Hunting for disease genes in multi-functional diseases

Disease genes may be identified through functional, positional, and candidate gene approaches. Although extensive and often labor-intensive studies such as family linkage analysis, functional investigation of gene products and genome database searches are usually involved, thousands of human disease genes, especially for monogenic diseases with Mendelian transmission, have been identified. However, in diseases caused by more than one gene, or by a combination of genetic and environmental factors, identification of the genes is even more difficult. Common examples include atherosclerosis, cancer, Alzheimer's disease, asthma, diabetes, glaucoma, and age-related macular degeneration. There have been conflicting reports on the roles of associated genes. Even with population-based case-control studies and new statistical methods such as the sib-ship disequilibrium test and the discordant alleles test, there is no agreement on whether alpha2-macroglobulin (A2M) is a gene for Alzheimer's disease. Another example is the inconsistent association between age-related macular degeneration and ATP-binding cassette transporter (ABCR). Ethnic variation causes further complications. In our investigation of LDL-receptor variants in familial hypercholesterolemia, and the trabecular meshwork inducible glucocorticoid response protein, or myocillin (TIGR-MYOC) mutation pattern in primary open angle glaucoma, we did find dissimilar results in Chinese compared to Caucasians. New information from the Human Genome Project and advancements in technologies will aid the search for and confirm identification of disease genes despite such challenges.

The apolipoprotein E epsilon4 allele is unlikely to be a major risk factor of age-related macular degeneration in Chinese

Apolipoprotein E (ApoE) is a major transporter of lipids and cholesterol in the nervous system. Age-related macular degeneration (ARMD), characterized by drusen containing lipids, was reported to show a lower frequency of the ApoE epsilon4 allele than control subjects. We sought to examine the association of this polymorphism with ARMD in Hong Kong Chinese. Among 98 ARMD subjects, the frequency of epsilon4 carriers showed a trend toward a decrease compared to controls, but it was not significant (11.2 vs. 15.0%, p < 0.52). The association of epsilon4 with an apparent reduced risk of ARMD was reported previously in the exudative form of the disease, however among 39 exudative ARMD patients there was also no significant difference in epsilon4 frequency (12.8%, p < 0.93). The lack of a statistically significant effect of epsilon4 may be due to the lower frequency of epsilon4 in Chinese than Europeans. Thus we cannot exclude a possible effect of this allele on ARMD risk, but we can conclude that this allele is likely not a major factor influencing ARMD risk in the Chinese.

Roles for lipoprotein lipase in Alzheimer's disease: an association study

Lipoprotein lipase (LPL) assists lipid transport by transferring lipids between lipoprotein particles and cells. LPL binds apolipoprotein E (apoE) lipoprotein particles and a major apoE receptor, low density lipoprotein receptor related protein (LRP). Because apoE and LRP polymorphisms alter Alzheimer's disease (AD) risk, and LPL itself is found in AD amyloid plaques, we examined whether LPL variants also affect AD risk. In case-control studies in the United States and Canada, the frequencies of two LPL alleles known to affect LPL enzymatic activity were measured in Caucasian AD or elderly normal (N) subjects. Pathologically confirmed subjects in both studies exhibited similar trends toward fewer 447Ter and more 291Ser alleles in AD. Combining results from both countries gave allele frequencies for 447Ter of 13.7% (26/190) in N and 9.4% (80/852) in AD (P = 0.10), and for 291Ser of 0.0% (0/184) in N and 1. 3% (8/636) in AD (P = 0.21). The trend appeared even greater for homozygous 447Ter subjects: 4.2% (4/95) of N vs. 1.4% (6/426) of AD (P = 0.09). These trends are consistent with a putative protective effect of 447Ter and causative effect of 291Ser on AD. Furthermore, brains of AD patients with 447Ter showed trends toward fewer plaques, tangles, and glia, and more neurons and cortical thickness than AD patients without 447Ter. Hippocampal plaques were significantly reduced. LPL might affect hippocampal function and thus dementia via its role as supplier of membrane components or antioxidants to neurons. Alternatively, LPL may play a part in plaque formation through its interaction with apoE and LRP.

