Drosophila Models of Simple and Complex Diseases

1
Drosophila Models of Simple and Complex Diseases
And their uses in Therapeutic Screening
Alice Schmid, PhD University of Utah Eccles In
stitute of Human Genetics
Test Smart DNT
2
Drug Discovery a brief history.
Until 1970, most discovery occurred in whole ani
mals In 1975, a shift to cell-based assays bega
n in Big Pharma Discovery. By 1982, Big Pharm
a was almost entirely focused on Cell Based
Assays, using robotic approaches to perform Hi
gh Throughput Therapeutic screens.
By 2000, high throughput cell-based assays had i
dentified more than 50 million data points (th
e effect of one compound at a particular dose in
a particular assay). (Drug Discovery A Historic
al Perspective Drew, J., (2000) Science 287 19
60-4).
3
The problem with these approaches was they had
separated
Discovery and Toxicology.
They had also become expensive and rather
unproductive!
4
In 2000, Mel Feany and Welcome Bender published
a landmark paper, reporting a Drosophila model of

Parkinsons Disease.

• This paper
• reported the most accurate animal model of PD to
  date.
  
• age-related neurodegeneration specific to
  dopaminergic cells
  
• Produced histological hallmarks of PD that had
  never been observed in vertebrates
  
• Subsequent investigators built on this model
• PD Flies on clinically relevant drugs responded
  in ways similar to human patients.
  
• Responses were consistent with known
  dose-response curves
  
• PD flies could be used to predict human
  pathology!
  
• Within 6 years, Drosophila geneticists have
  generated models of
  
• Huntingtons Disease
• Prion/prion-like neurodegeneration
• Fragile X Mental Retardation
• Spinocerebellar Ataxia 1
• Alzheimers Disease
• And more than 20 others.

5
Why are flies so amenable to models of human n
eurodegeneration? Are these models valid? How
similar to humans ARE flies? Are they useful i
n drug discovery? How efficient are they for dr
ug development, as compared to cell-based assays
?
6

• Why flies?
• We know a great deal about their
• simple nervous systems
• lineages,
• expression patterns,
• patterns of connectivity
• transcriptional regulation.
• We have more than 100 years of genetics--the
  Drosophila genome is the best annotated.
  
• Little genetic redundancy.
• Using the a misexpression system co-opted from
  yeast geneticists,
  
• we can misexpress human genes in the fly at
  particular times/places.
  

7
Ectopic Gene expression using GAl4
In yeast Galatose is an alternative energy sourc
e--
--One that requires new enzyme synthesis.
The intracellular portion of the Galactose R
Is translocated to the nucleus and acts as a TF.
Gal4
UAS Upstream Activating Sequence
UAS
promoter/enhancer-gal 4
UAS-reporter
UAS-reporters can include human genes for
misexpression AND fluorescent proteins, etc., for
cellular detection.
8
Are these models valid?
On the plus side, there are drugs for Huntington
s disease and other polyQ expansion
repeat diseases that are in clinical trials
following discovery in Drosophila drug
screens. The Drosophila PD model was highly
esteemed---until recently. It was discovered tha
t dopaminergic histology is not trivial in
insects, requiring more careful assessment of DA
losses.
9
Just how similar are human and Drosophila genes
and cells?
Dr. Ethan Biers group has compiled Homologies
into a searchable database known as Homophila.
http//superfly.ucsd.edu/homophila/ Their wor
k identified distinct homologs
for 77 of human disease causing genes
(i.e., 714/929 genes). As a comparison, 500 gen
es currently encode the targets
of all drugs currently on the market.
Of the 714 drosophila orthologs, 153 had availa
ble mutants. 56 more had p-element tags (209).
These can immediately be used in modifier screen
s!!! Note these are likely to be loss of fun
ction alleles. The gal4 system allows gain of fun
ction phenotypes. Most amyloidgenic diseases are
gain of function. The PD drug studies suggest s
ubcellular mechanisms may be surprisingly conserv
ed-- But the jury is still out.
10
Drug Screening in Flies---how does that work
exactly?
A large number of investigators have been
pursuing a laborious HTS approach in flies
Exelixis Use Drosophila genetics to isolate
drug targets. Performed scree
ns for many industry clients (i
n development of drugs, pesticides, technical
assays, etc.) Novartis Drosophila used in ge
netic screens for drug targets for more than ten
years. EnVivo Using Drosophila in high throu
ghput screens using adults, and developmental
stages. In academics, invertebrate drug sc
reens are becoming commonplace
Some of the best-known Peter Lansbury, a
t Harvard, has collaborated on HTS screens with
Mel Feany, using the PD paradigms.
Lawrence Marsh and Leslie Thompson at UC Irvin
e have screened for drugs using their Drosophi
la model of HD, and have HDAC inhibitors in cl
inical trials.
11
HTS drug screening is generally defined as
screens of 10,000
compounds or more per week.

• Transgenic flies expressing human peptides
• are fed instant fly food containing drugs and
  permeabilizing
  
• delivery vehicles.
• After several days, they are assayed for a
  phenotype
  
• Lifespan/viability
• Neurodegeneration of known neurons
• Behavior
• Histological defects
• Morphology

.
12
Drug Screens relying on lifespan extension are
not high throughput.
The amount of work involved makes HTS approaches
Expensive. (For the fly world, expensive is a
totally foreign concept!) The gene
rational time of the fly is 10 days
The wild type lifespan in the lab is
approxiamtely 50 days. Typica
lly, age-related neurodegeneration
shortens the lifespan by 50, to 30
days.
13
It is now possible to develop HTS approaches
using automated approaches.

• Transgenic backgrounds, expressing
• Mutant phenotypes or
• Human peptides
• In a cell-specific manner
• As well as fluorescent proteins

Using robotic approaches, fly phenotypes are
assayed in a high throughput manner
These approaches are currently in development
There are significant problems to overcome
Permeability Phenotypes can be complex As comple
xity increases, screening efficiency decreases.
Most of these HTS screens occur in the context of
development-- Whether or not developmental contex
ts can be used to model degenerative contexts?
---not yet known.
14
Comparing Flies Screens to Cell-Based Screens

• The advantages
• A return to combined toxicology and discovery
• Speed is comparable
• The ability to rapidly alter ones
  assay-genotype.
  
• High throughput efficiency
• Flies are CHEAP.

• The disadvantages
• Big deal. Its a fly. It gets you no closer to
  a drug in humans.
  
• Validation can require significanty more time.
• Drug delivery can be problematic.
• Some systems are more difficult to model (eg
  pulmonary diseases)
  

15

• What lies ahead?
• The answer is in computing.
• Biology has limits computing can extend them.
• Proof of concept studies.
• These require drugs that work, in humans AND
  flies,
  
• and for some disease models, no such drugs
  exist.
  
• Conventional (non-HTS) screens have generated
  hits.
  
• Will they be borne out?

16
Are they useful in drug discovery? How effici
ent are they for drug development,
as compared to cell-based assays?
The largest drawback of course is evolutionary
distance from humans. Is that severe enough to m
ake flies irrelevant? Flies are the single be
st forward genetic screening model (with all due
respect to worms and yeast). misexpression syst
em easy transgenic production extremely well
annotated genome There is considerable evidenc
e to suggest that drugs that work in humans work
in flies. (eg Mel Feanys PD fliesbut also dopa
minergic toxins) Flies are useful in identifyin
g drug classes. They can be used for rapid scr
eening. They can be used for general studies
of toxicity.