To test the possible functional role of REG3A in breast cancer cells, we utilized shRNA strategy to genetically silence REG3A. Three shRNAs led to substantial REG3A mRNA silencing, including shREG3A-Sq2, shREG3A-Sq3, and shREG3A-Sq6 (Fig. 2A) in pBC-1 TNBC cells. REG3A protein expression in pBC-1 cells was downregulated as well by the three shRNAs (Fig. 2B). REG3A shRNAs, however, failed to alter the expression of REG1 mRNA (Fig. 2A) and proteins (Fig. 2B).

Fig. 2

shRNA-induced knockdown of REG3A potently inhibits breast cancer cell proliferation and migration. The mRNA and protein expression of REG3A and REG1 in the stable pBC-1 primary breast cancer cells with the applied REG3A shRNA (“shREG3A-Sq2/3/6”, with non-overlapping sequences), the scramble control shRNA (“shC”), or in the parental control cells (“Ctrl”) was shown (A and B). Cells were further cultivated for indicated hours, cell proliferation (by measuring nuclear EdU incorporation, C), cell migration (D) and invasion (E) were tested. The pBC-2 primary cancer cells, MCF-7 and MDA-231 lines, the primary mammary epithelial cells (pMEC) or established MCF-10A epithelial cells with shC and shREG3A-Sq6 were formed, and REG3A mRNA expression tested (F and I). Cells were further cultivated for designated hours, cell proliferation (G and J) and in vitro cell migration (H) were examined similarly. Values were mean ± standard deviation (SD, n = 5). *P < 0.05 versus “shC” group. “N. S.” stands for non-statistical difference (P > 0.05). Experiments were repeated five times and similar results were obtained each time. Scale bar = 100 μm

The functional consequences of genetic silencing of REG3A were examined in breast cancer cells. As shown, REG3A silencing by the shRNAs largely inhibited pBC-1 cell proliferation and substantially decreased nuclear EdU incorporation (Fig. 2C). “Transwell” assay results showed that shREG3A-Sq2/3/6 treatment significantly reduced pBC-1 cell in vitro migration (Fig. 2D) and invasion (Fig. 2E). The lentiviral scramble control shRNA (“shC”), as expected, failed to alter REG3A/REG1 expression (Fig. 2A and B) and pBC-1 cell functions (Fig. 2C-E).

Among the tested shRNAs, shREG3A-Sq6 demonstrated superior activity in silencing REG3A. Thus, shREG3A-Sq6-expressing lentivirus was added to other primary TBNC cells, pBC-2, and also to the immortalized cell lines (MCF-7 and MDA-231). It again resulted in substantial REG3A mRNA downregulation after stable cell selection (Fig. 2F). Similar to the results in pBC-1 cells, shREG3A-Sq6-induced knockdown of REG3A largely inhibited proliferation (EdU incorporation in nuclei, Fig. 2G) and migration (Fig. 2H) in the primary and established breast cancer cells. Whereas in primary mammary epithelial cells (“pMEC”) and MCF-10 A established cells, REG3A silencing by shREG3A-Sq6 (Fig. 2I) failed to inhibit nuclear EdU incorporation, the indicator of cell proliferation (Fig. 2J). Thus REG3A silencing potently inhibited breast cancer cell proliferation and migration.

shRNA-induced silencing of REG3A provokes apoptosis in breast cancer cells

The results above demonstrated that genetic silencing of REG3A led to growth arrest. Whether cell apoptosis occurred was studied next. In pBC-1 TNBC cells, shREG3A-Sq2/3/6-induced silencing of REG3A (see Fig. 2) led to a significant viability (CCK-8 OD) reduction (Fig. 3A). Significant cell death, as evidenced by increased Trypan blue staining, was also detected in REG3A-silenced pBC-1 TNBC cells (Fig. 3B). In addition, REG3A shRNAs augmented Caspase-3 activity (Fig. 3C) and induced cleavage and activation of key apoptosis implementing proteins, including caspase-3, caspase-9, and poly(ADP-ribose) polymerase 1 (PARP) (Fig. 3D). Furthermore, the levels of Histone-bound DNA were increased in REG3A-silenced pBC-1 cells (Fig. 3E).

