**3. Preparing FFPE tissue sections for nucleic acids extraction**

All samples included in the studies conducted by our team result from biopsies performed in diffuse large B cell lymphoma patients, from lymph node samples or from reaction amyg‐ dala samples stored in the Pathology Department of Hospital das Clínicas, Medical School of University of São Paulo. All the herein used samples were formalin-fixed and paraffinembedded (FFPE) according to the standard methods described in the literature.

Four 20-µm thick cuts were performed in each sample using routine histological techniques. The sections were placed into 1.5 mL RNase- and DNase-free microtubes, and they were subsequently subjected to RNA and DNA extraction processes.

The first protocol used to prepare the cuts in the nucleic acid extraction process (suggested for the majority of commercial kits and protocols) consisted of deparaffinizing the sections with xylene and of rehydrating them with ethanol. In order to do so, 1 mL xylene PA was added to each sample (Synth® Diadema, SP, Brazil), and it was followed by homogenization using Vortex Genie 2T (Scientific Industries, Inc., Bohemia, NY, USA) and by incubation at 50°C, for 5 min, in digital thermomixer (Eppendorf AG, Hamburg, Germany). After incuba‐ tion, the samples were centrifuged at maximum speed for 5 min in the R-5418 microfuge (Eppendorf AG, Hamburg, Germany). The xylene was discarded and the cell button was washed two times with absolute ethanol (Merck KGaA, Darmstadt, Germany). The superna‐ tant was discarded after each wash. After the cell button was completely dried, the extrac‐ tion process started, as it is described in the following sections.

Moreover, even after some RNA extraction methods that allowed finding some viable op‐ tions were compared, it was observed that not all the extracted samples showed successful amplification in PCR reactions [17].

None of the previous studies available in the literature described the possible factors that could influence the amplification success. Therefore, the current study made the option of investigating some of the potential interferences in the nucleic acid obtainment process, namely: tissue fragment size, blocks' storage time, used fixative type, different cDNA syn‐ thesis primers and different primer sequences, among others [18].

After the aforementioned study, the tissue preparation protocol for RNA extraction process was modified by including a washing step. It consisted of using 1 ml phosphate buffer sal‐ ine (PBS) with 5 min incubation at room temperature, followed by full speed centrifugation for 5 min in the R-5418 microfuge (Eppendorf AG, Hamburg, Germany). This procedure is done to remove possible fixative residues that could work as PCR inhibitors [18]. Next, the same process was used in the DNA extraction performed by our team.

Such protocol change was done under the assumption that the amplification failure in PCR reactions could be caused by the presence of contaminants such as fixative waste working as PCR inhibitors.

The results show that the PBS washing step inclusion in the samples' extraction preparation process led to some statistically significant advantages such as the obtainment of better RNA concentration results (*p* = 0.00025), even when the same initial amount of tissue was used. In addition, the washing step allowed obtaining better sample purity levels (*p* = 0.0000001), increasing the samples amplification success (*p* = 0.018) both in the standard and in the real-time PCR reactions [18].

Two possible factors may have influenced the improved amplification efficiency of these samples. The first is based on the fact that formalin is water-soluble, thus the PBS washing step may have led to tissue-fixation residues' solubilization and removal, and it could possi‐ bly work as PCR inhibitors. Furthermore, previous studies suggested that pH values be‐ tween 6.5 and 9.0 are optimal for amplification [11, 9]. Then, the PBS solution may have possibly altered pH levels, thus increasing the quality of the obtained RNA and the amplifi‐ cation success.

Despite the influence of the PBS-based washing step addition on the samples' successful am‐ plification, we observed the fixative interference in this process too. Formalin-fixed samples showed more successful PCR amplification reactions than those fixed in formaldehyde or in Bouin's solution, even after further PBS washing (*p* = 0.000018) [18].

Another fact observed by our team refers to the paraffin-embedded tissue fragment size. All fragments size equals to or greater than 1.0 cm had the most successful samples' amplifica‐ tion (*p* = 0.034) [18]. This may be possibly due to the fact that smaller tissues may increase the fixative absorption and, consequently, cause greater nucleic acids degradation and chemical modification in the full extent of the tissue.

No impact was found on the amplification success when the other variables suggested in our study were tested (tissue type, block age, different primers used in the cDNA synthesis or the endogenous genes used in the PCR reaction), regardless of the use (or not) of the PBS washing step [18].

This protocol change was kept in the preparation of samples subjected to DNA extraction, due to the statistically significant impact caused by the PBS washing step inclusion on either the obtainment of better RNA concentrations or purity relations.

It is possible to conclude that this sample preparation is an essential step to obtain better quality nucleic acids for molecular biology studies.
