Linked linear amplification for simultaneous analysis of the two most common hemochromatosis mutations.
Biochemical screening for HH is not specific and does not detect all homozygotes or compound heterozygotes for C282Y and H63D because of incomplete penetrance of these mutations (10). Several molecular diagnostic techniques have been applied for screening mutations in HFE, including PCR amplification coupled with a wide variety of approaches, including restriction endonuclease analysis (11), reverse dot-blot analysis (12), allele-specific PCR (13), heteroduplex technology (14), denaturing gradient gel electrophoresis (15), hybridization with fluorescent probes (16, 17), and single-strand conformation polymorphism analysis (18).
A novel in vitro DNA amplification technology, linked linear amplification (LLA), has recently been described in detail (19). The key feature of LLA is the use of multiple primers containing nonreplicable elements during the extension reactions. LLA is a target amplification method comparable in yield to PCR, but it is more resistant to false-positive results caused by carryover amplicon contamination. LLA is less susceptible than PCR to single-nucleotide polymorphisms (SNPs) that may occur in the primer regions used for DNA amplification. A SNP can hinder PCR primer performance, leading to reduced or negligible yields (20) even when abundant DNA is available. LLA uses multiple primers; therefore, if the ability of one primer to function is impaired, other primers may be available to compensate.
We now report the evaluation of LLA coupled with allele-specific oligonucleotide (ASO) hybridization (21) for determining the presence of the C282Y and H63D mutations in human genomic DNA.
Materials and Methods
DNA SAMPLE PREPARATION
Whole-blood samples were obtained from 344 individuals for whom a clinical laboratory test was ordered to determine the HFE genotype for clinical diagnosis. Analyses were performed in the Molecular Diagnostics Laboratory at the University of Michigan, and the study was approved by the University's Institutional Review Board. DNA was prepared using the Gentra Pure Gene DNA isolation reagent set. We used 1-10 [micro]L of the resulting DNA solution (~50-250 ng of DNA) for the amplification reactions.
All samples were genotyped for both C282Y and H63D HFE mutations by PCR coupled with restriction fragment length polymorphism (PCR-RFLP) analysis (7, 22, 23). S65C testing was performed on a subset of samples by PCR-RFLP analysis as described by Mura et al. (8). The operators of the PCR and LLA tests were blinded to the results of the comparative method until all testing was completed.
Synthetic primer and allele-specific capture oligonucleotides were obtained from DNA TechnologyA/S. The LLA primers used for the amplification of both regions of the HFE gene and their sequences are shown in Table 1. The primers were designed to avoid the region of a polymorphism in intron 4 that could potentially cause an inaccuracy in the genotyping of the 282 codon in some methods (20). The nucleotide sequences for the four ASOs are also provided in Table 1.
GENOTYPING BY THE LLA ASSAY
The Bio-Rad LLA-based hemochromatosis test was used according to the product instructions. The HFE product includes an Amplification Master Mix, which contains most of the necessary components for amplification. The
only component not included in the product is AmpliTaq DNA polymerase (PE Applied Biosystems). We used 2 U of the polymerase in each LLA reaction. For the amplification reactions, 40 [micro]L of the Amplification Master Mix reagent and 10 [micro]L of template DNA, for a final volume of 50 [micro]L, were placed in a 200-[micro]L thin-walled thermal cycle tube. Reactions were performed in thermal cyclers programmed for 1 cycle of 3 min at 94 [degrees]C and 35 cycles of 45 s at 94 [degrees]C, 30 s at 50 [degrees]C, and 30 s at 72 [degrees]C, followed by 1 cycle for 5 min at 72 [degrees]C.
