ORIGINAL ARTICLE

ERK Signaling in Colorectal Cancer: A Preliminary Reporton the Expression of Phosphorylated ERK and the Effects of

Radiation Therapy

Benjamin W. Corn, MD,* Felix Kovner, MD, PhD,* Stella Bek, MD,† Isaiah Wexler, MD,‡Beatrice Lifschits, MD,† and Rony Seger, PhD§

Objectives: Phosphorylated extracellular signal-regulated kinase(ERK) activates cell survival pathways and may promote tumor cellproliferation. The purpose of this study was to determine the level ofphospho-ERK staining in pretreatment specimens, and the effect ofa short course of radiation therapy (RT) on the levels of phospho-ERK in specimens of colorectal cancer.Methods: Specimens of colorectal tumors were obtained frompatients beginning RT before surgery. Tumors were graded andhistologically staged. Subsequently, tumor samples were reactedwith antiphospho-ERK antibodies and the level of staining inten-sity was assessed.Results: Of the 12 subjects evaluated, 7 of 12 had tumors thatpositively stained for phospho-ERK before and/or after RT. Mostpatients who were phospho-ERK positive before RT converted toERK negativity after RT.Conclusions: The results of this pilot study indicate that phospho-ERK levels may serve as a prognostic marker in colorectal cancer,and further studies are warranted to determine how changes inphospho-ERK staining relate to clinical outcome.

Key Words: colorectal cancer, ERK, radiation therapy, MAPK

(Am J Clin Oncol 2008;31: 255–258)

The mitogen-activated protein kinase (MAPK)/extracellu-lar signal-regulated kinase (ERK) plays a central role in

mediating the transmission of growth-promoting signals frommultiple growth factor receptors. ERK is a relatively abun-dant protein kinase, which is activated by a specific kinase,MAPK/ERK kinase (MEK) that occupies a strategic down-stream position in critical intracellular signaling cascades.1

The MEK/ERK cascade has been tied to oncogenic transfor-mation and cancer progression.2 Constitutive activation of

MEK has been shown to bring about cellular transformation,whereas dominant negative MEKs reverse this process.3

The MEK/ERK pathway has been demonstrated to beconstitutively activated in a large number of tumor cell linesand primary human tumors.4 However, few investigators haveattempted to identify the role of this pathway in the patho-physiology of colorectal cancer. Several contemporary stud-ies have shown that ERK may be important for colon cancerprogression. Phorbol-ester-induced scattering of HT29-M6colon cancer cells (a model of cancer progression) is medi-ated by ERK1/2 and prevented by the MEK inhibitor, PD98059.5 RAS-activating mutations occur in a majority ofcolorectal carcinomas, and RAS has been shown to induceproliferative pathways by activating ERK in DLD colon cancercells.6 In the DLD colon cancer line, proline oxidase (POX)over-expression induced apoptosis, which was associated withlower levels of phosphorylated ERK, suggesting that acti-vated ERK has an antiapoptotic effect in colon cancer.7

Colorectal cancer remains the third most common ma-lignancy in the United States after tumors of the skin andlung.8 The cornerstone of treatment for rectal cancer contin-ues to be surgery; however, locally advanced cases (eg, AJCCstages II and III) are known to benefit from adjuvant treat-ment (eg, chemotherapy, radiation therapy). The molecularpathways responsible for the transformation of colorectalcells are not well understood, and there is an ongoing searchfor the signal transduction pathway that is disrupted in colo-rectal carcinoma.

Recently, evidence has been presented by Zhu and col-leagues implicating the MAPK pathway in the genesis of colo-rectal cancer.9 The T-cell-originated protein kinase (TOPK) is aMAPK kinase (MEK) oncogene that can constitutively acti-vate the MAPK signal transduction pathway leading to in-creased levels of phospho-ERK and cellular transformation.In human colorectal cancer cell line HT116, knockdown ofTOPK by siRNA reduces the rate of both in vitro and in vivocell proliferation. In the aforementioned study, TOPK wasfound to phosphorylate ERK2, and phosph-ERK2 phosphor-ylates TOPK (positive feedback loop). This feedback loopcan be blocked by knockdown of ERK2 using siRNA.9 Thisraises the possibility that the level of phospho-ERK expres-sion can be both a target and marker in colorectal cancer.

