Circadian genes have the potential to influence a variety of cancer-related biological pathways, including immunoregulation, which may influence susceptibility to non–Hodgkin's lymphoma (NHL). However, few studies have examined the role of circadian genes in lymphomagenesis. The current study examined Cryptochrome 2 (CRY2), a core circadian gene and transcriptional repressor, as a potential circadian biomarker for NHL. We first performed genetic association analyses of tagging single nucleotide polymorphisms (SNP) in CRY2 and NHL risk using DNA samples from a population-based case-control study (n = 455 cases and 527 controls). Three SNPs were found to be significantly associated with risk of NHL when combining all subtypes [dbSNP IDs, odds ratios (ORs), and 95% confidence intervals: rs11038689, OR, 2.34 (1.28–4.27), P = 0.006; rs7123390, OR, 2.40 (1.39–4.13), P = 0.002; and rs1401417, OR, 2.97 (1.57–5.63, P = 0.001)]. Each of these associations remained significant when restricting the analysis to B-cell cases and when further restricting to follicular lymphomas. An analysis of CRY2 diplotypes confirmed these significant findings. To further determine the functional effect of CRY2, we silenced the gene in vitro and performed a whole genome expression microarray. A pathway-based analysis showed that genes significantly altered by CRY2 knockdown formed networks associated with immune response and hematologic system development. In addition, these genes were predicted to have significant effects on several disease processes, including cancer (B-H P = 3.75E−9) and hematologic disease (B-H P = 8.01E−8). In conclusion, both genetic association and functional analyses suggest that the circadian gene CRY2 may play an important role in NHL development. [Cancer Res 2009;69(8):3605–13]

The human circadian rhythm is a fundamental aspect of human physiology, and a wide range of biological processes are influenced by the circadian clock, including body temperature, energy metabolism, hormone secretion, and sleep-wake cycles (1). Several observational studies indicate that individuals who do not maintain a normal sleep/wake cycle may be at increased risk for several cancer types, and after considering evidence from epidemiologic and experimental studies, the IARC recently concluded that shift work that involves circadian disruption is “probably carcinogenic to humans” (2). Although much of the current epidemiologic evidence has focused on breast cancer, it has recently been hypothesized that circadian disruption may also affect risk for non–Hodgkin's lymphoma (NHL), possibly through its influence on immunoregulation (3).

Limited indirect evidence suggests that genetic components of the circadian system may have a role in processes relevant for NHL tumorigenesis. For example, reduced expression of the circadian gene PER2 has been detected in lymphoma cell lines and in samples drawn from patients with acute myeloid leukemia (4), and a recent genetic association study showed that a nonsynonymous polymorphism in the core circadian gene NPAS2 is associated with decreased risk of NHL, especially B-cell lymphoma (5). In addition, several studies have established an important role for circadian rhythm in the maintenance of proper immune function. First, it has been shown that several key components of the immune system are under circadian regulation, with circadian rhythmicity present in nearly all aspects of immune response (611). Specifically, circadian rhythms have been observed in natural killer (NK) cells, which are an essential component of the innate immune system against infections and cancerous growth (12). Second, disruption of circadian rhythms can cause aberrant immune cell trafficking and abnormal cell proliferation cycles (13). Moreover, disruption of the circadian rhythms in NK cells and phagocytic activity has been observed in malignant melanoma cells, leading to a discoordination between the two immune system components that is not observed in healthy humans (14).

Overall, these preliminary studies suggest that circadian disruption has the potential to significantly affect a number of mechanisms that may influence NHL susceptibility, most notably through its role in influencing immune response. However, although immune dysfunction remains the only well-established risk factor for NHL (15, 16), immunodeficiency is seen only in a subset of NHL patients. As such, if an association between circadian disruption and lymphomagenesis can be firmly established, there remains the additional question of whether the relationship is maintained outside of pathways related to immune system function. Further study into these associations is therefore warranted and are apropos to investigations into the potential for circadian gene variants to serve as a novel panel of NHL risk biomarkers.

The current study investigates the role of the core circadian gene CRY2 in NHL tumorigenesis. CRY2 is essential to the maintenance of circadian rhythm through its role in the negative arm of the circadian feedback loop, and may have a broader regulatory role as a transcriptional repressor (17, 18). CRY2 has also been shown to be involved in cell cycle regulation, including roles in DNA damage checkpoint control (19) and regulation of genes important for cell cycle progression (20). Here, we report findings from an epidemiologic analysis of the association between genetic variants in CRY2 and risk of NHL. In addition, we performed a whole genome expression microarray to determine the effect of CRY2 silencing on the expression of cancer-related genes, and to determine whether CRY2 influences biological pathways that may be relevant for lymphomagenesis.

Case-control study of NHL. The study population has previously been described (21). Briefly, all participants were female residents of Connecticut, and cases were incident, histologically confirmed NHL (ICD-O, M-9590-9642, 9690-9701, 9740-9750) identified through Yale Cancer Center's Rapid Case Ascertainment between 1996 and 2000. Population-based controls younger than age 65 y were recruited by random digit dialing, and controls older than age 65 y were identified through Health Care Financing Administration files. Five-year age strata were constructed, and controls were frequency matched to cases by intermittently adjusting the number of controls selected from each stratum. Participation rates were as follows: 72% for cases, 69% for random digit dialing controls, and 47% for controls identified by health care financing records.

Data collection. The study was approved by Institutional Review Boards at Yale University, the Connecticut Department of Public Health, and the National Cancer Institute. Participation was voluntary, and written informed consent was obtained. Those who agreed were interviewed by trained study nurses either at the subject's home or at a convenient location, and after the administration of a questionnaire, subjects provided a 10-mL peripheral blood sample. Genomic DNA was isolated from peripheral blood lymphocytes for each study subject.

