Optimisation and Validation of PCR Method for HLA Gene Expression to Enable PCR System Transfer and Master Mix Change

(1)

Linköping University | Department of Physics, Chemistry and Biology Bachelor’s Thesis, 16 hp | Educational Program: Chemical Biology Spring term 2020 | LITH-IFM-G-EX—20/3892--SE

Optimisation and Validation of

PCR Method for HLA Gene

Expression to Enable PCR System

Transfer and Master Mix Change

Paulina Odlander

Examiner: Lars-Göran Mårtensson

(2)

i

Abstract

98% - 99,85% of patients diagnosed with celiac disease, gluten intolerance, carry either human leukocyte antigen (HLA) genes DQ2 or DQ8, which is why detection of these serotypes is the most relevant method in determining celiac disease risk or diagnosis. If suffering from celiac disease, getting diagnosed and being put on a gluten restrictive or gluten-free diet is a health profit. Not doing so can lead to deteriorating absorption of nutrients in the intestine leading to malnutrition, and also long-term complications such as cancer, infertility, autoimmunity and weakening of the bone tissue.

Health Tech company Dynamic Code AB provides a PCR test for determination of HLA DQ-genes connected to development of celiac disease. The PCR method is probe based and in real time and is at this time carried through on the, somewhat outdated, PCR instrument from Thermo Fisher/Applied Biosystems called 7300 Real-Time PCR System. The run time for this analysis on the instrument is 1 hour and 50 minutes. The Master Mix in use is TaqMan™ Gene Expression Master Mix, from the same manufacturer.

Moving on to a more modern PCR instrument is a natural step for the company and is favourable in several regards, one of them being the run time that will be cut by 50 minutes, allowing for more samples to be analysed in the same amount of time. The objective is to move the HLA analysis to Thermo Fisher’s QuantStudio™ 6 and 7 Flex Systems and at the same time change the Master Mix to SolisFast® Probe qPCR Mix (Purple) from Solis BioDyne, in order to achieve better accuracy as this Master Mix is more compatible with the latter instrument, along with reducing reagent cost as it is less expensive.

In order to find the optimal primer and probe concentration for each target included in the HLA analysis, their concentrations were varied and tested with the new Master Mix on the new instrument.

PCR instrument transfer and Master Mix change was successful and validation experiments showed a 98,9% accuracy for the new method compared to the original method.

(3)

(4)

iii

Abbreviations

ABI 7300 or 7300 – Applied Biosystems 7300 Real-Time PCR System HLA – Human leukocyte antibody

IPC – Internal positive control PCR – Polymerase chain reaction QS – QuantStudio™ 6 or 7 Flex System Solis - SolisFAST® Probe qPCR Mix (Purple) TaqMan - TaqMan™ Gene Expression Master Mix

(5)

iv

v 5.2. Mix 2: DQA1_02... 24 5.2.1. Mix 2: DQB1_02 ... 24 5.2.2. Mix 2: Probe 1 ... 25 5.2.3. Mix 2: Probe 2 ... 25 5.3. Mix 3: DQA1_03... 26 5.3.2. Mix 3: Probe 1 ... 27 5.3.3. Mix 3: Probe 2 ... 28 5.4. Mix 4: DQB1*0302 B ... 28 5.4.1. Mix 4: IPC ... 29 5.4.2. Mix 4: Probe 1 ... 29 5.4.3. Mix 4: Probe 2 ... 29 5.5. Validation ... 30 5.5.1. Mix 1 ... 30 5.5.2. Mix 2 ... 30 5.5.3. Mix 3 ... 30 5.5.4. Mix 4 ... 30

5.6. Final Primer and Probe Concentrations ... 31

5.7. Final Affirmation of Genotypes ... 32

6. Discussion ... 36 7. Conclusion ... 38 8. Acknowledgments ... 39 9. References ... 40 Appendix A ... 42 1. PCR-plate Setups ... 42 Appendix B ... 44

1. PCR Amplification Plots – Optimisation experiments ... 44

(7)

1

1. Introduction

The polymerase chain reaction (PCR) technique is still the most fundamental strategy being used for gene amplification by scientists all over the world. The technique enables detection of foreign pathogens e.g. virus and bacteria in biological material, such as human cells. PCR can also be used to perform genotyping, which refers to the process of determining differences in genes. Both gene amplification and genotyping are used for medical purposes in detection of infectious diseases or when determining risk factors for development of genetically caused disorders.

One genetically caused disorder in humans are in focus in this project - the risk for genetically disposed coeliac disease. Through genotyping of the specific genes playing the big role in development of the disease, it is possible to determine the potential risk for a specific individual to develop it.

Besides the PCR instrument itself, one of the central reagents used in a PCR is the so-called Master Mix, in which samples are placed, which contains some of the crucial components needed to perform the reaction.

The health-tech company Dynamic Code AB performs PCR tests for determining risk for coeliac disease on the 7300 Real-Time PCR System from Applied Biosystems (ABI 7300) which is a part of ThermoFisher Scientific, a PCR instrument that has been on the market for many years. Although the instrument still is reliable, switching to the more modern instruments QuantStudio™ 6 and 7 Flex System (QS) from Applied Biosystems, run time for each analysis can be cut almost in half, which means almost double the number of samples can analysed in the same amount of time. By changing the instrument to the QS, it also becomes possible to scale up the production thanks to the interchangeable well formats, from 96-well to 384-well, which is not possible for the ABI 7300, an upscale which is relevant for Dynamic Code AB. Besides this, the QS, compared to the ABI 7300, has a greater sensitivity in detecting small changes in reactions and overall well-to-well and instrument-to-instrument data is improved. Changing the PCR instrument will greatly improve the operation at Dynamic Code.

The two methods will be referred to as the ‘original method’, the one currently in use by the company, carried through on the ABI 7300, and the ‘new method’ or ‘optimised method’, which is going to be the optimised method, carried through on the QS.

Data obtained from a PCR can be greatly influenced by the Master Mix used in terms of sensitivity, specificity, accuracy and dynamic range, and specific Master Mixes are optimised for usage with specific instruments to achieve optimal accuracy in test results. The TaqMan™ Gene Expression Master Mix from Thermo Fisher (TaqMan) is optimised for and recommended to use with ABI 7300 and SolisFAST® Probe qPCR Mix (Solis) from Solis Biodyne has compatibility with the QS system. Thus, changing the Master Mix along with the instrument transfer is essential to attain accurate and precise results. The new Master Mix is, in addition to this, more affordable which together with the compatibility aspect makes the enzyme change an important matter for the company.

(8)

2 1.1. Goal

The goal of the project is to successfully transfer analysis for detecting human leukocyte antibody (HLA) genes, from an old PCR instrument to a latter one, changing the Master Mix and optimising the method through adjustments of parameters, in specific, primer and probe concentrations. The method will be validated, and results should be equal or at least comparable to the original method.

The objective is to not have the Ct-values generated from the analysis with the new method

differ too far off from the corresponding Ct-values generated from the original method, so that

patient diagnosis based on Ct-values does not have to be re-evaluated and changed.

1.2. Purpose

The purpose of the project is to streamline the testing procedure by cutting down the time for which a PCR run requires as well as cutting down on reagent cost. Moving on from the outdated PCR machine to the more relevant version is also eagerly awaited and a natural transition to make.

Table 1: Planned changes.

