BDGI/OKL38: A New Breast Cancer Target

The estimated worldwide annual incidence of breast cancer is approximately one million and rising. However, great strides have been made in our understanding of the risk factors associated with breast cancer. Several factors have been known for some time such as age, geography, and family history. Others, in particular reproductive factors, have been described more recently. For example, earlier age at first full-term pregnancy and higher number of pregnancies can cut breast cancer risk in half.1

Additionally, researchers are diligently working on characterizing mechanisms involved in breast cancer development in hopes of identifying therapeutic targets. For instance, Wang et al. have recently cloned and characterized a novel inhibitor of breast cancer cell growth that may represent a promising new target for breast cancer therapy. Bone marrow stromal cell (BMSC)-derived growth inhibitor (BDGI) was identified from a human BMSC cDNA library and encodes a 477 amino acid (aa), ~52 kDa, soluble, non-secretory protein that exhibits anti-tumor effects.2

A truncated version of BDGI had been described previously by Huynh et al. and named OKL38, or pregnancy induced growth inhibitor. OKL38 is a 317 aa, ~38 kDa protein that is also soluble, but not secreted. OKL38 is ubiquitously expressed (with highest levels in ovary, kidney, and liver) and up-regulated in normal breast epithelial cells during pregnancy. However, expression is low in human breast cancer cell lines and largely absent in DMBA-induced rat mammary tumors. When MCF-7 human breast adenocarcinoma cells are transfected with OKL38, in vitro cellular proliferation and in vivo tumor formation are reduced. Based on these data, the authors speculate that OKL38 may be a hormonally regulated growth inhibitor that becomes suppressed upon malignant transformation.3

Not surprisingly, Wang et al. report similar behavior for BDGI. Further, BDGI overexpression in MCF-7 cells reduced their in vitro metastatic activity, arrested the cells in S phase, and induced apoptosis. The cell cycle arrest and apoptosis effects appear to be mediated by changes in the expression of checkpoint and apoptosis regulators, as cyclin A is decreased, p27 is increased, and the anti-apoptotic factors, Bcl-2 and Bcl-xL, are decreased.2

Since the initial cloning and characterization of human OKL38, its genomic sequence has been described as spanning ~18 kb and containing 8 exons. Differential promoter usage and alternative splicing appear to produce as many as four different transcripts. The largest of these is 2.4 kb encoding a 560 aa, ~61 kDa protein. Two transcripts are 2.2 and 1.9 kb encoding 477 and 447 aa proteins, respectively, both of which are ~52 kDa. Another minor transcript was also observed, but only a partial sequence was obtained. While the original 38 kDa variant was not observed in these experiments, the authors suggest that it appears during translation as a result of internal start codon usage.4

Mouse and rat homologs of human OKL38 have also been identified and the sequences are highly conserved across species. Upon characterization of the rat homolog, in vivo expression appeared to be regulated by chorionic gonadotropin (CG). These data suggest a potential link between OKL38 and the observed pregnancy-associated reduction in breast cancer risk,5 as CG has been described previously as inhibiting DMBA-induced rat mammary tumor formation.6 Thus far, BDGI/OKL38 has been implicated not only in breast cancer, but also in ovarian,5 kidney,4 and liver cancer.7 Further, the human BDGI/OKL38 gene is localized to chromosome 16q23, a region known for its susceptibility to loss of heterozygosity and for its strong association with aggressive breast, ovarian, kidney, liver, and prostate cancers.7

Figure 1. Schematic depicting a model of BDGI/OKL38 actions in human breast epithelium that either protect against (left) or allow the development (right) of breast cancer.

Taken together, these data support a model in which OKL38 normally functions as a hormonally regulated growth inhibitor that contributes to the breast cancer protective effects of pregnancy, but perhaps due to inherent instability at its genomic location, its expression may become suppressed allowing cancer to develop (Figure 1).2-5


  1. Dumitrescu, R.G. & I. Cotarla (2005) J. Cell. Mol. Med. 9:208.
  2. Wang, T. et al. (2005) J. Biol. Chem. 280:4347.
  3. Huynh, H. et al. (2001) Endocrinology 142:3607.
  4. Ong, C.K. et al. (2004) J. Biol. Chem. 279:743.
  5. Ong, C.K. et al. (2004) Endocrinology 145:4763.
  6. Russo, I.H. et al. (1990) Carcinogenesis 11:1849.
  7. Riou, P. et al. (2002) Clin. Cancer Res. 8:3178.