An Improved Method to Isolate Mitochondrial Contact Sites

Summary

Mitochondrial contact sites are protein complexes that interact with mitochondrial inner and outer membrane proteins. These sites are essential for the communication between the mitochondrial membranes and, thus, between the cytosol and the mitochondrial matrix. Here, we describe a method to identify candidates qualifying for this specific class of proteins.

Abstract

Mitochondria are present in virtually all eukaryotic cells and perform essential functions that go far beyond energy production, for instance, the synthesis of iron-sulfur clusters, lipids, or proteins, Ca²⁺ buffering, and the induction of apoptosis. Likewise, mitochondrial dysfunction results in severe human diseases such as cancer, diabetes, and neurodegeneration. In order to perform these functions, mitochondria have to communicate with the rest of the cell across their envelope, which consists of two membranes. Therefore, these two membranes have to interact constantly. Proteinaceous contact sites between the mitochondrial inner and outer membranes are essential in this respect. So far, several contact sites have been identified. In the method described here, Saccharomyces cerevisiae mitochondria are used to isolate contact sites and, thus, identify candidates that qualify for contact site proteins. We used this method to identify the mitochondrial contact site and cristae organizing system (MICOS) complex, one of the major contact site-forming complexes in the mitochondrial inner membrane, which is conserved from yeast to humans. Recently, we further improved this method to identify a novel contact site consisting of Cqd1 and the Por1-Om14 complex.

Introduction

Mitochondria perform a variety of different functions in eukaryotes, with the most well-known being the production of ATP through oxidative phosphorylation. Other functions include the production of iron-sulfur clusters, lipid synthesis, and in higher eukaryotes, Ca²⁺ signaling, and the induction of apoptosis^1,2,3,4. These functions are inseparably linked to their complex ultrastructure.

The mitochondrial ultrastructure was first described by electron microscopy⁵. It was shown that mitochondria are rather complex organelles consisting of two membranes: the mitochondrial outer membrane and the mitochondrial inner membrane. Thus, two aqueous compartments are formed by these membranes: the intermembrane space and the matrix. The mitochondrial inner membrane can be even further divided into different sections. The inner boundary membrane stays in close proximity to the outer membrane, and the cristae form invaginations. So-called crista junctions connect the inner boundary membrane and the cristae (Figure 1). Furthermore, electron micrographs of osmotically shrunken mitochondria reveal that sites exist at which the mitochondrial membranes are tightly connected^6,7. These so-called contact sites are formed by protein complexes spanning the two membranes (Figure 1). It is thought that these interaction sites are essential for cell viability due to their importance for the regulation of mitochondrial dynamics and inheritance, as well as the transfer of metabolites and signals between the cytosol and the matrix⁸.

The MICOS complex in the mitochondrial inner membrane is probably the best characterized and the most versatile contact site-forming complex. MICOS was described in yeast in 2011, and it consists of six subunits^9,10,11: Mic60, Mic27, Mic26, Mic19, Mic12, and Mic10. These form a complex of approximately 1.5 MDa that localizes to the crista junctions^9,10,11. The deletion of either core subunit, Mic10 or Mic60, leads to the absence of this complex^9,11, meaning these two subunits are essential for the stability of MICOS. Interestingly, MICOS forms not only one but multiple contact sites with various mitochondrial outer membrane proteins and complexes: the TOM complex^11,12, the TOB/SAM complex^{9,12,13,14,15,16}, the Fzo1-Ugo1 complex⁹, Por1¹⁰, OM45¹⁰, and Miro¹⁷. This strongly indicates that the MICOS complex is involved in various mitochondrial processes, such as protein import, phospholipid metabolism, and the generation of the mitochondrial ultrastructure¹⁸. The latter function is probably the major function of MICOS, as the absence of the MICOS complex induced through the deletion of MIC10 or MIC60 leads to an abnormal mitochondrial ultrastructure that virtually completely lacks regular cristae. Instead, internal membrane vesicles without connection to the inner boundary membrane accumulate^{19, 20}. Importantly, MICOS is conserved in form and function from yeast to human²¹. The association of mutations in MICOS subunits with severe human diseases also emphasizes its importance for higher eukaryotes^22,23. Although MICOS is highly versatile, additional contact sites must exist (based on our unpublished observations). Indeed, several other contact sites have been identified, for instance, the mitochondrial fusion machineries Mgm1-Ugo1/Fzo1^24,25,26 or Mdm31-Por1, which is involved in the biosynthesis of the mitochondrial-specific phospholipid cardiolipin²⁷. Recently, we improved the method that led us to the identification of MICOS to identify Cqd1 as part of a novel contact site formed with the outer membrane complex Por1-Om14²⁸. Interestingly, this contact site also seems to be involved in multiple processes such as mitochondrial membrane homeostasis, phospholipid metabolism, and the distribution of coenzyme Q^28,29.

