Background: Alternative splicing of the Pkm gene product generates the PKM1 and PKM2 isoforms of the glycolytic enzyme pyruvate kinase. PKM2 expression is associated with embryogenesis, tissue regeneration, and cancer. PKM2 is also the pyruvate kinase isoform expressed in most wild-type adult tissues, with PKM1 restricted primarily to skeletal muscle, heart, and brain. To interrogate the functional requirement for PKM2 during tumor initiation in an autochthonous mouse model for soft tissue sarcoma (STS), we used a conditional Pkm2 allele fl (Pkm2 ) to abolish PKM2 expression. Results: PKM2 deletion slowed tumor onset but did not abrogate eventual tumor outgrowth. PKM2-null sarcoma cells expressed PKM1 with tumors containing a high number of infiltrating PKM2 expressing stromal cells. End-stage PKM2-null tumors showed increased proliferation compared to tumors with a wild-type Pkm2 allele, and tumor metabolite analysis revealed metabolic changes associated with PKM2 loss. Conclusions: While PKM2 is not required for soft tissue sarcoma growth, PKM2 expression may facilitate initiation of this tumor type. Because these data differ from what has been observed in other cancer models where PKM2 has been deleted, they argue that the consequences of PKM2 loss during tumor initiation are dependent on the tumor type. Keywords: PKM2, Soft tissue sarcoma, Cancer Background heart, brain, and skeletal muscle. PKM2, on the other Proliferating cells, including cancer cells, favor aerobic hand, includes exon 10 but not exon 9. Expression of glycolysis and display increased glucose consumption the PKM2 isoform is widespread in adult epithelial tis- and lactate production compared to many sues and has been associated with embryogenesis, tissue non-proliferating cells. Expression of the M2 isoform of regeneration, and cancer [1, 2]. Importantly, expression pyruvate kinase (PKM2), the glycolytic enzyme that cata- of PKM2 impacts the fate of glucose, in part through lyzes the last step in glycolysis, has been implicated in regulated enzymatic activity. While PKM1 is a constitu- promoting this metabolic state of cancer cells. tively active enzyme, PKM2 enzymatic activity can be The pyruvate kinase gene, Pkm, encodes the PKM1 inhibited by a variety of cellular signaling events, many and PKM2 isoforms, which are the result of alternative of which are associated with proliferation [3–7]. Consist- splicing of mutually exclusive exons. The PKM1 isoform ent with this notion, activation of PKM2 enzymatic ac- includes exon 9 but not exon 10 and shows a pattern of tivity in cancer cells through either exogenous expression that is restricted in the adult largely to the expression of PKM1 or use of small molecule activators leads to delayed xenograft tumor formation [8, 9]. PKM2 is highly expressed in many tumor types, and a * Correspondence: firstname.lastname@example.org; email@example.com number of metabolic and non-metabolic functions in David H. Koch Institute for Integrative Cancer Research and Department of cancer have been attributed to PKM2 [1, 5, 8, 10–15]. Biology, Massachusetts Institute of Technology, Cambridge, MA 02139, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Dayton et al. Cancer & Metabolism (2018) 6:6 Page 2 of 9 However, whether PKM2 is required for tumor forma- on a CryoStar NX70 (Thermo Scientific). PKM2 anti- tion remains a controversial question. Recent studies of body (1:500, CST 4053) was used. PKM2 function in autochthonous mouse models of can- cer have revealed a complicated picture of PKM2 func- Immunohistochemistry (IHC) tion in cancer. For example, loss of PKM2 in a leukemia IHC was performed on 4-μm-thick sections using ABC and breast cancer model led to contrasting results: while Vectastain kit (Vector Labs) with antibodies to PKM1 PKM2 loss limited progression in one case, it led to ac- (1:500, CST 7067), PKM2 (1:500, CST 4053), and celerated tumor growth in the other [16, 17]. PKM2 de- phospho-Histone H3 (1:200, CST 9701). The staining letion also accelerated tumor growth in a mouse was visualized with DAB (Vector Labs, SK-4100), and medulloblastoma model . Loss of PKM2 in a colon the slides were counterstained with hematoxylin. For cancer model driven by APC loss did not alter tumor dual staining, DAB and ImmPACT VIP (Vector Labs, initiation, growth, or progression . Finally, germline SK-4605) were used. Hematoxylin and eosin (H&E) loss of PKM2 led to late onset of hepatocellular carcin- staining was performed on a Varistain Gemini automated oma, at least in part through non-cell autonomous slide stainer (Thermo Shandon). Soft tissue sarcoma hu- means . In no cases was PKM2 absolutely required man tissue array, SO2081, was purchased from US Bio- for tumor formation; however, the different response to max Inc. (Rockville, MD USA). The TMA was scored PKM2 loss in various tumors might be indicative of dis- for PKM1 and PKM2 