The gene is located on chromosome 12 on the plus strand at 12202778 bp from pter to 12364018 from pter (size is 161241 bases). It was originally thought to comprise 6 exons (Guo et al., 2001) but 7 possible additional exons have subsequently been identified (Montpetit et al., 2002).

The gene encodes multiple transcripts; reference sequences exist for 3 transcript variants. These give rise to two different protein isoforms (Fig. 1). Transcript variants 1 and 4 differ only in their 5-UTR and both encode the same protein, termed isoform 1 or Bcl-GL. Usage of a different splice acceptor site results in the production of transcript variant 2. This contains extra nucleotides within the coding sequence, resulting in a frameshift and consequently a truncated protein product with distinct C-terminal amino acids, termed isoform 2 or Bcl-GS. Transcript variant 2 may be a candidate for nonsense-mediated mRNA decay (NMD), but the RefSeq record (NM_030766.1) has been retained, since there is evidence for the expression of protein product. However, the RefSeq (NM_138724.1) for a further variant, originally termed variant 3, which encodes Bcl-G-Median (Bcl-GM) (Montpetit et al., 2002), has been permanently suppressed, as this transcript is also a candidate for NMD and evidence for protein expression is lacking.

There are no known pseudogenes.

The two protein isoforms are members of the Bcl-2 family. Bcl-GL comprises 327 amino acids and contains BH2 and BH3 domains, whereas Bcl-GS comprises 252 amino acids and contains the BH3 domain only (Fig. 2). The two proteins are identical in sequence for the first 226 amino acids.

Highest levels of gene expression in the adult human are found in testis, hence the name Bcl-G, for Bcl-Gonad. Indeed Bcl-GS is uniquely expressed in this tissue (as also noted for Bcl-GM) (Guo et al., 2001; Montpetit et al., 2002). Bcl-GL is predominantly expressed in testis, with transcripts also present in lung, pancreas, prostate, bone marrow and colon (Guo et al., 2001; Montpetit et al., 2002). The expression of Bcl-GL has also been reported in freshly isolated human T-lymphocytes (CD4⁺ and CD8⁺ cells) and B-lymphocytes (Luo et al., 2009). Both Bcl-GL and Bcl-GS transcripts are expressed in a wide range of breast cancer cell lines and cultured mammary epithelial cells, but protein expression has been demonstrated for the former isoform only (Benito et al., 2006; Lin et al., 2007). Bcl-G transcript expression is more variable across prostate cell lines (Pickard et al., 2010). High levels of Bcl-GS transcripts are found in testicular embryonal cancer cell lines, with lower levels present in urinary bladder and mantle cell lymphoma cell lines, followed by myeloid and breast cancer cell lines, but protein was detected for the testicular embryonal cancer cell lines only (Benito et al., 2006).

Expression of GFP-tagged proteins in COS-7 cells has revealed a diffuse localization throughout the cell for Bcl-GL (similar to GFP alone), but a more punctate cytosolic localization, with partial co-localization to mitochondria, for Bcl-GS (Guo et al., 2001). The latter has been confirmed using Myc-tagged Bcl-GS (Liu et al., 2008). Deletion of the BH3 domain from Bcl-GS has no effect on its subcellular distribution, despite preventing interaction with Bcl-XL (Guo et al., 2001). However interaction with JAB1 results in a more diffuse cytoplasmic localisation for Bcl-GS (Liu et al., 2008).

