A review of room temperature storage of biospecimen tissue and nucleic acids for anatomic pathology laboratories and biorepositories

2.1 Dry chemical or polymer matrices can preserve DNA only or RNA only for years and potentially for decades

To some extent, desiccated chemical matrices mimic natural extremophile biology that allows organisms such as tardigrades or brine shrimp to protect cellular systems in a dried state and later revive via rehydration- a process known as anhydrobiosis [10]. In contrast to extremophiles which inclusively preserve cellular structure, nucleic acids, and protein, currently commercial chemical matrices permit DNA-only long-term (years) room temperature storage or RNA-only storage and not both simultaneously. For most medical center biorepositories, it may not be cost effective to prospectively extract DNA and RNA from all tissue biospecimens when only a subset of specimens may be used. These matrices may be best suited for specific situations such as storage of excess extracted nucleic acids, storage of specific samples with predicted high demand for nucleic acids, or back-up of frozen biospecimens.

There is little independent data, particularly long-term data, regarding most of the commercial matrices. The one matrix with extensive data is Flinders Technology Associates (FTA) filter paper, a macroporous cellulose matrix treated with uric acid, detergent, and chelating agent. Upon application of a biosample to FTA paper, cells are lysed and DNA is released [11–13]. Cell lysis results in loss of tissue morphology but the DNA binds to the paper and is preserved. In addition to the effect of chelating agents, the DNA is dried in order to slow water-dependent enzymatic activity. FTA technology can be used to store genomic, plasmid, whole blood, or intracellular DNA; however, the manufacturer does not recommend storage of RNA. Forensic and evolutionary biology studies indicate utility of FTA paper for PCR assays after greater than 7 years [14–16]. DNA from 60 surgical specimens stored on FTA cards for 49 days at room temperature performed equally compared to freshly extracted DNA when assessed by PCR and IgH gene rearrangement tests [12]. Formal studies of the utility of FTA paper samples for whole exome and whole genome sequencing is lacking. Storage formats, testing results, reusability features, maximum load specifications, and mechanism are summarized in Table 2.

Chemical matrices use various combinations of reagents that minimize or prevent oxidation, hydrolysis, and nuclease activity. The chemical matrix is typically layered in wells of a titer plate or tube. Most of the matrices are proprietary and only limited information is available as to their precise composition (See Table 2). Nucleic acids samples are added to the matrix then dried and stored. Addition of water or more commonly a buffer can elute the nucleic acids out of the matrices for use. Forensic buccal, blood, and semen DNA stored for 45-months (3.66 years) using Biomatrica DNAstable showed limited degradation compared to −20°C controls when assessed by real-time PCR [22]. Genomic DNA stored for 1 year using Biomatrica’s SampleMatrix® or at −20°C displayed equal DNA recovery and no detectable inhibition of amplification assessed by quantitative PCR, gel electrophoresis, and multiplex short tandem repeats analysis [20]. 336 RNA samples stored for 2 weeks and shipped in Biomatrica RNAstable, IntegenX GenTegra RNA, or −80°C were shown to be equally amenable to parallel RNA quality analyses and real-time PCR [28]. Room temperature DNA and RNA storage formats, testing results, reusability features, and maximum load specifications are summarized in Table 2 and Table 3 respectively. Peer reviewed studies that assess the utility of these chemical matrices for whole exome/genome sequencing and cDNA microarray analyses are highly desirable.

2.2 Stabilization Solutions may preserve whole tissues at room temperature up to 7 days and DNA for 6 months

One stabilization solution, QIAGEN’s AllProtect Tissue Reagent (QIAGEN, Valencia, CA, USA), preserves DNA, RNA, and protein in fresh tissue for up to 7 days at room temperature or one year in a refrigerator (see Table 1). The liquid reagent permeates cells, inactivates endogenous nucleases, and reduces microbial and reactive oxygen species activity but its composition is proprietary. AllProtect is of particular interest as specimens placed in AllProtect then frozen may potentially have short-term protection against freezer failure giving a few days to transfer samples safely to a working freezer. Five mice liver samples immersed in AllProtect before storage at −80°C (without snap-freezing) demonstrated better DNA/RNA quality than snap-frozen (without Allprotect) controls [29]. In contrast, 111 breast tissue samples immersed in AllProtect then snap-frozen and stored at −80°C did not show significant differences in DNA or RNA quality relative to fresh frozen controls [30]. Both studies measured DNA quality by A_260/280 outputs on Nanodrop 2000™ (Thermo scientific, Waltham, MA, USA) and 1.0% agrose gel electrophoresis, and RNA quality by Agilent 2100 Bioanalyzer^® (Agilent Technologies, Inc., Santa Clara, CA, USA) [20,21]. Tissue or blood cells can also be placed directly into Biomatrica’s DNAgard solution but this stabilizes DNA only though for 6 months at room temperature per the manufacturer’s data (see Table 4).

