Colorectal Neoplasia Goes with the Flow: Prostaglandin Transport and Termination

The past 10 to 15 years have witnessed major advances in our understanding of polyunsaturated fatty acid metabolism involving prostaglandin (PG)--especially PGE₂--synthesis, activity and degradation in neoplasia, especially colorectal neoplasia. Little is known, however, about the role of PG transport in these processes. The flow of PGs in and out of colorectal cells involves highly coordinated activities of PG transporters that become highly dysregulated in colorectal neoplasia. Recent work by various investigators supports key components of this flow for novel molecular-targeted approaches to prevent or treat colorectal neoplasia.

The DuBois lab first reported that cyclooxygenase-2 (COX-2) is overexpressed in colorectal adenomas and cancer (1), which led to seminal trials of COX-2 inhibitors in familial adenomatous polyposis (2) and sporadic colorectal adenomas (3). This research group and others showed that COX-2-derived PGE₂ promotes tumorigenesis by affecting angiogenesis, cell adhesion, invasion, proliferation and apoptosis (4-9). These effects of COX-2/PGE₂ are implemented by PGE₂ binding to and activating certain EP receptors and thus activating important pathways including epidermal growth factor receptor (EGFR), peroxisome proliferator-activated receptor δ (PPAR-δ), Ras, PI3K, and beta-catenin signaling. More recent work from this and other groups has shown that 15-hydroxyprostaglandin dehydrogenase (15-PGDH), which degrades intracellular PGE₂, is suppressed in colorectal neoplasia from pre-adenoma to invasive stages; suppressed growth of human CRC cells in immunodeficient mice and inhibited the development of murine intestinal neoplasia when its expression was restored by transfection; and was upregulated by certain agents including EGFR and histone deacetylase (HDAC) inhibitors (10-15).

Strong evidence suggests that two cellular transporters of PGE₂—the transmembrane influx PG transporter (PGT, carrying into the cytoplasm) and efflux multidrug resistance-associated protein 4 (MRP4, carrying out to the extracellular milieu)--are deeply implicated in colorectal neoplasia. Although it is the first-identified (in 1995) and best-studied prostaglandin transporter (16, 17), PGT has not previously been studied in any cancer setting. Effective termination of PGE₂ may require both PGT, which has a high affinity and specificity for PGE₂, and 15-PGDH (18). As for MRP4: In non-cancer model systems, MRP4 knockout or knockdown led to a pronounced reduction in extracellular PGE₂, and MRP4 was inhibited by certain nonsteroidal anti-inflammatory drugs (NSAIDs) (19, 20); also, MRP4 is overexpressed in colorectal and other cancers (21).

In their exciting new report “Regulation of the Prostaglandin Transporter in Colorectal Neoplasia” in this issue (22), Holla and his colleagues in the DuBois lab build on the foregoing and other related discoveries to address our limited understanding of PGE₂ transport and inactivation in neoplasia. They found that PGT was downregulated in CRC in the vast majority of their human specimens and cell lines. Although focused on colorectal neoplasia, this study indicated that PGT expression also is dysregulated in stomach, ovary, kidney and lung cancers. PGT downregulation occurred early, or at the level of adenomas in MIN mice. Forced PGT overexpression in vitro in HCA-7 cells reduced extracellular PGE₂ levels and increased intracellular levels of 15-keto PGE₂, the catabolic product of PGE₂; both events occurred in a dose-dependent fashion, which is consistent with PGT transporting PGE₂ into cells and suggests that the 15-PGDH enzyme was still available inside these cells for degrading PGE₂ into the keto product (11). These investigators also showed evidence of epigenetic regulation of PGT by an HDAC inhibitor and a demethylating agent. These PGT results are the first reported for any cancer setting.

These new results of Holla et al amplify on the known patterns of prostaglandin signaling events in colorectal neoplasia. We now can say that colorectal neoplasia throttles down PGT expression in addition to revving up the expressions of COX-2, mPGES1 and MRP4 and throttling down 15-PGDH (Fig. 1). Data now suggest that cellular efflux of PGE₂ in normal cells can occur via two mechanisms--diffusion and MRP transport--whereas PGE₂ influx occurs primarily via the PGT. MRP4 appears to increase its efflux role in neoplasia, although diffusion likely continues. This flexibility in efflux suggests that limiting PGE₂ influx through reduced PGT may play an important role in upregulating effects of the COX-2 pathway in colorectal neoplasia and in producing increased levels of extracellular PGE₂ to bind and activate EP receptors. The suggestive timing of these events involves stepwise progressions from early (adenomas) through late stages of colorectal neoplasia.

