Macrophage migration inhibitory factor (MIF) gene ... - Nature

Genes and Immunity (2007) 8, 646–652 & 2007 Nature Publishing Group All rights reserved 1466-4879/07 $30.00 www.nature.com/gene

ORIGINAL ARTICLE

Macrophage migration inhibitory factor (MIF) gene polymorphisms are associated with increased prostate cancer incidence KL Meyer-Siegler1,2, PL Vera1,2, KA Iczkowski3, C Bifulco4, A Lee5, PK Gregersen5, L Leng4 and R Bucala4 Research and Development 151, Bay Pines VA Healthcare System, Bay Pines, FL, USA; 2Departments of Surgery and Molecular Medicine, University of South Florida, Tampa, FL, USA; 3Department of Pathology, University of Colorado Health Science Center, Aurora, CO, USA; 4Internal Medicine Rheumatology, Yale University School of Medicine, New Haven, CT, USA and 5Robert S Boas Center for Genomics and Human Genetics, The Feinstein Institute for Medical Research, North Shore LIJ Health System, Manhasset, NY, USA 1

Recurrent or persistent inflammation has emerged as an important factor in cancer development. Overexpression of macrophage migration inhibitory factor (MIF), an upstream regulator of innate immunity with pleiotropic effects on cell proliferation, has been implicated in prostate cancer (CaP). Two polymorphisms in the promoter of the MIF gene (173G to C transition and seven copies of the 794 CATT repeat) are associated with increased MIF expression in vivo and poor prognosis in autoimmune diseases. We conducted a retrospective analysis of 131 CaP patients and 128 controls from a group of Veterans’ Administration patients undergoing routine prostate-specific antigen screening. Patients with CaP were enrolled regardless of treatment. Inclusion criteria for the control group were absence of documented diagnosis of cancer and/or chronic inflammation within patient computerized records. Logistic regression demonstrated a significant association between CaP and the 173G/C, the 173C/C and the 794 7-CATT MIF polymorphisms (Po0.001). Patients with the 794 7-CATT allele had an increased risk of CaP recurrence at 5 years. Individuals with 173G/C, 173C/C and 794 7-CATT MIF genotypes have an increased incidence of CaP and these genotypes may serve as an independent marker for cancer recurrence. Genes and Immunity (2007) 8, 646–652; doi:10.1038/sj.gene.6364427; published online 30 August 2007 Keywords: cytokines; genetic polymorphisms; inflammation; macrophage migration inhibitory factor

Introduction Prostate cancer (CaP) is the most common malignant neoplasm to occur in men and the second leading cause of cancer mortality in the United States.1 The biological aggressiveness of CaP is variable, with a spectrum that ranges from indolent cancer confined to the prostate, to cases with rapid, extra-prostatic extension and distant metastasis.2,3 The underlying mechanisms responsible for the clinical progression of CaP are poorly understood, but may involve changes in expression and response to cytokine and growth factor receptors, which can be modulated by inflammatory signals.4,5 There is emerging evidence that recurrent or persistent inflammation plays a role in cancer development.6–8 Inflammation-mediated events, such as production of reactive oxygen species and activation of growth factors and signal-transduction events that promote cell survival Correspondence: Dr KL Meyer-Siegler, Bay Pines VA Healthcare System, R&D 151, Building 23, Room 226, 10000 Bay Pines Boulevard, Bay Pines, FL 33744, USA. E-mail: [email protected] Received 18 June 2007; revised 2 August 2007; accepted 3 August 2007; published online 30 August 2007

and proliferation, are considered to be components of risk and a factor in cancer progression.6 This model is supported by the epidemiological relationship between chronic or recurrent local inflammation and cancer development.6 Examples include the association between Barrett’s esophagus and esophageal cancer, Helicobacter gastritis and gastric cancer, chronic pancreatitis and pancreatic cancer, cirrhosis and hepatic cancer, and inflammatory bowel disease and colon carcinoma.6 Epidemiological studies that examined expression of pro- and anti-inflammatory factors suggested an association between chronic intra-prostatic inflammation and CaP9,10 and documented an association between longterm nonsteroidal anti-inflammatory drugs use and reduced CaP risk.11,12 Inflammation is frequently present in prostate biopsies, radical prostatectomy specimens and tissue resected for the treatment of benign prostatic hyperplasia, and it is often a feature of atrophic foci characterized by an increased proliferative index.9 If chronic inflammation is an important factor in CaP development then logical candidate genetic determinants of CaP risk would be allelic variants of genes associated with inflammation. Variants of genes involved in innate and acquired immunity appear to play an important part in determining inherited CaP risk.13

