American Journal of Medical Genetics 42542-545 (1992)
Linkage Analysis in Juvenile Neuronal Ceroid Lipofuscinosis Jonathan L. Haines, Wei-Lang Yan, Rose-Mary Boustany, Ann Jewell, Cecile Julier, Xandra 0. Breakefield, a n d J a m e s F. Gusella Molecular Neurogenetics Laboratory, Massachusetts General Hospital, Charlestown (J.LH., W.-L.Y.,A.J., X.0.B ., J.F.G.); Division of Pediatric Neurology, Duke University Medical Center, Durham, North Carolina (R.-M.B.); C.E.P.H.,Paris, France (C.J.) Neuronal ceroid lipofuscinosis (NCL, Batten disease) is an autosomal recessive disease characterized by progressive mental retardation, cortical atrophy, seizures, and retinal degeneration. Several subtypes have been delineated on the basis of age-at-onset and histological characteristics;the most common is the juvenile (JNCL) form. Recently, the gene for JNCL was shown to reside on chromosome 16 through linkage studies to the haptoglobin locus and anonymous DNA markers using numerous European families. We have now examined &? families from North America with JNCL for linkage to markers in 16q21-23. Results in 3 families tend to support linkage to chromosome 16; 3 families remained uninformative, a n d 2 families produced negative lod scores in this region. A test of homogeneity was suggestive, but could not significantly reject the null hypothesis of homogeneity. We are continuing to collect families, particularly those with multiple living affecteds, and are identifying other probes in this region. Given close localization on chromosome 16 for JNCL, molecular strategies, including candidate gene strategies, are being explored. KEY WORDS: Batten disease, NCL, homogeneity, chromosome 16 INTRODUCTION Neuronal ceroid lipofuscinosis (NCL, Batten disease) is characterized by progressive mental retardation, cortical atrophy, seizures, and blindness resulting from retinal degeneration [Rider and Rider, 1988;Santavuori
et al., 19741. The pathology is characterized by an accumulation of autofluorescent lipopigment, usually in curvilinear bodies. Several subtypes have been delineated on the basis of age-at-onset and histological characteristics. The most common form is the juvenile (JNCL) form with onset generally between ages 5-10; an infantile (INCL) form with onset generally before one year, primarily found in Finland; a late infantile form (LINCL) with onset generally between ages 2 and 4; and adult forms (Kuf disease) with onset after age 15 also exist [Boustany et al., 1988; Hofman, 19901. With the exception of one adult form, all types of NCL are inherited as autosomal recessive traits. The incidence has been suggested to be 1-51100,000, indicating an allele frequency of about 11300, and a carrier frequency of about 1/150 [Rider and Rider, 1988; Hofman, 19901. The simple mode of inheritance of NCL allows the use of modern genetic techniques for identification of the defective gene. In 1989, Eiberg et al. presented initial evidence of linkage of JNCL to the haptoglobin locus on chromosome 16. Gardiner et al. [19901 followed up this report with a more complete analysis using anonymous DNA probes in this region in a series of European families and provided overwhelming evidence of linkage to this region. Jarvela et al. [19911have reported that INCL does not link to chromosome 16, and, in fact, resides on chromosome 1.The location of the LINCL gene has not yet been determined. We undertook studies to confirm the localization of JNCL in North American families, to examine the potential for heterogeneity, and to more closely localize the defective gene for NCL.
MATERIALS AND METHODS Families Families from North America were identified primarily through the Batten Disease Support and Research Association. All patients were examined by oneof us (R.M.B.) and met classical criteria for a diagnosis of either Received for publication April 5, 1991; revision received AuJNCL or LINCL. A breakdown of the number of patients gust 30, 1991. Address reprint requests to Jonathan L. Haines, Molecular Neu- and families is given in Table I. Eight families (all rogenetics Laboratory, Building 149, 6th floor, Massachusetts JNCL) were chosen for analysis based on the number of potentially informative events and availability of DNA. General Hospital, Charlestown, MA 02129.
0 1992 Wiley-Liss, Inc.
Linkage in NCL TABLE I. Distribution of Affected and Unaffected Individuals Collected Class JNCL LINCL
Affected individuals 35 15
Families 26 10
Unaffected individuals 27 5
Probes Three probes (D16S4, D16S148, D16S150) were tested in these families. D16S4 was obtained from the ATCC; D16S148 and D16S150 were generated by one of us (C.J.). Probe DNA was prepared as previously described [Rouleauet al., 19901,as was Southern blotting. Endonuclease digestion of genomic DNA was performed as recommended by the manufacturer.
