Bioinformatics and Evolution of Vertebrate Pancreatic Lipase and Related Proteins and Genes

Background: Pancreatic lipase (PTL) functions in the presence of colipase in the hydrolysis of emulsified fats in the small intestine following secretion from the pancreas. Pancreatic lipase related protein 1 (PLR1) is also found in pancreatic secretions and may perform a regulatory role in lipolysis; PLR2 catalyses pancreatic triglyceride and galactolipase reactions while PLR3 may serve a related but unknown lipase function. Comparative PTL, PLR1, PLR2 and PLR3 amino acid sequences and structures and gene locations and sequences were examined using data from several vertebrate genome projects. Methods: Sequence alignments and conserved predicted secondary and tertiary structures were studied and key amino acid residues and domains identified based on previous reports on human and pig PTL. Comparative analyses of vertebrate PTL-like genes were conducted using the UC Santa Cruz Genome Browser. Phylogeny studies investigated the evolution of these vertebrate PTL-like genes. Data: Human and mouse PTL sequences shared 78% identities but only 64-68% identities with human and mouse PLR1 and PLR2 sequences. Several vertebrate PTL and PTL-like protein domains were predicted using bioinformatics including an N-signal peptide; N-glycosylation site(s); an α/β hydrolase fold region containing a catalytic triad; a ‘lid’ region for the active site; a ‘hinge’ separating the lipase and PLAT regions; and a C-terminal PLAT region. Eutherian mammalian PLR1 sequences retained residues (196Val/198Ala) responsible for the loss of triacylglycerol lipase activity whereas lower vertebrate PLR1 sequences retained the ‘active’ lipase residues. In contrast, mammalian PTL, PLR2 and PLR3 sequences exhibited lipase ‘active’ residues (196Ala/198Pro); chicken and frog PLR1 sequences retained the lipase ‘active’ residues; and opossum and platypus PLR1 sequences exhibited 196Ala/Ser198 and 196Ser/198Pro residues. A phylogenetic tree analysis provided evidence for four distinct vertebrate PTL-like gene families. Conclusions: Pancreatic lipase (PTL) and related genes and proteins (PLR1 and PLR2) are present in all vertebrate genomes examined whereas PLR3 is found only in primate genomes. The ‘inactive’ form of vertebrate PLR1 is restricted to eutherian mammals. Vertebrate PTL-like genes apparently originated in a vertebrate ancestor following gene duplication events of an ancestral PTL-like ancestral gene. Two separate lines of PTL-like gene evolution are proposed in lower vertebrates (PTL/PLR1 and PLR2), with a further gene duplication event (PLR2/PLR3) for primate genomes. Holmes and Cox, J Data Mining in Genom Proteomics 2012, 3:1 DOI: 10.4172/2153-0602.1000112 Journal of Data Mining in Genomics & Proteomics J o u r n a l of D ata Mi ning in Gmics & rot e o m i c s

Biochemical studies of mammalian PTL lipases have shown that these enzymes are members of an esterase/lipase superfamily hydrolysing long-chain acyl-triglycerides at a water/lipid interface [4,5,10,11]. Structural and molecular modeling studies of human, pig and horse PTL [12][13][14][15][16] have shown that these enzymes contain several structural and functional domains, includng an N-terminal signal peptide; an N-glycosylation site [1,4]; an α/β hydrolase fold containing a Ser/Asp/His active site catalytic triad [17]; a 'lid' which typically covers the active site but in the presence of substrate, colipase and bile salts, PTL undergoes a conformational change enabling access to the active site by the emulsified lipid substrate [14,[16][17][18][19][20]; a C-terminal β-sandwich (PLAT) domain which participates in lipid and colipase binding; and a 'hinge' region which separates the α/β hydrolase fold from the PLAT domain. Structural and biochemical studies of human PLR1 have shown similarities in sequence and domain regions with mammalian PTL [21], although no detectable in vitro catalytic activity with PTL-like substrates has been reported [6,8]. Human PLR1 has been 'reactivated' by site-directed mutagenesis at two sites corresponding to human PTL residues (Ala196Val and Pro198Ala) representing the active and inactive forms of PLR1, respectively [7,8].
