report1.tex

% -*- mode: LaTeX; LaTeX-command: "latex -shell-escape" -*-
\documentclass[12pt]{article}

\usepackage[cm]{fullpage}
\usepackage{authblk}            % For affil{}.
\usepackage{datatool}           % Import csv file as table.
\usepackage{natbib}             % For bibliography{}.
\usepackage[nottoc,numbib]{tocbibind} % Enumerate bibliography section.
\usepackage{url}                % For processing url{} in .bib file.
\usepackage{graphicx}           % Figures.
\usepackage{minted}             % Python source code with highlighting.
\usepackage{listings}           % MM cfg source code.
\usepackage{epstopdf}           % Convert .TIF to .PNG
\epstopdfDeclareGraphicsRule{.tif}{png}{.png}{convert #1 \OutputFile}
\AppendGraphicsExtensions{.tif}

\title{Effects of adhesion and myosin II on \emph{Dictyostelium discoideum} proliferation and division}
\author[1]{Pariksheet Nanda}
\author[2]{Dinkar Ahuja}
\affil[1]{Department of Biomedical Engineering}
\affil[2]{Department of Molecular and Cellular Biology}
\date{February 20, 2015}

\begin{document}

\maketitle{}

\begin{abstract}
  This study considers the case of cell proliferatation separately from cell division.
  \emph{Dictyostelium discoideum} proliferate at the same rate, both, in suspension, and in conventional membrane adhesion allowing environments.
  Cells in suspension, suffer difficulty cleaving their cell membrane.
  Thus, they form multi-nucleated cells and, presumably, also replicate other cellular machinery, to exist as multi-unit, membrane fused cells.
  Myosin II does not affect proliferation, but its absence slows the rate of membrane cleavage by ten times or more.
\end{abstract}

\section{Introduction}
\label{sec:intro}

\subsection{Background}
\label{sec:bg}
\emph{Dictyostelium discoideum} proliferation depends on external environmental conditions, such as the availability of nutrition, and adhesion sites.
Adhesion is necessary for the cell membrane to create fission forces for division.
As the cells grown, in response to dimishing nutrition and space, they change behavior as they sense their density \citep{DGD:DGD1248} and have to endure the burden of their own waste products.

Left alone in these dimishing conditions, the cells eventually reach starvation which causes them collectively to organize to move to another location in search of food.
They aggregate to form slugs, which, in the wild, would migrate to a sunlit area.
Some of the cells of the slug sacrifice themselves as the support stalk, while other cells mature into a fruiting body.
These ambitious, expeditionary cells get dispersed as spores to, hopefully, continue the cycle of life.

Returning back to ideal growth conditions; cultures of these amoebae, or any cell line for that matter follow a characteristic semi-log growth curve.
Under ideal conditions, each of these amoeba cells are known to divide once every 12 hours.
Graphing cell density, during ideal conditions, on a base-2 logarithmic axis, yields a linear growth region, which is known as the \emph{log phase} of cell growth.
The ``semi-'' in the semi-log term mentioned earlier is in reference to the contrasting, linear time axis.

Observing this ideal log phase region can be accomplished by measuring cell density by counting cells from a known volume of culture, and then charting the length during which linearity is preserved.

Adhesion surface vailability can be reduced by maintaining the cells in suspension.
Normally, these amoebae start settling onto the bottom of a petri dish within a minute, and completely settle down in about 30-40 minutes.
When a cell settles, it more or less immediately starts crawling along the bottom surface of the dish.

Finally, an internal cell organelle used to create fission force is the cytoskeleton.
A \emph{myosin II} null mutant presents the opportunity to observe how weaker intracellular forces affect cell behavior and growth rates.

\subsection{Glossary}
\label{sec:glossary}
\begin{description}
\item[Adhesion]
  Attraction between dissimilar molecules.
  Contrasts with cohesion, where the attraction is between identical molecules.
\item[Cleavage]
  Furrow in the cytoplasm membrane during the beginning of cytokinesis.
\item[Cytokinesis]
  The final step in cell division where the cytoplasm and cell membrane divide to form two daughter cells.
\item[Cytoskeleton]
  Protein fiber network in the cytoplasm involved in cell shape, support, locomotion, and transport of materials within the cell.
\item[Dictyostelium discoideum]
  An amoeba used as a model organism.
  It is haploid and most often reproduces asexually by mitosis, but can also reproduce sexually by fusing with another cell for meiosis.
\item[Myosin II]
  Motor protein which uses ATP energy to move actin filaments for contraction, cytokinesis, and vesicle transport.
\end{description}

\subsection{Hypothesis}
\label{sec:hypo}
\begin{itemize}
\item Cells in suspension will not be able to divide easily, because they lack surface adhesion to generate the necessary cytokinetic forces.
\item Cells without myosin II will grow slower than normal cells with myosin II, since the motor protein plays a role in cytokinesis.
\end{itemize}

\subsection{Null Hypothesis}
\label{sec:nullhypo}
\begin{itemize}
\item Cells in suspension will grow at the same rate as cells on plates.
\item Cells without myosin II will grow at the same rate as normal cells with the protein.
\end{itemize}

\subsection{Objective}
\label{sec:exp}
Quantify the growth by culturing two cell lines: NC4A2, and HK321 cells, the latter being the myosin II null mutant.
Grow one set, of these two cell lines, in bacterial shakers to simulate suspension, and another set on 60 millimeter plates to be considered as the normal adhesion reference.

