Differential Transcription Factor Binding Research Articles

A statistical approach for identifying single nucleotide variants that affect transcription factor binding

Non-coding variants located within regulatory elements may alter gene expression by modifying transcription factor (TF) binding sites, thereby leading to functional consequences. Different TF models are being used to assess the effect of DNA sequence variants, such as single nucleotide variants (SNVs). Often existing methods are slow and do not assess statistical significance of results. We investigated the distribution of absolute maximal differential TF binding scores for general computational models that affect TF binding. We find that a modified Laplace distribution can adequately approximate the empirical distributions. A benchmark on invitro and invivo datasets showed that our approach improves upon an existing method in terms of performance and speed. Applications on eQTLs and on a genome-wide association study illustrate the usefulness of our statistics by highlighting cell type-specific regulators and target genes. An implementation of our approach is freely available on GitHub and as bioconda package.

Activator protein transcription factors coordinate human IL-33 expression from noncanonical promoters in chronic airway disease.

IL-33 is a cytokine central to type 2 immune pathology in chronic airway disease. This cytokine is abundantly expressed in the respiratory epithelium and increased in disease, but how expression is regulated is undefined. Here we show that increased IL33 expression occurs from multiple noncanonical promoters in human chronic obstructive pulmonary disease (COPD), and it facilitates production of alternatively spliced isoforms in airway cells. We found that phorbol 12-myristate 13-acetate (PMA) can activate IL33 promoters through protein kinase C in primary airway cells and lines. Transcription factor (TF) binding arrays combined with RNA interference identified activator protein (AP) TFs as regulators of baseline and induced IL33 promoter activity. ATAC-Seq and ChIP-PCR identified chromatin accessibility and differential TF binding as additional control points for transcription from noncanonical promoters. In support of a role for these TFs in COPD pathogenesis, we found that AP-2 (TFAP2A, TFAP2C) and AP-1 (FOS and JUN) family members are upregulated in human COPD specimens. This study implicates integrative and pioneer TFs in regulating IL33 promoters and alternative splicing in human airway basal cells. Our work reveals a potentially novel approach for targeting IL-33 in development of therapeutics for COPD.

Differential Motif and Transcription Factor Binding Predictions Link to ULK4 Gene Polymorphisms and Associated with Essential Hypertension in Caucasian Females

Differential Motif and Transcription Factor Binding Predictions Link to ULK4 Gene Polymorphisms and Associated with Essential Hypertension in Caucasian Females

A long noncoding RNA influences the choice of the X chromosome to be inactivated

X chromosome inactivation (XCI) is the process of silencing one of the X chromosomes in cells of the female mammal which ensures dosage compensation between the sexes. Although theoretically random in somatic tissues, the choice of which X chromosome is chosen to be inactivated can be biased in mice by genetic element(s) associated with the so-called X-controlling element (Xce). Although the Xce was first described and genetically localized nearly 40 y ago, its mode of action remains elusive. In the approach presented here, we identify a single long noncoding RNA (lncRNA) within the Xce locus, Lppnx, which may be the driving factor in the choice of which X chromosome will be inactivated in the developing female mouse embryo. Comparing weak and strong Xce alleles we show that Lppnx modulates the expression of Xist lncRNA, one of the key factors in XCI, by controlling the occupancy of pluripotency factors at Intron1 of Xist. This effect is counteracted by enhanced binding of Rex1 in DxPas34, another key element in XCI regulating the activity of Tsix lncRNA, the main antagonist of Xist, in the strong but not in the weak Xce allele. These results suggest that the different susceptibility for XCI observed in weak and strong Xce alleles results from differential transcription factor binding of Xist Intron 1 and DxPas34, and that Lppnx represents a decisive factor in explaining the action of the Xce.

Allelic imbalance of chromatin accessibility in cancer identifies candidate causal risk variants and their mechanisms.

