Sequencing and validation of housekeeping genes for quantitative real-time PCR during the gonadotrophic cycle of Diploptera punctata

Abstract

Background: Quantitative RT-PCR (q-RT-PCR) is a powerful tool that allows for the large scale analysis of small changes in gene expression. Accurate and reliable results depend on the use of stable reference genes for normalization. However, the expression of some widely used housekeeping genes can vary under different experimental setups. To our knowledge, no validation studies have been reported for reference genes in cockroaches. The aim of the current study is the identification and validation of a set of eight housekeeping genes during the first gonadotrophic cycle of the cockroach, Diploptera punctata. This study made use of two different algorithms (geNorm and Normfinder) to evaluate the stability of gene expression.

Results: Candidate housekeeping genes were sequenced: β-actin (Actin), elongation factor 1 alpha (EF1a), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), armadillo (Arm), ribosomal protein L32 (RpL32), succinate dehydrogenase (SDHa), annexin IX (AnnIX) and α-tubulin (Tub). The expression of these eight genes was analyzed in corpora allata (CA) and ovaries of adult female D. punctata. Both geNorm, as well as Normfinder characterized SDHa, EF1a and Arm as being the most stably expressed in the corpora allata. In the ovary, the geNorm calculation showed Tub, EF1a and RpL32 to be most stable, whereas Normfinder identified Tub, EF1a and Arm as the best. In ovary, the least stable gene was Actin, challenging its usefulness in normalization. As a proof of principle, the expression of follicle cell protein 3c and CYP15A1 was monitored during the first gonadotrophic cycle.

Conclusion: Arm and EF1a form the most stably expressed combination of two reference genes out of the eight candidates that were tested in the corpora allata. Our results show that the combined use of Tub, EF1a and RpL32 ensures an accurate normalization of gene expression levels in ovary of D. punctata. Our study has indicated that neither Actin nor AnnIX should be used for normalization of transcript levels when studying the first gonadotrophic cycle in CA or ovary of D. punctata. The results stress the necessity for validation of reference genes in q-RT-PCR studies in cockroaches.

Figure 1

JH and oocyte measurements of the test animals. The right-hand axis shows the length of basal oocytes dissected from D. punctata females during the first gonadotrophic cycle (day 0 to day 6 after the imaginal molt, oviposition took place at the start of day 7). The data represent means of 30 individual animals. Vertical bars indicate standard deviations. The left-hand axis represents the amount of JH released during the first gonadotrophic cycle (day 0 to day 7 after final molt). The data are represented in pmol/hour per individual CA and represent means of 12 independent biological replicates. The vertical bars indicate standard errors.

Figure 2

Boxplots showing the Ct variation of the reference genes. The variation in Ct values of the eight candidate reference genes in (A) the CA samples (3 biological replicates for each time point, n = 24) and in (B) the ovary (3 biological replicates for each time point, n = 24) as indicated by the raw Ct values. The values are given as the cycle threshold (Ct, mean of triplicate samples). The boxplots show the 25th quartile, median and the 75th quartile (horizontal lines) and the minimal and maximal values (whiskers).

Figure 3

Stability values in the CA as generated by the two algorithms, A) geNorm and B) Normfinder. (A) The average expression stability values (AESM) from least stable (left) to most stable (right). Threshold for an unstable gene is M ≥ 1.5. (B) The expression stability values from the candidate reference genes calculated by the Normfinder software.

Figure 4

Stability values in the ovary as generated by the two algorithms, A) geNorm and B) Normfinder. (A) The average expression stability measure (AESM) from least stable (left) to most stable (right). Threshold for an unstable gene is M ≥ 1.5. (B) The expression stability values from the candidate reference genes calculated by the Normfinder software.

Figure 5

Optimal number of reference genes for normalization as calculated by geNorm. Pairwise variation analysis determining the optimal number of reference genes required ensuring accurate normalization between normalization factors NF_n and NF_n+1 in (A) CA and (B) ovary during the gonadotrophic cycle.

Figure 6

Graphical representation of the relative transcript level of the target gene A) follicle cell protein 3c measured in the ovary and B) CYP15A1 in the CA during the first gonadotrophic cycle of D. punctata. Measurements were taken every day during the cycle (day 0 to day 7 after final molt). The data represent means of three independent pools of ten animals, run in triplicate using q-RT-PCR. Fcp 3c was normalized to Tub, RpL32 and EF1α transcript levels or to Actin transcript levels. CYP15A1 was normalized to EF1α and Arm transcript levels or to AnnIX transcript levels. Vertical bars indicate S.E.M.