\section{Results}
\label{sec:res}
The growth rates in all 4 cases were similar, as seen in Fig.~\ref{fig:logphase}, and close to the textbook 12 hour division rate of \emph{Dictyostelium discoideum}.

\DTLloaddb[omitlines=8,noheader,keys={Hours,Plate HK321,Plate NC4A2,Flask HK321,Flask NC4A2}]{exp3}{data/exp3.csv}
\begin{table}\centering
  \caption{Cell counts (10,000 cells per ml)}
  \DTLdisplaydb{exp3} 
  \label{tbl:counts}
\end{table}

\begin{figure}\centering
  \includegraphics{fig/exp3_plot}
  \caption{Log phase analysis of cell growth, between cultures in shaker flasks and growth plates.
The unexpected growth seen in flasks was due to a systematic measurement error of counting nuclei as individual cells instead of counting nuclei clusters as single cells.}
\label{fig:logphase}
\end{figure}

As expected, on the plate, the NC4A2 culture doubled every 11.9 hours, and the HK321 culture grew slower; doubling every 12.3 hours.
The flask growth rates of both, NC4A2 and HK321 cell lines, were nearly identical at 12.7 and 12.6 hours respectively.

The flask suspension cells qualitatively differed in their growth patterns, both, from the plate cultures, and between cell types.
The NC4A2 cells grown in suspension would form relatively small clusters of 5 nuclei or less, whereas the HK321 in suspension would form very large 3-dimensional clusters.
The plate cells did not noticably cluster.

Neither the NC4A2, nor the HK321 cells should not have been able to divide in suspension due to an adhesion surface missing.
The video micrograph recording of fluorescent nuclei shows a single undivided HK321 cell with 4 nuclei \citep[Cell Growth]{iweb}, indicating its ability to grow and duplicate its DNA, and its inability to divide.
Frame 4 of that video has been cropped and adjusted for contrast in Fig.~\ref{fig:hk321-4nuclei}.
This observation supports the statement that HK321 cells would have difficulty dividing in suspension.

\begin{figure}\centering
  \includegraphics{data/exp3_hk321-4nuclei.tif}
  \caption{Fluorescence image of 4 nuclei in a single undivided HK321 cell from the flask suspension (Left).
Phase image of surrounding mass of undivided cells (Right).
The image set was captured with the 20x objective, 2x2 CCD binning, and exposures of 500 ms and 20 ms, respectively.}
\label{fig:hk321-4nuclei}
\end{figure}

One expects the HK321 missing the myosin II protein function to suffered from impeded cytokinesis.
To observe this, the division mechaism for both the NC4A2 (Fig.~\ref{fig:nc4a2-division}) and HK321 (Fig.~\ref{fig:nc4a2-division}) cells were recorded by video microscopy \citep[Nuclear Staining]{iweb}.

\begin{figure}\centering
  \includegraphics{data/exp3_nc4a2-division.tif}
  \caption{Division of NC4A2 cell cultured in the petri dish, right before membrane starts to cleave (Left) and the resulting daughter cells (Right).
The time between images was 3 minutes and 20 seconds.
The timelapse image set was captured with the 20x objective, 1x1 CCD binning, and an exposure of 30 ms.}
\label{fig:nc4a2-division}
\end{figure}

\begin{figure}\centering
  \includegraphics[scale=0.5]{data/exp3_hk321-division.tif}
  \caption{Timelapse of division of HK321 cell cluster cultured in the petri dish over 1.5 hours.
Unlike the NC4A2 cells, these cells do not divide easily and have to stretch tens of microns away before the cell membrane ``snaps''.
The timelapse image set was captured with the 20x objective, 1x1 CCD binning, and an exposure of 35 ms.}
\label{fig:hk321-division}
\end{figure}

\section{Discussion}
\label{sec:dis}
The growth seen in flasks was unexpected.
Cells in flask suspension grew as mutli-nuclei clusters.
As this behavior was not known at the time of data collection, there was a systemic measurement error of counting nuclei as individual cells instead of counting the cluster as a single cell.
Additionally, the Chinese hemocytometer used for cell counting was not capable of producing phase images due to a curved surface in the glass disrupting the phase wavefront.
This brightfield illumination limitation made it difficult to distinguish cell clusters from individual cells.

The clustering observations in the flask cells, together with the fluorescence movies, indicate that both cell types experienced some level of difficulty in cleaving the cell membrane by being in suspension alone.
The larger clustering of the HK321 strain also makes one wonder about whether cells inside the 3-dimensional structure might suffocate and thus limit the maximum size of a cluster.
However, the HK321 maximum count and subsequent drop measured in the flask coincides closely with the NC4A2 strain.
Thus suffocation does not seem to be an issue.

