video1771932013

Transcript

• 01:23Yeah.
• 03:15Yeah.
• 04:19Good afternoon, everyone. I'm Jeremy
• 04:21Asens. I'm the chief of
• 04:22pediatric cardiology. You don't often
• 04:24get to see the the
• 04:25pediatric cardiologist up here, but,
• 04:28I'm I'm really happy to
• 04:29introduce Ben Landis.
• 04:31Ben is a new member
• 04:33of our section. We were
• 04:34lucky to recruit him this
• 04:35past summer,
• 04:36and he
• 04:37has expertise in aortopathy, and,
• 04:40we thought this would be
• 04:41a great way to bring
• 04:42folks together.
• 04:44So, Ben, I'm just gonna
• 04:45give you a little background.
• 04:46He did his pediatric cardiology
• 04:47fellowship at Cincinnati Children's, which
• 04:49for those of you who
• 04:50don't know is a really
• 04:52world renowned place for pediatric
• 04:54cardiology,
• 04:55cardiac surgery, and pediatric cardiac
• 04:56related research.
• 04:58While he was there in
• 04:59addition to his fellowship, he
• 05:00got a graduate, certificate in
• 05:01bioinformatics
• 05:02and did a fellowship in
• 05:04cardiovascular genetics as well. So
• 05:06he's got
• 05:07lots of letters and things
• 05:09behind his name or in
• 05:10front of his name, I
• 05:11guess.
• 05:12He joined the faculty at
• 05:14Indiana University in twenty fifteen.
• 05:17And while he was there,
• 05:18he developed a research program
• 05:19that focused on aortopathy and
• 05:21congenital heart disease.
• 05:22His lab identified a gene,
• 05:25called coq
• 05:26u eight b,
• 05:27that is now known to
• 05:28be a genetic modifier for
• 05:30aortopathy.
• 05:31He also
• 05:33established a multi institutional
• 05:35tissue and blood
• 05:36bank
• 05:38with tissue from,
• 05:39adults and children with aortopathy,
• 05:41and that bank now has
• 05:43over eleven hundred specimens.
• 05:45And he uses that repository
• 05:47to do genomic analysis and
• 05:49phenotype, genotype linkage
• 05:52studies.
• 05:54That tissue bank is in
• 05:55the process of making its
• 05:56way from Indiana to Yale,
• 05:59hopefully, not in the snowstorm.
• 06:02He, on the clinical side,
• 06:03developed and led
• 06:05a a multidisciplinary
• 06:06aortopathy clinic,
• 06:08that saw both children and
• 06:09adults with aortopathy.
• 06:11So my hope for the
• 06:13future here is to develop
• 06:14a similar model of care
• 06:16where we can have
• 06:18cross generational care where families
• 06:20would come to our center,
• 06:23both adults and children to
• 06:26receive their medical care, surgical
• 06:28care,
• 06:29and all of the
• 06:32knee and and sort of
• 06:33meet all of the needs
• 06:34that they have as families
• 06:35with arotopathy.
• 06:36And I think that would
• 06:37also serve as a really
• 06:39fertile ground for ongoing research
• 06:41and innovation in this space.
• 06:43So with all of that,
• 06:45here's Ben Landis.
• 06:53Great. Thank you very much,
• 06:54Jeremy. Thank you for the
• 06:55opportunity to talk today.
• 06:58It's a great opportunity to
• 07:00talk to a division of
• 07:01cardiovascular medicine as a pediatric
• 07:03cardiologist and,
• 07:04being new to to Yale.
• 07:08I hope, some of what
• 07:09I show, and talk about
• 07:11today,
• 07:12enhances your lunch experience. So,
• 07:16so, just a quick disclosure
• 07:18there. So,
• 07:19drastic aortic aneurysm and dissection
• 07:22is an aortopathy characterized by
• 07:24aortic dilation,
• 07:26histopathology
• 07:27that's comprised of smooth muscle
• 07:30cell abnormalities,
• 07:32accumulation
• 07:33of nucoid extracellular matrix, and
• 07:35degradation and disarray of the
• 07:37elastic fibers.
• 07:39Thoracic aortic aneurysm is typically
• 07:41asymptomatic,
• 07:43but poses a deadly risk
• 07:45of a thoracic aortic dissection
• 07:47in which there's a separation
• 07:48between the insimal
• 07:50and medial layers of the
• 07:51aorta can lead to,
• 07:54death, major complications,
• 07:56and including an aortic rupture.
• 07:59So, looking broadly at thoracic
• 08:01aortic aneurysm,
• 08:03you can define them as
• 08:04a heritable,
• 08:05bicuspid aortic valve associated
• 08:07sporadic, and we do see
• 08:09aortic dilation in the context
• 08:10of complex complex heart defects
• 08:12as well.
• 08:14So, this, list of genes
• 08:16is is taken from a
• 08:18a next generation sequencing panel
• 08:19that we would send, typically
• 08:21for patients who have thoracic
• 08:22aortic aneurysm,
• 08:24consisting of thirty five genes.
• 08:25And it, the list of
• 08:27genes gives some insight into
• 08:28the path pathophysiology
• 08:30that underlies the disease.
• 08:32And you can see here
• 08:33it's, includes genes important for
• 08:34the extracellular matrix
• 08:36such as FBN one and
• 08:37Markman syndrome associated,
• 08:40genes important for TGF beta
• 08:41signaling.
• 08:43Many of these patients will
• 08:44present with a syndrome of
• 08:46Loewe's Dietz syndrome,
• 08:48genes important for smooth muscle
• 08:50contraction,
• 08:51and then a a hodgepodge
• 08:52of other less common,
• 08:54genes.
• 08:55There's x linked associations with
• 08:57a, thoracic aortic aneurysm,
• 08:59and then as well as
• 09:00well, this panel will include
• 09:02a couple,
• 09:03conditions that are, autosomal recessive.
• 09:06Many of these genes are
• 09:08associated with, extra cardiac syndromic
• 09:10features,
• 09:11but some do not, such
• 09:12as the smooth muscle cell
• 09:14contractile genes typically will present
• 09:15with,
• 09:16minimal, if any,
• 09:18extra cardiac features.
• 09:20In addition to the single
• 09:21gene associations,
• 09:24there are some,
• 09:25abnormalities in copy number associated
• 09:28with disease, including Turner syndrome,
• 09:30monosomy x, seven q one
• 09:32one point two three duplication,
• 09:33which involves the gene elastin,
• 09:36as well as this duplication.
• 09:38So just to kind of
• 09:39give a picture for the
• 09:41a couple of conditions we're
• 09:42talking about today, Marfan syndrome,
• 09:44in addition to the aortopathy,
• 09:46has associated ocular findings,
• 09:48skeletal findings, and cutaneous findings.
• 09:52Loeys Dietz syndrome,
• 09:53has overlapping phenotypic features with
• 09:56Marfan syndrome. Again,
• 09:58commonly associated with,
• 10:00changes in TGF beta genes.
• 10:02And some of these are
• 10:02skeletal.
• 10:03Hypertilarism
• 10:04may be a distinctive
• 10:06feature compared to Marfan syndrome.
• 10:08Bifid uvula certainly is,
• 10:10in about fifty percent of
• 10:12patients, as well as a
• 10:13more extensive and and diffuse,
• 10:16arterial involvement that often,
• 10:19includes,
• 10:20arterial tortuosity
• 10:22and risk for complications of,
• 10:25our, the distillate or as
• 10:27well as,
• 10:28branches, arterial branches.
• 10:32Briefly, bicuspid aortic valve associated
• 10:34aortopathy,
• 10:36Really, we have very little
• 10:37understanding of the genetic basis
• 10:39of bicuspid aortic valve, as
• 10:41well as the aortopathy. There's
• 10:42a few single gene Mendelian
• 10:44causes that have been identified,
• 10:46but these wouldn't
• 10:48would not be routinely tested
• 10:49except for notch one in
• 10:50the context of a clinical
• 10:52evaluation for TAA.
• 10:55In thinking about, the importance
• 10:57of a genetic diagnosis,
• 11:00when it comes to aortic
• 11:01risk.
• 11:02The twenty twenty two guidelines
• 11:04from the,
• 11:06is a excellent review that,
• 11:09also a set of set
• 11:11of recommendations
• 11:12that that gave good,
• 11:14appreciation for the the risks
• 11:16associated with particular genetic abnormalities.
• 11:18So,
• 11:19that includes their guidelines for
• 11:21the thresholds for,
• 11:23performing a aortic a prophylactic
• 11:25aortic surgery,
• 11:27on the proximal aorta.
• 11:29And as you can see,
• 11:29there's,
• 11:31the thresholds will, be lower
• 11:33in patients who have Marfan
• 11:35syndrome,
• 11:36as well as those who
• 11:37have high risk features, even
• 11:38lower. Lowy's Dietz syndrome, because
• 11:41of the,
• 11:42data would suggest that many
• 11:44patients have a higher risk,
• 11:45and therefore, the thresholds are
• 11:47lower in, the majority of
• 11:49genes associated with Lowy's Dietz
• 11:50syndrome. And then smooth muscle
• 11:52contractile genes also will have
• 11:53a lower threshold.
• 11:55And when there's a heritable
• 11:57association in the family, but
• 11:59a genetic cause isn't identified,
• 12:00there's also,
• 12:01consideration for surgery at a
• 12:03lower threshold.
• 12:05You know, as as mentioned,
• 12:06the the guidelines are extensive
• 12:08here, and, the high risk
• 12:10features generally will include things
• 12:11like rapid aortic dilation, family
• 12:13history of dissection,
• 12:15morph morphology of the proximal
• 12:17aorta, whether it's root and
• 12:18ascending,
• 12:19involvement, for example,
• 12:21if there's arterial tortuosity,
• 12:23as well as some some,
• 12:25attention to non cardiovascular abnormalities
• 12:27in the context of Loeys
• 12:28Dietz syndrome, potentially indicating more
• 12:30severe phenotype.