After amplification, 1-10 [micro]L of the LLA products, 24-15 [micro]L of deionized water (total volume of sample and water was 25 [micro]L), and 25 [micro]L of Denaturing Solution reagent were mixed in a tube and incubated at room temperature for 10 min. For each DNA sample, a 40-[micro]L aliquot of Hybridization Buffer reagent was added to two wells of a color-coded (pink) strip containing the immobilized wild-type and mutant codon 282 ASOs and two wells of a light-green strip containing immobilized wild type and mutant ASOs for codon 63.  Subsequently, a 10-[micro]L aliquot of the denatured LLA products was added .to each of the four wells containing the Hybridization Solution. The solution in the wells was covered and mixed; the strips were then incubated at 37[degrees]C for 1 h. After hybridization, the wells were washed five times with Well Wash Buffer to remove any unbound amplification products from the ASOs immobilized in the wells. A 50-[micro]L aliquot of streptavidin-horseradish peroxidase conjugate reagent was then added to each well and incubated at 37 [degrees]C for 30 min. After a repeat of the wash procedure described above, a 50-[micro]L aliquot of 3,3',5,5' tetramethylbenzidine reagent was added to each well and incubated at room temperature for 10 min. The color reaction was terminated by the addition of 50 [micro]L of Stop Solution to each well.
[FIGURE 1 OMITTED]
For data analysis, the absorbance at 655 nm (background) was subtracted from the respective 450 nm absorbance. The minimum acceptable corrected absorbance values for N + M (wild-type + mutant alleles) was set previously at 0.60 based on earlier studies (data not shown). The analysis was repeated for any test that gave a N + M value < 0.60. The ratio of [A.sub.450nm] - 655nm for the wild-type allele (N) divided by the [A.sub.450nm] - 655nm for the mutant allele (M) was calculated for both mutations. The genotypic tests were interpreted as follows: a N/M ratio >8 corresponded to a wild type, a N/M ratio < 0.125 corresponded to a homozygous mutant, and a N/M ratio of 0.3-3.0 corresponded to a heterozygote. N/M ratios of 3-8 or 0.125-0.3 would indicate an indeterminate result and require that the assay be repeated. The cutoffs for the N/M ratios for the various HFE genotypes were determined in a previous study using several hundred specimens. From these previous studies, the tolerance ranges for the N/M ratios were determined so that no inaccurate genotype was reported. The present report was conducted in part to verify these cutoffs. In addition, previous studies indicated no relationship between the amount of input DNA into the LLA assay and the N/M ratio results (data not shown; internal study performed at Bio-Rad).
CONCORDANCE OF TEST RESULTS WITH THOSE OF PCRR-FLP ANALYSIS
The LLA method was used in combination with the ASO hybridization method to detect the two most common genetic mutations associated with hemochromatosis. After DNA purification, the HFE gene was amplified in a multiplex LLA reaction containing 27 LLA primers, the 4 innermost being biotin-labeled at the 5' terminus (Fig. 1). The LLA primers for in vitro amplification of the HFE gene were chosen to amplify two segments of the gene. A symmetric LLA reaction in which seven primers up stream and seven downstream of codon 63 were used provided adequate absorbance (N + M + 0.60). The genotype at codon 282 was determined with an asymmetric LLA reaction in which seven primers 5' of the codon and six primers 3' of the codon were used. The use of a seventh primer downstream of codon 282 did not increase sensitivity.
[FIGURE 2 OMITTED]
An aliquot of the LLA reaction was chemically denatured and added to each of four microwells in a 2 x 2 well grid. Each well contained a unique immobilized oligonucleotide probe designed to detect the wild-type codon 63 allele, mutant codon 63 allele, wild-type codon 282 allele, or mutant codon 282 allele. Immobilized ASO probes specifically captured complementary LLA products as detected by the colorimetric reaction described in the Materials and Methods. The ASO used for capturing LLA amplicons containing codon 63 did not contain a sequence domain complementary to codon 65. This design allowed amplicons containing either the wild-type codon 65 or the S65C mutation equally to hybridize to the microwells.
A diagram illustrating each of the six possible outcomes for analysis of the C282Y and H63D hemochromatosis mutations in the 2 2 microwell assay grid is shown in Fig. 2.5 In the case of a wild-type individual (Fig. 2A), a colored signal was generated only in the wells containing the wild-type ASO probes. An individual who was homozygous mutant for one of the two mutations produced a colored signal with the mutant ASO probe complementary to that mutated sequence, but not with its wild-type analog, whereas the other wild-type ASO showed the inverse pattern (Fig. 2, B and C). Individuals heterozygous for a single mutation generated a colored signal with both the wild-type and mutant probes specific for that mutation site and with the wild-type probe specific for the detection of the other mutation, but not with its mutant counterpart (Fig. 2, D and E). Compound heterozygous individuals generated colored signals with the two mutant ASO probes and their wild-type counterparts (Fig. 2F). All 320 analyses gave N + M absorbance values > 0.60, the minimum acceptable sum as a quality-control check.
