Embryogenesis, development, and solid tumor metastasis have many intersecting factors in regard to molecular events such as cell motility, changes in cellular differentiation, and changes in the interactions with cells of the local microenvironment (Kelleher et al., 2006; Heerboth et al., 2015). Indeed, these processes are tightly regulated and coordinated in part via gene regulation by transcription factors. Perhaps not surprisingly, a number of transcription factors have been shown to have overlapping roles in development, cancer progression, and cancer metastasis (Jackson et al., 1990; Schreiber et al., 2000; Hollern et al., 2014; Nowak et al., 2015; Zhang et al., 2016; To and Andrechek, 2018). Although some transcription factors have been well characterized for their roles in both development and in cancer; there remains many transcription factors with known significances in development that have not been examined for potential roles in carcinogenesis and metastatic progression. One such transcription factor with documented roles in development is FOXI3. Despite regulation of cancer-related processes such as epithelial to mesenchymal transition (Edlund et al., 2014; Shirokova et al., 2016); FOXI3 remains unexplored in tumorigenesis and metastasis. In this review, we summarize the role of FOXI3 in embryogenesis and bone development and examine its potential role in cancer progression using an informatics approach. Moreover, we discuss possible gene targets that might allow FOXI3 to facilitate solid tumor bone metastasis.

The transcription factors of the Forkhead box family of transcription factors consist of a highly conserved, 110-amino acid Forkhead or “winged helix” DNA binding domain (DBD) which usually binds to DNA through RYAAAYA motif (where R = purine and Y = pyrimidine) (Nakagawa et al., 2013; Chen et al., 2016). The FOX genes of this family play important roles in various biological processes including embryogenesis, development and metabolism (Lehmann et al., 2003; Coffer and Burgering, 2004; Tuteja and Kaestner, 2007a,b).

Given that many developmental pathways, transcription factors, and processes, are commonly implicated in cancer progression, we propose that FOXI3 may also play an important role in cancer. Indeed, with governance of developmental features such as stem cell maintenance and involvement in EMT, FOXI3 may enable resistance to chemotherapy or tumor metastasis (Anders and Carey, 2008; Chaffer et al., 2016). However, to date no studies have been published that demonstrate FOXI3's role in tumor progression. As such, we have utilized a variety of public data to examine FOXI3 across cancer types and test for association with key clinical measures of tumor progression.

One of the largest coordinated efforts to document the genomic features of cancer has been the work of The Cancer Genome Atlas. Using the cBioPortal to access multiple cancer types (Cerami et al., 2012; Gao et al., 2013), we first interrogated FOXI3 expression across several major cancer types, (Figure 2). While relatively few liver and thyroid cancers show elevated FOXI3 expression; other cancers display a wide range of FOXI3 expression across patient tumors (breast, lung, prostate, ovarian, and head and neck cancer). With this, we questioned whether changes in FOXI3 expression associate with features of tumor progression in several of these cancer types.

Within the TCGA prostate cancer cohort (Cerami et al., 2012; Gao et al., 2013; Cancer Genome Atlas Research Network, 2015), we find that FOXI3 expression increases with tumor stage (Figure 3A). While T4 tumors were infrequent in this cohort, many T3 stage tumors as compared to T2 stage have elevated FOXI3 expression. The progression from T2 to T3 stage is classified in part by the presence of tumor cell invasion to the surrounding tissues of the prostate gland (T3) (Buyyounouski et al., 2017). This marks the first signs of metastatic potential of FOXI3. Indeed, given the bounty of research demonstrating EMT as critical to the invasive potential of prostate cancer, future research examining whether FOXI3 expression functions in EMT, invasion, and metastasis could bring together these clinical associations to key processes of tumor progression controlled by FOXI3. Within the TCGA breast cancer cohort (Cancer Genome Atlas Research Network, 2012), some of the primary tumors were annotated for the site where distant metastasis was observed. Examining FOXI3 expression in this data showed that breast cancers that metastasized to the bone often had high FOXI3 expression (Figure 3B). While the number of cases are limited, additional experiments may illuminate the role of in facilitating bone metastasis. In support of this hypothesis, we noted that many of the genes shown to be downregulated with FOXI3 loss (Figure 1), also have demonstrated roles in tumor progression and metastasis (Table 2), including bone metastasis.

These putative targets of FOXI3 (Figure 1) also provide an additional tool for informatics inquiries. Initially, one of the limitations in our examination was the omission of probes for FOXI3 on the microarrays of a variety of clinically rich gene expression datasets. However, the use of FOXI3 putative target genes as a signature of FOXI3 activity bypasses this limitation and allows for additional analyses. One dataset of particular interest was the published gene expression dataset across distant prostate cancer metastatic sites (Haider et al., 2016). As shown in Figure 4A, FOXI3 signature levels were high in bone, lung, and lymph node metastases. FOXI3 target genes also showed an association with metastasis in breast cancer. Probing expression data from primary breast tumors annotated for time to distant metastasis, we observed that tumors with the highest expression of the FOXI3 signature genes present distant metastasis earlier than those with low expression (Figure 4B). Taken together with the observations in Figure 3A, FOXI3 may mark and function in tumor cell invasion and metastasis. Given the clues from developmental studies, it might be that this function occurs by EMT and is an area that should be investigated. Additionally, data shown in Figures 3B, 4A, suggests that FOXI3 may facilitate colonization of tumor cells to the bone. Indeed, a number of the putative targets of FOXI3 have been implicated in bone metastasis and supports the need for future studies to examine FOXI3 specifically in the bone metastasis setting.

In this review, we note that FOXI3 a regulator of early bone development may play a potential role in cancer progression (Table 3). We further note that several of the genes that are downregulated with the loss of FOXI3, such as SPOCK1, Dlx, MAF, PthrP, HAS2, SVEP1, and TGFβ3 (Jussila et al., 2015; Shirokova et al., 2016) have been shown to play important roles in distant metastasis to other sites including bone (Table 2). Based on our analysis of publicly available clinical data regarding FOXI3 expression in different cancer types and stages, we speculate that FOXI3 may have multiple roles throughout metastatic progression. At the primary tumor, FOXI3 may allow for stem cell maintenance or EMT; the latter in turn may implicate FOXI3 in tumor cell invasion into the local tissues and also into circulation. Further, in advanced cancer stage, during colonization of distant sites, FOXI3 may facilitate reconfiguring of the tumor microenvironment by expression of paracrine factors. These mechanisms may come into play to facilitate bone metastasis as evidenced by several of the above listed potential target genes having previously demonstrated roles in bone metastasis (Table 2). Thus, we propose that when FOXI3 in tumor cells modulate the secretion of other bone factors including SPOCK1, Dlx, MAF, PthrP, HAS2, SVEP1, TGFβ3, and FGF (Jussila et al., 2015; Meng et al., 2016; Shirokova et al., 2016) which further primes the microenvironment, creates a crosstalk to help tumors grow in the bone (Figure 5). Understanding the intersect of the regulatory behavior of FOXI3 on factors prominent in hallmark processes of tumor progression to bone is essential to elucidating its role when these processes are hijacked and go awry. As a whole, this review highlights promising directions for research of FOXI3 and its potential roles in cancer progression. Given no such studies exist to date, there is much to be discovered for this key developmental factor in the cancer setting and is a direction our lab plans to pursue.

JJ and AM devised the concept with CY. AM, JJ, and OW developed the theory. DH, AM and OW developed and analyzed the tables. AM and DH, developed and interpreted the figures. AM, JJ, DH, OW, TR, WB, and CY drafted the manuscript.