2.3 Current lyophilization (Freeze-drying) approaches can preserve DNA, RNA, and protein including enzymatic activity at room temperature albeit with some nucleic acid degradation

Lyophilization, also known as freeze-drying or cryodesiccation, is a three-step dehydration process [31]. First, the lyophilizer, a device capable of manipulating temperature and pressure, freezes extracellular and intracellular water within a sample into ice. Second, pressure is lowered and temperature is raised sufficiently under a vacuum to sublimate ice directly into water vapor removing 95% of cellular moisture. Third, elevated temperatures break physicochemical bonds binding residual water molecules to tissue leaving a final desiccated sample with residual water content of about 1–4% of original tissue moisture [32–33]. In our laboratory, brain tumor samples stored for one year at room temperature in the dark and in vacuum-sealed vials displayed modest DNA and moderate RNA degradation but were suitable for histology, immunohistochemistry, PCR, RT-PCR, gene methylation, copy number variation, Western blots, and enzymatic assays [34]. In a study by Matsuo and colleagues, lyophilized rat liver tissue stored for four years at ambient room temperature in nitrogen-filled vials demonstrated minimal degradation, but RNA degradation was seen near the one year time-point of storage [35]. Despite the RNA degradation, RT-PCR was successful. They noticed, as did we, variable degrees of vacuolization in the histological sections probably related to ice crystal formation but histologic details were generally of good quality. 140 lyophilized human colon tumor, breast tumor, kidney for epithelial tumors, lymph node tumor, and skin lymphomas for lymphoid tumors samples stored in the dark at room temperature for one year exhibited no marked DNA loss as assessed by a ND-1000 Spectrophotometer and 1% agrose gel electrophoresis, or RNA loss as evaluated with a RNA 6000 NanoAssay and qRT-PCR [36]. Additional work is required to optimize lyophilization protocols and storage conditions including whether additives such as antioxidants or RNase inhibitors may enhance stability of lyophilized tissues.

2.4 Fixation with formalin-free fixatives and paraffin embedding can affect nucleic acid quality

The search for formalin substitutes has been driven by two major factors: its potential carcinogenicity [6] and the fact that formalin fixation yields fragmented and cross-linked DNA and RNA [9]. At least seven alcohol-based, non-crosslinking, formalin-free fixatives-UMFIX (Sakura Finetek USA, Torrance, CA), FineFIX (Leica Biosystems, Buffalo Grove, IL), HOPE (Polysciences Inc, Warrington, PA), methacarn, modified methacarn, Paxgene Tissue (Qiagen, Valencia, CA), RCL2 (Alphelys SAR, Plaisir, France) are reported to have a satisfactory quality of histology and extracted macromolecules (nucleic acids and proteins) [37–39]. Fixative composition, mechanisms, and independent testing results are detailed in Table 5. Crosslinking and non-alcoholic fixatives generally do not perform as well as non-crosslinking and alcoholic fixatives [38,39]. Non-crosslinking fixatives do not require heat-induced epitope retrieval [38]. Of the fixatives in Table 5, modified methacarn is the favored fixative in six independent evaluations, though two others suggest that UMFIX and RCL2 are equal in performance [38,40]. Paxgene Tissue induces histological artifacts that outweigh its molecular protective benefits [41,42]. A recent study also implicates high temperatures during the paraffin embedding process as a cause of reduced RNA yields and poor RNA quality in FFPE tissues [43]. It is important to note that placing fresh tissue into a formalin-free fixative alone is insufficient to preserve nucleic acid integrity. Care must be taken to ensure that the tissue processor uses ethanol in lieu of formalin during its cycles. Further study of formalin-free fixatives taking into account the embedding and processing conditions may permit a better understanding of how and whether to integrate these fixatives into daily practice.

