Adrenergic regulation of brown adipocyte

Adipose Tissue Differentiation and Development

Adrenergic regulation of brown adipocyte differentiation B. Cannon and J. Nedergaard The Wenner-Gren Institute, The Arrhenius Laboratories F3, Stockholm University, S- I06 9 I Stockholm, Sweden

407

Introduction Brown and white fat-cells share a basic life pattern in developing from undifferentiated fibroblast-like precursors to becoming increasingly differentiated with time. Throughout the life of an organism, a pool of precursors remains available. It would seem that the number of white fatcells, at least, is not constant but is influenced by external factors, such as food supply, and so presumably is their degree of differentiation. However, a much more conspicuous response to external stimuli is found in brown adipose tissue: the process of recruitment. Physiologically, recruitment is the increase in the total heat-producing capacity of the tissue that occurs in response to a chronic environmental stimulus, especially chronic exposure to cold conditions. At the cellular and molecular level, it may be fair to define recruitment as the sum of those processes that accompany the development of the enhanced capacity for heat production. Although it has been possible to define and follow many of the recruitment-associated cellular and molecular processes in vivo, the present trend is to analyse these processes in vitro using brown fat-cell cultures. The advantage of utilizing in vitm systems is illustrated in Figure 1, where the complex situation in vivo is compared with the somewhat more defined in vitro experimental systems. As is also implied in Figure 1, the present understanding is that noradrenaline has a dual role in the tissue, being both the agent that induces the acute metabolic response in the tissue, i.e. heat production, and also the agent that initiates the recruitment process under discussion here. The recruitment process can appropriately be considered to consist of cell proliferation and cell differentiation events.

Proliferation In vivo

At normal environmental temperatures, brown adipose tissue is proliferatively dormant. It has been estimated that only about 40% of the total cell population of the tissue is present as mature adipocytes. However, a significant fraction of the remainder are precursor cells, committed to becoming brown adipocytes at some point in

time, but with an intact proliferative potential and an undifferentiated appearance. It is currently not clear what prevents these precursor cells from proliferating. There are presumably serum growth factors available but the cells remain unresponsive to these. It may be said that the cells are growth-inhibited by the presence of, and contact with, neighbouring cells, and, although this is an adequate description of the situation, the molecular basis for the growth inhibition is not known; it can apparently be overcome when physiologically necessary. When an increased capacity for non-shivering thermogenesis is required, proliferation commences rapidly (Figure 2 A ) . This increased cell proliferation was first observed 30 years ago by Cameron and Smith [4] who demonstrated a transient 70-fold increase in mitotic index in brown adipose tissue of animals exposed to cold. They concluded that the proliferating cells were originating from (or close to) the endothelium, as was later confirmed [5,6]. The cold-induced hyperplasia can be mimicked in vivo by noradrenaline injections, and the effect of cold can be blocked by P-adrenergic antagonists [7,8]. Thus cell proliferation in brown adipose tissue is apparently stimulated by noradrenaline released by the sympathetic nerves, just as is thermogenesis. In vitro

Although it appeared that cell proliferation was under sympathetic control, it was not unequivocal from studies in vivo that this was due to a direct interaction of noradrenaline with the brown preadipocytes, even if this was the implication. Studies on cell culture systems could potentially clarify this situation. Precursor cells, isolated by collagenase digestion of brown adipose tissue, can be plated on to plastic culture dishes. In the presence of newborn calf serum, the cells proliferate readily with a doubling time of about 2 4 h until they reach apparent confluence [ 1,2]. This ‘spontaneous’ proliferation requires the presence of serum and thus presumably of conventional growth factors. In serum-free conditions, when the cells are quiescent, certain of these conventional growth factors do indeed increase DNA synthesis [12]. The in

I996

Biochemical Society Transactions

vitro system therefore does not entirely mimic the in vivo situation where conventional growth 408

factors are presumably persistently present without proliferation commencing. One possible interpretation is that the presence of mature adipocytes exerts an inhibitory action on proliferation, although this has not been tested. The in vivo studies indicated that noradrenaline is the agent that, directly or indirectly, initiates the tissue hyperplasia. T h e addition of noradrenaline to postconfluent brown adipocyte cultures does not lead to any stimulation of DNA synthesis [Z]. There may be two interpretations of this. First, there are no precursors left in this

system, all of them having developed into mature cells. In this case, this experiment implies that in the in vitm system all cells lose their proliferative potential upon reaching confluence and initiate a differentiation programme. Secondly, the structural organization of the system in the Petri dish may be so different from that occurring in vivo that this limits proliferation: in the tissue, the cells grow in a three-dimensional array, whereas they become growth-inhibited when the plate is covered with a monolayer of cells. If, however, the noradrenaline is added to the cell cultures before confluence, ongoing DNA synthesis is accelerated [Z]. Also, when a

