Optimization of Analytical Procedure for Protein Analysis in Fresh Frozen Laser Microdissectioned Human Tissue

(1)

1

Bachelor Thesis Scheikunde

Optimization of Analytical Procedure for Protein Analysis

in Fresh Frozen Laser Microdissectioned Human Tissue

door

Naomi Uwugiaren

juni 2016

Studentnummer

10362053

Verantwoordelijk docent

Prof. dr. Garry Corthals

Begeleider

Dr. Irena Dapic

Onderzoeksinstituut

Van 't Hoff Institute for Molecular Sciences

Onderzoeksgroep

(2)

2

Abstract

Sample preparation in tissue proteomics represents a challenge due to the limited availability of the specimen, the low reproducibility of the current methods and possible sample loss during preparation steps. Extraction of proteins is easily influenced by the components of the extraction buffer and protein digestion might be incomplete.

Here we tested and optimized several protocols for protein extraction and digestion from fresh frozen laser capture microdissectioned human uterine tissue. The protocols varied in composition of extraction buffers by means of detergents (SDS or SDC), chaotropes (urea), organic solvents (ACN), and in the proteases used for digestion (trypsin or trypsin/Lys-C). The results were evaluated by means of the number of identified protein groups, number of identified peptides, and number of acquired spectra and by the physiochemical properties of the identified proteins.

The results showed that SDS and SDC were not completely removed from the samples. Samples extracted with SDS gave a lower number of identified protein groups compared to the extraction buffer containing urea. Samples extracted with SDC could not be analyzed due to precipitation of SDC. The extraction buffer containing 8 M urea and 60% ACN resulted in the highest number of protein group identifications. In addition, supplementing the trypsin digestion with Lys-C showed less protein group identifications than a trypsin digestion, whereas an addition of trypsin two consecutive times increased the protein group identifications by 17% compared to the digestion with the single addition of trypsin.

(3)

3

Populair wetenschappelijke samenvatting

De vraag naar het analyseren van eiwitten in weefsels is groot geworden, aangezien zij veel informatie kunnen bevatten over de processen die plaatsvinden in organismen op een moleculair niveau. De resultaten hiervan kunnen gebruikt worden voor het vinden van biomarkers en het ontwikkelen van medicijnen. Weefsels hebben echter een beperkte beschikbaarheid en zijn daardoor in kleine hoeveelheden beschikbaar. Een ander probleem dat kan optreden is het verlies van sample. Het doel van dit onderzoek was daarom om een analytische methode te optimaliseren voor het analyseren van de aanwezige eiwitten in kleine hoeveelheden menselijk baarmoederweefsel met behulp van een combinatie van vloeistofchromatografie en massaspectrometrie (MS).

Eerst zijn de eiwitten uit de samples gehaald door gebruik te maken van een extractiebuffer. Eiwitten zijn normaal gesproken gevouwen, maar ze moeten worden ontvouwen zodat ze goed toegankelijk zijn. Dit wordt gedaan door het verbreken van de ionbindingen, van der Waalsbindingen en waterstofbruggen tussen de eiwitten. Hiervoor kunnen detergentia en chaotropen worden gebruikt, aangezien zij die bindingen breken. Voorbeelden van detergentia zijn natriumdodecylsulfaat (SDS) en natriumdeoxycholaat (SDC) en een voorbeeld van een chaotroop is ureum. Het nadeel van SDS is dat deze moeilijk te verwijderen is uit de sample, wat kan leiden tot vervuilingen. Verder kan het toevoegen van een organisch oplosmiddel, zoals acetonitril, helpen in het ontvouwen van eiwitten. Daarna worden de eiwitten afgebroken tot peptiden met behulp van enzymen, ook wel proteasen genoemd. Dit wordt uitgevoerd omdat intacte eiwitten moeilijk te analyseren zijn met MS vanwege hun grootte. Trypsine wordt hiervoor veel gebruikt vanwege zijn hoge specificiteit en efficiëntie, maar deze knipt niet op alle mogelijke plaatsen. Daarom kan trypsine met andere proteasen worden gecombineerd, zoals Lys-C, zodat zij op de overige plaatsen knippen. Hierna worden de peptiden gescheiden met vloeistofchromatografie op basis van hun polariteit en geanalyseerd met MS, zodat hun sequenties bepaald kunnen worden. Door gebruik te maken van een database kan uiteindelijk worden bepaald welke eiwitten oorspronkelijk aanwezig waren in de sample. Deze stappen zijn geïllustreerd in Figuur 1.

(4)

4 In dit experiment zijn verschillende extractiebuffers getest die SDC, SDS of ureum bevatten. SDS en SDS konden niet volledig worden verwijderd, waardoor sommige samples niet konden worden geanalyseerd en andere samples een laag aantal eiwitten gaven. De extractiebuffer met ureum gaf de meeste aantal eiwitten en is daarom gekozen om verder te optimaliseren. Het toevoegen van acetonitril zorgde voor een verhoging in het aantal eiwitten. Het combineren van trypsine en Lys-C hielp niet in het verhogen van het aantal eiwitten, maar wat wel hielp was het twee keer toevoegen van trypsine.

(5)

5

List of abbreviations

ACN Acetonitrile

Arg Arginine

Asp Aspartic acid

BPC Base peak chromatogram

C18 Octadecyl hydrocarbon chain

CHAPS 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate CV Coefficient of variation Cys Cysteine DMSO Dimethylsulfoxide DTE Dithioerythritol DTT Dithiothreitol

ESI Electrospray ionization

FA Formic acid

FASP Filter-aided sample preparation

FF Fresh frozen

FFPE Formalin fixed paraffin embedded

Glu Glutamic acid

h Hour

HEPES 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid

His Histidine

IAA Iodoacetamide

ICAT Isotope-coded affinity tag

LC Liquid chromatography

LCM Laser capture microdissection

LC-MS/MS Liquid chromatography tandem mass spectrometry

LFQ Label-free quantification

LOD Limit of detection

Lys Lysine

M Molar

MALDI Matrix-assisted laser desorption/ionization

MeOH Methanol

min Minute

m-NBA meta-nitrobenzyl alcohol

MS Mass spectrometry

(6)

6

NP-40 Nonidet P-40

NSAF Normalized spectral abundancy factor

OCT Optimal cutting temperature

PEG Polyethylene glycol

Pro Proline

PVA Polyvinyl alcohol

QTOF Quadrupole time-of-flight

RIPA Radioimmunoprecipitation assay

ROI Region of interest

RP Reversed-phase

RT Room temperature

SDC Sodium deoxycholate

SDS Sodium dodecyl sulfate

SDS page Sodium dodecyl sulfate polyacrylamide gel electrophoresis

Ser Serine

SILAC Stable isotope labeling by amino acids in cell culture

SOP Standard operating procedure

SP3 Single-pot solid-phase-enhanced sample preparation

SPE Solid-phase extraction

TCA Trichloroacetic acid

TCEP Tris(2-carboxyethyl)phosphine

TFA Trifluoroacetic acid

TIC Total ion chromatogram

TOF Time-of-flight

Tris Tris(hydroxymethyl)aminomethane

(7)

7

8 2.2.3.5. Protocol 4.1, 4.3, 4.4, and 4.6 ... 25 2.2.3.6. Protocol 4.2 ... 25 2.2.3.7. Protocol 4.5 ... 25 2.2.3.8. Protocol 5.1.1 ... 25 2.2.3.9. Protocol 5.1.2 ... 26 2.2.3.10. Protocol 5.2.1 ... 26 2.2.3.11. Protocol 5.2.2 ... 26 2.2.3.12. Protocol 5.3 ... 26 2.2.3.13. Protocol 6 and 7 ... 27 2.2.4. Sample purification ... 27 2.3. Instrumental analysis ... 27 2.4. Software used... 29

