Development of an IMAC method for the enrichment of phosphopeptides from human kidney tissue

(1)

Research project

MSc Chemistry - Analytical Sciences

Development of an IMAC method for the enrichment of

phosphopeptides from human kidney tissue

Marieke Wingelaar

Date

14-04-2020

Supervisor

Prof. dr. Garry L. Corthals

Daily supervisor

Prof. dr. Garry L. Corthals

P.J. (Petra) Jansen, MSc

Research institute

(2)

I

Abstract

Protein phosphorylation is one of the most ubiquitous post translational protein modifications. The phosphorylation state of a protein determines its activity, reactivity and its ability to bind other molecules. Protein phosphorylation plays a role in regulating many essential cellular functions and is involved in almost all cellular signaling pathways. It has also been found to play a role in various diseases. In the context of an ongoing collaborative research on kidney disease and diagnostics, a method is required that can provide a view on the phosphoproteome of cells in kidney tissues. Ultimately, we are interested to know if and how phosphorylation is involved in disease development, kidney dysfunction and kidney rejection following transplantation.

However, to study phosphoproteins from a complex biological sample, an enrichment procedure is required to isolate the phosphopeptides. One of the methods that can be used is immobilized metal ion chromatography (IMAC), which binds phosphopeptides based on the interaction of the negatively charged phosphopeptides and a positively charged transition metal ion.

The aim of this project was to develop an IMAC phosphopeptide enrichment method for human kidney tissue samples. Two different strategies for fabricating IMAC columns have been explored. The first, GELoader tip columns, was found to be impractical due to the properties of the IMAC resin and, moreover, resulted in very few phosphopeptide identifications. Therefore, another approach was pursued, in which IMAC columns are constructed from fused silica capillaries. The method proposed in this thesis includes thoroughly tested procedures for frit making and column packing. The method also includes a protocol for activation of the IMAC resin with Fe3+_{-ions and for the actual} phosphopeptide enrichment experiment. The method has been developed until it was practically feasible, but its performance at phosphopeptide enrichment is still uncertain. There are still many parameters in the method that can be optimized.

The basis of an IMAC method for enrichment of kidney tissue sample has been established in this project, but the method still needs to be tested be fine-tuned. If the method is successful in the future, it may shed a light on the involvement of protein phosphorylation in kidney disease.

(3)

II

List of abbreviations

Abbreviation Definition

ABC Ammonium Bicarbonate

ACN Acetonitrile

CID Collision-Induced Dissociation

Da Dalton

DDA Data-Dependent Acquisition

DHB 2,3-Dihydroxybenzoic acid

DTT Dithiothreitol

EDTA Ethylenediaminetetraacetic acid

ESI Electrospray Ionization

FA Formic Acid

FDR False Discovery Rate

FF Fresh Frozen

FFPE Formalin Fixated Paraffin Embedded

H&E Haemotoxylin and Eosin

IAA Iodoacetamide

ID Inner Diameters

IMAC Immobilized Metal Ion Affinity Chromatography

LC Liquid Chromatography

m/z Mass to Charge Ratio

MALDI matrix assisted laser desorption/ionization

MOAC Metal Oxide Affinity Chromatography

MS Mass Spectrometry

MS/MS Tandem Mass Spectrometry

PolyMAC Polymer-Based Metal Ion Affinity Capture

PTM Post Translational Modification

SDS sodium dodecyl sulfate

TFA Trifluoroacetic Acid

(6)

1

1 Introduction

1.1 Proteomics: The study of proteins

Proteomics refers to the large-scale study of proteins, including their structure, biological function, expression or spatial distribution. The proteome is the entire set of proteins that is produced by or present in an organism or tissue. Proteomics plays an important role in understanding biological processes, protein functions and interactions, which in turn enables the discovery of biomarkers, the identification of new drug targets and improvement of disease diagnosis and treatment.

It is estimated that the human genome contains approximately 20,000 protein-coding genes (Figure 1).1_{These can be transcribed into 200,000 different RNA-transcripts, from which more than 1,000,000} different proteins can be formed. Besides, the proteome is different for each cell type and tissue and is very dynamic, increasing the complexity even further. It changes over time in response to internal and external conditions that induce alterations in protein expression, post-translational modifications (PTMs) and protein trafficking. Analysis of the proteome provides a unique molecular fingerprint of cells or tissues at the moment the sample was taken.2

Mass spectrometry (MS) is typically the method of choice for identifying proteins in a biological sample and has become an indispensable tool in proteomics.3_{MS is favoured for its high sensitivity, high} throughput capacity and ability to identify PTMs.4_{Over the years, various MS-based methods have} been developed for proteome analysis. Due to the proteome’s complexity, separation of the sample is required before analysis with MS. Both gel chromatography and liquid chromatography (LC) based methods have been successfully deployed. LC is more commonly used, since it has the advantage that it can be used in-line with MS. The approaches that are possible to study proteomics can be divided into three categories: ‘top-down’, ‘bottom-up’ and an intermediate, ‘middle-down’ proteomics (Figure 2). In the top-down approach proteins are characterized by MS analysis of the intact proteins. In contrast, in both middle-down and bottom-up proteomics (also known as “shotgun proteomics”),

Figure 1. The human proteome contains more than 1,000,000 proteins and is complex and very dynamic,

(7)

2

proteins are first digested by proteolytic enzymes into smaller fragments, called peptides. In bottom-up proteomics, these peptides are small (~500 – 3000 Da) while in middle-down proteomics the fragments are larger (2000 – 20000 Da). Protein identifications are made by comparing measured mass spectra of proteins or peptides to those stored in a database. In the case of bottom-up proteomics, MS/MS is used to fragment peptides even further, and information from multiple peptide identifications is assembled to obtain a protein identification.

The three approaches provide different information and selection of the approach should be based on the research question to be answered. Each method has its strengths and weaknesses. Advantages of analyzing intact proteins (top-down) are that samples are relatively less complex, so that proteins can be more easily identified, and that information about the intact proteins can be obtained. However, it has two important drawbacks: intact proteins have poor ionization efficiency and are difficult to separate.5_{Because of their bigger size compared to peptides, they are more difficult to} ionize and have low diffusion rates. Currently, bottom-up proteomics is the method that is most commonly used, although this method suffers from some difficulties as well.6_{A major disadvantage is} limited protein sequence coverage caused by incomplete digestion, resulting in missing peptides. Besides, there may be ambiguity of the parent protein for redundant peptide sequences.

1.2 Phosphoproteomics

As mentioned above, the proteome also encompasses PTMs, which are chemical alterations a native protein can undergo after biosynthesis. Such modifications may comprise of adding or removing specific functional groups to the protein, or cleaving bonds to remove certain sequences. More than 200 types of PTMs have been identified.7_{One very common PTM is phosphorylation, which is an} esterification reaction whereby a phosphate group (PO43−) is transferred from an energy-rich ATP molecule and covalently attached the hydroxyl side chain of an amino acid residue (Figure 3).8_The Figure 2. Overview of the different MS based strategies for protein identification. In top-down proteomics proteins are

characterized by MS analysis of the intact proteins. In middle-down and bottom-up proteomics proteins are first digested by proteolytic enzymes into smaller fragments, called peptides. Protein identifications are made by comparing measured masses of proteins or peptides to masses from a stored database. Reprinted from Switzar et al. (2013).

