Paracrine roles of NAD + and Cyclic ADP-ribose in increasing intracellular calcium and enhancing cell proliferation of 3T3 fibroblasts.

CD38 is a bifunctional ectoenzyme synthesizing from NAD(+) (ADP-ribosyl cyclase) and degrading (hydrolase) cyclic ADP-ribose (cADPR), a powerful universal calcium mobilizer from intracellular stores. Recently, hexameric connexin 43 (Cx43) hemichannels have been shown to release cytosolic NAD(+) from isolated murine fibroblasts (Bruzzone, S., Guida, L., Zocchi, E., Franco, L. and De Flora, A. (2001) FASEB J. 15, 10-12), making this dinucleotide available to the ectocellular active site of CD38. Here we investigated transwell co-cultures of CD38(+) (transfected) and CD38(-) 3T3 cells in order to establish the role of extracellular NAD(+) and cADPR on [Ca(2+)](i) levels and on proliferation of the CD38(-) target cells. CD38(+), but not CD38(-), feeder cells induced a [Ca(2+)](i) increase in the CD38(-) target cells which was comparable to that observed with extracellular cADPR alone and inhibitable by NAD(+)-glycohydrolase or by the cADPR antagonist 8-NH(2)-cADPR. Addition of recombinant ADP-ribosyl cyclase to the medium of CD38(-) feeders induced sustained [Ca(2+)](i) increases in CD38(-) target cells. Co-culture on CD38(+) feeders enhanced the proliferation of CD38(-) target cells over control values and significantly shortened the S phase of cell cycle. These results demonstrate a paracrine process based on Cx43-mediated release of NAD(+), its CD38-catalyzed conversion to extracellular cADPR, and influx of this nucleotide into responsive cells to increase [Ca(2+)](i) and stimulate cell proliferation.


Introduction
CD38, a type II transmembrane glycoprotein of 46 kDa, formerly known as a leukocyte activation antigen (1,2), has attracted increasing attention since it proved to be a bifunctional ectoenzyme involved in the metabolism of two signal molecules, i.e. cyclic ADP-ribose (cADPR) . and NAADP + (3,4). CD38 is able either to convert NAD + to cADPR (ADP-ribosyl cyclase) and then to hydrolyze cADPR (cADPR hydrolase), or to catalyze a base exchange reaction leading to NAADP + biosynthesis from NADP + and nicotinic acid (3,4). Cyclase and base exchange activities are common to other members of the CD38 family, the best known of which is a soluble protein purified and characterized from the marine mollusk Aplysia californica (5,6).
CD38-containing membrane vesicles proved to elude such compartmentation and to be causally associated to cADPR-dependent [Ca 2+ ] i increases (10,11). Elucidation of the topological paradox of the CD38/cADPR system came from some recent findings: i) the plasmamembrane of several cell types harbors a passive transport system for pyridine dinucleotides, which is responsible for NAD + fluxes through the membrane (11), thus providing NAD + substrate to the otherwise unaccessible active site of CD38. This dinucleotide transporter has been identified with connexin 43 hemichannels (12). ii) Transmembrane CD38 is an active transporter of catalytically produced cADPR across its oligomeric structure (13). iii) A third, CD38-unrelated mechanism of permeation of extracellular cADPR across cell membranes has been postulated in selected cell types (14,15).
Presence of multiple transport systems for NAD + and cADPR in the plasmamembrane raises the possibility of a paracrine exchange of these molecules between neighboring cells via Cx43, CD38 and eventually cADPR influx. The possibility of NAD + /cADPR-related paracrine mechanisms and their potential role in regulating intracellular calcium were experimentally addressed in the present study by means of co-cultures of CD38 sense-and antisense-transduced 3T3 fibroblasts.
CD38 -3T3 cells were found to respond to the paracrine production of cADPR by cocultured CD38 + 3T3 cells with a calcium-related increase of proliferation. This hitherto unrecognized interplay between extracellular NAD + and cADPR may represent a means for regulating intracellular calcium homeostasis and relevant cell 6 400 µl PBS containing 10 mM glucose (PBS-glucose) at 37°C. At different times, 60 µl aliquots of the incubation mixtures were centrifuged 30 s at 5,000 x g and the corresponding supernatants were deproteinized with TCA (10% final concentration) as described (14). HPLC analyses of nucleotides in the samples were performed as described (15). Protein content was determined according to Bradford (18). Determination of intracellular cADPR in co-culture conditions. After 48 h co-culture on 75 mm diameter plates, CD38target cells were washed with 10 ml PBS, detached with trypsin and washed twice with 1 ml of ice-cold PBS at 5,000 x g for 30 s.
Pellets were resuspended in 250 µl of cold water and frozen at -20°C, then thawed and sonicated in ice 1 min at 3W. A 50 µl aliquot was withdrawn for assay of protein (18), while the rest of the sample was deproteinized with 10% TCA (14). The cADPR content of the cell extracts was analyzed by two subsequent HPL chromatographies after addition of trace amounts of radiolabeled [H 3 ]cADPR (2 x 1,000 cpm) as internal standard (14). Identification of the cADPR peak in the cell extracts was confirmed by co-elution with the radioactive standard, by comparison of the absorbance spectrum and elution time with standard cADPR and by the disappearance of the corresponding peak in the matched CD38-hydrolyzed samples (14). The concentration of intracellular cADPR was calculated from the area of the HPLC peak, taking into account the percentage of nucleotide recovery obtained with the radioactive standard.
Determination of extracellular cADPR. At various times of co-culture of CD38over CD38 +/-3T3 feeder cells in complete medium (without phenol red), the medium was collected and clarified by three repeated centrifugations at 300 x g for 5 min.
The cell-free medium was TCA-deproteinized (14) and submitted to enzyme digestion to hydrolyze nucleotides potentially interfering with the cADPR assay (16).
cADPR content in the samples was determined by a sensitive and specific radioimmunoassay (RIA) (16), rather than by HPLC as for intracellular cADPR levels (see above), because high salt concentrations in these samples proved to interfere with the latter type of analyses. identical to those recorded in the same 3T3 fibroblasts exposed to oleamide, a known inhibitor of solute exchange across gap junctions (21) and of specific NAD + transport through Cx43 hemichannels in isolated 3T3 cells (12). Also, the  (Table I).

