I need the writer only to paraphrase this file. Please use an academic word and do not change the meaning of the sentence. Any word highlighted with yellow colour, keep it as it is.
The British English should be used in this order.
If there is any thing that is unclear please contact me .
Thanks in advance
General material and methods
A large list of 1,980 patients was provided by the METABRIC cohort (MB) for this study, with validated clinicopathological and transcriptomic data. These annotated primary fresh-frozen breast cancer specimens were collected from tumour banks in the UK and Canada. Primary tumours of female samples in the METABRIC study were used for mRNA extraction (Curtis, Shah, Chin, et al., 2012). Illumina TotalPrep RNA Amplification Kit and Illumina Human HT-12 v3 Expression BeadChips supplied by Ambion, Warrington, UK were used for evaluation of mRNA expression. The status of LVI was assessed firstly for 1565 patients from the METABRIC cohort through histological reviewing via haematoxylin and eosin slides (H&E). Then, LVI status was also assessed through immunohistochemistry (IHC) for Nottingham subset (n= 285) in METABRIC that utilised CD31, CD34 and D2-40 (Mohammed, Martin, Mahmmod, et al., 2011).Clinicopathological profiles for the MB cohort covered 1,980 unselected primary operable invasive BC female patients. The median of the patients’ age at diagnosis was 61.8 years, with a range of 74.36 years. Patients with grade-I tumours represented 9.0% of the whole cohort, whereas patients with grade-II and grade-III tumours accounted for 40.7% and 50.3%, respectively. Regional LN metastasis positivity was observed in 47.5%, where the LN involvement at stage-II and stage-III was 31.5% and 16.0%, respectively. The latter LN staging was based on the NPI staging system: stage-I = no lymph node involvement, stage-II = <3 positive lymph nodes, stage-III = > positive lymph nodes. LVI positives were 635 (out of 1565, i.e. 41%) and LVI negatives were 930 (59%). The BCSS for each patient was calculated in months, up to 300 months, and it started from the date of the primary surgical treatment to the time of death from BC. Last follow-up status revealed that 505 (25.5%) patients died of BC while 1,071 (54.1%) survived the disease. According to histological examination of the surgical specimens of the MB cohort, the histological phenotypes of these patients were as follows: 83.8% as IDC, 7.5% as ILC, and the rest had special types of invasive BC, including mucinous, medullary, and tubular carcinomas.The overall distribution of intrinsic subtypes of BC in the MB cohort was assessed with prediction analysis of microarray of 50 genes (PAM50), a technique based on RT-qPCR. Normal subtype accounted for 10.4% of the sample, with luminal-A, 36.3%, luminal-B, 24.6%, HER2-enriched, 12.1%, and basal-like, 16.6%. Integrating germline variants, somatic aberrations, and aberration with gene expression data augmented BC classification in this cohort by producing 10 IntClusts (1 to 10). These clusters revealed novel heterogeneous gene expression landscapes within molecular subtypes classified according to their expression of hormonal receptors. Clinical characteristics for the MB cohort are available in Table 2.1.
The specimens in this cohort were derived from the well-characterised NPS and consisted of 3,173 unselected primary operable invasive BC female patients. The NPS patients underwent surgery at Nottingham City Hospital between 1987 and 2006. Previously, patients’ biopsies from this cohort were used in the construction of tissue microarray (TMA) formalin-fixed paraffin-embedded (FFPE) blocks. Also, few FFPE full-face tissue sections were available and they were used for the general characterisation of protein expression patterns on different invasive BC tissues. Robust clinicopathological profiles for each patient included histological phenotype, molecular subtypes, primary tumour size, histological grade, tumour stage,nodal status, distant metastasis, LVI, and NPI. The DFS and BCSS were recorded for up to 20 years (median, 175 months) and included information on local and regional recurrences and DMFS. The total follow up time for this study was 10 years. Details of the clinical factors are shown in Table 2.4. A broad panel of IHC data of biomarkers is available for this cohort: ER; PR; androgen receptor (AR), HER2, proliferation marker Ki67; cytokeratins such as CK5/6, CK14 and CK17; EMT related markers including E-Cadherin, P-cadherin and N-Cadherin as previously published [160-163].The number of patients in statistics for some variables, could be different from the number mentioned above owing to the absence of valid data and the lack of tumour specimens in the sectioned tissues.
