The Nature of Intracrine Peptide Hormone Action
Abstract
Abstract—Current theory holds that peptide hormone action results from hormone binding to cell-surface receptors, with the generation of intracellular second messengers. However, a growing body of evidence suggests that intracellular peptide hormone, either internalized or synthesized in situ, can exert physiologically relevant effects. These effects are diverse and poorly understood. I propose that such intracrine action can serve to modulate cellular function over time and thereby play a role in biological memory of various sorts, in the maintenance of hormonal responsiveness, and in cellular differentiation.
Introduction
Our laboratory introduced the term “intracrine” in 1984 to describe the actions of a peptide hormone within its cell of synthesis.1 This concept was later explicitly expanded to include the intracellular action of internalized peptide hormones.2 These formulations were based on our studies of the actions of intracellular angiotensin II.1 2 3 4 5 6 7 8 9 10 11
Subsequently, evidence supporting intracrine action has been developed for a large number of peptide hormones and factors (for example, insulin, growth hormone, prolactin, nerve growth factor, interferon-γ, fibroblast growth factor [FGF], Tat protein, platelet-derived growth factor [PDGF], epidermal growth factor [EGF], parathyroid hormone–related protein [PTHrP], endogenous opiates, angiogenin, among others).12 13 14 15 16 17 18 19 20 21 22 23 24 Although the intracellular actions of steroid hormones and their receptors have been studied intensively, the implications, if any, of intracrine peptide hormone action for normal or abnormal cellular physiology remain unclear.23 Indeed, the default view appears to be that if such intracellular binding/action of peptide hormones exists at all, it represents a vestigial or unimportant aspect of biology.
This view, however, is beginning to change as evidence mounts for the binding of a large number of peptide hormones to nuclei and other intracellular structures and for biological change associated with that binding. It would therefore appear useful at this time to reconsider the possible implications of intracrine peptide action.
By way of review, one can note that peptide hormones have been reported to act intracellularly in the following ways: (1) binding to receptors in the endoplasmic reticulum soon after synthesis and generating second messengers, eg, PDGF-B/v-sis24 25 ; (2) binding to intravesicular receptors after internalization, with subsequent generation of second messengers, eg, EGF in the absence of its membrane-anchoring domain26 ; (3) binding to receptors on the nuclear membrane, with the subsequent generation of second messengers, eg, insulin17 27 28 ; (4) binding to nucleolar components, eg, FGF, PTHrP, angiogenin13 18 20 ; and (5) binding to chromatin, eg, angiotensin II, PTHrP, nerve growth factor (NGF), EGF, and PDGF.3 12 18 21 29
Given these observations, it appears important to determine the actions of intracellular hormones, whether they are synthesized intracellularly or internalized. More important is the determination of the physiological relevance, if any, of these intracrine effects. The possibility that intracrine hormones are homeostatic regulators of the intracellular milieu operating to stabilize the intracellular environment in ways not yet appreciated has been proposed.1 2 11 23 However, it also is possible that considerable complexity could develop in these intracrine systems, leading to the assumption of additional physiological functions. The following additional physiologically relevant intracrine actions of peptide hormones have been proposed or readily come to mind: (1) providing information essential for the transcriptional and other effects of hormone binding to cell-surface receptors (angiogenin20 ); (2) providing information necessary to modulate or “fine-tune” the actions of hormone bound to cell membrane receptors; (3) providing information to regulate the production of signal-transducing elements, such as hormone receptors, nuclear transcription factors, and downstream mediators of hormone action (angiotensin II30 31 ); (4) reinforcing or mitigating (either in intensity or duration) the effects of hormone binding to cell-surface receptors (PTHrH, angiotensin30 31 32 ); (5) regulating nucleolar and/or ribosomal functions and thereby cell functioning (FGF, PTHrP, angiogenin13 20 29 33 ); (6) altering intracellular calcium fluxes so as to generate a physiological signal (angiotensin II34 35 ); (7) causing cell-surface receptors, either alone or in association with ligand, to generate intracellular effects, such as the transport of JAK/STAT transcription factors to the nucleus (interferon-γ36 ); and (8) producing differentiation of some target cells
In fact, there is experimental evidence to support most of these possible modes of action in the case of specific peptide hormones. For example, angiogenin produces the endothelial cell proliferation necessary for angiogenesis only if the hormone reaches the nucleolus.20 Cell-surface binding is insufficient to generate this physiological effect. Thus, this observation suggests perhaps the most robust function for intracrine action proposed thus far: an obligate role in peptide hormone action. The mitogenic activity of another angiogenic growth factor, a modified acidic FGF, displays a similar requirement for nuclear translocation of growth factor.37 Although it does not appear that intracrine action is a required component of all peptide hormone or peptide growth factor action, this possibility cannot be excluded at this time. The example of angiogenin demonstrates that at least in some cases, such a requirement for intracrine action does indeed exist.
