Although CAIX and CAXII expression during mouse embryonic development has been previously described [28
], a systematic comparative study of CAIX, CAXII and HIF-1α expression in the embryonic, fetal and postnatal tissues during human development has not been reported. We find that the distributions of CAIX and CAXII immunoreactivity in the mouse tissues are similar but not identical to the immunoreactive patterns observed in the human organs. For example, staining for CAIX in the mouse embryos showed a relatively wide distribution pattern with moderate signals in the brain, lung, pancreas and liver, and weak expression in the kidney and stomach [28
]. In human fetal tissues, CAIX expression in those organ systems was restricted to certain cell types, such as the primitive ependymal cells of the brain, immature mesenchymal cells of the bronchial tree system, and epithelial cells of the pancreatic ducts and stomach. High levels of CAXII expression were seen in the choroid plexus in both mouse and human tissues, but CAXII expression in the embryonic mouse kidneys and other organs was weak and became negative in the adult mice. In contrast, the intensity of CAXII expression in the human fetal tissues persists postnatally throughout the adult lifespan.
Several intriguing features concerning CAIX expression have emerged from the study of the developing human embryo and fetus. As expected, HIF-1α and CAIX expression colocalized in many, but not all, of the embryonic and early fetal tissues. As fetal development progressed, and vascularization increased, the lack of co-expression of HIF-1α and CAIX became more apparent in multiple tissues. In those instances where HIF-1α is not expressed but CAIX is, one must assume that its expression is regulated by other transcriptional factors. We, and others [16
], have shown that SP1 is required.
Furthermore, transcription mediated indirectly by P13-kinase activity may also be a possibility [18
]. Of particular interest is the situation where HIF-1α is stably expressed but no CAIX expression is found. Stable inactive HIF-1α expression has been noted in the presence of proteasome inhibitors [30
]. One possibility is the expression of a co-repressor inhibiting the activity of HIF-1. Another important factor in the regulation of CAIX expression is the cell origin. CAIX expression was consistently observed in cells derived from the coelomic epithelium. The best examples are mesothelial cells, modified mesothelial cells (Müllerian epithelium), underlying mesenchymal cells lining the body cavity, and coelomic remnants, such as the rete ovarii, rete testis and tubuli reti, the hydatid of Morgagni and the appendix of the testis. Another feature of the regulation of CAIX expression is its pH control function. CAIX is a transmembrane carbonic anhydrase that possesses cell surface enzyme activity that catalyzes the conversion of CO2 into bicarbonate and protons [31
]. Thus, the CAIX expression observed in the gastrointestinal tract is more likely induced by cellular acidity and the requirement for proton transport.
Although CAXII expression was originally thought to be regulated by HIF-1 [10
], this is clearly not the case. There is no obvious co-localization of HIF-1α and CAXII.
Furthermore, there is also no co-localization of CAIX and CAXII. The distribution of CAXII expression would suggest that it plays an important physiological role in secretory/absorptive cells in different organ systems, primarily involving ion transport and fluid concentration. Corroborating this suggestion is a previous study that found that CAXII is overexpressed in the ciliary epithelial cells of glaucomatous eyes and may be involved in aqueous humour production [32
A particularly intriguing finding was the discrete distribution of CAIX expression in rare cells or niches in late stages of fetal development, and postnatally, that correspond to sites previously identified as harboring adult stem cells. For example, CAIX expression in the skin is restricted to the hair follicles, including the bulge, sebaceous gland, outer root sheath and infundibulum, plus rare cells in the inter-follicular zone. This corresponds to epidermal stem cell niches that have been described, using various stem cell biomarkers [33
]. Also, the distribution of CAIX expression in the small and large intestine is very similar to that described for the identification of the intestinal stem cell niche, using the Lgr5 biomarker [35
]. Finally, our observation of CAIX expression in Müllerian-type columnar cells and reserve cells of the cervix, during fetal development and postnatally, appears to be similar to that described in the study conducted by Martens and colleagues [36
]. These authors claim that the Müllerian epithelial cells represent the stem cells for endocervical reserve cells and columnar cells. Thus, we hypothesize that the rare reserve cells expressing CAIX in the cervix and other organ systems may correspond to putative stem cell regions. However, at this time, such distributions and their relevance to stem cell identity are speculative and will require more comparison and functional analyses. These studies are in progress. If CAIX is found to be a stem cell marker for certain tissues it will be an attractive one. Its transmembrane location will be extremely useful for cell sorting analyses. It may also provide a caveat for cancer therapeutic regimens that target CAIX-expressing cells for destruction [37