Search

Monolithic silicon photonic technology provides high bandwidth and large volume for datacom applications, while maintaining the low-cost assembly and waferscale test readiness characteristic of CMOS. Monolithic integration and high-yield CMOS processing enables complex, high-performance WDM architectures for CWDM4 transceivers.

Silicon photonics is rapidly becoming the key enabler for meeting the future data speed and volume required by the Internet of Things. A stable manufacturing process is needed to deliver cost and yield expectations to the technology marketplace. We present the key challenges and technical results from both 200mm and 300mm facilities for a silicon photonics fabrication process which includes monolithic integration with CMOS. This includes waveguide patterning, optical proximity correction for photonic devices, silicon thickness uniformity and thick material patterning for passive fiber to waveguide alignment. The device and process metrics show that the transfer of the silicon photonics process from 200mm to 300mm will provide a stable high volume manufacturing platform for silicon photonics designs.

We present a monolithic segmented Mach-Zehnder transmitter, fully integrated in a 90 nm CMOS process. The transmitter exhibits a link sensitivity of −14.1 dBm optical modulation amplitude (OMA) at bit error rate (BER) = 10−-12 at 12.5 Gbps.

CMOS Integrated Nano Photonics (CINPs) hold promise for bringing semiconductor industry efficiency and dramatic cost reductions to optical interconnect applications [1]. IBM has developed a CINP technology (TX) (CMOS9WG) in a sub-100nm CMOS platform where electronic and photonic components are monolithically integrated into the CMOS circuit analog and mixed-signal frontend. This minimizes processing steps, mask levels, component parasitics, and packaging/assembly complexity, and is designed to enabled multi-channel short reach optical interconnects up to 25 Gbaud rates [2]. Hybrid technology solutions also exist, where electrical and optical components are on separate chips [3]. Hybrid and monolithic platforms each offer unique advantages. However, the monolithic platform has the distinct advantage of being able to address both the hybrid and monolithic approaches.

We propose a flexible optical transmitter for the short-reach optical interconnects that includes a silicon photonic segmented Mach–Zehnder modulator driven by a distributed six-channel 32-nm silicon on insulator (SOI) complementary metal-oxide semiconductor driver integrated circuit. Optical equalization is demonstrated to extend the bandwidth limitation of the transmitter with non-return to zero signaling at ${\text{25 Gb/ s}}$ . We also generate four-level pulse amplitude modulation (PAM-4) signaling using the same transmitter architecture. Transmission of 46 Gb/s PAM-4 signal with bit error rate (BER) well below hard-decision forward error correction limit is experimentally demonstrated. Low driver power consumption of 130 mW at ${\text{46 Gb/ s}}$ PAM-4, corresponding to 2.8 pJ/bit power efficiency, is also achieved.

Monolithic CMOS photonics seeks to minimize total transceiver cost by simplifying packaging, design and test. Here we examine 25 Gb/s applications in the context of integrated transistor performance and demonstrate a 4λ×25 Gb/s reference design.

We present a novel silicon photonic transmitter including 90nm CMOS segmented modulator co-packaged with low power 32nm CMOS driver IC. Optical equalization is demonstrated for the first time with the multi-segment Mach-Zehnder modulator at 22Gb/s.

A manufacturable platform of CMOS, RF and opto-electronic devices fully PDK enabled to demonstrate a 4×25 Gb/s reference design is presented. With self-aligned fiber attach, this technology enables low-cost O-band data-com transceivers. In addition, this technology can offer enhanced performance and yield in hybrid-assembly for applications at 25 Gbaud and beyond.

A novel high-speed Mach-Zehnder modulator (MZM) fully integrated into a 90 nm CMOS process is presented. The MZM features 'double-pass' optical phase shifter segments, and the first use of integrated inductors in a 'velocity-matched' distributed-electrode configuration.