Fig. 3

shRNA-induced silencing of REG3A provokes apoptosis in breast cancer cells. The pBC-1 primary breast cancer cells with the applied REG3A shRNA (“shREG3A-Sq2/3/6”, with non-overlapping sequences), the scramble control shRNA (“shC”), as well as the parental control cells (“Ctrl”) were cultivated for indicated time periods, cell viability and death were tested by CCK-8 (A) and Trypan blue staining (B) assays, respectively; Caspase-PARP activation was examined as well (C and D); Levels of histone-bound DNA were measured via an ELISA assay (E), with mitochondrial depolarization tested by JC-1 fluorescence staining assay (F); Cell apoptosis was examined by nuclear TUNEL staining and TUNEL-positive nuclei percentage (% versus DAPI) was calculated (G). The pBC-1 primary breast cancer cells with shREG3A-Sq6 or shC were treated with z-DEVD-fmk (“zDEVD”, 30 µM), z-VAD-fmk (“zVAD”, 30 µM) or vehicle control (0.15% DMSO), cells were further cultivated for additional 96 h, cell viability (CCK-8 OD, H) and death (Trypan blue staining assays, I) were tested. The pBC-2 primary cancer cells, MCF-7 and MDA-231 established cancer cell lines, the primary mammary epithelial cells (pMEC) or established MCF-10A epithelial cells, with shC and shREG3A-Sq6, were cultivated for designated hours, cell viability (CCK-8 OD, J), the caspase-3 activity (K and M) and cell apoptosis (by measuring TUNEL-positive nuclei percentage, L and N) were tested. Values were mean ± standard deviation (SD, n = 5). *P < 0.05 versus “shC” group. #P < 0.05 (H and I). “N. S.” stands for non-statistical difference (P > 0.05). Experiments were repeated five times and similar results were obtained each time. Scale bar = 100 μm

Loss of mitochondrial membrane potential (MMP) is a characteristic marker of mitochondrial apoptosis cascade activation [29]. Here, shRNA-induced silencing of REG3A led to mitochondrial depolarization in pBC-1 cells, which was evidenced by the JC-1 transition from red fluorescence aggregates to green fluorescent monomers (Fig. 3F). Additional experimental results showed that apoptosis was induced by REG3A shRNAs and TUNEL-positively stained nuclei were increased (Fig. 3G). The control shC treatment, as expected, failed to provoke caspase-apoptosis activation in pBC-1 cells (Fig. 3A-G). Importantly, apoptosis activation should be the primary mechanism of REG3A shRNA-induced cytotoxicity in breast cancer cells. The caspase-3 specific inhibitor z-DEVD-fmk and the pan-caspase inhibitor z-VAD-fmk each significantly inhibited REG3A silencing (by shREG3A-Sq6)-induced viability reduction (Fig. 3H) and cell death (Fig. 3I) in pBC-1 TNBC cells.

In pBC-2 TNBC cells and immortalized cell lines (MCF-7 and MDA-231), genetic silencing of REG3A by shREG3A-Sq6 (see Fig. 2) led to dramatic viability reduction (Fig. 3J), caspase-3 activation (Fig. 3K), and apoptosis (Fig. 3L); the latter was again evidenced by increased nuclear TUNEL staining (Fig. 3L). Whereas in primary pMEC and established MCF-10 A epithelial cells, silencing of REG3A by shREG3A-Sq6 (see Fig. 2) failed to provoke caspase-3 and apoptosis activation (Fig. 3M and N).

