Metformin and Melatonin in Adrenocortical Carcinoma: Morphoproteomics and Biomedical Analytics Provide Proof of Concept in a Case Study
Robert E. Brown1, Jamie Buryanek1, and Mary F. McGuire2
1Department of Pathology and Laboratory Medicine at UT Health McGovern Medical School and 2Biomedical Analytics, Houston, TX, USA
Abstract. Metformin has been proposed as a novel anti-cancer drug for adrenocortical carcinoma (ACC) based upon Poli’s recent preclinical studies that 1. “in vitro” metformin modulates the ACC cell model H295R and 2. “in vivo” metformin inhibits tumor growth in a xenograft model as confirmed by a signifi- cant reduction of Ki67 [1]. Here we report on our prior clinical case study that provides proof of concept for Poli’s studies. We were requested to perform morphoproteomic analysis to further define the biology of, and raise targeted therapeutic options, for a case of post-treatment and chemoresistant ACC metastatic to the liver and the lung. Profiling the patient’s ACC from the liver resulted in the recommendation of met- formin as a maintenance therapy, which was supported by biomedical data analysis. The patient remains on maintenance therapy with metformin and melatonin and is free of disease some 7 years post diagnosis, thus underscoring the recommendation for clinical trials employing these therapeutic agents.
Key words: Adrenocortical carcinoma, metformin, melatonin, morphoproteomics, biomedical analytics, biology, insulin-like growth factor, mTORC/Akt, c-Met, beta-catenin, COX-2, EZH2 and SPARC path- ways, therapeutic options, metformin, melatonin, celecoxib, nab-paclitaxel.
Introduction
ACC is a relatively rare disease with a reported inci- dence of 0.7-2.0 cases/million habitants/year [2]. The 5-year survival of patients with ACC has been reported from 28% to 38.6%, the latter for those who underwent surgical resection [3,4]. Moreover, it has been noted that 30-70% of the patients with ACC present with advanced disease with very poor prognosis and without effective therapeutic op- tions, despite extensive genomic analyses that have revealed numerous signal transduction pathway ab- errations that play a central role in ACC’s patho- physiology [5].
Recently Poli and co-workers reported on the in vi- tro effects of metformin that included reducing the cell viability and proliferation of the ACC cell model H295R in association with: the activation of AMPK, significant inhibition of the phosphoryla- tion/activation of ERK1/2 and mTOR,
interference with the proliferative autocrine loop of IGF2/IGF-1R and triggering of the apoptotic path- way. In an in vivo, xenograft application of the ACC-H295R model, they also demonstrated that metformin administration inhibited tumor growth, as confirmed by a significant reduction of Ki67 [1].
The purpose of this report is fourfold: 1. to docu- ment the impact on cell cycle progression [1,3,6] and tumorigenic pathways in ACC, such as en- hancer of Zeste homolog 2(EZH2) that promote its progression [7], subsequent to the institution of a therapeutic regimen that included metformin and melatonin; 2. to employ biomedical analytics in quantifying the morphoproteomic findings in the patient’s tumors and in reinforcing the role of met- formin and melatonin in the biology of ACC as determined by morphoproteomic analysis and data from Ingenuity Pathways Analysis (IPA) and MEDLINE/PubMed; 3. to provide correlates of the anti-ACC effects of metformin in this case study and the preclinical studies; and 4. to illustrate the application of morphoproteomics in defining the biology of metastatic ACC and in the raising of additional therapeutic options to target its biology.
Address correspondence to Robert E. Brown, MD; Department of Pathology and Laboratory Medicine at UT Health McGovern Medical School, 6431 Fnnin Street, Houston, TX, USA; e mail: robert. brown@uth.tmc.edu
A
B
C
Materials and Methods
We reviewed the medical record of a patient with recur- rent ACC, whose tissue specimens were referred to the Department of Pathology and Laboratory Medicine, UT Health McGovern Medical School, Houston, Texas for morphoproteomic analysis. Informed consent, data col- lection, and molecular analysis were performed in accor- dance with the guidelines of the Institutional Review Board (IRB). Molecular analyses and biomedical analyt- ics were performed (RE Brown’s CLIA and CAP certified Consultative Proteomics Laboratory) in order to analyze the biology of the patient’s tumors, to ascertain targeted therapeutic options and to assess the response signature to morphoproteomic-guided therapies.
