Detection of MTRR 66A-->G polymorphism using the real-time polymerase chain reaction machine LightCycler for determination of composition of allele after restriction cleavage.

The MTRR gene codes for methionine synthase reductase, one of the enzymes involved in the conversion of homocysteine to methionine. This conversion influences the overall level of total plasma homocysteine (tHcy) and mutations, which reduces the enzyme activity and results in an increased concentration of tHcy. A high homocysteine level is a well-documented independent risk factor for cardiovascular disease. A polymorphism in the gene for methionine synthase reductase (MTRR 66 A>G) has been shown to be associated with the risk of giving birth to a child with Down's syndrome, and the risk of having a foetus with neural tube defects. We have established a method for analysing MTRR 66A>G on DNA from dried blood spots using melting temperature analysis. The DNA was extracted from dried blood spots using a fast procedure by boiling only.

Upregulation of interleukin-12 and -17 in active inflammatory bowel disease.

BACKGROUND: Cytokines are essential mediators of the intestinal inflammation during active episodes of inflammatory bowel disease (IBD). Interleukin (IL)-12 and IL-17 are potent immunoregulatory cytokines whose roles in the pathogenesis of IBD are unknown. The aim of this study was to evaluate the colonic expression of IL-12 and IL-17 genes in IBD. METHODS: Fifty-one patients (22 with ulcerative colitis (UC), 17 with Crohn disease (CD), and 12 controls) who underwent colonoscopy were included. IBD disease activity was determined using a clinical grading scale. The degree of inflammation, as well as the content of CD4+ T cells (synthesizing IL-17) and CD68+ macrophages (synthesizing IL-12) in colonic biopsies, was determined. The amounts of IL-12 and IL-17 mRNA were assessed by RT-PCR, using GAPDH as an internal standard. RESULTS: In colonic specimens, IL-17 mRNA expression was increased in moderately and severely active UC (P = 0.03) and in all degrees of activity in CD (P < 0.04). Levels of IL-12 mRNA were upregulated in both active UC and active CD compared to controls (P < 0.02). In cases of remission, IL-12 mRNA expression was similar to that found in control samples. Compared to controls, histological examination showed significant differences in signs of chronic and acute inflammation in UC (P < 0.01) and CD (P < 0.02), revealing a high correlation between clinical disease activity and histological scoring (r2 = 0.92, P < 0.005). Whereas CD4+ T cells were observed in lymphocyte aggregates located profound in the lamina propria, CD68+ macrophages were primarily found just underneath the surface epithelium. The density of CD4+ and CD68+ cells correlated significantly with the amounts of IL-17 and IL-12 mRNA, respectively (P < 0.05). CONCLUSION: The expression of both IL-12 and IL-17 mRNA is induced in active UC and CD and may thus be involved in sustaining the intestinal inflammation in IBD. Inhibition of IL-12 or IL-17 might be future therapeutic targets in IBD.

Polymeric membranes for aromatic/aliphatic separation processes.

For the membrane-based separation of benzene/cyclohexane mixtures, the pervaporation properties of different 6FDA (4,4'-hexafluoroisopropylidene diphthalic anhydride) based copolyimide membranes have been investigated. In order to obtain high permeability as well as high selectivity copolyimides were synthesised using a combination of 4MPD (2,3,5,6-tetramethyl-1,4-phenylene diamine) and 6FpDA (4,4'-hexafluoro-isopropylidene dianiline) as well as DABA (3,5-diaminobenzoic acid) as monomers. Cross-linking is possible with this type of copolyimides and necessary to reduce swelling effects, which often occur with polymeric membranes and lead to a deterioration of the separation characteristics in aromatic/aliphatic separation processes. In order to find the most suitable membrane material, the polymer structure, the crosslinking agents, as well as the crosslinking methods have been varied. The separation characteristics have been determined by sorption and pervaporation experiments. Sorption experiments have been carried out at 60 degrees C using benzene, toluene and ethylbenzene as aromatic components and cyclohexane, cyclohexene, hexane and heptane as aliphatic solvents. Pervaporation experiments have been performed at 60 degrees C using benzene/cyclohexane mixtures with benzene concentrations covering the whole concentration range. It has been found that crosslinked copolyimide membranes show excellent chemical resistance, strongly reduced swelling behaviour and higher selectivity in aromatic/aliphatic separation compared to conventional membrane materials.

Sudden infant death syndrome, childhood thrombosis, and presence of genetic risk factors for thrombosis.

