Efficient immortalization of primary human cells by p16INK4a-specific short hairpin RNA or Bmi-1, combined with introduction of hTERT

Abstract

Activation of telomerase is sufficient for immortalization of some types of human cells but additional factors may also be essential. It has been proposed that stress imposed by inadequate culture conditions induces senescence due to accumulation of p16(INK4a). Here, we present evidence that many human cell types undergo senescence by activation of the p16(INK4a)/Rb pathway, and that introduction of Bmi-1 can inhibit p16(INK4a) expression and extend the life span of human epithelial cells derived from skin, mammary gland and lung. Introduction of p16(INK4a)-specific short hairpin RNA, as well as Bmi-1, suppressed p16(INK4a) expression in human mammary epithelial cells without promoter methylation, and extended their life span. Subsequent introduction of hTERT, the telomerase catalytic subunit, into cells with low p16(INK4a) levels resulted in efficient immortalization of three cell types without crisis or growth arrest. The majority of the human mammary epithelial cells thus immortalized showed almost normal ploidy as judged by G-banding and spectral karyotyping analysis. Our data suggest that inhibition of p16(INK4a) and introduction of hTERT can immortalize many human cell types with little chromosomal instability.

Figure 1

Elevation of p16 during serial passage of several primary human cell types but not of neonatal skin fibroblasts. Whole‐cell protein extracts of human dermal papilla cells (HDPC), human small airway epithelial cells (HSAEC), human mesenchymal stem cells (HMSC), human prostate epithelial cells (HPrEC), human bronchial epithelial cells (HBEC), human dermal keratinocytes (HDK), human foreskin keratinocytes (HFK) and human foreskin fibroblasts (HFF) were collected at the indicated population doublings (PD). Expression levels of (A,C) p16 together with (B) Rb were analyzed by immunoblotting. HeLa cell extract was used as a positive control for p16 in some blots. β‐Actin was used as a loading control. (D) Growth curves for the different cell types.

Figure 2

Bmi‐1 extends the life span and downregulates p16 of human mammary epithelial cells (HMEC), human dermal keratinocytes (HDK) and human small airway epithelial cells (HSAEC) as well as E7. (A,D) HMEC were infected at population doublings (PD) = 15, designated as day 0. Infected cells were selected in the presence of 200 µg/mL G418 and harvested at PD = 24. (B,E) HDK were infected at PD = 12, designated as day 0. Infected cells were selected in the presence of 100 µg/mL G418, and harvested at PD = 21. (C,F) HSAEC were infected at PD = 15, designated as day 0. Infected cells were selected in the presence of 50 µg/mL G418 and harvested at PD = 21. Expression levels of Bmi‐1, p16 and p53 as well as β‐actin were analyzed by immunoblotting.

Figure 3

Weak activation of telomerase by Bmi‐1 in human mammal epithelial cells (HMEC) but not in human dermal keratinocytes (HDK) and human small airway epithelial cells (HSAEC). Telomerase activity of (A) HMEC, HDK and (B) HSAEC transduced with Bmi‐1 was analyzed by TRAP assay. Those without Bmi‐1 and HMEC transduced with the telomerase catalytic subunit hTERT, as well as eight tandem repeats of telomeric sequence (TSR8) and CHAPS buffer alone were used as controls.

Figure 4

Efficient immortalization of human dermal keratinocytes (HDK) and human small airway epithelial cells (HSAEC) by combination of Bmi‐1 and the telomerase catalytic subunit hTERT. (A) Bmi‐1 HDK were infected with hTERT, c‐Myc or control vector at PD = 51. Three days after infection, hygromycin (5 µg/mL) was added for selection. (B) Bmi‐1 HSAEC were infected with hTERT at population doublings (PD) = 18. Two days after infection, puromycin (0.5 µg/mL) was added for selection. (C) Telomerase activity of Bmi‐1 HDK after hTERT or c‐Myc transduction was analyzed as in Fig. 3. TSR8, eight tandem repeats of telomeric sequence.

Figure 5

Expression of p16‐specific short hairpin RNA (shRNA) 1 and 8 and Bmi‐1 as well as E7 allows human mammal epithelial cells (HMEC) to bypass mortality stage 0 (M0). (A) Growth of HMEC introduced with each retroviral construct. HMEC were infected at population doublings (PD) = 9, designated as day 0. Two days after infection, puromycin (0.5 µg/mL) was added to the medium for selection. H1R indicates HMEC infected with the MSCVpuroH1R control vector. (B) Phase contrast image of HMEC transduced with p16 shRNA1 (upper) or H1R (lower) at PD = 12–15. (C) Downregulation of p16 expression by p16 shRNA1, p16 shRNA8 and Bmi‐1. At PD = 12, whole‐cell protein extracts were analyzed by immunoblotting. Three different target sequences (1, 6 and 8) were set up, and were assessed. H1R indicates MSCVpuroH1R control vector‐infected. (D) Expression of p14^ARF was clearly detectable at later passage in HMEC expressing p53 shRNA plus the telomerase catalytic subunit hTERT. Whole‐cell protein extracts of each immortalized HMEC at PD = 18, 30 and 120 were analyzed by immunoblotting.

Figure 6

The telomerase catalytic subunit hTERT overcomes mortality stage 1 (M1) and induces immortalization of human mammal epithelial cells (HMEC). (A–D) Growth curve of the post‐mortality stage 0 (M0) HMEC with or without introduction of hTERT. Post‐M0 HMEC transduced with (A) p16 shRNA1, (B) p16 shRNA8 and (C) Bmi‐1, or (D) those that spontaneously escaped from M0 were infected with LXSN‐hTERT or LXSN retrovirus at population doublings (PD) = 15. Two days after infection, G418 (50 µg/mL) was added to the medium for selection. (E) TRAP assays with various post‐M0 HMEC after transduction of hTERT (+) or LXSN (–). (F,G) Expression levels of several cell cycle‐associated proteins in post‐M0 HMEC at the indicated PD were measured by immunoblotting. HMEC expressing (F) Bmi‐1 plus hTERT and (G) p16 shRNA8 plus hTERT were analyzed with the same samples from those expressing hTERT alone. Note that a mild increase of p16 levels together with some fluctuation of Bmi‐1 levels were observed in HMEC expressing Bmi‐1 plus hTERT at later passages, whereas there was a constant decrease of p16 levels in control cells (F). Also, a mild increase of p21 levels was observed in HMEC expressing p16 shRNA8 plus hTERT and hTERT alone at PD = 45 and later passages (G).

Figure 7

Repression of p16 by Bmi‐1 and p16 shRNA without promoter methylation. (A) Genomic map of the 5′‐CpG islands of the INK4a gene. The bar under the Ink4a locus presents a magnification of the region, from −159 to +233, analyzed in this study. The 35 CpG sites in this region are numbered according to their 5′ to 3′ order in the INK4a genomic sequence and positioned based on their location within the genomic sequence. This region was amplified from bisulfite‐treated DNA. White and black circles represent unmethylated and methylated CpG sites, respectively. For each cell type, 8–19 clones were analyzed. (B) 5‐Aza‐2′‐deoxycytidine treatment does not induce p16 re‐expression in human mammal epithelial cells (HMEC) expressing p16 shRNA plus hTERT. HMEC at population doublings (PD) = 81–132 were treated with 5‐aza‐2′‐deoxycytidine (5 µM) for 72 h and whole‐cell protein extracts were analyzed by immunoblotting.