No specific biochemical roles have as yet been reported for mammalian PLR1 although a possible role in modulating colipase binding to PTL has been suggested [5]. Ren and coworkers [22] have recently examined PLR1 knock-out (KO) mice and reported mature onset obesity with increased fat mass, and have proposed that the enzyme may function as a metabolic inhibitor of PTL-colipase catalyzed triglyceride metabolism.
Structural and biochemical studies of mammalian PLR2 have reported strong similarities with PTL although differences were observed in kinetic and colipase binding properties and significant species differences were also reported [4][5][6][7][23][24][25]. Significantly, this enzyme has been shown to be predominantly responsible for pancreatic lipase-like neonatal fat digestion in rats and mice [26] and participates in the dietary fat digestion in human newborn children [27,28]. PLR2 also functions as a galactolipase (these substrates are the major lipids in plant cells) and participates in the hydrolysis of triglycerides, phospholipids and vitamin A esters [5,[29][30][31]. There are no reports available concerning the functions of human PLR3 although a recent study [32] has demonstrated over-expression of this gene in hepatocellular carcinoma (HCC) indicating an association of PLR3 with HCC.
Evolutionary studies of vertebrate PTL-like genes and proteins have been largely restricted to structural and genomic comparisons with other members of the neutral lipase gene superfamily, including lipoprotein lipases (LPL), hepatic lipases (HL), endothelial lipases (EL) and phospholipases A1 [5,10,[33][34][35][36][37]. This paper reports predicted gene structures and amino acid sequences for several vertebrate pancreatic lipase (PTL) and related protein (PLR1, PLR2 and PLR3) genes and enzymes not previously reported. Predicted secondary and tertiary structures for vertebrate pancreatic lipase (PTL) and related proteins (PLR1, PLR2 and PLR3) are also described as well as the structural and evolutionary relationships of these genes and enzymes. Comparative and phylogeny studies also showed that amino acid sequences for the PTL-like family are highly conserved during vertebrate evolution and suggested that the genes encoding these enzymes were generated from gene duplication events of an ancestral PTL-like ancestral gene resulting in two separate lines of mammalian gene evolution: PTL/ PLR1 and PLR2/PLR3, with the latter gene being observed only in primate genomes.
BLAT (Blast Like Alignment Tool) analyses were subsequently undertaken for each of the predicted PTL-like amino acid sequences using the UC Santa Cruz genome browser [http://genome.ucsc. edu/cgi-bin/hgBlat] with the default settings to obtain the predicted locations for each of the vertebrate PTL-like genes, including predicted exon boundary locations and gene sizes for coding exons. Structures for human PTL, PLR1, PLR2 and PLR3 isoforms (splicing variants) were obtained using the AceView website to examine predicted gene and protein structures (http://www.ncbi.nlm.nih.gov/IEB/Research/ Acembly/index.html?human) [48]. The UC Santa Cruz web browser genome browser (http://genome.ucsc.edu) was used to examine comparative structures for vertebrate PTL-like genes [49].

Predicted structures, properties and alignments of vertebrate PTL-like sequences
Predicted secondary and tertiary structures for human and other vertebrate PTL-like proteins were obtained using the PSIPRED v2.5 web site tools [http://bioinf.cs.ucl.ac.uk/psipred/psiform.html] and SWISS MODEL web tools [http://swissmodel.expasy.org/], respectively. The reported tertiary structure for human PTL [13] (PDB entry 1n8s) was used as the reference for the frog PTL tertiary structure with a modeling range of residues 17-465; and the reported structure for human PLR1 (unpublished results: PDB entry 2ppl) served as a reference for frog PLR1 with a modeling range of 19 to 467). Alignments of vertebrate PTL-like sequences were assembled using the ClustalW2 alignment program (http://www.ebi.ac.uk/Tools/clustalw2/index.html) [50].

