Cat Submandibular Gland Research Articles

Differences in parasympathetic vasodilator and salivary responses in the cat submandibular gland between lingual and chorda-lingual nerve stimulation.

Experimental activation of parasympathetic vasodilatation and salivation in the submandibular gland by a "pseudoreflex" method has definite advantages over other commonly used methods. However, it is unclear which stimulation sites allow for avoidance of direct activation of the parasympathetic efferents to this gland. We examined this question in heavily anesthetized (alpha-chloralose and urethane), artificially ventilated, cervically vago-sympathectomized cats, using the intersection of the chorda-lingual nerve and the submandibular gland duct ("the intersection") as our reference point. Both vasodilatation and salivation evoked by electrical stimulation of the lingual nerve were abolished by section of the chorda-lingual nerve (10 mm proximal to the intersection), provided the stimulation site was distal to the intersection by 4 mm or more. This salivation was abolished by the muscarinic receptor antagonist, scopolamine. Thus, by careful choice of stimulation site, submandibular gland responses mediated solely by reflex parasympathetic mechanisms can be evoked by lingual nerve stimulation in this preparation.

Glycosylation of secretory proteins in salivary glands and saliva studied by lectin-probes.

Lectin-probes were used histochemically to study glycosylation patterns in prepackaged secretory material in cat and rat submandibular glands and to assess the secretory changes induced by nerve stimulations. The same probes were also used for correlative biochemical assessments of the constituents in glandular extracts and in saliva from the same experiments, after electrophoretic separations and immobilization on nitrocellulose membranes. These studies provided a broader understanding than the use of either procedure alone. Lectin binding patterns in saliva were more complex than anticipated from the histochemistry and this probably relates to the presence of different constituents in secretory granules and a greater availability of binding sites biochemically. In cats, parasympathetic stimulation caused depletion of lectin staining in acini and to some extent in demilunes, whereas sympathetic stimulation caused depletion of lectin binding in striated ducts and reduction in demilunar size. Biochemically after SDS-PAGE of saliva similar secretory effects were observed but each nerve also evoked some secretion from the cells not showing histochemical change. In rats, sympathetic stimulation caused depletion of lectin positive granules from both acini and granular tubules. After SDS-PAGE of the saliva one zone of lectin binding 94-150 kD (Mr) was identified as acinar mucin. Granular tubule proteins consisting mainly of kallikreins occurred in a zone 25-35 kD (Mr) and lectin bindings suggested that different kallikreins may be differently glycosylated. This unexpected possibility was confirmed on slot-blot preparations of separated kallikreins, which also revealed the novel finding that some kallikreins are O- as well as N-glycosylated.

Distribution of nitric oxide synthase containing elements in the feline submandibular gland

Combined nicotinamide adenine dinucleotide phosphate-diaphorase (NADPH-d) histochemistry and nitric oxide synthase (NOS) immunocytochemistry were used to study the distribution of nitric oxide synthesizing elements in the cat submandibular gland. A large number of thin varicose fibres, with intense staining for both markers, were seen around or in close contact with the acini. Some of the stained nerve fibres were associated with intra- and interlobular salivary ducts and blood vessels. All neurones in the submandibular ganglia showed intense staining for both NADPH-d and NOS. The epithelial layer of the salivary ductal branches and the endothelial lining of blood vessels were NOS immunonegative but NADPH-d positive. Our results suggest that NO might act as a neurotransmitter in the regulation of blood flow and secretion in the submandibular salivary gland.

Effects of clonidine and yohimbine on parasympathetic reflex salivation and vasodilatation in cat SMG.

The present experiments were designed to examine the site of action of the alpha 2-adrenoceptor agonist, clonidine, and the alpha 2-antagonist, yohimbine, on the salivary secretion response to parasympathetic and sympathetic stimulation in the cat submandibular gland (SMG). The parasympathetic reflex salivary secretion was significantly affected by both clonidine and yohimbine (clonidine reduced and yohimbine increased the amount of saliva secreted) while the reflex vasodilator response was not affected by either agent. No degree of methacholine-induced salivation was affected by prior treatment with either clonidine or yohimbine. The salivation elicited by facial nerve root stimulation was markedly reduced by clonidine but potentiated by yohimbine. Prior treatment with scopolamine completely abolished the salivary secretion induced by either methacholine or facial nerve root stimulation. Clonidine and yohimbine had no effect on SMG salivation elicited by sympathetic stimulation. These results indicate that the alpha 2-agonist and antagonist act specifically on parasympathetic salivary fibers but have no effect on parasympathetic vasodilator fibers. They thus suggest that postganglionic parasympathetic innervation is differentiated into salivary and vasodilator fibers.