2.4.1. Determination of the tissue areas ... 29

2.4.2. Database search ... 29

3. Results and discussion... 30

3.1. Optimization of the MS and MS/MS parameters ... 30

3.2. Evaluation of chromatographic conditions ... 31

3.2.1. Influence of packing material of the analytical columns on protein identifications ... 31

3.2.2. Influence of the chromatographic run time on protein identifications... 31

3.2.3. Influence of trap columns on protein identifications ... 32

3.2.4. Influence of the injected amount on protein identifications ... 33

3.3. Comparison and optimizing extraction buffers ... 33

3.3.1. Evaluation of different extraction buffers ... 33

3.3.2. Optimizing the extraction buffer ... 39

3.3. Distribution of identified proteins according to their molecular weight ... 41

3.4. Distribution of proteins according to their cellular location ... 43

3.5. Correlation of protein and peptide abundance among different protocols ... 44

4. Conclusion ... 46

Supplementary information ... 47

(9)

9

1. Introduction

Proteins as parts of cells, tissues, and organisms are important in metabolic pathways and biochemical processes. Since they are the products of genes, their study can play a role in the better understanding of the functions of genes and biological processes at molecular level. Proteomics which is the large scale study of proteins has the potential to lead to new tools in medical intervention. Retrieved information might be used for the development of personal responses in medical treatment and in biomedical research, such as drug screening and biomarker discovery.2

Tissues are a valuable source of information on processes involved in organisms at a molecular level. However, tissue samples are precious and unique and therefore their use is limited, leading to a need for analyzing small sample amounts. In addition, solving difficulties emerging from minimizing amount of starting material and sample loss during sample preparation remains a challenge.3_Therefore,

the aim of this study is to optimize the analytical procedure of the LC-MS method for the analysis of small tissue amounts from fresh frozen (FF) laser capture microdissectioned (LCM) human uterine tissue.

1.1. Tissue proteomics

In recent years, there has been an increasing interest in analyzing the protein content of tissue samples. Biobanks are used for the storage and collection of biological samples, including tissue samples. Advantages of biobanks are that they enable researchers to gain an easier access to samples and that the samples are stored according to standard operating procedures (SOPs).4_{SOPs assure that the samples}

are stored in a reproducible way with a high quality. Furthermore, research groups are able to compare samples, since their samples are preserved in the same way.5

Tissue samples are precious and therefore several methods for the preservation of tissues for later studies exist. First, tissues can be placed in formalin and then embedded with paraffin, which are defined as formalin fixed paraffin embedded (FFPE) tissues. This preservation method has been used routinely for a long time and an advantage of this method is that it is possible to store the samples at room temperature. However, it can lead to chemical modifications of the proteins due to crosslinking, because formaldehyde is able to react with the side chains of lysine (Lys), arginine (Arg), histidine (His), serine (Ser), and aspartic acid (Asp) residues.6_{Alternatively, tissues can be freshly frozen using optimal}

cutting temperature (OCT) compound or using liquid nitrogen or dry ice.5_{The samples are then stored}

at -80 °C, which can be disadvantageous to the costs when the samples are stored for a long time.7

Advantages of these methods are that the samples can be stored long-term and that the structures of the tissues are not changed as in FFPE tissues.8

OCT compound, which is used for the embedding of tissues, contains among other components polyvinyl alcohol (PVA) and polyethylene glycol (PEG), which can cause ion suppression and peaks in the spectrum that might cover the peaks of the peptides. Removing OCT compound would therefore

(10)

10 improve the MS analysis. Several studies investigating the removal of OCT compound have been conducted. Weston et al. carried out the removal of OCT compound with diethyl ether-MeOH precipitation, filter-aided sample preparation (FASP), and sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Inspection of the spectra after the removal confirmed that OCT compound was not present and that the peptide peaks were more evident after OCT removal, which resulted in an improved number of identified protein groups.8_{Another study by Enthaler et al. reported that washing}

tissue samples multiple times with EtOH and H2O was successful in the removal of OCT compound,

since it dissolves in polar solvents, but protein loss occurred.9

Small amounts of tissue samples can be obtained by LCM, which is a technique wherein a laser is operated that cuts tissues into sections.10

1.2. Techniques in protein research

Facing proteome complexity requires the implementation of different technologies for the analysis, with each of them showing advantages and limitations. A variety of techniques are used in proteomics for the profiling and characterization of single proteins or protein mixtures, the dynamic study of proteins, and the investigation of post-translational modifications.11_{Some of them are gel based (SDS-PAGE and}

2-D gel electrophoresis), which separate proteins according to their isoelectric points and/or molecular weights. However, these techniques have a low reproducibility and an insufficient separation of acidic, basic, and hydrophobic proteins.11_{Others of them are chemical isotope labeling methods (ICAT and}

SILAC), which allow the quantification of proteins.11

Proteomics using MS has enabled the detection of the protein content of tissues, due to the high sensitivity and efficiency of MS.10_{Proteins can be studied in their intact form or as a mixture of peptides}

obtained after digestion. While there is no single approach that provides all the necessary information, there are several strategies used in protein studies: top-down proteomics, middle-down proteomics, and bottom-up proteomics (Figure 1).

Figure 1. An overview of the three sub-groups of proteomics: top-down (a), middle-down (b), and

(11)

11 In top-down proteomics, intact proteins are analyzed, though the performance is limited as it is difficult to ionize and fragment entire proteins in MS due to their high molecular weight (MW). On the contrary, in bottom-up (shotgun proteomics) and middle-down proteomics proteins are digested into peptides.12

Since peptides are more easily ionized and fragmented than proteins, this approach faces less difficulties. After the digestion, the peptides are analyzed by MS and identified using bioinformatics tools. Subsequently, the proteins are identified using database searches.13_{The difference between}

middle-down and bottom-up proteomics is that middle-middle-down proteomics uses peptide fragments with a size of 2000-20000 Da, while bottom-up proteomics uses fragments of 500-3000 Da.12

1.3. Bottom-up proteomics

The past decade has seen the rapid development of bottom-up proteomics, mostly with the goal of discovering protein biomarkers. Identification of peptides from MS/MS spectra requires high resolution and highly accurate tandem mass spectra, since peptide identification is achieved by the comparison of experimental spectra of peptide fragments with theoretical ones. However, the main disadvantages of bottom-up proteomics, such as a low reproducibility and sensitivity, still need to be overcome. As a result, not all proteins which are present in the sample can be identified, as shown in Figure 2.

Figure 2. Overview of correlation between the number of proteins in a sample, protein identified, and

proteins quantified in MS-based proteomics.14

Bottom-up protein analysis is a multistep process and a consistent and reproducible procedure still represents an obstacle. Sample preparation is widely accepted as one of the critical steps, since small differences in the processing of specimen can largely influence results. In the next Section, an outline of the sample preparation and LC-MS analysis of bottom-up proteomics will be discussed as well as its difficulties.

(12)

12 Figure 3 provides the workflow of bottom-up proteomics. It consists of the following steps: extraction of the proteins, reduction of the disulfide bonds formed between cysteine (Cys) residues, alkylation of the Cys residues, digestion of the proteins into peptides, sample purification, and analysis with LC-MS.

Figure 3. An overview of the standard workflow in bottom-up proteomics. First, the proteins are

extracted, followed by reduction of the disulfide bonds and alkylation of the Cys residues. Then, the proteins are digested into peptides and sample purification takes place. The peptides are analyzed with LC-MS.

1.3.1. Protein extraction

Protein extraction is usually one of the first steps in sample preparation in bottom-up proteomics. The cells are lysed and the proteins are extracted from the cells using an extraction buffer. The solubility of the proteins needs to be promoted, because they are frequently insoluble in their native form. The solubility might be facilitated by the unfolding of proteins which is achieved by disrupting the hydrophobic interactions, hydrogen bonds, disulfide bonds, and ionic interactions between amino acid residues and proteins.15_{The solubilization and denaturation of proteins might be promoted by the}

addition of different chaotropes, detergents, organic solvents, reducing agents, and salts to extraction buffers.