Phylogenetic studies and sequence divergence
Alignments of vertebrate PTL-like sequences and human (Homo sapiens), mouse (Mus musculus), frog (Xenopus tropicalis) and zebrafish (Danio rerio) EL (endothelial lipase), LPL (lipoprotein lipase) and HL (hepatic lipase) sequences were assembled using BioEdit v.5.0.1 and the default settings [51]. Alignment ambiguous regions, including the amino and carboxyl termini, were excluded prior to phylogenetic analysis yielding alignments of 395 residues for comparisons of vertebrate PTL-like, EL, LPL and HL sequences with the fruit fly (Drosophila melanogaster) CG5966 lipase sequence (Table  1; Supplementary Table 1). Evolutionary distances were calculated using the Kimura option [52] in TREECON [53]. Phylogenetic trees were constructed from evolutionary distances using the neighborjoining method [54] and rooted with the fruit fly CG5966 lipase sequence. Tree topology was reexamined by the boot-strap method (100 bootstraps were applied) of resampling and only values that were highly significant (≥90) are shown [55].

Alignments of vertebrate PTL amino acid sequences
The deduced amino acid sequences for rhesus monkey (Macaca mulatta), dog (Canis familiaris) and opossum (Monodelphis domestica) PTL, which were identified in the NCBI database (see Table 1) (http://www.ncbi.nlm.nih.gov/), are shown in Figure 1 together with previously reported PTL sequences for human (Homo sapiens) [12,56] [15], chicken (Gallus gallus) [58] and catfish (Ictalurus punctatus) [59]. These PTL sequences were chosen as representatives of primate (human and rhesus monkey), eutherian (rat, horse and dog), marsupial (opossum), bird (chicken) and fish (catfish) species. Alignments of human and other vertebrate PTL sequences showed between 55-93% identities, suggesting that these are products of the same family of genes, whereas comparisons of sequence identities of vertebrate PTL proteins with human and mouse hepatic lipase (HL), lipoprotein lipase (LPL) and endothelial lipase (EL) exhibited much lower levels of sequence identities (24-27%), indicating that these are members of distinct lipase families ( Table 2).
The amino acid sequences for human, rhesus monkey, rat, horse, dog and opossum PTL contained 465 residues whereas chicken and catfish PTL contained 467 and 471 amino acids, respectively ( Figure  1). Previous structural studies on horse and human PTL [13,15] have identified key residues for these vertebrate PTL proteins (sequence numbers refer to human PTL). These included the catalytic triad for the active site (Ser169; Asp193; His280); the hydrophobic N-terminus signal peptides (see also Table 1) which assist secretion of the enzyme from acinar pancreatic cells [5]; disulfide bond forming residues (Cys20/Cys26; Cys106/Cys118; Cys254/Cys278; Cys302/Cys313; Cys316/Cys321; Cys448/Cys465); the predicted 'lid' region (255-277) which covers the active site and participates in lipid substrate binding in analogous lipases [12,15]; and a 'hinge' region for vertebrate PTL which spatially separates the α/β N-terminal hydrolase region from the C-terminal PLAT β-sandwich region (residues 354-465) [12,15]. With the exception of the N-terminus signal peptides, nearly half of these residues (42%) are strictly conserved or underwent conservative substitutions (19%) indicating the essential nature of these residues in contributing to PTL structure and function. In contrast, the N-terminal region (residues 1-16) underwent major changes in the number and sequence of amino acid residues but retained a predicted signal peptide property in each case ( Figure 1; Table 1). An N-glycosylation site, previously reported for human and pig PTL at 183Asn-184Gly-185Thr [12], was also predicted for the rhesus and dog PTL sequences ( Figure  1). In contrast, this site was absent from the horse PTL sequence which may explain why this enzyme lacks a glycosyl group [60]. Rat, opossum and chicken PTL sequences also lacked this human PTL N-glycosylation site whereas catfish PTL contained three predicted N-glycosylation sites for this enzyme: 55NAS; 306NKT; 394NGT ( Figure 1). The significance of these differences in N-glycosylation binding for vertebrate PTL remains to be investigated.

Alignments of vertebrate PTL-like amino acid sequences
Alignments of vertebrate PTL, PLR1 and PLR2, as well as primate PLR3 amino acid sequences examined, showed between 42-93% identities, suggesting that these are products of a conserved gene family during vertebrate evolution ( Table 2). In addition, comparisons of mammalian PTL, PLR1, PLR2 and PLR3 sequences also showed higher levels of sequence identities within each of the PTL-like subgroups of enzymes, suggesting that these are distinct members within the mammalian PTL gene family.