Sudden infant death syndrome or "cot death" has until the late eighties been a significant cause of death in children between the ages of 1 month and 1 year. Approximately two per 1000 children born alive dies of sudden infant death syndrome each year in Western Europe, North America, and Australia. The vulnerability of the infant brain stem to ischemia has been suggested to be a conceivable cause of sudden infant death syndrome. This is compatible with a hypothesis that genetic risk factors for cerebral thrombosis could cause microinfarction in the brain stem during the first month of life, affecting vital centers or their blood supply. The presence of three common point mutations seen in families with thrombophilia (1691G-->A in the coagulation factor V gene, 677C-->T in the methylenetetrahydrofolate reductase gene, and the 20210G-->A mutation in the prothrombin gene) could increase the risk for thrombosis in the child. This prompted us to investigate these genetic markers of thromboembolic disease in 121 cases of sudden infant death syndrome and in relevant controls, in the expectation of a more frequent occurrence of these markers if thrombosis is an etiological factor in sudden infant death syndrome. The frequency of homozygous 1691G-->A mutation in SIDS cases was higher than expected (odds ratio: 7.3, 95% confidence interval, 1.2-45.8). The allele frequencies (theta;) in cases of sudden infant death syndrome of the 1691G-->A, 677C-->T, and 20210G-->A alleles was 2.6% (1.0-5.5), 32.6% (26.8-38.9), and 0.9% (0.1-3.4), respectively. None of the allele frequencies found in the background population (3.4% for the 1691G-->A allele, 29% for the 677C-->T allele, and 1% for the 20210G-->A allele) differed significantly from that in cases of sudden infant death syndrome. In 5,251,027 inhabitants in Denmark, the incidence of venous thromboembolism was 0.9 per 1000 per year in the background population, and less than one-thousandth of these were children. Consequently it is not likely that venous thrombosis is a major cause of sudden infant death syndrome. On the other hand, this does not exclude other known or unknown risk factors for thrombosis as possible etiological factors for sudden infant death syndrome. It is likely that we must continuously employ the exclusion principle on possible etiological causes in genetic material from a large group of victims of sudden infant death syndrome if the phenomenon of sudden infant death syndrome is to be ascribed to a specific hereditary disorder.

Familial thrombophilia associated with homozygosity for the cystathionine beta-synthase 833T-->C mutation.

Severe hyperhomocysteinemia due to cystathionine beta-synthase (CBS) deficiency is a strong risk factor for premature cardiovascular disease. Among untreated patients, approximately 50% have suffered a thromboembolic event by 30 years of age. We report on 3 sisters with severe hyperhomocysteinemia due to homozygosity for the CBS 833T-->C mutation. These patients, who displayed no other known thrombophilic predisposition, had suffered single or multiple venous thrombosis before CBS deficiency was diagnosed relatively late in life. In this family, homozygosity for the 833T-->C mutation was associated with a mild phenotype with respect to other sequelae of CBS deficiency. Consequently, our results indicate that most cases with this genotype may remain undiagnosed. Investigated family members heterozygous for the 833T-->C mutation displayed normal total homocysteine in plasma (tHcy) levels, even when they were homozygous for the methylenetetrahydrofolate reductase 677C-->T polymorphism. The prevalence of homozygosity for the 833T-->C mutation has previously been estimated at no less than 1:20 500 in our population. Because a reduction of the severely elevated levels of tHcy in CBS deficiency reduces cardiovascular risk and because homozygosity for the 833T-->C mutation is more prevalent than previously thought, our results emphasize the importance of measuring tHcy routinely in thrombophilia screening.

Intermediate and severe hyperhomocysteinemia with thrombosis: a study of genetic determinants.