Change from Change to

PCR instrument Applied Biosystems 7300 Real-Time PCR System (Thermo Fisher)

QuantStudio™ 6 and 7 Flex Systems (Thermo Fisher)

Run time 1 hour 50 minutes 1 hour

Master Mix TaqMan™ Gene Expression

Master Mix from Thermo Fisher

SolisFast® Probe qPCR Mix (Purple) from Solis BioDyne

Passive Reference Dye ROX Mustang purple

DNA polymerase AmpliTaq Gold DNA Polymerase

SolisFAST® DNA Polymerase

Probe used TaqMan® TaqMan®

(9)

3

2. Theoretical Background

2.1. Celiac Disease

Celiac disease is a chronic autoimmune disease that causes ingestion of gluten protein to start an inflammation that damages the intestinal tract of the small intestine, which can lead to deteriorating absorption of nutrients in the intestine leading to malnutrition, and also long term complications such as cancer, infertility, autoimmunity and weakening of the bone tissue (Megiorni et al. 2009). Since celiac disease is a genetic condition, the spread of individuals carrying the genetic variation that causes celiac disease is uneven distributed throughout the world, although it exists on all continents (Lindfors et al. 2019) with an average occurrence of approximately 1%-2% in Europe (Catassi et al. 2007). Although the disease is common, a vast majority of genetically predisposed individuals do remain undiagnosed, ostensibly due to a vague symptom picture or a complete lack of symptoms (Megiorni et al. 2008). There is a health profit in getting diagnosed and being put on a gluten restrictive or gluten-free diet, which to this date is the only effective treatment (Tack et al. 2010), which is why Dynamic Code is providing a test for detection of celiac disease.

The test detects the presence of DQ2.5-, DQ2.2- and DQ8-alleles (genetic predisposition), by gene amplification through PCR. Specific antibodies for tTG-IgA and total-IgA in the sample (to rule out igA-sufficency) is also detected through enzyme-linked immunosorbent assay (ELISA)-based analysis. Anti-tissue transglutaminase, i.e. Anti-tTG, and anti-endomysial antibodies, i.e. EmA, is detected through serologic screening, e.g. ELISA, and the presence of these indicates on an active phase of celiac disease in the patient (Megiornia et al. 2008).

2.1.1. Human Leukocyte Antibody

The majority of the genetic risk factors for developing celiac disease lie within the over 200 genes located in the region for the major histocompatibility complex (MHC), in humans titled human leukocyte antibody (HLA), on chromosome 6.

The HLA complex, which is the most gene-dense region in the entire genome, codes for around ten thousand different cell surface peptides, i.e. antigens, especially located on dendritic cells, B lymphocytes and macrophages, that can be recognised by T cells that produce immune responses to fight foreign pathogens, thus playing a crucial role in the immune system (Shen et. al, 2010; Williams 2001). If immune cells recognise self-antigens as foreign antigens, autoimmune diseases, such as celiac disease, can occur.

The MHC is divided into three class regions, I, II and III. Within class II lies loci for HLA DQ-genes, which will be in focus in this report. The HLA-DQ gene complex furthermore consists of two genes, HLA-DQA1 and HLA-DQB1. Further, the gene complex and the genes combine in different serotypes and haplotypes, some of them listed in Table 2.

(10)

4

Table 2: Genetic variants in the HLA gene complex that can cause celiac disease.

Name Type Serotypes Haplotypes

HLA-DQ Gene complex 2, 8 -

HLA-DQA1 Gene - 03, 05

HLA-DQB1 Gene - 02, 0302

Table 3: Combinations of allelic variants of the DQA1 and DQB1 alleles and the risk the combinations compose in developing celiac disease (Megiorni et. al 2007).

HLA-DQ heterodimer Risk for developing

celiac disease

DQ2.5 and DQ8 Very high

DQ2.5 (with a double dose of DQB1*02) Very high

DQ8 High

DQ2.5 (single dose of DQB1*02) High DQ2.x (double dose of DQB1*02) High DQ2.x (single dose of DQB1*02) Low

DQX.5 Extremely low

DQX.x Extremely low

In almost all cases, genetically predisposed individual’s express HLA-DQ2 or HLA-DQ8 variants (Table 3), which makes detection of these serotypes the most relevant method in determining celiac disease risk or diagnosis (Megorini et. al 2008; Brown, Guandalini, Semrad & Kupfer 2019; Cecilio & Bonatto 2015). Both genes code for α/β heterodimers involved in mediating the activation of CD4+, gluten-specific T-cells, with gut-homing potential possibly leading to lymphocytic elimination of epithelial cells (Jabri & Sollid 2017).

The two HLA genes, DQ2 and DQ8, account for approximately 98% - 99,84% of patient diagnosis, based on a large European collaborative study, one American study and one Italian study (Karell et al. 2003; Megiorni et al. 2009; Pietzak et. al 2009), and approximately 95% of patients carry allelic variants DQA1*05 and DQB1*02 in DQ2 gene, or DQA1*03 and DQB1*0302 in DQ8 gene (Megiorni et. al 2007), although heterozygosity for either of the DQ2 or DQ8 alleles also induce a risk. Detection of the sensible DQ genes does not predict development of celiac disease but their absence means a negative prediction of the disease to close to 100% (Wolters & Wijmenga 2008).

The HLA complex is the most convoluted and polymorphic genetic system in the entire human genome to ever be discovered. One single locus can be possessed by (for HLA-A, HLA-B, HLA-C and DRB1) hundreds of different allelic variants (Williams 2001), each variant determining which antigens will be presented on the cell surface. Because of the redundancy in allelic variants, the chance that an individual will possess the same allele on both chromosomes is small, resulting in a wide variation of genotypes within the human population. Furthermore, the two alleles operate in co-dominance, making the variance even more complex. Some of these allelic variants (presented in Table 2 and Table 3) are known to cause celiac disease.

HLA class I and II genes show significant sequence homology due to their mutual descend from the same common ancestor as a result of former gene duplication events. The genes have closely related pseudogenes (defect, non-functional genes) which errors prevent successful transcription and translation. These facts make primer designing for PCR diagnosis a critically important matter to accomplish locus specific amplification in molecular diagnosis. HLA allele identification is to be considered the most complex problem in molecular diagnostics, with more than 1300 known HLA alleles (Williams 2001).

(11)

5 2.2. Polymerase Chain Reaction

The polymerase chain reaction (PCR) allows for exponential amplification of specific DNA fragments through repeating of cycles under specific conditions (Harris 1998). The reaction is continued until desired amount of product is acquired, often around 40 cycles. After 40 cycles, one single DNA molecule could theoretically have generated over a thousand billion copies (240 = 1,099,511,627,776). Although, in reality, amplification efficiency is not always 100% but rather 90-105% (Johnson et. al 2013).

PCR is an artificial mimic of when a cell, in vivo, replicates its DNA. Although a cell uses the aid of helicase to separate the double strand and RNA primase to add primers to the strand, the procedure is not far off from what occurs during a PCR.

PCR has several fields of applications within molecular biology, such as gene expression, genotyping (detecting), mutagenesis, sequencing, cloning, etc. In this study, the DNA yield is used to detect variants of the HLA genes. In fact, the study of allelic sequence variation within the HLA locus via PCR was in 1986, only the year after the PCR method first was invented, the inception to forensic analysis based on PCR (Saiki et. al 1986).

