Mitochondrial alignment in ATP gradients

(1)

Mitochondrial alignment in ATP gradients

A study of mitochondrial movements and distribution in polarized cells

Minna Green-Petersen

Master’s degree project in Biomedical physics Section of Cell Physics, Department of Applied Physics

KTH Royal Institute of Technology Science for Life Laboratory (SciLifeLab)

Supervisor: Miroslav Huliciak, PhD Examiner: Professor Hjalmar Brismar

TRITA-FYS 2016:50 ISSN 0280-316X ISRN KTH/FYS/–16:50—SE

(2)

Abstract

In this thesis project mitochondria in polarized cells were to be investigated.

The hypothesis was that mitochondria in polarized cells, which have high energy consumption at one side of the cell, will move to be close to the area of energy demand. This is a largely unexplored part of mitochondrial function connected to the question of how cells coordinate energy production and consumption. The first part of the project involved designing and optimizing the experimental setup, such as cell culture, cell transfection and imaging.

The thesis discusses these aspects in detail. After that results of experiments on Calu-3 and proximal tubule cells (PTCs) from rat kidneys are presented and discussed.

The most noteworthy result of the study is that of optical flow measurements of mitochondria in PTCs grown on membrane and on glass. Optical flow is a method used to measure movements in images, commonly in computer vision, that was used here to quantify the movements of mitochondria.

Polarized cells grown on membranes were compared to unpolarized cells on glass and the result of the optical flow was v_membrane = 119 nm/s and v_glass = 116 nm/s respectively. Mitochondria in polarized cells were thus faster than those in unpolarized cells and the difference was statistically significant with p = 0.042. This result was unexpected based on the initial hypothesis that mitochondria in polarized cells would be strictly located in one area.

The discussions of experimental setup and the end result of optical flow in this study could serve as a background for additional studies of mitochondrial movement in relation to cell energy consumption.

(3)

I den här studien undersöktes mitokondrier i polariserade celler. Hypotesen var att mitokondrier i polariserade celler, som har ojämn energiförbrukn- ing, flyttar sig närmare omr˚adet med hög energikonsumtion. Det är en till stora delar outforskad del av mitokondriens funktion som hör till den större fr˚ageställningen om hur cellen koordinerar produktion och konsumtion av energi. Till en början handlade projektet om att designa och optimera experimentet, p˚abörja en cellkultur, transfektera cellerna och uppställningen för mikroskopi och en stor del av uppsatsen redovisar de stegen. Resul- tat av experiment p˚a Calu-3 celler och celler fr˚an proximala tubuli i r˚attor presenteras och diskuteras.

Det viktigaste resultatet är mätningar av optiskt flöde i celler fr˚an proximala tubuli som vuxit p˚a membran och p˚a glas. Mätning av optiskt flöde är en metod som beräknar rörelse i bilder och vanligtvis används i datorseende, som här användes för att bestämma rörelsen hos mitokondrier. Polaris- erade celler som vuxit p˚a membran hade mitokondrier med optiskt flöde vmembran= 119 nm/s och de p˚a glas hade vglas= 116 nm/s. Allts˚a rörde sig mitokondrier i polariserade celler snabbare än de i opolariserade celler, och resultatet var statistiskt signifikant med p = 0.042. Resultatet var oväntat med utg˚angspunkt fr˚an den initiala hypotesen, att mitokondrier i polariserade celler skulle vara lokaliserade i ett omr˚ade.

Diskussionen kring designen av experimentet och resultaten av optiskt flöde i det här projektet är en startpunkt för fler studier av mitokondriers rörelse i relation till cellens energikonsumtion.

(4)

ii

Acknowledgements

I would like to thank Hjalmar Brismar for the opportunity to come and work at his lab with this exciting project, and for many ideas and suggestions during the course of it. I also want to thank Miroslav Huliciak for supervising me and teaching me lab work, Ying Fu and Huijan Yin for teaching me how to manage Calu-3 cell cultures and make TEER measurements, Jacopo Fontana, Liang Zhang and Linn´ea Nilsson for helping me with proximal tubule cell cultures and Kristoffer Bernhem for letting me use his code for optical flow calculations.

(5)

ADP Adenosine diphosphate.

ATP Adenosine triphosphate.

CyPet Cyan fluorescent protein.

EMEM Eagle’s minimum essential medium.

FBS Fetal bovine serum.

FWHM Full width at half maximum.

GFP Green fluorescent protein.

MTS Mitochondrial targeting sequence.

OXPHOS Oxidative phosphorylation.

P/S Penicillin/Streptomyacin.

PBS Phosphate-buffered saline.

PTC Proximal tubule cell.

RFP Red fluorescent protein.

TEER Transepithelial electrical resistance.

iii

(6)

Chapter 1 Introduction

This thesis concerns the mitochondrion, the organelle most known as the powerhouse of the cell, as it produces energy in the form of ATP. The mitochondrion also has other important functions and have been a hot topic for researchers in recent years, but for this project the mitochondrions function as energy supplier and distributer was to be investigated. In biology text books mitochondria are drawn as round or elliptical discrete units randomly distributed in the cytosol. In reality they exist both as smaller separate units and larger connected networks of many elongated mitochondria that can transfer membrane potential between each other. These networks of mitochondria are dynamic and change shape as a result of fusion and fission of individual mitochondria. A lot of research has focused on the mitochondrial networks seen in muscle, also called the mitochondrial reticulum [1],[2],[3].

These networks are believed to work as a power transmitting system to distribute energy quickly from where it is produced to where it is used in the cell. The advantage of a structure like that is obvious since muscle cells use a lot of energy in a short amount of time.