REG3A knockout produces significant anti-breast cancer cell activity

To completely deplete REG3A, the CRISPR/Cas9 knockout (KO) strategy was employed. As shown, mRNA and protein expression of REG3A were depleted in koREG3A pBC-1 TNBC cells (Fig. 4A and B). Contrarily, REG1 mRNA and protein expression was intact (Fig. 4A and B). Testing pBC-1 cell functions demonstrated that CRISPR/Cas9-induced REG3A KO substantially inhibited cell proliferation and decreased the ratio of EdU-positive nuclei (Fig. 4C). Moreover, pBC-1 cell in vitro migration (Fig. 4D) and invasion (Fig. 4E) were significantly suppressed with REG3A KO. Further studies showed that REG3A KO caused the accumulation of JC-1 green fluorescence monomers in pBC-1 cells, supporting mitochondrial depolarization (Fig. 4F). Moreover, the caspase-3 activity (Fig. 4G) and the TUNEL positively-stained nuclei percentage (Fig. 4H) were both significantly increased in koREG3A pBC-1 cells, supporting apoptosis activation. Together, these results showed that REG3A KO led to significant anti-cancer cell activity in primary breast cancer cells.

Fig. 4

REG3A knockout produces significant anti-breast cancer cell activity. The single stable pBC-1 cells, with the lentiviral CRISPR/Cas9-REG3A-KO construct (“koREG3A”) or the CRISPR/Cas9-control construct (“koC”), were established and expression of listed mRNAs and proteins was shown (A and B). Cells were further cultivated for indicated hours, cell proliferation (by measuring EdU incorporation, C), cell migration (D) and invasion (E) were tested; Mitochondrial depolarization (JC-1 green monomers formation, F); Caspase-3 activity (G) and cell apoptosis (TUNEL-positive nuclei percentage, H) were measured as well. Values were mean ± standard deviation (SD, n = 5). *P < 0.05 versus “koC” group. “N. S.” stands for non-statistical difference (P > 0.05). Experiments were repeated five times and similar results were obtained each time. Scale bar = 100 μm

REG3A overexpression results in cancer-promoting activity in breast cancer cells

Since REG3A silencing/KO resulted in potent anti-breast cancer cell activity, we hypothesized that ectopic overexpression of REG3A could then exert opposite functions. As compared with the control cells with the empty vector (“Vec”), REG3A mRNA expression increased over ten fold in oeREG3A-Slc1/-Slc2 pBC-1 TNBC cells (Fig. 5A). REG3A protein upregulation was detected as well (Fig. 5B). The mRNA and protein levels of REG1 were again unchanged (Fig. 5A and B). Cellular functional studies revealed that ectopic overexpression of REG3A strengthened pBC-1 cell proliferation and increased the EdU-incorporated nuclei ratio (Fig. 5C). Moreover, in vitro cell migration (Fig. 5D) and invasion (Fig. 5E) were augmented in oeREG3A-Slc1/-Slc2 pBC-1 cells.

Fig. 5

REG3A overexpression results in cancer-promoting activity in breast cancer cells. The pBC-1 primary breast cancer cells expressing the REG3A-overexpressing construct (oeREG3A-Slc1 and oeREG3A-Slc2, representing two stable selections) or the empty vector (“Vec”) were established, and expression of listed mRNAs and proteins was shown (A and B). Cells were further cultivated for indicated hours, cell proliferation (by measuring EdU incorporation, C), cell migration (D) and invasion (E) were tested; The pBC-2 primary breast cancer cells or established lines (MCF-7 and MDA-231) with the lentiviral REG3A-overexpressing construct (“oeREG3A”) or the empty vector (“Vec”) were established, and expression of REG3A mRNA was shown (F); Cells were further cultivated for designated time periods, cell proliferation (G) and migration (H) were tested similarly. Values were mean ± standard deviation (SD, n = 5). *P < 0.05 versus “Vec” group. “N. S.” stands for non-statistical difference (P > 0.05). Experiments were repeated five times and similar results were obtained each time. Scale bar = 100 μm

The REG3A-overexpressing lentivirus was also added to pBC-2 TNBC cells and immortalized lines (MCF-7 and MDA-231). Following stable cell selection by puromycin, stable cells (“oeREG3A”) were formed. These cells showed significantly upregulated REG3A mRNA expression (Fig. 5F). Similar to the results in pBC-1 cells, oeREG3A also augmented cell proliferation (Fig. 5G) and accelerated cell migration (Fig. 5H) in primary and immortalized breast cancer cells. Together, we showed that REG3A overexpression resulted in cancer-promoting activity in breast cancer cells.