(8)

3

branch of proteomics that focuses on identifying and characterizing phosphorylated proteins is called

phosphoproteomics. In eukaryotes, protein phosphorylation primarily occurs at serine, threonine, or

tyrosine residues, composing approximately 90%, 10%, and <1% of the total phosphoproteome, respectively (Figure 4 and 5).9,10_{Histidine phosphorylation has also been reported, but} phosphohistidine is very unstable and therefore rarely identified in phosphoproteomics studies.11 Protein phosphorylation is an ubiquitous PTM; it is estimated that 75-90% of the human proteome can be phosphorylated.12,13_{The number of phosphorylation sites in one protein can vary from 1 to} more than 100.14

Figure 3. Schematic representation of the phosphorylation reaction. In a

phosphorylation reaction a phosphate group is covalently attached to the hydroxyl group of a protein, at the cost of one molecule of ATP. Reprinted from

Chemistry LibreTexts, Soderberg (2019).

Phosphoserine

90% Phosphothreonine 10% Phosphotyrosine <1% Figure 5. Chemical structures of the most common phosphorylated amino acidfs: phosphoserine,

phosphothreonine and phosphotyrosine.

Figure 4. Chemical structures of the amino acids that are most commonly phosphorylated in eukaryotes:

serine, threonine and tyrosine.

(9)

4

Protein phosphorylation is catalyzed by a class of enzymes called kinases. The reverse reaction, protein dephosphorylation, in which a phosphate group is removed from the protein, is catalyzed by phosphatases (Figure 6).15_{The phosphorylation state of a protein determines its activity, reactivity} and its ability to bind other molecules. Due to its reversable nature, (de)phosphorylation can tightly control protein activity in response to cellular or external stimuli.16_{Over the years it has become} apparent that phosphorylation plays a crucial role regulating cell cycle progression, differentiation and development, metabolism, nerve activity, muscle constriction, transcription, translation, endocytosis, phagocytosis, and apoptosis.17,18_{Phosphoproteomics helps understand these signaling pathways and} provides insight into proteins that are important for regulating them.

From the three proteomic approaches, the bottom-up approach is the only method that is widely applied to studying protein phosphorylation, because it is the most suitable for complex samples. This is also the approach that will be used during this project. In bottom-up proteomics, proteins are first digested into smaller peptides, which are subsequently separated and analyzed using LC-MS/MS. In MS/MS (also called tandem MS or MS2_{) peptides are first ionized and these ions are}_{a separated by}

their m/z ratios. In the second stage, high-intensity ions of specific m/z values are isolated and undergo collision with an inert gas, causing peptide ion fragments to form. These fragments are then separated and detected. By comparing the tandem mass spectra with those from a peptide spectral library, peptides can be identified. Protein identification is accomplished by assigning multiple peptide sequences to proteins. The large number of MS/MSspectra generated requires automated search engines for the interpretation of peptide fragmentation data. Computer algorithms that are commonly used for this task are SEQUEST®1, Mascot™ and Byonic™.

Figure 4. Schematic representation of phosphorylation and dephosphorylation reactions with serine and tyrosine

as substrate. Kinases catalyze the phosphorylation reaction, in which a phosphate group is transferred from ATP molecule and covalently attached the hydroxyl side chain of an amino acid residue. The reverse reaction, the removal of the phosphate group from a protein under release of a release of hydrogen phosphate ion (HPO42-),

is catalyzed by kinases. Redesigned from Chromosite, Phosphorylation and Phospho Proteins.

P

_ATP

P

_ADP

P

P i

H

2

O

Kinase

Phosphatase

Phosphorylation

Dephosphorylation

Protein Phosphoprotein

(10)

5

1.3 Phosphoproteomics in kidney disease diagnostics

Since protein phosphorylation is involved in so many important cellular signaling pathways, it is no surprise that aberrant phosphorylation may lead to disease. For example, phosphorylation has been shown to play a role in various types of cancer, including breast cancer19_{, ovarian cancer}20_{, leukemia}21 and colon cancer22_{, chronic inflammatory diseases}23_{, diabetes}24_{, and neuropathologies such} Alzheimer’s disease25_{and Parkinson's disease.}26_{Given its involvement in such a diverse array of} diseases, it can be presumed that phosphorylation may play a role in kidney disease as well. So far, little is known about protein phosphorylation related to kidney disease and learning more about this subject may improve diagnostics and treatment of kidney disease.

Currently, the routine diagnostic process for kidney disease is based on analysis of urine and blood samples, and imaging tests like ultrasound, computed tomography (CT) and magnetitic resonance imaging (MRI)27_{. Sometimes removal of a small sample of kidney tissue for testing, called a biopsy, is} necessary to provide a diagnosis.28_{A biopsy can quickly diagnose common kidney diseases and help} determine causes of kidney damage. Other reasons to perform a kidney biopsy are to assess the condition of the kidneys and monitor the health of a transplanted kidney.

Figure 5. Overview of the different routine stains in nephropathology. A) HE stain for an overview of glomerular changes. B) PAS stain for more detailed analysis of the glomerular structure. C) Protein stain (SFOG), marker for subendothelial and intramembranous protein deposits. D) Silver stain. E) Immunofluorescence with antibodies against IgG. F) Immunohistochemistry using the antiperoxidase and an antibody against IgG. Reprinted from Amann et al. (2006).

(11)

6

To obtain clinical information from a renal biopsy, the biopsy is fixated and cut into very thin sections of 2-3 µm thick, which can be used for histochemical or immunohistochemical staining and studied with light or electron microscopy (Figure 7).28_{A standard staining is the haematoxylin and eosin (H&E)} staining, which shows the general structure of the tissue. In addition, other staining techniques are used to more selectively stain tissue structures, cells, cellular components, and specific substances like proteins or ions. These methods can be used to answer specific questions about the kidney samples.

This specificity is also a major limitation of these diagnostic methods. Each method can identify only a single target of interest. They will not identify an abnormality if you were not looking for it specifically and they cannot provide an overview of the total protein expression and cell signaling pathways that are activated in a disease. Consequently, diseases are sometimes overlooked or wrongly diagnosed, or can only be diagnosed if the disease has progressed to a more advanced stage.29_{As an alternative,} proteomics could be used as a non-targeted approach to study the kidney proteome.30_{Changes in} protein levels or PTMs occur before the patient shows symptoms and while other biomarkers are still in normal range. Early identification of these changes could allow for a treatment to interrupt disease progression and prevent kidney damage or failure (Figure 8).31

However, currently little information is available about protein phosphorylation in kidneys and its role in kidney diseases. Given its relevance in other diseases, it is worthwhile studying the phosphoproteome of kidney tissues cells from diseased and healthy individuals. Ultimately, it is of interest to know if and how phosphorylation is involved in disease development, kidney dysfunction and rejection following kidney transplantation. This knowledge could then be used to identify Figure 6. Early identification of patients at risk of kidney disease an appropriate treatment can interrupt disease progression

(12)

7

biomarkers and develop diagnostic tools to diagnose these diseases, identify therapeutic targets and enable early treatment and prevent negative outcomes.

1.4 Challenges in studying phosphoproteomics

Despite its importance as a post-translational modification, studying protein phosphorylation is still challenging. These challenges arise from the biochemical properties of phosphoproteins as well as the analytical methods that are available to study them.

