Communication Vol. 263, No. 3, Issue of January 25, pp. 1107-1110,1988 THE JOURNAL OF BIOLOGICAL CHEMISTRY

Q 1988 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.

Human Hepatic Lipase CLONED cDNA SEQUENCE, RESTRICTION FRAGMENT LENGTH POLYMORPHISMS, CHROMOSOMAL LOCALIZATION, AND EVOLUTIONARY RELATIONSHIPS WITH LIPOPROTEIN LIPASE AND PANCREATIC LIPASE*

(hceived for publication, October 2, 1987) Santanu DattaS, Chi-Cheng LuoQ, Wen-Hsiung LiQ, Peter VanTuinenll, David H. Ledbetterl, Mary A. Brown$, San-Hwan Chen$, Shyan-woei LiuS, and Lawrence Chan$ From the $Departments of Cell Biology and Medicine, and lllnstitute for Molecular Genetics, Baylor College of Medicine, Houston, Texas 77030 and §Center for Demographic and Population Genetics, University of Texas, Houston, Texas 77030

Human hepatic lipase is an important enzyme in high density lipoprotein (HDL) metabolism, being impli- cated in the conversion of HDL, to HDL3. Three human hepatic lipase cDNA clones were identified in two Xgtll libraries from human liver. The cDNA-derived amino acid sequence predicts a protein of 476 amino acid residues, preceded by a 23-residue signal peptide. Four potential N-glycosylation sites are identified, two of which are conserved in rat hepatic lipase. On align- ment with human, mouse, and bovine lipoprotein lip- ase, the same two sites were also conserved in lipopro- tein lipase in all three species. Stringent conservation of the cysteine residues was also evident. Comparative analysis of amino acid sequences shows that hepatic lipase evolves at a rapid rate, 2.07 X lo-* substitutions/ sitelyear, about four times that in lipoprotein lipase and half that in pancreatic lipase. Further, hepatic lipase and pancreatic lipase appear to be evolutionarily closer to each other than either of them is to lipoprotein lipase. Southern blot analysis revealed high frequency restriction fragment length polymorphisms of the he- patic lipase gene for the enzymes Hind111 and MspI. these polymorphisms will be useful for haplotype and linkage analysis of the hepatic lipase gene. Using cloned human hepatic lipase cDNA as a hybridization probe, we performed Southern blot analysis of a panel of 13 human-rodent somatic cell hybrids. Concordance analysis of the various hybrid clones indicates that the hepatic lipase gene is located on the long arm of human chromosome 16. Analysis of hybrids containing differ- ent translocations of chromosome 15 localized the gene to the region 15q15-22.

HL-16512 (to L. C.) , GM-30998 (to W-H. L.), HD-20619 (to D. H. * This work was supported by National Institutes of Health Grants

L.), and HD-06814 (to P. V. T.), and by a grant from the March of Dimes Birth Defects Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accord- ance with 18 U.S.C. Section 1734 solely to indicate this fact.

The nucleatide sequence(.) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) 503540.

Hepatic lipase catalyzes the hydrolysis of tri-, di-, and monoacylglycerol, acyl-CoA thioesters, and phospholipids (1- 3). Hepatectomy experiments implicate the liver as the main source of hepatic lipase in postheparin plasma (4). Direct evidence for the production of hepatic lipase has been dem- onstrated in liver parenchymal cells (5-7). Hepatic lipase has been postulated to catalyze the hydrolysis of IDL’ triacylglyc- erols to produce low density lipoprotein, and that of HDL, triacylglycerols and phospholipids to produce HDL, (8, 9). There is an inverse relationship between hepatic lipase activ- ity and plasma HDL levels (lo), supporting the hypothesis that hepatic lipase may be involved in the hepatic uptake of HDL cholesterol (11-13).

Because of its extremely low concentration, the amount of hepatic lipase that can be purified from liver or postheparin plasma has been insufficient for detailed structural analysis. Recently, the structure of rat hepatic lipase was deduced from its cloned cDNA sequence (14). Here, we report the cloning and complete sequence analysis of human hepatic lipase cDNA. Using the cloned hepatic lipase cDNAs as hybridiza- tion probes, we have mapped the human hepatic lipase gene to the long arm of chromosome 15. We have further detected high frequency RFLPs of the hepatic lipase gene for the enzymes MspI and HindIII. Sequence comparisons among hepatic lipase, lipoprotein lipase, and pancreatic lipase from various species allow us to infer the evolutionary relationships among these three lipolytic enzymes.

