Regular paper

Use of dissected paraffin block tissue as a source of mRNA for transcriptional profiling and biomarker identification: a review, with preliminary findings in adrenocortical carcinoma tissue

Urszula Waszut1✉ and Norman F. Taylor2

1Department of Nuclear Medicine, Medical University of Gdansk, Gdańsk, Poland; 2Department of Clinical Biochemistry, King’s College Hospital, London, SE5 9RS UK

Cancer tissue obtained by biopsy or at surgery is usually handled and stored as formalin-fixed, paraffin-embedded (FFPE) blocks. While primarily prepared for histological examination, this material represents a valuable resource that is stable and transportable, and which is ideal for examination in 3 dimensions. It has been used for DNA and protein analysis but the potential for transcriptional profiling has been little exploited: the quality of the extracted RNA is uncertain and there are few reports of its efficient use. We review the background and practicalities of this approach and describe a pilot study as a worked example in which RNA of good yield and quality could be consistently obtained from differentiated areas of adrenocortical carcinoma (ACC) tissue.

Keywords: FFPE, adrenocortical carcinoma, mRNA preparation, transcription profiling

Received: 09 February, 2021; revised: 23 January, 2022; accepted:
12 March, 2022; available on-line: 27 May, 2022

✉e-mail: waszut.urszula@gmail.com

Acknowledgements of Financial Support: This project was funded within the grant “Bridge” (POMOST/2012-5/3) awarded to Prof Dorota Dworakowska, entitled “Pre-clinical targeting of PI3K/Akt/mTOR and RAF/MEK/ERK signalling pathways in adrenocortical cancer: impact on steroidogenesis, cell proliferation and apoptosis”, co-financed by the Foundation for Polish Science and European Union under the auspices of the Department of Nuclear Medicine, Faculty of Nursing and Institute of Maritime and Tropical Medicine, Medical University of Gdansk, Head of the Department: Prof. Dr hab. n. med. Piotr Lass.

Abbreviations: ACA, Adrenocortical adenoma; ACC, adrenocortical carcinoma; FFPE, formalin-fixed, paraffin-embedded tissue

INTRODUCTION

The generation being born today can be expected to live on average more than 100 years. A consequence is that humanity is going to increasingly struggle with ageing-related disease, with cancer among the most important. Study of mammals that are more resistant to ageing has determined that gene regulation rather than gene sequence is critical, with a crucial role of RNA-RNA crosstalk in mammalian ageing, oncogenesis, and pathogenesis of diverse cancers. This involves an intricate interplay among diverse RNA species such as IncRNA and miRNAs. Transcriptomics thus can be expected to be an important tool in the search for potential biomarkers and therapeutic targets for precision oncology (Huang, 2020).

Patient tumour material is routinely obtained by biopsy or after surgery and is stored in the form of FFPE blocks. This is the standard way that hospitals manage diagnostic procedures on tumour tissues, but these samples are maintained in long term storage and are thus available for mutation profiling and precision oncology, with the developing prospect of using such findings for standard identification to aid the clinical decision-making process (Takeda et al., 2018).

Cancer precision medicine based on tumour tissue analysis by next-generation sequencing and other methods is already beginning to provide diagnostic and prognostic information for immediate clinical application and is playing a crucial role in genotype-matching a tumour case with available clinical trials (Gerratana et al., 2019).

Gene expression profiling from FFPE has been increasingly utilised in recent years. Applications have included classifying the tissue of origin of 32 different tumour classes (not including ACC) using a 92-gene panel, achieving 82% correct classification compared with 88% for preparations from frozen tissue (Ma et al., 2006).

Subtyping of cancers using data from a variety of platforms for gene expression, both microarray and RNA-Seq methods, has been recently outlined, using a data-driven reference approach, aiming to eliminate platform-specific bias (Zhang et al., 2019). Below we present an example of the use of FFPE as an investigative tool for the rare cancer ACC, a disorder that still presents major challenges for efficient diagnosis and prediction of potential for malignancy.

FFPE RNA integrity

FFPE-origin RNA is known to be degraded in a specific manner. The most commonly used means to assess this has been formaldehyde gel electrophoresis with ethidium bromide staining, which requires visual assessment of the ribosomal bands. However, this is currently regarded as of limited reliability in comparison with modern genomic techniques due to the effects of the electrophoretic conditions, ethidium bromide fluorescence, an influence of the amount of RNA being loaded, and a lack of sensitivity and specificity (Fleige & Pfaffl, 2006). A newer method for RNA analysis is based on the Bioanalyser 7500 (Agilent). This assesses ribosomal bands and the entire electrophoretic track of the sample. It enables the actual presence or absence of degraded products to be automatically determined and expressed as the RNA Integrity Number (RIN). However, this in turn is claimed to have limitations (Chung et al., 2008; Oberli et al., 2008). A review by Zhang et al. (2017) concludes that both microarray and RNA-seq methods perform better with FFPE–derived RNA than predicted by the RIN. They cite published comparisons of paired FFPE-derived and frozen material giving similar results. RNA chain lengths from FFPE-derived RNA are mostly in the 50-200 bp range.

FFPE material shows increasing degradation over time. A comparison of 10-year-old and recent samples showed lower RNA quantity in the older samples but of similar quality. However, while gene expression, using RT-PCR, of a 151 bp fragment was satisfactory in all samples, that of a 242 bp fragment was poor in the older material (Ribeiro-Silva et al., 2007). Niesen and others (Niesen et al., 2013) found that RNA yield decreased with storage time but, unexpectedly, the number of gene-present calls in their FFPE RNA tissue samples (using Affymetrix GeneChips) were comparable with freshly isolated human peripheral blood mononucleocyte cells. They suggested that transcripts are uniformly fragmented sequences across the full length, with the majority of the transcript being 75–100 bp and a small number of them retained at up to 3000–4000 bp in length. Other determinants of RNA quality are methods of fixation, storage temperature, and light. Base substitution by the fixative can diminish the accuracy of reads.

Tolerance to sample degradation has also been shown by Die & Román (2012), who scrupulously examined the influence of RNA quality on gene expression profiling by microarrays and found that amplification curves with exponential and plateau regions of properly sigmoidal shape could usually be acquired with qPCR, even from degraded templates. They studied the effects of varying amounts of deliberately induced RNA degradation on the expression profile of samples from human tumour and healthy tissue. Moderate degradation did not spoil microarray analyses and permitted meaningful results to be obtained if used carefully. They advised that with strongly degraded input RNA, small expression differences might give a false positive result. Accordingly, we have used in our example a normalisation method employing reference genes and have limited reporting of gene expression differences to 5-fold change or more.

