- Email: [email protected]
Time and space — novel aspects of hormone action
hasdiverse physiological roles
Time and space novel aspects of hormone action THE LAST FOUR years have seen widespread acceptance of the idea that receptor-stimulated polyphosphoinositide hydrolysis generates two intracellular messengers, inositol 1,4,5~sphospha~ [(1,4,5)I~~~ and diacyiglycerol (DG)1*2.A recent meeting* emphasized that identification of the messengers is only the ~ginning of the story. The response to hormone stimulation is often a series of transient oscillations in cytosolic Ca*’ concentrations; this digital coding of the Ca2* signal may be as important as the spatial organization of each part of the intracellular signalling pathway. One very important theme, the temporal aspects of intracellular signalling, was developed by Berridge (Cambridge) and Cobbold (Liverpool). In many cells measurements of intracellular Ca*+, either dire@ly using Ca*+sensitive dyes or indirectly by monitoring the behaviour of Ca2+-sensitive ion channels, have revealed that many hormones trigger oscillations in the cytosolic free Ca*+ concentration -’ oscillations that in some cells are mi,micked by intracellular injection of (1,4,5)IP3 (Ref. 3). The frequency of these oscillations, typically with periods of a few seconds, increases as the concentraticbn of hormone or injected (1,4,5)I& is raise the increased4. These observations lf~~stt~ttipids and trune~embrune s~m~tfi~‘, tht! &!@I s@Ci@, London, 24 December 1987.
TIPS - February 1988Wol. 91
interesting possibility that cells ma read a frequency encoded Ca Y+ signal in much the same way as many neurones relay information by varying the frequency of all-or-nothing action potentials. Such digitally encoded signals are relatively noise free and many therefore provide cells with a more reliable means of responding to changes in hormone concentration. Understanding of how this digital signal is decoded to provide a graded cell response further temporal must await analysis of the later steps in the signalling pathway. One possibility suggested by Berridge is that a transient Ca*+ spike elicits a response (phoslonger-lived phorylation of a protein, for example) which then slowly reverses until the next Ca2+ spike triggers the next response. The steadystate level of phosphoprotein would therefore be a function of the frequency of the Ca*+ oscillations and the cell would be able to mount a sustained graded response to an agonist without needing to maintain a constantly elevated cytoplasmic Cazt activity. In hepatocytes each Ca*+mobilizing agonist produces a characteristically shaped oscillation5. This led Cobbold to suggest that the mechanisms underlying the oscillations must be close to the receptor and may perhaps be a consequence of oscillatory changes in intracellular (1,4,5)IP, concentration. An alternative, though not incompatible, mode1 was proposed by Berridge who suggested the oscillations may arise from periodic discharge of an intracellular Cazt pool without the need for rapid oscillations in (1,4,5)IP, concentration. 0
0
cl
Michell (Birmingham) introduced a second theme that was to recur in subsequent presentations: the spatial organization of each component of the signalling pathway. Berridge and Fain6 working with blowfly salivary glands and, more recently, Monaco’ working with a mammary tumour cell line have shown that only a small fraction of the total cellular phosphatidylinositol 4,5-bisphosphate (I’&) pool is
available to the receptor-stimulated phospholipase C. Michell reported that even in the simple red blood cell there are distinct pools of polyphosphoinositides within the inner leaflet of the plasma membrane. Why do cells maintain multiple pools of phosphoinositides and how, within the same leaflet of the membrane, do they keep the pools apart? Irvine (Babraham) and Hanley (Cambridge) suggested that there must also be compartmentalization within the cytoplasm because inositol pentakisphosphate (I&,) and inositol hexakisphosphate (IE6) are too insoluble to occur freely in the cytosol at the high concentrations found in astrocytoma cells. A novel role for these higher inositol phosphates - as excitatory neurotransmitters - was suggested by the preliminary results of Hanley; such a role suggests that their sequestration in the cytoplasm may perhaps be within secretorv vesicles. q
0
0
A further example of strict spatial organization is the (1,4,5)IPssensitive intracellular Ca2+ pool and here the function seems very clear. Substantial evidence, marshalled by Putneya (NIEHS, North Carolina), suggests that this pool may reside in a fraction of the endoplasmic reticulum that is closely associated with the plasma membrane. Results from Petersen (Liverpool) on mouse lacrymal gland supported earlier work from Irvine’ suggesting that Cazt entry at the plasma membrane requires the cooperative actions of (1,4,5)IPs and its phosphorylated product (1,3,4,5)IPd. By combining these results with the ideas of Putney and the results of GilllO (Maryland) showing that GTE allows CaZt to pass between endoplasmic reticulum Cazt pools that are normally distinct, Irvine suggested a model for receptorregulated Cazt entry. He proposed that (1,3,4,5)1P4, perhaps in concert with GTP, may regulate the relationship between the (1,4,5)IPs-sensitive Ca*+ pool and the plasma membrane such that, when both inositol phosphates are present, Ca*+ can flow from the extracellular space into the endoplasmic reticulum and from there into the cytosol through a channel regulated by (1,4,5)IPs.
