Somatic DNA Alterations in Breast Carcinomas of Different Lymph-Node Status by DNA Fingerprint Analyses

Somatic DNA Alterations in Breast Carcinomas of Different Lymph-Node Status by DNA Fingerprint Analyses Regine Schneider-Stock, Thomas Günther, Albert Roessner, and Jörg T. Epplen

ABSTRACT: The purpose of this study was to screen for somatic changes in invasive breast tumors by multilocus DNA fingerprints comparing normal (blood) and malignant tissue samples from 34 patients. The comparison of lymph node–positive and node–negative breast carcinomas was of primary interest. After restriction enzyme digestion with HinfI and HaeIII, altered banding patterns were detected by using the oligonucleotide probe (GTG) 5 in 7 of 34 (20.5%) and in 3 of 34 (8.8%) tumors after hybridization with (GACA)4. The overall frequency of changes thus amounted to 29.4%. Because long (GACA) n repeat motifs, generating predominant DNA fingerprint bands, are localized on the short arms of the human acrocentric chromosomes, sequences that are important in breast carcinogenesis may be present in these regions. The overall methylation status of the DNA does not appear to be responsible for DNA fingerprint differences, as can be demonstrated with the restriction endonuclease HaeIII. DNA fingerprint differences did not correlate with tumor grade, stage, and hormone receptor status. Tumors with lymph-node metastases expressed DNA fingerprint differences more frequently. © Elsevier Science Inc., 1998

INTRODUCTION Common cancer often develops sequentially from a normal cell toward an invasive, metastatic tumor [1]. This process may include chromosomal aberrations, gene amplifications, somatic recombinations, point mutations, or all four [2]. Compared with the germ-line configuration, detection of DNA differences in tumors is of utmost importance for the identification of relevant genes and mutations involved in tumorigenesis. Such genomic areas may provide clues to relevant genes through loss of heterozygosity (LOH)—for example, as revealed by informative markers in the vicinity of tumor suppressor genes [3–5]. Multilocus DNA fingerprinting has the potential to efficiently screen for genetic changes in tumors owing to the somatic stability of the fingerprint patterns in one individual. By the comparison of DNAs from tumors and the corresponding nonneoplastic reference tissues, a variety of

From the Department of Pathology, Medical Faculty, Otto von Guericke University (R. S.-S., T. G., A. R.), Magdeburg, Germany; and Molecular Human Genetics, Ruhr-University (J. T. E.), Bochum, Germany. Address reprint requests to: Dr. Regine Schneider-Stock, Department of Pathology, Medical Faculty of the Otto-von-Guericke University, 39120 Magdeburg, Germany. Received April 7, 1997; accepted September 10, 1997. Cancer Genet Cytogenet 103:149–154 (1998)  Elsevier Science Inc., 1998 655 Avenue of the Americas, New York, NY 10010

different tumors have already been studied for DNA alterations: bladder [6], colon [7], leukemia [8], ovary [9], prostate [10], and renal cell carcinoma [11]. Furthermore, diagnostic applications of repetitive DNA sequences are useful in determining the clonality and the relation of primary tumors to metastases [12–14], screening for DNA amplifications [15, 16] and DNA methylation status [6], monitoring chromosomal losses and rearrangements [17–19], or differentiating between cell lines of the same tumor [20–22]. Numerous clinicopathologic prognosis factors have been established for breast carcinoma. Along with tumor extension, grading and lymph-node status, hormone receptor expression, and DNA ploidy are the most important. In addition to the diagnostic possibilities at the molecular level, an increasing number of tumor-associated genetic alterations can be detected, whose prognostic significance must be compared with the traditional factors. In our studies on 34 breast carcinomas, we correlated the results with those obtained by DNA fingerprint analysis by using conventional clinicopathologic prognostic factors. MATERIALS AND METHODS Tumors Our investigation material comprised 34 randomly selected, invasive carcinomas of the breast (Table 1). The