Single nucleotide polymorphism selection and genotyping. Single nucleotide polymorphisms (SNP) were identified using the Tagger algorithm (22), which is implemented in the Haploview interface (23) of HapMap's genome browser, Release 22.3

Five SNPs (rs10838524, rs11038689, rs11605924, rs2292912, and rs7123390) were identified as representative of all variations found within the exonic and intronic regions of the CRY2 gene using the CEU population returning SNPs with minor allele frequency of ≥0.2 and r2 of ≥0.8. In addition, one intronic SNP (rs1401417) that had been identified as significantly associated with prostate cancer risk in a previous study (24) was also included in the genotyping pool. Genotyping for all SNPs was performed at Yale University's W.M. Keck Foundation Biotechnology Research Laboratory using the Sequenom MassARRAY multiplex genotyping platform (Sequenom, Inc.) according to the manufacturer's protocol. Duplicate samples from 100 study subjects and 40 replicate samples from each of 2 blood donors were interspersed throughout each batch for all genotyping assays. Genotyping failed for one SNP (rs10838524). The concordance rates for QC samples were over 95% for the remaining assays. Genotyping call rates were as follows: 97.0% for rs11038689, 97.5% for rs11605924, 97.7% for rs2292912, 96.6% for rs7123390, and 97.8% for rs1401417. All genotyping scores, including quality control data, were rechecked by different laboratory personnel and the accuracy of each assay was confirmed.

Cell culture and treatments. Human breast adenocarcinoma cells (MCF-7) were used to determine the effect of CRY2 knockdown on pathways related to lymphomagenesis. MCF-7 cells were chosen rather than cells derived from lymphoma tissue, as lymphoma cells would likely begin with aberrant immune signaling, thus causing difficulty in interpreting the effects of CRY2 knockdown on immunoregulation. Cells were obtained from American Type Culture Collection and were maintained in DMEM (Invitrogen) supplemented with 10% fetal bovine serum (Invitrogen), 0.01 mg/mL bovine insulin, and 1% penicillin/streptomycin (Sigma-Aldrich). siRNA oligos targeting CRY2 (Sense, 5′-UGCUUCAUUCGUUCAAUGUUAAGCCGG-3′; Antisense, 5′-GGCUUAACAUUGAACGAAUGAAGCA-3′) and a scrambled sequence negative control siRNA (Sense, 5′-CUUCCUCUCUUUCUCUCCCUUGUGA-3′; Antisense, 5′-UCACAAGGGAGAGAAAGAGGGAAGGA-3′) were designed and manufactured by Integrated DNA Technologies. Each oligo was diluted in OPTI-MEM serum-free medium (Invitrogen), complexed with Lipofectamine RNAiMax transfection reagent (Invitrogen), and reverse transfected with ∼100,000 cells in 12-well plates at a final concentration of 10 nmol/L in growth medium without penicillin/streptomycin. Cells were harvested 48 h after transfection for subsequent analysis.

RNA isolation and quantitation.CRY2 silencing efficiency was determined by qPCR of RNA samples isolated using the RNA Mini kit (Qiagen), with on-column DNA digestion, according to the manufacturer's instructions for mammalian cells. First-strand cDNA was synthesized from purified RNA using the AffinityScript cDNA kit (Stratagene) with oligo-dT primers. Quantitative real-time PCR conditions were prepared using a SYBR Green PCR master mix (Applied BioSystems) with gene-specific primers, and a standard thermal cycling procedure on an ABI 7500 instrument (Applied BioSystems). The primers used for CRY2 amplification were as follows: L, ACCGGGGACTCTGTCTACTG; R, GCCTGCACTGCTCATGCT. CRY2 knockdown was assessed using the 2−ΔΔCt method with RNA content normalized to the housekeeping gene HPRT1.

Gene expression microarray and pathway analysis. Gene expression differences in CRY2 knockdown and normal cells were interrogated by whole genome microarray (Agilent, Inc., 44k chip, performed by MoGene). RNA was isolated in duplicate (biological replicates) from cells treated with CRY2-targeting or scrambled negative siRNA. Transcripts were identified as “differentially expressed” if they fit the criteria of Benjamini-Hochberg–adjusted P value of <0.01 in both biological replicates and mean fold change of >|2|. Transcripts with differential expression were investigated for network and functional interrelatedness using the Ingenuity Pathway Analysis software tool (Ingenuity Systems).4

This software scans the set of input genes to identify networks using information in the Ingenuity Pathways Knowledge Base, an extensive, manually curated database of functional interactions extracted from peer-reviewed publications (25). A Fisher's exact test, based on the hypergeometric distribution, is then performed to determine the likelihood of obtaining at least the same number of molecules by chance (i.e., from a random input set), as actually overlap between the input gene set and the genes present in each identified network. All microarray data have been uploaded to the Gene Expression Omnibus database. Raw and processed data related to this experiment can be accessed via their Web site5 by referencing accession #GSE14617.

Statistical analysis. All statistical analyses were performed using the SAS statistical software (SAS Institute), unless otherwise noted. For the case-control analyses, allelic distributions for all SNPs were tested by goodness-of-fit χ2 for compliance with Hardy-Weinberg equilibrium. A χ2 test was also used to determine whether any of the variants were associated with case-control status, using either the full table (homozygous wild-type, heterozygous, and homozygous variant), a dominant model, or a recessive model. Odds ratios (OR) and 95% confidence intervals (CI) were determined for each SNP disease association by unconditional multivariate logistic regression, including the following covariates: age, race, and family history of cancer in a first-degree relative. Haplotype estimates were calculated by the PHASE program, which reconstructs haplotypes from population genotyping data (26). Diplotypes were constructed based on the pair of haplotypes estimated for each individual, and ORs and 95% CIs for each diplotype were determined by unconditional multivariate logistic regression, using the same covariates as the main effects model but with all other diplotypes as the reference category.

Due to the multiple comparisons inherent in the microarray analysis, adjustments were made to control for false discoveries. A multiple comparisons correction was applied to each observation using the Benjamini-Hochberg method, as previously described (27), to obtain an adjusted P value (B-H P value).

Association between CRY2 variants and NHL risk. Compared with controls, NHL cases reported a higher proportion of NHL and other cancers among first-degree relatives. There were no significant differences in age and race between cases and controls (Table 1). Genotypic frequencies were determined at each locus, and no allelic distributions significantly departed from the values expected under Hardy-Weinberg equilibrium among the controls (P < 0.01). Three of the five SNPs, rs11038689, rs7123390, and rs1401417, were found to be significantly associated with case-control status in the overall three-by-two table of allele distribution versus case-control status (Supplementary Table S1), with P values of 0.005, 0.003, and 0.004, respectively, for the 2 degrees of freedom χ2 test. Each of the three SNPs significantly associated with disease status in the full table was also significant when a recessive model was assumed but not when assuming a dominant model. As such, ORs and CIs were determined by unconditional multivariate logistic regression under the assumption of a recessive model. Variant alleles in the same three SNPs (rs11038689, rs7123390, and rs1401417) that were associated with case-control status in the unadjusted χ2 analysis were also significantly associated with increased risk of NHL in the adjusted logistic regression model, with ORs (95% CIs) of 2.34 (1.28–4.27; P = 0.006), 2.40 (1.39–4.13; P = 0.002), and 2.97 (1.57–5.63; P = 0.001), respectively (Table 2).