Morphoproteomics. The use of bright field microscopy and immunohistochemistry directed against various pro- tein analytes can better define the biology of a neoplastic process and the pathogenetic occurrences that might be responsible for its development, chemoradioresistance, and propensity to recur. That is the application of mor- phoproteomics [8,9]. To that end, we applied immuno- histochemical probes against the following protein ana- lytes in unstained sections of the patient’s recurrent ACC metastatic to the liver and before morphoproteomic- guided targeted therapy, and in the subsequent metasta- sis to the lung while on morphoproteomic-guided, met- formin and melatonin: total insulin-like growth factor (IGF)-1 receptor (R), at tyrosine 1165/1166 (Gen Way Biotech Inc. San Diego, CA); phosphorylated (p)-c-Met (Tyr 1234/1235; Cell Signaling Technology, Inc, Danvers, MA), mammalian target of rapamycin (mTOR) phosphorylated on serine 2448(Cell Signaling Technology Inc), p-Akt (Ser 473) (Cell Signaling Technology Inc); Ki-67 (G1, S, G2 and M phases of the cell cycle; DakoCytomation Carpentaria, California), enhancer of zeste homolog2(EZH2; Cell Signaling Technology, Inc);nuclear beta-catenin(BDPharmingen, Becton Dickinson San Jose, California), cyclooxygenase(COX)-2( DakoCytomation), and secret- ed protein, acidic and rich in cysteine (SPARC, osteonectin;Novocastra, New Castle Upon Tyne, United Kingdom). The level of expression of the analytes was graded on a 0 to 3+ scale based on signal intensity indi- cated by a 3,3’- tetrahydrochloride (DAB) chromogenic (brown) signal and the nuclear estimation of Ki-67 and EZH2 percentages and mitotic index based on mitotic figures/10 high power fields was carried out by two sur- gical pathologists (REB and JB) with consensus. The de- tails of the morphoproteomic staining procedure have been previously described [8,9].
A
B
C
D
Biomedical Analytics. Biomedical analytics develops and applies methods from mathematics and computer science to gain insights into biological processes based on personalized data and published biomedical research. In this study, biomedical analytics integrated the mor- phoproteomic analysis of the patient’s adrenocortical carcinomas in the liver (pre-treatment) and lung (during treatment with metformin and melatonin in a mainte- nance mode), and generated an evoked pathway net- work to demonstrate the interconnection and efficacy of the metformin and melatonin therapies on key analytes in the patient’s pre-treatment (2012) molecular signature.
The patient’s data were normalized and weighted by an algorithm customized for the pathologist of record. The resulting score for each analyte was entered, along with its UNIPROT ID, into Ingenuity Pathway Analysis (IPA, www.ingenuity.com). Pathway networks and their interactions with the proposed therapies were evoked
based on existing IPA data. From these graphs and ad- ditional data mining of the National Library of Medicine’s MEDLINE database, a single network model was constructed using IPA Pathway Designer to repre- sent the key modulation and adaptive responses in the signal transduction processes.
Case Report
Patient clinical course and treatment history. A 31-year-old white female was originally diagnosed with adrenocortical carcinoma (ACC) at age 24. She subse- quently began chemotherapy that included doxorubicin, cisplatin, etoposide and mitotane and metyrapone at MD Anderson Cancer Center with some shrinkage of the tumor leading to its surgical removal via right adre- nalectomy and right hepatectomy. However, her ACC returned a few months later with metastatic disease to the left side of her liver and left lower lobe of the lung, and the recurrent tumor was resistant to standard
A
B
C
D
E
F
A
B
100
90
80
70
60
50
2012
2014
40
30
20
10
0
EZH2
KI67
Mitotic Index
chemotherapy. When the patient was 26 years old, the metastatic tumor in the liver was excised via partial left hepatectomy and a portion was submitted for morpho- proteomic analysis [8,9]. The morphoproteomic analysis resulted in the recommended therapeutic considerations that included a statin, metformin (500 mg twice per day) and melatonin (20 mg once per day at bedtime), and these agents were initiated as part of a personalized treatment plan of her medical oncologist. This brought the patient’s cancer under control clinically, and her im- proved health enabled her to start up her family. At age 27, the patient was pregnant, working part-time, and working out in the gym again. During her pregnancy and thereafter, she stopped the statin but continued
metformin at 500 mg 4 times per week and melatonin at 20 mg 3 times per week as a maintenance regimen. She and her husband welcomed their daughter, born later that year. Subsequently, at age 28, a “spot” was con- firmed in the right upper lobe of the lung and was re- moved and submitted for morphoproteomic analysis. By comparison with the pretreatment specimen from the liver, the metastatic tumor from the right upper lobe of the lung with the patient on maintenance therapy with metformin and melatonin showed molecular and mor- phometric evidence of growth inhibition (vide infra). At age 29, repeat scans revealed no evidence of tumor, and she and her husband welcomed the birth of their son. Some 7 years post diagnosis the patient has resumed a healthy clinical life and remains on melatonin and met- formin as a maintenance type therapy to reduce the risk of recurrent disease.
