This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yu, J. Articles by McLeod, H. L. Search for Related Content PubMed PubMed Citation Articles by Yu, J. Articles by McLeod, H. L.

Ferredoxin Reductase

Pharmacogenomic Assessment in Colorectal Cancer¹

Department of Medicine, Washington University School of Medicine and the Siteman Cancer Center, St. Louis, Missouri 63110

	ABSTRACT

Top ABSTRACT Introduction Materials and Methods Results Discussion REFERENCES

 
Ferredoxin reductase (FDXR) is a putative contributor to TP53-mediated apoptosis from 5-fluorouracil chemotherapy through the generation of oxidative stress. With TaqMan real-time quantitative reverse transcription-PCR, this study established a significant difference in FDXR relative RNA expression level between tumor (median, 212.9 units) and normal tissues (median, 113.8 units) from 51 colorectal cancer patients (P < 0.001). Seven single nucleotide polymorphisms (SNPs) in the FDXR gene were discovered, with no significant difference in variant allele frequency between colon tumor and normal tissues (P > 0.05), and the common haplotypes for FDXR were not different between colon tumor and normal samples. No correlation was observed between FDXR genotype and RNA expression implying that the polymorphisms described in this study are not regulating FDXR expression in colon cancer. This genomic characterization provides the foundation for pharmacogenetic analysis of the impact of FDXR on chemotherapy for colorectal cancer.

	Introduction

Top ABSTRACT Introduction Materials and Methods Results Discussion REFERENCES

 
After nearly 5 decades, 5-FU³ continues to play an important role in the treatment of common human solid tumors, including colorectal, cervix, head/neck, gastric, and breast (1) . With the development of capecitabine and other oral 5-FU drugs, the use of this agent is increasing (2) . Despite the considerable clinical experience with 5-FU, there is still insufficient information on the key genes regulating cell death. Recent studies have identified the gene encoding FDXR as a putative contributor to TP53-mediated apoptosis from 5-FU, through the generation of oxidative stress in the mitochondria (3 , 4) . FDXR is a kDa 50,000 mitochondrial flavoprotein attached to the matrix side of the inner mitochondrial membrane. FDXR transports electrons from NADPH via the soluble single electron shuttle ferredoxin to a membrane-integrated cytochrome P450 enzyme (CYP11A1), which catalyzes the side-chain cleavage of cholesterol to form pregnenolone in a multistep biosynthesis of steroid hormones (5) . Under substrate-limiting conditions, excess electrons can leak from this shuttling system, and overexpressed FDXR can deplete the reduced NADPH, and both processes will increase the amount of superoxide in cells (3 , 4 , 6 , 7) . These previously described biochemical properties, together with its induction by TP53, imply a role for FDXR in TP53-dependent apoptosis through oxidative stress.

Although the importance of genetic polymorphism on chemotherapy activity has been established (8 , 9) , no information thus far exists on FDXR genomics in human tumors. In the present study, RNA expression and genotype status of the FDXR gene in tumor and adjacent normal tissue from 51 colorectal cancer patients were investigated using TaqMan quantitative real-time reverse transcription-PCR and Pyrosequencing technologies. In addition, the correlations between gene expression of FDXR and other interacting death genes (TP53, BAX, BCL2, BCLX, and BCLX_L), and between the FDXR RNA expression and genotype were analyzed. This study provides an initial understanding of FDXR genomics in human colon cancer, in which 5-FU is heavily used.

	Materials and Methods

Top ABSTRACT Introduction Materials and Methods Results Discussion REFERENCES

 
Patients and Samples.
Tumor specimens and paired normal colon tissues used in this study were from 51 Dukes’ C colorectal cancer patients [28 male/23 female; age range, 39–96 (median,79) years]. Samples were snap-frozen in liquid nitrogen immediately after surgery and stored at -80°C. None of the patients had received preoperative radiation or chemotherapy. Tumor classification and differential grade were evaluated in all of the cases: 40 enteric adenocarcinoma, and 11 mucinous, and 1 Grade I, 38 Grade II, 12 Grade III. Twenty-seven tumors were localized in the right colon, 18 in the left colon, and the remaining 6 in the rectum. Written informed consent was obtained from all of the patients for the investigators to bank tumor tissue and to perform genomic analysis. This study was approved by the Washington University Human Subjects Committee.

