Ethanol inhibition of antigen-induced histamine release from guinea pig lung: correlation with intracellular levels of cyclic nucleotides.

The effect of ethanol on histamine release from lungs of sensitized guinea pigs was studied in conjunction with measurements of tissue concentrations of cyclic AMP and cyclic GMP. Addition of antigen $\text{in vitro}$ elicited a rapid increase in cyclic AMP and cyclic GMP and stimulated release of histamine. Ethanol (2%) inhibited antigen-induced release of histamine over 95% and completely inhibited the increase in both cyclic nucleotides. The activity of cyclic AMP-dependent protein kinase was only slightly affected by ethanol.Metiamide blocked the ovalbumin stimulated increase in cyclic AMP but not cyclic GMP. Pyrilamine did not prevent the rise in either cyclic nucleotide. This suggests that the antigen-induced rise in cyclic AMP is an indirect result of histamine released from the tissue. The inability of H₁ and H₂ receptor antagonists to affect antigen-induced elevation of cyclic GMP in sensitized lung fragments suggests that an elevation in cyclic GMP might be either a primary event in the mediator release sequence or secondary to the release of a mediator other than histamine. The ability of ethanol to inhibit mediator release might be due to its capacity to attenuate the antigen-induced elevation of cyclic GMP in sensitized lung.     

IN VIVO CHANGES OF CEREBRAL CYCLIC ADENOSINE 3'',5''-MONOPHOSPHATE INDUCED BY BIOGENIC AMINES: ASSOCIATION WITH PHOSPHORYLASE ACTIVATION

—The intravenous injection of adrenaline, isoprenaline and histamine to 4-6-day-old chicks resulted in a rapid increase in the cyclic AMP content of cerebral hemispheres that had been removed and frozen within 0·5 s using a freeze-blowing technique. Noradrenaline, dopamine, adenosine, 5-HT and acetylcholine did not significantly alter the nucleotide concentration in vivo. Addition of adrenaline, isoprenaline and histamine to incubated chick cerebral cortex slices also increased the cyclic AMP content of the tissue. Noradrenaline was considerably less potent than these amines and adenosine was ineffective. Low phosphorylase a levels (16 per cent of total activity) were observed in instantaneously frozen cerebral hemispheres of untreated chicks. The injection of adrenaline, isoprenaline and histamine resulted in a rapid conversion of phosphorylase b to a and a significant fall in tissue glycogen. Administration of noradrenaline was without effect on the relative forms of phosphorylase and also failed to influence cerebral glycogen. Phosphorylase activation was not observed in chick cerebral slices under conditions producing large increases in cyclic AMP. It is suggested that in vivo phosphorylase activation and subsequent glycogenolysis may occur, at least in part, in glia and that these cells may be damaged during preparation of cerebral slices. The results provide evidence of a metabolic role for cyclic AMP in cerebral tissue.     

Glutatmine synthetase activity in the chick brain during postnatal growth. Effect of MSO

Glutamine synthetase activity was estimated in the chick cerebral hemispheres, optic lobes and cerebellum between the 1st and the 30th day of postnatal growth. Glutamine synthetase activity is higher in the cerebellum than in the cerebral hemispheres and lowest in the optic lobes at 1 day after hatching; at 30 days after hatching, it is the same in the optic lobes and in the cerebellum and lowest in the cerebral hemispheres. The great increase of glutamine synthetase activity between the 1st and the 4th day after hatching corresponds to the appearance of the heterogeneity of the chick brain glutamate metabolism. The glutamine synthetase activity is inhibited by MSO $\text{in vivo}$ at a concentration of 100 mg kg ^?1 at values of 87, 90 and 89 % in cerebral hemispheres, optic lobes and cerebellum of 1, 2 and 4-day-old chicks. The enzyme inhibition is less pronounced $\text{in vitro}$ and reaches values of about 25 and 75 % for 1 and 10 mM MSO concentrations respectively in the three brain areas of the 1 to 4-day-old chick and values slightly lower in the 30-day-old chick brain.     

The dextro and levorotatory isomers of N-phenylisopropyladenosine: stereospecific effects on cyclic AMP-formation and evoked synaptic responses in brain slices.