One of the aims of this study was to identify a set of specific genes whose expression best identify drivers controlling a mechanism underlying LVI positive tumours. Candidate genes were selected based on the following criteria:
ANNs are a form of artificial intelligence inspired by learning in human neuronal systems and have been shown to be capable of modelling complex systems with high predictive accuracies on several large scale datasets (Ball et al., 2002). We have used the METABRIC cohort containing 47,293 transcripts and TCGA cohort containing 20,000 transcripts to develop an ANN model to identify novel genes associated with LVI positive status. Our aim was to identify; using a novel prediction method (ANN), a set of genes that show significant association with LVI positivity (high expression vs. low expression) and to validate the genes using protein expression. To study this, we have classified the patients into five random groups (three for METABRIC and two for TCGA) based on LVI status. Then, the ANN algorithm was run for each group, separately. Each run contains 20 loops. Then, the results was filtered to identify concordant genes with lowest test errors which present in multiple loops for each group. Then, the results of all groups was compared to identify similar genes which present in different groups ( the 100,200,300,400 and 500 genes ) respectively. Top 100 genes was used in this study for gene selection. These 100 genes are in Table 2.5. These data has been bioinfomatically analysed using the ANN analysis in collaboration with Dr Graham Ball from Nottingham Trent University.
Freely and accessible web-based resources were utilised to explore the common pathway of the chosen genes. The list of browsed resources included:
All of the FFPE blocks in this study were accessible through specific requisition applications approved and signed by the NHSB. The “Appendix 1” form in this application includes the research and development approval number that allowed us to use NUH patients’ specimens in this research. The retrieved blocks must be archived and their related records kept strictly confidential.
.2 Tissue sectioning
Once the requested FFPE blocks were obtained and indexed in special labelled boxes, tissue sectioning requests were completed. These forms are for human tissue section requests and release records. The form includes two sections, A and B. The ‘A’ section is to update the Breast Consortium records in the NHS, while the ‘B’ section contains checklists for the release and dispatch of biological materials by the NHSB. This form must include the Research Ethics Committee approval number, the Health Technology Assessment (HTA) study number, issued by the National Institute for Health Research (NIHR), and the NHSB ACP number. This form applies to full-face sectioning and TMA sectioning as well. This request is completed when IHC staining is required.
2.2 Sectioning in the microtome
Before sections were cut from the FFPE blocks, safety procedures and risk assessment guidelines were followed. Preparing sections from the FFPE tissues served both of the routine histological examinations with H&E and IHC. Instruments and tools used for this method included: a rotary microtome (Leica RM2125, Leica microsystems, United Kingdom); a paraffin section mounting bath (Thermo Scientific, MH8516, United Kingdom), filled with deionised water to eliminate distortions of sectioned tissue; and glass slides, with Leica’s X-tra® or Leica’s Apex adhesive slides preferred for use in H&E for IHC applications. Both types of glass slide exhibit positively charged surfaces, allowing maximum adhesion to occur under hydrophilic conditions. A plastic box containing crushed ice, disposable microtome blades (S35 Feather, CellPath, Newtown, Powys, Wales), curved-tip forceps, and xylene-resistant marker pen, or pencil, were also available.
FFPE blocks were chilled by placing them, face up, on ice. Then, the chilled blocks were inverted, face down, on ice for five minutes maximum. A block was carefully positioned in a microtome chuck; water droplets were removed prior to this step. The movement of the block’s face to the blade edge in both axes needed to be parallel; if required, the microtome chuck was unclamped and the block orientation was adjusted. Sectioning thickness was set routinely at 4μm. The microtome handle was turned clockwise until the sections began to be cut; the angle of the block face was adjusted if necessary. The block face was trimmed until a ‘full-face’ section was produced appropriately. For the TMA blocks, extra sections were cut for optimisation purposes. When the ribbon of full-face sections was produced, forceps were used to grasp its leading edge. Then, it was laid down carefully on the water surface of the water bath, with both ends of the ribbon supported at all times. As the sections expanded, the individual sections were separated with forceps. Single sections were mounted onto microscopic slides by submerging the slides into the water, beneath the tissue, at an angle of approximately 45°. Sections were realigned over the slide when necessary. Mounted slides were placed vertically in a drying rack, then all of the prepared slides were collected and dried overnight in an oven at 37°C. All slides were transferred to a labelled storage box and stored in a cold room at 4°C until needed.
Preparation of donor blocks and TMA construction
Archival FFPE blocks of invasive BC tissue were retrieved from the NHSB. H&E stained sections from these blocks were taken and assessed for sufficient invasive malignant cells for sampling. Representative tumour areas on each slide were marked alongside their corresponding spots on the FFPE blocks; only well-trained pathologists performed the latter step. Marked blocks were loaded onto the TMA Grand Master® (3D HISTECH®, Budapest, Hungary). To make a recipient block for the biological samples, molten paraffin wax (55–58°C) was poured into a mould, onto which a histopathological tissue cassette was mounted. Ice was used to solidify the paraffin and the mould was removed. For TMA construction, the blocks were trimmed. The TMA blocks were designed to hold a maximum of 150 cases. Marked areas of the tumours from the donor blocks were harvested using a 0.6mm needle. Maps were prepared for the TMA blocks, showing medical identifiers for each sample and on each well.