Quantitative Real-time Reverse Transcription-PCR.
Regions of high tumor cellularity were selected for RNA extraction (median 86.3%; range, 65–95%). Tissue total RNA was isolated from the colon tumor or adjacent normal colon mucosa with the TRIzol RNA isolation kit (Invitrogen, Carlsbad, CA), and the quality of RNA (i.e., A_260/280 > 1.8 and the clear RNA bands for 28S, 18S, and 5S were seen) was confirmed in the Siteman Cancer Center Tissue Procurement Core. After reverse-transcription into cDNA using Superscript II reverse transcriptase (Invitrogen, Carlsbad, CA), the mRNAs of FDXR, TP53, BAX, BCL2, BCLX, and BCLX_L were determined by real-time PCR, which was carried out in a 10-µl reaction mix containing 2 µl of cDNA (10 ng/µl), 5 µl of 2x TaqMan universal PCR master mix (ABI, Foster City, CA), and 3 µl of primer and probe mix (400 nM each forward and reverse primers⁴ and 200 nM TaqMan probe.⁴ All real-time PCR assays were performed in triplicate on an ABI PRISM 7700 Sequence Detector System (ABI, Foster City, CA) with the following program: 50°C for 2 min to activate uracil N-glycosylase enzyme, 95°C for 10 min to denature uracil n-glycosylase and activate DNA polymerase, 40 cycles at 95°C for 20 s and at 60°C for 1 min. The sequence detection program calculates a threshold cycle number (C_T) at which the reporter fluorescence generated by cleavage of the probe is statistically greater than that of the background signal (10) .

Measurement of Relative RNA Expression Level.
In this study, relative expression level was calculated using a modified comparative C_T method, which uses actual real-time PCR amplification efficiency instead of assuming all sets of gene primer and probe have approximately equal efficiency (10 , 11) . The gene encoding amyloid ß precursor protein was used as an internal reference gene, because it had near identical expressions between colon tumor and normal tissues in previous serial analysis of gene expression analysis (12) and less than a 3-fold change between all tumor and normal sample pairs in this study. The relative expression level of an individual target gene was normalized to the reference gene and to one of all 102 colon tumor and normal RNA samples that was with the maximum C_T value (i.e., the lowest expression level, called calibrator sample or 1x sample) in any target gene. A mathematical model was applied to determine the normalized relative expression level of a target gene in the individual sample, using the following formula:

where E_target is the real-time PCR efficiency of target gene transcript, and E_reference is the real-time PCR efficiency of reference gene transcript (10) . Therefore, relative RNA expression level of a gene in this study is a unitless number based on the reference gene and the 1x sample.

SNP Mining and Validation.
Genomic DNA was extracted from the colon tumor and normal tissue with the Qiagen DNA isolation kit. SNPs in the FDXR gene were mined using the publicly available SNP databases (13) . PCR primers were designed using Primer Express (v1.5; ABI, Foster City, CA), and the Pyrosequencing primers were designed using the Pyrosequencing SNP Primer Design software (v1.01; http://www.pyrosequencing.com). PCR was carried out using Amplitaq Gold PCR master mix (ABI), 5 pmol of each primer⁴ (IDT, Coralville, IA), and 1 ng of DNA. Pyrosequencing was carried out as described previously (14) using the Pyrosequencing PSQ HS96A instrument and software (Pyrosequencing, Uppsala, Sweden). The genotype was called variant if it differed from the Refseq consensus sequence at the specific single-nucleotide polymorphism position (http://www.ncbi.nlm.nih.gov/LocusLink/refseq.html).

Statistical Analysis.
Statistical analyses were performed with the software STATISTICA (StatSoft, Inc., Tulsa, OK). Significance of the difference of relative expression level between paired tumor and normal samples was evaluated by Wilcoxon matched-pairs test. The influence of gender, tumor location, pathological variables, or SNPs on FDXR expression was evaluated with the Mann-Whitney or Kruskal-Wallis test. Correlation between variables was observed with Spearman rank correlation. The significance level was set at P < 0.05. Analysis of Hardy-Weinberg equilibrium and haplotype was carried out using the Polymorphism and Haplotype Analysis Suite (15) .