While many germline cancer risk variants have been identified through genome-wide association studies (GWAS), the mechanisms by which these variants operate remain largely unknown. Here we used 406 cancer ATAC-Seq samples across 23 cancer types to identify 7,262 germline allele-specific accessibility QTLs (as-aQTLs). Cancer as-aQTLs had stronger enrichment for cancer risk heritability (up to 145 fold) than any other functional annotation across seven cancer GWAS. Most cancer as-aQTLs directly altered transcription factor (TF) motifs and exhibited differential TF binding and gene expression in functional screens. To connect as-aQTLs to putative risk mechanisms, we introduced the regulome-wide associations study (RWAS). RWAS identified genetically associated accessible peaks at >70% of known breast and prostate loci and discovered new risk loci in all examined cancer types. Integrating as-aQTL discovery, motif analysis and RWAS identified candidate causal regulatory elements and their probable upstream regulators. Our work establishes cancer as-aQTLs and RWAS analysis as powerful tools to study the genetic architecture of cancer risk.

Differential expression and transcription factor binding associated with genotype at a pharmacogenetic variant in OPRD1

ABSTRACT Background The functional mechanism is unknown for many genetic variants associated with substance use disorder phenotypes. Rs678849, an intronic variant in the delta-opioid receptor gene (OPRD1), has been found to predict regional brain volume, addiction risk, and the efficacy of buprenorphine/naloxone in treating opioid use disorder. The variant has also been implicated as an expression quantitative trait locus (eQTL) for several genes. Objectives The objective of this study was to identify functional differences between the two alleles of rs678849 in vitro. We hypothesized that the two alleles of rs678849 would have different effects on transcriptional activity due to differential interactions with transcription factors. Methods 15bp regions containing the C or T alleles of rs678849 were cloned into luciferase constructs and transfected into BE(2)C neuroblastoma cells to test the effect on transcription. Electrophoretic mobility shift assays (EMSA) using nuclear lysates from BE(2)C cell or human postmortem medial prefrontal cortex were used to identify proteins that differentially bound the two alleles. Results At 24 hours post-transfection, the C allele construct had significantly lower luciferase expression than the T allele construct and empty vector control (ANOVA p < .001). Proteomic analysis and supershift assays identified XRCC6 as a transcription factor specifically binding the C allele, whereas hnRNP D0 was found to specifically bind the T allele. Conclusion These functional differences between the C and T alleles may help explain the psychiatric and neurological phenotype differences predicted by rs678849 genotype and the potential role of the variant as an eQTL.

Landscape of allele-specific transcription factor binding in the human genome

Sequence variants in gene regulatory regions alter gene expression and contribute to phenotypes of individual cells and the whole organism, including disease susceptibility and progression. Single-nucleotide variants in enhancers or promoters may affect gene transcription by altering transcription factor binding sites. Differential transcription factor binding in heterozygous genomic loci provides a natural source of information on such regulatory variants. We present a novel approach to call the allele-specific transcription factor binding events at single-nucleotide variants in ChIP-Seq data, taking into account the joint contribution of aneuploidy and local copy number variation, that is estimated directly from variant calls. We have conducted a meta-analysis of more than 7 thousand ChIP-Seq experiments and assembled the database of allele-specific binding events listing more than half a million entries at nearly 270 thousand single-nucleotide polymorphisms for several hundred human transcription factors and cell types. These polymorphisms are enriched for associations with phenotypes of medical relevance and often overlap eQTLs, making candidates for causality by linking variants with molecular mechanisms. Specifically, there is a special class of switching sites, where different transcription factors preferably bind alternative alleles, thus revealing allele-specific rewiring of molecular circuitry.