The traditional Mach-Zehnder modulator (MZM) figure of merit (FOM) has been defined as (Vπ2)/υ3dBe, and works effectively for LiNbO3 long haul modulators. However, for plasma dispersion based electro-optic modulators, or any modulator that has an inherent relationship between its bandwidth, required drive voltage, and optical insertion loss/gain, this FOM is inappropriate. This is particularly true for short reach links with no optical amplification. In the following, we propose a new modulator FOM (M-FOM) based on device metrics that are essential for short-reach links, such as the peak-to-peak drive voltage, modulator rise-fall time, and relative optical modulation amplitude. Link sensitivity measurements from two MZMs that have different bandwidths and optical losses are compared using our M-FOM to demonstrate its utility. Furthermore, we present a novel application protocol of our M-FOM to provide deeper insight into the relative system impact that modulator performance has on data links with no optical amplification, by taking the ratio of M-FOMs from two modulators driven with the same radio frequency drive power.

A silicon microring modulator with independent resonance and coupling tunability is monolithically integrated with a stacked CMOS driver. At 25Gbps, the OMA is -1.9dBm, the extinction ratio is 3.5dB, and the excess loss is 2.6dB.

Summary form only given. This talk will give a general overview of the IBM Silicon Photonics program and specifically discuss CMOS compatible traveling wave electro-optic modulator design. A Non-Return-to-Zero Transmitter-link penalty calculation protocol for Mach-Zehnder Interferometric modulators based on the phase shifter efficiency-loss figure-of-merit will be presented. Our Transmitter-link penalty analysis protocol allows one to easily assess an expected penalty estimation from only the RF Vpp drive, modulator efficiency loss FOM, and the assumption that transmitter bandwidth is sufficient to support the link data rate, which allows system designers to better understand how device-level performance metrics impact system link budget.

A mid-infrared silicon nanophotonic integrated circuit platform can have broad impact upon environmental monitoring, personalized healthcare, and public safety applications. Development of various mid-IR components, including optical parametric amplifiers, sources, modulators, and detectors, is reviewed.

A brief overview of the IBM Silicon Photonics program will be given. Mach-Zehnder modulator linearization techniques as well as other transmitter focused issues will be presented. Article not available.

We demonstrate the first mid-IR silicon nanophotonic modulator, with a bitrate of 3Gbps. Our results show that similar design methodologies can be applied to silicon mid-IR active devices as for telecom devices.

Microring-based silicon electro-optic modulators allow high-speed and low-power devices to be fabricated in small footprints. A 30 Gbps reverse-biased microring modulator, and an 8 Gbps forward-biased microring modulator with hybrid-integrated preemphasis driver, are presented.

Pseudomonas aeruginosa exoenzyme S ADP-ribosylates p21ras and several related proteins. ADP-ribosylation of p21ras does not alter interactions with guanine nucleotides. The ras-related GTP-binding proteins, including Rab3, Rab4, Ral, Rap1A, and Rap2, are also substrates; given these results, we propose a model for the role of exoenzyme S in pathogenesis.

Clostridium botulinum type C and D strains produce exoenzyme C3, which ADP-ribosylates the Rho protein, a 21-kDa regulatory GTP-binding protein. In a previous work, we demonstrated that the C3 gene is encoded by bacteriophages C and D of C. botulinum by using DNA-DNA hybridizations with oligonucleotides deduced from the C3 protein N-terminal sequence. The C3 coding gene was cloned and sequenced, but its upstream DNA region could not be studied because of its instability in Escherichia coli. In this work, the upstream DNA region of the C3 gene was directly amplified by the polymerase chain reaction and sequenced. The C3 gene encodes a polypeptide of 251 amino acids (27,823 Da) consisting of a 40-amino-acid signal peptide and a mature protein of 211 amino acids (23,546 Da). The C3 mature protein was expressed in E. coli under the control of the trc promoter. The recombinant polypeptide obtained was recognized by C3 antibodies and ADP-ribosylated the Rho protein. The C3 gene nucleotide sequence is identical on C and D phage DNAs. At the amino acid sequence level, no similarity was found among C3, other ADP-ribosylating toxins, or tetanus or botulinal A, C1, and D neurotoxins.