Fig. 1.

The flow of colorectal neoplasia passes through the prostaglandin E₂ (PGE₂) transporters. A, the PGE₂ influx-dominated flow of normal colorectal cells; the dotted blue circular line represents the general pattern of this flow. B, the PGE₂ efflux-dominated flow of neoplastic colorectal cells; the dotted red circular line represents the general pattern of this flow. C, a conceptual approach for preventing or treating colorectal cancer via molecular targeting of the processes shown in panels A and B. The colorectal neoplasia risk factors shown at the upper left in panel C are key early events associated with upregulated COX-2 and neoplasia in the colorectal region. The agent classes shown at the lower right in panel C are supported by PGT and/or 15-PGDH upregulation data. All of the processes suggested in panels A–C are well described toward the end of the text. 15-PGDH, 15-hydroxyprostaglandin dehydrogenase; AA, arachidonic acid; COX-1, cyclooxygenase 1; DNMT, DNA methyltransferase; MRP, multidrug resistance-associated protein; mPGES1, microsomal PGE synthase 1; PGT, PG transporter; ППAP-γ, peroxisome proliferator-activated receptor-γ.

The novel findings of Holla et al support the conclusion that PGT may collaborate functionally with 15-PGDH to inactivate PGE₂. Fig. 1 provides a model summarizing the elements involved in the PG transport-related flow of PGE₂ into and out of normal and neoplastic colorectal cells, and highlights opportunities for using preventive or therapeutic agents to target this flow in neoplasia. The PGE₂ influx-dominated flow of normal colorectal cells is associated with a relatively low level of PGE₂ (compared with the level in neoplastic cells) originating in the cytoplasm and either undergoing degradation and inactivation by highly expressed 15-PGDH or escaping from the cell via diffusion and MRP4 [MRP1 and -2 also have been implicated in PGE₂ efflux in non-cancer experimental systems (23)] (Fig. 1A). The resultant extracellular PGE₂ then flows by means of highly expressed PGT back into the cytoplasm, where it can be inactivated by 15-PGDH. This cycle keeps the level of PGE₂ below that generally associated with dysregulated neoplastic cells. The PGE₂ efflux-dominated flow of neoplastic colorectal cells is associated with a relatively high level of PGE₂ (produced by upregulated COX-2 and mPGES1) in the cytoplasm that flows out of the cell via diffusion and upregulated MRP4 (Fig. 1B). This process is relatively unimpeded because 15-PGDH becomes downregulated in these cells. The PGT also is downregulated, and so extracellular PGE₂ is relatively free to bind with and activate EP receptors (especially EP-2 and -4), which are deeply involved in colorectal neoplasia. Findings of Holla et al and others show an intriguing potential for epigenetic or other regulation of PGT and 15-PGDH in various cancer systems (11, 14, 15, 22, 24-26) and raise the possibility that certain agents, especially EGFR (11, 14) and HDAC (22) inhibitors, may prevent or treat colorectal neoplasia via targeting PGT and/or 15-PGDH to reverse the PGE₂ efflux-dominated flow of colorectal neoplasia (Fig. 1C). It is conceivable that agents that upregulate PGT may be synergistic with agents that upregulate 15-PGDH in increasing the influx and catabolism (by 15-PGDH) of PGE₂ and thus in producing desirable effects on, for example, proliferation, apoptosis and angiogenesis. Prior work has shown that 15-PGDH functions as a tumor-suppressor gene (10-12). The new findings of Holla et al in this issue now nominate PGT as a tumor-suppressor gene, a hypothesis that is very likely to be tested in the near future.

It is clinically important to identify promising non-COX-2 targets within the polyunsaturated fatty acid metabolic signaling pathway (8, 27, 28) because COX-2-selective NSAIDs increased adverse cardiovascular events in long-term prevention trials also showing significant suppression of colorectal neoplasia (3, 29). As indicated in Fig. 1, promising such targets include mPGES1, EP-2 and -4, 15-PGDH, MRP4 and PGT (8-14, 22, 28). Future studies should examine whether the beneficial effects of epigenetic regulation in colorectal neoplasia (27, 30, 31) involve the PG transport-related mechanisms hypothesized above in this article. Mechanistic studies of PGT downregulation in neoplasia may identify additional new and promising prevention and therapy targets. We look forward to future studies that will assess the functional role of PGT in neoplasia, testing whether restoration of PGT expression suppresses tumorigenesis, and elucidating the interactions between PGT and 15-PGDH in antagonizing the activity of the COX-2 oncogenic pathway.