The Association between Telomerase, ~53, and Clinical Staging in Colorectal Cancer

CPT Tommy Brown, MD, CPT Wade Aldous, PhD, CPT Raymond Lance, MD, CPT Jason Blaser, MD, MAJ Thomas Baker, MD, LTC William Williard, MD, Ft. Lewis, Washington

BACKGROUND: A proposed etiology of tumor acti- cleotide sequence (TTAGGG)n onto the ends of telo- vation involves ~63 mutations while telomerase meres. This enzyme is normally present in immortalized may serve as a key enzyme for maintenance of cell lines, germ-line tissues, and tumor tissues. Kim et al7 tumor cell proliferation. demonstrated telomerase activity in 12 different tumor

METHODS: Telomerase activity levels were mea- types and none of 50 normal somatic tissue samples. In sured in colorectal adenocarcinomas and corre- studies of colorectal cancer, telomerase activity was de- sponding normal tissue using a modified telo- tected in more than 90% of tumor specimens.8’9 merit repeat amplification protocol, and ~63 Although much research has been presented on ~53 and mutations were identified using immunohisto- telomerase separately, few researchers have studied the chemical staining. Results were compared with relationship between the tw0.l’ We investigated the asso- staging data using regression analysis. ciation between telomerase activity and ~53 mutations in

RESULTS: Telomerase activity was present in 23 colorectal cancer. By stratifying telomerase activity and of 23 (100%) of the tumors and only 2 (9%) of comparing this with tumor staging and ~53 mutations, we normal specimens (f ~0.0001). The ~63 muta- hope to better define the sequence of events in the malig tions were present in 18 of 23 (78%) of the tu- nant process. mors. No significant correlation between ~63 mutations, telomerase activity levels, and staging METHODS was found.

CONCLUSIONS: Telomerase activity in 100% of the Specimens were prospectively collected from patients un-

tumors suggests telomerase activation is a uni- dergoing elective resection for colon cancer from March

versal event in colorectal tumor progression; 1996 to January 1997. All colon specimens were sent fresh

however, telomerase activity appears to be inde- to pathology. Within 2 hours of resection a sample of the

pendent of ~63 mutations and clinical staging. tumor and of corresponding normal tissue were harvested. Different instrumentation was used for harvest of normal

Am J Surg. 1998;175:364-366.0 ‘1998 by Ex- and tumor samples to avoid contamination. Both samples cerpta Medica, Inc. were then preserved immediately by rapid freezing in iso-

nudson’s’ “two hit” theory propcses a multifactorial pentane and stored at -70°C.

K

Total protein extraction of matched normal and tumor etiology to tumor activation. Oncogene mutations may represent potential “hits” leading to malignant

tissue samples were performed as described.7 Protein ex- tractions were measured using the BCA (bicinchoninic

progression. The ~53 oncogene has been investigated ex- tensively in colorectal cancer. There are multiple studies

acid) protein assay reagent (Pierce, Rockford, Illinois), and

supporting and refuting the prognostic significance of ~53 protein samples were diluted to 1.5 Fg/pL.

Fluorescent TRAP assay. Fifty microliter reactions con- in relation to survival, staging, and differentiation. Despite taining 6 pg of protein extract were assayed in the reaction these differences in outcome, ~53 mutations are consis- tently present in nearly half of the tumors studied.‘-”

mix 1.5 mM MgC12, 20 mM Tris-HCl (pH 8.3), 68 mM KCI, 1 mM EGTA, 0.05% Tween 20,2 mM spermidine, 50

Recently, telomerase has taken a lead role as a potential mM/dNTP, 100 ng F-TS (5’-FAATCCGTCGAGCA- enzyme responsible for tumor longevity. Telomerase is a GAGTT-3’) and CX (5’-CCCTTACCCTTACCCTAA- ribonucleoprotein that synthesizes a spc,cific repeating nu- 3’) and 2 U AmpliTaq DNA polymerase (Perkin-Elmer,

Branchburg, New Jersey). Five microliters of product were mixed with an equal volume of STOP buffer (Blue dextran

From the Department of General Surgery, Madigan Army Med- 2000, deionized formamide; Pharmacia Biotech, Piscat-

I ical Center, Tacoma, Washington. away, New Jersey) and loaded onto ALF DNA sequencing All financial support for this research has bl?en provided by the gels, OS mm thickness (Readymix, Pharmacia Biotech).

