Elevation of sulfatides in ovarian cancer: an integrated transcriptomic and lipidomic analysis including tissue-imaging mass spectrometry

Abstract

Background: Sulfatides (ST) are a category of sulfated galactosylceramides (GalCer) that are elevated in many types of cancer including, possibly, ovarian cancer. Previous evidence for elevation of ST in ovarian cancer was based on a colorimetric reagent that does not provide structural details and can also react with other lipids. Therefore, this study utilized mass spectrometry for a structure-specific and quantitative analysis of the types, amounts, and tissue localization of ST in ovarian cancer, and combined these findings with analysis of mRNAs for the relevant enzymes of ST metabolism to explore possible mechanisms.

Results: Analysis of 12 ovarian tissues graded as histologically normal or having epithelial ovarian tumors by liquid chromatography electrospray ionization-tandem mass spectrometry (LC ESI-MS/MS) established that most tumor-bearing tissues have higher amounts of ST. Because ovarian cancer tissues are comprised of many different cell types, histological tissue slices were analyzed by matrix-assisted laser desorption ionization-tissue-imaging MS (MALDI-TIMS). The regions where ST were detected by MALDI-TIMS overlapped with the ovarian epithelial carcinoma as identified by H & E staining and histological scoring. Furthermore, the structures for the most prevalent species observed via MALDI-TIMS (d18:1/C16:0-, d18:1/C24:1- and d18:1/C24:0-ST) were confirmed by MALDI-TIMS/MS, whereas, a neighboring ion(m/z 885.6) that was not tumor specific was identified as a phosphatidylinositol. Microarray analysis of mRNAs collected using laser capture microdissection revealed that expression of GalCer synthase and Gal3ST1 (3'-phosphoadenosine-5'-phosphosulfate:GalCer sulfotransferase) were approximately 11- and 3.5-fold higher, respectively, in the ovarian epithelial carcinoma cells versus normal ovarian stromal tissue, and they were 5- and 2.3-fold higher in comparison with normal surface ovarian epithelial cells, which is a likely explanation for the higher ST.

Conclusions: This study combined transcriptomic and lipidomic approaches to establish that sulfatides are elevated in ovarian cancer and should be evaluated further as factors that might be important in ovarian cancer biology and, possibly, as biomarkers.

Figure 1

Differences in the level of expression of genes for sphingolipid biosynthesis through sulfatides, ST, and the amounts of ST in epithelial ovarian tumors versus normal ovarian stromal tissues. (A) The relative levels of mRNA for the enzymes of the shown steps of the de novo sphingolipid biosynthesis pathway were imported into a KEGG style pathway heatmap [15] with fold differences in gene expression for ovarian cancer/normal stromal cells are represented by the color scale (red = higher; blue = lower). In this depiction, the pathway begins in the upper left corner with the condensation of serine and palmitoyl-CoA by serine palmitoyltransferase (SPT, shown for its three known genes) to form 3-ketosphinganine (3KSa) which is converted to sphinganine (Sa) by 3KSa reductase. Sa is N-acylated to dihydroceramide (DHCer) by a family of Cer synthases (CerS), desaturated (by DHCer desaturases, DES), and converted to sphingomyelin (SM) by SM synthases (SMS), phosphorylated to Cer 1-phosphate by Cer kinase (CERK) (this branch of sphingolipid biosynthesis is also thought to involve a Cer transport protein, CERT), or glycosylated to glucosylceramide (GlcCer) or galactosylceramide (GalCer), which can undergo sulfation to ST by GalCer sulfotransferase (Gal3ST1). Also shown are the steps of ST turnover, via aryl sulfatase (ARSA) and galactocerebrosidase (GALC) which operate in concert with prosaposins (PSAP), and GlcCer turnover via glucosylcerebrosidase (GBA1). The heat maps depict the normalized level of expression of these genes in the epithelial ovarian tumors divided by the level of expression on normal stromal tissues averaged for the 12 patients and 8 controls. (B and C) Amounts of the various N-acyl-chain length subspecies of ST as measured by LC ESI-MS/MS for clinically similar patients with ovarian carcinoma (B, n = 12) or normal ovary (C, n = 12).

