- Email: [email protected]
CHANGES IN COLLAGEN METABOLISM IN PROSTATE CANCER: A HOST RESPONSE THAT MAY ALTER PROGRESSION
0022-5347/01/1665-1698/0 THE JOURNAL OF UROLOGY® Copyright © 2001 by AMERICAN UROLOGICAL ASSOCIATION, INC.®
Vol. 166, 1698 –1701, November 2001 Printed in U.S.A.
CHANGES IN COLLAGEN METABOLISM IN PROSTATE CANCER: A HOST RESPONSE THAT MAY ALTER PROGRESSION N. BURNS-COX, N. C. AVERY, J. C. GINGELL
AND
A. J. BAILEY
From the Bristol Urological Institute, Southmead Hospital and Collagen Research Group, University of Bristol, Bristol, United Kingdom
ABSTRACT
Purpose: The interaction of cancer cells and the extracellular matrix is essential for cancer progression. However, little is known about the influence of cancer on the metabolism of collagen, which is the major constituent of the extracellular matrix. We studied changes in collagen metabolism in prostate cancer with increasing Gleason score and correlated them with clinical parameters and patient survival. Materials and Methods: Collagen content and clinical parameters in 3 types of prostatic tissue were compared, including the foci of prostatic cancer, unaffected tissue from the same cancerous prostates and prostatic tissue from patients with no evidence of cancer. In addition, to assess collagen metabolism tissue obtained prospectively from 45 patients undergoing prostatic biopsy was assessed for collagen content, the type and extent of collagen cross-linking, serine proteinase, matrix metalloproteinase and the C-terminal propeptide of type I collagen. Results: With increasing Gleason sum of the cancer collagen content at the focus decreased, while that of surrounding unaffected tissue from the same prostate increased to levels significantly above that from controls with no cancer. Markers of collagen synthesis in the prostate biopsy material were significantly increased in the presence of prostate cancer. Conclusions: In prostate cancer there are changes in collagen metabolism not only in the cancer focus but also in nearby histologically benign prostatic tissue. These observed changes are related to the Gleason score of the tumor and may represent a host response. Collagen content in the surrounding unaffected tissue may be a predictor of patient survival. KEY WORDS: prostate, prostatic neoplasms, collagen, extracellular matrix, disease progression
The hallmark of cancer is its ability to invade locally adjacent host structures and metastasize. Liotta et al described 3 basic steps essential to this process, namely adhesion, degradation and migration.1 Each step involves interactions of cancer cells with components of the host basement membrane and extracellular matrix, primarily collagen. Cancer cells adhere to components of the extracellular matrix through the specific interaction of cell adhesion integrins2 or to host basement membrane through laminin.3 Degradation occurs after the secretion of proteolytic enzymes capable of degrading specific components of basement membrane or extracellular matrix. The 3 major groups of such proteinases considered important for cancer cell invasion are the cysteine proteinases, the serine proteinases and the matrix metalloproteinases. A total of 14 matrix metalloproteinases have been identified, of which each has differing substrate specificities for components of the extracellular matrix or basement membrane.4 There are also complex interactions among the proteinase groups, for example the serine proteinase plasmin can activate matrix metalloproteinases.5 In addition, the groups of proteinases have specific activators, for example urokinasetype plasminogen activator,6 and inhibitors, such as tissue inhibitors of matrix metalloproteinases7 and plasmin activator inhibitor.6, 8 Although the system is complex, it tightly regulates the processes involved in normal healing and remodeling but it is exploited by cancer cells during invasion. In normal tissue cell mobility is rare after organogenesis and cellular differentiation. However, cellular migration is essential for invasion of the host extracellular matrix and subsequent metastasis by cancer. In addition to growth factors9 and tumor secreted factors,1 components of the extraAccepted for publication June 22, 2001. Supported by the Andrology Research Fund.