Table 1.

Distribution of selected characteristics by case-control status

VariableCases (n = 455)
Controls (n = 527)
P
n (%)n (%)
Mean age (y) 61.88 62.34 0.607 
Race    
Caucasian 438 (96.26) 496 (94.12)  
African-American 13 (2.86) 14 (2.66)  
Other 4 (0.88) 17 (3.22) 0.108 
Family history of cancer in first degree relatives    
No 96 (21.10) 130 (24.67)  
NHL 9 (1.98) 2 (0.38)  
Other cancer 350 (76.92) 395 (74.95) 0.030 
Case pathology    
All B cell 365 (80.22)   
    Diffuse large B cell 135 (36.99)   
    Follicular 105 (28.77)   
    SLL/CLL 54 (14.79)   
    Marginal Zone 30 (8.22)   
    Other 41 (11.23)   
All T cell 33 (7.25)   
NOS 58 (12.75)   
VariableCases (n = 455)
Controls (n = 527)
P
n (%)n (%)
Mean age (y) 61.88 62.34 0.607 
Race    
Caucasian 438 (96.26) 496 (94.12)  
African-American 13 (2.86) 14 (2.66)  
Other 4 (0.88) 17 (3.22) 0.108 
Family history of cancer in first degree relatives    
No 96 (21.10) 130 (24.67)  
NHL 9 (1.98) 2 (0.38)  
Other cancer 350 (76.92) 395 (74.95) 0.030 
Case pathology    
All B cell 365 (80.22)   
    Diffuse large B cell 135 (36.99)   
    Follicular 105 (28.77)   
    SLL/CLL 54 (14.79)   
    Marginal Zone 30 (8.22)   
    Other 41 (11.23)   
All T cell 33 (7.25)   
NOS 58 (12.75)   

Abbreviation: NOS, not otherwise specified.

Table 2.

Association of CRY2 variants with NHL risk

GenotypeAll
B cell
T cell
Cases nControls nOR* (95% CI)Cases nControls nOR* (95% CI)Cases nControls nOR* (95% CI)
rs11038689          
    A/A or A/G 408 495 — 326 495 — 29 495 — 
    G/G 33 17 2.34 (1.28–4.27) 26 17 2.35 (1.25–4.41) 17 2.69 (0.74–9.84) 
    P — — 0.006 — — 0.008 — — 0.135 
rs11605924          
    A/A or A/C 327 365 — 263 365 — 25 365 — 
    C/C 114 151 0.82 (0.62–1.10) 89 151 0.79 (0.58–1.08) 151 0.78 (0.34–1.77) 
    P — — 0.182 — — 0.142 — — 0.550 
rs2292912          
    C/C or C/G 418 489 — 324 489 — 32 489 — 
    G/G 23 29 1.00 (0.55–1.82) 18 29 1.03 (0.54–1.96) 29 0.48 (0.06–4.07) 
    P — — 0.990 — — 0.938 — — 0.502 
rs7123390          
    G/G or G/A 396 491 — 315 491 — 29 491 — 
    A/A 41 21 2.40 (1.39–4.13) 34 21 2.54 (1.44–4.46) 21 2.11 (0.59–7.54) 
    P — — 0.002 — — 0.001 — — 0.253 
rs1401417          
    G/G or G/C 409 503 — 331 503 — 28 503 — 
    C/C 34 14 2.97 (1.57–5.63) 26 14 2.87 (1.47–5.58) 14 4.64 (1.41–15.34) 
    P — — 0.001 — — 0.002 — — 0.012 
GenotypeAll
B cell
T cell
Cases nControls nOR* (95% CI)Cases nControls nOR* (95% CI)Cases nControls nOR* (95% CI)
rs11038689          
    A/A or A/G 408 495 — 326 495 — 29 495 — 
    G/G 33 17 2.34 (1.28–4.27) 26 17 2.35 (1.25–4.41) 17 2.69 (0.74–9.84) 
    P — — 0.006 — — 0.008 — — 0.135 
rs11605924          
    A/A or A/C 327 365 — 263 365 — 25 365 — 
    C/C 114 151 0.82 (0.62–1.10) 89 151 0.79 (0.58–1.08) 151 0.78 (0.34–1.77) 
    P — — 0.182 — — 0.142 — — 0.550 
rs2292912          
    C/C or C/G 418 489 — 324 489 — 32 489 — 
    G/G 23 29 1.00 (0.55–1.82) 18 29 1.03 (0.54–1.96) 29 0.48 (0.06–4.07) 
    P — — 0.990 — — 0.938 — — 0.502 
rs7123390          
    G/G or G/A 396 491 — 315 491 — 29 491 — 
    A/A 41 21 2.40 (1.39–4.13) 34 21 2.54 (1.44–4.46) 21 2.11 (0.59–7.54) 
    P — — 0.002 — — 0.001 — — 0.253 
rs1401417          
    G/G or G/C 409 503 — 331 503 — 28 503 — 
    C/C 34 14 2.97 (1.57–5.63) 26 14 2.87 (1.47–5.58) 14 4.64 (1.41–15.34) 
    P — — 0.001 — — 0.002 — — 0.012 
*

Adjusted for age (continuous), race, and family history of cancer in first- or second-degree relatives.

