A novel karyoskeletal protein: characterization of protein NO145, the major component of nucleolar cortical skeleton in Xenopus oocytes.

The nucleolus is a ubiquitous, mostly spheroidal nuclear structure of all protein-synthesizing cells, with a well-defined functional compartmentalization. Although a number of nonribosomal proteins involved in ribosome formation have been identified, the elements responsible for the shape and internal architecture of nucleoli are still largely unknown. Here, we report the molecular characterization of a novel protein, NO145, which is a major and specific component of a nucleolar cortical skeleton resistant to high salt buffers. The amino acid sequence of this polypeptide with a SDS-PAGE mobility corresponding to M(r) 145,000 has been deduced from a cDNA clone isolated from a Xenopus laevis ovary expression library and defines a polypeptide of 977 amino acids with a calculated mass of 111 kDa, with partial sequence homology to a synaptonemal complex protein, SCP2. Antibodies specific for this protein have allowed its recognition in immunoblots of karyoskeleton-containing fractions of oocytes from different Xenopus species and have revealed its presence in all stages of oogenesis, followed by a specific and rapid degradation during egg formation. Immunolocalization studies at the light and electron microscopic level have shown that protein NO145 is exclusively located in a cage-like cortical structure around the entire nucleolus, consisting of a meshwork of patches and filaments that dissociates upon reduction of divalent cations. We propose that protein NO145 contributes to the assembly of a karyoskeletal structure specific for the nucleolar cortex of the extrachromosomal nucleoli of Xenopus oocytes, and we discuss the possibility that a similar structure is present in other cells and species.

Soluble liver antigen: isolation of a 35-kd recombinant protein (SLA-p35) specifically recognizing sera from patients with autoimmune hepatitis.

Autoantibodies to soluble liver antigen (SLA) are considered a specific marker of autoimmune hepatitis. We have performed immunoscreening of a human liver gene expression library with an anti-SLA-positive serum. A reactive clone with a 35-kd open reading frame (ORF) and a 563 base pair (bp) 3' untranslated region (UTR) was isolated (soluble liver antigen [SLA]-p35), showing strong homology to an independently isolated putative SLA/liver-pancreas antigen (LP) sequence (Acc. No. AF146396), and a UGA serine tRNA-protein complex (tRNP)((Ser) Sec) related protein (AJ238617), as well as different expression sequence tag (EST)-clones from lymphatic and oncofetal tissues. Expressed in Escherichia coli, SLA-p35 showed dose-dependent and complete blocking of reactivity to native SLA antigen after preabsorption with the 35-kd recombinant protein. It recognized 67/85 (78.8%) precharacterized anti-SLA-positive sera in dilutions up to 1:40,000 in immunoblot, without detectable cross reactivity in the controls. The commercially available SLA/LP enzymelinked immunosorbent assay (ELISA), by comparison, recognized 63/85 samples (74.1%). Of the negative samples, 18% showed strong inhibition rates (80% and above) in the polyclonal inhibition ELISA. We conlude that the complementary DNA now isolated by 3 independent approaches encodes for the major but not sole antigenic component of soluble liver antigen. Although its truncated form presented here may serve to improve diagnostics based on the new recombinant polypeptide, it currently cannot fully replace the polyclonal inhibition ELISA.

Nuclear coactivator protein p100 is present in endoplasmic reticulum and lipid droplets of milk secreting cells.

We have identified the p100 protein, previously known as a novel cellular coactivator, as a constituent of endoplasmic reticulum and cytosolic lipid droplets from milk secreting cells. Cytosolic lipid droplets of terminally differentiated mammary epithelial cells are secreted as milk lipid globules. However, milk lipid globules did not have detectable amounts of p100 protein. The p100 protein was found also in cytosol from lactating mammary gland, in storage lipid droplets from mouse adipocytes, and in endoplasmic reticulum from liver. Immunofluorescence microscopy of mammary epithelial cells confirmed the presence of p100 in non-nuclear regions of these cells. Partial sequence analysis of tryptic peptides from p100 from cow mammary gland showed extensive homology with the reported sequence of p100 determined from a human cDNA. Antibodies against a peptide synthesized to duplicate a sequence in human p100 recognized a protein of the size of p100 in cow, mouse and rat cell fractions.