543
To more fully utilize the available marker information, and to give a possible indication of location of NCL relative to the markers, a 4-point multipoint analysis was performed. The overall multipoint curve is presented in Figure 1. The peak location score is 1.47 about 3 cM proximal to D16S150. The individual family multipoints are presented in Figure 2. A cursory examination of this curve suggests that a t least one family is showing substantial negative location scores throughout this region. To test the possibility of heterogeneity, the individual family location scores were entered into HOMOG for analysis. The estimated proportion of linked families is 0.72, but the significance level is only suggestive [x2(1) = 2.05, P = 0.083.
DISCUSSION Although the number of families we have studied to date is small, sufficient genetic information exists to Linkage Analysis detect linkage at over 10 cM if the tested marker is Two-point and multipoint linkage analyses were per- informative in every single meiotic event. Under the formed by using LINKMAP (V. 4.9) of the LINKAGE more realistic expectation of only a two-allele system, we package [Lathrop et al., 19841. Standard lod scores were could detect linkage up to 3 CMfrom the tested marker. generated for the two-point analyses. Location scores (log,, (likelihooddifferences))were generated for multi- Currently, we are intensifying our efforts to collect more point analyses. Homogeneity was tested using the families with both JNCL and LINCL and expect to douHOMOG program [Ott, 19851with results from the mul- ble or triple our sample size. Additionally, we have contipoint analyses. All calculations were done on a VAX centrated on using families with more than one sampled 8700 computer. Distances between markers were those affected individual. Although these are the most informative families, families with only a single sampled given in HGMlO [Keats et al., 19891. Simulation studies utilized the program SIMLINK affected individual and at least one unaffected sibling (version 4.1) [Ploughman and Boehnke, 19891. A two- will also contribute some information. These families allele marker system with equal allele frequencies was can be particularly helpful if recombination events are observed. assumed. Our two-point data are not particularly helpful in determining linkage andlor gene order. However, it is RESULTS clear that the defective gene for JNCL does not lie right The maximum potential lod scores a t various recom- next t o D16S148 or D16S4 (assuming homogeneity; see bination fractions for the 8 families under study are below). given in Table 11. The summary two-point results for The overall multipoint analysis does not improve the D16S4, D16S148, and D16S150 are presented in Ta- situation. The peak score of 1.47 (not surprisingly, near ble 111. All 3 markers have positive scores, but none the D16S150 locus) is barely different than the peak reach significance. D16S4 and D16S148 both show re- scores on either side of the map, and in any event, is not combination with the JNCL gene. enough to confirm linkage in these families. The most informative analysis is the test for homogeneity. Although we must caution that it is not formally signifiTABLE 11. Possible lod Scores From Simulation Studies cant, there is a suggestion that at least one family does on 8 Families not link to this region. There are several potential explanations for this result: First, simple laboratory errors Recombination fraction 0.00 0.05 0.10 (sample collection mix-up, mislabeled DNA samples, etc.) could result in apparent recombination. All samMaximum 7.47 6.69 4.51 ples have been rerun with the same results. We are now Mean exwcted 4.20 2.69 1.76 attempting to re-collect the family to eliminate the pos-
TABLE 111. Two-Point lod Scores for JNCL vs. Chromosome 16 Markers
e Marker D16S148 D16S150 D16S4
0.00 -m
1.36 -m
0.05 -0.05 1.29 0.24
0.10
0.15
0.20
0.28 1.15 0.45
0.37 0.96 0.51
0.35 0.79 0.49
0.30 0.22 0.41 0.32
0.40
e
2
0.06 0.12 0.10
0.38 1.36 0.53
0.17 0.00 0.17
544
Haineset al.
I
-0.3
1
-0.2
-0.1
0
0.2
0.1
0.3
0.4
Location S148
5150 54
defined. More markers on 16q need to be tested to localize the JNCL gene more closely and identify closely flanking markers. This will allow the use of linkage disequilibrium studies to identify polymorphisms residing very close to the gene. These can act as starting points for such techniques as walking and jumping to move toward the defective gene. As well, candidate genes identified from review of the biochemistry of NCL quickly can be tested for localization to the proper region of chromosome 16 once they are cloned. Obviously, any candidate gene residing in the proper region deserves intense scrutiny. Finally, the location of the gene for LINCL needs to be identified. It may be allelic to either the INCL or JNCL forms, or may be an entirely new locus. Efforts must be redoubled to collect enough families with LINCL to determine its location.
