Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2014 Jun 15.
Published in final edited form as: Int J Cancer. 2012 Dec 13;132(12):2928–2939. doi: 10.1002/ijc.27960

Genetic Variation in Bone Morphogenetic Proteins and Breast Cancer Risk in Hispanic and non-Hispanic white women: the Breast Cancer Health Disparities Study

Martha L Slattery 1, Esther M John 2, Gabriela Torres-Mejia 3, Jennifer S Herrick 1, Anna R Giuliano 4, Kathy B Baumgartner 5, Lisa M Hines 6, Roger K Wolff 1
PMCID: PMC3653321  NIHMSID: NIHMS423916  PMID: 23180569

Abstract

Bone morphogenetic proteins (BMP) are thoughtx to be important in breast cancer promotion and progression. We evaluated genetic variation in BMP-related genes and breast cancer risk among Hispanic (2111 cases, 2597 controls) and non-Hispanic white (NHW) (1481 cases, 1586 controls) women who participated in the 4-Corner’s Breast Cancer Study, the Mexico Breast Cancer Study, and the San Francisco Bay Area Breast Cancer Study. BMP genes and their receptors evaluated include ACVR1, AVCR2A, ACVR2B, ACVRL1, BMP1, BMP2, BMP4, BMP6, BMP7, BMPR1A, BMPR1B, BMPR2, MSTN and GDF10. Additionally, 104 ancestral informative markers were assessed to discriminate between European and Native American ancestry. The importance of estrogen on BMP-related associations was suggested through unique associations by menopausal status and estrogen (ER) and progesterone (PR) receptor status of tumors. After adjustment for multiple comparisons ACVR1 (8 SNPs) was modestly associated with ER+PR+ tumors [odds ratios (ORs between 1.18 and 1.39 padj< 0.05]. ACVR1 (3 SNPs) and BMP4 (3 SNPs) were associated with ER+PR− tumors (ORs 0.59 to 2.07 padj< 0.05). BMPR2 was associated with ER−PR+ tumors (OR 4.20, 95% CI 1.62, 10.91 padj< 0.05) as was GDF10 (2 SNPs ORs 3.62 and 3.85 padj< 0.05). After adjustment for multiple comparisons several SNPs remained associated with ER−PR− tumors (padj< 0.05) including ACVR1 BMP4, and GDF10 (ORs between 0.53 and 2.12). Differences in association also were observed by percentage of Native ancestry and menopausal status. Results support the hypothesis that genetic variation in BMPs is associated with breast cancer in this admixed population.

Keywords: BMP, ACVR1, BMPRIB, breast cancer, Hispanic, genetic admixture, survival, ER status

Introduction

Bone morphogenetic proteins (BMP) are thought to be involved in the initiation and progression of cancer 1, 2. As members of the TGFβ-signaling pathway they play a critical role in carcinogenesis through regulation of cell growth, differentiation, proliferation, and apoptosis 3. Members of the human BMP family include BMP1-7, growth differentiation factors (GDF) including GDF10 and GDF8 (also known as myostatin) 4 and their receptors. BMP ligands bind to type 1 and type 2 receptors. Type 1 receptors include BMPR1A, BMPR1B, activin A receptor type 1 (ACVR1), and activin receptor-like kinase 1 (ACVRL1) ; type II receptors include BMPR2, activin A receptor type IIA (ACVR2A) and type IIB (ACVR2B) 4. Both types of receptors are needed for BMP signaling, although type I receptors bind with a higher affinity than the type II receptors. BMPs have been shown to trigger a SMAD-signaling cascade that is linked to reduced cell proliferation and cellular growth kinetics of glioblastomas 5, 6 and may play a key role in regulating tumor initiation.

Little is known about the genetic variation in BMP genes and their associations with breast cancer risk, although these genes have been shown to be important in colorectal cancer7. Some studies suggest a link of BMPs to breast cancer disease progression and metastasis. Because of BMPs role in bone formation, they have been examined for their involvement in metastasis to the bone after diagnosis with breast cancer 8. Additionally BMPs have been associated with estrogen-induced proliferation of breast cancer cells 9. One study has shown that BMP-SMAD activation is involved in the progression of estrogen receptor positive (ER+) breast cancers specifically 10.

In this study, we examined genetic variation in BMP1, BMP2, BMP4, BMP6, BMP7, Growth Differentiation Factor 10 (GDF10) also known as BMP3B, myostatin (MSTN), and their relevant receptor genes BMPR1A, BMPR1B, BMPR2, and ACVR1, ACVR2A, ACVR2B, and ACVRL1. We assessed overall associations as well as associations by ER and progesterone receptor (PR) status of the tumors and by menopausal status. Given our ethnically diverse population we also evaluated associations by genetic admixture to determine whether differences in these genes may contribute to observed ethnic differences in breast cancer incidence rates.

Methods

The Breast Cancer Health Disparities Study includes participants from three population-based case-control studies, the 4-Corner’s Breast Cancer Study, the Mexico Breast Cancer Study, and the San Francisco Bay Area Breast Cancer Study11. All participants signed informed written consent prior to participation and each study was approved by the Institutional Review Board for Human Subjects at each institution.

Participants from the 4 Corner’s Breast Cancer Study were NHW, Hispanic, or Native American women living in non-reservation areas in the states of Arizona, Colorado, New Mexico, or Utah at the time of diagnosis or selection 12 and included female breast cancer cases between 25 and 79 years of age with a histological confirmed diagnosis of in situ (n=341) or invasive (n=1492) cancer between October 1999 and May 2004 and who provided a blood sample and completed the interview. Controls were selected from the target populations and were frequency matched to cases on ethnicity and 5-year age distribution. Of the 852 Hispanic, 22 American Indian, and 1683 NHW cases and 913 Hispanic, 23 American Indian, and 1669 NHW controls who participated in the interview, a blood sample for DNA extraction was provided for 76% of participants in Arizona, 71% of participants in Colorado, 75% of participants in New Mexico, and 94% of participants in Utah.

