phosphodiesterases in rat liver

Biochimica et Biophysica Acta 1450 (1999) 45^52

Di¡erential regulation of the expression of nucleotide pyrophosphatases/ phosphodiesterases in rat liver Cristiana Stefan, Rik Gijsbers, Willy Stalmans, Mathieu Bollen * Afdeling Biochemie, Faculteit Geneeskunde, Katholieke Universiteit Leuven, Campus Gasthuisberg, Herestraat 49, B-3000 Leuven, Belgium Received 8 January 1999; accepted 10 March 1999

Abstract We propose the name nucleotide pyrophosphatases/phosphodiesterases (NPP) for the enzymes that release nucleoside-5Pmonophosphates from various pyrophosphate and phosphodiester bonds. Three structurally related mammalian NPPs are known, i.e. NPPK (autotaxin), NPPL (B10/gp130RB13ÿ6 ) and NPPQ (PC-1). We report here that these isozymes have a distinct tissue distribution in the rat but that they are all three expressed in hepatocytes. In FAO rat hepatoma cells only the level of NPPQ was stimulated by TGF-L1 . In rat liver, the concentration of the transcripts of all three isozymes was found to increase manyfold during the first weeks after birth, but the increased expression of the NPPK mRNA was transient. The level of the NPP transcripts transiently decreased after hepatectomy, but NPPK mRNA was also lost after sham operation, which suggests that it may belong to the negative acute-phase proteins. The loss of the L- and Q-transcripts after hepatectomy was not due to a decreased NPP gene transcription or an increased turnover of the mature transcripts. However, hepatectomy also caused a similar loss of the nuclear pool of the NPPL and NPPQ mRNAs. We conclude that a deficient processing and/or an increased turnover of the NPP pre-mRNAs underlies the hepatectomy-induced decrease of the L- and Q-transcripts. A similar loss of nuclear NPPQ mRNA was also noted after treatment with cycloheximide, indicating that protein(s) with a high turnover control the stability and/or processing of the immature NPPQ transcript. ß 1999 Elsevier Science B.V. All rights reserved. Keywords: Nucleotide pyrophosphatase; Phosphodiesterase I; PC-1; Autotaxin; B10

1. Introduction Type I phosphodiesterases (EC 3.1.4.1) were originally described as enzymes that generate nucleoside5P-monophosphates from a variety of oligonucleotides and nucleotide derivatives, as opposed to the type II phosphodiesterases, which release nucleoside-3P-monophosphates [1]. However, the latter enAbbreviations: NPP, nucleotide pyrophosphatase/phosphodiesterase * Corresponding author. Fax: +32 (16) 345995; E-mail: [email protected]

zymes are now classi¢ed as 3P-exonucleases (EC 3.1.16) and, instead, the name `phosphodiesterase II' is now often used to denote a structurally unrelated subgroup of cyclic nucleotide phosphodiesterases [2]. Moreover, it has been established that type I phosphodiesterases can also hydrolyse phosphoanhydride bonds and are identical to the enzymes that have been classi¢ed as nucleotide pyrophosphatases (EC 3.6.1.9) [3]. To avoid a double and confusing nomenclature, we propose to name these enzymes henceforward nucleotide pyrophosphatases/phosphodiesterases (NPP). They release nucleoside-5P-monophosphates from various pyrophosphate bonds (in

0167-4889 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 8 9 ( 9 9 ) 0 0 0 3 1 - 2