EXAMPLE: pilot study of mRNA preparation from ACC

Background

ACC accounts for 2–11% of adrenal masses. It is responsible for approximately 0.2% of all cancer deaths in the US, with a mean survival time of only months. Available treatments are controversial; outcomes are poor, with only 35% survival at five years.

Imaging of adrenal tumours often reveals a large mass with a central area of necrosis or haemorrhage; heterogeneous enhancement favours malignancy (Moreira & Pow-Sang, 2002; Bharwani et al., 2011). Histological diagnosis of ACC requires specialist expertise (Libe, 2019). The scoring system of Weiss is universal and may be improved using more criteria (Blanes & Diaz-Cano, 2007). It is difficult to predict malignancy based on clinical, biochemical or histologic grounds in the absence of capsular invasion and distant metastases (Fonseca et al., 2012). These are also poor mortality indicators (Calissendorff et al., 2016) but careful control of techniques offers some improvement (Jouinot & Bertherat, 2018). There is a high degree of intertumoural heterogeneity, both in terms of histology, biochemistry (Taylor et al., 2017) and molecular biology (Libe, 2019), which may change with tumour progression (Vatrano et al., 2018). This variability stands in strong contrast to the H295R ACC cell line model, used universally for in vitro study and also to the relatively more common ACA, which shows a limited range of frequent mutations (Backman et al., 2019).

Few studies have yet attempted to differentiate portions of these heterogeneous tumours. Separating zones within either fresh or frozen tissue would be problematic. Many method descriptions do not specify which was used, although Giordano used frozen tissue (Giordano et al., 2003; Giordano et al., 2009). For these rare carcinomas, gathering and processing of frozen tissue for gene expression studies on a large scale would not be feasible (Akerstrom et al., 2015; Backman et al., 2019). In contrast, long-term storage of tumour tissue as FFPE blocks is universal. If RNA of adequate quality and quantity for gene expression analyses could be isolated from this material, even blocks in long-term storage from a variety of centres could be studied. Since it is solid, direct comparison of histology and gene expression in tumour areas is theoretically possible, although limited by quantity of tissue available in some ACC cases. This has the potential to provide a much more comprehensive picture of the disruption of control pathways that lead to tumour development.

The pilot study described below aimed to efficiently isolate nucleic acids from differentiated portions of FFPE material of sufficient quality for gene expression profiling. We know of only one published study of FFPE from patients with ACC. This did not differentiate portions of the tumour and showed a poor success rate, with only 4 of 14 blocks producing useable material (Lombardi et al., 2006). We first assessed the yield of RNA and then similarly used RT-PCR for β-actin to check the adequacy of the RNA template in some samples. We then compared performance of an ACC sample and normal tissue in an mTOR pathway array.

Although ACC tissue features accumulation of disturbances of several genes, increased expression of IGF-2 signalling is the most frequent change (Barlaskar et al., 2009; Fonseca et al., 2012). IGF-2 is an upstream promotor of the mTOR pathway (Fassnacht et al., 2013; Szyszka et al., 2016). Molecular biology has the potential to be a precision oncology tool which can identify alterations in different signalling pathways and offer single biomarkers or clusters that can be utilised in clinical practice (Durand et al., 2011; Fonseca et al., 2012; Fassnacht et al., 2013; Assie et al., 2014; Vatrano et al., 2018).

METHODS

Figure 1 summarises the preparation steps.

Preparation of portions of FFPE histopathology sections

The FFPE blocks used had been prepared 2-5 years previously from tissue being surgically removed from patients with ACC (n=10) or ACA (n=7). Microsections were prepared by an expert histopathologist; 9 slides per case were obtained, cut at 7 µm thick sections. Only the first from each series was stained with hematoxylin–eosin (because staining of FFPE sections may enhance RNA degradation) and designated as central or peripheral by histopathology expert Dr Salvador Diaz-Cano. Regions of the tissue in the succeeding sections with morphological differences matching those in the stained section were identified using a single lens or binocular microscope at the lowest magnification that would enable distinction of areas while providing a wide enough field of view. The working space and tools were cleaned with RNase AWAY (Thermo Fisher Scientific). These sections (7 for RNA extraction, and one for DNA extraction) were dissected using a thin scalpel blade and the differentiated portions were pooled in two separate Eppendorf tubes for each extraction.

Isolation of RNA from FFPE tissue sections

The RNeasy FFPE Kit (Qiagen) was used, following the RNeasy FFPE Handbook protocol with modifications. It is specifically designed to, as much as possible, reverse the chemical modifications of RNA caused by formaldehyde, such as crosslinking, which cannot be detected by standard quality analysis, such as agarose gel electrophoresis and ‘lab on-a-chip’ (the Agilent Bioanalyzer system) and isolates RNA fragments longer than 70 nucleotides, which are suitable for gene expression analysis (Qiagen, 2013).

A maximum of 2 sections 5–20 μm thick is recommended in the protocol. Our 7 sections on slides, divided into 2 zones, equated to the middle of this range. Each sample was first dewaxed in a fume cupboard using 1 ml xylene. After thorough vortex mixing and centrifugation at maximum speed for 5 min (2 min in the protocol), the xylene containing the paraffin was carefully poured off, taking care not to disturb the pellet remaining at the bottom of the tube. Deparaffinization was repeated, followed by a similar wash with 96% aqueous ethanol. The tubes were left in the fume cupboard until the pellet was completely dry. The next step, Proteinase K digestion, was extended from the time suggested in the Qiagen protocol from 15 min to 12 hours at 56°C, as it was observed that this had a very significant impact on the total RNA concentration and quality, in accordance with advice provided on the GenomeWeb site (Johnson, 2014). This was followed by 15 min at 80°C, according to the protocol, a step optimised to reverse formaldehyde-induced cross-linking. Cooling and then DNase treatment followed the protocol.

For RNA purification, the protocol recommends two options for the volumes of ethanol and Buffer RBC that are passed through the RNeasy MiniElute spin column. The higher volumes were found to give higher yield of the total RNA: Buffer RBC (500 µl) and ethanol (1200 µl). Buffer RPE ×2 was added to the column according to the protocol. The RNA was eluted using 17 µl RNase-Free Water (for use with RNAase-Free DNase I, provided by kit) with centrifugation at full speed for 3 min (1 min in protocol). The Qiagen columns used for FFPE RNA purification are claimed to remove fragments shorter than 200 bp.

After extraction, 2 μl of the obtained RNA was immediately taken from the original tube and combined with 2 µl RNase-free deionised water in a mini tube and kept on ice, ready for measurement of the concentration of total RNA using a NanoDrop 1000 Spectrophotometer (ThermoFisher). The parameters A260/A230 A260/A280 were in the range of 1.89–2.05; these values indicate satisfactory purity. Total RNA samples were stored in –80°C.