Because NHL is composed of various subtypes with the potential for distinct etiologies, we also performed a stratified analysis including B-cell and T-cell lymphomas only. These results were qualitatively and quantitatively similar to those obtained in the overall analysis with ORs (95% CIs) for homozygous variants of rs11038689, rs7123390, and rs1401417 among B-cell lymphomas only of 2.35 (1.25–4.41; P = 0.008), 2.54 (1.44–4.46; P = 0.001), and 2.87 (1.47–5.58; P = 0.002), respectively. Although similar point estimates for the ORs were observed among T-cell lymphomas only, only rs1401417 reached statistical significance [OR, 4.64 (1.41–15.34; P = 0.012)] due to the small sample size. Further stratification was performed to analyze four common subtypes of B-cell lymphoma (Table 3). Each of the SNPs significantly associated with lymphoma in the full population (rs11038689, rs7123390, and rs1401417) was also significantly associated with follicular lymphoma, with ORs (95% CIs) of 3.17 (1.39–7.19; P = 0.006), 3.67 (1.79–7.56; P < 0.001), and 3.06 (1.24–7.54; P = 0.015), respectively. Two SNPs, rs7123390 and rs1401417, were significantly associated with B-cell chronic lymphocytic leukemia/prolymphocytic leukemia/small lymphocytic lymphoma, with ORs (95% CIs) of 3.38 (1.35–8.45; P = 0.009) and 4.23 (1.53–11.67; P = 0.005), respectively. Only rs1401417 was significantly associated with diffuse large B-cell lymphoma [OR, 2.67 (1.13–6.33; P = 0.026)]. Very few patients were classified as having marginal zone B-cell lymphoma, and no CRY2 SNPs were significantly associated with this NHL subtype.

Table 3.

Association of CRY2 variants with NHL risk by B-cell subtype

GenotypeDLBCL
FL
CLL/SLL
MZBL
Controls nCases nOR* (95% CI)Cases nOR* (95% CI)Cases nOR* (95% CI)Cases nOR* (95% CI)
rs11038689          
    A/A or A/G 495 126 — 90 — 48 — 27 — 
    G/G 17 1.41 (0.54–3.65) 10 3.17 (1.39–7.19) 2.75 (0.96–7.86) 2.30 (0.50–10.61) 
    P — — 0.484 — 0.006 — 0.060 — 0.285 
rs11605924          
    A/A or A/C 365 93 — 79 — 31 — 19 — 
    C/C 151 39 0.98 (0.64–1.50) 21 0.61 (0.36–1.04) 12 0.72 (0.36–1.40) 10 1.21 (0.54–2.70) 
    P — — 0.931 — 0.068 — 0.329 — 0.643 
rs2292912          
    C/C or C/G 489 127 — 97 — 52 — 27 — 
    G/G 29 0.53 (0.17–1.65) 0.95 (0.31–2.93) 0.27 (0.03–2.26) 1.60 (0.34–7.60) 
    P — — 0.274 — 0.932 — 0.227 — — 
rs7123390          
    G/G or G/A 491 124 — 85 — 45 — 27 — 
    A/A 21 1.31 (0.54–3.16) 14 3.67 (1.79–7.56) 3.38 (1.35–8.45) 1.86 (0.41–8.43) 
    P — — 0.547 — <0.001 — 0.009 — 0.423 
rs1401417          
    G/G or G/C 503 122 — 95 — 47 — 29 — 
    C/C 14 2.67 (1.13–6.33) 3.06 (1.24–7.54) 4.23 (1.53–11.67) 1.30 (0.16–10.33) 
    P — — 0.026 — 0.015 — 0.005 — 0.806 
GenotypeDLBCL
FL
CLL/SLL
MZBL
Controls nCases nOR* (95% CI)Cases nOR* (95% CI)Cases nOR* (95% CI)Cases nOR* (95% CI)
rs11038689          
    A/A or A/G 495 126 — 90 — 48 — 27 — 
    G/G 17 1.41 (0.54–3.65) 10 3.17 (1.39–7.19) 2.75 (0.96–7.86) 2.30 (0.50–10.61) 
    P — — 0.484 — 0.006 — 0.060 — 0.285 
rs11605924          
    A/A or A/C 365 93 — 79 — 31 — 19 — 
    C/C 151 39 0.98 (0.64–1.50) 21 0.61 (0.36–1.04) 12 0.72 (0.36–1.40) 10 1.21 (0.54–2.70) 
    P — — 0.931 — 0.068 — 0.329 — 0.643 
rs2292912          
    C/C or C/G 489 127 — 97 — 52 — 27 — 
    G/G 29 0.53 (0.17–1.65) 0.95 (0.31–2.93) 0.27 (0.03–2.26) 1.60 (0.34–7.60) 
    P — — 0.274 — 0.932 — 0.227 — — 
rs7123390          
    G/G or G/A 491 124 — 85 — 45 — 27 — 
    A/A 21 1.31 (0.54–3.16) 14 3.67 (1.79–7.56) 3.38 (1.35–8.45) 1.86 (0.41–8.43) 
    P — — 0.547 — <0.001 — 0.009 — 0.423 
rs1401417          
    G/G or G/C 503 122 — 95 — 47 — 29 — 
    C/C 14 2.67 (1.13–6.33) 3.06 (1.24–7.54) 4.23 (1.53–11.67) 1.30 (0.16–10.33) 
    P — — 0.026 — 0.015 — 0.005 — 0.806 
*

Adjusted for age (continuous), race, and family history of cancer in first- or second-degree relatives.

Association between CRY2 diplotypes and NHL risk. To further explore the relationship among SNPs, haplotypes were estimated for these 5 SNPs and 13 different haplotypes were identified among all subjects (Supplementary Table S2). Nine haplotypes carried at least one risk allele (i.e., a variant associated with increased risk in the main effects analysis). Because the individual risk estimates indicated a recessive disease model, we used the haplotype information to construct diplotypes for each individual. Figure 1 depicts the location of each marker within the CRY2 gene, and shows the NHL subtypes that were significantly associated with each homozygous variant diplotype. Full results are provided in the supplementary material (Supplementary Table S3). ORs (95% CIs) for homozygous variant diplotypes of rs11065924, rs11038689, rs2292912, rs1401417, and rs7123390 using the full population were as follows: 0.81 (0.61–1.08), 2.32 (1.27–4.24), 0.96 (0.53–1.75), 2.81 (1.51–5.23), and 2.36 (1.39–4.02), respectively. The OR associated with having the homozygous variant diplotype at one or more of the risk loci (rs11038689, rs1401417, or rs7123390) was 2.72 (1.61–4.59).

Figure 1.