1.3.1.1. Buffers and salts

Buffers are added to extraction buffers, since they maintain an alkaline pH which favors protein stability.16_{Commonly used buffers for protein extraction are ammonium bicarbonate,}

4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES), radioimmunoprecipitation assay (RIPA), and trisaminomethane hydrochloride (Tris HCl).17_{The often used RIPA buffer contains 150 mM NaCl, 0.5%}

sodium deoxycholate (SDC), 0.1% sodium dodecyl sulfate (SDS), and 50 mM Tris HCl and is highly efficient in the solubilization of proteins due to the combination of the detergents SDC and SDS.18

Protein extraction Reduction Alkylation Digestion Sample purification LC-MS analysis

(13)

13 Salts are used to control the pH and ionic strength of the extraction buffer.19_{Furthermore, they}

can either increase or decrease the solubility of proteins due to the ionic strength of the buffer. At a high ionic strength, water molecules form interactions with the salt ions instead of the proteins. As a result, proteins form hydrophobic interactions with each other, followed by their precipitation (salting-out).17

At a low ionic strength however, the solubility of the proteins is increased. The ions surround the charged amino acid residues and as a result, disruption of the ionic interactions between the residues will take place and ionic interactions will form between the residues and ions. Consequently, the stabilization and solubilization of the proteins is increased (salting-in).17_{Salts can be divided into salting-in ions, neutral}

ions, and salting-out ions. Salting-in ions weaken hydrophobic interactions and therefore increase the solubility of proteins. Salting-out ions strengthen hydrophobic interactions between proteins and are therefore used to precipitate proteins, whereas salting-in or neutral ions are used for the solubilization of proteins. Examples of salting-in salts used in extraction buffers are ammonium bicarbonate, KCl, NaCl, and Tris HCl.7,17,20,21

1.3.1.2. Detergents

Detergents are used to disrupt hydrophobic interactions, since they form micelles in aqueous media. Hydrophobic molecules dissolve in the micelles, leading to an increased solubility of the proteins.16

Detergents are classified into three types: ionic, non-ionic, and zwitterionic. The latter type of detergents means that the compound contains both positive and negative charged groups. Ionic detergents are more effective in the solubilization of proteins than non-ionic and zwitterionic detergents, because their charged head is able to interact with charged amino acid residues and is therefore able to break ionic interactions.22_{In addition, non-ionic and zwitterionic detergents form adducts with proteins and give}

signal suppression making them less preferred than ionic detergents.23_{SDS is a commonly used ionic}

detergent and is very efficient in solubilizing hydrophobic and amphipathic membrane proteins15_;

though a major drawback is that it is incompatible with LC-MS analysis and that it can give significant interferences during the analysis. In addition, the concentration of SDS must be reduced prior to digestion, since the activity of the commonly used protease is decreased in the presence of concentrations higher than 0.1% SDS.24

The removal of SDS often leads to sample loss and thus, there has been an increasing interest in the use of other detergents. Proc and co-workers have reported the ionic detergent SDC as a good alternative for SDS with a similar efficiency, but with an easier removal.25_{This is in agreement with the}

results of Leon et al. They carried out the digestion with extraction buffers containing 5% SDC, and 8 M urea, and found that SDC resulted in more identified proteins compared to urea.26_{The structures of}

(14)

14

Figure 4. Structures of the ionic detergents SDS and SDC.

Other alternatives for SDS are the zwitterionic detergent CHAPS and the non-ionic detergents Nonidet P-40 (NP-40) and Triton X-100 (Figure 5), but they have a lower efficiency in the solubilization of proteins compared to SDS. Studies have shown that they can however be added to the buffer in order to complement the protein extraction.7,27

Figure 5. Structures of the zwitterionic detergent CHAPS and the non-ionic detergents NP-40 and Triton

X-100.

Also, several MS-compatible detergents have been developed, such as RapiGest SF28_,

MaSDeS22_{and ProteaseMAX}29_{(Figure 6). These detergents are acid-labile, meaning that they hydrolyze}

under acidic conditions. The products do not interfere with MS analysis and do not show ion suppression.29_{Several studies have reported a similar or higher protein extraction efficiency of the}

MS-compatible detergents compared to SDS, but their main disadvantage is their cost.30,31

(15)

15

1.3.1.3. Chaotropes

Chaotropes disrupt intermolecular hydrogen bonds and hydrophilic interactions and are therefore used to promote the denaturation of proteins. Examples of chaotropes include urea, thiourea, and guanidine HCl. However, the activity of trypsin is decreased in the presence of urea with a concentration higher than 4 M or guanidine HCl with a concentration higher than 0.1 M.32,33_{Therefore, the concentration of}

chaotropes needs be reduced before the addition of trypsin.

Furthermore, incubation of samples containing urea at a high temperature should be avoided, since proteins are more prone to carbamylation. Urea is in equilibrium with ammonium cyanate and high temperatures will shift the equilibrium to the side of ammonium cyanate. The cyanate anion can be protonated to form isocyanic acid, leading to the carbamylation of the amino and sulfhydryl groups of amino acids present in proteins (Scheme 1).34

Scheme 1. Reaction scheme of the carbamylation of an amino group with isocyanic acid. R: side chain.

Consequently, the digestion of carbamylated proteins is hindered, which affects the digestion efficiency. In addition, the retention times and intensity of the peaks and masses of the ions are changed, resulting in a more complicated analysis of the peptides. In order to prevent this, cyanate scavengers are added to the extraction buffer, such as ammonium bicarbonate, methylamine, and Tris HCl.35

Urea has been successfully used in extraction buffers for the protein extraction of FF and FFPE tissues.36–38_{A combination of SDS and urea can be used to supplement each other, since they extract}

different proteins.39–41

1.3.1.4. Solvent-assisted digestion

The addition of an organic solvent to the extraction buffer has the benefit that the denaturation of proteins is promoted, leading to an increased solubility of the proteins. Another advantage is that the removal of organic solvents is easy to accomplish through lyophilization. Furthermore, the addition of an organic solvent can help in the extracting of membrane proteins, because they have an increased solubility in organic-aqueous buffers compared to aqueous buffers.42

Russell et al. investigated the tryptic digestion of multiple proteins using various concentrations of MeOH, ACN, 2-propanol, and acetone as organic solvents. They found that the sequence coverage of the proteins was increased using the organic buffers compared to an aqueous buffer.43_{Another study}

by Zhang et al. used two extraction buffers containing 60% MeOH or 1% SDS for the tryptic digestion of E. coli. They reported that using the extraction buffer containing 60% MeOH resulted in a higher

(16)

16 number of protein identifications and a higher number of unique identified proteins compared to the extraction buffer containing 1% SDS.44

It is still unclear what the optimal concentrations of organic solvents are, since several studies have shown different results. A study comparing the tryptic digestion efficiency of extraction buffers containing 6 M guanidine HCl, 80% ACN, or 0.1% RapiGest was carried out by Hervey and co-workers. They found that the number of identified peptides was the highest when using 80% ACN.45_{In contrast,}

Wall et al. demonstrated that the enzyme activity of trypsin was decreased in the presence of 80% ACN. They concluded that this was the result of autolysis and denaturation of trypsin, resulting in deactivation of the enzyme.46

1.3.2. Reduction and alkylation

Reduction of disulfide bonds between Cys residues to thiol groups might be performed using reducing agents, such as dithiothreitol (DTT), tris(2-carboxyethyl)phosphine (TCEP), β-mercaptoethanol, or dithioerythritol (DTE). Breaking of disulfide bonds leads to the denaturation of proteins, which in turn improves their solubilization. DTT is one of the mostly used reducing agents, because it is a strong reducing agent and it prevents the reformation of disulfide bonds.47

Subsequently, the thiol groups of the Cys residues are alkylated to prevent the reformation of disulfide bonds. This is generally achieved by the addition of iodoacetamide (IAA) or iodoacetic acid.16,47_{The reaction of the reduction of disulfide bonds and the alkylation of Cys residues is shown in}

Scheme 2.