Hyperhomocysteinemia is an independent risk factor for cardiovascular disease. In search of genetic factors causing elevated levels of total homocysteine in plasma (tHcy), we investigated a cohort of consecutively identified, unrelated thrombosis patients (n = 28) having intermediate or severe hyperhomocysteinemia (30 micromol/l 100 micromol/l, respectively). The methylene-tetrahydrofolate reductase (MTHFR) 677C-->T genotype, and the complete cystathionine beta-synthase (CBS) genotype was determined in all patients. We found that the MTHFR T/T genotype was strongly correlated with intermediate hyperhomocysteinemia, being present in 73.9% of those cases (17 of 23). In three of five patients with severe hyperhomocysteinemia, compound heterozygosity for CBS mutations was detected. Among the mutations, two novel missense mutations: 1265C-->T (S422L) and 1397C-->T (S466L) were detected. The phenotype in those patients was quite mild, thromboembolism apart. This indicates that a search for CBS mutations in patients with severe hyperhomocysteinemia is important to ensure the detection of a possible CBS deficiency, thus enabling treatment. Co-existence of the MTHFR T/T genotype and the common CBS 844ins68 variant was significantly higher among patients (10.7%) as compared to controls (1.2%), indicating that this genotype combination is a thrombotic risk factor (P <0.05). In a few patients, hyperhomocysteinemia could not be explained by this genetic approach, suggesting that other genetic risk factors were implicated.

Thrombophilic predisposition in stroke and venous thromboembolism in Danish patients.

The present study describes 403 patients with thrombosis, from a uniform ethnic and geographical background. Two-hundred-and-seven individuals had suffered mild or moderate stroke and 196 individuals suffered venous thromboembolism. We recorded levels of antithrombin, protein C and protein S, plasminogen and plasma homocysteine, and the presence of the factor V Leiden mutation, the prothrombin 20210G-->A variant, and the methylenetetrahydrofolate reductase (MTHFR) 677C-->T polymorphism. Controls for the mutation frequencies consisted of Guthrie card blood spots from a cohort of new-born babies. The cumulative prevalence of deficiencies in antithrombin, protein C, protein S or plasminogen was 2.4% in patients with stroke and 11.2% in patients with venous thrombosis. The factor V Leiden mutation was present in 11.1% of patients with stroke and 26.5% of patients with venous thrombosis, compared with 6.6% of controls (n = 4188; P < 0.05 and P < 0.0001, respectively). The prevalence of the prothrombin 20210A variant was 3.1% in patients with venous thrombosis, 1.9% in patients with stroke and 2.0% in controls (n = 500; P > 0.05). Hyperhomocysteinemia was present in 16.0% of patients with stroke and 17.6% of patients with venous thrombosis. The prevalence of the MTHFR 677T/T genotype was no different in patients with stroke (10.6%) and venous thrombosis (8.7%) than in controls (8.3%; n = 1084; P > 0.05); thus, it apparently contributed to thrombosis only via its influence on total plasma homocysteine, which was significantly increased in patients with the T/T genotype (P < 0.001). The MTHFR T/T genotype did not further increase the risk for thrombosis in carriers of the factor V Leiden mutation. Overall, thrombotic events were associated with a known risk factor in 27% of patients with stroke and 55% of patients with venous thrombosis.

Indications of immunological tolerance in kidney transplantation.

As a complication to immunosuppressive treatment in allotransplantation, malignant diseases such as post-transplant lymphoproliferative disorder (PTLD) may occur. The patient in the present case is a 21-year-old man transplanted at the age of 11 with a kidney from his mother and at the age of 15 with a kidney from his father. During the immunosuppressive treatment the patient developed PTLD resulting in the withdrawal of the immunosuppressive drugs. At the time of writing, the immunosuppressive drugs have been withdrawn for more than 3 years. We report the findings of a state of donor-specific tolerance occurring after transplantation. Post-transplant cells from the patient show a non-reactive response in mixed lymphocyte cultures (MLCs) to cells from both the mother and the father. We demonstrate a reduction in the mRNA expression of the Thl cytokines IL-2 and IFN-gamma in the very same MLCs. The expression of Th1 cytokine mRNA was measured semi-quantitatively using competitive reverse transcription-polymerase chain reaction (RT-PCR). The reduction in the Th1 cytokine mRNA expression is not seen in the MLCs with patient cells against cells from a paternal HLA-A, B and DR-matched individual, suggesting the influence of other allorecognition factors than HLA-A, B and DR. Detection in vitro of a lowered expression of Th1 cytokine mRNA supports the notion of these mRNAs as indicators of post-transplant tolerance. Further studies will reveal whether the cytokine mRNA measurements on short time stimulated lymphocytes can be used more generally as a monitoring parameter of tolerance in kidney transplantation.

Detection of a novel deletion in the cystathionine beta-synthase (CBS) gene using an improved genomic DNA based method.

