A Novel Variant of Lysosomal Acid Lipase (Leu336→Pro) Associated With Acid Lipase Deficiency and Cholesterol Ester Storage Disease
Abstract Cholesterol ester storage disease (CESD) is associated with premature atherosclerosis, hepatomegaly, elevated LDL cholesterol levels, and in most cases, low HDL cholesterol levels. Previous studies have shown a G→A mutation at the 3′ splice junction of exon 8 (E8SJM) of the gene encoding lysosomal acid lipase (LAL) in two kindreds with CESD. In a Canadian-Norwegian kindred with this disease, we show this mutation in conjunction with an as yet unknown T→C transition in exon 10 predicting a Leu336→Pro (L336P) replacement and an A→C transversion in exon 2 predicting a T-6P replacement in the prepeptide. Identification of the L336P rather than the T-6P replacement as the second defect underlying CESD in our patient is deduced from three lines of evidence. First, the E8SJM allele is located in cis with the mutation predicting the T-6P–encoding allele but in trans with the L336P-encoding allele; second, the L336P but not the T-6P replacement cosegregates with low LAL activity in the family; third, the T-6P replacement was found in 6 of 28 alleles from subjects with normal lysosomal acid lipase activity, suggesting that this variant represents a frequent nonfunctional polymorphism. Since the residual LAL activity is higher and the clinical phenotype based on plasma lipid values and severity of hepatosplenomegaly is milder in this case than in a previously studied case who was homozygous for the E8SJM allele, we conclude that the L336P variant appears to be associated with a phenotypically mild form of CESD.
- Received January 9, 1995.
- Accepted March 3, 1995.
Deficient activity of lysosomal acid lipase (LAL) is expressed as two major phenotypes: Wolman disease and cholesterol ester storage disease (CESD) (for review, see Reference 11 ). Wolman disease occurs in infancy and is nearly always fatal before the age of 1 year. Hepatosplenomegaly, steatorrhea, abdominal distension, adrenal calcification, and failure to thrive are observed in the first weeks of life. In contrast, CESD has a more benign clinical course. CESD, initially identified by Fredrickson,2 may not be detected until adulthood, and moderate hyperbetalipoproteinemia, hypertriglyceridemia, and hepatomegaly may be the only clinical signs. In addition, low plasma HDL cholesterol (HDL-C) and premature atherosclerosis occur in most cases with CESD.
The enzymatic defect has been demonstrated in several types of cells and tissues, including liver, spleen, lymph nodes, aorta, peripheral blood leukocytes, and cultured skin fibroblasts. Goldstein et al3 found higher residual activity of LAL in intact fibroblasts from patients with CESD than in those with Wolman disease, providing a biochemical explanation for the less severe phenotype associated with CESD. Cloning of a cDNA encoding human LAL revealed the primary structure of the enzyme.4 The deduced sequence is related to gastric and lingual lipases; it lacks significant homologies with neutral lipases. The gene is located on chromosome 10q23.2 through 10q23.35 and consists of 10 exons spanning approximately 45 kb.6 Klima et al7 have shown that combination of a null allele with an allele harboring a splice-junction mutation in exon 8 (E8SJM) of the gene encoding LAL may cause CESD. They showed that the G→A mutation in the last nucleotide of exon 8 leads to exon skipping, causing deletion of codons 254 through 277 in the LAL mRNA. However, current knowledge regarding genotype-phenotype relations in CESD is rather limited. We have diagnosed CESD in a Spanish subject homozygous for the E8SJM allele.8 Although his residual LAL activity was higher than in the compound heterozygous case containing the null allele, his clinical phenotype based on plasma lipid values was unexpectedly severe.