Diplotype analysis and location of CRY2 variants. Haplotypes were estimated for each individual using the PHASE program, and diplotypes were constructed by combining the two haplotype estimates for each individual. Because the main effects associations indicated a recessive disease model, risk estimates were calculated for all combinations of haplotypes resulting in a homozygous variant diplotype at each locus, with all other diplotypes as the reference category. +, Estimates were obtained for each NHL subtype represented in the population; +, homozygous diplotypes that were significantly associated with each subtype (P < 0.05); −, nonsignifcant associations. 1+ risk variant, any haplotype combination that resulted in a homozygous variant diplotype for at least one of the following makers: rs11038689, rs1401417, or rs7123390. DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; CLL/SLL, B-cell chronic lymphocytic leukemia/prolymphocytic leukemia/small lymphocytic lymphoma; MZBL, marginal zone B-cell lymphoma.

Figure 1.

Diplotype analysis and location of CRY2 variants. Haplotypes were estimated for each individual using the PHASE program, and diplotypes were constructed by combining the two haplotype estimates for each individual. Because the main effects associations indicated a recessive disease model, risk estimates were calculated for all combinations of haplotypes resulting in a homozygous variant diplotype at each locus, with all other diplotypes as the reference category. +, Estimates were obtained for each NHL subtype represented in the population; +, homozygous diplotypes that were significantly associated with each subtype (P < 0.05); −, nonsignifcant associations. 1+ risk variant, any haplotype combination that resulted in a homozygous variant diplotype for at least one of the following makers: rs11038689, rs1401417, or rs7123390. DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; CLL/SLL, B-cell chronic lymphocytic leukemia/prolymphocytic leukemia/small lymphocytic lymphoma; MZBL, marginal zone B-cell lymphoma.

Close modal

Microarray reveals several cancer- and immune-related genes that are influenced by CRY2. To determine which genes and biological pathways might be influenced by CRY2, we silenced the gene in vitro, and performed a whole genome array comparing gene expression in cells with reduced CRY2 to cells with normal CRY2 levels. Differentially expressed transcripts (734) were identified (B-H P < 0.01 in both biological replicates and mean fold change of >|2|), and these genes were further explored through pathway and functional analyses using the IPA software package. Twenty-three networks were identified as significantly associated with the differentially expressed genes at P value of <1.0E−10. Among these, seven were associated with Cancer, six networks were important for Immune Response, and four were important for Hematological System Development or Hematological Disease (Supplementary Table S4).

The top network operative in Hematological System Development and Function (P = 1.0E−36; Fig. 2) also contained several cancer-related transcripts, including molecules important for DNA repair, cell migration and metastasis, apoptosis, cell proliferation, and angiogenesis. The details of each transcript in the network, including fold changes, adjusted P values, and a short description of function with relevant citations, can be found in the supplementary material (Supplementary Table S5). Briefly, 35 genes were present in the network, 33 of which could be uniquely associated with a single transcript in the microarray. Of these, 27 were significantly up-regulated after CRY2 silencing, 4 were not significantly altered, and only 2 were significantly down-regulated, which is consistent with the role of CRY2 as a transcriptional repressor. The probability of obtaining at least this many differentially expressed molecules in one network by chance alone is reflected by the P value for the network (P = 1.0E−36), indicating that it is very unlikely that the network was obtained randomly and, therefore, CRY2 is likely to be important for immune response coordination and processes related to hematologic system development. At the center of this network is interleukin (IL)-6, an immunoregulatory cytokine that is important for cell growth. This transcript was strongly up-regulated after CRY2 silencing, with a mean fold increase of 8.17 (B-H P = 0). In addition, a member of the vascular endothelial growth factor family (VEGFC), which is also important for cell growth, as well as angiogenesis, was up-regulated >6-fold (B-H P = 6.99E−4). Interestingly, a number of transcripts with proapoptotic, antiproliferative, or other tumor suppressor capabilities were also up-regulated in this network. These include IFNβ1 that was up-regulated >3-fold (B-H P = 2.96E−9), thioredoxin-interacting protein, >11-fold up-regulated (B-H P = 2.47E−4), retinoic acid receptor responder 3, >2-fold up-regulated (B-H P = 1.86E−4), and 2 members of the chemokine (C-X-C motif) ligand family (CXCL9 and CXCL11), >10- and 4-fold up-regulated, respectively (B-H P = 1.01E−5 and = 3.04E−6).

Figure 2.

Network of interactions among genes differentially expressed after CRY2 knockdown. This network was identified as operative in Immune Response, Cell-to-Cell Signaling and Interaction, and Hematological System Development and Function, as determined by Ingenuity Pathway Analysis software. Each shaded transcript was significantly altered after CRY2 silencing, with a mean fold change of >|2| and B-H P value of <0.01 in both biological replicates of the whole genome microarray. Red, genes that are up-regulated; green, genes that are down-regulated. Higher intensity color corresponds to greater magnitude of the fold change. Of the 29 molecules in the network that were significantly altered, 27 were up-regulated, whereas only 2 were down-regulated, which is consistent with the role of CRY2 as a transcriptional repressor.

Figure 2.

Network of interactions among genes differentially expressed after CRY2 knockdown. This network was identified as operative in Immune Response, Cell-to-Cell Signaling and Interaction, and Hematological System Development and Function, as determined by Ingenuity Pathway Analysis software. Each shaded transcript was significantly altered after CRY2 silencing, with a mean fold change of >|2| and B-H P value of <0.01 in both biological replicates of the whole genome microarray. Red, genes that are up-regulated; green, genes that are down-regulated. Higher intensity color corresponds to greater magnitude of the fold change. Of the 29 molecules in the network that were significantly altered, 27 were up-regulated, whereas only 2 were down-regulated, which is consistent with the role of CRY2 as a transcriptional repressor.

Close modal

Apart from the transcripts in this network, several additional genes that may be relevant for lymphomagenesis were identified as significantly altered after CRY2 silencing. These include three members of the chemokine (C-C motif) ligand family (CCL3, CCL4, and CCL5), which are important for immunoregulation and inflammatory processes, several additional members of the interleukin family of cytokines (including IL-11, IL-15, IL-18, IL-28a, IL-28b, and IL-29), two insulin-like growth factor binding proteins (IGFBP3 and IGFBP6), and genes in the MHC, class I (HLA-A, HLA-B, HLA-C, and HLA-E). Of note, CRY2 itself was significantly down-regulated in both microarray replicates (B-H P < 0.05), with exactly the same fold change of −3.17 observed in both samples.

