Mitochondrial dysfunction and death in motor neurons exposed to the glutathione-depleting agent ethacrynic acid.

This study investigated the mechanisms of toxicity of glutathione (GSH) depletion in one cell type, the motor neuron. Ethacrynic acid (EA) (100 microM) was added to immortalized mouse motor neurons (NSC-34) to deplete both cytosolic and mitochondrial glutathione rapidly. This caused a drop in GSH to 25% of the initial level in 1 h and complete loss in 4 h. This effect was accompanied by enhanced generation of reactive oxygen species (ROS) with a peak after 2 h of exposure, and by signs of mitochondrial dysfunction such as a decrease in 3-(4,5-dimethyl-2-thiazoyl)-2,5-diphenyltetrazolium bromide (MTT) (30% less after 4 h). The increase in ROS and the MTT reduction were both EA concentration-dependent. Expression of heme oxygenase-1 (HO-1), a marker of oxidative stress, also increased. The mitochondrial damage was monitored by measuring the mitochondrial membrane potential (MMP) from the uptake of rhodamine 123 into mitochondria. MMP dropped (20%) after only 1 h exposure to EA, and slowly continued to decline until 3 h, with a steep drop at 5 h (50% decrease), i.e. after the complete GSH loss. Quantification of DNA fragmentation by the TUNEL technique showed that the proportion of cells with fragmented nuclei rose from 10% after 5 h EA exposure to about 65% at 18 h. These results indicate that EA-induced GSH depletion rapidly impairs the mitochondrial function of motor neurons, and this precedes cell death. This experimental model of oxidative toxicity could be useful to study mechanisms of diseases like spinal cord injury (SCI) and amyotrophic lateral sclerosis (ALS), where motor neurons are the vulnerable population and oxidative stress has a pathogenic role.

Modulation of cerebellar and hepatic nitric oxide synthase by exogenous arginine and endotoxin.

We have studied in mice the effect of treatment with exogenous arginine and/or LPS by monitoring serum nitrite/nitrate levels and by investigating the response of cerebellar and liver nitric oxide synthase (NOS). We measured NOS activity in cerebellar extracts while changes in iNOS mRNA were followed in the liver since direct assay of NOS activity proved unreliable with this tissue. In fact, liver and cerebellum extracts were both very active in converting arginine into a citrulline-like metabolite, but only cerebellum conversion was dependent on addition of NADPH and inhibitable by N(G)-methyl-l-arginine. Treatment with LPS, on its own, increased serum nitrite/nitrate levels at 5 and 20 h after injection, while treatment with LPS and arginine produced nitrite/nitrate levels in the serum even greater at 5 h, but significantly lower at 20 h. Liver iNOS mRNA levels were markedly increased by LPS, and this effect was significantly decreased when mice were also given exogenous arginine. A stimulatory effect of LPS was also found on NOS activity in the cerebellum, where a very small stimulation may have also been caused by arginine feeding. These findings indicate that LPS stimulates NOS expression/activity both in the cerebellum and in the liver and suggest a complex pattern of modulation of iNOS by arginine, with NO being first produced in excess and then downregulating iNOS expression.

Kupffer cell depletion partially prevents hepatic heme oxygenase 1 messenger RNA accumulation in systemic inflammation in mice: role of interleukin 1beta.