Figure 2

Individual variation in the microarray data for the synthases for ST and GalCer in ovarian carcinoma versus normal stromal tissue. Each letter represents the microarray data from an individual subject. The relative mRNA levels for the synthases for ST (Gal3ST1) and GalCer were analyzed for normal human ovarian stromal tissues (n = 8) and epithelial ovarian tumor cells (n = 12) using Affymetrix HG U133 Plus 2 Gene Chips.

Figure 3

A representative analysis of a histological thin section from an ovarian carcinoma sample by MALDI tissue imaging mass spectrometry (MALDI TIMS). (A) A thin section (10 μm) with H & E staining. Regions distinguished by "a" were histologically identified as non-malignant stroma, "b" as serous papillary epithelial carcinoma, and "*" are regions where there were no cells. The spots represent the sites where the MALDI TIMS spectra were collected (i.e., each spot represents the region targeted by the laser to generate one spectrum) with the regions histologically scored as serous papillary epithelial carcinoma in black; blue spots were from regions scored as stromal cells; the red dots are regions where there were no cells or the histology was undefined. (B and C) Spectra obtained from the shown regions of the adjacent thin section (10 μm). (D) Ion intensities of the m/z 778.6 (d18:1/C16:0 ST) species from regions scored as normal (n = 23) or ovarian cancer (n = 20) (P = 0.00003, by one-tailed Wilcoxon rank sum test).

Figure 4

Identification of ions from MALDI TIMS by precursor ion fragmentation and product ion scan. A thin section from the ovarian tumor in Fig. 3 was analyzed using a hybrid quadrupole time-of-flight mass spectrometer (ABI Q-STAR) in negative ionization mode with selection of the shown parent ions for fragmentation and analysis of the product ion spectra. (A and B) Product ion spectrum for each compound (m/z 778.6 ion was identified as a d18:1/C16:0 ceramide monohexosylsulfatide and m/z 885.6 was most consistent with the shown phosphatidylinositol).

Figure 5

Comparison of ion intensities for ST in non-malignant stroma or carcinoma as identified by H & E staining. (A) Regions of H & E staining thin section from Figure 3. The spots represent the sites where the MALDI TIMS spectra were collected in adjacent sections. (B) Distribution of the intensities of the ions (m/z 778.6 (d18:1/C16:0 ST)) in stromal regions and cancerous regions.

Figure 6

Visualization of the localization of ST in a thin section of ovarian carcinoma tissue using MALDI TIMS. (A) H & E stained thin section from Figure 3. Distinctive features in (A) have been manually traced with dashed lines, which have been superimposed on (B-D) to aid in comparison of the distribution of the ions with the H & E stained thin section. (B-D) Pseudo-color ion images where the relatively intensity of the labeled m/z 778.6 (B), m/z 888.6 (C), and m/z 890.6 (D) using the heatmap scale.

Figure 7

Difference in the level of expression of genes for sphingolipid biosynthesis through ST of human ovarian normal surface epithelial cells and epithelial tumor cells. The relative mRNA levels were analyzed for normal human ovarian surface epithelial cells (n = 12) and epithelial ovarian tumor cells (n = 12) using Affymetrix HG U133 Plus 2 Gene Chips. The fold differences of mRNA level for the shown steps of the de novo sphingolipid biosynthesis pathway were imported into a KEGG style pathway heatmap. (A) Fold difference in gene expression for ovarian epithelial carcinoma/normal epithelial cells are represented by the color scale (red = higher; blue = lower). (B) Relative mRNA levels for Gal3ST1 and GalCer synthase in each of the normal ovarian surface epithelial cells (n = 12) and ovarian epithelial carcinoma (n = 12).

Figure 8

Visualization of phosphatidylinositol and ST in a thin section of normal ovarian tissue using MALDI TIMS. Adjacent thin sections of normal ovarian tissue was prepared as described in the text for H & E staining (A) and pseudo-color ion images for phosphatidylinositol (B) as a positive control (m/z 885.6 at the instrument calibration used for this image) and ST (at m/z 778.6, C; m/z 888.6, D; and m/z 890.6, E) using the shown heat map scale. The ovarian tissue in (A) has been manually traced with dashed lines, which have been superimposed on panels B-E to aid in comparison.