To evaluate the analytical accuracy of the LLA hemochromatosis assay, a total of 344 purified and archived DNA specimens that had previously been genotyped by PCR-RFLP (7) were analyzed in the LLA test. The DNAs had been purified over the timeframe of 1997 to 2002 and had been stored at -20 to -80 [degrees]C. Ninety-three percent (320 of 344) of the DNA samples could be amplified by the LLA reaction. The DNA specimens that could not be amplified were presumably either degraded or of insufficient quantity for the LLA reaction. LLA can detect 200 copies of a DNA sequence, and this is typically adequate for genotyping clinical specimens. The limitation to the copy number derives from the number of primers that must be used to obtain the desired sensitivity. Additional primers require additional sequences flanking the region of interest. If the flanking region is degraded, then LLA performance will be reduced. Increased LLA yields can be obtained by increasing the number of cycles, but that is also possible for PCR. Modifying the LLA detection scheme (data not shown) can also significantly increase sensitivity. A PCR version of the HH assay was performed on the samples that failed in the LLA version of the HH assay. Of the 24 samples retested, 2 did not produce results by either method, and 22 gave positive results by PCR. The samples were probably degraded and fragmented, although this was not confirmed by gel analysis. The specimens analyzed by LLA in this study were up to 5 years old. On the basis of other experimentation (data not shown), fresh samples provided LLA-based results for all cases.
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
The amplified samples included the following reported genotypes as determined by PCR-RFLP analysis: wild-type 282 and 63 codons (n = 105), C282Y homozygous mutant (n = 54), C282Y heterozygous (n = 52), H63D homozygous mutant (n = 17), H63D heterozygous (n = 59), and compound H63D and C282Y heterozygous mutant (n = 33). No specimen gave an indeterminate result in any of the experiments reported here.
The ratios of N/(N + M) and M/(N + M) (see Materials and Methods) in the LLA-based assay were calculated and used to graph the results of the analyses of codons 282 and 63 for all of the specimens that amplified. The graph of N/(N + M) vs M/(N + M) for each specimen is shown in Fig. 3. The data obtained from the analysis of the 282 and 63 codons are shown in panels A and B of Fig. 3, respectively. Samples analyzed as wild type, heterozygous, or homozygous for the C282Y mutation clustered into three groups (Fig. 3A), based on their genotypes. The rectangular boundaries of the clusters were mathematically defined using the maximum allowed N/M ratios as provided in Materials and Methods. The three genotypic groups were clearly distinct from each other. In no case was a specimen found to be located outside of the cluster containing other samples of the same genotype. Samples analyzed as wild type, heterozygous, or homozygous for the H63D mutation were also clustered in three distinct groups (Fig. 3B), based on their genotypes. As with the C282Y mutation, the three genotypic groups were clearly distinct from each other, and no single sample graphed in a manner consistent with a genotype different from that expected. The genotype assignments determined by LLA and PCR-RFLP assays agreed for all 320 specimens (100%) tested.
EFFECT OF S65C MUTATION ON GENOTYPING OF CODON 63
All specimens genotyped as wild type for codons 282 and 63 (n = 105) and those genotyped as wild type for codon 282 and either heterozygous or homozygous for H63D (n = 76) were analyzed by PCR-RFLP for the presence of an S65C mutation. Three specimens (1%) had wild-type 282 and 63 codons but were heterozygous for the mutant S65C codon. These three specimens gave identical genotypes in the LLA-based test for the 282 and 63 codons, indicating that the LLA-based assay was unaffected by the presence of a S65C mutation.