• 12:33So I was gonna go
• 12:34through, some recent studies that
• 12:36highlight,
• 12:37some associations
• 12:38between genes and risk.
• 12:41And so on the left
• 12:42here,
• 12:43is is a a plot
• 12:45that's, from the from a
• 12:46university,
• 12:48in Osaka in which they
• 12:49looked at five hundred eighteen
• 12:50patients. And the these studies
• 12:52have group genes based on
• 12:54on classes, essentially, especially when
• 12:56it comes to Loewe's Dietz
• 12:57syndrome.
• 12:57And I think what what
• 12:58we can appreciate here is
• 13:00that in patients who have,
• 13:02changes in genes
• 13:04that are involving the TGF
• 13:06beta signaling pathway
• 13:07or in the smooth muscle
• 13:09contractile
• 13:10tended to have a higher,
• 13:11an earlier onset of aortic
• 13:13events, including dissections need for
• 13:15aortic surgery.
• 13:18Kind of corroborating that data
• 13:20would be a larger study,
• 13:21international study, the month from
• 13:23the Montalcino aortic consortium
• 13:25in which they've looked at,
• 13:27patients who have vascular EDS,
• 13:30patients with TGF beta signaling,
• 13:32gene abnormalities,
• 13:33and Marfan syndrome. And you
• 13:35can pay attention. They looked
• 13:36at arterial complications as well
• 13:37as aortic
• 13:39complications. So you can pay
• 13:40attention to the dash lines
• 13:41here with the aortic. And
• 13:42again, what we're what we
• 13:44can appreciate here,
• 13:46that, you know, we started
• 13:47to appreciate here as well
• 13:48is at some point, really,
• 13:50the risk for
• 13:51a dissection
• 13:52starts to become fairly
• 13:54similar between Marfan syndrome and
• 13:56Loeys Dietz syndrome despite there
• 13:58being potentially an earlier,
• 14:00risk.
• 14:01And so,
• 14:02you know, I think that
• 14:03also we can look at
• 14:04these things, these risks, and
• 14:06it's been looked at likewise
• 14:07with the Montalcino,
• 14:08aortic consortium by looking at
• 14:10specific genes. And here we're
• 14:12seeing aortic events occurring in
• 14:14a similar,
• 14:15age dependency between smooth muscle
• 14:17contractile genes and TGF beta
• 14:19genes. Then when you look
• 14:20at the the specific genes
• 14:21within the TGF beta signaling
• 14:23pathway, what's starting to emerge
• 14:24is, patients who have mutations
• 14:26in TGF beta receptor one
• 14:28or TGF beta receptor two
• 14:30tend to have earlier complications,
• 14:31aortic events, than those with
• 14:33the other genes.
• 14:36Likewise, with smooth muscle cell
• 14:37genes,
• 14:38ACTA two seems to be,
• 14:41sorry, PRKG one seems to
• 14:43be particularly,
• 14:44prone to an early complication,
• 14:47and then followed by ACTA
• 14:48two changes and then MYLK,
• 14:51myosin like chain kinase. So
• 14:53these are giving some level
• 14:54of,
• 14:55insights into how we could
• 14:56stratify a patient's risk based
• 14:58on genes.
• 15:00So,
• 15:01you know, with the rationale
• 15:02there, as well as other,
• 15:05pieces of rationale for genetic
• 15:06testing, of course, I took
• 15:08a look at the literature
• 15:09in terms of what's the
• 15:10yield when patients are coming
• 15:11in for testing with the
• 15:12next generation sequencing panel.
• 15:14And, you know, I think
• 15:15that over the course of
• 15:16time, there's different selection criteria
• 15:18in these studies and they're
• 15:19retrospective.
• 15:21But overall,
• 15:22you can see that the
• 15:23the likelihood of identifying a
• 15:25pathogenic likely pathogenic variant in
• 15:27in in these genes, you
• 15:29know, ranging from twenty to
• 15:30thirty six genes depending on
• 15:31the panel,
• 15:32could be four percent, but
• 15:33upward of eighteen percent.
• 15:35And I wanted to highlight
• 15:36here too that there's commonly
• 15:38variants of uncertain
• 15:40significance identified. And that's a
• 15:41real challenge when it comes
• 15:43to the management,
• 15:44and something that, you know,
• 15:46warrants further studies in terms
• 15:47of, developing novel ways to
• 15:50functionally interpret variants, for example.
• 15:53And this further shows so
• 15:54this is from the,
• 15:55an aorta clinic in in
• 15:57Canada, and they looked at
• 15:58two hundred fifty patients. And
• 15:59you can see that the
• 16:00variants of uncertain significance,
• 16:03which you typically wouldn't be
• 16:04clinically actionable, but you have
• 16:06to, consider some of them
• 16:08as potentially disease contributing.
• 16:11You can see that some
• 16:12of these are in genes
• 16:12that are have high significant
• 16:14importance at least, FBM one,
• 16:16you know, TGF beta two,
• 16:18TGF beta r one. So,
• 16:20you know, there's a real
• 16:20need for triaging or classification
• 16:23of variants of uncertain significance.
• 16:26The genetic complexity of of
• 16:28aortopathy also is highlighted by
• 16:30Marfan syndrome,
• 16:31so high locus heterogeneity,
• 16:34in this condition. So,
• 16:35this is data I extracted
• 16:37from ClinVar,
• 16:39this month. And what we're
• 16:40highlighting here is these are
• 16:41all
• 16:42variants that were reported in
• 16:44the ClinVar database, including pathogenic,
• 16:46likely pathogenic. And so when
• 16:48we look at these likely
• 16:49pathogenic, pathogenic variants, you can
• 16:50see that
• 16:52three thousand six hundred fifteen
• 16:53different variants have been associated
• 16:55with Marfan syndrome in FBN
• 16:57one. And there's a range
• 16:58of types of mutation there
• 16:59when it comes to deletion
• 17:01duplications
• 17:01as well as frame shift,
• 17:03missense changes, nonsense, and splice
• 17:05sites.
• 17:07Likewise,
• 17:08we see in the databases
• 17:10lots of variance of uncertain
• 17:11significance in FBN one. FBN
• 17:13one sixty five exon, so
• 17:14a large gene.
• 17:16But even in a condition
• 17:17like Marfan syndrome where there's
• 17:18lots of experience, there's still
• 17:20tons of uncertainty when it
• 17:21comes to variant interpretation,
• 17:24and its contribution to to
• 17:26disease.
• 17:28So the
• 17:29going a little bit deeper
• 17:30into trying to understand genetic
• 17:32classification
• 17:33and risk,
• 17:34this study looked at the,
• 17:36cumulative
• 17:36risk when it came to
• 17:37the types of FBN one
• 17:39variants in in patients who
• 17:40have Marfan syndrome.
• 17:42So, the thinking is that,
• 17:44changes in FBN one, can
• 17:45have a dominant negative effect,
• 17:47or be a haploinsufficiency
• 17:49mechanism. And then these these
• 17:51people as well identified a
• 17:52certain regions,
• 17:53in the gene where cysteine
• 17:55residues could be affected and
• 17:56may have had a more
• 17:58severe phenotype. And I think
• 17:59this data corresponds with other
• 18:01studies as well in which,
• 18:03in general, patients who have
• 18:04a mutation leading to haploinsufficiency
• 18:06have a higher risk for
• 18:08a complication
• 18:09compared to those who have,
• 18:11what's presumed to be a
• 18:13dominant negative effect based on
• 18:15it being, for instance, a
• 18:16missense change.
• 18:19Likewise, we've started to be
• 18:20able to stratify bay in
• 18:22other genes based on variant
• 18:23type. And so, in TGFBR
• 18:25two patients, you can see
• 18:26that, an arginine five twenty
• 18:28eight had a really high
• 18:29risk in the Montechino
• 18:32for early complications.
• 18:34Whereas, SMAD three, when you
• 18:36look at the different types
• 18:37of changes that were reported
• 18:38in that data set, we
• 18:40really don't see a clear
• 18:41stratification,
• 18:42with risk.
• 18:44When it came to smooth
• 18:45muscle contraction genes,
• 18:47this, variant, affecting residue one
• 18:49seventy nine,
• 18:51in ACTA two seems to
• 18:52be particularly
• 18:53predisposing to complications.
• 18:55And and you can also
• 18:56start to identify others that
• 18:58that may also confer an
• 18:59increased risk.
• 19:01And then, they also looked
• 19:02at myosin light chain kinase
• 19:03variants, and and interestingly, missense
• 19:05variants tended to be,
• 19:07a a higher risk than
• 19:09those that were predicted to
• 19:10be protein truncated leading to
• 19:12nonsense media decay.
• 19:14So there are efforts out
• 19:15there to try to stratify
• 19:17risk based on gene, gene
• 19:19class, and variants.
• 19:20I I just wanted to
• 19:21highlight though that, you know,
• 19:23as you can see, people
• 19:23who are living to age
• 19:24fifty, for for instance, that
• 19:26have a,
• 19:27dominant negative predicted dominant negative
• 19:30variant. You can see it's
• 19:31not it's around fifty percent
• 19:33of individuals are having complications.
• 19:35And I think this highlights
• 19:36the clear variability in the
• 19:38severity of disease even, among
• 19:41patients who have the same,
• 19:42for instance, gene or even
• 19:43variants,
• 19:44change.
• 19:46So I mentioned twenty percent
• 19:47heritable,
• 19:48genes.
• 19:49Twenty percent of the disease
• 19:50can be associated with heritable,
• 19:52conditions. There's also been recent
• 19:55data to try to understand
• 19:57aortic dilation and aneurysm
• 19:59in the sense of a
• 20:01complex disease.
• 20:02And so GWAS studies have
• 20:03been conducted
• 20:05using UK Biobank data
• 20:08and associating that with ascending
• 20:10aorta diameter values on MRIs
• 20:12and, you know, eighty two
• 20:13GWAS loci were were identified.