Since the lipid profile in CESD has a relatively high prevalence in the general population, it cannot be excluded at present that certain mutations in the LAL gene result in milder phenotypic forms than does the rare classic type of CESD normally associated with severe hepatomegaly. These mutations could contribute to mixed hyperlipidemia associated with increased risk for premature atherosclerosis in a larger fraction of the general population. Moreover, since heterozygotes can hardly be detected based on measurements of LAL activity alone, it cannot be excluded that heterozygosity for defects in the LAL gene also leads to an increased risk for hyperlipidemia and/or premature atherosclerosis. To answer these questions, it appears important to identify more heterozygotes and additional molecular defects in the gene encoding LAL, thus elucidating genotype-phenotype relations. We here report the underlying cause in a Canadian-Norwegian patient with an apparently mild form of CESD.
All persons gave their informed consent prior to their inclusion in the study. Patient A.S.P. was a 19-year-old woman with a diagnosis of CESD (Table 1⇓). Her mother (G.S.P.) and her sister (N.S.P.) had elevated fasting triglycerides and cholesterol values. Her father (D.R.), who originates from Canada, and one of her uncles, who lives in California, were not available for the analyses. The mother’s family comes from Norway, thus making consanguinity unlikely.
Determination of LAL and Acid β-Galactosidase Activities
Blood samples were collected with EDTA as anticoagulant and transported on ice from Oslo, Norway, or Sønderburg, Denmark, to the laboratory in Münster, Germany. White blood cells were isolated from 5 mL blood by centrifugation over a cushion of Ficoll-Paque ET (Pharmacia) according to the supplier’s instructions. The cells were washed once with 5 mL phosphate-buffered saline (137 mmol/L NaCl, 3 mmol/L KCl, 10 mmol/L Na2PO4, and 2 mmol/L KH2PO4, pH 7.4) and resuspended with 250 μL phosphate-buffered saline containing 1% Triton X-100. A cell extract was obtained by three freeze-thawing cycles followed by centrifugation in an Eppendorf bench-top centrifuge at maximal speed for 5 minutes at 4°C. Immediately afterward, the activities of LAL9 and acid β-galactosidase10 were measured by using p-nitrophenyl myristate or p-nitrophenyl-β-d-galactopyranoside (Sigma) as substrates.
Skin fibroblasts were obtained by cultivating a skin biopsy from the index patient (A.S.P.) according to standard procedures. The cells were scraped from the plates with a rubber policeman, and extracts were prepared as described above. LAL activity was measured with various p-nitrophenyl acyl esters9 (Sigma) and with cholesteryl[14C]oleate and [14C]triolein as described by Haley et al.11
Direct Sequencing of the LAL Gene
Genomic DNA was extracted from peripheral blood leukocytes following standard procedures.12 Each exon, including the intron boundary regions of the LAL gene, was individually amplified by polymerase chain reaction (PCR) followed by direct sequencing on a solid support.13 The genomic nucleotide sequences used in the construction of primers were derived.6 14 All 3′ primers were labeled with biotin and fluorescein; nested primers were used for sequencing. PCR was performed with 10 pmol nonbiotinylated and 5 pmol biotinylated primer in a total volume of 50 μL containing 0.1 mmol/L of each dNTP, 0.1 μg genomic DNA, and 0.5 U of SuperTaq DNA-polymerase (HT Biotechnology Ltd) in a buffered solution (50 mmol/L KCl, 10 mmol/L Tris-HCl, pH 9.5, 1.5 mmol/L MgCl2 , 0.1% [vol/vol] Triton X-100, and 0.01% [wt/vol] gelatin). The reaction was carried out by using a Perkin Elmer GeneAmp PCR System 9600 (Perkin Elmer) with an initial hot-start technique15 followed by a touchdown-PCR cycling protocol16 with an annealing temperature dropping from 69°C to 62°C within 40 cycles.
To prepare single-stranded DNA, 35 μL magnetic beads (Dynabeads M-280; Dynal AS) were washed and resuspended in 50 μL binding buffer (10 mmol/L Tris-HCl, pH 7.5, 1 mmol/L EDTA, and 2.0 mol/L NaCl) before immobilizing the DNA by adding the complete PCR reaction to the beads. Strand separation was performed as recommended by the supplier of the magnetic beads. The immobilized template was sequenced with one of the fluorescein-labeled primers following the instructions of the AutoRead T7 sequencing kit (Pharmacia Biotechnology). DNA electrophoresis and sequence analysis were performed on an automated laser fluorescence DNA sequencer.