cellular matrix can promote cancer cell motility, for example laminin, fibronectin and collagen type IV.10 In the prostate cancer cell line LNCaP Murphy et al observed that the growth rate, prostate specific antigen (PSA) production and sensitivity to androgen depended on collagen type I in the culture medium.11 Therefore, it seems that the clinical outcome of cancer is profoundly influenced by components of the host extracellular matrix. Cancer cells injected into athymic mice do not develop into tumors unless provided with constituents of the extracellular matrix.12, 13 Collagen is the most abundant protein in mammals, and a major constituent of extracellular matrix and basement membranes.14 It also has an important role in the response of the body to chronic injury. The major characteristic of such repair is the deposition of scar tissue, which is predominantly composed of fibrous collagen. Because cancer invasion is a chronic and often progressive condition, it seems likely that such healing may form part of the host response with associated collagen deposition. If the host response is vigorous, it may impede or even prevent cancer cell invasion of healthy tissue but little is known about the influence of cancer on collagen metabolism. We report changes in collagen metabolism in prostate cancer, which is a host response that may alter progression. MATERIALS AND METHODS
We analyzed the collagen content of unstained prostate sections obtained for histological testing from patients with prostate cancer and determined the levels of collagen metabolism from fresh samples obtained separately at biopsy. Quantification of collagen in benign and malignant prostatic tissue sections. A computer search identified 101 patients with prostate cancer who had undergone transurethral
1698
CHANGES IN COLLAGEN METABOLISM IN PROSTATE CANCER
prostatic resection at The Bristol Urological Institute between 1994 and 1997. We obtained sections of resected tissue before hematoxylin and eosin staining was done. In each case we identified a portion of tissue containing prostate cancer and a portion without evidence of prostate cancer. As a control group, we also obtained similar tissue samples from another 22 patients with benign prostatic hyperplasia (BPH) alone and no evidence of prostate cancer. A histopathologist with experience in prostate cancer performed tissue identification and categorization as well as repeat Gleason scoring. To assess patient survival additional data on age, clinical stage, prostate specific antigen (PSA) and bone scans were obtained from the medical records. To test the reproducibility of the repeat Gleason scoring 10 specimens were randomly selected and evaluated by another experienced histopathologist. To determine whether small cancerous foci may have been concealed within the noncancerous portion of the tissue from patients with prostate cancer we performed multiple sectioning and staining in 10 tissue portions to detect evidence of deep seated foci. Determining collagen content. Each portion of tissue was de-waxed using repeat changes of xylene and then lyophilized and weighed before acid hydrolysis in 6N hydrochloric acid at 112C for 24 hours. Hydroxyproline content of each hydrolysate was determined using the method of Bannister and Burns.15 That is, since hydroxyproline represents approximately 14% of fibrous collagen dry weight, multiplication by 7.14 yields the percent of collagen of each portion of tissue. Collagen metabolism in prostatic biopsies from malignant and benign prostates. After obtaining informed consent biopsy was prospectively obtained from the posterolateral aspect of each lobe of the prostate in 45 consecutive patients referred for transrectal ultrasound and biopsy due to the suspicion of prostate cancer based on increased PSA and/or clinically suspicious digital rectal examination. This biopsy was performed in addition to routine sextant biopsies, which were sent for histological examination. It was snap frozen in liquid nitrogen and stored at ⫺80C until analyzed. Each biopsy sample was washed in phosphate buffered saline, blot dried and unequally divided in 2 portions. The larger portion was used to determine the amount and type of intermediate and mature collagen cross-links after sodium borohydride reduction16 and acid hydrolysis, the amount of advanced glycation end product and total collagen content, as
1699
described. The smaller portion was agitated overnight at 4C in 20 mM. triethanolamine containing 0.1% polyoxyethylene lauryl ether. The extract was used to determine intracellular proteinase and the level of carboxy-terminal propeptide of type I collagen. Extracts were characterized by reference to standards after gelatin or casein substrate polyacrylamide gel electrophoresis, and quantified by computer linked transmittance densitometry and by a commercially available immunoassay. Statistical analysis. The paired and independent samples t tests were done as appropriate. Levene’s test was used to determine equal or unequal variance. We performed linear regression and calculated the Pearson correlation coefficient to determine the relationships of 2 continuous variables. Univariate Cox regression analysis was also done to determine whether the clinical or biochemical parameters measured predicted patients survival.
RESULTS
We analyzed 3 types of prostatic tissue. In the 101 patients in the cancer malignant group a prostatic tissue sample was known to contain a cancer focus. In the 90 patients in the cancer benign group a paired prostatic tissue sample from a malignant prostate had no evidence of cancer on histological examination. In the 22 patients in the BPH control group with no evidence of prostate cancer the prostatic tissue sample had the appearance of BPH only. Correlating collagen content with Gleason sum. Figure 1 shows the change in mean collagen content in the 3 groups compared with the Gleason sum. In tissue with a Gleason sum of less than 5 there was no significant difference in collagen content in the groups. When the Gleason sum was 5 and greater cancer malignant group tissue collagen content was always lower than in cancer benign group tissue and overall there was a significant difference (paired t test p ⬍ 0.001). Over the same Gleason sum range the collagen content of cancer benign group tissue was also higher than that of control tissue but it was not statistically significant until Gleason sum 8 and greater. Clinical data were available for 90 of the 101 patients 55 to 94 years old at diagnosis (mean age 74) with prostate cancer. Mean followup after diagnosis was 41 months. Digital rectal examination demonstrated stages T1/T2 and T3/T4 disease
FIG. 1. Collagen content of cancer benign (CBG), cancer malignant (CMG) and BPH prostate tissue
1700
CHANGES IN COLLAGEN METABOLISM IN PROSTATE CANCER
0.03 0.065 0.08
loproteinase 2 expression and activation increased in cancer malignant group biopsies, so did carboxy-terminal propeptide of type I collagen and hydroxylysinoketonorleucine (R ⫽ 0.04, p ⫽ 0.008 and R ⫽ 0.58, p ⬍ 0.001, respectively). These data demonstrated coordinated increased levels of collagen synthesis and degradation in the malignant group.