Morphoproteomic Findings. The comparative study of the liver metastasis of her ACC prior to morphopro- teomic-guided therapy and the “spot” removed by wedge resection from the right upper lobe of the lung while she was on maintenance therapy with metformin and mela- tonin revealed the following in her tumors on microana- tomical and morphoproteomic analysis: apparent “lipo- matous metaplasia” [10] of the tumor in the post-treatment lung specimen on hematoxylin-eosin (H&E) staining with more apparent progression into the mitotic phase in the pretreatment specimen (Figure 1A- C); and reduced nuclear expressions of EZH2(enhancer of zeste homolog2), and Ki-67 percentage and mitotic index in the lung specimen while on treatment with metformin and melatonin vis-à-vis the pretreatment
Pri4a1
PARM1
PDGF-DD
PDGFD
WNT16
Immunoglobulin
IL12 (family)
PROKR1 THEM4
DENND3
FUT6
AKT1
tubulin (complex)
UX51
Cgm4/Psg16
INPP4B
Proinsulin
P38 MAPK
melatonin
Akt
2’-fucosyllactose
inositol 1-phosphate
CASP3
Jnk
CRISPLD2
metformin
9
FGD2
mir-101
miR-101-3p
MTMR7
GTPase
miR-26a
Mmp
S100a7a
CTNNB1
mit (
Ybx1-ps3
DENND2A
MKI67
EZH2
liver metastasis (i.e., EZH2 at a range of 13% to 25% versus 50% to 70%, Ki-67 at -11% versus -33% and mitotic index at <2 mitotic figures/ 10 high power fields versus up to 13 mitotic figures/10 high power fields (Figure 2A-D, respectively for EZH2 and Ki-67 expres- sion, and normalized bar graph in the Biomedical Analytics section, vide infra). However, comparable ex- pression of the following protein analytes was retained in both the pretreatment and post-treatment specimens with metformin and melatonin: insulin-like growth fac- tor -1 receptor (IGF-1R), phosphorylated (p)-c-Met (Tyr 1234/1235), mammalian target of rapamycin (mTOR) phosphorylated on serine 2448, p-Akt (Ser 473), nuclear beta-catenin, cyclooxygenase(COX)-2 (Figure 3A-F), and secreted protein, acidic and rich in cysteine (SPARC, osteonectin) (Figure 4A&B).
Biomedical Analytics. In order to provide a visual com- parison of the pre-treatment and post-treatment ACC specimens (the latter while the patient was on mainte- nance doses of metformin and melatonin), biomedical analytics generated a normalized score of the compara- tive analyses for EZH2, Ki67, and mitotic index. This is
illustrated in the bar graph (Figure 5). Pathway analysis incorporating the results from the post-treatment versus pretreatment specimens and focusing on EZH2 and pro- liferation indices, Ki-67 (MK167) and the interrelation- ships with the beta-catenin (CTNNB1) and the mTORC/Akt pathway graphically demonstrate the in- terconnection and efficacy of the metformin and melato- nin therapies (Figure 6).