RT-PCR for β-actin

The TaqMan double-dye probe RT-PCR assay (Applied Biosystems) for β-actin was used to check for the presence of amplifiable RNA template. qRT-PCR fragment targets should ideally be 50–150 bp. The maximum amplicon size should not exceed 400 bp, and for probe assay, a study of the relationship between length and efficiency, has shown increase of Ct above a length of about 200 bp (Debode et al., 2017). The β-actin TaqMan amplicon size is 171 bp.

The assay was carried out with the QuantiTect Reverse Transcription Kit (Qiagen), according to the QuantiTect Probe PCR Handbook (Qiagen, 2011a). This first eliminates genomic DNA contamination before RT-PCR. A Rotor-Gene 6000 real-time PCR cycler (ThermoFisher) was set, according to the protocol, to 50 repeats with: denaturation 10 s 94°C, annealing 57°C, hold 20°C. An amplification plot and table with Ct results was finally generated with the supplied software.

Pathway-targeted preamplification following cDNA synthesis

The extracted RNA was tested for performance in gene expression profiling using a pathway-specific Human mTOR Signalling RT² Profiler PCR Array (Qiagen 2015). One sample (peripheral ACC case 1) and control (normal cortex) were prepared in duplicate.

For the reverse transcription reaction, 200 ng total RNA for each sample was processed using the RT² PreAMP cDNA Synthesis Kit (Qiagen). This was less than advised in the protocol (500 ng) since we wanted to check whether it was possible to save RNA.

The cDNA synthesis and preamplification reaction procedures were carried out according to the RT² PreAMP cDNA Synthesis Handbook (Qiagen, 2011b). Incubations and cycling were processed with a Veriti Thermal Cycler (ThermoFisher). Lack of presence of genomic DNA is monitored by a quality control on the Qiagen PCR array, which also contains a reverse transcription (RT) control and a positive PCR control, which provides a check on cDNA synthesis efficiency. These checks were passed during our assessments.

PCR amplification on the Human mTOR Signalling RT² Profiler PCR Array (Qiagen 2015)

The main amplification reactions were carried using a ViiA 7 Real-Time PCR System (ThermoFisher) instrument equipped with ViiA™ 7 Software. The reaction on the RT² Profiler PCR Array 4×96 well array was initially prepared according to the RT² Profiler PCR Handbook. This was shown to give poor results for the amplification plot obtained using the supplied software. The cycling program was then changed to an alternative FFPE ViiA 7 protocol specially obtained from Qiagen, which used the following cycling conditions: hold Stage, 50°C, 2.00 min; PCR Stage, 40 cycles 95°C, 0.15 min, 60°C, 1.00 min, melt curve stage with default settings. On completion, the Ct values obtained were much improved but with some delay, which indicated that it was better to use the amount of template specified in the protocol.

Data normalisation and result calculation

Gene expression data was normalized using the arithmetic average of the three most stable of a panel of 5 housekeeping genes according to current practice (Pfaffl, 2007). The duplicates showed good agreement for each gene and the means were used for calculation. The ΔΔCT method (to calculate expression differences between tumour tissue and normal) was applied using RT² PCR Array Data Analysis version 3.5 software. Fold changes relative to normal tissue higher than 5 were considered significant.

RESULTS

Extracted FFPE-derived total RNA

Total RNA obtained from the on-slide FFPE designated tumour areas was in the range of 30–380 ng/µl (Table 1); one area from a normal cortex was extracted, at a concentration of 280 ng/µl.

The probe RT-PCR for the β-Actin gene was carried out on the first 5 pairs of samples to provide an assessment of whether the total RNA quality was adequate for amplification. All were found to give positive results. The amplification curves showed the expected shape and plateau regions, according to qPCR gold standards. Ct values for duplicates are shown in Table 1. These are all well below 29 cycles: higher values would have indicated ineffective nucleic acid isolation or PCR inhibition, inefficient RT-reaction, RNA/cDNA degradation, or too little template. The FFPE-originated templates were thus confirmed as sufficient for gene expression analysis with qPCR arrays.

Most samples provided a quantity of RNA that would be enough for 4 PCR array runs on a 94 times 4 well plate format (or two runs in duplicate) and dozens of qPCR validation reactions. The protocol recommends an RNA concentration of 500 ng per plate. Samples would need to be run in duplicate, which would require at least 68 ng/μl. For a few samples in which the concentration obtained was smaller than this, it would only be possible to run one PCR array.

Table 2 lists those genes in the mTOR pathway that were differently expressed. Appendix A (at https://ojs.ptbioch.edu.pl/index.php/abp) shows expression of all genes graphically. ACC peripheral tumour area versus normal adrenal tissue is expressed as fold change, together with brief detail on their role in the mTOR pathway, to help expand understanding of malignancy or biomarker potential.

Discussion of the findings

The 84 genes targeted by the array are relevant to the mTOR pathway and so are related to the regulation of growth and differentiation and stress responses. Of these, 19 showed significant change in expression in the ACC tissue. Interpreting a given gene product as promoting or inhibiting tumour growth is not always straightforward, with, in some cases, opposite effects being reported for different malignant tumour types.

mTOR is a core component of two complexes, mTORC1 and mTORC2, each associated with several other components and subject to different stimulatory and inhibitory influences. In mTORC1, it acts as a serine-threonine protein kinase. Activation of this complex stimulates cell growth and proliferation in nutrient-rich conditions favourable to cell growth (Pópulo et al., 2012). In mTORC2, mTOR acts as a tyrosine protein kinase, controlling the actin cytoskeleton, with major effects on cell proliferation and survival. Under hypoxic conditions in a normal cell, mTORC1 activity is down-regulated by mechanisms which include activation of AMPK and by tuberous sclerosis complex 1 and 2 (TSC1/2). The latter deactivates Rheb (Ras homologue enriched in brain), a protein pivotal in the activation of mTORC1. The expression of hypoxia-modulated genes can differ between normal and tumour cells and may be affected by the tumour microenvironment (Cella et al., 2015).

Potential significance of these major changes is discussed in the following paragraphs, in the same sequence as in Table 2.