The heme oxygenase 1 (HO-1) gene is rapidly activated in the liver after lipopolysaccharide (LPS) treatment. Ninety minutes after LPS treatment (0.1 mg/kg, intraperitoneally) hepatic HO-1 messenger RNA (mRNA) of mice was 40 times the control value. To investigate the hepatic cellular source of the increased HO-1 transcript, we treated mice with LPS and galactosamine (700 mg/kg, intraperitoneally), a selective transcriptional inhibitor of hepatocytes. Galactosamine prevented the LPS-mediated increase of HO-1 mRNA in the liver, indicating that hepatocytes are the main cell type in which HO-1 mRNA accumulates after LPS treatment. We then tested in vitro and in vivo the hypothesis that LPS-mediated hepatic accumulation of HO-1 mRNA is caused by intercellular communication between Kupffer cells and hepatocytes. Isolated rat hepatocytes showed an increase in HO-1 mRNA compared with controls after 90 minutes of exposure to a LPS stimulated Kupffer cell-conditioned medium. This suggests that soluble mediators from Kupffer cells were responsible for this effect. To study the role of Kupffer cells in vivo, we treated mice with Kupffer cell-inactivating or -depleting agents and LPS. Gadolinium chloride and liposome-encapsulated dichloromethylene diphosphonate lowered LPS-mediated HO-1 mRNA accumulation (by about 50%); in these groups hepatic levels of interleukin (IL)-1beta were decreased, by more than 75%. Methylpalmitate hardly affected hepatic HO-1 mRNA accumulation or IL-1beta content after LPS treatment. There was no relationship between HO-1 mRNA and serum TNF or IL-6 levels. These results suggest that LPS-mediated hepatic HO-1 mRNA accumulation is a hepatocyte response partly caused by soluble mediators, particularly IL-1beta, released from Kupffer cells.

Prion protein fragment 106-126 differentially induces heme oxygenase-1 mRNA in cultured neurons and astroglial cells.

Heme oxygenase (HO), which catalyzes the degradation of heme, has two isozymes (HO-1 and HO-2). In brain the noninducible HO-2 isoform is predominant, whereas the inducible HO-1 is a marker of oxidative stress. Because brain oxidative stress might be present in prion-related encephalopathies (PREs), as in other neurodegenerative diseases, we investigated whether HO-1 mRNA was induced in neuronal and astroglial cell cultures by a peptide corresponding to residue 106-126 of human prion protein (PrP). This peptide is amyloidogenic, and when added in vitro to cultured cells it reproduces the neuronal death and astroglial proliferation and hypertrophy occurring in PREs. HO-1 mRNA did not accumulate in rat cultured neurons from hippocampus or cortex exposed to PrP 106-126 (50 microM for 5 days). PrP 106-126 induced HO-1 mRNA accumulation in rat astroglial cultures depending on the exposure time and concentration, being maximal (33-fold) after 7 days of exposure at 50 microM. The nonamyloidogenic amidated or amidated-acetylated PrP 106-126 was ineffective, as was a scrambled peptide used as control. N-Acetylcysteine reduced (50%) the accumulation of HO-1 mRNA in astroglial cells after PrP 106-126 (25 microM) given for 5 days. Thus, oxidative stress is apparently a feature of the toxicity of PrP 106-126, and it might also occur in PREs; induction of HO-1 could contribute to the greater resistance of astrocytes compared with neurons to PrP 106-126 toxicity.

Modulation of constitutive and inducible hepatic cytochrome(s) P-450 by interferon beta in mice.

AIMS/METHODS: Interferon beta is used as a therapeutic agent, but its effects on the hepatic cytochrome P-450-dependent drug metabolizing system have not yet been characterized. We investigated the effect of interferon beta on cytochrome P-450 in mice. RESULTS: Interferon beta (2 x 10(5) units/mouse) significantly reduced total hepatic cytochrome P-450 (20%) and the activity of NADPH cytochrome C reductase (12%) 24 h after administration; lower doses had no such effect. Various monooxygenase activities were slightly reduced, the one most affected being 7-ethoxycoumarin O-deethylase (29%). In phenobarbital-treated mice, interferon beta reduced the induction of total cytochrome P-450 (22%), the activities of pentoxyresorufin O-dealkylase (38%), benzyloxyresorufin O-dealkylase (30%), erythromycin N-demethylase (30%), 7-ethoxycoumarin O-deethylase (16%) and cytochrome P-450 2B1 (33%) and 3A (45%) proteins. In beta-naphthoflavone-treated mice, interferon beta lowered the induction of total cytochrome P-450 (18%), the activities of ethoxyresorufin O-deethylase (31%) and of 7-ethoxycoumarin O-deethylase (25%) and of cytochrome P-450 1A1 protein (31%). CONCLUSIONS: Thus it appears that induced cytochrome(s) P-450 were susceptible to interferon beta, this effect not being influenced by the type of inducer. Since various members of the same cytochrome P-450 subfamilies catalyze oxidation of drugs in humans, our findings have potential significance as regards the fate of drugs or exogenous compounds given to patients receiving interferon beta.