Salivary secretion in cat submandibular gland mediated by chorda tympani afferents.

The aim of the present study was to investigate whether the afferent traffic from the tongue mediated only via the chorda tympani nerve (CTN) can still elicit reflex salivary and vasodilator responses in the cat submandibular gland (SMG) after section of the lingual nerve proper (LNP). Electrical stimulation of the chorda lingual nerve (CLN) at a site approximately 5 mm distal to the intersection of the CLN and the SMG duct elicited salivary and vasodilator responses in the SMG in sympathectomized cats. Both responses were unaffected by section of the LNP. The optimal frequency of CLN stimulation for submandibular salivation and vasodilation was 20 Hz, regardless of whether the LNP had been cut. Prior treatment with the autonomic ganglion blocker hexamethonium (10 mg/kg iv) virtually abolished the salivation and the blood flow increase in SMG. Prior treatment with scopolamine (0.1 mg/kg iv) almost abolished the salivary secretions but had no effect on the vasodilator responses in the SMG elicited by CLN stimulation after LNP section. The mechanism underlying the reflex submandibular salivation mediated by chorda tympani afferents appears to involve parasympathetic muscarinic receptors, but the mechanism for the vasodilator response has yet to be established. These results indicate that afferent traffic passing through the CTN on CLN stimulation is importantly involved in the parasympathetic reflex secretory and vasodilator responses in the cat SMG.

The vasodilator and secretory effects elicited by sympathetic nerve stimulation in cat submandibular gland

The effects of electrical stimulation of the peripheral cut ends of the ascending cervical sympathetic trunk on vasomotor, particularly vasodilator, response were studied in relation to salivary secretion in the cat submandibular gland. The vasodilator and salivary responses were compared by stimulating the peripheral cut ends of the sympathetic nerve at various intensities (1–8 V), durations (5–60 s) and frequencies (1–100 Hz) using a 2 ms pulse duration. Electrical stimulation of the cervical sympathetic nerves caused vasoconstriction followed by vasodilatation and salivation. There were certain differences in the maximal responses of vasodilatation and salivation during sympathetic stimulation. For example, optimal frequencies for vasodilatation and salivation were 10 Hz and 20 Hz, respectively. Time-dependent increases in vasodilatation and salivary secretion were seen for periods of up to 30 s and 60 s, respectively. The volume of salivation was not necessarily correlated with the magnitude of vasodilator response. Prior treatment with an α-adrenoceptor blocking agent phentolamine largely reduced vasoconstriction ( P < 0.01) and almost completely abolished salivary secretion ( P < 0.01), but had only a slight inhibitory effect on vasodilatation (0.1 < P < 0.05). Propranolol (a β-adrenoceptor blocking agent) pretreatment significantly decreased vasodilatation ( P < 0.05) but had no statistically significant effect on vasoconstriction or salivation. Scopolamine (a muscarinic cholinoceptor blocking agent) had no effect on the responses followed by sympathetic nerve stimulation but abolished the vagal mediated reflex salivation, indicating that the salivation, but not vasodilatation, elicited by activation of the afferent fibers of the vagus nerve is mediated via parasympathetic muscarinic fibers. The present results indicate that α-adrenoceptors participate exclusively in salivary and vasoconstrictor responses, whereas β-adrenoceptors are in part involved in sympathetic-induced vasodilatation in the cat submandibular gland.

The mechanism of histamine release induced by pilocarpine from different tissues: studies on rat peritoneal mast cells.