Determination of extracellular NAD
On the contrary, concentrations of extracellular cADPR, measured by means of a specific RIA (16), were progressively increasing in the CD38 + /CD38cocultures, until reaching 6.0± ± ± ±0.8 nM at 72 h (Table I) The results shown in Fig. 3 and in Table I  In order to better correlate extracellular cADPR to intracellular calcium release, the specific cADPR antagonists 8-NH 2 -cADPR and 8-Br-cADPR were separately added to the co-culture medium. As shown in Fig. 4 co-culturing with CD38 + feeder cells demonstrate a paracrine role of extracellular NAD + and cADPR in the mechanism underlying these changes. Specifically, NAD + release from cells followed by CD38-catalyzed generation of extracellular cADPR seem to be the required steps. Since NAD + release has now been shown to take place in isolated 3T3 cells across hexameric hemichannels of Cx43 (12), we attempted to disrupt the paracrine effects of co-culture by inhibiting the NAD + -exporting activity of Cx43 hemichannels. Oleamide proved to block the [Ca 2+ ] i increase in our coculture setting almost completely (Fig.4). Moreover, the specific anti-Cx43 oligodeoxynucleotide inhibited the [Ca 2+ ] i increase in the target CD38cells, while the corresponding sense deoxynucleotide was totally uneffective (Fig.4). These results give further support to the idea of Cx43-mediated export of cellular NAD + and of subsequent generation of extracellular cADPR at the outer surface of the CD38 + feeder cells followed by influx of cADPR into the target CD38cells across a Cx43-unrelated transport system (Fig.2).
Role of cADPR in the changes observed in mixed CD38 +/co-cultures. In an effort to directly demonstrate this paracrine mechanism, we measured intracellular cADPR in the target CD38cells during the co-culture experiments. CD38cells were incubated on 75 mm diameter transwell plates over pre-established CD38 -/+ feeder layers for 48 h in the same conditions used for cell cycle analysis (see below). Cell extracts were then analyzed by HPLC. The intracellular cADPR concentration was undetectable in the CD38target cells co-cultured over homologous CD38layers (controls), while it was estimated to be 2.1±0.1 picomoles/mg in the CD38grown on CD38 + feeders.
This value is in the range of reported intracellular concentrations of cADPR in constitutively CD38 + human lymphoid and myeloid cell lines (27,28). .
Effect of the co-culture over CD38 +/feeders on the proliferation of CD38 -3T3 cells.
De novo expression of CD38 has been demonstrated to enhance the rate of proliferation of some cell types, including 3T3, via increases of [Ca 2+ ] i elicited by intracellular cADPR (10). In order to investigate whether the calcium mobilization in CD38cells which is induced by cADPR generated and provided by CD38 + cell feeders could interfere with cell growth, we assayed proliferation of CD38 -3T3 target cells co-cultured with CD38 +/-3T3 feeder cells. A significant (p<0.05) increase in cell proliferation was observed in CD38fibroblasts co-cultured over CD38 + cells until 72 h as compared with the same CD38cells grown on homologous CD38feeders (Fig.5A). This increase was inhibited by addition of NAD + -ase to the medium, with a maximum effect being recorded after 24 h culture (p<0.05).