A marked downregulation of PRKAG3 and PRKAB2 is a feature of some tumour cells, which appear to be under selection pressure to downregulate AMPK; this would overcome cell growth being stopped in hypoxic conditions (Hardie, 2015). Normally, environmental stressors, such as heat shock, nutrient deprivation, hypoxia, and ischemia inhibit the mTOR pathway via action of AMPK (Rubio et al., 2013). This protein generally promotes glucose-sparing, oxidative metabolism, rather than the rapid glucose uptake and glycolysis used by quickly proliferating cells. Imaging of adrenal tumours often reveals a large mass with a central area of necrosis, likely indicating that they have outgrown their blood supply and so must have developed in a relatively hypoxic environment. Their patterns of steroidogenesis frequently indicate relative deficit of the enzyme activities that are mitochondrial, especially of 11β-hydroxylase, which also may reflect hypoxia (Moreira & Pow-Sang, 2002; Bharwani et al., 2011; Taylor et al., 2017).

Downregulation of GSK3B impairs phosphorylation of β-catenin, leading to the constitutive activation of the β-catenin-mediated Wnt pathway (Nault & Zucman-Rossi, 2014; Glibo et al., 2021). β-Catenin accumulation in the nucleus promotes a large number of oncogenes and is a frequent characteristic of adrenocortical cancers. It is associated with a dismal prognosis. Study of candidate genes deregulated in β-catenin-mutated adrenocortical tumours may lead to a better understanding of the role of the pathway in adrenal cancer onset and maintenance (Durand et al., 2011). Among other functions, GSK3B also acts on the mitochondria to decrease the threshold at which apoptosis occurs in response to stress stimuli and to prevent the development of the Warburg effect, so that decreased expression would further the capacity of the tumour to survive hypoxic stress. Hypoxia is not effective in suppressing mTOR once other negative regulators are lost during malignant transformation (Cella et al., 2015).

MAPK signalling was found to be upregulated, as was the insulin receptor. This pathway mediates direct cellular responses to many proliferative stimuli and transduces signal from IGF2 (El-Shewy et al., 2007). The mTORC1 pathway is entwined with insulin signalling, with a regulatory feedback loop. IGF2 binding with the insulin receptor protects it from downregulation, which causes continuous mitogenic stimulation and thus sustained carcinogenic effects (Djiogue et al., 2013; Fraenkel et al., 2008). In chemoresistant Ewing’s sarcoma, the IGF1/IGF receptor is switched to IGF2/insulin receptor dependency, which results in continuous activation of MAPK signalling, leading to proliferation and migration. This switch may also partially explain cases of chemoresistance to various anticancer drugs. Failure of efficacy of the mTOR suppressor everolimus in ACC was reported as related to the insulin signalling loop (Djiogue et al., 2013). Thus, our present results appear to reflect this mechanism in ACC. In up to 90% of ACC, there is a loss of heterozygosity leading to bi-allelic expression of IGF2 (Barlaskar et al., 2009; Fassnacht, et al., 2013, Fonseca et al., 2012). Dependence of recurrence-free survival on the level of expression of the IGF-II gene cluster has been shown (de Fraipont et al., 2005). A comprehensive genomic study of ACC has differentiated three subtypes with divergent clinical outcomes, with all strongly expressing IGF-2 but with disparate regulation of miRNA (Mohan et al., 2018).

In our study MYO1C was strongly downregulated in tumour cells vs normal cells. It is a tumour suppressor candidate, acting to suppress both expression and activation of AKT. Thus, a reduction results in activation of AKT and stimulation of the mTOR pathway. In transfection experiments in cultured HEK-293 cells, the levels of MYO1C and AKT were negatively correlated and also endometrial carcinoma grade was worst in those with lowest MYO1C expression (Visuttijai et al., 2016). The PI3K pathway can be activated by diverse oncogenes or upstream growth factors that bind receptor tyrosine kinase (RTK), causing stimulation of protein translation and metabolic processes associated with cell proliferation, enhanced motility and, finally, inhibition of apoptosis. The molecular product of PI3K, AKT, activates mTOR. All the members of the PI3K/Akt/mTOR pathway, from the upstream RTK inducers to the final effectors, have been associated with cancer development and progression. Tian and collegaues (Tia et al., 2019) quote that >70% of cancers have these increased.

PI3Kδ expression was the most strongly downregulated. Although it is closely related to PI3K, it has a specific function in promotion of the immune system (Nunes-Santos et al., 2019). Mutations are present in some cancers: this would be expected to diminish immune response and so enable tumour progression.

VEGFs were all markedly overexpressed in this study. Hypoxia plays a major role in increased vascular VEGF production, but activation of the PI3K/AKT/mTOR pathway in tumour cells can also increase VEGF secretion. VEGF A has been reported to be highly expressed in ACC (Fassnacht et al., 2013), with correlation to the degree of malignancy (Foltyn et al., 2012, Kolomecki et al., 2001, Weddell et al., 2018). VEGF B and C were especially strongly upregulated in our study. VEGFs interact with the VEGF receptors 1-3, which are RTKs. Several suppressive agents have been developed in prospective ACC treatment but with disappointing results (Fassnacht et al., 2013). Inhibition of VEGFR 1 and 2 using a designed peptide diminished PI3K/AKT and MAPK/ERK1/2 signalling pathway activity and blocked angiogenesis, tumour growth, and metastasis in cultured cells (Sadremomtaz et al., 2018). IGFs and insulin also interact with RTKs. Sorafenib, a multi-RTK inhibitor, is effective against some cancers but has a limited antiproliferative effect in ACC. A recent study of Sorafenib incubation with the ACC cell line H295R confirmed its action but some cells survived (Cerquetti et al., 2021).

Increased expression of STK11 is not easily interpreted. Although initially shown to be a tumour suppressor in some models, recent studies suggest that certain mutations of STK11 do lead to gain of oncogenic properties, such as in lung cancer, where mutations in exons 1 & 2 are associated with an aggressive subtype (Marcus & Zhou, 2010; Pécuchet et al., 2017). Peutz-Jeghers syndrome features STK11 mutations and a high rate of cancer, especially those of breast and gastrointestinal tract. Activation of mTORC1 downstream targets has been shown in affected families (Li et al., 2018).

Decreased expression of eIF4E is initially surprising since it is overexpressed in many cancers but underexpression does also occur, indicating it may have either an oncogenic or tumour suppressor function (Sobočan et al., 2020).

Similarly, decreased expression of PRKCG is superficially unexpected since it is overexpressed in many cancers and had been thought to be tumour-promoting since its ligands, the phorbol esters, have unambiguous oncogenic properties. However, an analysis of PRKCG mutations in human cancers has shown that most were loss-of-function, and none were activating, indicating that it is a tumour suppressor. A depletion of its function promotes anchorage-independent growth (Antal et al., 2015; Garg et al., 2014).

RhoA is overexpressed in this study. It has also been associated with malignant cancer proliferation and metastasis. The latter may be enhanced by its action at the rear of migrating cells to promote detachment (Gordon et al., 2014). It is seen as having a great potential as a drug target, as well as a candidate biomarker (Nam et al., 2019).