MIF polymorphisms in prostate cancer KL Meyer-Siegler et al

Macrophage migration inhibitory factor (MIF) is an upstream regulator of innate immunity implicated in tumor growth based on in vitro and murine models of tumorigenesis.14 The mechanisms underlying MIF’s contribution to the malignant phenotype appear diverse, and include promotion of tumor progression by neoangiogenesis,15 as well as direct enhancement of growth factor-stimulated cell cycle progression by mitogenactivated protein (MAP) kinase activation.16 MIF also inhibits the pro-apoptotic and growth inhibitory function of the p53 tumor suppressor in neoplastic cell lines.17 Breast,18–20 colon20–22 and lung-derived tumors23,24 all contain significantly higher levels of MIF mRNA and/or protein than their noncancerous, normal tissue counterparts. Studies also reported MIF expression closely correlated with tumor aggressiveness and metastatic potential.21–23 Migration inhibitory factor is produced by prostate epithelial cells in culture, and its secretion is upregulated in metastatic CaP.25 Associations between high-serum MIF levels or increased intratumor MIF expression, and CaP biologic aggressiveness are reported.25,26 Additionally, there is evidence for a relationship between increased MIF expression and the acquisition of drug resistance by hormone-independent CaP cells.27 Two functional promoter polymorphisms in the 50 flanking region of the MIF gene are strongly associated with protein production. The first polymorphism comprises a tetra-nucleotide (CATT) repeat at position 794 of the gene that is present between 5 and 8 times within a Pit-1 transcription factor binding site.28 Functional gene reporter studies in defined cell lines (for example, monocytes, epithelial cells) demonstrate that MIF gene expression increases proportionally with the number of CATT repeats in the promoter.28 A second single nucleotide polymorphism (173G/C) influences MIF gene expression, either by linkage disequilibrium with the CATT site, or by an independent action involving a transcriptional activator site.29 There is accruing evidence for the clinical relevance of these polymorphisms as high-expression MIF alleles may influence susceptibility and severity of different autoimmune inflammatory disorders.28–30 We hypothesized that MIF polymorphisms, which occur commonly in the population,31 are also associated with incidence and outcome of CaP.

Methods Approval for this study was obtained from the Bay Pines VA Healthcare System and Yale Institutional Review Boards. MIF genotyping A retrospective study of MIF genotype in patients undergoing routine prostate-specific antigen (PSA) screening was performed using random serum samples obtained directly from the Bay Pines Veterans Administration Health Care System (VAHCS) clinical laboratory. At Bay Pines VAHCS all patients are subject to routine PSA screening on an annual basis. The serums were collected the day that they were drawn and stored at 80 1C After the serum sample was obtained individual patient computerized records were searched for all relevant information (PSA, cancer and inflammatory disease diagnosis, Gleason sum, age),

which was used to determine if the patient met the appropriate inclusion criteria. Inclusion criteria for the CaP group (n ¼ 131) were diagnosis of CaP with no other concurrent cancer or inflammatory disease as documented within computerized records, PSA amounts from the randomly obtained serum sample and availability of serum samples for MIF protein, and genotype determinations. Gleason evaluation of prostate biopsies was documented within the computerized records of 77 of the 131 CaP patients. Fiveyear follow-up PSA amounts were available for 42 of the 131 patients. Inclusion criteria for the control group (n ¼ 128) was absence of cancer and/or any inflammatory disease as documented within the individual patient computerized record, PSA amounts from the randomly obtained serum sample and availability of serum samples for MIF protein, and genotype determinations. Controls were enrolled regardless of serum PSA amounts. Genomic DNA was isolated from serum using DNAzol (Invitrogen, Carlsbad, CA, USA) then amplified by multiple displacement amplification (GenomiPhi, GE Healthcare, Piscataway, NJ, USA).31 DNA quality was assessed by agarose gel electrophoresis. The yield of amplified genomic DNA was between 10 and 15 mg in a final volume of 50 ml. MIF genotype was determined as described previously.28–31 Briefly, PCR for the CATT repeat polymorphism was performed using the forward primer (50 -TGCAGGAACCAATACCCATAGG-30 ) and a TET fluorescent reverse primer (TET lab50 -AATGG TAAACTCGGGGAC-30 ) or G/C single nucleotide polymorphism (SNP) regions.31 The resulting PCR reaction products were denatured and then resolved using an ABI 310 Genetic Analyzer (PerkinElmer-Applied Biosystems, CA, USA). Serum MIF was determined by enzyme-linked immunosorbent assay as described (R&D Systems, Minneapolis, MN, USA).26