The reduced extent of inhibition afforded by NAD + -ase at 48 and 72 h, despite
the appearance of detectable levels of extracellular cADPR (Table I) In Table II   site of CD38 in their plasmamembrane (12). Subsequent cADPR generation is followed by its channelling across oligomeric CD38 to reach the cytosol of the feeder cell, thus completing an autocrine loop (13), and also by appearance of cADPR in the extracellular medium. The third step is permeation of cADPR across the plasmamembrane of target CD38cells (Fig. 1), as previously suggested to occur in several cell types responding to extracellular cADPR with calcium mobilization and in some cases with remarkable changes in cell functions (14, 15,24,25).
The paracrine model summarized in Fig.6 envisages new roles for NAD + as a cell-to-cell communication signal mimicking a hormone, while extra/intracellular cADPR represents its second messenger and intracellular calcium behaves as a third messenger regulating selected cell functions (3,9,29). Recently, a comparable yet intracellularly localized loop has been described in rat heart mitochondria, where opening of the permeability transition pore (PTP) is followed by release of intramichondrial NAD + . This dinucleotide can accordingly behave as substrate for the NAD + -glycohydrolase located outside the matrix space (30) which has been shown to express ADP-ribosyl cyclase activity (31). Therefore, release of NAD + through mitochondrial PTP is expected to produce cytosolic cADPR with consequent calcium release from sarcoplasmic reticulum (30).  (Table I), since its reported Km is 14 µ µ µ µM NAD + (19). Accordingly, generation of cADPR in the extracellular space could play a limiting role in this complex process.

Although the paracrine process involving NAD
Finally, the next step, i.e. clearance of extracellular cADPR by the CD38target cells requires elucidation of the transport system responsible for cADPR influx ( Fig.1), whose molecular properties are as yet unknown.
With respect to this, an interesting feature is the remarkably high efficiency in the co-colture system of extracellular cADPR concentrations as low as 4.0-6.0 nM (Table I) (Table II) (Table I). Finally, the time-dependent increase of the proliferation rate measured at 24, 48 and 72 h in CD38target cells co-cultured over CD38 + feeders (Fig. 5A), which likely reflects a progressive shortening of the S phase of the cell cycle until reaching a T S value of 10± ± ± ±5.4 h at 72 h (Table II), parallels the increase of [cADPR] e in the media (Table I).
The growth-enhancing effect featured by CD38 + cell feeders on CD38target fibroblasts by virtue of a paracrine NAD + /cADPR mechanism may have important functional consequences which should extend beyond our model system of coculture. For instance, in bovine tracheal strips, we were able to show that coincubation of mucosa CD38 + fragments with smooth myocytes induces the NAD + /cADPR-mediated increase of [Ca 2+ ] i in these cells (15). Moreover, a cADPRdependent expansion of human hemopoietic progenitors grown on CD38 + stroma cells has been recently observed in our laboratory ** . The mechanism underlying the calcium-related stimulation of cell growth proved to be a significant shortening of the S phase of the cell cycle (Table II) fibroblasts and human HeLa cells (10). Therefore, [Ca 2+ ] i increases that follow either enhanced intracellular traffic of NAD + and cADPR (9) or an extracellular exchange of both signal metabolites can trigger an increased cell proliferation via a significant shortening of the S phase of cell cycle. Figures.   Fig. 1 (3), or 1 µM 8-Br-cADPR (4), or 50 µM oleamide (5). CD38target cells pre-treated with Cx43 antisense (6) or sense (7) oligodeoxynucleotide were co-cultured 24 h on CD38 + feeder cells pre-treated with the same oligodeoxynucleotide. [Ca 2+ ] i measurements were carried out as described (10). Values are means ± SD of 5 different experiments.  For details see text.