Participants from the Mexico Breast Cancer Study were between 28 and 74 years of age, living in one of three states, Monterrey, Veracruz and Mexico City, for the past five years as previously described13. Eligible cases were women diagnosed with either a new histologically confirmed in situ or invasive breast cancer between January 2004 and December 2007 at 12 participating hospitals from three main health care systems in Mexico. Controls were randomly selected from the catchment area of the 12 participating hospitals using a probabilistic multi-stage design. A total of 1000 cases and 1074 controls completed the in-person interview, and blood was collected and DNA extracted for 85% and 96% of women, respectively.

The San Francisco Bay Area Breast Cancer Study included women aged 35 to 79 years from the San Francisco Bay Area diagnosed with a first primary histologically confirmed invasive breast cancer between 1995 and 2002; controls were identified by random-digit dialing (RDD) and frequency-matched to cases based on the expected race/ethnicity and 5-year age distribution.14, 15 This analysis included subjects with bio-specimen collection, including Hispanic cases diagnosed between April 1997 and April 2002 and a 10% random sample of NHW cases diagnosed between April 1997 and April 1999 and their matched controls. A total of 1105 cases (793 Hispanic, 312 NHW) and 1318 controls (998 Hispanic, 320 NHW) completed the in-person interview, and blood or mouthwash samples were collected and DNA extracted for 93% of cases and 92% of controls.

Data Harmonization

Data were harmonized across all study centers and questionnaires. Variables used in the analyses included body mass index (BMI) calculated as weight (kg) divided by height squared [meters squared (m2)], based on measured height (or self-reported height if the measurement was declined) and self-reported weight during the referent year or more distantly recalled weight if referent year weight was not available or measured weight if neither were available. The referent year was defined as the year prior to diagnosis for cases or selection for controls. Parity was defined as the number of full-term pregnancies, age at first birth was defined as age at first live birth or still birth, and race/ethnicity in the U.S. studies was based on self-report (all women in Mexico were classified as Hispanic since information on race/ethnicity was not collected). Women were classified as either pre -menopausal or post-menopausal based on responses to questions on menstrual history. Women who reported still having periods during the referent year were classified as pre-menopausal. Center-specific definitions were used to define post-menopausal women. Women were classified as post-menopausal if they reported a natural menopause. If they reported taking hormone therapy (HT) and were still having periods and were at or above the 95th percentile of age for race/ethnicity of those who reported having a natural menopause (i.e., ≥ 12 months since their last period) within their study center they were classified as post-menopausal. This age was 58 for NHW and 56 for Hispanics from the 4-Corner’s Breast Cancer Study, 54 for the Mexico Breast Cancer Study, and 55 for NHW and 56 for Hispanics from the San Francisco Bay Area Breast Cancer Study. Mean daily grams of alcohol intake consumed over the lifetime were available for all but about 600 cases and controls from California. For those women we used alcohol consumption during the referent year as an adjustment variable. Physical activity was harmonized as hours of vigorous activity performed at leisure and chores during the referent year and analyzed using center-specific cut-points to accommodate the level of inquiry of each study questionnaire

Genetic Data

DNA was extracted from either whole blood or mouthwash samples; 7287 blood-derived and 634 mouthwash-derived samples were available. Whole Genome Amplification (WGA) was applied to the mouthwash-derived DNA samples prior to genotyping. A tagSNP approach was used to characterize variation across candidate genes. TagSNPs were selected using the following parameters: linkage disequilibrium (LD) blocks were defined using a Caucasian LD map and an r2=0.8; minor allele frequency (MAF) >0.1; range= −1500 bps from the initiation codon to +1500 bps from the termination codon; and 1 SNP/LD bin. Additionally, 104 Ancestral Informative Markers (AIMs) were used to distinguish European and Native American ancestry in the study population11. All markers were genotyped using a multiplexed bead array assay format based on GoldenGate chemistry (Illumina, San Diego, California). A genotyping call rate of 99.93% was attained (99.65% for WGA samples). We included 132 blinded internal replicates representing 1.6% of the sample set. The duplicate concordance rate was 99.996% as determined by 193,297 matching genotypes among sample pairs.

In the current analysis, we evaluated 138 SNPs in 14 genes: ACVR1 (16 SNPs), AVCR2A (6 SNPs), ACVR2B (3 SNPs), ACVRL1 (4 SNPs), BMP1 (10 SNPs), BMP2 (6 SNPs), BMP4 (4 SNPs), BMP6 (23 SNPs), BMP7 (24 SNPs), BMPR1A (9 SNPs), BMPR1B (18 SNPs), BMPR2 (8 SNPs), MSTN (1 SNPs,) and GDF10 (6 SNPs). Online Supplement 1 describes the SNPs in detail, including the minor allele frequency (MAF) and Hardy-Weinberg equilibrium (HWE) p value.

Tumor Characteristics

Cancer registries in Utah, Colorado, Arizona, New Mexico, and California provided information on stage at diagnosis and ER and PR status. Information on ER and PR status was available for 1019 (69%) NHW cases and 977 (75%) Hispanic cases.

Statistical Methods

The program STRUCTURE was used to compute individual ancestry for each study participant assuming two founding populations16, 17. A three-founding population model was assessed but did not fit the population structure with the same level of repeatability and correlation among runs as the two-founding population model. Participants were classified by level of percent Native American ancestry. Assessment across categories of ancestry was done using cut-points based on the distribution of genetic ancestry in the control population with the goal of creating distinct ancestry groups that had sufficient power to assess associations. Three strata, 0-28%, 29 to 70%, and 71 to 100%, were used to evaluate associations by level of Native ancestry. Genetic ancestry was used as a continuous variable when used to adjust for possible confounding. Genes were assessed for their association with breast cancer risk by strata of genetic ancestry, ER/PR status, and menopausal status. All statistical analyses were performed using SAS version 9.2 (SAS Institute, Cary, NC). Logistic regression models were used to estimate the age, genetic ancestry, and study center-adjusted odds ratios (OR) and 95% confidence intervals (CI) for breast cancer risk associated with SNPs. Except for the California study, both invasive and in situ cases were included given associations that the Mexico Study did not distinguish between these categories. Additionally, we adjusted for potential confounding variables of BMI, parity, age at first birth, hours per week of vigorous physical activity, alcohol consumption, and genetic admixture. SNPs were assessed assuming a co-dominant model. Based on the initial assessment, SNPs which appeared to have a dominant or recessive mode of inheritance were evaluated with those inheritance models in subsequent analyses. Interactions between genetic variants and genetic ancestry and menopausal status were assessed using p values from a Wald chi-square test.