Discussion
This report documents the impact on cell cycle pro- gression [1,3,6] and tumorigenic pathways in ACC, such as enhancer of Zeste homolog 2(EZH2) that promote its progression [7], subsequent to the in- stitution of a therapeutic regimen that included metformin and melatonin. Specifically, the latter therapies guided by morphoproteomic analysis was associated with a reduction in the Ki-67 (which re- flects the G1, S and G2 and M phases of the cell cycle), the mitotic index and the level of EZH2 ex- pression in the metastatic site subsequent to the
ADRENOCORTICAL CARCINOMA
1 IGF 2 gene
(+)
Celecoxib
(-)
JIGF 2
Celecoxib
(-)
(-)
IGF-1R (Tyr 1165/1166)
Metformin
Nab-Paclitaxel. SPARC
IRS-1
(-)
Metformin p-mTOR
Nab-Paclitaxel +
(-)
(+)
(Ser 2448)
Metformin
ß-tubulin
(-)
(+)
Celecoxib
Celecoxib
Mitotic Arrest + Apoptosis
p-mTOR (Ser 2448)
p-c-Met (Tyr 1234/1235)
(-)
(+)
(+)
COX-2
p-ERK 1/2 (Thr 202/Tyr 204)
(+)
ACC survival
(+)
p-Akt
(Ser 473)
(+)
COX-21
(-)
Celecoxib
Metformin (-)
(+)
(+)
(-)
Celecoxib
Beta-catenin
(+)
(+)
ACC tumor progression
COX-2
(-)
Celecoxib
implementation of metformin and melatonin as maintenance therapy; and was associated with a complete clinical remission in the patient some 7 years after the initial diagnosis. This accords with the preclinical studies of Poli, et al. [1] in support- ing a role for metformin in the treatment of ACC. Moreover, it coincides with the literature regarding the ability of metformin to cause a partial cell cycle arrest, including at the G1/G0 phase of the cycle in some tumors[11,12] and to inhibit EZH2 via phar- macogenomic upregulation of miR-26a and miR101 [13,14] which suppress EZH2.
The retention of expression of IGF-1R, p-c-Met, the mTORC/Akt pathway, COX-2, nuclear beta- catenin, and SPARC expression in the patient’s specimen treated with metformin and melatonin provides the opportunity to consider other combi- natorial therapies with metformin and melatonin in ACC in those not responsive to metformin and melatonin alone and in conjunction with the ex- pression of the aforementioned pathways which contribute to their biology. Specifically, although metformin has been shown to inhibit both the mTORC1 and components of the mTORC2/Akt
pathway [12], it does not appear that it was able to accomplish this alone. Agents such as celecoxib which has both COX-2 dependent and indepen- dent effects may have application against ACC for the following reasons: celecoxib induces apoptosis, in part by blocking Akt activation [15,16]; celecox- ib inhibits the insulin-like growth factor pathway with downregulation of both IGF-1R expression and phosphorylation of Akt [17] and also up-regu- lates Insulin-like growth factor binding protein (IGFBP-3) in some studies to antagonize such sig- naling[18]; and celecoxib inhibits both the nuclear translocation and expression of beta-catenin and the transcription of beta-catenin-associated genes [19,20] and also inhibits c-Met kinase activity and expression and beta-catenin-associated transcrip- tion [21,22]. Finally, the expression of SPARC in both specimens in this patient is consistent with the fact that it acts as a mechanism for the increased efficacy of nab-paclitaxel [23] and prior preclinical studies using nab-paclitaxel in SPARC-expressing adrenal cortical cancer cells that showed in vitro in- hibition at lower IC50 concentrations compared with mitotane and a greater decrease in tumor weight in xenograft models than mitotane [24]. Such combinatorial therapies with metformin, cele- coxib and nab-paclitaxel are depicted in Figure 7.
In analyzing the therapeutic response, biomedical analytics evoked 15 molecules unique to the patient (Figure 5) and 15 unique to the data mining con- trol and lacking in the patient in the pretreatment model. The large proportion of unique molecules implies that the patient’s underlying molecular pathways may hold clues as to why this person re- sponded well to this therapy; other patients, with similar morphoproteomic signatures and evoked pathways, may also benefit from this therapy. We acknowledge that this is an N of 1 study, and data need to be confirmed in a larger patient sample. Future research is planned to investigate these findings.