647

Statistical analyses Differences in population characteristics (age, serum MIF and PSA) were analyzed using the Student’s t-test or Mann–Whitney U-test, where appropriate. Differences in allele frequencies were analyzed by w2-test. Backwards stepwise age-adjusted logistic regression (using SPSS v.14) examined the incidence of CaP associated with MIF polymorphisms and whether these polymorphisms were associated with increased Gleason sum. We examined the effects of single (SNP, CATT) polymorphisms and also of having a combination of these two as there may be a synergistic effect between the SNP and CATT repeats.32 For the analysis of the combination of MIF polymorphisms on the incidence of CaP and Gleason sum we examined the following three categories: (1) G/G and 5/5, 5/6, 6/6; (2) G/C, C/C or 5/7, 6/7, 7/7; and (3) C/C and 5/7, 6/7, 7/7. We performed a two-way analysis of variance (ANOVA) to determine the effect of SNP and CATT repeats on MIF serum level. The patients were divided into two categories of SNP: þ C (G/C or C/C) and C (G/G) and two categories of CATT: þ 7 (5/7, 6/7, 7/7) and 7 (5/5, 5/6, 6/6). We also performed internal cross-validation to estimate the predictive utility of the MIF polymorphisms in determining CaP risk. Using SPSS we randomly divided the data set into a 70% for training a model (using binary logistic regression as above) and 30% for a test of the Genes and Immunity


648

model to determine specificity (true-positive) and sensitivity (true-negative) estimates. A total of 10 random estimates were averaged to obtain an error of the estimate. In addition, differences in the ability of either polymorphism alone or the combination (SNP/CATT) on specificity and sensitivity were assessed using ANOVA followed by post hoc (Newman–Keuls) tests. Haplotype frequencies were determined using Phase v2.1.1 (www.stat.washington.edu/stephens/software. html). P-values p0.05 were considered statistically significant in all analyses.

Results Study population The demographic characteristics of the CaP patients and the control group are shown in Table 1. The median age of the CaP group (71, range 43–87 years) was higher (Po0.0001, Mann–Whitney) than the controls (65, range 36–87 years). The ethnic background for the entire patient group studied was 95.3% Caucasian, 2.7% African Americans and 2% other. There was no difference in mean serum PSA at the time of sampling between CaP (median 1.5 ng ml1, range 0.1–24.7 ng ml1) and control groups (median 1.7 ng ml1, range 0.2–43 ng ml1; Mann–Whitney, P ¼ 0.16). As documented previously,26 median serum MIF levels in the CaP group (9.8, interquartile range (IQR) 6.6 ng ml1) were higher (Po0.0001, Mann–Whitney) when compared to control group (2.9, IQR 2.5 ng ml1). Gleason sum evaluation of prostate biopsies was available in the computerized records of 77 of the 131 CaP patients. MIF allele and haplotype frequencies in prostate cancer patients The allelic distributions in either CaP or control groups for MIF 173C polymorphism and/or the 794 MIF

Table 1 Demographic and clinical data of CaP subjects and controls

Age (year) Mean7s.d. Range Ethnicity (%) Caucasian African American Hispanic Asian/Pacific Islander

CaP

Controls

70.1670.89 43–87

64.3971.09 36–87

125 (95.4%) 5 (3.8%) 1 (0.8%) 0

122 (95.3%) 2 (1.6%) 3 (2.3%) 1 (0.8%)

PSA (ng ml ) Mean7s.d.

2.8974.35

2.6974.61

MIF (ng ml1) Mean7s.d.

9.3274.99

3.1871.47

1

Gleason sum p6 X7 No data available

20 57 54

Abbreviations: CaP, prostate cancer; MIF, macrophage migration inhibitory factor; PSA, prostate-specific antigen. Genes and Immunity

CATT polymorphism were in accordance with Hardy– Weinberg equilibrium. Allelic frequencies for both markers were significantly different between the CaP and the control groups (Table 2; Po0.0001). There is linkage disequilibrium between the two polymorphisms (D0 ¼ 0.87),33 with the common haplotype distributions varying significantly between the two groups (Po0.0001). The 173G/6-CATT haplotype was the most common haplotype in the control group (44.9%), while the 173C/6-CATT haplotype was the most common in the CaP group (31.4%, Po0.001). Although these results point to a possible difference in the allelic distribution of MIF genotypes in CaP versus control groups, sampling bias is possible because the samples were from a greater than 95% Caucasian veteran population undergoing routine PSA screening. We therefore compared the allelic and genotype frequency of our control and CaP group with that of a previously reported Caucasian control group.31 The non-Caucasian patients in our group were not included in this analysis because of concerns of racial variation in the population distribution of MIF alleles.31 There was no difference in the allelic frequency of 173C (P ¼ 0.11) or the 794 CATT (P ¼ 0.09) polymorphisms between our control group and those previously reported for normal Caucasian patients.28 However, the allelic frequency of the CaP group for both polymorphisms was significantly different from that reported previously for cancer-free Caucasian controls (Po0.0001).28 MIF 173C and 794 MIF 7-CATT alleles and serum MIF in prostate cancer patients An association between elevated serum MIF levels and CaP was previously reported.34 In addition, highexpression MIF alleles are associated with higher circulating MIF levels in patients with autoimmune inflammatory arthritis.29 Therefore, we examined the