Haplotypes were developed to help define risk associated with genes. SNPs were selected based on their individual significance overall or within a genetic admixture group as well as for those considered in higher linkage disequilibrium (LD) or an R-squared linkage disequilibrium measure greater than or equal to 0.5. Haplotypes were estimated using the EM algorithm. Per-copy haplotype risk estimates were obtained using logistic regression and adjusted for age, study, genetic admixture, BMI in referent year, age at first birth, parity, alcohol consumption and vigorous physical activity.

The p values were based on one degree of freedom (1-df) Wald test statistics and were adjusted for multiple comparisons taking into account tagSNPs within the gene using the step-down Bonferroni correction (i.e., Holm method). This method isbased on the effective number of independent SNPs as determined using the SNP spectral decomposition method proposed by Nyholt 18 and modified by Li and Ji 19. We provide an online Supplement 2 that shows the number of SNPs analyzed and the number of SNPs considered for adjustment based on correlation between SNPs within a gene. The interaction p values, based on 1-df Wald chi-square tests, were adjusted using the step-down Bonferroni correction or the Holm’s test 20. This method of correction for multiple comparisons is very conservative, especially for correlated variables such as SNPs within a gene. Given that we are assessing hypothesized associations within a candidate pathway and candidate genes, we considered an adjusted p value of 0.15 or less as potentially important. However, we include in the text only those SNPs with adjusted p values of <0.05 and highlight those associations using bold font in the tables.

Results

The majority of participants were Hispanic (59% of cases and 62% of controls) (Table 1). More Hispanic women than NHW were diagnosed prior to 40 years of age (9% vs. 6%). Hispanic cases were more likely to be pre-menopausal at diagnosis than NHW cases (41% vs. 34%). Hispanic cases were more likely to have ER−PR− tumors than NHW women (23% vs. 18%).

Table 1.

Description of Study Population by Self-reported Race/Ethnicity

NHW Hispanic
Controls Cases Controls Cases
N % N % N % N %
Total 1586 37.9 1481 41.2 2597 62.1 2111 58.8
Study Site
 4 Corner’s States 1322 83.4 1227 82.8 723 27.8 597 28.3
 Mexico 0 0.0 0 0.0 994 38.3 816 38.7
 California 264 16.6 254 17.2 880 33.9 698 33.1
Age (years)
 <40 116 7.3 89 6.0 311 12.0 200 9.5
 40-49 408 25.7 409 27.6 831 32.0 713 33.8
 50-59 409 25.8 413 27.9 756 29.1 617 29.2
 60-69 350 22.1 361 24.4 526 20.3 430 20.4
 70+ 303 19.1 209 14.1 173 6.7 151 7.2
 Mean 56.6 56.0 52.3 52.7
Menopausal Status
 Pre-menopausal 494 31.5 489 33.5 1027 40.7 836 40.9
 Post-menopausal 1076 68.5 970 66.5 1499 59.3 1210 59.1
Estimated Native American Ancestry
 Low (0 - 28%) 1578 99.5 1472 99.4 278 10.7 275 13.0
Intermediate (29 - 70%) 7 0.4 7 0.5 1686 64.9 1393 66.0
High (71 - 100%) 1 0.1 2 0.1 633 24.4 443 21.0
ER/PR Status*
 ER+PR+ NA 695 68.2 NA 605 61.9
 ER+PR− NA 121 11.9 NA 115 11.8
 ER−PR+ NA 15 1.5 NA 28 2.9
 ER−PR− NA 188 18.4 NA 229 23.4
Open in a new tab
*

Tumor information was unavailable for the Mexico study sites.

There were few associations with breast cancer risk overall. However, when we examined cases defined by ER/PR status, several genes appeared to be associated with specific breast cancer subtypes (Table 2 shows those that remained significant at the 0.05 level after multiple comparison adjustment, whereas Supplement 3 shows those that were initially significant at the 0.05 level but after multiple adjustment testing had adjusted p values of >0.05). Variants in ACVR1 (8 SNPs) and BMP4 (1 SNP) were associated with ER+PR+ tumors, although the level of association was modest with ORs between 1.19 and 1.33. ACVR1 (4 SNPs), BMP4 (3 SNPs), and GDF10 (1 SNP) were associated with ER+PR− tumors, with OR estimates ranging from 0.59 to 2.07. BMPR2 (1 SNP) and GDF10 (2 SNPs) were associated with ER−PR+ tumors (ORs between 3.62 and 4.20). ACVR1 (1 SNP), ACVRL1 (1 SNP), BMP4 (1 SNP), and GDF10 (1 SNP) were associated with ER−PR− tumors.

Table 2.

Associations between BMP-related genes and risk of breast cancer overall and by ER/PR status