From the Departments of *Oncology and †Pathology at the Tel Aviv SouraskyMedical Center, Tel Aviv, Israel; ‡Department of Pediatrics at HadassahUniversity Medical Center, Jerusalem, Israel; and §Department of BiologicalRegulation at the Weizmann Institute of Science, Rehovoth, Israel.

Reprints: Benjamin W. Corn, MD, Department of Pathology at the Tel AvivSourasky Medical Center, 6 Weizman Street, Tel Aviv, Israel 64239.E-mail: bencorn@tasmc.health.gov.il.

Copyright © 2008 by Lippincott Williams & WilkinsISSN: 0277-3732/08/3103-0255DOI: 10.1097/COC.0b013e31815a5647

American Journal of Clinical Oncology • Volume 31, Number 3, June 2008 255

The aim of the present pilot study was to extend thefindings of Zhu et al by determining whether colorectal tumorspecimens positively stain for phospho-ERK. We also eval-uated the impact of radiation therapy on MEK/ERK signalingamong a relatively homogenous group of patients in terms ofsite (low-lying rectal carcinoma), stage (T3–4), and treatmentscheme. The results of our study suggest that phospho-ERKstaining is detected in more than half of the subjects studied,and that radiation induces changes in the expression ofphospho-ERK even after a relatively brief period of time. Ourdata suggest that the intensity of phospho-ERK staining, if itcorrelates with tumor progression and outcome, may be auseful prognostic marker in colorectal cancer.

MATERIALS AND METHODSThe study was conducted at the Tel Aviv Sourasky

Medical Center (TASMC). Between May 1996 and June1999, TASMC established a policy of preoperative radiother-apy for resectable rectal cancer following the publication ofthe second Swedish Rectal Cancer Trial (Stockholm II).10

The protocol required obtaining tissue samples before andafter radiation treatment. Thus, it was possible to determinethe effect of therapy on the MEK/ERK pathway on patho-logic specimens of colon cancer after site-directed irradiation.

Ultrasonography was used for preoperative staging. Aspart of the protocol, patients underwent a pretreatment biopsyof the tumor, and during surgery, specimens of resectedtumor were stored for future studies. To be included in thestudy, subjects had to have a primary, histologically proven,low-lying rectal cancer without metastatic spread. Beforeinitiating this retrospective study, approval was obtained bythe Helsinki Committee (eg, Institutional Review Board) ofthe TASMC as a discarded tissue protocol.

Once a diagnosis of adenocarcinoma of the rectum wasestablished by biopsy, patients received an abbreviated

course of pelvic irradiation one week before scheduled ante-rior resection or abdominoperineal excision of the tumor. Onthe basis of the protocol, patients received 25 Gy in 5fractions with megavoltage photons. The clinical target vol-ume included the primary tumor, the mesorectal and presacrallymph nodes, the lymph nodes along the internal iliac vessel,the lumbar lymph nodes up to the level of the 5th lumbarvertebral body, and the lymph nodes at the obturator foram-ina. Patients were treated in the prone position with a 3-fieldtechnique. Shielding of tissues that were not considered to beat risk for tumor dissemination was achieved with customizedcerrobend blocks.

The diagnostic material from the biopsy was evaluatedwith hematoxylin and eosin staining, and specimens weregraded and histologically typed (eg, mucinous vs. adenocar-cinoma, etc). Resected specimens were examined by thestudy pathologist and reviewed by a senior pathologist bothof whom were blinded to the clinical data. Each specimenwas classified according to the staging system of the AJCC.Representative samples from either the pretreatment biopsyor the surgical specimens were immunostained with antibod-ies to phosphorylated ERK (cat No. M8159, Sigma Co., St.Louis, MO). This was done by standard cytohistochemicalstaining in the Department of Pathology at TASMC. Aqualitative 4-point scale was developed to classify the inten-sity of staining based on the numbers of cells positive forphospho-ERK staining (negative, weakly positive, moder-ately positive, and strongly positive). In addition, the exam-ining pathologist noted whether the staining was diffuse,primarily nuclear, or cytoplasmic. An example of phospho-ERK staining is displayed in Figure 1.