Another cell type that has special energy requirements are polarized cells in epithelium. This is a type of cell that form thin layers and transport substances from one side of the cell to the other. To make this possible the cell sets up ion gradients to drive transport of solutes against their concentration gradients. This is done with the ion pump Na⁺/K⁺− ATPase (NKA) which pumps 3 sodium ions out of the cell and 2 potassium ions into the cell while consuming 1 molecule of ATP. NKA is strictly located to one side of the cell membrane in polarized cells, which gives a high need for ATP at this side of the cell.

In the cell, ATP can be produced either by the mitochondria in oxidative phosphorylation or directly in the cytosol by glycolysis. After ATP is produced it diffuses, eventually reaching the point where it is needed. The hypothesis that this thesis is based on was that in polarized cells, where the energy consumption is high locally at one side of the cell, mitochondria will

1

(9)

relocate closer to the NKA to provide energy where it is needed. In a larger perspective the aim was to investigate how the distribution of energy works in the cell, and how the cell manages varying energy needs.

(10)

Chapter 2 Background

2.1 Energy generation in cells

Cells need energy to perform tasks like synthesizing new proteins, replicating DNA, dividing and serving cell type specific functions. A lot of the energy needed by a cell is used to maintain ion gradients over the cell membrane.

These ion gradients are used by the cell to transport nutrients, building blocks and waste products in and out of the cell. Energy is derived through breakdown of food molecules, which the cell can store in the form of high energy bonds in molecules called activated carriers. The most useful activated carrier in the cell is ATP (adenosine triphosphate), which stores energy in the form of a phosphate bond. The energy of the phosphate bond is used by splitting ATP into ADP (adenosine diphosphate) and inorganic phosphate, and the opposite reaction is catalyzed in ATP synthesis

ADP + P_i *) ATP.

ATP can be generated directly from breakdown of glucose in a process called glycolysis, but it can also be generated using an intermediate activated carrier through oxidative phosphorylation. Glycolysis takes place in the cytosol of the cell whereas oxidative phosphorylation takes place in the mitochondria. Oxidative phosphorylation is the more efficient method of breakdown since it generates more molecules of ATP per molecule of glucose, but both methods have advantages and are used in cells more or less depending on the energy demands of that cell [7].

2.1.1 Glycolysis

The first way for cells to generate ATP is through glycolysis. In glycolysis one glucose molecule is split into two pyruvate molecules while generating two types of activated carriers in the process. The reaction takes place in the cytosol without the involvement of oxygen and is the energy generating

3

(11)

pathway in anaerobic bacteria. At the start of the reaction two molecules of ATP are consumed, but at the end the outcome is four ATP molecules and two NADH molecules, so the net result is two ATP molecules and two NADH for each glucose molecule that is broken down.

The pyruvate molecules that remain after glycolysis are transported into the mitochondria where they are broken down further to release energy that is used to drive oxidative phosphorylation [7].

2.1.2 Mitochondria and oxidative phosphorylation

As stated above the mitochondrion is the organelle in the cell where ATP synthesis takes place in the process oxidative phosphorylation, also called OXPHOS. The mitochondrion is a double membrane organelle, it has an outer and inner membrane which create two separate spaces, the intermembrane space between the membranes and the matrix inside the inner membrane, see figure 2.1. The outer membrane contains pores big enough to let through molecules up to 5 kDA in size, so it is permeable to ions and other solutes, even smaller proteins. The inner membrane on the other hand is impermeable to ions and small solutes, which allows an electrochemical gradient to be established over it. It is this gradient that is used for ATP synthesis. The inner membrane is very folded, forming cristae, which increases the area of membrane that can be used for ATP production. Mitochondria can exist both as small separate units in the cell or form large networks of tubular mitochondria. These networks change continuously as a result of fusion and fission of individual mitochondria.

Outer membrane

Matrix

Inner membrane Intermembrane space

Figure 2.1: The structure of the mitochondrion.

Under most conditions the majority of ATP in cells is generated by oxidative phosphorylation. The pyruvate that is left after glycolysis in the cytosol is transported into the matrix of the mitochondrion where is is broken

(12)

2.2. POLARIZED CELLS 5

down further in the citric acid cycle. The breakdown of pyruvate produces the activated carriers NADH and FADH2, that store energy in the form of high energy electrons. In the next step of oxidative phosphorylation the high energy electrons are passed through the electron-transport chain. This is a number of proteins in the inner membrane of the mitochondrion that the electrons are passed along, gradually losing energy while hydrogen ions are pumped out of the matrix and into the intermembrane space. The energy of the electrons is thus transformed into an ion gradient of hydrogen, which is used to drive the synthesis of ATP by the protein with the fitting name ATP synthase. ATP synthase lets hydrogen ions flow back into the matrix while using the energy to form ATP from ADP and inorganic phosphate.

The ATP produced in the matrix is transported out via an antiporter that exchanges ATP for ADP, also utilizing the hydrogen ion gradient over the inner membrane. The net result of ATP in oxidative phosphorylation is 30 molecules of ATP per molecule of glucose [7].

2.2 Polarized cells

The cell type of interest in this project was polarized epithelium, which is a cell type that form layers that act as a barrier between two environments and transport molecules vectorially across the cell layer. These cells exist in a few places in the body, for example the kidney, where they serve the important function of concentrating and determining the composition of urine.

Polarized epithelial cells are distinguished by an apical and basolateral side of the cell membrane, separated by tight junctions, see figure 2.2. The tight junctions run all along the circumference of the cell membrane and prevent solutes from passing freely from one side of the cell to the other, but they also work as a divider of the cell membrane. Tight junctions separate the apical and basolateral part of the membrane so that no membrane proteins can diffuse between them. This allows different compositions of membrane proteins on the different sides of the cell and is what gives polarized epithelial cells the ability to transport molecules across cell layers. The NKA is one of the proteins that is located only in the basolateral membrane in polarized epithelial cells [7].