0.32 0.84
DISCUSSION
The significance of biochemical and clinical parameters as predictors of survival Variable PSA % Ca benign group collagen % Ca benign/malignant group collagen Clinical stage Gleason score
p Value (univariate Cox regression)
in 53 and 37 cases, respectively. Only 33 patients had undergone a bone scan. There was no relationship of the collagen content in each group and the clinical stage of cancer. However, the ratio of cancer malignant-to-benign collagen content may have been greater in clinically localized disease but the difference did not reach clinical significance at the 95% confidence level (p ⫽ 0.087). Univariate Cox regression analysis showed that absolute PSA was the strongest predictor of survival but the percent collagen in the cancer benign group and the percent collagen ratio of the 2 groups were stronger predictors of survival than the Gleason sum or clinical stage (see table). Gleason sum reproducibility was tested by the second histopathologist, who down graded a single sample from a high to a moderate score and up graded a single moderate score to a high score. This degree of Gleason sum re-grading in moderate and high grade tumors was reasonable and may have been expected. Reassuringly there was no down or up grading from low to high grade disease. Multiple sectioning and staining of cancer benign group tissue showed no deep seated cancerous foci. Changes in collagen metabolism in prostatic biopsies from malignant and benign prostate glands. From the 45 consecutive patients who underwent transrectal ultrasound and biopsy we obtained a total of 80 samples for analysis. No significant differences in mature or glycated collagen crosslinks were observed in the groups. Gelatin zymography revealed increased expression and activation of matrix metalloproteinase 2 but not matrix metalloproteinase 9 in biopsies from malignant versus benign biopsies, although neither expression nor activation achieved a significant level (fig. 2, A). Conversely casein zymography demonstrated a decrease in all serine proteinases in the malignant group, which attained significance for plasminogen. An increase in carboxy-terminal propeptide of type I collagen in the malignant group indicated new onset type I collagen synthesis, which was confirmed by a significant increase in the level of collagen intermediate cross-link hydroxylysinoketonorleucine (fig. 2, B). As matrix metal-
In our study the metabolism of collagen was greatly increased in biopsies of prostate glands containing cancer. In these biopsies there were increased levels of collagen type I propeptides and collagen intermediate cross-links as markers of synthesis and increased degradation based on matrix metalloproteinase levels. In addition, histological sections revealed increased collagen content in prostatic tissue distant to the cancer focus. This differential increased with tumor grade and cannot be explained by a relative increase in cellularity and loss of extracellular matrix in the cancer focus alone since the increased level of collagen remote from the cancer focus was also significantly greater than the collagen concentration of the control tissue. These data imply upregulation of collagen synthesis throughout a cancerous prostate but with up-regulation of collagen degradation at the cancerous foci. The decrease in collagen content of the cancer extracellular matrix may lead to decreased cellular nutritional support and, therefore, explain the increased number of necrotic foci in high grade tumors. In addition, high grade prostate cancer is associated with the loss of structural strength, increased cellularity, decreased extracellular matrix and increased friability, which confirm our finding of collagen depletion. In situ hybridization studies have indicated evidence of increased type I collagen synthesis by fibroblasts in benign tissue near the invasive edge of breast cancer and increasing synthesis with increasing cancer grade.17, 18 This finding would support our data on increased collagen content in prostatic biopsies remote from cancer foci and it contrasts with reports of a desmoplastic response in which excessive deposition of collagen occurs in the stroma of colorectal cancer, for example.19 Preferential collagen deposition at a tumor edge may develop before invasion and prevent a noninvasive tumor from becoming invasive. Encapsulation of an inflammatory site or foreign object by fibrous collagen is a common feature in the body and a similar response may be elicited due to the cancer focus. Development of a collagen capsule around a noninvasive tumor may represent a host defense mechanism in prostate cancer. Barsky and Gopalakrishna reported that the size and number of pulmonary metastases were increased in mice
FIG. 2. Prostatic tissue from benign and malignant prostate glands. A, matrix metalloproteinase (MMP2) expression and activation. OD, optical density. B, carboxy-terminal propeptide of type I collagen level.