Table 2 Allele frequencies for the MIF 173 SNP and MIF 794 CATT polymorphisms and estimated haplotype frequencies in patients with CaP (cases, n ¼ 131) and benign prostate (controls, n ¼ 128) Cases (%)

Controls (%)

MIF 173 genotype Allele frequency G C

42.0 58.0

77.7 22.3

MIF 794 CATT genotype Allele frequency 5 6 7

13.4 54.2 32.4

35.2 57.8 7.0

MIF haplotype Estimated frequency G/5 G/6 G/7 C/5 C/6 C/7

5.6 22.8 13.6 7.8 31.4 18.8

27.4 44.9 5.4 13.6 12.9 1.5

P-value

0.0001

0.0001

0.0001

Abbreviation: MIF, macrophage migration inhibitory factor.


association of high-expression MIF alleles (173C and/ or 7-CATT) with serum MIF concentration in our CaP group. The means and standard errors of the mean are presented in Table 3. The two-factor ANOVA showed a significant main effect for the SNP factor, F (1,128) ¼ 5.6273, P ¼ 0.019; and a significant main effect for the CATT factor, F (1,128) ¼ 9.4475, P ¼ 0.0026; but the interaction between SNP and CATT was not significant, F (1,128) ¼ 0.0263, P405. Thus, the effects of SNP and CATT on serum MIF levels are additive and the combination of the 173C and 7-CATT polymorphisms resulted in the highest serum MIF amounts (9.9670.7 ng ml1). Association between MIF 173C and the 794 MIF 7-CATT polymorphisms and prostate cancer Comparison of the allele and estimated haplotype frequencies between the CaP and the control group documented a significant difference (Table 2). Backward stepwise logistic regression was performed to determine the association between these genotypes and CaP. Since our control group was significantly younger than the CaP group and because increasing age is a CaP risk factor, all calculated odds ratios (OR) were adjusted for age. Logistic regression demonstrated a statistically significant association between the G/C (OR 5.12, 95% CI 2.73–9.61, Po0.001) and C/C (OR 10.20, 95% CI 4.78– 21.75, Po0.001) genotypes and CaP incidence (Table 4). Similar results were obtained with 794 CATT genotype in which logistic regression demonstrated statistically significant associations between the 5/7, 6/6, 6/7 and 7/ 7 genotypes and CaP incidence (Table 4). Analysis of MIF genotype combinations determined that the C/C genotype and 7-CATT (5/7, 6/7 or 7/7 genotype) were associated with CaP incidence (OR 24.65, 95% CI 8.26– 73.56, Po0.001; Table 4). Internal cross-validation was used to estimate the predictive utility of the MIF polymorphisms in determining CaP incidence (Table 5). The sensitivity of the combined generated model was 85% with 79.8% accuracy in predicting normal and 72.5% accuracy in predicting CaP patients. Association between MIF 173C and the 794 MIF 7-CATT polymorphisms and high Gleason sum tumors Migration inhibitory factor circulates in increased concentrations in patients with CaP, and serum levels

Table 3 Association of increased serum MIF and MIF genotype in

show a correlation with increasing Gleason sum.32 The correlation between MIF polymorphisms and CaP histology was determined in tumors with Gleason sum p6 (low grade) and X7 (high grade). Due to the small sample size, the G/C and C/C genotypes were pooled and compared to the G/G genotype and the presence of a single 794 7-CATT allele was compared to its absence. A strong association between high-grade tumors and the 173C allele was determined (OR 9.69, 95% CI 2.20– 42.66; Table 6). The overall predictive utility of the generated model was 79.2% with 94.7% accuracy in predicting high-grade tumors and 35% accuracy in predicting low-grade tumors. There was no correlation between the presence of a single 7-CATT allele and Gleason sum. In addition, the combination of a single 173C allele and 7-CATT allele was no better at predicting high-grade tumors than the 173C allele alone (Table 6). These data suggest that the highexpression 173C MIF genotype closely correlates with CaP aggressiveness. Table 4 Age-adjusted OR for the MIF 173 SNP or MIF 5-, 6-, 7-CATT polymorphisms and CaP incidence Odds ratio (95% CI)