In considering the other proposed modes of intracrine action, one can note that EGF, after artificial modification by removal of its membrane-anchoring domain, stimulates the generation of second messengers in cytoplasmic vesicles after hormone internalization. Similarly, intracellular insulin can stimulate nuclear membrane receptors and result in the generation of second messengers.17 28 Intracellular angiotensin II has also been shown to interact with intracellular receptors and generate second messengers.34 35 These observations suggest that in some cases, the intracellular action of a hormone can modulate or amplify a signal at the cell surface, because the intracellular hormone is seen to release the same (or similar) second messengers as binding of hormone to cell-surface receptors.
Yet another reported intracrine action involves the modulation of components of the intracrine system itself. Intracellular angiotensin II, for example, has been reported to upregulate the transcription of components of the renin-angiotensin system itself as well as downstream mediators of angiotensin action. Also, in cardiac myocytes, inhibition of κ-opioid receptor ligand binding to its nuclear receptor by the antagonist dynorphin B is associated with increased opioid gene transcription by these cells. This effect is mediated by nuclear protein kinase C. These actions of both angiotensin and the opioid ligand have the effect of influencing the strength of the respective intracrine system.19 30 31
A more complex intracrine role is illustrated by the action of PTHrP. This protein has been shown to stimulate mitogenesis after binding to vascular smooth muscle cell nuclei, but it inhibits mitogenesis after cell-surface binding.32 Possibly the nuclear action of PTHrP could mitigate, in either time or intensity, the growth-inhibiting action of PTHrP binding to cell-surface receptors. This observation of discordant effects of nuclear as opposed to cell-surface binding also raises the possibility of truly dichotomous functioning of the protein (as opposed to simple mitigation at the nucleus of effects induced at the cell surface), because the protein can apparently be synthesized either in forms destined for extracellular transport or in forms lacking the sequences required for secretion and therefore destined for retention in the cell.
Angiogenin and FGF operate at the nucleolus, and it has recently been suggested that nucleolar physiology can regulate overall cell function.13 20 33 Like extracellularly administered angiogenin and FGF, extracellularly delivered PTHrP migrates to nucleoli. Indeed, readily detectable amounts of PTHrP can be found in the nuclei of some cells even in the absence of external administration of hormone.18 29 This nucleolar hormone probably results from synthesis of forms of the protein lacking a secretory signal and therefore confined to the cell interior. However, secretion and reuptake with subsequent localization of hormone to the nucleolus cannot be excluded. Although these and other peptide hormones are frequently found in nucleoli after extracellular administration, and although arguments have been advanced regarding possible nucleolar effects on cell cycling, ribosomal physiology, and protein synthesis, it is also important to note that in most studies showing nucleolar hormone localization, some chromatin localization is also seen, raising the possibility that a direct effect of internalized or locally synthesized hormone on nonribosomal gene transcription could also be occurring.12 18 29 38
Alteration in calcium transport has been shown to be an important mediator of hormone action at the cell membrane, and similar findings have been reported in relation to intracrine hormone action. Intracellular angiotensin regulates calcium influx via a protein kinase C pathway in cardiac myocytes and also stimulates inward calcium flux in vascular smooth muscle cells.34 35 These findings suggest that agents that affect calcium transport could in some cases influence both cell membrane receptor–mediated and intracellular hormone action.
Recently, a novel role of internalized hormones has been suggested. It has been proposed that some hormones (eg, interferon-γ) use nuclear localization signals on their cell-surface receptors to translocate transcription factors, such as STATs, to the nucleus.36 That is, the internalization of hormone and receptor that is produced by ligand binding appears to be capable of causing the binding of transcription factors to the hormone-receptor complex, with subsequent transport of the entire complex to the nucleus. In this process, putative nuclear localization signals in the hormone receptors are used to effect nuclear translocation of the associated transcription factor. In this way, transcription factors lacking nuclear import signals could be delivered to the nucleus as a consequence of hormone binding to cell-surface receptors. This could provide a mechanism by which internalized hormone could influence gene transcription.