CHANGES IN COLLAGEN METABOLISM IN PROSTATE CANCER
injected with melanoma cells when collagen synthesis was inhibited.20 If a tumor is invasive, its behavior and, therefore, prognosis depend on tumor (grade, stage and so forth) and host factors. When the host mounts a vigorous collagen response to repair damage or form a tissue barrier to impede invasion, the clinical outcome may be altered. In our series there appeared to be an increase in host collagen synthesis, which was greater according to the aggressiveness of the tumor (higher Gleason sum). We know that high Gleason sum tumors are associated with a poorer prognosis and any host collagen response cannot alter this fact. We do not know whether the collagen increases as a result of or before invasion as an attempt to prevent invasion. However, it is important that evidence exists to indicate such a graduated response by the host to prostate cancer. Our data show that collagen content of the prostatic tissue remote from cancer is a stronger predictor of survival than Gleason score or clinical stage. CONCLUSIONS
Novel markers, such as carboxy-terminal propeptide of type I collagen and hydroxylysinoketonorleucine, obtained from biopsies identical to those used for histology may be used to assess the strength of this host collagen response. In turn this response may be an important factor in assessing the need for radical treatment or determining the likely outcome of prostate cancer. REFERENCES
1. Liotta, L. A., Steeg, P. S. and Stetler-Stevenson, W. G.: Cancer metastasis and angiogenesis: an imbalance of positive and negative regulation. Cell, 64: 327, 1991 2. Paul, R., Ewing, C. M., Jarrard, D. F. et al: The cadherin cell-cell adhesion pathway in prostate cancer progression. Br J Urol, Suppl., 79: 37, 1997 3. Paulsson, M.: Basement membrane proteins: structure, assembly and cellular interactions. Crit Rev Biochem Mol Biol, 27: 93, 1992 4. Parson, S. L., Watson, S. A., Brown, P. D. et al: Matrix metalloproteinases. Br J Surg, 84: 160, 1997 5. Okada, Y., Gonoji, Y., Naka, K. et al: Matrix metalloproteinase 9 (92-kDa gelatinase/type IV collagenase) from HT 1080 human fibrosarcoma cells. Purification and activation of the precursor and enzymic properties. J Biol Chem, 267: 27712, 1992 6. Andreasen, P. A., Kjoller, L., Christensen, L. et al: The urokinase-type plasminogen activator system in cancer metastasis: a review. Int J Cancer, 72: 1, 1997
1701
7. Murphy, G. and Willenbrock, F.: Tissue inhibitors of matrix metalloendopeptidases. Meth Enzymol, 248: 496, 1995 8. Kruithof, E. K. O.: Plasminogen activator inhibitors: a review. Enzyme, 40: 113, 1988 9. Jouanneau, L., Gavrilovic, J., Caruelle, D. et al: Secreted or nonsecreted forms of acidic fibroblast growth factor produced by transfected epithelial cells influence cell morphology, motility and invasive potential. Proc Natl Acad Sci USA, 88: 2893, 1991 10. Aznavoorian, S., Stracke, M. L., Krutzsch, H. et al: Signal transduction for chemotaxis and haptotaxis by matrix macromolecules in tumor cells. J Cell Biol, 110: 1427, 1990 11. Murphy, B. C., Pienta, K. J. and Coffey, D. S.: Effects of extracellular matrix components and dihydrotestosterone on the structure and function of human prostate cancer cells. Prostate, 20: 29, 1992 12. Passaniti, A., Isaacs, J. T. and Haney, J. A.: Stimulation of human prostatic carcinoma tumour growth in athymic mice and control of migration in culture by extracellular matrix. Int J Cancer, 51: 318, 1992 13. Fridman, R., Kibbey, M. C., Royce, L. S. et al: Enhanced tumor growth of both primary and established human and murine tumour cells in athymic mice after coinjection with Matrigel. J Natl Cancer Inst, 83: 769, 1991 14. Royce, P. M. and Steinmann, B. U.: Connective Tissue and its Heritable Disorders: Molecular, Genetic and Medical Aspects. New York: Wiley-Liss, 1993 15. Bannister, D. W. and Burns, A. B.: Adaptation of the Bergman and Loxley technique for hydroxyproline determination to the autoanalyzer and its use in determining plasma hydroxyproline in the domestic fowl. Analyst, 95: 596, 1970 16. Sims, T. J., Avery, N. C. and Bailey, A. J.: Quantitative determination of collagen crosslinks. In: Methods in Molecular Biology: Extracellular Matrix Protocols. Edited by C. H. Streuli and M. E. Grant. Totawa, New Jersey: Humana Press, p. 11, 2000 17. Gilles, C., Polette, M., Seiki, M. et al: Implication of collagen type I-induced membrane-type I 1-matrix metalloproteinase expression and matrix metalloproteinase-2 activation in the metastatic progression of breast carcinoma. Lab Invest, 76: 651, 1997 18. Wapnir, I. L., Southard, H., Chen, G. et al: Collagen gene expression in the neomatrix of carcinoma of the breast. Invasion Metastasis, 16: 308, 1996 19. Hewitt, R. E., Powe, D. G., Carter, G. I. et al: Desmoplasia and its relevance to colorectal tumour invasion. Int J Cancer, 53: 62, 1993 20. Barsky, S. H. and Gopalakrishna, R.: Increased invasion and spontaneous metastasis of BL6 melanoma with inhibition of the desmoplastic response in C57/BL6 mice. Cancer Res, 47: 1663, 1987
Vol. 166, 1698 –1701, November 2001 Printed in U.S.A.