The technique advanced into what is called quantitative PCR (qPCR) (also known as real time PCR but not to be confused with reverse transcriptase PCR as both tend to be abbreviated RT-PCR). qPCR allows for real time analysis of quantity of PCR product by measuring relative fluorescence emitted from unspecific fluorescent dye or fluorescent-labelled target-specific probes, which will be directly proportional to the amount PCR product being produced (more on probe-based qPCR under ‘Probe’) (Maddocks & Jenkins 2016).

From here on, mentions of “PCR” in this report will be referring to quantitative PCR.

(12)

6 2.2.1. The Components

Needed to carry out a quantitative polymerase chain reaction are, besides the PCR instrument itself and the PCR plate (often 96-well) or PCR tube, following reagents: template DNA, primers, dNTPs, Mg2+, reaction buffer, polymerase and nuclease free water, and some sort of reporter molecule. These components together are what is commonly called a Master Mix. First of all, a template of what is actually being replicated and amplified is necessary. This is often referred to as the ‘sample’. Template DNA can be derived from a eukaryotic cell, in that case often in the shape of linear DNA originated in the nucleus, or from a prokaryotic cell in the shape of a circular bacterial plasmid, or even a virus (the template can in this case even be derived from RNA and must be converted into complementary DNA through reverse transcription)(Pfaffl 2010; Maddocks & Jenkins 2016; National Laboratory of Enteric Pathogens et. al 1991).

Either form, the genomic material of interest must be extracted from whatever cell or capsid it resides in and purified, before it can be used as a template in PCR analysis.

In a cell, when single stranded DNA is exposed and ready to be transcribed, a primer needs to be attached to the strand to act as an indicator on where in the genome the transcription is to be initiated and simply as a primer for nucleotides to further be added to during elongation of the strand. The primer, a short nucleic acid strand, is added by the enzyme RNA primase in the cell. The same is true for DNA amplification through PCR: it is only possible under the aid of primers, a forward and a reversed. This is where synthetic primers, also called oligonucleotides, come in use.

These are synthetic DNA segments, around 15-30 nucleotides in length, designed to complementary attach to those particular parts in the template DNA where the start of transcription is desired (Harris 1998). The primers will run in the opposite direction to the template DNA, and in turn also run in opposite directions to each other. This means one primer will run in the forward direction, from the 3’ end, and one primer also 3’ to 5’ but in the reverse direction to the other primer.

Amplification of two or more target sequences simultaneously is also possible through multiplex PCR by adding multiple sets of primers to the Master Mix along with different probes to detect each sequence through its unique fluorescence signal (Basu 2015).

Attaining PCR product is rarely the problem, it is attaining specific PCR product that is the tricky part. Primer design is a crucial part in PCR (especially in multiplex PCR where several targets are being amplified in a single reaction) as primers need to be specific enough to not attach to unwanted parts in the genome (or even to another primer in the mixture, called a primer dimer) so that random regions in the DNA get amplified. Each primer pair (forward primer and reverse primer) should enable effective amplification of strictly one amplicon. Very short primers are often not specific as they risk to complement not only the target sequence and thus misprime, while long primers mean hybridisation rate will be reduced and by extension, yield will be small (Basu 2015).

When primers have attached to the exposed single stranded DNA, it is time for the hybridisation, that is, the annealing of complementary nucleic acids to the template DNA to extend the forming DNA strand. The hybridisation is catalysed, both in the cell and in a polymerase chain reaction, by an enzyme called a polymerase (Ishino & Ishino 2014).

The polymerase commonly used in PCR, Taq polymerase, is a thermostable DNA polymerase isolated from the extreme thermophile bacteria Thermus aquaticus which, thanks

(13)

7

to its resistance to high temperatures (while DNA denatures during denaturation steps in PCR, Taq polymerase stays intact and active), small size, and absence of nuclease activity, makes it applicable for the field of application (Chien et. al 1976).

Taq polymerase binds to the 3’ end of the primer and synthesises approximately 1 kb each minute (2 kb the first, initial minute) at an optimal elongation temperature of 70-80 °C (Lorenz 2012), through extension of both primers in 3’ to 5’ direction to the template.

Figure 2: Primers (blue and orange) annealed to template DNA (bold), and Taq polymerase elongating the chain in 3' to 5' direction to the template DNA (arrows).

Deoxynucleotide triphosphates (dNTPs) come in the form of dATP, dTTP, dGTP and dCTP, each representing one of the four nitrogenous bases adenine, thymine, guanine and cytosine, and each binding through hydrogen bonding to its corresponding base in the template DNA. These are the building blocks required to build the new DNA string and need to be in equal concentration in the Master Mix (Harris 1998).

For proper function, the polymerase needs magnesium cations, which act as a cofactor to the enzyme as the ions, through interactions with the active site in DNA polymerase and the dNTPs, catalyses the formation of phosphodiester bonds between the dNTP and the primer on the 3’ end. In addition, the magnesium ions stabilise the structure through bindings and decreases the negative charge.

As for the other components in the Master Mix, the concentration of magnesium ions can greatly impact the outcome of the PCR product. An excessive amount of magnesium may lead to non-specific amplicons due to increase in polymerisation activity, but decrease in overall reliability and consistency, while an insufficient amount of magnesium will do the opposite and reduce polymerisation but increase infallibility.

Contribution of magnesium ions to the mixture often come in the form of MgCl2 or MgSO4

and are usually included in the reaction buffer (Slack et. al 2011).

KCl and Tris are often main reagents in the reaction buffer, which key tasks are to contain the pH balance and maintain ion levels. pH value for the buffer at room temperature is 8.2 to 9 but when temperature increases, pH for Tris (and thus the buffer) decreases. When the reaction temperature rises to 72 °C, which is optimal extension temperature for Taq polymerase, pH decreases to 7.0-7.5, which also is the optimal pH for Taq (Harris 1998). This way the buffer maintains optimal reaction conditions throughout the PCR.

(14)

8

As stated by the name, quantitative PCR gives the opportunity for quantitative analysis of DNA, i.e. it is possible to determine how much amplicon has been created in real time throughout the reaction, in comparison to regular PCR, where yield has to be analysed post reaction through, for example, gel electrophoresis.

Measurement of quantity of the amplicon in real time is made possible through fluorescent labelling of the PCR product, either with fluorescent dye or fluorescent probes. Fluorescence emission is measured after each cycle and the signal is proportional to the amount of PCR product generated.

As opposed to dye-based qPCR, where fluorescent dye binds non-specifically to DNA and therefore gives non-specific signals, probe-based qPCR is designed to complementary bind to desirable sequences in the DNA. Just like primer designing is a crucial part of a successful PCR, so is probe designing. Properly designed probes enable accurate and dependable detection, which is the basis for qPCR (Basu 2015).

The probe is tagged with a fluorescent reporter. Its emission gives rise to fluorescence resonance energy transfer (FRET), which is the basis of real-time detection. On the probe, close to the reporter, is a quencher molecule. When the quencher is present, fluorescens emission from the reporter is quenched. Then Taq polymerase extends the DNA chain, it hydrolyses the probe with its exonuclease activity. When the reporter dissociates from the primer and distance itself from the quencher, fluorescence is emitted and recorded by the detector. Every tagged molecule sends out its own signal which means signal generated is in direct proportion to the number of molecules being synthesised – the more molecules, the more intense signal (Wong et. al 2015).

2.3. Internal Positive Control

An internal positive control (IPC) is sometimes included in the PCR mix so as to detect any PCR inhibiting substances in the DNA samples (naturally present or added as a result of contamination in the lab during sample processing and nucleic acid purification), a faulty sampling or otherwise degradation of the DNA, minimising the risk for false-negative results or decreased sensitivity (Schrader et. al 2012).

