As indicated earlier in this chapter, a variety of humoral or circulating factors influence the de- velopment of cancer cachexia and many nutritional aspects of the host-tumor relationship. Fur- thermore, the interactions and interrelations of many of these factors add to the complexity of the mechanisms of cachexia seen in the tumor-bearing host. No single amino acid, cytokine, steroid, or polypeptide hormone can account for the syndrome of cancer cachexia. A list of some of the major humoral factors and their interrelationships are seen in Table 17.6.
Table 17.6 Humoral Factors and Their Interrelations in Cancer Cachexia
Table 17.6 is not meant to be exhaustive, as may be noted later in the chapter, but rather indicates the complexity of the various hormones, cytokines, and amines that contribute to the clinical picture of cachexia in the cancer patient. The reader is referred to various reviews on the effects of other factors noted in the table (Tisdale, 1997; Inui, 1999; Laviano et al., 1995; Mat- thys and Billiau, 1997; Roubenoff, 1997). The cytokines interleukin-1, interleukin-6, interferon- γ, and tumor necrosis factor-α all appear to have some degree of effectiveness in inducing some aspects of the syndrome of cancer cachexia (Tisdale, 1997). However, in some instances little or no change is seen in the serum levels of one or the other of these cytokines, again suggesting that no single factor is effective in inducing the syndrome of cancer cachexia. Of these cytokines, the best studied, which may serve to model the others as well as many other components of the syndrome of cancer cachexia, is tumor necrosis-α (TNF-α). Prostaglandins are produced both by host cells and in some cases by neoplastic cells. Thus, administration of inhibitors of prosta- glandin synthesis will have variable effects. However, in mice in which several cytokines were eliminated by gene targeting, inhibition of prostaglandin production by the neoplasm and host resulted in improvement of the cachectic state (Cahlin et al., 2000).
Tumor Necrosis Factor-α and Related Cytokines in Cancer Cachexia
The effects of TNF were first observed in the latter part of the nineteenth century by Coley (1893) based on the observation that patients with streptococcal infections might also exhibit a partial remission of some neoplasms. Subsequent investigations demonstrated that injection of bacterial products, specifically a lipopolysaccharide, could induce hemorrhagic necrosis of transplantable neoplasms in mice. Subsequently, a serum factor associated with administration of the bacterial product was isolated and named tumor necrosis factor (Carswell et al., 1975). Since that time, there has been a dramatic increase in our knowledge of TNF-α, its genetics, and its function. TNF-α is a member of a superfamily of genes of which more than 10 are presently known, along with their respective receptors (cf. Gruss, 1996). Neoplasms engineered to express high levels of TNF when grown in mice rapidly induce a syndrome of cancer cachexia (Oliff
et al., 1987). An interesting comparison of the clinical and metabolic alterations seen in cancer cachexia with those of the in vivo effects of TNF-α are listed in Table 17.7. As noted, there are relatively few differences in the syndrome and the effects of the cytokine, at least in those parameters listed. However, in some model systems it was possible to dissociate some of the effects of TNF-α from the cachectic syndrome. In particular, a number of studies have failed to detect elevated circulating levels of TNF-α in cachectic cancer patients or to associate an eleva- tion of this cytokine with the development of cachexia (cf. Tisdale, 1997). Still, the effects of the cytokine serve as an excellent model with which to consider the mechanisms whereby such hu- moral substances may exert effects leading to cachexia in the cancer patient.
TNF-α is produced predominantly by tissue macrophages, but other cell types—including endothelial cells, fibroblasts, and epithelium—synthesize significant amounts of this polypep- tide when subjected to appropriate stimuli (cf. Luster et al., 1999). A closely related member of the TNF-α family, lymphotoxin-α, is produced exclusively by lymphocytes, but its functions are not as well known as those of TNF-α, although both ligands interact with the same receptors (see below). TNF-α is synthesized as a large transmembrane molecule (Figure 17.6). It may be activated by a proteolytic cleavage at the surface of the cell membrane, releasing the 17-kDa active ligand, which may circulate in an endocrine manner or interact with adjacent cells in a paracrine manner.