Mechanisms of interleukin-2-induced depression of hepatic cytochrome P-450 in mice.

Interleukin-2 (15 micrograms/mouse, i.p. twice daily for 4 days and once on the 5th day) significantly lowered cytochrome P-450 and heme content and increased heme oxygenase mRNA accumulation; the activities of 7-ethoxycoumarin O-deethylase, ethoxy- and pentoxyphenoxazone O-dealkylases were decreased. The activity of the type O form of hepatic xanthine oxidase increased, but there was no increase in lipid peroxide, expressed in terms of microsomal malondialdehyde. In vivo inactivation of xanthine oxidase activity by feeding mice with tungstate did not substantially change the degree of interleukin-2-induced cytochrome P-450 depression, suggesting that the two processes are not causally linked. Induction of tolerance to endotoxin by a 4-day pretreatment with lipopolysaccharide resulted in 50% protection against this depression despite inhibition of the interleukin-2 induced formation of tumor necrosis factor. This suggests that the release of tumor necrosis factor per se does not fully account for the depression of cytochrome P-450. Dexamethasone, already used in patients to reduce the toxicity of interleukin-2 therapy, provided full protection against the cytochrome P-450 depression.

Mechanisms of endotoxin-induced haem oxygenase mRNA accumulation in mouse liver: synergism by glutathione depletion and protection by N-acetylcysteine.

In in vitro systems haem oxygenase-1 (HO-1) mRNA increases after exposure to agents causing oxidative stress. We lowered cellular antioxidant defence systems in vivo by giving mice increasing doses (0.15 g/kg-1.6 g/kg) of DL-buthionine-(S,R)-sulphoximine (BSO), a specific inhibitor of glutathione synthesis. Maximum glutathione depletion (80%) coincided with maximum hepatic HO-1 mRNA accumulation (about 20 times), whereas with 50% depletion, accumulation was only doubled. It has been suggested that reactive oxygen and nitrogen intermediates are involved in hepatic toxicity of endotoxin (lipopolysaccharide, LPS); LPS even at low doses [0.1 mg/kg, intraperitoneally (i.p.)] induces HO-1 mRNA about 25-fold after 1 h. Hepatic glutathione depletion (respectively 40% and 80%) after a low (0.3 g/kg) or a high (1.6 g/kg) BSO dose, resulted in potentiation of the HO-1 mRNA accumulation induced by LPS (0.1 mg/kg, i.p.). In the absence of BSO, N-acetylcysteine (NAC) (1 g/kg orally) reduced LPS-induced HO-1 mRNA accumulation to one fourth. Under the same experimental conditions S-adenosylmethionine (SAM) was not effective. NAC also reduced HO-1 mRNA accumulation when administered to mice in which glutathione was depleted and its synthesis blocked by BSO (1.6 g/kg). Thus reactive oxygen intermediates are likely mediators of LPS-induced HO-1 mRNA accumulation, and glutathione content appears to be one of the factors regulating this accumulation in the liver. Our findings are compatible with the theory that HO-1 induction might have a protective function in vivo when defence mechanisms against oxidants are challenged.

Depression of liver metabolism and induction of cytokine release by diphtheria and tetanus toxoids and pertussis vaccines: role of Bordetella pertussis cells in toxicity.

A 24-h pretreatment of mice with diphtheria and tetanus toxoids and whole-cell pertussis vaccines depressed liver cytochrome P-450 and therefore prolonged hexobarbital-induced sleeping time in mice. The depression of liver drug metabolism by a cellular vaccine containing a mutated pertussis toxin was less marked than that induced by the wild-type vaccines, indicating that the mutated vaccine might have lower toxicity in this regard. The wild-type vaccines decreased microsomal P-450 levels by 50%, while the mutated whole-cell vaccine had a less marked effect (a decrease of 30%), paralleling the results obtained in sleeping time experiments. Furthermore, an acellular mutated vaccine did not affect liver drug metabolism, indicating a role of the whole bacterial cell in this side effect. All the cellular vaccines studied induced high serum interleukin-6 levels; on the other hand, the acellular mutated vaccine induced very low interleukin-6 levels, indicating that the whole bacterial cell is also important for interleukin-6 induction. All vaccines studied were very poor tumor necrosis factor inducers.