The meeting also brought out other novel and exciting work. l On the diacylglycerol limb of the signalling pathway, the identification of a large family of protein kinases C each encoded by distinct genesir and of the further diversity that may come from post-translational modifications of these proteins has prompted speculation on the possible functional differences between these kinases. Recent work from Nishizuka (Kobe) provides some of the answers. He demonstrated that different protein kinases C differ in both their tissue and their subcellular distribution; they differ in their sensitivity to the known activators Ca’+, phosphatidylserine and DG; and they differ too in their substrate specificity.
2 It has long been recognized that lithium ions have teratogenic effects during early embryonic development, and more recently Warner (London) has observed that these effects are manifested at intracellular lithium concentrations similar to those that inhibit phosinositol mono hosphate phatase activity Ip2. This has hinted at a role for phosphoinositide metabolism in regulating morphogenesis. More direct evidence comes from the elegantly simple experiments of Busa (Johns Hopkins). He injected lithium ions into a single cell in the vegetal pole of a 32-cell stag2 Xenopus embryo and found that the tadpole developed abnormally - it had two heads! Remarkably, these dramatic effects of lithium were specifically overcome by coinjection of myo-inositol, suggesting that the teratogenic effects of lithium may be due to inhibition of inositol recycling. While the nature of the long-sought morphogens is stil: obscure, these exciting results suggest that one of their actions may be to regulate phosphoinositide metabolism or perhaps that an inositol phosphate is itself a morphogen. 2 Another surprise came from Whitman (Boston) who found that the product of phosphorylation of phosphatidylinositol (PI) by the complex of type-l PI kinase with middle T/pp60’-~“c is distinct from phosphatidylinositol 4-monophosphate [PI(4)P], the lipid known to be a part of the receptor-regulated
TIPS - February 1988 IVol. 91
45
phosphoinositide cycle. Stevens (Welwyn) has identified the new PIP as PI(3)P. It is not yet known whether PI(3)P can be further phosphorylated or if it is a substrate for hydrolysis by a phospholipase C. In view of the transforming properties of middle T antigen and of mutant src products, the results raise the fascinating possibility that PI(3)P may mark the beginning of a signalling pathway that is specificicallyassociated with control of cell proliferation.
_-_ 17 ri
-. 1 I I r7
COLIN
W. TAYLOR
AFRC Unit of Insect Neurophysiology and Pharmacology, Department of Zoology, Downing Street, Cambridge CB2 3EJ, UK.
References 1 Benidge, M.