Figure I

A schematic view at the cellular level of the recruitment process in brown adipose tissue in vivo compared with in vitro models ( A ) The complexity of the recruitment process in vivo. In unstimulated conditions, the tissue contains brown adipocytes at a given level of differentiation [here indicated by a fat droplet and a mitochondrion with the uncoupling protein (UCP)], as well as less differentiated preadipocytes and fully undifferentiated precursor cells. During the recruitment process, the following parallel events may be assumed to take place. The existing brown adipocytes are stimulated by noradrenaline to initiate heat production. At the same time, it is likely that these cells also respond to adrenergic stimulation by advancing in their degree of differentiation (here illustrated by more mitochondria and more UCP in the mitochondria). The preadipocytes will advance to at least the state of being true brown adipocytes. The precursors will start to divide and some of the daughter cells will develop into preadipocytes and mature brown adipocytes. As these processes are not taking place with the same time constants, the characteristics of the tissue, as normally investigated by biochemical means, will alter in a complex manner during this process, e.g. the increase in the number of cells may initially exceed the increase in total UCP content, making the tissue seem less fully recruited than it really is. Further, the picture of differentiation observed will be heavily influenced by the methods of study, i.e. membrane preparations from the tissue will include membranes from differentiating cells whereas mature (floating) cell preparations even in the fully recruited state will select for populations of cells that are not necessarily much different from the original mature cells. (6) The recruitment process as studied in vitro. Three types of developmental studies are illustrated. ( I ) The spontaneous developmental process of brown fat-cells grown in culture. Precursors isolated from brown adipose tissue, when plated, undergo a series of morphological and biochemical changes. First, a ‘spontaneous’ proliferation process is observed, then a partial increase is seen in the degree of differentiation, but with time, these cells tend to dedifferentiate [I]. It is not certain what physiological conditions this in vitro developmental pattern mirrors: during the recruitment process in vivo, the cells are always under some degree of adrenergic stimulation, in contrast with the case here. (2) The acute effects of noradrenaline addition. With the use of acute addition of noradrenaline, only such processes that are directly under adrenergic control can be studied. In the precursors, adrenergic stimulation leads to increased cell proliferation [2]; in the mature cells, it leads to an increase in expression of the gene for the UCP [3]. (3) The effects of chronic noradrenaline stimulation. If the chronic presence of noradrenaline induces changes that are principally different from those observed after acute addition, it may be said that noradrenaline also affects the degree of differentiation of the cells. Undoubtedly, this is the condition that most resembles the conditions of recruitment in vivo (A ).

A. Recruitment in-vivo

B. Differentiation in-vitro

Adrenerpic stimulation Unstimulated

Volume 24

Recruitment transition

Fully recruited

Time in culture Adrenergicallyunstimulated cell growth

h


Figure 2 Parameters of recruitment as studied in vivo and in vitro The same time scale is used in all curves. (A) The increase in total DNA content of the interscapular brown adipose tissue during the recruitment process ( 8 ) The effect of noradrenaline on total DNA of brown fat-cells in a serum-free culture (v. Golozovbova, unpublished work, but see [2] for methodology). (C) The increase in the content of the uncoupling protein thermogenin (UCP) during the recruitment process (adapted from [9]).(D) The increase in UCP in cells chronically exposed to noradrenaline (adapted from [ 101). The slope between (C) and (D) is not directly comparable, ( E ) The ratio between UCP and DNA content of brown adipose tissue during the recruitment process (adapted from [ I I]).

PI.