Cytokine induction of haem oxygenase mRNA in mouse liver. Interleukin 1 transcriptionally activates the haem oxygenase gene.

Accumulation of the mRNA coding for haem oxygenase (HO, EC 1.14.99.3) was stimulated by treating mice with endotoxin (lipopolysaccharide, LPS; 20 micrograms/mouse intraperitoneally), suggesting that haem catabolism is a target of infection and inflammation in vivo. Therefore various cytokines, possible mediators for the biological responses to LPS, were administered intraperitoneally to mice, and the levels of HO mRNA were measured by Northern-blotting analysis using the rat HO cDNA as a probe [Shibahara, Müller, Taguchi and Yoshida (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 7865-7869]. Marked induction of HO mRNA was observed 2 h after administration of interleukin 1 (IL-1) (34-fold) and tumour necrosis factor (19.5-fold) (5 micrograms/mouse), whereas interleukin 6 (6.2 micrograms/mouse) was much less active (3.5-fold) and interleukin 2 (25 micrograms/mouse) and interferon-gamma (3 micrograms/mouse) were ineffective. HO mRNA induced by the cytokines of LPS accumulated rapidly (maximum at 1-2 h after administration), preceding the elevation of HO enzymic activity. Treatment of mice with IL-1 stimulated the transcription of the HO gene by 4-fold, as assessed by in vitro nuclear-run-on assay. These results indicate that enzymic haem catabolism in the liver is a process inducible in vivo by inflammatory cytokines, which up-regulate HO synthesis at the transcriptional level. Increased removal of haem might be part of the protective mechanisms elicited by the acute-phase response, possibly to reduce the pro-oxidant state of the cell.

Hepatic protoporphyria is associated with a decrease in ligand binding for the mitochondrial benzodiazepine receptors in the liver.

Protoporphyrin IX (PP) and N-methylprotoporphyrin IX (N-MePP) added in vitro to liver membranes reduced dose-dependently the affinity of [3H]PK 11195 for the mitochondrial benzodiazepine receptors (MBRs), the latter being about 20 times more potent (Ki 4.5 and 0.25 microM). Preincubation of these two porphyrins with liver homogenates for 120 min at 4 degrees resulted in significant inhibition of [3H]PK 11195 binding even after repeated washings of the membranes due to the residual presence in the membranes of about 35 and 5% of PP and N-MePP, respectively. Thus, the hypothesis that an in vivo increase in the hepatic porphyrin content modifies the binding of the isoquinoline PK 11195 to the MBRs was investigated in an experimental model of protoporphyria. PP and N-MePP were allowed to accumulate in vivo through treatment with 3,5-diethoxycarbonyl-1, 4-dihydrocollidine (DDC) (100 mg/kg i.p., once), and rats were killed 5 h after treatment when hepatic porphyrin accumulation was marked (10-fold increase), PP predominating. In the liver, treatment reduced the affinity (Kd) of [3H]PK 11195 for MBRs (from 3.56 to 15.37 nM, P < 0.01) and the maximum number of binding sites (Bmax) (55% decrease, P < 0.05); the affinity (Ki) of RO 5-4864 for [3H]PK 11195 binding sites was also reduced (from 23.9 to 72.99 nM, P < 0.05). No significant differences were found in the brain cortex. Liver and brain diazepam binding inhibitor levels and plasma corticosterone levels were unchanged. The reduction in [3H]PK 11195 binding to MBRs in the liver of DDC-treated rats thus appears to be attributable to a specific effect of the DDC-induced formation of the two protoporphyrins; this conclusion suggests that in hepatic protoporphyria processes modulated by MBRs may be altered.

Interleukin-1 and tumour necrosis factor induce hepatic haem oxygenase. Feedback regulation by glucocorticoids.