P-value

MIF 173 genotype G/G G/C C/C

1.0 (reference) 5.12 (2.73–9.61) 10.20 (4.78–21.75)

— o0.001 o0.001

MIF CATT genotype 5/5 5/6 5/7 6/6 6/7 7/7

1.0 2.20 7.00 4.06 71.92 51.73

— 0.264 0.013 0.040 o0.001 o0.002

MIF –173/CATT genotype Category 1 G/G and 5/5, 5/6 or 6/6 Category 2 G/C, C/C or 5/7, 6/7, 7/7 Category 3 C/C and 5/7, 6/7 or 7/7

C+

1.0 (reference)

—

8.52 (4.50–16.14)

o0.001

24.65 (8.26–73.56)

o0.001

Table 5 Internal cross-validation (70/30 split) of age-adjusted MIF polymorphism CaP prediction models Sensitivity

Specificity

PPV (%)

NPV (%)

73.8% (69.5–78.1) 70.3% (65.4–75.0) 85.0% (83.2–86.8)

68.7% (64.6–72.7) 73.8% (66.1–81.9) 63.5% (59.6–67.5)

70.8

71.8

74.7

70.2

72.5

79.8

CATT 7

7+

4.7070.5 ng ml1 (n ¼ 19) 7.1370.8 ng ml1 (n ¼ 38)

7.1771.0 ng ml1 (n ¼ 9) 9.9670.7 ng ml1 (n ¼ 65)

MIF 173 G/C genotype MIF 794 CATT genotype

C

(reference) (0.55–8.83) (1.51–32.48) (1.07–15.44) (14.63–363.54) (4.51–593.67)

Abbreviation: MIF, macrophage migration inhibitory factor.

CaP (n ¼ 131) SNP

Abbreviation: SNP, single-nucleotide polymorphisms. Data are expressed as mean serum MIF amounts7s.e. for each of the SNP and CATT categories. C is defined as G/G, C+ is defined as G/C or C/C; 7 is defined as 5/5, 5/6 or 6/6; 7+ is defined as 5/7, 6/7 or 7/7.

649

SNP/CATT combined

Abbreviations: MIF, macrophage migration inhibitory factor; NPV, negative predictive value; PPV, positive predictive value; SNP, single-nucleotide polymorphism. The data are mean percentages of 10 random 70/30 splits of the total data set. Numbers in parentheses are lower and upper limits of 95% CIs. Genes and Immunity


650

Table 6 Age-adjusted OR for the high-expressing MIF genotype and prostate tumor Gleason sum Genotype

MIF genotype and Gleason sum Gleason sum p6 No. (%)

G/G G/C or C/C 5/5, 5/6 or 6/6 5/7, 6/7 or 7/7 G/G and 5/5, 5/6 or 6/6 G/C, C/C or 5/7, 6/7 or 7/7 G/C, C/C and 5/7, 6/7 or 7/7

7 13 9 11 5 6 9

(9.1) (16.9) (11.7) (14.3) (6.5) (7.8) (11.7)

Odds ratio (95% CI)

P-value

X7 No. (%) 3 54 20 37 3 17 37

(3.9) (70.1) (26.0) (48.0) (3.9) (22.1) (48.0)

1.0 9.69 1.0 1.51 1.0 4.72 6.85

(reference) (2.20–42.66) (reference) (0.54–4.26) (reference) (0.86–26.04) (1.38–34.14)

— 0.003 — 0.433 — 0.075 0.019

The Gleason sum data were available for a subset of the CaP group (n ¼ 77) that was divided into low (p6) and high (X7) Gleason sum categories.

Prostate cancer recurrence predicted by MIF high-expression polymorphisms The relationship between high-expression MIF polymorphisms and CaP recurrence (biochemical failure defined as two PSA values X0.4 ng ml1 and rising, or initiation of adjuvant therapy within 5 years post-CaP diagnosis) was examined in 42 of the CaP patients for which repeated PSA determinations were available. We initially grouped these patients into four categories: G/G and 5/5, 5/6 or 6/6 (n ¼ 8), G/C or C/C and 5/5, 5/6 or 6/6 (n ¼ 21), G/G and 5/7, 6/7 or 7/7 (n ¼ 0), and G/C or C/C and 5/7, 6/7 or 7/7 (n ¼ 13). Since there was no recurrence in the G/G and 5/5, 5/6 or 6/6 category, the patients were grouped such that the high-expression MIF alleles (G/C or C/C and 5/7, 6/7 or 7/7, n ¼ 13) were compared to the remaining patients (n ¼ 29). Of the highMIF expression CaP patients (G/C or C/C and 5/7, 6/7 or 7/7), 46.2% were diagnosed with recurrent cancer within 5 years, whereas only 10.3% of the remaining patients experienced recurrence (OR 4.80, 95% CI 1.20– 19.25, P ¼ 0.027; Figure 1).