Determination of binding specificity for selected antibodies with western blotting
Western blotting (WB) was performed to assess the binding specificity of each purchased antibody to its corresponding antigenic sequence. The qualitative detection of a single protein, or an epitope, in various BC cell lysates, or lysates from other cell lines, was our tool to describe the binding specificity for each used antibody in this study. Occasionally, the detection of multiple bands in WB infers the presence of non-specific binding or a specific binding with a modified protein with two or more recognisable epitopes by the same primary antibody. The term ‘blotting’ denotes the transferring of the isolated protein from a separation gel medium onto a surface of non-reactive membrane. This technique was first introduced in transferring proteins from polyacrylamide gels to nitrocellulose membranes, as detailed in .
Cell line lysates
WB was performed on cell lysates of different human cell lines (MCF7, MDA-MB-231, SKBR3, MDA-MB-468 and HeLa) to validate the specificity of the antibodies used in IHC. The specificity were reflected as a specific band.
Growing of cells and preparation of cell lysates
MCF-7, MDA-MB-231 and BT474 were grown using RPMI media (Sigma, Catalogue code…..) supplied with 10% Foetal Bovine Serum (FBS) (Sigma Catalogue code…..). SKBR3 cell line was grown in McCoy’s media (Sigma Catalogue code…..) which was also supplied with 10% FBS..
Cells were grown to about 80-90% confluency. After removing growth media and washing with PBS, the cells trypsinised by adding 3mL 10% trypsin(Sigma Catalogue code…..) and incubated at 37°C for less than 5 minutes to dissociate cells. Detached cells were neutralised by adding 7 ml fresh media and centrifuged at 1,000 rpm for 5 minutes. The pellet was then re-suspended in 5-10 ml of media. Fresh media were then used to suspend the cells at a concentration of 1×106 cells per ml and centrifuged for 5 minutes then placed on ice. For each 106 cell pellet (on ice) 1ml of RIPA buffer(25mM Tris, 150mM NaCl, 1% Nonidet P-40, 1% sodium deoxycholate and 0.1% Sodium Dodecyl Sulfate (SDS)) (Thermo Fisher Scientific Catalogue code…..) and protease and phosphatase inhibitors cocktail (Catalogue code …Pierce, Thermo Fisher Scientific, UK) were added and the pellet re-suspended. After wards, the cell lysis was incubated for 15 minutes on ice with frequent gentle shaking and then centrifuged at 13,000 rpm for 10 minutes, in order to remove cell debris. The supernatant was removed and stored either in -20°C or – 80°C for long term storage.
Cell lysate protein quantification (Pierce™ BCA protein assay)
The Pierce™ BCA Protein Assay Kit (catalogue code: NP0321BOX, Invitrogen, Thermo Fisher Scientific, UK)was used to quantify proteins in the cell lysate using the standard curve method of pre-determined protein concentrations of BSA (bovine serum albumin). The dilutions of BSA were prepared in nuclease-free water to obtain the following concentrations: 2000 µg/ml, 1500 µg/ml, 1000 µg/ml, 750 µg/ml, 500 µg/ml, 250 µg/ml, 125 µg/ml, 25 µg/ml and 0 which is the blank. 8μl of diluted protein samples (diluted 1:5 in nuclease free water), as well as the standards, were pipetted in triplicates in a 96-well plate. They were then mixed with 200 µl of the working solution of BCA reagent and incubated at 37°C for 30 minutes before reading on a FLUOstar OPTIMA, UK/ Infinite® F50 (UK) microplate reader (catalogue code: NP0321BOX, Invitrogen, Thermo Fisher Scientific, UK) at a wavelength of 595 nm.
Preparation of western blotting reagents and tools
Prior to sample preparation, all required reagents and experimental tools were prepared and ready to use. The XCellSureLockTM Mini-Cell and XCell IITM Blot Module (catalogue code: EI0002, Invitrogen, Thermo Fisher Scientific, UK) were assembled to use in protein isolation and protein transferring. A NuPAGETM 4–12% Bis-Tris Protein Gels, ten well(catalogue code: NP0321BOX, Invitrogen, Thermo Fisher Scientific, UK) was purchased and used as pre-casted polyacrylamide gelsfor denatured and reduced protein separations.
The working solution of the running buffer was freshly prepared by adding 50ml of NuPAGETM MOPS SDS Running Buffer 20X (catalogue code: NP000102, Invitrogen, Thermo Fisher Scientific, UK) to 950ml of distilled water. A standard preparation of the transfer buffer for one gel was then prepared by mixing 50ml of NuPAGETM Transfer Buffer20X (catalogue code: NP0006, Invitrogen, Thermo Fisher Scientific, UK), 100ml of 10% methanol, 849ml of distilled water, and the addition of 1ml from NuPAGETM Antioxidant (catalogue code: NP0005, Invitrogen, Thermo Fisher Scientific, UK).If two gels were used for protein separation and blotting, methanol concentration was adjusted up to 20% and the distilled water volume was reduced to 749ml. The remaining volumes of the other solution stocks were the same. The wash solution was prepared by dissolving 1ml of Tween-20, non-ionic detergent, in 1,000ml of phosphate-buffered saline (PBS). The blocking solution was prepared by dissolving 2.5gm of Marvel’s dried and fat-free milk at 5%, or in 50mL of washing solution. All primary antibodies were prepared upon use and according to their previously optimised conditions.