REG3A is vital for Akt-mTOR activation in breast cancer cells

Due to various genetic mutations, Akt-mTOR cascade is often hyper-activated in TNBC, serving as a key protein for cancer progression [30,31,32,33,34,35,36]. We thus analyzed the potential effect of REG3A on Akt-mTOR cascade activation. pBC-1 TNBC cells were PTEN-null cells and showed high basal phosphorylation of Akt (at Ser-473) and S6K (at Thr-389) (Fig. 6A). Importantly, REG3A silencing by different shRNAs (shREG3A-Sq2/3/6, see Figs. 2 and 3) largely inhibited Akt and S6K phosphorylation in pBC-1 cells (Fig. 6A). Total Akt1 and S6K protein levels were unchanged with REG3A knockdown (Fig. 6A). Moreover, in pBC-1 cells, CRISPR/Cas9-induced KO of REG3A (see Fig. 4) also remarkably decreased Akt1 and S6K phosphorylation (Fig. 6B), without affecting total Akt1 and S6K expression (Fig. 6B). Contrarily, in REG3A-overexpressed pBC-1 cells, oeREG3A-Slc1/2 (see Fig. 5), Akt-S6K phosphorylation was significantly augmented (Fig. 6C). The results support that REG3A is important for Akt-mTOR activation in breast cancer cells.

Fig. 6

REG3A is vital for Akt-mTOR activation in breast cancer cells. The pBC-1 primary breast cancer cells, with the applied REG3A shRNA (“shREG3A-Sq2/3/6”, with non-overlapping sequences), the scramble control shRNA (“shC”), the lentiviral CRISPR/Cas9-REG3A-KO construct (“koREG3A”), the CRISPR/Cas9-control construct (“koC”), REG3A-overexpressing construct (oeREG3A-Slc1 and oeREG3A-Slc2, two stable selections) or the empty vector (“Vec”), were cultured and expression of listed proteins was shown (A-C). The pBC-1 primary breast cancer cells with shREG3A-Sq6 were further stably transduced with the viral constitutively-active mutant S473D Akt1 (caAkt1) construct or the empty vector (“Vec”), expression of listed proteins was shown (D). Cells were further cultivated for designated hours, cell proliferation, migration and apoptosis were examined respectively through nuclear EdU staining (E), “Transwell” (F) and nuclear TUNEL staining (G) assays. The oeREG3A-Slc1 pBC-1 primary cells were treated with LY294002 or vehicle control (“Veh”, 0.15% DMSO) for designated hours, expression of listed proteins was shown (H); Cell proliferation (I) and migration (J) were tested using the same methods. Values were mean ± standard deviation (SD, n = 5). *P < 0.05 versus “shC”/“koC”/“Vec” group. #P < 0.05. Experiments were repeated five times and similar results were obtained each time. Scale bar = 100 μm

To understand the role of Akt-mTOR activation in REG3A-mediated breast cancer cell growth, a viral constitutively-active mutant S473D Akt1 (“caAkt1”) construct was stably transduced to shREG3A-Sq6-expressing pBC-1 cells. As shown, caAkt1 restored Akt and S6K phosphorylation in REG3A-silenced pBC-1 cells (Fig. 6D). Expectably, caAkt1 did not affect REG3A protein expression (Fig. 6D). The functional studies showed that caAkt1 largely ameliorated REG3A silencing-induced proliferation arrest (EdU assays, Fig. 6E), migration inhibition (Fig. 6F), and apoptosis (Fig. 6G) in pBC-1 cells. Thus, mediating Akt-mTOR activation could be a primary mechanism of REG3A-driven breast cancer cell growth.