0165-4608/98/$19.00 PII S0165-4608(97)00395-6

150

R. Schneider-Stock et al.

study included only female patients between 45 and 89 years of age (mean age: 60.8). Representative tissue was prepared, shock-frozen in liquid nitrogen, and stored at 2708C until use. Peripheral venous blood was obtained from patients before surgery. Their tumors were removed before irradiation or chemotherapy. Staging, Typing, and Grading The maximal tumor diameter was recorded. If necessary, measurements were corrected after histologic assessment. The number, origin, and metastases in regional lymph nodes were determined according to UICC classification [23]. Histologic typing of the tumors was performed according to the World Health Organization [24]. The grade of the tumors was according to the three-stage malignancy grading inaugurated by Bloom and Richardson [25]. Immunohistochemistry The presence of estrogen receptors and progesterone receptors was monitored by monoclonal antibodies (Abbott Laboratories). The antibodies were bound to formaldehyde–phosphate-buffered-saline-fixed frozen sections (4 mm). The peroxidase–antiperoxidase complex was linked with a bridge antibody after incubation with a chromogen substrate solution. The estrogen receptor monoclonal antibody complexes were visualized with a light microscope. Both positive and negative control sections were treated similarly. The immunoreactive score was established according to Remmele [26]. Proliferation Marker MIB-1 For immunohistochemistry, the deparaffinized sections were boiled in 0.01 M citrate buffer (pH 6.0) under increased pressure for 30 minutes, cooled afterward, and rinsed with running water for 15 minutes. Sections were incubated with primary antibodies anti-Ki-67 and MIB-1 (Oncogene Science, USA) and diluted in a ratio of 1:30 for 60 minutes at 378C in a humidified chamber. Staining was revealed by applying the Alkaline Phosphatase Anti Alkaline Phosphatase technique [27]. To determine the percentage of MIB-1-positive cells (proliferation index), 40 high-power fields in representative areas with approximately 1,000 cells each were examined.

In case labeled cells were distributed heterogeneously, the region with the highest density of MIB-1-positive nuclei was regarded as the area of proliferative activity of the tumor. The cutoff point used for MIB-1 staining is as follows: high proliferation index is greater than 20% tumor cells stained. DNA Ploidy Measurements DNA ploidy was measured on tumor samples representing the least-differentiated part by means of the AHRENS Image Cytometry System data-bank system with the use of 8-mmthick Feulgen-stained sections. We used 30 fibrocytes beyond the tumor for calibration. The most important proportional errors caused by glare were optimally rectified by a glare correction of 25% [28]. The thickness of the section having been chosen, nuclear cut and overlapping could be avoided by thorough focusing. At least 200 tumor cells were measured from each specimen. We used only DNA stem-line ploidy to compare the results obtained by DNA fingerprint analysis, facilitating a subdivision into diploid and nondiploid tumors. With preliminary experiments on 50 normal tissues taken into account, DNA indices of 1.0 6 0.2 were classified as diploid, the remaining ones as nondiploid. DNA Preparation and Oligonucleotide Fingerprinting Genomic DNA was isolated according to standard procedures [29] from peripheral blood lymphocytes and freshly frozen tumor tissues. A control Hematoxylin and Eosin (H & E) slide was stained to monitor the contamination with normal epithelial cells and stromal cells. Only areas with more than 50% of tumor cells were included in the study. The quality of high-molecular DNA was controlled in 0.7% agarose gels stained with ethidium bromide. From each sample, 10 mg of DNA was digested with the restriction endonucleases Hinf I and HaeIII according to the manufacturer’s recommendations. The DNA fragments were separated on 0.8% horizontal agarose gels (GNA 200, Pharmacia, Biotech) in Tris-borate-EDTA buffer for as long as 40 hours at 35 V. Gels were dried and used for in gel hybridization with the radiolabeled probes (GTG)5 and (GACA)4 [30, 31].

Table 1 Alterations in DNA fingerprint patterns observed with probe (GTG)5 with two different restriction enzymes in DNA from patients with breast tumors Altered DNA fragment size (kb) Hinf I Patient