The control sequence is amplified along with the other target sequences and detected with a different probe to rule out failed amplification in cases where the target sequence could not be detected.

2.4. The Reaction

The reaction consists of three stages, as displayed in Figure 3: an initial denaturation, performed just at the beginning of the reaction, followed by a cycling of three temperatures during the replication step, and ended with a final extension.

(15)

9 Figure 3: A typical PCR program with its cycles and temperatures.

2.4.1. Initial Denaturation

During the initial denaturation, double-stranded DNA denatures under high temperature to form two single strands of DNA. The temperature is set to around 94 °C to 98 °C. For G-C rich template DNA, the temperature is set closer to 98 °C due to a phenomenon called base-stacking, which bases guanine and cytosine contributes to. Base-stacking energy contributes to DNA stabilisation which is why G-C rich DNA will require a higher temperature to denature (Yakovchuk et. al 2006). Initial denaturation temperature is also dependent on the optimal temperature for the DNA polymerase activity.

Formation of secondary structures within the template DNA due to unbroken hydrogen bonds may prevent proper primer annealing and disruption of the polymerase. However, too high melting temperature may also cause the polymerase to denature, resulting in lower product yield.

The initial denaturation, performed once before the cycling begins, is of importance because it ensures that all double stranded DNA is separated into single strands so that primers can anneal to the target sequence. The high temperature also helps with inactivating possible nucleases in the sample, that otherwise risk to degrade DNA and primers (Lorenz 2012).

2.4.2. Replication step

DNA amplification takes place during the cycling process in the replication step. Denaturation, primer annealing and extension/elongation is alternated in thermal cycles, around 40 of them. After denaturation for approximately a minute, temperature is lowered to around 50-65 °C (5 °C below the melting temperature for the primers is optimal), thus entering the annealing step. Now, primers have the time to anneal to their target sequences on the DNA template and extension of the new DNA strand can begin. The temperature is once again increased, this time to around 72 °C, to meet the optimal annealing temperature for the polymerase. During the elongation, the polymerase extends the primer in 5’ to 3’ direction, creating a complementary DNA strand to the template.

After each cycle, the amount of product created is doubled (at least theoretically). Cycling is ended when desired amount of product is reached (Lorenz 2012).

(16)

10 2.4.3. Final extension step

When desired amount of product is reached, the final extension takes place, during which uncompleted amplicon is given extended time to finish. During this time, if Taq polymerase is used, the addition of 3’ adenine residue overhang is added to the amplicons – a useful modification for some subsequent steps when gene cloning (Lorenz 2012).

2.5. Amplification Plot

Figure 4: A typical amplification plot showing eight amplification curves generated from five samples. Two target sequences were amplified, “purple” and “orange”. All samples showed substantial amplification for target “purple” and all of them did at some point cross the threshold value, 0.05. Only in three samples was target “orange” amplified, though not efficient enough for any of the samples to cross the threshold value, 0.2.

From the amplification plot generated by the QuantStudio™ Real-Time PCR Software, each sample will appear as a sperate curve (as long as they give rise to any fluorescence emission) and plotted with the number of cycles in the PCR on the x-axis against the normalised fluorescence intensity on the y-axis. Rn (normalised reporter) is the fluorescence of the reporter dye divided by the fluorescence of a passive reference and ΔRn is given when subtracting the baseline (the initial cycles where little change in fluorescence signal is reported) from Rn. In Figure 4, two thresholds have been set, one to 0.2 and one to 0.05 (orange and purple). The thresholds are specific to the targets analysed. In this particular case, five samples were analysed for two target sequences. All the five samples showed successful amplification of one of the target sequences but only three of those samples also showed amplification of the other target sequence. Although, none of these three samples (orange) had a high enough amplicon concentration to cross the 0.2 threshold.

The higher concentration of the copied target sequence in the beginning of the reaction, the sooner the fluorescence signal will be detected. It will require less cycles of amplification to detect a signal from a sample that contain high concentrations of the target sequence.

(17)

11 2.6. Ct-value

Ct-values are obtained from a table in the PCR software. The Ct-value for a sample equals the

x-value (cycle number) at which the curve crosses the threshold and is a relative measure of the targets’ concentration. If a sample generates a low Ct-value, thus showing substantial

amplification in an early cycle, it means the concertation of the target sequence initially was high, compared to a sample with a higher Ct-value. A sample with a high Ct-value needed more

cycles to amplify the target sequence, hence the target sequence initially existed in a lower concentration.

2.7. Passive Reference Dye, Reporter Dye and Master Mix: Changes on Ct-values

A passive reference dye’s function is to provide an internal fluorescence to act as a reference to the reporter dyes to be normalised against. Normalisation is needed to account for fluctuations in fluorescence due to changes in volume or concentration.

Depending on the Master Mix composition, there is a great chance fluorescence signals will vary, even though there is no change in concentration of probe, reference dye or template DNA. This is because factors such as pH and salt concentrations within the master mixes can vary and thereby affect the fluorescence signals generated. This is displayed in Figure 5, accessed from Applied Biosystems application note “Real-time PCR: Understanding Ct”, which can be

downloaded through their website.

Although, in this project, changing the Master Mix also meant changing the passive reference dye, from ROX to MP, and since the Rn value is defined as the ratio of the reporter dye (FAM/NED/VIC) divided by the reference dye (ROX/MP), this change in reference dye most definitely will have an impact on the Rn values. Having the Rn values changed (and thereby also the Ct-values) does not mean sensitivity of the reaction was compromised.

Figure 5: Difference in fluorescence (VIC and ROX) due to different Master Mixes. The y-axis shows the emission intensity and the x-axis shows the emission wavelength of the fluorophore. Picture from ABI’s “Real-time PCR: Understanding Ct”.

(18)

12

As fluorescence intensity may vary with different Master Mixes, so may the Ct-values, even

though the sensitivity of the Master Mixes is the same (Figure 6). It is important to notice that variations in Ct-values do not reflect the performance of the reaction, however, when

implementing a different method of analysis, it is beneficial to not have the Ct-values change

since patient diagnosis is based on Ct-values and that would mean that algorithms and manual

evaluations will have to be changed, or else diagnosis will be incorrect.

Figure 6: Ct-values (x-intercept) differ between the two Master Mixes even though the target concentration was identical.

Threshold are in both reactions set the same. Ct-value for Master Mix A is 14 and Ct-values for Master Mix B is around

12,5. Picture from ABI’s “Real-time PCR: Understanding Ct”.

In addition, a change in Master Mix and/or passive reference dye might also lead to an increase in standard deviation of the Ct-values of the samples in the PCR assay, which is not favourable

as that will result in lower confidence when distinguishing between target concentrations with small differences.

In the original method with ABI 7300 and TaqMan, the passive internal reference dye was ROX which is supposed to give superb precision on Applied Biosystems qPCR instruments (such as the 7300), according to the manufacturer ThermoFisher. In the new method with QS and Solis, Mustang Purple (MP) was used as passive reference. In each duplex reaction, regardless method, either FAM, VIC or NED dyes were used to detect the two target sequences in each assay.

Table 4: Passive reference dye and reporter dye in use for each method and PCR instrument. Emission maxima for each dye is specified.