Mechanisms of TNF-α Action—Signaling Pathways
As noted above, TNF-α is a member of a superfamily of ligands each having its own receptor, although TNF-α may interact with either of two different receptors. This is shown schematically
Table 17.7 Some Clinical and Metabolic Alterations in Cancer Cachexia: Comparison with the In Vivo Effects of Tumor Necrosis Factor
Figure 17.6 Proposed mechanism for priming and activation of TNF production from the macrophage. (Reproduced from Mizuno, 1992, with permission of the author and publisher.)
in Figure 17.7, which indicates that most cells of the organism possess receptors for TNF-α. As noted in the figure, interaction with the 55-kDa receptor (now known as CD120a) involves sig- nificantly more functions than interaction with the 75-kDa receptor. In the latter case, it appears that activation is preferentially by the cell-bound form of TNF, and it is likely that many of its functions are as yet unknown (cf. Wallach et al., 1999). As with many other receptors, the active form is the dimer, and the predominant form of the soluble TNF-α ligand is a homotrimer, which likely has as one of its functions the induction of dimerization and subsequent signal transduc- tion from the receptor.
A schematic representation of the 55- and 75-kDa (CD1206) receptors for TNF and lymphotoxin-α is seen in Figure 17.8. In addition, another receptor, Fas, is shown because of its similarity to the 55-kDa TNF receptor. It has a separate ligand and associates with a kinase that is involved in the induction of apoptosis by the interaction of Fas with its ligand (cf. Baker and Reddy, 1996). The interesting characteristic of the Fas and 55-kDa TNF-α receptor are domains in the intracellular portion of the molecule known as death domains. These domains interact with proteins as shown, having the names of TRADD and FADD. While the exact molecular mechanisms of these proteins is unknown, it is likely that they recruit other pro- teins, possibly activating such to ultimately result in apoptosis or, in the case of the 55-kDa TNF-α receptor, activation of the nuclear transcription factor NF-κB (Beyaert and Fiers,1998). The 75-kDa receptor, on the other hand, associates with different proteins known as TRAFs, which apparently have specific regions involved in the ultimate signaling pathway to NF-κB activation as shown (cf. Wallach et al., 1999). Not shown in this diagram but appar-
Figure 17.7 Diagram of TNF-α-receptor interactions leading to various effects within the organism. The receptors occur on virtually all cells throughout the organism, thus accounting for the generalized or- ganismal response resulting from circulating levels of TNF. (Adapted from Bazzoni and Beutler, 1996, with permission of the authors and publisher.)
ently involved in TNF-α induction of apoptosis is the sphingomyelin signaling pathway, which involves lipid intermediates, notably ceramide, and NF-κB translocation to the nucleus (Yang et al., 1993).
Thus, TNF-α has effects that induce both cell replication, presumably through signal transduction pathways to activation of nuclear transcription factors, and apoptosis, via the death domains of the receptor and associated proteins, leading to activation of the apoptotic pathway. It is of interest that in the mouse skin, TNF-α acts as a tumor promoter (Fujiki and Suganuma,1994) and may also serve as an autocrine growth factor for human B lymphocytes (Boussiotis et al., 1994). All of this illustrates the complexity with which these humoral factors exert their effects at a mechanistic level, leading to their part in the induction of cancer cachexia in the whole organism.
Figure 17.8 Diagrammatic representation of the proposed interaction between three members of the TNF receptor superfamily and their associated proteins. Both CD120a (55 kDa) and Fas transduce apop- totic signals via a C-terminal death domain following activation by the ligand. The death domain region, in addition to its function as an effector in apoptosis, also appears to mediate protein-protein interactions re- sulting in nuclear events including NF-κB activation. In contrast, the other TNF receptor, CD120b (75 kDa), recruits a different set of proteins, the TRAF protein, resulting in signal transmission. This in turn transduces nuclear events that may result in the modulation of cellular responses and gene expression. The figure depicts the receptors as monomers for clarity, but it should be remembered that it is their dimeric form that is active. (Modified from Baker and Reddy, 1996, with permission of the authors and publisher.)