Finally, some evidence for an intracrine hormone role in differentiation is provided by the finding that nuclear angiotensin causes changes in chromatin conformation consistent with gene transcription and/or differentiation.4 Similar findings have been reported for PDGF, NGF, and EGF.12 Of perhaps more relevance, however, is the observation that constitutive expression of a nucleolar-localizing form of PTHrP in a chondrocyte cell line inhibits terminal differentiation and apoptosis.29 Also, stable transfection of keratinocytes with a retroviral vector expressing PTHrP antisense blocks the differentiation of these cells.39 Whether this results from the loss of PTHrP intracrine action, as opposed to cell-surface receptor–mediated effects, is unknown.
One interpretation of these findings is that intracrine peptides can function in a wide variety of ways reminiscent of the actions of hormones at the cell surface. However, one might ask whether there are any underlying principles of intracrine action.
Hypothesis
To address this question, it may be helpful to conceive of the cell interior as one traditionally thinks of the exterior environment. This view may be particularly relevant if the eukaryotic cell owes any part of its existence to the symbiotic integration of simpler cells. According to this formulation, peptides and other factors that originally regulated the functions of earlier prokaryotic cells developed into the peptide systems that now regulate eukaryotic cell function. Intracrine hormones could still serve as intracellular homeostatic regulators in eukaryotic cells.1 2 11 23 Following this line, it would not be surprising to find elements of diverse signaling systems associated with intracellular organelles. On the simplest level, similarities in the membranes associated with the cell surface and intracellular organelles suggest the existence of common signaling systems on each. But there may be other parallels between cells and organelles. In particular, the genomic intracrine actions of peptide hormones could be viewed as functioning like the operon of bacterial cells.40 This view is similar to the idea that orphan steroid hormone receptors are biologically ancient receptors for cellular metabolic ligands that continue to function in the eukaryotic cell.23 To be sure, the eukaryotic cell and its genome are significantly more complex than prokaryotic forms, and eukaryotic signaling systems are also more complex; for example, the simple linked genes of the classic bacterial operon are not found in eukaryotes. Nonetheless, it is possible that intracellular signaling systems have developed in complexity along with cell-surface receptor–based systems, and it is also possible that peptide hormones first functioned in an intracrine fashion before assuming an endocrine/paracrine role. According to the operon concept, regulators of gene function, such as repressors, interact with genomic elements to direct gene expression and secondarily cellular metabolism, with the activity of these regulators determined by environmental or other factors. In a sense, the feedback loops established represent a form of differentiation in that the presence of a repressor at the operon alters the metabolism of the cell as long as the environment does not change. In the case of the eukaryotic cell nucleus (or other intracellular organelles), the “external environment” is the cytoplasm. The cytoplasm represents an environment to which the nucleus responds by altering gene transcription so as to affect that environment. However, because the external environment of these organelles is in large measure homeostatically controlled and indeed is in large part under control of the genome itself, feedback loops, once established, will be long-lived and will mimic other forms of cell differentiation in that they will persist in the face of changes in the extracellular environment. It is here hypothesized that peptide hormones functioning in an intracrine fashion within eukaryotic cells can establish long-lived feedback loops, with the end result that the cellular response to future stimuli is altered because of the intracrine hormone action. It is noteworthy that this form of differentiation is an active and to some degree a plastic or modulatable process. The apparent ability of some intracrine hormones, such as angiotensin II, to regulate transcription of components of their own effector pathways or of their synthetic pathways is consistent with this view. Indeed, it may be that in all cells responsive to a peptide hormone, an intracellular system exists based on the local production of that same hormone so as to control the genome-directed synthesis of the elements involved in the response to that hormone and to mediate long-term hormonal effects. In some cells, the establishment of these intracellular hormone loops may serve a memory function, such that long-term cellular responses to stimuli can be conditioned by all factors that influence the intracrine hormone loop. This cellular memory could be neurological, immunological, or otherwise. However, this form of memory is distinctly characterized by the creation and the continuous modulation of intracellular hormone loops. This is a plastic (modulatable) and active (in the sense that it requires intracellular hormone action) form of cellular differentiation or memory.