RESULTSSpecimens of low-lying rectal carcinoma were obtained

from 12 patients before radiotherapy treatment and during

FIGURE 1. Changes in phospo-ERKstaining in response to radiationtreatment of colorectal cancer. Rep-resentative samples from the pre-treatment biopsy and the surgicalspecimens were subjected to stan-dard histochemical procedures forparaffin embedding and immunohis-tochemical analyses using mouseanti phospho-ERK Ab, relevant bioty-nilated 2nd Ab and routine detec-tion. In this example, positive immu-nostaining was markedly reduced bythe radiation therapy.

Corn et al American Journal of Clinical Oncology • Volume 31, Number 3, June 2008

© 2008 Lippincott Williams & Wilkins256

surgery. Table 1 lists the histologic characteristics of thetumor, and the AJCC staging before and after radiotherapy.The AJCC stage in 7 subjects did not change, 2 advancedfrom T2 to T3, and 3 regressed with 1 subject havingcomplete eradication of the tumor.

Samples were immunostained for phospho-ERK beforeinitiation of radiation treatment and after completion of ra-diotherapy (ie, during the definitive surgical procedure). Anexample for the intensity of phospho-ERK staining beforeand after radiation therapy for 2 of the patients is shown inFigure 1. The summary of these parameters is also listed inTable 1. For 7 of the 12 subjects, there was positive phospho-ERK staining before and after radiation therapy. In 1 case(patient 5), a preirradiation biopsy could not be evaluated forphosphorylated ERK staining because the material had beenexhausted during the initial histologic evaluation. In anothercase (patient 4), a complete pathologic response was achievedafter the delivery of preoperative radiotherapy thereby pre-cluding determination of phospho-ERK status in the opera-tive specimen. Of the 10 remaining patients, 5 patients whowere positive for phospho-ERK converted to a reduced signal(4 of 5 completely negative signal) after radiation therapy.Only 1 patient advanced from phospho-ERK negativity to aweak signal. In 4 subjects, no phospho-ERK signal wasdetected either before or after radiotherapy. In terms of thecellular location of phosphor-ERK, the intensity of the stainwas greater in the nucleus reflective of the fact that ERK ismost active in the nucleus. Staining was more intense in thecase of the poorly differentiated carcinoma when comparedwith the moderately and well differentiated sample.

Tumors that manifested positive nuclear staining forERK tended to be higher grade (mucinous, grade 3, grade 2).No grade 1 specimen stained positively for phosphorylatedERK. Conversely, tumors that were negative for phosphory-lated ERK staining tended to be from the well-differentiatedend of the spectrum (grade 1, grade 1–2, grade 2). No grade3 specimen had negative phospho-ERK status.

Among the patients studied, no correlative patternswere identified between AJCC stage and the staining forphosphorylated ERK. As for the staining with anti-ERKantibodies not specific for phospho-ERK, our results indicateno significant differences in the amount of ERK between thedifferent patients irrespective of the stage of the cancer or thetreatment.

DISCUSSIONRecent studies involving colorectal cancer cell lines

and specimens have shown that MEK/ERK signal transduc-tion pathway has a critical role in the invasiveness andproliferative properties of these tumors.9 On the basis of thesefindings, we undertook to retrospectively review colorectaltumor specimens to determine whether these tumors havesignificant levels of phosphorylated ERK (activated), andwhether irradiation of the tumor site has an effect on the levelof phospho-ERK staining. Our results show more than half ofthe tumor specimens stained positively for phospho-ERK,and in 5 of the 12 cases, treatment with radiation therapyresulted in a conversion from positive to negative phospho-ERK staining.

Various tyrosine kinase cascades have been implicatedin tumorogenesis and even as predictors of histopathologicaldamage resulting from radiation therapy. For instance, Yeohand colleagues11 have demonstrated that part of the chain ofinjurious events in irradiated colorectal tissue occurs as aresult of the direct activation by ionizing radiation of NFKB.NFKB is another signaling factor that is often activatedtogether with ERK, and it is therefore possible that similarbehavior to that of NFKB holds true for the ERK cascade.