Amino acid sequence alignments for previously reported human (Homo sapiens) PTL [56], PLR1 [61], PLR2 [6] and PLR3 [9] and mouse PTL [42,62], PLR1 [61] and PLR2 [26,42] sequences are shown in Figure 2 (also see 1). The amino acid sequence for the human PLR1 subunit contained 467 residues while the mouse PLR1 sequence contained 473 amino acids, due to an extended C-terminus sequence. Human and mouse PLR2 sequences were similar in length to PTL and PLR1 with 469 and 468 amino acids respectively, while human PLR3 contained 467 amino acids. Structural studies of mammalian PLR1 and PLR2 sequences have enabled the identification of key residues for these enzymes identical to or comparable with those previously described for human PTL (see Figure 1). Key residues for each of the human and mouse PTL-like sequences included the following (human PTL sequence numbers were used) ( Figure 2): the catalytic triad for the active site (Ser169; Asp193; His280); the predicted N-signal peptide regions (residues 1-16) with a glycine-↓-lysine/arginine peptide bond recognized for hydrolysis in each case; 12 disulfide bond forming residues (Cys20/Cys26; Cys106/Cys118; Cys254/Cys278; Cys302/Cys313; Cys316/Cys321; Cys448/Cys465); distinct predicted N-glycosylation sites, including human PTL at 183Asn-184Gly-185Thr [12], mouse PTL at 293Asn-294Pro-295Thr, mouse PLR1 at 159Asn-   Table 1 for sources of vertebrate PTL sequences; * shows identical residues for proteins; : one similar alternate residue; . two similar alternate residues; key PTL active site residues: Ser169, Asp193 and His280 (human PTL); N-signal peptide residues are in red; N-glycosylation residues are in shaded green; five disulphide bonds are represented as |----| with Cys residues in shaded blue; β-sheets (β1-β22) are numbered according to human PTL sequences [12] in shaded grey; α-helices (α1-10) are also numbered according to the human PTL structure and are in shaded yellow; |####| represents 'lid' region covering active site; colipase binding residues are in purple; hinge refers to region connecting the 'lipase' and 'plat' domains; bold underlined font shows residues corresponding to known or predicted exon start sites; exon numbers refer to human PTL gene coding exons. 160Tyr-161Ser (none predicted for human PLR1), human PLR2 at 353Asn-354Phe-355Thr and four predicted N-glycosylation sites for human PLR3 at 74Asn-75Ser-76Ser, 125Asn-126Gly-127Ser, 338Asn-339Gly-340Ser and 412Asn-413Ile-414Thr; and residues previously recognized for forming salt bridges between human PTL and colipase [21], including 385Gln, 406Asp and 416Lys, which are retained for the human and mouse PTL and PLR2 sequences but replaced with lysine for the second of these residues for the human and mouse PLR1 sequences and with glutamate and glutamine residues respectively, for the first and third residues of the human PLR3 sequence. Figure 3 presents amino acid sequence alignments for vertebrate PTL, PLR1, PLR2 and primate PLR3 sequences for a region of the PTL active site previously shown to contribute significantly to lipase activity. Human PLR1 has been 'reactivated' by site-directed mutagenesis at two sites corresponding to human PTL residues (Ala196Val and Pro198Ala), representing the active and inactive forms of PLR1, respectively [7,8]. Comparisons of vertebrate PTL, PLR2 and primate PLR3 sequences in this region show that the 'active' Ala196/ Pro198 residues were observed in each case, whereas 'inactive' Val196/ Ala198 residues were retained for each of the 11 eutherian PLR1 sequences examined. In contrast, the predicted opossum and platypus PLR1 sequences contained Ala196/Ser198 and Ser196/Pro198 residues respectively, while the predicted chicken and frog PLR1 sequences retained the 'active' PTL-like residues, Ala196/Pro198. This suggests that the 'inactive-lipase' PLR1 property has been conserved for all eutherian PLR1 sequences examined, whereas PLR1 sequences from earlier mammalian and vertebrate species may be 'active-lipase' forms of this enzyme.
