High-quality total RNA extraction from early-stage lamprey embryos

Obtaining high-purity total RNA from animal tissues is important for next-generation sequencing (NGS)-based transcriptomics and high yields of RNA are also crucial for single-cell RNA-seq analyses. Lampreys are one of two extant jawless vertebrates (cyclostomes) that, along with hagfish, diverged from gnathostomes in the earliest period of vertebrate evolution. Therefore, the early evolutionary changes of the vertebrate body plan can be explored by comparing cyclostomes to gnathostomes [1]. Lampreys are not laboratory animals, such as mice or zebrafish, but obtaining their eggs is relatively easy. In the mating season (in the northern hemisphere, March–June), more than 10,000 eggs can be obtained from a single mature female via artificial spawning [2]. Embryos can be raised in the laboratory in plastic dishes with 10% Steinberg's solution until the ammocoetes larval stage (Figure 1A). Eggs initially have an adhesive gel mass on the vegetal side of the egg membrane surface, which anchors them to the sand at the river bottom and then the gel mass disappears during the early cleavage stage [3]. Additionally, lamprey eggs have a double chorion [4], allowing them to be collected in 2-ml tubes for RNA extraction and for water to be removed without destroying them (Figure 1B). Hatching enzymes then weaken the egg membrane before hatching, which occurs at Tahara's stage 24: the midpharyngeal stage (Figure 1G) [3]. Lamprey embryonic cells contain an abundance of yolk granules composed of the lipovitellin–phosvitin lipid–protein complex [5]. These yolk granules are consumed during embryonic development but can still be found in the endodermal cells of posthatched larvae.

Two major protocols derived from the acid guanidinium thiocyanate-phenol-chloroform (APGC) extraction method are generally used for total RNA extraction from cells or tissues [6]. The first involves centrifugation with isopropanol after separating protein/DNA and RNA using a solution containing guanidine thiocyanate and acid phenol (including the commercially available TRIzol reagent [Thermo Fisher, MA, USA], ISOGEN [Nippon Gene, Japan] and QIAzol [QIAGEN, MD, USA]). The second protocol uses a silica membrane to bind RNA in the presence of chaotropic ions, following the same removal protein and DNA as the first protocol (commercial kits include the RNeasy kit, QIAGEN).

However, neither of these protocols works well for extracting RNA from early-stage lamprey embryos. The isopropanol precipitation method has abundant contamination in the final RNA solution, and low optical density (OD) 260/280 and 260/230 ratios (<1.5 and <1.0, respectively). These contaminants cannot be removed by extra phenol/chloroform extraction or additional alcohol precipitation. In addition, the RNA integrity number (RIN) value of RNA extracted from early-stage embryos tends to be low with this method [7]. For transcriptome analysis of mRNA, it is generally necessary to use total RNA with a RIN score >8 to avoid bias toward the 3′ end region of the mRNA (see TruSeq Stranded mRNA reference guide, Illumina^®). In contrast, the method using silica filters yields significantly less RNA (e.g., <10 ng/100 eggs). This may be due to contaminants preventing the binding of RNA molecules to silica. To date, there have been very few published studies on transcriptome analysis using early-stage lamprey embryos [8]. Additionally, while single-cell transcriptome analysis has been reported for the spinal cord and brain of ammocoete larvae [9,10], to our knowledge, there are none utilizing early-stage embryos. This may be attributed to the challenges associated with extracting RNA from early-stage embryos of lamprey with high quality and high efficiency. To overcome these problems, the QIAzol/isopropanol precipitation protocol was modified to improve the yield and purity of extracted total RNA as follows (step-by-step procedure available in Supplementary Appendix 1).

Embryos at different stages (stages 2, 16, 20 and 24; Figure 1D–G) of the Arctic lamprey, Lethenteron camtschaticum, were collected in 2-ml microtubes (Figure 1B, up to 500 μl) and homogenized with QIAzol reagent and stainless-steel beads. The homogenized solution was divided equally into three groups to evaluate the protocol. The first group was treated using the conventional APGC method (called ‘Normal’ in Figures and Tables) following the QIAzol Lysis Reagent instruction manual (QIAGEN), the second using a silica membrane filter according to RNeasy Lipid Tissue Mini Kit (QIAGEN), and the third group was treated using the modified precentrifugation and salt addition protocol (Supplementary Appendix 1).