BBAMCR 14464 21-4-99

46

C. Stefan et al. / Biochimica et Biophysica Acta 1450 (1999) 45^52

e.g. nucleoside diphosphates and triphosphates, NAD, FAD, UDP-glucose) and phosphodiester bonds (in oligonucleotides and arti¢cial substrates like the p-nitrophenyl ester of TMP). Three structurally related mammalian NPPs are known that are encoded by di¡erent genes. These isozymes have all been discovered in a di¡erent context and for none of them their function as an NPP was clear from the beginning, which explains their inconsistent and illogical nomenclature (Table 1). NPPK or autotaxin was discovered as a protein that is secreted by melanoma cells and stimulates cell motility [4]. A splice variant of NPPK cloned from a rat brain library was named phosphodiesterase IK [5]. NPPL was initially described as gp130RB13ÿ6 , i.e. a plasma membrane glycoprotein of 130 kDa that is recognized by the monoclonal antibodies RB13-6, which visualize a subpopulation of neural precursor cells that are sensitive to malignant conversion by N-ethyl-N-nitrosourea [6]. More recently, it has been reported that the latter protein is identical to B10, a membrane protein on the apical surface of hepatocytes [7]. The same isozyme was also independently cloned from a prostate library and named phosphodiesterase IL [8]. The best studied NPP is the one we now term the Q-isozyme, which was initially discovered as a surface marker of antibody-secreting B-lymphocytes, hence the name plasma-cell di¡erentiation antigen-1 (PC-1) [9]. It is now known, however, that NPPQ is expressed in a large variety of tissues, including the liver, where it is present on the basolateral surface of the hepatocytes [7]. NPPQ has been implicated in the control of various processes such as extracellular nucleotide metabolism [10], protein glycosylation [11] and insulin signalling [12], although conclusive evidence for these functions is lacking. Table 1 Nomenclature of mammalian nucleotide pyrophosphatases/ phosphodiesterases Proposed name

Other names

Refs.

NPPK

Autotaxin Phosphodiesterase IK gp130RB13ÿ6 B10 Phosphodiesterase IL PC-1

[4] [5] [6] [7] [8] [9]

NPPL

NPPQ

NPPs are transmembrane proteins, but for all three isozymes there also appear to exist soluble and secreted forms that are generated by proteolysis [9,13]. The NPPs in the plasma membrane consist of a short N-terminal intracellular domain, a single transmembrane domain and a large extracellular C-terminal domain that contains the catalytic NPP site. A fraction of NPPQ is present in the membranes of the endoplasmic reticulum, where its C-terminal domain faces the lumen of the endoplasmic reticulum [3,14]. The activity and expression of NPPs are tightly regulated. NPPK [15] and NPPQ [10,16] are inactivated by autocatalytic phosphorylation at submicromolar concentrations of ATP. This inactivation results from phosphorylation of the same active site Thr that also forms a covalent, nucleotidylated intermediate during the pyrophosphatase/phosphodiesterase reaction. Re-activation of NPPQ results from nucleotide-stimulated autodephosphorylation [10,16]. This autoregulatory mechanism of NPPK and NPPQ may prevent the hydrolysis of extracellular nucleotides when they become scarce. The expression of NPPs is tightly controlled by growth factors and hormones. NPPK is downregulated by interferon-Q in ¢broblasts [17]. In smooth muscle cells, the NPPL transcript level is decreased by PDGF and angiotensin II [18]. The expression of NPPQ in osteosarcoma cells and chondrocytes, but not in ¢broblasts, was reported to be stimulated by TGF-L1 and, in chondrocytes, this e¡ect was antagonized by interleukin-1L [19,20]. NPPQ in osteosarcoma cells and T-lymphocytes is induced by agents that activate protein kinase A or C [21,22]. We have previously reported that the expression of NPPQ in rat liver is growth-related, i.e. the level of NPPQ increases dramatically during the ¢rst weeks after birth and is transiently decreased following a 70% hepatectomy [14]. It has also been reported that the level of NPPL in rat serum increases substantially following bile duct ligation and during the development of cholangiocarcinoma, probably as a result of proteolytic cleavage of the membrane-associated form [13]. We show here that all three NPP isozymes are expressed in rat hepatocytes and that their expression in the liver is di¡erentially regulated. In addition, we report that the transient loss of the NPPL and NPPQ mRNAs during liver regeneration is due to a de-