To further support the role of Akt-mTOR cascade in REG3A-mediated actions, the pan PI3K-Akt-mTOR inhibitor LY294002 [37] was utilized. In oeREG3A-Slc1 pBC-1 primary cells (see Fig. 5), treatment with LY294002 (at 1 µM) blocked Akt-S6K1 phosphorylation (Fig. 6H). Once again, REG3A protein expression as well as total Akt1 and S6K1 expression were unaffected by LY294002 (Fig. 6H). After LY294002 treatment, pBC-1 cell proliferation (by calculating EdU-incorporated nuclei percentage, Fig. 6I) and migration (Fig. 6J) were substantially decreased.

REG3A is important for maintaining the integrity of mTOR complexes

Next, we examined the possible underlying mechanism of REG3A-promoted Akt-mTOR activation. The mTOR complex 1 (mTORC1), a multiple-protein complex including mTOR, Raptor, and mLST8, phosphorylates S6K1 and 4EBP1 [38,39,40,41,42,43]. The other mTOR protein complex, mTORC2, is assembled with mTOR, Rictor, Sin1, mLST8, and several others, and phosphorylates Akt (Ser-473) and other AGC kinases [38,39,40,41,42,43]. The co-immunoprecipitation assay results, Fig. 7A, demonstrated that REG3A silencing (by “shREG3A-Sq6”, see Figs. 2 and 3) or knockout ( “koREG3A”, see Fig. 4) potently inhibited mTOR-Raptor association (mTORC1 assemble) and mTOR-Rictor association (mTORC2 assemble) in pBC-1 TNBC cells. However, there was no change in the expression levels of the mTOR, Rictor, and Raptor proteins of the mTOR complex (Fig. 7A, “Input”). Contrarily, in REG3A-overexpresed pBC-1 cells (“oeREG3A-Slc1” and “oeREG3A-Slc2”, see Fig. 5), mTOR-Raptor association (Fig. 7B) and mTOR-Rictor association (Fig. 7B) were both augmented. Once again, mTOR, Rictor, and Raptor expression was unchanged with REG3A overexpression (Fig. 7B, “Input”). Therefore, REG3A could be important for maintaining the integrity of both mTORC1 and mTORC2, thereby promoting Akt-mTOR activation in breast cancer cells.

Fig. 7

REG3A is important for maintaining the integrity of mTOR complexes. The pBC-1 primary breast cancer cells, with shREG3A-Sq6 (“shREG3A”), the lentiviral CRISPR/Cas9-REG3A-KO construct (“koREG3A”), REG3A-overexpressing construct (oeREG3A-Slc1 and oeREG3A-Slc2, two stable selections) or the empty vector (“Vec”), were cultured, mTOR-Raptor-Rictor associations were examined via co-immunoprecipitation (“IP”) assays (A and B), with expression of the listed proteins tested as “Inputs” (A and B). “Ctrl” stands for the parental control cells (A). Values were mean ± standard deviation (SD, n = 5). *P < 0.05 versus “Ctrl”/“Vec” group. Experiments were repeated five times and similar results were obtained each time

ZNF680 is a potential transcription factor of REG3A in breast cancer cells

Given the observed elevation in both mRNA and protein expression of REG3A in breast cancer tissues and cells, our hypothesis centered on the possibility of a transcriptional mechanism contributing to the upregulation. As limited studies have explored the transcription factors associated with REG3A, we conducted a search using the JASPAR transcription factor database. Five transcription factors exhibiting the highest potential binding affinity to REG3A were identified: Foxl2, ZNF680, ZSCAN31, Nr1H2, and Foxo1 (Fig. 8A). To explore their impact on REG3A mRNA expression, we designed siRNAs targeting each of these transcription factors and individually transfected them into pBC-1 TNBC cells. Only Foxl2 siRNA and ZNF680 siRNA led to a noteworthy silencing of REG3A in pBC-1 cells (Fig. 8B). Silencing other transcription factors had no significant effect (Fig. 8B). Notably, ZNF680 siRNA exhibited a greater potency than Foxl2 siRNA in downregulating REG3A (Fig. 8B).