The stereoisomers of N⁶-phenylisopropyladenosine elicit accumulations of cyclic AMP in brain slices via interaction with adenosine-receptors. The response in guinea pig cerebral cortical slices and in rat hippocampal slices is blocked by theophylline and potentiated by biogenic amines. A chelator, EGTA, potentiates the response to phenylisopropyladenosine in guinea pig cerebral cortical slices. The $\text{1}$-isomer (EC₅₀ 25 μM) is four- to five-fold more potent than the $\text{d}$-isomer in eliciting accumulations of cyclic AMP in brain slices. In a rat coronal hippocampal slice $\text{in vitro}$, $\text{1}$-phenylisopropyladenosine (IC₅₀ ～ 0.7 μM) reduces the amplitude of evoked synaptic responses generated $\text{via}$ a monosynaptic pathway to the CA1 pyramidal neurons. The $\text{d}$-isomer is nearly one hundred-fold less potent. Thus, the adenosine-receptors involved in the electrophysical response appear much more stereoselective for the $\text{1}$-isomer of phenylisopropyladenosine than the adenosine-receptors involved in cyclic AMP-generation in brain slices.     

Direct and indirect actions of impromidine (a new H2-receptor agonist) on atrial tissue

The maximum chronotropic response of rabbit atria to impromidine, a new H₂ agonist is $\text{?72\%}$ of the histamine response, but the dose needed to cause a half-maximum increase in rate is 1,000 times less for impromidine than for histamine. Pharmacological data using propranolol strongly suggests that the release of norepinephrine is partially responsible for this response, and the effect of diphenhydramine suggests that histamine may be released as well. Improidine also appears to release both compounds when guines pig atria are tested.Before impromidine can be characterized as a specific and potent H₂-receptor agonist, additional data concerning some of its indirect actions must be obtained.     

CYCLIC ADENOSINE 3'',5''-MONOPHOSPHATE FORMATION IN GUINEA-PIG BRAIN SLICES: EFFECT OF H1- AND H2-HISTAMINERGIC AGONISTS

—A variety of histamine analogs elicit accumulations of radioactive cyclic AMP in guinea-pig neocortical and hippocampal slices labelled during a prior incubation with [¹⁴C]adenine. The H₁agonist, 2-aminoethylthiazole, elicits accumulation of cyclic AMP in neocortical and hippocampal slices both in the absence or presence of adenosine. The presence of adenosine increases the maximum response to 2-aminoethylthiazole and decreases the EC₅₀ by nearly 10-fold. In the absence of adenosine the effects of 2-aminoethylthiazole are antagonized in hippocampal slices by both d-brompheniramine and metiamide, while in the presence of adenosine only d-brompheniramine is an effective antagonist. The H₂-agonist, 4-methylhistamine, elicits a somewhat smaller accumulation of cyclic AMP than does 2-aminoethylthiazole in both cortical and hippocampal slices. In the presence of adenosine the response to 4-methylhistamine is enhanced, but is markedly lower than that seen with the combination of adenosine and 2-aminoethylthiazole. The dose-response relationship for 4-methylhistamine in the presence of adenosine appears in hippocampal slices to consist of two components. The response to 4-methylhistamine in the absence of adenosine is blocked by metiamide, while in the presence of adenosine the response is partially blocked by both H₁ and H₂-antagonists. The accumulation of cyclic AMP elicited by histamine is greatly increased by adenosine but the EC₅₀ is not significantly decreased. The results suggest that (i) both H₁- and H₂-receptors regulate cyclic AMP-formation in the central nervous system, (ii) the synergism between adenosine and histamine is mediated primarily by interaction with H₁-receptors and (iii) that adenosine greatly increases the affinity of the H₁-receptors for both H₁ and H₂-agonists without affecting its affinity for histamine.     

Histamine suppression of human lymphocyte responses to mitogens.

Exogenously added histamine in non-cytotoxic concentrations (10^?5?10^?3M) suppresses in vitro proliferation of lymphocytes induced by PHA or Concanavalin A. This suppressive effect was observed when histamine was present for as short as $\text{1}\text{2}$ hr in the beginning of the culture. Histamine, in concentrations as high as 10^?3M, did not cause increased release of isotope from ⁵¹Cr-labeled lymphocytes following 4 hr of incubation. The histamine H₂ receptor antagonist, metiamide, but not the H₁ receptor antagonists diphenhydramine or chlorpheniramine, blocked the histamine suppressive effect. Some of the biological implications of these findings are discussed.     

