Statistical modeling of real time PCR data for membrane transporter expression in Manduca sexta larvae


Meeting Abstract

P3.93  Tuesday, Jan. 6  Statistical modeling of real time PCR data for membrane transporter expression in Manduca sexta larvae. WOODS, K.M.*; DOWNER, A.N.; COWGILL, J.A.; GILLEN, C.M.; HARTLAUB, B.A.; Kenyon College, Gambier, OH; Kenyon College, Gambier, OH; Kenyon College, Gambier, OH; Kenyon College, Gambier, OH; Kenyon College, Gambier, OH hartlaub@kenyon.edu

We examined the relationship between midgut transporter expression and larval instar in the tobacco hawkmoth Manduca sexta, testing the hypothesis that increased expression of midgut transporters compensates for reduced surface area to volume ratio as larvae grow. We used real-time PCR to investigate expression of two transporters: masBSC, a cation-chloride cotransporter, and KAAT, a potassium amino acid cotransporter. masBSC mRNA was detected in all tested tissues including foregut, midgut, hindgut, salivary gland, and Malpighian tubule, with no significant differences in expression level between tissues. KAAT mRNA was also detected in all tested tissues with approximately 500 fold higher expression in middle and posterior midgut than other tissues. Midgut expression of both genes was increased by 2 to 9 fold in the fifth compared to fourth instar larvae. We compared standard parametric methods (t tests and ANOVA procedures) with Mann-Whitney-Wilcoxon and Kruskal-Wallis tests, competing nonparametric procedures. With the exception of a few cases, the parametric and nonparametric tests yielded similar test statistics resulting in no change in significance. Ansari-Bradley tests showed no significant differences in the variability between instars. Sensitivity analyses showed that outliers did not have a significant effect on our inferences. For masBSC, multiple linear regression models explained approximately 89 percent of the variability in Ct. Similar models were fit and analyzed for KAAT, using Ct and delta Ct as response variables. The multiple regression model for KAAT explains approximately 96 percent of the variability in delta Ct, with significant instar, tissue, and interaction effects. (This project was supported by Kenyon College.)

the Society for
Integrative &
Comparative
Biology