MATERIALS AND METHODS

Construction and Sequence Analysis of cDNA Clones-A rat hepatic lipase cDNA clone was identified in a rat liver cDNA library in Xgtll by oligonucleotide hybridization using oligonucleotides of published rat hepatic lipase cDNA sequence (14). The rat cDNA clone was then used to screen two human liver cDNA libraries in Xgtll (15, 16) by cross-hybridization. Three overlapping human hepatic lipase cDNA clones were identified and confirmed by complete sequencing using the chain termination method of Sanger et al. (17) following subclon- ing in the M13 vectors mp18 and mp19. Of the three cDNA clones, XHLl extends from the codon for Gln-39 to the poly(A); XHL2 extends from the codon for isoleucine at position -8 to the poly(A); XHL3 extends from the 5’-end of the sequence to the codon for Lys- 329 (Fig. 1).

Southern Blot Analysis-The insert DNAs from XHL2 were labeled with 32P by nick translation and used for hybridization in Southern blots (18) of human genomic DNA or somatic cell hybrid DNA.

Somatic Cell Hybrids-The construction and characterization of most of the Chinese hamster-human somatic cell hybrids and the mouse-human hybrids have been reported previously (19, 20). In addition to these, hybrids 15.2 and LS-6 were included to confirm the chromosomal and regional localization of hepatic lipase. Clone 15.2 is the product of fusion between GM3200 (46,XY; fragile X male) and Chinese hamster, and contains only human chromosomes 8, 15, and the X. LS-6 is a human-mouse hybrid constructed from a patient with Prader-Willi syndrome and a balanced translocation 45, XX,-15,-17,+der(l7)t(l$l7)(ql3;p13) (case reported by Elder et al. (21)). It retained many chromosomes including the der(l5), but not the normal 15. Chromosome analysis was performed by trypsin G banding of all hybrids and G-11 staining of selected clones. A human chromosome was scored as present if 20% or more of the cells contained the chromosome and absent if less than 20% of the cells contained the chromosome.

Estimation of the Number of Amino Acid Substitutions-The num-

I The abbreviations used are: IDL, intermediate density lipopro- teins; HDL, high density lipoproteins; RFLP, restriction fragment length polymorphism; kb, kilobase pair(s).

1107

1108 Human Hepatic Lipase cDNA Sequence

ber of amino acid substitutions between homologous protein se- quences is estimated by Kimura's method (22), which makes a cor- rection for multiple substitutions at the same residue site.

RESULTS AND DISCUSSION

Human Hepatic Lipase cDNA and Deduced Amino Acid Sequence-The complete sequence of human hepatic lipase cDNA is deduced from three overlapping clones (Fig. 1). Both strands were sequenced in entirety, and all restriction sites were crossed. The DNA sequence comprises 1,550 nucleotides plus a poly(A) tail. I t includes 4 nucleotides in the 5"untrans- lated region, 1,497 nucleotides in the coding region, and 49 nucleotides in the 3"untranslated region. The polyadenyla- tion signal AATAAA precedes the poly(A) tail by 15 nucleo- tides. The amino acid sequence of human hepatic lipase deduced from the nucleotide sequence is shown in Fig. 1. The mature protein is a 476-amino acid polypeptide preceded by a 23-residue signal peptide. The calculated molecular weight of 53,249 is consistent with previous estimates of 64,000- 69,000 of the enzyme purified from human postheparin plasma (6, 23,24), human hepatic lipase being a glycoprotein. Intracellular hepatic lipase precursors in the rat have appar- ent molecular weights of 47,000 and 53,000 compared to a 59,000 mature protein (5). The latter form differs from the two precursors in that it contains carbohydrate. Two hydro- phobic domains are identified in the hepatic lipase sequence (underlined by interrupted lines in Fig. 1) that show high similarity to a putative lipid-binding region of lipoprotein lipase. As pointed out by previous workers (14), they are also present in rat hepatic lipase, as well as mammalian lipoprotein lipase, lingual lipase, pancreatic lipase, and lecithin-choles- terol acyltransferase.

The deduced amino acid sequence has four potential N- glycosylation sites (Fig. 1). Two of these sites (residues 55-57 and 374-376) are conserved in the rat sequence (Fig. 2). The other two sites (residues 19-21 and 339-341) are present only in human hepatic lipase, but are not conserved in rat hepatic lipase (14). It is interesting that when hepatic lipase sequences are aligned with lipoprotein lipase sequences (Fig. 2), the two common N-glycosylation sites in human and rat hepatic lipase are completely conserved in lipoprotein lipase in man, cow, and mouse. Further, they constitute the only potential N-

