mycofluidics - the fluid mechanics of fungal chimerism and spore...

MycofluidicsThe Fluid Mechanics of Fungal Chimerism and Spore Dispersal

Presenters: Devin Bowers, Trevor Caldwell, Francesca Grogan, LisaYamada

Mentors: Prof. Marcus Roper, Prof. Emilie Dressaire

UCLA Applied Math REU

August 8th, 2012

Introduction to Mycology

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 2 / 25

Notable Fungi

Rice blast

Cordyceps

White nose syndrome

Figure: Cordyceps versus an ant

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 3 / 25

Introduction to Mycology

A chimera is an organism that is made up of cells with geneticallydifferent nuclei.

Can reproduce asexually via spores

Most fungi grow as hyphae, which are filamentous structures thatgrow at their tips.

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 4 / 25

Moran Process

A Moran process is a stochastic process that models processes affectinggenetic diversity.The dynamics of two different populations of nuclei withina fungi can be modeled as a two species Moran process.

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 5 / 25

Roper M, Simonin A, Hickey PC, Leeder A, Glass NL (2012) Nuclear dynamics in afungal chimera (preprint).

Previous Work

Fungal chimerism gives rise to the internal genetic diversity of fungi.It has been found that nuclear mixing within the mycelium keepsgenetically different nuclei well-mixed and in stable proportionsthroughout the fungus.

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 6 / 25

Roper M, Simonin A, Hickey PC, Leeder A, Glass NL (2012) Nuclear dynamics in afungal chimera (preprint).

Our Experiment: Question

Is a fungus able to maintain its genetic diversity? If so, how?

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 7 / 25

Our Experiment: Method

Neurospora crassa (N. crassa), atype of bread mold, with GFP orDsRed proteins

GFP and DsRed cause nuclei toemit a green and redfluorescence, respectively.

Figure: N. crassa Cultures

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 8 / 25

Our Experiment: Method

Figure: Race Tubes

We built and innoculatedrace tubes with bothGFP and DsRed.

When the coloniesstarted sporulating, wetook images with amicroscope.

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 9 / 25

Our Experiment:Method

Figure: DIC Images of Spores

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 10 / 25

Our Experiment:Method

Figure: Fluorescent DsRed and GFP

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 11 / 25

Our Experiment: Method

Image Analysis

Find the percentage of spores that have only red nuclei (pR), onlygreen nuclei (pG ), and both (pY ).

Find the frequency of the number of nuclei present in a spore to getP(Nnucl = n).

From this, we can solve∑n

P(Nnucl = n)pnr = pR

∑n

P(Nnucl = n)(1− pr )n = pG

for pr , the probability that a nucleus is red.

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 12 / 25

Our Experiment: Method

Pr Fitting

Estimate pr from likelihood function

L(NR ,NG ,NY ) =N!

NR !NG !NY !pNRR pNG

G pNYY

Find maximum of log-likelihood

NR

pR

∂pR∂pr

+NG

pG

∂pG∂pr− NY

pY

(∂pR∂pr

+∂pG∂pr

)= 0

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 13 / 25

Our Experiment: MethodImage Analysis

Otsu’s Method–histogram-based thresholdingI Finds optimum value in a bi-modal histogram such that the intra-class

variance is minimal.

Figure: Initial Intensity Histogram

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 14 / 25

Our Experiment: MethodImage Analysis

Modified Otsu’s Method to take in the original histogramUsed this second threshold to identify nuclei within spores

Figure: Normalized Intensity Histogram

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 15 / 25

Thresholding

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 16 / 25

Thresholding

Figure: Otsu’s thresholding

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 17 / 25

Our Experiment: Results

Figure: pr values along race tube

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 18 / 25

Our Experiment: Results

Figure: pr values along race tube

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 19 / 25

Our Experiment: Results

Figure: pr values along race tube

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 20 / 25

Our Experiment: Results

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 21 / 25

Our Experiment: Results

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 22 / 25

Our Experiment: Results

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 23 / 25

Our Experiment: Results

Interpretation of pr Fluctuations

Moran process

Effect of shaking and accelerated growth

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 24 / 25

Our Experiment: Future Work

Improve pr fittingI Multinomial distribution with dependency parameters

Segmentation issuesI ClumpingI Illumination/Saturation sensitivity

D. Bowers, T. Caldwell, F. Grogan, L. Yamada (REU) Mycofluidics August 8th, 2012 25 / 25