Linkage of Plasma Ab42 to a Quantitative Locus on Chromosome 10 in Late-Onset Alzheimer's Disease Pedigrees Nilufer Ertekin-Taner, Neill Graff-Radford, Linda H. Younkin, Christopher Eckman, Matthew Baker, Jennifer Adamson, James Ronald, John Blangero, Michael Hutton, and Steven G. Younkin

a: The probands for all families except 7 and 12 were seen by one of us (NGR) and diagnosed as probable AD based on NINCDS-ADRDA criteria. The probands for families 12 and 7 were determined to have AD through histories from multiple family members. None of the LOAD families collected had a family history suggesting autosomal dominant inheritance. This study was approved by the Mayo Clinic Institutional Review Board.

b: Four families were ascertained via an AD proband with extremely high plasma Ab42 and/or Ab40 levels (top 10th percentile of the 545 AD patients in our series).

c: The remaining six LOAD families were ascertained via an AD proband prior to measurement of Ab, and one of these families (d: Family 11) was subsequently found to have an AD patient with extremely high Ab.

e: Total number of family members with complete phenotypic information / Total number of people in the pedigrees.

f: Family members between the ages of 20 and 65 with complete phenotypic information.

g: Total number of LOAD cases in the pedigrees. (Total number of LOAD cases with DNA samples). The average number of LOAD cases was 4.7 and we had DNA samples from an average of 1 case (range: 0-3).

h: Ages of onset of the LOAD patients in the pedigrees. We had age of onset information for all of the probands and 3 non-proband ADs in these families. The values for families 9, 11, and 17 are an average of two ADs (1 proband and 1 non-proband).

Supplemental Table 1. Maximum multipoint lod scores
	ALL LOAD FAMILIES		EXTREME LOAD FAMILIES
CHROMOSOMEb	MLSa, c, e (cM)	p-valued	MLSa, c, f (cM)	p-valued
10	1.82 (82)	0.008	3.93 (81)	0.0001
19	0.49 (84)	0.11	0.12 (83)	0.27
1	0.07 (139)	0.33	0	-
5	0.11 (60)	0.29	0.20 (30)	0.22
9	0	-	0	-
12	0	-	0	-
21	0	-	0.35 (20)	0.16

a: Variance component method implemented in the software package SOLAR, was used to perform multipoint linkage analysis to detect quantitative trait loci determining the level of plasma Ab42. This method estimates the amount of variance in a quantitative trait due to a major gene (h²_q), residual genetic factors (h²_r) and environmental effects (e²), based on the covariance between arbitrary relative pairs. SOLAR employs an identity-by-descent allele sharing linkage analysis method that is applicable to multigenerational extended pedigrees of arbitrary size and structure. We tested for linkage by comparing the likelihood of the restricted model where the variance due to the QTL (h²_q) is fixed to 0 to that of a model where h²_q is estimated. The difference between the two log₁₀ likelihoods produces a lod score that is equivalent to the classical lod score of linkage analysis. The linkage analysis method implemented in SOLAR is model-free, and therefore does not require the specification of disease allele frequency, penetrance or mode of inheritance. DNA was extracted from peripheral blood leukocytes using routine methods. Genotypes were obtained using an ABI 377 sequencer and associated Genescan/Genotyper software packages. The pedigree structure, phenotypic and genotypic information was maintained in PEDSYS (http://www.sfbr.org/sfbr/public/software/pedsys/pedsys.html), which also produced output files for polygenic and linkage analysis with SOLAR. Plasma Ab42 was used as the quantitative trait, and analysis was restricted to the 20- to 65-year-old group to avoid confounding changes in plasma Ab42 that occur in subjects over age 65 in association with aging and cerebral Ab deposition (7). Since the variance component method is sensitive to outliers and non-normality of distribution, we used 10*log(Ab42) values in the analysis, and we excluded two values that were 4 SD above the mean to achieve a normal distribution compatible with the assumptions of the program. For all analyses, we calculated robust lod scores as descibed by Blangero et al. (13). We incorporated sex, age, age², sex*age, and sex*age² as covariates in the heritability estimations. None of these covariates were significant and they were, therefore, excluded from the multipoint linkage analysis.

b: Each of the five regions with MLS > 1 in the sib pair study of Kehoe et al. (12) was assessed initially using the polymorphic marker corresponding to the sib-pair peak along with 2 flanking markers within an average distance of 5.9 cM (2.7 to 10.2 cM). Two other regions were also analyzed with three markers: a region on chromosome 12 based on reports of linkage and association from several groups and the region surrounding the APP gene to test for changes in the promoter area.

c: When multipoint lod scores (MLS) were calculated using SOLAR, the regions on chromosome 10 and 19 gave maximum MLS of 1.08 and 0.99, respectively, whereas the other regions tested gave scores below 0.30. To pursue this result, we genotyped 4 additional markers in the chromosome 10 region and 2 more in the chromosome 19 region. Thus seven polymorphic markers were genotyped for locus 10, five for locus 19 and three for the rest of the loci.

d: For empirical calculations of the p-value, 10, 000 simulations under the null hypothesis of no linkage were carried out using the "lodadjust" option in SOLAR.

e: Maximum MLS were obtained for 10 LOAD pedigrees composed of 203 informative individuals and

f: 5 "Extreme" LOAD pedigrees composed of 124 informative individuals.

Supplemental Table 1. Twopoint LOD scores
		ALL LOAD FAMILIESa		EXTREME LOAD FAMILIESb
MARKER	LOCUS (cM)	TLS	p-valuec	TLS	p-valuec
D10S539	72.9	0.01	0.45	0.00	-
D10S1227	76.4	0.02	0.41	0.06	0.33
D10S589	79.2	0.15	0.25	1.92	0.01
D10S1225	80.9	1.03	0.04	0.57	0.10
D10S1211	82.5	0.58	0.10	0.69	0.08
D10S522	86.2	0.82	0.06	1.99	0.01
D10S1670	86.3	0.88	0.06	0.69	0.08

Twopoint LOD scores on chromosome 10, obtained from (a) LOAD families and (b) the "extreme families" subset. The locations of the marker loci were determined from the MAP-O-MAT program (URL: http://linkage.rockefeller.edu/mapomat). (c) Empirical p-values.