Fig. 8

ZNF680 is a potential transcription factor of REG3A in breast cancer cells. The JASPAR database predicted potential transcription factors for REG3A (A). Following transfection of pBC-1 cells with the specified siRNAs targeting various transcription factors or a scramble non-sense siRNA (siC) for 48 h, the expression of REG3A mRNA was assessed (B). pBC-1 cells were then subjected to lentiviral ZNF680 shRNA (shZNF680-Sq1 or shZNF680-Sq2) and scramble control shRNA (“shC”) treatments (C and D), lentiviral CRISPR/Cas9-ZNF680-KO construct (“koZNF680”) and CRISPR/Cas9 control construct (“Cas9-C”) treatments (E and F), as well as lentiviral ZNF680-expressing construct (oeZNF680) or an empty vector (“Vec”) treatments (G and H), resulting in the establishment of stable cells. The expression of the listed mRNAs and proteins was then evaluated (C-H). Chromatin Immunoprecipitation (ChIP) assay results demonstrated the relative levels of ZNF680-bound REG3A promoter in specified breast tumor tissues (“T”) and matched adjacent normal tissues (“N”) (I) as well as in listed breast cancer cells and pMEC/MCF-10 A cells (J). “Ctrl” stands for the parental control cells. Values were mean ± standard deviation (SD). * P < 0.05 versus “siC” (B). * P < 0.05 versus “shC”/“Cas9-C”/“Vec” cells (C-H). * P < 0.05 versus “N” tissues or pMEC cells (I and J). Experiments were repeated five times and similar results were obtained each time

Subsequently, pBC-1 cells were treated with the lentivirus carrying ZNF680 shRNA (shZNF680-Sq1 and shZNF680-Sq2, utilizing different sequences), and stable cells were generated. In contrast to control pBC-1 cells treated with scramble control shRNA (“shC”), pBC-1 cells with shZNF680 exhibited a marked reduction in ZNF680 mRNA (Fig. 8C) and protein levels (Fig. 8D), accompanied by a significant downregulation of REG3A mRNA (Fig. 8C) and protein (Fig. 8D) expression. As an alternative strategy, pBC-1 cells expressing Cas-9 were stably transduced with a lentiviral CRISPR/Cas9-ZNF680-KO construct, resulting in the development of ZNF680 knockout cells (“koZNF680”). In these koZNF680 pBC-1 cells, both the mRNA (Fig. 8E) and protein (Fig. 8F) expression of ZNF680 was depleted, and REG3A expression was substantially reduced.

To further substantiate our hypothesis, pBC-1 cells were exposed to the lentivirus carrying the ZNF680-expressing construct. Stable cells, “oeZNF680,” were established through puromycin-based selection. These cells exhibited pronounced upregulation of ZNF680 mRNA (Fig. 8G) and protein levels (Fig. 8H). Notably, the overexpression of ZNF680 resulted in a concurrent upregulation of REG3A mRNA (Fig. 8G) and protein expression (Fig. 8H). Importantly, ChIP assay results revealed a significant increase in the binding between ZNF680 protein and the putative REG3A promoter region (as identified in the JASPAR database) in breast cancer tissues from local patients (Fig. 8I). Furthermore, this enhanced binding was consistently observed in various primary and immortalized breast cells (pBC-1, pBC-2, MCF-7, and MDA-231) (Fig. 8J). In contrast, the ZNF680-REG3A promoter binding affinity was relatively low in normal breast tissues (“N”) (Fig. 8I) and in pMEC/MCF-10 A cells (Fig. 8J). These findings indicate that ZNF680 is an important transcription factor for REG3A, and the enhanced binding between ZNF680 and the REG3A promoter may represent a key mechanism contributing to the upregulation of REG3A in breast cancer.

