UPTEC X 09 027
Examensarbete 30 hp Maj 2010
Development of a positive control kit for in situ Proximity Ligation Assay
Daniel Ekman
Molecular Biotechnology Programme
Uppsala University School of Engineering
UPTEC X 09 027 Date of issue 2009-11-10 Author
Daniel Ekman
Title (English)
Development of a positive control kit for in situ PLA
Title (Swedish)
Abstract
Product development of a positive control kit for in situ Proximity Ligation Assay (in situ PLA). In situ PLA is a novel method for a more sensitive and specific measurement of protein, protein interactions and protein modifications. Duolink control kit is intended to be used as a first time start up kit and as for support help. The control kit contains cells, antibodies and PLA probes, besides the duolink detection kit it is all needed for a complete in situ PLA experiment. The development procedure consists of the deciding and optimization of the control kits different parts.
Keywords
Duolink Control kit, In situ PLA, proximity, thymidine kinase (TK1), Olink Bioscience, Supervisors
Erik Nyström Olink Bioscience Scientific reviewer
Göran Holmquist Olink Bioscience
Project name Sponsors
Language
English
Security
ISSN 1401-2138 Classification
Supplementary bibliographical information Pages
23
Biology Education Centre Biomedical Center Husargatan 3 Uppsala Box 592 S-75124 Uppsala Tel +46 (0)18 4710000 Fax +46 (0)18 555217
Development of a positive control kit for in situ Proximity Ligation Assay
Daniel Ekman
Sammanfattning
Forskningen av proteiner och deras uttrycksmönster utgör idag en stor del av det forskningsarbete som kartlägger människans celler och vilka processer som styr dess funktioner. Idag anser man att uppkomsten av många svåra sjukdomar beror på avvikelser i uttrycksmönstret av kroppens proteiner. Cancerforskningen är idag kraftigt fokuserad på var och när proteiner är aktiva.
Länge har man studerat proteiner genom att färga dem med hjälp av specifika antikroppar.
Till dessa sätts enzym eller flouroecenta markörer som kan ses i t.ex. mikroskop. Nyligen har ett nytt verktyg för att hjälpa forskarna i sitt arbete utvecklats. In situ proximity ligation assay (in situ PLA) är en metod som kraftigt ökar informationen som tas fram vid proteinfärgning.
Eftersom som in situ PLA är en ny metod så kan det behövas hjälp med analysutförandet samt resultattolkningen. Därför valde Olink Bioscience, som säljer och äger metoden, att utveckla ett kontroll kit som skall användas som en förstagångshjälp samt hjälpa befintliga kunder som hamnat i tekniska problem. Detta examensarbete var att ta fram ett sådant kit, innehållande alla delar som krävs för en in situ PLA assay, som på ett användarvänligt sätt skall ge ett lättolkat resultat. På detta sätt vill Olink Bioscience visa upp sin nyutvecklade metod och alla dess fördelar.
Examensarbete 30hp
Civilingenjörsprogrammet Molekylär bioteknik Uppsala universitet maj 2010
1
1. Introduction ... 3
1.1 Duolink and the In situ proximity ligation assay (in situ PLA) ... 3
1.1.1 Aim of study ... 4
1.1.2 In situ PLA applications ... 4
1.2 Other methods for protein detection ... 5
1.2.1 Immunofluorescence (IF) ... 5
1.2.2 Immunohistochemistry (IHC) ... 5
1.2.3 Fluorescence resonance energy transfer (FRET) ... 5
1.2.4 Bioluminescence Resonance Energy Transfer (BRET) ... 6
1.3 The control kit ... 6
1.3.1 The way to the control kit ... 7
1.4 Thymidine kinase (TK) ... 8
2. Materials and methods ... 8
2.1 The Cells ... 8
2.1.1 Cell lines ... 8
2.1.2 Cell preparation and fixation ... 10
2.1.3 Cell stability ... 10
2.1.4 Cell durability ... 10
2.2 The Primary Antibody ... 11
2.2.1 Anti‐TK1‐XPA 210‐A ... 11
2.2.2 Blocking ... 11
2.2.3 Antibody incubation ... 11
2.2.4 Antibody stability ... 11
2.2.5 Antibody storage and durability ... 11
2.3 PLA probes ... 11
2.3.1 PLA probe storage ... 12
2.4 RCA product formation ... 12
2.5 Detection ... 12
2.6 Washing ... 12
2.7 Verification ... 12
2.8 Analysis ... 13
3. Results ... 13
3.1 The cells ... 13
3.1.1 Cell comparison ... 13
2
3.1.2 Cell stability after epitope retrieval ... 13
3.1.3 Cell durability ... 14
3.2 The antibody ... 14
3.2.1 Antibody dilution ... 14
3.2.2 Antibody storage ... 15
3.2.3 Antibody durability ... 15
3.2.4 Batch comparison ... 16
3.3 In situ PLA protocol ... 16
3.31 Protocol ... 16
3.32 Washing step ... 16
3.4 Verification ... 17
4. Discussion ... 17
5. Conclusion ... 20
6. Acknowledgements ... 21
7. References ... 22
3
Figure1: Graphic illustration of the in situ PLA method. A. Binding of two primary antibodies to the protein or protein complex. B.