Overall ER + PR +1 ER + PR − ER − PR + ER − PR −
Controls Cases (N=1292 cases) (N=235 cases) (N=41 cases) (N=411 cases)
N N OR2 (95% CI) OR (95% CI) OR (95% CI) OR (95% CI) OR (95% CI)
ACVR1 (rs2033962)
 GG 2940 2452 1.00 1.00 1.00 1.00 1.00
 GT/TT 1159 1081 1.09 (0.99, 1.20) 1.27 (1.11, 1.46) 0.96 (0.71, 1.29) 0.94 (0.46, 1.90) 1.13 (0.91, 1.42)
ACVR1 (rs4380178)
 GG 3046 2538 1.00 1.00 1.00 1.00 1.00
 GA/AA 1053 997 1.12 (1.01, 1.24) 1.25 (1.08, 1.44) 0.96 (0.71, 1.30) 1.01 (0.50, 2.04) 1.25 (1.00, 1.57)
ACVR1 (rs920522)
 TT 3501 2970 1.00 1.00 1.00 1.00 1.00
 TC/CC 601 565 1.13 (1.00, 1.28) 1.33 (1.12, 1.59) 1.43 (1.01, 2.03) 0.79 (0.31, 2.04) 0.94 (0.70, 1.27)
ACVR1 (rs17182166)
 GG/GT 4051 3466 1.00 1.00 1.00 1.00 1.00
 TT 51 68 1.40 (0.97, 2.03) 1.47 (0.93, 2.33) 1.27 (0.49, 3.24) 1.66 (0.22, 12.62) 2.12 (1.13, 3.97)
ACVR1 (rs1146035)
 GG 2958 2467 1.00 1.00 1.00 1.00 1.00
 GT/TT 1144 1066 1.08 (0.98, 1.20) 1.26 (1.10, 1.45) 0.97 (0.72, 1.30) 1.02 (0.52, 2.03) 1.10 (0.88, 1.38)
ACVR1 (rs10497191)
 CC 3050 2578 1.00 1.00 1.00 1.00 1.00
 CT/TT 1050 957 1.09 (0.98, 1.21) 1.20 (1.04, 1.39) 1.23 (0.92, 1.66) 0.92 (0.45, 1.89) 1.06 (0.84, 1.33)
ACVR1 (rs10497192)
 TT 2206 1852 1.00 1.00 1.00 1.00 1.00
 TC 1606 1409 1.04 (0.95, 1.15) 1.20 (1.05, 1.38) 1.01 (0.76, 1.34) 0.86 (0.45, 1.66) 1.08 (0.87, 1.34)
 CC 290 274 1.13 (0.94, 1.35) 1.39 (1.09, 1.77) 1.13 (0.68, 1.87) 0.84 (0.25, 2.85) 1.02 (0.68, 1.53)
ACVR1 (rs4233672)
 GG 2764 2303 1.00 1.00 1.00 1.00 1.00
 GA/AA 1337 1231 1.10 (1.00, 1.21) 1.22 (1.07, 1.40) 0.99 (0.75, 1.32) 0.95 (0.48, 1.85) 1.14 (0.92, 1.41)
ACVR1 (rs10933443)
 TT 2341 1994 1.00 1.00 1.00 1.00 1.00
 TC/CC 1756 1538 1.02 (0.93, 1.12) 1.19 (1.04, 1.35) 1.00 (0.76, 1.31) 1.21 (0.65, 2.26) 1.07 (0.87, 1.32)
ACVR1 (rs2883605)
 GG 3568 3047 1.00 1.00 1.00 1.00 1.00
 GT/TT 483 438 1.00 (0.87, 1.15) 1.06 (0.87, 1.28) 0.59 (0.37, 0.95) 0.95 (0.37, 2.48) 1.11 (0.82, 1.49)
ACVR1 (rs10497193)
 AA/AG 4002 3447 1.00 1.00 1.00 1.00 1.00
 GG 99 87 1.04 (0.77, 1.39) 1.01 (0.66, 1.53) 2.07 (1.08, 3.98) 1.76 (0.41, 7.51) 0.80 (0.40, 1.61)
ACVR1 (rs4664901)
 TT/TC 3891 3338 1.00 1.00 1.00 1.00 1.00
 CC 207 195 1.10 (0.90, 1.35) 1.11 (0.84, 1.48) 1.94 (1.21, 3.11) 1.31 (0.40, 4.32) 1.01 (0.64, 1.58)
ACVRL1 (rs11169953)
 CC 2003 1787 1.00 1.00 1.00 1.00 1.00
 CT/TT 2098 1747 0.91 (0.83, 1.00) 0.93 (0.81, 1.06) 1.05 (0.80, 1.37) 1.31 (0.70, 2.46) 0.73 (0.60, 0.90)
BMP4 (rs17563)
 TT 1449 1056 1.00 1.00 1.00 1.00 1.00
 TC/CC 2376 2159 1.16 (1.05, 1.29) 1.29 (1.10, 1.51) 1.26 (0.90, 1.77) 1.39 (0.64, 3.02) 1.38 (1.06, 1.80)
BMP4 (rs762642)
 TT/TG 3391 2957 1.00 1.00 1.00 1.00 1.00
 GG 710 578 0.93 (0.82, 1.05) 0.96 (0.80, 1.14) 0.62 (0.40, 0.94) 1.26 (0.58, 2.77) 0.90 (0.68, 1.20)
BMP4 (rs2761887)
 AA 1295 1067 1.00 1.00 1.00 1.00 1.00
 AC/CC 2805 2466 1.07 (0.97, 1.18) 1.10 (0.95, 1.27) 1.54 (1.13, 2.11) 1.10 (0.55, 2.16) 1.19 (0.95, 1.50)
BMP4 (rs4898820)
 TT/TG 3147 2743 1.00 1.00 1.00 1.00 1.00
 GG 955 790 0.95 (0.85, 1.06) 0.95 (0.81, 1.11) 0.62 (0.43, 0.90) 1.00 (0.47, 2.11) 0.95 (0.74, 1.22)
BMPR2 (rs12621870)
 TT 2525 2171 1.00 1.00 1.00 1.00 1.00
 TC 1354 1178 1.00 (0.91, 1.10) 1.04 (0.91, 1.20) 0.94 (0.70, 1.25) 1.93 (0.98, 3.77) 0.90 (0.72, 1.13)
 CC 212 174 0.94 (0.76, 1.17) 1.06 (0.79, 1.42) 1.24 (0.71, 2.16) 4.20 (1.62, 10.91) 0.76 (0.45, 1.28)
GDF10 (rs7093975)
 CC 2346 2001 1.00 1.00 1.00 1.00 1.00
 CT 1526 1285 0.97 (0.88, 1.06) 0.93 (0.81, 1.06) 0.90 (0.67, 1.20) 1.61 (0.82, 3.15) 0.96 (0.77, 1.20)
 TT 229 244 1.22 (1.01, 1.48) 1.19 (0.91, 1.56) 1.42 (0.86, 2.33) 3.62 (1.40, 9.41) 1.24 (0.82, 1.89)
GDF10 (rs762454)
 AA 2228 1802 1.00 1.00 1.00 1.00 1.00
 AG 1553 1432 1.12 (1.02, 1.23) 1.13 (0.99, 1.30) 1.26 (0.95, 1.68) 0.65 (0.32, 1.31) 1.02 (0.82, 1.26)
 GG 312 293 1.09 (0.92, 1.30) 1.07 (0.84, 1.36) 1.77 (1.15, 2.71) 1.31 (0.49, 3.50) 0.53 (0.33, 0.87)
GDF10 (rs2853838)
 CC 2554 2197 1.00 1.00 1.00 1.00 1.00
 CA 1379 1182 1.00 (0.90, 1.10) 0.97 (0.84, 1.12) 0.94 (0.71, 1.26) 2.04 (1.06, 3.91) 1.03 (0.82, 1.28)
 AA 169 155 1.09 (0.87, 1.36) 1.07 (0.77, 1.48) 1.24 (0.67, 2.30) 3.85 (1.27, 11.71) 1.33 (0.82, 2.15)
Open in a new tab
1