IRDye® 680RD Donkey (polyclonal) anti-Rabbit IgG(product number: 925-68073, LI-COR Biotechnology, St. John’s Innovation Centre, Cowley Road, Cambridge, UK) and IRDye® 800CW Donkey (polyclonal) anti-Mouse IgG (product number: 925-32212, LI-COR Biotechnology, St. John’s Innovation Centre, Cowley Road, Cambridge, UK) were used, as appropriate, to detect the primary antibodies of the studied proteins, whether the raising host was mouse or rabbit. The ratio of 1:15000 was applied to dilute fluorescent secondary antibodies in the blocking solution. Primary anti-β-actin antibody produced in (product details) and primary anti-β-actin antibody produced in mouse (product details) were also used, as β-actin protein is endogenously expressed in all lysates, and considered as the sample’s normal control. The fluorescent secondary antibodies were prepared at the dilution ratio of 1:5000 in the blocking solution, as explained above.
Each cell line lysates, the commercially purchased protein standards, MagicMarkTM XP western protein standard (catalogue code: LC5602, Invitrogen, Thermo Fisher Scientific, UK) and NovexTM sharp pre-stained protein standard (catalogue code: LC5800, Invitrogen, Thermo Fisher Scientific, UK), were thawed in a bucket of crushed ice.Then, based on protein quantification of each lysate, the detected amount of the lysate was mixed with its corresponding amount of NuPAGETM LDS sample buffer 4x (catalogue code: NP0008, Invitrogen, Thermo Fisher Scientific, UK) and NuPAGETM sample reducing agent 10x (catalogue code: NP0004, Invitrogen, Thermo Fisher Scientific, UK. All tubes were spun gently using a laboratory vortex mixer and the protein mixture was digested at 100°C on a heating dry block.
Loading samples and separation of proteins mixture
A ten-well precast NuPAGEBis-Tris 4-12% protein gel cassette (catalogue code: NP0321BOX, Invitrogen, Thermo Fisher Scientific, UK) was inserted inside a running tank, the XCellSureLockTM Mini-Cell (catalogue code: EI0002, Invitrogen, Thermo Fisher Scientific,UK). The previously prepared running buffer was poured slowly into the running tank until the maximum fill line was reached by the liquid. No leakage was confirmed. NuPAGE antioxidant reagent (catalogue code: NP0005, Invitrogen, Thermo Fisher Scientific, UK) was added to the running buffer at 1ml to protect the protein from possible oxidation.
Starting from the first well on either left- or right-hand sides, the prepared samples as describe was loaded onto the wells. If the number of samples was less than the number of wells, the empty wells were filled with the loading buffer alone at the same volume. The running tank lid was positioned and the electrodes’ cables and plugs were fastened securely. Electrical power supply (Invitrogen, Life Technologies, UK) was connected and set to run for 90 minutes at 150 volts.
Transferring protein bands onto nitrocellulose membrane
The blotting module was assembled in the last 10 minutes of the running time of the protein separation step. The precast gel cassette was opened using a specific metal opener and the gel was removed with extreme caution and merged into a weighting boat filled with running buffer. Sponges and filter papers were cut to the exact size of the protein separation gel and were soaked thoroughly in the transfer buffer, as explained previously. The blotting module was opened and, from the cathode chamber, the gel was layered with blotting membrane, soaked sponges, and the cut filter papers in the following order: 2 x sponges, 1 x filter paper, gel, 1 x nitrocellulose membrane, 1 x filter paper, 2 x sponges. Each layer was flattened with a cylindrical roller to remove any air bubbles that may distort the transfer of protein bands. The blotting module was merged firmly into the previously prepared transfer buffer contained in the running tank, cables and plugs were connected, and the power supply was turned on for 60 minutes at 30 volts.
The blotting module was dismantled and the membrane was laid carefully, facing upward, in a weighting boat filled with proper blocking solution, 5% of Marvel Milk, and left for 60 minutes at room temperature on a rocker. Then, the membrane was rolled carefully, facing inward, and inserted into a skirted 50ml Falcon tube filled with 5ml of a primary antibody solution for the selected biomarkers at the previously optimised dilution. The membrane was incubated overnight in the cold room. After it had equilibrated to room temperature, the excess blocking solution was poured out carefully and the membrane was unrolled in a weighting boat filled with washing solution without touching the blotting face. The membrane was washed three time for five minutes with rocking. Then, the secondary antibodies were prepared at 1:15000 by diluting 1μl in 15ml of blocking solution. The membrane was incubated for 60 minutes with continuous rocking. The membrane was kept covered to avoid direct exposure to light at this stage. Finally, the membrane was washed three times for five minutes in the washing buffer.