The program also computes a P value based on the likelihood of obtaining the observed number of differentially expressed molecules in a given process by chance alone. The top three disease processes associated with the differentially expressed gene set were “Cancer” (B-H P = 3.75E−9), “Inflammatory Disease” (B-H P = 5.19E−8), and “Hematological Disease” (B-H P = 8.01E−8; Fig. 3). In addition, 144 of the differentially expressed transcripts were related to “Tumorigenesis” (B-H P = 4.57E−9), 64 molecules were associated with “Immune Response” (B-H P = 7.03E−10), and 49 were involved in “Proliferation of Lymphocytes” (B-H P = 6.76E−9). The large number of cancer- and immune-related molecules influenced by CRY2 knockdown provides further evidence suggesting a role for CRY2 in lymphoproliferative processes.

Figure 3.

Diseases most strongly associated with the significantly altered genes. The Ingenuity software scans its Knowledge Base of manually curated relationships for instances in the literature that can link a specific molecule to a functional disease process. The software also assigns a P value based on the likelihood of obtaining the observed number of disease-related molecules in a given data set by chance alone. Due to the large number of disease processes represented in the database, B-H P values are presented to correct for multiple comparisons. Threshold line, a B-H P value of 0.05.

Figure 3.

Diseases most strongly associated with the significantly altered genes. The Ingenuity software scans its Knowledge Base of manually curated relationships for instances in the literature that can link a specific molecule to a functional disease process. The software also assigns a P value based on the likelihood of obtaining the observed number of disease-related molecules in a given data set by chance alone. Due to the large number of disease processes represented in the database, B-H P values are presented to correct for multiple comparisons. Threshold line, a B-H P value of 0.05.

Close modal

The observed connection between circadian disruption (e.g., shift work) and cancer risk in epidemiologic studies has led to the circadian gene hypothesis, which suggests that genetic variants in genes responsible for maintaining circadian rhythm may affect an individual's susceptibility to human cancers. This hypothesis is supported by results from recent genetic association studies of breast cancer (28, 29), prostate cancer (24), and NHL (5). The findings from the current study of CRY2 provide more evidence demonstrating a role for the circadian genes in NHL susceptibility.

To the best of our knowledge, genetic variants in CRY2 have not been previously examined in NHL, and only one of the SNPs genotyped in the case-control portion of this analysis had been studied previously; in a population-based case-control study conducted in China, which showed a significant association between the variant allele of rs1401417 and increased risk of prostate cancer (24). This finding is consistent with the significant association we observed in the NHL population, and further investigations into the nature of these relationships are warranted to determine whether CRY2 has a global effect on cancer susceptibility.

The observed associations between CRY2 and NHL risk provide additional molecular epidemiologic evidence supporting the proposed role of circadian genes as tumor suppressors (30). Circadian genes have been shown to affect expression of 2% to 10% of mammalian genes (31) including many cancer-related genes (30). Emerging data from animal models have further shown a substantial effect of circadian genes on tumor-related biological pathways such as cell proliferation, cell cycle control, DNA damage response, and apoptosis (30). Mice with mutations in the circadian gene PER2 have deficiencies in DNA damage response and are more prone to tumorigenesis (32). Altered expression of circadian genes also occurs in human tumors, as studies have shown that the period (PER) genes fail to maintain daily rhythmic expression patterns in cancer cells (33). Although CRY2 has also been shown to be involved in cell cycle regulation (34), explicit mechanisms for its role in cancer susceptibility, especially NHL tumorigenesis, are currently unknown.

The microarray analysis, which implicated CRY2 as having the potential to influence gene expression in a number of pathways, including those with relevance for cancer and immune function, provides the first mechanistic evidence suggesting that CRY2 may be important for NHL susceptibility. The findings relative to IL-6 are of particular interest. It has been shown that IL-6 can inhibit DNA synthesis by preventing cell cycle progression into S phase (35). It has also been fairly well-established that IL-6 plays an important role in B-cell proliferation (36), and antibodies against IL-6 or its receptor have been explored as treatments for B-lymphoproliferative disorder (37) and inflammatory autoimmune diseases (38). In addition, a previous study has shown that mice expressing an IL-6 transgene exhibit lymphoproliferation and plasmacytosis, and nearly a third of these mice developed follicular and diffuse large cell B-cell lymphomas (39). Moreover, serum levels of IL-6 and VEGF have been shown to be of value in predicting clinical outcome in some patients with NHL (40, 41). The observation that silencing of CRY2 results in >8-fold induction of IL-6, and 6-fold induction of VEGFC, is therefore highly relevant in understanding the etiology of hematologic malignancies, and may have important implications for predicting NHL prognosis. Furthermore, cell signaling by IL-6 is partially mediated by its effects on the Janus-activated kinase (JAK)/signal transducers and activators of transcription 3 (STAT3) pathway (42). Although JAK was unaffected after CRY2 silencing, STAT3 was up-regulated >2-fold (B-H P = 7.06E−5). Because IL-6 has been shown to confer increased survival and chemotherapeutic resistance on lymphoma cells, which is at least partially mediated by STAT3, STAT3 has been proposed as a potential therapeutic target for patients with NHL (43). As such, although the pleiotropic biological effects of CRY2 may make it a poor candidate to be directly targeted by therapeutic agents, its effects on both IL-6 and STAT3 may lead future investigations to consider circadian timing when administering cytokine-targeting chemotherapies.

Apart from IL-6, several other interleukins were also significantly altered in CRY2 knockdown cells. IL-18, which was up-regulated by >10-fold, may play an important but paradoxical role in cancer risk (44). It has been implicated in cancer protection, through its role in activating immune cells to eliminate sporadic cancers but may also promote tumor progression by encouraging angiogenesis, tumor growth, and local invasion. Interestingly all three members of a newly described class of IFN λs (IL-28a, IL-28b, and IL-29) were significantly up-regulated after CRY2 silencing. These immunoregulatory cytokines have antiviral and antitumor activity, and may also have potential as therapeutic agents in the treatment of cancer (45).