Addition of two secondary antibodies targeting the first to antibodies. C. Connector oligos en the ligation mixture creates the complete DNA circle. D. Addition of polymerase to the DNA circle. E. The RCA process begins. F. Addition of fluorescent labeled oligos visualizes each rolling circle product as one signal.
1. Introduction
1.1 Duolink and the In situ proximity ligation assay (in situ PLA) Today, when the interest of the proteome and its vast complexity
and enormous capability to affect all cells of the body increases, the demands for better and more advanced methods for protein detection grow. The ability to detect single proteins, or protein interactions, and the possibility to quantify the result has been made possibly by the recently developed in situ Proximity Ligation Assay (PLA) technology published Ulf Landegren and co-workers at Rudbeck laboratory in Uppsala.
In situ PLA is based on the rolling circle
amplification (RCA) technique. In RCA, circular DNA strands are formed and give rise to single stranded amplification products composed of hundreds of complements of the circular DNA strands [1].
The protein, or protein complexes, to be detected are first identified by two primary antibodies directed to epitopes on the protein or the adjacent proteins. Secondary antibodies directed against the different primary antibodies and are then added. The in situ PLA protocol uses two different secondary antibodies called PLUS and MINUS probes. All PLA probes used were produced by Olink Bioscience through conjugation of an oligionucleotide and antibody forming so called proximity probes. To ensure the
high specificity of the method, binding of both PLUS and
MINUS probes are necessary for creation of the RCA circle. The difference, between the oligos used on the PLUS and the MINUS probes is that the PLUS probe oligo is priming and the MINUS
probe oligo is non priming. To form the nucleotide circle, the proximity binding of both probes is necessary.
A
B
C
D
E
F
4 To form the DNA circle, two connector oligonucleotides first hybridize to the PLUS and MINUS arms of two different PLA probes. The two connectors are ligated together and a circle is formed. When polymerase and free nucleotides is added the circulized DNA can act as a template for the RCA process, and the PLUS arm acts as a primer for the ampification reaction [2]. DNA probes conjugated with a fluorophore, complementary to the repeated sequence, make it possible to detect the interaction in a fluoroesent microscope. Each
proximity event gives rise to a fluoresent signal thus making the method highly sensitive. The need for two specific primary antibodies reduces the background which increases the
specificity of the method.
1.1.1 Aim of study
The aim of this project was to develop a kit composed of antibodies, cells and PLA probes that could be a sold as a product in the duolink product series. The control kit should be both functional and attractive on the protein staining market which could be harsh for a newly developed product series as the duolink one. Another demand on the product was that it could bring in profit; the cost of goods sold (COGS) must be large enough. The studies
experimental part was both to decide the different product parts and optimization of them.
1.1.2 In situ PLA applications
There are mainly tree different applications of in situ PLA.
1.1.2.1 Duolink IQ
Duolink IQ is used to detect and quantify interacting proteins. Duolink IQ uses two different primary antibody directed to the two proteins of interest [3]. See Figure 2.
1.1.2.2 Duolink PQ
Also Duolink PQ uses two primary antibodies, one directed against the target protein and one against the phosphorylation site on the same protein [3]. See Figure 2.
1.1.2.3 Duolink Q
The third kind of in situ PLA is Duolink Q. In Duolink Q it is the protein expression that is of interest. You can either use only one primary antibody which will increase the sensitivity of the method or use two primary antibodies directed against two different epitopes on the same protein much like Duolink IQ, which will increase the specificity of the method [3]. See Figure 2.
5
1.2 Other methods for protein detection
1.2.1 Immunofluorescence (IF)
Immunofluorescence (IF) normally uses two antibodies, one primary antibody, and a
fluorophore-coupled secondary antibody, that recognizes the primary antibody. In some cases it could be advantageous to couple the fluorophore directly to the primary antibody. This direct labeling decreases the number of steps needed and avoids cross-reactivity and background problems. The presence of the fluorophore can be detected in a fluorescence microscope thus making it possible to study proteins [4].
1.2.2 Immunohistochemistry (IHC)
Immunohistochemistry (IHC) uses the same fundamental grounds as IF, with one primary antibody, directed to the protein of interest, and one secondary antibody directed against the primary antibody. Once an antigen-antibody binding occurs, it is demonstrated with a colored histochemical reaction visible by light microscope [5]. Commonly the secondary antibody is coupled with horseradish peroxidase (HRP) which is the enzyme that reacts with
diaminobenzidine (DAB) and produces a brown color.
There are two different kinds of IHC-methods, direct and indirect. The direct method uses one HRP labeled primary antibody, reacting in a one step reaction, with its antigen. This procedure is simpler and a more rapid assay compared with the indirect method, which uses both primary and secondary antibodies.
1.2.3 Fluorescence resonance energy transfer (FRET)
Fluorescence resonance energy transfer (FRET) is a technique for measuring protein
interactions in vivo [6]. FRET is based on the efficient Resonance Energy Transfer (RET) [7].
The FRET method uses two fluorophores fused to each one of the two interacting proteins and
Figure2: A. Detection of a protein‐protein complex using two different antibodies. B. Detection of one protein and its specific phosphorylation site using two different antibodies. C. Detection of one protein using one antibody. D. Detection of one protein using two different antibodies.