CHANGES IN COLLAGEN METABOLISM IN PROSTATE CANCER: A HOST RESPONSE THAT MAY ALTER PROGRESSION N. BURNS-COX, N. C. AVERY, J. C. GINGELL
AND
A. J. BAILEY
From the Bristol Urological Institute, Southmead Hospital and Collagen Research Group, University of Bristol, Bristol, United Kingdom
ABSTRACT
Purpose: The interaction of cancer cells and the extracellular matrix is essential for cancer progression. However, little is known about the influence of cancer on the metabolism of collagen, which is the major constituent of the extracellular matrix. We studied changes in collagen metabolism in prostate cancer with increasing Gleason score and correlated them with clinical parameters and patient survival. Materials and Methods: Collagen content and clinical parameters in 3 types of prostatic tissue were compared, including the foci of prostatic cancer, unaffected tissue from the same cancerous prostates and prostatic tissue from patients with no evidence of cancer. In addition, to assess collagen metabolism tissue obtained prospectively from 45 patients undergoing prostatic biopsy was assessed for collagen content, the type and extent of collagen cross-linking, serine proteinase, matrix metalloproteinase and the C-terminal propeptide of type I collagen. Results: With increasing Gleason sum of the cancer collagen content at the focus decreased, while that of surrounding unaffected tissue from the same prostate increased to levels significantly above that from controls with no cancer. Markers of collagen synthesis in the prostate biopsy material were significantly increased in the presence of prostate cancer. Conclusions: In prostate cancer there are changes in collagen metabolism not only in the cancer focus but also in nearby histologically benign prostatic tissue. These observed changes are related to the Gleason score of the tumor and may represent a host response. Collagen content in the surrounding unaffected tissue may be a predictor of patient survival. KEY WORDS: prostate, prostatic neoplasms, collagen, extracellular matrix, disease progression
The hallmark of cancer is its ability to invade locally adjacent host structures and metastasize. Liotta et al described 3 basic steps essential to this process, namely adhesion, degradation and migration.1 Each step involves interactions of cancer cells with components of the host basement membrane and extracellular matrix, primarily collagen. Cancer cells adhere to components of the extracellular matrix through the specific interaction of cell adhesion integrins2 or to host basement membrane through laminin.3 Degradation occurs after the secretion of proteolytic enzymes capable of degrading specific components of basement membrane or extracellular matrix. The 3 major groups of such proteinases considered important for cancer cell invasion are the cysteine proteinases, the serine proteinases and the matrix metalloproteinases. A total of 14 matrix metalloproteinases have been identified, of which each has differing substrate specificities for components of the extracellular matrix or basement membrane.4 There are also complex interactions among the proteinase groups, for example the serine proteinase plasmin can activate matrix metalloproteinases.5 In addition, the groups of proteinases have specific activators, for example urokinasetype plasminogen activator,6 and inhibitors, such as tissue inhibitors of matrix metalloproteinases7 and plasmin activator inhibitor.6, 8 Although the system is complex, it tightly regulates the processes involved in normal healing and remodeling but it is exploited by cancer cells during invasion. In normal tissue cell mobility is rare after organogenesis and cellular differentiation. However, cellular migration is essential for invasion of the host extracellular matrix and subsequent metastasis by cancer. In addition to growth factors9 and tumor secreted factors,1 components of the extraAccepted for publication June 22, 2001. Supported by the Andrology Research Fund.