The reproducibility of the LLA assay results was examined with four specimens. One sample each of a wild-type genotype, a C282Y wild-type/H63D homozygous mutant genotype, a C282Y homozygous mutant/H63D wild-type genotype, and a H63D/C282Y compound heterozygous genotype was assayed on each of 5 days. The starting material for each analysis was whole blood so that the entire process, including DNA purification, was included in each replicate assay. The results of this study are summarized in Fig. 4, in which the N/M ratio for each analysis was graphed for each of the assay days. The results obtained for codon 282 are shown in panels A and B of Fig. 4, and the results for codon 63 are shown in panels C and D. In all cases, the samples gave consistently unambiguous genotypic analyses; in no case did a specimen give a N/M ratio that approached a cutoff such that the assignment of the genotype was questionable or incorrect
Morbidity and mortality in individuals affected with HH could be significantly reduced if the disease were diagnosed before the onset of hepatic, cardiac, and endocrine dysfunction (2). HH is readily treated with periodic phlebotomy to remove excess iron in the blood. Early detection of the C282Y mutation is important because > 65% of individuals affected with hemochromatosis are homozygous for this mutation. The H63D mutation should also be detected, although it appears to be clinically significant only when associated with a C282Y mutation (compound heterozygote) (5, 23, 24). The relative penetrance of the compound heterozygote is ~0.5% of the homozygous C282Y genotype (25, 26). The need thus exists for an accurate test to rapidly determine the HFE genotypes at both codons 63 and 282.
We have demonstrated that the LLA method accurately genotypes the H63D and C282Y HFE mutations and that the presence of S65C does not interfere with genotyping codon 63. The LLA method is less labor-intensive and requires less time to complete than the commonly used PCR-RFLP assay. The HH-LLA assay as described involves the same amplification times and steps as PCRRFLP analysis. The ELISA can be assembled and run immediately after amplification. Data interpretation is much simpler than for PCR-RFLP analysis because color ratios in the wells indicate the presence or absence of a particular sequence in the sample. In contrast, PCR-RFLP analysis involves restriction enzyme digestion and analysis of the size of the resulting restriction fragments, usually by gel or capillary electrophoresis.
LLA has also been used to detect other SNPs associated with diseases, such as the factor V Leiden, factor II G20210A, and [beta]-globin gene mutations (19). Amplification of genomic DNA with LLA offers several advantages over PCR that make it an attractive approach in molecular diagnostics: lower carryover contamination, less likelihood of amplification failure in the presence of a SNP in the priming region (20), and high specificity attributable to the use of nested primers. The results reported here also clearly illustrate that LLA can be multiplexed for simultaneous testing of more than one mutation in the same amplification tube. Detecting HFE mutations by LLA does not require gel electrophoresis and requires less than one 8-h shift to complete. A single operator can simultaneously process up to three microwell plates, each with a capacity for 20 specimens, for a total of 60 samples in one batch with minimum instrumentation (i.e., plate washer and reader). Moreover, the microwell plate format is amenable to full automation and the batch handling of many clinical specimens.
In conclusion, we developed a new HFE genotype method that simultaneously detects H63D and C282Y mutations. The method is accurate, reproducible, and suitable for the direct detection of mutations from genomic DNA.
We thank Jennifer Howard for her expertise in conducting the PCR-RFLP testing. Dr. Killeen is a consultant for Bio-Rad, manufacturer of the Linked Linear Amplification system.
Received December 31, 2002; accepted April 25, 2003.
(1.) Burke W, Thomson E, Khoury MJ, McDonnell SM, Press N, Adams PC, et al. Hereditary hemochromatosis: gene discovery and its implications for population-based screening. JAMA 1998;280: 172-8.
(2.) Bacon BR, Powell LW, Adams PC, Kresina TF, Hoofnagle JH. Molecular medicine and hemochromatosis: at the crossroads. Gastroenterology 1999;116:193-207.
(3.) Online Mendelian Inheritance in Man. [TM] MIM No. 235200. Last edited April 30, 2003. Baltimore, MD: Johns Hopkins University. http://www.ncbi.nlm.nih.gov/omim/.
(4.) Beutler E, Bothwell TH, Charlton RW, Motulski AG. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease, 8th ed. New York: McGraw-Hill, 2001:2237-69.
(5.) Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA, Basava A, et al. A novel MHC class I-like gene is mutated in patients with hereditary hemochromatosis. Nat Genet 1996;13:399-408.
(6.) Trombini P, Mauri V, Salvioni A, Corengia C, Arosio C, Piperno A. S65C frequency in Italian patients with hemochromatosis, porphyria cutanea tarda and chronic viral hepatitis with iron overload. Haematologica 2001;86:316-7.
(7.) Wallace DF, Dooley JS, Walker AP. A novel mutation of HFE explains the classical phenotype of genetic hemochromatosis in a C282Y heterozygote. Gastroenterology 1999;116:1409-12.