A B C D
6 is dependend on the distance between the two fluorophores. One fluorophore absorbs
electromagnetic energy of one wavelength (the excitation frequency) and re-emits that energy at a different wavelength (the emission frequency), which result in a two peaked spectrum. To obtain the combined FRET effect, the emission peak of the donor fluorophore must overlap with the excitation peak of the acceptor fluorophore. In FRET light energy is added at the excitation frequency for the donor fluorophore. This will transfer some of its energy to the acceptor which then re-emits the light at its own emission wavelength. This will result in that the acceptor fluorophore emits more light energy than normal, due to the extra energy input.
This light energy can be detected at the acceptors excitation frequency [6].
1.2.4 Bioluminescence Resonance Energy Transfer (BRET)
The BRET assay technology is, as FRET, based on RET between a donor moiety and a acceptor moiety. BRET is a naturally occurring phenomenon and differs from FRET in that it uses a luciferase as the donor and a fluorescent acceptor [7]. When the acceptor and the donor comes in close contact to each other, as in molecular interactions, RET will occur thus making it possible to monitor protein interactions.
1.3 The control kit
The control kit is indented to be used, as a positive control, by first- time in situ PLA users to learn how to perform the in situ PLA procedure [8]. For first- time users it can be hard to determine what real PLA signal is since the fluorescence microscope detects many other more or less fluorescent materials. The control kit is further intended to be used in support to help costumers that in one way or another found themselves to be in trouble.
Figure3: A. Example of the positive control reaction using the duolink control kit ((PLA signals are stained with red color and nucleus are stained with blue color). B. Example of the negative control reaction using duolink control kit (PLA signals are stained with red color and nucleus are stained with blue color).
A B
7 The control kit contains reference slides, primary antibodies and also secondary antibodies called PLA probes. The primary antibodies and the cells are in normal cases provided by the costumers. Combined with the Duolink Detection kit, which contains all of the necessary reagents, this is all that is needed for the in situ PLA procedure.
1.3.1 The way to the control kit
The mere thought of making a control kit was born out of a wish to help first time users or to support already existing customers with troubles in performing or analyzing in situ PLA. A control kit is intended to be used both as a positive and negative control which can be necessary while in situ PLA still is a new and not completely established method in the scientific community.
The first step in designing the product, was determining the composition of the kit. The kit shall include cells, primary antibodies and PLA probes. All of these components have to be tested, easy to handle and be able to sell. The Primary antibody should be commercially available and directed against a protein expressed in a large number of cells, robust and easy to store and dilute. The cells physical appearance was of some importance because the results would be easy visualized and understandable. The ratio between the size of the nucleus and the size of the cytoplasm had to be large enough so that the PLA signals could be nicely separated thus making it easier to count and separate the PLA signals. The ability to count the relative levels of proteins or protein -protein interactions present in a cell is one of the main advantages with the in situ PLA technique. This ability was of course a desirable property in the control kit. Furthermore, the cells must be easy to grow and harvest so that large quantities of the glass slides (carrying the cells) could be made and making the product reproducible and affordable. The cells have to be fixed onto the slides so that they can survive transport and other stress factors due to handling and shipping, that involves several temperature changes.
The control kit must also be durable enough to be stored in the freezer during a longer period.
At last, a legal agreement giving the company rights to sell both cells and antibodies must be made with the suppliers.
When all the components are decided and durability tested the optimization process giving satisfactory results will begin. After this the validation of the product and beta testing remains. Validation is the processes where the product is tested within the company.
Acceptance criteria are set for the different components which must be fulfilled. These demands are necessary when the results can alternate from experiment to experiment. The
A B
8 very last step before releasing is beta testing where people outside of the company test the product to simulate real costumers. This is to ensure that the product gives the same results independently of handling.
1.4 Thymidine kinase (TK)
Thymidine kinase is a phosphotransferase and can be found in most living cells, which makes it a perfect protein for the control kit as it makes it possible for the costumers to apply the control kit antibody onto their own cell material. It is naturally present in two different forms, TK1 and TK2. TK1 is a cell- cycle regulated cytoplasmic enzyme and TK2 is a constitutively expressed mitochondrial
enzyme. Both TK1 and TK2 serve to maintain sufficient dTTP for DNA replication and repair during cellular metabolism [9]. Thymidine kinases thereby have a key function in the
synthesis of DNA and thus in cell division [9].
The TK1 has a tetrameric structure that contains an alpha/beta domain much like the ATPase domain of the RecA structural family with a domain containing zinc [10]. The zinc ion connects beta-structures that form into a lasso-type loop [10]. The thymidine from dTTP is hydrogen bonded to atoms of the lasso loop on the main-chain [10]. The structure of TK1 differs from other deoxyribonucleoside kinasesin this matter which usually bind to the side chain thus indicating a different evolutionary origin [10].
Thymidine kinases are today used in many antiviral and anticancer drugs. It is also used in clinical chemistry as a proliferation marker in diagnosis, control of treatment and follow-up of malignant diseases.
2. Materials and methods 2.1 The Cells
2.1.1 Cell lines
These are the different cell-lines that were tested during the project.