cellular matrix can promote cancer cell motility, for example laminin, fibronectin and collagen type IV.10 In the prostate cancer cell line LNCaP Murphy et al observed that the growth rate, prostate specific antigen (PSA) production and sensitivity to androgen depended on collagen type I in the culture medium.11 Therefore, it seems that the clinical outcome of cancer is profoundly influenced by components of the host extracellular matrix. Cancer cells injected into athymic mice do not develop into tumors unless provided with constituents of the extracellular matrix.12, 13 Collagen is the most abundant protein in mammals, and a major constituent of extracellular matrix and basement membranes.14 It also has an important role in the response of the body to chronic injury. The major characteristic of such repair is the deposition of scar tissue, which is predominantly composed of fibrous collagen. Because cancer invasion is a chronic and often progressive condition, it seems likely that such healing may form part of the host response with associated collagen deposition. If the host response is vigorous, it may impede or even prevent cancer cell invasion of healthy tissue but little is known about the influence of cancer on collagen metabolism. We report changes in collagen metabolism in prostate cancer, which is a host response that may alter progression. MATERIALS AND METHODS
We analyzed the collagen content of unstained prostate sections obtained for histological testing from patients with prostate cancer and determined the levels of collagen metabolism from fresh samples obtained separately at biopsy. Quantification of collagen in benign and malignant prostatic tissue sections. A computer search identified 101 patients with prostate cancer who had undergone transurethral
1698
CHANGES IN COLLAGEN METABOLISM IN PROSTATE CANCER
prostatic resection at The Bristol Urological Institute between 1994 and 1997. We obtained sections of resected tissue before hematoxylin and eosin staining was done. In each case we identified a portion of tissue containing prostate cancer and a portion without evidence of prostate cancer. As a control group, we also obtained similar tissue samples from another 22 patients with benign prostatic hyperplasia (BPH) alone and no evidence of prostate cancer. A histopathologist with experience in prostate cancer performed tissue identification and categorization as well as repeat Gleason scoring. To assess patient survival additional data on age, clinical stage, prostate specific antigen (PSA) and bone scans were obtained from the medical records. To test the reproducibility of the repeat Gleason scoring 10 specimens were randomly selected and evaluated by another experienced histopathologist. To determine whether small cancerous foci may have been concealed within the noncancerous portion of the tissue from patients with prostate cancer we performed multiple sectioning and staining in 10 tissue portions to detect evidence of deep seated foci. Determining collagen content. Each portion of tissue was de-waxed using repeat changes of xylene and then lyophilized and weighed before acid hydrolysis in 6N hydrochloric acid at 112C for 24 hours. Hydroxyproline content of each hydrolysate was determined using the method of Bannister and Burns.15 That is, since hydroxyproline represents approximately 14% of fibrous collagen dry weight, multiplication by 7.14 yields the percent of collagen of each portion of tissue. Collagen metabolism in prostatic biopsies from malignant and benign prostates. After obtaining informed consent biopsy was prospectively obtained from the posterolateral aspect of each lobe of the prostate in 45 consecutive patients referred for transrectal ultrasound and biopsy due to the suspicion of prostate cancer based on increased PSA and/or clinically suspicious digital rectal examination. This biopsy was performed in addition to routine sextant biopsies, which were sent for histological examination. It was snap frozen in liquid nitrogen and stored at ⫺80C until analyzed. Each biopsy sample was washed in phosphate buffered saline, blot dried and unequally divided in 2 portions. The larger portion was used to determine the amount and type of intermediate and mature collagen cross-links after sodium borohydride reduction16 and acid hydrolysis, the amount of advanced glycation end product and total collagen content, as
1699
described. The smaller portion was agitated overnight at 4C in 20 mM. triethanolamine containing 0.1% polyoxyethylene lauryl ether. The extract was used to determine intracellular proteinase and the level of carboxy-terminal propeptide of type I collagen. Extracts were characterized by reference to standards after gelatin or casein substrate polyacrylamide gel electrophoresis, and quantified by computer linked transmittance densitometry and by a commercially available immunoassay. Statistical analysis. The paired and independent samples t tests were done as appropriate. Levene’s test was used to determine equal or unequal variance. We performed linear regression and calculated the Pearson correlation coefficient to determine the relationships of 2 continuous variables. Univariate Cox regression analysis was also done to determine whether the clinical or biochemical parameters measured predicted patients survival.
RESULTS
We analyzed 3 types of prostatic tissue. In the 101 patients in the cancer malignant group a prostatic tissue sample was known to contain a cancer focus. In the 90 patients in the cancer benign group a paired prostatic tissue sample from a malignant prostate had no evidence of cancer on histological examination. In the 22 patients in the BPH control group with no evidence of prostate cancer the prostatic tissue sample had the appearance of BPH only. Correlating collagen content with Gleason sum. Figure 1 shows the change in mean collagen content in the 3 groups compared with the Gleason sum. In tissue with a Gleason sum of less than 5 there was no significant difference in collagen content in the groups. When the Gleason sum was 5 and greater cancer malignant group tissue collagen content was always lower than in cancer benign group tissue and overall there was a significant difference (paired t test p ⬍ 0.001). Over the same Gleason sum range the collagen content of cancer benign group tissue was also higher than that of control tissue but it was not statistically significant until Gleason sum 8 and greater. Clinical data were available for 90 of the 101 patients 55 to 94 years old at diagnosis (mean age 74) with prostate cancer. Mean followup after diagnosis was 41 months. Digital rectal examination demonstrated stages T1/T2 and T3/T4 disease
FIG. 1. Collagen content of cancer benign (CBG), cancer malignant (CMG) and BPH prostate tissue
1700
CHANGES IN COLLAGEN METABOLISM IN PROSTATE CANCER
0.03 0.065 0.08
loproteinase 2 expression and activation increased in cancer malignant group biopsies, so did carboxy-terminal propeptide of type I collagen and hydroxylysinoketonorleucine (R ⫽ 0.04, p ⫽ 0.008 and R ⫽ 0.58, p ⬍ 0.001, respectively). These data demonstrated coordinated increased levels of collagen synthesis and degradation in the malignant group.