Histamine modulation of eosinophil migration.

Preincubation of eosinophils with 10(-5) M or higher concentrations of histamine inhibited the eosinophil chemotactic response to endotoxin-activated serum whether by using the nucleopore filter assay and counting the cells migrating through the filter, or by using the Zigmond-Hirsch assay and counting the cells at each 10-mum interval. When the H2-receptor sites on the eosinophils were blocked by metiamide, the inhibitory capacity of histamine was prevented. Preincubation of eosinophils with 10(-6) M histamine increased the number of responding eosinophils to endotoxin-activated serum and this enhancement was blocked by an H1-receptor antagonist. Isoproteronol and aminophylline inhibited eosinophil movement and increasing concentrations of dibutryl cyclic AMP inhibited eosinophil migration. Concentrations of histamine that consistently resulted in inhibition of eosinophil movement stimulated an increase in cyclic AMP that was prevented by blocking the H2-receptor but not the H1-receptor. Thus, histamine-dependent inhibition of the eosinophil chemotactic response to other agents is mediated through the H2-receptor and is associated with an increase in the intracellular level of cyclic AMP whereas histamine dependent enhancement of eosinophil migration to other agents appears to be mediated through the H1-receptor. Eosinophils behave as a heterogeneous population as assessed by the ability of histamine to augment or inhibit cell migration. This may reflect differences in H1 to H2 receptor density or cell responsiveness to receptor stimulation. The chemoattractant activity of histamine itself is not influenced by H1 or H2 receptor antagonists, thus it is possible that an eosinophil has a third type of histamine receptor.     

Stimulatory Effect of Histamine on Cyclic AMP Formation in Chick Pineal Gland

Abstract: Histamine (HA) potently stimulated cyclic AMP accumulation in intact pineal glands taken from light-exposed chicks. The action of HA was stronger in the presence of forskolin and the phosphodiesterase inhibitor 3-isobutyl-1-methylxanthine (IBMX). The effect of HA was mimicked by HA H₁- and H₂-receptor-selective agonists in the following order of potency: HA > 4-methylhistamine (H₂) > 2-methylhistamine (H₁) > 2-thiazolylethylamine (H₁) ≫ dimaprit (H₂). The HA H₃-receptor-selective agonist (R)α-methylhistamine was poorly active. The effect of HA was antagonized by selective H₂-receptor blockers (tiotidine > oxmetidine > cimetidine = ranitidine) and was not significantly affected by the selective H₁- and H₃-receptor blockers mepyramine and thioperamide. A detailed analysis of an antagonistic action of ranitidine (versus HA) revealed a noncompetitive mode of action of the H₂ blocker. The stimulatory action of the H₁ agonist 2-thiazolylethylamine (both under basal conditions and in the presence of forskolin or IBMX) was not significantly influenced by three H₁-receptor-selective blockers (mepyramine, triprolidine, and diphenhydramine), but it was totally counteracted by ranitidine. Using accepted selective agonists and antagonists of the HA H₁, H₂, and H₃ receptor we were unable to identify clearly the receptor subtype mediating the HA action on the cyclic AMP-generating system of the chick pineal. It is suggested that the receptor under consideration may represent either an H₂-like (in terms of mammalian criteria) or avian-specific HA receptor. The data suggest that HA may be considered a modulator of the pineal activity in chicks.     

Accumulation of inositol phosphates and cyclic AMP in guinea-pig cerebral cortical preparations: Effects of norepinephrine,histamine, carbamylcholine and 2-chloroadenosine

Norepinephrine and serotonin augment by about 2-fold the accumulation of cyclic [3H]AMP elicited by 2-chloroadenosine in [³H]adenine-labeled guinea-pig cerebral cortical slices. Histamine causes a 3-fold augmentation. The first two agents have no effect on cyclic AMP alone, while histamine has only a small effect alone. The augmentation of the 2-chloroadenosine response appears to be mediated by α₁-adrenergic, 5HT₂-serotonergic and H₂-histaminergic receptors. VIP-elicited accumulations of cyclic AMP are also augmented through stimulation of α₁-adrenergic, 5HT₂-serotonergic and H₁-histaminergic receptors. Activation of these amine receptors also increases the turnover of phosphatidylinositols in [³H]inositol-labeled guinea pig cerebral cortical slices. Norepinephrine causes a 5-fold, serotonin a 1.2-fold, and histamine a 2.5-fold increase in accumulations of [³H]inositol phosphates. 2-Chloroadenosine, vasoactive intestinal peptide, baclofen, and somatostatin have no effect on phosphatidylinositol turnover, nor do the last two agents augment accumulations of cyclic AMP elicited by 2-chloroadenosine. The data suggest a possible relationship between turnover of phosphatidylinositol and the augmentations of the cyclic AMP accumulations elicited by biogenic amines in brain slices.     