intensity. Tumors that showed no tinct tissue metabolic requirements. positive staining were given a score of 0, those with weak fl In this study, we used a conditional Pkm2 allele (M2 ) staining were given a score of 1, those with strong posi- to assess the requirement for PKM2 in a mouse model tive staining were given a score of 2, and those with very of soft tissue sarcoma. In this model, Cre-mediated acti- strong staining were given a score of 3. vation of oncogenic Kras and loss of p53 (KP) in the ImageJ software was used for quantification of pHH3 hind limb of mice leads to the formation of soft tissue staining, and four to five 10× fields were counted per sarcomas within 3 months . Loss of PKM2 in this tumor. context led to delayed tumor initiation and decreased fl/fl tumor penetrance. The tumors that formed in KP M2 RNA isolation and qRT-PCR mice expressed PKM1 instead of PKM2. Metabolite ana- RNA was isolated from flash frozen crushed sarcoma tis- −/− lysis of KP M2 sarcoma tissue showed changes in the sue following manufacturer’s instructions for Trizol pool sizes of several metabolites, arguing that metabol- (Invitrogen). One to 2 μg of RNA was reverse tran- ism in these tumors changes as a result of PKM2 dele- scribed following manufacturer’s instructions for High tion and consequent PKM1 expression. Capacity cDNA Reverse Transcription Kit (Applied Bio- systems). KAPA SYBR Fast mix (Kapa Biosystems, Methods KK4604) was used for quantitative PCR with primers Mouse strains and treatments listed below. Expression levels were calculated relative to All animal work was conducted in accordance with a β-actin and normalized to WT samples. protocol approved by the MIT Committee on Animal Care. KP and M2 flox mouse strains have been described Target Forward Reverse gene previously [17, 21]. All animals were maintained on a mixed C57BL/6J × 129SvJ genetic background. β-actin GGCATAGAGGTCTTTACGGATGTC TATTGGCAACGAGCGGTTCC Sarcoma tumorigenesis was initiated as previously de- Pkm TGACACCTTCCTGGAACACA TTCAGCATCTCCACAGATCG scribed by intramuscular infection of mice with Pkm1 GTCTGGAGAAACAGCCAAGG TCTTCAAACAGCAGACGGTG Ad5-CMV-Cre (University of Iowa Gene Transfer Core) Pkm2 GTCTGGAGAAACAGCCAAGG CGGAGTTCCTCGAATAGCTG [20, 22]. For histologic analysis, tissues were fixed in 4% formalin, embedded in paraffin, and sections were stained with hematoxylin and eosin (H&E). Immunoblotting Immunofluorescence Sarcoma tissue lysates were generated from flash frozen For immunofluorescence analysis, mice were perfused ground whole tumors following lysis with ice cold RIPA with PBS, followed by perfusion with 2% paraformalde- buffer, supplemented with HALT phosphatase and hyde (PFA). After an overnight fixation in 4% PFA at protease inhibitors (Thermo-Scientific, PI-78420 and 4 °C, tissues were placed in 30% sucrose overnight at 87786). The following antibodies were used for immuno- 4 °C with shaking. Tissue was allowed to equilibrate in blotting: anti-Hsp90 (1:1000; CST 4877), anti-PKM1 OCT:PBS(1:1) solution andembeddedinOCT (1:5000; Cell Signaling, 7067), and anti-PKM2 (1:5000; (Tissue-Tek). Frozen sections (10 μm thick) were cut Cell signaling, 4053). Dayton et al. Cancer & Metabolism (2018) 6:6 Page 3 of 9 Metabolite extractions and LC/MS analysis Results Metabolites were extracted from ground flash frozen PKM2 is expressed in mouse and human sarcomas whole tumors by methanol extraction. For polar To inform our study of PKM isoform requirements in metabolite analysis, dried metabolite samples were soft tissue sarcomas (STS), we examined the expression resuspended in 100 mL of water, centrifuged at 15,000×g patterns of PKM1 and PKM2 in normal adult skeletal at 4° C for 10 min, and 1 mL of supernatant was muscle and STS tissue. This was assessed by injected for LC/MS analysis using QExactive Orbitrap immunohistochemistry (IHC) staining using antibodies mass spectrometer as previously described [23, 24]. specific to either PKM1 or PKM2. IHC staining of both mouse and human skeletal muscle confirmed previous reports that PKM1 is the primary isoform expressed in EdU incorporation this tissue (Fig. 1a, b,Additional file 1: Figure S1A) . Mice were given one pulse of intraperitoneally injected We next stained 48 primary human rhabdomyosarcomas EdU, and tissue was harvested after 3 h. EdU was including embryonal, spindle cell, alveolus, and visualized using Click-iT EDU AlexFluor-647 kit for IF pleomorphic rhabdomyosarcomas (undifferentiated (Invitrogen C10340). For EdU incorporation in cell lines, sarcomas) for PKM1 and PKM2 (Fig. 1c). Consistent with cells were grown on a monolayer and treated with EdU the PKM expression pattern of their tissue of origin, the for 1 h. majority of the examined rhabdomyosarcomas showed positive staining for PKM1 (32/48, 67%). Less than half of the examined rhabdomyosarcomas showed positive M2/M1 expression-associated survival