ADDENDUM Since the time that this paper was presented, we have had an opportunity to examine additional families and use additional markers. We have now studied a total of sibility of a sample switch at collection. No evidence of 20 families segregating JNCL and have examined them nonpaternity has been found. Second, the family could using the ordered set of markers Dl6Sl47-Dl6S67have a non-genetic form of JNCL. However, multiple D16S148. A multipoint analysis of these data provided cases have been observed in this family, which, given peak location scores for each interval of +2.0, +2.7, the rarity of this disorder, is virtually a guarantee of an - 0.07, and + 1.68. Thus we continue to see evidence for underlying genetic etiology. Third, this family might linkage of JNCL to this region of chromosome 16. Four not have classical JNCL. However, the family has been families continue to provide negative lod scores in this examined extensively, and all results are completely region, but a test of homogeneity is not significant. The coincident with the classical JNCL findings. Fourth, the family providing the strongest negative location scores markers tested are not close enough to the JNCL gene. has been recollected and the data regenerated with the There is no way to differentiate this possibility from same results. Thus laboratory error as the source of the true heterogeneity until more closely linked and/or negative scores can be ruled out. flanking markers to the JNCL gene are found. Clearly, further genetic analysis in NCL is necessary, ACKNOWLEDGMENTS and several avenues of investigation can be followed. We are indebted to the BDSRA and the families who The level of heterogeneity (if any) must be defined. Alagreed to participate in this study. This work was supready, the loci for JNCL and for INCL are known to ported in part by NIH grant PO1 NS 24279-05. reside on different chromosomes (16 and 1, respectively). The collection of more families will allow the REFERENCES extent (if any) of heterogeneity in JNCL to be closely Fig. 1. Total location scores for JNCL vs. D16S148, D16S150, D16S4. Locations are in cM.
1
I
I
I
0
-0.3
-0.2
-0.1
0.1
0
0.2
0.3
0.4
Location S148
S150 S4
Fig. 2. Location scores for each individual family for JNCL vs. D16S148, D16S150, D16S4. Locations are in cM.
Boustany RN, Alroy J , Kolodny EH (1988):Clinical classification of neuronal ceroid lipofuscinosis subtypes. Am J Med Genet [Suppll 5:47-58. Eiberg H, Gardiner RM, Mohr J (1989): Batten disease (SpielmeyerSjogren disease) and haptoglobins (HP): hint of linkage and assignment to chromosome 16. Clin Genet 36:217-218. Gardiner M, Sandford A, Deadman M, Poulton J, Cookson W, Reeders S, Jokaiho I, Peltonen L, Eiberg H, Julier C (1990): Batten disease (Spielmeyer-Vogt disease, juvenile onset neuronal ceroidlipofuscinosis) gene (CLN3) maps to human chromosome 16. Genomics 8:387-390, 1990. Hofman IL (1990):“The Batten-Spielmeyer-Vogt Disease.” Doorn, The Netherlands: Bartimeus Foundation. Jarvela I, Schleutker J , Haataja L, Santavuori P, Puhakka L, Manninen T, Palotie A, Sandkujili LA, Renlund M, White R, Aula P, Peltonen L. (1991): Infantile form of neuronal ceroid lipofuscinosis (CLN1) maps to the short arm of chromosome 1. Genomics 9:170173. Keats B, Ott J , Conneally M (1989): Report ofthe committee on linkage and gene order. Cytogenet Cell Genet 51:459-502. Lathrop GM, Lalouel JM, Julier J , Ott J (1984): Strategies for multilocus linkage analysis in humans. Proc Natl Acad Sci USA 81:3443-3446. Ott J (1985): “Analysis of Human Genetic Linkage.” - Baltimore: Johns Hopkins University F’ress, pp 111-119.
Linkage in NCL Ploughman LM, Boehnke M (1989):Estimating the power ofa proposed linkage study for a complex genetic trait. Am J Hum Genet 44543551. Rider JA, Rider DL (1988): Batten disease: Past, present, and future. Am J Med Genet ISupplI 5:21-26.
545
Rouleau GA, Bazanowski A, Gusella JF, Haines JL (1990): A genetic map of chromosome 1:comparison of different data sets and linkage programs. Genomics 7:313-318. Santavuori P, Haltia M, Rapola J, Raitta C, Keranen A (1974):Infantile type of so-called neuronal ceroidlipofuscinosis. Acta Genet Med Gemellol (Roma) 23:197-200.