It is interesting that PLD1 is overexpressed but PLD2 is underexpressed. PLD1 is activated by RhoA. Both PLDs are frequently overexpressed in cancers. Increased levels of choline-containing precursors are reported to be emerging markers of tumour progression (Glunde et al., 2011). PLD1 integrates upstream signalling of mTOR, exerts control of cell size and is involved in amino acid availability sensing for mTOR (Fang et al., 2003). PLD1 but not PLD2 is required for Rheb activation of the mTOR pathway. The TSC/Rheb pathway is an integrator at upstream signals such as cellular energy level and growth factors which regulate mTORC1 (Sun et al., 2008). The extent to which these isozymes have separate roles remains uncertain; however, development of selective inhibitors is being actively pursued, both to elucidate this and as potential cancer therapeutic agents (Brown et al., 2017).

RRAGA (RRAGagA) was upregulated. This activates mTORC1 signalling via the GTPase Rheb. Rag GTPases are upstream effectors that are sensitive to nutrients (Hong-Brown et al., 2012, Sancak et al., 2010)

CONCLUSIONS

We have shown that differentiated portions of FFPE sections of ACCs can reliably yield RNA of adequate quality for gene expression profiling. Our analysis of mTOR pathway gene expression using microarray analysis in one sample has shown changes that are already reported for this tissue or are plausible in the context of cancer, which further supports the potential of this material for clinical applications and research.

The yield is sufficient in most cases for running expression arrays in duplicate. However, this has been largely superseded by RNA-seq methods, which allow the use of much less material (Krogsdam-Christensen 2019). This may enable finer differentiation within the block. An example is the use of single sections of archival breast cancer samples (Pennock et al., 2019).

The RNA obtained from FFPE material using the method we have described or others that are commercially available should be suitable for all current approaches to gene expression studies. A comparison of nine mRNA profiling methods for FFPE versus frozen breast cancer material showed a generally good comparability (Turnbull et al., 2020), with little influence on the ability to identify differentially expressed transcripts.

A painstaking approach to sample preparation, using dissection with a scalpel, appears worthwhile and also enables any normal tissue in the tumour sample to be eliminated or harvested separately. The single published study of ACC tissue (Lombardi et al., 2006) discussed the potential for confusion caused by the presence of non-neoplastic tissue in the tumour samples and advised microdissection but considered that this could cause further degradation. A recent study of canine mammary tumours did use laser microdissection successfully, with the extracted RNA being analysed with NGS. These authors also achieved an up to tenfold improvement in yield with the use of focused ultrasonication with proteinase K to assist tissue disruption and perhaps also cleavage of RNA-protein crosslinks (Amini et al., 2017). Simultaneous analysis of differentiated portions of the tumour using both metabolomic analysis by GC-MS and gene expression could be very powerful for improving the fundamental understanding of tumour progression, prognosis, or a new therapy option of this challenging disorder. Gene expression profiling may enable determination of tissue origin, with potential for clinical routine use.

We conclude that use of archival FFPEs offers a greatly expanded potential for transcriptomic studies of these rare adrenal carcinomas on a population-wide basis.

Acknowledgements

The research study was conducted at the Molecular Pathology lab, King’s College Hospital, London, UK, in collaboration with the Richard Dimbleby Laboratory of Cancer Research, King’s College London, London UK. FFPE block and slide preparation was carried out in the Department of Histopathology, King’s College Hospital. We thank Dr Salvador Diaz-Cano for providing this material and for carrying out the histological examination of it.

REFERENCES

Akerstrom T, Willenberg HS, Cupisti K, Ip J, Backman S, Moser A … Hellman P (2015) Novel somatic mutations and distinct molecular signature in aldosterone-producing adenomas. Endocr Relat Cancer 22: 735–744. https://doi.org/10.1530/ERC-15-0321

Amini P, Ettlin J, Opitz L, Clementi E, Malbon A, Markkanen E (2017) An optimised protocol for isolation of RNA from small sections of laser-capture microdissected FFPE tissue amenable for next-generation sequencing. BMC Mol Biol 18: 22. https://doi.org/10.1186/s12867-017-0099-7

Antal CE, Hudson AM, Kang E, Zanca C, Wirth C, Stephenson NL, Trotter EW, Gallegos LL, Miller CJ, Furnari FB, Hunter T, Brognard J, Newton AC (2015) Cancer-associated protein kinase C mutations reveal kinase’s role as tumor suppressor. Cell 160: 489–502. https://doi.org/10.1016/j.cell.2015.01.001

Assié G, Jouinot A, Bertherat J (2014) The ‘omics’ of adrenocortical tumours for personalized medicine. Nat Rev Endocrinol 10: 215–228. https://doi.org/10.1038/nrendo.2013.272

Backman S, Akerstrom T, Maharjan R, Cupisti K, Willenberg HS, Hellman P, Bjorklund P (2019) RNA Sequencing provides novel insights into the transcriptome of aldosterone producing adenomas. Sci Rep 9: article 6269. https://doi.org/10.1038/s41598-019-41525-2

Barlaskar FM, Spalding AC, Heaton JH, Kuick R, Kim AC, Thomas DG, Giordano TJ, Ben-Josef E, Hammer GD (2009) Preclinical targeting of the type I insulin-like growth factor receptor in adrenocortical carcinoma. J Clin Endocrinol Metab 94: 204–212. https://doi.org/10.1210/jc.2008-1456

Bharwani N, Rockall AG, Sahdev A, Gueorguiev M, Drake W, Grossman AB, Reznek RH (2011) Adrenocortical carcinoma: the range of appearances on CT and MRI. Am J Roentgenol 196: W706–W714. https://doi.org/10.2214/AJR.10.5540

Blanes A, Diaz-Cano SJ (2007) Histologic criteria for adrenocortical proliferative lesions: value of mitotic figure variability. Am J Clin Pathol 127: 398–408. https://doi.org/10.1309/MCGUQ3R4A4WWN3LB

Brown HA, Thomas, PG, Lindsley CW (2017) Targeting phospholipase D in cancer, infection and neurodegenerative disorders. Nat Rev Drug Discov 16: 351–367. https://doi.org/10.1038/nrd.2016.252

Calissendorff J, Calissendorff F, Falhammar H (2016) Adrenocortical cancer: mortality, hormone secretion, proliferation and urine steroids – experience from a single centre spanning three decades. BMC Endocr Disord 16: 15. https://doi.org/10.1186/s12902-016-0095-9

Cella CA, Minucci S, Spada F, Galdy S, Elgendy M, Ravenda PS ... Fazio N (2015) Dual inhibition of mTOR pathway and VEGF signalling in neuroendocrine neoplasms: from bench to bedside. Cancer Treat Rev 41: 754–760. https://doi.org/10.1016/j.ctrv.2015.06.008