-21 ._. me u c Y L -7 mc or 7 m n C x< 17, nr n c m, n r TGC A X n r A X uu T C A -1 D C D l -c "P 111 s.7 Pro 14" 0. 1". *. ,I . Y" U" V., Y" <". ,I. .n. I , . 0," *, 5, A,. Y"

a uu ux C1T YL cc. u a i c 1 m cw yi* +zk CFI u1 -1 o7 * Y A uc - U T "0 u, u s Arc .AG Y C -1 n c CIC C1" Cln a.r Y" w. ?ID Ti" 110 m. 5 1 Y u o AZO &I. rln u. ".I rl" m r Y Y " I . GI" ,I., t". ,)Ir I.s .*. uI ,D D C n r cw CU Y L u1 u e ux nt u c .TI C u A X u, u, mc uc Y C n1 u c u c IGC n c u< x c x c O r m, oc, L." .D. C1" GI" mr U" CL" GI" CY. rln ,I. Y S , I . U" "I. .ID U O m. Y" St" $1" CI. 01" Ih. Y" .,- Y" '0

mc A m A n Am us ax rw Xcc mc cu 02 OT 0. c11 PA< ty n x XG w 1.c mc CZC -0 n C Y G x1 ur CEr wc O r ".l 8.. I , . XI. "I. =I" 7ro a.r 11.1 U P CIY "*1 Y U C I V " 0 .ro *I. rrp C l n -I ".I L1. AI. u u L". s.r C l " ,re .I. C l n w

bl G I G Y C me Lsc =a -0 uc n C are Y< n o Y C u c u5 u c ?A< Y C M F Y C C I C c_ PAC a -7 o c mc U r Y Y .=* ".* U" v.1 Ti" Y" ".1 "0 I Z S x,. 7"Z Y" I,* "5. us X,. *. I". 11.11. n, Y. "I 1*1 us I." Y., c,, 8. CI" 110

mc mc -1 n 7 D C cm XG nc U C * x= o, nc x7 cc. u c u/ m7 u< R. n, ax m c ,.x o c my D. ut mc

x. cw m Dc Loc -7 TCC A% mr G5. Y C u c Y G a71 ccc -1 "IC Y A mc CIG 0, CFC mr G5. cc, n C n7 ur cw -I I.1 .I. AI. Y Y Y" -0 .rv Y Y n u 51" *r ".I m" Y Y * , u s *. "i. ".I "L. Y" %.-Gi" n. *I I." C i y l , . "I. I.1 111

Y C == uz Y1 c- m X I CCL u 7 u r UX -I nr mr U, ccc u . 0 1 Y C n, *cc CCG U G u c *,r u c o C -c mc m c

*r C1" m. "1. a 1 *. 5 1 I,. GI" 01" 7 h l "I. L". I, . 61" YQ I,. I h l C," Y" UD A i . AI. r,, I j 0 Y" ,I. C," r , , *. 1.0

*I. ,=a s.I L.. u s Y Y S I -re U P k.0 A,. s . r ,I" Y.l ..D A>. I , . *I. mr Ph. i h i "0 GI" *I. l i D,(l Y" I, "., Dlll > io *IC - CIT ccc ma - CLC TAT u c ne TM ccc YC ccc a xcc nc u c rcr mc t ~ c c x n c 01 uL n c TLC +.=& nl AI, 11. L". CL" .ro 11. GLY ML. nr U P m. wr Pro l.D G," C l r *r tn. r / n PrD C I " C". *I. Ih. U" C I " MY mi AZO *I. I > . 240

&I. =In "I. CIY .I. "I .I. 11. m r r l l r.r r l . L". 0. a.r *I. m u "0 *I "., *,. Y" Ph. I , . u s *"Y2Uy. *x. .,. I70 D E CAT u c mc m U I CFC A X Y C us Y C 111 - r.x x c u c u c - x c mr u c CII n c >IC U C _ E n C Or C1E Y C

ccc m u c YFF .15 =c rAc m G m _I u< m.5 Y C m m .zx O r mc C I 5 n c n r c€ T u - nc * E= m u c Y r 51" m. 'Ln I. I C Ai. nr Fro 0. el l "I -I "a I.I 111. *I CL" GI" YV 0. Y" l.c 0. Y. L,. c,, u* 0. ". ,lr ,no o c aE T Y ut G I C C Y u= - -0 m.2 .zx nc ..% Y Y mc n r n C n C mL Y C cu ULE u r x c mr nr - - TAT CAI r x - =A u c &IC u r n c AX Y< u1 m ACC c c ~ ma C*L x n XI m m I ~ T 1u =A n c cw YA - uc - _G

Y U n r nr 81. v.3 u* n n ClU P r m c,, *r w. I C L". u n L." Ph. Y V v., r)lr uo A,. Cl" *I rrD )in. L". ".2 *, "1. 111)

nr C1. Y Y w. I,. CI. m. I,. r i u mr Pra 1,. 01" 111. I h r .I. ,hr n.I s.r Y" Y" C, , ,hi L". rlU w. *.I 1'0 CIT - A77 m .x - == = - cw u1 mr +.=I AAI Y T 1m X E m m .IC Ycc D C U1 m U, a x w o c