Discussion We report a significant association between high-production MIF alleles and both incidence and recurrence of CaP. The MIF alleles defined by the 173C SNP and the 794 7-CATT promoter polymorphisms were associated independently with CaP, and this association increased for the 173C/7-CATT haplotype (Table 2), in accord with prior observations in autoimmune inflammatory diseases.29 We also observed a significant association between the 173C/C and 173G/C genotypes with Gleason sum tumors X7 (Table 6). An independent contribution of the CATT repeat to Gleason sum was not observed, which may reflect the more significant effect of the 173 SNP or the greater heterogeneity of the CATT locus, which exists in four allelic forms.27,28 We also observed a significant association between the presence of the 173C and 794 7-CATT alleles and recurrent cancer within 5 years unrelated to initial Gleason sum (Figure 1). This analysis was restricted to Caucasian patients, which comprised the largest ethnic group (495%) within our study population, because of the known population stratification in the MIF locus.31 Genes and Immunity

Figure 1 Cox proportional hazard model estimates of remaining free from PSA recurrence based upon migration inhibitory factor (MIF) genotype. Biochemical recurrence (two PSA values of 0.4 ng ml1 or greater and rising or initiation of adjuvant therapy) was entered into the analysis as the event of interest, and patients not experiencing treatment failure by 60-month post-initial diagnosis were censored based upon the date of last observation. Patients with a 173C/7-CATT carrier genotype had a significantly shorter biochemical recurrence interval than the remaining patients (Po0.01). 173C/7-CATT carrier genotype: 6 events, 7 censored versus remaining genotypes: 3 events, 26 censored.

Because of limitations in sample size and follow-up clinical data, we could not assess with greater precision the contribution of the different alleles to long-term survival, and such an inquiry will be important to perform. Both the 173C and the 7-CATT polymorphisms were associated with elevated levels of circulating MIF (Table 3). These findings support the functional role of these alleles in regulating MIF expression in the studied population, and suggest that MIF’s pro-tumorigenic action in prostate tissue may result from either local tissue overexpression or from its action as a circulating mediator. These data are the first evidence supporting the relevance of these polymorphisms in cancer. Intratumor MIF production has been previously reported to correlate in pathologic studies, with the biological aggressiveness of colon and lung-derived


tumors.22–24 MIF expression was observed in premalignant lesions such as colonic tubular adenomas,35 but it has not, so far, been linked to the earliest steps of carcinogenesis. Several studies identified an association between linked high-expression MIF alleles to different autoimmune inflammatory diseases.28–30 The present data are the first evidence supporting the relevance of these polymorphisms in cancer. Molecular epidemiology studies identified several genetic markers of disease susceptibility for prostatic carcinoma, including genes involved with androgensignaling cascade, vitamin D receptor and those involved in the activation and detoxification of environmental chemicals.35 The causality and strength of these findings are uncertain because of inconsistency across studies and the relative weakness of the statistical associations.35 Case–control studies examining genetic polymorphisms in different inflammatory cytokines also have yielded inconsistent results or are biologically difficult to explain.36–40 The association of MIF polymorphisms with CaP appears biologically plausible given the observed relationships between high-expression alleles and CaP incidence and outcome, and the correlation with serum MIF levels. Whether MIF has a direct action in carcinogenesis, as suggested by its ability to sustain MAP (ERK-1/2) kinase phosphorylation,16and to override p53 tumor suppressor activity,17,21,41 or if it acts primarily to maintain inflammatory cell activation and production of other downstream mediators within the tumor microenvironment, remains to be elucidated. MIF genotype nevertheless may serve as a risk marker in CaP, and therapies directed at MIF, which are presently in clinical development for inflammatory disorders, may offer a useful approach to the treatment of CaP.

Acknowledgements This study was funded by grants from VA Merit Review Program (KLMS, PLV), the NIH (1R21DK075059-01-A1, PLV; R01AR050498 RB), Donaghue Foundation and the Yale Department of Pathology (CB).