Meiotic lamin C2: the unique amino-terminal hexapeptide GNAEGR is essential for nuclear envelope association.

Meiotic lamin C2 is the only A-type lamin expressed during mammalian spermatogenesis. Typical for this short lamin is the unique hexapeptide GNAEGR, which substitutes the nonhelical amino terminus and part of the alpha-helical rod domain present in somatic lamins. Meiotic lamin C2 also lacks a carboxyl-terminal CaaX box, which is modified by isoprenylation and involved in nuclear envelope (NE) association of somatic isoforms. The mechanism by which lamin C2 becomes localized in the NE is totally unknown. Here we demonstrate that the hexapeptide GNAEGR is essential for this process: (i) Its deletion resulted in a diffuse distribution of lamin C2 within nuclei of transfected COS-7 cells; (ii) Mutated somatic lamin C, containing the sequence GNAEGR at its amino terminus, was located at the NE. The mass spectrometric analysis of the amino terminus of lamin C2 revealed that it is modified by myristoylation. Correspondingly, the substitution of the first glycine residue abolishes the NE association of lamin C2. We conclude that NE association of lamin C2 is achieved by a mechanism different from that of somatic lamins.

Molecular characterization reveals identity of microtubule-associated proteins MAP3 and MAP4.

The identity of two microtubule-associated proteins, MAP3 and MAP4, was verified both immunologically and biochemically. MAP3 was enriched from the heat-stable fraction of rat brain extracts by reverse-phase HPLC and preparative two-dimensional gel electrophoresis. Both MAP3 and MAP4 antibodies reacted with the corresponding spots on two-dimensional Western blots. Amino acid sequences of internal peptides derived from rat MAP3 matched with corresponding sequence stretches of mouse MAP4. In the kidney cortex, the MAP3 antibody stained not only glomerular podocytes but also interstitial cells. This distribution pattern of MAP3 is identical to that of MAP4 reported previously. These results indicate that MAP3 and MAP4 are identical.

Casein kinase 2 binds and phosphorylates the nucleosome assembly protein-1 (NAP1) in Drosophila melanogaster.

The nucleosome assembly protein-1 (NAP1) was originally identified in HeLa cells as a factor facilitating the in vitro assembly of nucleosomes. However, in yeast cells NAP1 is required in the control of mitotic events induced by the Clb2/p34(CDC28). Here, we show that Drosophila NAP1 is a phosphoprotein that is associated with a kinase able to phosphorylate NAP1. By using an in-gel kinase assay we found that this kinase displays a molecular mass of 38 kDa. Following purification and peptide microsequencing, we identified the kinase phosphorylating NAP1 as the alpha subunit of casein kinase 2 (CK2). With the help of a series of NAP1 segments and synthetic peptides, we assigned the CK2 phosphorylation sites to residues Ser118, Thr120, and Ser284. Interestingly, Ser118 and Thr120 are located within a PEST domain, while Ser284 is adjacent to the nuclear localization signal. Substitution of the identified phosphoresidues by alanine was found to reduce considerably the ability of CK2 to phosphorylate NAP1. The enhanced ability of CK2 to phosphorylate phosphatase-treated NAP1 extracted from Drosophila embryos and the similar tryptic phospho-peptide pattern of in vivo labelled NAP1 and in vitro labelled NAP1 with CK2 indicate that NAP1 is a natural substrate of CK2. Further analysis revealed that both CK2alpha and beta subunits are associated with NAP1 but we found that only the catalytic alpha subunit establishes direct contact with NAP1 on two distinct domains of this protein. The location of CK2 phosphorylation sites in NAP1 suggests that their phosphorylation can contribute to a PEST-mediated protein degradation of NAP1 and the translocation of NAP1 between cytoplasm and nucleus.

Aspartyl proteases in Caenorhabditis elegans. Isolation, identification and characterization by a combined use of affinity chromatography, two-dimensional gel electrophoresis, microsequencing and databank analysis.