Here, we used a variation of the previously described fractionation of mitochondria^{9,30,31,32,33}. Osmotic treatment of mitochondria leads to the disruption of the mitochondrial outer membrane and to a shrinkage of the matrix space, leaving the two membranes only in close proximity at contact sites. This allows for the generation of vesicles that consist exclusively of mitochondrial outer membrane or mitochondrial inner membrane or at contain contact sites of both membranes through mild sonication. Due to the mitochondrial inner membrane possessing a much higher protein-to-lipid ratio, mitochondrial inner membrane vesicles exhibit a higher density compared to mitochondrial outer membrane vesicles. The difference in density can be used to separate the membrane vesicles through sucrose buoyant density gradient centrifugation. Thus, the mitochondrial outer membrane vesicles accumulate at low sucrose concentrations, while the mitochondrial inner membrane vesicles are enriched at high sucrose concentrations. The vesicles containing contact sites concentrate at intermediate sucrose concentrations (Figure 2). The following protocol describes this improved method, which requires less specialized equipment, time, and energy compared to our previously established one³², in detail and provides a useful tool for the identification of possible contact site proteins.

Protocol

1. Buffers and stock solutions

1. Make a 1 M 3-morpholinopropane-1-sulfonic acid (MOPS) solution in deionized water, pH 7.4. Store at 4 °C.
2. Prepare 500 mM ethylenediaminetetraacetic acid (EDTA) in deionized water, pH 8.0. Store at room temperature.
3. Prepare 2.4 M sorbitol in deionized water. Store at room temperature after autoclaving.
4. Prepare 2.5 M sucrose in deionized water. Store at room temperature after autoclaving.
5. Prepare 200 mM phenylmethylsulfonyl fluoride (PMSF) in isopropanol. Store at −20 °C.
   CAUTION: PMSF is toxic if swallowed and causes severe skin burns and eye damage. Do not breathe in the dust, and wear protective gloves and goggles.
6. Prepare 72% trichloroacetic acid (TCA) in deionized water. Store at 4 °C.
   CAUTION: TCA may cause respiratory irritation, severe skin burns, and eye damage. Do not breathe in the dust, and wear protective gloves and goggles.
7. Prepare the SM buffer: 20 mM MOPS and 0.6 M sorbitol, pH 7.4.
8. Prepare the swelling buffer: 20 mM MOPS, 0.5 mM EDTA, 1 mM PMSF, and protease inhibitor cocktail, pH 7.4.
9. Prepare the sucrose gradient buffers: 0.8 M, 0.96 M, 1.02 M, 1.13 M, and 1.3 M sucrose in 20 mM MOPS, and 0.5 mM EDTA, pH 7.4.
   NOTE: All the buffers can be stored at 4 °C; however, it is recommended to store the sucrose-containing buffers at −20 °C to avoid fungal growth. The protease inhibitors have to be added freshly prior to use.

2. Generation of submitochondrial vesicles

1. Isolate crude yeast mitochondria freshly according to established protocols^34,35.
   NOTE: For this experiment, mitochondria were isolated from Saccharomyces cerevisiae. The generation of gradient pure mitochondria devoid of other organelles is not necessary.
2. Resuspend 10 mg of freshly isolated mitochondria in 1.6 mL SM buffer (4 °C) by pipetting (Figure 2, step 1).
3. Transfer the mitochondrial suspension to a pre-cooled 100 mL Erlenmeyer flask.
4. Slowly add 16 mL of swelling buffer using a 20 mL glass pipette. Apply constant mild stirring on ice during the addition. This osmotic treatment results in water uptake into the matrix space. The swelling of the mitochondrial inner membrane will disrupt the outer membrane. However, both membranes will stay in contact at the contact sites⁹(Figure 2, step 2).
5. Incubate the samples under constant mild stirring for 30 min on ice.
6. Slowly add 5 mL of 2.5 M sucrose solution using a 5 mL glass pipette to increase the sucrose concentration in the samples to approximately 0.55 M. This osmotic treatment results in a water efflux from the matrix³⁶. The shrinkage of the inner membrane is intended to maximize its distance to the residual fragments of the outer membrane. This decreases the probability of generating artificial hybrid vesicles of inner and outer membranes through sonication (Figure 2, step 3).
7. Incubate the samples under mild stirring for 15 min on ice.
8. Subject the mitochondria to sonication to generate submitochondrial membrane vesicles (Figure 2, step 4).
   1. Transfer the mitochondrial suspension to a pre-cooled rosette cell.
   2. Sonicate the mitochondrial suspension at a 10% amplitude for 30 s while cooling the rosette cell in an ice bath.
   3. Rest the suspension for 30 s in an ice bath.
   4. Repeat step 2.8.2 and step 2.8.3 three more times.
      NOTE: The sonication conditions will vary according to the sonicator used. Optimization will be necessary for individual machines. Sonication that is too harsh will lead to the artificial formation of vesicles consisting of mitochondrial inner and outer membranes.