References 1 Cancer Facts & Figures 2007. In: http://www.cancer.org/ downloads/STT/CAFF2007PWSecured.pdf. American Cancer Society, 2007. 2 George NJ. Conservative management of prostate cancer. Lancet 1988; 1: 1001–1010. 3 Hamdy FC. Prognostic and predictive factors in prostate cancer. Cancer Treat Rev 2001; 27: 143–151. 4 Debes JD, Tindall DJ. Mechanisms of androgen-refractory prostate cancer. N Engl J Med 2004; 351: 1488–1490. 5 Zhu P, Baek SH, Bourk EM, Ohgi KA, Garcia-Bassets I, Sanjo H et al. Macrophage/cancer cell interactions mediate hormone resistance by a nuclear receptor derepression pathway. Cell 2006; 124: 615–629. 6 Coussens LM, Werb Z. Inflammation and cancer. Nature 2002; 420: 860–867. 7 Cordon-Cardo C, Prives C. At the crossroads of inflammation and tumorigenesis. J Exp Med 1999; 190: 1367–1370. 8 Karin M, Lawrence T, Nizet V. Innate immunity gone awry: linking microbial infections to chronic inflammation and cancer. Cell 2006; 124: 823–835.

651 9 De Marzo AM, Marchi VL, Epstein JI, Nelson WG. Proliferative inflammatory atrophy of the prostate: implications for prostatic carcinogenesis. Am J Pathol 1999; 155: 1985–1992. 10 Palapattu GS, Sutcliffe S, Bastian PJ, Platz EA, De Marzo AM, Isaacs WB et al. Prostate carcinogenesis and inflammation: emerging insights. Carcinogenesis 2005; 26: 1170–1181. 11 Mahmud S, Franco E, Aprikian A. Prostate cancer and use of nonsteroidal anti-inflammatory drugs: systematic review and meta-analysis. Br J Cancer 2004; 90: 93–99. 12 Platz EA, Rohrmann S, Pearson JD, Corrada MM, Watson DJ, De Marzo AM et al. Nonsteroidal anti-inflammatory drugs and risk of prostate cancer in the Baltimore Longitudinal Study of Aging. Cancer Epidemiol Biomarkers Prev 2005; 14: 390–396. 13 De Marzo AM, Platz EA, Sutcliffe S, Xu J, Gronberg H, Drake CG et al. Inflammation in prostate carcinogenesis. Nat Rev Cancer 2007; 7: 256–269. 14 Mitchell RA, Bucala R. Tumor growth-promoting properties of macrophage migration inhibitory factor (MIF). Semin Cancer Biol 2000; 10: 359–366. 15 Chesney J, Metz C, Bacher M, Peng T, Meinhardt A, Bucala R. An essential role for macrophage migration inhibitory factor (MIF) in angiogenesis and the growth of a murine lymphoma. Mol Med 1999; 5: 181–191. 16 Mitchell RA, Metz CN, Peng T, Bucala R. Sustained mitogenactivated protein kinase (MAPK) and cytoplasmic phospholipase A2 activation by macrophage migration inhibitory factor (MIF). Regulatory role in cell proliferation and glucocorticoid action. J Biol Chem 1999; 274: 18100–18106. 17 Hudson JD, Shoaibi MA, Maestro R, Carnero A, Hannon GJ, Beach DH. A proinflammatory cytokine inhibits p53 tumor suppressor activity. J Exp Med 1999; 190: 1375–1382. 18 Bando H, Matsumoto G, Bando M, Muta M, Ogawa T, Funata N. Expression of macrophage migration inhibitory factor in human breast cancer: association with nodal spread. Jpn J Cancer Res 2002; 93: 389–396. 19 Bini L, Magi B, Marzocchi B, Arcuri F, Tripodi S, Cintorino M et al. Protein expression profiles in human breast ductal carcinoma and histologically normal tissue. Electrophoresis 1997; 18: 2832–2841. 20 Hagemann T, Wilson J, Kulbe H, Li NF, Leinster DA, Charles K et al. Macrophages induce invasiveness of epithelial cancer cells via NF-kappa B and JNK. J Immunol 2005; 175: 1197–1205. 21 Yao K, Shida S, Selvakumaran M, Zimmerman R, Simon E, Schick J et al. Macrophage migration inhibitory factor is a determinant of hypoxia-induced apoptosis in colon cancer cell lines. Clin Cancer Res 2005; 11: 7264–7272. 22 Legendre H, Decaestecker C, Nagy N, Hendlisz A, Schu¨ring MP, Salmon I et al. Prognostic values of galectin-3 and the macrophage migration inhibitory factor (MIF) in human colorectal cancers. Mod Pathol 2003; 16: 491–504. 23 White ES, Flaherty KR, Carskadon S, Brant A, Iannettoni MD, Yee J et al. Macrophage migration inhibitory factor and CXC chemokine expression in non-small cell lung cancer: role in angiogenesis and prognosis. Clin Cancer Res 2003; 9: 853–860. 24 Tomiyasu M, Yoshino I, Suemitsu R, Okamoto T, Sugimachi K. Quantification of macrophage migration inhibitory factor mRNA expression in non-small cell lung cancer tissues and its clinical significance. Clin Cancer Res 2002; 8: 3755–3760. 25 Meyer-Siegler K, Fattor RA, Hudson PB. Expression of macrophage migration inhibitory factor in the human prostate. Diagn Mol Pathol 1998; 7: 44–50. 26 Meyer-Siegler KL, Iczkowski KA, Vera PL. Further evidence for increased macrophage migration inhibitory factor expression in prostate cancer. BMC Cancer 2005; 5: 73. 27 Yu DS, Hsieh DS, Chang SY. Increasing expression of GST-pi MIF, and ID1 genes in chemoresistant prostate cancer cells. Arch Androl 2006; 52: 275–281. 28 Baugh JA, Chitnis S, Donnelly SC, Monteiro J, Lin X, Plant BJ et al. A functional promoter polymorphism in the macrophage migration inhibitory factor (MIF) gene associated with disease severity in rheumatoid arthritis. Genes Immun 2002; 3: 170–176. Genes and Immunity