Apolipoprotein E isoforms in Alzheimer's disease pathology and etiology

The apolipoprotein E (apoE) epsilon4 allele increases risk of Alzheimer's disease (AD), perhaps by accelerating plaque formation, or by impairing neuron repair. Considerable evidence supports both mechanisms. AD patients with epsilon4 have more and earlier amyloid deposits than do patients without epsilon4. The same is true of non-demented control subjects. In vitro, all apoE isoforms inhibit amyloid beta protein (Abeta) aggregation, but apoE4 less effectively than apoE3. Transgenic amyloid-producing mice expressing apoE3 or apoE4 develop less Abeta deposition than apoE knockout mice. These observations are consistent with an effect of apoE isoforms on Abeta aggregation in AD. ApoE is important for neurite maintenance since apoE knockout mice lose neurites and suffer behavioral deficits with aging or treatment with excitotoxins. ApoE4 mice show similar defects, but apoE3 mice are normal. AD patients with epsilon4 show more neuritic deficits than epsilon3 carriers. ApoE epsilon4 also worsens neurological impairment in head injury, stroke, and multiple sclerosis. Thus, apoE4 is less effective at neurite maintenance. Perhaps epsilon4 increases AD risk by both mechanisms: allowing amyloid deposition and failing to repair neurites. In either case, introducing apoE3 or apoE2 into the brain, for example by gene therapy or cell grafts, might delay AD progression.

Truncations in the TIGR gene in individuals with and without primary open-angle glaucoma

PURPOSE

To investigate the coding exons in the trabecular meshwork-induced glucocorticoid response protein (TIGR) gene for mutations in primary open-angle glaucoma (POAG) in Chinese subjects.

METHODS

Ninety-one Chinese patients with POAG and 113 of their family members without glaucoma were screened for sequence alterations in the TIGR gene by polymerase chain reaction, conformation-sensitive gel electrophoresis, and DNA sequencing. One hundred thirty-two unrelated individuals without glaucoma, aged 50 years or more, were studied as control subjects.

RESULTS

Five sequence variants that lead to amino acid changes were identified. One was novel: Arg91Stop in one patient with POAG. Four had been reported: Arg46Stop in subjects with and without POAG, including an unaffected 77-year-old woman homozygous for Arg46Stop; Gly12Arg in subjects without glaucoma; and Asp208Glu and Thr353Ile in subjects with and without POAG. The previously reported 1-83(G-->A) and Arg76Lys polymorphisms were detected in both patients and controls and always occurred together.

CONCLUSIONS

A different pattern of TIGR sequence variants exists in the Chinese than in non-Chinese populations. No common TIGR mutation that causes POAG was found. The occurrence of subjects without glaucoma who are heterozygous or homozygous for Arg46Stop suggests that reduction in the amount of TIGR protein does not cause glaucoma. Thus, the TIGR missense mutations known to cause POAG probably do not cause glaucoma by inactivating a normal TIGR function, but rather through the gain of a pathologic function.

Compound heterozygosity of Leu252Val and Leu252Arg causing lipoprotein lipase deficiency in a chinese patient with hypertriglyceridemia

BACKGROUND

We investigated lipoprotein lipase (LPL) gene mutations in a Chinese male with severe hypertriglyceridemia and recurrent pancreatitis.

METHODS

We screened for LPL sequence mutation in the LPL gene in this patient, his relatives and 160 unrelated hypertriglyceridaemic subjects. We determined the postheparin plasma LPL activity of subjects carrying a LPL mutation and studied the in vitro expression of mutant LPL in COS-1 cells.

RESULTS

The proband was found to be a compound heterozygote for a novel Leu252Val and a reported Leu252Arg mutation in the LPL gene. He had low plasma levels of postheparin LPL activity and mass. The two mutations segregated independently in his family. In vitro expression analysis showed that Leu252Arg abolished both the catalytic function and secretion of LPL, while Leu252Val abolished the catalytic function but only reduced secretion by about half. We have also detected heterozygous Leu252Val and Leu252Arg mutations each in one hypertriglyceridaemic individual.

