**7. Expression of Survivin 3***γ*

Survivin 3 expression in primary cancer tissues was determined by quantitative RT-PCR on primary breast and prostate tumor samples obtained in an array format. In contrast to most cancers where wild-type Survivin is the predominant transcript, Survivin 3 was expressed at higher levels in benign prostate tissue and at equivalent levels in breast cancer cells. In primary breast cancer samples a statistically significant increase in wild-type Survivin expression at stages 1, 2 and 3 compared to control normal breast tissue was observed, with

Survivin: Identification of Selective Functional Signaling Pathways

**8. Survivin 3***γ* **demonstrates antiapoptotic activity** 

antibody that recognizes predominantly human Survivin.

cell maintenance.

in Transformed Cells and Identification of a New Splice Variant with Growth Survival Activity 195

breast tissues and that the expression of Survivin increased at stage 1 and remained high at all stages, whereas a recent report (Sohn et al., 2006) showed Survivin expression increased gradually at all disease stages. This difference could again be due to different methodologies and sensitivity between RNA analyses versus immunohistochemistry. Since the Survivin 3 splice variant was identified in the human HL-60 promyelocytic cell line, we evaluated its expression in other hematopoietic cells by quantitative RT-PCR using SYBR Green. While wild-type Survivin was the predominant transcript identified in all cell lines tested, Survivin 3 transcripts represented ~0.7-6% of the total Survivin transcripts detected. The absolute expression of Survivin 3 was lowest in growth factor dependant cells and highest in growth factor independent cells. Analysis of Survivin 3 in primary umbilical cord blood (UCB) CD34+ cells that are enriched for hematopoietic stem and progenitor cells showed that Survivin 3 was expressed at significantly higher levels than wild-type Survivin. This is in contrast to most splice variants that are expressed at lower levels than wild-type Survivin in these cells. However, growth factor stimulation induced a dramatic increase in expression of wild-type Survivin after 48 hours consistent with our previous findings (Fukuda et al., 2002; Fukuda & Pelus, 2001; Fukuda & Pelus, 2002), whereas the expression of Survivin 3 increased only slightly, therefore relative levels compared to wild-type Survivin decreased by ~2-fold. This suggests perhaps that Survivin 3 may play a role in hematopoietic stem

To evaluate the function of the novel Survivin 3 splice variants, we stably expressed human (Hu) wild-type Survivin with or without an N-terminal Flag-tag or Survivin 3 with an Nterminal HA-tag in murine YAC-1 lymphoma cells. Wild-type HuSurvivin, wild-type Flag-HuSurvivin and HA-Survivin 3 were all expressed, although their levels of expression varied, with wild-type HuSurvivin expressing at 31-fold, whereas wild-type Flag-HuSurvivin and HA-Survivin 3 were expressed at 76- and 63-fold higher levels, respectively than the empty vector control cells at the RNA level. Western blot analysis using a rabbit anti-Survivin polyclonal antibody that recognizes both the human and the murine Survivin showed background levels of murine Survivin as well as the presence of each of the transduced Survivins, indicating that each of the splice variants was translated. Analysis of band intensity indicated that wild-type HuSurvivin was expressed at the highest level followed by wild-type Flag-HuSurvivin with HA-Survivin 3 expressed at a lower level. The same pattern of expression was observed with a mouse anti-Survivin monoclonal

As a measure of anti-apoptotic function we evaluated resistance of YAC-1 cells transduced with wild-type Survivin, wild-type Flag-Survivin and HA-Survivin 3 to the chemotherapeutic drugs Etoposide and Taxol at their optimal therapeutic dose. Treatment of cells with Taxol showed that, as expected, over expression of either wild-type Survivin or wild-type Flag-Survivin provided 2.7- and 2.8-fold resistance to Taxol-induced cell death, respectively, compared to YAC-1 cells transduced with vector control (Figure 6). Ectopic expression of HA-Survivin 3 provided 2.6-fold resistance. In multiple experiments, the IC50's for wild-type HuSurvivin, wild-type Flag-HuSurvivin and HA-Survivin 3 were significantly higher than control (2.85 ± 0.32; P<0.05). YAC-1 cells expressing wild-type

no significant difference observed between wild-type Survivin expression at stage 1 compared to later stages. In 7 normal breast tissue samples wild-type Survivin and Survivin 3 were expressed at the same level. However, in contrast to expression of wild-type Survivin, Survivin 3 expression remained relatively constant at all stages of disease, resulting in statistically significant (p<0.005) lower relative expression to wild-type Survivin.

Fig. 5. Agarose gel electrophoresis of cDNA amplified from HL60 cells (lane 2). The Survivin-3B variant is indicated by a solid arrow, the dotted arrow points to the slower migrating band. Amino acid sequence of Survivin-3 is shown on the right. Bold letters indicate novel sequence distinct from wild-type Survivin. The stop codon is indicated by the \*.