The recent study by Zhu et al underscores the impor-tance of ERK.9 TOPK, a putative oncogene that may beresponsible for transformation in colorectal cancer phosphor-ylates ERK2, and ERK2 activates TOPK. The preliminarydata from the present study indicate that radiation therapy hasan effect on the levels of phospho-ERK, and the effect onphospho-ERK may be related to the beneficial effects of radi-ation in this form of cancer. Alternatively, phospho-ERK may bea prognostic marker with lower levels reflecting a better re-sponse to radiation therapy and an improved prognosis.

From a different perspective, the reversion from posi-tive to negative phospho-ERK staining may have clinicalsignificance. In tumor cell lines, irradiation of the cells resultsin the activation of several cell survival cascades (mitogenic,proliferation) including the RAF-MEK-ERK pathway.12 Incertain cell lines, ERK along with other mediators defendcells against radiation induced cytotoxicity by increasingradioresistance.13,14 Other researchers have found that acti-vation of ERK by irradiation increases clonogenic radiosen-sitivity.15 It has been suggested that inhibition of ERK acti-vation by specific MEK inhibitors such as PD98059 mightenhance the effects of radiation therapy providing that thetumor is a type in which activated ERK induces radioresis-tance.16 In the present study, in a majority of patients,radiation therapy resulted in a decrease in phospho-ERKlevels, and this may correlate with a better outcome due to theinactivation of ERK mediated survival pathways.

TABLE 1. Tumor Histology and Immunocytochemistry

PatientInitial

T Stage HistologyPost RTStatus

InitialPhopsho-ERK

Post RTPhospho-ERK

1 T2 Mucinous T3 � �

2 T3 Grade 3 T3 � �

3 T2 Grade 2 T3 ��� �

4 T3 Grade 2 CR � CR

5 T3 Grade 2 T3 No tissue �

6 T3 Grade 1–2 T3 �� �

7 T3 Grade 2 T2 � �

8 T3 Grade 2 T3 � �

9 T3 Grade 2 T3 � �

10 T2 Grade 2 T1 � �

11 T3 Grade 2 T3 � �

12 T3 Grade 1 T3 � �

CR indicates complete remission; �, no phospho-ERK staining; �, weak phospho-ERK staining; ��, moderate phospho-ERK staining; ���, strong phospho-ERKstaining.

American Journal of Clinical Oncology • Volume 31, Number 3, June 2008 ERK Signaling in Colorectal Cancer

© 2008 Lippincott Williams & Wilkins 257

Recently, members of the Ras-MAPK cascade haveemerged as attractive targets for pharmacologic interventionin cancer. For instance, CI-1040 became the first MEK-targeted agent to enter clinical trials and was found to be ahighly potent and selective inhibitor of both MEK isoforms.17

The drug was found to inhibit the clonogenic growth of adiverse panel of tumor cell lines including colorectal can-cer.18 Furthermore, a second generation oral MEK inhibitor(PD0325901) seems to be 50-fold more potent than CI-1040while possessing improved bioavailability.19 These observa-tions are likely to spur the development of additional MEKinhibitors, especially those targeting TOPK, which may beinstrumental in combating those tumors that seem to beresistant to radiation alone. Thus assessing the status ofphospho-ERK at various stages may guide therapy.

This study was limited by the fact that there were nocontrol specimens of normal colorectal tissue. However, ourresults were comparable to those of Zhu et al as 64% of thecolorectal cancer specimens in our study were positive forphospho-ERK staining when compared with 81% of thespecimens that were positive for TOPK staining. Althoughthe present study was a retrospective study done in a limitednumber of patients, the increased levels of phospho-ERKseen initially in most patients, and the reduction in phospho-ERK in the majority of patients after a short course ofradiation therapy would suggest that phospho-ERK couldserve as an important surrogate marker in terms of responseto therapy.

This study shows the feasibility of assessing phospho-ERK levels in tumor specimens and the potential utility ofphospho-ERK as a prognostic indicator. Additional studiesare necessary to determine which clinical and molecularfactors relate to pretreatment phospho-ERK levels, andwhether phospho-ERK levels are related to clinical outcome.In addition, it would be important to assess the levels ofphospo-ERK and other signaling molecules in the MAPKpathway including TOPK in both colorectal cancer and nor-mal colorectal tissue to determine which factors best correlatewith phospho-ERK levels.