BBAMCR 14464 21-4-99

C. Stefan et al. / Biochimica et Biophysica Acta 1450 (1999) 45^52

47

creased processing and/or an increased lability of their precursors. 2. Materials and methods Full-length NPPK cDNA from a human breast adenocarcinoma cell line (MDA-MB-435) and NPPL cDNA from rat brain were gifts from Dr. M. Stracke (Bethesda, MD, USA) and Dr. H. Deissler (Essen, Germany), respectively. Hepatocytes were isolated from normally fed rats as described by Bollen et al. [23]. A 70% hepatectomy, involving the removal of the large median and left lateral liver lobes, was performed under ether anaesthesia on Wistar rats weighing approx. 200 g [14]. Sham surgery consisted of the same manipulations except that the liver was left intact. Total RNA was isolated from freeze-clamped livers by CsCl centrifugation [24]. For the experiment illustrated in Fig. 6, the livers were ¢rst fractionated by the citric acid method of Miller [25], with slight modi¢cations. Brie£y, 2 g of liver was homogenized in 20 ml of a bu¡er containing 75 mM citric acid at pH 2.3, 0.25 M sucrose and 0.5% (v/v) Triton X-100. The homogenate was layered on top of two cushions, the ¢rst one containing 10 ml of 25 mM citric acid at pH 2.3, 1.8 M sucrose and 0.1% Triton X-100, and the second one containing 5 ml of 75 mM citric acid at pH 2.3, 1 M sucrose and 0.5% Triton X-100. Following centrifugation for 75 min at 25 000Ug, the nuclear pellet was resuspended in 50 mM citric acid at pH 2.3, 0.25 M sucrose, 40% (w/v) glycerol, 0.5% Triton X-100 and washed three times in this bu¡er. RNAs from the homogenates and the nuclei were prepared by the CsCl method [24]. All other materials and experimental procedures were as described previously [14]. 3. Results and discussion 3.1. Tissue distribution of NPPs Northern blot analysis of various rat tissues revealed a quite di¡erent distribution of the three NPP isozymes (Fig. 1). A single NPPK transcript of 3.5 kb was detected, which was predominantly ex-

Fig. 1. Northern blot analysis in rat tissues. Each lane contains 30 Wg of total RNA isolated from the indicated tissues. The blot was consecutively hybridized with full-length cDNAs encoding NPPK, NPPL, NPPQ and 18S rRNA, the latter to ascertain equal loading. In between the blot was dehybridized each time in 0.5% SDS at 95³C for 15 min. Dehybridization was veri¢ed by autoradiography.

pressed in brain, lung, duodenum and adrenals. However, the NPPK mRNA was also clearly present in rat liver. Our results agree with those of Narita et al. [5], who found a single transcript in various rat brain regions, but are at variance with those of Fuss et al. [26], who detected several NPPK transcripts in rat tissues, including a much smaller `liver-speci¢c' species. We found that the NPPL transcript (3.0 kb) was scarce in brain and adrenals, but abundant in liver, heart, kidney and duodenum (Fig. 1). The NPPQ-cDNA hybridized in rat tissues with two transcripts (4.1 and 3.2 kb) that were mainly present in liver, heart and kidney. Remarkably, the testis expressed L- and Q-transcripts that were considerably smaller than those in other tissues, which indicates the existence of testis-speci¢c splice variants. Northern analysis of RNA from either total liver or freshly isolated hepatocytes showed that the mRNAs encoding all three NPPs are present in hepatocytes at a similar relative abundance as in intact liver (Fig. 2). However, in FAO rat hepatoma cells only NPPL and NPPQ were expressed.

BBAMCR 14464 21-4-99

48

C. Stefan et al. / Biochimica et Biophysica Acta 1450 (1999) 45^52

Fig. 2. The expression of NPPs in hepatocytes and FAO cells. (A) Northern blots with total RNA (30 Wg/lane) from rat liver, freshly isolated hepatocytes or cultured FAO rat hepatoma cells. Also shown is the e¡ect of an incubation of FAO cells with 3 ng/ml TGF-L1 for 24 h on the NPP mRNA level. General procedure as described in the legend to Fig. 1. (B) E¡ect of an incubation of FAO cells for up to 48 h, in the absence or presence of TGF-L1 , on the level of NPPQ, as revealed by Western analysis with NPPQ-speci¢c antibodies.

been shown to increase the level of NPPQ in T-lymphocytes and osteosarcoma cells through activation of protein kinases A and C, respectively [21,22], did not a¡ect the expression of NPPQ in FAO hepatoma cells within 24 h (not shown). The expression of the NPP mRNAs in rat liver was also found to be age-related (Fig. 3). All isozyme mRNAs were present at birth but their levels increased severalfold during the ensuing weeks. The NPPK transcript level increased only transiently during the ¢rst weeks after birth and returned to neonatal levels after 5^6 weeks. The concentration of the NPPL transcript increased quite abruptly between 2 and 3 weeks of age and remained stable thereafter. The level of the NPPQ transcript was very low at birth, but increased gradually between birth and 3 weeks of age, as reported previously [14]. We have also used liver regeneration after 70% hepatectomy as a model to study the growth-related expression of NPPs (Fig. 4). The level of all NPP transcripts decreased manyfold after partial hepatectomy, similarly to what we demonstrated previously for NPPQ [14]. The minimal NPP mRNA levels were