	Results

Top ABSTRACT Introduction Materials and Methods Results Discussion REFERENCES

 
RNA Expression of the FDXR Gene in Colon Tumor.
There was a significant difference in FDXR relative RNA expression level between tumor (median, 212.9 units) and normal tissues (median, 113.8 units) in 51 colorectal cancer patients (P < 0.001). The colon tumor had a median 1.60-fold higher FDXR RNA expression than did the adjacent normal colon sample, and tumor was higher than adjacent normal tissue in 37 (73%) of 51 patients. Interindividual variability of FDXR RNA expression was large, with a coefficient of variance of 97.6% and a 40.4-fold range in the tumor tissue and 92.4% and 25.6-fold in the normal tissue. The relative expression level of FDXR along with TP53, BAX, BCL2, BCLX, and BCLX_L is shown in Fig. 1 . Correlations between FDXR and TP53, BAX, BCL2, BCLX, or BCLX_L RNA expressions were determined in the colon tumor and normal tissues (Table 1) . FDXR demonstrated statistically significant correlations with TP53, BAX, and BCL2 in both colon tumor and normal tissues, but these were of borderline biological significance (R = 0.33–0.37; P = 0.007–0.019). All four of the BCL2 family members had a high level of correlation with each other in colon tumor and normal RNA expression (Table 1 ; P < 0.001 for all).

View larger version (22K): [in this window] [in a new window]   Fig. 1. The relative expression levels (log scale of median) of six genes in the colon tumor () and the adjacent normal tissues (). All of the assay value was normalized to an internal reference gene amyloid ß precursor protein (APP) and a 1x sample with the largest CT value (subtracting target gene CT value from reference gene CT value).

FDXR Genotyping.
In silico mining identified two SNPs in the amino acid coding region (Q123R and A256A), and five intronic SNPs (IVS7+38C>T, IVS7+49G>A, IVS7–98A>C, IVS9+15A>G, IVS9–97T>C) in the FDXR gene. Variant allele frequencies in the colon DNA samples ranged from 0.01 to 0.43 for the tumors, and from 0.02 to 0.40 for the normal samples (Table 2) , and statistically, there was no significant difference in variant allele frequency between the colon tumor and normal DNA samples (P > 0.05). Using the Polymorphism and Haplotype Suite (15) , linkage and haplotype analysis for the tumor DNA samples demonstrated significant linkage between Q123R and the intron 7 and 9 SNPs (P < 0.001 for all). Three haplotypes described 91% of samples (93% for the tumor, 90% for the normal; Table 3 ). The small number of haplotypes represents the high degree of linkage found between four of the seven SNPs. Only one discordant genotype was seen between the normal and tumor samples: a patient heterozygous for FDXR Q123R in the normal tissue had only the variant allele in the tumor sample. This small level of variation between the tumor and normal tissue is consistent with a low level of loss of heterozygosity in FDXR and that region of chromosome 17 (17q24–25) in colon cancer.

	Discussion

Top ABSTRACT Introduction Materials and Methods Results Discussion REFERENCES

 
FDXR has been implicated as a mediator of 5-FU-driven apoptosis, but there is little information on expression in human tissues or its gene structure. This study demonstrated that FDXR RNA expression in colon tumor is significantly higher than that in matched normal mucosa, independent of demographic, anatomical, or pathological variables. The increased RNA expression of FDXR in tumors may enhance the initiation of tumor cell apoptosis and may increase the sensitivity to chemotherapy through the resulting oxidative stress (3 , 4) . In the majority of human solid tumors, including colon cancer, expression of the TP53 gene is increased, but some TP53 proteins may lose tumor suppressor function because of gene mutation (16 , 17) . Whether overexpressed mutated TP53 induces the increased expression of downstream elements remains unclear. However, Liu and Chen (4) have found a p53 response element within the promoter of FDXR, and showed that the p53 response element is responsive to various p53 family proteins. Our study has demonstrated that the RNA expression of FDXR in both tumor and normal colon tissues was positively correlated with TP53, and tumor TP53 expression was also significantly higher than normal colon (Fig. 1) , suggesting that increased FDXR RNA expression in colon tumors may be induced at least in part by TP53 (i.e., FDXR is a downstream effector in the TP53-mediated apoptosis pathway; Refs. 3 , 4 ). It is notable that several recent studies have shown that the selection of TP53 or BAX status will be an important determinant to the sensitivity of tumor cells to chemotherapy (18 , 19) . Therefore, the relationship between the TP53 or/and BAX gene status and the activation of FDXR gene expression needs to be evaluated to further clarify the role of FDXR in apoptotic tumor cell death under oxidative stress. However, a single gene is unlikely to be the only key regulator of sensitivity to chemotherapy because the regulation of drug activity is complex in biological systems.