Crude homogenates of the nematode Caenorhabditis elegans exhibit maximal proteolytic activity under acidic pH conditions. About 90% of this activity is inhibited by the oligopeptide pepstatin, which specifically inhibits the activity of aspartyl proteases such as pepsin, cathepsins D and E or renin. We have purified enzymes responsible for this proteolytic activity by a single-step affinity chromatography on pepstatin-agarose. Analysis of the purified fraction by 1D SDS gel electrophoresis revealed six bands ranging from 35 to 52 kDa. After electrotransfer to poly(vinylidene difluoride) membranes, all bands were successfully subjected to N-terminal microsequencing. On 2D gels, the purified protein bands split into 19 spots which, after renewed microsequencing, were identified as isoelectric variants of the six proteins already described. The N-termini obtained for these proteins could be correlated to genomic DNA sequences determined in the course of the C. elegans genome sequencing project. All these sequences were predicted to code for expressed proteins as collected in the WORMPEP database. Five of the six coding sequences identified in this study were found to contain the typical active-site consensus sequence of aspartyl proteases and displayed an overall amino acid identity between 25 and 66% as compared to aspartyl proteases from other organisms. In addition to the five aspartyl proteases detected at the protein level, we have identified the coding sequences for seven other enzymes of this protease family by a similarity search in the genomic DNA of C. elegans which has recently been completely sequenced.

Molecular characterization of proteolytic processing of the Gag proteins of human spumavirus.

Spumaviruses, or foamy viruses, express Gag proteins that are incompletely processed by the viral protease in cell cultures. To delineate the proteolytic cleavage sites between potential Gag subdomains, recombinant human spumaretrovirus (HSRV) Gag proteins of different lengths were expressed, purified by affinity chromatography, and subjected to HSRV protease assays. HSRV-specific proteolytic cleavage products were isolated and characterized by Western blotting. Peptides spanning potential cleavage sites, as deduced from the sizes of the proteolytic cleavage products, were chemically synthesized and assayed with HSRV protease. The cleaved peptides were then subjected to mass spectrometry. In control experiments, HSRV protease-deficient mutant proteins were used to rule out unspecific processing by nonviral proteases. The cleavage site junctions identified and the calculated sizes of the cleavage products were in agreement with those of the authentic cleavage products of the HSRV Gag proteins detectable in viral proteins from purified HSRV particles and in virus-infected cells. The biological significance of the data was confirmed by mutational analysis of the cleavage sites in a recombinant Gag protein and in the context of the infectious HSRV DNA provirus.

Analysis of intracytoplasmic hyaline bodies in a hepatocellular carcinoma. Demonstration of p62 as major constituent.

Intracytoplasmic hyaline bodies (IHBs) resemble inclusions in hepatocellular carcinoma cells, which so far have escaped further characterization. A relationship to Mallory bodies was suggested on the basis of light microscopy and filamentous ultrastructure. A hepatocellular carcinoma containing numerous IHBs was studied. Our studies revealed immunoreactivity of IHBs with the monoclonal antibodies SMI 31 and MPM-2, which recognize hyperphosphorylated epitopes present on paired helical filaments in Alzheimer's disease brains (SMI 31) or on diverse proteins hyperphosphorylated by mitotic kinases in the M-phase of the cell cycle (MPM-2). One- and two-dimensional gel electrophoresis of tumor extracts followed by immunoblotting with SMI 31 and MPM-2 antibodies revealed a major immunoreactive protein with an apparent molecular weight between 62 and 65 kd, which was resolved into several highly acidic (pH 4.5) protein components in two-dimensional gels. This protein was undetectable in non-neoplastic liver tissue. Sequence analysis identified the SMI 31 and MPM-2 immunoreactive material as p62, indicating that p62 is a major constituent of IHBs. p62 is an only recently discovered protein that is a phosphotyrosine-independent ligand of the SH2 domain of p56(lck), a member of the c-src family of cytoplasmic kinases. Moreover, p62 binds ubiquitin and may act as an adapter linking ubiquitinated species to other proteins. These features suggest a role of p62 in signal transduction and possibly also carcinogenesis. IHBs observed in the hepatocellular carcinoma cells presented are the first indications of a role of p62 in disease.

Molecular characterization of a protease secreted by Erwinia amylovora.

