Comparing and Optimizing RNA Extraction from the Pancreas of Diabetic and Healthy Rats for Gene Expression Analyses

Abstract

Advanced differential gene expression analysis requires high-quality RNA. However, isolating intact pancreatic RNA is challenging due to abundant pancreatic ribonucleases, which limits efficient downstream gene expression analysis. RNAlater treatment reduces endogenous ribonucleases effects through either pre-organ excision via organ mass or bile duct direct injection or organ mass injection post-isolation. We compared RNA extraction protocols to establish a reproducible and effective pancreatic RNA extraction method to obtain high RNA integrity number (RIN) values from healthy and streptozotocin (STZ)-induced diabetic rats for gene expression analyses. Different methods were tested focusing on RNase activity inhibition using RNAlater (Qiagen) pre-harvest of the pancreatic tissue, and extracted RNA quality and concentration were analyzed using NanoDrop spectrophotometer, Agilent Bioanalyzer, and RT-PCR. Inclusion of several pre- and post-excision modifications in the RNeasy Mini Kit (Qiagen) protocol resulted in RIN values more than two-fold higher compared to those using the standard protocol. Additionally, RT-PCR amplification of the housekeeping gene, β-actin, revealed no differences in extracted RNA quality from healthy and STZ-induced diabetic rats. We compared and developed a more effective and reproducible pancreatic RNA extraction method from healthy and diabetic rats, which resulted in RNA of superior quality and integrity and is suitable for complex molecular investigations.

Keywords: RIN value; RNA extraction; RNA isolation; STZ-diabetic rat model; pancreas.

Figure 1

RNAlater (Qiagen) injection into the bile duct (BD). (A) The liver was flipped over and pushed aside to expose the pancreas (P). (B) The BD was located and traced distally to its point of joining with the pancreatic duct (hepatopancreatic duct: HPD) at the sphincter of Oddi and occluded with a clamp at the clamp site (CS). RNAlater (Qiagen) was injected by inserting a diabetic syringe needle into the BD. (C) To confirm the effectiveness of the RNAlater (Qiagen) perfusion technique, Indian ink was injected into a test pancreas, showing good perfusion. The diagrams in the figure were generated using mindthegraph.com. Accessed on 10 December 2021.

Figure 2

The optimized RNA extraction protocol workflow. The diagrams in the figure were generated using mindthegraph.com, accessed on 21 April 2022. RIN, RNA integrity number.

Figure 3

Comparison of RNA integrity number (RIN) values of pancreatic RNA samples from healthy rats following different excision techniques and extraction protocols demonstrated as (A) a heat map with the three RIN values of each protocol reported inside boxes and (B) a bar graph showing the significant difference in RNA integrity between the optimized and standard protocols using the standard one-way analysis of variance (ANOVA) for RNA analysis followed by Fisher’s LSD test. ns, non-significant (p > 0.05) or significant at ** p < 0.01, *** p < 0.001.

Figure 4

Electropherograms generated by an Agilent 2100 Bioanalyzer using the pico kit showing (A) RNA extracted using a standard protocol with an RIN = 3.6 and high peaks in the 5S and fast region, which indicates RNase degradation as compared with (B) RNA extracted using our optimized protocol that had low peaks in the same regions with two clear 18S and 28S regions, indicating intact RNA with RIN = 8.8.

Figure 5

Virtual gel-like images generated by an Agilent 2100 Bioanalyzer using the pico and nano kits. (A) There was a high level of RNA degradation in samples extracted using the standard RNeasy mini protocol (Qiagen). Samples isolated from diabetic rats (DR) are presented in lanes 1–3, which had an RIN range from 4.2–7, while those from normal rats (NR) in lanes 7–11 had an RIN value between 2.3–7.3. Red highlighted lanes 4–6 show samples with low RIN values that could not be calculated due to degradation or sample contamination. (B) Intact RNA with very low levels of degradation in samples extracted using our optimized method. Samples isolated from diabetic rats (DR) are presented in lanes 1–7, which had an RIN value range from 6.7–8.8, while those from normal rats (NR) in lanes 8–12 had an RIN value ≥8.0 except for NR2 and NR3 (RIN = 7.4, 7.2), which was due to a technical difficulty during the homogenization step.

Figure 6

The quantity (ng/µL), quality (absorbance ratio A260/A280 nm), and integrity (RIN) of samples extracted from (A) normal rats (NR) (n = 6) and (B) diabetic rats (DR) (n = 6) using our optimized protocol. The dashed horizontal line represents the mean RIN.

Figure 7

Agarose gel electrophoresis of β-actin amplicons (289 bp) produced from cDNA that was synthesized from RNA extracted from rat pancreas isolates. RNA was extracted from the pancreas of normal rats using the standard RNeasy Mini Kit protocol (Qiagen) (lanes 1–3); RNAlater (Qiagen) injection without clamping and treatment with 5 mL QIAzol (Qiagen) (lanes 4 and 5); clamping and RNAlater (Qiagen) injection followed by the optimized protocol with 7 mL QIAzol (Qiagen) (lanes 6 and 7). As a positive control, RNA was extracted from rat liver using the standard protocol (lane 8). β-actin was also amplified from cDNA that was synthesized from RNA isolated from pancreas of diabetic rats using clamping and RNAlater (Qiagen) injection followed by the optimized protocols with 5 mL QIAzol (Qiagen) (lanes 9–11), or from that of normal rat pancreas samples after clamping and RNAlater (Qiagen) injection using the optimized protocol with 5 mL QIAzol (Qiagen) (lanes 12–14). A 100-bp DNA size marker (Agilent Technologies, Santa Clara, CA, USA) and a 123-bp DNA ladder (left) were used as references.