Instrument Passive reference dye

Reporter dye type 1 Reporter dye type 2 Reporter dye type 3 Applied Biosystems

7300

ROX (~610 nm) FAM (~520 nm) NED (~575 nm) VIC (~554 nm)

(19)

13

3. Methodology

3.1. DNA Samples

The samples used in the experiments was already extracted and purified DNA obtained from Dynamic Code’s costumers and samples were picked randomly for the experiments where they all throughout remained anonymous (i.e. could never be linked to a person).

Purification of DNA was done with Quick-DNA™ 96 Kit from the brand Zymo Research which is a technique based on a series of lysis buffer wash steps on a silica-based plate and where the finished product finally is eluted through the silica into an elution plate. After eluted, the product is ready to be directly placed into a PCR-mix in a PCR-plate.

3.2. Negative and Positive Controls

Controls are used to supervise the method of analysis to make sure it works properly and to warn against possible errors as well as for follow up of the method over a longer period of time. To better understand where in the process contamination took place, or where possible inhibitors were introduced in the sample, negative and positive controls can be specified into extraction negative and positive and amplification negative and positive. This makes it possible to more or less pinpoint where in the process contamination was introduced and what steps needs to be redone. For example, if the extraction negative control was contaminated, the entire extraction process and the following amplification process has to be redone, but if only the amplification negative control failed, solely the amplification has to be redone.

For this project, only amplification negatives and positives were of relevance since no extraction was done.

With all analyses involved in this project, a negative control along with a positive control unique for each mix (four of them), were amplified together with the rest of the samples. Amplification negative and positive were placed directly to the PCR-plate along with the other samples right before amplification.

3.2.1. Negative Controls

Negative controls verify sample-to-sample contamination, PCR contamination and reagent contamination. The negative control should only contain those reagents necessary for the polymerase chain reaction, but no template. It should be ‘blank’ and should not be showing any kind of amplification in the amplification plot. If there is considerable noise from the negative control sample, the control was at some point in the process contaminated and chances are all or some of the other samples on that plate also were, meaning none of the results can be trusted. In the negative control wells, 3 µl ddH2O was placed instead of the 3 µl template DNA.

Accepted Ct-values for the negative controls are everything >35 or “undetermined” (not enough

amplification to cross the given threshold).

3.2.2. Positive Controls

Positive controls verify that the DNA yield is good enough to draw any conclusions from. If the positive control, which should be of known concentration, is ‘as positive’ as expected, it

(20)

14

affirms that DNA was not degraded by inhibitors thereby risking false-negative or faulty results, and that in all the reaction was carried through completely and correctly.

In the positive control wells, 3 µl purified DNA, extracted with Quick-DNA™ 96 Kit, from specific individuals of Dynamic Code’s staff positive for the given genotypes being analysed, was placed instead of the 3 µl template DNA. At some point in the project, amplification positive related to two of the mixes ran out and a new positive control was created and used. The new positive control was created through DNA pooling of equal amounts of samples previously analysed that showed strong amplification (corresponding to the Ct-values in table

of the target genes).

Accepted Ct-values for the positive controls are the same as what is considered a positive

result for costumer samples (Table 5). Table 5: Accepted Ct-values for the positive controls.

Gene Ct-value DQA1*02 <29 DQA1*05 <27 DQB1*02 <28 DQA1*03 <30 DQB1*0302 <27 DQB1*0302B# <27 IPC <32 3.3. Reaction Components

The total reaction volume was always 15 µl per well. 12 µl of these were the PCR-mix (Master Mix) and 3 µl were the DNA sample (or 3 µl ddH2O if negative control/3 µl positive control

DNA template if positive control).

PCR-mixes consists of double-distilled water (from here on labelled only as ‘water’), HLA-mix (1-4 depending on reaction) and Master Mix, either TaqMan™ Gene Expression Master Mix (original method) or SolisFast® Probe qPCR Mix (Purple) (new method). Volumes vary depending on the Master Mixes stock concentration where one of them (specified further down) is more concentrated and needs to be diluted with more water.

The double distilled water, ddH2O, is deionized water with a resistance better or equal to

18MΩcm and it is obtained from a dispenser called ELGA PureLab Flex and aliquoted into sterile Eppendorf tubes before use.

3.3.1. HLA-mix

What is labelled as “HLA-mix” is the oligonucleotide-mix containing the primers and probes. The oligo-mix was for this optimisation project the most central part of the method optimisation where volumes of the different components (Table 9) were varied and tested.

Each one of the four oligo-mixes contains primers and probes for two target sequences (so called duplex). A total of six unique targets are amplified in pairs in this analysis:

(21)

15 Table 6: Target genes amplified in each of the four HLA-mixes.

HLA-mix 1 DQA1_05 IPC

HLA-mix 2 DQA1_02 DQB1_02

HLA-mix 3 DQA1_03 DQB1_0302

HLA-mix 4 DQB1*0302 B IPC

Mix 4 is a confirmative analysis performed on any samples positive for only HLA-DQ8 (DQA1_03 & DQB1_0302) that have been analysed with mixes 1-3 prior.

3.3.2. PCR-mix: TaqMan™ Gene Expression Master Mix

In the original method used for HLA gene amplification via qPCR within Dynamic Code, TaqMan Master Mix is used and samples are placed in the ABI 7300 instrument.

Standard protocol for the TaqMan PCR mixes is as followed (12 µl per sample, 15 µl in total when template is added):

Table 7: Protocols for PCR-mixes based on the TaqMan Master Mix, original method. Mix 1 (DQA1_05 & IPC)

Stock concentration End concentration Volume TaqMan™ Gene

Expression Master Mix

2X 1X 7,5 µl

HLA-mix 1 25X 1X 0,6 µl

ddH2O 3,9 µl

Mix 2 (DQA1_02 &DQB1_02)

2X 1X 7,5 µl

ddH2O 3,9 µl

Mix 3 (DQA1_03 & DQB1_0302)

2X 1X 7,5 µl

ddH2O 3,9 µl

Mix 4 (DQB1*0302 B & IPC)

2X 1X 7,5 µl

ddH2O 3,9 µl

The polymerase included in TaqMan is called AmpliTaq Gold DNA Polymerase, UP (Ultra Pure). Passive internal reference is ROX dye.

(22)

16

3.3.3. PCR-mix: SolisFast® Probe qPCR Mix (Purple)

When running the reaction with Solis Master Mix, composition of PCR-mix components is slightly changed because of the 5X-concentration (compared to TaqMan’s 2X-concentration), therefore more water is added.

Standard protocol for the Solis PCR mixes is as followed (12 µl per sample, 15 µl in total when template is added):

Table 8: Protocols for PCR-mixes based on the Solis Master Mix, new method. Mix 1 (DQA1_05 & IPC)

Stock concentration End concentration Volume SolisFast® Probe qPCR

Mix (Purple)

5X 1X 3,0 µl

ddH2O 8,4 µl

Mix 2 (DQA1_02 &DQB1_02)

Mix (Purple)

5X 1X 3,0 µl

ddH2O 8,4 µl

Mix 3 (DQA1_03 & DQB1_0302)

Mix (Purple)

5X 1X 3,0 µl

ddH2O 8,4 µl

Mix 4 (DQB1*0302 B & IPC)

Mix (Purple)

5X 1X 3,0 µl

ddH2O 8,4 µl

The polymerase included in Solis is called SolisFAST® DNA Polymerase. Passive internal reference is MP (Mustang Purple).

(23)

17 3.4. Oligonucleotide Mixes

Six different sequences were, together with an internal positive control targeting the human RNA polymerase II gene in reaction 1 and 4, amplified to express and detect HLA-DQ2 and HLA-DQ8 in samples analysed for celiac disease.