2.2.1 Polarized epithelium in the kidney

The kidney has several functions, the main one being to filtrate the blood of toxins and waste products. It consists of smaller units called nephrons and a human kidney contains about 10⁶ nephrons. Each nephron consist of the malpighian body and the tubule. The malpighian body is where the blood is filtered. It consists of a web of capillaries, the glomeruli, enclosed in a double walled capsule called Bowman’s capsule. The capillaries of the glomerulus

(13)

Tight junctions

Na+/K+-ATPase

Apical membrane

Basolateral membrane

Figure 2.2: The figure shows a schematic of polarized cells, with the apical and basolateral membrane separated by tight junctions and the sodium- potassium pumps at the basolateral membrane.

are covered with endothelium that contains 50 − 100 nm sized pores. The pressure in the glomeruli is high which causes water and solutes to leak from the blood in the glomeruli into the capsular space between the glomerulus and Bowman’s capsule, this first filtered liquid is called the primary urine.

The primary urine that is filtered to the capsular space leaves the capsule and enters the tubule where it is processed further. Most of the water in the primary urine is reabsorbed, and some substances are reabsorbed while others are secreted to form the final urine, that is excreted. It is this part of the nephron that utilizes polarized cells.

The first part of the tubule is called the proximal tubule, here salts, proteins and glucose are reabsorbed. Proximal tubule cells are polarized to allow vectorial transport across the cell layer. The side facing the lumen, in this case the urine, is the apical side and the side facing the blood is the basolateral side. The cell uses the gradient of sodium over the cell membrane as well as the membrane potential, which is negative inside, to drive transport from the lumen and into the cell. The sodium that enters the cell is transported out of the cell on the basal side via NKA, to maintain the sodium gradient over the membrane [8].

2.2.2 The role of the sodium-potassium pump

The NKA is one of the most common ion pumps in the body, present in all cells and its function is to maintain the membrane potential and concentration gradients of sodium and potassium over the cell membrane. The concentration of these two ions intracellularly and extracellularly are opposite and typical values for mammalian cells can be seen in table 2.1.

NKA transports two potassium ions into and three sodium ions out of the cell while consuming the energy of one molecule of ATP. Both ions travel against their concentration gradients. Since it uses the energy stored

(14)

2.2. POLARIZED CELLS 7

Ion Intracellular concentration [mM] Extracellular concentration [mM]

Sodium 1-15 145

Potassium 140 5

Table 2.1: Typical concentrations of sodium and potassium in mammalian cells.

in ATP directly, NKA is an active transporter. The sodium and potassium gradients established by the pump can be used by other membrane proteins to transport energy and building blocks into the cell and waste products out of the cell, or to set up other ion gradients in secondary and tertiary transport processes. NKA is thus responsible for a big part of the cells transport processes and consumes a lot of the energy required by the cell.

The way NKA works can be described as a six state model, as in figure 2.3 and described below.

1) The pump has high affinity to sodium and binds three sodium ions on the intracellular side of the membrane.

2) The pump is phosphorylated by one molecule of ATP, which becomes ADP.

3) The phosphorylation causes a conformal change so that the sodium ions are exposed to the extracellular side of the cell, it also changes the affinities of the ions so that the sodium ions dissociate from the pump, and the affinity for potassium ions is increased.

4) Two potassium ions bind to the pump on the extracellular side.

5) The pump is dephosphorylated.

6) The conformation changes back to expose the potassium ions to the intracellular side and the affinity for potassium changes to lower again, so the potassium ions dissociate. At the same time the affinity of sodium is higher and the process can start over again.

The turnover rate of this process depends on the intra- and extracellular concentrations of sodium and potassium, the intracellular concentration of ATP and the dissociation constants of all of the involved solutes [8], [9], [10].

(15)

P

Intracellular space

Intracellular space Extracellular space

Extracellular space

ATP ADP

K⁺

Na⁺

P P P

1 2

4 3 5

6

Figure 2.3: A schematic of the six state model of NKA. 1) Three sodium ions bind to the pump on the intracellular side of the membrane, 2) the pump is phosphorylated by the transfer of the phosphate group of one molecule of ATP, which becomes ADP, 3) the pump changes conformational state so that it faces the extracellular space and the sodium ions dissociate, 4) two potassium ions bind to the pump, 5) the pump is dephosphorylated, 6) dephosphorylation leads to the conformational change back to facing the intracellular space, where the potassium ions dissociate.

(16)

Chapter 3 Materials and methods

3.1 Cell culture

The first part of this project was to establish a polarized cell culture. A few cell types form polarized monolayers when cultured on semipermeable membranes. Transwell 6-well, 24 mm polyester membranes treated with poly- carbonate and with pore diameter 0.4µm (Corning NY) was used. These consist of an insert with the membrane that is lowered into a well, see figure 3.1. Cells are seeded onto the membrane, with direct access to the upper compartment and holes on the side of the insert that give access to the lower compartment. Two cell types were used for the experiments in this project, initially Calu-3, and later proximal tubule cells from rat.

For comparison with unpolarized conditions, cells were seeded onto 18 mm glasses, in 12-well plates.

Polyester membrane

Lower compartment Upper

compartment

Figure 3.1: Insert with a polyester membrane that separates the upper and lower compartment.

3.1.1 Calu-3 cells

Calu-3 is a human airway epithelial carcinoma cell line. Calu-3 was pur- chased from ATCC and stored frozen, thawed and seeded in petri dishes

9

(17)

before seeding on membranes. Protocols were adapted from [4] and [5].