652 29 Donn RP, Shelley E, Ollier WE, Thomson W, British Paediatric Rheumatology Study Group. A novel 50 -flanking region polymorphism of macrophage migration inhibitory factor is associated with systemic-onset juvenile idiopathic arthritis. Arthritis Rheum 2001; 44: 1782–1785. 30 Barton A, Lamb R, Symmons D, Silman A, Thomson W, Worthington J et al. Macrophage migration inhibitory factor (MIF) gene polymorphism is associated with susceptibility to but not severity of inflammatory polyarthritis. Genes Immun 2003; 4: 487–491. 31 Zhong XB, Leng L, Beitin A, Chen R, McDonald C, Hsiao B et al. Simultaneous detection of microsatellite repeats and SNPs in the macrophage migration inhibitory factor (MIF) gene by thin-film biosensor chips and application to rural field studies. Nucleic Acids Res 2005; 33: e121. 32 Wu S-P, Leng L, Feng Z, Liu N, Zhao H, McDonald C et al. Macrophage migration inhibitory factor promoter polymorphisms and the clinical expression of scleroderma. Arthritis Rheum 2006; 54: 3661–3669. 33 Stephens M, Scheet P. Accounting for decay of linkage disequilibrium in haplotype inference and missing-data imputation. Am J Hum Genet 2005; 76: 449–462. 34 Meyer-Siegler KL, Bellino MA, Tannenbaum M. Macrophage migration inhibitory factor evaluation compared with prostate specific antigen as a biomarker in patients with prostate carcinoma. Cancer 2002; 94: 1449–1456.

Genes and Immunity

35 Coughlin SS, Hall IJ. A review of genetic polymorphisms and prostate cancer risk. Ann Epidemiol 2002; 12: 182–196. 36 Wu HC, Chang CH, Chen HY, Tsai FJ, Tsai JJ, Chen WC. p53 gene codon 72 polymorphism but not tumor necrosis factoralpha gene is associated with prostate cancer. Urol Int 2004; 73: 41–46. 37 McCarron SL, Edwards S, Evans PR, Gibbs R, Dearnaley DP, Dowe A et al. Influence of cytokine gene polymorphisms on the development of prostate cancer. Cancer Res 2002; 62: 3369–3372. 38 Michaud DS, Daugherty SE, Berndt SI, Platz EA, Yeager M, Crawford ED et al. Genetic polymorphisms of interleukin-1B (IL-1B), IL-6, IL-8, and IL-10 and risk of prostate cancer. Cancer Res 2006; 66: 4525–4530. 39 Lindmark F, Zheng SL, Wiklund F, Bensen J, Ba¨lter KA, Chang B et al. H6D polymorphism in macrophage-inhibitory cytokine-1 gene associated with prostate cancer. J Natl Cancer Inst 2004; 96: 1248–1254. 40 Hayes VM, Severi G, Southey MC, Padilla EJ, English DR, Hopper JL et al. Macrophage inhibitory cytokine-1 H6D polymorphism, prostate cancer risk, and survival. Cancer Epidemiol Biomarkers Prev 2006; 15: 1223–1225. 41 Fingerle-Rowson G, Petrenko O, Metz CN, Forsthuber TG, Mitchell R, Huss R et al. The p53-dependent effects of macrophage migration inhibitory factor revealed by gene targeting. Proc Natl Acad Sci USA 2003; 100: 9354–9359.