4277 Functional consequences of the juvenile idiopathic arthritis risk variant at 1q24.3

OBJECTIVES/GOALS: Juvenile idiopathic arthritis (JIA) is the most common childhood rheumatologic disease childhood and a cause of pain and potential disability. JIA has a strong genetic component and no known cure. The goal of this study is to evaluate allele-dependent effects of a novel JIA risk variant at 1q24.3. METHODS/STUDY POPULATION: JIA patients meeting criteria for the two most common disease subtypes (oligoarticular and RF neg polyarthritis) were genotyped using the Immunochip, an Illumina array with dense coverage of the HLA region and 186 other loci previously reported in autoimmune diseases. Phase I association findings (Hinks, 2013) and Phase II analysis (unpublished) of an expanded cohort (4,271 JIA and 14,390 controls) identified new risk loci, including rs78037977 at 1q24.3. We prioritized rs78037977 and predicted possible impacted mechanisms based on Bayesian predictions of attributable risk, the surrounding chromatin landscape, and transcription factor binding data. A luciferase reporter assay was used to assess allele-dependent enhancer activity. RESULTS/ANTICIPATED RESULTS: rs78037977 is located between FASLG and TNFSF18 at chromosome 1q24.3 is associated with JIA (p = 6.3x10−09), and explains 94% of the posterior probability at this locus; no other SNPs in linkage disequilibrium (r2&gt;0.6). The chromatin landscape around rs78037977 contains H3K4Me1 and H3K27Ac marks, which are indicative of enhancer activity. Further, &gt;160 transcription factors have chromatin immunoprecipitation followed by sequencing (ChIP-seq) peaks overlapping rs78037977 in various cellular contexts. In luciferase reporter assays, the region around rs78037977 containing the reference A allele had ~2-fold increased enhancer activity compared to the non-reference allele. DISCUSSION/SIGNIFICANCE OF IMPACT: This work provides in vitro evidence to support allele-dependent enhancer activity of a novel JIA-risk variant at 1q24.3. Our ongoing work investigates the effect of the DNA-containing region of rs78037977 on gene expression and differential transcription factor binding at rs78037977.

Gene targets for engineering osmotolerance in Caldicellulosiruptor bescii

BackgroundCaldicellulosiruptor bescii, a promising biocatalyst being developed for use in consolidated bioprocessing of lignocellulosic materials to ethanol, grows poorly and has reduced conversion at elevated medium osmolarities. Increasing tolerance to elevated fermentation osmolarities is desired to enable performance necessary of a consolidated bioprocessing (CBP) biocatalyst.ResultsTwo strains of C. bescii showing growth phenotypes in elevated osmolarity conditions were identified. The first strain, ORCB001, carried a deletion of the FapR fatty acid biosynthesis and malonyl-CoA metabolism repressor and had a severe growth defect when grown in high-osmolarity conditions—introduced as the addition of either ethanol, NaCl, glycerol, or glucose to growth media. The second strain, ORCB002, displayed a growth rate over three times higher than its genetic parent when grown in high-osmolarity medium. Unexpectedly, a genetic complement ORCB002 exhibited improved growth, failing to revert the observed phenotype, and suggesting that mutations other than the deleted transcription factor (the fruR/cra gene) are responsible for the growth phenotype observed in ORCB002. Genome resequencing identified several other genomic alterations (three deleted regions, three substitution mutations, one silent mutation, and one frameshift mutation), which may be responsible for the observed increase in osmolarity tolerance in the fruR/cra-deficient strain, including a substitution mutation in dnaK, a gene previously implicated in osmoresistance in bacteria. Differential expression analysis and transcription factor binding site inference indicates that FapR negatively regulates malonyl-CoA and fatty acid biosynthesis, as it does in many other bacteria. FruR/Cra regulates neighboring fructose metabolism genes, as well as other genes in global manner.ConclusionsTwo systems able to effect tolerance to elevated osmolarities in C. bescii are identified. The first is fatty acid biosynthesis. The other is likely the result of one or more unintended, secondary mutations present in another transcription factor deletion strain. Though the locus/loci and mechanism(s) responsible remain unknown, candidate mutations are identified, including a mutation in the dnaK chaperone coding sequence. These results illustrate both the promise of targeted regulatory manipulation for osmotolerance (in the case of fapR) and the challenges (in the case of fruR/cra).