300

-

250

-

A. In-vivo recruitment of cells

409

B. In-vitro recruitment of cells

* s 4- 200

C. In-vivo recruitment of UCP

-

1 cold

600

K 400

6 200

3

control 0

I

I

I

3

10

15

I

20

I 2s

0

Days in the cold

D

I996


410

similar experiment is performed in the absence of serum when the cells are quiescent, noradrenaline can significantly increase DNA content (Figure 2 B ) . T h e tentative conclusion must be that noradrenaline can function as a genuine mitogen in this system. It is also possible that there is a noradrenaline-induced inhibition of apoptosis, which would contribute to the elevated DNA content. T h e in vivo experiments indicated that the noradrenergic stimulation of DNA synthesis is ,8-adrenergically mediated. Further studies in the cell culture system have confirmed that this is indeed the case and that any agent that elevates intracellular cyclic AMP increases DNA synthesis [ 2 ] .This is a rather unusual system, as cyclic AMP in other systems is often associated with inhibition of proliferation and increased differentiation. Analysis of the ,8-adrenoceptor involved in mediation of the proliferative signal has demonstrated that it is the ,!lI-adrenoceptor that is responsible for the elevation of cyclic AMP leading to proliferation. T h e acute action of noradrenaline on mature brown adipocytes - the induction of thermogenesis - has been shown to be a predominantly ,!13-adrenoceptor-mediated event (at least in rodents). These receptors are expressed most strongly in adipose tissues, and, like other /?-receptors, are coupled to cyclic AMP. However, it appears that these P3-receptors are not expressed in preadipocytes, where only ,8,-receptors seem to be coupled to enhanced cyclic AMP synthesis. An increased expression of the proto-oncogene c-myc has generally been associated with increased cell proliferation in numerous systems. In brown adipose tissue in vivo, there is no increase in c-myc expression when animals are placed in the cold and cell proliferation is initiated; nor is an increased expression seen when noradrenaline is added to preconfluent brown adipocytes in order to stimulate DNA synthesis (P. Tvrdik, unpublished work). These experiments may be interpreted to indicate that the serum-free quiescent cells enter Go, whereas cells in the animal are in an arrested GI phase with an adequate c-myc level and require no further c-myc for progression through the cell cycle. These results should be contrasted with the proto-oncogene c-fos, the expression of which is elevated under all conditions studied [ 131. This dissociation in the expression of these protooncogenes is rather unusual.

Volume 24

Increased differentiation In vivo

It is of importance to know whether the individual mature brown adipocyte has achieved a higher degree of differentiation after the recruitment process. A satisfactory definition of degree of differentiation must thus be established. This could be defined as an increase in the oxidative capacity per cell, implying an increased mitochondrial complement. This seems to be fulfilled, although the magnitude of the increase is more marginal than would initially be anticipated. An increase in the presence of the mitochondrial uncoupling protein (UCP; thermogenin) [14] is also a prerequisite for the thermogenic process. However, although an initial dramatic increase in expression at the mRNA level of the UCP gene is noted immediately upon cold exposure and this level of expression is maintained [ 9 ] , this leads only slowly to an increase in the total content of UCP (Figure ZC), and it would appear that at the end of the transition phase, when acclimation has been achieved, the increase in UCP and in mitochondrial complement per cell is rather moderate (Figure 2 E ) , and that the most dramatic increase is in cell number, with only a modest increase in the capacity of the individual cell. It is possible that different species adopt different modes of acclimation and thus different means of recruitment. For example, it appears that the guinea pig markedly increases its oxidative capacity per cell and has thus a more moderate requirement for cell proliferation as a means of increasing non-shivering thermogenesis.

In vitro

In the cell culture system where the spontaneously proliferating preadipocytes reach confluence, ‘spontaneous’ differentiation is also initiated. T o a higher degree than proliferation, this process indeed appears to be spontaneous, since it can proceed, at least to some extent, in a serum-free medium. One of the earliest events appears to be the expression of [j3-adrenoceptor mRNA, a switch to the f13-adrenoceptor being the one that is coupled to adenylate cyclase (T. Bengtsson, unpublished work). Further, the UCP gene becomes responsive to adrenergic stimulation, leading on noradrenaline addition to a large increase in UCP mRNA [ 131 and a consecutive increase in the amount of UCP (Figure 20).


The adrenergic pathways and the recruitment process As has been pointed out above, the recruitment process seems to be governed by increased sympathetic stimulation and mediated via direct effects of noradrenaline on the cells. This is a unique situation which makes evaluation of what are true characteristics of the advanced degree of differentiation difficult. Especially with respect to the effects of recruitment on the adrenergic transmission system itself, it may be difficult to distinguish what is part of a ‘normal’ desensitization due to chronic adrenergic stimulation and what is a ‘true’ recruitment effect. This distinction may, however, be artificial and we may have to accept that all differences observed after recruitment should be considered as part of the recruitment process.

