Supplementary Information

Figure S1. Senescence markers for shMTH1-transduced A549 and validation of MTH1 suppression-related molecular effects via two additional Sigma Mission shRNA constructs.

(A) Representative senescence associated beta-galactosidase (SA-beta-gal) images from the set used for the quantitation in Fig. 1C.

(B) A549 cells develop persistent DSB foci upon MTH1 suppression. Images from the indicated samples represent DAPI-stained nuclei merged with gamma H2AX (red) and 53BP1 (green) foci. Staining was carried out as described in Refs. 9,15. Colocalized foci appear yellow. Quantitation of foci number is shown below the images for experimental n=2. Greater than 50 cells were counted over five separate fields per sample.

(C) A549 cells enter a G1/S arrest upon MTH1 suppression. Processing and analysis was carried out as described in Ref. 51. A representative flow cytometric profile following fixation and propidium iodide staining is shown. Cell cycle fractions were calculated using the BD Accuri software.

(D) Immunoblotting to show efficiency of MTH1 knockdown by the two additional shMTH1 constructs in the indicated cell lines.

(E), (F) Induction of senescence-associated molecular markers, p53/p21cip1, and SA-beta-gal staining in shMTH1-2 and shMTH1-3 transduced A549 cells.

(G) Proliferation assay. Proliferation curves indicate that shMTH1-2 and shMTH1-3 induce very similar proliferation defects in the three indicated cell lines when compared to the original shMTH1 construct.

Figure S2. MTH1 suppression induces cellular senescence in p53-competent H460 cells but its knockdown is not tolerated in the bulk population.

(A) Western blotting to determine extent of MTH1 suppression at the indicated time points. Note that incompletely MTH1-suppressed cells dominate the H460 culture within 12 days following transduction.

(B) Proliferation curves. Cells were seeded in triplicate at a density of 5x104 at the indicated time points following transduction. Cells were derived from counterpart cultures to those assayed in (A). Note the rapid proliferation rate of these cell lines enables outgrowth of cells with incomplete MTH1 expression.

(C) SA-beta-gal staining at the indicated timepoints. Percentage of stained cells is noted below representative images. Note loss of staining at 13d corresponds with loss of MTH1 suppression and its associated proliferation defect in the bulk population.

Figure S3. MTH1 suppression does not induce in vitro cell death.
(A) Cell death was assessed by propidium iodide (PI)/Annexin V staining as per instructions on the ApoAlert Cell Death Assay kit (Clontech). Untransduced A549 cells treated with 2.5 µg/ml puromycin for 18 hours were used as a positive apoptosis control. Cells were analyzed by flow cytometry in the FL1 (FITC for Annexin) and FL3 (PI) channels on a BD Accuri C6 cytometer. Unstained samples were used as controls to set up the four-quadrant gate shown. Note that there appears to be no major cell death or differences in shGFP vs. shMTH1 in the NSCLC cell lines analyzed. This is consistent with the lack of visual apoptotic morphology in the cultures.

(B) Western blotting. The indicated samples were probed for cleaved PARP and caspase-3. Note the lack of pro-apoptotic cleaved caspase-3 bands below the main caspase bands.

Figure S4. In vitro and in vivo validation of the plko-Tet-on. shMTH1 system in A549 cells and xenograft tumor formation by A549 cells with incomplete constitutive MTH1 knockdown.

(A) A549 cells transduced with the plko-Tet-on.shMTH1 construct or the control plko-Tet-on.shLuc construct were immunoblotted for MTH1 expression under the indicated conditions. Note progressive suppression of MTH1 protein over 3 days of 100 µg/ml doxycycline hyclate (Dox) treatment.

(B) Protein lysates derived from A549 xenograft tumors formed either in the absence or presence of doxycycline hyclate were immunoblotted for MTH1 expression.

(C) Western blotting to indicate extent of MTH1 knockdown. Approximately 30 µg protein was immunoblotted.

(D) Proliferation curves from counterpart cells in (C) show that the A549 shMTH1 from this transduction were able to proliferate in vitro.

(E) Tumor growth rates from 106 A549 shGFP or shMTH1 cells measured at indicated time points following subcutaneous injection into Nu/Nu mice.

(F) Tumor volumes per injection site are tabulated. Note the complete lack of tumor growth by the shMTH1 cells at the day 19 timepoint, indicating a role for MTH1 in initiating tumor formation.

Figure S5. Representative tissue sections from xenograft tumors showing cleaved caspase-3 staining

Immunostaining for apoptotic (cleaved caspase-3, Cell Signaling Technology; 1/100 dilution) cells in the indicated samples was performed using rabbit polyclonal and biotinylated goat anti-rabbit IgG secondary antibodies (Vector Laboratories, Burlingame, CA; 1/100 dilution) as previously described in Ref. 53. Representative images are shown from H23 and H358 xenograft tumors.

Figure S6. Representative tissue sections from xenograft tumors showing MTH1 and Ki67 staining

Representative sections are shown from immunohistochemical staining of xenograft tumors in Fig. 3. Note lower MTH1 and Ki67 staining in the shMTH1 tumors relative to shLuc or shGFP tumors (with the exception of the A549 uninduced tumors).