Biochemical challenges

Protein phosphorylation is a dynamic process, tightly controlled by kinases and phosphatases, adding and removing phosphate groups to proteins in response to all kinds of internal and external cellular stimuli. Even small alterations to the cell’s environment occurring during sample collection and storage may induce great changes in cellular signaling networks, altering the phosphoproteome.16 Besides, special care must be taken during sample preparation. Cells lyses will activate many proteases and phosphatases that can degrade proteins and phosphoproteins. The phosphate group is very labile and is easily lost. Controlling temperature and pH during sample preparation and using protease- and phosphate inhibitors may help prevent this.32

Analytical challenges

As discussed earlier, MS-based analysis is the most commonly used method to study phosphopeptides, allowing the identification of phosphopeptides and phosphorylation sites, and bottom-up proteomics is practically the only applicable method for the characterization of phosphopeptides. However, this is not straightforward, since the intrinsic properties of protein phosphorylation complicate identification with MS for several reasons. The primary challenge is the low stoichiometry of phosphorylation and therefore the low abundance of phosphorylated proteins in cells and tissues.33_{Proteins can also be phosphorylated on different sites, resulting in even lower} abundancy of each phosphorylated species. This makes detection and identifications of phosphoproteins by MS difficult.

Secondly, phosphopeptides carry a net negative charge due to their negatively charged phosphate groups. This causes severe suppression of the phosphopeptide signal when positive ion mode and soft ionization methods such as ESI and MALDI are used.34_{However, positive ion mode and soft ionization} techniques are required for identification of amino acids and determining their sequence, but ionization efficiency of phosphopeptides in this mode is low. Negative ion mode cannot be used, since the resulting MS/MS spectra are too complex to interpret 35_{Furthermore, ionization efficiency is also} suppressed by the presence of abundant unphosphorylated peptides.36

Another challenge is that fragmentation of phosphopeptides in MS/MS is more difficult than of non-phosphorylated peptides, because of the lability of the phosphate group.37_{In collision-induced} dissociation (CID), the neutral loss of the phosphate group via β-elimination is the predominant fragmentation pathway. As a result, the amount of peptide backbone fragments that are formed is limited, while they are required for subsequent determination of the amino acid sequence and peptide identification.

(13)

8

1.5 Phosphopeptide enrichment strategies

Although there are considerable challenges in detecting and identifying phosphopeptides with MS/MS, this is not an impossible task. A sensible approach to address these problems is to try to reduce sample matrix complexity and to increase the relative concentration of the phosphopeptides in the sample. This is achieved by carrying out a phosphopeptide enrichment procedure before LC-MS/MS analysis, in which as many as possible non-phosphorylated peptides are removed and phosphopeptides are concentrated.

So far, many approaches have been developed and successfully employed to enrich phosphopeptides or phosphoproteins from complex biological samples prior to MS-based analysis. Examples include: strong anion and/or cation exchange chromatography,38–42_{chemical derivatization of} phospho-residues,43–45_{immuno-precipitation with specific antibodies,}46,47_{and affinity chromatographic} methods such as: polymer-based metal ion affinity capture (PolyMAC),48–50_{immobilized metal ion} affinity chromatography (IMAC),37,51–54_{and metal oxide affinity chromatography (MOAC).}55–57_Figure 916_{shows a selection of the diverse enrichment strategies that have been developed.}

Figure 7. Overview of methods that are available for the enrichment of phosphopeptides prior to MS-based analysis. The methods

can be grouped into three categories: I) methods based on immunoprecipitation, II) methods using chemical derivatization and II) affinity chromatography-based methods. Reprinted from Thingholm et al. (2009).

(14)

9

Currently, IMAC and MOAC (especially TiO2) are the phosphopeptide enrichment techniques that are most commonly used. These methods are comparable as they both exploit the affinity of phosphate groups for metal ions (Figure 10).58_{In IMAC metal ions are immobilized on a carrier resin by a chelating} agent, while MOAC uses the affinity of metals in metal oxide particles. Although both methods rely on the same mechanism, there are differences in their performance. IMAC appears to have a stronger selectivity for multiply phosphorylated peptides51_{, whereas TiO}₂ _{is more efficient for} monophosphorylated peptides.54_{Although IMAC is generally assumed to suffer more from insufficient} specificity, the consensus in the field is that no method is superior to the other and that both methods isolate a different subset of the phosphoproteome.51,59,60_{If a more complete coverage of the} phosphoproteome is desired, it is recommended to use a combination of both methods. This project will deploy IMAC as enrichment method for phosphopeptides, focusing on developing a method for the analysis of kidney tissue samples.

1.6 Immobilized Metal Ion Affinity Chromatography (IMAC)

IMAC is a form of affinity chromatography, in which phosphopeptides are retained by the strong affinity between positively charged metal ions and the negatively charged phosphate groups of the phosphopeptides (Figure 11). Metal ions that have shown affinity for oxygen containing functional groups and can therefore be used in IMAC include Cu2+_{, Ni}2+_{, Zn}2+_{, Co}2+_{, Fe}3+_{, Al}3+_{and Ca}2+_{. The metal} ions are immobilized by chelation to a support matrix, such as chromatographic resins or magnetic beads. The chelating ligand is covalently attached to the matrix. Unoccupied coordination sites are left on the metal ion, which can interact with negatively charged peptides. Bound phosphopeptides can be eluted from the IMAC material using various approaches. Alkaline buffers, EDTA solutions,61_highly acidic solutions54_{or solutions containing phosphate}62_{or phosphoric acid}63_{have all been reported as} eluants for phosphopeptides.

Figure 8. Overview of the mechanism of IMAC and MOAC, the two most commonly used enrichment strategies for

phosphopeptides. Both methods exploit the affinity of phosphate groups for metal ions. In IMAC the metal ions are immobilized on a carrier resin, while MOAC exploits the affinity of metals in metal oxide particles. Reprinted from Arrington et al. (2017).

(15)

10 Chelating agents

The two most commonly used metal chelating agents for IMAC are nitrilotriacetic acid (NTA) and iminodiacetic acid (IDA). NTA resin coordinates with the metal ion with four valencies, whereas IDA only has three (Figure 12). In most publications about IMAC IDA resin was used, although NTA resin is generally considered superior.64,65_{During this project IDA resin will be used, attached to POROS™} 20 MC beads. This was the type of resin that was available in our lab and it is believed that any of the resins would perform well enough for this application.

Figure 9. Schematic representation of the binding mechanism of IMAC. A metal ion is immobilized on

the IMAC resin by a chelating agent. Phosphopeptides are bound to the metal ion by interaction of the negatively charged phosphate groups and the positively charged metal ion.

P

O

Chelating ligand Peptide Phosphate group Metal

ion

M

+ IMAC resin Peptide IMAC resin IMAC resin Peptide

Figure 10. Schematic representation of IDA (A) and NTA (B) metal chelating agents. IDA resin

coordinates with three valencies to the metal ion, NTA resin with four.

A. IDA resin

(16)

11 Non-specific binding

One of the major limitations of IMAC is its poor selectivity. The positively charged metal ions that are used in IMAC do not only retain negatively charged phosphopeptides, but also non-phosphorylated peptides containing acidic residues, such as glutamic acid, aspartic acid, histidine and cysteine residues. In order to improve specificity, highly selective and stringent column washing, sample loading and elution conditions are required. Especially the pH and the composition of the sample loading buffer are important. The pH determines the protonation state of acidic functional groups and consequently the proportion of negative charges on phosphate groups, which are responsible for binding to the metal ion on the IMAC resin (Figure 13).66_{The pH also determines the extent to which} acidic residues are protonated, which have a higher pKa value than the phosphate group on phosphopeptides. For optimal phosphopeptide binding, the pH must be high enough to ensure that the majority of phosphopeptides are deprotonated, but not so high that acidic amino acids become protonated too. The ideal pH for the sample loading buffer has been shown to be around pH 2-3.67 Besides, it has been shown that the addition of 50% ACN to the sample loading buffer reduces non-specific binding.62

IMAC is a method with many variables that can be adapted to the specific requirements of the application. Choice of metal ion, chelating agent, loading conditions and elution buffers are among the factors that can have great impact on the efficiency and specificity of the enrichment. However, optimizing these is not always straightforward. A comprehensive review of all factors affecting IMAC performance that should be taken into account when developing an IMAC method is provided in Appendix A, as well as additional strategies that can be used to reduce non-specific binding.