J. and Irvine, R. F. (1984) Nature 312,3X-321 2 Nishizuka, Y. (1986) Science 233, 305312 3 Parker, I. and Miledi, R. (1987) Prof. R. Sot. London Ser. 3 232,59-70 4 Woods, N. M., Cuthbertson. K. S. R. and Cobbold, P. H. (1986) Nature 319, 600-602 5 Wood, N. M., Cufhbertson, K. S. and Cobbold, P. H. (1987) Cell Calcium 8,79-
10
353-357
- r-u 2-n-I
Antisense nucleic acids as a potential class of pharmaceuticalagents A !ong-standing goal of pharmacology is to produce rationally designed drugs - drugs which are engineered to selectively interact with or affect a specific cellular target. This is rarely the way drugs have been discovered: serendipity has lent a hand in the development of most pharmaceuti.cal agents. Recent advances in molecular biology suggest that the development of such rationally designed drugs might be at hand. The basis for this optimism is experiments using antisense RNA or DNA molecules. The strand of RNA which is translated into a protein is termed the sense or the plus strand; the strand complementary to the sense strand is the antisense or minus strand. The introduction of a specific antisense RNA into a cell can markedly inhibit the functioning of the sense strand mRNA. The precise mechanism of this inhibition is not clear, but most evidence suggests that intracellular formation of highly (Fig. 1) specific sense-antisense hybrids impairs the translation, stability or transport of the sense mRNA. Several recent studies suggest that this antisense technology could be exploited to develop a new class of pharmaceutical agents. Potentially these drugs could be used as antiviral or anticancer agents for diseases as diverse as AIDS or Burkitt’s lymphoma. As the sequences of additional viral, oncogenie, and regulatory genes and
messenger RNAs become available - currently the amount of sequence data is doubling approximately every 14 months* - the possibilities of employing antisense nucleic acids as therapeutic agents increase as well. Antisense RNAs which act as negative regulators are known to occur in nature. In prokaryotes, DNA replication is regulated in part by an endogenous antisense RNA which is complementary to, and hybridizes with, the short piece of primer RNA that is required for the initiation of DNA synthesit?. In addition, the levels ef porins (major membrane proteins in bacteria which are involved in osmoregulation*) and of the catabolite activator protein3, are regulated by endogenous antisense RNAs. In these organisms the antisense RNAs appear to impede the translation of the sense strand into protein. Endogenous antisense RNAs have not yet been described in eukaryotes but several examples exist where precise complementary base pairing is essential for normal biological function. The simple folding of tRNA and rRNAs depends on intramolecular hybridization; the folded structure of these molecules is essential for their function. The splicing of the primary transcript of most eukaryotic messenger RNAs into mature messenger RNA requires a small nuclear RNA called Ul. The 5’-end of Ul lDufa from Intelli#enetics, Pab Alto.
6 Fain, J. N. and Berridge, M. J. (1979) Biochem. J. 180,655-&l 7 Monaco, M. E. and Woods, D. (1983) J. Biol. Gem. 258,15125-15129 8 Putney, J. W. Jr (1986) Cell Calcium 7, 1-12 9 Irvine, R. F. and Moor, R. M. (1986) Biochem. J. 240,917-920 10 Chueh, F. H., Mullaney, J. M., Chosh, T. K., Zachary, A. L. and Gill, 0. L. (1987) J. Bid. Chem. 262,13857-13864 11 Parker, P. J., Couasens, L., Totiy, N., Rhee, L., Young, S., Chen, E., Stabel, S., Waterfield, M. D. and Ullrich, A. (1986) Science 233.853-859 12 Breckenridge, L. J,, Warren, R. L. and Warner, A. E. (1987) Development 99,
CA.
RNA contains a short sequence which is complementary to the conserved sequences found at the 5’-end of splice junctions4. It appears that the short complementary sequence of Ul RNA hybridizes to the 5’-end of the region to be removed (the intron). Removal of the eight complementary bases from Ul using a nuclease abolishes splicing in vitro5. Intramolecular base pairing may also play a role in the regulation of certain genes. For example, the c-myc gene contains a sequence in the first noncoding exon which can hybridize with complementary sequences in the second exon. The stem and loop structure thus formed may impede the transcription or translation of c-myc mRNA6. That antisense DNA or RNA can impair the expression of a specific target mRNA is now well established (reviewed in Refs 7 and 8). Two approaches have been taken to ‘deliver’ antisense nucleic acids into cells. Antisense DNA or RNA can be introduced directly into cells by microinjectiong*’ . Alternatively, DNA molecules which will produce antisense RNAs can be introduced into the process of target cells b
transfection7*g,1Y*lz.In this case the antisense RNA is synthesized from the transfected DNA within the target cell. Izant and Wein-
traub microinjected or transfected cells with a viral thymidine kinase (vTK) gene which produced antisense vTK mRNA when expressed in recipient cells9. Expression of
the antisense vTK gene markedly decreased the activity of viral thymidine kinase. Additional experiments established the specificity of the system: expression of the anti-vTK gene inhibited viral TK
0 1988,Ebcvler Publlcatlon~, CambrIdge 0169 - 6147/8U/W2.00