Estrogen Receptor (ER) and Progesterone Receptor (PR) data were compared to 3125 controls from sites where cases have ER/PR data

2

Odds Ratios (OR) and 95% Confidence Intervals (CI) adjusted for age, study, BMI in referent year, vigorous activity in referent year, parity, age at first birth, alcohol consumption and genetic admixture; bold text designates significant associations at the 0.05 level after adjustment for multiple comparisons

Several genes were associated with breast cancer risk within genetic admixture groups (Table 3). BMP4 rs17563 was associated with breast cancer among women with greater European ancestry at the 0.05 level or less after adjustment for multiple comparisons. However, among Hispanic women with low Native American ancestry (0-28%) BMP4 rs17563 was associated with increased risk (OR 2.42 95% CI 1.40,4.16 for CC vs. TT genotypes) and BMP6 rs270417 was associated with reduced risk (OR 0.55 95% CI 0.38,0.79 dominant model) (data in Supplement 4 that shows associations for NHW women and for Hispanic women by admixture category).

Table 3.

Associations between candidate genes and breast cancer by genetic admixture

0 - 28% Native American Ancestry 29 - 70% Native American Ancestry 71 -100% Native American Ancestry Interaction
Controls Cases Controls Cases Controls Cases Raw Holms
N N OR1,2 (95% CI) N N OR (95% CI) N N OR (95% CI) P-value P-value
ACVR2A (rs1014064)
 AA 848 837 1.00 552 435 1.00 141 129 1.00 0.20 0.39
 AG/GG 971 897 0.93 (0.82, 1.06) 1109 935 1.08 (0.92, 1.26) 481 302 0.69 (0.51, 0.92)
ACVR2A (rs2161983)
 CC 835 822 1.00 545 432 1.00 138 129 1.00 0.13 0.39
 CT/TT 983 911 0.94 (0.82, 1.07) 1116 938 1.07 (0.92, 1.25) 484 302 0.67 (0.50, 0.90)
ACVR2A (rs3768687)
 GG 847 834 1.00 550 435 1.00 140 129 1.00 0.16 0.39
 GA/AA 968 896 0.94 (0.82, 1.07) 1109 931 1.07 (0.91, 1.25) 480 302 0.68 (0.51, 0.92)
ACVR2A (rs10497025)
 CC 1024 1013 1.00 1097 837 1.00 400 300 1.00 0.60 0.60
 CG/GG 795 720 0.92 (0.80, 1.05) 564 533 1.23 (1.05, 1.43) 222 131 0.75 (0.57, 0.99)
ACVR2B (rs928813)
 GG 359 353 1.00 685 531 1.00 360 217 1.00 0.005 0.01
 GT/TT 1454 1369 0.97 (0.82, 1.15) 973 834 1.11 (0.95, 1.29) 262 210 1.35 (1.04, 1.76)
ACVR2B (rs2276541)
 AA 672 600 1.00 452 371 1.00 139 119 1.00 0.02 0.03
AG/GG 1147 1132 1.10 (0.95, 1.26) 1209 999 1.01 (0.86, 1.20) 483 312 0.69 (0.51, 0.93)
BMP4 (rs17563)
 TT 396 316 1.00 692 500 1.00 361 240 1.00 0.82 1.00
 TC/CC 1365 1330 1.24 (1.05, 1.47) 785 660 1.16 (0.99, 1.36) 226 169 1.13 (0.86, 1.49)
BMP4 (rs762642)
 TT 635 587 1.00 550 495 1.00 226 147 1.00 0.659 1.00
TG 878 857 1.05 (0.91, 1.22) 814 665 0.89 (0.76, 1.05) 288 206 1.09 (0.81, 1.45)
 GG 305 290 1.01 (0.83, 1.24) 297 210 0.77 (0.62, 0.96) 108 78 1.08 (0.74, 1.57)
BMP4 (rs2761887)
 AA/AC 1468 1399 1.00 1373 1088 1.00 480 345 1.00 0.91 1.00
 CC 350 334 1.01 (0.86, 1.20) 287 281 1.27 (1.05, 1.53) 142 86 0.85 (0.62, 1.16)
BMP7 (rs7273197)
 CC 839 792 1.00 949 854 1.00 381 269 1.00 0.26 0.84
 CT/TT 980 942 1.02 (0.90, 1.17) 712 516 0.79 (0.68, 0.92) 241 162 0.88 (0.67, 1.15)
BMPR2 (rs17199235)
 AA 1426 1348 1.00 1441 1236 1.00 598 413 1.00 0.12 0.73
 AG/GG 393 386 1.03 (0.88, 1.21) 220 134 0.67 (0.53, 0.84) 24 18 0.93 (0.48, 1.81)
Open in a new tab
1

Bold indicates adjusted odds ratio (OR) estimates remained significant at ≤0.05 after adjustment for multiple comparisons.

2

Odds Ratios (OR) and 95% Confidence Intervals (CI ) adjusted for age, study, genetic admixture, BMI in referent year, vigorous activity in referent year, parity, age at first birth and alcohol consumption.

ACVR2A rs1014064, rs2161983, rs10497025, and rs3768687, and ACVR2B rs928813 and rs2276541 were associated with breast cancer among those with the greatest Native American ancestry; the ACVR2B associations were statistically different from those observed for women with more European ancestry. ACVR2A rs10497025, BMP4 rs762642, and rs2761887, BMP7 rs7273197, and BMPR2 rs17199235 were associated with breast cancer in the intermediate ancestry group.