Scheme 2. Reaction scheme of the reduction of a disulfide bond between two Cys residues, followed by

the alkylation of the thiol groups with an alkylating agent to prevent the reformation of disulfide bonds.47

1.3.3. Protein digestion

The cleavage of proteins into peptide fragments can be achieved with proteases or through chemical cleavage and this Section will only focus on proteolytic cleavage. Additionally, several methods that have been developed to accelerate the digestion time will be explored.

Proteases are used to cleave proteins into peptides. Trypsin is the most widely used protease for protein digestion due to its high specificity and efficiency, which is essential for protein identification. Furthermore, peptides produced by trypsin are in favorable length for MS fragmentation. The optimal pH for trypsin activity is 7.5-8.5. Trypsin cleaves at the carboxyl side of Lys and Arg residues, but not when proline (Pro) is linked on the carboxyl side of Lys and Arg or when Asp is N-linked to Lys and

(17)

17 Arg.13_{Lys and Arg are one of the most abundant amino acids in the human body}12_{and are well}

distributed in a protein,33_{creating peptides with a favorable length for MS fragmentation.}48

The susceptibility of trypsin to autolysis has been reduced using laboratory modified trypsin that is highly resistant to autocatalytic reactions.49_{Moreover, the autolysis of trypsin can be decreased by}

the addition of calcium ions when their natural concentration in samples in low.47

However, a key problem is that the efficiency of trypsin might be easily affected by the other reagents, such as urea, and that it sometimes misses cleavage sites, resulting in miscleaved peptide fragments that are not reproducible and difficult to predict. This can lead to the miscalculation of the actual occurrence of the peptides since it is possible that the miscleaved fragments are too long to detect in the MS.49

Several attempts have been carried out to improve the digestion efficiency. Fang and co-workers tried to reduce the number of nonspecific trypsin cleavages by investigating the effect of the protein to trypsin ratio. They found that a higher ratio resulted in a higher number of nonspecific trypsin cleavages. This may be explained by the fact that the chance of a trypsin molecule encountering another trypsin molecule is higher at an increased concentration of trypsin, resulting in a higher autolytic rate.50_In

addition, several studies have investigated the effects of the digestion time on the efficiency of trypsin, as autodigestion of trypsin can occur with a long digestion time. Klammer et al. and Proc et al. suggested a digestion time shorter than the usual overnight digestion, since it resulted in the same number of protein identifications. 25,32

1.3.3.1. Multiple enzyme digestion

However, sometimes trypsin digestion leaves cleavage sites uncleaved and therefore, two mostly used proteases to supplement the trypsin digestion are chymotrypsin and Lys-C.48_{Furthermore, trypsin has a}

lower cleavage efficiency towards Lys compared to Arg and to prevent this, trypsin might be combined with Lys-C, which cleaves at the carboxyl end of Lys residues.48_{Another useful characteristic of Lys-C}

is that it has a similar pH range of activity as trypsin and that it maintains its activity in harsh conditions as 8 M urea.32_{As a consequence, the digestion conditions do not need to be changed and the same}

extraction buffer can be used. Multiple studies have shown that a combination of trypsin and Lys-C results in a higher digestion efficiency and more identified proteins compared to a trypsin only digestion.20,51–53_{Additionally, the number of miscleaved peptides decreased when using trypsin/Lys-C}

instead of trypsin only.20,53

1.3.3.2. Techniques to accelerate the digestion

Several studies have been carried out on accelerating the tryptic digestion in order to shorten the overall preparation time, since the digestion is usually conducted overnight. In this Section, several techniques to accelerate and improve the protein extraction and digestion will be described.

(18)

18 It has been shown that microwave-assisted digestion decreases the tryptic digestion time to less than 30 min and that it increases the digestion efficiency.54–56_{Microwave irradiation accelerates the}

digestion in the first minutes of the reaction, followed by the denaturation and inactivation of trypsin after 30 min. The mechanism remains unclear, but it is thought that microwave irradiation helps in the unfolding of proteins which provides better accessibility to the cleavage sites.57

In addition, it has been demonstrated that ultrasonic energy speeds up the digestion by increasing the diffusion rate.58_{An ultrasound bath is not powerful and efficient enough to accelerate the enzymatic}

digestion59_{, but developed alternatives are ultrasonic probes and sonoreactors.}58,60_{They are able to}

reduce the digestion time to less than five minutes.

Yang et al. carried out a pressure-assisted tryptic digestion using a syringe with a pressure of 6 atm that completed the digestion in 30 min.61_{An advantage of this method is that the same sample}

preparation steps were used as without the syringe. An explanation for the decreased digestion time is that an elevated pressure increases the number of collisions between proteins and trypsin molecules, speeding up the digestion.

1.3.4. Sample purification

Following the digestion, the samples need to be purified from compounds as buffers, salts, and detergents, since they can interfere with LC-MS. Presence of these contaminants might cause ion suppression, which negatively influences the sensitivity, accuracy, and precision of the analysis. Sample purification might be achieved by solid-phase extraction (SPE) or C18 ZipTips. ZipTips are pipette tips consisting of C18 material, which allows purification of the sample. Disadvantages are that sample loss might occur and that a limited amount of sample can be loaded.47

As mentioned before, the removal of SDS from protein samples remains a challenge. SDS can be removed using precipitation methods or dialysis, however protein recoveries are often low. Therefore, several new methods to remove SDS have been developed. For example, Zhou et al. reported that no sample losses occurred when using KCl to precipitate SDS in the form of KDS,62_{and Wisniewski and}

co-workers developed filter-aided sample preparation (FASP) which implements a filtration device to remove SDS.63_{Puchades et al. compared three methods to remove SDS: protein precipitation with}

acetone, SDS precipitation with chloroform/MeOH/water, and SPE. They found that the method using SPE was not able to sufficiently remove SDS and gave a protein recovery of 50%. The acetone and chloroform/MeOH/water methods were able to reduce SDS to a concentration that allowed MS analysis, giving protein recoveries of 80% and 50%, respectively. Their conclusion was to use acetone since it showed the largest recovery. Additionally, in a study carried out by Kachuk et al., seven methods for SDS removal were compared: SDS page, protein precipitation with acetone, protein precipitation with TCA (trichloroacetic acid), detergent precipitation with KCl, strong cation exchange, SPE, and FASP.64

(19)

19 lowest protein recoveries (< 40%). FASP removed SDS completely, but the protein recovery was shown to be 24-40%. SDS page and strong cation exchange removed SDS to a large extent, but the protein recoveries were lower than using acetone, which had a SDS removal efficiency similar to SDS page and strong cation exchange. Therefore, they suggested protein precipitation with acetone, which is consistent with the results of Puchades et al. However, another disadvantage of acetone precipitation besides sample loss is that the resulting pellet has a low solubility making it hard to dissolve the precipitated proteins.65

A method was developed to prevent sample loss in which the entire preparation is carried out in a single tube, called single-pot solid-phase-enhanced sample preparation (SP3).66_{The tube is filled with}

paramagnetic beads, which are coated with hydrophilic carboxylate material. The carboxylate surface has a neutral charge in acidic conditions and a negative charge in basic conditions, which will allow the binding of charged proteins. Furthermore, when an organic solvent is added to the aqueous buffer, a hydrophilic layer will form on the magnetic beads, since the aqueous buffer and beads are attracted to each other. The analytes will distribute between the hydrophilic layer and the organic buffer. Proteins and peptides bind to the hydrophilic layer through hydrophilic interactions and ionic interactions with the beads. Then, contaminants, such as detergents and chaotropes, are removed by changing the composition of the organic buffer. Proteins and peptides are eluted by decreasing the organic component of the buffer and the pH. The principle of SP3 is shown in Figure 7.