HaeIII

Loss of band

New band

Intensity shift

Loss of band

New band

Intensity shift

5 11 18

— — —

1 4

— — —

— 5 —

— — —

13

5.8 8 — — 2.2

—

—

4.8

3.5 — 1 1.5 —

— — —

— — —

5 — —

— 5 —

— — —

29 26 22

—

151

DNA Fingerprinting in Breast Carcinomas Statistical Analyses Significance of associations between the occurrence of DNA fingerprint differences, clinicopathologic parameters, DNA ploidy, and MIB-1 proliferation index, respectively, were estimated by using Fisher’s exact test [32]. A P value of less than 0.05 was considered statistically significant. RESULTS We analyzed DNA from invasive breast carcinomas and the corresponding constitutional DNA by using genetic fingerprinting with the oligonucleotide probes (GTG)5 and (GACA)4 after restriction digestion with Hinf I and HaeIII. In the multilocus banding patterns obtained, clearly resolvable, hypervariable fragments ranged from more than 23 kb to 1 kb (Fig. 1). The DNA fingerprint evaluations are given in Tables 1 and 2. All in all, differences were observed in 10 of 34 tumors (29.4%) and the corresponding constitutional DNA. These differences included apparent deletions [6 with (GTG)5 and three with (GACA)4] and appearance of novel bands [eight with (GTG)5]. Once, a significant increase in intensity of a band in tumor DNA was detected after Hinf I digestion and hybridization with (GACA)4. Most of the band alterations in the breast tumors (n 5 11) were discovered after HaeIII digestion, especially when analyzed with (GTG)5. As detailed in Tables 1 and 2, changes with the use of (GTG)5 were seen only after digestion with Hinf I in one patient, whereas six band changes in five patients were detected only after digestion with HaeIII. Three tumors (patients 5, 13, and 18) showed changes with both restriction enzymes. Alterations with both oligonucleotide probes were not observed. The summarized data on clinicopathologic parameters and DNA fingerprint analyses were given in Table 3. Seven of 19 (37%) invasive ductal carcinomas and 5 of 11 invasive lobular carcinomas expressed DNA fingerprint differences. In two medullary carcinomas, no band alterations were identified. Five of 12 (41.7%) of the grade II tumors and 4 of 17 (23%) of the grade III carcinomas showed altered DNA fingerprints (grade I 1 II versus grade III: P 5 0.363). Band alterations could be observed in 3 of 11 pT1 (27%), 5 of 13 pT2 (38%), and 2 of 5 pT3 tumors (40%) (P 5 0.882). Seven of 12 tumors (58.3%) with somatic changes had developed lymph-node metastases (P 5 0.015). Only 3 of 19 lymph node–negative tumors (15.8%) showed DNA fingerprint differences. In 7 of 19 estrogen receptor–positive tumors (36.8%) and 5 of 19 progesterone-receptor-expressing tumors (26.3%), somatic changes were observed (P 5 0.279 and P 5 0.606, respectively). Only 3 of 15 estrogen receptor–negative tumors (20%) and 4 of 15 progesterone receptor–negative tumors (26.6%) showed banding alterations. Somatic changes were found in patients between 47 and 89 years of age (average age: 59.5). Patients without any alteration in the DNA fingerprint patterns were between 48 and 83 years old (average age: 62.9). From the total of 20 aneuploid and 1 tetraploid tumors, 5 of 10 tumors (50%) with banding alterations were aneuploid (data not shown). The remaining 50% of tumors with altered fingerprint patterns showed diploid stem lines (total number of

Figure 1 Hybridization patterns of normal (N) and tumor (T) DNA from breast carcinoma patients. Size markers are given on the left-hand side. Changes are indicated by arrows. (A) (GTG)5 hybridization after Hinf I (a) and HaeIII (b, c, d) digestion: a, d, new band (tumor 18 and 5, respectively); b, loss of band (tumor 13); c, no detectable changes (tumor 24). (B) (GACA)4 hybridization after HaeIII digestion: loss of two bands (tumor 1).

diploid tumors 10; P 5 0.094). Tumors with DNA fingerprint differences had an average MIB-1 index of 20% (data not shown). Tumors without any altered banding patterns exhibited a lower proliferation index (12%; P 5 0.020). For 3 tumors, data on ploidy and proliferation index were not available. DISCUSSION To monitor genomic changes during carcinogenetic processes in breast tumors, DNA fingerprinting was used. The somatic stability of the oligonucleotide-generated patterns

152

R. Schneider-Stock et al. Table 2 Band alterations in DNA fingerprints observed with probe (GACA)4 with two different restriction enzymes in DNA from patients with breast tumors Altered DNA fragment size (kb) Hinf I Patient