1.7 Human kidney tissue samples

During this project, human tissue samples will be used for the development of a phosphopeptide enrichment method. The tissue specimens were collected from the cortex/medulla region of the kidney (Figure 14).68_{After tissue samples have been collected, they must be fixated to persevere the} material as close to its natural state as possible. To achieve this, it is necessary to prevent autolysis and putrefaction, which are degenerative processes that occur as soon as a tissue is deprived of its blood supply.69_{Fixation also improves the physical properties of the tissue, allowing to cut thin slices} from it that can be histologically stained and studied. Tissues must be fixated quickly after collection, or a tissue specimen can be irreversibly damaged and information that can be obtained from it can be lost.

Figure 11. Chemical structures of the phosphate functional group at different pH. Shown are the

corresponding experimentally determined pKa, values (pKa,1 = 1.5, pKa,2 = 6.5). At low pH the molecule is

neutral, when the pH is increased the phosphate group has a net negative charge, and when the pH is increased above 6.5 the phosphate group carries a net double negative charge. The phosphate groups on peptides are more acidic than phosphoric acid (pKa,1 = 2.2, pKa,2 = 7.2).

pKa,1 = 1.5 pKa,2 = 6.5

+ H+ _{+ H}+

- H+ - H+

(17)

12

Figure 12. Illustration of the kidney atomy. The samples used during this study were obtained from cortex and

medulla region of the kidney (marked with blue box). Reprinted from Urology care foundation website.

The standard tissue preservation procedure for pathological examination is formalin-fixed paraffin-embedded (FFPE) tissue. Immediately after the tissue sample is taken, it is emerged in a formalin solution, leading to the formation of stable and irreversible methylene bridge cross-links to stabilize the tissue. The tissue is then embedded into a paraffin wax block for protection. FFPE samples can be stored at room temperature, which is convenient and cost-effective, and they can be preserved for a very long time without quality reduction.70_{FFPE tissue is very suitable for histological and} immunohistochemical staining and morphology analysis. For molecular analysis however, such as proteomics studies, FFPE is less suited. The main reason is that while proteins are preserved, they are also denatured during formalin fixation. This limits protein extraction efficiency and may hamper protein identification in MS analysis.71_{Besides, formalin does not inactivate phosphatases, so FFPE} samples contain less phosphorylated proteins than the original tissue.72

An alternative method of tissue preservation is fast frozen (FF) tissue. FF samples are prepared by a flash freezing the tissue in liquid nitrogen and immediately storing them in a –80 °C freezer. This method is very fast and considerably less labor intensive than the FFPE procedure. In FF samples proteins do not undergo any modification and are preserved in a native, undegraded state, making these samples excellent for biochemical analysis. Therefore, FF tissues have become the standard for MS based proteomics studies. However, FF specimens are not frequently collected, since FFPE is and old and well-established method and it is routine practice for pathologists. Hence, there is an increasing interest in analyzing the proteome of FFPE tissue samples. Furthermore, FFPE samples have been collected and archived for a long time and vast libraries exist containing the samples along with their corresponding clinical data, including histology, diagnosis, treatment history and response, and outcome.73_{For FF samples such data are usually not available, limiting the utility of information that} can be obtained from proteomic analysis of these samples.

Several protocols have so far successfully used FFPE tissues for bottom-up MS based proteomics. In these publications a heat-induced antigen retrieval procedure was performed to remove the formalin induced crosslinks, showing promising results in terms of protein identifications by subsequent MS/MS analysis.74_{In this project both FF and FFPE tissue samples will be used. FF samples are easier} to work with and may be more suitable to develop and enrichment method, whereas FFPE tissues have more clinical relevance.

(18)

13

Figure 13. Overview of the workflow that will be followed during this project. The focus of this project will be on the development

of an IMAC method for phosphopeptide enrichment. Designed using images from Edith Cowan University website,

IconArchive.com and GenoSkin website.

1.8 Aim of this project

The goal of this project is to develop a method for the analysis of phosphorylated proteins in human kidney tissues. More specifically, this project focuses on establishing an IMAC-based enrichment method for phosphopeptides from tissue samples. The method must be applicable to the minute sample sizes obtained from pathology, which range from 4 to 20 µm thickness and 5 mm2_{to 20 mm}2 in width. Although successful IMAC enrichment methods have been described previously, a method for such small tissue samples has yet to be established. Commercially available methods exist but are not suitable for microscopic sample sizes and modifying these is not an option as they are expensive. Therefore, a new method must be developed based on previous methods developed in our group and in the literature, but in a way that it is applicable to small samples. Eventually, the method that will be developed during this project is to be used to study protein phosphorylation related to kidney disease and ultimately to learn if and how phosphorylation is involved in disease development, kidney dysfunction and kidney rejection, but this is beyond the scope of the project..

I will propose an IMAC method for the enrichment of phosphopeptides from human kidney tissue samples. For the development of this method tissue homogenate samples (larger quantities of homogenized tissue) will be used. When the IMAC method is fully optimized it is to be applicable for microscope tissue samples. Figure 15 shows the total workflow from tissue sample till data analysis. All steps from protein extraction till data-analysis will be performed during the project, but the focus is on phosphopeptide enrichment.

Phoshopeptide

enrichment LC–MS/MS data acquisition Database search Protein

extraction Proteolitic digestion Tissue sample IMAC LC MS MS/MS

(19)

14

2 Materials & Methods

2.1 Materials

Chemicals

Table 1. List of chemicals that were used

Instrumentation

Table 2. List of instrumentation that was used.

Name Supplier Extra information

Acetic acid Merck Purity > 99%

Acetonitrile, ACN Biosolve Purity > 99.8%

Ammonium bicarbonate, NH4HCO3 Honeywell Fluka Purity > 99% Ammonium hydroxide, 10% Honeywell Fluka

Disodium hydrogen phosphate, Na2HPO4 Merck Purity 99.95%, #7558-79-4

Ethanol (absolute) Scharlau # ET005

Ethylenediaminetetraacetic acid, EDTA Sigma-Aldrich Purity > 99%

Formic acid, FA Sigma-Aldrich purity 98% - 100%

Iodoacetmide, IAA Sigma-Aldrich purity >99%; #I6125 Iron(III)chloride, FeCl3 Sigma-Aldrich Purity 99.9%, #7705-08-0

L-Dithiothreitol Sigma-Aldrich Purity > 99%

Sodium hydroxide Sigma-Aldrich Pellets, Purity > 98%

Trifluoroacetic acid, TFA Merck purity ~100%

Trypsin Sigma-Aldrich # T260000

Urea Sigma-Aldrich purity > 99%; # 17-1319-01

Water (ULC/MS) Biosolve 232141

Name Manufacturer Instrument

Atrium 611 Sartorius Ultrapure water system

BioTek SynergyMx BioTek Microplate Reader

Centrifuge 5804R Eppendorf Benchtop Centrifuge

Heto PowerDry LL 1500 Thermo Electron Freeze Dryer

Ika Vortex 4 Basic Ika Works Benchtop vortex shaker

Mettler AE 240 Mettler Toledo Balance scale

Thermomixer Compact Eppendorf Thermomixer

TripleTOF 5600+ system AB Sciex Mass Spectrometer

LEGATO® 210 kdScientific Syringe Pump

Eksigent® EkspertTM nanoLC 425

(20)

15 Materials

Table 3. List of materials that were used.