We elucidated the intron-exon boundaries of the 15 coding exons of the human cystathionine beta-synthase (CBS) gene in order to establish an improved method based on PCR and direct sequencing for detection of CBS mutations. Using this method we identified the pathogenic mutations in two Danish siblings with CBS deficiency. Patients were compound heterozygotes: we detected the 833T-->C mutation and a novel 22 bp deletion of exon 4 (493-514del) that introduces a frameshift and a stop codon immediately after the deletion. The deletion resulted in no detectable mRNA from this allele, as assessed by sequencing of cDNA. The established method represents an improvement of the existing method based on sequencing of cDNA because it permits the detection of mutations within the entire coding region of the CBS gene from a peripheral blood sample, including splice mutations and mutations resulting in the lack or a reduced amount of transcript.

Comparison of the level of cytokine mRNA in buffy coat-derived platelet concentrates prepared with or without white cell reduction by filtration.

BACKGROUND: The generation of proinflammatory cytokines in platelet concentrates (PCs) by white cells is thought to be implicated in febrile nonhemolytic reactions. Furthermore, other cytokines could be generated in the PCs as signs of white cell activation when PCs are prepared from a pool of buffy coats (BCs). The number of white cells in the PCs is crucial to cytokine generation. STUDY DESIGN AND METHODS: Each of the PCs (n = 12) was prepared from a pool of BCs from four donors. Before storage, half of the BC-derived PCs (BC-PCs) were white cell reduced by filtration. The BCs and the unfiltered and filtered BC-PCs were assayed for the presence of cytokine mRNA (i.e., interleukin [IL]-1beta, IL-6, tumor necrosis factor alpha [TNFalpha], IL-8, IL-2, and interferon gamma [IFN-gamma]) by the use of competitive reverse transcription-polymerase chain reaction. RESULTS: No mRNA of cytokines with pyrogenic activity, that is, IL-1beta, IL-6 and TNFalpha, was detected in either the filtered or the unfiltered BC-PCs. Likewise, IL-2 mRNA was not found in any of the BC-PCs. IFN-gamma mRNA, however, was detected in the unfiltered BC-PCs but not in the filtered BC-PCs. IL-8 mRNA was found in both the unfiltered and the filtered BC-PCs, but with a lower frequency in the filtered BC-PCs. CONCLUSION: The unfiltered BC-PCs produced in the top-and-bottom bag system contain traces or undetectable levels of the investigated cytokine mRNA. The results can be used in defining consensus recommendations for the use of filtered and unfiltered blood components.

Intestinal interleukin-8 concentration and gene expression in inflammatory bowel disease.

BACKGROUND: Interleukin-8 (IL-8) is an important cytokine for recruitment and activation of polymorphonuclear neutrophils (PMNs), cells that are abundant in the intestinal lesions of ulcerative colitis (UC) and Crohn's disease (CD). The present investigation was conducted to evaluate intestinal IL-8 concentration and IL-8 gene expression in parallel in inflammatory bowel disease (IBD) patients and a non-inflammatory control group. METHODS: The intestinal concentration of IL-8 was measured with a sandwich enzyme-linked immunosorbent assay (ELISA) technique (detection limit, 17.4 pg/mg protein), and relative quantitation of IL-8 mRNA transcript levels was done with a reverse transcription polymerase chain reaction (RT-PCR)-based method. Biopsy specimens from 66 humans who underwent colonoscopy--28 with UC, 18 with CD and colonic involvement, and 20 non-inflammatory disease-specific controls who subsequently were found to fulfill the diagnostic criteria for irritable bowel syndrome (IBS)--were included. None had received glucocorticoids within 3 months. RESULTS: Using a one-tailed variance analysis, a significant concordance between increasing IL-8 protein concentrations and disease activity was found both in UC and CD (P < 0.001), and only trace amounts were detected in IBS biopsy specimens. No differences were found between the two groups of UC and CD patients (P > 0.05), and no differences were found between quiescent IBD and IBS (P > 0.05). However, the PCR method showed IL-8 mRNA in 8 of 18 CD patients (44.4%; 95% confidence limits, 21.5-69.2%) and 7 of 28 UC patients (25.9%; 95% confidence limits, 11.1-46.3%), as compared with 0 of 20 IBS (P < 0.005). Increased IL-8 mRNA levels were found only in active CD, which was not the case in UC. No correlation was found between intestinal IL-8 ELISA and IL-8-mRNA levels (r = 0.24, P > 0.05). CONCLUSIONS: The observed correlation between disease activity and expression of the IL-8 gene in active CD colitis but not in UC and the increased IL-8 protein concentrations in affected intestinal segments of IBD as compared with the non-inflamed IBS indicate a possible transient IL-8 gene expression or altered mRNA stability in UC and CD, as is well known for other cytokines, such as IL-2. If so, it may form the basis of new therapeutic regimens for IBD like IL-10.