A X AI0 LIC BAG m u c mc yc I. cc. o r rw ex YI o c rw U F Y C _F u c llcc n c arc _A rw ..% Y L a u c

ern w. 11. Pro I,. 71. Y Y 0," w. 6," I,. I,. I r * c *r 1h. U" I,. n r YV UD Y., US I , , G I Y G," Y" 1,o

*I. 8.f 11. *. .h. LY* rze sa" "D a.r *I. ".* zrs LZ. -11 "*I 7.0 "I 7hr "., Cis /"r 1,. 1,. ,_ tro s.r I", m y .10 ' 2 0

C Y u C X 1 ~ C I F - C X ~ Y C ~ I C " L U m 1 ~ * Y C U T ~ 1 W , . ~ ~ l m m X ~ c r r w r r u w r u r u* "L. ur 011 U" ".I Y" L". I.= I,. u o ".I *. a,. r l l C l " mr n. Cl" wg -r m r ,". 01. *, G I " U" m, "S .IO

RA ~n m ctc ma Y C UT cu ATC n c mr - m GU ATA TCT YL YT w AX - m mllt..Lw. I." Y" Y Y Ax0 .rD Thr GI" GI" LI. TI. m. ".I Y. 0. Cl" 11. Y. 2.r w. IhC s.r w. us w. I,. us . r r r u ~ ~ . . . ~ ~ s u r L t r . c r L n r r r u u \ .....

FIG. 1. Nucleotide sequence and deduced amino acid se- quence of cloned human hepatic lipase cDNA. The putative N - glycosylation sites and the polyadenylation signal are underlined. The two hydrophobic domains showing similarity to the lipid-binding regions of other lipases are underlined by interrupted lines.

hHL rHL

rnLPL hLPL

bLPL dPL PPL

hHL rHL hLPL SLPL bLPL dPL PP L

hHL rHL hLPL rmPL bLPL dPL POL

hHL THL

rnLPL hLPL

bLPL dP L PP L

hHL IHL hLPL rnLPL bLPL dPL PP L

hHL rHL

InLPL hLPL

bLPL apL PPL

hHL rHL hLPL "LPL bLPL dPL PPL

.......... LGHVHLLGKSLGAHMGIAGSLT--NKKVNRITGWPAGPNFEKAEAPS-RLSPDDIONWLHTFTRCS-PGRS 193 PSQVQLIGHSLGAHVAGEAGSRTPC----LGRITGWPV~SFQ-GIPEEVRLDPTDMW~IHTDMPLIPFLC 215 PSWHVIGHSLGSHMCEAGRRTNG--IIER~TGWPA~PCF~GTPELVRLDPSDAK~DVIHTDMPIIPNLG 214 . . . . . . . " . . . . . 0 VGIKQPIGHYDFKPNGGSFQPGCHFLELYRHIAOHGFNAII-~TIKCSHERSVHLFlDSLLHAGTOYUYPCGDU 202 VCIKOPfAHKDFKPNGGSFOPGCH~LELKKHlAEHGLNAlT-QT~KCAHERSVHL~lDSLQHSNWNTGFHCS~ 284 ICIOKPVGHVDIYPNGGTFOPGRIICUIRVIAERGLGDVD-OLVKCSHERS~HLF~DSL~E~NPSMKUC~~K 261 IGIQKPVCHVDIKPNGGTFQPGCNICUIRVIAERG~DVD-OLVKCSHERSIHLFIDSLUl~ENPSMYR~CY~~ 767

I U F B S N K C F P C P D Q G C P Q ~ H K M K F A V K I S D E T O K K F ~ T G D S S N F A R W R K ~ S I T L ~ K ~ I C O A ~ A ~ - - - 360 HVFTWKCFPCPSEGCPPHYMRFPGKTNGVSWFK~TGDASNFARMRK~SVILSGKK~GHILVSL---- 359

~ T K E M Q K I P I T L G M G I A - S ~ Y S F L I T W W I G E L I M I K F K Y E N S A V V A ~ ~ Q T I I P Y S T G P R H S G L V L 4 2 1

YG~AESENIPFTLP~-VS-TNKTYSFLIYTEW~GEL~~LKU-KS----------DSKFSUSD~SSPCFA~ 401 KG~AESENIPFILPE-VS-IN~IKSFLIKTE~lGEL~KLK~-~S----------DSYFS~PD~SSPSNl 401 YG~AESENIPFILPE-VS-INKIKSFLLKTE~ICEL~~LKW-IS----------DSKFSWSNlSSPCPDI 401 FGSKGNTH--QFNIFKGILKPGSIHSNEFDAKWVGT~EKVKFLY-NNHV-------VNPTF--------PK~A 417 FGNEGNSR--OY~IKKCT14PDNTHSD~FDSDVEVGDLOKVKFI~NNHV-------INPTL--------PRVGA 411