Samples

All samples were provided by Dr. J. Kers. from the Amsterdam UMC (Amsterdam Universitair Medische Centra, The Netherlands). Samples that were used during this project were human kidney tissue homogenate samples, both FF and FFPE, collected from the cortex/medulla region of the kidney. The kidneys were collected from ten patients with unknown medical histories and fixated as FF or FFPE tissue blocks. Tissue blocks were immersed in a 0.1 M ammonium bicarbonate buffer and homogenized using a tissue homogenizer. FFPE samples were deparaffinized before homogenization. The homogenates were aliquoted and stored at – 80 °C until use. The tissue contents of the samples were 5 mg tissue in 17.24 μL solution for FF samples and 5 mg tissue in 33.33 μL solution for FFPE samples.

Name Supplier Information

EmporeTM _{C18-SD 4 mm/1 mL cartridges} _SUPELCO® _{# 77871-U, Lot # 97002}

Frit kit Next Advance Kasil frit kit

GF/CTM (Glass Microfiber Filters) GE Healthcare Life

Science # 1822-021; Diameter 21 mm

Glass Gastight syringe, 1mL Hamilton # 81320

IMAC resin Thermo Fischer

Scientific POROS™ 20 MC Perfusion Chromatography™ Bulk Media for Metal Chelate Affinity Chromatography PEEK MicroTight® tubing sleeves and

connections Upchurch Scientific® PEEK sleeve (.009” ID), PEEK MicroTight union and fitting (#P-771)

PierceTM Quantitative Fluorometric

(21)

16

Figure 14. Overview of workflow from tissue sample till data analysis. Yellow and grey steps are sample

preparation steps (Yellow is only for FFPE samples) Phosphopeptide enrichment is the focus of the project. For all other steps protocols will be used that had already been developed by previous students and were already available in the Corthals lab.

2.2 Methods

Workflow

In Figure 16 an overview is provided of the total workflow from sample preparation to IMAC enrichment, LC-MS/MS and data analysis. The yellow and grey rectangles represent the sample preparation. The steps in yellow are only required for FFPE samples and can be skipped for FF samples. After desalting of the samples, each sample was split in two parts: 90% of the sample was used for phosphopeptide enrichment, while the remaining 10% did not undergo any further treatment and was analyzed directly with LC-MS/MS, serving as a control. The focus of this project is the development of an IMAC method. For all other steps in the workflow, methods had already been developed by previous students or were already available in the Corthals lab. These methods have been adopted without changing them in this project.

Deparaffinization

Anitgen retrieval

Protein extraction

Reduction & alkylation

Desalting Lyophilisation LC-MS/MS analysis Data analysis Phosphopeptide enrichment Digestion Kidney tissue

(22)

17

2.2.1 Sample preparation

The complete protocol for sample preparation can be found in Appendix B. Table 4. Buffers used for sample preparation.

!_{FFPE samples. FFPE tissue samples need to be deparaffinized followed by an antigen} retrieval step, which reverses the protein crosslinking induced by formalin fixation. The FFPE homogenates that were used were already deparaffinized, so this step could be skipped, and the procedure was commenced with antigen retrieval. FF samples are not formalin fixated, so sample preparation can start at the protein extraction step in the protocol.

Antigen retrieval

To each FFPE sample 50 µl of 30% ACN in 100 mM NH4HCO3 (10x diluted stock solution) was added. The samples were briefly vortexed and incubated for 90 min at 95°C in the thermomixer under moderate mixing (around 600 rpm).

Protein extraction and reduction

For each sample, a reduction solution was prepared as follows: first, a 30% ACN v/v in 100 mM NH4HCO3 solution and a solution of 700 mM Dithiothreitol (DTT) in 25 mM NH4HCO3 were prepared. In a new vial, 100 µL of the 100 mM NH4HCO3 solution was added for each sample, followed by for each sample 72 mg urea, 10 µL 1M NH4HCO3 solution and 2.8 µL of the DTT solution. 100 µL of this solution was added to each sample and incubated in a thermomixer for 30 min at 37°C.

Alkylation

100µl of a solution of 700 mM Iodoacetamide (IAA) in 25 mM NH4HCO3 was added to each sample and incubated in a thermomixer for 30 min at 37°C, covered with aluminum foil to protect the samples from light.

Digestion

To each sample was added 120 µL 1M NH4HCO3 solution and 880 µL H2O. The amount of trypsin to add to each sample was calculated as follows: Trypsin works optimally when added to the sample in a 1:30 ratio (trypsin:proteins) and earlier experiments in our lab had shown that the amount of protein extracted from the tissue samples was roughly 7% (w/w). Each sample contained 2.5 mg tissue, giving 0.175 mg protein, so 5.83 µg trypsin should be added per sample. A stock solution of 1 1µg/µL trypsin was prepared and 5.8 µL was added to each sample. The samples were incubated overnight in a thermomixer at 37°C.

To quench the digestion, 50 µL of a 5% Trifluoroacetic acid (TFA) solution was added to each sample, and the samples were centrifuged for 45 min. at 12500 rpm at 20°C.

1 M ammonium bicarbonate 0.7906 g NH4HCO3 dissolved in 10 mL H2O 30% (v/v) ACN/0.1 M

ammonium bicarbonate 300 μL ACN added to 700 μL of 0.1 M ammonium bicarbonate 0.7 M DTT 10.8 mg dithiothreitol dissolved in 100 μL of 25 mM ammonium

bicarbonate

Solution A 72 mg urea dissolved in 100 μL of 30% ACN/0.1 M ammonium bicarbonate, to which 2.8 μL of 0.7 M DTT was added

0.7 M IAA 12.9 mg iodoacetamide dissolved in 100 μL of 25 Trypsin buffer 120 μL of 1 M ABC mixed with 880 μL H2O

(23)

18 Desalting

Samples were desalted using 4mm/1ml C18 SPE cartridges. A revolver tang was used to punch holes in the lids of Eppendorf tubes, through which the small bottom parts of the SPE tubes could be inserted. This way, the desalting could be performed in a bench top centrifuge. The solution were pipetted into the SPE tubes and the flow-through was collected in the Eppendorf tubes. A new set of Eppendorf tubes was used for the elution.

Table 5. Overview of desalting procedure after sample preparation.

At this point each peptide sample was divided into two fractions: 90% destined for IMAC enrichment, while the other 10% would not be enriched and analyzed with LC-MS/MS for comparison.

Lyophilization

The lid of each Eppendorf containing peptide sample was replaced for one with a small hole in it, and the samples were lyophilized overnight in a freeze drier.

2.2.2 Peptide essay

A fluorometric essay was used to determine the peptide concentration after sample preparation, using Pierce™ Quantitative Fluorometric Peptide Assay. In this assay an amine-reactive fluorescent reagent is used that uniquely labels the N-terminus of peptides. The unenriched fraction of each sample was dissolved in a 35 µL 3% ACN/0.1% FA solution. To determine the peptide concentration, 10 μL of the sample solution, 70 μL of the assay buffer, and 20 μL of the fluorescent reagents were added to a well of a black 96 well plate. A calibration line was constructed ranging from 7.8 - 1000 μg/mL by diluting a series of a standard tryptic digest, using the same loading buffer as the peptide samples. Calibration lines were performed in triplicate and each sample in duplicate. As soon as the fluorescent reagent was added, the plate was kept in the dark and incubated for 30 min. Fluorescence was measured with a fluorescence plate reader using 390 nm excitation and 475 nm emission. The gain was set as 100, the read speed was normal, the read height was 8 mm and each well was measured 10 times with an accumulation delay of 100 msec. The complete protocol can be found in Appendix C.