Haplotype analysis showed rare haplotypes to carry some risk beyond SNPs themselves (Table 4). For the ACVR1 the rare A-T haplotype of rs4380178 and rs17182166 (0.028% of the population) was associated with an OR of 1.51 (95% CI 1.14,2.00). The rare C-T-A-G haplotype of BMP6 rs10498671, rs1225929, rs1107495, and rs911749 which occurred in 0.012% of the population was associated with reduced breast cancer risk (OR 0.34, 95% CI 0.19,0.62). Among the highest Native ancestry group, two haplotypes of BMP7 were associated with increased breast cancer risk: BMP7 rs174080735, rs6025446, and rs6025468 A- A-A (0.014% of the population) OR 2.25 (95% CI 1.11,4.58) and BMP7 rs6127983 and rs6025468 T-G (0.008% of the population) (OR 4.55 95% CI 1.09,19.00). Additional haplotypes based on SNPs in higher LD also were identified as being statistically significant. These haplotypes were observed in less than 5% of the population and were primarily in BMP6 and BMP7.

Table 4.

Associations between haplotypes and breast cancer risk by Native American ancestry

Haplotype Frequency OR2 (95% CI) Haplotype Frequency OR2 (95% CI) Haplotype Frequency OR2 (95% CI)
  0 to 28% Native American Ancestry 29 - 70% Native American Ancestry 71 - 100% Native American Ancestry
BMP7 (rs1475000 A>G; rs4811822 T>C)1 ACVR1 (rs2165436 C>T; rs6437117 A>T)1 BMP7 (rs17480735 G>A; rs6025446 A>G; rs6025468 A>G)
 A-T 0.480 0.98 (0.89, 1.08)  C-A 0.837 0.98 (0.85, 1.12)  G-A-A 0.427 1.12 (0.93, 1.35)
 G-C 0.367 0.96 (0.87, 1.06)  T-T 0.115 1.07 (0.91, 1.26)  G-G-A 0.342 0.94 (0.78, 1.14)
 A-C 0.117 1.05 (0.91, 1.22)   T-A 0.035 0.73 (0.55, 0.98)  G-G-G 0.188 0.82 (0.65, 1.03)
 G-T 0.036 1.32 (1.01, 1.73)   C-T 0.014 1.69 (1.07, 2.68)  G-A-G 0.021 1.63 (0.73, 3.64)
BMP7 (rs4811822 T>C; rs2180780 C>G)1 ACVR1 (rs4380178 G>A; rs17182166 G>T) A-A-A 0.014 2.25 (1.11, 4.58)
 T-C 0.474 0.97 (0.88, 1.07)  G-G 0.801 0.98 (0.85, 1.12)  A-G-A 0.005 2.83 (0.45, 17.68)
 C-G 0.397 0.95 (0.86, 1.04)  A-G 0.115 0.89 (0.75, 1.05)  A-A-G 0.002 0.35 (0.03, 3.64)
 C-C 0.087 1.12 (0.95, 1.33)  G-T 0.056 1.01 (0.80, 1.27)  A-G-G 0.002 <0.001 <0.001 >999
  T-G 0.042 1.34 (1.05, 1.72)   A-T 0.028 1.51 (1.14, 2.00) BMP7 (rs6127983 T>C; rs6025468 A>G)
BMP7 (rs12481628 A>G; rs2180780 C>G)1 BMP6 (rs10498671 T>C; rs1225929 A>T; rs1107495 A>G; rs911749 G>A)  T-A 0.474 1.17 (0.97, 1.40)
 A-C 0.523 0.97 (0.88, 1.07)  T-T-A-G 0.229 1.28 (1.13, 1.45)  C-A 0.314 0.96 (0.79, 1.17)
 G-G 0.390 0.96 (0.87, 1.06)  T-A-A-G 0.217 0.95 (0.84, 1.08)  C-G 0.204 0.81 (0.64, 1.01)
 A-G 0.048 1.17 (0.93, 1.46)  T-A-G-G 0.155 0.95 (0.82, 1.09)   T-G 0.008 4.55 (1.09, 19.00)
  G-C 0.039 1.35 (1.05, 1.73)  T-A-G-A 0.094 0.96 (0.81, 1.14) BMP2 (rs1979855 T>C; rs3178250 T>C)1
 C-A-G-A 0.067 0.95 (0.78, 1.15)  T-T 0.920 1.04 (0.73, 1.46)
 T-T-G-G 0.060 0.89 (0.69, 1.14)  C-C 0.046 0.65 (0.41, 1.02)
 T-A-A-A 0.048 0.91 (0.69, 1.21)   T-C 0.029 1.83 (1.04, 3.22)
 C-A-A-A 0.036 0.91 (0.70, 1.19)  C-T 0.005 1.03 (0.25, 4.16)
 C-T-A-A 0.028 0.78 (0.54, 1.12) BMP6 (rs267190 T>G; rs267806 C>T)1
 C-A-A-G 0.017 1.06 (0.71, 1.58)  T-T 0.723 0.86 (0.70, 1.07)
 T-T-A-A 0.016 1.09 (0.61, 1.97)  G-C 0.189 1.05 (0.83, 1.34)
 T-T-G-A 0.014 1.71 (0.99, 2.97)   T-C 0.066 1.49 (1.04, 2.14)
  C-T-A-G 0.012 0.34 (0.19, 0.62)  G-T 0.021 0.75 (0.38, 1.46)
BMP7 (rs1475000 A>G; rs4811822 T>C)1 BMP7 (rs1475000 A>G; rs162317 G>A)1
 A-T 0.502 0.97 (0.87, 1.07)  A-G 0.576 0.90 (0.75, 1.09)
 G-C 0.366 1.02 (0.92, 1.14)  G-A 0.396 1.09 (0.91, 1.31)
 A-C 0.095 0.91 (0.76, 1.09)  A-A 0.015 0.55 (0.24, 1.25)
  G-T 0.037 1.41 (1.06, 1.88)   G-G 0.012 2.71 (1.14, 6.47)
BMP7 (rs1475000 A>G; rs6025446 A>G)1 BMP7 (rs1475000 A>G; rs2180780 C>G)1
 A-G 0.465 0.90 (0.82, 1.00)  A-C 0.568 0.91 (0.76, 1.09)
 G-A 0.371 1.03 (0.92, 1.14)  G-G 0.399 1.09 (0.91, 1.32)
 A-A 0.132 1.09 (0.93, 1.27)  A-G 0.023 0.64 (0.33, 1.22)
  G-G 0.032 1.47 (1.07, 2.02)   G-C 0.010 3.11 (1.16, 8.33)
BMP7 (rs4811822 T>C; rs2180780 C>G)1 BMP7 (rs12438 A>G; rs3787380 T>C)1
 T-C 0.491 0.97 (0.87, 1.07)  A-C 0.551 0.84 (0.70, 1.01)
 C-G 0.373 1.02 (0.92, 1.14)  G-T 0.392 1.05 (0.87, 1.27)
 C-C 0.088 0.90 (0.75, 1.09)   A-T 0.052 1.78 (1.19, 2.67)
  T-G 0.048 1.30 (1.01, 1.67)  G-C 0.005 1.16 (0.31, 4.30)
BMP7 (rs162315 G>A; rs172983 G>A)1 BMP7 (rs6127983 T>C; rs162317 G>A)1
 G-G 0.656 0.93 (0.84, 1.04)  C-G 0.518 0.84 (0.70, 1.00)
 A-A 0.244 1.02 (0.91, 1.16)  T-A 0.412 1.05 (0.88, 1.27)
 A-G 0.097 1.10 (0.93, 1.31)   T-G 0.070 1.62 (1.14, 2.32)
  G-A 0.002 6.17 (1.31, 29.05) BMP7 (rs6127983 T>C; rs2180780 C>G)1
BMP7 (rs172983 G>A; rs6014967 G>A)1  C-C 0.506 0.85 (0.71, 1.02)
 G-G 0.679 0.95 (0.85, 1.06)  T-G 0.410 1.07 (0.89, 1.29)
 A-A 0.238 1.01 (0.89, 1.14)   T-C 0.072 1.50 (1.06, 2.14)
 G-A 0.074 1.06 (0.87, 1.29)  C-G 0.013 0.63 (0.25, 1.56)
  A-G 0.008 2.14 (1.17, 3.91) BMP7 (rs3787380 T>C; rs6014949 G>A)1
 C-G 0.552 0.85 (0.71, 1.02)
 T-A 0.399 1.07 (0.89, 1.29)
  T-G 0.045 1.76 (1.15, 2.69)
 C-A 0.004 0.73 (0.17, 3.06)
BMP7 (rs12481628 A>G; rs2180780 C>G)1
 A-C 0.530 0.88 (0.74, 1.06)
 G-G 0.383 1.09 (0.91, 1.32)
  G-C 0.047 1.57 (1.03, 2.40)
 A-G 0.040 0.75 (0.45, 1.24)
BMPR1A (rs7088641 T>C; rs2168730 A>G)1
 T-A 0.582 1.01 (0.84, 1.22)
 C-G 0.398 0.94 (0.78, 1.14)
  C-A 0.017 2.36 (1.15, 4.86)
 T-G 0.004 0.42 (0.08, 2.22)
Open in a new tab
1