• 20:16Likewise,
• 20:17a large study of eight
• 20:18thousand TA dissection cases compared
• 20:21to four hundred fifty thousand
• 20:22non
• 20:23thoracic aortic aneurysm dissection cases,
• 20:26in the million veterans program
• 20:27identified twenty one,
• 20:30loci that were associated
• 20:32with disease.
• 20:33So trying to put together
• 20:34maybe a polygenic
• 20:35contribution
• 20:36to disease,
• 20:37either development or or progression.
• 20:40And so as doctor Aznes,
• 20:42alluded to, you know, my
• 20:44research has has tried to
• 20:46utilize human samples in order
• 20:48to ask questions that are
• 20:49clinically relevant.
• 20:50In order to to pursue
• 20:52this, we developed this, study
• 20:54in which, we enroll participants,
• 20:56collect comprehensive data.
• 20:58When the needing an aortic
• 21:00surgery, we collected,
• 21:01aortic tissue and processed in
• 21:03many ways, including,
• 21:05specimens,
• 21:07processed for histology,
• 21:08electron micro
• 21:09microscopy,
• 21:10flash freezing,
• 21:11and we also cultured primary
• 21:13smooth muscle cells directly from
• 21:15the aorta using an explant
• 21:16outgrowth method and then extracted
• 21:18RNA and protein at early
• 21:19passage.
• 21:20In addition to that, all
• 21:21participants,
• 21:23would provide a blood sample,
• 21:24and we have processed those
• 21:25broadly as well for transcriptome,
• 21:29you
• 21:30DNA extraction, plasma studies, as
• 21:32well as, frozen aliquots of
• 21:34whole blood.
• 21:36So, success was was, kinda
• 21:38indicated by the large number
• 21:40of patients. We have enrolled
• 21:41over fourteen hundred,
• 21:43collected aortic tissue samples from
• 21:44greater than four hundred individuals.
• 21:47This includes cases and controls,
• 21:49undergoing a heart transplant
• 21:51or or, organ donors. And
• 21:53then we cultured smooth muscle
• 21:55cells from greater from over
• 21:56a hundred, individuals.
• 21:58And so, I mentioned before
• 22:00that, the,
• 22:02effect of an FBN one
• 22:03variant may have clinical significance
• 22:06on course,
• 22:08in patients who have Marfan
• 22:09syndrome. And so we conducted
• 22:11a study in which we,
• 22:13try to utilize the patient's
• 22:14own samples in order to
• 22:15understand the transcriptional effects of
• 22:18FBN one variance. And so,
• 22:19we we studied in this
• 22:21here, twelve patients, five with
• 22:23Marfan syndrome and seven controls,
• 22:25collected a blood sample for
• 22:27whole genome sequencing,
• 22:29and then we cultured the
• 22:30smooth muscle cells and did
• 22:31mRNA sequencing,
• 22:33at greater than, typical depths.
• 22:36We we attempted, and then
• 22:37we sought to integrate
• 22:39understand the transcriptional effects of
• 22:41the variance.
• 22:42You can see that relatively
• 22:44young patients with Marfan syndrome
• 22:45and the and the controls
• 22:46were reasonably,
• 22:48matched to age as well.
• 22:49And so, one of the
• 22:50things we did, we first
• 22:51had mRNA seek, data. And
• 22:53so we asked the question
• 22:54of whether we could identify
• 22:56pathogenic variants directly from sequencing
• 22:58of the mRNA seek reads.
• 22:59And indeed indeed, we did,
• 23:01and then we confirmed these
• 23:02with genome sequencing.
• 23:04And we wanted to use
• 23:05that data also to understand
• 23:07what is the functional effect
• 23:08on the transcript.
• 23:10So, what we observed was
• 23:11that in the patients who
• 23:12had non synonymous,
• 23:13single nucleotide variants that the
• 23:15fraction of the alternative allele
• 23:17reads in the mRNA seek
• 23:18data was similar to to
• 23:19the reference,
• 23:21whereas,
• 23:22in a patient who had
• 23:23a stop gain variant, we
• 23:24saw a decrease,
• 23:25in the fraction of reads,
• 23:27with the alternative allele in
• 23:28that patient's smooth muscle cells,
• 23:31indicating likely non sense media
• 23:32decay.
• 23:34Amongst the data was also
• 23:35we identified a, deletion in
• 23:37exon forty seven in one
• 23:38patient,
• 23:40and in in trying to
• 23:40understand what was the, effect
• 23:42of allelic transcription on this
• 23:44individual. You can see that
• 23:45the number
• 23:46of reads that overlapped the
• 23:48normal exon exon junctions,
• 23:50was relatively similar to the
• 23:52number of reads that aligned
• 23:53over the abnormal,
• 23:55exon exon junctions,
• 23:56again, suggesting that this, variant
• 23:59did not lead to
• 24:00significant,
• 24:01pretranslational
• 24:02transcriptional abnormality.
• 24:04We further explored allelic expression
• 24:07in in these samples.
• 24:08Here, we've plotted across all
• 24:10all samples,
• 24:12the single the single nucleotide
• 24:14variants that were identified. And
• 24:15we're graphing here the fraction
• 24:16of reads with the alternative
• 24:18allele, and then we've labeled,
• 24:19according to samples. So some
• 24:21patients would have multiple snips,
• 24:23in this gene, and then
• 24:23we can look at what
• 24:24the proportion of reads are.
• 24:26And you can see that
• 24:26for the patient who had
• 24:27the nonsense,
• 24:29who had the, nonsense variant,
• 24:31we,
• 24:32observe that in all for
• 24:33all SNPs, a skewing of
• 24:35the, of the ratio,
• 24:37of the of the fraction
• 24:38of reads with the alternative
• 24:39allele, again, adding additional support
• 24:41to the,
• 24:42likelihood of nonsense mediated decay.
• 24:44In addition to that, I
• 24:45would suggest that we're detecting
• 24:47the,
• 24:49what is a truncated
• 24:52allele
• 24:53transcript.
• 24:54And, you know, I think
• 24:55we're suggesting that, you know,
• 24:56this isn't necessarily
• 24:58rapidly degraded and maybe could
• 25:00have a combination of a
• 25:01haploinsufficiency
• 25:02as well as potentially dominant
• 25:04negative effects if that transcript
• 25:05goes on to translation, for
• 25:06example.
• 25:07We looked at the gene
• 25:08expression level overall. And, again,
• 25:10with the patient who had
• 25:10nonsense median decay, we saw
• 25:12a low level of FBN
• 25:13one relative to controls in
• 25:15other in the majority of
• 25:16other cases.
• 25:17Again,
• 25:18indicating that, you know, in
• 25:19this patient, there wasn't, inadequate,
• 25:22for example, compensatory
• 25:23increase in expression of the
• 25:25reference allele.
• 25:27And and and so we've
• 25:28kind of more completely characterized
• 25:30this patient's and others' transcriptional
• 25:32effects using this. And so,
• 25:34you know, this is kind
• 25:35of a test case example.
• 25:36I see an opportunity for
• 25:38us to utilize these types
• 25:39of techniques in order to
• 25:41improve our clinical diagnostic pipelines,
• 25:44when cells
• 25:46and and, and DNA
• 25:48is available in order to
• 25:49to look at FBN one
• 25:51genes as well as improve
• 25:52our, interpretation of variants and
• 25:54other, single gene causes of
• 25:56aortopathies.
• 25:58In these data, we did
• 25:59a differential expression
• 26:00analysis of Marfan syndrome compared
• 26:01with controls.
• 26:02We saw an enrichment of
• 26:03genes important for glycerophospholipid
• 26:06metabolism,
• 26:07and, and and and, specifically,
• 26:10genes that are important for
• 26:11the
• 26:15generation and processing of of
• 26:17of lysophosphatidic
• 26:18acid, a fatty acid.
• 26:20And what we observed here
• 26:22is a pattern in which
• 26:24the genes that lead to
• 26:26LPA,
• 26:27so lysophosphatidic
• 26:28acid production
• 26:29were decreased, and those that
• 26:30converted lysophosphatidic
• 26:32acid to phosphatidic
• 26:33acid was increased. So this
• 26:35is some preliminary data suggesting
• 26:37potentially dysregulation of this pathway,
• 26:39specifically in smooth muscle cells
• 26:40and Marfan syndrome.
• 26:43As further exploration in these
• 26:45data, we did single cell
• 26:47gene expression profiling of the
• 26:49cells in culture. So we
• 26:51have always presumed and many
• 26:52have presumed that,
• 26:54the cells that are cultured
• 26:55as in an x plane
• 26:56outgrowth method are smooth muscle
• 26:58cells. So we did single
• 26:59cell profiling and labeling with
• 27:01single r,
• 27:03a computational program
• 27:04confirmed that these these cells
• 27:06do have the characteristics of
• 27:08smooth muscle cells in in
• 27:09four different samples.
• 27:11The pseudo bulk data from
• 27:12the single cell correlated directly
• 27:14with the mRNA seek data,
• 27:16for the gene expression profiling,
• 27:18validating this, single cell, fixed
• 27:21RNA profiling approach to the
• 27:22cultured cells.
• 27:25And then we further delved
• 27:26into the single cell data,
• 27:28thinking about how we may
• 27:29be able to use,
• 27:31cluster analysis,
• 27:33in order to subcategorize
• 27:34the expression states of smooth
• 27:36muscle cells in culture,
• 27:38knowing that there is likely
• 27:39to be heterogeneity.
• 27:41And then being able to
• 27:42look at subpopulations
• 27:43and perform differential expression analysis.
• 27:45And you can see we
• 27:46identified,
• 27:47based on canonical markers,
• 27:49a variety of subtypes of
• 27:50smooth muscle cells,
• 27:53similar proportions between Marfan syndrome
• 27:55controls.
• 27:56We identified a gene called
• 27:57TRPD two,
• 27:59transient receptor potential
• 28:01lineloid
• 28:01two, that was increased in
• 28:03Marfan syndrome compared with controls
• 28:05in this,
• 28:06in the in these, in
• 28:08these, single cell,
• 28:10data
• 28:11and and specifically highest in
• 28:12the genes that were characterized
• 28:14as as contractile.