Visualisation of protein bands
Image capturing of the immunodetected protein bands was performed at two, near-infrared, different emission wavelengths. The infrared dye that was used to label the IRDye® 680RD secondary antibody has an emission wavelength at 694nm whereas the IRDye® 800CW secondary antibody has an emission wavelength at 795nm.The Odyssey Fc Imager (model number: 2800, LI-COR Biotechnology, St. John’s Innovation Centre, Cowley Road, Cambridge, UK) was used to capture digital images of the membranes showing red and green fluorescent bands. If the molecular weight of the β-actin overlapped with the targeted antigen of the studied biomarker, a multiplexed image of both anti-rabbit and anti-mouse secondary antibodies was generated digitally. Yellow bands indicated the overlapping proteins with molecular weight around 42 kilo Daltons (kDa). All images were labelled and stored on a personal hard drive.
IHC is an indispensable practice in the clinical management of BC patients. From molecular subtyping to the detection of validated prognostic biomarkers, IHC plays a critical role in enriching the reliability of pathology reports in many neoplastic diseases, including invasive BC. The first IHC experiment was reported in the 1940s . Fluorescein isothiocyanate-labelled antibodies (FITC-labelled Abs) was the first material that allowed researchers to visualise pneumococcal antigens with a fluorescent dye. Continuous efforts were dedicated to developing IHC methods over the next few years, leading to the substitution of fluorophores with enzymes in labelling Abs, for example, peroxidase and alkaline phosphatase.
The IHC procedure is based on a reversible reaction between an antibody (Ab) and an antigen (Ag). Utilising this biochemical bonding between Ab and Ag in the clinical practice of IHC improves patients’ clinical profiles and prognosis accuracy by combining the observed histological features of a tumour with its protein expression patterns. In BC pathology, the immunostaining of selected myoepithelial proteins such as smooth muscle actin (SMA), tumour protein 63 (TP63), and smooth muscle myosin protein (MYH11) is to differentiate between invasive and benign, or normal, proliferative compartments of the breast , and to identify adenoid cystic carcinoma if the anti-calponin antibody exhibits negative staining . Also, IHC staining of adhesion molecules such as epithelial cadherin (E-cadherin) added high relativity to this technique in revealing discohesive and highly metastatic phenotypes of ILC . Finally, ER, PR, HER2, Ki67, tumour protein 53 (p53), and the apoptosis regulator Bcl-2 were all detected by IHC  and helped in the individualisation of therapeutic regimens for each BC patient, especially before and after neoadjuvant chemotherapy.
Immunohistochemistry tools, chemicals, and procedures
In this study, IHC was used to semi-quantify the levels of protein expression on invasive BC TMA sections, 4μm in thickness. In addition, full-face sections, a random number of samples, from 10 to 15, were IHC stained to investigate the heterogeneity of the expressed target; therefore, the use of TMA sections needed to be validated subsequently. All cut tissues were obtained from invasive or invasive-mixed FFPE specimens. The IHC procedure was performed by using NovolinkTM Max Polymer Detection System (product code: RE7280-K, Leica, Biosystems, UK), which is suitable for antigen detection by using mouse immunoglobulin G (IgG) and rabbit IgG.
Labelling of slides
All tissue sections were loaded onto Leica’s X-tra® slides. Then, the slides were labelled specifically with a marker pen for IHC applications. Date, symbol of targeted protein, user name, dilution factor of applied antibody, and specimen’s identification number were the basic labelling data for each slide.
Deparaffinisation and rehydration of tissue sections
Starting with dewaxing the tissues, all TMA sections were placed on a hot plate (Leica HI1220 flattening table, product code: 14042321474, Leica Biosystems, UK), face up, at 60˚C for 10 minutes. Deparaffinisation continued with sorting the slides into the Leica’s slide rack (product code: 14047533750, Leica Biosystems, UK), then the rack holding the slides was placed into a reagent vessel filled with fresh xylene (product code: X/0250/PB17, Fisher Scientific, UK) and contained inside the chamber of the Leica AXL. The dewaxing programme, previously introduced into the machine, was activated and this programme kept the slides soaked in xylene for five minutes, followed by automated transfer of the holding rack into another fresh xylene vessel, where it was incubated for another five minutes. Three baths of IMS were applied to all slides, as the first two baths were in 100% IMS for two minutes and the third bath was in 70% IMS. The rack was then rehydrated in a vessel filled with flowing water for five minutes.