3. Separation of submitochondrial vesicles

1. Separate the generated vesicles from the remaining intact mitochondria by centrifugation at 20,000 x g for 20 min at 4 °C. The intact mitochondria will pellet while the vesicles will stay in the supernatant.
2. Concentrate the vesicle mixture.
   1. Transfer the supernatant to fresh ultracentrifugation tubes.
   2. Load a cushion of 0.3 mL of 2.5 M sucrose solution at the bottom of the tube using a 1 mL syringe equipped with a 0.8 mm x 120 mm cannula.
   3. Centrifuge at 118,000 x g for 100 min at 4 °C.
      NOTE: Here, a swinging bucket rotor with a maximum radius of 153.1 mm was used. The centrifugation conditions strongly depend on the rotor and will have to be adapted when another rotor is used.
3. The vesicles will appear as a disc on the top of the sucrose cushion after the centrifugation. Discard approximately 90% of the supernatant. Now, harvest the concentrated vesicles by resuspending them in the remaining buffer including the 2.5 M sucrose by pipetting up and down. Transfer the suspension to an ice-cold Dounce homogenizer.
4. Homogenize the suspension with at least 10 strokes using a polytetrafluoroethylene potter.
5. Prepare the sucrose gradient.
   1. For an 11 mL step gradient with five steps (1.3 M, 1.13 M, 1.02 M, 0.96 M, and 0.8 M sucrose in 20 mM MOPS and 0.5 mM EDTA, pH 7.4), each layer has 2.2 mL of sucrose solution.
      NOTE: A refractometer is recommended to measure and adjust the sucrose concentrations; however, it is not essential. Apply 10 µL of the buffer to the refractometer, and detect the respective refractive indices. The calculation of sucrose concentration is as follows:
      
      1.3333 represents the refractive index of water. 0.048403 is a sucrose specific conversion index obtained from the linear scaling of molar refractivity to sucrose concentration in aqueous solutions ³⁷.
   2. Add the highest sucrose concentration to the centrifugation tube, and put the tube at −20 °C. Wait until the layer is completely frozen before applying the next one. Proceed accordingly until the last layer is added, and store the gradients at −20 °C.
      NOTE: Remember to transfer the frozen gradients to 4 °C when starting the experiment to allow the gradients to thaw. This will take approximately 3 h. For the centrifugation here, a swinging bucket rotor with a capacity of 13.2 mL was used. When a different rotor is used, the gradient step volumes must be adapted accordingly.
6. Measure the sucrose concentration of the samples using a refractometer. If there is not access to a refractometer, one could alternatively assume that the sucrose concentration is 2 M. This supposed sucrose concentration will then be the basis for the next step.
7. Adjust the sucrose concentration to 0.6 M by the addition of an appropriate amount of 20 mM MOPS, 0.5 mM EDTA, 1 mM PMSF, and protease inhibitor cocktail, pH 7.4. If the sucrose concentration is estimated as 2 M instead of being measured, it is recommended to test whether the concentration is low enough after adjustment. Apply a small aliquot of the sample (ca. 50 µL) on top of 200 µL of 0.8 M sucrose in a test tube. The sample must stay on top of the 0.8 M sucrose.
8. Load the samples (approximately 1 mL) on top of the sucrose gradient (keep 10% for later reference). Careful pipetting is important to avoid the disturbance of the gradient (Figure 2, step 5).
9. Separate the vesicles by centrifugation at 200,000 x g and 4 °C for 12 h. If possible, set the centrifuge to slow acceleration and deceleration to avoid the disturbance of the gradient (Figure 2, step 6).
   NOTE: Here, a swinging bucket rotor with a maximum radius of 153.1 mm was used. The centrifugation conditions strongly depend on the rotor and will have to be adapted when another rotor is used.
10. Harvest the gradient from top to bottom in 700 µL fractions using a 1 mL pipette. This will result in 17 fractions, which allows for a sufficient resolution.