The measures of the calibration line were averaged and the background signal was subtracted to calculate the calibration line. This was used to estimate the peptide content of the samples from the average of the duplicate signals.

Step Solution Amount (µL) Centrifugation Repeats

RPM time (min)

Wash 50% ACN in H2O 300 1000 2 2

0.1% TFA 300 1000 2 2

Load sample max. 1000 1000 5 until all sample

has been loaded

Wash 0.1% TFA 300 1000 2 2

Elute 0.1% formic acid (FA)

(24)

19

2.3 IMAC method development

In the Corthals lab, a method for phosphopeptide enrichment had already been developed several years ago. Due to practical difficulties (a more detailed description of which will be provided in the discussion section), this method could not be used during this project. Therefore, a new method had to be developed, including a new way to fabricate columns. Several approaches have been tested. The most successful methods are adopted in the final protocol (Appendix D and E).

2.3.1 _{Approach 1: Batch incubation and GELoader tip IMAC columns}

Overview

In this approach an IMAC column is fabricated inside a GELoader tip®. The POROS™ 20 MC media is first activated by incubation with FeCl3. The IMAC beads are subsequently incubated with the peptide mixture to allow binding of the phosphopeptides. An IMAC column is constructed by squeezing the constricted end of a GELoader® tip. The IMAC beads with bound phosphopeptides are packed into the tip to form a column from which the phosphopeptides can be eluted. An overview of this method is provided in (Figure 17)

Method Solutions

Table 6. Solutions that were used during the enrichment protocol

50 mM EDTA in 1 M NaCl 1.46 g EDTA was added to 100 mL H2O. Pellets of NaOH were added until the pH reached 8 and EDTA dissolved. 5.84 g NaCl was added to this solution.

0.1 M FeCl3 in 0.1 M acetic acid 81.1 mg FeCl3 dissolved in 5 mL acetic acid

Loading buffer 0.1% TFA in 50% ACN

Elution buffer 1% ammonia water (20 μL ammonia solution (10 %), 180 μL H2O (pH ~11)) IMAC beads Fe3+_ions Non-phosphorylated peptides Phosphorylated peptides Activating IMAC

beads with Fe3+_-ions Mixing IMAC beads _{with peptide sample}

Incubation Incubation

Packing beads in GELoader® tip

Elution

(25)

20 Samples

Both FF and FFPE homogenate samples were used for this experiment. Based on the peptide concentration determined by the fluorometric assay, samples were pooled to contain ca. 56 µg peptide per sample for FF samples and 17 µg peptide per sample for FFPE samples.

Activating the IMAC resin

A slurry of IMAC resin was prepared by weighing 50 mg resin and suspending it in 1 mL MiliQ H2O. The resin was then washed, activated and equilibrated as shown in Table 7. After adding a reagent to the resin, the beads were spun down and the supernatant was removed carefully, while trying to avoid removing any of the resin. Different centrifugation speeds were tested to find the lowest speed at which the IMAC resin forms a pellet at the bottom of the Eppendorf tube.

Table 7. Overview of procedure for activating IMAC beads with Fe3+_{. Protcol adapted from Lee et al. (2007).}37

Step Reagent Amount (µL) Centrifugation Repeats

Speed (rpm) Duration (min) Washing 50 mM EDTA in 1

M NaCl 500 100 - 6000 5 3

Washing H2O 500 100 - 6000 5 3

Washing 0.1 M acetic acid 500 100 - 6000 5 3

Activation 0.1 M FeCl3 in 0.1

M acetic acid 500 6000 5 3

Washing 0.1 M acetic acid 500 6000 5 2

Equilibration loading buffer 500 6000 5 1

Sample incubation

The beads were resuspended in 500 µL loading buffer. For each separate enrichment 100 µL of resin was transferred to a new 1.5 mL Eppendorf tube. The peptide sample to be enriched was suspended in 30 µL loading buffer and added to the resin. The sample was incubated with the IMAC beads in a Thermomixer for 30 min at room temperature while shaking at 650 rpm.

Packing the IMAC column

To construct the IMAC column, the constricted end of an Eppendorf GELoader® tip was squeezed to create a passage narrow enough to hold the IMAC resin inside the column but allow the liquid to flow though. From the wide end of the tip ca. 0.5 cm was cut off, so that the tip of a plastic syringe just fitted inside. The tip was rinsed once with loading buffer. A maximum 50 µL of the resin with the sample was added to the tip and with a 1 mL plastic syringe air pressure was applied onto the open end of the tip. The resin was packed inside the narrow end of the tip and the liquid flowed through. This step was repeated to load the remaining amount of slurry. Figure 18 shows the IMAC column after loading the resin. The resin was washed once by adding 40 µL loading buffer and pushing this through the column with air pressure. Figure 1916_{gives a step by step demonstration of the procedure} of preparing the column and loading the resin.

Elution of phosphopeptides

The phosphorylated peptides bound to the column were eluted into a clean Eppendorf tube using 30 μL elution buffer. This step was performed slowly (~1 drop/s) and repeated 2 times.

(26)

21 GELoader Tip

1 cm

IMAC beads (with bound phosphopeptides)

Constricted end of tip

Figure 18. Photo of a packed GeLoader tip IMAC column. The constricted end of the tip is squeezed to create a very

narrow openings, through which the IMAC resin cannot pass. The white area that is visible is the IMAC resin, to which the phosphopeptides are bound.

Figure 19. IMAC column construction in GELoader tips. A) The constricted end of a GELoader tip is

squeezed to prevent the IMAC beads from leaking. B) The top of the GELoader tip is cut to make a plastic syringe fit into the opening. C) The IMAC beads are loaded onto the tip. D) The IMAC beads are packed by applying air pressure using a plastic syringe to form a column. E) The same principle can also be used to make larger columns. Reprinted from Thingholm & Jensen (2009).

(27)

22 Desalting

Small reversed phase microcolumns used for desalting and concentration were fabricated as follows: with the end of a glass pipet, small plugs of C8 material were stamped out of a 3M Empore C18 extraction disk. Three layers of these were inserted with the pipet into the constricted end of a GELoader® tip, to create a small column.

The desalting was performed as described in Table 8. In each step the reagent was added to the GELoader tip and then pushed through the column by centrifugal force. The GELoader® tip was placed into an Eppendorf tube through a small hole that was cut out of the cap. This way the column could be used in a bench-top centrifuge. The peptide samples were acidified with 150 µL of 2% FA before loading onto the column.

Table 8. Overview of desalting procedure after phosphopeptide enrichment, using a homemade C18 column.

Step Reagent Volume

(µL) Speed (rpm) Centrifugation Duration (min)

Wetting ACN 50 5000 1

Equilibration 0.1% TFA 50 5000 1

Loading sample eluate from enrichment

experiment 200 5000 2

Washing 10% FA 50 5000 1

2x 5000

Elution 0.1% FA, 80 % ACN 50 2000 2

Samples were then freeze-dried and stored at -20˚C until LC-MS analysis. LC-MS/MS

Samples (both enriched and unenriched) were analyzed with a nano-LC system coupled to an ESI- tripleTOF MS/MS system. Before analysis samples were lyophilized and resuspended in a sampling buffer containing 0.1% FA and 3% ACN, the buffer volume depending on the peptide amount in the sample. For non-enriched samples, the injection amount was around 1 μg, based on peptide concentrations estimated by the fluorometric peptide assay. Enriched samples, which contained small amounts of peptides, were resuspended in 12 μL of loading buffer, from which 10 μL was injected into the LC-MS/MS system.