Pseudomonas aeruginosa exoenzyme S ADP-ribosylates several GTP-binding proteins of apparent Mr = 23,000-25,000. Exoenzyme S absolutely requires a soluble eukaryotic protein, which we have named FAS (Factor Activating exoenzyme S), in order to ADP-ribosylate all substrates. The rate of ADP-ribosylation of all exoenzyme S substrates increases linearly with time and with the FAS concentration. FAS is wide-spread in eukaryotes but appears to be absent from prokaryotes. We have estimated the molecular mass of the protein to be approximately 29,000 daltons and its pI to be 4.3-4.5. Several bacterial toxins share this sort of requirement for the presence of a eukaryotic protein for enzymic activity. In particular, FAS resembles ADP-ribosylation factor, a 21,000-dalton GTP-binding protein which performs an analogous function for cholera toxin. However, we can find no evidence that FAS binds GTP. In the presence of FAS, exoenzyme S ADP-ribosylates several proteins in lysates of P. aeruginosa. The requirement for a eukaryotic protein for enzymic activity, which is common to several bacterial toxins, may be a device to identify the eukaryotic environment and to ensure that the enzymes cannot function within and harm the toxin-producing bacteria.

Cholera toxin and other heat-labile enterotoxins have the same subunit structure (A5B) and all catalyse the mono ADP-ribosylation of Ns, a regulator of adenylate cyclase, probably at an arginine residue. They also ADP-ribosylate a variety of other membrane and soluble proteins at much slower rates. The rates differ from protein to protein but it may be that every arginine residue in every protein is ADP-ribosylated at some slow rate. A guanine nucleotide triphosphate is required for the ADP-ribosylation of the major (Ns) and minor substrates alike. It used to be thought that all the substrates were GTP-binding proteins but this cannot be so. Rather, the GTP is required because it has to bind to some additional site on the membrane, termed 'S', in a cooperative event that involves a soluble protein called cytosolic factor (CF). If we expose erythrocyte membranes to CF and the GTP analogue Gpp(NH)p we can later extract in detergent a factor or complex that confers upon naive erythrocyte membranes the ability to be ADP-ribosylated. Pertussis toxin also has an A5B structure and acts on an intracellular substrate for ADP-ribosylation, namely the negative regulator of adenylate cyclase, called Ni. ADP-ribosylation prevents the reduction of cyclase activity by inhibitory hormones. The ADP-ribosylation of Ns or Ni does not affect the rate of ADP-ribosylation of the other protein.

The recognition of double-stranded DNA breaks and single-stranded nicks by human poly(ADP-ribose) polymerase and the consequent enzymic activation were examined using derivatives of the enzyme expressed in Escherichia coli. The N-terminal 162 residues encompass two zinc fingers. Deletion or mutation of the first finger results in a loss of activation by DNA with either single-stranded or double-stranded damage. Destruction of the second finger reduces activation by double-stranded DNA breaks only slightly, but eliminates activation by single-stranded DNA nicks. These data suggest that activation by single-stranded DNA nicks requires two zinc fingers, but activation by double-stranded DNA breaks requires only the finger closer to the N terminus. Variant proteins that lack both zinc fingers are enzymically inactive but still exhibit weak DNA binding, which is independent of DNA damage. Thus, other regions are also capable of binding intact DNA, but the recognition of a strand nick or break which occasions the synthesis of poly(ADP-ribose) specifically requires the zinc fingers.

LiNbO3 is an active host medium for guided wave electro-optics and has been the basis for the largest, most sophisticated integrated optic structures to date.

Optical amplification has been demonstrated in waveguides fabricated in Nd-doped LiNbO3 grown from a melt1 and in planar diffused Er: LiNbO3.2 Such substrates are, however, not compatible with guided wave integrated optics since absorption will occur in every section of the waveguide which is not optically pumped to transparency. Lateral confinement of the gain medium is also necessary to avoid absorption of the evanescent field. We report a method for local incorporation of Er3+ which significantly enhances the diffusivity of the rare earth ion and is compatible with standard Ti: LiNbO3 photolithographic processing.