Allele-Specific PCR With the LAL-Encoding cDNA
A fibroblast culture from the index patient was used to isolate total RNA according to the method described by Chirgwin et al17 followed by isolation of poly(A)-RNA via affinity chromatography over oligo-(dT) cellulose according to Aviv and Leder.18 The LAL cDNA was obtained after reverse transcription with M-MuLV reverse transcriptase (Biolabs) according to standard procedures. The oligonucleotide primer (5 μmol/L) used for cDNA synthesis was complementary to positions 1304 through 1333 of the cDNA sequence (3′ nontranslated region) published by Anderson and Sando.4 A DNA fragment consisting of 340 nucleotides containing the coding information for the carboxyl terminus of LAL was amplified from the cDNA derived from the allele not harboring the E8SJM allele by two-step PCR. The first step was performed with a pair of oligonucleotide primers (0.5 μmol/L each) corresponding to nucleotide positions 668 through 700 (5′ primer) and 1217 through 1240 (3′ primer) of the cDNA using the same conditions as described for direct sequencing above in a total volume of 100 μL. Five microliters was removed and used for a second amplification with the same 3′ primer and a 5′ primer corresponding to positions 901 through 924. Since the region between nucleotides 863 and 934 is deleted from cDNA as a consequence of the presence of the mutation in exon 8 on one allele of the patient, this amplification scheme results in specific amplification of the cDNA fragment derived from the other allele. Sequencing was performed with [α-35S]dATP by cycle sequencing (Cyclist Taq DNA sequencing kit; Stratagene) according to the instruction manual of the supplier.
Secondary Structure Analysis
The effect of the Leu336→Pro (L336P) replacement on secondary structure elements of LAL was predicted by using the program alb as described by Ptitsyn and Finkelstein.19 Analysis was performed with the predict program package present in the HUSAR biocomputer. This program was used since it takes into account the local interactions along each chain region and long-range interaction between different regions.
Clinical Phenotype and Laboratory Findings of Patient A.S.P.
CESD was suspected in a 19-year-old woman of Norwegian-Canadian ancestry who had hepatomegaly without splenomegaly and altered liver function without jaundice. Although most of her plasma lipids were within the normal range upon initial analysis at our clinic in Münster in 1992 (Table 1⇑), her HDL-C of only 0.41 mmol/L was clearly below the fifth percentile. Moreover, apoB levels were found elevated to 1.42 g/L. Between 1984 and the time of this article, her cholesterol ranged between 5 and 9 mmol/L. The clinical anomalies had become evident at least 5 years previously when she presented with elevated serum cholesterol and hepatomegaly.
To confirm the diagnosis of CESD we measured the activity of LAL in extracts of lymphocytes isolated from blood samples. Activity measured with the substrate p-nitrophenyl myristate was only 4.66 U/mg protein compared with 32.81±4.72 U/mg protein obtained in three normal control subjects (Table 1⇑). Thus, the residual activity was 14.2%. In contrast, the activity of acid β-galactosidase was only slightly reduced, to 81%, of the mean of six normal control subjects (A.S.P., 3.34 and control subjects, 4.12±1.30 U/mg protein), thus excluding the possibility of LAL inactivation during transport to our laboratory. The activity ratio of LAL to β-galactosidase was 5.7 times higher in normal control subjects than in the patient, suggesting a selective deficiency of LAL. Likewise, the residual activities measured with the more natural substrates cholesteryl[14C]oleate and [14C]triolein in extracts prepared from fibroblasts were only 8.2% and 8.9% of the control value, whereas β-galactosidase activity was normal (9.6 U/mg protein).