Recently, the so-called “emergent properties” of biological signaling systems have come under investigation.41 It is contended that the multiple secondary messengers generated by hormone binding to receptor can interact, with the result that persistently high levels of some mediators are generated by interactions between signaling systems. It is assumed that this behavior can account, at least in part, for the learned behavior of some biological systems. One possible problem with generalizing this formulation is that the nonlinearity in the concentration of secondary messengers necessary to produce these emergent properties appears to be inconsistent with the precision and specificity with which many hormonal systems operate. The localization of mediators to one or another intracellular organelle adds another layer of complexity to the proposed system, but it does not eliminate this concern.42 An intracrine hormone system, however, could serve as the substrate for these emergent properties without grossly interfering with hormone action as usually assessed.
For example, in hepatocyte nuclei, intracrine angiotensin II appears to upregulate angiotensinogen as well as renin gene transcription and in the presence of angiotensin-converting enzyme could cause intracellular angiotensin concentrations to increase.30 31 Thus, exposure of cells to angiotensin would set up a cycle in which internalized angiotensin II exerted an intracrine effect to enhance endogenous production of angiotensin II in a positive-feedback fashion. The effect of this would be to change the set point of the intracrine system such that cellular angiotensin II would remain above basal levels for some time after external angiotensin concentrations had returned to normal. Thus, a memory of the initial angiotensin exposure would reside in the cells. In the case of parathyroid hormone–related protein, variable amounts of hormone have been detected in the nucleoli of cells unexposed to external PTHrP. The frequency with which nucleolar PTHrP is detected in unexposed cells suggests that the PTHrP intracrine system is set at a high but variable level in many cell types. Vascular smooth muscle cells synthesize PTHrP and possess PTHrP cell-surface receptors.32 Binding of hormone to surface receptors inhibits proliferation, whereas intracrine hormone arguably offsets this action by stimulating proliferation. PTHrP-secreting vascular cells whose intracrine PTHrP system is set at high levels could be protected against the antiproliferative effects of PTHrP, whereas target cells lacking such exuberant concentrations of intracellular PTHrP are more susceptible to the antiproliferative effects of the protein. The vascular cells possessing active intracrine systems could be said to be in a different state of responsiveness (resistance) to the hormone or in a differentiated state from other target cells. These examples cannot be precise because the detailed workings of either the angiotensin II or the PTHrP intracrine systems are unknown. These scenarios are offered to sketch in a conceptual way the possible workings of intracrine systems according to the hypothesis proposed here. They lend support to the notion that intracrine hormone action can affect cellular function over time in a fashion similar to receptor regulation and other established mechanisms influencing cellular responsiveness. It is also interesting to note that another level of complexity could be introduced into intracrine hormone signaling by the interaction of intracrine hormones. For example, angiotensin II has been reported to stimulate PTHrP gene expression in cultured vascular smooth muscle cells, and exogenously administered PTHrP has been reported to blunt the mitogenic response of the cells to angiotensin II.43 44 The effect of exogenous PTHrP is most likely mediated by cell-surface receptors, but it is possible that PTHrP synthesized endogenously by the cells operates in an intracrine fashion at the cell nucleus. This possibility and the possible interactions of intracrine PTHrP and intracrine angiotensin are potential areas for further study, as are the possible interactions of other intracrine hormones. Finally, although the hypothesis that intracrine peptide hormone action can produce long-lived alterations in cellular responsiveness has been advanced here, it is also possible that steroid and thyroid hormones and their intracellular receptors could operate in a similar fashion.
If this hypothesis is correct, biologically active hormone (whether synthesized locally or internalized), along with an intracrine signaling system, will most likely be found in every cell responsive to a given peptide hormone. Moreover, the elimination of intracellular hormone (or the elimination of the intracellular action of a hormone) would be expected to lead to altered cellular responsiveness to that hormone (and possibly to others) over time. In other cells, such interruption of intracrine systems could be manifested as loss of memory. It is likely that this idea will soon be testable, given the ongoing development of novel hormone inhibitors, some of which may be effective inside cells. Also of note is the possibility that in certain circumstances, the augmentation of intracellular intracrine loops could restore or improve some forms of biological memory.
Footnotes
-
Reprint requests to Richard Re, MD, Division of Research, Alton Ochsner Medical Foundation, 1516 Jefferson Highway, New Orleans, LA 70121.
- Received April 23, 1999.
- Revision received June 7, 1999.
- Accepted June 14, 1999.