Briefly, the modified protocol began with centrifuging the homogenized solution at 12,000 g for 15 min at 4°C and then transferring the solution while avoiding the lipids floating on the surface and the precipitates. Chloroform was then added to the solution, followed by 250 μl of a mixed salt solution composed of 0.8 M sodium citrate and 0.2 M sodium chloride for every 1 ml of QIAzol solution, and then half the usual volume of isopropanol (250 μl for every 1 ml of QIAzol). Although the solution becomes cloudy at this step in the normal protocol, it remains transparent in this protocol (Figure 1D), suggesting that precipitation is impeded by the high concentration of salts. The subsequent steps follow the normal protocol, with the final elution in RNase-free water. The results of the three protocols are summarized in Table 1. The normal precipitation method produced the highest amount of total RNA, but the absorbance peak shifted from 260 nm with a large peak observed at approximately 230 nm, resulting in extremely low 260/280 and 260/230 ratios, indicating contamination by proteins and other substances (Figure 2). Conversely, the RNeasy kit produced significantly lower yields at stages 2, 16 and 20 (Table 1). Bioanalyzer analysis showed that the RNeasy samples do not produce an rRNA peak until stage 20, resulting in an unmeasurable RIN (Figure 3). The modified protocol, with precentrifugation and the addition of salt, greatly improved the total RNA yield and quality (Table 1). Moreover, the highest RIN values, indicating RNA integrity, were obtained by using the modified method. As mentioned, high RIN values are crucial to obtain unbiased results in transcriptome analysis. This method yields significantly higher RNA recovery than the filter-based method, making it applicable for RNA extraction in single-cell analysis. Therefore, this method is expected to be useful for future single-cell transcriptome analysis in lamprey embryos.

The protocol for alcohol precipitation in the presence of high amounts of salts was used for total RNA extraction from plants containing proteoglycans and polysaccharides and RNA extraction from mouse liver, where these substances are abundant [11]. As shown here, this method can also improve the purity and yield of extractions from early-stage lamprey embryos. However, bioanalyzer results show that very little RNA of less than 200 bp was extracted using this method (Figure 3) and thus, the method should be avoided in small RNA research. In the presence of chaotropic salts, DNA or RNA can form hydrogen bonds with the -OH on the silica membrane. The reason RNA from lamprey eggs are not extracted by conventional methods (e.g., RNeasy kit) was likely due to the interference of contaminants that are not removed by guanidine-phenol extraction and subsequently inhibited the binding between RNA and silica in the following steps. These contaminants were present in the final product obtained by precipitation methods and inhibited RNA binding in filter-based methods. Although we did not investigate the identity or origin of these contaminants or the specific region of the egg from which they originated, it is certain that their influence decreases as development progresses, as shown in the Figure 2. Although we have not investigated the identity of the substance responsible for contaminants or its specific origin within the egg, it is evident from the results that the level of contamination decreases as development proceeds (Figure 2). Interestingly, at stage 24, RNA can be extracted successfully by any of these methods. Indeed, the reduction of yolk granules in the cells throughout development might also decrease the amount of lipids during stages 20–24, but the most significant developmental change during this period is hatching [3]. Thus, it is highly likely that the contaminants in the early stages are derived from the egg membrane. Additionally, the presence of highly adhesive substances that facilitate adherence of the egg to the river sand and double chorions may also impede RNA extraction. Although the manual removal of egg membranes from hundreds of eggs using forceps has the potential to reduce contamination, it is not a practical procedure during the rapid early stages of embryonic development in lampreys. Thus, in the case of early lamprey embryos, this new method is considered the most effective solution. This study was conducted only on the Arctic lamprey, L. camtschaticum, but is probably applicable to all other lamprey species since they also spawn at the river bottom in the same manner.