RIIRVMGCT~RMTFCS~NT--DDLLLRPIQEKIFVKCCIKSKTSKRKIR 416

OKIRV~GETQKI(I(VIFCSREK--VSHWKGIUPAVFVKC~DKSL--NKKSG 4 4 8 KTIWVMGET(RRMTFCPDNV--DDLQLHPIQEKVFVKCDLKSKD------ 472

ERIRVMGETQK~IFCAREK- -VSHMKGISAVFVKCHDKSL- - -KKSG 441 RKIRVMCETQK~IFCSREK--USK14UGKSPVlFVKCHDKSL--NRKSG 448 TKITVQKGEEKTVHSFCSESNLECVLLTLTP------C------------ 450 SKITVEIUIDGK-~DFCSQE~REEVLLTLNP------C------------ <a9

.~~

. . . .. - L G T K E E I K K I P I I L C E G I I - S N K T Y S L L I T W I ~ G E L ~ U L K F K W E N S A ~ A ~ ~ Q T ~ M ~ ~ P H K A ~ L I L 429

0 . . .. FIG. 2. Alignment of human ( I t ) and rat (r) hepatic lipase

(HL) , human, mouse (m), and bovine (b) lipoprotein lipase (LPL), and dog ( d ) and pig ( p ) pancreatic lipase (PL). The sources of the sequence are: rat hepatic lipase, Komaromy and Schotz (14); human lipoprotein lipase, Wion et al. (25); mouse lipoprotein lipase, Kirchgessner et al. (29); bovine lipoprotein lipase, Senda et al. (32); dog pancreatic lipase, Kerfelec et al. (33); and pig pancreatic lipase, DeCaro et al. (34). Residues that are identical in all sequences are marked by asterisks. The putative N-glycosylation sites in human hepatic lipase are marked by a horizontal bar. The two sites that are conserved among all hepatic lipase and lipoprotein lipase sequences are marked by double bars. The cysteine residues that are conserved in all sequences are marked by circles around the asterisks.

glycosylation sites in lipoprotein lipase; one additional se- quence Asn-Pro-Ser pointed out by Wion et al. (25) is also present in the lipoprotein lipase of all three species but cannot be a glycosylation site because proline is not allowed in N- glycosylation sequences (26). If the two highly conserved sites are indeed N-glycosylated in vivo, their stringent conservation suggests that they are important to the function of lipoprotein lipase and hepatic lipase. Their absence in another lipolytic enzyme pancreatic lipase indicates that they are not essential to pancreatic lipase action.

Another remarkable feature of the hepatic lipase, lipopro- tein lipase, and pancreatic lipase sequences is the conservation of the cysteine residues. There are 10 cysteine residues each in hepatic lipase and lipoprotein lipase in all species examined (Fig. 2). These residues are thus completely conserved. Fur- ther, although there are a few insertions or deletions between hepatic lipase and lipoprotein lipase, the distances between these cysteine residues are also conserved. There is a cluster of 6 cysteine residues in the middle of the molecule in both hepatic lipase and lipoprotein lipase; these 6 residues are separated from one another by identical numbers (22, 24, 10, 2, and 4) of intervening amino acid residues. Of the 10 cysteine residues in hepatic lipase and lipoprotein lipase, 8 are com- pletely conserved in pancreatic lipase, which has, in total, 11 (dog) or 12 (pig) cysteine residues. The conservation of cys-

Human Hepatic Lipase cDNA Sequence 1109

teine residues among the three enzymes across species sug- gests that these residues might be important in disulfide bridge formation and the maintenance of secondary structures which are potentially required for optimal function of these enzymes.

Restriction Fragment Length Polymorphism of the Human Hepatic Lipase Gene-Of the restriction enzymes we tested on DNAs from a minimum of six unrelated Caucasians, six (TaqI, EcoRI, PstI, HincII, SacI, and BamHI) detected no variation, and two (HindIII and MspI) detected unequivocal RFLPs (Fig. 3A). The RFLPs detected with HindIII and MspI are independent of each other. Therefore, they arose from the loss or gain of individual restriction sites rather than a single insertion-deletion event.

Upon digestion with HindIII, Southern blots of genomic DNAs display three constant bands at 10.5, 7.0, and 0.6 kb. In addition, there are three variable bands a t 3.8,2.8, and 1.0 kb. The appearance of the 2.8-kb band is always coupled to that of the 1.0-kb band. The patterns are thus consistent with two alleles consisting of the 3.8-kb band on the one hand, and the 2.8 + 1-kb band on the other. The latter allele is the result of the presence of an additional HindIII site in the 3.8-kb allele. The frequencies of the two alleles among 15 Caucasians are: 0.73 for the 3.8-kb allele and 0.27 for the 2.8 + 1.0-kb allele.