Linkage Analysis in Juvenile Neuronal Ceroid Lipofuscinosis Jonathan L. Haines, Wei-Lang Yan, Rose-Mary Boustany, Ann Jewell, Cecile Julier, Xandra 0. Breakefield, a n d J a m e s F. Gusella Molecular Neurogenetics Laboratory, Massachusetts General Hospital, Charlestown (J.LH., W.-L.Y.,A.J., X.0.B ., J.F.G.); Division of Pediatric Neurology, Duke University Medical Center, Durham, North Carolina (R.-M.B.); C.E.P.H.,Paris, France (C.J.) Neuronal ceroid lipofuscinosis (NCL, Batten disease) is an autosomal recessive disease characterized by progressive mental retardation, cortical atrophy, seizures, and retinal degeneration. Several subtypes have been delineated on the basis of age-at-onset and histological characteristics;the most common is the juvenile (JNCL) form. Recently, the gene for JNCL was shown to reside on chromosome 16 through linkage studies to the haptoglobin locus and anonymous DNA markers using numerous European families. We have now examined &? families from North America with JNCL for linkage to markers in 16q21-23. Results in 3 families tend to support linkage to chromosome 16; 3 families remained uninformative, a n d 2 families produced negative lod scores in this region. A test of homogeneity was suggestive, but could not significantly reject the null hypothesis of homogeneity. We are continuing to collect families, particularly those with multiple living affecteds, and are identifying other probes in this region. Given close localization on chromosome 16 for JNCL, molecular strategies, including candidate gene strategies, are being explored. KEY WORDS: Batten disease, NCL, homogeneity, chromosome 16 INTRODUCTION Neuronal ceroid lipofuscinosis (NCL, Batten disease) is characterized by progressive mental retardation, cortical atrophy, seizures, and blindness resulting from retinal degeneration [Rider and Rider, 1988;Santavuori
et al., 19741. The pathology is characterized by an accumulation of autofluorescent lipopigment, usually in curvilinear bodies. Several subtypes have been delineated on the basis of age-at-onset and histological characteristics. The most common form is the juvenile (JNCL) form with onset generally between ages 5-10; an infantile (INCL) form with onset generally before one year, primarily found in Finland; a late infantile form (LINCL) with onset generally between ages 2 and 4; and adult forms (Kuf disease) with onset after age 15 also exist [Boustany et al., 1988; Hofman, 19901. With the exception of one adult form, all types of NCL are inherited as autosomal recessive traits. The incidence has been suggested to be 1-51100,000, indicating an allele frequency of about 11300, and a carrier frequency of about 1/150 [Rider and Rider, 1988; Hofman, 19901. The simple mode of inheritance of NCL allows the use of modern genetic techniques for identification of the defective gene. In 1989, Eiberg et al. presented initial evidence of linkage of JNCL to the haptoglobin locus on chromosome 16. Gardiner et al. [19901 followed up this report with a more complete analysis using anonymous DNA probes in this region in a series of European families and provided overwhelming evidence of linkage to this region. Jarvela et al. [19911have reported that INCL does not link to chromosome 16, and, in fact, resides on chromosome 1.The location of the LINCL gene has not yet been determined. We undertook studies to confirm the localization of JNCL in North American families, to examine the potential for heterogeneity, and to more closely localize the defective gene for NCL.
MATERIALS AND METHODS Families Families from North America were identified primarily through the Batten Disease Support and Research Association. All patients were examined by oneof us (R.M.B.) and met classical criteria for a diagnosis of either Received for publication April 5, 1991; revision received AuJNCL or LINCL. A breakdown of the number of patients gust 30, 1991. Address reprint requests to Jonathan L. Haines, Molecular Neu- and families is given in Table I. Eight families (all rogenetics Laboratory, Building 149, 6th floor, Massachusetts JNCL) were chosen for analysis based on the number of potentially informative events and availability of DNA. General Hospital, Charlestown, MA 02129.
0 1992 Wiley-Liss, Inc.
Linkage in NCL TABLE I. Distribution of Affected and Unaffected Individuals Collected Class JNCL LINCL
Affected individuals 35 15
Families 26 10
Unaffected individuals 27 5
Probes Three probes (D16S4, D16S148, D16S150) were tested in these families. D16S4 was obtained from the ATCC; D16S148 and D16S150 were generated by one of us (C.J.). Probe DNA was prepared as previously described [Rouleauet al., 19901,as was Southern blotting. Endonuclease digestion of genomic DNA was performed as recommended by the manufacturer.