A protease with a molecular mass of 48 kDa is secreted by the fire blight pathogen Erwinia amylovora in minimal medium. We characterized this activity as a metalloprotease, since the enzyme was inhibited by EDTA and o -phenanthroline. A gene cluster was determined to encode four genes connected to protease expression, including a structural gene (prtA) and three genes (prtD, prtE, prtF) for secretion of the protease, which are transcribed in the same direction. The organization of the protease gene cluster in E. amylovora is different from that in other Gram-negative bacteria, such as Erwinia chrysanthemi, Pseudomonas aeruginosa and Serratia marcescens. On the basis of the conservative motif of metalloproteases, PrtA was identified to be a member of the metzincin subfamily of zinc-binding metalloproteases, and was confirmed to be the 48 kDa protease on gels by sequencing of tryptic peptide fragments derived from the protein. The protease is apparently secreted into the external medium through the type I secretion pathway via PrtD, PrtE and PrtF which share more than 90% identity with the secretion apparatus for lipase of S. marcescens. A protease mutant was created by Tn 5 -insertions, and the mutation localized in the prtD gene. The lack of protease reduced colonization of an E. amylovora secretion mutant labelled with the gene for the green fluorescent protein (gfp) in the parenchyma of apple leaves.

Ganglioside GM2-activator protein and vesicular transport in collecting duct intercalated cells.

This study describes the molecular characterization of an antigen defined by an autoantibody from a woman with habitual abortion as GM2-activator protein. The patient showed no disorder of renal function. Accidentally with routine serum screening for autoantibodies, an immunoreactivity was found in kidney collecting duct intercalated cells. Three distinct patterns of immunostaining of intercalated cells were observed: staining of the apical pole, basolateral pole, and diffuse cytoplasmic labeling. Ultrastructurally, the immunoreactivity was associated with "studs," which represent the cytoplasmic domain of the vacuolar proton pump in intercalated cells. This pump is subjected to a shuttling mechanism from cytoplasmic stores to the cell membrane, which exclusively occurs in intercalated cells. Peptide sequences of a 23-kD protein purified from rat kidney cortex showed complete identity with corresponding sequences of GM2-activator protein. In the brain, GM2-activator protein is required for hexosaminidase A to split a sugar from ganglioside GM2. Because neither ganglioside GM2 nor GM1 (its precursor) is present in significant amounts in the kidney, the previous finding that this tissue contains the highest level of activator protein in the body was confusing. In this study, a novel role for GM2-activator protein in intercalated cells is proposed, and possible roles in the shuttling mechanism are discussed.

Identification of p62, a phosphotyrosine independent ligand of p56lck kinase, as a major component of intracytoplasmic hyaline bodies in hepatocellular carcinoma.

Intracytoplasmic hyaline bodies (IHBs) resemble a peculiar type of cytoplasmic inclusions in cells of hepatocellular carcinoma (HCC) which so far have escaped further characterization. In order to determine protein composition of IHBs we investigated tissue of a HCC containing numerous IHBs by immunohistochemistry and Western blot analysis using a large panel of different antibodies. Our studies revealed immunoreactivity of IHBs with the monoclonal antibodies SMI 31 and MPM-2 which recognize hyperphosphorylated epitopes present on paired helical filaments in Alzheimer's disease brains (SMI 31) and on proteins hyperphosphorylated by mitotic kinases (MPM-2), respectively. In two-dimensional Western blots of HCC extracts SMI 31 and MPM-2 antibodies detected a 62 to 65 kD protein with an isoelectric point around 4.5. Microsequencing identified this protein as p62, a recently identified phosphotyrosine-independent ligand of the SH2 domain of tyrosine kinase p56lck. Immunoreactivity of p62 protein spots with antibodies to phosphorylated epitopes (i.e. SMI 31 and MPM-2) suggest that p62 is highly phosphorylated in IHBs. This is the first report on accumulation of p62 as cellular inclusions and its association with human disease.

Adipophilin is a specific marker of lipid accumulation in diverse cell types and diseases.

We report the human DNA and protein sequence of adipophilin and its association with the surface of lipid droplets. The amino acid sequence of human adipophilin has been determined by using cDNA clones from several tissues and confirmed by the reverse transcription/polymerase chain reaction method and Edman sequencing. The open reading frame of adipophilin encodes a polypeptide with a calculated molecular weight of 48.1 kDa and an isoelectric point of 6.72. By immunofluorescence and electron-microscopic localization with newly raised specific poly- and monoclonal antibodies, we show that this protein is not restricted to adipocytes as previously indicated by studies of the mouse homologous protein, adipose-differentiation-related protein. Adipophilin occurs in a wide range of cultured cell lines, including fibroblasts and endothelial and epithelial cells. In tissues, however, expression of adipophilin is restricted to certain cell types, such as lactating mammary epithelial cells, adrenal cortex cells, Sertoli and Leydig cells of the male reproductive system, and steatosis or fatty change hepatocytes in alcoholic liver cirrhosis. Our results reveal adipophilin as a possible new marker for the identification of specialized differentiated cells containing lipid droplets and for diseases associated with fat-accumulating cells.