(20)

20

1.3.5. LC-MS analysis

After the sample preparation, the instrumental analysis is an important part. Different instruments have been used for the separation and detection of peptides in bottom-up proteomics. In this thesis, nano LC-ESI-MS has been used for all the sample analysis and therefore, only this technique will be discussed in more details.

To decrease sample complexity, peptides might be separated with reversed phase (RP) nano liquid chromatography (LC) due to their different chromatographic behavior arising from their amino acid sequences.68_{In addition, the small flow rates ensure a better ionization of the peptides, because the}

created droplets have a higher surface-to-volume ratio.69

The LC-system is hyphenated with MS where ionization and fragmentation of the peptides occur. Peptides are ionized using electrospray ionization (ESI), since it is a soft ionization technique causing little fragmentation of the ions due to their low internal energy.70_{ESI might be used for the}

analysis of complex samples, such as peptides.71_{Several studies have suggested that the sensitivity of}

ESI might be increased by the addition of the supercharging reagents dimethylsulfoxide (DMSO)72,73

and meta-nitrobenzyl alcohol (m-NBA)72_{to the LC buffers. Supercharging reagents enhance charging}

on the peptides and thereby the fragmentation.72_{Meyer and Komives digested five proteins using pepsin,}

elastase, and trypsin and carried out the LC-MS analysis in the presence of 5% DMSO and found that the number of identified peptides increased.72_{DMSO causes charge state reduction and coalescence in}

one charge state, resulting in an increase of the signal of that state, which leads to a simpler precursor spectrum.72,73_{ESI is mostly coupled to hybrid tandem MS systems as quadrupole time-of-flight (QTOF)}

systems, which show a high resolution and mass accuracy. In this work, TripleTOF 5600+ MS was used, a hybrid quadrupole time-of-flight system that shows a high resolution, fast acquisition rates, and a high sensitivity.74_{An initial scan of the peptide fragments (MS1) is acquired in Q1 and then these fragments}

are further fragmented in Q2 and scanned (MS/MS).74_{Subsequently, the MS/MS spectra are used by}

bioinformatics tools for structure elucidation and amino acid sequence determination.

The peptides can be fragmented on their amino terminus or carboxyl terminus. The fragmentation of peptides can take place at three different cleavage sites (Figure 8). The bond between the alpha carbon and the carbonyl carbon can be broken, which creates a-ions and x-ions. Next, the bond between the carbonyl carbon and amide nitrogen can be broken creating b-ions and y-ions. Peptides are usually fragmented in this manner. Finally, the bond between the amide nitrogen and alpha-carbon can be broken, which creates c-ions and x-ions.75

(21)

21

1.3.6. Protein identification

Acquired data is further processed using bioinformatics tools. Obtained MS/MS spectra are compared with theoretical fragment ion MS/MS spectra using a search engine to identify peptides. Several search engines exist, such as Protein Pilot (AB Sciex), Mascot76_{, and SEQUEST}77_{. These programs have}

different algorithms, and therefore the identifications might differ. After the peptides have been identified, their sequences are compared to the sequences of known proteins to determine which proteins were present in the sample.78

Label-free quantification (LFQ) is one of the mostly used approaches in quantitative proteomics. LFQ can be divided into two strategies. In the first strategy, the peaks of the extracted ion chromatograms (XICs) are integrated over the retention time to calculate the peak area, which indicates the abundancy of the peptides.14_{The second strategy is known as spectral counting. The number of}

MS/MS spectra are counted which are a measurement of the abundancy of the peptides. However, the spectral count of a large peptide will be higher than the spectral count of a small peptide and therefore must be corrected. This is accomplished by using the normalized spectral abundancy factor (NSAF), which divides the spectral counts by the length of the protein and then divides it by the sum of all spectral counts divided by the length of all proteins.79_{An example of software packages to calculate LFQ values}

is MaxQuant.80

1.4. Aim of the research

The aim of this study was to optimize an analytical method for the determination of proteins from small amounts of laser microdissectioned OCT embedded human uterine tissue. Sample preparation will be optimized by testing several extraction buffers, consisting of different detergents, chaotropes, and a varied amount of organic solvent for the protein extraction. Furthermore, the impact of different proteases on the protein digestion will be examined. After sample preparation, samples will be analyzed on nano LC-ESI-MS using optimized instrumental parameters. Finally, the different protocols will be compared in terms of protein identifications, physiochemical properties, and acquired LFQ values.

(22)

22

2. Materials and methods

2.1. Chemicals

ACN (LC-MS grade) and formic acid (FA, 99%) were obtained from Biosolve and NH4HCO3 was

delivered by Fluka Analytical. Lys-C (Mass Seq Grade) was bought from Promega. KCl was purchased from Merck and Tris HCl was delivered by Roche. Acetone (≥ 99.9%), CaCl2, DTT (≥ 99.0%), IAA (≥

99%), NaCl, SDC (≥ 98.0%), SDS (≥ 98.5%), thiourea (≥ 99.5%), trifluoracetic acid (TFA, 99%), trypsin (European Pharmacopoeia Reference Standard), and urea were obtained from Sigma Aldrich.

2.2. Sample preparation

2.2.1. Uterus tissue samples

Cryosectioned fresh frozen tissues of smooth muscle from uterus with a thickness of 16 µm were used for the optimization of the analytical procedure. Using a scalpel (Martor KG, Germany), small regions of interest (ROIs) were cut from the tissue samples (Figure 9). Variations between the ROIs selected for the optimization of the analytical procedure were kept to a minimum in order to ensure that they were homogeneous.

Figure 9. Example of a glass slide of 16 µm cryosectioned uterine tissue. Each spot was cut into 6

(23)

23

2.2.2. Protein extraction and digestion

Proteins from uterine tissue samples were extracted using the buffers shown in Table 1 and the experimental protocols are in details explained in Section 2.2.3. Information on the area of the samples, the protocols used for sample preparation, sample purification methods, amount of sample analyzed in LC-MS, and chromatographic conditions is summarized in Supplementary table 1.

Table 1. An overview of the extraction buffers that were tested with their composition. Protocol Composition of the extraction buffer

1 1. 5% SDC 2. 7.5 mM DTT 2 1. 50 mM Tris HCl 2. 150 mM NaCl 3. 0.25% SDC 3 1. 4% SDS 2. 0.1 M Tris HCl 4 1. 1.44 g urea 2. 10 µL 1 M NH4HCO3 3. 2.8 µL 700 mM DTT

4. 30% ACN in 100 mM NH4HCO3 to a total volume of 2 mL (4.1)

5 1. 1.44 mg urea

2. 200 µL of 1 M NH4HCO3

3. 56 µL of 700 mM DTT

4. 30% ACN in 100 mM NH4HCO3 to a total volume of 2 mL

6 1. 80% ACN in 50 mM NH4HCO3

7 1. 60% MeOH in 50 mM NH4HCO3

2.2.3. Protocols

Protocols tested for sample processing are explained in detail here. Centrifugation was carried out with a centrifuge (Centrifuge 5804 R, Eppendorf), incubation was performed with an incubator (Thermomixer Compact, Eppendorf), sonicating was performed with in an ultrasonic bath (Transsonic Digital S, Elma Schmidbauer GmbH), and vortexing was carried out with a shaker (Vortex 4 basic, IKA).

(24)

24

2.2.3.1. Protocol 1

100 µL of Buffer 1 was added to the sample, followed by sonication for 30 min at RT and incubation for 10 min at 85 °C. Then, 50 µL of 45 mM DTT was added and incubated for 20 min at 60 °C, followed by the addition of 50 µL of 100 mM IAA and incubation in the dark for 30 min at RT. Subsequently, the sample was diluted with H2O up to a final volume of 1 mL. 0.7 µL of 0.1 µg/µL of trypsin solution

was added. The digestion was carried out for 17h at 37 °C. Next, 5 µL of 10% TFA was added to the sample. The sample was centrifuged for 45 min at 13000 rpm and the supernatant was collected.