Direct Sequencing of the LAL-Encoding Gene Reveals Three Sequence Deviations
LAL deficiency may be caused by mutations within the gene encoding the enzyme. Therefore, we sequenced the entire coding portion of the LAL gene, including the flanking intron sequences. This analysis revealed three apparently heteroallelic deviations from the nucleotide sequences of the LAL gene.6 14 They consist of a G→A substitution in the last nucleotide position of exon 8, a T→C mutation in exon 10 predicting an L336P missense mutation, and an A→C transversion in exon 2 predicting a Thr→Pro missense mutation in the prepeptide (amino acid −6) (Fig 1⇓). The G→A substitution at the last nucleotide position of exon 8 was previously found in two other patients with CESD. Klima et al7 found it in conjunction with a null allele, and Muntoni et al8 showed homozygosity in another patient. To distinguish whether the L336P or the T-6P or the combination of the two mutations is the underlying second reason for CESD in the present patient, we performed a family study.
Inheritance of the Sequence Deviations in Exons 2, 8, and 10 and Frequency of the T-6P Replacement
Direct sequencing of exons 2, 8, and 10 of the LAL gene, including the flanking intron regions, revealed that the mother, sister, and grandfather of the patient are heterozygous for the allele encoding the L336P replacement (Fig 2⇓). In contrast, the allele encoding the T-6P replacement was not found in her sister but in all other family members of the maternal branch (Fig 2⇓). This leads to two conclusions: the mutation encoding the L336P replacement present in the patient must have been transmitted via her mother, and the T-6P missense mutation is most likely not located on the same allele as the mutation encoding the L336P replacement. The main reason for the latter conclusion is that H.S.P. has inherited the allele carrying the T-6P replacement but not the allele encoding the L336P variant from his father (J.P.; Fig 2⇓). Moreover, T-6P replacement is not found in N.S.P., although she is heterozygous for the L336P replacement. If the unlikely possibility of recombination (likelihood P<.0005 assuming a frequency of 0.01 per 106 nucleotides) is excluded, the E8SJM is therefore most likely located in cis with the allele encoding the T-6P replacement and in trans with the allele encoding the L336P replacement in the patient. Although the precise genotype of her father is unknown, since he was unfortunately not available for this study, it is very likely that he has transmitted the two sequence deviations in exon 8 and exon 2 to A.S.P.
The LAL activities of the individuals who were available for this analysis are shown in Table 2⇓. It is obvious that low LAL activity (<20 U/mg) segregates with the presence of the allele encoding the L336P replacement and the mutation in exon 8 but not with the allele encoding the T-6P replacement. This is best illustrated for H.S.P. and N.S.P. H.S.P., who is homozygous for the T-6P replacement but does not have the L336P replacement, has LAL activity within the normal range. Conversely, N.S.P., who is heterozygous for the L336P replacement but does not have the T-6P replacement, has nearly half-normal LAL activity. Since the T-6P replacement is present in 7 of 14 alleles in the family, we also investigated the frequency of the corresponding allele in a randomly selected population from Westfalia. Whereas the allele encoding L336P was not present in 28 sequenced alleles, we found the allele predicting T-6P replacement in 21% with no difference in mean LAL activity in either group as measured in peripheral white blood cells (data not shown).
Allele Encoding the L336P Replacement Is Not a Null Allele
Since the E8SJM allele has been found in conjunction with a null allele in a patient with CESD,7 we further wished to exclude the possibility that L336P is also associated with a null allele in the patient of this study. For this purpose, we performed an allele-specific PCR after reverse transcription of RNA from fibroblasts. Since we used a 5′ primer that was complementary to sequences within the 72-bp deletion due to the heterozygous presence of the E8SJM allele,7 only the cDNA corresponding to the transcript encoded by the other allele should be amplified. As shown in Fig 3⇓, direct sequencing of the resulting DNA fragment showed C instead of T at the corresponding nucleotide position predicting the L336P replacement. This result confirmed the trans localization of E8SJM and L336P deduced from the family study. In addition, the possibility that L336P is present in a null allele was excluded.