CONCLUSION

These results indicated that the leucine 252 is critical for the catalytic activity and secretion of LPL. Why the substitution by valine instead of arginine resulted only in a partial suppression of LPL secretion, remains to be investigated. Leu252Val and Leu252Arg are the likely cause of hypertriglyceridemia in these subjects because of their deleterious effects on LPL activity or secretion. Leu252Val/Leu252Arg is the first compound heterozygous mutation known to occur in the same codon of the LPL gene. So far they are found only in Chinese.

Apolipoprotein E genotype and its pathological correlation in Chinese Alzheimer's disease with late onset

In this study, we attempted to find a relationship between apoliprotein E (ApoE) genotypes and Alzheimer's disease (AD) pathology in different areas of the brain in Chinese. We also studied the borderline group of possible AD (Poss). There were 34 definite or probable AD (Ad), 18 Poss, and 123 brains from age-matched normal subjects (N). ApoE genotype was determined by nested polymerase chain reaction on genomic DNA extracted from archival paraffin-embedded materials. Hippocampus (including entorhinal cortex), amygdala, superior temporal lobe, middle frontal gyrus, and inferior parietal lobule of the brains of Ad and Poss were examined with beta amyloid (A beta) immunostaining, and the same regions plus medial occipital lobe were examined with tau immunostaining. The percentage of plaque area stained for A beta in each brain region was obtained by an image analyzer, and the average number of neurofibrillary tangles stained for tau was counted with an eyepiece graticule. ApoE epsilon4 frequency was increased in both Ad (22.1%, chi2, df = 1, P = .00005), and Poss (33.3%, P = .000005) compared with N (5.3%). A beta load was significantly increased in the neocortex in Ad examined with at least 1 copy of epsilon4 compared with subjects without epsilon4 (Mann-Whitney, P = .014). The same trend, though not statistically significant, occurred in Poss (P = .15). Tau expression was associated with ApoE epsilon4 in neither Ad nor Poss. Poss is genetically and histologically similar to Ad, although the overall A beta load is significantly increased in the latter. These findings support the recent Consensus Report's findings that all Alzheimer-type pathology may be significant.

Low-density lipoprotein receptor-related protein (LRP) gene 766T polymorphism and Parkinson's disease

The C766T polymorphism in exon 3 of the low-density lipoprotein receptor-related protein (LRP) gene is underrepresented in Alzheimer's disease (AD) compared with normal subjects. We examined this polymorphism in 186 patients with Parkinson's disease (PD) and 187 age-matched normal Chinese subjects in addition to 227 newborns representing the general population. The fraction of individuals with 766T was 12.8% in normal subjects and 11.3% in patients with PD, not a significant difference (p = 0.77). The odds ratio was 0.86 with a 95% confidence interval of 0.44-1.69, thus the LRP C766T polymorphism does not play a major role in risk for PD, although the possibility cannot be excluded that it plays a minor role or is a significant risk factor in other ethnic groups.

Apolipoprotein E promoter and alpha2-macroglobulin polymorphisms are not genetically associated with Chinese late onset Alzheimer's disease

In this study, we investigated two newly reported polymorphisms in association with late onset Alzheimer's disease (AD) in Chinese. They were a -491 A/T polymorphism in the Apolipoprotein E (APOE) promoter region and a five base pair deletion at exon 18 of alpha2-Macroglobin (A2M). There were 196 AD and 180 normal controls (N), which were age- and sex-matched. APOE epsilon4 alleles were significantly increased in AD vs. N (chi2 = 33.3, P < 0.000001). However, neither the -491 A/T (chi2 = 1.13, P = 0.29) nor A2M (chi2 = 0.18, P = 0.67) polymorphism was associated with AD risk, suggesting that these polymorphisms do not represent risk factors for AD in the Chinese population.