Figure S7. Validation of p53 function in MTH1 suppression-induced DNA damage and OIS.

(A) Alkaline comet assay. As seen with the original shMTH1 construct (shMTH1-1), the two Sigma Mission shMTH1 constructs also induce DNA strand breaks in the A549 but not the H358 cells.

(B) Extent of p53 suppression in A549 cells. Cells were transduced with a puromycin-selectable shp53 or shGFP construct and harvested following selection.

(C) The shp53 cells in (B) were co-transduced with a hygromycin-selectable version of shMTH1 or shGFP. Immunoblotting against the indicated proteins was carried out on approximately 25 µg protein lysates following selection.

(D) Proliferation curve for the shGFP vs. shMTH1-transduced A549-shp53 cells.

(E) Quantitation of SA-beta-gal activity for the indicated A549-shp53 samples.

(F) Alkaline comet assay results are shown for the indicated A549-shp53 samples.

Figure S8. Characterization of MTH1 suppression effects in H1563 cells with and without activated KRAS expression.

(A) Immunoblotting to show differences in MTH1 and pAkt levels between KRAS-mutant A549 cells and KRAS-wt H1563 cells.

(B) ROS levels in A549 vs. H1563 cells. Note lower level of ROS in the wt KRAS-sustaining H1563 relative to mutant KRAS A549 cells.

(C) qPCR analysis of KRAS and MTH1 message levels in the pBp and KRAS isogenic H1563 cells. Note the increase in MTH1 message levels in the KRAS-transduced cells.

(D) Oncogenic KRAS increases proliferation of H1563 cells. Note the significantly increased proliferative rate in the H1563 KRAS (transduced with pBp.KRASV12) cells relative to the pBp vector-transduced counterpart cells.

(E) Oncogenic KRAS elevates total cellular ROS levels. Representative flow FITC profiles run on a BD Accuri cytometer following CM-DCF-DA staining of the indicated samples is shown. Quantitation from two independent experiments is shown to the right.

(F) Total cellular 8-oxoG staining in the indicated H1563 samples. Staining was carried out as described in Refs. 9,15. Note that KRASV12 induces an increase in 8-oxoG staining, consistent with the elevated ROS levels in the KRAS-transduced cells in (D).

(G) H1563 pBp.KRAS cells were cultured for approximately two weeks in either ambient (21%) or 5% O2. Both sets were subsequently transduced with a hygromycin-selectable shGFP or shMTH1 construct. Following approximately one week in selection and culture in either ambient or 5% O2, DNA strand breaks in the two sets of cells were determined via the comet assay. Note that shMTH1 cells cultured at 5% O2 do not exhibit increased DNA strand breaks relative to their shGFP counterparts.

(H) SA-beta-gal assay. Counterpart cultures from (G) were stained for SA-beta-gal activity. Note the lower beta-gal staining in shMTH1 cells cultured at 5% O2.

Figure S9. Characterization of baseline KRAS and ROS levels among the three NSCLC cell lines used in this study

(A) Relative KRAS protein expression among the indicated cell lines.

(B) Relative cellular ROS levels among the three KRAS-mutant cell lines. Note that the relative ROS levels correlate positively with the respective KRAS levels among the three lines.

(C) KRAS levels in the p53-nonfunctional H358 and H23 cells decline under MTH1 suppression. As seen with the original shMTH1 construct, the two Sigma Mission shRNA constructs also lead to a decrease in KRAS protein expression.

(D) H1563 KRAS cells, which are p53-functional and undergo shMTH1-induced OIS, do not show a decrease in KRAS levels upon MTH1 suppression.

(E) KRAS levels decrease in H358 shMTH1 xenograft-derived protein lysates relative to H358 shGFP xenografts lysates. Immunoblotting using 20 µg lysate from the indicated xenograft H358 tumor samples was carried out. Coomassie blue staining was used to indicate protein loading.

(F) Western blotting to determine KRAS expression levels approximately 6 days following shRNA transduction. Note that the KRAS levels are equivalent in the H358 and H23 shGFP/shMTH1 pairs in contrast to the KRAS reduction observed in the 14 day post-transduction blot in Fig. 5A.

(G) Levels of pErk1/2 and total Erk1/2 from Fig. 5A.

(H) Effect of reduced oxygen tension culture on H358 shMTH1 cells. Indicated cell lines were seeded at 5 x 104, following 7 days in 5% oxygen culture. Media was changed every two days on cells not being harvested for counting. Each point was counted in triplicate.

(I) Western blotting to determine KRAS expression levels in indicated H358 cells cultured at 5% or ambient oxygen. Cells were harvested approximately 14 days after being moved to 5% or ambient oxygen culture. There is a marker lane between the ambient and 5% oxygen samples.

(J) Difference in ROS levels between H358 pBn vs. H358 pBn.Myr-Akt cells under MTH1 suppression. Measurement of ROS levels in the H358 pBn and H358 pBn.Myr-Akt cells (left) and H358 pBn and H358 pBn.Myr-Akt cells transduced with either shGFP or shMTH1 (right). Measurements were carried out 18d post transduction with shGFP or shMTH1.