Time and space novel aspects of hormone action THE LAST FOUR years have seen widespread acceptance of the idea that receptor-stimulated polyphosphoinositide hydrolysis generates two intracellular messengers, inositol 1,4,5~sphospha~ [(1,4,5)I~~~ and diacyiglycerol (DG)1*2.A recent meeting* emphasized that identification of the messengers is only the ~ginning of the story. The response to hormone stimulation is often a series of transient oscillations in cytosolic Ca*’ concentrations; this digital coding of the Ca2* signal may be as important as the spatial organization of each part of the intracellular signalling pathway. One very important theme, the temporal aspects of intracellular signalling, was developed by Berridge (Cambridge) and Cobbold (Liverpool). In many cells measurements of intracellular Ca*+, either dire@ly using Ca*+sensitive dyes or indirectly by monitoring the behaviour of Ca2+-sensitive ion channels, have revealed that many hormones trigger oscillations in the cytosolic free Ca*+ concentration -’ oscillations that in some cells are mi,micked by intracellular injection of (1,4,5)IP3 (Ref. 3). The frequency of these oscillations, typically with periods of a few seconds, increases as the concentraticbn of hormone or injected (1,4,5)I& is raise the increased4. These observations lf~~stt~ttipids and trune~embrune s~m~tfi~‘, tht! &!@I s@Ci@, London, 24 December 1987.
TIPS - February 1988Wol. 91
interesting possibility that cells ma read a frequency encoded Ca Y+ signal in much the same way as many neurones relay information by varying the frequency of all-or-nothing action potentials. Such digitally encoded signals are relatively noise free and many therefore provide cells with a more reliable means of responding to changes in hormone concentration. Understanding of how this digital signal is decoded to provide a graded cell response further temporal must await analysis of the later steps in the signalling pathway. One possibility suggested by Berridge is that a transient Ca*+ spike elicits a response (phoslonger-lived phorylation of a protein, for example) which then slowly reverses until the next Ca2+ spike triggers the next response. The steadystate level of phosphoprotein would therefore be a function of the frequency of the Ca*+ oscillations and the cell would be able to mount a sustained graded response to an agonist without needing to maintain a constantly elevated cytoplasmic Cazt activity. In hepatocytes each Ca*+mobilizing agonist produces a characteristically shaped oscillation5. This led Cobbold to suggest that the mechanisms underlying the oscillations must be close to the receptor and may perhaps be a consequence of oscillatory changes in intracellular (1,4,5)IP, concentration. An alternative, though not incompatible, mode1 was proposed by Berridge who suggested the oscillations may arise from periodic discharge of an intracellular Cazt pool without the need for rapid oscillations in (1,4,5)IP, concentration. 0
0
cl
Michell (Birmingham) introduced a second theme that was to recur in subsequent presentations: the spatial organization of each component of the signalling pathway. Berridge and Fain6 working with blowfly salivary glands and, more recently, Monaco’ working with a mammary tumour cell line have shown that only a small fraction of the total cellular phosphatidylinositol 4,5-bisphosphate (I’&) pool is
available to the receptor-stimulated phospholipase C. Michell reported that even in the simple red blood cell there are distinct pools of polyphosphoinositides within the inner leaflet of the plasma membrane. Why do cells maintain multiple pools of phosphoinositides and how, within the same leaflet of the membrane, do they keep the pools apart? Irvine (Babraham) and Hanley (Cambridge) suggested that there must also be compartmentalization within the cytoplasm because inositol pentakisphosphate (I&,) and inositol hexakisphosphate (IE6) are too insoluble to occur freely in the cytosol at the high concentrations found in astrocytoma cells. A novel role for these higher inositol phosphates - as excitatory neurotransmitters - was suggested by the preliminary results of Hanley; such a role suggests that their sequestration in the cytoplasm may perhaps be within secretorv vesicles. q
0
0
A further example of strict spatial organization is the (1,4,5)IPssensitive intracellular Ca2+ pool and here the function seems very clear. Substantial evidence, marshalled by Putneya (NIEHS, North Carolina), suggests that this pool may reside in a fraction of the endoplasmic reticulum that is closely associated with the plasma membrane. Results from Petersen (Liverpool) on mouse lacrymal gland supported earlier work from Irvine’ suggesting that Cazt entry at the plasma membrane requires the cooperative actions of (1,4,5)IPs and its phosphorylated product (1,3,4,5)IPd. By combining these results with the ideas of Putney and the results of GilllO (Maryland) showing that GTE allows CaZt to pass between endoplasmic reticulum Cazt pools that are normally distinct, Irvine suggested a model for receptorregulated Cazt entry. He proposed that (1,3,4,5)1P4, perhaps in concert with GTP, may regulate the relationship between the (1,4,5)IPs-sensitive Ca*+ pool and the plasma membrane such that, when both inositol phosphates are present, Ca*+ can flow from the extracellular space into the endoplasmic reticulum and from there into the cytosol through a channel regulated by (1,4,5)IPs.