Secondary Wall Regulating NACs Differentially Bind at the Promoter at a CELLULOSE SYNTHASE A4 Cis-eQTL.

Arabidopsis thaliana CELLULOSE SYNTHASE A4/7/8 (CESA4/7/8) are three non-redundant subunits of the secondary cell wall cellulose synthase complex. Transcript abundance of these genes can vary among genotypes and expression quantitative trait loci (eQTL) were identified in a recombinant population of the accessions Bay-0 and Shahdara. Genetic mapping and analysis of the transcript levels of CESAs between two distinct near isogenic lines (NILs) confirmed a change in CESA4 expression that segregates within that interval. We sequenced the promoters and identified 16 polymorphisms differentiating CESA4Sha and CESA4Bay. In order to determine which of these SNPs could be responsible for this eQTL, we screened for transcription factor protein affinity with promoter fragments of CESA4Bay, CESA4Sha, and the reference genome CESA4Col. The wall thickening activator proteins NAC SECONDARY WALL THICKENING PROMOTING FACTOR2 (NST2) and NST3 exhibited a decrease in binding with the CESA4Sha promoter with a tracheary element-regulating cis-element (TERE) polymorphism. While NILs harboring the TERE polymorphisms exhibited significantly different CESA4 expression, cellulose crystallinity and cell wall thickness were indistinguishable. These results suggest that the TERE polymorphism resulted in differential transcription factor binding and CESA4 expression; yet A. thaliana is able to tolerate this transcriptional variability without compromising the structural elements of the plant, providing insight into the elasticity of gene regulation as it pertains to cell wall biosynthesis and regulation. We also explored available DNA affinity purification sequencing data to resolve a core binding site, C(G/T)TNNNNNNNA(A/C)G, for secondary wall NACs referred to as the VNS element.

Genomic meta-analysis of the interplay between 3D chromatin organization and gene expression programs under basal and stress conditions

BackgroundOur appreciation of the critical role of the genome’s 3D organization in gene regulation is steadily increasing. Recent 3C-based deep sequencing techniques elucidated a hierarchy of structures that underlie the spatial organization of the genome in the nucleus. At the top of this hierarchical organization are chromosomal territories and the megabase-scale A/B compartments that correlate with transcriptional activity within cells. Below them are the relatively cell-type-invariant topologically associated domains (TADs), characterized by high frequency of physical contacts between loci within the same TAD, and are assumed to function as regulatory units. Within TADs, chromatin loops bring enhancers and target promoters to close spatial proximity. Yet, we still have only rudimentary understanding how differences in chromatin organization between different cell types affect cell-type-specific gene expression programs that are executed under basal and challenged conditions.ResultsHere, we carried out a large-scale meta-analysis that integrated Hi–C data from thirteen different cell lines and dozens of ChIP-seq and RNA-seq datasets measured on these cells, either under basal conditions or after treatment. Pairwise comparisons between cell lines demonstrate a strong association between modulation of A/B compartmentalization, differential gene expression and transcription factor (TF) binding events. Furthermore, integrating the analysis of transcriptomes of different cell lines in response to various challenges, we show that A/B compartmentalization of cells under basal conditions significantly correlates not only with gene expression programs and TF binding profiles that are active under the basal condition but also with those induced in response to treatment. Yet, in pairwise comparisons between different cell lines, we find that a large portion of differential TF binding and gene induction events occur in genomic loci assigned to A compartment in both cell types, underscoring the role of additional critical factors in determining cell-type-specific transcriptional programs.ConclusionsOur results further indicate the role of dynamic genome organization in regulation of differential gene expression between different cell types and the impact of intra-TAD enhancer–promoter interactions that are established under basal conditions on both the basal and treatment-induced gene expression programs.