(8.) Mura C, Raguenes O, Ferec C. HFE mutations analysis in 711 hemochromatosis probands: evidence for S65C implications in mild form of hemochromatosis. Blood 1999;93:2502-5.
(9.) Arya N, Chakrabrati S, Hegele RA, Adams P. HFE S65C variant is not associated with increased transferrin saturation in voluntary blood donors. Blood Cells Mol Dis 1999;25:354-7.
(10.) Beutler E, Felitti V, Gelbart T, Ho N. The effect of HFE genotypes on measurements of iron overload in patients attending a health appraisal clinic. Ann Intern Med 2000;133:329-37.
(11.) Jouanolle AM, Gandon G, Jezequel P, Blayau M, Campion ML, Yaouanq J, et al. Haemochromatosis and HLA-H. Nat Genet 1996;14:251-2.
(12.) Chehab FF. Molecular diagnostics: past, present, and future. Hum Mutat 1993;2:331-7.
(13.) Guttridge MG, Thompson J, Worwood M, Darke C. Rapid detection of genetic mutations associated with haemochromatosis. Vox Sang 1998;75:253-6.
(14.) Jackson HA, Bowen DJ, Worwood M. Rapid genetic screening for haemochromatosis using heteroduplex technology. Br J Haematol 1997;98:856-9.
(15.) Christiansen L, Bygum A, Thomsen K, Brandrup F, Horder M, Petersen NE. Denaturing gradient gel electrophoresis analysis of the hemochromatosis (HFE) gene: impact of HFE gene mutations on the manifestation of porphyria cutanea tarda. Clin Chem 1999;45:2025-6.
(16.) Mangasser-Stephan K, Tag C, Reiser A, Gressner AM. Rapid genotyping of hemochromatosis gene mutations on the Light-Cycler with fluorescent hybridization probes. Clin Chem 1999;45: 1875-8.
(17.) Ugozzoli LA, Chinn D, Hamby, K. Fluorescent multicolor multiplex homogeneous assay for the simultaneous analysis of the two most common hemochromatosis mutations. Anal Biochem 2002; 307:47-53.
(18.) Hertzberg MS, McDonald D, Mirochnik O. Rapid diagnosis of hemochromatosis gene Cys282Tyr mutation by SSCP analysis. Am J Hematol 1998;57:260-1.
(19.) Reyes AA, Ugozzoli LA, Lowery JD, Breneman JW 3rd, Hixson CS, Press RD, et al. Linked linear amplification (LLA): a new method for the amplification of DNA. Clin Chem 2001;47:31-40.
(20.) Jeffrey GP, Chakrabarti S, Hegele RA, Adams PC. Polymorphism in intron 4 of HFE may cause overestimation of C282Y homozygote prevalence in haemochromatosis. Nat Genet 1999;22:325-6.
(21.) Wallace RB, Shaffer J, Murphy RF, Bonner J, Hirose T, Itakura K. Hybridization of synthetic oligodeoxyribonucleotides to [phi][chi]174 DNA: the effect of single base pair mismatch. Nucleic Acids Res 1979;6:3543-57.
(22.) Carella M, D'Ambrosio L, Totaro A, Grifa A, Valentino MA, Piperno A, et al. Mutation analysis of the HLA-H gene in Italian hemochromatosis patients. Am J Hum Genet 1997;60:828-32.
(23.) Bartolo C, McAndrew PE, Sosolik RC, Cawley KA, Balcerzak SP, Brandt JT, et al. Differential diagnosis of hereditary hemochromatosis from other liver disorders by genetic analysis. Arch Pathol Lab Med 1998;122:633-7.
(24.) Beutler E, Gelbart T, West C, Lee P, Adams M, Blackstone R, et al. Mutation analysis in hereditary hemochromatosis. Blood Cells Mol Dis 1996;22:187-94.
(25.) Jouanolle AM, Fergelot P, Gandon G, Yaouanq J, Le Gall JY, David V. A candidate gene for hemochromatosis: frequency of the C282Y and H63D mutations. Hum Genet 1997;100:544-7.
(26.) Risch N. Haemochromatosis, HFE and genetic complexity. Nat Genet 1997;17:375-6.