In addition, recent studies have shown that proapoptotic members of the BCL2 family proteins are induced by TP53 and affect mitochondrial permeability (20, 21, 22) . The data presented here demonstrate that FDXR is correlated statistically with TP53, BAX, and BCL2 RNA expression in both tumor and normal colon tissues, and all four BCL2 family members have a high level of correlation with each other in colon tumor and normal mucosa. It suggests that the activation and/or regulation of FDXR gene expression may be closely associated with that of the member(s) of BCL2 family, as well as that of the TP53 gene. It may be that excess oxidative stress in cellular mitochondria, at least partially from FDXR overexpression, may be a significant event in 5-FU chemotherapy activity in colon tumors. However, further study is needed to elucidate whether high FDXR expression will translate into the generation of excess superoxide, and to what degree strong oxidative stress will influence apoptosis of tumor cells.

This is the first study to describe the presence of SNPs in the FDXR gene with the confirmed occurrence of seven database SNPs. No significant difference in variant allele frequency between DNAs from colon tumor and normal tissues was found, and the most common haplotypes are the same for the colon tumor and normal DNAs. Therefore, the SNPs detected in our study seem unlikely to be associated with the development of colon cancers. Although a recent study (23) has shown that the unique interactions of multiple SNPs within a haplotype, rather than any individual SNPs, ultimately can affect biological and therapeutic phenotype, the lack of correlation between tumor genotype (SNP) and phenotype (RNA expression) implies that the polymorphisms described in this study are not regulating FDXR expression in colon cancer. With the increasing number of SNPs being identified across the genome, further haplotype analysis may be required to confirm the real impact of FDXR SNPs on cancer chemotherapy.

	ACKNOWLEDGMENTS

 
We acknowledge the technical assistance of Christi Ralph and the haplotype analysis of Dr. Robert Freimuth. The clinical material evaluated in this study was collected and processed by the Siteman Cancer Center Tissue Procurement Core (Director, Mark Watson, M.D., Ph.D.).

	FOOTNOTES

 
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.

¹ Supported in part by the NIH Pharmacogenetics Research Network [U01 GM63340 (http://pharmacogenetics.wustl.edu)] and the Siteman Cancer Center (P30CA091842).

² To whom requests for reprints should be addressed, at Washington University School of Medicine, 660 South Euclid Avenue, Campus Box 8069, St. Louis, MO 63110. Phone: (314) 747-5183; Fax: (314) 362-3764; E-mail: hmcleod{at}im.wustl.edu

³ The abbreviations used are: 5-FU, 5-fluorouracil; FDXR, ferredoxin reductase; SNP, single nucleotide polymorphism.

⁴ The sequence information for primers and probes in this study is available upon request.

Received 6/25/03. Revised 7/24/03. Accepted 8/ 6/03.

	REFERENCES

Top ABSTRACT Introduction Materials and Methods Results Discussion REFERENCES

 