During the acute-phase response to bacterial endotoxins [lipopolysaccharide (LPS)] in mice, the hepatic activity of haem oxygenase (HO) is increased. We investigated the effects of the potential humoral mediators of inflammation, interleukin-1 (IL-1) and tumour necrosis factor (TNF), on hepatic HO activity. In mice, IL-1 or TNF (5 micrograms) caused an elevation of HO activity comparable with that after LPS exposure (20 micrograms). The induction of HO by both cytokines was more pronounced in adrenalectomized mice. In the intact mice induction of HO activity by cytokines was observed earlier than depression of 7-ethoxycoumarin O-de-ethylase, a cytochrome P-450-dependent enzyme activity. Pretreatment with dexamethasone of the intact mice (3 mg/kg) or of the adrenalectomized mice (0.4 mg/kg) prevented the induction of HO activity caused by LPS and IL-1 respectively. These results suggest that: (1) HO activity is increased during an IL-1- or TNF-mediated acute-phase response, so haem metabolism might be a potential target of inflammation, and (2) HO induction by IL-1 and TNF does not require glucocorticoids, which in fact act as antagonists of this cytokine-induced effect.

[Postnatal growth of very low birth weight newborn (LBW infant). Anthropometry after a period of 3 years, longitudinal study]. / Crecimiento postnatal del recién nacido de muy bajo peso (RNMBP). Antropometría a tres años plazo, estudio longitudinal.

Growth of 230 very low birth weight infants (VLBWI) admitted to the neonatal wards of a metropolitan pediatric hospital at Santiago, Chile was studied prospectively up to 36 months of age, in the period 1980-1988. For further analysis patients were separated in groups A, 60 newborn infants with birth-weight below 1,001 g and B, 170 newborn infants with birth-weight between 1,001 and 1,500 g. We used Patri's growth charts, to compare the results with full term healthy newborns of 3,318 g average birth-weight from the same socio-economical status. The average weight of group A infants was below 2 SD at age one year and between one and two SD at 2 and 3 years of age. In group B infants weight was between one and 2 SD at one year of age and below 1 SD at 2 years. At age 3 years average weight was very close to normal. Group A infants were not successful to achieve the average height of the standard at age 3 year, but this same goal was obtained at 2 years of age in group B infants. Head circumference was within the normal average at ages 3 and 2 year in group A and B infants, respectively.

Bile and biliary lipid secretion in rats with hexachlorobenzene-induced porphyria. Effect of S-adenosyl-L-methionine administration.

We investigated liver morphology and biliary function in vivo in rats made porphyric by hexachlorobenzene (HCB). In one group of HCB rats we also evaluated whether S-adenosyl-L-methionine (SAMe), administered during the last 15 days of HCB treatment, attenuated liver injury and the accumulation of porphyrins (HCB + SAMe group). In HCB rats we found: (a) a 100% increase in liver weight; (b) a 500-fold increase in total liver porphyrins (TLP); (c) significantly increased serum bilirubin and cholesterol levels; (d) unchanged total bile flow (TBF) but enhanced levels of the bile acid independent fraction (BAIF); and (e) decreased excretion in bile of bile acids (BA), phospholipids (PL) and cholesterol (CHO) (58, 65 and 47%, respectively, expressed as mmol/min per kg liver). SAMe was found to partially reverse HCB-related effects. TLP levels were about 65% lower in HCB + SAMe treated rats than in HCB rats. However, while SAMe restored bile CHO excretion to control values, it did not influence bile excretion of BA, PL, or BAIF. In conclusion, HCB-induced porphyria was characterized by a complex derangement of liver morphology and biliary function that was unrelated to the extent of porphyrin accumulation in the liver.

Protective effect of S-adenosyl-L-methionine in hepatic uroporphyria. Evaluation in an experimental model.

The potential use of S-adenosyl-L-methionine (SAMe) as therapy for human porphyria cutanea tarda was investigated in an experimental model of hepatic porphyria--that is, chronic treatment of female rats with 0.2% hexachlorobenzene (HCB) in the diet. Administration of SAMe (25 mg/kg subcutaneously twice daily) during the last 15 days of HCB administration halved porphyrin accumulation in the liver but did not alter HCB-induced massive inhibition of uroporphyrinogen decarboxylase. Equally unaffected were inhibition of glutathione peroxidase and stimulation of lipid peroxide formation induced by HCB. Hypothetically, the beneficial effect of SAMe on hepatic porphyrin accumulation might be linked to modifications of the cellular availability of adenosine triphosphate.