Functional desensitization to noradrenaline

Brown fat-cells isolated from recruited tissue are about 10-fold less sensitive to noradrenaline than are cells from non-recruited animals [15]. The molecular background for the desensitization is still not known. A down-regulation of the P3-adrenoceptor has been reported to occur both in vivo and in v i m [16-191, but further studies have demonstrated that this down-regulation is transient [ 191 and thus cannot explain the desensitization. The content of G,a is decreased during recruitment [ZO] , and cyclic AMP phosphodiesterase activity is increased [21], but these effects do not seem to be sufficient to explain the desensitization adequately. Although desensitization in general is a well-described phenomenon, its physiological significance is not immediately evident. For example, it would seem less physiologically efficient to decrease the sensitivity of the brown fatcells to noradrenaline while the animal still remains in the cold and thus needs a constant and high level of heat production. It may, however, be possible to discuss the desensitization as a physiologically meaningful process in connection with brown adipose-tissue function. It has been found that the intensity of sympathetic stimulation of the tissue is unaltered during recruitment [ZZ] . As the total heat-producing capacity of the tissue increases during this time, the animal may need a system to allow a smaller fraction of the recruited capacity of the tissue to be used with the same intensity of nervous stimulation; desensitization may play this role.

Functional desensitization to adenosine

Adenosine can partly inhibit noradrenalineinduced thermogenesis. One effect of recruitment is that isolated brown fat-cells lose their ’ may responsiveness to adenosine [23,24]. T h is be explained by a decrease in the content [25] of Gia subunits that transfer the inhibitory effect of adenosine to the adenylate cyclase. The functional role of this desensitization is not known. Possible Frictional up-regulation of a ,-adrenergic pathw-

Concerning the a I-adrenergic receptors, an unexpected up-regulation during recruitment has been observed. There is an increase in a l adrenoceptor density [26,27] which seems to be due to an increase in the expression of the aIAadrenoceptor gene (the a lB-adrenoceptor gene is poorly expressed in the tissue, and the alDadrenoceptor gene shows a transient decrease in expression) (K. Kikuchi-Utsumi, M. KikuchiUtsumi, B. Cannon and J. Nedergaard, unpublished work). The direct contribution to thermogenesis of al-adrenoceptors is low or absent in control cells [29], and there is no published indication that the contribution is higher in cells from recruited tissue, despite the increased density of a I-adrenoceptors. It is, however, clear that the al-adrenoceptors are very active (synergistically with P-adrenoceptors) in the regulation of, for example, the expression of important transcription factors, such as c-fos [13] and C/EBPa (S. Rehnmark, unpublished work). This may thus mean that the al-adrenoceptor pathway constitutes a positive feed-forward loop for the recruitment process in brown adipose tissue. 1 Ntchad, M., Kuusela, P., Carneheim, C., Bjorntorp, P., Nedergaard, J. and Cannon, B. (1983) Exp. Cell Res. 149, 105-118 2 Bronnikov, G., Houstek, J. and Nedergaard, J . (1992) J. Biol. Chem. 267,2006-2013 3 Rehnmark, S., NCchad, M., Herron, D., Cannon, B. and Nedergaard, J. (1990) J. Biol. Chem. 265, 16464-16471 4 Cameron, I. L. and Smith, R. E. (1964) J. Cell Biol. 23, 89-100 5 Hunt, T. E. and Hunt, E. A. (1967) Anat. Rec. 157, 537-546 6 Bukowiecki, L. J., GCloCn, A. and Collet, A. J. (1986) Am. J. Physiol. 250, C880-C887 7 GCloCn, A., Collet, A. J., Guay, G. and Bukowiecki, L. J. (1988) Am. J. Physiol. 254, C175-Cl82 8 Rehnmark, S. and Nedergaard, J. (1989) Exp. Cell