Figure3: Graphic illustration of the thymidine kinase protein.
9 2.1.1.1 A431
A431 is a human epidermoid carcinoma cell line. A431 has a large number of EGF binding sites and has been used as indicator for anti-TGF binding. A431 is grown and harvested at Rudbeck laboratory by Ola Söderberg.
2.1.1.2 K562
K562 is a cell line from human chronic myeloid leukemia cells (CML). The cells are highly sensitive targets for killer cells and have been used in in vitro natural killer assays. K562 is grown and harvested by 3H Biomedical.
2.1.1.3 U2973
U-2973 is an Epstein-Barr virus negative cell line. U-2973 is grown and harvested at Rudbeck laboratory by Ola Söderberg [11].
2.1.1.4 ULA
ULA is, as U-2973, cultured at Rudbeck laboratory by Ola Söderberg. It is an Lymphoma cell line established from ascites fluid [12].
Figure 4: Examples of the different cell lines. A. A431. B. K562. C. U2973. D. ULA. E. MDA 175. F. SK‐BR‐3. G. MDA 231. H. Skin tissue.
Cells shown in picture A, B, C, D, E, G, H are stained with a antibody targeting the TK1 protein. Cells on picture F are stained with antibodies targeting the MAPK3s phosphorylated site.
10 2.1.1.5 Her2 cell array
SK-BR-3, MDA 175 and MDA 231 are well characterized Her2-expressing breast cancer cells, from ATCC. These cell lines are used for characterization of breast tumors. All three are derived fromhumans. They express Her2 at different rates where the SK-BR-3 cell line over expresses the HER2/c-erb-2 gene product and MDA 231 under expresses it.
2.1.1.6 Skin tissue cells
Human epidermal cells derived at Akademiska sjukhuset in Uppsala.
2.1.2 Cell preparation and fixation
The A431, U2973 and ULA cell lines were first cultured then centrifuged to the microscope slides and fixed in either paraformaldehyde (PFA) or with Zinc (Zn) fixative solution. The cell lines grown directly on slides, K562 (3H Biomedical, Uppsala) were fixed with either PFA or Zn. In the case of the PFA fixed cells the cell membrane was permeabilized by 0,2%
Triton-100X in PBS for 3 min at room temperature. This treatment is not necessary for the Zn fixed cells. The sectioned cells lines (SK-BR-3, MDA 175, MDA 231) were formalin fixed and paraffin embedded (FFPE) as described by Andersson et al [13]. 4 µm sections of these three cell lines were cut on to microscope slides by Ulrica Larsson and Inga Hansson at the Rudbeck laboratory, Uppsala. The paraffin was removed by a wash series in xylen followed by a decreasing alcohol series (2 x 10min xylen, 2 x 3min 99% alcohol and 2 x 3min 96%
alcohol). After the paraffin was removed the epitopes are retrieved using HIER (Heat Induced Epitope Retrieval) in a citrate buffer (pH 6). PFA fixed cells were as a final step quenched with 400µl 1M glycine in 90ml PBS for 2 times 5min.
2.1.3 Cell stability
The sectioned cells were tested if they endured freezing after the retrieval process. They were frozen both when they were still wet after the last PBS wash and after they had dried. A test where the cells were thawed, and kept in the fridge (+4°C) for three days, before they were frozen again was also performed.
2.1.4 Cell durability
The cells were kept at -20 C, +4 C and +37 C, after the retrieval process. The cells
durability was tested first after 2 weeks and then again after 1 month by performing an in situ PLA and comparing the amount of signals per cell. The cell durability has been continually after that.
A B
D E F
11
2.2 The Primary Antibody
2.2.1 AntiTK1XPA 210A 2.2.1.1 Background
Anti XPA-210-A (AroCell) Uppsala, which is derived from chicken, has earlier been used for cytoplasmic IHC (see 1.2.1) staining in sectioned tissue proliferation cells. The antibody is a designed, affinity purified, polyclonal IgY product. The uses have often been cancer research, both basic and clinical [14].
2.2.1.2 Storage
The Anti XPA-210-A is stored in 0,15M NaCl, 0,01M phosphate (pH 7,4), 0,02% NaN3 at 4 C. The storage concentration is 0,2mg/mL [15].
Two different antibody batches, Gf012 and Ha005, were used during the project.
2.2.2 Blocking
Blocking was done by using Duolink standard block for 30 min at 37 C.
2.2.3 Antibody incubation
The anti-TK1-XPA 210-A antibody was incubated overnight at +4 C at the diluitions 1:200, 1:400, 1:800, 1:1000, 1:1200, 1:1250 or incubated for 30 min at either room temperature (RT) or at +37C at the dilutions 1: 200, 1:300, 1:500, 1:750, 1:1000, 1:1200, 1:2000.
2.2.4 Antibody stability
To ensure the robustness of Anti-TK1-XPA 210-A the antibody was freeze-thawed ten times.
Then processed in an in situ PLA and compared to an antibody reference sample that had undergone no stress.
2.2.5 Antibody storage and durability
The antibodies were stored partly in the concentration that later was to be used in the reaction (1x), partly in a five times higher concentration (5x). These two storage possibilities were compared to the non diluted reference antibody. Both two different storage possibilities were kept at 37 C and 45 C. The antibody durability was tested after 5 weeks.