Each mix is duplexed and each target has one forward primer (FP1 and FP2), one reversed primer (RP1 and RP2) and one probe (P1 and P2), as listed below. In tables below are concentrations and volumes optimised for the original method, with TaqMan Master Mix run on ABI 7300.

Table 9: Protocols for oligonucleotide mixes (HLA-mixes), original method. Mix 1 (DQA1_05 & IPC)

Oligo Tag Stock

concentration End concentration Volume HLAmix1_FP1 100 µM 300 nM 45 µl HLAmix1_RP1 100 µM 300 nM 45 µl HLAmix1_P1 FAM 100 µM 200 nM 30 µl HLAmix1_mix4_FP2 100 µM 50 nM 7,5 µl HLAmix1_mix4_RP2 100 µM 50 nM 7,5 µl HLAmix1_mix4_P2 NED 100 µM 200 nM 30 µl ddH2O 435 µl

Mix 2 (DQA1_02 & DQB1_02)

Oligo Tag Stock

concentration End concentration Volume HLAmix2_FP1 100 µM 300 nM 45 µl HLAmix2_RP1 100 µM 300 nM 45 µl HLAmix2_P1 VIC 100 µM 200 nM 30 µl HLAmix2_FP2 100 µM 300 nM 45 µl HLAmix2_mix3_RP2 100 µM 300 nM 45 µl HLAmix2_P2 FAM 100 µM 200 nM 30 µl ddH2O 360 µl

Mix 3 (DQA1_03 & DQB1_0302)

Oligo Tag Stock

concentration End concentration Volume HLAmix3_FP1 100 µM 300 nM 45 µl HLAmix3_RP1 100 µM 300 nM 45 µl HLAmix3_P1 VIC 100 µM 200 nM 30 µl HLAmix3_FP2 100 µM 300 nM 45 µl HLAmix2_mix3_RP2 100 µM 300 nM 45 µl HLAmix3_P2 FAM 100 µM 200 nM 30 µl ddH2O 360 µl

(24)

18 Mix 4 (DQB1*0302 B & IPC)

Oligo Tag Stock

concentration End concentration Volume HLAmix1_FP1 100 µM 300 nM 45 µl HLAmix1_RP1 100 µM 300 nM 45 µl HLAmix1_P1 FAM 100 µM 200 nM 30 µl HLAmix1_mix4_FP2 100 µM 50 nM 7,5 µl HLAmix1_mix4_RP2 100 µM 50 nM 7,5 µl HLAmix1_mix4_P2 NED 100 µM 200 nM 30 µl ddH2O 435 µl

The concentrations above are what have been optimised in order to make the HLA analysis run properly with the Solis Master Mix and QS instruments.

3.5. PCR Optimisation and Validation

Although most manufacturers provide solid and robust assays designed to meet all necessary criteria, ThermoFisher included, assays need to be adjusted depending on the selection of instrument, reagents, thermal PCR conditions and primer and probe concentrations.

Although optimisation can be done through adjustment of many different components, it is favourable to maintain standard buffer conditions and instead solely focus on modifying primer and/or probe concentrations, which will affect primer binding kinetics.

Primer design optimisation is sometimes a valid way of approach, for example if primer dimers are a problem, however, in this project, focus was solely on optimisation through modifications in primer/probe concentrations. Primers were already design and working properly.

The approach was a one-factor-at-a-time approach where only one of the components (forward and reverse primer for target one, forward and reverse primer for target two, probe for target one or probe for target two) concentration was varied while remaining components had a fixed concentration. For DQ-targets, the original primer concentration, 300 nM, was tested together with 100 nM and 900 nM. For IPC, 100 nM and 300 nM were tested for primers, and for each probe, 200 nM, 100 nM and 400 nM were tested.

Figure 7: Overview of experiments on primer (blue squares) and probe (grey squares) optimisation during the project. The so called one-factor-at-a-time approach. Probe 1 is connected to target 1 and probe 2 is connected to target 2.

(25)

19

When all mixes were completely optimised, each mix was placed on a full PCR-plate and compared to the original method by placing those same samples on yet another PCR-plate which was analysed according to the original, already optimised method. If results added up with the already validated method, the optimisation was successful.

(26)

20

4. Method

4.1. Optimisation of Primers and Probes

An oligonucleotide mix (oligo-mix) was prepared with the concentrations being evaluated in the particular experiment. When doing this, the original protocol for preparing oligo-mixes (Table 9) was used as a guideline and volumes of the components were adjusted to create the desired concentrations.

The PCR-mixes were prepared, one for each concentration being examined in the particular experiment. For experiments on ABI 7300 according to the original method, the protocols in Table 7 were followed, and for experiments on QS 6 or 7, the protocols in Table 8 were followed. The previously prepared oligo-mix was combined with the Master Mix and water, 12 µl for each sample in total.

12 µl of the PCR-mix was pipetted into each well of a PCR-plate. 3 µl template DNA was then pipetted in each well, adding up to a total of 15 µl reaction solution per well. One amplification positive control and one amplification negative control was added for each different concentration. See appendix A for PCR-plate setup.

Initially, one full PCR-plate containing all four HLA-mixes (24 wells of each mix) was pipetted and analysed according to the original method, and one with the new Solis Master Mix on the QS, i.e. two identical plates but with different Master Mixes and placed in different instruments. The samples and their positions on the plates matched up between the two plates, so as to make a direct comparison of the results possible. This experiment was conducted to get an overview of the differences right away from just running the experience non modified, with a different Master Mix and on a different instrument.

The experiments following this initial experiment was conducted based on the one-factor-at-a-time approach described in chapter 3.6.

4.2. Validation of Modified PCR Method

After optimisation of each primer and each probe in all the mixes was completed, validation experiments were performed, one for each of the four HLA-mixes, where the final concentration for each of the components (primers and probes) was tested on a full PCR-plate. The same samples were placed on two PCR-plates, one according to the original method and one according to the optimised method. This was done for each of the four mixes.

The data obtained from the validation experiments were compared, and the Ct-values obtained

from each method were inserted into an algorithm that determines final HLA diagnosis. The algorithm compared the values from each method, and determined final diagnosis (positive or negative for the DQ-genes, or not proven) so as to see if final results were the same regarding method. If so, the goal of the project was reached – final results were the same disregarding of method.

(27)

21

5. Results

From the initial experiment, where the original method was compared to the new method, but with no modifications to any of the primer or probe concentrations, the differences in Ct-values

between the methods were calculated (Solis QS – TaqMan 7300). The values varied from a median of -1,8 (for DQB1_0203 in mix 3) up to 6,6 (for DQB1_02 in mix 2).

Table 10: Mean and median of the difference between Ct-values from each sample on QS versus 7300 before optimisation.

Mix 1 Mix 2 Mix 3 Mix 4

DQA1_05 IPC DQA1_02 DQB1_02 DQB1_0203 DQA_03 DQB1*0203 B IPC Mean _4,207 _2,866 _-0,239 _6,603 _-1,812 _2,563 _3,913 _3,286 Median _4,119 _3,018 _0,794 _6,563 _-1,897 _2,631 _5,744 _3,635

Also found from this experiment was the fact that 12 of the 88 costumer samples on the plate that was valued as “undetermined” (was not assigned a Ct-value due to too little amplification)

on ABI 7300, was in fact detected on QS, although with high Ct-values. An indication on a

greater sensitivity on the QS compared to ABI 7300. 5.1. Mix 1: DQA1_05

Figure 9: Amplification plot for mix 1 DQA_05 300 nM (red), 100 nM (pink) and 900 nM (purple), and IPC 50 nM (purple). Probe 1 200 nM and probe 2 200 nM.

