of follow-up.
• 08:30There's, about seven hundred fifty
• 08:32million prescriptions
• 08:34given to these patients, and
• 08:36so we can now
• 08:38algorithmically
• 08:40test for associations
• 08:42between any drug approved in
• 08:43Norway
• 08:44and the
• 08:45future risk of healthy Norwegians
• 08:48of developing Parkinson's disease. And
• 08:50so you do this over
• 08:51and over for each drug,
• 08:53to identify drugs linked to
• 08:55reduce risk.
• 08:57Then we're taking these drugs
• 08:59into into clinical trials in
• 09:01a dish animal models
• 09:03and to medicinal chemistry. And
• 09:05one
• 09:06one,
• 09:07class of drugs
• 09:09that was very strong
• 09:11strongly associated with reduced risk
• 09:13are,
• 09:14asthma drugs, surprisingly.
• 09:16Those are beta two,
• 09:18adrenoreceptor
• 09:20agonists.
• 09:21And we have since shown
• 09:22that, actually,
• 09:23the longer acting they are,
• 09:25the more lipophilic and brain
• 09:27penetrant they are, the stronger
• 09:28the effect is. This has
• 09:30now been, replicated in more
• 09:32than eight countries
• 09:33and in,
• 09:35ten
• 09:36toxic and genetic models of
• 09:38Parkinson's disease. So there is
• 09:40an association
• 09:41between these asthma drugs and
• 09:43reduced risk of Parkinson's disease.
• 09:46And excitingly,
• 09:47what Monica Sharmer in,
• 09:50the Adam Center and instructor
• 09:52in the Adam Center has
• 09:53found is
• 09:55that if you use,
• 09:57stem cells of Parkinson's patients
• 09:59as as avatars in a
• 10:00test tube and look at
• 10:01the mitochondrial
• 10:03networks.
• 10:04So this is a healthy
• 10:05nit mitochondrial network with this
• 10:07nice tubular structure.
• 10:09In patients carrying the synuclein
• 10:11triplications,
• 10:13the mitochondrial network is busted
• 10:15into this,
• 10:17disjointed
• 10:21spherical forms. But treatment with
• 10:25beta two agonists
• 10:26partially restores the mitochondrial network.
• 10:29And in, you know, an
• 10:31immense body of work,
• 10:34Monica has shown that the
• 10:36beta two agonist
• 10:38actually
• 10:38affect,
• 10:40modulate
• 10:41mitochondrial respiration,
• 10:42remodel
• 10:43mitochondria
• 10:44to exactly counter,
• 10:47the effects
• 10:48conferred by uncoupled uncoupled,
• 10:50such as PM twenty d
• 10:52one. So so we think
• 10:54that these drugs actually meant
• 10:56to be excellent
• 10:58candidates for,
• 11:00as precision therapeutics
• 11:02for patients with the PM
• 11:04twenty one,
• 11:06risk variant. Here is the
• 11:08sea, seahorse,
• 11:10respirometry
• 11:11data where,
• 11:13respiration is reduced in Parkinson's
• 11:17neurons,
• 11:18be the two agonists, partially
• 11:19restore it here to help
• 11:21healthy neurons.
• 11:22And,
• 11:24with that,
• 11:25here's what we are ho
• 11:26where we are hoping to
• 11:28be in the future in
• 11:29twenty thirty four. When a
• 11:31patient comes to the clinic,
• 11:33ask the discovery engine, says
• 11:35the patient says, hi, discovery
• 11:36engine. What medication
• 11:38works for me?
• 11:40Patient inputs three drops of
• 11:42blood. The engine scans the
• 11:44genome and sucks in the
• 11:45health data
• 11:49and and spits out the
• 11:50result. Hi. Your bioscan suggests
• 11:53that gene x drives your
• 11:55disease progression.
• 11:56May I recommend the following
• 11:58precision,
• 11:59drug and precision biomarker to
• 12:01correct this? And don't forget
• 12:03to discuss this with your
• 12:05physician. So thank you for
• 12:07listening to me, and thank
• 12:08you, everybody in the lab
• 12:09and the Adam Center for
• 12:11this awesome work and to
• 12:12the ASAP team,
• 12:14who is working with us
• 12:15on that.