For MspI digestion, we identified four constant bands with mobilities of 5.8, 3.8, 0.7, and 0.5 kb. There are two variable bands at 1.2 and 1.0 kb which represent the two different alleles. The frequencies among 16 Caucasians for these alleles are: 0.56 for the 1.0-kb allele and 0.44 for the 1.2-kb allele.

Human hepatic lipase is important in HDL (8-13) and

A 1 2 3 4 1 2 3 4 3.8r

*1.2

2.8- 4 .0

1.0*

FIG. 3. A, RFLPs of the human hepatic lipase gene showing homo- zygous (lanes 1, 2, and 3) and heterozygous (lane 4) patterns for HindIII (left) and MspI (right). B, Southern blots of somatic cell hybrid DNA using the enzyme BamHI. M, mouse DNA; H, human DNA; h, hamster DNA. The human specific bands are identified by solid triungks, the hamster bands by an arrow. The mouse bands are barely detectable using this enzyme. Lanes 1-10 are, respectively, hybrids MH-18, blank lane, SA-5, MR1.21, MR7.11, MR2.2, 16.1, 8.2, 1.4, and CJA, of these, lanes 1, 3, 6, 8, 9, and 10 are positive for the human-specific hepatic lipase bands. EcoRI was used in some of the mapping studies. It also gave a clearly different pattern between human and rodent blots (data not shown).

possibly IDL metabolism (27, 28). The identification of high frequency RFLPs for two commonly used enzymes in the human hepatic lipase gene will be useful in the linkage analy- sis of patients with disorders of HDL or IDL metabolism.

Chromosomal Mapping of the Hepatic Lipase Gene-The patterns of Southern blots using human hepatic lipase cDNA probes and the enzymes BamHI and EcoRI are quite different between human and mouse or hamster DNA (Fig. 3B). We have used blots of somatic cell hybrid DNA digested with these enzymes and correlated the presence of the human- specific DNA bands with the presence of specific human chromosomes in the various cell hybrids. Table I shows the results of such analysis on DNAs from 13 hybrids. Synteny analysis of the various hybrid clones indicates concordance between the hepatic lipase gene and human chromosome 15. This observation allows assignment of the human hepatic lipase gene to chromosome 15 and to no other chromosome. Regional mapping of the gene was performed by using three cell hybrids, LS-6, SA-5, and SP-3, which contained, among other human chromosomes, only portions of chromosome 15. All three hybrids gave a positive hybridization signal. Com- parison of the chromosomal regions present in these clones demonstrates that the human HL gene is located on 15q15- q22.

Evolution of Lipolytic Enzymes-Fig. 2 shows an alignment of human and rat hepatic lipase, human, mouse, and bovine lipoprotein lipase, and dog and pig pancreatic lipase. It is clear from the alignment that all these proteins are homolo- gous to one another (29, 30); the homology is at least 29%, which occurs between mouse lipoprotein lipase and pig pan- creatic lipase. We note that some regions have been well conserved among all the sequences shown in Fig. 2.

To study the rate of evolution in these proteins, we have estimated the number ( d ) of amino acid substitutions/site between each protein pair (Table 11), using Kimura’s method (22). The d value between human and rat hepatic lipase is 0.31, and the divergence time between the two species is about 75 million years (31), so the rate of amino acid substitution in hepatic lipase is estimated to be 0.31/(2 X 75 X lo6) = 2.07 x IO-’ substitutions/site/year. This rate is substantially higher than the average rate (1.4 x lo-’) in mammalian hemoglobin chains (31), suggesting that hepatic lipase is not a conservative protein. In a similar manner, we calculate that the rate of amino acid substitution is 0.44 X lo-’ for the divergence between human and mouse lipoprotein lipase, 0.50 x lo-’ between human and bovine lipoprotein lipase, and 0.63 x lo-’ between mouse and bovine lipoprotein lipase; we assume a divergence time of 80 million years between bovine and human or mouse (31). The average for these three rates is 0.52 x lo-’, which is only about one-third of that in mammalian hemoglobin chains, suggesting that lipoprotein lipase is a fairly conservative protein. Assuming a divergence time of 60 million years between dog and pig (31), we estimate a rate of 3.67 X lo-’ for the divergence between dog and pig pancreatic lipase, which evidently is a rapidly evolving pro- tein, like the immunoglobulins (31).