Pseudomonas aeruginosa exoenzyme S ADP-ribosylates several GTP-binding proteins of apparent Mr = 23,000-25,000. Exoenzyme S absolutely requires a soluble eukaryotic protein, which we have named FAS (Factor Activating exoenzyme S), in order to ADP-ribosylate all substrates. The rate of ADP-ribosylation of all exoenzyme S substrates increases linearly with time and with the FAS concentration. FAS is wide-spread in eukaryotes but appears to be absent from prokaryotes. We have estimated the molecular mass of the protein to be approximately 29,000 daltons and its pI to be 4.3-4.5. Several bacterial toxins share this sort of requirement for the presence of a eukaryotic protein for enzymic activity. In particular, FAS resembles ADP-ribosylation factor, a 21,000-dalton GTP-binding protein which performs an analogous function for cholera toxin. However, we can find no evidence that FAS binds GTP. In the presence of FAS, exoenzyme S ADP-ribosylates several proteins in lysates of P. aeruginosa. The requirement for a eukaryotic protein for enzymic activity, which is common to several bacterial toxins, may be a device to identify the eukaryotic environment and to ensure that the enzymes cannot function within and harm the toxin-producing bacteria.

There are two functionally and physically distinct types of guanyl nucleotide site associated with the adenylate cyclase system of pigeon erythrocytes. One is on the well known regulatory protein, N, that mediates the adenylate cyclase response to hormones, guanyl nucleotides and fluoride, and is the substrate for ADP-ribosylation by cholera toxin. We now describe a second site that must be occupied by GTP or an analog of GTP before N can be ADP-ribosylated. We call this second site S. It differs from the site on N in many respects. GTP appears to be rapidly hydrolyzed when it is bound to N but not when bound at S. GTP analogs such as guanosine 5'-O-(3-thiotriphosphate) (GTP gamma S) bind stably to both sites but the binding of GTP gamma S to N is more sensitive to EDTA and is more easily prevented by guanosine 5'-O-(2-thiodiphosphate). The nucleotide binding only to S is promoted by the cytosolic protein required by cholera toxin. Isoproterenol decreases GTP gamma S binding to S while indirectly increasing GTP gamma S binding to N. By adjusting the binding conditions, the nucleotides bound functionally to N and S can be varied independently and then the effect of ADP-ribosylation upon the adenylate cyclase activity can be seen to depend on the type of nucleotide bound to N. This activity rises, falls slightly, or remains at zero, if N is occupied by GTP, GTP gamma S, or guanosine 5'-O-(2-thiodiphosphate, respectively.

The ability of DNA is support poly(ADP-ribose) synthesis is completely dependent upon the number and type of strand breaks it contains and is independent of the sequence. Single-stranded DNA is ineffective. Covalently closed circular plasmid DNA is ineffective, but when it is enzymatically digested it activates poly(ADP-ribose) polymerase in proportion to the number of strand breaks, suggesting that the polymerase recognizes DNA ends. Double-stranded restriction fragments with flush ends are approximately 3 times more effective than are fragments with unpaired nucleotides extending from the 3' termini and about 10 times more effective than are either fragments with unpaired nucleotides extending from the 5' termini or plasmids with single-strand breaks. All types of restriction fragments become more effective upon removal of terminal 5'-phosphate groups. This specificity profile may relate to the proposed role of poly(ADP-ribose) synthesis in the repair of DNA strand breaks, for those which are assumed to be more difficult to repair in vivo are the more effective stimulators. Poly(ADP-ribose) polymerase has no divalent cation requirement when supplied with flush-ended DNA fragments, but magnesium may enhance the effective of other types of DNA by activating magnesium-dependent nucleases. Ineffective DNAs, such as covalently closed plasmids or synthetic homopolymers that are unable to form Watson-Crick duplexes, apparently compete with effective DNA and weakly inhibit poly(ADP-ribose) synthesis.