3.1.2 Proximal tubule cells

Proximal tubule cells (PTCs), were taken from the outer layer of the rat kidney. The tissue was treated to separate the PTCs from the rest of the structures in the outer kidney. The result is a solution that contains the PTCs, glomeruli and other cell debris. The cells were seeded by allowing the cell debris to settle at the bottom of a tube, pipetting PTC cell suspension and seeding directly on the membrane. PTCs grow quickly but dedifferentiate into fibroblasts and are to be used for experiments by day 3-4 after seeding.

3.2 Transfection of mitochondria

To label mitochondria for fluorescent microscopy the cells were transfected using a mitochondrial targeting sequence (MTS) connected to green fluorescent protein (GFP). Transfection was done with Lipofectamine LTX with PLUS reagent (Invitrogen) and Opti-MEM (Life technologies). The protocol was adapted to cell type and well size according to table 3.1.

Membrane, 6 well

Cell type Medium OPTI-MEM DNA Lipid reagent Plus reagent

Calu-3 2 ml 2 × 100µl 2.5µg 10µl 2.5µl

PTC 2 ml 2 × 100µl 2.5µg 5µl 2.5µl

Glass, 12 well

Calu-3 1 ml 2 × 50µl 1µg 5µl 1µl

PTC 1 ml 2 × 50µl 1µg 2µl 1µl

Table 3.1: Table of the most common transfections in this project.

3.3 Cell polarization measurements

The next step in the project was to evaluate the polarization of the cells, since this was important for the hypothesis. The polarization of cells result in a few characteristics that can be measured. One method that is commonly used to check the formation of tight junctions is to measure the transepithelial electrical resistance (TEER). This has been done on Calu-3 cultures by others in this lab [4]. The other idea was to investigate the localization of NKAs, since this is of interest and also an indicator of polarization if the pumps are located only on the basolateral side of the cell membrane.

(18)

3.4. TURNOVER RATE OF NKA 11

3.3.1 TEER measurements

TEER is used to determine the formation and integrity of tight junctions in a monolayer of cells. For the measurements a pair of chopstick STX2 electrodes, together with an EVOM2 volt/ohm meter was used. To measure the electrodes are placed with one stick in the upper compartment and one in the lower compartment and the electrical resistance over the membrane can be read out on the EVOM2.

For Calu-3 TEER was not measured since there was a good source of the development of TEER on identical cultures [4]. For PTCs TEER was measured from the day after seeding and for a total of four days.

3.3.2 Cotransfection of mitochondria and NKA

To label NKAs as well as mitochondria cells were cotransfected with MTS- GFP and a fluorescent protein coupled to the NKA α1 subunit. Three different constructs were used, red fluorescent protein (RFP) coupled to human NKA α1 subunit, CyPet (cyan fluorescent protein) coupled to rat α1 subunit and mCerulean3 (cyan fluorescent protein) coupled to human α1

subunit. Cotransfection was done the same way as transfection with only MTS, with the difference that the DNA was half of each construct.

3.4 Turnover rate of NKA

Part of the hypothesis of this project was that mitochondria will be located closer to the NKAs when the pump is in need of more ATP. The turnover rate depends on both the intra- and extracellular concentration of sodium and potassium. The idea was thus to lower the basolateral concentration of sodium to increase the consumption of ATP by the pump. To avoid any osmotic effects the sodium had to be replaced by a solute that exerts the same osmotic pressure.

3.5 Imaging

A few problems arise when imaging cells grown on a membrane. The sim- plest way to image living cells is to take the insert out and place it on a cover glass of larger diameter than the bottom of the insert. The cells can then be imaged in an inverted, confocal microscope through the cover glass and the membrane. A problem with this approach is that you have almost no liquid on the basal side of the cells and thus cannot do any experiments with the medium composition on this side of the cells.

To image cells in an inverted microscope but with basal medium and the option to change this medium during the experiment a chamber had to be designed. The challenge was to have enough space underneath the

(19)

membrane to have liquid that can flow there and at the same time focus on the cells with an objective. A specialized long distance objective or one with lower magnification which also has a longer working distance is needed in this setup. The chamber that was designed for this purpose was based on ones made in other studies [12], [13]. It has the insert resting on a raised edge with the cover glass glued to the underside of the chamber. The space between the membrane and glass was 0.15 mm, with access to change medium through holes on the side of the chamber. This chamber proved to work in imaging cells with a 40× water objective but there was not enough time to use it on polarized cells.

The next option is to use an upright microscope, where the objective is dipped into the liquid in the upper compartment of the insert. This allows live imaging, the option to have basal medium and to change this medium.

The cells can also be fixed and mounted on a cover slip. This allows imaging from the cell side and gives nice images. The disadvantages are that the cells are dead so that you cannot study changes and behaviors of one cell over time and that the fixing can disturb the cell structure so that what you see is not representative of the living cell.

Objective Glass

Figure 3.2: Chamber for perfusion experiments. With this holder for the inserts cells can be imaged with an inverted microscope while controlling the medium on the basal side of the cells.

3.5.1 Z-stacks

To investigate the distribution of mitochondria z-stack images were taken, which means several images with a shift in focus in the z-direction, see figure 3.3 for an example. One idea to investigate the mitochondria was to plot the intensity profile of such a stack in the z-direction, see figure 3.4. This gives an idea of how the mitochondria are distributed in the cell, if they tend to cluster to one side of the cell, and how wide their extent is. The full width at half maximum, FWHM, was calculated for these distributions to quantify the shape further.

(20)

3.5. IMAGING 13

z = 1.48μm z = 1.85μm z = 2.22μm

z = 2.59μm z = 2.96μm z = 2.96μm

Figure 3.3: Part of a z-stack of images of a fixed Calu-3 sample.