Table I1 is also useful for inferring the evolutionary rela- tionships among the three types of lipase. We note that pancreatic lipase has evolved about 7 times faster than lipo- protein lipase and about 2 times faster than hepatic lipase. Therefore, although the average d value is only 0.91 between hepatic lipase and lipoprotein lipase, but 1.50 between hepatic lipase and pancreatic lipase, the latter are evolutionarily closer to each other than either of them is to lipoprotein lipase. This conclusion is noteworthy because hepatic lipase and pancreatic lipase have apparently assumed widely differ-

1110 Human Hepatic Lipase cDNA Sequence

TABLE I Synteny analysis of human hepatic lipase (HL) gene by Southern blot analysis of somatic cell hybrids

Hybrids" Human chromosomes

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Y X HLb

I I1

+ + + + + +

I11 + + + IV V VI VI1 VI11 IX + + + + + ' + + + + + d + + X

+ XI

+ + + + XI1

+ + + + + + + + + + e + + + + + + + + XI11

+ + +

96 concord- 38 38 69 54 46 38 62 62 38 46 46 62 46 51 100 54 69 31 38 62 54 46 23 46

+ + + + + + + + + + + + + + + + + +

+ + + + + + + + + + + + + + + + +

+ + + + + + + + + + + + + + + +

+ + + + + + +

+

+

+ + ' ' # g

ance "The 13 hybrids used for analysis are: I, CJA; 11, 1.4; 111, 8.2; IV, 16.1; V, MR2.2; VI, MR7.11; VII, MR1.21;

VIII, 4.12; IX, MH-18; X, 15.2; XI, LS-6; XII, SA-5; XIII, SP-3. Hybrids IX, XI, XII, and XI11 are human-mouse hybrids. The rest are human-Chinese hamster hybrids.

Hybridization to 32P-labeled XHL2 cDNA probe.

Indicates the presence of a portion of chromosome 9 (9qll-qter) and a portion of 17 (17pll-qter). e Chromosome 15 observed by prior DNA hybridizations in our laboratory but not by direct G-banding analysis.

e Indicates the presence of part of the long arm (15ql3+qterj of chromosome 15. 'Indicates the presence of part of the long arm ( 1 5 q l M t e r ) of chromosome 15 and part of 17 (17~13-qter). #Indicates the presence of a portion (15pter-22) of chromosome 15 and the long arm (17qll.2-qter) of

chromosome 17.

TABLE I1 Proportion of amino acid differences (above diagonal) and estimated

number of amino acid substitutions/site (below diagonal) between lipolytic protein sequences

The abbreviations used are: h, human; r, rat; m, mouse; b, bovine; d, dog; p, pig; HL, hepatic lipase; LPL, lipoprotein lipase; PL, pan- creatic IiDase.

7.

8.

9.

10.

hHL

hHL rHL 0.31 hLPL 0.87 mLPL 0.89 bLPL 0.89 dPL 1.41 pPL 1.48

rHL hLPL mLPL bLPL

0.25 0.53 0.54 0.54 0.54 0.55 0.56

0.91 0.06 0.07 0.95 0.07 0.09 0.96 0.08 0.10 1.54 1.55 1.56 1.55 1.55 1.62 1.64 1.59

dPL pPL

0.67 0.68 0.69 0.69 0.69 0.70 0.69 0.71 0.69 0.70

0.33 0.44

ent functions during evolution. Like lipoprotein lipase, he- patic lipase is an intravascular enzyme anchored on the endo- thelial lining where it is essential for normal lipoprotein metabolism by the hydrolysis of various lipids on lipoprotein particles. It appears that in the course of evolution, hepatic lipase has maintained a structure and function closely related to those of lipoprotein lipase, which is a very conservative protein. In contrast, pancreatic lipase, a rapidly evolving protein, has assumed an entirely different metabolic role as a digestive enzyme whose site of action is exclusively in the lumen of the gastrointestinal tract.

REFERENCES 1. Kuusi, T., Nikkila, E. A,, Taskinen, M.-R., Somerharju, P. & Ehnholm, C.

2. Jensen, G. L., Daggy, B. & Bensadoun, A. (1982) Biochim. Biophys. Acta

3. Laboda, H. M., Glick, J. M. & Phillips, M. C. (1986) Biochim. Biophys.

4. Ehnholm, C., Schroder, T., Kuusi, T., Bang, B., Kinnunen, P., Kahma, K.

5. Laposata, E. A,, Laboda, H. M., Glick, J. M. & Strauss, J. F. (1987) J. Biol.

6. Ehnholm, C., Shaw, W., Greten, H. & Brown, W. V. (1975) J. Biol. Chem.

(1982) Clin. Chim. Acta 122,39-45

710,464-470

Acta 876,233-242

& Lempinen, M. (1980) Biochim. Biophys. Acta 617,141-149

Chem. 262,5333-5338

250,6756-6761

11.