Heat-induced epitope retrieval: Citrate buffer, pH 6.00
Formalin or aldehyde fixation and paraffin-embedding procedures may lead to immense physical and chemical modifications in any surgical biopsies. Transposition or alterations of the antigenic site are a few effects of this method on FFPE tissues. Poor exposure to epitope may deteriorate the positivity of IHC staining. The optimum condition to re-expose the antigenic sites of the FFPE tissues was selected in this study. Various heat-induced epitope retrieval (HIER) methods were tested under different conditions. The commonly used protocol of the citrate buffer antigen retrieval method was designated as the optimum antigen retrieval method.
Preparation of sodium citrate buffer, pH 6.00
The sodium citrate buffer was prepared by the complete dissolving of 2.94gm of Tris-sodium citrate dihydrate in 1,000ml of distilled water. A few drops of 1 Normal of HCl were added to adjust the pH to exactly 6.00 using a daily calibrated laboratory pH meter. The solution was stored in a fridge at 4˚C. A plastic Leica AXL vessel was filled with 500ml of citrate buffer, pH 6.00 solution; the rack from step 126.96.36.199.was immersed in it, and a plastic lid with no metal handles was placed on the top of the vessel. The slides, settled in the citrate (pH 6.00) vessel, were placed inside a laboratory microwave; the microwave’s door was locked and it was operated at maximum power for 20 minutes. Protective glasses were worn, the vessel was removed from the turned-off microwave with a heat-resistant pair of gloves, the lid was removed in the sink, not on the bench, and flowing water was applied with a rubber hose into the vessel, displacing the citrate buffer and cooling down the slides for five minutes.
Blocking of endogenous peroxidases
Endogenous peroxidase activity is a biochemical feature tissues, which is still chemically active even if the tissue is fixed in formalin and embedded in paraffin wax. Applying a horseradish peroxide (HRP) conjugated antibody may interact with peroxidases presented in the tissues, leading to high and non-specific staining noise. The need for the endogenous peroxidases blocking step could be tested by incubating the studied tissues with DAB substrate; specific details are given in step 188.8.131.52. If the tissues turned a brown colour, the blocking of endogenous peroxidases was obligatory. Calibrated micropipettes, with a volume range of 1-1,000μl, were available. 300μl of Novolink’s peroxidase blocking reagent was applied to each slide, containing 3–4% (v/v) hydrogen peroxide (H2O2)(product number: RE7101). All slides were left for five minutes to react, then they were washed thrice with TBS at intervals of five minutes.
Blocking non-specific reacting proteins
Non-specific staining could also occur due to the presence of binding sites, peptide sequences that mimic antigenic sites recognisable by the applied antibody. The non-specific protein–protein bindings between the applied antibody and the tissue surface may take different forms such as charge-based and hydrophilic/hydrophobic attractions. Using a new micropipette plastic tip, 300μl of Novolink’s protein block reagent, 0.4% casein in PBS, with stabilisers, surfactant, and 0.2% Bronidox L as a broad spectrum anti-bacterial and anti-fungal preservative (product number: RE7102) were applied and the tissue slides were incubated for five minutes, then washed with TBS as in step 184.108.40.206.
Applying the optimised primary antibody
Several primary Abs were purchased and optimised for this study.All the antibodies were optimised on TMA before the staining on the whole series, by changing different variables in the protocol of the staining until the optimum result was obtained; such as no/less background staining, and high degree of expression heterogeneity, starting with the dilution recommended by the supplier’s datasheet in addition to three or more dilutions above and below the recommended dilution.In the forthcoming chapters, the optimisation conditions, positive and negative controls, binding specificity, and dilution factor for each primary Ab will be discussed in detail. However, BondTM primary Ab diluent (catalogue number: AR9352, Leica Biosystems, UK) was used in this study as our universal diluent for all primary Abs. Using a fixed volume micropipette at 300μl, one load of 300μl of the optimally diluted primary antibody was applied to the corresponding tissues only A few slides were nominated for this study as blank slides to test the user’s investigational quality; negative control slides were used to test the binding specificity of the post primary reagent, or secondary antibody, with the fragment crystallisable region (Fc region) of the primary Ab; the application of primary Abs for all experiments was skipped intentionally on those slides.In addition, specific positive controls, as advised by the antibody manufacturer or choosing a specific tissue from the human protein atlas (http://www.proteinatlas.org/) to observe the pattern and intensity of the protein expressions on TMA were included in the IHC experiments. However, all negative controls used in this study were applied on BC tissue. The incubation time for the primary antibodies was either 60 minutes on the bench or left overnight in a cold room at 4˚C. Washing cycles with TBS were applied as in previous steps after incubation periods.
Applying post-primary antibody
Post-primary antibody was applied at 300μl to all TBS-washed slides from the previous step. This contains rabbit anti-mouse IgG <10μg/ml in 10% (v/v) animal serum in TBS (0.09%), with ProClin™ 950 as the preservative (product number: RE7111). The slides were left for a further incubation time of 30 minutes, then washed thrice in TBS for five minutes.
