Transcript Investigating Plant Growth Using AVS Dr. R. P. Fletcher University of

November 8th 1999
UAUUG Birmingham
1
Investigating Plant Growth
using AVS
Presentation to the
UK AVS and Uniras User Group Meeting
University of Birmingham
November 8th 1999
Dr. R. P. Fletcher
University of York
A report on work done by:
•
•
•
•
Dr. S. M. Bougourd, University of York
Dr. C. L. Wenzel, University of York
… in collaboration with
Dr. J . Haseloff, MRC Laboratory of Plant
Science, Cambridge
• …and me
November 8th 1999
UAUUG Birmingham
3
Outline
•
•
•
•
•
•
Which part of plant growth?
Which plant?
Why?
How?
How we use AVS
What we want to do (with AVS?)
November 8th 1999
UAUUG Birmingham
4
Which part of the plant?
•
•
•
•
•
•
Above or below ground?
For us … below
This means the ROOTS
Specifically …
How do the root cells differentiate?
Which cells elongate and why?
November 8th 1999
UAUUG Birmingham
5
Which Plant?
•
•
•
•
•
•
•
Aribidopsis thalinana
A member of the brassica family
Also known as:
Thale cress
or …
Mouse Eared cress
It’s a weed!
November 8th 1999
UAUUG Birmingham
6
Just so you know what it looks like
Whole Plant
Flowers
November 8th 1999
UAUUG Birmingham
7
… and there’s more ...
November 8th 1999
UAUUG Birmingham
8
Why use this weed?
•
•
•
•
•
•
Small size and rapid life cycle
Prolific seed production
Simple genome
Many mutants and transformed populations
Perturb the behaviour of targeted cells
Monitor phenotypic expression
November 8th 1999
UAUUG Birmingham
9
The goal
“To understand the genetical and
cellular interactions that co-ordinate
the development of the root meristem”
November 8th 1999
UAUUG Birmingham
10
How we acquire the data
• Roots are visualised using Laser Scanning
Confocal Microscopy (LSCM)
• Also known as Confocal Scanning Laser
Microscopy (CSLM)
November 8th 1999
UAUUG Birmingham
11
Quick tutorial on CLSM
• A scanning laser beam is focussed onto a
fluorescent specimen
• Mixture of reflected and emitted light is
captured by a photo-multiplier via beam
splitter
November 8th 1999
UAUUG Birmingham
12
Tutorial continued
• Arranged so only the emitted light enters
the photo-multiplier
• A confocal aperture (pin-hole) placed in
front of the photo-multiplier
• The effect is to only allow emitted light
from the “in focus” area to pass into the
photo-multiplier
November 8th 1999
UAUUG Birmingham
13
Principles
November 8th 1999
UAUUG Birmingham
14
Typical System
November 8th 1999
UAUUG Birmingham
15
The real thing
November 8th 1999
UAUUG Birmingham
16
Interesting problem?
• Its all very well staining specimens so that
they fluoresce, but ...
• We need to see whole root tip, not just
sections and ...
• We need same level of staining throughout,
but ...
• Normal stains kill the cells and are bleached
by the laser scanning process
November 8th 1999
UAUUG Birmingham
17
The Solution!
• Everybody’s buzzword these days
• Genetic Modification!
• The idea is to get the plant to manufacture
its own fluorescent stain
• So, we will borrow a gene from somewhere
else in the natural world
November 8th 1999
UAUUG Birmingham
18
Obtaining the Gene
• Plenty of naturally fluorescent plants and
animals out there
• The oceans are full of them
• The jellyfish, Aequorea victoria, from the
Pacific Ocean has been used.
• They produce the protein, Green
Fluorescent Protein (GFP).
November 8th 1999
UAUUG Birmingham
19
Wibbly Wobbly Jellyfish
November 8th 1999
UAUUG Birmingham
20
Pretty, Pretty
November 8th 1999
UAUUG Birmingham
21
… and they can swim
November 8th 1999
UAUUG Birmingham
22
Getting the Gene into the Plant
• A quick tutorial about genetic modification
• … gene extracted ... put in vector, a soil
bacterium … isolate “infected cells” and
regenerate whole plants.
• Can even link “instructions” to the GFP
gene to make the plant only produce the
fluorescent protein in certain parts of the
plant
November 8th 1999
UAUUG Birmingham
23
A Single Image
November 8th 1999
UAUUG Birmingham
24
An Image Stack
November 8th 1999
UAUUG Birmingham
25
Getting this Stack into AVS
• The old nutshell!
• First, find out the format of the Bio-Rad
PIC files.
• Hunt round for some “v” … IAC maybe?
• Got some code, but was developed for
ALPHA
• Had “endian” problems
November 8th 1999
UAUUG Birmingham
26
Fix the code and develop
Visualisation Modules
• Fix the “v” code to read the correct “endianness” of the data
• Amount of data can be a problem
• 512 * 768 * stack size (loadsa data!)
• Hope the decimation modules in Version 5
will help here
• Even running on 350Mhz PC or SGI 02,
both with 128 Mb of memory, AVS is slow
November 8th 1999
UAUUG Birmingham
27
Network for preliminary viewing
November 8th 1999
UAUUG Birmingham
28
Using AVS to view along a
different axis
tip
Single frame
November 8th 1999
Back a bit
UAUUG Birmingham
29
Movie view along the axis
November 8th 1999
UAUUG Birmingham
30
What are we actually seeing?
• GFP fluorescing in the cell walls
• The higher the intensity the more GFP
• Would be better to invert the images
November 8th 1999
UAUUG Birmingham
31
Inverted Image Stack
November 8th 1999
UAUUG Birmingham
32
Non-invasive non-lethal
• The use of the GFP means we can study the
plant root growth “in vivo”
• The aim is to understand the fate of the
different root tip cells
• Need to find a way to “tag” cells from one
image stack to another
• Time dimension
November 8th 1999
UAUUG Birmingham
33
Cell fate?
Divide
Root tip cell
Differentiate
Some just grow
November 8th 1999
Some elongate and grow
UAUUG Birmingham
34
Need to see 3D view
•
•
•
•
3D reconstruction from “cloud of points”
Need to “cut away”
Need to “identify” cells
Need to track “fate”
November 8th 1999
UAUUG Birmingham
35
Preliminary 3D Investigation
Orthoslices
November 8th 1999
UAUUG Birmingham
36
Animate the orthoslices
November 8th 1999
UAUUG Birmingham
37
Complex Network
November 8th 1999
UAUUG Birmingham
38
Add in some “real” 3D
Volume
November 8th 1999
UAUUG Birmingham
39
Another View
November 8th 1999
UAUUG Birmingham
40
Animated volume cutaway
November 8th 1999
UAUUG Birmingham
41
So just how useful is AVS?
•
•
•
•
Using AVS can really help to see the data
Reconstructing different orthogonal views
Volume visualisation will help
Data volume is a problem on “small”
systems
• Decimation routines will be welcome
November 8th 1999
UAUUG Birmingham
42
Future Work
• Need to work out how to mark cell volumes
in order to track specific cells
• Create new fields from marked data
• Visualise these “new” fields with time “n”
images
• Difference frames may help from time “n”
to time “N+1”
• Big data processing effort here needed
November 8th 1999
UAUUG Birmingham
43
THAT’S
ALL
FOLKS
November 8th 1999
UAUUG Birmingham
44