Cerquetti L, Bucci B, Raffa S, Amendola D, Maggio R, Lardo P, Petrangeli E, Torrisi MR, Toscano V, Pugliese G, Stigliano A (2021) Effects of sorafenib, a tyrosin kinase inhibitor, on adrenocortical cancer. Front Endocrinol 12: 667798. https://doi.org/10.3389/fendo.2021.667798

Chung JY, Braunschweig T, Williams R, Guerrero N, Hoffmann KM, Kwon M, Song YK, Libutti SK, Hewitt SM (2008) Factors in tissue handling and processing that impact RNA obtained from formalin-fixed, paraffin-embedded tissue. J Histochem Cytochem 56: 1033–1042. https://doi.org/10.1369/jhc.2008.951863

Debode F, Marien A, Janssen É, Berben G, Bragard C (2017). The influence of amplicon length on real-time PCR results. Biotechnology, Agronomy and Society and Environment 21: 3–11. https://doi.org/10.25518/1780-4507.13461

Die JV, Román B (2012) RNA quality assessment: a view from plant qPCR studies. J Exp Bot 63: 6069–6077. https://doi.org/10.1093/jxb/ers276

Djiogue S, Nwabo Kamdje AH, Vecchio L, Kipanyula MJ, Farahna M, Aldebasi Y, Seke Etet PF (2013) Insulin resistance and cancer: the role of insulin and IGFs. Endocr Relat Cancer 20: R1–R17. https://doi.org/10.1530/ERC-12-0324

Durand J, Lampron A, Mazzuco TL, Chapman A, Bourdeau I (2011) Characterization of differential gene expression in adrenocortical tumors harboring beta-catenin (CTNNB1) mutations. J Clin Endocrinol Metab 96: E1206-11. https://doi.org/10.1210/jc.2010-2143

El-Shewy HM, Lee MH, Obeid LM, Jaffa AA, Luttrell LM (2007) The insulin-like growth factor type 1 and insulin-like growth factor type 2/mannose-6-phosphate receptors independently regulate ERK1/2 activity in HEK293 cells. J Biol Chem 282: 26150–26157. https://doi.org/10.1074/jbc.M703276200

Fang Y, Park IH, Wu AL, Du G, Huang P, Frohman MA, Walker SJ, Brown HA, Chen J (2003) PLD1 regulates mTOR signaling and mediates Cdc42 activation of S6K1. Curr Biol 13: 2037–2044. https://doi.org/10.1016/j.cub.2003.11.021

Fassnacht M, Kroiss M, Allolio B (2013) Update in adrenocortical carcinoma. J Clin Endocrinol Metab 98: 4551–4564. https://doi.org/10.1210/jc.2013-3020

Fleige S, Pfaffl MW (2006) RNA integrity and the effect on the real-time qRT-PCR performance. Mol Aspects Med 27: 126–39. https://doi.org/10.1016/j.mam.2005.12.003

Foltyn W, Strzelczyk J, Marek B, Kajdaniuk D, Sieminska L, Rosiek V, Kos-Kudla B (2012) The usefulness of determining the serum concentrations of vascular endothelial growth factor (VEGF) and its soluble receptor type 2 (sVEGF-2) in the differential diagnosis of adrenal incidentalomas. Endokrynol Pol 63: 22–28. https://journals.viamedica.pl/endokrynologia_polska/article/viewFile/25195/20024

Fonseca AL, Healy J, Kunstman JW, Korah R, Carling T (2012) Gene expression and regulation in adrenocortical tumorigenesis. Biology 2013: 26–39. https://doi.org/10.3390/biology2010026

de Fraipont F, El Atifi M, Cherradi N, Le Moigne G, Defaye G, Houlgatte R, Bertherat J, Bertagna X, Plouin PF, Baudin E, Berger F, Gicquel C, Chabre O, Feige JJ (2005) Gene expression profiling of human adrenocortical tumors using complementary deoxyribonucleic Acid microarrays identifies several candidate genes as markers of malignancy. J Clin Endocrinol Metab 90: 1819–1829. https://doi.org/10.1210/jc.2004-1075

Fraenkel M, Ketzinel-Gilad M, Ariav Y, Pappo O, Karaca M, Castel J, Berthault MF, Magnan C, Cerasi E, Kaiser N, Leibowitz G (2008) mTOR inhibition by rapamycin prevents β-cell adaptation to hyperglycemia and exacerbates the metabolic state in type 2 diabetes. Diabetes 57: 945–957. https://doi.org/10.2337/db07-0922

Garg R, Benedetti LG, Abera MB, Wang H, Abba M, Kazanietz MG (2014) Protein kinase C and cancer: what we know and what we do not. Oncogene 33: 5225–5237. https://doi.org/10.1038/onc.2013.524

Gerratana L, Davis AA, Shah AN, Lin C, Corvaja C, Cristofanilli M (2019) Emerging role of genomics and cell-free DNA in breast cancer. Curr Treat Options Oncol 20: 68. https://doi.org/10.1007/s11864-019-0667-9

Giordano TJ, Thomas DG, Kuick R, Lizyness M, Misek DE, Smith AL, Sanders D, Aljundi RT, Gauger PG, Thompson NW, Taylor JMG, Hanash SM (2003) Distinct transcriptional profiles of adrenocortical tumors uncovered by DNA microarray analysis. Am J Pathol 162: 521–531. https://doi.org/10.1016/S0002-9440(10)63846-1

Giordano TJ, Kuick R, Else T, Gauger PG, Vinco M, Bauersfeld J, Sanders D, Thomas DG, Doherty G, Hammer G (2009) Molecular classification and prognostication of adrenocortical tumors by transcriptome profiling. Clin Cancer Res 15: 668–676. https://doi.org/10.1158/1078-0432.CCR-08-1067

Glibo M, Serman A, Karin-Kujundzic V, Vlatkovic IB, Miskovic B, Vranic S, Serman L (2021) The role of glycogen synthase kinase 3 (GSK3) in cancer with emphasis on ovarian cancer development and progression: a comprehensive review. Bosn J Basic Med Sci 21: 5–18. https://doi.org/10.17305/bjbms.2020.5036

Glunde K, Bhujwalla ZM, Ronen SM (2011) Choline metabolism in malignant transformation. Nat Rev Cancer 11: 835–848. https://doi.org/10.1038/nrc3162

Gordon BS, Kazi AA, Coleman CS, Dennis MD, Chau V, Jefferson LS, Kimball SR (2014) RhoA modulates signaling through the mechanistic target of rapamycin complex 1 (mTORC1) in mammalian cells. Cell Signal 26: 461–467. https://doi.org/10.1016/j.cellsig.2013.11.035