In the presence of ATP and a cytosolic factor, cholera toxin fragment A1 catalyzes the transfer of ADP-ribose from NAD to a number of soluble and membrane-bound proteins of the pigeon erythrocyte. Evidence is presented that suggests that the most readily modified membrane protein (Mr 42,000) is the adenylate cyclase-associated GTP-binding protein. Its modification by toxin is stimulated by guanine nucleotides. Adenylate cyclase activity increases in parallel with the addition of ADP-ribose to this protein and decreases in parallel with the subsequent reversal of ADP-ribosylation by toxin and nicotinamide. The protein is only accessible to toxin A subunits if the erythrocytes are lysed. When adenylate cyclase activity reaches a maximum, the number of ADP-ribose residues bound to this protein (about 1500 per cell) is similar to the reported number of beta-adrenergic receptors.

Binding of the beta-adrenergic agonist [3H]hydroxybenzylisoproterenol to the beta-adrenergic receptor of rat reticulocyte membranes results in the coupling of the receptor to the guanine nucleotide regulatory protein associated with the adenylate cyclase system. This regulatory component, referred to as the G-protein, was identified by its specific [32P]-ADP-ribosylation catalyzed by cholera toxin. Incubation of [32P]ADP-ribosylated rat reticulocyte membranes with the [3H]hydroxybenzylisoproterenol agonist prior to membrane solubilization and gel exclusion chromatography resulted in the coelution of the 42,000 Mr [32P]ADP-ribosylated G-proteins with the agonist-occupied beta-adrenergic receptors. The receptor-G-protein complex was not formed when receptors were unoccupied or occupied with antagonists at the time of solubilization. Incubation of rat reticulocyte membranes with [3H]hydroxybenzylisoproterenol in the presence of guanine nucleotides reversed or prevented the formation of this receptor-G-protein complex. These data provide direct evidence for the molecular interactions promoted by agonist occupancy of beta-adrenergic receptors. It is probable that the formation of a receptor-G-protein complex is crucial for catecholamine stimulation of the adenylate cyclase enzyme and, hence, transmembrane information transfer.

Cholera toxin-catalyzed ADP ribosylation of the membrane-bound GTP-binding protein that regulates adenylate cyclase activity requires certain components of the cytosol. In broken cells it is prevented by depletion of endogenous nucleotides and is restored by the provision of GTP (1 to 3 PM half-maximum) or ITP (30 PM) although not by ATP, CTP, TTP, or UTP. An hydrolysis-resistant GTP analog, guanyl-5’-yl imidodiphosphate (Gpp(NH)p, 1 to 3 p ~ ) , also supports the toxin-catalyzed reaction, but adenyl-5’-yl imidodiphosphate is only weakly effective. The ability of nucleoside triphosphates to support toxin-catalyzed ADP ribosylation or the toxin-catalyzed activation of adenylate cyclase correlate well. Even with excess GTP or Gpp(NH)p, the rate of ADP ribosylation and adenylate cyclase activation depends upon the presence of a macromolecular component of the erythrocyte cytosol. If GTP is employed, ADP ribosylation requires the simultaneous presence of this nucleotide, cytosolic macromolecules, NAD’, toxin, and membranes. However, membranes may be predisposed to respond to the toxin by preincubating them with Gpp(NH)p and cytosolic macromolecules. Preincubated membranes, washed free of unbound nucleotide, can then be ADP ribosylated by incubation with cholera toxin and NAD+ without additional cofactors. Both Gpp(NH)p and the macromolecular fraction of the cytosol are required for such preactivation. It is complete in 15 min at 37°C and is about three times slower at 25°C. It requires magnesium, is accelerated by isoproterenol, and is inhibited by GTP or sodium fluoride. The conditions for preactivation resemble those for the specific binding of Gpp(NH)p to the regulatory component of adenylate cyclase and consequent quasi-permanent activation of the cyclase, except that direct activation of the cyclase does not depend upon a cytosolic macromolecule. We conclude that the substrate for cholera toxin is a complex of the GTP-binding protein with GTP, in some way modified by a macromolecule of the cytosol.