In general, the shape of the curves is similar with some shifting along the x and y axis between the different concentrations. The 900 nM concentration of DQA1_05 did result in the highest amplitude amongst the three different concentrations, especially higher than 100 nM, but also slightly higher than 300 nM. Although, when studying the different concentrations individually, it seems like DQA1_05 samples with 900 nM concentration did somewhat suppress the IPC signals (purple) the most. Regarding Ct-values, samples appear in pretty much the same range

when it comes to concentrations 300 nM and 900 nM. Ct-values for the 100 nM samples was

(28)

22

The IPC amplification was, in general, far too low when using the original concentration from Table 9 of 50 nM. IPC should be in the Ct-range 23-32 while Ct-values from this experiment

were 33 and above. Not only was the amplification delayed, also the efficiency was poor. No clear exponential phase or even linear phase was achieved. IPC primer concentration had to be increased drastically in order to get more sensible amplification.

For DQA1_05 in Mix 1, 900 nM concentration was at this point considered to be the most optimal and following optimisation experiments on Mix 1 was based on this concentration for DQA1_05.

5.1.1. Mix 1: IPC

Figure 10: Amplification plot for mix 1 IPC 100 nM (green) and 300 nM (blue), and DQA1_05 900 nM. Probe 1 200 nM and probe 2 200 nM.

When IPC primer concentration was increased to 100 nM (green) and 300 nM (blue), signals were stronger and appearance of the curves made more sense. There was a significant difference between the stronger and the weaker concentration, both in amplitude and in amplification efficacy, where samples analysed with the higher concentration were detected with higher signals as well as after fewer cycles.

However, the goal is not to obtain the most efficient amplification of the IPC, since the purpose of the internal control is to indicate whether any inhibitors might be present in the sample. If the IPC concentration is made too high, results of a quantitative amplification might be faulty as the signals could be exaggerated due to the excessive concentration.

Although, at this point, 300 nM IPC was considered giving the best results but would later be reconsidered due to the fact just stated.

(29)

23 5.1.2. Mix 1: Probe 1

Figure 11: Amplification plot for mix 1 probe 1 in 100 nM (yellow), 200 nM (red) and 400 nM (grey), DQA1_05 900 nM, IPC 300 nM, signals are hidden. Probe 2 200 nM.

For probe 1, connected to DQA1_05 in mix 1, signals were stronger the higher the concentration of the probe. Also, Ct-values decreased when concentration increased, i.e. samples were

amplified earlier in the PCR when probe concentration was higher.

400 nM gave objectively speaking slightly better results, but 200 nM was not far off. Therefore, due economic reasons, 200 nM was chosen.

5.1.3. Mix 1: Probe 2

Figure 12: Amplification plot for mix 1 probe 2 in 100 nM (orange), 200 nM (lilac) and 400 nM (blue), DQA1_05 900 nM, IPC 300 nM, probe 1 200 nM.

In this case, regarding probe 2 connected to IPC in mix 1, the higher the concentration the better the result was not the case.

The lowest probe concentration, 100 nM, clearly gave the worst results – amplification started many cycles into the reaction, fluorescence signals were weak and there was a large scattering between samples regarding Ct-values.

(30)

24

Figure 13: Amplification plot for mix 1 probe 2 in 100 nM (orange), 200 nM (lilac) and 400 nM (blue), DQA1_05 900 n, IPC 300 nM, probe 1 200 nM. Only samples 1-8 are previewed for concentrations 200 nM and 400 nM.

If comparing a smaller selection of samples of 200 nM and 400 nM (Figure 13), it becomes clearer as to which one of these concentrations was more successful.

200 nM was considered the optimal concentration for probe 2. 5.2. Mix 2: DQA1_02

Comparing the three variations of concentrations of the forward and reversed primers for DQA1_02 target in mix 2 (see appendix B for amplification plots), 100 nM had by far the worst amplification. While six samples from the 300 nM concentration and five samples from the 900 nM concentration was clearly positive, none of the samples from the 100 nM was successfully amplified; they appear very late into the reaction and amplification curves do not adapt desirable shape.

When comparing 300 nM to 900 nM, there is a better distinction between the two clusters of curves in the lower concentration, 300 nM. When comparing samples individually, one can see that DQB1_02 signals have been supressed by raising the concentration on DQA1_02 primers. In some samples, the DQB1_02 is barely detected at all.

Considering all this, 300 nM was chosen as the most optimal primer concentration for DQA1_02.

5.2.1. Mix 2: DQB1_02

When comparing the three concentrations for the second target in mix 2, DQB1_02, what is most obvious is the spread amongst the signals in the lowest concentration, 100 nM. The non-concurrent signals along with poor fluorescence signal rules out this concentration to be the best.

(31)

25

Figure 14: Amplification plot for mix 2 DQB1_02 in 300 nM (green) and DQB1_02 900 nM (blue), probe 1 and probe 2 both 200 nM.

When looking at 300 nM and 900 nM, displayed together in the figure above, there are no apparent differences. Therefore, due to economic reasons, 300 nM was considered the optimal primer concentration of DQB1_02.

5.2.2. Mix 2: Probe 1

The most apparent difference between the different concentrations on probe 1 connected to DQA1_02 is the spread in Ct-values amongst the samples. Raising the concentration up to 400

nM seems to have resulted in current signals amongst the samples. In addition, the non-specific amplification visible in the far right of the plot (see appendix B) was a bigger issuer with this concentration where many of those non-specific curves crossed the threshold quite early in the reaction.

For 100 nM, the non-specific amplicons did barely cross the threshold, and if they did, it was at a high Ct-value. Results were not much better for 300 nM; therefore 100 nM was chosen for

probe 1.

5.2.3. Mix 2: Probe 2

100 nM concentration was not successful for probe 2. Amplification was just not efficient enough.

400 nM clearly resulted in stronger signals of DQB1_02, but at the same time DQA1_02 signals were supressed. Over all, curves in the plot showing probe 2 concentration 200 nM showed great results and thus, this concentration was chosen as the optimal.

(32)

26 5.3. Mix 3: DQA1_03

Figure 15: Amplification plot for mix 3 DQA1_03 in 100 nM (red), 300 nM (yellow) and 900 nM (blue), and DQB1_0302 300 nM (grey), probe 1 and probe 2 both 200 nM.

Moving on to mix 3, even displaying all three concentrations in the same plot, one can clearly notice the differences between them.

100 nM primer concentration resulted in very poor amplification of DQA1_03 where all samples positive for DQA1_03 was detected late and with weak fluorescence.

300 nM and 900 nM curves adapt the same appearance and amplification seems to be efficient. However, there is a significant difference in at what point samples start amplifying as well as when they intercept with the threshold. Considering these aspects, 900 nM amplification did perform better and decision was at this point made to move forward with this concentration.

5.3.1. Mix 3: DQB1_0302

Figure 16: Amplification plot for mix 3 DQB1_0302 in 100 nM (red), 300 nM (green) and 900 nM (blue), and DQA1_03 900 nM (yellow), probe 1 and probe 2 both 200 nM.

(33)

27

Generally, between all three concentrations, amplification of non-specific sequences was a bigger problem than with previously targets, especially when using 300 nM and 900 nM primer concentration.

