The promyelocytic leukemia protein PML has a pro-apoptotic activity mediated through its RING domain

Abstract

The promyelocytic leukemia protein PML is known to form nuclear multiprotein complexes which are compromised in several pathogenic conditions including acute promyelocytic leukemia. We show that in cells infected with a single stranded RNA virus, which relocates PML bodies to the cytoplasm, the infected cells are more resistant to serum starvation induced apoptosis than their uninfected counterparts. Antisense PML oligonucleotides increase cell survival under serum deprivation conditions indicating that PML is directly involved in the apoptotic activity. Transient transfection studies have indicated that this pro-apoptotic activity of PML is mediated through the zinc binding region known as the RING finger. Viral attack of PML nuclear bodies appears to allow the virus to deregulate host cell apoptotic machinery in order to establish chronic infection.

Fig. 1

The effect of LCMV infection on the subcellular distribution of the cell cycle marker PCNA. Panels A and B show uninfected and infected NIH 3T3 cells respectively. Cells were infected with 1 plaque forming unit per cell of LCMV Armstrong for 100 h as described previously [18]. The images were collected on a Leica con-focal laser microscope. Magnification was 100X.

Fig. 2

LCMV infection increases survival in serum starved NIH 3T3 cells. Cells were infected as in Section 2. Live cells were counted using trypan blue exclusion. Each data point represents the average trypan blue cell count of three wells in a 24 well plate. Four fields of view were counted per well. Variation is no greater than ± 5% for each point. Data from infected cells are indicated by filled circles and normal cells by open squares.

Fig. 3

A: The effect of antisense PML on survival of serum starved NIH 3T3 cells. Oligonucleotides were added to a total concentration of 3 μM. Cells were grown in 12 well plates in 10% FBS/DMEM, pre-treated for 1 h with the appropriate oligonucleotide or with media only. Cells were then rinsed with serum free media and new serum free media was added. After 18 h, cells were harvested, and counted. Experiments were carried out in quadruplicate and four fields were counted for each. Error bars indicate standard deviations. B: Protein expression levels of PML were monitored using standard techniques in cells as treated in A. Total cell lysates were used where the total protein concentration was normalised prior to loading onto gels. Standard Western blots were obtained and probed with a PML polyclonal antibody (see Section 2). –Se indicates serum withdrawn, +Se, with serum, S, serum withdrawn cells treated with sense oligonucleotide and AS with antisense.

Fig. 4

Over-expression of PML decreases survival of serum starved cells. Cells were transfected with the constructs described in Section 2. A summary of constructs is shown in A. Arrows indicate the position of point mutations and lines represent deletions. Results of cell counts are shown in B. Transfected cells were allowed to recover overnight before serum withdrawal. For full details of mutations see text. Each experimental point is the result of at least 16 measurements. The data shown are from cells serum starved for three days before counting. Error bars indicate the standard deviations.

Fig. 5

NIH 3T3 cells apoptose under serum deprivation conditions. Cells were stained with YOYO, a nucleic acid binding dye, and images are from a confocal laser microscope. White arrows indicate apoptotic bodies. Cells were treated identically to those in Fig. 4. Cells were treated as follows: plus serum control (A), serum withdrawn cells mock transfected (B), serum withdrawn cells transfected with the RING mutant (C) and serum withdrawn cells transfected with PML. The objective was 40 × with zoom of 3.282 (A), 3.282 (B), 3.282 (C) and 3.313 (D).