Molecular characterization of proteolytic processing of the Pol proteins of human foamy virus reveals novel features of the viral protease.

Spumaviruses, or foamy viruses, express a pol-specific transcript that codes for a Pol polyprotein that consists of the protease, reverse transcriptase, ribonuclease H, and the integrase domains. To delineate the proteolytic cleavage sites between the Pol subdomains, recombinant human foamy virus (HFV) Pol proteins were expressed, purified by affinity chromatography, and subjected to either HFV protease assays or autocatalytic processing. In control experiments, HFV protease-deficient mutant proteins in which the active site Asp was replaced by an Ala residue were used to rule out unspecific processing by nonviral proteases. Specific proteolytic cleavage products were isolated, and the cleavage sites were analyzed by amino acid sequencing. Peptides spanning the resulting cleavage sites were chemically synthesized and assayed with HFV protease, and the cleaved peptides were subjected to mass spectrometry. The cleavage site sequences obtained were in complete agreement with the amino-terminal sequences from amino acid sequencing of authentic cleavage products of the HFV Pol proteins. Analysis by fast-protein liquid chromatography of a short version of the active HFV protease revealed that the enzyme predominantly formed dimeric molecules.

Synaptopodin: an actin-associated protein in telencephalic dendrites and renal podocytes.

Synaptopodin is an actin-associated protein of differentiated podocytes that also occurs as part of the actin cytoskeleton of postsynaptic densities (PSD) and associated dendritic spines in a subpopulation of exclusively telencephalic synapses. Amino acid sequences determined in purified rat kidney and forebrain synaptopodin and derived from human and mouse brain cDNA clones show no significant homology to any known protein. In particular, synaptopodin does not contain functional domains found in receptor-clustering PSD proteins. The open reading frame of synaptopodin encodes a polypeptide with a calculated Mr of 73.7 kD (human)/74.0 kD (mouse) and an isoelectric point of 9.38 (human)/9. 27 (mouse). Synaptopodin contains a high amount of proline ( approximately 20%) equally distributed along the protein, thus virtually excluding the formation of any globular domain. Sequence comparison between human and mouse synaptopodin revealed 84% identity at the protein level. In both brain and kidney, in vivo and in vitro, synaptopodin gene expression is differentiation dependent. During postnatal maturation of rat brain, synaptopodin is first detected by Western blot analysis at day 15 and reaches maximum expression in the adult animal. The exclusive synaptopodin synthesis in the telencephalon has been confirmed by in situ hybridization, where synaptopodin mRNA is only found in perikarya of the olfactory bulb, cerebral cortex, striatum, and hippocampus, i.e., the expression is restricted to areas of high synaptic plasticity. From these results and experiments with cultured cells we conclude that synaptopodin represents a novel kind of proline-rich, actin-associated protein that may play a role in modulating actin-based shape and motility of dendritic spines and podocyte foot processes.

Cytosolic factors mediate protein insertion into the peroxisomal membrane.

Following in vitro translation of the 22 kDa peroxisomal membrane protein (Pmp22p), gel filtration analysis of the post-ribosomal supernatant revealed that Pmp22p forms two complexes. Complex I is of high molecular weight, results in a crosslinking product of 80 kDa, and by co-immunoprecipitation with anti-TCP1 antibody was identified as TRiC. In complex II Pmp22p was crosslinked to a yet unknown polypeptide of 40 kDa (P40). This complex exhibited much higher efficiency to insert Pmp22p into the peroxisomal membrane compared to complex I. In a model we suggest that newly synthesized Pmp22p is first bound to TRiC before being transferred to P40 which may function as a cytosolic Pmp22p receptor.

Molecular characterization and expression pattern of XY body-associated protein XY40 of the rat.