Activation of protein kinase in the guinea pig fundic gastric mucosa by histamine.

The effect of histamine, 1,4-methylhistamine and ethanol on cyclic AMP levels and protein kinase activation was measured in tissue strips from the fundic region of guinea pig gastric mucosa. Histamine induced a significant elevation of tissue cyclic AMP levels and also $\text{in situ}$ activation of the protein kinase. 1,4-methylhistamine, an inactive analog of histamine, and ethanol had no effect on these two parameters. Results suggest that protein kinase activation is involved in the cyclic AMP-mediated action of histamine on the gastric fundic mucosa.     

Methionine enkephalin and morphine alter monoamine and cyclic nucleotide levels in the cerebral ganglia of the freshwater bivalve Anodonta cygnea.

Methionine enkephalin and morphine increased dopamine levels in the cerebral ganglia of $\text{Anodonta cygnea}$. Both agents increased the levels of C'GMP and depressed the levels of C'AMP. The pharmacological effect on dopamine and cyclic nucleotide levels were blocked by prior treatments of the cerebral ganglia with naloxone. The study demonstrates pharmacologically the possible existence of an opiate receptor mechanism.     

Histamine H2-receptors in the gastric mucosa: role in acid secretion.

Currently, two major hypotheses dominate thinking about the role of histamine in the regulation of gastric acid secretion. Code has proposed that histamine is the final common mediator of secretagogue action on the parietal cell while Konturek and Grossman have suggested a multi-receptor control of the secretory process. Experimental results derived from the use of recently synthesized histamine H₂-receptor antagonists have been used by both groups to support their hypotheses. Paradoxically, these hypotheses depend on the presumed specificity of the H₂-antagonists in blocking histamine mediated acid secretion while the apparent lack of such secretagogue specificity of the H₂-antagonists is an important basis for the development of the hypotheses. Our review will analyze the experimental evidence which implicates the histamine H₂-receptor in the control of hydrogen ion secretion as well as evidence for and against receptor specificity in the gastric mucosa of histamine H₂-receptor antagonists.     

Interaction of cyclic nucleotides and ecdysterone in breaking the pupal diapause of the bertha armyworm, Mamestra configurata WLK.

Cyclic GMP and cyclic AMP have distinct and opposite effects upon the action of ecdysterone in diapausing pupae of the Bertha armyworm, $\text{Mamestra}$$\text{configurata}$. Cyclic GMP enhanced the effectiveness of suboptimal doses of ecdysterone in breaking diapause; the amount of cyclic GMP required to lower the ED50 of ecdysterone by half was 80 μg/g. Dibutyryl cyclic GMP had no apparent effect on the action of ecdysterone over a wide dose range (0.07 – 70 μg/g). On the other hand, cyclic AMP and dibutyryl cyclic AMP effectively blocked the diapause-breaking action of ecdysterone when administered simultaneously with the steroid hormone. The amount of cyclic AMP required to reduce the incidence of diapause termination from 100% to 50% was 60 μg/g; for dibutyryl cyclic AMP the amount required was only 14 μg/g. No cyclic nucleotide tested in the study could by itself break the pupal diapause of $\text{M.}$$\text{configurata}$. The concept that cyclic GMP and cyclic AMP provide at least different if not opposing regulatory influences in certain insect systems is discussed briefly in the light of these observations.     