Pilocarpine releases histamine from mast cells of cat submandibular gland and rat liver. In the salivary gland, histamine is released into the saliva and venous outflow. Atropine blocks the salivation, but not histamine release from the submandibular gland into the blood. Histamine release from the gland could be due to a direct action of pilocarpine on tissue mast cells or to an indirect action of mediators (acetylcholine and peptides). These hypotheses were further investigated in our present studies on rat peritoneal mast cells. Our results show: (1) histamine release from rat peritoneal mast cells induced by pilocarpine (ED50 = 1.7 x 10(-2) mol/l) occurs at 1000-fold higher concentrations than by substance P (ED50 = 1.7 x 10(-5) mol/l) and in 6.5-fold higher concentrations than by atropine (ED50 = 2.6 x 10(-3) mol/l), (2) pilocarpine injected directly into the rat peritoneal cavity causes histamine release from peritoneal mast cells in 1.8-fold higher concentrations than from isolated rat peritoneal mast cells. These results would support the hypothesis that histamine release from cat submandibular gland is caused by peptidergic co-transmission during the stimulation of the organ.

Nerve-induced secretion of glycoconjugates from cat submandibular glands: a correlative study with lectin probes on tissues and saliva.

Horseradish peroxidase-conjugated lectins were used on tissue sections to localize the main secretory glycoproteins in cat submandibular glands and on Western blots to evaluate their movement into saliva with selective nerve stimulation. Central acinar cells bound lectins from Arachis hypogaea (PNA) specific for the terminal disaccharide Gal beta 1, 3GalNac, Griffonia simplifolia (GSA I-B4) specific for terminal alpha Gal, and Lotus tetragonolobus (LTA) specific for fucose. Lectins from Limax flavus (LFA) specific for sialic acid and Dolichos biflorus (DBA) specific for terminal alpha GalNac reacted preferentially with demilunar cells, whereas apical granules in striated ducts were recognized principally by LTA. Parasympathetic stimulation promoted the release of lectin-reactive glycoconjugates from both central and demilunar cells. In contrast, sympathetic stimulation caused almost complete release of LTA-reactive granules in striated ducts and only moderate secretion from demilunar cells. Lectin blots of stimulated saliva discriminated many of the constituent bands, providing information about their glycosylation. Several bands were common to both parasympathetic and sympathetic saliva, and many bands gave wider ranges of lectin binding than anticipated from the histochemistry. The major component in parasympathetic saliva was a glycoconjugate of less than 12 KD which reacted with every lectin tested. Lectin blots of sympathetic saliva showed a prominent diffuse LTA-reactive band around 33 KD, which was attributed to tissue kallikrein. The identity and cellular origin of most bands in stimulated submandibular saliva are still unclear but the technique shows considerable promise for improving the recognition and characterization of individual glycoconjugates.

Effects of acute and long-term atropine treatment on levels, release and response to VIP and PHI in the submandibular gland of cat and rat.

We have studied the effects of acute and long-term treatment of cats and rats with atropine on the levels, release and effects of two peptides, vasoactive intestinal polypeptide (VIP) and peptide histidine isoleucine (PHI), that probably co-exist with acetylcholine in the parasympathetic nerves supplying the submandibular gland. Atropine treatment (progressively increasing doses from 2 to 15 mg kg-1 injected s.c.) for 14 days did not alter the contents of VIP- or PHI-like immunoreactivity (-IR) in the cat submandibular gland or in three other tissues (nasal mucosa, trachea and tongue). Acute as well as long-term atropine treatment decreased the vasodilation following low-, but not high-, frequency parasympathetic nerve stimulation. During prolonged stimulation (60 min) there was a decreased vasodilatation response following both acute and long-term atropine treatment. The overflow of VIP-IR and PHI-IR following parasympathetic nerve stimulation was markedly increased by acute, but not by long-term atropine treatment. The VIP- or PHI-induced stimulation of cyclic AMP (cAMP) accumulation in the cat submandibular gland was not altered after long-term atropine treatment. Similarly, treatment of male Sprague-Dawley rats with atropine (20 mg kg-1) or imipramine (20 mg kg-1) for 14 days did not alter the sensitivity to VIP or to PHI of cAMP accumulation in the submandibular gland, nor was there any change in VIP-IR or PHI-IR content. In conclusion, although atropine treatment causes an acute increase in the overflow of VIP and PHI evoked by parasympathetic nerve stimulation, there is no depletion of peptide stores upon long-term treatment, nor is there any change in the effect of exogenous VIP and PHI on cAMP-accumulation.

Enzymatic degradation and quantitative lectin labeling for characterizing glycoconjugates which act as lectin acceptors in cat submandibular gland.