analyses in human staining for PKM2, arguing against an absolute sarcoma patients requirement for PKM2 expression in this tumor type RNA-seq gene expression profiles (n = 265) and relevant (Additional file 1:Figure S1B). PKM1 andPKM2 IHC clinical data (n = 261) for sarcoma (SARC) patients were intensity scoring revealed heterogeneity in the distribution obtained from the Cancer Genome Atlas data portal of PKM2-expressing tumors according to rhabdomyosar- (TCGA; cancergenome.nih.gov/). Patients with primary coma subtypes (Fig. 1d). In contrast to the other three ex- tumor samples and associated survival data (n = 259) amined subtypes, pleomorphic rhabdomyosarcomas were ranked by the ratio of the RPKM (reads per Kb per contained a high percentage of PKM2-positive tumors million mapped reads) values of PKM2 and PKM1 (10/12, 83%). The majority of these tumors also expressed isoform-specific exons. Values greater than one indicate PKM1 (10/12, 83%). higher expression of the M2 isoform compared to the In addition to our IHC analysis of PKM1 and PKM2 M1 isoform, with higher values indicating greater M2 expression in primary human sarcomas, we also isoform expression. Kaplan–Meier survival analysis was analyzed human sarcoma gene expression data from The conducted to compare patients with high M2/M1 ex- Cancer Genome Atlas (TCGA) to determine whether pression ratio (top 20th percentile) to those in the bot- the expression ratio of PKM2 to PKM1 could predict tom 20th percentile (n = 51 within each group). The patient outcome . Across > 200 sarcoma patients log-rank test was used to assess significance. Cox pro- within TCGA, those within the highest quintile for portional hazards analysis was conducted across all pa- PKM2/PKM1 expression ratio had a significantly shorter tients in the TCGA SARC cohort (n = 261) to assess the median 5-year survival when compared to either those prognostic significance of the M2/M1 expression ratio. within the lowest quintile for PKM2/PKM1 expression Univariate Cox regression analysis for individual charac- ratio, or when compared with all other patients com- teristics was performed to benchmark single-variable bined (Fig. 2a, Additional file 2: Figure S2). Furthermore, hazard ratios. Subsequently, a multivariable model was Cox regression analysis using this TCGA sarcoma pa- implemented to estimate the prognostic value of the tient cohort showed that a high PKM2/PKM1 ratio was M2/M1 ratio while controlling for other patient charac- independently prognostic of shorter survival when con- teristics (age, gender). Hazard ratio proportionality as- trolling for age and gender (Fig. 2b). sumptions for the Cox regression model fit were validated by testing for all interactions simultaneously PKM2 loss constrains sarcoma tumor initiation (p = 0.0856). Interaction between significant covariates We were intrigued by the disparate PKM expression (M2/M1 ratio and age) was tested using a likelihood ra- patterns in normal and transformed skeletal muscle tio test (LRT) to contrast a model consisting of both co- tissue. While normal skeletal muscle exclusively variates with another model consisting of both expressed PKM1, a sizeable number of human primary covariates and an interaction term. All survival analyses tumors expressed both PKM1 and PKM2. This is in were conducted within a 5-year survival timeframe using contrast to breast cancer, leukemia, and colon cancer the survival package in R (www.r-project.org). where the normal and tumor tissue both express PKM2 Dayton et al. Cancer & Metabolism (2018) 6:6 Page 4 of 9 Fig. 1 Human and mouse soft tissue sarcomas express PKM2. Representative images of IHC for PKM1 and PKM2 in human (a) and mouse (b) skeletal muscle and human rhabdomyosarcoma (c). Corresponding H&E images are shown. Scale bars, 20 μm. d Quantification of PKM1 and PKM2 staining intensities in human rhabdomyosarcoma tumors relative to rhabdomyosarcoma subtype (embryonal, spindle cells, alveolus, pleomorphic). Score 0 = no staining, score 1 = weak, score 2 = positive, or score 3 = strong LSL-G12D/+ fl/fl [16, 17, 19]. This, in addition to the prognostic value of Kras ; p53 (KP) mouse model by viral deliv- PKM2/PKM1 revealed by our analysis of the TCGA ery of Cre recombinase to skeletal muscle tissue initiates sarcoma cohort, led us to postulate that expression of the development of high-grade sarcomas with myofibro- PKM2 may play a role in malignant transformation of blastic differentiation . Cross-species gene expression skeletal muscle tissue to form STS. analysis has shown that tumors from the KP model To directly test this hypothesis in a well-defined gen- closely resemble human undifferentiated pleomorphic etic context, we used a genetically engineered mouse sarcomas  and our IHC analysis of human pleo- model of soft tissue sarcoma. Conditional activation of morphic sarcomas showed that this subtype of STS is oncogenic