3.1 Room temperature storage has lower fiscal, labor, and environmental costs over the long-term

After relatively modest upfront costs, room temperature storage requires minimal space and basic overhead for maintaining room temperature. Conversely, the upfront costs of freezers are substantial. Outfitting a room to store 10 freezers may require appropriate air-conditioning and installing circuitry connected to back-up generators and dedicated lines for each freezer. Costs to outfit the room may be as much as 20 – 40,000 United States Dollars (USD). Additionally, a freezer has a life span and may have to be replaced after 10–15 years. Furthermore, there are expenses associated with space needs. For example, the National Neurologic AIDS Bank, Los Angeles, CA maintains multiple of its own freezers in leased space that is subject to periodic rent increases and also rents freezer space at a commercial biobank. Annual costs for frozen storage such as warranties, maintenance, electricity, back-up power sources, back-up carbon dioxide, or liquid nitrogen tanks accumulate over the long-term. It has been estimated that the one year storage cost for a lyophilized sample is 3 euros compared to 24 euros at −80°C and 31 euros in liquid nitrogen [36]. In a vendor sponsored study at Stanford University, it was estimated from 12 pilot laboratories that converting from freezer to room temperature storage would save 1.2 – 1.4 million USD per year [62]. While we take the absolute amount of savings with a grain of salt, we do believe that cost savings do accrue particularly the longer the more specimens are retained. In addition to the financial cost, the environmental cost of freezers is substantial. In comparison to the 4000 lbs CO₂ and 15,000 lbs CO₂ of a −80 °C and −150 °C freezer respectively, a typical household refrigerator ejects approximately 1,700 lbs CO₂ annually [63]. Currently, shipping of frozen biospecimens requires the use of dry ice, specially trained personnel to package the specimens, and high freight costs related to shipping a hazard (dry ice). If the biospecimen does not arrive in time, it will thaw completely. Room temperature preservation approaches avoids these issues [17,26].

3.2 Less space is required for room temperature storage

In a vendor-sponsored study conducted at Stanford University, storage in dry multi-well plates was estimated to reduce nucleic acid storage space requirements by about 70% over frozen storage [63]. Frozen tissues are stored in tubes, which is less spatially efficient than storing tissue derivatives on a high density plate at room temperature. A typical 96-well plate dimensions may be 127 mm by 86 mm by 8mm thick. In comparison, 96 purified DNA samples stored within standard 1.7ml microcentrifuge tubes (9.2 mm diameter, 38.7 mm length)(Sorenson Biosciences, Salt Lake City, UT, USA) aligned adjacently occupies approximately 883 mm by 883 mm by 39 mm- a substantially larger volume. For −150°C chest freezers where compressor and other mechanical requirements are particularly high, the actual storage space for containers may be as little as 17.3% of the total freezer volume (0.23 m³ storage capacity; 1.34 m³ total volume- Sanyo MDF-C2156VANC, Sanyo North America, Wooddale, IL, USA). In our experience, space needs for a freezer exceed the base footprint of the freezer as there is a need for adequate spacing in between units. To cool their contents, freezers generate substantial heat and packing large numbers of freezers too closely can result in chronic mechanical failures related to the stressor of elevated local temperatures. Liquid nitrogen tanks are also inefficient space-wise due to height limitations and the need for attached liquid nitrogen supply tanks.

4.1 Issues in replacing formalin fixation

Why haven’t formalin-free fixatives been more widely accepted in the pathology community? After all, the new fixatives offer potential advantages over FFPE. The higher cost of the fixatives is one impediment. Also, many current testing modalities such as PCR or RT-PCR work reasonably well with FFPE specimens reducing the impetus for change. In the United States, intra-laboratory validation for microscopy and routine testing is a major barrier. According to College of American Pathologists (CAP) guidelines, changes in biospecimen processing that can impact testing must be validated. To switch to a new formalin-free fixative, immunohistochemistry procedures for each antibody in a laboratory’s repertoire and diverse molecular assays would need to be revalidated [58]. Furthermore, formalin-free fixatives and lyophilization face the challenge of producing histology equal or superior to that of FFPE in the subjective eyes of demanding pathologists. Earlier generation formalin-free fixatives Glyo-Fixx, STF-Streck, Omnifix, Histochoice, and Histofix were reported to have a lower quality of histology [65]. RCL2 is reported to allow good morphology, though pigment deposition, degranulation, and RBC lysis can occur [39]. FineFIX can lead to partial tissue disintegration during fixation, shrinkage artifacts, degranulation, and RBC lysis [39]. One review concludes however that there are relatively minor differences that do not interfere with establishing the correct diagnostic conclusion [66]. The growth of whole exome and whole genome sequencing as well as other next-generation molecular approaches that require a higher quality of nucleic acids could tip the balance towards formalin alternatives.

4.2 Issues in replacing frozen tissues

In terms of replacing frozen tissue, no modalities exist currently that permit simple room-temperature preservation of tissue while preserving histology, DNA, RNA, and protein for the long-term (years long). Most current modalities preserve DNA or RNA alone. In addition, existing published real time stability data for the commercial matrix technologies is fairly limited except for FTA paper. Nevertheless, the experience with FTA paper and commercial accelerated aging studies of newer matrices suggest room temperature stability in the order of decades is likely to be achievable. There are large provincial or national genetic repositories that use matrices in multi-well plates for room temperature DNA storage but these laboratories also have significant resources such as robotic high-throughput nucleic extraction that lower their cost for doing so. Lyophilization has an intriguing potential to provide a means to preserve all tissue components including enzymatic activity; however, further research on lyophilized specimens is needed to better stabilize nucleic acids for long term storage [34].