The meeting also brought out other novel and exciting work. l On the diacylglycerol limb of the signalling pathway, the identification of a large family of protein kinases C each encoded by distinct genesir and of the further diversity that may come from post-translational modifications of these proteins has prompted speculation on the possible functional differences between these kinases. Recent work from Nishizuka (Kobe) provides some of the answers. He demonstrated that different protein kinases C differ in both their tissue and their subcellular distribution; they differ in their sensitivity to the known activators Ca’+, phosphatidylserine and DG; and they differ too in their substrate specificity.
2 It has long been recognized that lithium ions have teratogenic effects during early embryonic development, and more recently Warner (London) has observed that these effects are manifested at intracellular lithium concentrations similar to those that inhibit phosinositol mono hosphate phatase activity Ip2. This has hinted at a role for phosphoinositide metabolism in regulating morphogenesis. More direct evidence comes from the elegantly simple experiments of Busa (Johns Hopkins). He injected lithium ions into a single cell in the vegetal pole of a 32-cell stag2 Xenopus embryo and found that the tadpole developed abnormally - it had two heads! Remarkably, these dramatic effects of lithium were specifically overcome by coinjection of myo-inositol, suggesting that the teratogenic effects of lithium may be due to inhibition of inositol recycling. While the nature of the long-sought morphogens is stil: obscure, these exciting results suggest that one of their actions may be to regulate phosphoinositide metabolism or perhaps that an inositol phosphate is itself a morphogen. 2 Another surprise came from Whitman (Boston) who found that the product of phosphorylation of phosphatidylinositol (PI) by the complex of type-l PI kinase with middle T/pp60’-~“c is distinct from phosphatidylinositol 4-monophosphate [PI(4)P], the lipid known to be a part of the receptor-regulated
TIPS - February 1988 IVol. 91
45
phosphoinositide cycle. Stevens (Welwyn) has identified the new PIP as PI(3)P. It is not yet known whether PI(3)P can be further phosphorylated or if it is a substrate for hydrolysis by a phospholipase C. In view of the transforming properties of middle T antigen and of mutant src products, the results raise the fascinating possibility that PI(3)P may mark the beginning of a signalling pathway that is specificicallyassociated with control of cell proliferation.
_-_ 17 ri
-. 1 I I r7
COLIN
W. TAYLOR
AFRC Unit of Insect Neurophysiology and Pharmacology, Department of Zoology, Downing Street, Cambridge CB2 3EJ, UK.
References 1 Benidge, M.