2.3 PLA probes
Both the PLUS and MINUS PLA probes in the control kit are anti-chicken antibodies and thereby bind to the TK1 antibody. The PLA probes used were produced by Olink Bioscience through conjugation of an oligonucleotide to an antibody.
12 2.3.1 PLA probe storage
The probes are stored in the freezer concentrated as 5x stocks in a stabilization buffer. In the final in situ PLA reaction they only require a dilution step, in highly purified water and antibody diluents, to obtain the 1x working concentration. The probes are incubates on the slides for 2h at 37°C.
2.4 RCA product formation
The cells are washed 2x 5min in TBS-T (Tris-Buffered Saline Tween-20) before adding the hybridization mix and then incubated for 15 min at 37⁰C. After a 1 min wash in TBS-T the ligation mix containing ATP and ligase, is applied to the sample and incubated for 15 min at 37⁰C. This solution is washed away, 2x2 min in TBS-T, and the amplification mixture, which contains free nucleotides and polymerase is subsequently added to the slide. The sample is now incubated for 90 min at 37⁰C.
2.5 Detection
Two different detection stocks were used during the project. Both contain Hoechst 33342 as nuclear stain which visualizes the cell nuclei by staining the DNA. The PLA signals are in one case stained by a TexasRed analog (excitation 598nm, emission 613nm) and in the other case by 563 (excitation 557 nm, emission 563nm). The samples are incubated for 60 min at 37⁰C with the detection solution.
2.6 Washing
The last step is a washing step when the cells are cleaned from the remaining reagents. Four different washing methods were tested, each with a series of decreasing concentrations.
a) SSC with Tween 2min 2x SSC‐T; 2min 1x SCC‐T; 2min 0.2x SSC; 2min 0.02x SCC; 1min EtOH.
b) SSC without Tween 2min 2x SSC; 2min 1x SCC; 2min 0.2x SSC; 2min 0.02x SCC; 1min EtOH.
c) TBS‐T 3min 1x TBS‐T 3min 0.1x TBS‐T; 3min 0.01 TBS‐T; 1min EtOH.
d) Both TBS and SCC 3min 1x TBS; 3min 1x TBS; 3min 0.2x SSC; 2min 0.02x SCC.
2.7 Verification
When all the parameters of the control kit were decided and tested a beta kit was produced, containing two batches cell slides and two batches of primary antibodies. Certain acceptance demands regarding the components durability and stress stability were set up for all of the different components. Verification tests comparing the two cell and antibody batches were performed. Furthermore the new protocol, with the new length and new washing method, was tested.
13
2.8 Analysis
Images for analysis were obtained by using an epifluorescence microscope (Axio Imager M1, Zeiss) with an AxioCam MRm-camera. Each image consists of one picture from the nucleus channel and 20 pictures from the signal channel. The nucleus is stained with Hoechst
(excition 350nm and emission 461nm) is found by the cameras autofocus. The images of the PLA signals, which are stained with the TexasRed fluorophore , is composed of 20 different pictures, 10 below the nucleus and 10 above, and cover thereby the whole cell depth. Images were processed in the Axio Vision 4.5 (Carl Zeiss) software and converted from zvi format to tif format. The tif images are analyzed by the free software Blobfinder v3.2 which separately counts the nucleus and the PLA signals. The Blobfinder software is developed at the Centre for Image Analysis, Uppsala University.
During this project the result images were analyzed by counting the number of signals per cell, these values are given by the Blobfinder software. To obtain the background signals (unspecific signals) the number of signals that are not located in neither cytoplasm nor nucleus are divided by the picture area that are not covered by cells.
3. Results 3.1 The cells
3.1.1 Cell comparison
Of all seven cell lines that were tested MDA 175 showed the most satisfying results regarding stability and visual demands. The MDA 175 breast cancer cells are well separated without forming clusters seen with the SK-BR-3 cells and A431. They shows good proportion between the cytoplasm and the nucleus, where the ULA and U2973 cells have very small cytoplasm. MDA 175 cells are together with SK-BR-3 and MDA 231 cells section cells and are have thereby almost no difference between samples. Grown cells such as ULA, U2973 and A431 show much greater variability in number of cells per reaction. Auto fluorescence was seen in many cases with the K562 cell line.
3.1.2 Cell stability after epitope retrieval
To freeze the cell slides after epitope retrieval (ER) treatment caused no specificity problems.
Fine PLA results were obtained both when the cell slides were frozen wet and frozen dried.
The test when the cells were thawed and then refrozen caused no losses for the primary specificity and both the positive and the negative controls gave satisfactory results.
14 3.1.3 Cell durability
Cells that were stored in +4°C and +37°C began to show unspecific signals after 1month while the cells that were stored in the freezer showed no such tendencies. Durability of the retrieved cells has been continually tested and only a slight decrease in the number of PLA signals was observed when the cells had been incubated for 4 month when stored at -20°C.