The mechanism of elongation of poly(ADP-ribose) on poly(ADP-ribose) polymerase was examined in two ways. The first technique involved a pulse-chase protocol. Poly(ADP-ribose) polymerase was labeled with radioactive NAD, excess precursor was removed by rapid gel filtration chromatography, and nonradioactive NAD was supplied for a second incubation. The products were released with alkali and digested with venom phosphodiesterase which generates AMP uniquely from the distal terminus. The distal residue that was labeled during the pulse remained at the distal terminus and was not converted to an internal residue during the chase. The second technique employed the NAD analog, 2'-deoxyNAD (dNAD), which can engage in mono-ADP-ribose addition reactions but lacks the 2'-OH that is required for polymer formation. dNAD inhibits ADP-ribose incorporation competitively but is not incorporated at the enzyme-distal chain terminus. These findings are inconsistent with a model of poly(ADP-ribose) synthesis in which new residues are added to the 2'-OH terminus of the growing chain, distal to the polymerase attachment. They are consistent with the alternative possibility that new residues are added at the 1" terminus, adjacent to the polymerase. Any such "proximal addition" model requires that there be at least two active center sites (akin to the ribosomal A and P sites), which at a certain stage of each elongation cycle will be occupied by ADP-ribose monomers and ADP-ribose polymers, respectively. Although dNAD does not enter poly(ADP-ribose), it does engage in a slow side reaction whereby a single dADP-ribose residue is added covalently to the polymerase itself, thereby inactivating the enzyme.

The maturation of rat reticulocytes to erythrocytes is paralleled by a marked loss of adenylate cyclase responsiveness to catecholamines and guanine nucleotides without significant loss of basal enzyme activity. These changes in adenylate cyclase activity are accompanied by a parallel %fold decrease in both the number of p-adrenergic binding sites and the quantity of M, = 42,000 protein substrates for cholera toxin-catalyzed [32P]ADP ribosylation. In addition, the maturation of the rat erythrocyte is accompanied by a loss of the ability of guanine nucleotides to modulate receptoragonist interactions. Thus, whereas guanine nucleotides promote tritiated hydroxybenzylisoproterenol agonist dissociation from the P-adrenergic receptor and reduce the potency of isoproterenol in competing for dihydroalprenolol antagonist binding in rat reticulocyte membranes, these regulatory effects of guanine nucleotides are virtually absent in membranes derived from mature rat erythrocytes. Furthermore, the ability of agonist occupancy of the P-adrenergic receptor of rat reticulocyte membranes to promote an increase in receptor size, as assessed by gel exclusion chromatography of solubilized preparations, is not observed in mature erythrocyte preparations. This agonist-induced increase in apparent receptor size has been previously observed in frog erythrocyte membranes (Limbird, L. E., and Lefkowitz, R. J. (1978) Proc. NatZ. Acad. Sei. U. S. A. 75,228-232) and is thought to reflect the formation of a receptor guanine nucleotide regulatory complex. Consistent with this hypothesis is the ability of guanine nucleotides to reverse the formation of this larger molecular size form of the receptor in rat reticulocyte membranes. The inability to promote this receptorguanine nucleotide regulatory protein complex in rat erythrocyte membranes suggests that one molecular concomitant of rat erythrocyte maturation is a decrease in the effectiveness of receptor-guanine nucleotide regulatory protein communication.

Pseudomonas aeruginosa exoenzyme S has appeared to be relatively indiscriminate in its choice of substrates, but in fact it ADP-ribosylates only a small subset of cellular proteins and exhibits a marked preference for several different membrane-associated proteins of apparent Mr = 23,000-25,000, at least some of which appear to bind GTP. One of these is the p21 product of the proto-oncogene c-H-ras, which can be labeled to completion. ADP-ribosylation does not alter the interaction of p21c-H-ras with guanyl nucleotides, but does cause a shift in electrophoretic mobility that implies a large conformational change. Exoenzyme S modifies all of its substrates at arginine residues.