12.

13.

14.

15.

16.

17.

18. 19.

20.

21.

22.

23.

24.

25.

26. 27.

28. 29.

30.

31.

32.

33.

34.

Leitersdorf, E., Stein, 0. & Stein, Y. (1984) Biochim. Biophys. Acta 7 9 4 , 961 -%X

Rao, S. N., Cortese, C., Miller, N. E., Levy, Y. & Lewis, B. (1982) FEES

Kinnunen, P. K. J. (1984) in Lipases (Borgstrom, B. & Brockman, H. L., Lett. 150, 255-259

Kuusi, T., Nikkila, E. A,, Tikkanen, M. J., Taskmen. M.-R. & Ehnholm, eds) pp. 307-328, Elsevier/North-Holland, New York

C. (1983) in Atherosclerosis VI. Proceedinm of The Sixth International

_l_ "_

J. R. & Jurutka, K. R., eds) pp. 628-632, Springer-Verlag New York Inc., Syzizposibz (Schetter, F. G., Gotto, A. M., Uiddelhoff, G., Habenicht, A.

Kuusi, T., Kinnunen, P. K. J. & Nikkila, E. A. (1979) FEES Lett. 104 , New York

Jansen, H., van Tol, A. & Hulsmann, W. C. (1980) Biochem. Biophys. Res. 384-388

Bamberger, M., Lund-Katz, S., Phillips, M. C. & Rothblat, G. H. (1985) Commun. 92,53-59

Komaromy, M. C. & Schotz, M. C. (1987) Proc. Natl. Acad. Sci.U. S. A. 8 4 , Biochemistry 24,3693-3701

Young, R. A. & Davis, R. W. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 1526-1530

Wei, C. F., Chen, S. H., Yang, C. Y., Marcel, Y. L., Milne, R., Li, W. H., 1194-1198

Sparrow, J. T., Gotto, A. M., Jr. & Chan, L. (1985) Proc. Natl. Acad. Sci. U. S. A. 82,7265-7269

Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. Acad. Sei. U. S. A. 74, 5463-5467

Humphreys-Beher, M. G., Bunnell, B., VanTuinen, P., Ledbetter, D. H. & Southern, E. M. (1975) J. Mol. Biol. 98,503-517

VanTuinen, P., Johnson, K. R., Ledbetter, S. A., Nussbaum, R. L., Rovera,

Elder, F. F. B.. Nichols, M. M., Hood, 0. J. & Harrison, R. W. (1985) Am.

Kimura, M. (1983) ?he Neutral Theory of Molecular Evolution, Cambridge

Cheng, C. F., Bensadoun, A,, Bersot, T., Hsu, J. S. T. & Melford, K. H.

Aueustin, J.. Freeze. H.. Teiada, P. & Brown. W. V. (1978) J. Bid. Chem.

Kidd, V. J. (1986) Proc. Natl. Acad. Sci. U. S. A. 83,8918-8922

G. & Ledbetter, D. H. (1987) Oncogene 1,319-322

J. Med. Genet. 20 519-524

University Press, Cambridge, Great Britain

(1985) J. Biol. Chem. 260, 10720-10727

2 5 3 , 2912-2920

M. (1987) Science 235, 1638-1641 Wion, K. L., Kirchgessner, T. G., Lusis, A. J., Schotz, M. C. & Lawn, R.

Marshall, R. D. (1974) Biochem. SOC. Symp. 4 0 , 17-26 Goldberg, I. J., Le, N.-A,, Paterniti, R., Ginsberg, H. N., and Lindgren, F.

Murase, T. & Itakura, H. (1981) Atherosclerosis 39,293-300 Kirchgessner, T. G., Svenson, K. L., Lusis, A. J. & Schotz, M. C. (1987) J.

Ben-Zeev, O., Ben-Avram, C. M., Wong, H., Nikazy, J., Shively, J. E. &

Dayboff, M. 0. (1972) Atlas of Protein Sequence and Structure, Vol. 4,

Senda, M., Oka, ,K., Brown, W. V., Qasba, P. & Furuichi, Y. (1987) Proc.

Kerfelec, B., LaForge, K. S., Puigserver, A. & Scheele, G. (1986) Pancreas

DeCaro, J., Boudonard, M., Bonicel, J., Guidoni, A., Desnuelle, P. & Rovery,

T. (1982) J. Clin. Inuest. 70, 1184-1192

Biol. Chem. 262,8463-8466

Schotz, M. C. (1987) Biochim. Biophys. Acta 9 1 9 , 13-20

National Biomedical Research Foundation, Silver Spring, MD

Natl. Acad. Scz. U. S. A. 84,4369-4373

1 , 430-437

M. (1981) Biochim. Biophys. Acta 671, 129-138