3.2. Regulation of the expression of NPPs As has already been shown for NPPQ in bone and cartilage cells [19,20], we found that the expression of the Q-isozyme in FAO cells was stimulated by TGFL1 (Fig. 2B). In contrast, the expression of the Kand L-isozymes was not a¡ected by this growth factor. The addition of TGF-L1 , i.e. 3 ng/ml (0.1 nM) for 24 h increased the level of NPPQ mRNA (Fig. 2A) and protein (Fig. 2B) by 2- and 4-fold, respectively, as compared to the corresponding conditions without the growth factor. It is also shown in Fig. 2B that a prolonged incubation per se already resulted in an increased level of NPPQ protein, in agreement with the report that the concentration of NPPQ in FAO cells increases with the cell density [14]. cAMP and the phorbol ester PMA, which have

Fig. 3. The expression of NPP isozymes in the liver of neonatal and young rats. The ¢gure shows Northern blots with total RNA (30 Wg/lane) that was isolated from the liver of rats of the indicated age. All blots were prepared with RNA from the same series of livers.

BBAMCR 14464 21-4-99

C. Stefan et al. / Biochimica et Biophysica Acta 1450 (1999) 45^52

49

Fig. 4. The e¡ects of sham operation or 70% hepatectomy on the NPP transcript levels. Rats were sacri¢ced at the indicated times after sham surgery or partial hepatectomy and the hepatic NPP mRNA concentrations were estimated by Northern blot analysis (30 Wg total RNA/lane).

reached 16^20 h after hepatectomy. At 20 h the levels of the K-, L- and Q-transcripts were decreased by 90 þ 8%, 73 þ 11% and 74 þ 12% (n = 3), respectively. The NPP transcript concentrations gradually normalized during the ensuing 2 weeks, when liver re-

generation became nearly complete. Sham operation did not signi¢cantly a¡ect the level of the NPPQ and NPPL transcripts but, unexpectedly, caused an equally important decrease in the level of the NPPK transcript as did hepatectomy. This shows

Fig. 5. E¡ects of inhibitors of protein synthesis on the level of NPP mRNAs. Rats were injected for the indicated times with actinomycin D (2 mg/kg body weight; i.p.) plus K-amanitin (0.5 mg/kg body weight; i.v.) or with cycloheximide (50 mg/kg body weight; i.p.). Northern blot analysis with the indicated probes was done on total RNA (30 Wg/lane). Each blot was prepared with RNA from the same series of animals.

BBAMCR 14464 21-4-99

50

C. Stefan et al. / Biochimica et Biophysica Acta 1450 (1999) 45^52

that the loss of NPPK was not associated with the regeneration process itself but is most likely caused by the acute-phase response, which represents the early set of reactions to in£ammatory or stressful stimuli, such as surgery [27^30]. Therefore, NPPK may belong to the negative acute-phase proteins, i.e. a group of intrahepatic or plasma proteins that are downregulated during the acute-phase response. This is in accordance with the report that NPPK in ¢broblasts is downregulated by interferon-Q, which has anti-in£ammatory properties [17]. In further agreement with this view, it should be noted that the time course of downregulation of the NPPK transcript after sham operation (Fig. 4) is similar to that of other negative acute-phase proteins [27,28]. Remarkably, the restoration of the NPPK transcript level in sham-operated animals was reproducibly much slower than in the hepatectomized rats and remained incomplete after 2 weeks. A similar dissociation of the curves describing mRNA levels in acute-phase and regenerating liver has been observed for other proteins [28]. We have previously demonstrated that the steadystate level of the NPPQ transcripts decreases rapidly