3.2 The antibody
3.2.1 Antibody dilution
The amount of signals per cell varies between different cell lines and cell preparation. Same tendencies can however be found. Higher concentration give saturated results were the signals
cannot be separated from each other. Lower concentrations give results so weak that the protein expression pattern is hard to distinguish. Cells with more than 200 signals per cell cannot be analyzed with the Blobfinder software which measures intensity maxima and need thereby signals separated from each other. Of the different dilution the 1:1200 was found to be the best, in regards of the number of PLA signals per cell, for the overnight incubation. For the overnight incubation the 1:500 dilutions was the best match. All of these dilutions were tried on the Gf 012 antibody batch.
Figure 5: Examples from PLA results using different concentrations of the anti TK1 antibody. All results are from the overnight incubation protocol. A:
1:200 dilution on A431 cells. B: 1:800 dilution on A431 cells. C: 1:1200 dilution on MDA 175 cells.
D: 1:2000 dilution on MDA 175 cells.
A B
C D
15
Figure6: Results from the antibody storage experiment. The bars labeled with 1x represent the
antibodies that have been stored in the lower concentration that doesn’t need dilution. Bars labeled with 5x represent the antibodies that have been stored in a higher concentration that need to be diluted before use. Antibodies that have been stored for two month are compared with antibodies that have been stored for two days and a newly made antibody dilution directly from the antibody stock.
3.2.2 Antibody storage
Highest storage stability was found when the antibody was stored at the higher concentrated 5 x stocks. Almost no loss of signals was seen when a two month old and a newly diluted antibody stock were compared at three different storage temperatures. The antibodies that were stored at the right 1x working concentration showed more instability regarding the number of PLA signals.
3.2.3 Antibody durability
The antibody durability test that was first executed after 5 weeks gave no significant signal loss in neither ‐20°C freezer or in the +4°C fridge storage conditions. Antibody durability for the 5x diluted stock has been continually tested and has not showed any sign of activity loss after 6 month at ‐20°C.
A small activity loss has been observed in the antibodies that has been stored 5x concentrated at +4°C fridge.
Figure7: Antibody durability experiment with antibodies, in stock dilution, from different incubation temperatures and different incubation time. Antibody stocks that had been kept partly at ‐20°C and partly at +4°C for 3 and 5 weeks and was compared to a newly made stock.
A B
16 3.2.4 Batch comparison
Five different dilutions, 1:100; 1:200; 1:250; 1:300 and 1:400, of the Ha005 antibody batch were tested to match the activity of the 1:500 diluted Gf012 batch. An overlap in number of signals was found with the 1:200 dilutions. The 1:250, 1:300 and 1:400 of the Ha005 batch gave to few signals and the 1:100 dilutions gave to many signals compared to the 1:500 dilutions with the Gf012 batch.
3.3 In situ PLA protocol
3.31 Protocol
Shortening the protocol by decreasing the incubation time for the primary antibody, from an
overnight incubation to 30 min incubation, was possible but decreased the number of signals per cell.
The same tendency was observed for all tree cell lines (U2973, ULA and MDA175). The signal loss was compensated by increasing the antibody concentration from a 1:1200 dilution to a 1:500 dilution for the MDA 175 cell line. Incubation at room temperature and incubation at 37°C gave approximately the same amount of signals. A slightly higher number of unspecific signals were seen in the samples that had been incubated in a higher temperature.
3.32 Washing step
Washing with TBS‐T showed a higher number of signals compared to washing with SSC both with and without Tween. Performing the last wash step with SCC and SCC‐T gave roughly the same result.
Removing the ethanol washing step as the final washing step gave no signal loss and resulted not in any other unwanted effects such as washing buffer artefacts or unspecific signals on the material.
Removing the ethanol step gave some indications that it would counteract the presence of unwanted auto fluorescence. These results were not seen in repeated experiments.
Figure8: Result from the experiments where the two antibody batches were concentration matched. The best overlap was found when the Ga005 was diluted 1:200 and the Gf012 was diluted 1:500. Results are taken from different experiments.
17
Figure9: Results from the verification experiment of the two glass batches and the two antibody batches. The left bars represent the first antibody batch and the bars to the right represent the second antibody batch. The blue bars are the results from the first glass batch and the red are the results from the second glass batch.
3.4 Verification
Both batches glass slides and both batches antibodies gave acceptable results. They were also cross tested and gave results that match the demands. When the two different glass batches were compared it was seen that glass batch 1
gave slightly more signals than glass batch 2 and that antibody batch 2 had a little more activity than antibody batch 1.
The modification made in the protocol with the new length and new washing step showed no problems.
4. Discussion
The anti TK1 antibody was the first component chosen in the control kit. This
choice was based on the antibody’s robust qualities and that the TK1 protein, which is a cytoplasmic marker expressed during the cell proliferation step, is common in almost all cells. The expression pattern of the TK1 protein helps to show the specificity of the in situ PLA method as you clearly can see different number of fluorescent signals between different cells, due to which step in the proliferation cycle they are in affecting the expression rate of the TK1 protein.
Two major proposals were made for the antibody storage. Either the antibody should be stored at the concentration that was meant for the final reaction or it should be kept at a higher concentration that had to be diluted, in highly purified water, before use. The other options that had to be
investigated were at which temperature the antibody should be kept. Antibody stability studies showed that the antibody survived best when kept at ‐20⁰C and in the higher concentration. The samples that were stored in the lower concentration showed a higher signal unspecificity when kept at +4⁰C. The choice of ‐20°C storage of the TK1 antibody and in a higher concentration was also desirable when the rest of the regents in the detection kit were all stored in the same temperature.