The original paper on BCL2L14 reported that GFP- or FLAG-tagged Bcl-GL and Bcl-GS have pro-apoptotic activity in a range of cell lines (COS-7, HEK293T and PC3), and that the latter is the more potent isoform (Guo et al., 2001). Deletion of the BH2 domain from Bcl-GL allowed its interaction with Bcl-XL and enhanced apoptotic activity, whereas the BH3 domain of Bcl-GS was required for interaction with Bcl-XL and apoptotic activity (Guo et al., 2001).
The pro-apoptotic activity of Bcl-GL in COS-7 cells has been independently confirmed, and it has been further demonstrated that this activity is negatively regulated by phosphorylation, catalyzed by MELK (Lin et al., 2007). Similarly, Bcl-GS has been reported to induce basal apoptosis in HeLa cells; an effect that is potentiated by JAB1 (Liu et al., 2008). The latter is thought to out-compete Bcl-2/Bcl-XL for binding of Bcl-GS (Liu et al., 2008). On the other hand, transfection of Huh-7 cells with Bcl-GL has no effect on basal apoptosis (Zhang et al., 2006), while the survival of HCT116 and NTERA2 cells is unaffected by elevated levels of Bcl-GS consequent upon ectopic TEF expression (Benito et al., 2006), suggesting that cell context is important for the apoptotic activity of Bcl-G protein isoforms. Crucially lentiviral transfection of Bcl-GL results in enhanced apoptosis of CD4⁺ T-cells from healthy subjects, demonstrating that pro-apoptotic activity is not restricted to transformed cells (Luo et al., 2009).
Furthermore, siRNA-mediated silencing of BCL2L14 gene expression attenuates apoptosis induction in several cell lines in response to diverse stimuli, including: ultraviolet-C irradiation (22Rv1 and HEK293T cells) (Pickard et al., 2010; Pickard et al., 2011), etoposide (HCT116 cells) (Benito et al., 2006), neratinib (SKBR-3 cells) (Seyhan et al., 2012), p53-induction (Saos-2 cells) (Miled et al., 2005), and ectopic expression of the putative tumour suppressor gene FAU (HEK293T cells) (Pickard et al., 2011). BCL2L14 is itself a target gene for the apoptosis master regulator p53 (Miled et al., 2005), and it is one of a subset of pro-apoptotic genes whose expression is hyperactivated in response to interferon-α and -γ co-treatment of hepatoma cells (Zhang et al., 2006). Its expression is also increased in mammalian cells following exposure to a diverse range of apoptotic stimuli (including chemotherapeutic agents) such as: gamma- radiation therapy (Finnberg et al., 2008), UV-C irradiation (Pickard et al., 2011), arsenic trioxide (Galimberti et al., 2012), nanoparticulate tetraiodothyroacetic acid (Glinskii et al., 2009), etoposide and cisplatin (Benito et al., 2006).
Mutations in pancreatic/duodenum homeobox protein 1 (PDX1) can result in human type 2 diabetes. Haploinsufficiency of this gene in mice also causes diabetes and is associated with enhanced apoptosis/necrosis of β-cells. In this regard, BCL2L14 has been identified as one of several genes encoding pro-apoptotic Bcl-2 family members whose expression is markedly up-regulated following shRNA-mediated depletion of Pdx1 in mouse insulinoma MIN6 cells (Fujimoto et al., 2010). Additionally BCL2L14 is one of several pro-apoptotic genes that is up-regulated upon exposure of human fibroblasts to pressure-induced oxidative stress (Oh et al., 2008).
The recent production of a mouse in which BCL2L14 is inactivated has demonstrated no apparent deleterious effects on cell types that normally express this gene (Giam et al., 2012b), perhaps suggesting some degree of functional redundancy. These authors further demonstrated that mBclG had negligible pro-apoptotic activity in human 293T cells and that the BH3 domain was incapable of binding Bcl-2 family members. Rather mBclG was capable of binding several novel partners, including transport particle protein (TRAPP) complex proteins, suggesting a potential role for mBclG in vesicle trafficking (Giam et al., 2012b). As regards pig Bcl-G, its ectopic expression in a porcine kidney cell line stimulates apoptosis induced by polyinosinic: polycytidylic acid (synthetic analogue of dsRNA and mimic of viral infection) (Jiang et al., 2012).

Homologues of BCL2L14 have been identified in a wide range of species (e.g. see Jiang et al., 2012). However, the Bcl-GS protein isoform is specific to humans (Giam et al., 2012a). Mouse and pig Bcl-G share 68% and 71% amino acid identity, respectively, with human Bcl-GL.