0 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Depth [µm]

0 1000 2000 3000 4000 5000 6000 7000

Intensity [a.u.]

FWHM

Figure 3.4: The intensity profile corresponding to the z-stack above with the FWHM marked.

(21)

3.6 Optical flow

The next idea for the project was to compare the movements of mitochondria in polarized and unpolarized cells and draw conclusions based on that. The optical flow is a measure of the apparent movement of a brightness pattern in an image in relation to the viewer. It can be used to discern objects from each other and to calculate their velocities and is used in for example computer vision. Here it was used to quantify the movement of mitochondria in living cells grown on membrane and glass.

3.6.1 Theory

The calculation of optical flow is based on the assumption that the brightness of a point in an object does not change in time. Calling the brightness at this point E(x, y, t) this means that

dE dt = 0,

which can be expanded with the chain rule of differentiation into δE

δx dx

dt +δE δy

dy dt +δE

δt = 0.

Simplifying this expression by introducing the velocities u = ^dx_dt and v = ^dy_dt and the partial derivatives Ex, Ey and Et means that this equation can be written

Exu + Eyv + Et= 0.

The partial derivatives can be estimated from a series of images which leaves two unknowns, u and v. One additional equation is needed to estimate these velocities. There are several different methods that can be used to solve this problem but for this project the Horn-Schunck method was used, which assumes that the brightness pattern varies smoothly in the image. This constraint can be met by minimizing the square of the magnitude of the optical flow velocity,

E_smoothness² = δu δx

2

+ δu δy

2

+ δv δx

2

+ δv δy

2

.

Along with minimizing

E_brightness= E_xu + E_yv + E_t this gives the velocities u, v [11].

(22)

3.6. OPTICAL FLOW 15

3.6.2 Measurements and calculation

The calculation of optical flow can be done in MATLAB using the Computer vision system toolbox, were the default method for the calculation is Horn- Schunck described above. To calculate the optical flow a time series of images were taken with the interval 5 − 10 seconds, for a total of 10 minutes.

To compensate for the cells drifting out of focus over the course of the imaging a small z-stack of images were taken for each frame with 0.8 − 1µm between the images in the z-direction, see figure 3.5. These three images were merged into one as a z-projection of the maximal intensity in each pixel, basically locating the mitochondria in the z-stack at each time point and using only those pixels to form the image.

Figure 3.5: A z-stack of images with slightly different focus points were made into one image to compensate for drift during imaging.

(23)

Results

4.1 Calu-3

The initial experiments were conducted on Calu-3 cultures. Calu-3 grow slowly and form a monolayer in about 5 days and the tight junctions have formed properly after 10-14 days. At the same time they proved difficult to transfect, with transfection efficiency decreasing with number of days after seeding. It was found that they could be transfected between day 5-7 after seeding but not after this point. At best the transfection efficiency would be around 10 %.

4.1.1 Low-sodium experiments

To do experiments with low sodium medium in the lower compartment of the inserts a cell medium of only the salts, HEPES (biological buffer) and glucose of EMEM (Eagle’s minimum essential medium) was prepared. Sodium was replaced by calcium by replacing NaCl with CaCl₂ to give the same osmolarity. Five different concentrations of sodium of this medium was prepared, one with the sodium concentration of the normal medium, which had 134mM, and four with lower sodium concentrations, 60mM, 80mM, 100mM and 120mM, see appendix for formulations.

Calu-3 cells were transfected on day 5 and on day 6 the medium was changed to the minimum medium with normal sodium in the apical compartments, and one insert with each of the lower sodium concentrations in the lower compartment. The last insert was left with normal EMEM in both compartments to act as a control. Cells were then left overnight in the incubator and the medium was changed again the next morning, after that the cells were fixed and mounted on coverslips. After fixing and mounting z-stack images were taken of each fixed sample and the intensity in z calculated for each stack. The results can be seen in figure 4.1. These intensity curves have been normalized and centered to make comparison easier. The full width at half maximum (FWHM) was also calculated for each set of

16

(24)

4.1. CALU-3 17

distributions. Unfortunately it is hard to draw any conclusions based on these plots as there is no clear trend in the shapes of the curves or in the FWHM.

(25)

−100 −5 0 5 10 0.2

0.4 0.6 0.8 1 1.2

Depth [µm]

Normalized intensity

FWHM = 4.47 120mM sodium

−100 −5 0 5 10

0.2 0.4 0.6 0.8 1 1.2

Depth [µm]

−100 −5 0 5 10

0.2 0.4 0.6 0.8 1 1.2

Depth [µm]

−100 −5 0 5 10

0.2 0.4 0.6 0.8 1 1.2

Depth [µm]

Figure 4.1: Normalized and centered intensity profiles of mitochondria in Calu-3 after incubation with low-sodium medium overnight.

(26)

4.2. PROXIMAL TUBULE CELLS 19

4.2 Proximal tubule cells

Proximal tubule cells proved easier to transfect than Calu-3, and showed higher transfection efficiency, see figure 4.2. However, it was difficult to seed them at high enough density to get them confluent by day 3-4 after seeding, which was needed to avoid them dedifferentiating to fibroblasts.

4.2.1 TEER

TEER was measured for PTCs from the day after seeding, called day 1 and for a total of 4 days, see figure 4.3. The result for all seeding densities was only slightly above the baseline by day four and under the microscope it could be seen that the cells were still not confluent.