Indicates haplotypes were generated from SNPs where R2 values was greater than 0.50

2

Adjusted for age, study, genetic admixture, BMI in referent year, vigorous activity in referent year, parity, age at first birth and alcohol consumption; no important haplotypes were identified for lowest Native American ancestry group.

We observed differences in association by menopausal status for several genes; Figure 1 shows those associations that remained significant at the 0.05 level after adjustment for multiple comparisons. Supplement 5 shows all associations that were significant prior to adjustment along with the associated unadjusted and adjusted p values. For women with low Native American ancestry, ACVR1 rs1220134 and ACVR2A rs10497025 were associated with reduced breast cancer risk and BMRP1B rs1863652 was associated with increased risk among pre-menopausal women, whereas BMP4 rs17563 was associated with increased breast cancer risk among post-menopausal women. Breast cancer risk was different for pre-menopausal and post-menopausal women for several genes among women with the highest level of Native American ancestry. Among pre-menopausal women, ACVR2B rs2276541 was associated with reduced risk whereas rs928813 was associated with increased risk. Among post-menopausal women ACVR2B rs503327 and GDF10 rs1902725 were associated with increased risk. ACVR2A rs1097025, BMP4 rs762642 and rs2761887, BMP7 rs7273197, and BMPR2 rs17199235 showed differences in risk by menopausal status among women in the intermediate ancestry category.

Figure 1.

Figure 1

Open in a new tab

Associations between BMP-related genes and breast cancer risk by menopausal status and Native American Ancestry

Discussion

BMPs are extracellular signaling molecules and are the largest group of the TGFβ superfamily. Because of their role in cell regulation, proliferation, apoptosis, and migration, they have been implicated as potentially important in cancer etiology. Our data suggest that genetic variation in BMPs and their receptors are involved in breast cancer etiology and prognosis. Few SNPs were associated overall with breast cancer risk after adjustment for multiple comparisons and associations were very modest. However, several BMPs and their receptors were associated with specific ER and PR tumor phenotype and the associations were generally stronger for specific subtypes than for breast cancer risk. Additionally we observed unique associations by genetic ancestry as well as by menopausal status.