Nonenterotoxigenic strains no. 1196-78 and no. 1074-78 of Vibrio cholerae serogroup O1 (biotype El Tor, serotype Ogawa) were isolated from sewage water in Brazil and fed to 20 volunteers. Neither strain caused diarrhea. None of the seven volunteers who ingested Ogawa strain no. 1074-78 (10(6) organisms) excreted the organism whereas eight of the 13 volunteers who ingested Ogawa strain no. 1196-78 (10(6) or 10(8) organisms) did excrete the organism in their stools. None of 114 stool-culture isolates yielded cholera enterotoxin, and none of the 20 volunteers had significant increases in serum titers of antitoxin as measured by enzyme-linked immunosorbent assay although six of the volunteers had slightly elevated vibriocidal antibody levels. Six volunteers used as controls and four volunteers who had stool cultures positive for strain no. 1196-78 of V. cholerae were challenged with pathogenic El Tor Ogawa strain no. E7946 (10(6) organisms) to determine if previous ingestion of the Brazilian strain would induce protective immunity. All 10 of the volunteers developed diarrhea, and the severity of the illness was similar in both groups.

The structure and products of the two cistrons encoding the Escherichia coli heat-labile toxin (LT) were studied. The LT deoxyribonucleic acid (DNA) region had been isolated as part of a DNA fragment from the plasmid P307, and this fragment was joined to the cloning vector pBR313. Deletion mutations of various lengths were introduced into the LT DNA region and into the adjacent DNA sequences. Analysis of the deletions indicated that the maximum size of the LT DNA region was 1.2 x 10(6) daltons. Two proteins of 11,500 daltons and 25,500 daltons had been shown to be encoded by the LT DNA region. The functions of these LT gene products were investigated. The 11,500-dalton protein had an adsorption activity for Y-1 adrenal cells, and this protein was shown to form aggregates of four or five monomers. The 25,500-dalton protein was shown to have an adenylate cyclase-activating activity. The two cistrons encoding for each of the LT proteins have been located on a genetic map of the LT DNA region. Both cistrons are probably transcribed from the same promoter.

We have recently shown that poly(ADP-ribose) polymerase forms poly(ADP-ribose) by adding ADP-ribose residues to the polymerase-proximal end of an enzyme-bound nascent chain. In this light we have reexamined the mode of hydrolysis of enzyme-bound poly(ADP-ribose) by poly(ADP-ribose) glycohydrolase. When the substrate has been labeled by a pulse-chase protocol, soluble glycohydrolase releases a significant amount of labeled oligomer which can only come from the enzyme-distal (2') end of the polymer. This constitutes additional evidence for the proximal growth of chains. Oligomer is infrequently released from the proximal (1") end of enzyme-bound chains. Rather, the bulk of the poly(ADP-ribose) is digested directly to ADP-ribose monomers. We conclude that poly(ADP-ribose) glycohydrolase starts digestion with an endonucleolytic incision and then removes ADP-ribose residues processively in the 2'----1" direction. Therefore, in contrast to earlier models of polymer growth and hydrolysis, a single poly(ADP-ribose) chain may be extended at one end and simultaneously degraded at the other end. The balance between synthesis and degradation may control the quantity and distribution of polymer around the DNA break which occasions its synthesis.

Exoenzyme C3 from Clostridium botulinum types C and D specifically ADP-ribosylated a 21-kilodalton cellular protein, p21.bot. Guanyl nucleotides protected the substrate against denaturation, which implies that p21.bot is a G protein. When introduced into the interior of cells, purified exoenzyme C3 ADP-ribosylated intracellular p21.bot and changed its function. NIH 3T3, PC12, and other cells rapidly underwent temporary morphological alterations that were in certain respects similar to those seen after microinjection of cloned ras proteins. When injected into Xenopus oocytes, C3 induced migration of germinal vesicles and potentiated the cholera toxin-sensitive augmentation of germinal vesicle breakdown by progesterone, also as caused by ras proteins. Nevertheless, p21.bot was immunologically distinct from p21ras.

NAD is a necessary cofactor for the activation of adenylate cyclase (ATP pyrophosphate-lyase (cyclizing), EC 4.6.1.1) by cholera toxin. Lysates of certain types of cell that hydrolyze their endogenous store of NAD after cell disruption respond poorly or not at all to cholera toxin. Lysates of pigeon erythrocytes, which lack enzymes that degrade NAD, provide a convenient and reproducible system for assaying the activity of cholera toxin in vitro and allow investigation of the mechanism of action of the toxin upon broken cells.