4.2.2 NKA cotransfection

Cells were cotransfected with MTS-GFP and three different kinds of fluorescent NKA α₁ subunit, RFP, CyPet and mCerulean3. In both Calu-3 cells and PTCs mitochondria were transfected and imaged through transfection but NKA was not, see figure 4.4. As can be seen in these pictures the RFP gave almost no signal except for some debris in Calu-3 ((a) and (b)) and the cyan fluorescent proteins gave only what can be assumed to be autofluores- cence from the cells ((c), (d)). As a control cells were also transfected with only NKA, without better results.

4.2.3 Optical flow

The optical flow of mitochondria was calculated for PTCs grown on membrane or glass, 30 cells of each split over 2 membranes and 2 glasses. Time series images were taken on day 3 or 4 after seeding and the confluence was

< 80% for both cases. The results can be seen as a cumulative distribution function in figure 4.5 and as a box plot in figure 4.6. The median optical flow in the two cases was vmembrane = 119 nm/s for the mitochondria on membranes and v_glass = 116 nm/s for the mitochondria on glass. This was a significant difference with p = 0.042, i.e. the probability of this outcome given there is no difference between the two groups. The result was hence that the mitochondria in cells on membranes move faster than those in cells on glass, which was the opposite of what was expected.

(27)

Figure 4.2: Fixed PTCs transfected with MTS-GFP.

Figure 4.3: TEER measurements of PTCs with different seeding volumes on day 1-4 after seeding on membranes.

(28)

4.2. PROXIMAL TUBULE CELLS 21

(a) (b)

(c) (d)

Figure 4.4: Results of cotransfections on Calu-3 and PTCs with MTS-GFP and a NKA fluorescent protein. (a) Calu-3 with RFP, (b) Calu-3 with mCerulean3, (c) PTC with RFP, (d) PTC with CyPet.

(29)

100 105 110 115 120 125 130 135

Optical flow [nm/s]

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

Fraction of cell count

Membrane Glass

Figure 4.5: Cumulative distribution functions of optical flow for mitochondria in cells grown on membrane and glass.

Membrane, 30 cells Glass, 30 cells 100

105 110 115 120 125 130 135

Optical flow [nm/s]

Figure 4.6: Box plot representation of the optical flow data. The median is marked with a red line, the blue box signifies 25-75% of the distribution, black lines show minimum and maximum and red crosses show outliers.

(30)

Chapter 5 Discussion

This project was a study of the movements and distribution of mitochondria in cells with locally high energy consumption. Many aspects of the experimental setup had to be optimized, such as determining the polarization of the cells, transfecting them, and imaging on the membrane that they were grown on. In the end the initial hypothesis has not been shown and instead we have an unexpected result from the optical flow measurements.

Each of the difficulties presented in this project report can be solved with enough time and resources. The first was that Calu-3 cells were difficult to transfect and grew slowly, therefore the change was made to PTCs. PTCs are to be preferred over Calu-3 cells anyway because they are primary cells, and cell lines often rely more on glycolysis for energy production. This means that the mitochondria in Calu-3 might not have much or any ATP production. PTCs are easier to transfect than Calu-3, grow faster and are more suited for the project.

Transfection of NKA as well as mitochondria would be interesting since the dynamic between the two is investigated here. Unfortunately NKA transfection never worked, the reason might be that the DNA constructs were faulty or some mistake in the transfection or imaging, but there is no reason to believe that it cannot be done. The issue of growing PTCs to confluence could be solved by seeding more cells, or use smaller membranes to seed on. You could also make a change to a fully functioning nephron as a model system, and thus avoid having to grow cells to polarization.

The imaging setup was improved by designing the chamber that can also be used for perfusion experiments, alternatively making the change to an upright microscope and dip-in objective.

The design of the perfusion chamber or change to an upright microscope was an important step since these both allow imaging of living cells. It was clear from the imaging of fixed samples and intensity distribution plots that no conclusions could be drawn. The shape of cells vary significantly and it is difficult to see what corresponds to individual cells, which makes

23

(31)

comparison meaningless. If enough individual cells could be imaged it might give information, but z-stack imaging is time consuming and this approach is not ideal for this purpose. A more ideal experiment would instead be to observe one living cell as the activity of the NKA is changed and see the mitochondria respond to this.

In line with the original hypothesis that mitochondria would gather and stay at the NKAs in polarized cells we anticipated that the optical flow would be lower for these than for cells grown on glass. The measurements show the opposite, that the median optical flow for mitochondria on membranes was v_membrane = 119 nm/s and for glass v_glass = 116 nm/s, with strong statistical significance (p = 0.042). This raises the question of what drives the movements of mitochondria. In a different study done in this lab is was shown that high glucose in PTCs slows down mitochondria. This is contributed to a shift in energy production from oxidative phosphorylation to more glycolysis, so the mitochondria slow down when they are less useful.

An important factor for the function of mitochondria is the cytosolic ratio of ATP to ADP, since ATP that is produced in the matrix is transported out of the mitochondria via the ATP/ADP exchanger. If the concentration of ADP is low in the cytosol newly synthesized ATP can not leave the matrix, and no new ADP is transported into the matrix to be phosphorylated by ATP synthase. A new hypothesis based on these results could be that it is the ATP/ADP ratio that drives the movements of mitochondria, so that they move to places where this is low. In the cells grown on membrane the energy consumption is uneven and may be changing whereas the cells on glass might have reached an equilibrium where the ratio is the same everywhere in the cell, giving these results. It is worth to note that the optical flow measurements only measure movement in the x-y-plane, movements in the z-plane would also be interesting to examine.

In the end one can conclude that this was an exploratory study of cell energy metabolism, where much is yet to be understood. The main part of the project consisted of designing and optimizing the experiments needed.

The final result of the optical flow raises several questions and there are many ways to go about to continue studies in this area.