 The specificity of the analyses for the wild-type and mutant codons of HFE is conferred exclusively by use of unique complementary ASO capture sequences in the wells. See Table 1 for the four sequences used in the HFE test.
 Nonstandard abbreviations: HH, hereditary hemochromatosis; nt, nucleotide; LLA, linked linear amplification; SNP, single-nucleotide polymorphism; ASO, allele-specific oligonucleotide; and RFLP, restriction fragment length polymorphism.
 Theoretically, there are three other possible codon 282/63 genotypes: compound homozygous, compound homozygous for C282Y/heterozygous for H63D, and heterozygous for C282Y/compound homozygous for H63D. No cases of any one of these three genotypes have been reported. For this reason, the patterns in the ELISA-ASO grid that would be obtained with these three genotypes are not given in Fig. 2. Any of these genotypes would, however, be readily evident from its expected unique pattern if such a genotype were present.
ANTHONY A. KILLEEN,  * JOHN W. BRENEMAN III,  ARLENE R. CARILLO,  JASON LIU,  AND CRAIG S. HIXSON  #
 Department of Pathology, University of Michigan, Ann Arbor, MI 48109.
 Clinical Diagnostics Group, Nucleic Acid Technology, Bio-Rad Laboratories, Hercules, CA 94547.
* Address correspondence to this author at: Department of Laboratory Medicine & Pathology, University of Minnesota, Mayo Mail Code 609, 420 Delaware St. SE, Minneapolis, MN 55455. E-mail firstname.lastname@example.org.
# Send reprint requests to this author at: Bio-Rad Laboratories, 5500 East Second St., Benicia, CA 94510. E-mail: email@example.com.
Table 1. Primer and ASO sequences. Primer no. Codon Primer sequence, (a) 5'-3' 1 282 Biotin-AGGGCTGGATAACCTTGXCTG 2 282 TCGAACCTAAAGACGTATTXCCC 3 282 AGCCAATGGATGCCAAGXAGT 4 282 CCATGAAGTGGCTGAAGGXTAA 5 282 GAACTACTACCCCCAGAAXATC 6 282 ATCATGTGACCTCTTCAGTXACC 7 282 CTCTTTCCTGTCAAGTGCXTCC 8 282 Biotin-ATCACAATGAGGGGCTGAXCCA 9 282 CTCTCATCAGTCACATACXCCA 10 282 CCCAATAGATTTTCTCAGXTCC 11 282 CTCATCTCACTGCCATAATXACC 12 282 CCCATCCCCTAACAAAGXGCA 13 282 GTTAAAGCCTTTGATTGCXACC 14 63 Biotin-AGCTTTGGGCTACGTGGXTGA 15 63 GACCTTGGTCTTTCCTTGXTTG 16 63 CTGTCTCCAGGTTCACACXCTC 17 63 TCCCCTCCTACTACACATXGTT 18 63 TGCAGGTGTGTGGAGCCXCAA 19 63 CTTCACAAAATGAGGACCXGAC 20 63 GGAAGAAAGACAGGACTGXAAC 21 63 Biotin-TGGCTTGAAATTCTACTGGXAAC 22 63 CCCACCCTTTCAGACTCXGAC 23 63 CCATAATAGTCCAGAAGTCAXCAG 24 63 TACCCTTGCTGTGGTTGTGXTTT 25 63 AAAAGCTCTGACAACCTCAXGAA 26 63 TGCTCCCACAAGACCTCXGAC 27 63 AGCAAATTCCTTCCCTCTXCCC Codon no. Genotype ASO sequence, 5'-3' 282 Wild type CCACCTGGCACGTATAT 282 Mutant ATATACGTACCAGGTGG 63 Wild type CTCATGATCATAGAACAC 63 Mutant ACTCTCATCATCATAGA (a) X in sequence indicates position of the nonreplicable element (propanediol).
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Molecular Diagnostics and Genetics|
|Author:||Killeen, Anthony A.; Breneman, John W., III; Carillo, Arlene R.; Liu, Jason; Hixson, Craig S.|
|Date:||Jul 1, 2003|
|Previous Article:||Biological threat detection via host gene expression profiling.|
|Next Article:||DNA integrity as a potential marker for stool-based detection of colorectal cancer.|