Involvement of interleukin-4 and -10 in inflammatory bowel disease.

The pathogenesis of ulcerative colitis (UC) and Crohn's disease (CD) may be associated with a decreased production of cytokines suppressing macrophage and T-cell functions: interleukins (IL) -4 and IL-10. Serum concentrations of IL-4 and IL-10 were measured using an ELISA technique, and intestinal IL-4 and IL-10 mRNA was detected by a reverse transcriptase polymerase chain reaction (RT-PCR) in 34 patients with inflammatory bowel disease (IBD) (20 with UC and 14 with CD) and compared to 12 control subjects. The superoxide production was measured spectrophotometrically in activated PMNs initially incubated in the presence of IL-4 or IL-10. No differences were found in numbers of cells that might be potential IL-4 or IL-10 producers (T cells, macrophages, B cells, and mast cells) in biopsy specimens using immuno- and histochemistry. IL-4 mRNA was detectable in specimens from 77.8% of the UC patients (P > 0.05) and 0% of the CD patients (P < 0.05), as compared to 81.8 in controls, and was significantly different (P < 0.0001) between UC and CD patients. The IL-10 amplification product was detectable in specimens from 30.0% UC patients (P < 0.003), but not in CD patients (78.6%, P > 0.05) as compared to controls (91.7%). The circulating protein levels of IL-4 were below the detection limit in all groups (detection limit 4 pg/ml), while the median IL-10 concentration was 12.5 pg/ml in UC, 18.1 pg/ml in CD, and 19.5 pg/ml among controls (detection limit 3 pg/ml), which did not differ in any of the three groups (P > 0.05). Finally, the superoxide production was inhibited and delayed by the addition of IL-10 (P < 0.01), whereas IL-4 only delayed this parameter. In conclusion, apart from the well-known suppressive effect on proinflammatory cytokine production, IL-4 delays and IL-10 inhibits superoxide generation. IL-4 mRNA expression is decreased in intestinal tissue from CD patients, while IL-10 mRNA expression is decreased in majority of UC patients, suggesting different immunopathogenesis of the two diseases.

Human blood eosinophils produce and secrete interleukin 4.

Interleukin 4 (Il-4) is an immunoregulatory cytokine which induces T-cell proliferation and differentiation into a Th2 phenotype, and is of particular importance for the induction of IgE synthesis. In the present study, the capability of human peripheral blood eosinophils from allergic and non-allergic donors to produce Il-4 was examined. Using reverse transcribed polymerase chain reaction (RT-PCR), it was shown that highly purified eosinophils from allergic patients express mRNA for Il-4. Resting eosinophils also gave specific immunoreactivity with anti-Il-4 antibodies, consistent with translation of Il-4 mRNA. Light and electron microscopic immunocytochemistry revealed that Il-4 was prestored in the eosinophilic granules. These results were confirmed by Il-4 specific ELISA which showed that Il-4 production could be upregulated in the eosinophils and released from the eosinophils following stimulation with the calcium ionophore A23187. These data indicate that eosinophils may be an important source of Il-4 at sites of allergic inflammation. Thus, eosinophils may act as immunomodulatory cells enhancing the allergic response through formation of Th2-cells and inducing the isotype switching to IgE in human B-cells.

A polymerase chain reaction-based method for the semiquantitative study of interleukin-8 mRNA in human basophil leukocytes.

We examined a potential method for quantitative analysis of cytokine expression patterns in purified human basophil leukocyte preparations. Basophil mRNA was reverse transcribed and cytokine cDNA levels determined by competitive polymerase chain reaction (PCR) with internal standard cDNAs constructed by site-directed mutagenesis. Co-reverse transcribed glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) cDNA levels were used as an internal control to correct for unequal efficiencies in RNA isolation and reverse transcription. The method was subjected to a statistical validation giving the within series precision of the analysis. This method was used to examine interleukin-8 expression patterns in basophils from different donors. The results from this study revealed increased interleukin-8 mRNA levels after in vitro challenge with a variety of stimuli.

One short well conserved region of Alu-sequences is involved in human gene rearrangements and has homology with prokaryotic chi.