2.2.3.2. Protocol 2

100 µL of Buffer 2 was added to the tissue sample, followed by sonication for 30 min at RT and incubation at 4 °C (Protocol 2.1) or 85 °C (Protocol 2.2). Then, the sample was centrifuged at 13000 rpm for 30 min. The proteins were precipitated with 20% TFA for 30 min at 4 °C, followed by centrifugation at 13000 rpm for 10 min and the decantation of the supernatant. 50 µL of 45 mM DTT and 50 µL of H2O were added to the precipitated proteins. The sample was incubated for 30 min at 37

°C, followed by the addition of 50 µL of 12 mM IAA and incubation for 30 min at 37 °C in the dark. Subsequently, 100 µL of a digestion buffer was added (100 mM Tris HCl, 6 M urea, 2 M thiourea, and 0.5% SDS) and the sample was diluted to a final volume of 1 mL with a dilution buffer (100 mM Tris HCl, 10 mM CaCl2). 0.7 µL of 0.1 µg/µL trypsin was added and the digestion was carried out for 17h

at 37 °C. 5 µL of 10% TFA was added to the sample. The sample was centrifuged for 45 min at 13000 rpm and the supernatant was collected.

2.2.3.3 Protocol 3.1

The sample was mixed with 100 µL of Buffer 3 and sonicated for 30 min. Subsequently, the sample was incubated for 10 min at 85 °C. 50 µL of 10 mM DTT was added and the incubation was carried out for 60 min at 37 °C. Next, 50 µL of 20 mM IAA was added and the sample was incubated for 30 min in the dark at RT. 1.2 mL of cold acetone was added, followed by incubation for 20 min at 4 °C and centrifugation for 10 min at 13000 rpm. 100 µL of a denaturing buffer (8 M urea and 50 mM NH4HCO3)

was added to the pellet and the sample was diluted with 50 mM NH4HCO3 to a final volume of 1 mL.

0.7 µL of 0.1 µg/µL trypsin was added and the digestion was carried out for 17h at 37 °C. Next, 5 µL of 10% TFA was added to the sample. The sample was centrifuged for 45 min at 13000 rpm and the supernatant was collected.

2.2.3.4. Protocol 3.2

The sample was mixed with 100 µL of Buffer 3, followed by sonication and incubation for 10 min at 85 °C. Subsequently, 50 µL of 10 mM DTT was added and the incubation was carried out for 60 min at 37 °C. Next, 50 µL of 20 mM IAA was added and the sample was incubated for 30 min in the dark at RT. 200 µL of 0.5 M KCl was added and the sample was incubated for 5 min at RT and centrifuged for 10 min at 13000 rpm. 100 µL of a denaturing buffer (8 M urea and 50 mM NH4HCO3) was added to the

(25)

25 supernatant. Subsequently, the sample was diluted with 50 mM NH4HCO3 to a final volume of 1 mL.

0.7 µ of 0.1 µg/µL trypsin was added and the digestion was carried out for 17h at 37 °C. 5 µL of 10% TFA was added. The sample was centrifuged for 45 min at 13000 rpm and the supernatant was collected.

2.2.3.5. Protocol 4.1, 4.3, 4.4, and 4.6

100 µL of Buffer 5 was added to the tissue sample, followed by sonication for 30 min at RT and incubation for 30 min at 37 °C. Next, 9.2 µL of 700 mM IAA was added and the incubation was carried out for 30 min at 37 °C in the dark. The sample was diluted with 120 µL of 1 M NH4HCO3 and 880 µL

H2O. 7 µL of 0.01 µg/µL trypsin was added and the digestion took place for 17h at 37 °C. Next, the

sample was centrifuged for 45 min at 13000 rpm and 48 µL of 5% TFA was added to the supernatant.

2.2.3.6. Protocol 4.2

The tissue sample was mixed with 100 µL of Buffer 5 and the sample was sonicated for 30 min and incubated for 30 min at 37 °C. Next, 9.2 µL of 700 mM IAA was added and the sample was incubated for 30 min at 37 °C in the dark. The sample was diluted with 12 µL of 1 M NH4HCO3 and 88 µL H2O.

10 µL of 0.1 µg/µL Lys-C was added and the digestion took place for 3h at 37 °C. Next, the sample was further diluted with 108 µL of 1 M NH4HCO3 and 792 µL H2O. 7 µL of 0.01 µg/µL trypsin was added

and the digestion took place for 17h at 37 °C. The sample was centrifuged for 45 min at 13000 rpm and 48 µL of 5% TFA was added to the supernatant.

2.2.3.7. Protocol 4.5

100 µL of Buffer 5 was added to the tissue sample, followed by sonication for 30 min at RT and incubation for 30 min at 37 °C. 9.2 µL of 700 mM IAA was added and the incubation was carried out for 30 min at 37 °C in the dark. 120 µL of 1 M NH4HCO3 and 880 µL H2O were added to the sample. 7

µL of 0.01 µg/µL trypsin was added and the digestion took place for 17h at 37 °C. Subsequently, another 7 µL of 0.01 µg/µL trypsin was added and the digestion took place for 6h at 37 °C. Subsequently, the sample was centrifuged for 45 min at 13000 rpm. 48 µL of 5% TFA was added to the supernatant.

2.2.3.8. Protocol 5.1.1

The tissue sample was mixed with 100 µL of Buffer 5 and the sample was sonicated for 30 min at RT and incubated for 30 min at 37 °C. Next, 9.2 µL of 700 mM IAA was added and the incubation was carried out for 30 min at 37 °C in the dark. 120 µL of 1 M NH4HCO3 and 880 µL H2O were added to

the sample. 7 µL of 1 µg/µL trypsin was added and the digestion took place for 17h at 37 °C. Subsequently, the sample was centrifuged for 45 min at 13000 rpm and 48 µL of 5% TFA was added to the supernatant.

(26)

26

2.2.3.9. Protocol 5.1.2

100 µL of Buffer 5 was added to the tissue sample, followed by sonication for 30 min at RT and incubation for 30 min at 37 °C. Next, 9.2 µL of 700 mM IAA was added and the incubation was carried out for 30 min at 37 °C in the dark. The sample was diluted with 120 µL of 1 M NH4HCO3 and 880 µL

H2O. 0.7 µL of 0.1 µg/µL trypsin was added and the digestion took place for 18.5h. After the digestion

had completed, 1 (sample 5c-2) or 2 µL (sample 5c-3) of 0.1 µg/µL Lys-C was added. The digestion took place for 4h at 37 °C. Subsequently, the sample was centrifuged for 45 min at 13000 rpm and 48 µL of 5% TFA was added to the supernatant.

2.2.3.10. Protocol 5.2.1

100 µL of Buffer 3 was added to the tissue sample, followed by sonication for 30 min at RT and incubation for 10 min at 85 °C. 100 µL of 0.5 M KCl was added and the incubation was carried out for 5 min at RT. Next, the sample was centrifuged at 13000 rpm for 10 min and the supernatant was collected. 100 µL of buffer 5 was added to the sample. Subsequently, the sample was sonicated for 30 min at RT and incubated for 30 min at 37 °C. Next, 9.2 µL of 700 mM IAA was added and the sample was incubated for 30 min at 37 °C in the dark. 120 µL of 1 M NH4HCO3 and 880 µL H2O were added,

followed by the addition of 0.7 µL of 0.1 µg/µL trypsin and the digestion took place for 17h at 37 °C. The sample was centrifuged for 45 min at 13000 rpm and 48 µL of 5% TFA was added to the supernatant.