Phenotypic Consequences of Heterozygosity for the L336P Replacement
The plasma lipid values of the family members are shown in Table 3⇓. It is obvious that the patient A.S.P., except for her decreased HDL-C, has normal lipid parameters, an unusual finding in CESD. Moreover, the individuals who are heterozygous for the L336P replacement (J.P., G.S.P., and N.S.P.) have relatively high fasting total cholesterol and triglyceride values, ranging from 5.7 to 6.9 mmol/L (mean, 6.37±0.60 mmol/L) and from 1.13 to 2.12 mmol/L (mean, 1.78±0.55 mmol/L), respectively. The mean values for cholesterol (5.54±1.63 mmol/L) and triglycerides (1.17±0.26 mg/dL) are lower in the group not affected by the mutation. However, since the pedigree is rather small the differences are not significant.
We have identified a new variant of LAL containing an L336P replacement associated with LAL deficiency and CESD. The second allele in this patient harbors a G→A splice-junction mutation that was shown to result in expression of a truncated variant of LAL (Δ254 through 277) via exon 8 skipping.7 Since the same mutation was shown to be the underlying defect in two other independent cases of CESD,7 8 we focused this investigation on the L336P variant. Several lines of evidence indicate the functional role of the mutation. First, by family study and allele-specific PCR, we showed trans location relative to the mutation in exon 8. Second, by PCR amplification after reverse transcription of RNA from the patient’s fibroblasts, we showed that the L336P allele does not represent a null allele.7 Third, the L336P-encoding allele cosegregates with low LAL activity in the family. Finally, we showed that a third mutation in the patient’s LAL gene predicting a T-6P replacement does not seem to play a functional role for CESD but represents a relatively common polymorphism. The main evidence for this conclusion is that T-6P does not cosegregate with low LAL activity in the family. As shown by the family study, the sequence deviation predicting T-6P replacement is most likely localized on the E8SJM allele in the proband. Conversely, a previously studied patient who was homozygous for E8SJM did not reveal the T-6P replacement. This can be explained by assuming that recombination has occurred between exons 2 and 8, suggesting that either both mutations are relatively old or that a hot spot for recombination is located within the LAL gene locus.
The nonconservative L336P replacement predicted from the second allele would be expected to have a disrupting effect on an α-helical segment that according to secondary structure prediction is most likely located between residues 332 and 343 of LAL (Fig 4⇓). Further indication for the importance of the leucine at position 336 is suggested by the fact that this residue is found at the identical corresponding positions in all the known enteric lipases.20 21 On the other hand, this site is relatively far away from what is presumably the active-site serine located at position 153 of LAL. This assignment is based on the position of the residue in a conserved esterase pentapeptide motif and the observation that the equivalent serine at position 153 has been implicated in the catalytic mechanism of human gastric lipase.22 Although the precise three-dimensional structure of LAL is unknown, we currently believe that the distal location of the L336P replacement with respect to the active site is the reason for the relatively high residual LAL activity measured in extracts from the patient’s leukocytes and fibroblasts.
The proband’s normal cholesterol and triglyceride levels are not normally seen in CESD. Therefore, diagnosis initially relied on detecting hepatomegaly, hyperbetalipoproteinemia, and partial HDL deficiency. The only other clear support for the diagnosis was very low LAL activities in lymphocytes and fibroblasts. Partial HDL deficiency has now been observed in all three carriers of the E8SJM allele who have developed CESD.7 8 One possible explanation for this finding is that exchange of cholesterol esters for triglycerides coming from VLDL and/or chylomicron remnants mediated by cholesterol ester transfer protein would lead to depletion of cholesterol esters in HDL. However, this appears to be rather unlikely since the patient of this study has very low HDL-C levels even in the absence of elevated triglycerides. Thus, partial HDL deficiency may be directly related to the abnormal hepatic lipid metabolism due to the inefficient lysosomal catabolism of triglycerides and cholesterol esters. Increased activity of hepatic lipase has been reported in one case of CESD,23 but the precise mechanisms of how LAL deficiency can lead to partial HDL deficiency are not yet understood.