Three members of the chemokine (C-C motif) ligand family (CCL3, CCL4, and CCL5) were also up-regulated in CRY2 knockdown cells. These genes are clustered together on the long arm of chromosome 17 and are important for immunoregulation and inflammation (46). CCL3, which was up-regulated >15-fold in CRY2 knockdown cells, has been shown to be elevated in patients with multiple myeloma and other hematologic cancers compared with healthy controls, and increased serum levels of CCL3 were associated with advanced multiple myeloma stage (47) and poorer prognosis in multiple myeloma (48). Of additional interest was the identification of four genes in the MHC, class I (HLA-A, HLA-B, HLA-C, and HLA-E), which were all significantly up-regulated after CRY2 silencing. Apart from being central to immunoregulation, these genes have been associated with several cancers, including Hodgkin's disease (49) and non–Hodgkin's lymphoma (50). Taken together, these results suggest that reductions in CRY2 have the potential to significantly effect processes relevant for lymphomagenesis, and represent a first step in understanding the mechanism by which SNPs in the CRY2 gene might influence NHL susceptibility.

It is important to recognize some of the limitations of the study. First, it is unclear how closely the effect of CRY2 silencing in vitro might mimic the conditions that would arise in vivo. At the level of the organism, circadian rhythmicity is made even more complex by environmental cues, such as light exposure, and hormonal pathways, which may influence circadian gene expression. In addition, because one aim of the study was to examine the influence of CRY2 silencing on immune response pathways, we chose not to use cells derived from lymphoma tissue, which may begin with some level of aberrant immune-related gene expression. As such, it will be necessary for future studies to confirm that the findings we observed are applicable to lymphocytes, and that the changes in gene expression after CRY2 silencing that we observed at the cellular level are representative of those that would occur at the organismal level.

In summary, the findings from our case-control analysis suggest a novel association between the circadian gene CRY2 and risk of NHL, supporting the hypothesis that genetic variations in circadian genes may confer inherited susceptibility to NHL. The subsequent loss-of-function analysis and whole genome expression microarray suggest that the observed association could potentially be attributed to the effect of CRY2 on several genes important for cancer in general, as well as a number of genes with known relevance for hematologic malignancies. Our findings provide a novel panel of promising biomarkers for NHL risk and prognosis, which warrant further investigation. In addition, because previous studies have suggested that circadian-related environmental exposures, such as light at night or rotating shift work, may influence cancer susceptibility, future investigations into potential interactions between circadian gene variants and environmental exposures may be of interest in developing cancer prevention strategies.

No potential conflicts of interest were disclosed.

Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/).

Grant support: NIH (grants CA122676, CA110937, and CA108369).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Irina Tikhonova at Yale University's W.M. Keck Foundation Biotechnology Research Laboratory for Sequenom genotyping analysis.