The XY body is a structure formed by the partially synapsed chromosomes X and Y that is located at the nuclear periphery of mammalian pachytene spermatocytes. In contrast to the autosomal bivalents of the same nucleus, the XY body is characterized by its differential chromatin condensation and transcriptional inactivity. In order to shed some light on the biological significance of these differences we have been characterizing XY body-associated proteins. We present here the cDNA sequence and expression pattern of XY40, a protein that is associated with the axial elements of the XY body. RNA blot analysis revealed that during spermatogenesis the transcript that encodes protein XY40 was highly enriched in pachytene spermatocytes. This transcript was also detectable in brain and, to a lesser extent, in liver and kidney. Although the signal in brain was as strong as in spermatocytes, protein XY40 could be detected only in the latter. The nucleic acid sequence reveals that XY40 is a novel protein with a few similarities to already known nucleotide sequences. Among these similarities the most interesting is a box that is shared by the 3' untranslated region of XY40 and the 5' untranslated region of Munc-18c, a member of a protein family involved in synaptic vesicle exocytosis. Since the transcripts of both XY40 and Munc18-c show a similar expression pattern, it is tempting to speculate that this common sequence is involved in translation regulation.

Two-dimensional gel electrophoresis of Caenorhabditis elegans homogenates and identification of protein spots by microsequencing.

Employing isoelectric focusing on immobilized pH gradients followed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) we have obtained a map of C. elegans proteins, from a mixed culture containing all developmental stages, presenting over 2000 spots within the window of isoelectric points (pI) 3.5-9 and a molecular mass of 10-200 kDa. Edman microsequencing yielded successful results in 12 out of 24 analyzed spots. All but one of the N-terminal sequences retrieved C. elegans sequences in cosmid and/or expressed sequence tag clones. Structurally related protein sequences found in data banks included enzymes in energy metabolism (cytochrome oxydase, ATP synthase, enolase), a fatty acid-binding protein, a translationally controlled tumor protein, an unknown C. elegans protein, an acidic ribosomal protein, a titin-like protein, a G-protein beta chain, cyclophilin, and cathepsin D. Experimental determination of N-termini allowed us to define sites of signal cleavage providing further information on the physiological role of the newly found C. elegans proteins. This report demonstrates the possibility of two-dimensional gel electrophoresis and Edman microsequencing in the elucidation of C. elegans proteome.

Ether lipid biosynthesis: isolation and molecular characterization of human dihydroxyacetonephosphate acyltransferase.

In this paper we describe isolation and molecular characterization of human dihydroxyacetonephosphate acyltransferase (DAP-AT). The enzyme was extracted from rabbit Harderian gland peroxisomes and isolated as a trimeric complex by sucrose density gradient centrifugation. From peptide sequences matching EST-clones were obtained which allowed cloning and sequencing of the cDNA from a human cDNA library. The nucleotide-derived amino acid sequence revealed a protein consisting of 680 amino acid residues of molecular mass 77187 containing a C-terminal type 1 peroxisomal targeting signal. Monospecific antibodies raised against this polypeptide efficiently immunoprecipitated DAP-AT activity from solubilized peroxisomal preparations, thus demonstrating that the cloned cDNA codes for DAP-AT.

Adipocyte differentiation-related protein is secreted into milk as a constituent of milk lipid globule membrane.

Milk lipid globules from humans, cows and rats contained a protein identified as adipocyte differentiation-related protein (ADRP) associated with the globule surface membrane material. This protein, previously believed to be specific to adipocytes, was a major constituent of the globule surface and was present in a detergent-insoluble complex that contained stoichiometric amounts of butyrophilin and xanthine oxidase. Identification of ADRP was by sequence similarity of tryptic peptides from cow and human proteins with the sequence inferred from the cDNA for mouse ADRP. The putative ADRP of lipid globules from cow, human and rat milk was recognized specifically by antisera raised against a peptide synthesized to duplicate the N-terminal 26 residues of the mouse protein. In homogenates of lactating mammary gland, ADRP was found only in endoplasmic reticulum and in lipid droplet fractions. ADRP was modified, apparently post-translationally, and one modification apparently was acylation, primarily with C14, C16 and C18 fatty acids. Two isoelectric variants of ADRP were present in cow globule membrane material. In vitro, ADRP served as a substrate for protein kinases associated with milk lipid globule membrane, but this protein did not seem to become phosphorylated intracellularly.