Twelve years ago I believed that the enigma of renal scarring accompanying glomerulonephritis was soon to be solved. Border et al. [1] demonstrated that experimental nephritis could be suppressed by antiserum against transforming growth factor (TGF) β1. The rats in their study were given an antibody directed against Thy-1, an antigen present on mesangial cells, into a tail vein. The resultant mesangial cell injury initiates a model quite reminiscent of mesangio-proliferative glomerulonephritis in humans. The condition is subacute in nature and the animals can recover. Border et al. showed in that study that TGF-β1, an inducer of extracellular matrix production, is increased in the model. Furthermore, the administration of anti-TGF-β1 antibody at the time of disease induction suppressed the increase in extracellular matrix production. In a follow-up study Border et al. [2] studied decorin, a natural inhibitor of TGF-β1 that also protects against scarring in the same model of rat glomerulonephritis. Decorin is a proteoglycan that is induced by TGF-β, and that can neutralize its biological activity [3]. I had assumed that by bottling decorin and marketing the material, a handy treatment for chronic renal failure would soon be at hand. I erred.

A decade, and much bewilderment, later I remain confused. I have suffered through "mothers against decapentaplegic," so-of-a (you guessed it), and signaling pathways that make Theseus' job in the labyrinth on Crete appear a trivial exercise. Actually, the term Smad comes from an amalgamation of the "Mad"-ness of the Drosophila researchers and the Sma protein discovered by Caenorhabditis elegans researchers, the respective homolog in the worm. In this labyrinth there is not merely one Minotaur. Instead, the beasts reside at every corner. Therefore, I laid anti-Thy1 glomerulonephritis and TGF-β whatever to rest and moved to other things. The contribution to Journal of Molecular Medicine by Hartner et al. [4] have rekindled my interest in this topic.

Hartner et al. [4] studied anti-Thy1 nephritis in rats and performed a descriptive study of TGF-β1, TGF-β2, and TGF-β receptors I, II, and III (yes, unfortunately there are three receptors). I would have thought that all these tedious details would have been long worked out, particularly since the disease can be cured (see above), but this proved not to be the case. They found that TGF-β2 (not TGF-β1) expression increased in mesangial cells and podocytes. Western blotting showed a tenfold upregulation early in the course of the disease. In cultured mesangial cells TGF-β1 and TGF-β2 had similarly potent effects in terms of stimulating nuclear Smad 2/3 accumulation, inhibiting DNA synthesis, and inducing β1-integrin or type I collagen synthesis. The receptor localization studies showed that the TGF-β receptor I expression is increased remarkably by day 6 in the model. The TGF-β receptor II was also strongly enhanced during the course of the disease, while the TGF-β receptor III was upregulated only around day 6. The authors have pretty good circumstantial evidence that the TGF-β machinery is upregulated in the model, confirming what we knew 12 years ago. They have supplied additional details, regarding TGF-β2. Earlier studies implicated TGF-β1, but perhaps the antibodies able to distinguish between these isoforms have improved. Does the study shed additional light on mechanisms or treatment possibilities? Unfortunately, the answer is no.

Has our knowledge of TGF-β signaling increased in terms of understanding this and other models? The answer to that question is affirmative. Böttinger and Bitzer [5] have prepared a scholarly review on TGF-β signaling in renal disease. They reviewed the Smad family of signaling proteins, the TGF-β/Smad signaling axis, bone morphogenic protein receptor kinases, and the actions of phospho-Smads. With the help of homology domains and linkers, we finally get around to gene transcription, fascinating reading to say the least. In the nucleus the action finally steps up. I counted 30 transcription factors that participate to activate at least 25 genes, most of which were unfamiliar to me. I hope to get around to these target genes, especially SLUG. Slug is a snail member family. I should have known, since slugs are snails without a shell. SLUG is a down-stream target of myoD. I spare you the list of coactivators, corepressors, and Smad binding elements. Böttinger and Bitzger then introduced me to the notion of epithelial-mesenchymal transition. TGF-β is thereby able to influence desmosomal disassembly, cell-matrix adhesion remodeling, and activation of progenitor cell factors. In these processes other signaling molecules, including MEK/ERK, RoaA, PI3 K, and Akt are involved, to name just a few. TGF-β therefore participates in entire signaling networks. These networks can also regulate and lead to apoptosis.

If decorin is not the therapeutic answer, are there other avenues? Chen et al. [6] showed that a Smad itself can block the effects of TGF-β on mesangial cells. As eloquently explained by Schnaper et al. [7], the Smads are divided into various categories. There are receptor-activated Smads that respond to TGF-β or other ligands. The ligands bind to the TGF-β II receptor that associates with and phosphorlyates the TGF-β I receptor that in turn phosphorylates the receptor-activated or "R" Smads. These Smads then assume an activated form and form a heteromultimer with various molecules including a second type of Smad, the Co-Smad. This complex then translocates into the nucleus (Fig. 1). The third Smad category comprises the inhibitory "I" Smads that are homologues of "R" Smads but that lack the carboxy-terminal sequences essential for activation. Chen et al. [6] raised the possibility that the "I" Smad7 is overexpressed to dampen TGF-β signaling. Along this vein, Bitzer et al. [8] showed that NF-κB, when activated, can inhibit TGF-β-induced phosphorylation, nuclear translocation, and DNA binding. This NF-κB induced inhibition comes about by the up-regulation of Smad7. However, the activation of NF-κB through proinflammatory cytokines or whatever mechanism would not appear to be a good method to inhibit renal scarring. Every inflammatory renal disease to date has shown evidence of NF-kB activation. Perhaps that genie should be left in the bottle.

Fig. 1.
figure 1

Occupation of the TGF-β RII results in phosphorylation of the TGF-β I receptor and phosphorylation of the "R" Smads, as shown, as well as the "Co" Smad4. The heteromultimer complex can translocate into the nucleus. However, "I" Smads, such as Smad7 can inhibit the process and thereby dampen TGF-β signaling. Smad7 can also inhibit NF-κB activation. TGF-β target genes include a litany of target genes, including MAD, SLUG, and TGIF (I wish it were). Far better information is available in [5, 7, 11]

Full size image

TGF-β regulates a plethora of processes that include cancer. In this sense TGF-β is a tumor suppressor. TGF-β1 deficient mice show an increased incidence of chemically induced and spontaneous tumors. A dominant negative construct targeted to breast or skin epithelium also enhances tumorigenicity. However, before we conclude that TGF-β inhibition might facilitate cancer spread, we should look at some recent evidence. Muraoka et al. [9] inhibited TGF-β by means of a soluble Fc:TGF-β type II receptor fusion protein to test the breast tumor hypothesis. The fusion protein TGF-β inhibitor reduced pulmonary metastasis in their model tenfold. Interestingly, the primary tumor cells were not affected. In a similar study Yang et al. [10] examined transgenic mice that stably express soluble Fc:TGF-β RII. These mice, when subjected to melanoma, also showed reduced pulmonary metastasis. Similarly, the primary tumors were unchanged. Thus, although diminished TGF-β activity may promote tumor development, inhibition of TGF-β signaling seems to reduce metastasis. Suppressing a tumor suppressor thus appears to make some sense [11]. As the current contribution suggests, this complicated signaling pathway will occupy us long into the future. The therapeutic consequences are potentially great, but nevertheless remain uncertain.

Respectfully,

Friedrich C. Luft