(32)

Chapter 6 Conclusion

In this project mitochondria in polarized cells were investigated. Focus was initially on developing methods to grow polarized cells, determine the polarization, and visualize the mitochondria. A suitable experimental setup was developed and a number of changes to the original plan were made, like using PTCs instead of Calu-3 and designing a chamber for imaging living cells on membranes. The results of the movements of mitochondria calculated as optical flow showed that mitochondria in cells on membranes move faster than those in cells grown on glass. This is presumed to be due to the polarization of the cells on membranes and is a starting point for additional studies of mitochondrial movements in relation to energy consumption in cells.

25

(33)

[1] L. E. Bakeeva, Yu. Chentsov, V. P. Skulachev, Mitochondrial frame- work (reticulum mitochondriale) in rat diaphragm muscle, Biochimit- ica et Biophysica Acta 501, 349-369 (1978).

[2] V. P. Skulachev, Mitochondrial filaments and clusters as intracellular power-transmitting cables, TRENDS in Biochemical Sciences 26, (2001).

[3] B. Glancy, L. M. Hartnell,D. Malide,Z. X. Yu, C. A. Combs, P. S. Con- nelly, S. Subramaniam, R. S.Balaban, Mitochondrial reticulum for cellular energy distribution in muscle, Nature 523, 617-620 (2015).

[4] A. Turdalieva, J. Solandt, N. Shambetova, H. Xu, H. Blom, H. Bris- mar, M. Zelenina, Y.Fu, Bioelectric and morphological response of liquid-covered human airway epithelial Calu-3 cell monolayer to pe- riodic deposition of colloidal 3-mercaptopropionic-acid coated CdSe- CdS/ZnS core-multishell quantum dots, PloS one 11.2, (2016):

e0149915.

[5] J. L. Harcourt, L. M. Haynes, Establishing a liquid-covered culture of polarized human airway epithelial Calu-3 cells to study host cell response to respiratory pathogens in vitro, Journal of visualized experiments: JoVE 72, (2013).

[6] D. G. Nicholls, S. Ferguson, Bioenergetics 4th edition, (Burlington:

Elsevier Science, 2013).

[7] Alberts, Bray, Hopkin, Johnson, Lewis, Raff, Roberts, Walter, Essen- tial cell biology, 4th ed (Garland Science, 2014).

[8] S. Silbernagl, A. Despopoulos, Color atlas of physiology, 6th ed (Thieme publishing group, 2009).

[9] T.F. Weiss, Cellular biophysics, Volume 1: Transport, (The MIT Press, 1996).

[10] W. D. Stein, T. Litman, Channels, carriers and pumps, 2nd ed (Else- vier, 2015).

26

(34)

BIBLIOGRAPHY 27

[11] B. K. P. Horn, B. G. Schunck, Artificial intelligence 17, 185-203 (1981).

[12] D. Olteanu, X. Liu, W. Liu, V. C. Roper, N. Sharma, B. K. Yoder, L. M. Satlin, E. M. Schwiebert, M. O. Bevensee, Increased Na⁺/H⁺ exchanger activity on the apical surface of a cilium-deficient cortical collecting duct principal cell model of polycystic kidney disease, Am J Physiol Renal Physiol 302, C1463-C1451 (2012).

[13] B. J, Siroky, W. B. Ferguson, A. L. Fuson , Y. Xie, A. Fintha, P. Kom- losi, B. K. Yoder, E. M. Schwiebert, L. M. Guay-Woodford, P. D. Bell, Loss of primary cilia results in deregulated and unabated apical calcium entry in ARPKD collecting duct cells, Am J Physiol Renal Phys- iol 290, F1320-F1328 (2006).

(35)

Protocols

Thawing and seeding Calu-3 in petri dishes

1. Get cells from the freezer, keep them on lice until ready to thaw.

2. Thaw the cells by stirring in 37 degrees water bath.

3. When thawed, transfer cells to a tube and add 30ml of warmed medium (500 ml EMEM+50ml FBS+5ml P/S).

4. Centrifuge the cells for 5 min at 1000-1200 rpm.

5. Remove the supernatant from the tube and resuspend the pellet in 5ml EMEM full medium by pipetting up and down.

6. Add the solution to a larger petri dish and add another 5ml of medium.

Gently shake the plate so distribute the cells evenly.

7. Put the dish with cells in the incubator at 37 degrees, 5% CO2.

8. The next day, remove the medium from the petri dish and wash with 10ml of warmed PBS by swirling it in the dish, remove the PBS. Do this washing twice before adding new warm medium. The cells are ready to be reseeded into petri dishes or on membranes in about 5 days after thawing and seeding in the petri dish.

28

(36)

29

Seeding Calu-3 onto membrane from petri dish

1. Check under the microscope that the cells have grown to 70 − 80%

confluence, they are then ready to be seeded.

2. Empty the petri dish of medium. Wash the cells with by pouring 10 ml of PBS over the cells and gently turning the dish back and forth. Then remove the PBS from the dish with a pipette. This is done to wash away all traces of medium, which contains FBS that can inhibit the trypsin that will be added to the dish in the next step.

3. Add 1 ml of trypsin to the dish and turn the dish to distribute it.

Remove the trypsin.

4. Add another 3 ml of trypsin to the dish and put the dish in the incubator for 3 min to allow for the cells to detach from the bottom of the dish.

5. After incubation, pipette the liquid up and down and pour it over the dish to help detach the cells. To suspend the cells evenly in the liquid pipette up and down and make sure there are no visible clumps.

6. Next add 3 ml or more of EMEM medium which will neutralize the trypsin. Again, make sure that no clumps of cells are forming.

7. Transfer most of the liquid into a conical tube but save a small amount, about one drop in an Eppendorf tube for counting the cell density.