Lipoprotein lipase mutations and Alzheimer's disease

Lipoprotein lipase (LPL) helps transfer lipids from lipoprotein particles to cells. In the brain, LPL is present in Alzheimer's disease (AD) amyloid plaques. LPL binds apolipoprotein E (ApoE) lipoprotein particles and low-density lipoprotein receptor-related protein (LRP), an ApoE receptor. Since polymorphisms in both ApoE and LRP influence AD risk, we sought to determine whether LPL mutations also affect AD risk. In a case-control study, the frequencies of two of the most common known LPL mutations were measured in European-Americans either clinically diagnosed or pathologically confirmed as AD or normal control (N) subjects. In clinically diagnosed subjects, the Ser447Ter mutation comprised 9.8% (62/630) of alleles in N and 3.8% (9/238) in AD, a significant difference (P = 0.0057), while the Asn291Ser mutation comprised 1.1% (5/460) of alleles in N and 5.1% (8/158) in AD, also a significant difference (P = 0.0073), though in pathologically confirmed subjects the allele frequencies for AD did not significantly differ from N for either mutation. In clinically diagnosed subjects, LPL mutations were associated with altered AD risk, suggesting a potential role for LPL in the causation of AD. Further studies in different populations should help clarify the questions raised by these results.

Lipoprotein lipase mutations and Alzheimer's disease

Lipoprotein lipase (LPL) helps transfer lipids from lipoprotein particles to cells. In the brain, LPL is present in Alzheimer's disease (AD) amyloid plaques. LPL binds apolipoprotein E (ApoE) lipoprotein particles and low-density lipoprotein receptor-related protein (LRP), an ApoE receptor. Since polymorphisms in both ApoE and LRP influence AD risk, we sought to determine whether LPL mutations also affect AD risk. In a case-control study, the frequencies of two of the most common known LPL mutations were measured in European-Americans either clinically diagnosed or pathologically confirmed as AD or normal control (N) subjects. In clinically diagnosed subjects, the Ser447Ter mutation comprised 9.8% (62/630) of alleles in N and 3.8% (9/238) in AD, a significant difference (P = 0.0057), while the Asn291Ser mutation comprised 1.1% (5/460) of alleles in N and 5.1% (8/158) in AD, also a significant difference (P = 0.0073), though in pathologically confirmed subjects the allele frequencies for AD did not significantly differ from N for either mutation. In clinically diagnosed subjects, LPL mutations were associated with altered AD risk, suggesting a potential role for LPL in the causation of AD. Further studies in different populations should help clarify the questions raised by these results.

Low density lipoprotein receptor related protein gene amplification and 766T polymorphism in astrocytomas

Low density lipoprotein receptor related protein (LRP) is a receptor for protease complexes, and may function in cell growth and repair, and in tumor invasiveness. LRP expression increases in glioblastomas compared to lower grade astrocytomas. Two potential mechanisms for this increased expression were investigated. The LRP C766T polymorphism is protective against Alzheimer's disease, perhaps through alteration of LRP expression. The frequency of the polymorphism was measured in astrocytoma patients and controls, but no significant difference was found. Differential PCR revealed LRP gene amplification in four of 25 high-grade gliomas and 0 of 23 other brain tumors. Co-amplification with epidermal growth factor receptor (EGFR) occurred in all four of the LRP-amplified tumors. Thus, LRP amplification may be partly responsible for increased LRP expression in astrocytomas, and may often occur in conjunction with EGFR amplification.

Alterations in cell surface carbohydrates on T cells from virally infected mice can distinguish effector/memory CD8+ T cells from naive cells

Glycosylation changes on surface molecules of T cells affect cell trafficking and function and may be useful in discriminating between naive, effector, and memory T cells. To analyze oligosaccharide structures on T cells activated in vivo, we examined alterations in sialic acid residues on T cells following infection of mice with lymphocytic choriomeningitis (LCMV), vaccinia virus, and vesicular stomatitis virus. We found that the majority of CD8 T cells from mice acutely infected with these viruses showed increased binding to peanut agglutinin (PNA). All of the PNAhighCD8 T cells from infected mice were CD44high, indicating that glycosylation changes were occurring on activated T cells. There was also an increase in the PNAhighCD4 T cell population in virally infected mice. Increased PNA binding to activated CD8 T cells correlated with higher endogenous neuraminidase levels in these cells. This higher neuraminidase activity most likely contributed to the PNAhigh phenotype by cleaving sialic acid residues off the core-1 O-glycans or glycoproteins destined for the cell surface. A PNAhighCD8 T cell population persisted in immune mice that had cleared the LCMV infection. When spleen cells from immune mice were sorted into PNAhigh and PNAlow populations, >95% of the LCMV-specific memory CD8 T cells segregated with the PNAhigh population. This shows that virus-specific memory CD8 T cells remain hyposialylated and can be distinguished from naive CD8 T cells based on PNA binding. Thus, PNA can be used as a marker for Ag-experienced T cells.