Video recordings of membrane division of plate grown NC4A2 takes a little over 3 minutes.
Fig.~\ref{fig:nc4a2-division} shows such an NC4A2 cell division event.
HK321 cells employ a ``rubberband'' system to finalize division, where cells travel tens of microns away from eachother before the membrane, stretched like a rubberband, snaps to form separate cells.
In some cases, video recording of cell clusters attempting division showed such a membrane cleavage event lasting well over an hour \citep[Nuclear Staining]{iweb}.
Fig.~\ref{fig:hk321-division} shows frames from the video where the cells separate 50 microns, but the membrane remains in tact at 33 minutes.
The cells keep moving about their positions till 1 hour 22 minutes, until finally the membrane snaps at 1 hour 36 mintues.

There are a few outlier measurements in the palte count data, per Table~\ref{tbl:counts}.
Specifically, even during the linear phase there is a bimodal distribution of cell growth, but which was averaged out in the analysis curve fit.
This bimodal distribution may have been caused by uneven trituration, or uneven spread of cells on the hemocytometer.
It is tempting detract from the cell measurement protocol of using 10 ul of liquid from a micro-pipette, and instead use a remaining drop from the tip end of the 10 ml pipette used for tituration.
Later analysis comparing the protocol method with this lazy practise showed a significantly different count results.

\section{Materials and Methods}
\label{sec:matmeth}
HK321 is a myosin II heavy chain null mutant derived from NC4A2 \citep{CM:CM5}.

HL5 media, treated with antibiotics, was used to feed the cells.
The antibiotics protect the media from bacterial, but not fungal, contamination.

Plates used were standard 60 mm plastic cell culture dishes and were seeded with 10 ml of media and cells.
Flasks had 125 ml capacity and seeded with 20 ml of media and cells.
The flask lids were kept loose to avoid suffocating the culture.

Cells were counted twice a day per the protocol in Supplement~\ref{sec:celldens}; in the morning by the first author and in the evening by the second author.
Ideally, samples would have been recorded every 12 hours, but regular sampling deviation was due to two factors.
Some measurements are missing due to inclement weather.
A few measurements are also only a few hours apart since neither author is a ``morning person''.

Data analysis was programmed using a Python script run inside an IPython shell \citep{PER-GRA:2007} on Gentoo GNU/Linux.% Reference these.

Microscopy videos were captured using the Micro-Manager software \citep{umgr,JBM36} version 1.4.19 controlling a Scion CFW-1310M CCD detector on a manual Nikon Ti inverted microscope.
Thorlabs TLED devices provided both, diascopic and epifluorescence, illumination modalities.
An Arduino microcontroller, controlled by Micro-Manager, TTL triggered the LEDs.
The configuration file is included in Supplement \ref{sec:cfg}.

This manuscript was prepared using \LaTeX{} with the GNU/Emacs editor and its accompanying AUCTex package.
All source files are freely available on GitHub (\url{https://github.uconn.edu/pan14001/cell-bio-lab-2015}), and have been released into the public domain.
While the website has the most up-to-date file versions, the script used to generate the graph has also been included in Supplement~\ref{sec:code}.
Future reports will explore better Literate Programming \citep{KNUTH:LP:1992} formats, specifically the IPython notebook.

\bibliographystyle{biochem}
\bibliography{report1}

\section{Supplements}
\label{sec:sup}

\subsection{Protocol to determine cell density}
\label{sec:celldens}
\begin{enumerate}
\item Putting adhered cells back into suspension, requires agitation.
  The agitation method depends on the container; for cells in a glass conical flask, a bacterial shaker is used to keep the cells moving and prevent settling.
  For flasks and test tubes that have been sitting and require agitation:
  \begin{enumerate}
  \item Make sure the container cap is  to screwed on tight.
  \item Press the container on a benchtop vortex mixer for a few seconds at a time.
  \item Release the tension on the screw cap to prevent suffocation.
  \end{enumerate}
  For cells on a plate, we cannot shake the container as nothing secures media from spilling out of the low wall height.
  For growth plates only:
  \begin{enumerate}
  \item Agitate, or \emph{titurate}, plate bottom using 5 ml pipette.
  \item Pipette out a few ml of media using a Pipettman, and release the media back across the bottom of the plate at maximum velocity, but do not empty all liquid as that would blow in bubbles.
  \item Keep the pipette mouth perpendicular and move it around regions of the bottom surface to clear areas of adhered amoebae.
  \item Verify tituration was effective by viewing plate bottom using 10x lens in microscope.
  \item Retiturate until satisfactory.
  \end{enumerate}
\item Pipette 10 ul of titurated solution onto haemocytometer.
\item Count at least 100 cells, or count all etched squares using mechanical counter.
\item Clean the haemocytometer and coverslip with de-ionized water.
  Use benchtop microscope to confirm they all cells have been washed away from the grid areas.
  This inspection is especially important to prevent errors when measuring low cell densities.
\end{enumerate}

\subsection{Source code for Figure \ref{fig:logphase}}
\label{sec:code}
\inputminted[linenos]{python}{fig/exp3_plot.py}

\subsection{Micro-Manager Configuration File}
\label{sec:cfg}
\inputminted[linenos]{python}{data/exp3_scion-tled.cfg}

\end{document}