following the in vivo administration of the translational inhibitor cycloheximide, but is not a¡ected within 8 h by the injection of the transcriptional inhibitors actinomycin D plus K-amanitin [14]. In Fig. 5 it is shown that this also applies to NPPK. Indeed, both the NPPK and NPPQ transcripts gradually disappeared after the administration of cycloheximide, with an estimated half-life of 1^3 h. On the other hand, the NPPL transcript was not a¡ected within 8^9 h by the inhibitors of protein synthesis. Collectively, these data indicate that the half-life of the NPP transcripts is rather long ( s 10 h), but that the maintenance of the NPPK and NPPQ transcripts depends on protein(s) with a high turnover. 3.3. Cycloheximide and hepatectomy also a¡ect nuclear NPP mRNA Partial hepatectomy did not a¡ect the transcription rates of the NPPL (not illustrated) and NPPQ genes [14], as revealed by nuclear run-on assays. Moreover, the level of the remaining NPPL and NPPQ transcripts at 16 h after partial hepatectomy was not further decreased by the administration of

Fig. 6. The e¡ects of partial hepatectomy or cycloheximide on the total and nuclear levels of the NPPL and NPPQ transcripts. Total RNA (30 Wg/lane) was prepared both from liver homogenates and from nuclei that were isolated from the same homogenates, as detailed in Section 2. The RNA was obtained from control rats and from animals that either had undergone a partial hepatectomy 20 h earlier or had been intraperitoneally injected with cycloheximide (50 mg/kg body weight) 6 h previously.

BBAMCR 14464 21-4-99

C. Stefan et al. / Biochimica et Biophysica Acta 1450 (1999) 45^52

actinomycin D plus K-amanitin for up to 8 h (not shown), indicating that the e¡ect of hepatectomy cannot be explained by an increased turnover of these mRNAs. On the other hand, the level of the L- and Q-transcripts was a¡ected similarly in the total and nuclear pools of RNA by partial hepatectomy or by an injection with cycloheximide (Fig. 6). This indicates that these treatments did not cause a de¢ciency in the nucleocytoplasmic transport of L- and Q-transcripts, which should have resulted in their accumulation in the nucleus. Collectively, the above data suggest that a nuclear event underlies the hepatectomy-induced loss of the NPPL and NPPQ transcripts and most likely involves an increased turnover and/or a de¢cient processing of their pre-mRNAs. The same conclusion may apply to the loss of NPPQ transcript after the administration of cycloheximide. 4. Conclusions We showed here that all three members of the mammalian NPP family are present in rat hepatocytes (Figs. 1 and 2), but that their expression is di¡erentially regulated. Thus, the expression of the three isozymes increased substantially during the hepatic growth phase in the ¢rst weeks after birth, albeit with a di¡erent time course (Fig. 3), showing that their expression is controlled by di¡erent stimuli. Conversely, the three NPP transcripts were transiently lost during liver regeneration following 70% hepatectomy (Figs. 4 and 6). For the NPPL and NPPQ mRNAs this loss appears to be accounted for by an intranuclear post-transcriptional process, involving the stability or processing of the immature transcripts. On the other hand, NPPK was also downregulated following sham surgery (Fig. 4), which suggests that NPPK represents a negative acute-phase protein and that its expression is controlled by anti-in£ammatory agents. We also observed that the NPPK and NPPQ transcripts, in contrast to the NPPL transcript, rapidly decreased after the administration of cycloheximide (Fig. 5), indicating that the stability of the K- and Q-transcripts is controlled by one or more proteins with a short half-life.

51

Acknowledgements This work was ¢nancially supported by the Fund for Scienti¢c Research-Flanders (Grant G.0237.98). Karolien Nelissen and Cindy Verwichte provided expert technical assistance. We thank Dr. H. Deissler (Essen, Germany) and Dr. M. Stracke (Bethesda, MD, USA) for the generous gifts of NPPL and NPPK cDNAs, respectively. Dr. Veerle Vulsteke is acknowledged for advice on the culture of FAO cells.