Hong-Brown, LQ, Brown CR, Kazi AA, Navaratnarajah M, Lang CH (2012) Rag GTPases and AMPK/TSC2/Rheb mediate the differential regulation of mTORC1 signaling in response to alcohol and leucine. Am J Physiol Cell Physiol 302: C1557–C1565. https://doi.org/10.1152/ajpcell.00407.2011

Hardie DG (2015) Molecular Pathways: Is AMPK a friend or a foe in cancer? Clin Cancer Res 21: 3836–3840. https://doi.org/10.1158/1078-0432.CCR-14-3300

Huang Z (2020) Non-coding RNA regulation potentially contributes to the exceptional longevity evolved in bats. Biodiversity Genomics 2020 from 5-9 October,virtual conference. https://stream.venue-av.com/e/biodiversity/Talk_Abstracts

Johnson M (2014) Study Shows FFPE Pre-PCR Processing Problems Plague Pathology Labs. Genomeweb[website]. GenomeWebhttps://www.genomeweb.com/sample-prep/study-shows-ffpe-pre-pcr-processing-problems-plague-pathology-labs-0#.Xunoo-cwiiM, acessed August 3, 2018

Jouinot A, Bertherat J (2018) Adrenocortical carcinoma: Differentiating the good from the poor prognosis tumors. Eur J Endocrinol 178: R215–R230. https://doi.org/10.1530/EJE-18-0027

Kolomecki K, Stepien H, Bartos M, Kuzdak K (2001) Usefulness of VEGF, MMP-2, MMP-3 and TIMP-2 serum level evaluation in patients with adrenal tumours. Endocr Regul 35: 9–16. https://www.sav.sk/journals/endo/full/er0101b.pdf

Krogsdam Christensen AM (2019) Reliable RNA-Seq Expression Profiling from Low-Quality FFPE Biobank Samples In Lexogen [webinar]. Available from https://www.lexogen.com/webinar-reliable-rna-seq-expression-profiling-from-low-quality-ffpe-biobank-samples/?gclid=Cj0KCQiAuefvBRDXARIsAFEOQ9GYTyoQD8pKbKzgFlCNQHTWKdct-fdhBvSGqIN_MGnhZkpcpPJI7IIaAq1pEALw_wcB

Li R, Wang Z, Liu S, Wu B, Zeng D, Zhang Y, Gong L, Deng F, Zheng H, Wang Y, Chen C, Chen J, Jiang B (2018) Two novel STK11 missense mutations induce phosphorylation of S6K and promote cell proliferation in Peutz Jeghers syndrome. Oncology Lett 15: 717–726. https://doi.org/10.3892/ol.2017.7436

Libe R (2019) Clinical and molecular prognostic factors in adrenocortical carcinoma. Minerva Endocrinol 44: 58–69. https://doi.org/10.23736/S0391-1977.18.02900-0

Lombardi CP, Raffaelli M, Pani G, Maffione A, Princi P, Traini E, Galeotti T, Rossi ED, Fadda G, Bellantone R (2006) Gene expression profiling of adrenal cortical tumors by cDNA macroarray analysis. Results of a preliminary study. Biomed Pharmacother 60: 186–90. https://doi.org/10.1016/j.biopha.2006.03.006

Ma X-J, Patel R, Wang X, Salunga R, Murage J, Desai R, Tuggle JT, Wang W, Chu S, Stecker K, Raja R, Robin H, Moore M, Baunoch D, Sgroi D, Erlander M (2006) Molecular classification of human cancers using a 92-gene real-time quantitative polymerase chain reaction assay. Arch Pathol Lab Med 130: 465–473. https://doi.org/10.1043/1543-2165(2006)130[465:MCOHCU]2.0.CO;2

Marcus AI, Zhou W (2010) LKB1 regulated pathways in lung cancer invasion and metastasis. J Thorac Oncol 5: 1883–1886. https://doi.org/10.1097/JTO.0b013e3181fbc28a

Mohan DR, Lerario AM, Hammer GD (2018) Therapeutic targets for adrenocortical carcinoma in the genomics era. J Endocr Soc 2: 1259–1274. https://doi.org/10.1210/js.2018-00197

Moreira SGJ, Pow-Sang JM (2002) Evaluation and management of adrenal masses. Cancer Control 9: 326–334. https://doi.org/10.1177/107327480200900407

Nam S, Kim JH, Lee DH (2019) RHOA in gastric cancer: functional roles and therapeutic potential. Front Genet 10: 438. https://doi.org/10.3389/fgene.2019.00438

Nault JC, Zucman-Rossi J (2014) Physiopathology of hepatocellular carcinoma. In Pathobiology of Human Disease a dynamic encyclopedia of disease mechanisms. McManus LM, Mitchell RN eds, pp 1881–1886. Academic Press, Cambridge, Mass. https://www.sciencedirect.com/referencework/9780123864574/pathobiology-of-human-disease

Niesen CE, Xu J, Fan X, Li X, Wheeler CJ, Mamelak AN, Wang C (2013) Transcriptomic profiling of human peritumoral neocortex tissues revealed genes possibly involved in tumor-induced epilepsy. PLoS One 8: e56077. https://doi.org/10.1371/journal.pone.0056077

Nunes-Santos CJ, Uzel G, Rosenzweig SD (2019) PI3K pathway defects leading to immunodeficiency and immune dysregulation. J. Allergy Clin Immunol 143: 1676–1687. https://doi.org/10.1016/j.jaci.2019.03.017

Oberli A, Popovici V, Delorenzi M, Baltzer A, Antonov J, Matthey S, Aebi S, Altermatt HJ, Jaggi R (2008) Expression profiling with RNA from formalin-fixed, paraffin-embedded material. BMC Med. Genomics 1: 9. https://doi.org/10.1186/1755-8794-1-9

Pécuchet N, Laurent-Puig P, Mansuet-Lupo A, Legras A, Alifano M, Pallier K, Didelot A, Gibault L, Danel C, Just PA, Riquet M, Le Pimpec-Barthes F, Damotte D, Fabre E, Blons H (2017) Different prognostic impact of STK11 mutations in non-squamous non-small-cell lung cancer. Oncotarget 8: 23831–23840. https://doi.org/10.18632/oncotarget.6379

Pennock ND, Jindal S, Horton W, Sun D, Narasimhan J, Carbone L, Fei SS, Searles R, Harrington CA, Burchard J, Weinmann S, Schedin P, Xia Z (2019) RNA-seq from archival FFPE breast cancer samples: molecular pathway fidelity and novel discovery. BMC Med Genomics 12: 1–18 https://doi.org/10.1186/s12920-019-0643-z