Peptides were firstly loaded onto a trap column (1 cm long, 75 μm ID, 5 μm particle size, 100 Å) to remove any remaining salt or contaminants. Peptides were eluted from the trap column and separated on the analytical column at a flow rate of 300 nL/min. The method that was used for LC-MS was developed by a previous student (D. Patel) for a phosphopeptide enrichment study and had not been changed for this study. Two gradients were used, with mobile phase A consisting of 0.1% FA in H2O and the mobile phase B consisting of 0.1% FA in ACN. Protocol 1, which has a steeper gradient, was used for enriched samples, which have lower peptide contents, while protocol 2, with a less steep gradient, was used for non-enriched control samples.

Mass spectrometry was used in positive ion mode, using DDA and scanning precursor ions with a m/z range of 400 – 1250 for 500 ms. The 30 most intense precursor ions with charge state between 2 and 4, and with at least 100 counts per second were selected for subsequent fragmentation. Each MS/MS scan (m/z 100 – 1800) was accumulated for 100 ms per precursor ion with an isolation window of 50 mDa in high sensitivity mode. The total cycle time was around 3.5 s.

(28)

23 Table 9. Gradients used for LC separation.

Protocol 1 Protocol 2

Gradient 5% - 34% 5% - 40%

Gradient time 0 - 24 min 0 - 45 min

Gradient slope 1.21 %B/min slope 0.78 %B/min Total analysis time 46 min 60 min

Data analysis

Raw MS/MS data was analyzed using ProteinPilot (version 5.0, AB Sciex), searching against the human Swiss-Prot database (downloaded on 23-01-2019, 20412 entries) with the search parameters set to the following values: Identification; Iodoacetamide for cystine alkylation; Trypsin digestion; TripleTOF 5600; Homo sapiens; Thorough search effort; Biological modifications; Detected protein threshold 0.05; conducting FDR analysis. Phosphorylation emphasis was selected.

The output generated by ProteinPilot is an Excel file containing the number of identifications at spectra, distinct peptide and protein levels with global or local 1%, 5%, and 10% FDR. Only peptide identifications with global 1% FDR were considered as identified peptides.

(29)

24

Figure 16. An IMAC column made by pulling the capillary apart over a flame to

create a narrow tip, which functions as a frit.

2.3.2 Approach 2: Capillary IMAC columns

In this approach an IMAC column is constructed from a fused silica capillary. In contrast to the batch incubation method, the capillary column is first packed with the IMAC resin, which is subsequently activated by flowing Fe3+_{ions through the column. The IMAC column can then be used for the} enrichment of phosphopeptides.

2.3.2.1 Column construction

Frits

Frits are produced at one end of the column to keep the resin inside of the column when pressure is applied but allow liquid to flow through. For construction of all columns, fused silica capillaries were cut to a length of ca. 15 cm. The optimal inner diameter for the column was not known, so capillaries with IDs of 75, 100, 150 and 200 µm were tested.

Method 1: Corthals lab standard procedure

50 µL formamide and 150 µL water were added together and vortexed. To this mixture 200 µL Kasil®1624 was added and it was vortexed again. 2 µL of this solution was pipetted onto a glass fiber disk. On end of the capillary was put on the Kasil spot on the disk and twisted until it got through. This way, a small plug of glass fiber was inserted in the capillary. This procedure was repeated until there were 5 layers of glass fiber in each capillary. The capillaries were put in a clean Eppendorf tube with the frit end down and left in a Thermomixer at 80˚C overnight.

Method 2: Pulling capillaries over a flame

The middle of a fused silica capillary was heated over a blue Bunsen flame. First, the capillary’s coating was burned away. Then, when the middle became hot enough and pliable, the ends were carefully pulled away from each other, drawing the middle into a very thin string until it broke. This resulted in two capillaries with very narrow endings, ideally small enough to hold resin inside the capillary and let the liquid flow though (Figure 20)

Method 3: Kasil®1624 frits

Kasil®1624 and formamide were mixed in a 3:1 ratio. Usually 150 µL Kasil®1624 and 50 µL is used. This is enough to produce at least 20 frits. It is not recommended to use less, for this would be more difficult to work with. First, Kasil®1624 was added to a glass vial (it is very important to use glass vials, to prevent polypropylene from Eppendorf tubes to leach and polymerize into the frit). While vortexing the vial, the formamide was added and mixed by pipetting up and down. To make the frits, the capillaries were dipped shortly (~1 s) into the mixture, one at a time, and placed with the frit end down in clean glass vials. It is important that these steps are performed swiftly, before the mixture start polymerizing. The vial with capillaries was placed in an oven of 100˚C to let the frits polymerize overnight.

1 cm

Narrow tip

Part of the capillary without coating

(30)

25

The next day the frits were cut to a length of 5-10 mm and flushed for 5 minutes with ethanol, the solvent that would later be used for column packing. Finally, the columns were dried with a 3 min helium flush. The columns could then be stored until further use.

Column packing

To pack the IMAC resin inside the capillary, a helium pressure vessel was used as a pressure source to push a slurry of IMAC beads into the column. Figure 21 shows this system when used to pack a flame pulled capillary column. The pressure vessel will be discussed in more detail in the next section. In short, an Eppendorf tube containing the slurry is placed inside the vessel, into which the capillary is inserted. When pressure is applied, the only way the slurry can go is into the capillary, where it is retained by the frit, while the liquid flows out of the column.

Preparing the slurry

5 mg IMAC resinwas weighed and resuspend in 1 mL MiliQ H2O. The suspension was homogenized by shaking the Eppendorf tube, and 200 µL of the suspension was transferred to a 2 mL Eppendorf tube and 800 µL ethanol was added.

Packing columns

The suspension was homogenized well before it was placed inside the pressure vessel. A capillary was placed inside the pressure vessel, with the frit end upwards and the other end inserted in the Eppendorf containing the IMAC slurry, approximately 1 mm above the bottom of the tube. A pressure of 200 psi was applied, forcing the resin into the capillary. When the column was packed to a length ~0.5 cm shorter than the desired length, the pressure vessel was closed. After 5 minutes, the column could be removed from the pressure vessel. During these 5 minutes, the column was packed a little bit further be pressure that was still in the system. The pressure escapes from the system slowly and removing the column too soon from the pressure vessel can cause gaps in the column from the sudden pressure drop.

Figure 17 The helium pressure vessel when

(31)

26 Storing columns

To prevent dehydration of the resin, an Eppendorf tube was filled with ethanol and covered with parafilm. The end with frit was inserted into the Eppendorf through the parafilm and the column was stored vertically, to prevent resin from falling out. When columns with a flame-pulled tip were used, only the intact end was inserted into the Eppendorf to prevent damaging the tip and the column was kept horizontally.

2.3.2.2 Pressure sources

Different pressure sources can be used to force the liquid through an IMAC capillary column. Several methods were tested.

Syringe pump

To use a syringe pump, the liquid to be passed through the column is first taken in by the syringe. The syringe is then inserted into the pump and the column is connected to the tip of the syringe. The syringe pump moves the plunger towards the tip to dispense the liquid and push it through the connected column. The flow rate and volume to be dispensed can be set on the pump.