8. Centrifuge the conical tube for 5 min at 1400 rpm.

9. Place a drop of cell suspension from the Eppendorf tube on a counting glass (hemacytometer) and place a cover glass on top. Count the cells under a microscope, the total number of cells within one square with a three-line border corresponds to X · 10⁴cells/ml.

10. When the conical tube is ready, remove and discard the supernatant.

Resuspend the cells by adding 6 ml or more EMEM and turning the tube upside down repeatedly until the pellet has dissolved.

11. Fill each well in the Transwell 6-well plate with 1 ml of cell suspension in the apical compartment and 2 ml in the basolateral compartment.

12. Put the Transwell plate in an incubator at 37°and 5 % CO2.

13. The petri dish still contains cells that can be used for further seeding so add 10 ml to the dish and place it in the incubator as well.

(37)

TEER measurements

For the electrodes to be submerged in liquid the inserts have to have 1.5 ml medium in the upper compartment and 2.5 ml in the lower compartment.

Do not change medium sooner than half an hour before measuring since this affects the TEER.

1. Prepare the equipment for TEER measurements, EVOM2 volt/ohm meter, STX2 chopstick electrodes and heating plate by wiping it with ethanol, placing it in the hood and plugging it in.

2. Sterilize a pair of tweezers with ethanol, and sterilize the chopsticks by emerging in ethanol. Wait a few minutes for the heating plate to warm up.

3. Put the plate with cells on the heating plate, wait a few minutes for it to get the same temperature.

4. Make sure the electrodes are dry before starting to make measurements. Take the lid of the plate and lower the electrodes in so that one arm is in the upper compartment and one is in the lower. Wait approximately ten seconds for the value to stabilize on the ohm meter display before noting it.

5. Remove the electrodes and use the tweezers to spin the insert to measure in the next hole. Remember to spin it the same way so that you measure in each hole once.

6. Measure TEER as before and repeat for the three holes in an insert.

7. After finishing an insert, put the lid back on and sterilize the chopsticks by submerging in ethanol again. Wait until dry before measuring in the next insert.

(38)

31

Low sodium medium for Calu-3 cell culture This medium is based on the formulation of EMEM.

Component, base medium Chemical formula g/100 ml

Calcium chloride CaCl₂ 0.020

Magnesium chloride monohydrate MgCl₂· H₂O 0.024

Potassium chloride KCl 0.040

Sodium bicarbonate NaHCO₃ 0.150

Sodium phosphate monobasic NaH₂PO₄· H₂O 0.014

HEPES - 0.477

D-Glucose - 0.100

Additional component, 60 mM Na⁺

Sodium chloride NaCl 0.245

Calcium chloride CaCl2 0.252

Calcium chloride CaCl2 0.104

Table A.1: Formulation of low-sodium media used for experiments on Calu-3 cells.

(39)

Fixing and mounting cells grown on membrane onto coverslips What you need:

Formaldehyde Medium PBS

Microscope slides

Coverslips D 25mm, thickness 1.5 (0.17mm) Immu-Mount

Surgical scalpel and tweezers

1. Wash the cells 3 times with pre-warmed medium to remove dead cells.

2. Fix the cells with around 4% formaldehyde in medium. Since the formaldehyde I used is 36.5 − 38% I dilute it by ten. So I add 900µl of pre-heated medium and 100µl formaldehyde to a tube and mix. The formaldehyde is added in the safety hood since it is toxic. I found that it is sufficient to fix only from the upper compartment of the inserts so the total amount of fixing solution needed is 1 ml for each insert.

3. Replace the medium in the upper compartment of the inserts with the fixing solution and put it in the incubator for 10min.

4. Remove the fixing solution and wash with PBS 2-3 times.

5. The membrane should be placed with the cell side up on a microscope slide. Use a surgical scalpel to cut along the edges of the membrane. I found that cutting almost the whole way around and then lifting the insert and resting the membrane partly on the microscope slide before cutting the last part loose was an easy way to place the membrane on the slide without the edges folding up.

6. After the membrane is placed on the microscope slide, use a piece of paper to suck up any extra liquid around the edges.

7. Place a drop of Immu-Mount on the membrane and cover it with a coverslip. Leave it to set. Store the slides in the dark in a refrigerator.

(40)

Appendix B

Sample MATLAB code

1

2 % 2016-06-17, Minna Green-Petersen

3 % A script to calculate the FWHM of mitochondria intensity profile

4

5 files = dir('Data 60mM/*.xls');

6 hold on;

7 FWHM 60mM = zeros(length(files),1);

8 for i = 1:length(files)

9 searchPath = sprintf('Data 60mM/%s',files(i).name);

10 data = dlmread(searchPath,'\t',1, 0);

11 [maxVal,maxInd] = max(data(:,2));

12 halfMax = maxVal/2;

13 diff = abs(data(:,2)-halfMax);

14 sorted = sort(diff);

15 ind1 = find(diff == sorted(1));

16 ind2 = find(diff == sorted(2));

17 xpoint1 = data(ind1,1);

19 xMax = data(maxInd,1);

20 j = 3;

21 while xpoint1 > xMax && xpoint2 > xMax

22 ind2 = find(diff == sorted(j));

24 j = j + 1;

25 end

26 while xpoint1 < xMax && xpoint2 < xMax

27 ind2 = find(diff == sorted(j));

29 j = j + 1;

30 end

31 fwhm = abs(xpoint1-xpoint2);

32 FWHM 60mM(i,1)= fwhm;

33 plot(data(:,1),data(:,2),data(ind1,1),data(ind1,2),'*', ...

34 data(ind2,1),data(ind2,2),'*');

35 end

33

Mitochondrial alignment in ATP gradients