Pfaffl MW (2007) Real-time PCR. Tevfik Dorak M ed, pp 63–82. Taylor & Francis, New York, Abingdon. http://gmo-qpcr-analysis.com/dorak-book-real-time-pcr-2006.pdf#page=90

Pópulo H, Lopes JM, Soares P (2012) The mTOR signalling pathway in human cancer. Int J Mol Sci 13: 1886–1918. https://doi.org/10.3390/ijms13021886

Qiagen (2011) QuantiTect® Probe PCR Handbook. https://www.qiagen.com/jp/resources/download.aspx?id=3d850c15-ccb3-4ecf-843f-6bbdaeccc713&lang=en Qiagen (2011b) RT 2 PreAMP cDNA Synthesis Handbook. https://www.qiagen.com/us/resources/resourcedetail?id=a1282711-57b0-4411-b496-4500c54e148b&lang=en

Qiagen (2013) RNeasy® FFPE Handbook. https://www.qiagen.com/de/resources/resourcedetail?id=7b8ed707-c2d6-4b9e-a1ba-8893935622f9 Sample & Assay Technologies QIAGEN Sample and Assay Technologies

Qiagen (2015) RT² Profiler PCR Array (96-Well Format and 384-Well [4 x 96] Format) Human mTOR Signaling. http://6309039.s21d-6.faiusrd.com/61/ABUIABA9GAAg2eWH3QUo59jWhQE.pdf

Ribeiro-Silva A, Zhang H, Jeffrey SS (2007) RNA extraction from ten year old formalin-fixed paraffin-embedded breast cancer samples: a comparison of column purification and magnetic bead-based technologies. BMC Mol Biol 8: 118. https://doi.org/10.1186/1471-2199-8-118

Rubio T, Vernia S, Sanz P (2013) Sumoylation of AMPKbeta2 subunit enhances AMP-activated protein kinase activity. Mol Biol Cell 24: 1801–1811. https://doi.org/10.1091/mbc.E12-11-0806

Sadremomtaz A, Mansouri K, Alemzadeh G, Safa M, Rastaghi AE, Asghari SM (2018) Dual blockade of VEGFR1 and VEGFR2 by a novel peptide abrogates VEGF-driven angiogenesis, tumor growth, and metastasis through PI3K/AKT and MAPK/ERK1/2 pathway. Biochim Biophys Acta Gen Subj 1862: 2688–2700. https://doi.org/10.1016/j.bbagen.2018.08.013

Sancak Y, Bar-Peled L, Zoncu R, Markhard AL, Nada S, Sabatini DM (2010) Regulator-Rag complex targets mTORC1 to the lysosomal surface and is necessary for its activation by amino acids. Cell 141: 290–303. https://doi.org/10.1016/j.cell.2010.02.024

Sobočan M, Smolle MA, Schatz C, Haybaeck J (2020) The interplay of tumor stroma and translational factors in endometrial cancer. Cancers 12: 2074. https://doi.org/10.3390/cancers12082074

Sun Y, Fang Y, Yoon MS, Zhang C, Roccio M, Zwartkruis FJ, Armstrong M, Brown HA, Chen J (2008) Phospholipase D1 is an effector of Rheb in the mTOR pathway. Proc Natl Acad Sci U S A 105: 8286–829. https://doi.org/10.1073/pnas.0712268105

Szyszka P, Grossman A, Diaz-Cano S, Sworczak K, Dworakowska D (2016) Molecular pathways of human adrenocortical carcinoma — translating cell signalling knowledge into diagnostic and treatment options. Endokrynologia Polska 67: 427–450. https://doi.org/10.5603/EP.a2016.0054

Takeda H, Nishikawa H, Osaki Y (2018) The new era of precision medicine in hepatocellular carcinoma: the urgent need for promising biomarkers. Hepatobiliary Surg Nutr 7: 490–491. https://doi.org/10.21037/hbsn.2018.08.06

Taylor DR, Ghataore L, Couchman L, Vincent RP, Whitelaw B, Lewis D, Diaz-Cano S, Galata G, Schulte KM, Aylwin S, Taylor NF (2017) A 13-Steroid serum panel based on LC-MS/MS: use in detection of adrenocortical carcinoma. Clin Chem 63: 1836–1846. https://doi.org/10.1373/clinchem.2017.277624

Tian T, Li X, Zhang J (2019) mTOR signaling in cancer and mTOR inhibitors in solid tumor targeting therapy. Int J Mol Sci 20: 755 https://doi.org/10.3390/ijms20030755

Turnbull AK, Selli C, Martinez-Perez C, Fernando A, Renshaw L, Keys J, Figueroa JD, He X, Tanioka M, Munro AF, Murphy L, Fawkes A, Clark R, Coutts A, Perou CM, Carey LA, Dixon JM, Sims AH (2020) Unlocking the transcriptomic potential of formalin-fixed paraffin embedded clinical tissues: comparison of gene expression profiling approaches. BMC bioinformatics 21: 1–10. https://doi.org/10.1186/s12859-020-3365-5

Vatrano S, Volante M, Duregon E, Giorcelli J, Izzo S, Rapa I, Votta A, Germano A, Scagliotti G, Berruti A, Terzolo M, Papotti AM (2018) Detailed genomic characterisation identifies high heterogeneity and histotype -specific genomic profiles in adrenocortical carcinomas. Mod Pathol 31: 1257–1269. https://doi.org/10.1038/s41379-018-0042-6

Visuttijai K, Pettersson J, Mehrbani Azar Y, van den Bout I, Örndal C, Marcickiewicz J, Nilsson S, Hörnquist M, Olsson B, Ejeskär K, Behboudi A (2016) Lowered expression of tumor suppressor candidate MYO1C stimulates cell proliferation, suppresses cell adhesion and activates AKT. PLoS One 11: e0164063. https://doi.org/10.1371/journal.pone.0164063

Weddell JC, Chen S, Imoukhuede PI (2018) VEGFR1 promotes cell migration and proliferation through PLC gamma and PI3K pathways. NPJ Syst Biol Appl 4: 1. https://doi.org/10.1038/s41540-017-0037-9

Zhang P, Lehmann BD, Shyr Y, Guo Y (2017) The utilization of formalin fixed-paraffin-embedded specimens in high throughput genomic studies. Int J Genomics 2017: 1926304. https://doi.org/10.1155/2017/1926304

Zhang L, Thapa I, Haas C, Bastola D (2019) Multiplatform biomarker identification using a data-driven approach enables single-sample classification. BMC Bioinformatics 20: 1–13. https://doi.org/10.1186/s12859-019-3140-7