Department of Clinical Investigations, Madigan Army Medical Center, United States Army. The opinions or assertions contained

Telomeric repeats were detected with the ALF DNA anal-

herein are the private views of the authors and are not to be ysis system under normal sequencing conditions of 1500 V,

construed as official or reflecting the views cf the Department of 38 W, at a constant temperature of 45°C. Raw data were

the Army or the Department of Defense. analyzed using the Fragment Manager to detect peaks and Requests for reprints should be addressed to MCHJ-CPT determine the height and area of the resulting chromato-

1 rommy Brown, MD, Department of General Surgery, Madigan I Vmy Medical Center, Tacoma, Washington 98431.

grams. The activity level of lo-fold serially diluted samples

Presented at the 84th Annual Meeting of the North Pacific were stratified according to the total area under the peaks

: Surgical Association, Portland, Oregon, November 14-15, 1997. derived from the telomerase assay compared with a known

5 telomerase positive testis standard and measured in per-

I

364 0 1998 by Excerpta Medica, Inc. 0002-961 O/98/$1 9.00

All rights reserved. PII SOOO2-9610(98)00057-9

[ ~53, TELOMEFIASE, AND CANCER STAGING/BROWN ET AL

centages. Low, medium, and high ranges for the telomerase activities were assigned by the spectrum of percentages compared with the standard (low = 0% to 25%, medium = 26% to 50%, high >50%)

Immunohistochemistry. Parrafin-emE,edded blocks of tu- mor adjacent to the tissue harvested in the fresh state were used for immunohistochemistry to detect the presence of p53 mutations. Cut sections were deparaffinized in xylene and rehydrated with alcohol and distilled water. Antibod- ies to p53 (PAb 1801) were supplied by Oncogene Science (Uniondale, New York) and were used according to the manufacturer’s specifications. All stained slides were re- viewed by light microscopy by the same pathologists. The pathologists reviewing the slides were blinded to telomer- ase activity and clinical data. Staining was identified as positive or negative if nuclear stain was present or not present respectively. Slides staining positive were further stratified from 1 + to 4+ based upon overall intensity and percent tumor cells staining, with an estimated scale of ~25% cells positive = l+, 25% to 53% = 2+, 50% to 75% = 3+, and >75% = 4+.

Analysis Tumors were staged according to the standard TNM

staging system.’ ’ Individual variables analyzed included differentiation, nodal status, and stag:. The associations between telomerase activity levels, ~‘53 mutations, and pathological data were analyzed using Spearman rank order and regression analysis. Informed consent was not deemed necessary by the investigational board for this protocol.

RESULTS Twenty-three patients underwent l,esection for colon

cancer. All cases were spontaneous colorectal cancers. The average age was 71 years, and there were 11 women and 12 men. The average size of tumors was 5 cm, and the range was 2 to 13 cm. The staging was as follsws: stage I, 1; stage II, 10; stage III, 6; stage IV, 6. Three tumors were well differentiated, 14 were moderate, and 6 were poorly differ- entiated.

Telomerase activity was detected in 23 of 23 (100%) of the tumor specimens and in 2 of 23 (9%) of corresponding normal colon samples (P = 0.0001). A representative chromatogram showing the low, medium, and high activity levels is shown in Figure 1. The relationships among telomerase activity levels, p53 mutations, and pathological data are shown in the Table Telomerase activity levels were not significantly associated with staging, differentia- tion, or nodal status.

A p53 mutation was detected in 18 (78%) of 23 tumor specimens by nuclear staining. Figure 2 demonstrates nu- clear staining specific for p53 mutations. Of these tumors, 6 had 1 + expression, 4 had 2+ expression, 7 had 3+ expression, and 1 had 4+ expression. The p53 mutations were not significantly associated with clinical stage, differ- entiation, nodal status, or telomerase activity levels.

COMMENTS Telomerase activity has been demonstrated in multiple

tumor lines and is absent in corresponding normal tissues7 In our study of colorectal carcinoma, we found telomerase activity present in all tumor samples. Activity found in the

Figure 1. Telomeric repeats of selected patient’s samples shown by chromatogram. Three different patient samples of varying activity were diluted lo-fold and assayed along with the testis standard. Percentages of enzymatic activity of the samples were calculated by measuring the area underneath the curves as com- pared with the reference testis standard.

two normal tissues likely was due to the incomplete micro- dissection of tumorous tissue from the normal margin, resulting in contamination. Telomerase activity in all stages of colorectal cancer supports the hypothesis that telomerase activation is an early event in the malignant process. Chadeneau et al’ found telomerase activity in 96% of colorectal tumor specimens and in none of 15 adeno- matous polyps. Tahara et al9 demonstrated telomerase ac- tivity in 95% of tumors studied, but also found telomerase activity in 10 of 10 tubular adenomas sampled. Both of these previous studies included both early and late stage tumors, and telomerase activity was rarely found in normal adjacent mucosa, supporting the concept of early activa- tion.