In summary, the morphoproteomic/morphometric findings in the patient’s metastatic ACC specimens before and after morphoproteomic-guided targeted therapy with metformin and melatonin in a main- tenance mode, reveal a reduction in the cell cycle progression (Ki-67 and mitotic index) and the
tumorigenic EZH2 pathway in the post-treatment specimen. This provides proof of concept for the preclinical studies of Poli and co-workers [1] in sug- gesting a role for metformin in the treatment of ACC. Moreover, morphoproteomics has identified additional pathways in this study to support the combinatorial targeting of the IGF-1R, c-Met, mTORC/Akt, COX-2, and beta-catenin pathways in ACC with metformin, celecoxib, and melatonin, and as a chemotherapeutic adjunct, the use of nab- paclitaxel based on SPARC expression in their ACC’s for those individual patients whose tumors do not respond to metformin, celecoxib and melatonin.
Acknowledgments
The authors thank Pamela Johnston, HT (ASCP) for technical assistance and Ms. Bheravi Patel for secretarial support and help with the graphics.
References
1. Poli G, Cantini G, Armignacco R, Fucci R, Santi R, Canu L, Nesi G, Mannelli M, Luconi M. Metformin as a new anti- cancer drug in adrenocortical carcinoma. Oncotarget. 2016 Aug 2; 7(31):49636-49648.
2. Libé R. Adrenocortical carcinoma (ACC): diagnosis, progno- sis, and treatment. Front Cell Dev Biol. 2015 Jul 3; 3:45.
3. Choi YM, Kwon H, Jeon MJ, Sung TY, Hong SJ, Kim TY, Kim WB, Shong YK, Lee JL, Song DE, Kim WG. Clinicopathological Features Associated With the Prognosis of Patients With Adrenal Cortical Carcinoma: Usefulness of the Ki-67 Index. Medicine (Baltimore). 2016 May; 95(21):e3736.
4. Bilimoria KY, Shen WT, Elaraj D, Bentrem DJ, Winchester DJ, Kebebew E, Sturgeon C.Adrenocortical carcinoma in the United States: treatment utilization and prognostic factors. Cancer. 2008 Dec 1; 113(11): 3130-3136.
5. Costa R, Carneiro BA, Tavora F, Pai SG, Kaplan JB, Chae YK, Chandra S, Kopp PA, Giles FJ. The challenge of developmental therapeutics for adrenocortical carcinoma. Oncotarget. 2016 Jul 19; 7(29):46734-46749.
6. Beuschlein F1, Weigel J, Saeger W, Kroiss M, Wild V, Daffara F, Libé R, Ardito A, Al Ghuzlan A, Quinkler M, Oßwald A, Ronchi CL, de Krijger R, Feelders RA, Waldmann J, Willenberg HS, Deutschbein T, Stell A, Reincke M, Papotti M, Baudin E, Tissier F, Haak HR, Loli P, Terzolo M, Allolio B, Müller HH, Fassnacht M. Major prognostic role of Ki67 in localized adrenocortical carcinoma after complete resection. J Clin Endocrinol Metab. 2015 Mar; 100(3):841-849.
7. Ip JC, Pang TC, Glover AR, Soon P, Zhao JT, Clarke S, Robinson BG, Gill AJ, Sidhu SB. Immunohistochemical vali- dation of overexpressed genes identified by global expression microarrays in adrenocortical carcinoma reveals potential pre- dictive and prognostic biomarkers. Oncologist. 2015 Mar; 20(3):247-256.
8. Brown RE. Morphoproteomics: exposing protein circuitries in tumors to identify potential therapeutic targets in cancer pa- tients. Expert Rev Proteomics. 2005 Jun; 2(3):337-348.
9. Brown RE. Morphogenomics and morphoproteomics: a role for anatomic pathology in personalized medicine. Arch Pathol Lab Med. 2009 Apr; 133(4):568-579.
10. . Izumi M, Serizawa H, Iwaya K, Takeda K, Sasano H, Mukai K. A case of myxoid adrenocortical carcinoma with extensive lipo- matous metaplasia. Arch Pathol Lab Med. 2003 Feb; 127(2):227-230.
11. Alimova IN, Liu B, Fan Z, Edgerton SM, Dillon T, Lind SE, Thor AD. Metformin inhibits breast cancer cell growth, colony formation and induces cell cycle arrest in vitro. Cell Cycle. 2009 Mar 15; 8(6):909-915.
12. Liu B, Fan Z, Edgerton SM, Deng XS, Alimova IN, Lind SE, Thor AD. Metformin induces unique biological and molecular responses in triple negative breast cancer cells. Cell Cycle. 2009 Jul 1; 8(13):2031-2040.