Applying NovolinkTM polymer
The Novolink™ polymer binds to rabbit IgGs, and it also detects the post-primary and any rabbit-hosted primary antibodies. The chemical composition of this reagent is anti-rabbit Poly-HRP-IgG <25μg/ml containing 10% (v/v) animal serum in TBS (0.09 %), with ProClin™ 950 as a preservative (product number: RE7112).300μl of this reagent was loaded onto each slide and left for 30 minutes, followed by three washing cycles with TBS for five minutes.
Blotting with 3,3’-diaminobenzidine (DAB)
DAB peroxidase substrate solution is prepared by mixing DAB chromogen, containing 1.74% (w/v) 3,3’-diaminobenzidine, in a stabiliser solution (product number: RE7105) with Novolink™ DAB substrate buffer polymer, containing ≤0.1% H2O2 in a buffered solution and preservative (product number: RE7143). The mixing ratio was 1:20, and it was prepared freshly each time it was needed. The mixed solution was always kept in the dark until use. The application of this solution was at 300μl for five minutes. Excessive chromogen was washed off with TBS thrice for five minutes.
Counterstaining with haematoxylin and tissue dehydration
Although the counterstaining step might seem optional, it provides clear and sharp nucleus staining images that facilitate semi-quantification of the IHC-stained protein. 300μl of haematoxylin reagent, containing <0.1% haematoxylin (product number: RE7107), was applied to each section and left for six minutes to react. The slide coverplateswere disassembled inside the initial water tank and excessive haematoxylin staining was removed. The slides were re-assorted into the Leica AXL rack, which was put into its plastic vessel filled with distilled water. The vessel was then placed inside the Leica AXL system and the dehydration programme was activated. The dehydration step included washing the slides in three baths of ethanol (EtOH), for two minutes each, followed by two washes in xylene for five minutes each. The stained tissues were then mounted in DPX medium and sealed with a transparent cover glass, or coverslip. DPX is a synthetic resin mountant mixture of distyrene (a polystyrene), a plasticiser (tricresyl phosphate), and xylene.
Imaging of tissue sections and histological scoring (H-score) system
Whole slide imaging
The quality of IHC staining intensity is not permanent. Staining details can fade due to long periods of storage, direct exposure to light or, simply, the slide itself can be damaged for various reasons. Digitalising histological tissue sections is not only a fancy terminology accompanying modern clinical practices; in fact, it has helped pathologists to keep their patients’ highly valued FFPE specimens for a longer time, in a smaller space, with convenient and private sharing and editing options for patients’ clinical data.
In this study, all IHC-stained tissue sections were sent to an electron microscopy and imaging unit, at the Cellular Pathology Department, Queen’s Medical Centre (QMC), A Floor, West Block, NUH. A specific request form for slides scanning and imaging was requested and filled in appropriately. It includes the total number of slides intended for scanning, name of stained protein, and slide labelling data. A non-specimen request form was completed for billing purposes. High-resolution digital images of the whole slides were generated with the Nanozoomer Digital Pathology scanner (NDP C9600series; Hamamatsu Photonics, Welwyn Garden City, Hertfordshire, UK). The scanning at 20-times magnification power was applied on all slides, and NDP.view2 software (Hamamatsu Photonics, Systems Division) was used to facilitate the storage of the TMA cores using a high-resolution screen (24 inches, 1920×1080 pixels). All scanned images of the studied biomarkers were duplicated and one copy of each TMA set was kept in a labelled folder in the Breast Cancer Research Group (BCRG) local drive.
The selection of the scoring methods used to statistically interpret IHC staining results requires credibility and reproducibility. Hence, the semi-quantitative scoring method in IHC also requires expert and well-trained observers who score the TMA cores with high degrees of scoring concordance. H-score is a common method that was applied to the IHC-stained tissue sections in this study. The H-score is generated by setting four levels of scoring intensities, negative=0, weak=1, moderate=2, and strong=3, by examining various and randomly selected TMA cores’ scanned images from different slides. The sum of individual intensity multiplied by the total percentage of cells expressing it is the mathematical definition of an H-score. An H-score is calculated according to this formula: [0 × (% cells expressing intensity 0)] + [1 × (% cells expressing intensity 1)] + [2 × (% expressing intensity 2)] + [3 × (% expressing intensity 3)].
First, a pathologist besides the main researcher were asked to be the second scorers for at least 25% of the whole cohort under investigation. The selection and scoring of the TMA slides that represented the previous percentage was done blindly and individually. The scoring results were concealed until the main researcher completed the H-scoring for all slides. Intra-class correlation among the scores was calculated and the threshold of H-score agreement was >75%. Any slide with a lower scoring concordance was re-scored and the discrepancies were discussed. Heterogeneous protein expression may occur and all of the staining intensities could be presented in one TMA core; these biomarkers with this pattern of expression are not suitable for TMA scoring in this scenario. For this reason, a panel of a small number of full-face tissue sections was IHC stained to assess the applicability of TMA.
Several statistical parameters were accredited in this study. Nominal variables, including categorical and ordinal variables, continuous scale data, and independent or dependent variables exemplify this variety. Each statistical test requires specific conditions and variable type(s); therefore, multiple statistical analyses were conducted. All statistical tests were performed with the Statistical Package for the Social Sciences (SPSS).
The description of the available data in this study was essential to identify specific features of a particular dataset. For example, scale H-score data for a population, mean, median, and mode were investigated and depicted into histogram charts illustrating the normal distribution curve. These three statistics characterise the central tendency statistics for datasets of interest. Cut-off points for continuous variables were decided according to their normal distribution status. If the data was normally distributed, the cut-off value was determine around the mean. Otherwise, median was designated as a proper cut-off point. Other figures, such as percentage and total number of cases, were also employed in this study using descriptive analyses.
Identification of Cut-Off point
One method to determine an appropriate dichotomisation for biomarkers, the cut-off point between populations, was dependent on the distribution of the protein expression using IHC data where either the mean or median were chosen based on normally or not normally distributed data respectively. Additionally, frequency histograms were used for visualisation of the distribution and for detection of apparent cut-off points. Furthermore, cut-off points from the published literature were considered. An alternative method for dichotomisation was derived using X-tile software (version 3.6.1, 2003-2005, Yale University, USA). X-tile randomly splits the data into two groups and the optimal cut-off point, based on prediction of BCSS is determined in a training group and validated in the second group. Information regarding the stained antibodies in this study, their expression and their cut-offs are listed in the next chapter.
Intra-class correlation coefficient (ICC)
As part of the quality control and H-score assessment procedures, the continuous H-score data for an immune-stained protein generated by two individual observers were subjected to the ICC test. This test quantitatively measures the resemblance between two datasets and assigns a degree of scoring reliability between the two scorers. The ICC agreement test was applied to all stained protein, and for at least 25% of all valid TMA cores.
Chi-square test (goodness to fit)
The chi-square test was one of the non-parametric statistical tests that was used in this study to determine the statistical relationships between two, or more, nominal independent variables. A chi-square test could expect that the difference of rates distribution for a selected and categorised biomarker of interest against valid equivalent data belongs to the other nominal variable(s), LVI, tumour grade, LN stage, or other previously stained biomarkers, mathematically. Output data was presented in cross-tabulation matrices of rows and columns. The threshold of significance (p-value) was set at <0.05.
Mann-Whitney U test (median analysis)
As with the chi-square test, the Mann-Whitney-Wilcoxon (MWW) test is another non-parametric test that was used for univariate analysis with clinicopathological variable and selected biomarkers. This test allowed us to evaluate the statistical associations between the nominal data and the distributed median of the continuous data. The postulation of normalised scaling data for this test is not required, in contrast with the student’s t-test. The p-value was set at <0.05.
Student’s t-test (mean analysis)
The student’s t-test was used prior to this study to assess the statistical difference between the means of two normally distributed variables. The p-value was set at <0.05.
One way ANOVA test was used to find out which of different BC classes ( by IHC or cell lines) were significantly different from each other (post hoc test; Tukey).
Pearson rank test (model/line of best fit) and Spearman rank
The Pearson correlation coefficient (r) was used in this study to determine the strength of the linear association between two normalised transcripts of continuous variables. While the spear man rank was used for the correlation in non-normalised data. The value of r takes a range of values from +1 to -1. The value of 0 indicates that there is no association between the two interrogated variables; the closest value to +1 indicates a strong co-expression in the same direction, whereas the negative values suggest an inverse correlation of co-expression. The correlations were considered to be significant if the p-value was <0.05. The correlations were considered to be strong, in either direction, if the values of r were between (-1.0 to -0.5) or (1.0 to 0.5), moderate if the values of r were between (-0.5 to -0.3) or (0.3 to 0.5), weak if the values of r were between (-0.3 to -0.1) or (0.1 to 0.3), and non-existent if the values of r were between (-0.1 to 0.1).
Kaplan-Meier survival analysis and log-rank test
The log-rank test was applied to estimate the survival distribution for two populations of BC patients statistically, when survival observations were censored. The associated estimations of BC patients’ survival were plotted onto Kaplan-Meier curves. The survival status for a patient who had died from BC was numerated as (1) in SPSS to indicate this event. The p-value was set at <0.05.
Proportional hazards model (Cox regression model)
The prognostic independence of an investigated differentially expressed gene or protein was assessed against standardised prognostic factors in Cox’s proportional hazard model. Tumour grade, LN stage, and tumour size are examples of the used predictive factors. The p-value was set at <0.05.
Try it now!
How it works?
Follow these simple steps to get your paper done
Place your order
Fill in the order form and provide all details of your assignment.
Proceed with the payment
Choose the payment system that suits you most.
Receive the final file
Once your paper is ready, we will email it to you.