Sites of binding of eight different lectins (LTA, UEA I, WGA, SBA, DBA, CON A, PNA, RCA I) to cat submandibular gland were studied after exposure of tissue sections to sialidase, alpha-fucosidase, beta-galactosidase, alpha-mannosidase, beta-N-acetylglucosaminidase. All lectins were affected by enzymatic predigestion and the labeling of individual lectins was highly dependent upon the glycosidase used to pretreat the sections. Glycoconjugates of demilunar, acinar and ductal cells exhibited a different composition of terminal sequences. For example, fucose proved to form the disaccharide fucose-galactose in demilunar and acinar cells, whereas it was present with the sequence fucose-N-acetyl-D-glucosamine in striated duct cells. Sialic acid participated both to the terminal sequence sialic acid-galactose and sialic acid-N-acetyl-D-galactosamine either in demilunar or in ductal cells. Lectin labeling combined with glycosidase digestion was also helpful in verifying the influence of neighbouring oligosaccharides on the affinity of lectins for the respective sugars.

Kallikrein release by rat kidney slices: effect of norepinephrine.

A lot of studies report the effect of norepinephrine (NE) on the concentration and release of kallikrein in rat and cat submandibular salivary glands (1, 2, 3, 4, 5). However very few data are available about the effect of NE on the renal kallikrein kinin system.KeywordsCortical SliceTissue KallikreinUrinary KallikreinRenal Cortical SliceRenal Nerve StimulationThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Regulation of amino acid influx and efflux at the basolateral plasma membrane of the salivary gland epithelium: effects of parasympathetic nerve stimulation.

Basolateral amino acid transport systems in the salivary epithelium of resting and secreting cat submandibular glands were characterized by means of a rapid paired-tracer dilution technique. Amino acid uptake was measured by comparison of venous tracer concentration profiles for a labeled amino acid and D-mannitol (an extracellular tracer of similar size) following an intra-arterial bolus injection of both radioactive molecules. Unidirectional uptake of 21 amino acids, dopamine, noradrenaline, and serotonin was quantified in non-secreting glands. During 8 Hz parasympathetic nerve stimulation, significant epithelial uptakes were measured for L-[3H] alanine and L-[3H] phenylalanine, but less than 0.2% of the injected amino acid was recovered in the collected saliva. In non-secreting glands, cross-inhibition studies of L-[3H] alanine, L-[3H] phenylalanine, and L-[3H] lysine uptake by unlabeled amino acid competitors and detailed kinetic influx experiments indicated that epithelial uptake was mediated by three distinct parallel transport systems: ASC (short-chain neutral), L (branched-chain and aromatic neutral), and y+ (cationic). Rapid metabolism of alanine was inhibited by aminooxyacetate, and the metabolic uncoupler dinitrophenol selectively accelerated the efflux of transported large neutral amino acids and L-lysine. Concurrent autoradiographic experiments suggest that transport sites for small and large neutral amino acids are localized in the basolateral plasma membrane of acinar, demilunar, and striated ductal cells.

Differential binding sites of peroxidase-labelled lectins in the submandibular gland of sucking and adult cats

The glycosidic residues in the submandibular gland of cat at different ages were studied using a battery of 6 lectins conjugated with horseradish peroxidase as histochemical reagents. The submandibular glands of sucking and adult cats showed diversities in the reaction intensity or in the localization of binding sites. The morphologically distinct secretory tracts appear to produce different types of oligosaccharides in sucking and adult subjects. The sugars that can be defined as growth markers in the cat submandibular gland seem to be alpha-L-fucose, N-acetyl-D-glycosamine and beta-galactose. The data originated from this research have been compared with biochemical data previously obtained by the authors on the cat submandibular gland during growth.

Autoradiographic localization of transported neutral amino acids in epithelia of cat submandibular and sublingual salivary glands.

Light-microscopic autoradiography was used to localize the cellular sites for neutral amino acid uptake in submandibular and sublingual salivary gland epithelia. The vasculature of isolated glands was perfused for 3-5 min with either L-(3-3H)serine or L-(4-3H)phenylalanine and then fixed by perfusion with buffered glutaraldehyde. In the submandibular gland the small neutral amino acid L-serine and the aromatic amino acid L-phenylalanine were localized to central acinar cells, demilunar cells and ductal cells. In the sublingual gland silver grains associated with each of these tritiated amino acids were localized to central acinar and ductal cells. Perfusion of both submandibular and sublingual glands with unlabelled L-serine (25 mM) or L-phenylalanine (30 mM) resulted in a significant decrease in the silver grain density associated with each labelled amino acid. The absence of silver grains in the lumina of acinar and ductal cells and the presence of tight junctions near the apical surface of the epithelium strongly suggest that the initial uptake of these amino acids was mediated by basolateral plasma membrane carriers.