Supplementary Figure 1 shows amino acid alignments for the 'lid' regions for vertebrate PTL, PLR1, PLR2 and primate PLR3 sequences (residues 255-277 for human PTL) [12]. High levels of sequence identities were observed for each of the vertebrate PTL-like family sequences for this region, indicating that this is a highly conserved region for all vertebrate (PTL, PLR1 and PLR2) and primate (PLR3) sequences. In addition, the 'lid' sequences separated into 2 groups according to their comparative sequence identities: vertebrate PTL and PLR1; and vertebrate PLR and primate PLR3, which are very similar sequences in each case. The surface loop has been shown to cover the active site of human PTL and influences interfacial activation and lipid binding for this enzyme [18]. Different specificities for these properties are likely given the existence of two discrete PTL-like 'loop' sequences for the PTL/PLR1 and PLR2/PLR3 enzymes.

Secondary and tertiary structures for vertebrate PTL-like lipases
Predicted secondary structures for rhesus monkey, rat, dog, opossum, chicken and catfish PTL sequences were compared with those previously reported for human and horse PTL [12][13][14][15][16] (Figure 1). Alpha-helix and β-sheet structures for these sequences were numbered as for those described for horse PTL [15] which were predominantly retained for all vertebrate PTL sequences examined. Similar predicted secondary structure analyses were also undertaken for human and mouse PTL, PLR1, PLR2 and human PLR3 sequences (Figure 2) which demonstrated that the major structural features previously reported have been retained for each of these predicted sequences. Figure 4 compares previously reported structures for human PTL [12] and PLR1 (submitted to PDB data bank: 2PPL) protein sequences with predicted tertiary structures for the frog (Xenopus tropicalis) PTL and PLR1 subunits. Major features for these proteins, based on previous reports [12][13][14][15][16], demonstrated separation into two major functional domains: a 'lipase' domain containing several parallel β-sheets, three large α-helices and an 'active-site' zone covered by a 'lid' and a 'plat' domain, which are separated by a 'hinge' region. These overall features are readily observed for the predicted frog PTL and PLR1 three-dimensional structures. Table 1 and Supplementary Table 1 summarize the predicted locations for vertebrate PTL-like genes based upon BLAT interrogations of genomes using the reported sequences for human PTL, PLR1, PLR2 and PLR3 [6,9,56,61] and the predicted sequences for other vertebrate PTL-like proteins and the UC Santa Cruz Genome Browser [49]. The mammalian PTL-like genes were predominantly transcribed on the positive strand and located in the following gene order: PLR3-PTL-PLR1-PLR2 for human (see Supplementary Figure 2), orangutan and rhesus genomes; and PTL-PLR1-PLR2 for other mammalian genomes, including mouse, rat, dog, pig, rabbit, opossum and platypus genomes (Table 1 and Supplementary Table 1). Horse PTL-like genes were however located on the negative strand but in the same gene order as for other eutherian mammalian genomes (Supplementary Table 1). In contrast, PTL-like genes were located on chicken chromosome 6 on the positive strand in the reverse order (PLR2-PLR1-PTL), which suggests that a chromosomal rearrangement has occurred in the region of the PTL-like genes, although in each case, the three genes are proximally localized on mammalian and bird genomes. Figures 1 and 2 summarize the coding exon start sites for vertebrate PTL genes, human PLR1, PLR2 and PLR3 genes and mouse PLR1 and PLR2 genes showing 12 coding exons in identical or similar positions which are consistent with previous reports [9,[63][64]. This is consistent with a common evolutionary origin for these genes. Figure 5 shows the predicted structures of mRNAs for human PTL, PLR1, PLR2 and PLR3 transcripts for the major isoform in each case [48]. These human mRNA transcripts varied in length from 18.8 kilobases to 50.1 kilobases for these PTL-like genes. The human PLR2 and PLR3 gene transcripts contained extended untranslated 3'-UTR regions, which in the case of the PLR2 transcript, was encoded by 3 distinct exons. The levels of expression for these human PTL-like genes have been compared with the average level of gene expression observed in the human genome (see Table 1). Three of these