BMPs have been associated with estrogen in several studies, some of which have linked BMPs to expression of estrogen receptors and estrogen signaling 21. BMP6 and BMP7 have been shown to inhibit estrogenic enzyme expression 9. Helms and colleagues 10 have reported that BMPR1B is involved in the progression and differentiation of ER-positive breast cancer. BMP6 and BMP7 preferentially bind to BMPR1B and ACVR1. Estrogen also has been shown to reduce expression of BMPR1A, BMPR1B, ACVR2A, and ACVR2B 9. BMP7 has been shown to be associated with expression of both the estrogen and progesterone receptor 22. In another study, hypermethylation of BMP6 was observed in all ER-negative breast cancers, but in only 18% of ER-positive breast cancers, suggesting a correlation between BMP6 expression and ER status of breast cancers23. BMP6 also has been shown to inhibit ER –induced mitosis 9; BMP2 has been associated with ER-negative tumors 24. Our results add to the knowledge about the association between BMPs and their receptors and risk of breast cancer defined by hormone receptors. Overall we observed few associations with BMP6 and BMP7 that remained significant after adjustment for multiple comparisons given the size of the genes and number of SNPs assessed. However, rare haplotypes of these genes were associated with breast cancer risk, primarily among women with intermediate and high Native American ancestry. Given associations observed between menopausal status and ER/PR tumor status, we believe that our data support the findings by Takahashi and Wang that showed the importance of BMP receptors and estradiol9, 25. Many of our associations by ER and PR status were observed for type 1 receptors (ACVR1 and ACVRL1). Most SNPs in these genes appeared to be more strongly associated with ER+ tumors than ER− tumors. However, for some SNPs associations were stronger for ER− tumors, and for some the associations were similar for ER− PR− tumors (e.g. ACVR1 rs17182166, ACVRL1 rs11169953, BMP4 rs17563, GDF10 rs762454). Several SNPs were associated with ER− PR+ tumors (BMPR2 rs12621870, and GDF10 rs7093975 and rs2853838).

In our assessment of genetic associations by menopausal status, we observed differences in risk by genetic admixture within menopausal groups. While no associations were observed for BMP7 and ER/PR tumor status, three SNPs were associated with decreased risk of pre-menopausal breast cancer risk among women in the intermediate admixture group and one SNP was associated with post-menopausal breast cancer risk among women with more Native American ancestry. BMP6 rs1225929 was associated with post-menopausal breast cancer among women in the middle admixture group, and BMP6 rs270417 was associated with pre-menopausal breast cancer for those with the most Native ancestry. Most associations found were with type 1 and type 2 BMP receptors, again stressing the importance of receptors in the estrogen-related associations with BMPs 9.

We and others have previously reported that women with higher Native ancestry were at reduced risk of breast cancer11, 26-28. Given this difference in risk, it is of interest to determine whether unique genetic factors are associated differently with breast cancer based on genetic admixture. For the most part we found that associations with BMP-related genes did not differ by genetic admixture. Among those that did, we observed generally stronger risk estimates and more associations with increasing Native American ancestry, with the exception of BMP4 rs17563 which was associated among women with more European ancestry. Others have not examined these genes in similar admixed populations.

The study is the largest to date reporting associations with breast cancer in a genetically admixed population of European and Native American ancestry. We were able to evaluate associations by genetic admixture as well as by ER/PR status of tumors. However, the study has some limitations. Data on tumor characteristics were not available for the entire study population, which restricted our ability to evaluate these characteristics across the same genetic admixture spectrum as we did for the analysis of SNP and breast cancer risk associations. Additionally, we were limited in power to evaluate some tumor phenotypes, for instance there were only 43 cases of ER−PR+ tumors. We evaluated SNPs in several candidate genes. While we hypothesized associations with specific genes, we were limited in our ability to make similar hypotheses regarding specific SNPs. Although we adjusted for multiple comparisons, it is possible that associations are spurious, and thus replication in other studies is needed. Additionally, we have limited information on functionality of these SNPs and our interpretation of findings is greatly guided by the literature on BMPs and their association with cancer in general.

Our findings support the role of genetic variation in BMPs in the etiology of breast cancer. Associations of BMP-related SNPs were in some instances influenced by menopausal status and resulted in associations that were specific to ER and PR status of tumors. Overall, BMP genes were more commonly associated with breast cancer in women with more Native American ancestry, suggesting the importance of genetic ancestry to understanding risk associated with breast cancer. Studies to confirm these findings are needed.

Supplementary Material

Supplement 01-02
Supplement 03-05

Acknowledgments

The Breast Cancer Health Disparities Study was funded by grant CA14002 from the National Cancer Institute to Dr. Slattery. The San Francisco Bay Area Breast Cancer Study was supported by grants CA63446 and CA77305 from the National Cancer Institute, grant DAMD17-96-1-6071 from the U.S. Department of Defense and grant 7PB-0068 from the California Breast Cancer Research Program. The collection of cancer incidence data used in this study was supported by the California Department of Public Health as part of the statewide cancer reporting program mandated by California Health and Safety Code Section 103885; the National Cancer Institute’s Surveillance, Epidemiology and End Results Program under contract HHSN261201000036C awarded to the Cancer Prevention Institute of California; and the Centers for Disease Control and Prevention’s National Program of Cancer Registries, under agreement #1U58 DP000807-01 awarded to the Public Health Institute. The 4-Corner’s Breast Cancer Study was funded by grants CA078682, CA078762, CA078552, and CA078802 from the National Cancer Institute. The research also was supported by the Utah Cancer Registry, which is funded by contract N01-PC-67000 from the National Cancer Institute, with additional support from the State of Utah Department of Health, the New Mexico Tumor Registry, and the Arizona and Colorado cancer registries, funded by the Centers for Disease Control and Prevention National Program of Cancer Registries and additional state support. The contents of this manuscript are solely the responsibility of the authors and do not necessarily represent the official view of the National Cancer Institute or endorsement by the State of California Department of Public Health, the National Cancer Institute, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors. The Mexico Breast Cancer Study was funded by Consejo Nacional de Ciencia y Tecnología (CONACyT) (SALUD-2002-C01-7462). We would also like to acknowledge the contributions of the following individuals to the study: Sandra Edwards for data harmonization oversight; Erica Wolff and Michael Hoffman for laboratory support; Carolyn Ortega for her assistance with data management for the Mexico Breast Cancer Study, Jocelyn Koo for data management for the San Francisco Bay Area Breast Cancer Study. Dr. Tim Byers for his contribution to the 4-Corner’s Breast Cancer Study, Dr. Josh Galanter for assistance in selection of AIMS markers for the study, Dr. Elad Ziv for his input into the study, and Drs. Sue Ingles and Wei Wang for contribution to the study.

References

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 01-02
NIHMS423916-supplement-Supplement_01-02.docx (43.8KB, docx)
Supplement 03-05
NIHMS423916-supplement-Supplement_03-05.docx (119.7KB, docx)

RESOURCES