TGFβ3, MSX1, and MMP3 as Candidates for NSCL±P in an Indian Population.

To evaluate the association of transforming growth factor β3 ( TGFβ3), muscle segment homeobox 1 ( MSX1), Metalloproteinases 3 ( MMP3), and MMP9 genes as candidates for nonsyndromic cleft lip and/or palate in an Indian population. Case-control association study, mutational screening, and functional evaluation of obtained mutations. Mutational screening of the developmental genes, TGFβ3 and MSX1, along with functional evaluation and association of promoter region SNPs-one each in MMP3 and MMP9. Two hundred forty five NSCL±P cases from G. S. Memorial Plastic Surgery Hospital and Trauma Center, Varanasi and 201 healthy controls without a family history of congenital malformations from nearby schools, primary health centers, and the university hospital. Sequencing, SSCP, and PCR-RFLP were used for candidate gene screening. MatInspector and electrophoretic mobility shift assay (EMSA) were used to check the differential transcription factor binding of the variants at promoter region. Luciferase assay was used to test the transcriptional potential of the variant, and evaluation of the alternative splice site was carried out using exon-trapping experiment. Metalloproteinases3 -1171 5A/6A was associated with NSCL±P, whereas MMP9 -1562 C/T did not show association. A rare variant in the promoter region of TGFβ3 (rs117462711) creates a differential binding site, confirmed by EMSA. Luciferase assay showed 3.7-fold increased expression level in mutant construct. A synonymous change in MSX1 (rs34165410) showed association with NSCL±P, which may create an alternative splice site or lead to low codon usage. Exon-trapping experiment failed to confirm alternative splicing, indicating low codon usage frequency of the mutant affecting the gene function. TGFβ3, MSX1, and MMP3 are candidates for NSCL±P.

Parallel factor ChIP provides essential internal control for quantitative differential ChIP-seq.

A key challenge in quantitative ChIP combined with high-throughput sequencing (ChIP-seq) is the normalization of data in the presence of genome-wide changes in occupancy. Analysis-based normalization methods were developed for transcriptomic data and these are dependent on the underlying assumption that total transcription does not change between conditions. For genome-wide changes in transcription factor (TF) binding, these assumptions do not hold true. The challenges in normalization are confounded by experimental variability during sample preparation, processing and recovery. We present a novel normalization strategy utilizing an internal standard of unchanged peaks for reference. Our method can be readily applied to monitor genome-wide changes by ChIP-seq that are otherwise lost or misrepresented through analytical normalization. We compare our approach to normalization by total read depth and two alternative methods that utilize external experimental controls to study TF binding. We successfully resolve the key challenges in quantitative ChIP-seq analysis and demonstrate its application by monitoring the loss of Estrogen Receptor-alpha (ER) binding upon fulvestrant treatment, ER binding in response to estrodiol, ER mediated change in H4K12 acetylation and profiling ER binding in patient-derived xenographs. This is supported by an adaptable pipeline to normalize and quantify differential TF binding genome-wide and generate metrics for differential binding at individual sites.

DNase-capture reveals differential transcription factor binding modalities.

We describe DNase-capture, an assay that increases the analytical resolution of DNase-seq by focusing its sequencing phase on selected genomic regions. We introduce a new method to compensate for capture bias called BaseNormal that allows for accurate recovery of transcription factor protection profiles from DNase-capture data. We show that these normalized data allow for nuanced detection of transcription factor binding heterogeneity with as few as dozens of sites.

P4471The 4q32.1 coronary artery disease locus influences disease risk via differential transcription factor binding

P4471The 4q32.1 coronary artery disease locus influences disease risk via differential transcription factor binding

A Bayesian Approach for Analysis of Whole-Genome Bisulfite Sequencing Data Identifies Disease-Associated Changes in DNA Methylation.

