Figure 1.
The distribution of cytoplasmic rods and rings was independent of the Golgi complex and centrosomes, and these structures were not enriched in tubulin or vimentin.
(A) Merged image of HEp-2 co-stained with human anti-RR prototype serum 604/Alexa 488 goat anti-human Ig (green) and rabbit anti-giantin (Golgi marker)/Alexa 568 goat anti-rabbit Ig (red). Rods are often presented adjacent (short arrows) or perpendicular (long arrows) to the nucleus while rings (arrowheads) are found either 1 or 2 to a cell. M, mitotic cell. HEp-2 cells were also co-stained with serum 604/Alexa 488 goat anti-human Ig (green, B,E,H) and different cytoplasmic markers using mouse anti-tubulin (C), anti-vimentin (F), anti-pericentrin (I), followed by Alexa 568 goat anti-mouse Ig (red). Nuclei were counterstained with DAPI (blue). Bar, 10 µm.
Figure 2.
Examples of various intermediate RR structures in HEp-2 cells stained with human serum 604 (green) and counterstained with DAPI (blue).
(A) Figure “8” structure (arrowhead) with a curved rod adjacent to the nuclear envelope; (B) elongated ring; (C) elongated, twisted ring (short arrow); (D) rod with a pin-loop (long arrow); (E) Figure “8” structure (arrowhead); (F) another rod with pin-loop (long arrow). Nuclei were counterstained with DAPI (blue). Bar, 5 µm.
Figure 3.
The expression of rods versus rings was not correlated with the cell cycle.
HEp-2 cells were co-stained with rabbit anti-CENP-F/Alexa 568 goat anti-rabbit IgG (A) and human anti-RR serum 604/Alexa 488 goat anti-human IgG (B). G1 cells had little or no CENP-F staining, whereas late S/G2 and mitotic (M) cells showed strong staining for CENP-F. Nuclei counterstained with DAPI (blue). Bar, 10 µm.
Figure 4.
CTPS1 and IMPDH2 were highly enriched in RR and human anti-RR prototype serum It2006 recognized IMPDH2.
(A) Enrichment of CTPS1, detected by rabbit anti-CTPS1 (green), to RR (arrows) identified by It2006 (red). Nuclei were counterstained with DAPI (blue). (B) IMPDH2 stained by rabbit anti-IMPDH2 (red) localized to RR (arrows) detected by It2006 (green). Bar, 10 µm. (C) Immunoprecipitation (IP) analysis using an extract of [35S]-methionine-labeled K562 cells and rabbit anti-IMPDH2, It2006, and a second human anti-RR serum 609. It2006 recognized a 55 kDa protein band that co-migrated with IMPDH2 immunoprecipitated by rabbit anti-IMPDH2. Serum 609 did not immunoprecipitate the 55 kDa protein. (D) IP-Western blot demonstrated that serum It2006 immunoprecipitated IMPDH2, which was recognized by both mouse monoclonal and rabbit anti-IMPDH2 antibodies. NHS, control normal human serum.
Figure 5.
Inhibition of CTPS1- or IMPDH2-induced formation of rods and rings.
Untreated HEp-2 cells (A) and cells treated with inhibitors 2 mM DON (CTPS inhibitor, B), 2 mM Acivicin (CTPS inhibitor, C), or 2 mM Ribavirin (IMPDH2 inhibitor, D) for 24 h were co-stained with human anti-RR serum It2006 (green) and rabbit anti-giantin (red). Nuclei were counterstained with DAPI (blue). The percentage of cells with RR displayed for each experiment is shown in the lower right corner with the total number of cells counted indicated in parentheses. Bar, 10 µm.
Table 1.
Concentration-dependent induction of RR in HEp-2 cells using different CTPS1 and IMPDH2 inhibitors for 24 h.
Table 2.
Differential time requirements for RR induction in HEp-2 cells treated with different CTPS1 and IMPDH2 inhibitors at 2 mM concentration.
Figure 6.