REG3A silencing impedes breast cancer xenograft growth in nude mice

At last, the potential role of REG3A on the growth of breast cancer cells in vivo was explored. The pBC-1 TNBC cells (at five million cells per mouse), with REG3A shRNA (“shREG3A-Sq6”) or control shRNA (“shC”), were injected into the nude mice flanks. The xenograft recordings started 20 days after cell injection, every six days after. The xenograft growth curve results in Fig. 9A demonstrated that the growth of shREG3A-Sq6-expressing pBC-1 xenografts was remarkably inhibited. The volumes of shREG3A-Sq6 pBC-1 xenografts were significantly lower than those of shC-expressing ones (Fig. 9A). Fifty-six days (“Day-56”) after initial pBC-1 cell injection, all xenografts were isolated and measured. As demonstrated, shREG3A-Sq6-expressing pBC-1 xenografts were significantly lighter and smaller than shC ones (Fig. 9B). There was, however, no significant difference in the animal body weights of the two groups of nude mice (Fig. 9C).

Fig. 9

REG3A silencing impedes breast cancer xenograft growth in nude mice. The pBC-1 primary breast cancer cells, with REG3A shRNA (“shREG3A-Sq6”) or scramble control shRNA (“shC”), were subcutaneously (s.c.) injected to the flanks of the nude mice to form xenografts. The pBC-1 xenograft volumes (A) and animal body weights (C) were recorded starting at Day-22 (22 days after cell injection) and every six days after. At Day-56, all pBC-1 xenografts were isolated and weighted (B); At Day-32 and Day-44, one pBC-1 xenograft in each group was isolated and total four xenografts were obtained. Expression of listed mRNAs and proteins in fresh xenograft tissue lysates was tested (D, E, F and J); The relative caspase-3 activity was measured as well (I). Alternatively, pBC-1 xenograft slides were subject to immunohistochemistry (IHC) staining of pAkt (Ser-473) (G). Moreover, nuclear Ki-67/DAPI fluorescence staining (H) and nuclear TUNEL/DAPI fluorescence staining (K) were carried out in the described pBC-1 xenograft slides. Values were mean ± standard deviation (SD). In A-C, ten mice per group (n = 10). For D-K, five random tissue pieces in each xenograft were tested. *P < 0.05 versus “shC” group. “N. S.” indicated no statistical difference (P > 0.05). Scale bar = 100 μm

Thirty-two (“Day-32”) and forty-four (“Day-44”) days after initial pBC-1 cell injection, one pBC-1 xenograft per group was isolated, and a tal of four pBC-1 xenografts were obtained. Part of each xenograft was cut into small pieces, and signaling changes were detected. As demonstrated, REG3A mRNA and protein expression were substantially decreased in shREG3A-Sq6-expressing pBC-1 xenograft tissues (Fig. 9D and E), while REG1A mRNA and protein expression were unchanged (Fig. 9D and E). Activation of Akt-mTOR cascade, or Akt-S6K phosphorylation, was largely inhibited in REG3A-silenced pBC-1 xenograft tissues (Fig. 9F). Part of the xenograft tissues were sectioned into tumor slides, and immunohistochemistry (IHC) results further confirmed Akt (S473 phosphorylation) inhibition after REG3A silencing (Fig. 9G).

The xenograft tissue immunofluorescence images showed that Ki-67-positive nuclei percentage was significantly decreased in pBC-1 xenografts with shREG3A-Sq6, supporting proliferation inhibition in REG3A-silenced xenografts (Fig. 9H). Further studies demonstrated that the Caspase-3 activity was significantly increased in shREG3A-Sq6-expressing pBC-1 xenograft tissues (Fig. 9I), where cleaved Caspase-3 levels were increased (Fig. 9J). Moreover, TUNEL-positive nuclei were increased in shREG3A-Sq6-expressing pBC-1 xenograft slides (Fig. 9K). These results further supported apoptosis activation in REG3A-silenced pBC-1 xenografts. Together, Akt-mTOR inactivation, proliferation inhibition, and apoptosis activation were detected in pBC-1 xenografts after REG3A silencing.