We further defined telomerase activity using a serial di- lution technique. Hiyama et al” showed a significant as- sociation of relevant activity of telomerase with outcome measures in neuroblastoma. We did not find a significant correlation between levels of telomerase activity and other clinicopathological variables to include stage, differentia- tion, and nodal status. We used chromatograms and were

T

Medium, 0.6 pg extract

High, 6.0 pg extmt

High, 0.6 pg extract

Testis, 0.6 pg extract

THE AMERICAN JOURNAL OF SURGERY@’ VOLUME 175 MAY 1998 365

53, TELOMERASE, AND CANCER STAGING/BROWN ET AL 1 1

TABLE Association between Telomerase Activity, p!53 Mutations,

and Clinicopathologic Data

Telomerase Activitv

n Low Moderate High P Value

age I II Ill IV

Differentiation Well Moderate Poor

Nodes negative positive

P= 1+ 2+ 3+ 4+

1 10 3

6 1 6 2

3 0 3 0 14 4 5 5

6 2 3 1 0.74

11 3 6 2 12 3 5 4 0.13

6 2 4 2 7 1 1 0

2 3 1 0.23

2 0 3 0 .70

Telomefase actiwty in fumor samples: 0% to 25% = low; 26% to 50% = moderate; MO% = high. ~53 mutations correspond to the area of nuclear staining: 7 + = ~25%. 2+ = 25% to 50%, 3+ = 50% to 75%, 4+ = >75%. Nodes positive or negative correspond to nodal metzstasis of tumor.

Figure 2. lmmunohistochemical stain for p53 showing nuclear staining of tumor (closed arrow) with adjacent negative staining normal colonic mucosa (open arrow). Magrification xl 00.

therefore able to generate defined values to determine telomerase activity levels, whereas the :3revious study de- fined levels based on serial dilution and detection alone. This difference in defining telomerase .activity as well as the different tumor types may explain the variation in our results from those of the previous study.

We found ~53 mutations in 78% of tumors studied. We did not find any significant correlation between ~53 mu- tations and tumor staging or differentiation, consistent with other studies. 2~4 Wynford-Thomas et al” proposed a model for tumor development in which ~53 causes growth arrest after telomeres reach a critically shortened length, allowing time for genomic repair. They proposed that, in

those cells with a mutation of p53, further telomere erosion results in destabilization and eventual activation of telom- erase to prevent cell death. Based on this hypothesis, one would expect ~53 mutations to be present in all tumors with telomerase activity. Based on our present findings, this is not the case. Furthermore, there were no other signifi- cant associations between telomerase activity and ~53 mu- tations even when accounting for variation in the degree of ~53 expression and telomerase activity. The presence of telomerase activity in all colorectal tumor samples supports the idea that telomerase activation is a universal event in the malignant progression of colorectal cancer. The etiol- ogy of oncogene mutations may be a direct result of telom- erase activation or a random event. Our data support no direct association between telomerase activity and ~53 mutations; however, further studies are clearly needed to define these relationships.

CONCLUSIONS Telomerase is present in a significant number of colorec-

tal tumors as compared with the normal adjacent tissue.

Telomerase activation in both early and late stage tumors supports the idea that telomerase activation is a universal event in the malignant progression of colorectal cancer. Telomerase activity is not associated with ~53 mutations or clinical staging.

REFERENCES 1. Knudson AG. Mutation and cancer: statistical studv of retino- blastoma. Proc Nutl Acad Sci USA. 1971;68:820-823. ’ 2. Yamaguchi A, Kurosaka Y, Fushida S, et al. Expression of p53 protein in colorectal cancer and its relationship to short term prognosis. Cancer. 1992;70:2778-2784. 3. Purdie CA, O’Grady J, Piris J, et al. ~53 expression in colorectal tumors. Am J Pa&l. 1991;138:807-813. 4. Zeng 2, Sarkis AS, Zhang 2, et al. ~53 nuclear overexpression: an independent predictor of survival in lymph node-positive colo- rectal cancer patients. .I Clin Oncol. 1994;12:2043-2050. 5. Nathanson SD, Linden MD, Tender P, et al. Relationship among ~53, stage, and prognosis of large bowel cancer. Dis Colon Rectum. 1994;37:527-534. 6. Bell SM, Scott N, Cross D, et al. Prognostic value of ~53 overexpression and c-Ki-ras gene mutations in colorectal cancer. Gustroenterology. 1993;104:57-64. 7. Kim MW, Piatyszek MA, Prowse KR, et al. Specific association of human telomerase activity with immortal cells and cancer. Science (Washington DC). 1994;266:2011-2015. 8. Chadeneau C, Hay K, Hirte HW, et al. Telomerase activity associated with aquisition of malignancy in human colorectal can- cers. Cancer Res. 1995;55:2533-2536. 9. Tahara H, Kuniyasu H, Yokozaki H, et al. Telomerase activity in preneoplastic and neoplastic gastric and colorectal lesions. Clin Cancer Res. 1995;1:1245-1251. 10. Wynford-Thomas D, Bond JA, Wyllie FS, Jones CJ. Does telomere shortening drive selection for ~53 mutation in human cancer? Mel Carcinogen. 1995;12:119-123. 11. Sabistan DC Jr, Lyerly HK. Essentials of Surgery. 2nd ed. Philadelphia: WB Saunders; 1994:355. 12. Hiyama E, Hiyama K, Yokoyama T, et al. Correlating telom- erase activity levels with human neuroblastoma outcomes. Nature Med. 1995;1:249-255.

366 THE AMERICAN JOURNAL OF SURGERY@ VOLUME 175 MAY 1998