1
Kondratov RV, Gorbacheva VY, Antoch MP. The role of mammalian circadian proteins in normal physiology and genotoxic stress responses.
Curr Top Dev Biol
2007
;
78
:
173
–216.
2
IARC Shift-work, painting and fire-fighting. In: IARC monographs on the evaluation of carcinogenic risks to humans. Lyon: IARC; 2007.
3
Zhu Y, Zheng T. Clock-cancer connection in non-Hodgkin's lymphoma.
Med Hypotheses
2008
;
70
:
788
–92.
4
Gery S, Gombart AF, Yi WS, et al. Transcription profiling of C/EBP targets identifies Per2 as a gene implicated in myeloid leukemia.
Blood
2005
;
106
:
2827
–36.
5
Zhu Y, Leaderer D, Guss C, et al. Ala394Thr polymorphism in the clock gene NPAS2: a circadian modifier for the risk of non-Hodgkin's lymphoma.
Int J Cancer
2007
;
120
:
432
–5.
6
Arjona A, Sarkar DK. Circadian oscillations of clock genes, cytolytic factors, and cytokines in rat NK cells.
J Immunol
2005
;
174
:
7618
–24.
7
Arjona A, Sarkar DK. Evidence supporting a circadian control of natural killer cell function.
Brain Behav Immun
2006
;
20
:
469
–76.
8
Chacon F, Cano P, Lopez-Varela S, et al. Chronobiological features of the immune system. Effect of calorie restriction.
Eur J Clin Nutr
2002
;
56
Suppl 3:
S69
–72.
9
Esquifino AI, Cano P, Jimenez-Ortega V, Fernandez-Mateos P, Cardinali DP. Neuroendocrine-immune correlates of circadian physiology: studies in experimental models of arthritis, ethanol feeding, aging, social isolation, and calorie restriction.
Endocrine
2007
;
32
:
1
–19.
10
Esquifino AI, Selgas L, Arce A, Maggiore VD, Cardinali DP. Twenty-four-hour rhythms in immune responses in rat submaxillary lymph nodes and spleen: effect of cyclosporine.
Brain Behav Immun
1996
;
10
:
92
–102.
11
Seres J, Herichova I, Roman O, Bornstein S, Jurcovicova J. Evidence for daily rhythms of the expression of proopiomelanocortin, interleukin-1-β and interleukin-6 in adenopituitaries of male long-evans rats: effect of adjuvant arthritis.
Neuroimmunomodulation
2004
;
11
:
316
–22.
12
Arjona A, Sarkar DK. Are circadian rhythms the code of hypothalamic-immune communication? Insights from natural killer cells.
Neurochem Res
2008
;
33
:
708
–18.
13
Mormont MC, Levi F. Circadian-system alterations during cancer processes: a review.
Int J Cancer
1997
;
70
:
241
–7.
14
Gamaleia NF, Skivka LM, Fedorchuk AG, Shishko ED. Circadian rhythms of cytotoxic activity in peripheral blood mononuclear cells of patients with malignant melanoma.
Exp Oncol
2006
;
28
:
54
–60.
15
Filipovich AH, Mathur A, Kamat D, Shapiro RS. Primary immunodeficiencies: genetic risk factors for lymphoma.
Cancer Res
1992
;
52
:
5465
–7s.
16
Kersey JH, Shapiro RS, Filipovich AH. Relationship of immunodeficiency to lymphoid malignancy.
Pediatr Infect Dis J
1988
;
7
:
S10
–2.
17
Griffin EA, Jr., Staknis D, Weitz CJ. Light-independent role of CRY1 and CRY2 in the mammalian circadian clock.
Science
1999
;
286
:
768
–71.
18
Vitaterna MH, Selby CP, Todo T, et al. Differential regulation of mammalian period genes and circadian rhythmicity by cryptochromes 1 and 2.
Proc Natl Acad Sci U S A
1999
;
96
:
12114
–9.
19
Unsal-Kacmaz K, Mullen TE, Kaufmann WK, Sancar A. Coupling of human circadian and cell cycles by the timeless protein.
Mol Cell Biol
2005
;
25
:
3109
–16.
20
Matsuo T, Yamaguchi S, Mitsui S, et al. Control mechanism of the circadian clock for timing of cell division in vivo.
Science
2003
;
302
:
255
–9.
21
Zhang Y, Holford TR, Leaderer B, et al. Blood transfusion and risk of non-Hodgkin's lymphoma in Connecticut women.
Am J Epidemiol
2004
;
160
:
325
–30.
22
de Bakker PI, Yelensky R, Pe'er I, et al. Efficiency and power in genetic association studies [see comment].
Nat Genet
2005
;
37
:
1217
–23.
23
Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps.
Bioinformatics
2005
;
21
:
263
–5.
24
Chu LW, Zhu Y, Yu K, et al. Variants in circadian genes and prostate cancer risk: a population-based study in China.
Prostate Cancer Prostatic Dis
2008
;
11
:
342
–8.
25
Calvano SE, Xiao W, Richards DR, et al. A network-based analysis of systemic inflammation in humans.
Nature
2005
;
437
:
1032
–7.
26
Stephens M, Smith NJ, Donnelly P. A new statistical method for haplotype reconstruction from population data.
Am J Hum Genet
2001
;
68
:
978
–89.
27
Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing.
J R Stat Soc Ser B (Methodological)
1995
;
57
:
289
–300.
28
Zhu Y, Brown HN, Zhang Y, Stevens RG, Zheng T. Period3 structural variation: a circadian biomarker associated with breast cancer in young women.
Cancer Epidemiol Biomarkers Prev
2005
;
14
:
268
–70.
29
Zhu Y, Stevens RG, Leaderer D, et al. Non-synonymous polymorphisms in the circadian gene NPAS2 and breast cancer risk. Breast Cancer Res Treat 2007.
30
Fu L, Lee CC. The circadian clock: pacemaker and tumour suppressor.
Nat Rev Cancer
2003
;
3
:
350
–61.
31
Storch KF, Lipan O, Leykin I, et al. Extensive and divergent circadian gene expression in liver and heart.
Nature
2002
;
417
:
78
–83.
32
Fu L, Pelicano H, Liu J, Huang P, Lee C. The circadian gene Period2 plays an important role in tumor suppression and DNA damage response in vivo.
Cell
2002
;
111
:
41
–50.
33
You S, Wood PA, Xiong Y, et al. Daily coordination of cancer growth and circadian clock gene expression.
Breast Cancer Res Treat
2005
;
91
:
47
–60.
34
Gauger MA, Sancar A. Cryptochrome, circadian cycle, cell cycle checkpoints, and cancer.
Cancer Res
2005
;
65
:
6828
–34.
35
Shen WH, Zhou JH, Broussard SR, et al. Proinflammatory cytokines block growth of breast cancer cells by impairing signals from a growth factor receptor.
Cancer Res
2002
;
62
:
4746
–56.
36
Kishimoto T. The biology of interleukin-6.
Blood
1989
;
74
:
1
–10.
37
Haddad E, Paczesny S, Leblond V, et al. Treatment of B-lymphoproliferative disorder with a monoclonal anti-interleukin-6 antibody in 12 patients: a multicenter phase 1–2 clinical trial.
Blood
2001
;
97
:
1590
–7.
38
Ding C, Jones G. Anti-interleukin-6 receptor antibody treatment in inflammatory autoimmune diseases.
Rev Recent Clin Trials
2006
;
1
:
193
–200.
39
Kovalchuk AL, Kim JS, Park SS, et al. IL-6 transgenic mouse model for extraosseous plasmacytoma.
Proc Natl Acad Sci U S A
2002
;
99
:
1509
–14.
40
Kurzrock R. Cytokine deregulation in hematological malignancies: clinical and biological implications.
Clin Cancer Res
1997
;
3
:
2581
–4.
41
Pedersen LM, Klausen TW, Davidsen UH, Johnsen HE. Early changes in serum IL-6 and VEGF levels predict clinical outcome following first-line therapy in aggressive non-Hodgkin's lymphoma.
Ann Hematol
2005
;
84
:
510
–6.
42
Brasier AR. Expanding role of cyclin dependent kinases in cytokine inducible gene expression.
Cell Cycle
2008
;
7
:
2661
–6.
43
Alas S, Bonavida B. Inhibition of constitutive STAT3 activity sensitizes resistant non-Hodgkin's lymphoma and multiple myeloma to chemotherapeutic drug-mediated apoptosis.
Clin Cancer Res
2003
;
9
:
316
–26.
44
Dinarello CA. The paradox of pro-inflammatory cytokines in cancer.
Cancer Metastasis Rev
2006
;
25
:
307
–13.
45
Lasfar A, Lewis-Antes A, Smirnov SV, et al. Characterization of the mouse IFN-λ ligand-receptor system: IFN-λs exhibit antitumor activity against B16 melanoma.
Cancer Res
2006
;
66
:
4468
–77.
46
Wolpe SD, Davatelis G, Sherry B, et al. Macrophages secrete a novel heparin-binding protein with inflammatory and neutrophil chemokinetic properties.
J Exp Med
1988
;
167
:
570
–81.
47
Choi SJ, Cruz JC, Craig F, et al. Macrophage inflammatory protein 1-α is a potential osteoclast stimulatory factor in multiple myeloma.
Blood
2000
;
96
:
671
–5.
48
Terpos E, Politou M, Szydlo R, et al. Serum levels of macrophage inflammatory protein-1 α (MIP-1α) correlate with the extent of bone disease and survival in patients with multiple myeloma.
Br J Haematol
2003
;
123
:
106
–9.
49
Diepstra A, Niens M, te Meerman GJ, Poppema S, van den Berg A. Genetic susceptibility to Hodgkin's lymphoma associated with the human leukocyte antigen region.
Eur J Haematol Suppl
2005
;
66
:
34
–41.
50
Choi HB, Roh SY, Choi EJ, et al. Association of HLA alleles with non-Hodgkin's lymphoma in Korean population.
Int J Hematol
2008
;
87
:
203
–9.

Supplementary data