Detergent-lysed BS-C-1, HeLa, and mouse L cells incorporate ADP-ribose from NAD+ into two classes of macromolecules. Metabolically stable products, which appear to be a variety of proteins to which are attached one or a few ADP-ribose residues, predominate when the cellular DNA remains intact. In addition, ghost cells have a potentially much greater capacity to synthesize poly(ADP-ribose), which is completely dependent upon the introduction of strand breaks into their DNA. The initial rate of poly(ADP-ribose) synthesis increases linearly with prior x-ray dose or with the concentration of endonuclease added and, once synthesized, the polymer is rapidly degraded with a half-life of 10 min or less. It appears that sites on the DNA capable of supporting a certain amount of poly(ADP-ribose) synthesis are created as a result of x-irradiation or nucleolytic cleavage and are rapidly eliminated, or "repaired," during subsequent incubation. The sites accumulate if cells are irradiated at 0 degree C; further incubation of the lysed cells with NAD+ at 35 degrees C results in both a burst of poly(ADP-ribose) synthesis and the elimination of the sites. NAD+ enhances the elimination of x-ray-induced sites. Thus, the synthesis of poly(ADP-ribose) may be required for the repair of DNA strand breaks.

By screening possible ADP-ribosyltransferase activities in culture supernatants from various Clostridium species, we have found one Clostridium difficile strain (CD196) (isolated in our laboratory) that is able to produce, in addition to toxins A and B, a new ADP-ribosyltransferase that was shown to covalently modify cell actin as Clostridium botulinum C2 or Clostridium perfringens E iota toxins do. The molecular weight of the CD196 ADP-ribosyltransferase (CDT) was determined to be 43 kilodaltons, and its isoelectric point was 7.8. No cytotoxic activity on Vero cells or lethal activity upon injection in mice was associated with this enzyme. CDT was neither related to C. difficile A or B toxins nor to C. botulinum C2 toxin component I. However, Vero cells cultivated in the presence of C. difficile B toxin had a lower amount of actin able to be ADP-ribosylated by CDT or C2 toxin in vitro. Antibodies raised against CDT reacted by immunoblot analysis with a 43-kilodalton protein of C. perfringens type E culture supernatant producing the iota toxin.

The blood clearance and distribution in the tissues of 125I after intravenous injection of small doses (1.5–5 MLD or 0.08–0.25 µg) of 125I-labeled diphtheria toxin has been followed in guinea pigs and rabbits and compared with the fate of equivalent amounts of injected 125I-labeled toxoid and bovine serum albumin. Toxoid disappeared most rapidly from the blood stream and label accumulated and was retained in liver, spleen, and especially in kidney. Both toxin and BSA behaved differently. Label was found widely distributed among all the organs except the nervous system and its rate of disappearance from the tissues paralleled its disappearance from the circulation. There was no evidence for any particular affinity of toxin for muscle tissue or for a "target" organ. Previous reports by others that toxin causes specific and selective impairment of protein synthesis in muscle tissue were not confirmed. On the contrary, both in guinea pigs and rabbits, a reduced rate of protein synthesis was observed in all tissues that had taken up the toxin label. In tissues removed from intoxicated animals of both species there was an associated reduction in aminoacyl transferase 2 content. It is concluded that the primary action of diphtheria toxin in the living animal is to effect the inactivation of aminoacyl transferase 2. The resulting inhibition in rate of protein synthesis leads to morphologic damage in all tissues reached by the toxin and ultimately to death of the animal.

Exoenzyme S, which had been thought to be unselective, catalyzes the ADP-ribosylation of only a subset of cellular proteins. The intermediate filament protein vimentin is one of the more abundant substrates. Disassembled vimentin, and proteolytic fragments of vimentin that cannot form filaments, is more readily ADP-ribosylated than is filamentous vimentin.