2.2.3.11. Protocol 5.2.2

100 µL of Buffer 3 was added to the tissue sample, followed by sonication for 30 min at RT and incubation for 10 min at 85 °C. 100 µL of 0.5 M KCl was added and the incubation was carried out for 5 min at RT. The sample was centrifuged at 13000 rpm for 10 min and then, 100 µL of Buffer 5 was added to the supernatant. Subsequently, the sample was sonicated for 30 min at RT and incubated for 30 min at 37 °C. Next, 9.2 µL of 700 mM IAA was added and incubated for 30 min at 37 °C in the dark. 120 µL of 1 M NH4HCO3 and 880 µL H2O were added, followed by the addition of 0.7 µL of 0.1 µg/µL

trypsin and the digestion took place for 17h at 37 °C. 1 µL of 0.1 µg/µL C was added and the Lys-C digestion took place for 4h at 37 °Lys-C. The sample was centrifuged for 45 min at 13000 rpm. 48 µL of 5% TFA was added to the supernatant.

2.2.3.12. Protocol 5.3

100 µL of Buffer 5 was added to the tissue sample, followed by sonication for 30 min at RT and incubation for 10 min at 37 °C. Subsequently, 50 µL of the sample was taken out and mixed with 100 µL of Buffer 3, followed by sonication for 30 min at RT and incubation for 10 min at 85 °C. 150 µL of 0.5 M KCl was added and incubated for 5 min at RT. The sample was centrifuged for 10 min at 13000 rpm and the supernatant was added to the first 50 µL of the sample. Next, 9.2 µL of 700 mM IAA was added and incubated for 30 min at 37 °C in the dark. 120 µL of 1 M NH4HCO3 and 880 µL H2O were

(27)

27 added, followed by the addition of 0.7 µL of 0.01 µg/µL trypsin and the digestion took place for 17h at 37 °C. Subsequently, the sample was centrifuged for 45 min at 13000 rpm and 48 µL of 5% TFA was added to the supernatant.

2.2.3.13. Protocol 6 and 7

100 µL of Buffer 6 (Protocol 6) or Buffer 7 (Protocol 7) was added and the sample was sonicated for 30 min, followed by incubation for 20 min at 37 °C. 10 µL of 200 mM DTT was added and incubation was carried out for 1h at 37 °C. Next, 16 µL of 100 mM IAA was added and the sample was incubated for 15 min at 37 °C in the dark. 7 µL of 0.01 µg/µL trypsin was added and the digestion took place for 17h at 37 °C. Subsequently, the sample was centrifuged for 45 min at 13000 rpm and 48 µL of 5% TFA was added to the supernatant.

2.2.4. Sample purification

Subsequently, SPE or lyophilization proceeded with the samples, as described in Supplementary table 1. Samples were subjected to SPE using Empore 1 mL columns, following the protocol in Table 2. Eluent was collected and the samples were evaporated with SpeedVac (miVac DNA concentrator GeneVac, SP Scientific).

Table 2. Procedure used for SPE.

Step Solution Volume

Conditioning 50% ACN/H2O 0.1% TFA 2 x 300 µL 2 x 300 µL Sample loading - - Washing 0.1% TFA 2 x 300 µL

Sample elution 0.1% HCOOH in 60% ACN 2 x 150 µL

The samples that were not subjected to SPE were lyophilized overnight using a freeze dryer (ThermoHetoPowerDry LL1500 Freeze Dryer, Thermo Fisher Scientific).

Dry residues were stored at -20 ⁰C until the analysis. Prior to LC-MS analysis, the samples were reconstituted in 25% ACN in 1% FA or 3% ACN in 1% FA (Supplementary table 1) to a final concentration of 0.2 mm2_/µL.

2.3. Instrumental analysis

LC-MS/MS analysis was performed with Eksigent ekspert nanoLC 425 system coupled to a TripleTOF 5600+ System. Three trap columns were used for the concentrating and desalting of the samples: an AB Sciex 0.5 mm long, 350 µm inner diameter trap column (Chrom XP C18 resin, 120 Å pore size, 3 µm), an in-house packed 1.5 cm long, 75 µm inner diameter trap column (Magic C18 resin, 100 Å pore size, 5 µm), and an in-house packed 5.0 cm long, 200 µm inner diameter trap column (Magic C18 resin, 100

(28)

28 Å pore size, 5 µm). An in-house packed 10 cm long, 75 µm inner diameter column (Magic C18 resin, 100 Å pore size, 5 µm) was used for peptide separation. Peptides were eluted using a 60 or 95 min long gradient composed of solvent A (0.1% FA in H2O) and solvent B (0.1% FA in ACN). The 60 min

gradient consisted of 5% to 40% B over 45 min, followed by 40% B to 100% B over 5 min and was constant at 100% B for 9 min. Then, the gradient was changed from 100% B to 5% B in 1 min. The 95 min gradient consisted of 5% to 30% B over 65 min and was then changed from 30% to 80% B in 1 min. It was constant at 80% B for 8 min, and subsequently changed from 80% to 5% B in 1 min and was constant at 5% B for 20 min.

Peptides were detected in positive ionization mode. The MS and MS/MS parameters for the detection of the peptides were optimized to yield the highest number of protein identifications, resulting in the parameters shown in Table 3.

Table 3. MS optimized parameters used for the detection of the peptides. MS Start mass and end mass (Da) 400-1250

Accumulation time (ms) 500

From charge state to charge state 2 to 4 With intensity greater than (cds) 100

Mass tolerance (mDa) 40

Switch after 30 spectra

Exclude for (s) 30

Exclude after occurrences 1

MS/MS Start mass and end mass (Da) 200-1800

Accumulation time (ms) 100

From charge state to charge state 2 to 4 With intensity greater than (cds) 100

Mass tolerance (mDa) 50

Switch after 30 spectra

Exclude for (s) 30

(29)

29

2.4. Software used

2.4.1. Determination of the tissue areas

Areas of the tissue samples were determined using ImageJ software v 1.50i (National Institutes of Health, USA).

2.4.2. Database search

Proteins were identified with Protein Pilot Software v 5.0 (AB Sciex, Singapore) with the parameters given in Table 4.

Table 4. Parameters used for the database search with Protein Pilot.

Sample type Identification

Cysteine alkylation Iodoacetamide

Digestion Trypsin / trypsin + Lys-C

Species Homo sapiens

Database Uniprot human thorough

Search effort Thorough ID

Detected protein threshold 0.05 (10%)

Proteins were quantified with MaxQuant v 1.5.4.1 (Max-Planck Institute for Biochemistry, Germany) and the results were viewed with Perseus (Max-Planck Institute for Biochemistry, Germany) v 1.5.4.1. The parameters used for data processing with MaxQuant that were changed from their initial value are shown in Table 5.

Table 5. Parameters used for the database search with MaxQuant that were changed from their initial

value.

Group-specific parameters

Digestion mode Enzyme Trypsin

Modifications Variable modifications Carbamidomethyl (C)

Label-free quantification Label-free quantification LFW

Instrument Instrument type AB Sciex Q-TOF

Max. charge 4

Intensity threshold 100

Global parameters

Sequences Fixed modifications Carbamidomethyl (C)

Max. peptide mass 5500

Label free quantification Separate LFQ in parameter groups Checked

(30)

30

3. Results and discussion

Sample preparation in bottom-up proteomics remains an obstacle. This is caused by the low reproducibility and sensitivity of the current methods together with the limited availability of the tissue samples. While many methods exist for the analysis of large amounts of proteins, the analysis of small amounts of tissue sample still represents a challenge. Here several sample preparation protocols for the analysis of small FF uterine tissue amounts have been examined and evaluated on the outcome on peptide and protein level. Several different protocols have been tested and due to time limitations, many of them have only been carried out once. For the total method validation, the protocol that showed the best performances has to be repeated and examined on analytical parameters as reproducibility, repeatability, and limit of detection (LOD). An overview of the number of identified protein groups, identified peptides, and acquired spectra for all performed experiments is given in Supplementary table 1.