• 28:16This is a feature plot
• 28:17showing that in general, higher
• 28:18levels of t r p
• 28:19v two expression.
• 28:20We then went to the
• 28:22tissue, the primary tissue from
• 28:23which these cells were cultured
• 28:25and and and observed increased
• 28:26t TRP v two expression
• 28:28in in in the tissues.
• 28:30We've done a single cell
• 28:31transcriptome analysis
• 28:33of a larger cohort of
• 28:34ten cases.
• 28:36This is primary,
• 28:37frozen tissues and five controls
• 28:39and also and these data
• 28:40showed increased
• 28:42expression of TRPV2.
• 28:44And so what is this
• 28:44gene? It's a mechanosensitive
• 28:47calcium permeable
• 28:48channel. Looking in literature about
• 28:50this gene, TRPV1
• 28:51is increased
• 28:53in tissue of Marfan patients
• 28:55in the insomel layer.
• 28:57In a prior report, this
• 28:59gene seems to be important
• 29:00in rats for aortic tone.
• 29:02And then also,
• 29:03this gene is regulated
• 29:05or
• 29:06altered by
• 29:09activation
• 29:10of the lysophosphatidic
• 29:11acid receptor one by lysophosphatidic
• 29:14acid. So potentially
• 29:15observing some connections there. And
• 29:17so I wanted to make
• 29:18a a point about, you
• 29:20know, one aspect of the
• 29:21pathophysiology
• 29:23of aortic aneurysms,
• 29:25and that is oxidative stress.
• 29:26So, there's
• 29:28abundant
• 29:29data and studies in in
• 29:30mouse models as well as
• 29:31in human tissues to indicate
• 29:33that oxidative stress may be
• 29:36a downstream
• 29:37mediator of the pathogenesis or
• 29:38at least involved in the
• 29:39pathogenesis
• 29:41of human
• 29:42and and and animal model
• 29:43TAA.
• 29:44We added to that literature
• 29:45with the largest
• 29:47collection of fixed,
• 29:49tissues,
• 29:50in which we stained for
• 29:51a marker of oxidative stress,
• 29:53nitrotyrosine,
• 29:54and blinded grading of the
• 29:56intensity
• 29:57observed an increase in TAA
• 29:59samples compared with controls.
• 30:01We also used our samples,
• 30:03to look in smooth muscle
• 30:04cells in situ, using electron
• 30:06microscopy and characterize the mitochondria
• 30:09using a semi quantitative score,
• 30:11again, blinded analysis.
• 30:13And we have, for the,
• 30:15ultra structural defects in the
• 30:17mitochondrial cristae. And, again, we
• 30:18saw higher defect scores in
• 30:20TAA for the majority, six
• 30:21out of the seven, cases
• 30:23compared compared with the controls.
• 30:25And most recently, we've looked
• 30:27at
• 30:28a series of cases, who
• 30:29had TAA,
• 30:30did on targeted metabolomics,
• 30:32and
• 30:33and compared those two controls.
• 30:35These are relatively young patients,
• 30:36average age in the thirties,
• 30:38and we're observing in this
• 30:40amongst these data, we we
• 30:42saw,
• 30:43a decrease in the ratio
• 30:44of reduced glutathione to to
• 30:45oxidized glutathione disulfide.
• 30:47Again, another marker of oxidative
• 30:49stress. So I think trying
• 30:50to paint a picture here
• 30:51and and,
• 30:53and and thinking about oxidative
• 30:54stress as a prevalent,
• 30:57aspect of the pathophysiology,
• 30:59including across, different etiologies, which,
• 31:03different etiologies, which, these these,
• 31:04these samples consisted of. And
• 31:07so,
• 31:08you know, as I pointed
• 31:10to in the,
• 31:11Kaplan Meier type curves that
• 31:13we observe with genes,
• 31:14and variance in gene types,
• 31:16you know, there is substantial
• 31:18inter individual variability
• 31:20in the progression and the
• 31:21outcomes in patients.
• 31:23We looked at in the
• 31:24young. So so in children,
• 31:25we echocardiography
• 31:26is the standard way we
• 31:27monitor them, and we calculate
• 31:29a z score to index
• 31:30their
• 31:31diameter to body size in
• 31:33order to grade whether there's
• 31:34dilation
• 31:35or
• 31:36and present, and if present,
• 31:37how severe.
• 31:39And we looked at retrospectively
• 31:40at a large group of
• 31:42relatively large group of patients,
• 31:44who were followed for at
• 31:44least five years, and the
• 31:46average follow-up
• 31:47was was ten years. And,
• 31:48you know, I think that
• 31:49this is kind of kind
• 31:50of,
• 31:51a lot of information, right
• 31:53here. But I think what
• 31:53you can see is, a
• 31:55rate of change in z
• 31:56score from baseline to to
• 31:57last
• 31:58of of zero would mean,
• 31:59you know, no no evidence
• 32:01for progression.
• 32:02There's some who improved and
• 32:03some who,
• 32:04who who progressed over time.
• 32:06And and that was highly
• 32:07variable within groups, including, you
• 32:09know, within Marfan, within bicuspid
• 32:10aortic valve patients who
• 32:15variability in pediatric progression of
• 32:16disease.
• 32:18In addition, there's pedigrees,
• 32:20that clearly highlight the intrafamilial
• 32:22variability. So patients who, have
• 32:24the same family members who
• 32:25have the same, pathogenic
• 32:27mutation, quite substantial differences in
• 32:30their outcomes. You know, this
• 32:31is a nice study that
• 32:32they looked at a a
• 32:33a family who had TGF
• 32:34beta r two. And you
• 32:35can see that, you know,
• 32:36there were complications in many,
• 32:38but some lived, you know,
• 32:39to later ages, without aortic
• 32:42events. And so the
• 32:43reasons that underlie this variation
• 32:46is unclear. And so all
• 32:47of this put together when
• 32:48I'm seeing patients is we
• 32:49were thinking about are we
• 32:50going to start a patient
• 32:51on beta blocker or AGTense
• 32:52receptor blocker,
• 32:53which is all we have
• 32:55right now. Are we going
• 32:56to recommend some activity restrictions
• 32:58to prevent their progressive dilation?
• 32:59How frequently are we gonna
• 33:00see them? And so you
• 33:01think about this dial. You
• 33:02know, we have a genetic
• 33:03diagnosis. There's some data that
• 33:04we can, you know, population
• 33:06wide,
• 33:07make some,
• 33:08make some decisions,
• 33:09have some rationale for our
• 33:10decision. But largely, it's a
• 33:12very much a
• 33:13kind of standard approach,
• 33:15that's not individualized.
• 33:16And you can see this
• 33:17gets worse as we're looking
• 33:19at the patients who don't
• 33:19have a genetic diagnosis with
• 33:21no real way to stratify
• 33:22their risk.
• 33:23So we thought that maybe
• 33:24genetic modifiers could contribute to
• 33:26the
• 33:27severity of disease.
• 33:29We did exome sequencing in
• 33:30three different families, first degree
• 33:32relatives who had divergence in
• 33:33their TA severity. And we
• 33:35looked at what variants were
• 33:36different, rare coding variants were
• 33:37different between the patients who
• 33:38had mild phenotype versus severe
• 33:40phenotype when it came to
• 33:41their aorta.
• 33:42We crossed all these variants,
• 33:44across the different pedigrees and
• 33:45found this gene CoQAB,
• 33:47this particular variant,
• 33:49that was present segregating with,
• 33:50with disease severity in all
• 33:52three
• 33:53families.
• 33:54What do we know about
• 33:55CoQAB?
• 33:56It's associated with an autosomal
• 33:57recessive
• 33:58kidney disorder,
• 34:00but it's nuclear encoded, translates
• 34:01to mitochondria. And there, it's
• 34:03important for the synthesis of
• 34:04coenzyme Q10.
• 34:06And, of course, coenzyme q
• 34:07ten as a head group
• 34:08and, you know, a isoprenoid,
• 34:10tail.
• 34:11It's important for, mitochondrial electron
• 34:13transport as well as, acts
• 34:15as a lipophilic antioxidant.
• 34:17So, you know, potentially,
• 34:19this gene could be acting
• 34:20in a mechanism of of
• 34:21oxidative stress and and,
• 34:23abnormalities in mitochondrial,
• 34:25function, for example. So, we
• 34:27looked in our smooth muscle
• 34:28cells and identified that, expression
• 34:30of CoQAB was decreased in
• 34:32the small series of patients,
• 34:33in smooth muscle cells. We've
• 34:35later, gone on to
• 34:37observe this in a much
• 34:38larger number of samples using
• 34:39mRNA seek that I'll show
• 34:40in a little bit.
• 34:42And then we did some
• 34:43experiments in which we,
• 34:45knocked down CoQAB expression.
• 34:48You can see that CoQAB
• 34:49localized to mitochondria and that
• 34:50we effectively knocked it down
• 34:52with our siRNA. And we
• 34:53observed, functional changes in the
• 34:55smooth muscle cells, including decreased
• 34:57aerobic respiration,
• 34:58increased
• 34:59oxidative stress,
• 35:00including lipid peroxidation,
• 35:02protein carbonylation,
• 35:04and changes in the expression
• 35:06of genes important for smooth
• 35:07muscle cell function, contractile genes.
• 35:09And then more recently, we've
• 35:11done some experiments where we've
• 35:12knocked down and did mRNA
• 35:13sequencing as well as in
• 35:14the context of additional stressors
• 35:16in order to more completely
• 35:18characterize what's the effect of
• 35:19loss of coQA b and
• 35:20aortic smooth muscle cells.
• 35:22During this time, I we
• 35:23saw a study, that was
• 35:25done in yeast,
• 35:26in in which they expressed
• 35:27a missense snips and constructs.