genes exhibited higher levels of  Table 1 for sources of human and mouse PTL, PLR1, PLR2 and human PLR3 sequences; * shows identical residues for proteins; : one similar alternate residue; . two similar alternate residues; key PTL active site residues: Ser169, Asp193 and His280 (human PTL); N-signal peptide residues are in red; predicted or known N-glycosylation residues are in shaded green; five disulphide bonds are represented as |----| with Cys residues in shaded blue; β-sheets (β1-β22) are numbered according to human PTL sequences [12] in shaded grey; α-helices (α1-10) are also numbered according to the human PTL structure and are in shaded yellow; |####| represents 'lid' region covering active site; colipase binding residues are in pink; hinge refers to region connecting the 'lipase' and 'plat' domains; bold underlined font shows residues corresponding to known or predicted exon start sites; exon numbers refer to human PTL gene coding exons.  Table 1  expression [PTL (x1.8); PLR1 (x1.5); PLR2 (x1.5)], whereas human PLR3 exhibited a lower than average level of expression (x0.2). Higher levels of gene expression were also observed for the mouse PTL-like genes: Ptl (x4.5 times); Plr1 (x2.9); and Plr2 (1.6) [48]. These high levels of gene expression are consistent with the major roles for these genes and encoded enzymes in lipid digestion in the body. Figure 6 presents 'heat maps' showing comparative gene expression for various human tissues obtained from GNF Expression Atlas Data using GNF1M chips [64] (http://genome.ucsc.edu; http://biogps.gnf.org). These data supported differential tissue expression for human PTL-like genes, with PTL, PLR1 and PLR2 showing highest levels in pancreatic islet tissue, which is consistent with previous reports for these genes ( Figure  5). In contrast, the human PLR3 gene was most highly expressed in bronchial epithelial cells with an as yet unknown function in the body.

Phylogeny and divergence of PTL-like sequences
A phylogenetic tree (Figure 7) was calculated by the progressive alignment of 55 vertebrate PTL-like amino acid sequences with twelve vertebrate hepatic lipase (HL), lipoprotein lipase (LPL) and endothelial lipase (EL) amino acid sequences. A sea squirt (Ciona intestinalis) lipase served as a 'root' for the vertebrate and invertebrate PTL-like sequences, and a fruit fly (Drosophila melanogaster) lipase sequence (CG5966) served as the 'root' for the tree overall (see Table 1 and Supplementary Table 1). The phylogram showed clustering of the lipase sequences into several groups, including mammalian PTL, PLR1 and PLR2 sequences; primate PLR3 sequences; chicken (Gallus gallus) and lizard (Anolis carolinensis) PTL sequences; frog (Xenopus tropicalis) PTL, PLR1 and PLR2 sequences; fish (seabream: Chrysophrys major; catfish: Ictalurus punctatus) PTL sequences; as well as distinct groups for the three neutral lipases: HL, LPL and EL. In addition, the phylogram suggested a sequence of gene duplication events for an ancestral PTLlike gene during vertebrate evolution (see Figure 7): (1) an ancestral PTL gene duplication generating the PTL/PLR1 and PLR2 precursor genes; (2) duplication of the vertebrate PTL/PLR1 precursor gene to form the PTL and PLR1 genes; and (3) duplication of the PLR2 gene prior to primate evolution to form the PLR2 and PLR3 genes, currently observed in primate genomes. It is also apparent that vertebrate neutral lipase genes (HL; EL; and LPL) have been generated during vertebrate evolution from a distinct ancestral gene, which is consistent with previous studies on the enzymes and genes [33][34][35][36][37].

Conclusions
These results demonstrate that mammalian and other vertebrate pancreatic-like (PTL) genes and encoded enzymes comprise at least four distinct forms, designated as PTL, PLR1, PLR2 and PLR3, which is consistent with previous studies [5,9,31]. PTL functions in the presence of colipase in the hydrolysis of emulsified fats in the small intestine following secretion from the pancreas [1][2][3][4][5]; PLR1 is also found in pancreatic secretions but has an as yet unknown function and exhibits no detectable triacylglycerol, cholesterol ester or galactolipase activities [6,8]; PLR2 functions as a pancreatic secretion galactolipase but may also contribute to T cell cytotoxicity in the body [6,25,65]; while PLR3 serves a related, but as yet unknown, lipase function.
PTL, PLR1 and PLR2 were each encoded by single genes among most vertebrate genomes. These genes are highly expressed in human and mouse pancreatic islet cells, which is consistent with their roles in lipid digestion in the intestine. In contrast, PLR3 was expressed at lower than average levels for human genes overall and was highly expressed in bronchial epithelial cells. Predicted secondary and tertiary structures for frog PTL and PLR1 proteins showed similarities with human and pig PTL [12][13][14][15][16] with two major functional domains: a 'lipase' domain containing several parallel β-sheets, three large α-helices and an 'active-site' zone covered by a 'lid' and a 'plat' domain separated by a 'hinge' region. Comparisons of PTL, PLR1, PLR2 and PLR3 amino acid sequences from vertebrates representative of mammals, birds, amphibians and bony fish, demonstrated that these are highly conserved proteins during evolution, not only for active site  [12] and from a submitted report to the PDB data bank (2PPL) respectively, and using SWISS MODEL methods for the frog PTL and PLR1 sequences; the rainbow color code describes the known tertiary structures from the N-(blue) to C-termini (red color); arrows indicate directions for β-sheets; known or active site for human PTL; N-terminal and C-terminal regions are shown, as are predicted structures and locations for 'lipase' and 'plat' domains; the 'lid' covering the active site; and the 'hinge' separating the 2 domains.  residues but also for those involved in lipid binding ('plat' domain) and shielding the active site from the aqueous environment near the active site (the 'lid'). Vertebrate PTL, PLR1, PLR2 and PLR3 (the latter enzyme was restricted to primates) shared between 42-93% sequence identities but <30% with the vascular lipases, HL (hepatic lipase), EL (endothelial lipase) and LPL (lipoprotein lipase), demonstrating that the vertebrate PTL-like lipases are a distinct family of related genes and proteins. Sequence alignments, key amino acid residues and conserved predicted secondary and tertiary structures were examined, including active site residues; disulfide bonds; predicted N-glycosylation and signal peptide sites; 'plat' and 'lid' domains; and a 'hinge' region.
Phylogeny studies of vertebrate PTL, PLR1, PLR2 and PLR3 genes and enzymes suggested that they originated in a vertebrate ancestor from successive gene duplication events of an ancestral PTL-like gene, forming initially two precursor genes (PTL/PLR1 and PLR2); followed by duplication of the former precursor gene to form the PTL and PLR1 genes; and (3) duplication of the PLR2 gene prior to primate evolution to form the PLR2 and PLR3 genes in primate genomes.
Overall, these results significantly contribute to our knowledge concerning the evolution of pancreatic lipase-like genes, enzymes with their major roles within pancreatic secretions in the digestion of emulsified fats in the duodenum, and in the case of PLR2, the digestion of plant lipids by serving as a galactolipase. In contrast, PLR3 genes and proteins were restricted to primates, serve an as yet unknown function(s) in the body and exhibit highest expression levels in human bronchial epithelial cells. Comparisons of vertebrate PTL, PLR2 and primate PLR3 sequences showed that the 'catalytically active' PL-like sequence (Ala196/Pro198) was observed for all species examined whereas a 'catalytically inactive' Val196/Ala198 PL-like sequence was observed for all eutherian mammalian PLR1 sequences examined. In contrast, chicken and frog PLR1 sequences retained 'active' PTLlike residues, Ala196/Pro198, whereas marsupial (opossum) and monotreme (platypus) PLR1 sequences contained Ala196/Ser198 and Ser196/Pro198 residues respectively. This suggests that the 'inactivelipase' PLR1 form, which plays a specialized metabolic inhibitor role [22], may function only in eutherian mammals, whereas PLR1 sequences from earlier mammalian and vertebrate species are 'activelipase' forms of this enzyme and may serve as additional forms of PTLlike lipases in these species. Further biochemical and genetic studies are needed to describe any specific role(s) for PLR1 in lower vertebrate species.