543
To more fully utilize the available marker information, and to give a possible indication of location of NCL relative to the markers, a 4-point multipoint analysis was performed. The overall multipoint curve is presented in Figure 1. The peak location score is 1.47 about 3 cM proximal to D16S150. The individual family multipoints are presented in Figure 2. A cursory examination of this curve suggests that a t least one family is showing substantial negative location scores throughout this region. To test the possibility of heterogeneity, the individual family location scores were entered into HOMOG for analysis. The estimated proportion of linked families is 0.72, but the significance level is only suggestive [x2(1) = 2.05, P = 0.083.
DISCUSSION Although the number of families we have studied to date is small, sufficient genetic information exists to Linkage Analysis detect linkage at over 10 cM if the tested marker is Two-point and multipoint linkage analyses were per- informative in every single meiotic event. Under the formed by using LINKMAP (V. 4.9) of the LINKAGE more realistic expectation of only a two-allele system, we package [Lathrop et al., 19841. Standard lod scores were could detect linkage up to 3 CMfrom the tested marker. generated for the two-point analyses. Location scores (log,, (likelihooddifferences))were generated for multi- Currently, we are intensifying our efforts to collect more point analyses. Homogeneity was tested using the families with both JNCL and LINCL and expect to douHOMOG program [Ott, 19851with results from the mul- ble or triple our sample size. Additionally, we have contipoint analyses. All calculations were done on a VAX centrated on using families with more than one sampled 8700 computer. Distances between markers were those affected individual. Although these are the most informative families, families with only a single sampled given in HGMlO [Keats et al., 19891. Simulation studies utilized the program SIMLINK affected individual and at least one unaffected sibling (version 4.1) [Ploughman and Boehnke, 19891. A two- will also contribute some information. These families allele marker system with equal allele frequencies was can be particularly helpful if recombination events are observed. assumed. Our two-point data are not particularly helpful in determining linkage andlor gene order. However, it is RESULTS clear that the defective gene for JNCL does not lie right The maximum potential lod scores a t various recom- next t o D16S148 or D16S4 (assuming homogeneity; see bination fractions for the 8 families under study are below). given in Table 11. The summary two-point results for The overall multipoint analysis does not improve the D16S4, D16S148, and D16S150 are presented in Ta- situation. The peak score of 1.47 (not surprisingly, near ble 111. All 3 markers have positive scores, but none the D16S150 locus) is barely different than the peak reach significance. D16S4 and D16S148 both show re- scores on either side of the map, and in any event, is not combination with the JNCL gene. enough to confirm linkage in these families. The most informative analysis is the test for homogeneity. Although we must caution that it is not formally signifiTABLE 11. Possible lod Scores From Simulation Studies cant, there is a suggestion that at least one family does on 8 Families not link to this region. There are several potential explanations for this result: First, simple laboratory errors Recombination fraction 0.00 0.05 0.10 (sample collection mix-up, mislabeled DNA samples, etc.) could result in apparent recombination. All samMaximum 7.47 6.69 4.51 ples have been rerun with the same results. We are now Mean exwcted 4.20 2.69 1.76 attempting to re-collect the family to eliminate the pos-
TABLE 111. Two-Point lod Scores for JNCL vs. Chromosome 16 Markers
e Marker D16S148 D16S150 D16S4
0.00 -m
1.36 -m
0.05 -0.05 1.29 0.24
0.10
0.15
0.20
0.28 1.15 0.45
0.37 0.96 0.51
0.35 0.79 0.49
0.30 0.22 0.41 0.32
0.40
e
2
0.06 0.12 0.10
0.38 1.36 0.53
0.17 0.00 0.17
544
Haineset al.
I
-0.3
1
-0.2
-0.1
0
0.2
0.1
0.3
0.4
Location S148
5150 54
defined. More markers on 16q need to be tested to localize the JNCL gene more closely and identify closely flanking markers. This will allow the use of linkage disequilibrium studies to identify polymorphisms residing very close to the gene. These can act as starting points for such techniques as walking and jumping to move toward the defective gene. As well, candidate genes identified from review of the biochemistry of NCL quickly can be tested for localization to the proper region of chromosome 16 once they are cloned. Obviously, any candidate gene residing in the proper region deserves intense scrutiny. Finally, the location of the gene for LINCL needs to be identified. It may be allelic to either the INCL or JNCL forms, or may be an entirely new locus. Efforts must be redoubled to collect enough families with LINCL to determine its location.