Kras and inactivation of p53 in a enriched for PKM2-expressing tumors (Fig. 1d). Fig. 2 A high PKM2/PKM1 expression ratio predicts poor prognosis. a Kaplan–Meier 5-year survival analysis comparing patients in the top quintile of PKM2/PKM1 expression ratio (n = 51; red) and those in the bottom quintile (n = 51; blue). Log-rank test p value is shown. b Results of univariate and multivariable Cox proportional hazards model on overall survival in the TCGA sarcoma cohort (all patients). Increasing PKM2/PKM1 expression ratio shows a significant association with poor survival after controlling for other characteristics Dayton et al. Cancer & Metabolism (2018) 6:6 Page 5 of 9 +/+ Importantly, IHC for PKM1 and PKM2 on STS derived lysates compared to KP M2 tumor lysates (Fig. 4c). from the KP model showed that most cells in all of the Interestingly, normal skeletal muscle tissue showed a examined tumors expressed both PKM1 and PKM2 significantly higher V for pyruvate kinase enzyme max (Additional file 1: Figure S1C). activity than tumors of either genotype. These findings To investigate the role of PKM2 in sarcoma are consistent with what has been described in adult development in the KP model, we crossed mice with a tissue isolated from germline Pkm2 knockout mice  fl conditional allele of Pkm2 (M2 )to KP mice and and suggest a common mechanism of decreased total performed intramuscular infections of cohorts of KP Pkm expression to compensate for replacement of fl/fl +/+ M2 and KP M2 mice with Ad-Cre (Fig. 3a). In con- PKM2 by the more active PKM1 isozyme. trast to similar experiments performed in other autoch- To further characterize the PKM expression patterns +/+ fl/fl thonous mouse cancer models, we found that deletion of sarcomas from KP; M2 and KP; M2 mice at the of PKM2 concomitant with sarcoma initiation delayed cell and tissue level, we performed IHC for PKM1 and tumor onset and decreased penetrance of the model PKM2 (Fig. 4d). Consistent with Western blot analysis, +/+ (Fig. 3b)(p = 0.003 by log-rank test). Deletion of PKM2 KP M2 tumors primarily and uniformly expressed −/− delayed the date of tumor onset by approximately 8 days PKM2. In contrast, KP M2 tumors consisted +/+ flfl (56.45 days for KP M2 versus 64.05 days for KP M2 ) primarily of PKM1-expressing cells and also contained a (Fig. 3c). number of cells that had high expression of PKM2. To To characterize the PKM expression patterns of determine whether the PKM2-expressing cells in KP +/+ fl/fl −/− sarcomas arising in KP M2 and KP M2 mice, we M2 tumors were cancer cells, we performed immuno- performed Western blot analysis of PKM1 and PKM2 fluorescence (IF) for PKM2 (Fig. 4e). We used KP mice +/+ expression in primary tumors. While KP M2 tumors for this study that carried a fluorescent Cre-reporter al- fl/fl LSL-tdTomato primarily expressed PKM2, tumors from KP M2 mice lele (Rosa26 ), allowing us to distinguish −/− (KP M2 tumors) lost PKM2 expression and expressed tumor cells from stroma through tumor-specific expres- PKM1 (Fig. 4a). Consistent with our protein analysis, sion of the red-fluorescent protein, tdTomato . qRT-PCR analysis of the same tumors showed very low Consistent with efficient deletion of PKM2 and transcript levels for Pkm2 accompanied by a dramatic consequent expression of PKM1 in tumor cells, we −/− increase in the transcript levels for Pkm1 in KP M2 found that PKM2 was expressed exclusively in the +/+ tumors relative to KP M2 tumors (Fig. 4b). Of note, tdTomato-negative stromal cells. Collectively, these data −/− the levels of Pkm1 in KP M2 tumors were lower than argue that while loss of PKM2 concomitant with KP sar- the levels of Pkm1 in wild-type untransformed skeletal coma initiation delays tumor onset, it does not fully ab- muscle tissue. rogate the eventual outgrowth of tumors in most cases. This quantitative assessment of mRNA abundance also revealed that the levels of total Pkm transcripts were Expression of PKM1 in late-stage soft tissue sarcomas −/− +/+ reduced in KP M2 tumors compared to KP M2 does not limit proliferation tumors. The observed decrease in total Pkm transcripts Expression of PKM1 in tumor cells has been shown to −/− in KP M2 tumors was accompanied by a reduction in limit proliferation and progression in some mouse the maximum rate of reaction (V ) when pyruvate models of cancer . Furthermore, in breast cancers max −/− kinase enzyme activity was assessed in KP M2 tumor with PKM2 deletion, PKM1 expression was only +/+ fl/fl Fig. 3 PKM2 loss delays KP soft tissue sarcoma initiation. KP M2 or KP M2 mice were injected intramuscularly with Ad5-CMV-Cre (a) and the onset of palpable sarcomas was monitored, p value by log-rank test is 0.003 (b). The percentage of total mice (n) with sarcomas by 62 days (gray box) is indicated. c Time for palpable tumor formation, p value using unpaired t test is 0.04 Dayton et al. Cancer & Metabolism (2018) 6:6 Page 6 of 9 +/+ −/− Fig. 4 PKM2-deficient sarcoma tumor cells express PKM1. a Western blot analysis for PKM1 and PKM2 on lysates from KP M2 or KP M2 sarcoma +/+ −/− tissue. Hsp90 was used as a loading control. b Expression of Pkm1, Pkm2,and total Pkm mRNA in KP M2 , KP M2 , or wild-type skeletal muscle tissue was assessed by qRT-PCR. Mean values ± SEM and p values using unpaired t test are shown, *p < 0.05; **p < 0.01; ****p < 0.0001. c Pyruvate +/+ −/− kinase enzymatic activity of KP M2 , KP M2 , or wild-type skeletal muscle tissue, p value using unpaired t test is shown, *p <0.05. d Representative +/+ −/− images for PKM1 and PKM2 IHC in KP M2 and KP M2 sarcoma tissue. Asterisks (*) indicate skeletal muscle fibers. Scale bars, 20 μm. −/− e Representative images of IF for tdTomato and PKM2 on KP M2 sarcoma tissue observed in the non-proliferating population of tumor Despite the longer time to palpable tumor formation fl/fl cells . To ask whether the proliferating cells in KP and the decreased tumor penetrance in KP M2 mice, −/− M2 sarcomas expressed PKM1, we performed there was no significant difference in the tumor weight fl/fl +/+ two-color IHC for PKM1 or PKM2 and the proliferative at time of death of KP M2 mice relative to KP M2 marker, phospho-histone H3 (pHH3) (Fig. 5a). Contrary mice (Additional file 3: Figure S3C). To determine −/− to the conclusions from breast cancer, we found that a whether KP M2 tumors had higher rates of proliferation +/+ large portion of PKM1-expressing tumor cells in KP than KP M2 tumors, we quantified pHH3-positive cells −/− M2 sarcomas also expressed pHH3. Thus, once these in tumors collected at the time of death from mice of each −/− tumors have been initiated, PKM1 expression does not genotype. Unexpectedly, we found that KP M2 tumors +/+ appear to abrogate tumor cell proliferation. Analysis of had more pHH3-positive cells compared to KP M2 tu- 5-ethynyl-2′-deoxyuridine (EdU) incorporation in mors, consistent with the notion that PKM1 expression in tdTomato-positive tumor cells was also consistent with late-stage KP sarcomas does not abrogate tumor prolifera- this conclusion (Fig. 5b), as was an analysis of proliferat- tion or progression (Fig. 5c). These findings are also con- ing cells in cell lines derived from primary tumors in KP sistent with PKM2 loss promoting proliferation in some +/+ fl/fl M2 and KP M2 mice (Additional file 3: Figure other tissues [2, 17, 18]. S3A). Furthermore, expression of PKM1 in sarcoma cell −/− lines derived from KP M2 tumors did not affect the PKM2 loss in soft tissue sarcomas leads to metabolic changes rates of in vitro proliferation as measured through cu- The fact that deletion of PKM2 in our KP model of soft mulative populations doublings, which were variable tissue sarcoma led to delayed tumor onset and across cell lines but did not correlate with Pkm genotype decreased penetrance but did not appear to limit (Additional file 3: Figure S3B). proliferation or progression of late-stage tumors raised Dayton et al. Cancer & Metabolism (2018) 6:6 Page 7 of 9 Fig. 5 PKM2-deficient sarcoma tumors have increased proliferation and an altered metabolic profile. a Representative images of dual-color IHC for +/+ −/− PKM1 or PKM2 (brown) and pHH3 (purple) on KP M2 and KP M2 sarcoma tissue. Scale bars, 20 μm. b Representative images of IF for −/− +/+ −/− tdTomato, PKM1, and EdU on KP M2 sarcoma tissue. c Quantification of pHH3+ cells in KP M2 and KP M2 sarcomas. Representative images of the IHC for pHH3 are shown. Scale bars, 20 μm, n = 6 tumors per genotype. p value using unpaired t test is shown, *p < 0.05. +/+ −/− d Heatmap showing measured metabolites that were significantly different between KP M2 and KP M2 sarcoma tissue. A full list of the measured metabolites is included in Additional file 4: Table S1 the possibility that KP M2−/− tumors had arisen While the exact cell of origin for soft tissue sarcoma in through an adaptive process that might modulate the ef- this mouse model is not well defined, several studies fects of PKM2 loss and consequent expression of PKM1. suggest that a diverse array of skeletal muscle cell types, Therefore, we hypothesized that the adaptive process including mature skeletal muscle cells and skeletal −/− that had allowed for KP M2 tumor outgrowth might muscle progenitor cells, can give rise to sarcomas with involve changes in glucose metabolism that could be ap- comparable frequency in this model [29, 30]. The het- parent as changes in metabolite pool sizes. Therefore, erogeneity in potential cells of origin for sarcomas raises we conducted targeted metabolomics analysis of primary the possibility that the consequences of PKM2 loss in +/+ fl/fl tumors derived from KP M2 and KP M2 mice. The this cancer type are dependent on the specific cell type metabolite profiles of primary tumor tissue revealed sig- that gives rise to the tumor. We