• 35:29And it was a surprising
• 35:30finding,
• 35:31that this particular snip,
• 35:33which is common,
• 35:35was had an association. So
• 35:37specifically, the
• 35:38the the variant that leads
• 35:39to the histidine,
• 35:40residue here was associated with
• 35:42decreased activation activity of the,
• 35:45coQ levels basically. So a
• 35:46complex two three assay measures
• 35:48decreased
• 35:49mitochondrial protein levels and decreased
• 35:51aerobic respiration in yeast. So,
• 35:53potentially a functional common SNP.
• 35:55And so, given our interest
• 35:57in CoQAB,
• 35:58I looked at forty eight
• 35:59patients who had longitudinal aortic
• 36:01follow-up and tested using a
• 36:02mixed model association for
• 36:05this SNP and the rate
• 36:06of aortic dilation
• 36:08and identified that the genotype
• 36:10of g
• 36:11compared to a was associated
• 36:13with a lower
• 36:14rate of aortic dilation.
• 36:19And then also, we looked
• 36:20at a second cohort, a
• 36:22cohort of patients who had
• 36:23early onset aortic dissection and
• 36:25saw the same pattern in
• 36:26which patients,
• 36:27who have the AA,
• 36:30genotype have more significant disease,
• 36:32day aortic dissection compared to
• 36:34those with the g,
• 36:36homozygous g,
• 36:37genotype.
• 36:39From a functional standpoint, I
• 36:40showed some functional data in
• 36:41yeast. We extracted
• 36:44protein from aortic smooth muscle
• 36:45cells at early passage. We
• 36:47genotype patients for this SNP,
• 36:49and then we measured CoQAB
• 36:50levels using a western blot.
• 36:52And we observed the same
• 36:53pattern in which the the
• 36:54patients who were homozygous,
• 36:56for the
• 36:58for this, allele that appears
• 37:00to have a protective effect,
• 37:02we saw higher levels of,
• 37:04CoQAB
• 37:04protein,
• 37:05in those cells.
• 37:07And then we confirmed this
• 37:08again. We did an additional
• 37:10six patients and put it
• 37:11put it all together to
• 37:12show that this gene,
• 37:14this variant, that was associated
• 37:16with
• 37:18less severe disease has higher
• 37:20levels of CoQAB.
• 37:22And so,
• 37:24thinking about how that may
• 37:25have a role in the
• 37:26in in the mechanism oxidative
• 37:27stress and disease.
• 37:29So, this is one data
• 37:30set. We're currently doing a
• 37:31study in which we're enrolling
• 37:33three hundred patients and doing
• 37:34whole genome sequencing.
• 37:35It's a multicenter study and
• 37:36we're going to be investigating
• 37:37whether the SNP is replicated
• 37:40for association
• 37:41with rate of dilation as
• 37:42well as looking at other
• 37:43candidate snips such such as
• 37:44some of those
• 37:45genes that were identified in
• 37:46GWAS.
• 37:48And so,
• 37:50transitioning a little bit here.
• 37:51So, we've recently done a
• 37:52study in patients who have
• 37:53Marfan syndrome in Loewe's Dietzen,
• 37:55which we've taken a frozen
• 37:56piece of tissue, split it,
• 37:58and done single cell transcriptome
• 37:59analysis using a fixed RNA
• 38:00profiling
• 38:01assay, and then also in
• 38:03parallel done, untargeted metabolomics,
• 38:06in those tissues.
• 38:07And this is kind of
• 38:08just gives you a map
• 38:09of the overall
• 38:10broad
• 38:18aortopathy,
• 38:19in terms of their proportions
• 38:20of small cell fibroblast may
• 38:22have been slightly different, but
• 38:23the remainder was similar.
• 38:25And then when we looked
• 38:26at the, untargeted metabolomics data,
• 38:28what we're observing in this,
• 38:30set of patients as well
• 38:31is increased,
• 38:32evidence for, oxidized glutathione
• 38:35relative to reduced glutathione.
• 38:36We also observed increased levels
• 38:38of long chain fatty acids.
• 38:40And then when we looked
• 38:41at the suitable data in
• 38:42the transcriptome,
• 38:43also identified decreased expression of
• 38:45genes that are very important
• 38:46for this, for the acylation
• 38:48of long chain fatty acids
• 38:50suggesting potentially a transcriptome
• 38:52metabolome connection as well as
• 38:54decreased levels of acylcarnitines,
• 38:57that were medium and long
• 38:58chain
• 38:59that, overall may indicate a
• 39:00decreased,
• 39:01activation of beta oxidation in
• 39:03in aortic aneurysm patients.
• 39:07And so,
• 39:08you know, thinking about our
• 39:10prior study in which we
• 39:11integrated,
• 39:12smooth muscle cell genome data
• 39:13with their transcriptome data from
• 39:15the,
• 39:16smooth muscle cells, we've expanded
• 39:18upon that and and by
• 39:19doing a larger cohort. And
• 39:21so,
• 39:22we're, we've done
• 39:24genome and transcriptome
• 39:25analysis for
• 39:27sixty three cases and fourteen
• 39:29controls
• 39:30using the same approach where
• 39:31we did whole genome sequencing
• 39:32for the patient and then
• 39:33mRNA sequencing of the smooth
• 39:34muscle cells that were extracted.
• 39:36And in this,
• 39:38study, we're hypothesizing
• 39:40that differences in allelic expression,
• 39:43between,
• 39:44cases and controls may be
• 39:46a clue to the, mechanisms
• 39:48of TA development and progression.
• 39:50And so the the approach
• 39:51here is we we called
• 39:52bio allelic SNPs using genome
• 39:54data. We did, we counted
• 39:56up the number of reads
• 39:57using a a GATK ASE
• 39:59recounter in the mRNA seek
• 40:01data, and then you compare
• 40:02those using we we performed
• 40:04an analysis of this differential
• 40:06allelic expression using,
• 40:07this this score. And then
• 40:09we also did
• 40:10a a differential gene expression
• 40:12analysis using ADJAR.
• 40:14And so the results of
• 40:15this differential allele specific expression
• 40:17analysis, these are,
• 40:20you know, recent recent results.
• 40:21So the the way that
• 40:22we are are measuring,
• 40:25differential allele
• 40:26specific expression,
• 40:28is this parameter in ASC
• 40:30score, which is really, testing
• 40:32the degree of disproportion
• 40:34between, the sick reads between
• 40:36alleles,
• 40:37and then taking that score
• 40:38and doing a case control
• 40:39comparison.
• 40:40And that's the top five
• 40:42most significant loci,
• 40:44are listed here in this
• 40:45table.
• 40:46And
• 40:47and of interest, you know,
• 40:48we we see second here
• 40:50is another gene that's important
• 40:51for CoQ biosynthesis,
• 40:53CoQ
• 40:55seven.
• 40:58And then, you know, thinking
• 40:59about how if we see
• 41:00differential allelic expression, what could
• 41:03be the functional
• 41:04effect of that?
• 41:05As a really high level,
• 41:07test, we,
• 41:08we look for overlap between
• 41:10genes that were differentially expressed
• 41:11in TAA compared with controls
• 41:13and those that were
• 41:15that had significant
• 41:17loci
• 41:18that can
• 41:19that contained a loci that
• 41:20was significantly different between TAA
• 41:22and and and controls. And
• 41:24what you can see here
• 41:25is there's overlap of a
• 41:26hundred sixty seven genes that
• 41:28were differentially expressed in TAA
• 41:30and also displayed
• 41:32at least one loci with
• 41:33differential allele specific expression.
• 41:36Considering around thirteen thousand genes
• 41:38were tested,
• 41:39that's a significant overlap.
• 41:41And and of interest as
• 41:42well, you know, amongst this
• 41:44overlapping group is is is
• 41:46CoQ a b as well
• 41:47as another, the,
• 41:49homolog of CoQ a b,
• 41:51CoQ a a.
• 41:53And when we looked at,
• 41:54these a hundred sixty seven
• 41:55genes in terms of the
• 41:56pathways, we see that amongst
• 41:58those that have differential allele
• 41:59specific expression
• 42:00and increased gene expression levels,
• 42:03actin filament binding,
• 42:05was enriched amongst those genes,
• 42:07cytoskeletal
• 42:08binding. And amongst those that
• 42:09had decreased expression as well
• 42:11as, allele specific differential allele
• 42:13specific expression between
• 42:14cases and controls. We see,
• 42:16genes important for oxidoreductase
• 42:18activity,
• 42:19alcohol metabolic process, and isoprenoid
• 42:21metabolic process.
• 42:23And so,
• 42:25wanted to show the data
• 42:26specifically for the CoQAB. So,
• 42:29so here we see, as
• 42:30I mentioned, this is a
• 42:31larger subset of patients, seventy
• 42:32seven patients in which we
• 42:34see decreased expression of CoQAB
• 42:35and TAA smooth muscle cells.
• 42:38And and it was interesting
• 42:39because it was this the
• 42:40specific SNP,
• 42:42that we identified as a
• 42:43as a candidate,
• 42:45genetic modifier
• 42:46of the progression of disease
• 42:48that also that showed the
• 42:49allelic imbalance. And you can
• 42:50see here that, the fraction
• 42:51of reads with the alternative
• 42:52allele was higher in cases
• 42:54compared compared with controls.
• 42:56So potentially another, piece of
• 42:58evidence and try to trying
• 42:59to understand what is the
• 43:00mechanism by which a common
• 43:02SNP,
• 43:03could lead to, to, contribute
• 43:05to the disease pathogenesis.
• 43:07So there's limitations to to
• 43:09these data. It's it's cultured
• 43:11cells.
• 43:12It's, short read genome sequencing.
• 43:14So, the ability to phase
• 43:16variance is is challenging, if
• 43:18not
• 43:20if not impossible.
• 43:21And then we did short
• 43:22read mRNA sequencing.
• 43:24And so a better approach
• 43:25when it comes to phasing
• 43:27and allelic expression analysis would
• 43:28be long read. So, we
• 43:30are doing a study in
• 43:31which we're looking at the
• 43:32tissue,
• 43:33frozen tissue,
• 43:35extracting DNA, and performing,
• 43:37a long read whole genome
• 43:38sequencing using Oxford Nanopore technology,
• 43:41which gives you also base,
• 43:43base modification data,
• 43:45and integrating that with the
• 43:46short read mRNA sequencing data.
• 43:48And we have collected enough
• 43:50patients where about a hun
• 43:51a hundred patients are are
• 43:53done now for this. And
• 43:54so we're gonna use this
• 43:55as another way to investigate
• 43:57differential allele specific expression as
• 43:59well as other possibilities.
• 44:00We're
• 44:01combining, those data,
• 44:03and,
• 44:05and taking you know, trying
• 44:06to prioritize what may be
• 44:08observed in the human,
• 44:10endogenous
• 44:11setting
• 44:12by, using a massively parallel
• 44:14reporter assays that are going
• 44:16to be determining in smooth
• 44:18muscle cells,
• 44:19what's the transcriptional
• 44:20effect of variants that are
• 44:21look localizing in, three prime
• 44:23UTRs
• 44:24and putative noncoding
• 44:26elements as a way to
• 44:27begin to,
• 44:28sort through what whole genome
• 44:30data looks like and how
• 44:31that integrates with RNA Seq
• 44:32data.
• 44:33Okay. So,
• 44:36very much switching gears from
• 44:38from the from the tissue
• 44:39studies,
• 44:40but connected
• 44:42clinically and on a research
• 44:43basis is I wanted to
• 44:44talk about a technique that
• 44:45we've developed in collaboration with
• 44:47engineers
• 44:48at Purdue University
• 44:50to try to improve our
• 44:51ability
• 44:52to phenotype patients
• 44:54using transthoracic
• 44:55echocardiography,
• 44:57including more accurate measurements, reproducible
• 45:00measurements,
• 45:01greater throughput, as well as
• 45:03extracting
• 45:04more functional data than what
• 45:06is the standard approach.
• 45:08Our standard approach, as we
• 45:09know, when it comes to
• 45:10aortic characterization would be to
• 45:12make measurements at the annulus,
• 45:14aortic group, the San Diego
• 45:15Junction ascending aorta.
• 45:16Calculate z scores in kids,
• 45:18and there's your phenotype. So
• 45:19it's it's kind of,
• 45:21woefully,
• 45:22simple, I would say, when
• 45:23it comes to how we're
• 45:24classifying or or characterizing our
• 45:26our patients' disease. So that
• 45:27was kind of a motivation
• 45:29for trying to do this.
• 45:30So, I think what you
• 45:31can see here, right, is
• 45:32so, this is a b
• 45:33mode. Right? And we've just
• 45:35put a plane here, you
• 45:36know, to highlight the fact
• 45:37that there's tons of translation,
• 45:39right, of the aortic root
• 45:40through cardiac cycles. When would
• 45:41you make the measurement? Know,
• 45:42what borders are you using?
• 45:43What what's at what plane
• 45:45are you measuring? All these
• 45:46things are confounding
• 45:47factors when it comes to
• 45:48research and clinical care.
• 45:50And and so we've developed
• 45:51this algorithm that's designed to
• 45:53track the translation
• 45:55of the aortic root. So
• 45:56it's tracking, the translation in
• 45:58the x,
• 45:59direction, y direction, and rotation,
• 46:02in the theta.
• 46:06Okay. And you can see
• 46:07here, this is a this
• 46:08is a representation of that
• 46:09data,
• 46:10for this sample I mean,
• 46:11for this, for this series.
• 46:14And then what we get
• 46:15out,
• 46:16after the, algorithm runs and
• 46:18when this runs
• 46:22Okay. Is that the the
• 46:23algorithm, what it's doing is
• 46:25is using these parameters to
• 46:26stabilize the aortic root, within
• 46:28the image, and then you
• 46:29can take these data that's
• 46:30now stabilized and extract,
• 46:33diameter information
• 46:34that is in a consistent
• 46:36plane,
• 46:37through the aortic root. And
• 46:38you can see here we're
• 46:39starting to detect,
• 46:41you know, kind of subtle
• 46:42deflections in the aortic root
• 46:44diameter,
• 46:45through the course of cardiac
• 46:46cycles.
• 46:47And
• 46:48the way this works, just
• 46:49just briefly,
• 46:50is that we we take
• 46:51a, a DICOM file, convert
• 46:53it to a MATLAB,
• 46:55file,
• 46:56and then there's user input
• 46:57when it comes to this.
• 46:58So a user will will
• 47:00right will pull up the
• 47:01program, define the the plane
• 47:03of the annulus, define the
• 47:04plane of the sinotubular junction.
• 47:06The algorithm will then rotate,
• 47:07the aorta so that we're
• 47:08perpendicular to the longitudinal axis
• 47:11of axis,
• 47:12and then generate these contours
• 47:13that can be fine tuned
• 47:14by the user.
• 47:17From there, it's a, iterative
• 47:19frame by frame difference minimization
• 47:21algorithm that will be tracking
• 47:23the aortic translation,
• 47:24and adjusting the parameters of
• 47:26x, y, and theta.
• 47:29And and one of the
• 47:30outputs from this is, aortic
• 47:31diameter time course tracing through
• 47:33cardiac cycles.
• 47:35And so what we've observed,
• 47:37is bimodal
• 47:38behavior. So in systole,
• 47:40aortic
• 47:41root diameter will increase.
• 47:43In end systole, there's a
• 47:44a a recoil. And then
• 47:46in diastole,
• 47:47a re expansion. And then,
• 47:49you know, through the course
• 47:50of diastole then,
• 47:52further contraction or or or
• 47:53recoil
• 47:54of the diameter.
• 47:56So from these curves, we're
• 47:58able to extract the maximum
• 47:59systolic diameter,
• 48:02quantitatively
• 48:03and unbiased way,
• 48:05the end diastolic diameter,
• 48:07which,
• 48:08is notoriously tricky, I think,
• 48:10to to to to capture.
• 48:13And then,
• 48:16and so when it comes
• 48:16to diameter measurements, we
• 48:18we
• 48:19ran the algorithm and then,
• 48:21and then validated,
• 48:22compared those to manual measurements,
• 48:26and and saw a good
• 48:27agreement,
• 48:28maybe a slight bias for
• 48:29higher diameter measurements with the
• 48:31algorithm compared to to the
• 48:33manual, but overall, a good
• 48:34interclass correlation coefficient,
• 48:36between the algorithm's output and
• 48:38the manual measurement.
• 48:40And then also from these
• 48:41data, which I'll show, we
• 48:43with our you know, with
• 48:44the availability of a maximum
• 48:46systolic diameter and a minimum
• 48:48and diastolic diameter, able to
• 48:50calculate,
• 48:51biomechanical
• 48:52properties as well
• 48:54using this. So, you know,
• 48:55we looked at these patients
• 48:57twenty controls, fifteen Marfan Syndrome,
• 48:59aged ten to fifteen years.
• 49:01As expected, the diameters
• 49:02extracted by the algorithm were
• 49:04larger,
• 49:05in Marfan syndrome compared to
• 49:07controls.
• 49:08And then interestingly,
• 49:09when we use the delta,
• 49:12in diastolic to maximum systolic
• 49:14data, we're seeing increased stiffness
• 49:16of the of the aortic
• 49:18root in Marfan,
• 49:19decreased strain, and decreased,
• 49:21distensibility.
• 49:22So so, you know, as
• 49:23a as a pilot study
• 49:24to say, we might be
• 49:25able to extract
• 49:27more comprehensive biomechanical properties using
• 49:29this tracking algorithm.
• 49:33And then also, you know,
• 49:34we've thought about how else
• 49:36we may be what other
• 49:37data may be useful here.
• 49:39So,
• 49:40you know, we see a
• 49:40rate of systolic expansion,
• 49:42a rate of systolic recoil,
• 49:44a rate of,
• 49:46diastolic expansion, and the rate
• 49:47of diastolic recoil. And we
• 49:49compare those between the Marfan
• 49:50syndrome,
• 49:52cases and the controls. And
• 49:53we're seeing a slower
• 49:55rate of recoil,
• 49:57at the end of in
• 49:58in the end systole in
• 49:59patients with a Marfan syndrome
• 50:01compared to controls. And, you
• 50:02know, trying to,
• 50:05you know, kind of think
• 50:06about how that may relate
• 50:07to intrinsic,
• 50:09elastic fiber differences in a
• 50:10Marfan syndrome, for example.
• 50:13And so, you know, with
• 50:15this algorithm, we're, you know,
• 50:16seeking to establish normative values
• 50:18for these,
• 50:19metrics, which are currently not
• 50:21not available,
• 50:22across age ranges. We'll do
• 50:24more case control comparisons,
• 50:26you know, thinking about how
• 50:27we predict risk. You know,
• 50:28is there a way to
• 50:29identify subtle biomarker
• 50:32maybe subtle, maybe just unassertainable
• 50:34previously biomechanical
• 50:36properties that could be predictive
• 50:37in a patient who has
• 50:38a an aortic root diameter
• 50:40of future progression, for example.
• 50:42Probably, this will improve our
• 50:44technical reproducibility
• 50:45between users,
• 50:47in terms of, you know,
• 50:48extracting,
• 50:49reliable
• 50:49information,
• 50:51and then, you know, working
• 50:51to translate the the algorithms
• 50:53used to animal models would
• 50:55would be powerful.
• 50:56And engineers love to keep
• 50:57developing, so they they've,
• 50:59started to develop additional kind
• 51:01of techniques to to do
• 51:02similar,
• 51:05data analysis, and that includes,
• 51:07using, NURBS curves for their
• 51:09ability to, make a continuous
• 51:11parametric curve,
• 51:13smoother data,
• 51:14less noise, and then sub
• 51:15pixel diameter measurements.
• 51:17And then expanding further upon
• 51:19that there,
• 51:21we've been working on machine
• 51:22learning,
• 51:23development,
• 51:24in order to automatically,
• 51:26segment the aortic root, for
• 51:29analysis,
• 51:30that would be much higher
• 51:31throughput than than what we're
• 51:32doing, currently. So,
• 51:34much work is is going
• 51:35on in that regard.
• 51:37And so,
• 51:39you know,
• 51:40putting things together with what
• 51:41I've shown you today, kind
• 51:43of other data and our
• 51:44data,
• 51:45thinking about how this ties
• 51:46into patient care. Right?
• 51:47And so, you know, trying
• 51:50to to think about how
• 51:51can we develop a better
• 51:52assessment of our patients, especially
• 51:54at early ages,
• 51:55and then inform our our
• 51:57cardiac management decision making
• 51:59accordingly.
• 52:00And so, one approach, you
• 52:02know, we are commonly doing
• 52:03a a TA panel in
• 52:04our patients who come into
• 52:05clinic, with a buccal swab.
• 52:07And we're always doing an
• 52:08echocardiogram,
• 52:09and we're currently measuring, aortic
• 52:11diameters.
• 52:12And so some of the
• 52:13tools that we wanna use,
• 52:15based on our data and
• 52:17research and then ultimately
• 52:19translating into clinic would be
• 52:20genome sequencing combined with transcriptome
• 52:22analysis in order to, you
• 52:23know, twenty only to you
• 52:24know, you know, it's a
• 52:26fraction, you know, anywhere from
• 52:27five to twenty percent of
• 52:28patients who have erotopathy
• 52:30that we are identifying genetic
• 52:32causes. So thinking about how
• 52:34genome plus transcriptome could be
• 52:35a more, robust,
• 52:36a way to, make diagnoses
• 52:38as well as, you know,
• 52:39our development of a variant
• 52:40functional assays.
• 52:42To do that, we you
• 52:42know, I showed you our
• 52:43echo tracking method, which could
• 52:45be, implemented,
• 52:47in,
• 52:48you know, in the coming
• 52:49years.
• 52:50And then, you know, from
• 52:51these data, we're getting information
• 52:52about genetic cause. We're identifying
• 52:55potential genetic modifiers through our
• 52:57studies,
• 52:58and then also identifying structural
• 53:00parameters. So can you, over
• 53:02time, accumulate data that,
• 53:04can be integrated into a
• 53:05risk classifier, and then we
• 53:06can start to stratify cardiac
• 53:08management?
• 53:09And really all this is,
• 53:10you know,
• 53:11you know, would be potentially
• 53:13doable when it comes to,
• 53:15the standard clinical workflows.
• 53:17Also, from these data, you
• 53:18know, we're learning about what
• 53:20could be the path of
• 53:20biology of aortic aneurysm and
• 53:22identifying therapeutic targets. And I
• 53:24went through a series of
• 53:25studies in which, overall, I
• 53:26think we're, you know, kind
• 53:27of
• 53:29the picture is one of,
• 53:31you know, metabolic,
• 53:32dysfunction, generally, you know, with
• 53:34oxidative stress,
• 53:35the role for CoQAB,
• 53:37in the smooth muscle cells
• 53:38and other evidence for CoQ
• 53:40ten synthesis genes,
• 53:42the changes in long chain
• 53:43fatty acid acylation that we've
• 53:45identified,
• 53:46lysophosphatidic
• 53:47acid metabolism, and and this
• 53:48candidate gene.
• 53:49So these are, you know,
• 53:51areas of further investigation.
• 53:54And so,
• 53:55these are, many, many people
• 53:57who I've worked with, collaborated,
• 53:59you know, have mentored me.
• 54:00I wanted to point out,
• 54:02Joel Corvera,
• 54:03aortic surgeon at Indiana University,
• 54:05was essential
• 54:06to the collection of aortic
• 54:07tissues, for example. Craig Orgin,
• 54:09who's a a biomedical engineer
• 54:11at Purdue University,
• 54:12cardiovascular
• 54:13imaging research lab.
• 54:15Glenn Iannucci,
• 54:16pediatric cardiologist who leads the
• 54:18aortic center at Emory University,
• 54:19which is our collaborator for
• 54:21a longitudinal study.
• 54:22Freddie Damon is a a
• 54:24pediatrics resident at Stanford who's,
• 54:27who wrote the code for
• 54:28our tracking algorithm. Shubh is
• 54:29a PhD student who's gonna
• 54:30be coming, next year here,
• 54:32is working on the algorithm
• 54:34as well. And KB was
• 54:34a medical student who did
• 54:36the a lot of the
• 54:36transcriptome
• 54:37data.
• 54:39So,
• 54:41yeah. So thanks for your
• 54:42attention. I really appreciate it.
• 54:43Happy to answer any questions.
• 54:46If if
• 54:52That was really
• 55:09the the hot spot and
• 55:09the variability that we're gonna
• 55:11have to deal
• 55:12with something.
• 55:14And then we just find
• 55:15that to better understand who,
• 55:19segments of the population will
• 55:21have a differential response
• 55:25their, like, their standard treatment
• 55:26should be modified, you know,
• 55:28in other words, are
• 55:29there gene variant treatment
• 55:32interaction
• 55:33that
• 55:35could then divide those who
• 55:37are more likely
• 55:46Super super interesting.
• 55:49And I'm very curious if
• 55:50you have
• 55:51any
• 55:55early data on whether there
• 55:58are complications of those patterns
• 56:01that predate likelihood
• 56:03of rupture in other words.
• 56:04You know, I think of
• 56:05the standard way or I
• 56:06think you're fortunate in terms
• 56:08of how to
• 56:10narrow the measurements
• 56:11and our reporting,
• 56:12which is
• 56:13cool. But I'd be very
• 56:15curious to know if you
• 56:16could use that to kinda
• 56:18say, well, this person is
• 56:19likely,
• 56:20you
• 56:21know, there's a change in
• 56:22that pattern that says this
• 56:24is working that needs to
• 56:25go quicker.
• 56:28So with respect to the
• 56:29first question, you are I
• 56:30think the data,
• 56:32tying genotype to the outcome
• 56:34response to therapy
• 56:36is Right. Was a trial
• 56:37called the compare trial,
• 56:39done in Netherlands
• 56:41years ago, maybe, where they
• 56:43they did some some in
• 56:44vitro work to classify the
• 56:46FBN one variance.
• 56:48This
• 56:49is,
• 56:50like, we've done.
• 56:53And then they they did
• 56:54they did,
• 56:55suggest
• 56:56that the, wasartan would patients
• 56:59would likely
• 57:01be.
• 57:07So that's an example. I
• 57:09think,
• 57:12you know, it's,
• 57:13when it comes to genotype
• 57:15outcomes,
• 57:16also, you know, some of
• 57:17these more recent multitudes,
• 57:20consortium studies have started to
• 57:22look at what's, you know,
• 57:24after the pair, what's the
• 57:25likelihood of of population, and
• 57:27how does a genetic diagnosis
• 57:29have the the
• 57:31those workplace
• 57:33or a sort of simple.
• 57:40Probably emerging data in the.
• 57:44You know, in terms of
• 57:44the you know? Yes. Absolutely.
• 57:46So I think everybody knows
• 57:48that you would for issues
• 57:49that have an aortic rupture,
• 57:50this section is
• 57:52more complicated than just the
• 57:54size, which is the the
• 57:55approach.
• 57:56And and so we'd be
• 57:58very interested to begin to
• 58:00look at those,
• 58:01you know, who have echo
• 58:03data and the fact that
• 58:05with outcomes. You know, echo.
• 58:07You know, there's other data.
• 58:09We We started to discuss,
• 58:11and and outcome distribution. They're
• 58:12very interesting. We have, recently,
• 58:15you know, we're just not
• 58:16to to start, but instead
• 58:18of transthoracic
• 58:19echo,
• 58:20using TDE data,
• 58:22in in
• 58:23the
• 58:25OR,
• 58:28run the run run our
• 58:29test, our math,
• 58:31on the air vision
• 58:33properties.
• 58:35Maybe some ability to definitely
• 58:37ability to monitor blood pressure
• 58:38real time,
• 58:39you know, for extraction of
• 58:41of of properties such,
• 58:43but also,
• 58:44you know, a very robust
• 58:46approach to tissue collection. You
• 58:50know, as we're thinking about,
• 58:51does,
• 58:52a certain
• 58:53dysfunction
• 58:54in the aortic,
• 58:55dynamics,
• 58:56correlate with a certain type
• 58:58of tissue abnormality, whether it's.
• 59:06Yeah. That that that that's,
• 59:08Here he is. I'm walking
• 59:09over to Jeff here. You
• 59:10mentioned in one slides the
• 59:12potential for IPS.
• 59:13Well, so I'm curious what
• 59:15I'm very curious what you
• 59:16do, and
• 59:17we have, obviously, connection within
• 59:19the CRC with the to
• 59:20support that could work.
• 59:23Yeah. No. I think, absolutely.
• 59:24I think, you know, if
• 59:25we wanted to
• 59:28to optimize our
• 59:30the approach when it comes
• 59:31to scriptional
• 59:32analysis or the interpretation of
• 59:34variants, you know, we need,
• 59:36you know, we can't get
• 59:37all the information we would
• 59:38need with a blood sample.
• 59:39So in patients who are
• 59:40preoperative,
• 59:41you know, we could we
• 59:42could,
• 59:43generate the packing cells,
• 59:45for the the two part
• 59:47molecular assessments
• 59:48to try to to correlate.
• 59:55As well as, you know,
• 59:56obviously, the ability to to
• 59:58to to preserve the system.
• 01:00:10Congratulations.
• 01:00:11That's so
• 01:00:12and you're with with this
• 01:00:14really
• 01:00:15pretty much work.
• 01:00:17No questions about a lot
• 01:00:19of these genetic in connections
• 01:00:21to air top feet.
• 01:00:24I I guess
• 01:00:25and I have a lot
• 01:00:26of questions, but I'll ask
• 01:00:27one, sort of general question.