In 48 prostate cancers at different stages of the disease, wild-type Survivin expression was higher in stage 1 prostate cancer compared to benign hyperplasia; however sample size was too small to evaluate statistical significance. No clear over-expression of wild-type Survivin was seen at higher stages, except for a single stage 4 sample. The expression of Survivin 3 was higher than wild-type Survivin in benign hyperplasia tissue and comparison of the relative expression of wild-type Survivin and Survivin 3 at each stage of prostate cancer indicated that Survivin 3 transcripts remained higher than wild-type Survivin at all stages except stage 1. To our knowledge, the higher expression of Survivin 3 in benign prostate hyperplasia is the first demonstration of a Survivin splice variant expressed at higher levels than wild-type Survivin and suggest that it may play a role in normal cell function.

The finding that Survivin 3 expression was higher than wild-type Survivin in benign prostate hyperplasia (BPH) samples and did not change with stage of disease is consistent with one study (Kaur et al., 2004) where Survivin expression was detected by immunohistochemistry. However, other reports show increased Survivin with stage of disease either at the RNA level by standard RT-PCR and gel electrophoresis (Kishi et al., 2004; Krajewska et al., 2003) or by immunohistochemistry (Krajewska et al., 2003). The difference between our study and those reported using standard RT-PCR could be due to differences in sensitivity of the method used. SYBR green quantitative RT-PCR is more sensitive and efficient in amplifying smaller amplicons than conventional RT-PCR. We also show that expression of wild-type Survivin and Survivin 3 were comparable in normal

no significant difference observed between wild-type Survivin expression at stage 1 compared to later stages. In 7 normal breast tissue samples wild-type Survivin and Survivin 3 were expressed at the same level. However, in contrast to expression of wild-type Survivin, Survivin 3 expression remained relatively constant at all stages of disease, resulting in statistically significant (p<0.005) lower relative expression to wild-type Survivin.

> MGAPTLPPAWQPFLKDHRISTFKN WPFLEGCACTPERMAEAGFIHCPT ENEPDLAQCFFCFKELEGWEPDDD PIEEHKKHSSGCAFLSVKKQFEELT LGEFLKLDRERAKNKI**GLTLSPRL**

Sur3B

Fig. 5. Agarose gel electrophoresis of cDNA amplified from HL60 cells (lane 2). The Survivin-3B variant is indicated by a solid arrow, the dotted arrow points to the slower migrating band. Amino acid sequence of Survivin-3 is shown on the right. Bold letters indicate novel sequence distinct from wild-type Survivin. The stop codon is indicated by the \*. In 48 prostate cancers at different stages of the disease, wild-type Survivin expression was higher in stage 1 prostate cancer compared to benign hyperplasia; however sample size was too small to evaluate statistical significance. No clear over-expression of wild-type Survivin was seen at higher stages, except for a single stage 4 sample. The expression of Survivin 3 was higher than wild-type Survivin in benign hyperplasia tissue and comparison of the relative expression of wild-type Survivin and Survivin 3 at each stage of prostate cancer indicated that Survivin 3 transcripts remained higher than wild-type Survivin at all stages except stage 1. To our knowledge, the higher expression of Survivin 3 in benign prostate hyperplasia is the first demonstration of a Survivin splice variant expressed at higher levels

than wild-type Survivin and suggest that it may play a role in normal cell function.

The finding that Survivin 3 expression was higher than wild-type Survivin in benign prostate hyperplasia (BPH) samples and did not change with stage of disease is consistent with one study (Kaur et al., 2004) where Survivin expression was detected by immunohistochemistry. However, other reports show increased Survivin with stage of disease either at the RNA level by standard RT-PCR and gel electrophoresis (Kishi et al., 2004; Krajewska et al., 2003) or by immunohistochemistry (Krajewska et al., 2003). The difference between our study and those reported using standard RT-PCR could be due to differences in sensitivity of the method used. SYBR green quantitative RT-PCR is more sensitive and efficient in amplifying smaller amplicons than conventional RT-PCR. We also show that expression of wild-type Survivin and Survivin 3 were comparable in normal

**SAVV\***

1000bp

Lane 1 2

300bp

Size marker

500bp

breast tissues and that the expression of Survivin increased at stage 1 and remained high at all stages, whereas a recent report (Sohn et al., 2006) showed Survivin expression increased gradually at all disease stages. This difference could again be due to different methodologies and sensitivity between RNA analyses versus immunohistochemistry. Since the Survivin 3 splice variant was identified in the human HL-60 promyelocytic cell line, we evaluated its expression in other hematopoietic cells by quantitative RT-PCR using SYBR Green. While wild-type Survivin was the predominant transcript identified in all cell lines tested, Survivin 3 transcripts represented ~0.7-6% of the total Survivin transcripts detected. The absolute expression of Survivin 3 was lowest in growth factor dependant cells and highest in growth factor independent cells. Analysis of Survivin 3 in primary umbilical cord blood (UCB) CD34+ cells that are enriched for hematopoietic stem and progenitor cells showed that Survivin 3 was expressed at significantly higher levels than wild-type Survivin. This is in contrast to most splice variants that are expressed at lower levels than wild-type Survivin in these cells. However, growth factor stimulation induced a dramatic increase in expression of wild-type Survivin after 48 hours consistent with our previous findings (Fukuda et al., 2002; Fukuda & Pelus, 2001; Fukuda & Pelus, 2002), whereas the expression of Survivin 3 increased only slightly, therefore relative levels compared to wild-type Survivin decreased by ~2-fold. This suggests perhaps that Survivin 3 may play a role in hematopoietic stem cell maintenance.