Adenosine-elicited accumulation of cyclic AMP in brain slices: potentiation by agents which inhibit uptake of adenosine

The uptake and incorporation of low concentrations of radioactive adenosine into guinea pig cerebral cortical slices is effectively inhibited by dipyridamole, hexobendine, papaverine, 6-($\text{p}$-nitrobenzylthio) guanosine, 5′-deoxy-adenosine and N⁶-phenyladenosine and ineffectively inhibited by other adenosine analogs such as 2-chloroadenosine, 3′-deoxyadenosine and tubercidin or by phosphodiesterase inhibitors such as theophylline, isobutylmethylxanthine, and N, 0-dibutyrylcyclic AMP. When uptake of 10–20

adenosine is inhibited 50–70% by dipyridamole, hexobendine, papaverine or 6-($\text{p}$-nitrobenzylthio)-guanosine, the adenosine-elicited accumulation of cyclic AMP is potentiated 2–3 fold. Potentiation of the effects of low concentrations of adenosine by various agents parallels more closely their efficacy as inhibitors of adenosine uptake rather than their potency as phosphodiesterase inhibitors. Amine-elicited accumulations of cyclic AMP are enhanced by hexobendine, dipyridamole, papaverine and 6-($\text{p}$-nitrobenzylthio) guanosine and this enhancement is blocked by an adenosine antagonist, theophylline. The stimulatory effects of the adenosine analogs, 5′-deoxyadenosine, 2-chloroadenosine and N⁶-phenyladenosine are blocked by theophylline and potentiated by hexobendine. The results are compatible with the hypothesis that the specific inhibition of uptake of adenosine potentiates adenosine or amine-elicited accumulations of cyclic AMP by increasing the effective extracellular concentration of adenosine within the slice. The inhibition or stimulation of cyclic AMP accumulation by adenosine analogs is consonant with differential activities as agonist or antagonist at an extracellular adenosine receptor.     

Isolated adrenal cortex cells: ACTH agonists, partial agonists, antagonists; cyclic AMP and corticosterone production

Isolated adrenal cortex cells respond to the addition of ACTH_1–39 or analogs with increased production of cyclic AMP and corticosterone. It is estimated that cyclic AMP production need proceed at less than 20% of maximum to induce maximum corticosterone production. ACTH_1–24, [Lys¹⁷, Lys¹⁸]ACTH_1–8 amide, and ACTH_1–16 amide induce a maximum rate of cyclic AMP and of corticosterone production equal to those of ACTH_1–39. The relative potencies as determined by cyclic AMP and by corticosterone production are in excellent agreement. The analog, ACTH_5–24, induces maximum cyclic AMP production equal to 45% of that of the natural hormone, but as predicted, induces maximum corticosterone production equal to that of ACTH_1–39. The derivative, [Trp(Nps)⁹]ACTH_1–39 induces 77% of maximum corticosterone production and less than 1% of maximum cyclic AMP production. The fragment ACTH_11–24 is a competitive antagonist of ACTH_1–39 for both cyclic AMP and corticosterone production. The observations on agonists, a partial agonist and a competitive antagonist are in harmony with the “second messenger” role assigned to cyclic AMP. A provisional model, based on the fit of the experimental observations to a set of equations, provides expressions of “intrinsic activity,” “receptor reserve”, “sensitivity”, and “amplification” in terms of maximum cyclic AMP production, concentration of ACTH which induces $\text{1}\text{2}$ maximum cyclic AMP production and concentration of cyclic AMP which induces $\text{1}\text{2}$ maximum corticosterone production.     

Adenosine 3′, 5′-cyclic monophosphate in Schistosoma mansoni

Homogenates of adult $\text{Schistosoma mansoni}$ (blood flukes), isolated from the porto-mesenteric veins of infected mice, contain substantial activities of adenylyl cyclase, cyclic AMP phosphodiesterase, and a cyclic AMP stimulated protein kinase. The adenylyl cyclase, which is largely sedimentable at 10,000xg, is stimulated 20-fold by 10mM sodium fluoride and 1.4 to 2-fold by serotonin, glucagon, prostaglandins E₁, E₂ or B₁. The phosphodiesterase, which is largely sedimentable at 10,000xg, is inhibited by both aminophylline and papaverine but is not influenced by 10mM sodium fluoride. The protein kinase, which is present in the 10,000xg supernatant is stimulated 4 to 8-fold by either cyclic AMP or cyclic GMP. There is a preference for cyclic AMP ($\text{K}{}_{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}\text{= 1.1×10}{}^{\mathrm{?7}}\text{M}$) over cyclic GMP ($\text{K}{}_{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}\text{= 4.5×10}{}^{\mathrm{?6}}\text{M}$). If intact worms are incubated in a glucose free medium there is a mobilization of glycogen stores which is preceded by a rise in cyclic AMP concentration. In a medium with 5mM glucose there is neither a rise in cyclic AMP nor mobilization of glycogen.     