I996

41 I


412

Res. 180,574-579 9 Jacobsson, A., Muhleisen, M., Cannon, B. and Nedergaard, J. (1994) Am. J. Physiol. 267, R999R 1007 10 Puigserver, P., Herron, D., Gianotti, M., Palou, A., Cannon, B. and Nedergaard, J. (1992) Biochem. J. 284, 393-398 11 Nedergaard, J., Unelius, L., Jacobsson, A., Muhleisen, M., Svoboda, P. and Cannon, B. (1993) in Life in the Cold. Ecological, Physiological, and Molecular Mechanisms (Carey, C., Florant, G. L., Wunder, B. A. and Honvitz, B., eds.), pp. 345-359, Westview Press, Boulder 12 Lorenzo, M., Valverde, A. M., Teruel, T. and Benito, M. (1993) J. Cell Biol. 123, 1567-1575 13 Thonberg, H., Zhang, S.-J., Tvrdik, P., Jacobsson, A. and Nedergaard, J. (1994) J. Biol. Chem. 269, 33 179-33186 14 Nedergaard, J. and Cannon, B. (1992) in New Comprehensive Biochemistry, vol. 23, Molecular Mechanisms in Bioenergetics (Ernster, L., ed.), pp. 385-420, Elsevier, Amsterdam 15 Nedergaard, J. (1982) Am. J. Physiol. 242, (2250C257 16 Granneman, J. G. and Lahners, K. N. (1992) Endocrinology 130, 109-1 14 17 Revelli, J.-P., Muzzin, P. and Giacobino, J.-P. (1992) Biochem. J. 286, 743-746 18 Klaus, S., Muzzin, P., Revelli, J.-P., Cawthorne,

M. A., Giacobino, J.-P. and Ricquier, D. (1995) Mol. Cell. Endocrinol. 109, 189- I95 19 Bengtsson, T., Redegren, K., Strosberg, A. D., Nedergaard, J. and Cannon, B. (1996) J. Biol. Chem., in the press 20 Svoboda, P., Unelius, L., Cannon, B. and Nedergaard, J. (1993) Biochem. J. 295, 655-661 21 Unelius, L., Bronnikov, G., Mohell, N. and Nedergaard, J. (1993) Am. J. Physiol. 265, C1340C 1348 22 Young, J. B., Saville, E., Rothwell, N. J., Stock, M. J. and Landsberg, L. (1982) J. Clin. Invest. 69, I061- 1071 23 Woodward, J. A. and Saggerson, E. D. (1986) Biochem. J. 238, 395-403 24 Unelius, L., Mohell, N. and Nedergaard, J. (1990) Am. J. Physiol. 258, C818-C826 25 Svoboda, P., Unelius, L., Dicker, A., Cannon, B., Milligan, G. and Nedergaard, J. (1996) Biochem. J., 314, 761-768 26 Raasmaja, A., Mohell, N. and Nedergaard, J. (1985) Eur. J. Pharmacol. 106, 489-498 27 Raasmaja, A. (1990) Acta Physiol. Scand. (Suppl. 590) 139, 1-61 28 Reference deleted 29 Mohell, N., Nedergaard, J. and Cannon, B. (1983) Eur. J. Pharmacol. 93, 183-193 Received 27 November 1995

P3-Adrenoceptorsand the regulation of metabolism in adipose tissues J. R. S. Arch and S. Wilson SmithKline Beecham, The Frythe, Welwyn, Herts. AL6 9AR, U.K.

Introduction Evidence that B-adrenoceptors (BAR) could be divided into and p2 subtypes was first presented in 1967 by Lands et al. [ l ] , but it was evident by 1972, when Furchgott [2] reviewed the classification of adrenoceptors, that gut and adipose tissue BARS from certain species did not fit well into this classification. T h e main problem was that the potencies of B1- and B2-AR antagonists as antagonists of gut and adipose-tissue responses were lower than expected for these receptors. Work by Harms, Zaagsma and their colleagues during the 1970s [3,4] emphasized Abbreviations used: AR, adrenoceptor; CHO, Chinese hamster ovary; NEFA, non-esterified fatty acids; pAz, -log,,, concentration of antagonist required to shift concentration-response curve to agonist twofold to the right.

Volume 24

beyond any reasonable doubt that both the potencies and stereoselectivities of antagonists in blocking B-agonist-stimulated lipolysis in white adipocytes were far too low for the functional receptors to be PI- or B2-ARs. It was not, however, until the mid-l980s, when selective P3AR agonists were identified and shown to have potential in the treatment of obesity and noninsulin-dependent diabetes [S, 61, that the existence of B3ARs began to be widely recognized. T h e cloning of the B3AR from human genomic D N A [7] gave a further boost to the area. T h e B3AR is a seven-transmembrane G-proteincoupled receptor consisting of a single polypeptide chain [8]. Its second messenger is cyclic AMP. T h e existence of an additional second messenger has been hypothesized to account for the steep relationship between [j3AR-generated cyclic AMP levels and lipolysis [9] and for the