4. Analysis of submitochondrial vesicles

1. Concentrate the proteins by subjecting each fraction to two sequentially performed TCA precipitations³⁸ to prepare the SDS-PAGE samples.
   1. Add 200 µL of 72% TCA to the individual fractions, and mix until the solution is homogeneous. Incubate the fractions for 30 min on ice, and pellet the precipitated proteins by centrifugation at 20,000 x g and 4 °C for 20 min. Discard the supernatant, add 500 µL of 28 % TCA solution, mix well, and repeat the centrifugation step.
   2. Wash the pellets with 1 mL of acetone (−20°C), and centrifuge for 10 min at 20,000 x g and 4 °C. Discard the supernatant, and let the pellets air dry. Resuspend the pellets in 60 µL of SDS sample buffer, and incubate the samples for 5 min at 95 °C.
      NOTE: A large amount of sucrose will remain, particularly in the high-density fractions, after the first TCA precipitation. This will be removed through the additional TCA precipitation.
2. Analyze 20 µL of each fraction by SDS-PAGE³⁹ and immunoblotting^40,41,42,43.

Results

It is relatively easy to separate mitochondrial inner and outer membranes. However, the generation and separation of contact site-containing vesicles are much more difficult. In our opinion, two steps are critical and essential: the sonication conditions and the gradient used.

Usually, linear gradients are thought to have a better resolution compared to step gradients. However, their reproducible production is tedious and requires special equipment. Therefore, we established a method to genera...

Discussion

Mitochondrial subfractionation is a complicated experiment with several highly complex steps. Therefore, we aimed to further improve and, to a certain degree, simplify our established method³². Here, the challenges were the requirement for complicated and highly specialized equipment, which are often individual constructions, and the enormous time and energy consumption. To this end, we tried to remove the pumps and individual constructions used for casting and harvesting the linear gradient and c...

Materials

Name	Company	Catalog Number	Comments
13.2 mL, Open-Top Thinwall Ultra-Clear Tube, 14 x 89mm	Beckman Instruments, Germany	344059
50 mL, Open-Top Thickwall Polycarbonate Open-Top Tube, 29 x 104mm	Beckman Instruments, Germany	363647
A-25.50 Fixed-Angle Rotor- Aluminum, 8 x 50 mL, 25,000 rpm, 75,600 x g	Beckman Instruments, Germany	363055
Abbe refractometer	Zeiss, Germany		discontinued, any pipet controller will suffice
accu-jet pro Pipet Controller	Brandtech, USA	BR26320	discontinued, any pipet controller will suffice
Beaker 1000 mL	DWK Life Science, Germany	C118.1
Branson  Digital Sonifier W-250 D	Branson Ultrasonics, USA	FIS15-338-125
Branson Ultrasonic 3mm TAPERED MICROTIP	Branson Ultrasonics, USA	101-148-062
Branson Ultrasonics 200- and 400-Watt Sonifiers: Rosette Cooling Cell	Branson Ultrasonics, USA	15-338-70
Centrifuge Avanti JXN-26	Beckman Instruments, Germany	B37912
Centrifuge Optima XPN-100 ultra	Beckman Instruments, Germany	8043-30-0031
cOmplete Proteaseinhibtor-Cocktail	Roche, Switzerland	11697498001
D-Sorbit	Roth, Germany	6213
EDTA (Ethylendiamin-tetraacetic acid disodium salt dihydrate)	Roth, Germany	8043
Erlenmeyer flask, 100 mL	Roth, Germany	X747.1
graduated pipette, Kl. B, 25:0, 0.1	Hirschmann, Germany	1180170
graduated pipette, Kl. B, 5:0, 0.05	Hirschmann, Germany	1180153
ice bath	neoLab, Germany	S12651
Magnetic stirrer RCT basic	IKA-Werke GmbH, Germany	Z645060GB-1EA
MOPS (3-(N-Morpholino)propanesulphonic acid)	Gerbu, Germany	1081
MyPipetman Select P1000	Gilson, USA	FP10006S
MyPipetman Select P20	Gilson, USA	FP10003S
MyPipetman Select P200	Gilson, USA	FP10005S
Omnifix 1 mL	Braun, Germany	4022495251879
Phenylmethylsulfonyl fluoride (PMSF)	Serva, Germany	32395.03
STERICAN cannula 21 Gx4 4/5 0.8x120 mm	Braun, Germany	4022495052414
stirring bar, 15 mm	VWR, USA	442-0366
Sucrose	Merck, Germany	S8501
SW 41 Ti Swinging-Bucket Rotor	Beckman Instruments, Germany	331362
Test tubes	Eppendorf, Germany	3810X
Tissue grinders, Potter-Elvehjem type, 2 mL glass vessel	VWR, USA	432-0200
Tissue grinders, Potter-Elvehjem type, 2 mL plunger with serrated tip	VWR, USA	432-0212
Trichloroacetic acid (TCA)	Sigma Aldrich, Germany	33731	discontinued, any TCA will suffice (CAS: 73-03-9)
TRIS	Roth, Germany	4855