have found that while nificant differences in levels of five metabolites in KP isolated Pax7+ and Sca1+ skeletal muscle progenitor −/− +/+ M2 tumors compared to KP M2 tumors: FAD, cells express appreciable levels of Pkm1 mRNA, they NADP, and allantoin, which were increased, and aspar- also express high levels of Pkm2 mRNA (data not tate and fumarate, which were decreased (Fig. 5d, shown). −/− Additional file 4: Table S1). Although it is unclear how The fact that KP M2 sarcomas were ultimately +/+ these metabolic changes are related to PKM2 loss, it is comparable in size to KP M2 sarcomas shows that notable that allantoin is a purine catabolite and aspartate the consequences of PKM2 loss can be overcome. plays an important role in purine biosynthesis. Moreover, the metabolic changes that we observe in KP −/− M2 tumors might be indicative of a compensatory Discussion change in the metabolic program of these cells +/+ Here, we show that PKM2 loss concomitant with tumor compared to KP M2 tumor cells. It is interesting to initiation leads to delayed sarcoma tumor onset, which note that while PKM1 expression in PKM2-null breast is ultimately overcome and accompanied by changes in tumors was shown to be limited to a subpopulation of metabolism. The delay in tumor initiation that we non-proliferating cells , PKM1 was uniformly −/− observe upon loss of PKM2 suggests that tumor expressed in KP M2 sarcomas and did not abrogate initiation in this model is partially dependent on the proliferation in these cells. This may represent a better activation of PKM2 expression in tumor-initiating cells. tolerance for PKM1 expression in non-epithelial cancers. Dayton et al. Cancer & Metabolism (2018) 6:6 Page 8 of 9 Consistent with these in vivo data, sarcoma cell lines de- Additional file 4: Table S1. Raw data for LC-MS metabolomics on KP −/− rived from KP M2 tumors also expressed PKM1, and M2+/+ and KP M2-/- sarcomas. Related to Fig. 5. This table contains all the LC-MS metabolomics data. Includes normalization and statistical this did not affect their rates of in vitro proliferation. analysis for each measured metabolite. (XLSX 140 kb) +/+ The fact that the in vitro proliferation of KP M2 and −/− KP M2 cell lines were not significantly different even Acknowledgements −/− though end-stage KP M2 tumors contained a higher We thank Kim Mercer for technical assistance; Elizaveta Freinkman and the +/+ number of proliferating cells than end-stage KP M2 Whitehead Metabolomics Core for the metabolite analysis. We also thank David Feldser and the entire Jacks and Vander Heiden labs for helpful tumors may point to a role for the microenvironment in discussions and experimental assistance. We thank the Koch Institute Swanson −/− allowing end-stage KP M2 tumors to catch up to their Biotechnology Center (SBC) for the technical support, specifically Denise +/+ KP M2 counterparts. Further investigation is needed Crowley and Kathleen Cormier in the Hope Babette Tang (1983) Histology −/− Facility. The authors wish to dedicate this paper to the memory of Officer Sean to delineate the exact mechanism by which KP M2 Collier for his caring service to the MIT community and for his sacrifice. sarcoma tumors overcome PKM2 loss. This study adds to a growing body of evidence that the Funding This work was partially supported by the Cancer Center Support Grant importance of PKM2 expression in tumor cells varies P30CA14051 from the National Cancer Institute and by grants from the across tumor types and might be indicative of different Howard Hughes Medical Institute (T.J., M.G.V.H.), the Burroughs Wellcome tumor-specific metabolic requirements [16, 17, 19]. Fund (M.G.V.H.), the Ludwig Center at MIT (M.G.V.H.), Stand-up to Cancer (M.G.V.H.), the MIT Center for Precision Cancer Medicine, and the Department Importantly, these results have general implications for of Health and Human Services R01CA168653 (M.G.V.H.). T.L.D. was partially the use of therapeutic strategies that target PKM2. The supported by a National Defense Science and Engineering Graduate (NDSEG) success of PKM2-targeted therapies is likely to depend Fellowship Award. V.G. was supported by a Jane Coffin Childs Memorial Fund Postdoctoral Fellowship. T.J. is a Howard Hughes Investigator, the David H. on both the tumor type and the stage at which a tumor Koch Professor of Biology, and a Daniel K. Ludwig Scholar at MIT. is treated. Availability of data and materials All data generated or analyzed during this study are included in this Conclusions published article and its supplementary information files. Our study found that PKM2 loss at the time of tumor Authors’ contributions initiation in a mouse model of STS driven by activation TLD, VG, CAL, and KMM performed and/or supervised experiments. RTB and of oncogenic Kras and loss of p53 leads to delayed AB were involved in the data analysis. VG and CAL gave constructive