An exfoliation and enrichment strategy results in improved transcriptional profiles when compared to matched formalin fixed samples

Abstract

Background: Identifying the influence formalin fixation has on RNA integrity and recovery from clinical tissue specimens is integral to determining the utility of using archival tissue blocks in future molecular studies. For clinical material, the current gold standard is unfixed tissue that has been snap frozen. Fixed and frozen tissue however, both require laser capture microdissection to select for a specific cell population to study. The recent development of a sampling method capable of obtaining a viable, enriched cell population represents an alternative option in procuring cells from clinical material for molecular research purposes. The expression profiles of cells obtained by using this procurement approach, in conjunction with the profiles from cells laser capture microdissected from frozen tissue sections, were compared to the expression profiles from formalin fixed cells to determine the influence fixation has on expression profiles in clinical material.

Methods: Triplicate samples of non-neoplastic colonic epithelial cells were recovered from a hemicolectomy specimen using three different procurement methods from the same originating site: 1) an exfoliation and enrichment strategy 2) laser capture microdissection from formalin fixed tissue and 3) laser capture microdissection from frozen tissue. Parameters currently in use to assess RNA integrity were utilized to assess the quality of recovered RNA. Additionally, an expression microarray was performed on each sample to assess the influence each procurement technique had on RNA recovery and degradation.

Results: The exfoliation/enrichment strategy was quantitatively and qualitatively superior to tissue that was formalin fixed. Fixation negatively influenced the expression profile of the formalin fixed group compared to both the frozen and exfoliated/enrichment groups.

Conclusion: The exfoliation/enrichment technique represents a superior alternative in tissue procurement and RNA recovery relative to formalin fixed tissue. None of the deleterious effects associated with formalin fixation are encountered in the exfoliated/enriched samples because of the absence of its use in this protocol. The exfoliation/enrichment technique also represents an economical alternative that will yield comparable results to cells enriched by laser capture microdissection from frozen tissue sections.

Figure 1

Expression of ber-Ep4 in colon cells. Immunohistochemistry of section of non-neoplastic colon tissue demonstrating expression by all colonic epithelial cells to the ber-Ep4 antibody. Note absence of expression in cells in the lamina propria. The difference in expression is exploited in the enrichment strategy for colonic epithelial cells from fresh tissue. (10×).

Figure 2

Recovery of an enriched colon cell population using the exfoliation/enrichment strategy. Colonic epithelial cells devoid of inflammatory cells and other cell types recovered using the exfoliation and enrichment strategy on fresh tissue. The yellow dots are the immunomagentic beads (H&E, 20×).

Figure 3

Viability of recovered cells by the exfoliation/enrichment strategy. The absence of blue staining in the colonic epithelial cells recovered after using exfoliation/enrichment indicates the cells recovered by this approach are viable. The yellow dots are the immunomagnetic beads (40×).

Figure 4

Study design. From a freshly resected colectomy specimen, three types of procurement were performed from tissue originating from the same site. In the fresh state, cells in Group 1 are exfoliated and enriched for using ber-Ep4(+) magnetic beads. The area from where these cells came from was then sectioned into 2 mirror pieces, with one fixed in formalin and the other frozen. For the latter two, colon cells were obtained by laser capture microdissection. After exfoliation, RNA was extracted in parallel from all the samples and afterwards underwent expression profiling on an Affymetrix microarray chip. The results were compared between the samples within each group and against the samples in other groups. EE = exfoliation/enrichment, FFPE = formalin fixed, paraffin embedded, LCM = laser capture microdissection, EP = expression profile assay.

Figure 5

RNA integrity. Representative electropherograms of 18S and 28S ribosomal RNA peaks from each sample procurement group. 5A = Group 1, 5B = Group 2, 5C = Group 3. See text for details.

Figure 6

Reproducibility of expression arrays between replicates. Scatterplots of samples within each group to assess reproducibility of the procurement method. 6A = Representative scatterplot between two of the samples in Group 1. 6B = Representative scatterplot between two samples in Group 2. 6C = Representative scatterplot between two of the samples in Group 3.

Figure 7

Comparison of expression profile for selected genes between the three different procurement methods. Expression profile of selected genes between the 3 different methods show a greater degree of similarity between Groups 1 and 3 than either with Group 2. Redundant genes indicate the presence of multiple probes, with the genes in the lower half of the cluster diagram indicating degeneration at one end of the mRNA species for that gene.