The higher antibody concentration also decreased the total volume of the antibody vessel.
Antibody durability was first tested after 5 weeks and showed no loss of signals regardless of storage temperature. A slight tendency that the number of signals increased after time could be noticed but were discarded as acceptable differences within the experiment. One unexpected problem was that the two different antibody batches gave different result. It seemed like the Ga005 batch had lower activity than the Gf012 batch. This was a problem since the control kit always should give the same
18 number of signals, 20 – 30 signals per cell. It was solved by deciding the right dilution for the Gf012 batch and then performing a dilution series where different concentrations of the HA005 batch were matched against Gf012. When the right corresponding concentrations were established, the two batches were diluted separately, so that the final product had a five times higher concentration then the working reaction concentration.
The next component that had to be decided was the cell line in the control kit. The MDA 175 cell line gave the most satisfying results mostly based on the cells visual qualities. The cell cytoplasm size was large enough to give nicely separated signals. Another advantage with a sectioned cell line is that the number of cells per reaction always is the same, a characteristic that can be hard to control with cells cultured directly on slides. One potentially disadvantage with paraffin embedded cells is that they have to be retrieved before use. To make the large batches with cell slides, which were needed to keep the production cost as low as possible, the cell slides had to survive in the freezer after the deparaffinization and antigen retrieval process. After retrieval the cells were therefore dried and then frozen to test durability. The slides were also exposed to temperature differences. The
durability studies showed that the cells that were kept in the refrigerator, which were one alternative for the final product, started to get fewer and unspecific signals but the cells that were kept in the freezer showed no such tendencies. When the freeze ‐thaw test, that simulates the shipping to the customer, gave good results it told us that the cells both could be kept in the freezer and endure shipping. Together these tests gave us the satisfying answer that the cell slides could survive storage after retrieval and thereby could be made in large batches. Combined these results gave the final solution that the cells should be kept at ‐20⁰C. When all of the legal agreements were made the MDA 175 cell line was definitely decided to be the control kit cells.
The control kit protocol should be optimized so that it would be as easy to perform as possible but it should also follow the normal in situ PLA protocol. One parameter that differs between experiments is the incubation time and temperature for the primary antibody. These parameters vary between different antibodies. The previous used incubation for the anti TK1 antibody was an overnight
incubation at 4°C. One wish was that the whole control kit protocol could be performed and analyzed in one day. When we lowered the incubation time to 30 min we saw a decrease in the number of signals. This had to be compensated for by increasing the antibody concentration.
Two different incubation temperatures were studied, +37°C and +20°C. The two temperatures gave approximately the same number of signals but we saw slightly more unspecific signals in the samples that had been incubated at the higher temperature, therefore we found that 30 min incubation at room temperature was the optimal choice.
19 One other deviation from the ordinary in situ PLA protocol was to change the buffer used in the last washing step. We wanted to try out to wash in TBS‐T which is the same buffer that is used
throughout the rest of the protocol. In the same way as for the standard final washing procedure, which is a mix of both SSC‐T and SCC, we used a decreasing concentration ladder but with TBS‐T. We observed an increasing number of signals compared to the old wash when we used the TBS‐T wash.
This was seen in all of our test samples. We believed these signals were unspecific singles and not real legal signals. This was maybe a consequence of the presence of Tween in all of the washing steps and not just in the first two ones. When this was not completely investigated we tried out another washing solution. This time we tested to wash all steps with SSC without Tween. This washing method proved to be equivalent to the older washing procedure. This new slightly different solution was not as elegant as the first one but was still simplifying the protocol.
When all components had been defined and their concentration decided only the last step remained.
This was the verification, in which the whole product was tested. The testing was performed both internally and externally. Certain demands were set which had to be fulfilled to enable the Control kit to pass as a commercial product. These demands were set to double check the products durability and reproducibility. Tests were made to investigate whether the different batches, of both control antibodies and control slides, gave the same number of signals. This was tested by a criss‐cross experiment, in which each antibody batch was applied to each glass batch. Small differences
between the batches were observed, one glass batch seemed to give more PLA signals and one batch of control antibodies seemed to have slightly higher activity. The differences were however not large enough to cause any concern, so all batches passed the verification demands.
Moreover, the new modified washing protocol was put to one more test with several antibodies and several cell materials, with different pre‐treatment and fixations. Just as previously the wash with TBS‐T gave the highest number of PLA signals, so it was finally discarded but the other new wash protocol gave once again the same result as the old protocol. Because of this the SSC wash without Tween was chosen for the control kit protocol.
The final validation was, once again, to look at the durability for the glass slides and the pre diluted (into the product 5x stock) control antibody. The control antibodies did also have to pass a freeze‐
thaw experiment. When these experiments gave no sign of signal loss, it convinced us that the control kit worked in all of its components. The external testing was performed by two PhD students at Karolinka Institutet, Stockholm. This testing was done to confirm that persons that never had worked with the in situ PLA technology method would get the same results as a more experienced user.