The syringe pump was used was the in combination with a 1 mL Hamilton glass syringe. The metal needle from the syringe was removed to avoid sample losses from adhesion of phosphopeptides to the metal, to which they can bind easily. The needle was replaced with a MicroTight tubing PEEK sleeve, which was secured in the plastic needle hub with an epoxy resin adhesive (Figure 22A & 22B). This is a type of adhesive that is known for its high strength binding properties and chemical resistance, and can be used on many types of materials. To fill the syringe, liquids could be drawn up directly through the PEEK sleeve. The column was attached to the syringe be inserting it into the sleeve and securing it with a PEEK MicroTight union and fitting (Figure 22C).The column was held in place by a support stand with clamps and could stay in place while the syringe was removed from the pump to change its content (Figure 23). When there was leakage between the union and fitting, metal pieces were used to make a tighter connection.

A

B

C

Figure 18. A) Original needle hub with metal needle. B) Original needle hub (a new, clean one), where the metal

needle is replaced by a PEEK sleeve. C) 1 mL Hamilton syringe, with an IMAC capillary column attached, using a PEEK MicroTight union and fitting.

(32)

27

As an attempt to remedy leakage of liquid through the needle hub, the surface of the hub was sealed with epoxy adhesive (24B). As an alternative, a new hub was made from PEEK, a material s known to be able to withstand high pressure and to be chemically inert (24A). An o-ring was placed between the hub and the needle to make a tight connection.

Helium pressure vessel

The helium pressure vessel (Figure 25A & 25B) consists of a metal cylinder with a long but narrow cavity inside, and a metal lid, which can be secured onto the cylinder with bolds and screws. The cylinder has a tap to which a cylinder is connected containing pressurized helium gas. An Eppendorf tube (with the lid cut off) containing the liquid to be passed through the column is inserted into the cavity and the lid is closed securely. A capillary column can be inserted through an inlet in the lid and into the Eppendorf tube. When opening the valve of the tap, helium gas flows into the cavity in the vessel and pushes the liquid from Eppendorf tube through the column. The pressure can be controlled by adjusting the taps on the pressure reducer on the helium tank (Figure 25C).

Figure 20. A) 1 ml Hamilton syringe with custom made PEEK hub and sleeve. B) 1 ml Hamilton syringe with PEEK sleeve and original needle hub, sealed with

epoxy adhesive to reduce leakage.

A

B

Figure 19. Syringe pump with syringe inserted. An IMAC column is connected to the syringe with metal

(33)

28

Figure 21. A) Picture of the helium pressure vessel that

was used. B) Schematic representation of the pressure vessel. C) Picture of the helium tank that was used, with attached pressure reducer, which can be used to adapt the pressure to the desired value.

Liquid outlet Helium Eppendorf tube containing solution

A

B

C

(34)

29

2.3.2.3 Phosphopeptide enrichment

This procedure consists of two parts: preparing the IMAC column and the actual enrichment experiment. Once a column is prepared, it can be stored until further use, and the same column can be used for multiple IMAC experiments.

Set-up

Figure 26 shows the set-up that was used for this experiment. The helium pressure vessel, as described above, was used as a pressure source for the entire protocol. For the enrichment experiments IMAC capillary columns were used with IDs of 75, 100 and 150 µm. A piece of capillary (ca. 20 cm, same ID as the used column) was connected to the IMAC column with a PEEK MicroTight union and fittings, using a PEEK sleeve to form a tight connection. This extension was inserted into the pressure vessel and served as an inlet to the column. Both the column and the extension capillary were held up by support stands and clamps. Small pieces of Styrofoam were placed between the capillaries and the clamps to hold them in place. The column was bended in a downward direction to facilitate collection of the column outflow. To change the sample inside the pressure vessel, the extension capillary could be removed from the pressure vessel while staying connected to the column, which did not need to be moved. The other reason for using an extension is to avoid bending of the column for collecting the outflow, for this may disturb the IMAC resin and affect the experiment.

Figure 22. A) Overview picture of the setup that was used. The helium

pressure vessel was used as a pressure source. A 15 cm piece of fused silica capillary was connected to the column and served as an inlet. The other end was inserted into the bomb to load solutions onto the column. The column was held in place by two support stands and clamps, using pieces of Styrofoam between the capillary and the clamps. The column was oriented in a downward direction and an Eppendorf tube was placed underneath the column outlet to collect flow through. B) Close-up of column outlet and Eppendorf tube to collect column flow through.

A

(35)

30 Part I: Preparing the IMAC column

Solutions

Table 10. Solutions that were used for preparing the IMAC column. * was prepared freshly on the day of the experiment

Method

The IMAC beads need to be activated with FeCl3 before the column can be used for phosphopeptide enrichment. The column was washed, activated and equilibrated by letting the different solutions flow through the column (Table 11). When each solution had been flowed through the column for the indicated time, the valve on pressure vessel was closed. After allowing 5 minutes for the pressure to slowly escape from the system, the extension capillary was removed from the pressure vessel. The lid was then opened and the Eppendorf inside replaced with an Eppendorf tube containing the next solution, closing it securely afterwards. The valve on the pressure vessel was then opened again to flow the next solution through the column. This procedure was repeated for all solvents.

Table 11. Procedure for the activation of an IMAC capillary column. *Activation: Load FeCl3 solution during 10 min

at 600 psi. Then close pressure vessel and wait 5 min to let incubate. Repeat 3 times.

Part II: Phosphopeptide enrichment Solutions

Table 12. Solutions that were used during phosphopeptide enrichment in an IMAC capillary column. All buffers were

prepared freshly on the day of the experiment.

Samples

For this experiment only FF homogenate samples were used, no FFPE samples. For each experiment approximately 100 µg peptide was dissolved in 40 µL loading buffer and transferred to a 2.0 mL Eppendorf tube.

50 mM EDTA 1.46 g EDTA was added to 100 mL H2O. Pellets of NaOH were added until the pH reached 8 and EDTA dissolved. 5.84 g NaCl was added to this solution.

0.1 M acetic acid 57 µL acetic acid was dissolved in 10 mL H2O 0.1 M FeCl3 in 0.1 M acetic acid* 81.1 mg FeCl3 dissolved in 5 mL 0.1M acetic acid

Step Solution Pressure (psi) Time (min.)

Washing H2O 600 20

50 mM EDTA 600 10

H2O 600 15

0.1 M acetic acid 600 15

Activation 0.1 M FeCl3 in 0.1 M

acetic acid 600 10 + 5 min waiting repeat 3 times*

Equilibration 0.1 M acetic acid 600 40

Loading buffer 0.1% TFA in 50% ACN

0.1 M acetic acid 57 µL acetic acid was dissolved in 10 mL H2O

(36)

31 Method

The enrichment was carried out in a similar manner as the preparation of the column, but using lower pressure. The procedure is displayed in Table 13.

Table 13. Procedure for phosphopeptide enrichment. *This washing step is not necessary if the enrichment is

performed directly after equilibrating the column. **Washing the column after the experiment is only required if the column is to be reused.

LC-MS/MS

Because no LC-MS was available, the enriched samples could not be analyzed in our lab. A small number of samples was analyzed in another lab, using a method similar to the one described previously. Unfortunately, more details are not available.

Step Solution Pressure (psi) Time (min.)

Washing* 0.1 M acetic acid 200 5

Load sample peptide sample 100-200 ~20

Washing loading buffer 200-400 20

H2O 200-400 20

Elution elution buffer 100-200 50

Washing** H2O 200 10

Development of an IMAC method for the enrichment of phosphopeptides from human kidney tissue

Research project

MSc Chemistry - Analytical Sciences