input tumor onset and decreased tumor penetrance. on the experimental design and data analysis. TLD, TJ, and MGVH conceived and designed the study and were involved in the interpretation of datasets Nonetheless, PKM2 was not required for eventual tumor −/− and writing the manuscript. All authors read and approved the final formation or outgrowth. KP M2 tumors express manuscript. PKM1 and contain infiltrating stromal cells that are Ethics approval PKM2-positive. PKM1 expression in these tumors did All animal work was conducted in accordance with a protocol approved by not abrogate tumor cell proliferation and, in fact, the MIT Committee on Animal Care. −/− end-stage KP M2 tumors contain more cancer cells Competing interests that stain positive for proliferation markers than +/+ −/− M.G.V.H. is a consultant and scientific advisory board member for Agios end-stage KP M2 tumors. We show that KP M2 Pharmaceuticals and Aeglea Biotherapeutics. The other authors declare that tumors have an altered metabolite profile, suggesting they have no competing interests. metabolic changes accompany loss of PKM2 expression in these tumors. Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Additional files Author details Additional file1: Figure S1. Quantification of PKM1 and PKM2 staining David H. Koch Institute for Integrative Cancer Research and Department of intensities shown as percent of tissue cores scored in 16 normal human Biology, Massachusetts Institute of Technology, Cambridge, MA 02139, USA. skeletal muscle samples (A) and 48 primary human rhabdomyosarcomas Rodent Histopathology Core, Harvard Medical School, Boston, MA 02111, (B). Score 0 = no staining, Score 1 = weak, Score 2 = positive, or Score 3 = USA. Department of Medical Oncology, Dana-Farber Cancer Institute, strong. (C) Representative images of IHC for PKM1 and PKM2 in KP Boston, MA 02115, USA. Howard Hughes Medical Institute, Massachusetts mouse sarcoma tissue. Corresponding H&E images are shown. Scale bars, Institute of Technology, Cambridge, MA 02139, USA. 20 μm. (JPG 1314 kb) Received: 11 February 2018 Accepted: 9 May 2018 Additional file 2: Figure S2. (A) Kaplan–Meier 5-year survival analysis comparing patients in the top quintile of PKM2/PKM1 expression ratio (n = 51; red) and all other patients combined (n = 208; blue). Log-rank test References p value is shown. (JPG 683 kb) 1. Mazurek S. Pyruvate kinase type M2: a key regulator of the metabolic Additional file 3: Figure S3. (A) Representative images of IF for +/+ −/− budget system in tumor cells. Int J Biochem Cell Biol. 2011;43:9 69–980. tdTomato, PKM1 or PKM2, and EdU on KP M2 or KP M2 sarcoma cell +/+ −/− 2. Dayton TL, Gocheva V, Miller KM, Israelsen WJ, Bhutkar A, Clish CB, Davidson lines. (B) Cumulative populaion doublings of KP M2 or KP M2 +/+ −/− SM, Luengo A, Bronson RT, Jacks T, Vander Heiden MG. Germline loss of sarcoma cell lines, n =3 KP M2 cell lines and n =4 KP M2 cell lines. PKM2 promotes metabolic distress and hepatocellular carcinoma. Genes (C) Sarcoma weight in grams. (JPG 2694 kb) Dev. 2016;30:1020–33. Dayton et al. Cancer & Metabolism (2018) 6:6 Page 9 of 9 3. Macintyre AN, Rathmell JC. PKM2 and the tricky balance of growth and 23. Sullivan LB, Gui DY, Hosios AM, Bush LN, Freinkman E, Vander Heiden MG. energy in cancer. Mol Cell. 2011;42:713–4. Supporting aspartate biosynthesis is an essential function of respiration in 4. Hitosugi T, Kang S, Vander Heiden MG, Chung T-W, Elf S, Lythgoe K, Dong proliferating cells. Cell. 2015;162:552–63. S, Lonial S, Wang X, Chen GZ, Xie J, Gu T-L, Polakiewicz RD, Roesel JL, 24. Birsoy K, Wang T, Chen WW, Freinkman E, Abu-Remaileh M, Sabatini DM. Boggon TJ, Khuri FR, Gilliland DG, Cantley LC, Kaufman J, Chen J. Tyrosine An essential role of the mitochondrial electron transport chain in cell phosphorylation inhibits PKM2 to promote the Warburg effect and tumor proliferation is to enable aspartate synthesis. Cell. 2015;162:540–51. growth. Sci Signal. 2009;2:ra73. 25. Cancer genome atlas research network. Electronic address: elizabeth. firstname.lastname@example.org, Cancer genome atlas research network. 5. Lv L, Li D, Di Zhao LR, Chu Y, Zhang H, Zha Z, Liu Y, Li Z, Xu Y, Wang G, Comprehensive and integrated genomic characterization of adult soft tissue Huang Y, Xiong Y, Guan KL, Lei QY. Acetylation targets the M2 isoform of sarcomas. Cell. 2017;171:950–965.e28. pyruvate kinase for degradation through chaperone-mediated autophagy 26. Mito JK, Riedel RF, Dodd L, Lahat G, Lazar AJ, Dodd RD, Stangenberg L, and promotes tumor growth. Mol Cell. 2011;42:719–30. Eward WC, Hornicek FJ, Yoon SS, Brigman BE, Jacks T, Lev D, Mukherjee S, 6. Christofk HR, Vander Heiden MG, Wu N, Asara JM, Cantley LC. Pyruvate Kirsch DG. Cross species genomic analysis identifies a mouse model as kinase M2 is a phosphotyrosine-binding protein. Nat Publ Group. 