HaeIII

Loss of band

New band

Intensity shift

Loss of band

New band

Intensity shift

8

—

—

1.5

—

—

1 14

— —

— —

— —

— 4 5 2.5

— —

— —

was demonstrated by using, for example, the probes (GTG)5 and (GACA)4 [16]. Here, 34 invasive breast tumors were analyzed, 10 of which (28.6%) exhibited DNA fingerprint patterns that differed from those of corresponding constitutive DNA. The observed changes mainly include gain or loss of DNA fragments containing large amounts of simple repeats or, only rarely, shifts in band intensity. Increased signal intensity in a specific pattern position may be due to overlay of a second band or related to localized amplification of DNA in a given segment, including hypervariable simple repetitive DNA. The appearance of a new band in tumor DNA might be associated with gain or loss of restriction enzyme recognition sites, with their differential methylation, or with genetic rearrangements, or it

Table 3 Correlation between clinico-pathologic parameters and DNA fingerprinting results Fingerprint differences Histologic parameter Typing IDC ILC MC tub./lob. ca Grading 1 2 3 unknown Staging pT1 pT2 pT3 unknown Lymph-node metastasis nodal positive nodal negative unknown Estrogen receptor positive negative Progesterone receptor positive negative

n

Yes

No

19 11 2 2

7 5 0 0

12 6 2 2

1 12 17 4

0 5 4 1

1 7 13 3

11 13 5 5

3 5 2 0

8 8 3 5

12 19 3

7 3 0

5 16 3

20 14

7 3

13 11

20 14

5 5

15 9

Abbreviations: tub./lob. ca, tubular/lobular carcinoma; IDC, intraductal carcinoma; ILC, intralobular carcinoma; MC, medullar carcinoma.

could be a derivative of unequal mitotic recombination. Any of these events could play an important role in the pathogenesis of the breast tumor or they could occur “secondarily” during the clonal expansion of the tumor cells. Whereas (GTG)5 fingerprints revealed most genomic alterations (12 of 16 band alterations, 75%), the detection rate for the oligonucleotide probe (GACA)4 was only 25%. Nevertheless, (GACA)4 revealed changes in the tumor DNA where (GTG)5 failed to demonstrate any deviation in three tumors. Hinf I sites terminating in methylated cytosine are resistant to enzymatic cleavage, and differences in methylation of DNA may alter the Hinf I cleavage, leading to changes in restriction fragments [33]. However, the DNA fingerprint alterations are not due to generalized tumor-specific methylation changes in DNA, because HaeIII digestion (not affected by methylation) also demonstrated corresponding DNA fingerprint changes. In addition, we observed a slight trend toward a more advanced stage in tumors with DNA fingerprint differences. The analysis of other clinicopathologic parameters revealed a significant association with somatic changes in DNA fingerprinting: lymph node-positive tumors expressed fingerprint differences more frequently than node–negative carcinomas (P , 0.0105, logistic regression). The probability of recognizing lymph node-positive tumors correctly was 70%; 78.3% of these tumors were correctly recognized as tumors without lymph-node metastases. Tavassoli [34] reported that survival, recurrence, speed of recurrence, and treatment failure correlate with the number of positive axillary nodes. Consequently, patients with lymphnode metastases have a poor prognosis. Thus DNA fingerprint shifts in the tumor might be indicative of a worse clinical course. There was no significant correlation between genetic data and tumor grade or type. The MIB-1 proliferation index is significantly higher in patients with somatic changes (P , 0.05). Consequently, both markers seem to be associated with each other in breast carcinomas. Tavassoli [34] reported that a high proliferation index correlates with the degree of tumor differentiation, vascular invasion, and lymph-node metastases. We could not confirm this statement, because our results suggest that node-negative breast carcinomas had a mean proliferation index similar to node-positive ones (13.9% and 14.9%, respectively). These results have to be confirmed on the basis of more cases. In contrast with previous reports [35, 36], we did not observe an association between somatic instability and ploidy of breast tumors.

DNA Fingerprinting in Breast Carcinomas Genetic instability, as demonstrated by DNA fingerprinting, seems to play a role in the carcinogenesis of breast tumors because it is present in a significant proportion of the carcinomas investigated (29.4%). Otherwise, changes remain undetected in the surplus DNA of the constitutional patterns. Alternatively, the acquisition of somatic alterations may be a relatively late event in the development of breast carcinomas, as suggested by Burke et al. [37]. To be demonstrable, such cell clones with altered DNA fingerprint patterns have to overgrow other tumor cells with constitutional patterns. In conclusion, our study demonstrates that DNA fingerprinting is useful for detecting and initially characterizing genomic instability. The significant differences between tumors developing lymph-node metastases and those that do not should be the subject of further investigation, especially with respect to the worse prognosis on breast tumors.

REFERENCES 1. Vogelstein B, Kinzler KW (1992): P53 function and dysfunction. Cell 70:523–526. 2. Garrett CT, Liscia DS, Nasim S, Ferreira-Gonzalez A (1995): Genetics of colorectal and breast cancer. Clin Lab Med 15:957–971. 3. Dorion-Bonnet F, Mautalen S, Hostein I, Longy M (1995): Allelic imbalance study of 16q in human primary breast carcinomas using microsatellite markers. Genes Chromosom Cancer 14:171–181.

153 various human tumors and their metastases. Cancer Res 52:2174–2179. 14. Sheu JC, Huang GT, Chou HC, Lee PH, Wang JT, Lee HS, Chen DS (1993): Multiple hepatocellular carcinomas at the early stage have different clonality. Gastroenterology 105: 1471–1476. 15. Nürnberg P, Zischler H, Fuhrmann E, Thiel G, Losanova T, Nisch G, Witkowski R, Epplen JT (1991): Coamplification of simple repetitive DNA fingerprint fragments and the EGFR gene in human gliomas. Genes Chromosom Cancer 3:79–88. 16. Nürnberg P, Barth I, Fuhrmann E, Lenzner C, Losanova T, Peters C, Pöche H, Thiel G (1991): Monitoring genomic alterations with a panel of oligonucleotide probes specific for various simple repeat motifs. Electrophoresis 12:186–192. 17. Haffty BG, Carter D, Flynn SD, Fischer DB, Brash DE, Simons J, Ziegler AM, Fischer JJ (1993): Local recurrences versus new primary: clinical analysis of 82 breast relapses and potential applications for genetic fingerprinting. Int J Radiat Oncol Biol Phys 27:575–583. 18. Hoffner L. Shen-Schwarz S, Deka R, Chakravarti A, Surti U (1992): Genetics and biology of human ovarian teratomas III: cytogenetics and origins of malignant ovarian germ cell tumors. Cancer Genet Cytogenet 62:58–65. 19. Thein SL, Jeffreys AJ, Cotter F, Flint J, O’Connor NTJ, Weatherall DJ, Wainscoat JS (1987): Detection of somatic changes in human cancer by fingerprint analysis. Br J Cancer 55:353–356. 20. Hane B, Tummler M, Jager K, Schleithoff L, Janssen JW, Drexler HG (1992): Differences in DNA fingerprintings of continuous leukemia-lymphoma cell lines from different sources. Leukemia 6:1129–1133. 21. Pakkala S (1992): DNA fingerprints in leukemia. Int J Clin Lab Res 21:269–272.

4. Eiriksdottir G, Sigurdsson A, Jonasson JG, Agnarsson BA, Sigurdsson H, Gudmundsson J, Bergthorsson JT, Barkardottir RB, Egilsson V, Ingvarsson S (1995): Loss of heterozygosity on chromosome 9 in human breast cancer: association with clinical variables and genetic changes at other chromosome regions. Int J Cancer 64:378–382.

22. Speth C, Epplen JT, Oberbäumer I (1991): DNA fingerprinting with oligonucleotides can differentiate cell lines derived from the same tumor. In Vitro Cell Dev Biol 27A:646–650.

5. Radford DM, Phillips NJ, Fair KL, Ritter JH, Holt M, DonisKeller H (1995): Allelic loss and the progression of breast cancer. Cancer Res 55:5180–5183.