Low density lipoprotein receptor related protein gene exon 3 polymorphism association with Alzheimer's disease in Chinese

Since apolipoprotein E4 (apoE4) is the major genetic risk for late onset Alzheimer's disease (AD), proteins that interact with apoE might be involved in AD pathogenesis. Low density lipoprotein receptor related protein (LRP) is an apoE receptor in the brain. In exon 3 of the LRP gene a polymorphism was found to be underrepresented in AD compared to normal Caucasian subjects (N). We examined this polymorphism in Chinese AD and N subjects. The polymorphism frequency in N was roughly half that reported for Caucasians. Compared to N, the frequency was significantly decreased in pathologically diagnosed, but not in clinically diagnosed AD patients. Thus, the role of the LRP exon 3 polymorphism in AD has now been demonstrated in two ethnic groups, suggesting the importance of LRP in AD pathogenesis.

No association detected between very-low-density lipoprotein receptor (VLDL-R) and late-onset Alzheimer's disease in Hong Kong Chinese

The epsilon4 allele of apolipoprotein E (ApoE) is a risk factor in late-onset Alzheimer's disease (AD). As a receptor for ApoE, very-low-density lipoprotein receptor (VLDLR) might be involved in AD pathogenesis. A Japanese study [Okuizimi, K., et al., Nature Genet., 11 (1995) 207-209] has shown an increased 5 and decreased 8 CGG-repeat allele frequency in the 5' untranslated region of VLDLR in Japanese AD versus normal controls (N). Subsequent studies in Caucasian Americans failed to duplicate the result. We examined this polymorphism in pathologically- or clinically-diagnosed Chinese late-onset AD. Our data did not show a significant increase in the 5 CGG-repeat in AD, thus suggesting no association to VLDLR. However, our data did show that the allele frequencies for each CGG-repeat were similar in both Chinese and Japanese.

Glycogen synthase kinase 3 alteration in Alzheimer disease is related to neurofibrillary tangle formation

Highly phosphorylated tau protein is the main component of paired helical filaments (PHF), which comprise the neurofibrillary tangles (NFT) in some neurons of patients with Alzheimer disease (AD). Glycogen synthase kinase 3 (GSK3) phosphorylates tau in vitro at several sites also found to be phosphorylated in PHF-tau; tau is phosphorylated at these sites in both AD and normal control (NC) brains, although the extent of phosphorylation is far greater in tau from AD. If GSK3 levels are increased in AD, then tau phosphorylation and perhaps PHF formation may occur. To quantify GSK3, blots of AD and NC brain supernatant and particulate fractions were probed with antibodies to GSK3. In particulate fractions of AD compared to NC, GSK3 alpha immunoreactivity did not increase, but in fact, decreased 40%, and GSK3 beta immunoreactivity decreased 30%. GSK3 alpha and GSK3 beta levels correlated well with each other. GSK3 levels correlated negatively with numbers of NFT.

Effect of intravenous immunoglobulin G on natural killer cell cytotoxicity in vitro in women with recurrent spontaneous abortion