• 01:00:29When I hear a talk
• 01:00:31where there are ten, sometimes
• 01:00:33hundreds of genes and gene
• 01:00:35modifiers
• 01:00:36that result in not identical
• 01:00:38but similar
• 01:00:40pathology or pathologic phenotypes.
• 01:00:42I always wonder, like, there's
• 01:00:44got to be a common
• 01:00:46pathophysiologic
• 01:00:47driver
• 01:00:49of the of the phenotype
• 01:00:50of the aortic dilatation
• 01:00:52and that and the dissection.
• 01:00:54You
• 01:00:55mentioned,
• 01:00:57issues.
• 01:00:59You mentioned oxidative stress.
• 01:01:01Is there anything about,
• 01:01:04mechanosensing
• 01:01:05of the
• 01:01:07aorta
• 01:01:07in its abnormal state that
• 01:01:10drives,
• 01:01:11you know, the oxidative stress
• 01:01:12or might drive inflammation, which
• 01:01:14you didn't talk about much
• 01:01:16that I have to mention?
• 01:01:25Yeah. So absolutely. So Jay
• 01:01:27Humphrey,
• 01:01:29here at at Yale and
• 01:01:30then
• 01:01:31she did a really
• 01:01:33nice paper with Diane and
• 01:01:34Melowitz
• 01:01:36thinking about how the smooth
• 01:01:38muscle cells, using these these
• 01:01:39things could be,
• 01:01:41when you look at the
• 01:01:42spectrum of genes,
• 01:01:44could be a
• 01:01:45kind of functionality
• 01:01:47or a degree of how
• 01:01:49mechanics and
• 01:01:51I mean, it's,
• 01:01:56remains to be done. So
• 01:01:57this was this was just
• 01:01:58a site I've used in
• 01:01:59the past, but kinda kinda
• 01:02:00highlights some of that where
• 01:02:01you you
• 01:02:02have the smooth muscle cells,
• 01:02:04sensing force generating force
• 01:02:07in the tissue, but then,
• 01:02:09you know, obviously,
• 01:02:11do the extra tether matrix
• 01:02:12and
• 01:02:13how is how is that
• 01:02:14contributing? So I think that
• 01:02:16that's,
• 01:02:17yeah, I think that that's,
• 01:02:18like, path
• 01:02:20to shore. Right? You're absolutely
• 01:02:22right. It'd be terrific to
• 01:02:23find a common,
• 01:02:25happiness.
• 01:02:27And I think as we're
• 01:02:29it's a it's a challenge
• 01:02:31when our when our human
• 01:02:32studies,
• 01:02:33because we have such, heterogeneity
• 01:02:35as genetic heterogeneity.
• 01:02:37Patient heterogeneity.
• 01:02:39But we
• 01:02:40see, you know, associations.
• 01:02:42We can hone in. This
• 01:02:44is genetic.
• 01:02:45We're not. And it see
• 01:02:46associations. It it raises the
• 01:02:48possibility that what you guys
• 01:02:49are it's,
• 01:02:50relatively common.
• 01:02:52You know, so I I
• 01:02:53think, there's there's a lot
• 01:02:54of work. Yeah. Yeah. To
• 01:02:56be done. Just try to
• 01:02:57to solve solve that problem.
• 01:02:59I mean, you know, eighty
• 01:03:01percent in kids, we use
• 01:03:02eighty percent.
• 01:03:06Which is really very rationalist.
• 01:03:08Rationalist developed in markets.
• 01:03:10It's a big trap.
• 01:03:12And
• 01:03:13and if probably affect it.
• 01:03:15You know? Your day is
• 01:03:15not. But, I mean, it's
• 01:03:17you know? You can get
• 01:03:18at it, you know, that
• 01:03:19type of question.
• 01:03:20I'm sorry.
• 01:03:23It's.
• 01:03:25And inflammation?
• 01:03:26Yeah. I, I see kids,
• 01:03:28so we don't see a
• 01:03:28lot of the you know,
• 01:03:30it's not a delayed
• 01:03:32age largely. I know there's
• 01:03:34some House house, man, which,
• 01:03:36you know, I'm not sure
• 01:03:37how physiologically relevant they are.
• 01:03:38It's a
• 01:03:40very, kind of, robust,
• 01:03:42mandatory,
• 01:03:44and
• 01:03:45then it's
• 01:03:47just also mechanism dissection.
• 01:03:50Maybe last question, Bayaria.
• 01:03:52Yeah. Actually, my question is
• 01:03:54very similar to this. And,
• 01:03:55you mentioned that this, calcium
• 01:03:57sensing I mean,
• 01:03:59calcium channel TRP v two,
• 01:04:01which is a mechanosensing
• 01:04:02channel,
• 01:04:03is increased,
• 01:04:05in expression of that. Have
• 01:04:06you looked at the other
• 01:04:07diseases you see in similar
• 01:04:09pattern you find so that
• 01:04:10you can have a common
• 01:04:11pathway? And another thing is
• 01:04:12a mitochondrial
• 01:04:13disease that you have. The
• 01:04:15link is actually unclear.
• 01:04:16But have you looked at
• 01:04:17the other ones such as,
• 01:04:18you know, TGF beta receptor
• 01:04:20mutations?
• 01:04:21If they have the same,
• 01:04:22and can you link that
• 01:04:23to, like, your TGF signaling
• 01:04:26as a common because there
• 01:04:27are a lot of them
• 01:04:29in that pathway,
• 01:04:31to see. And and then
• 01:04:32finally, the question is that
• 01:04:33if there is a mechanosensing
• 01:04:35and it has some you
• 01:04:36know, channel is actually a
• 01:04:37mechanosensing channel,
• 01:04:39why did you fail with
• 01:04:40the treatment of these? And
• 01:04:42have you ever looked to
• 01:04:43see, actually, these these genotype
• 01:04:44specific to people who like
• 01:04:45to have higher expression of
• 01:04:47TRP v two? Do you
• 01:04:48see an association in response
• 01:04:50to treatment that reduces
• 01:04:52blood pressure?
• 01:04:53Thank
• 01:04:57you.
• 01:04:58So
• 01:04:59Right. So so with respect
• 01:05:00to TRPV two, it's, you
• 01:05:01know, that's a observation in
• 01:05:03in Marfan syndrome,
• 01:05:05and when you see mutations.
• 01:05:07All these. So I think
• 01:05:08it warrants
• 01:05:09more investigation
• 01:05:11in in
• 01:05:12across populations.
• 01:05:15You know, I think,
• 01:05:16we can continue to investigate
• 01:05:18that with our, you know,
• 01:05:19larger cohort of mRNA seek
• 01:05:21data, for example.
• 01:05:23We just haven't gotten that
• 01:05:25got to that point yet.
• 01:05:29You know, mitochondrial the role
• 01:05:30for mitochondria, you know, there's
• 01:05:31some evidence in mouse models
• 01:05:33that was pretty compelling when
• 01:05:34it came to the,
• 01:05:36connection between the,
• 01:05:38each other matrix
• 01:05:40dysfunction
• 01:05:41and the links between
• 01:05:43the across the membrane to
• 01:05:44the bladder.
• 01:05:46And
• 01:05:47that's that remains to these.
• 01:05:49I mean, I would say,
• 01:05:50you know, we do see
• 01:05:51it's mild mild aortic root
• 01:05:53dilation in in
• 01:05:55kids who have, genetic mitochondrial
• 01:05:57disease. You know, thinking about
• 01:05:59how, you know, a piece
• 01:06:00of evidence for, you know,
• 01:06:03cut out is of the
• 01:06:04situation.
• 01:06:06And
• 01:06:07so
• 01:06:10Ben, can I ask a
• 01:06:11quick question?
• 01:06:12Yes.
• 01:06:14Exciting talk.
• 01:06:15Yeah. As Eric mentioned, we
• 01:06:16could provide IPSC engineering, tissue
• 01:06:19engineering to generate a small
• 01:06:21muscle tissue for you to
• 01:06:22study your disease.
• 01:06:24I have a question about,
• 01:06:25oxidative phosphorylation.
• 01:06:27So we also observe in
• 01:06:29our early diagnosis,
• 01:06:30iPS cells, small cells. We
• 01:06:32observe abnormal loss production. What
• 01:06:34are your thoughts about why
• 01:06:36this aneurysm is to know
• 01:06:38stenosis is small cells. They
• 01:06:40tend to have,
• 01:06:54a reasonable place to start
• 01:06:55with with that is is,
• 01:06:58an abrogation
• 01:06:59in,
• 01:07:00gene expression that,
• 01:07:02is, you know,
• 01:07:05required to mitigate,
• 01:07:08you know, to to
• 01:07:10temper
• 01:07:11oxidative stress and reactive oxygen
• 01:07:13species?
• 01:07:14You know, is there a
• 01:07:16mitochondrial,
• 01:07:17dysfunction,
• 01:07:18dysfunction
• 01:07:19in aerobic respiration
• 01:07:21that leads to,
• 01:07:23spurious generation
• 01:07:24of, of, free radicals.
• 01:07:27There's some data that, you
• 01:07:28know,
• 01:07:29could be potentially related to,
• 01:07:31angiotensin
• 01:07:32receptor signaling,
• 01:07:34you know, that cascade,
• 01:07:36leading to increased generation of
• 01:07:39of of ROS.
• 01:07:41Those are my thoughts right
• 01:07:42now, but I'm excited about
• 01:07:43the, IPS cell, as you
• 01:07:44mentioned.
• 01:07:46Great. Well, critical Thank you.
• 01:07:47Thank you for, teaching us,
• 01:07:49here, and
• 01:07:51brave enough to work.
• 01:07:53I'll do it now. So
• 01:07:54thank you for the, and,
• 01:07:56hopefully, I believe you heard
• 01:07:58some technical collaborations
• 01:07:59between our section and departments
• 01:08:01on its work,
• 01:08:02and, representing here how it
• 01:08:04proceeds over the. Thank you,
• 01:08:06Luis.
• 01:08:06Thank you so much. Thanks,
• 01:08:07Adrian. Thank you.