ADDENDUM Since the time that this paper was presented, we have had an opportunity to examine additional families and use additional markers. We have now studied a total of sibility of a sample switch at collection. No evidence of 20 families segregating JNCL and have examined them nonpaternity has been found. Second, the family could using the ordered set of markers Dl6Sl47-Dl6S67have a non-genetic form of JNCL. However, multiple D16S148. A multipoint analysis of these data provided cases have been observed in this family, which, given peak location scores for each interval of +2.0, +2.7, the rarity of this disorder, is virtually a guarantee of an - 0.07, and + 1.68. Thus we continue to see evidence for underlying genetic etiology. Third, this family might linkage of JNCL to this region of chromosome 16. Four not have classical JNCL. However, the family has been families continue to provide negative lod scores in this examined extensively, and all results are completely region, but a test of homogeneity is not significant. The coincident with the classical JNCL findings. Fourth, the family providing the strongest negative location scores markers tested are not close enough to the JNCL gene. has been recollected and the data regenerated with the There is no way to differentiate this possibility from same results. Thus laboratory error as the source of the true heterogeneity until more closely linked and/or negative scores can be ruled out. flanking markers to the JNCL gene are found. Clearly, further genetic analysis in NCL is necessary, ACKNOWLEDGMENTS and several avenues of investigation can be followed. We are indebted to the BDSRA and the families who The level of heterogeneity (if any) must be defined. Alagreed to participate in this study. This work was supready, the loci for JNCL and for INCL are known to ported in part by NIH grant PO1 NS 24279-05. reside on different chromosomes (16 and 1, respectively). The collection of more families will allow the REFERENCES extent (if any) of heterogeneity in JNCL to be closely Fig. 1. Total location scores for JNCL vs. D16S148, D16S150, D16S4. Locations are in cM.
1
I
I
I
0
-0.3
-0.2
-0.1
0.1
0
0.2
0.3
0.4
Location S148
S150 S4
Fig. 2. Location scores for each individual family for JNCL vs. D16S148, D16S150, D16S4. Locations are in cM.
Boustany RN, Alroy J , Kolodny EH (1988):Clinical classification of neuronal ceroid lipofuscinosis subtypes. Am J Med Genet [Suppll 5:47-58. Eiberg H, Gardiner RM, Mohr J (1989): Batten disease (SpielmeyerSjogren disease) and haptoglobins (HP): hint of linkage and assignment to chromosome 16. Clin Genet 36:217-218. Gardiner M, Sandford A, Deadman M, Poulton J, Cookson W, Reeders S, Jokaiho I, Peltonen L, Eiberg H, Julier C (1990): Batten disease (Spielmeyer-Vogt disease, juvenile onset neuronal ceroidlipofuscinosis) gene (CLN3) maps to human chromosome 16. Genomics 8:387-390, 1990. Hofman IL (1990):“The Batten-Spielmeyer-Vogt Disease.” Doorn, The Netherlands: Bartimeus Foundation. Jarvela I, Schleutker J , Haataja L, Santavuori P, Puhakka L, Manninen T, Palotie A, Sandkujili LA, Renlund M, White R, Aula P, Peltonen L. (1991): Infantile form of neuronal ceroid lipofuscinosis (CLN1) maps to the short arm of chromosome 1. Genomics 9:170173. Keats B, Ott J , Conneally M (1989): Report ofthe committee on linkage and gene order. Cytogenet Cell Genet 51:459-502. Lathrop GM, Lalouel JM, Julier J , Ott J (1984): Strategies for multilocus linkage analysis in humans. Proc Natl Acad Sci USA 81:3443-3446. Ott J (1985): “Analysis of Human Genetic Linkage.” - Baltimore: Johns Hopkins University F’ress, pp 111-119.
Linkage in NCL Ploughman LM, Boehnke M (1989):Estimating the power ofa proposed linkage study for a complex genetic trait. Am J Hum Genet 44543551. Rider JA, Rider DL (1988): Batten disease: Past, present, and future. Am J Med Genet ISupplI 5:21-26.
545
Rouleau GA, Bazanowski A, Gusella JF, Haines JL (1990): A genetic map of chromosome 1:comparison of different data sets and linkage programs. Genomics 7:313-318. Santavuori P, Haltia M, Rapola J, Raitta C, Keranen A (1974):Infantile type of so-called neuronal ceroidlipofuscinosis. Acta Genet Med Gemellol (Roma) 23:197-200.