Intravenous immunoglobulin (IVIg) has been used to treat women with recurrent spontaneous abortion (RSA), particularly for women with elevated natural killer (NK) cells. We investigated the effect of IVIg on peripheral blood NK cell activity in vitro in women with RSA. 51Cr-release assays using K562 in the presence of varying concentrations of IVIg were performed using PBL from 16 women with RSA. Antibody dependent cellular cytotoxicity (ADCC) was evaluated using Daudi cells. Effectors and targets were preincubated with IVIg. Binding of IVIg to K562 and Daudi was evaluated by flow cytometry. The effect of K562 absorbed IVIg on NK activity was compared to that of non-absorbed IVIg. NK cytotoxicity and ADCC in the presence of F(ab')2 fragments were compared with those in the presence of intact IVIg. IVIg produced a significant, dose dependent inhibition of NK activity in vitro. Inhibition of NK activity occurred when effectors but not targets were preincubated with IVIg. IVIg binds to K562 and Daudi. IVIg increased ADCC when targets but not effectors were incubated with IVIg. K562 absorbed IVIg produced more inhibition of NK cytotoxicity than non-absorbed IVIg. Suppression of NK cytotoxicity by F(ab')2 was as effective as that of IVIg. However, F(ab')2 did not increase ADCC. IVIg effectively reduces peripheral blood NK cytotoxicity in vitro. Inhibition of NK cytotoxicity is mediated at the effector cell level through the antigen binding portion of the immunoglobulins. Women with RSA and elevated NK cells may benefit from IVIg treatment.

Intravenous immunoglobulin inhibits natural killer cell activity in vivo in women with recurrent spontaneous abortion

We previously reported elevation of natural killer (NK) cells in women with recurrent spontaneous abortion (RSA) of immune etiology. In this study, we investigated the effect of intravenous immunoglobulin G (IVIg) on peripheral blood NK activity in vivo in women with RSA. Blood was drawn prior to and 7-11 days after IVIg therapy in eight women with RSA. NK activity was measured using K562 as target cells for 51Cr-release assays. Serum IgG concentrations were also measured. All received 400 mg/kg/day of IVIg for 3 consecutive days. 1) Seven of eight women became pregnant. Five delivered a live born infant. Three out of five women (60%) who delivered a live born infant showed a significant inhibition of NK cytotoxicity post IVIg and the rest did not show any changes; 2) NK cytotoxicity was significantly increased in a woman who miscarried again; 3) A woman who miscarried a chromosomally abnormal fetus showed a significant inhibition of NK cytotoxicity after IVIg; and 4) Serum IgG concentration increased significantly from 9.3 +/- 3.0 mg/ml to 23.5 +/- 5.1 mg/ml post IVIg therapy. IVIg effectively inhibits peripheral blood NK activity in vivo. These results are consistent with our previous finding showing that IVIg inhibits NK cell activity in vitro. Women with RSA and elevated NK cells may benefit from IVIg treatment.

Overexpressed tau protein in cultured cells is phosphorylated without formation of PHF: implication of phosphoprotein phosphatase involvement

Pyramidal neurons in affected regions of Alzheimer's disease (AD) brain contain neurofibrillary tangles (NFT), aggregates of paired helical filaments (PHF) composed mainly of phosphorylated microtubule-associated protein tau. To explore the role of tau phosphorylation in the aggregation of tau into PHF, we constructed mammalian cell culture systems producing high levels of intracellular phosphorylated tau. COS-1 fibroblast-like cells were transiently transfected to simultaneously express tau, MAP kinase (MAPK), and MAP kinase kinase (MAPKK), or alternatively to express tau and glycogen synthase kinase 3 (GSK3). B103 neuron-like cells (which contain MAPK but little tau or GSK3) were stably transfected to express tau or tau and GSK3. In both systems, GSK3-transfected cells contained tau AT8/M (defined by AT8 staining and tau PHF-like mobility), but MAPK-transfected cells required phosphatase inhibitors, such as okadaic acid (OKA) or calyculin (CAL), to produce tau AT8/M. In vitro, the same concentrations of CAL and OKA inhibit phosphatases 1 and 2A (PP1 and PP2A), except that 100-1000 times as much OKA is needed to inhibit PP1. Inducing tau phosphorylation at the AT8 site in MAPK-transfected cells required 2-10 times more OKA than CAL, suggesting both PP1 and PP2A helped block the phosphorylation. Though levels of tau AT8/M reached 2-8% of total cellular proteins in COS-1 cells, the ratio of particulate to supernatant tau levels did not increase, and no tangles were observed; perhaps post-translational modifications or co-aggregating proteins are needed to induce PHF.