Inhibition of CTPS1-induced formation of RR in several human cancer cell lines.
Induction of RR observed in human cancer cell lines HeLa (A), CAL 27 (B), THP-1 (C), and HCT116 (D) when treated with 2 mM DON in culture for 24 h, fixed, and co-stained with human anti-RR serum It2006 (green) and rabbit anti-giantin (red). Untreated controls showed few or no RR. The percentage of cells with RR displayed for each condition is shown in the upper right corner with the total number of cells counted indicated in parentheses. Bar, 5 µm.
Figure 7.
Inhibition of CTPS1-induced formation of RR in mouse primary cardiomyocytes, fibroblasts, and endothelial cells.
Mouse primary cardiomyocytes prepared together with fibroblasts and endothelial cells, were treated with 2 mM DON and cultured for 24 h. Cells were co-stained with human anti-RR serum IT2006 (green) and mouse anti-actinin monoclonal antibody (red). Nuclei counterstained with DAPI (blue). Actinin-positive cardiomyocytes (A, B, red) as well as actinin-negative fibroblast or endothelial cells (A, C) all show distinct rods. The percentage of cells with RR displayed is shown in the lower right corner with the total number of cells counted indicated in parentheses (A, all cells; B, actinin-positive cardiomyocytes only; C; actinin-negative fibroblast and endothelial cells). Bar, 10 µm.
Figure 8.
Overexpression of IMPDH2 prevented the induction of RR.
HeLa cells were transfected with either GFP-IMPDH2 (A–C) or a GFP vector (D–F) for 24 h and followed by 2 mM Ribavirin incubation for 24 h. Cells were stained with rabbit anti-IMPDH2 (red) and counterstained with DAPI (Blue). GFP-IMPDH2 transfected cells (T) did not express RR while untransfected cells showed rods (arrows) and rings (arrowheads). RR were detected in both GFP vector transfected and untransfected cells. Bar, 10 µm.
Figure 9.
RR formation increasingly sensitive to Ribavirin when IMPDH protein level is reduced by specific siRNA knockdown.
HeLa cells were either untransfected, transfected with siCTPS1, or siIMPDH2 for 44 h and followed by incubation of Ribavirin at various concentrations for an additional 4 h. (A) IMPDH2 expression was monitored by Western blot using rabbit anti-IMPDH2 antibody. Serial dilutions of untreated HeLa cell lysates (100%, 50%, and 25%) were used to help to comparatively quantitate the degree of protein knockdown. The level of actin was used as a loading control. Relative IMPDH2 protein levels were quantified using image J and normalized to the serial dilution standards. (B) Percentage of cells with RR were quantified using Mayachitra Imago with a range of 131 to 298 cells counted per data point. Cells treated with siIMPDH2 had a significant increase in percentage of cells with RR at 0.2 µM and 0.05 µM of Ribavirin analyzed by Fisher's exact test. (C–D) Representative image of cells at the lowest concentration of Ribavirin transfected with siCTPS1 or siIMPDH2. Arrows, rods; arrowhead, ring. Bar, 5 µm.
Figure 10.
Uninduced mouse embryonic stem cells expressed predominantly cytoplasmic rings that were disassembled during retinoic acid-induced differentiation and reassembled when treated with Acivicin.
(A) Mouse 3T3 cells for comparison shown with both rods (80–90%, arrowhead) and rings (10–20%, arrows). (B) Undifferentiated mouse ESCs shown with ∼90% rings (arrows) and few rods. (C) ESCs treated with RA for 4 days showed no RR. (D) RR were induced in these RA-differentiated cells by treatment with 2 mM Acivicin for 24 h. Cells were stained with human anti-RR serum It2006 (green, A–D) and co-stained with rabbit anti-giantin (red, A, B only). Nuclei counterstained with DAPI (blue). The percentage of ESCs with RR displayed for each condition is shown in the lower right corner with the total number of cells counted indicated in parentheses. Bar, 5 µm.