REFERENCES1. Yoon S, Seger R. The extracellular signal-regulated kinase: multiple

substrates regulate diverse cellular functions. Growth Factors. 2006;24:21–44.

2. Shaul YD, Seger R. The MEK/ERK cascade: From signaling specificityto diverse functions. Biochim Biophys Acta. 2006; Epub ahead of print.

3. Wolf I, Seger R. The mitogen-activated protein kinase signaling cas-cade: from bench to bedside. Isr Med Assoc J. 2002;4:641–647.

4. Hoshino R, Chatani Y, Yamori T, et al. Constitutive activation of the41-/43-kDa mitogen-activated protein kinase signaling pathway in hu-man tumors. Oncogene. 1999;18:813–822.

5. Diaz VM, Hurtado M, Kort EJ, et al. Requirement of the enzymatic andsignaling activities of plasmin for phorbol-ester-induced scattering ofcolon cancer cells. Exp Cell Res. 2006;312:2203–2213.

6. Park KS, Jeon SH, Kim SE, et al. APC inhibits ERK pathway activationand cellular proliferation induced by RAS. J Cell Sci. 2006;119:819–827.

7. Liu Y, Borchert GL, Surazynski A, et al. Proline oxidase activates bothintrinsic and extrinsic pathways for apoptosis: the role of ROS/superox-ides, NFAT and MEK/ERK signaling. Oncogene. 2006;25:5640–5647.

8. Landis SH, Murray T, Bolden S, et al. Cancer statistics, 1998. CACancer J Clin. 1998;48:6–29.

9. Zhu F, Zykova TA, Kang BS, et al. Bidirectional signals transduced byTOPK-ERK interaction increase tumorigenesis of HCT116 colorectalcancer cells. Gastroenterology. 2007;133:219–231.

10. Swedish Rectal Cancer Trial. Improved survival with preoperativeradiotherapy in resectable rectal cancer. N Engl J Med. 1997;336:980–987.

11. Yeoh AS, Bowen JM, Gibson RJ, et al. Nuclear factor kappaB (NFkappaB)and cyclooxygenase-2 (Cox-2) expression in the irradiated colorectum isassociated with subsequent histopathological changes. Int J Radiat OncolBiol Phys. 2005;63:1295–1303.

12. Bonner JA, Vroman BT, Christianson TJ, et al. Ionizing radiation-induced MEK and Erk activation does not enhance survival of irradiatedhuman squamous cell carcinoma cells. Int J Radiat Oncol Biol Phys.1998;42:921–925.

13. Dent P, Yacoub A, Fisher PB, et al. MAPK pathways in radiationresponses. Oncogene. 2003;22:5885–5896.

14. Wang T, Hu YC, Dong S, et al. Co-activation of ERK, NF KappaB andGADD45beta in response to ionizing radiation. J Biol Chem. 2005;280:12593–12601.

15. Watanabe H, Kurabayashi T, Miura M. Inhibition of the extracellularsignal-regulated kinase (ERK) pathway and the induction of radioresis-tance in rat 3Y1 cells. Int J Radiat Biol. 2004;80:451–457.

16. Seong J, Kim S, Suh C. Enhancement of radioresponse of murine tumorsby ERK inhibitor. Ann NY Acad Sci. 2002;973:371–373.

17. LoRusso PM, Adjei AA, Varterasian M, et al. Phase I and pharmaco-dynamic study of oral MEK inhibitor CI-1040 in patients with advancedmalignancies. J Clin Oncol. 2005;23:5281–5293.

18. Sebolt-Leopold JS, Dudley DT, Herrera R, et al. Blockade of the MAPkinase pathway suppresses growth of colon tumors in vivo. Nat Med.1999;5:810–816.

19. LoRusso PM, Krishnaumurthy S, Rinhart J, et al. A phase 1–2 clinicalstudy of a second generation oral MEK inhibitor PD 0325901 in patientswith advanced cancer. Proc Am Soc Clin Oncol. 2005;23:194s(abstract3011).

Corn et al American Journal of Clinical Oncology • Volume 31, Number 3, June 2008

© 2008 Lippincott Williams & Wilkins258