1. Longley D. B., Harkin D. P., Johnston P. G. 5-fluorouracil: mechanisms of action and clinical strategies. Nat. Rev. Cancer, 3: 330-338, 2003.[Medline]
2. Kurtz J. E., Andres E., Natarajan-Ame S., Noel E., Dufour P. Oral chemotherapy in colorectal cancer treatment: review of the literature. Eur. J. Intern. Med., 14: 18-25, 2003.[Medline]
3. Hwang P. M., Bunz F., Yu J., Rago C., Chan T. A., Murphy M. P., Kelso G. F., Smith R. A., Kinzler K. W., Vogelstein B. Ferredoxin reductase affects p53-dependent, 5-fluorouracil-induced apoptosis in colorectal cancer cells. Nat. Med., 7: 1111-1117, 2001.[Medline]
4. Liu G., Chen X. The ferredoxin reductase gene is regulated by the p53 family and sensitizes cells to oxidative stress-induced apoptosis. Oncogene, 21: 7195-7204, 2002.[Medline]
5. Muller J. J., Lapko A., Bourenkov G., Ruckpaul K., Heinemann U. Adrenodoxin reductase-adrenodoxin complex structure suggests electron transfer path in steroid biosynthesis. J. Biol. Chem., 276: 2786-2789, 2001.[Abstract/Free Full Text]
6. Hanukoglu I., Rapoport R., Weiner L., Sklan D. Electron leakage from the mitochondrial NADPH-adrenodoxin reductase-adrenodoxin-P450scc (cholesterol side chain cleavage) system. Arch. Biochem. Biophys., 305: 489-498, 1993.[Medline]
7. Rapoport R., Sklan D., Hanukoglu I. Electron leakage from the adrenal cortex mitochondrial P450scc and P450c11 systems: NADPH and steroid dependence. Arch. Biochem. Biophys., 317: 412-416, 1995.[Medline]
8. McLeod H. L., Yu J. Cancer pharmacogenomics: SNPs, chips, and the individual patient. Cancer Investig., 21: 628-638, 2003.
9. Watters J. W., McLeod H. L. Cancer pharmacogenomics: current and future applications. Biochem. Biophys. Acta, 1603: 99-111, 2003.[Medline]
10. Pfaffl M. W. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res., 29: 2002-2007, 2001.
11. Shirota Y., Stoehlmacher J., Brabender J., Xiong Y. P., Uetake H., Danenberg K. D., Groshen S., Tsao-Wei D. D., Danenberg P. V., Lenz H. J. ERCC1 and thymidylate synthase mRNA levels predict survival for colorectal cancer patients receiving combination oxaliplatin and fluorouracil chemotherapy. J. Clin. Oncol., 19: 4298-4304, 2001.[Abstract/Free Full Text]
12. Velculescu V. E., Madden S. L., Zhang L., Lash A. E., Yu J., Rago C., Lal A., Wang C. J., Beaudry G. A., Ciriello K. M., Cook B. P., Dufault M. R., Ferguson A. T., Gao Y., He T. C., Hermeking H., Hiraldo S. K., Hwang P. M., Lopez M. A., Luderer H. F., Mathews B., Petroziello J. M., Polyak K., Zawel L., Zhang W., Zhang X., Zhou W., Haluska F. G., Jen J., Sukumar S., Landes G. M., Riggins G. J., Vogelstein B., Kinzler K. W. Analysis of human transcriptomes. Nat. Genet., 23: 387-388, 1999.[Medline]
13. Marsh S., Kwok P., McLeod H. L. SNP databases and pharmacogenetics: great start but a long way to go. Hum. Mutat., 20: 174-179, 2002.[Medline]
14. Ronaghi M. Pyrosequencing sheds light on DNA sequencing. Genome Res., 11: 3-11, 2001.[Abstract/Free Full Text]
15. Schaid D. J., McDonnell S. K., Wang L., Cunningham J. M., Thibodeau S. N. Caution on pedigree haplotype inference with software that assumes linkage equilibrium. Am. J. Hum. Genet., 71: 992-995, 2002.[Medline]
16. Oren M. Regulation of the p53 tumor suppressor protein. J. Biol. Chem., 274: 36031-36034, 1999.[Free Full Text]
17. Unger T., Sionov R. V., Moallem E., Yee C. L., Howley P. M., Oren M., Haupt Y. Mutations in serines 15 and 20 of human p53 impair its apoptotic activity. Oncogene, 18: 3205-3212, 1999.[Medline]
18. Lowe S. W., Bodis S., McClatchey A., Remington L., Ruley H. E., Fisher D. E., Housman D. E., Jacks T. p53 status and the efficacy of cancer therapy in vivo. Science (Wash. DC), 266: 807-810, 1994.[Abstract/Free Full Text]
19. LeBlanc H., Lawrence D., Varfolomeev E., Totpal K., Morlan J., Schow P., Fong S., Schwall R., Sinicropi D., Ashkenazi A. Tumor-cell resistance to death receptor–induced apoptosis through mutational inactivation of the proapoptotic Bcl-2 homolog Bax. Nat. Med., 8: 274-281, 2002.[Medline]
20. Lee J. M. Inhibition of TP53-dependent apoptosis by the KIT tyrosine kinase: regulation of mitochondrial permeability transition and reactive oxygen species generation. Oncogene, 17: 1653-1662, 1998.[Medline]
21. Li P. F., Dietz R., von Harsdorf R. TP53 regulates mitochondrial membrane potential through reactive oxygen species and induces cytochrome c-independent apoptosis blocked by Bcl-2. EMBO J., 18: 6027-6036, 1999.[Medline]
22. Violette S., Poulain L., Dussaulx E., Pepin D., Faussat A. M., Chambaz J., Lacorte J. M., Staedel C., Lesuffleur T. Resistance of colon cancer cells to long-term 5-fluorouracil exposure is correlated to the relative level of Bcl-2 and Bcl-X(L) in addition to Bax and TP53 status. Int. J. Cancer, 98: 498-504, 2002.[Medline]
23. Drysdale C. M., McGraw D. W., Stack C. B., Stephens J. C., Judson R. S., Nandabalan K., Arnold K., Ruano G., Liggett S. B. Complex promoter and coding region ß2-adrenergic receptor haplotypes alter receptor expression and predict in vivo responsiveness. Proc. Natl. Acad. Sci. USA, 97: 10483-10488, 2000.[Abstract/Free Full Text]