3.1. Optimization of the MS and MS/MS parameters

The MS and MS/MS parameters for the detection of peptides need to be optimized to maximize the number of peptides selected for MS detection. Instrumental parameters as the accumulation time, scanned mass range, charge state, intensity threshold, number of top ions, and exclusion time were chosen for optimization. Hence, the MS and MS/MS parameters were first optimized to obtain the highest number of protein groups (Figure 10).

1 2 3 4

MS

Accumulation time (ms) 50 250 250 500

Mass range (Da) 400-1250 400-1250 400-1250 400-1250 Charge range 2 to 4 2 to 4 2 to 4 2 to 4 Intensity threshold (cds) 100 100 100 100

Top ions 30 30 30 30

Exclusion time (s) 30s/1occ 30s/1occ 30s/1occ 30s/1occ

MS/MS

Accumulation time (ms) 100 100 250 100 Mass range (Da) 200-1800 200-1800 200-1800 200-1800

Figure 10. The optimization of the MS and MS/MS parameters for the analysis of peptides using

samples prepared by Protocol 5.1.1 (n=1).

0 2000 4000 6000 8000 10000 12000 14000 16000 18000 0 50 100 150 200 250 300 350 400 450 Number of spectra and identified peptides Number of identified protein groups

(31)

31 The initial MS analysis parameters were as follows: an accumulation of 50 ms, a mass range of 400-1250 Da, a charge range of 2+ to 4+, an intensity threshold of 100 cds, an exclusion time of 30 s per 1 occurrence, and a collection of the 30 top ions. The MS/MS parameters consisted of an accumulation time of 100 ms and a mass range of 200-1800 Da (1). After changing the accumulation time of the MS analysis from 50 to 250 ms (2), the number of protein groups was almost doubled. This may be explained by the fact that a higher amount of ions is accumulated when the accumulation time is broadened, which improves the signal intensities and therefore the number of spectra and peptides. Based on this, the accumulation time of the MS/MS analysis was changed from 100 to 250 ms (3). However, this did not result in an increase of the number of identified protein groups which may be the effect of overloading. Increasing the accumulation time of the MS analysis to 500 ms increased the number of identified protein groups and was therefore further used in the experiments (4). As a result, the parameters that were used for the remaining experiments are those of 4 and they are further detailed in Table 3.

3.2. Evaluation of chromatographic conditions

To evaluate the influence of the chromatographic conditions on the outcome of the analysis, analytical columns with 3 and 5 µm particle sizes, different chromatographic run times, various amounts of analyzed tissue, and several trap columns were tested.

3.2.1. Influence of packing material of the analytical columns on protein identifications

Columns with a lower particle size have a higher efficiency compared to columns with larger particles, which could potentially lead to an increase in the number of protein identifications. For example, two columns packed with 3 and 1.9 µm material were compared on the digestion of bovine protein mixture and the 1.9 µm column had a higher peak capacity and resolution.81_{Therefore, another column with a}

particle size of 3 µm was tested besides the column packed with 5 µm. The number of identified protein groups, identified peptides, and acquired spectra of the 5 and 3 µm column were 296, 2724, 14500 and 227, 1921, 10268, respectively. Thus, the number of identified protein groups, spectra, and peptides was not improved using smaller particles of 3 µm. Therefore, all further analysis was carried out using columns with 5 µm particles.

3.2.2. Influence of the chromatographic run time on protein identifications

The chromatographic run time was changed from 60 to 95 min in order to determine its effect, since a longer run time might improve the separation of the peptides and an elution of the peptides which were originally retained on the column. As can been seen from Figure 11, the number of spectra increased by 66%, which is in agreement with the study of Hsieh et al., who found that increasing the run time from 60 min to 90 min doubled the number of spectra.82_{However, an increase in the number of identified}

peptides and protein groups was not observed, suggesting that extra information on unique peptides and proteins was not retrieved with the additional spectra. These results are consisted from those of Richards

(32)

32 and co-workers who reported that the number of uniquely identified peptides decreases with an increasing chromatographic run time.83

Figure 11. Number of identified protein groups, spectra, and identified peptides obtained using a 60

min gradient and a 95 min gradient time using samples prepared by Protocol 4.1. The injected amount of tissue was 1 mm2_{, the chromatographic conditions were as described in Section 2.3, and the 75 µm}

trap column was used (n=1).

3.2.3. Influence of trap columns on protein identifications

Three trap columns were compared on their loading capacity using Protocol 4.6: an AB Sciex trap column, a 200 µm trap column, and a 75 µm trap column. Figure 12 shows the number of identified protein groups, spectra, and identified peptides of the 3 trap columns. The AB Sciex and 75 µm trap column showed similar results, but the 200 µm trap column had approximately 13% protein group identifications less compared to the other two trap columns. During the analysis, the pressure with the 200 µm trap column was lower than when other trap columns were used, which might have affected the results.

Figure 12. Number of identified protein groups, spectra, and identified peptides obtained using an AB

Sciex trap column, a 200 µm trap column, and a 75 µm trap column using samples prepared by Protocol 4.6. The injected amount was 2 mm2_{, the chromatographic run time was 95 min, and the}

chromatographic conditions were as described in Section 2.3 (n=1).

0 5000 10000 15000 20000 25000 0 50 100 150 200 250 300 350 60 min 95 min Number of spectra and identified peptides Number of identified protein groups

Protein groups Spectra Peptides

0 5000 10000 15000 20000 25000 30000 0 50 100 150 200 250 300 350 400 AB Sciex 200 µm 75 µm Number of spectra and identified peptides Number of identified protein groups

(33)

33

3.2.4. Influence of the injected amount on protein identifications

In addition, samples with an area of 1 and 2 mm2_{were injected in order to determine the effect of the}

injected amount on the protein identifications. It is hypothesized that a higher amount of sample would result in more protein identifications, since their concentration is higher, resulting in more peptides which might be analyzed. This was proven to be correct as shown in Figure 13, since the number of protein group identifications increased. The increase of protein identifications when injected a higher amount was in agreement with previous research, but the number of spectra decreased, which is in contrast with similar studies.84,85_{However, previous research injected various amounts of FFPE tissue}

in range of 0.1 and 1.0 µg and it was shown that the difference in spectra between 0.5 and 1.0 µg was smaller compared to 0.1 and 0.25 µg, which might indicate that a limit of spectra is reached around 1.0 µg.84

Figure 13. Number of identified protein groups, spectra, and identified peptides obtained with an

injected amount of 1 and 2 mm2 _{using samples prepared by Protocol 4.6. The chromatographic run time}

was 95 min, the chromatographic conditions were as described in Section 2.3, and the AB Sciex trap column was used (n=1).

3.3. Comparison and optimizing extraction buffers

3.3.1. Evaluation of different extraction buffers

Next, different extraction buffers containing various concentrations of SDC, SDS, ACN, and urea were tested: 0.25% SDC (Protocol 1), 5% SDC (Protocol 2), 4% SDS (Protocol 3.1), 80% ACN (Protocol 6), 8 M urea with 30% ACN (Protocol 5.1.1) followed by sample purification with lyophilization and SPE. However, precipitation occurred in the samples containing SDC after sample purification with SPE indicating that the removal of SDC was not successful. As a result, the samples containing SDC were not analyzed.

The number of identified protein groups, spectra, and peptides using the described extraction buffers containing 4% SDS, 80% ACN, and 8 M urea with 30% ACN are shown in Figure 14.

0 5000 10000 15000 20000 25000 30000 0 50 100 150 200 250 300 350 400 1 mm^2 2 mm^2 Number of spectra and identified peptides Number of identified protein groups

Optimization of Analytical Procedure for Protein Analysis in Fresh Frozen Laser Microdissectioned Human Tissue

Bachelor Thesis Scheikunde