J. and Irvine, R. F. (1984) Nature 312,3X-321 2 Nishizuka, Y. (1986) Science 233, 305312 3 Parker, I. and Miledi, R. (1987) Prof. R. Sot. London Ser. 3 232,59-70 4 Woods, N. M., Cuthbertson. K. S. R. and Cobbold, P. H. (1986) Nature 319, 600-602 5 Wood, N. M., Cufhbertson, K. S. and Cobbold, P. H. (1987) Cell Calcium 8,79-
10
353-357
- r-u 2-n-I
Antisense nucleic acids as a potential class of pharmaceuticalagents A !ong-standing goal of pharmacology is to produce rationally designed drugs - drugs which are engineered to selectively interact with or affect a specific cellular target. This is rarely the way drugs have been discovered: serendipity has lent a hand in the development of most pharmaceuti.cal agents. Recent advances in molecular biology suggest that the development of such rationally designed drugs might be at hand. The basis for this optimism is experiments using antisense RNA or DNA molecules. The strand of RNA which is translated into a protein is termed the sense or the plus strand; the strand complementary to the sense strand is the antisense or minus strand. The introduction of a specific antisense RNA into a cell can markedly inhibit the functioning of the sense strand mRNA. The precise mechanism of this inhibition is not clear, but most evidence suggests that intracellular formation of highly (Fig. 1) specific sense-antisense hybrids impairs the translation, stability or transport of the sense mRNA. Several recent studies suggest that this antisense technology could be exploited to develop a new class of pharmaceutical agents. Potentially these drugs could be used as antiviral or anticancer agents for diseases as diverse as AIDS or Burkitt’s lymphoma. As the sequences of additional viral, oncogenie, and regulatory genes and
messenger RNAs become available - currently the amount of sequence data is doubling approximately every 14 months* - the possibilities of employing antisense nucleic acids as therapeutic agents increase as well. Antisense RNAs which act as negative regulators are known to occur in nature. In prokaryotes, DNA replication is regulated in part by an endogenous antisense RNA which is complementary to, and hybridizes with, the short piece of primer RNA that is required for the initiation of DNA synthesit?. In addition, the levels ef porins (major membrane proteins in bacteria which are involved in osmoregulation*) and of the catabolite activator protein3, are regulated by endogenous antisense RNAs. In these organisms the antisense RNAs appear to impede the translation of the sense strand into protein. Endogenous antisense RNAs have not yet been described in eukaryotes but several examples exist where precise complementary base pairing is essential for normal biological function. The simple folding of tRNA and rRNAs depends on intramolecular hybridization; the folded structure of these molecules is essential for their function. The splicing of the primary transcript of most eukaryotic messenger RNAs into mature messenger RNA requires a small nuclear RNA called Ul. The 5’-end of Ul lDufa from Intelli#enetics, Pab Alto.
6 Fain, J. N. and Berridge, M. J. (1979) Biochem. J. 180,655-&l 7 Monaco, M. E. and Woods, D. (1983) J. Biol. Gem. 258,15125-15129 8 Putney, J. W. Jr (1986) Cell Calcium 7, 1-12 9 Irvine, R. F. and Moor, R. M. (1986) Biochem. J. 240,917-920 10 Chueh, F. H., Mullaney, J. M., Chosh, T. K., Zachary, A. L. and Gill, 0. L. (1987) J. Bid. Chem. 262,13857-13864 11 Parker, P. J., Couasens, L., Totiy, N., Rhee, L., Young, S., Chen, E., Stabel, S., Waterfield, M. D. and Ullrich, A. (1986) Science 233.853-859 12 Breckenridge, L. J,, Warren, R. L. and Warner, A. E. (1987) Development 99,
CA.
RNA contains a short sequence which is complementary to the conserved sequences found at the 5’-end of splice junctions4. It appears that the short complementary sequence of Ul RNA hybridizes to the 5’-end of the region to be removed (the intron). Removal of the eight complementary bases from Ul using a nuclease abolishes splicing in vitro5. Intramolecular base pairing may also play a role in the regulation of certain genes. For example, the c-myc gene contains a sequence in the first noncoding exon which can hybridize with complementary sequences in the second exon. The stem and loop structure thus formed may impede the transcription or translation of c-myc mRNA6. That antisense DNA or RNA can impair the expression of a specific target mRNA is now well established (reviewed in Refs 7 and 8). Two approaches have been taken to ‘deliver’ antisense nucleic acids into cells. Antisense DNA or RNA can be introduced directly into cells by microinjectiong*’ . Alternatively, DNA molecules which will produce antisense RNAs can be introduced into the process of target cells b
transfection7*g,1Y*lz.In this case the antisense RNA is synthesized from the transfected DNA within the target cell. Izant and Wein-
traub microinjected or transfected cells with a viral thymidine kinase (vTK) gene which produced antisense vTK mRNA when expressed in recipient cells9. Expression of
the antisense vTK gene markedly decreased the activity of viral thymidine kinase. Additional experiments established the specificity of the system: expression of the anti-vTK gene inhibited viral TK
0 1988,Ebcvler Publlcatlon~, CambrIdge 0169 - 6147/8U/W2.00