2008;452: undifferentiated pleomorphic sarcoma/malignant fibrous histiocytoma. PLoS 181–6. One. 2009;4:e8075. 7. Anastasiou D, Poulogiannis G, Asara JM, Boxer MB, Jiang J-K, Shen M, 27. Madisen L, Zwingman TA, Sunkin SM, Oh SW, Zariwala HA, Gu H, Ng LL, Bellinger G, Sasaki AT, Locasale JW, Auld DS, Thomas CJ, Vander Heiden MG, Palmiter RD, Hawrylycz MJ, Jones AR, Lein ES, Zeng H. A robust and high- Cantley LC. Inhibition of pyruvate kinase M2 by reactive oxygen species throughput Cre reporting and characterization system for the whole mouse contributes to cellular antioxidant responses. Science. 2011;334:1278–83. brain. Nat Neurosci. 2009;13:133–40. 8. Christofk HR, Vander Heiden MG, Harris MH, Ramanathan A, Gerszten RE, 28. Israelsen WJ, Vander Heiden MG. Pyruvate kinase: function, regulation and Wei R, Fleming MD, Schreiber SL, Cantley LC. The M2 splice isoform of role in cancer. Semin Cell Dev Biol. 2015;43:43–51. pyruvate kinase is important for cancer metabolism and tumour growth. 29. Hettmer S, Liu J, Miller CM, Lindsay MC, Sparks CA, Guertin DA, Bronson RT, Nat Publ Group. 2008;452:230–3. Langenau DM, Wagers AJ. Sarcomas induced in discrete subsets of 9. Anastasiou D, Yu Y, Israelsen WJ, Jiang J-K, Boxer MB, Hong BS, Tempel W, prospectively isolated skeletal muscle cells. Proc Natl Acad Sci. 2011;108: Dimov S, Shen M, Jha A, Yang H, Mattaini KR, Metallo CM, Fiske BP, 20002–7. Courtney KD, Malstrom S, Khan TM, Kung C, Skoumbourdis AP, Veith H, 30. Blum JM, Añó L, Li Z, Van Mater D, Bennett BD, Sachdeva M, Lagutina I, Southall N, Walsh MJ, Brimacombe KR, Leister W, Lunt SY, Johnson ZR, Yen Zhang M, Mito JK, Dodd LG, Cardona DM, Dodd RD, Williams N, Ma Y, KE, Kunii K, Davidson SM, Christofk HR, et al. Pyruvate kinase M2 activators Lepper C, Linardic CM, Mukherjee S, Grosveld GC, Fan C-M, Kirsch DG. promote tetramer formation and suppress tumorigenesis. Nat Chem Biol. Distinct and overlapping sarcoma subtypes initiated from muscle stem and 2012;8:839–47. progenitor cells. CellReports. 2013;5:933–40. 10. Dayton TL, Jacks T, Vander Heiden MG. PKM2, cancer metabolism, and the road ahead. EMBO Rep. 2016;17:1721–30. 11. Lu Z. Nonmetabolic functions of pyruvate kinase isoform M2 in controlling cell cycle progression and tumorigenesis. Chin J Cancer. 2013;32:5–7. 12. Yang W, Zheng Y, Xia Y, Ji H, Chen X, Guo F, Lyssiotis CA, Aldape K, Cantley LC, Lu Z. ERK1/2-dependent phosphorylation and nuclear translocation of PKM2 promotes the Warburg effect. Nat Cell Biol. 2012;14:1295–304. 13. Yang W, Xia Y, Ji H, Zheng Y, Liang J, Huang W, Gao X, Aldape K, Lu Z. Nuclear PKM2 regulates b-catenin transactivation upon EGFR activation. Nat Publ Group. 2011;480:118–22. 14. Yang W, Xia Y, Hawke D, Li X, Liang J, Xing D, Aldape K, Hunter T, Yung WKA, Lu Z. PKM2 phosphorylates histone H3 and promotes gene transcription and tumorigenesis. Cell. 2012;150:685–96. 15. Gao X, Wang H, Yang JJ, Liu X, Liu Z-R. Pyruvate kinase M2 regulates gene transcription by acting as a protein kinase. Mol Cell. 2012;45:598–609. 16. Wang Y-H, Israelsen WJ, Lee D, Yu VWC, Jeanson NT, Clish CB, Cantley LC, Vander Heiden MG, Scadden DT. Cell-state-specific metabolic dependency in hematopoiesis and leukemogenesis. Cell. 2014;158:1309–23. 17. Israelsen WJ, Dayton TL, Davidson SM, Fiske BP, Hosios AM, Bellinger G, Li J, Yu Y, Sasaki M, Horner JW, Burga LN, Xie J, Jurczak MJ, DePinho RA, Clish CB, Jacks T, Kibbey RG, Wulf GM, Di Vizio D, Mills GB, Cantley LC, Vander Heiden MG. PKM2 isoform-specific deletion reveals a differential requirement for pyruvate kinase in tumor cells. Cell. 2013;155:397–409. 18. Tech K, Tikunov AP, Farooq H, Morrissy AS, Meidinger J, Fish T, Green SC, Liu H, Li Y, Mungall AJ, Moore RA, Ma Y, Jones SJM, Marra MA, Vander Heiden MG, Taylor MD, Macdonald JM, Gershon TR. Pyruvate kinase inhibits proliferation during postnatal cerebellar neurogenesis and suppresses medulloblastoma formation. Cancer Res. 2017;77:3217–30. 19. Lau AN, Israelsen WJ, Roper J, Sinnamon MJ, Georgeon L, Dayton TL, Hillis AL, Yilmaz ÖH, Di Vizio D, Hung KE, Vander Heiden MG. PKM2 is not required for colon cancer initiated by APC loss. Cancer Metab. 2017;5:10. 20. Kirsch DG, Dinulescu DM, Miller JB, Grimm J, Santiago PM, Young NP, Nielsen GP, Quade BJ, Chaber CJ, Schultz CP, Takeuchi O, Bronson RT, Crowley D, Korsmeyer SJ, Yoon SS, Hornicek FJ, Weissleder R, Jacks T. A spatially and temporally restricted mouse model of soft tissue sarcoma. Nat Med. 2007;13:992–7. 21. Jackson EL, Willis N, Mercer K, Bronson RT, Crowley D, Montoya R, Jacks T, Tuveson DA. Analysis of lung tumor initiation and progression using conditional expression of oncogenic K-ras. Genes Dev. 2001;15:3243–8. 22. DuPage M, Mazumdar C, Schmidt LM, Cheung AF, Jacks T. Expression of tumour-specific antigens underlies cancer immunoediting. Nature. 2012;482: 405–9.
Cancer & Metabolism – Springer Journals
Published: May 31, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera