Molecular cloning and characterization of a novel mammalian endo-apyrase (LALP1).

Here we describe the cloning, localization, and characterization of a novel mammalian endo-apyrase (LALP1) in human and mouse. The predicted human LALP1 gene encodes a 604-amino acid protein, whereas the mouse Lalp1 gene encodes a 606-amino acid protein. The human and mouse genes have 88% amino acid sequence identity. These genes share considerable homologies with hLALP70, a recently discovered mammalian lysosomal endo-apyrase. The human LALP1 gene resides on chromosome 10q23-q24 and contains 12 exons and 11 introns covering a genomic region of approximately 46 kilobase pairs. The subcellular localization and enzymatic activity of LALP1 indicated that LALP1 is indeed an endo-apyrase with substrate preference for nucleoside triphosphates UTP, GTP, and CTP.

Apyrases are hydrolytic enzymes that cleave nucleoside triand diphosphates in a calcium-or magnesium-dependent manner but are insensitive to P-, F-, or V-type ATPase inhibitors (1). Both apyrases and ATPases can use ATP as substrates; however, there are substantial differences (1,2). For instance, the ATPases are highly selective toward their substrate ATP with ADP and phosphate as reaction products. On the contrary, apyrases can use different nucleoside tri-and diphosphates as their substrates, with phosphates and nucleoside monophosphates as the main reaction products. Additionally, sequence comparison among various apyrases revealed the presence of apyrase conserved regions, which are highly conserved among all apyrases from various organisms such as plants, parasites, and mammals. Thus far, two types of apyrases, ecto-apyrases and endoapyrases, have been described (2). Ecto-apyrases such as CD39 are apyrase enzymes with their catalytic domain exposed on the cell surface (1)(2)(3) and are believed to be involved in many processes such as neurotransmission (4), platelet aggregation, and blood pressure regulation (5). On the other hand, endo-apyrases such as uridine diphosphatase and the 700-kDa lysosomal apyrase-like protein (LALP70) are apyrase enzymes with their catalytic domain localized intracellularly (6,7). The biological functions of the endo-apyrases are unknown, although it was suggested that they are important for regulating the level of activated sugar during protein glycosylation.
Recently, Wang and Guidotti (6) identified the first mammalian endo-apyrase, human uridine diphosphatase. This enzyme is predicted to be a 610-amino acid protein with two putative transmembrane domains. Using a myc-tagged version of this protein, this enzyme was found to be in the luminal side of the Golgi apparatus and had the highest catalytic activity using UDP as its substrate and had lower activities using GDP, CDP, UTP, GTP, CTP, and TTP. Interestingly, Biederbick et al. (7) reported altered substrate preferences of LALP70, which is a splicing variant of uridine diphosphatase and has an additional 8 amino acids (VSFASSQQ) resulting from the inclusion of an alternate exon. The enzymatic properties of this splice variant revealed a broad substrate specificity, with CTP, UDP, CDP, GTP, and GDP as preferred substrates. Using antibodies and a green fluorescent protein-tagged version of this splicing variant, LALP70 was co-localized with the autophagic marker monodansylcadaverine and the lysosomal protein lamp1, suggesting a lysosomal/autophagic vacuole subcellular location (8). Although most of the transiently expressed LALP70/GFP 1 fusion protein was co-localized with lamp1-positive vacuoles, an association with the Golgi apparatus and the endoplasmic reticulum could not be ruled out. These studies clearly demonstrated that LALP70/uridine diphosphatase is the first mammalian member of the endo-apyrase gene family. Here we report the molecular cloning and characterization of the LALP1 gene, which closely resembles LALP70 in both structure and enzymatic property. Hence, we propose that LALP1 is a second member of the endo-apyrase gene family.

EXPERIMENTAL PROCEDURES
5Ј-RACE and 3Ј-RACE-5Ј-and 3Ј-RACE reactions were conducted with the SMART RACE kit (CLONTECH) according to the manufacturer's instructions, with a slight modification as described previously (11,12). Briefly, for 5Ј-RACE, the first-strain cDNA synthesis is primed using a gene-specific primer and a SMART oligonucleotide with human brain total RNA. After reverse transcription reaches the end of the mRNA template, several dC residues are added to the end of the cDNA. The SMART oligonucleotide, which contains several 3-dG at its 3Ј end, anneals to the tail of the newly synthesized cDNA and then serves as a template for further extension of the cDNA by RT. After the RT reaction, an internal gene-specific reverse primer and a UP primer, which is complementary to SMART oligonucleotide, were used to perform PCR using the RT products as templates. To increase the specificity and * This research was supported in part by National Institutes of Health Grants 1R01DK53266-01 (to J.-X. S) and AG015688 (to D. H. W). 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 accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ These authors contributed equally to this work. product yield of 5Ј-RACE, nested PCR was then performed using another internal gene-specific primer and NUP primer (internal primer of UP primer). For 3Ј-RACE, the first-strand cDNA was synthesized using a modified oligo(dT) with a UP oligonucleotide tail.
The UP primer and a gene-specific forward primer were used for first-round PCR. Nested PCR was performed using NUP and an internal gene-specific forward primer. PCRs were carried out in a final volume of 20 l. After the RT-PCR, the samples were denatured at 94°C for 5 min; amplifications were carried out with 5 cycles of 30-s denaturing at 94°C, 30-s annealing at 68°C, and a 4-min extension at 72°C followed by 30 cycles of 30-s denaturing at 94°C, 30-s annealing at 62°C, and a 4-min extension at 72°C. PCR products obtained from nested PCR were loaded onto a 2% agarose gel, and individual bands were excised from the gel for direct sequencing. The PCR product was subcloned into TA vector (Invitrogen, San Diego, CA), and the sequence was determined with ABI 310 DNA sequencer as described previously (12)(13)(14).
Expression Analysis by Competitive RT-PCR-Total RNA was isolated from various mouse tissues using the RNeasy kit (Qiagen) according to the manufacturer's instructions. Total RNA (5 g) was used for RT reactions in a total volume of 20 l using a poly(T)-plus primer as described previously (13,14). Denatured RNA samples and the primer were incubated with reverse transcriptase at 42°C for 1 h. The RT reaction was then stopped by heating the samples at 95°C for 10 min. 2 l of the cDNA was used as template for a subsequent PCR in a total volume of 20 l. To assess the relative expression level of the mouse Lalp1 gene, competitive RT-PCR was performed. Briefly, a 750-bp Lalp1 fragment (mExon8F and 5ЈUTR-R primers) was co-amplified with a 450-bp ␤-actin fragment (Table  I) in the same PCR. The annealing temperature was set to 58°C, and the PCR was run for 30 cycles. The PCR products were analyzed on a 2% agarose gel in Tris-borate EDTA buffer.
Expression Constructs-Two expression constructs were made to study the enzymatic activity and cellular localization of the human LALP1 protein. First, full-length LALP1 cDNA was subcloned into mammalian expression vector pEGFP N1 (CLONTECH) using engineered HindIII restriction endonuclease sites via PCR. This construct is designated pEGFP/LALP1. In the second construct (pEGFP/LALP1-GFP), LALP1 cDNA is fused with a C-terminal GFP to produce LALP1-GFP fusion protein.
Cell Culture and Transfection-HEK297 cells were cultivated in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum and 50 g/ml gentamycin and incubated at 37°C in a humidified chamber equilibrated with 5% CO 2 . Cells were transfected with mammalian pEGFP expression constructs using LipofectAMINE (Life Technologies, Inc.) as transfection reagent following the manufacturer's instructions. Transfected cells were harvested for the apyrase assay 24 h after transfection.
Subcellular Localization with GFP Fusion-The cellular localization of LALP1 was determined by transfection experiments with the LALP1-GFP plasmid. After transfection, cells were cultured in chamber slides (LAB-TEK) for 12-18 h. Transfected cells were then fixed with 4% paraformaldehyde for 20 min and mounted with coverslips. The fluorescent images were obtained using a confocal microscope (Bio-Rad) equipped with an argon laser with excitation at 488 nm and detection at 510 -530 nm bandpass for GFP.
Nucleotide Phosphatase Activity Assay-Nucleotide phosphatase activity was measured in crude membranes from HEK293 cells transfected with pEGFP vector or pEGFP/LALP1 construct. Ten flasks (T75) of transfected HEK293 cells were homogenized with a Dounce homogenizer, and nuclei were separated according to the method of Wang and Guidotti (6). To separate the crude membrane fraction from the cytosol, the postnuclear supernatant was centrifuged at 100,000 ϫ g. The pellets were resuspended in 500 l of 20 mM Hepes, pH 7.4, 120 mM NaCl, 5 mM KCl, and 0.2 mM EDTA containing 0.1% Triton X-100. The protein concentration in each sample was determined with bicinchoninic acid (Pierce) according to the manufacturer's instructions.
To measure apyrase activity, HEK297 membrane suspension containing 10 g of total protein was adjusted to 45 l with reaction buffer containing 20 mM Hepes, pH 7.4, 120 mM NaCl, 5 mM KCl, 0.2 mM EDTA, 1 mM NaN 3 , and 0.5 mM Na 3 VO 4 , with or without 5 mM CaCl 2 . After preincubation for 5 min at 37°C nucleotide phosphatase reactions were initiated by the addition of 5 l of the same buffer containing 10 mM nucleotide phosphate substrates to give a final concentration of 1 mM. Samples were incubated for 30 min at 37°C. NTP/NDP hydrolysis under these conditions was linear up to 30 min. Apyrase activity was determined by measuring the inorganic phosphate released as described previously (6 -7). Values obtained from samples without CaCl 2 or MgCl 2 were subtracted from those obtained with CaCl 2 or MgCl 2 . All measurements were done in triplicate. Student's t test was performed to assess statistical significance.

RESULTS AND DISCUSSION
Cloning of the Human LALP1 Gene-The human LALP1 gene was cloned using SMART RACE technology. Based on the sequence of expressed sequence tag N35618 in the human chromosomal region 10q23-q24 (9, 10), several gene-specific reverse primers (hLALP1-RT, hLALP1-R1, and hLALP1-R2) and forward primers (hLALP1-F1 and hLALP1-F2) were designed for 5Ј-RACE and 3Ј-RACE, respectively. The resulting full-length sequence (Fig. 1) was 2962 bp (GenBank TM accession number AF269255). A GenBank TM search indicated that the human LALP1 gene is contained in clone RP11-483F11 (GenBank TM accession number AL133353). Comparison of cDNA and genomic sequences suggested that LALP1 has 12 exons that expand about 46 kilobase pairs of genomic sequence (Fig. 2). The cloned cDNA coincides almost exactly with the computer-predicted structure. The translation initiation codon ATG is located near the end of exon 1, which contains only eight coding nucleotides. Multiple

Cloning and Characterization of LALP1
in-frame stop codons are present upstream of the putative first ATG. A CpG island was identified at Ϫ124 bp upstream of the putative first ATG. A putative polyadenylation signal sequence (AATAAA) is found at 28 bp upstream of the poly(A) tail. The LALP1 gene contains 1,812 nucleotides of coding sequence and encodes a protein of 604 amino acids with a molecular mass of 69.88 kDa. Interestingly, the exon/intron structure of LALP1 is almost identical to that of the LALP70 gene (7).

Cloning of the Mouse Lalp1
Gene-The mouse Lalp1 gene was cloned based on homology with its human homologue. Briefly, two human LALP1 primers (hExon4-F and hExon8-R) were used to amplify the mouse gene from mouse brain cDNA. A fragment of 450 bp encompassing the region from exon 4 to exon 8 was obtained and directly sequenced. This mouse fragment shares 86.8% homology with the human LALP1 homologous region, confirming that the fragment is indeed derived from the mouse Lalp1 gene. A mouse reverse primer (mLALP-R1) was designed based on the partial mouse Lalp1 sequence and used together with a human forward primer (hExon1-F) in RT-PCR. This RT-PCR generated a fragment of 730 bp, which shares 89.5% homology with the human LALP1 sequence. To obtain the complete RT-PCR was conducted using total RNA from various mouse tissues and the mExon8-F/5ЈUTR-R primer pair (752 bp). A 450-bp ␤-actin fragment was coamplified in the same PCR. PCR products were electrophoresed on a 2% agarose gel. The ratio of Lalp1/␤-actin reflects the relative expression level of the Lalp1 gene. sequences at the 5Ј and 3Ј ends of the gene, 5Ј-and 3Ј-RACE were performed using several reverse and forward primers (mLALP1-RT, mLALP1-R2, mLALP1-F1, and mLALP1-F2).
Mouse Lalp1 cDNA contains 3268 bp (GenBank TM accession number AF288221) with an open reading frame of 1818 bp and encodes a protein of 606 amino acids (with a molecular mass of 69.99 kDa). The mouse and human cDNA sequences share 87% nucleotide sequence similarity in the entire coding region, whereas the sequence homology in 5Ј-UTR and 3Ј-UTR is poor, further supporting the open reading frame assignment. The deduced protein sequences are also highly conserved between human and mouse (88% identity).
To compare the relative expression levels in different tissues, competitive RT-PCR was performed with RNA from different mouse tissues. As shown in Fig. 3, Lalp1 is highly expressed in most tissues including the brain, kidney, liver, and testis. The expression is relatively lower in the lung, thymus, and heart. Among all tissues analyzed, the expression is lowest in the spleen. These results suggest that the LALP1 protein must play critical roles in the cellular functions of many different tissues.
Nucleotide Phosphatase Activities-A BLAST search revealed that the LALP1 and Lalp1 proteins have significant sequence homologies with human LALP70 and guanosine diphosphatase protein family members (Fig. 4). The human LALP1 shares 59% identity and 71% similarity to LALP70, 58% identity and 70% similarity to guanosine diphosphatase, and 25% identity and 42% similarity to CD39. These results suggest that LALP1 may be an apyrase.
To determine whether LALP1 is indeed an apyrase, nucleotide phosphotase activity was measured in HEK293 cells transfected with the pEGFP vector or the full-length LALP1 cDNA constructs (pEGFP/LALP1). The experiments were done in the presence of 1 mM azide (inhibitor of F-type ATPase) and 0.5 mM vanadate (inhibitor of P-type ATPase), which do not inhibit the activities of E-ATPases (1-2). As shown in Fig. 5, the mean activities were higher for cells containing LALP1 expression vector than in cells containing control vector. For example, the CTPase activity (24.9 Ϯ 1.46 nmol P i /min/mg) with the LALP1 construct is 5-fold higher than the CTPase activity (4.34 Ϯ 2.02 nmol P i /min/mg) for cells transfected with control vector. Student's t test suggested that the mean activities were significantly different between LALP1 and control vector with UTP (p Ͻ 0.02), GTP (p Ͻ 0.002), and CTP (p Ͻ 0.0001) as the substrates (Fig. 5). Furthermore, the nucleotide phosphatase activities are absolutely dependent on the presence of Ca 2ϩ ions (data not shown). Therefore, our data indicate that human LALP1 is truly an apyrase with a substrate preference for UTP, GTP, and CTP. It is interesting to note that the HEK293 cells have very high enzymatic activities for the GDP and UDP nucleotide substrates (controls in Fig. 5) due to endogenous apyrases. However, GDP and UDP do not seem to be preferred substrates for the LALP1 protein because LALP1-overexpressing cells do not have significantly higher enzymatic activities compared with control cells.
We then determined whether LALP1 is an ecto-apyrase or an endo-apyrase. Two sets of experiments were performed. First, the enzymatic activity of intact and disrupted cells was compared to determine whether it is located intracellularly or extracellularly. The CTPase activity of LALP1 cDNA-transfected HEK293 cells increased significantly after the cells were treated with 0.1% Triton X-100 or disrupted by mechanical forces (data not shown). These results suggest that human LALP1 is located inside the cell.
Second, we determined the cellular localization of LALP1 us-ing a LALP1-GFP fusion construct. The expression of LALP1-GFP fusion protein could be clearly observed as a punctuate distribution pattern under a confocal fluorescence microscope (Fig. 6A), indicating an intracellular vesicular compartmentalization as compared with a typical soluble GFP intracellular distribution (Fig. 6B). However, we did not observe any significant plasma membrane fluorescence, even under prolonged transfection conditions. Therefore, the LALP1-GFP fusion protein was associated with intracellular membrane compartments instead of the plasma membrane. This result further supports the notion that LALP1 is an endo-apyrase.
In conclusion, we have cloned the human LALP1 and mouse Lalp1 genes that share a high degree of sequence homology with mammalian endo-apyrases. Studies of the enzymatic activities and intracellular localization strongly suggest that these genes encode a novel member of the mammalian endoapyrase gene family. FIG. 5. Nucleotide phosphatase activity of human LALP1. Activity for each of the eight nucleotides was assayed with crude membrane preparations from cells transfected with the LALP1 cDNA construct (pEGFP/LALP1) (u) and the same control vector without a LALP1 insert (pEGFP) (f). The difference in activities between the LALP1 construct and control vector represents the increased activity of the transfected LALP1 expression vector. Activities for each of the eight nucleotides and each of the two cell lines were assayed in three independent experiments. The mean and S.D. of the three experiments are presented.
FIG. 6. Subcellular localization of human LALP1 using LALP1-GFP fusion protein. Full-length human LALP1 cDNA is fused with GFP at its C terminus and transfected into HEK293 cells. The protein localization was examined 18 h after transfection. A, LALP1-GFP; B, GFP.