Effects of VIP, PHM and substance P on blood vessels and secretory elements of the human submandibular gland

The effects of the neuropeptides VIP, PHM and substance P (SP) on vascular smooth muscle tone, K + secretion from exocrine elements and tissue content of cyclic AMP (cAMP) in the human submandibular gland were studied in vitro. All three peptides caused relaxation of noradrenaline contracted human submandibular arteries at nM concentrations. SP was slightly more active than VIP and PHM which had a similar potency as vasodilators. Only carbachol but not VIP, PHM or SP stimulated K + secretion from exocrine elements of the human submandibular gland. Principally similar in vitro effects on K + secretion were obtained on the cat submandibular gland, but in the rat not only carbachol but also SP stimulated K + secretion. VIP and PHM increased cAMP production of exocrine elements in the human submandibular gland in nM concentrations. VIP was about 5-fold more potent than PHM with regards to cAMP production. In conclusion, VIP, PHM and SP relaxed human submandibular arteries in vitro. Both VIP and PHM stimulated cAMP production in glandular tissue but none of the three peptides induced K + secretion from human submandibular gland tissue. This suggests that, in contrast to the situation in the rat, SP does not cause watery salivation in man, while VIP and PHM may modulate protein e.g. amylase content of the saliva.

Effects of Avian Pancreatic Peptide on Smooth Muscle Relaxations by Vasoactive Intestinal Peptide and Inhibitory Nerve Stimulation

Experiments were carried out to investigate whether avian pancreatic peptide (APP) antagonized relaxations of vascular and gastrointestinal smooth muscles induced by vasoactive intestinal peptide (VIP) and non-adrenergic, non-cholinergic inhibitory nerve stimulation using the cat submandibular gland, cat colon and dog stomach in vivo and several isolated smooth muscles in vitro. APP caused a dose-dependent inhibition of an atropine-resistant vasodilatation induced by chorda-lingual and pelvic nerve stimulation and by VIP. APP did not reduce VIP output induced by pelvic nerve stimulation. APP shifted the dose-response curve for VIP and the frequency-response curve for pelvic nerve stimulation to the right. The vasodilatation induced by isoproterenol or bradykinin was inhibited by somewhat higher doses of APP. APP failed to inhibit the vago-vagal reflex relaxation of the dog stomach. In guinea pig and rat fundic strips and in isolated chick rectum, APP did not exert any significant effects on relaxation caused by VIP or transmural stimulation in the presence of atropine and guanethidine. The result indicates that APP inhibits VIP-mediated relaxation of the vascular smooth muscle, but not that of the gastrointestinal smooth muscle. The inhibitory effect of APP on VIP-mediated vasodilatation is suggested to be due to the physiological competition.

Use of different derivatives ofd-Val-Leu-Arg for studying kallikrein activities in cat submandibular glands and saliva

Glandular kallikrein shows a special selectivity for D-Val-Leu-Arg-4-methoxy-2-naphthylamide in comparison with other potential oligopeptide substrates and it provides a useful histochemical substrate, although the reaction may not always be specific. However, in cat submandibular saliva, a biochemical assay using the closely related D-Val-Leu-Arg-7-amino-4-trifluoromethylcoumarin (AFC) as substrate, which affords more sensitive detection, showed that soya bean trypsin inhibitor causes no inhibition. This indicates that there are unlikely to be contaminating enzymes competing for the substrate in this body fluid. Support for this observation has been gained by the useful new enzyme overlay membrane technique for fluorescent assessment of reactive bands of enzymes after isoelectric focusing, using membranes of cellulose acetate impregnated with D-Val-Leu-Arg-AFC. Comparison of results after isoelectric focusing of purified cat submandibular kallikrein with samples of cat submandibular saliva confirmed that the substrate is monospecific for kallikrein in saliva of the cat. This knowledge has enabled us to start assessing the dynamics of the secretion of kallikrein by the gland. Testing individual drops of saliva has shown that an amazingly rapid mobilization of kallikrein occurs in high concentrations on sympathetic nerve stimulation. The corresponding oligopeptide-based inhibitor D-Val-Leu-Arg-chloromethyl ketone was found to be strongly inhibitory of the amidase reaction by kallikrein but showed a low specificity for kallikrein. Nevertheless, its effects have been tested in vivo by the intravascular route and it caused an increase in the resting salivary vascular resistance whether administered close-arterially or intravenously. Thus, it would seem that a kallikrein-like protease does influence the background tone in the vessels and the source of this enzyme is thought to be mast cells.