13. Bao B, Wang Z, Ali S, Ahmad A, Azmi AS, Sarkar SH, Banerjee S, Kong D, Li Y, Thakur S, Sarkar FH. Metformin inhibits cell proliferation, migration and invasion by attenuat- ing CSC function mediated by deregulating miRNAs in pan- creatic cancer cells. Cancer Prev Res (Phila). 2012 Mar; 5(3):355-364.
14. Zheng, M, Jiang, Y. P, Chen, W, Li, K. D, Liu, X, Gao, S. Y, Feng, H, Wang, S. S, Jiang, J, Ma, X. R, Cen, X, Tang, Y. J, Chen, Y, Lin, Y. F, Tang, Y. L, Liang, X. H. Snail and Slug col- laborate on EMT and tumor metastasis through miR-101-me- diated EZH2 axis in oral tongue squamous cell carcinoma. Oncotarget. 2015 Mar 30; 6(9):6797-6810.
15. Hsu AL, Ching TT, Wang DS, Song X, Rangnekar VM, Chen CS. The cyclooxygenase-2 inhibitor celecoxib induces apopto- sis by blocking Akt activation in human prostate cancer cells independently of Bcl-2. J Biol Chem. 2000 Apr 14; 275(15):11397-11403.
16. Arico S, Pattingre S, Bauvy C, Gane P, Barbat A, Codogno P, Ogier-Denis E. Celecoxib induces apoptosis by inhibiting 3-phosphoinositide-dependent protein kinase-1 activity in the human colon cancer HT-29 cell line. J Biol Chem. 2002 Aug 2; 277(31):27613-27621.
17. Liu CH, Bao HG, Ge YL, Wang SK, Shen Y, Xu . Celecoxib inhibits insulin-like growth factor 1 induced growth and inva- sion in non-small cell lung cancer. Oncol Lett. 2013 Jun; 5(6):1943-1947.
18. Levitt RJ, Buckley J, Blouin MJ, Schaub B, Triche TJ, Pollak M. Growth inhibition of breast epithelial cells by celecoxib is associated with upregulation of insulin-like growth factor binding protein-3 expression. Biochem Biophys Res Commun. 2004 Apr 2; 316(2): 421-428.
19. Sareddy GR, Kesanakurti D, Kirti PB, Babu PP. Nonsteroidal anti-inflammatory drugs diclofenac and celecoxib attenuates Wnt/ß-catenin/Tcf signaling pathway in human glioblastoma cells. Neurochem Res. 2013 Nov; 38(11):2313-2322.
20. Xia JJ, Pei LB, Zhuang JP, Ji Y, Xu GP, Zhang ZP, Li N, Yan JL. Celecoxib inhibits ß-catenin-dependent survival of the hu- man osteosarcoma MG-63 cell line. J Int Med Res. 2010 Jul- Aug; 38(4):1294-1304.
21. Tuynman JB, Vermeulen L, Boon EM, Kemper K, Zwinderman AH, Peppelenbosch MP, Richel DJ. Cyclooxygenase-2 inhibi- tion inhibits c-Met kinase activity and Wnt activity in colon cancer. Cancer Res. 2008 Feb 15; 68(4):1213-1220.
22. Yin J, Liu B, Li B, Liu Z, Xie X, Lv Z, Gao S, Guang J. The cyclooxygenase-2 inhibitor celecoxib attenuates hepatocellular carcinoma growth and c-Met expression in an orthotopic mouse model. Oncol Res. 2011; 19(3-4):131-139.
23. Desai N, Trieu V, Damascelli B, Soon-Shiong P. SPARC Expression Correlates with Tumor Response to Albumin- Bound Paclitaxel in Head and Neck Cancer Patients. Transl Oncol. 2009 May; 2(2):59-64.
24. Demeure MJ, Stephan E, Sinari S, Mount D, Gately S, Gonzales P, Hostetter G, Komorowski R, Kiefer J, Grant CS, Han H, Von Hoff DD, Bussey KJ. Preclinical investigation of nanoparticle albumin-bound paclitaxel as a potential treat- ment for adrenocortical cancer. Ann Surg. 2012 Jan; 255(1):140-146.