0.32 0.84
DISCUSSION
The significance of biochemical and clinical parameters as predictors of survival Variable PSA % Ca benign group collagen % Ca benign/malignant group collagen Clinical stage Gleason score
p Value (univariate Cox regression)
in 53 and 37 cases, respectively. Only 33 patients had undergone a bone scan. There was no relationship of the collagen content in each group and the clinical stage of cancer. However, the ratio of cancer malignant-to-benign collagen content may have been greater in clinically localized disease but the difference did not reach clinical significance at the 95% confidence level (p ⫽ 0.087). Univariate Cox regression analysis showed that absolute PSA was the strongest predictor of survival but the percent collagen in the cancer benign group and the percent collagen ratio of the 2 groups were stronger predictors of survival than the Gleason sum or clinical stage (see table). Gleason sum reproducibility was tested by the second histopathologist, who down graded a single sample from a high to a moderate score and up graded a single moderate score to a high score. This degree of Gleason sum re-grading in moderate and high grade tumors was reasonable and may have been expected. Reassuringly there was no down or up grading from low to high grade disease. Multiple sectioning and staining of cancer benign group tissue showed no deep seated cancerous foci. Changes in collagen metabolism in prostatic biopsies from malignant and benign prostate glands. From the 45 consecutive patients who underwent transrectal ultrasound and biopsy we obtained a total of 80 samples for analysis. No significant differences in mature or glycated collagen crosslinks were observed in the groups. Gelatin zymography revealed increased expression and activation of matrix metalloproteinase 2 but not matrix metalloproteinase 9 in biopsies from malignant versus benign biopsies, although neither expression nor activation achieved a significant level (fig. 2, A). Conversely casein zymography demonstrated a decrease in all serine proteinases in the malignant group, which attained significance for plasminogen. An increase in carboxy-terminal propeptide of type I collagen in the malignant group indicated new onset type I collagen synthesis, which was confirmed by a significant increase in the level of collagen intermediate cross-link hydroxylysinoketonorleucine (fig. 2, B). As matrix metal-
In our study the metabolism of collagen was greatly increased in biopsies of prostate glands containing cancer. In these biopsies there were increased levels of collagen type I propeptides and collagen intermediate cross-links as markers of synthesis and increased degradation based on matrix metalloproteinase levels. In addition, histological sections revealed increased collagen content in prostatic tissue distant to the cancer focus. This differential increased with tumor grade and cannot be explained by a relative increase in cellularity and loss of extracellular matrix in the cancer focus alone since the increased level of collagen remote from the cancer focus was also significantly greater than the collagen concentration of the control tissue. These data imply upregulation of collagen synthesis throughout a cancerous prostate but with up-regulation of collagen degradation at the cancerous foci. The decrease in collagen content of the cancer extracellular matrix may lead to decreased cellular nutritional support and, therefore, explain the increased number of necrotic foci in high grade tumors. In addition, high grade prostate cancer is associated with the loss of structural strength, increased cellularity, decreased extracellular matrix and increased friability, which confirm our finding of collagen depletion. In situ hybridization studies have indicated evidence of increased type I collagen synthesis by fibroblasts in benign tissue near the invasive edge of breast cancer and increasing synthesis with increasing cancer grade.17, 18 This finding would support our data on increased collagen content in prostatic biopsies remote from cancer foci and it contrasts with reports of a desmoplastic response in which excessive deposition of collagen occurs in the stroma of colorectal cancer, for example.19 Preferential collagen deposition at a tumor edge may develop before invasion and prevent a noninvasive tumor from becoming invasive. Encapsulation of an inflammatory site or foreign object by fibrous collagen is a common feature in the body and a similar response may be elicited due to the cancer focus. Development of a collagen capsule around a noninvasive tumor may represent a host defense mechanism in prostate cancer. Barsky and Gopalakrishna reported that the size and number of pulmonary metastases were increased in mice
FIG. 2. Prostatic tissue from benign and malignant prostate glands. A, matrix metalloproteinase (MMP2) expression and activation. OD, optical density. B, carboxy-terminal propeptide of type I collagen level.