Effect of indomethacin on the bradykinin-induced vasodilation in the submandibular gland.

Effects of the local infusion of bradykinin or isoproterenol were studied on the systemic blood pressure, blood flow, and vascular resistance of the submandibular gland before and after indomethacin treatment in cats. Both bradykinin and isoproterenol were shown to potently increase the glandular perfusion rate. The systemic arterial blood pressure fell by about 20% in response to indomethacin. The enhancement of blood flow by bradykinin or isoproterenol was not, however, affected by indomethacin treatment. The present results appear to suggest that the endogenous prostaglandins are not essentially involved in the development of responses produced by bradykinin or isoproterenol in the cat submandibular gland.

Submandibular salivary secretion in the cat and associated potassium movements: dependence on temperature and perfusate flow rate.

Cat submandibular glands were perfused with Locke solution in a thermostated chamber and intermittently stimulated with 10(-5)M acetylcholine (ACh). In one series of experiments the perfusion pressure was varied within the range 90-60 mm Hg, and secretory flow rate, active K+-reuptake, passive K+-release, and resting and ACh-induced venous flow rates were measured. The ACh-induced secretory flow rate and the maximal K+-fluxes were related to the simultaneous ACh-induced venous flow rates. A proportionality was found between the maximal rate of ACh-induced K+-release and ACh-induced venous flow rates below 8 ml/min, while at higher flow rates the K+-release leveled off. The maximal rate of the post-stimulatory K+-reuptake increased proportionally to the ACh-induced perfusate flow rate throughout the range studied. The secretory flow rate was much less affected by changes in ACh-induced perfusate flow rate. In another series of experiments the gland temperature was varied within the range 12-37 degrees C, and the same parameters were measured. All parameters decreased with cooling being reduced to 50% of their 37 degrees C values at: 24 degrees C for secretion, 19 degrees C for K+-reuptake, and 14 degrees C for K+-release. It is concluded: that the rate of ACh-induced K+-release is limited by the ACh-induced perfusate flow rate (within the physiological range), the capacity of the K+-reuptake mechanism is at least one order of magnitude larger than the maximal rate of K+-reuptake in vivo, the marked temperature sensitivity of the secretory flow rate reflects the high complexity of the mechanisms involved.

Corelease of vasoactive intestinal polypeptide and peptide histidine isoleucine in relation to atropine-resistant vasodilation in cat submandibular salivary gland.

Parasympathetic nerve stimulation of the submandibular salivary gland in the cat caused salivary secretion, vasodilation and a corelease of vasoactive intestinal polypeptide (VIP) and peptide histidine isoleucine (PHI) immunoreactivities (IR) into the venous effluent, as indicated by an increase in output. The ratio between the released VIP-IR and PHI-IR was close to 1:1. Gel-permeation chromatography of plasma from the submandibular venous effluent indicated that the released VIP-IR and PHI-IR were very similar to porcine VIP and PHI, respectively. Atropine pretreatment enhanced output of both VIP-IR and PHI-IR during the parasympathetic nerve stimulation to a similar extent (about 5-fold) compared to control stimulations. This increase could be due to an inhibitory presynaptic muscarinic receptor regulation of VIP and PHI release. Since VIP and PHI are present in the same postganglionic parasympathetic nerves in the gland and both peptides have vasodilator activity, the present data suggest that both VIP and PHI may contribute to the atropine-resistant vasodilation seen upon stimulation of the chorda-lingual nerve. The parasympathetic control of salivary gland function may thus involve, a multimessenger system with the classical transmitter acetylcholine and the peptides VIP and PHI.