This article has been cited by other articles:

				I. Shureiqi, X. Zuo, R. Broaddus, Y. Wu, B. Guan, J. S. Morris, and S. M. Lippman The transcription factor GATA-6 is overexpressed in vivo and contributes to silencing 15-LOX-1 in vitro in human colon cancer FASEB J, March 1, 2007; 21(3): 743 - 753. [Abstract] [Full Text] [PDF]

				G. Kroemer, L. Galluzzi, and C. Brenner Mitochondrial Membrane Permeabilization in Cell Death Physiol Rev, January 1, 2007; 87(1): 99 - 163. [Abstract] [Full Text] [PDF]

				L. Gao, Y. Feng, R. Bowers, M. Becker-Hapak, J. Gardner, L. Council, G. Linette, H. Zhao, and L. A. Cornelius Ras-Associated Protein-1 Regulates Extracellular Signal-Regulated Kinase Activation and Migration in Melanoma Cells: Two Processes Important to Melanoma Tumorigenesis and Metastasis Cancer Res., August 15, 2006; 66(16): 7880 - 7888. [Abstract] [Full Text] [PDF]

				I. Shureiqi, Y. Wu, D. Chen, X. L. Yang, B. Guan, J. S. Morris, P. Yang, R. A. Newman, R. Broaddus, S. R. Hamilton, et al. The Critical Role of 15-Lipoxygenase-1 in Colorectal Epithelial Cell Terminal Differentiation and Tumorigenesis Cancer Res., December 15, 2005; 65(24): 11486 - 11492. [Abstract] [Full Text] [PDF]

				L. Gao, H. Deng, H. Zhao, A. Hirbe, J. Harding, L. Ratner, and K. Weilbaecher HTLV-1 Tax transgenic mice develop spontaneous osteolytic bone metastases prevented by osteoclast inhibition Blood, December 15, 2005; 106(13): 4294 - 4302. [Abstract] [Full Text] [PDF]

				E. A. Kidd, J. Yu, X. Li, W. D. Shannon, M. A. Watson, and H. L. McLeod Variance in the Expression of 5-Fluorouracil Pathway Genes in Colorectal Cancer Clin. Cancer Res., April 1, 2005; 11(7): 2612 - 2619. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Yu, J. Articles by McLeod, H. L. Search for Related Content PubMed PubMed Citation Articles by Yu, J. Articles by McLeod, H. L.