CHANGES IN COLLAGEN METABOLISM IN PROSTATE CANCER
injected with melanoma cells when collagen synthesis was inhibited.20 If a tumor is invasive, its behavior and, therefore, prognosis depend on tumor (grade, stage and so forth) and host factors. When the host mounts a vigorous collagen response to repair damage or form a tissue barrier to impede invasion, the clinical outcome may be altered. In our series there appeared to be an increase in host collagen synthesis, which was greater according to the aggressiveness of the tumor (higher Gleason sum). We know that high Gleason sum tumors are associated with a poorer prognosis and any host collagen response cannot alter this fact. We do not know whether the collagen increases as a result of or before invasion as an attempt to prevent invasion. However, it is important that evidence exists to indicate such a graduated response by the host to prostate cancer. Our data show that collagen content of the prostatic tissue remote from cancer is a stronger predictor of survival than Gleason score or clinical stage. CONCLUSIONS
Novel markers, such as carboxy-terminal propeptide of type I collagen and hydroxylysinoketonorleucine, obtained from biopsies identical to those used for histology may be used to assess the strength of this host collagen response. In turn this response may be an important factor in assessing the need for radical treatment or determining the likely outcome of prostate cancer. REFERENCES
1. Liotta, L. A., Steeg, P. S. and Stetler-Stevenson, W. G.: Cancer metastasis and angiogenesis: an imbalance of positive and negative regulation. Cell, 64: 327, 1991 2. Paul, R., Ewing, C. M., Jarrard, D. F. et al: The cadherin cell-cell adhesion pathway in prostate cancer progression. Br J Urol, Suppl., 79: 37, 1997 3. Paulsson, M.: Basement membrane proteins: structure, assembly and cellular interactions. Crit Rev Biochem Mol Biol, 27: 93, 1992 4. Parson, S. L., Watson, S. A., Brown, P. D. et al: Matrix metalloproteinases. Br J Surg, 84: 160, 1997 5. Okada, Y., Gonoji, Y., Naka, K. et al: Matrix metalloproteinase 9 (92-kDa gelatinase/type IV collagenase) from HT 1080 human fibrosarcoma cells. Purification and activation of the precursor and enzymic properties. J Biol Chem, 267: 27712, 1992 6. Andreasen, P. A., Kjoller, L., Christensen, L. et al: The urokinase-type plasminogen activator system in cancer metastasis: a review. Int J Cancer, 72: 1, 1997
1701
7. Murphy, G. and Willenbrock, F.: Tissue inhibitors of matrix metalloendopeptidases. Meth Enzymol, 248: 496, 1995 8. Kruithof, E. K. O.: Plasminogen activator inhibitors: a review. Enzyme, 40: 113, 1988 9. Jouanneau, L., Gavrilovic, J., Caruelle, D. et al: Secreted or nonsecreted forms of acidic fibroblast growth factor produced by transfected epithelial cells influence cell morphology, motility and invasive potential. Proc Natl Acad Sci USA, 88: 2893, 1991 10. Aznavoorian, S., Stracke, M. L., Krutzsch, H. et al: Signal transduction for chemotaxis and haptotaxis by matrix macromolecules in tumor cells. J Cell Biol, 110: 1427, 1990 11. Murphy, B. C., Pienta, K. J. and Coffey, D. S.: Effects of extracellular matrix components and dihydrotestosterone on the structure and function of human prostate cancer cells. Prostate, 20: 29, 1992 12. Passaniti, A., Isaacs, J. T. and Haney, J. A.: Stimulation of human prostatic carcinoma tumour growth in athymic mice and control of migration in culture by extracellular matrix. Int J Cancer, 51: 318, 1992 13. Fridman, R., Kibbey, M. C., Royce, L. S. et al: Enhanced tumor growth of both primary and established human and murine tumour cells in athymic mice after coinjection with Matrigel. J Natl Cancer Inst, 83: 769, 1991 14. Royce, P. M. and Steinmann, B. U.: Connective Tissue and its Heritable Disorders: Molecular, Genetic and Medical Aspects. New York: Wiley-Liss, 1993 15. Bannister, D. W. and Burns, A. B.: Adaptation of the Bergman and Loxley technique for hydroxyproline determination to the autoanalyzer and its use in determining plasma hydroxyproline in the domestic fowl. Analyst, 95: 596, 1970 16. Sims, T. J., Avery, N. C. and Bailey, A. J.: Quantitative determination of collagen crosslinks. In: Methods in Molecular Biology: Extracellular Matrix Protocols. Edited by C. H. Streuli and M. E. Grant. Totawa, New Jersey: Humana Press, p. 11, 2000 17. Gilles, C., Polette, M., Seiki, M. et al: Implication of collagen type I-induced membrane-type I 1-matrix metalloproteinase expression and matrix metalloproteinase-2 activation in the metastatic progression of breast carcinoma. Lab Invest, 76: 651, 1997 18. Wapnir, I. L., Southard, H., Chen, G. et al: Collagen gene expression in the neomatrix of carcinoma of the breast. Invasion Metastasis, 16: 308, 1996 19. Hewitt, R. E., Powe, D. G., Carter, G. I. et al: Desmoplasia and its relevance to colorectal tumour invasion. Int J Cancer, 53: 62, 1993 20. Barsky, S. H. and Gopalakrishna, R.: Increased invasion and spontaneous metastasis of BL6 melanoma with inhibition of the desmoplastic response in C57/BL6 mice. Cancer Res, 47: 1663, 1987