References [1] W.E. Razell, Methods Enzymol. 6 (1963) 236^258. [2] C. Delporte, P. Poloczek, J. Winand, Am. J. Physiol. 270 (1996) C286^C292. [3] N.F. Rebbe, B.D. Tong, S. Hickman, Mol. Immunol. 30 (1993) 87^93. [4] M.L. Stracke, T. Clair, L.A. Liotta, Adv. Enzyme Regul. 37 (1997) 135^144. [5] M. Narita, J. Goji, H. Nakamura, K. Sano, J. Biol. Chem. 269 (1994) 28235^28242. [6] H. Deissler, F. Lottspeich, M.F. Rajewsky, J. Biol. Chem. 270 (1995) 9849^9855. [7] L.J. Scott, D. Delautier, N.R. Meerson, G. Trugnan, J.W. Goding, M. Maurice, Hepatology 25 (1997) 995^1002. [8] P. Jin-Hua, J.W. Goding, H. Nakamura, K. Sano, Genomics 45 (1997) 412^415. [9] J.W. Goding, R. Terkeltaub, M. Maurice, P. Deterre, A. Sali, S.I. Belli, Immunol. Rev. 161 (1998) 11^26. [10] C. Stefan, W. Stalmans, M. Bollen, Eur. J. Biochem. 241 (1996) 338^342. [11] N.F. Rebbe, S. Hickman, Biochem. Biophys. Res. Commun. 175 (1991) 637^644. [12] B.A. Maddux, P. Sbraccia, S. Kumakura, S. Sasson, J. Youngren, A. Fisher, S. Spencer et al., Nature 373 (1995) 448^451. [13] N.R. Meerson, D. Delautier, A.-M. Durand-Schneider, A. Moreau, M.L. Schilsky, I. Sternlieb, G. Feldmann, M. Maurice, Hepatology 27 (1997) 563^568. [14] C. Stefan, W. Stalmans, M. Bollen, Hepatology 28 (1998) 1497^1503. [15] T. Clair, H.Y. Lee, L.A. Liotta, M.L. Stracke, J. Biol. Chem. 272 (1997) 996^1001. [16] M. Uriarte, W. Stalmans, S. Hickman, M. Bollen, Biochem. J. 306 (1995) 271^277. [17] A.N. Santos, D. Riemann, A.N. Santos, A. Kehlen, K. Thiele, J. Langner, Biochem. Biophys. Res. Commun. 229 (1996) 419^424. [18] R. Kettenhofen, M.-K. Meyer zu Brickwedde, Y. Ko, H. Vetter, A. Sachinidis, J. Mol. Biol. 279 (1998) 323^329.

BBAMCR 14464 21-4-99

52

C. Stefan et al. / Biochimica et Biophysica Acta 1450 (1999) 45^52

[19] R. Huang, M. Rosenbach, R. Vaughn, D. Provedini, N. Rebbe, S. Hickman, J. Goding, R.J. Terkeltaub, Clin. Invest. 94 (1994) 560^567. [20] M. Lotz, F. Rosen, G. McCabe, J. Quach, F. Blanco, J. Dudler, J. Solan et al., Proc. Natl. Acad. Sci. USA 92 (1995) 10364^10368. [21] J.L. Solan, L.J. Deftos, J.W. Goding, R.A. Terkeltaub, J. Bone Miner. Res. 11 (1996) 183^192. [22] P. Deterre, L. Gelman, H. Gary-Gouy, C. Arrieumerlou, V. Berthelier, J.-M. Tixier, S. Ktorza, J. Goding, C. Schmitt, G. Bismuth, J. Immunol. 157 (1996) 1381^1388. [23] M. Bollen, L. Hue, W. Stalmans, Biochem. J. 210 (1983) 783^787.

[24] J.M. Chirgwin, A.E. Przybyla, R.J. MacDonald, W.J. Rutter, Biochemistry 18 (1979) 5294^5299. [25] L. Miller, Anal. Biochem. 100 (1979) 166^173. [26] B. Fuss, H. Baba, T. Phan, V.K. Tuohy, W.B. Macklin, J. Neurosci. 17 (1997) 9095^9103. [27] G. Schreiber, A. Tsykin, A.R. Aldred, T. Thomas, W.-P. Fung, P.W. Dickson, T. Cole, H. Birch, F.A. De Jong, J. Milland, Ann. NY Acad. Sci. 557 (1989) 61^85. [28] J. Milland, A. Tsykin, T. Thomas, A.R. Aldred, T. Cole, G. Schreiber, Am. J. Physiol. 259 (1990) G340^G347. [29] H. Baumann, J. Gauldie, Immunol. Today 15 (1994) 74^80. [30] H. Moshage, J. Pathol. 181 (1997) 257^266.

BBAMCR 14464 21-4-99