20
5. Conclusion
The control kit became a commercially product and it is today sold as an accessory complement together with the standard Duolink Detection kit. Even though the way to the control kit was not as straight as we first hoped, the final product in many ways corresponds with our first expectations.
We had to make a few compromises that we did not plan from the start. Maybe the greatest compromise was the choice to work with sectioned cells with an FFPE pre‐treatment and not with cultured grown cells that probably would have been the easiest choice. But we noticed some advantages with sectioned cells as long the project went along. The control kits release date was postponed more than once. However the last changes that we had to make were mostly for the best.
I believe and hope that the control kit is used and appreciated by as many costumers as possible. It is also my personal belief that the need for a control kit as a start up help for new costumers will grow by time as the in situ PLA technique grows to a established and well known method throughout research community.
21
6. Acknowledgements
First of all I would like to thank my supervisor Erik Nyström for all good help, encouragement and experiment planning. I would also like to give extra thanks to Eva Göhl who in much drove the project forward and gave us all inspiration. Furthermore I would extend my deepest thanks to the entire research and development group, Andrea Reyes, Ann‐ Catrin Andersson, Gabriella Edfeldt, and Göran Holmquist, who have with their knowledge and experience helped me much with my
laboratory work.
I am also very thankful for me research supervisor Göran Holmquist important and very thoroughly help with my report which have helped me very much.
Last but not least I would like to thank all employees at Olink Bioscience; Charlotta Göransson, Björn Ekström, Bonnie Tran, Heidi Schlingeman, Jörg Schlingeman, Maria Jungnelius, Mats Gullberg, Peter Karlberg and Simon Fredriksson. Together you make this company to a pleasant and stimulating place to work.
22
7. References
1. Söderber, O. Leuchowius, K‐J. Kamali‐Moghaddam, M. Jarvius, M. Gustafsdottir, S. Schallmeiner, E.
Gullberg, M. jarvius J. Landegren, U. Proximity Ligation: a specific and versatile tool for the proteonomic era. Genetic Engineering; 28 (2007); 85‐93.
2. Nilsson, M. Lock and roll: single‐molecule genotyping in situ using padlock probes and rolling –circle amplification. Histochem Cell Biol; 126 (2006); 159–164.
3. Duolink in situ PLA user manual; Art.no. 80010; revision: 3.0; 5. Uppsala.
4. Goldsby, R. Kundt, T. Osborne, B. Immunology fourth edition, Kuby (2001); ISBN 0716733315. New York.
5. J.A. Ramos‐vara. Technical Aspects of Immunohistochemistry, Vet Pathol; 42 (2005); 405‐426.
6. Rice, K, Thematic Review Application of Fluorescence Resonance Energy Transfer to Analyze Carbohydrates. Analytical Biochemistry; volume297, Issue 2 (2007); 117‐122.
7. Joly, E. Bioluminescence Resonance Energy Transfer (BRET 2 TM). Principle, Applications and Products.
Packard bioscience, biosignal packard INC; Online 2001; 1‐6; Retrieved from the Internet.
8. Duolink in situ PLA Control kit instruction;Doc. No. 0313; revision 1.2; 4. Uppsala.
9. Munch‐Petersen, B. Cloos, L. Jensen, HK. Tyrsted, G. Human thymidine kinase 1. Regulation in normal and malignant cells, Adv Enzyme Regul; 35(1995); 69‐89.
10. Welin, M. Kosinska, U. Mikkelsen, N‐E. Carnrot, C. Zhu, C. Wang, L. Eriksson, S. Munch‐Petersen, B.
Eklund, H. Structures of thymidine kinase 1 of human and mycoplasmic origin, Proc Natl Acad Sci U S A; 101(52) (2004); 17970–17975.
11. Boström, H. Leuchowius, K‐J. Hallböök, H. Nordgren, A. Thörn, I. Thorselius M. Rosenquist, R.
Söderberg O. Sundström, C. U‐2973, a novel B‐cell line established from a patient with a mature B‐
cell leukemia displaying concurrent t(14;18) and MYC translocation to a non‐IG gene partner.
European Journal of Haematology; 81; 3; 218‐225.
12. Berlund, M. Thunberg, U. Fridberg, M. Gjoörloff Wingren, A. Gullbo, J. Leuchowius, K‐J. Amini, R‐M.
Lagercrantz, S. Horvat, A. Enblad, A & Söderberg, O . Establishment of a cell line from a chemotherapy resistant diffuse large B‐cell lymphoma, Leukemia & Lymphoma; 48(5)(2007); 1038–1041.
23
13. Andesson, A C. Strömberg, S. Bäckwall, H. Kampf, C. Uhlen, M. Wester, K. Pontén, F. Analysis of protein expression in cell microarrays: a tool for antibody‐based proteomics, J. Histochem. Cytochem;
12(2006); 1413‐1423.
14. Larsson, U. Wanders, A. Wang, N. Tribukait, B. XPA‐210 – a new, universal, cytoplasmic cell proliferation marker for immunohistochemistry; staining data sheet (2007).
15. Stålberg, J. Affinitetsrenade antikroppar, IMS data sheet (2007); www.immunsystem.se; 1.
24