DNA methylation is a key epigenetic modification involved in gene regulation whose contribution to disease susceptibility remains to be fully understood. Here, we present a novel Bayesian smoothing approach (called ABBA) to detect differentially methylated regions (DMRs) from whole-genome bisulfite sequencing (WGBS). We also show how this approach can be leveraged to identify disease-associated changes in DNA methylation, suggesting mechanisms through which these alterations might affect disease. From a data modeling perspective, ABBA has the distinctive feature of automatically adapting to different correlation structures in CpG methylation levels across the genome while taking into account the distance between CpG sites as a covariate. Our simulation study shows that ABBA has greater power to detect DMRs than existing methods, providing an accurate identification of DMRs in the large majority of simulated cases. To empirically demonstrate the method's efficacy in generating biological hypotheses, we performed WGBS of primary macrophages derived from an experimental rat system of glomerulonephritis and used ABBA to identify >1000 disease-associated DMRs. Investigation of these DMRs revealed differential DNA methylation localized to a 600bp region in the promoter of the Ifitm3 gene. This was confirmed by ChIP-seq and RNA-seq analyses, showing differential transcription factor binding at the Ifitm3 promoter by JunD (an established determinant of glomerulonephritis), and a consistent change in Ifitm3 expression. Our ABBA analysis allowed us to propose a new role for Ifitm3 in the pathogenesis of glomerulonephritis via a mechanism involving promoter hypermethylation that is associated with Ifitm3 repression in the rat strain susceptible to glomerulonephritis.

Genetic prevention of hepatitis C virus-induced liver fibrosis by allele-specific downregulation of MERTK.

Infection by hepatitis C virus (HCV) can result in the development of liver fibrosis and may eventually progress into cirrhosis and hepatocellular carcinoma. However, the molecular mechanisms for this process are not fully known. Several genome-wide association studies have been carried out to pinpoint causative variants in HCV-infected patient cohorts, but these variants are usually not the functional ones. The aim of this study was to identify the regulatory single nucleotide polymorphism associated with the risk of HCV-induced liver fibrosis and elucidate its molecular mechanism. We utilized a bioinformatics approach to identify a non-coding regulatory variant, located in an intron of the MERTK gene, based on differential transcription factor binding between the alleles. We validated the results using expression reporter assays and electrophoresis mobility shift assays. Chromatin immunoprecipitation sequencing indicated that transcription factor(s) bind stronger to the A allele of rs6726639. Electrophoresis mobility shift assays supported these findings and suggested that the transcription factor is interferon regulatory factor 1 (IRF1). Luciferase report assays showed lower enhancer activity from the A allele and that IRF1 may act as a repressor. Treatment of hepatitis C with interferon-α results in increased IRF1 levels and our data suggest that this leads to an allele-specific downregulation of MERTK mediated by an allelic effect on the regulatory element containing the functional rs6726639. This variant also shows the hallmarks for being the driver of the genome-wide association studies for reduced risk of liver fibrosis and non-alcoholic fatty liver disease at MERTK.

High-Fat Diet During Mouse Pregnancy and Lactation Targets GIP-Regulated Metabolic Pathways in Adult Male Offspring.

Maternal obesity is a worldwide problem associated with increased risk of metabolic diseases in the offspring. Genetic deletion of the gastric inhibitory polypeptide (GIP) receptor (GIPR) prevents high-fat diet (HFD)-induced obesity in mice due to specific changes in energy and fat cell metabolism. We investigated whether GIP-associated pathways may be targeted by fetal programming and mimicked the situation by exposing pregnant mice to control or HFD during pregnancy (intrauterine [IU]) and lactation (L). Male wild-type (WT) and Gipr(-/-) offspring received control chow until 25 weeks of age followed by 20 weeks of HFD. Gipr(-/-) offspring of mice exposed to HFD during IU/L became insulin resistant and obese and exhibited increased adipose tissue inflammation and decreased peripheral tissue substrate utilization after being reintroduced to HFD, similar to WT mice on regular chow during IU/L. They showed decreased hypothalamic insulin sensitivity compared with Gipr(-/-) mice on control diet during IU/L. DNA methylation analysis revealed increased methylation of CpG dinucleotides and differential transcription factor binding of promoter regions of genes involved in lipid oxidation in the muscle of Gipr(-/-) offspring on HFD during IU/L, which were inversely correlated with gene expression levels. Our data identify GIP-regulated metabolic pathways that are targeted by fetal programming.