24. World Health Organization (1981): Histological Typing of Breast Tumors. ed 2. World Health Organization, Geneva.

6. Agurell E, Li R, Rannug U, Norming U, Tribukait B, Ramel C (1992): Detection of DNA alterations in human bladder tumors by DNA fingerprint analyses. Cancer Genet Cytogenet 60:53–60. 7. Bocker T, Schlegel J, Kullmann F, Stumm G, Zirngibl H, Epplen JT, Rüschoff J (1996): Genomic instability in colorectal carcinomas: comparison of different evaluation methods and their biological significance. J Pathol 179:15–19. 8. Pakkala S (1991): DNA fingerprinting in the detection of residual disease in acute leukemia. Leukemia 5:437–440. 9. Chen P, Hurst T, Khoo SK (1991): Detection of somatic DNA alterations in ovarian cancer by DNA fingerprint analysis. Cancer 67:1551–1555. 10. Werely J, Heyns F, Van Velden DJJ, Van Helden PD (1996): DNA fingerprint detection of somatic mutations in benign prostatic hyperplasia and prostatic adenocarcinoma. Genes Chromosom Cancer 17:31–36. 11. Bock S, Epplen JT, Noll-Puchta H, Rotter M, Hofler H, Block T, Hartung R, Jakse G, Wilmanns W, Petrides PE (1993): Detection of somatic changes in human renal cell carcinomas with oligonucleotide probes specific for simple repeat motifs. Genes Chromosom Cancer 6:113–117. 12. Bettink S, Wullich B, Christmann A, Zwergel T, Zang KD, Unteregger G (1992): Genetic heterogeneity of prostatic carcinoma-derived cell lines as emphasized by DNA fingerprinting. Electrophoresis 13:644–646. 13. Matsumura Y, Tarin D (1992): DNA fingerprinting survey of

23. Hermanek P, Scheibe O, Spiessl B, Wagner G (1987): TNMKlassifikation maligner Tumoren (UICC) 4. Auflage, Springer, Berlin, Heidelberg, New York.

25. Bloom HJG, Richardson WW (1957): Histologic grading and prognosis in breast cancer. Br J Cancer 11:359–377. 26. Remmele W (1986): Mammakarzinom: Prognostisch wichtige Rezeptorbestimmungen. Dtsch Ärztebl 83:3359–3362. 27. Vojtesek B, Bartek J, Midgley CA, Lane DP (1992): An immunochemical analysis of the human nuclear phosphoprotein p53: new monoclonal antibodies and epitope mapping using recombinant p53. J Immunol Methods 152:237–244. 28. Kindermann D, Hilgers CH (1994): Glare correction in DNA image cytometry. Anal Cell Pathol 6:165–180. 29. Sambrook J, Fritsch EF, Maniatis T (1989): Molecular Cloning: A Laboratory Manual. ed 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor. 30. Epplen JT, Ammer H, Epplen C, Kammerbauer C, Mitreiter R, Roewer L, Schwaiger W, Steimle V, Zischler H, Albert E, Andreas A, Beyermann B, Meyer W, Buitkamp J, Nanda I, Schmid M, Nürnberg P, Pena SDJ, Pöche H, Sprecher W, Schartl M, Weising K, Yassouridis A (1991): Oligonucleotide fingerprinting using simple repeat motifs: a convenient, ubiquitously applicable method to detect hypervariability for multiple purposes. In: DNA Fingerprinting: Approaches and Applications. T Burke, G Dolf, AJ Jeffreys, and R Wolff, eds. Birkhäuser, Basel, pp. 50–69. 31. Epplen JT (1992): DNA fingerprinting using oligonucleotide probes specific for simple repeat motifs. In: Advances in Electrophoresis. A Chrambach, MJ Dunn, BJ Radola, eds. VCH-Verlag, Weinheim, pp. 59–114.

154 32. Sachs L (1992): Angewandte Statistik. ed 7. Springer, Heidelberg, Berlin, New York. 33. Nelson M, McClelland M (1987): The effect of site specific methylation on restriction-modification enzymes. Nucleic Acids Res 15(Suppl.):219–230. 34. Tavassoli FA (1992): Pathology of the breast. Appleton & Lange, Norwalk. 35. Hovig E, Mullaart E, Borresen AL, Uitterlinden AG, Vijg J

R. Schneider-Stock et al. (1993): Genome scanning of human breast carcinomas using micro- and minisatellite core probes. Genomics 17:66–75. 36. Smith FB, Zappi ME (1993): Relationships between image cytometric DNA index, proliferation fraction and multiploidy and conventional nuclear grade in breast carcinoma. Mod Pathol 6:606–611. 37. Burke HB, Hutter RVP, Henson DE (1995): Breast carcinoma. In: Prognostic Factors in Cancer. Hermaneck et al., eds. Springer, Geneva, pp. 165–177.