Compared with a recently studied patient who was homozygous for E8SJM, the proband of this study has higher residual LAL activity, less pronounced abnormalities of her plasma lipid values, and less severe hepatomegaly. Therefore, it appears that the L336P variant is associated with a relatively mild phenotype of CESD. Anderson et al14 have described two LAL mutations causing Wolman disease, the more severe type of LAL deficiency that leads to neonatal death. One predicts a Leu179→Pro replacement, located close to the active site at Ser153; the other is a frame-shift mutation at nucleotide position 634 leading to a premature stop codon. Goldstein et al3 have introduced the hypothesis that higher residual activity of LAL in intact cells is the cause for the less severe phenotype associated with CESD compared with Wolman disease. In principle, there are two possibilities of how mutations in the LAL gene could lead to different levels of residual activity in whole cells. Either mutations in LAL resulting in complete acid lipase deficiency cause Wolman disease, whereas those associated with some residual activity predict CESD, or the activity level of another enzyme with low hydrolyzing side activity for triglycerides and cholesterol esters, such as phospholipase A2, is responsible for the level of residual activity and thus determines whether a patient has Wolman disease or CESD. To discriminate between the two possibilities, studies are under way in which all four known naturally occurring LAL variants in bacteria are being expressed in order to study their enzymatic properties after protein purification.
This study was supported in part by grants from the Deutsche Forschungsgemeinschaft (Se 459/2-1), Bristol Myers Squibb, and the Land Nordrhein-Westfalen and was performed in accordance with the ethical standards laid down in the 1964 Declaration of Helsinki. We thank Ulla Olthoff for excellent technical assistance.
Assmann G, Seedorf U. Acid lipase deficiency: Wolman disease and cholesterol ester storage disease. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The Metabolic and Molecular Basis of Inherited Disease. 7th ed. New York, NY: McGraw-Hill. 1995:2563-2587.
Fredrickson DS. Newly recognized disorders of cholesterol metabolism. Ann Intern Med. 1963;58:718. Abstract.
Goldstein JL, Dana SE, Faust JR, Beaudet AL, Brown MS. Role of lysosomal acid lipase in the metabolism of plasma low density lipoprotein: observations in cultured fibroblasts from a patient with cholesteryl ester storage disease. J Biol Chem. 1975;250:8487-8495.
Anderson RA, Sando GN. Cloning and expression of cDNA encoding human lysosomal acid lipase/cholesterol ester hydrolase. J Biol Chem. 1991;266:22479-22484.
Mahadevan S, Tappel AL. Lysosomal lipases of rat liver and kidney. J Biol Chem. 1968;243:2849-2854.
Hultman T, Stahl S, Hornes E, Uhlen M. Direct solid phase sequencing of genomic and plasmid DNA using magnetic beads as a solid support. Nucl Acids Res. 1989;17:4937-4946.
Anderson RA, Byrum RS, Coates PM, Sando GN. Mutations at the lysosomal acid cholesteryl ester hydrolase gene locus in Wolman disease. Proc Natl Acad Sci U S A. 1994;91:2718-2722.
Chou Q, Russell M, Birch DE, Raymond J, Bloch W. Prevention of pre-PCR mis-priming and primer dimerization improves low-copy-number amplifications. Nucl Acids Res. 1992;20:1717-723.
Don RH, Cox PT, Wainwright BJ, Baker K, Mattick JS. Touchdown PCR to circumvent spurious priming during gene amplification. Nucl Acids Res. 1991;19:4008.
Aviv H, Leder P. Purification of biologically active globin mRNA by chromatography on oligothymidylic acid cellulose. Proc Natl Acad Sci U S A. 1972;69:1409-1412.
Docherty AJP, Bodmer MW, Angal S, Verger R, Riviere C, Lowe PA, Lyons A, Emtage JS, Harris TJR. Molecular cloning and nucleotide sequence of rat lingual lipase cDNA. Nucl Acids Res. 1985;13:1891-1903.