Alu elements have repeatedly been found involved in gene rearrangements in humans. Although these elements have been suggested to stimulate gene rearrangements, sparse information is available for the possible mechanism(s) of these events. Here we present a compilation of Alu elements that have been involved in recombinational events leading to gene rearrangements, indicating the presence of a common 26 bp core sequence at or close to the sites of recombination. Besides the obvious possibility of retrotransposition, gene rearrangements may be induced by sequences that stimulate genetic recombination. We suggest that the core sequence stimulates recombination and may thereby cause the frequent involvement of these elements in gene rearrangements. Curiously, the core sequence contains the pentanucleotide motif CCAGC, which is also part of chi, an 8 bp sequence known to stimulate recBC mediated recombination in Escherichia coli.

A PvuII polymorphism of the low density lipoprotein receptor gene is not associated with plasma concentrations of low density lipoproteins including LP(a).

Lipoprotein(a) [Lp(a)] is a low density lipoprotein (LDL), in which apolipoprotein B-100 (apo B-100) is attached to apolipoprotein(a) [apo(a)], a glycoprotein of variable size. Lp(a) may be as atherogenic as LDL. In normal populations, Lp(a) concentrations in plasma are largely determined by the apo(a) gene locus on chromosome 6, but regulation of synthesis and catabolism of Lp(a) is poorly understood. In some studies, a PvuII restriction fragment length polymorphism (RFLP) in the LDL receptor gene seems to affect concentrations of LDL in plasma, and other studies have indicated that Lp(a) catabolism could be mediated by the LDL receptor. We therefore expected that the PvuII polymorphism in the LDL receptor gene might be associated with Lp(a) levels in 170 Caucasian men aged 40 years, selected to have a high representation of low molecular weight apo(a) phenotypes. However, plasma concentrations of cholesterol, LDL-cholesterol, HDL-cholesterol, triglycerides and Lp(a) were all unrelated to the LDL receptor gene PvuII polymorphism both in the group as a whole and when it was subgrouped by apo(a) phenotype. Therefore our data do not support the concept that this particular LDL receptor gene polymorphism is associated with LDL receptor function, and our data therefore neither support nor rule out a role for the LDL receptor in Lp(a) catabolism.

Detection of the apoB-3500 mutation (glutamine for arginine) by gene amplification and cleavage with MspI.

A single primer-template mismatch 2 bp from the apoB-3500 (G to A) mutation permits introduction of a cleavage site for MspI (C/CGG) in normal alleles but not in mutant alleles (CCAG). After amplification, cleavage, and polyacrylamide gel electrophoresis, normal and mutant alleles could be unambiguously distinguished. We constructed a positive (homozygous mutant) standard by site-directed mutagenesis. A negative standard was DNA from a homozygous normal subject. The method enables us to screen for the mutation with 12 microliters of spotted whole blood as the source of DNA.

DNA deletions in the low density lipoprotein (LDL) receptor gene in Danish families with familial hypercholesterolemia.

DNA samples from 25 unrelated Danish patients with familial hypercholesterolemia (FH) were screened by Southern blot hybridization to detect gross alterations in the low density lipoprotein (LDL) receptor gene. Three FH-patients were found to have a deletion. Two of these delete part of the cysteine rich domain, which comprises the ligand binding region of the LDL-receptor. The third deletion encompasses coding regions for the cytoplasmic part of the receptor. As two of these deletions could be equivalent to previously described LDL-receptor gene alterations, these data seem to support a notion of recombination hot spots which involve Alu-sequences.

Repetitive sequences involved in the recombination leading to deletion of exon 5 of the low-density-lipoprotein receptor gene in a patient with familial hypercholesterolemia.

Alu sequences in the low-density-lipoprotein (LDL) receptor gene are suspected of being of importance for the creation of gene defects leading to familial hypercholesterolemia (FH). One potential mechanism is that Alu sequences undergo homologous recombination, producing deletions or duplications of DNA segments on genomic DNA. In at least four cases (FH626, PO, JA and FH-DK3), a deletion of exon 5 of the LDL receptor gene has been reported. Only one of these (FH626) have so far been characterized in detail by sequence analysis and shown to involve two of the Alu repeated sequences, which are present in introns 4 and 5. We here report the complete characterization of FH-DK3 and show that the cross-over break points involve sequences similar, but not at identical positions in the 5' end, to those reported for FH626. The recombinations in both FH-DK3 and FH626 are suggested to have occurred within a 22-bp repeated sequence found in both junction alleles.