Parallel activation of cyclic AMP phosphodiesterase and cyclic AMP-dependent protein kinase in two human gut adenocarcinoma cells (HT 29 and HRT 18) in culture,by vasoactive intestinal peptide (VIP) and other effectors activating the cyclic AMP system

Vasoactive intestinal peptide (VIP), secretin, catecholamines and prostaglandin E₁ (PGE₁) in the presence of a cyclic nucleotide phosphodiesterase inhibitor stimulate the accumulation of cyclic AMP in two colorectal carcinoma cell lines (HT 29 and HRT 18) with subsequent activation of the cyclic AMP-dependent protein kinases. In HT 29 cells incubated without phosphodiesterase inhibitor, 10^?9 M VIP promotes a rapid and specific activation of the low $\text{K}{}_{\mathrm{<mtext>m</mtext>}}$ cyclic AMP phosphodiesterase (1.7-fold); at 25°C the effect is maintained for more than 15 min, while at 37°C the activity returns to basal value within 15 min. As shown by dose-response studies, VIP is by far the most effective inducer ($\text{K}{}_{\mathrm{<mtext>a</mtext>}}\text{= 4 · 10}{}^{\mathrm{?10}}\text{M}$) of the cyclic AMP phosphodiesterase activity; partial activation of the enzyme is obtained by 3 · 10^?7 M secretin, 10^?5 M isoproterenol and 10^?5 M PGE₁; PGE₂ and epinephrine are without effect. In HRT 18 cells VIP is less active ($\text{K}{}_{\mathrm{<mtext>a</mtext>}}\text{= 2 · 10}{}^{\mathrm{?9}}\text{M}$) whereas 10^?6 M PGE₁, 10^?6 M PGE₂ and 10^?5 M epinephrine are potent inducers of the phosphodiesterase activity. The positive cell response to dibutyryl-cyclic AMP further indicates that cyclic AMP is a mediator in the phosphodiesterase activation process. The incubation kinetics and dose response effects of the various agonists on the cyclic AMP-dependent protein kinase activity determined for both cell types in the same conditions show a striking similarity to those of phosphodiesterase. Thus coordinate regulation of both enzymes by cyclic AMP was observed in all incubation conditions.     

The effect of prostacyclin on intestinal ion transport in the rat

The actions of PGI₂ and PGE₂ on electrically monitored ion transport in rat jejunum and colon have been determined both $\text{in vivo}$ and $\text{in vitro}$. Whilst PGE₂ was shown to induce a marked change in ion transport PGI₂ was relatively ineffective. The ability of the prostanoids to influence ion transport is related to their capacity to change mucosal cyclic AMP levels since in isolated small intestinal enterocytes PGE₂ caused a marked stimulation in cyclic AMP levels whilst PGI₂ had little effect. In colonic mucosal scrapes PGE₂ was more effective than PGI₂ in stimulating changes in cyclic AMP levels. It appears doubtful that PGI₂ plays a direct role in the regulation of intestinal ion transport.     

Malaria enhances cyclic AMP production by immature erythrocytes in vitro.

Infection with a chloroquine-susceptible line of $\text{Plasmodium}$$\text{berghei}$ NYU-2 enhanced the production of cyclic AMP in response to isoproterenol only in immature mouse erythrocytes, which possess a functional hormone-receptor-adenylate cyclase complex. With 10^?7 M isoproterenol the increases in cyclic AMP concentrations (picomoles per 10⁸ erythrocytes) were 19 for a preparation of mature erythrocytes, 81 for a preparation enriched with immature erythrocytes, 25 for a preparation of infected mature erythrocytes, and 900 for a preparation enriched with infected immature erythrocytes. Beta-adrenergic receptor blockade with propranolol prevented this response to isoproterenol but had no effect on the stimulation produced by prostaglandin E₁. These findings indicate that the malaria parasite enhances the responsiveness of a fundamental regulatory system that is intrinsic to immature erythrocytes.