Regarding the signals that are obviously DQB1_0302, the target sequence, there is a correlation in primer concentration, amplitude and cycle number. The higher the concentration, the higher the amplitude and the lower the cycle number.

However, there was no greater difference between 300 nM and 900 nM, as both primer concentration performed well and due to economic reasons, 300 nM was therefore considered the optimal concentration for DQB1_0302.

5.3.2. Mix 3: Probe 1

Figure 17: Amplification plot for mix 3 probe 1 in 100 nM (green), 200 nM (yellow) and 400 nM (lilac), DQA1_03 900 nM and DQB1_0302 300 nM, probe 2 200 nM.

The lowest concentration of probe 1 connected to DQA1_03, 100 nM, in mix 3 clearly gave the weakest fluorescence signals and also the highest cycle numbers. In addition, when looking at the targets individually, 100 nM made non-specific amplification signals more wide spread than the higher concentrations. Non-specific amplification was reduced with the higher probe concentrations.

(34)

28

Figure 18: Amplification plot for mix 3 probe 1 in 200 nM (yellow) and 400 nM (lilac), DQA1_03 900 nM and DQB1_0302 300 nM, probe 2 200 nM.

DQA1_03 signals from the two higher concentrations were similar, in fact pretty much overlapping each other, see Figure 18. Due to economic reasons, the lower concentration, 200 nM was chosen.

5.3.3. Mix 3: Probe 2

Results from amplification with all three probe concentrations are similar and all yield strong signals implying an efficient amplification. 100 nM was considered the optimal concentration for probe 2 in mix 3.

5.4. Mix 4: DQB1*0302 B

IPC signals from all three mixes were similar and was therefore not considered when evaluation the results from the primer concentrations of DQB1*0302 B in mix 4.

As in most other previous cases, the lowest primer concentration did perform unsatisfactory in regards to efficiency.

(35)

29

Figure 19: Amplification plot for mix 4 DQB1*0302 B in 300 nM (lilac) and 900 nM (blue), IPC has been hidden in the plot, probe 1 and probe 2 both 200 nM.

300 nM and 900 nM are displayed together in the amplification plot above to show the accordance with each other. With both concentrations, there is a clear distinction between what are non-specific amplification signals and what are DQB1*0302 B signals. The target signals are clustered nicely together and amplification efficiency was good. Even though the sample size is small, at least within these samples 300 nM signals appear at a lower Ct and with even

more narrow spread.

300 nM was chosen as the most optimal concentration for primers connected to this target. 5.4.1. Mix 4: IPC

IPC in mix 4 was initially only tested in 300 nM. This concentration was considered too high for the IPC as curves appeared with Ct-values as low as the Ct-values, and explained in chapter

5.1.1., the goal is not to have IPC signals as high as possible, at as low Ct-values as possible.

In a later experiment, 100 nM was confirmed as the optimal IPC primer concentration, but in the following experiments regarding mix 4, 300 nM IPC is what is being used.

5.4.2. Mix 4: Probe 1

Generally, in this assay signals were strong and amplification efficient. The lowest concentration clearly gave the weakest signals and highest Ct-values. Results obtained from

200 nM and 400 nM were similar in regards to signal strength and especially in regards to Ct

-values where all samples of these concentrations fall within the same range. Due to economic reasons 200 nM was chosen to move forward with.

5.4.3. Mix 4: Probe 2

For the probe connected to the IPC in mix 4, 400 nM showed the worst performance out of the three concentrations. IPC signals were discordant and a real plateau phase was not reached

(36)

30

before the 40 cycles in the reaction, if compared to both 100 nM and 200 nM, where samples gather and plateau around Ct 36.

Considering the two lower concentrations, results are more satisfying in 200 nM samples where amplitude is higher and spacing along the threshold is slightly smaller. Results are not bad with 100 nM probe either, but since 200 nM was chosen for this probe in mix 1, choosing the same concentration in this mix will ease laboratory work and reduce complications when preparing the mixes.

5.5. Validation

5.5.1. Mix 1

When comparing the two targets within mix 1 run with the new optimised method on QS with the old method on 7300, Ct-values from each method align well.

Ct-values of DQA1_05 range from 21,3 to 26,7 on 7300, and between 22,9 and 28,4 on QS.

For IPC the interval is 23,2 to 33,4 on 7300, and 24,8 to 35,6 on QS. 5.5.2. Mix 2

Samples positive for DQA1_02 range from 22,2 to 27,7 on 7300 and from 22,9 to 27,4 on QS. For DQB1_02 the interval is 21,4 to 26,8 on 7300 and 22,6 to 27,7 on QS. Hence, there is a great accordance between the Ct-values obtained from the different methods, for both targets included in mix 2.

5.5.3. Mix 3

Besides the deviating sample at around Ct 36 in the DQA1_03 plots, samples range from 22,8

to 34,7 on 7300 and from 23,5 to 28,6 on QS. If the deviating sample at 30,0 on 7300 is disregarded, the sample range is 22,8 to 28,5. This particular sample is “undetermined” on QS. For DQB1_0302, Ct-values for positive samples fall within the interval 21,0 to 25,3 on 7300

and 21,2 to 26,5 on QS.

5.5.4. Mix 4

Samples positive for DQB1*0302 B in mix 4 fall within the Ct range 19,3 to 25,5 on 7300 and

21,0 to 26,4 on QS. For the IPC, values are 25,2 to 36,2 on 7300 and 22,3 to 32,7 on QS. After optimisation of the primer and probe concentrations along with lowering of some of the thresholds, the differences between Ct-values from each sample obtained from both methods was calculated (Solis QS – TaqMan 7300). The medians of the differences are presented in

(37)

31

Table 11: Medians of the difference between Ct-values from each sample on QS versus 7300 after optimisation and

adjustments to the thresholds compared to before optimisation and with old thresholds.

Mix 1 Mix 2 Mix 3 Mix 4

DQ A1_05 IPC DQ A1_02 DQ B1_02 DQ B1_0203 DQ A1_03 DQB1 *0203 B IPC Median before: 4,119 3,018 0,794 6,563 -1,897 2,631 5,744 3,635 Median after: 0,504 0,404 0,368 -0,453 0,374 0,794 0,566 -3,678

Table 12: Mean values of medians from Table 11, before and after optimisation for all targets and for only DQ-targets. Before optimisation After optimisation

Mean difference DQ-targets 2,992 0,359

Mean difference all targets 3,076 -0,140

5.6. Final Primer and Probe Concentrations Mix 1 (DQA1_05 & IPC)

Oligo End concentration, original method End concentration, new method HLAmix1_FP1 300 nM 300 nM* HLAmix1_RP1 300 nM 300 nM* HLAmix1_P1 200 nM 200 nM HLAmix1_mix4_FP2 50 nM 100 nM* HLAmix1_mix4_RP2 50 nM 100 nM* HLAmix1_mix4_P2 200 nM 200 nM

Mix 2 (DQA1_02 & DQB1_02)

Oligo End concentration End concentration, new method HLAmix2_FP1 300 nM 300 nM HLAmix2_RP1 300 nM 300 nM HLAmix2_P1 200 nM 100 nM HLAmix2_FP2 300 nM 300 nM HLAmix2_mix3_RP2 300 nM 300 nM HLAmix2_P2 200 nM 200 nM

Optimisation and Validation of PCR Method for HLA Gene Expression to Enable PCR System Transfer and Master Mix Change