Genetic predisposition to neuroblastoma mediated by a LMO1 super-enhancer polymorphism.

Neuroblastoma is a paediatric malignancy that typically arises in early childhood, and is derived from the developing sympathetic nervous system. Clinical phenotypes range from localized tumours with excellent outcomes to widely metastatic disease in which long-term survival is approximately 40% despite intensive therapy. A previous genome-wide association study identified common polymorphisms at the LMO1 gene locus that are highly associated with neuroblastoma susceptibility and oncogenic addiction to LMO1 in the tumour cells. Here we investigate the causal DNA variant at this locus and the mechanism by which it leads to neuroblastoma tumorigenesis. We first imputed all possible genotypes across the LMO1 locus and then mapped highly associated single nucleotide polymorphism (SNPs) to areas of chromatin accessibility, evolutionary conservation and transcription factor binding sites. We show that SNP rs2168101 G>T is the most highly associated variant (combined P = 7.47 × 10(-29), odds ratio 0.65, 95% confidence interval 0.60-0.70), and resides in a super-enhancer defined by extensive acetylation of histone H3 lysine 27 within the first intron of LMO1. The ancestral G allele that is associated with tumour formation resides in a conserved GATA transcription factor binding motif. We show that the newly evolved protective TATA allele is associated with decreased total LMO1 expression (P = 0.028) in neuroblastoma primary tumours, and ablates GATA3 binding (P < 0.0001). We demonstrate allelic imbalance favouring the G-containing strand in tumours heterozygous for this SNP, as demonstrated both by RNA sequencing (P < 0.0001) and reporter assays (P = 0.002). These findings indicate that a recently evolved polymorphism within a super-enhancer element in the first intron of LMO1 influences neuroblastoma susceptibility through differential GATA transcription factor binding and direct modulation of LMO1 expression in cis, and this leads to an oncogenic dependency in tumour cells.

Detection of functional polymorphisms influencing the promoter activity of the SAA2 gene and their association with milk production traits in Chinese Holstein cows.

Our previous RNA sequencing experiment showed that the serum amyloid A2 (SAA2) gene was one of the most promising candidates for milk protein and fat traits in dairy cattle. The SAA2 gene encodes an apolipoprotein related to high-density lipoproteins. To further validate its genetic effects, genotype-phenotype associations were performed in this study. Through resequencing of the entire coding region and the 5'-regulatory region of the SAA2 gene using pooled DNA of 12 unrelated sires, one novel 3-bp insertion-deletion and five previously reported SNPs were detected. These identified SNPs were genotyped and tested for association with five milk production-related traits in 717 Chinese Holstein cows. After Bonferroni correction for multiple t-tests, five of them were found to be statistically significant for milk yield, fat yield and protein yield (P<0.0001~0.0053). Haplotype-based association analysis revealed a similar effect on fat yield and protein yield (P=0.0005, P=0.0032 respectively). Then, using luciferase report assay, the regulatory effect of the three SNPs located in the promoter region (c.-22G>A; c.17G>C; c.114G>A) was evaluated on transcriptional activity. In HEK-293 cell lines, we found that constructs GCG and AGG showed higher luciferase activity compared with GCA (P<0.01, P<0.01 respectively). Meanwhile, the prediction of the putative differential transcription factor binding site revealed that c.17G>C and c.114G>A caused the alteration in the transcription factor. Overall, the findings presented here provide the first evidence for associations of the SAA2 gene with milk fat and protein traits, which appears to be a key candidate for milk production traits in dairy cattle.