Biochemical, morphological and flow-cytometric evaluation of the effects of hexachlorobenzene on rat liver.

This study investigated the ability of HCB (0.1% in the diet for 15 days) to cause early changes in the cellular ploidy of rat liver. Treatment caused marked hepatomegaly, increase of microsomal proteins and cytochrome P-450 content and reduction of hepatocyte microviscosity. Microscopic examination showed that the hepatocytes were enlarged, with hyaline cytoplasm and vacuoles. The size distribution of the isolated hepatocytes showed a larger percentage of bigger cells. Flow-cytometric DNA/protein analysis was performed on whole (fixed) cells and on nuclei. From the combined results of both analyses it was possible to exclude significant changes in the percentage sof diploid, mononucleated tetraploid, binucleated tetraploid and octoploid hepatocytes. The DNA and protein content of each subpopulation remained unchanged. Our results suggest that HCB does not cause early diploidization of liver cells and that hepatomegaly and cytochrome P-450 induction seem not to be correlated with effects on total DNA and total protein contents.

[Height, weight and head circumference according to gestational age in newborn infants born before 35 weeks]. / Talla, peso y perímetro craneano según edad gestacional en recién nacidos de menos de 35 semanas.

A retrospective and collaborative study was done in Santiago, Chile, in order to obtain national data on birth-weight, height and head circumference of babies born at 24 to 34 weeks of gestation: 370 babies with reliable gestational age, single pregnancies and no maternal nor fetal morbidity were included in the study. Babies were born in three government and one private hospitals from 1982 to 1987. Mean birthweight, height and head circumference for each gestational age from 24 to 34 weeks are presented in tables with their S.D. and charts +/- 1.5 S.D. The national use of these tables and curves is recommended.

[Postnatal growth of very low birth weight infants. II. Anthropometry at one year of age, a longitudinal study]. / Crecimiento postnatal del recién nacido de muy bajo peso: II. Antropometría a un año plazo, estudio longitudinal.

A prospective study up to one year of age, was done in 154 newborns with birth weight less than 1,501 g at a Metropolitan Hospital of Santiago, Chile, between 1983 and 1987. The patients were divided in groups A: 44 patients with birth weight less than 1,001 g; B: 68 patients with birth weights from 1,000 to 1,250 g and C: 48 patients with birth weight 1,251 to 1,500 g. At age one year weight of group A patients was 6,724 +/- 804 g; group B 7,782 +/- 927 g and group C 7,941 +/- 903 g. Weight and height of group A babies at one year of age were below 2 DS of the average weight and height for normal chilean babies of the same socioeconomic status and geographic area with 3,318 g mean birth weight. Group B and C were between 1 and 2 DS. Head circumference was within normal ranges in groups B and C patients, but in group A patients if was below 2 DS of the average.

Investigations on the role of free radical processes in hexachlorobenzene-induced porphyria in mice.

Male C57Bl/10 mice were chronically fed hexachlorobenzene (HCB) (0.02% of the diet) alone or in combination with a single subcutaneous dose of iron (12.5 mg iron per mouse). After eight weeks the group of mice pretreated with the iron overload was highly sensitized to the porphyrogenic effect of HCB, as shown by liver porphyrin accumulation. A synergistic effect of iron was evident on other parameters too, such as HCB-induced hepatic damage, activation of type O of xanthine oxidase, and decreased activity of copper zinc superoxide dismutase and glutathione peroxidase(s). None of these parameters was affected by iron alone. Iron alone and in association with HCB markedly raised the level of lipid peroxides, the increase in the HCB group being smaller. The combined treatment resulted in a significant reduction of HCB's inductive effects on microsomal heme and cytochromes P-450 and b5 and on the activity of aryl hydrocarbon hydroxylase. The content of nonprotein sulfhydryl groups was reduced to the same extent in mice treated with HCB or HCB plus iron. The results suggest that reactive intermediates such as are formed by lipid peroxidation are not sufficient on their own to create the conditions for uroporphyrinogen decarboxylase impairment, as evident in the group of mice receiving iron overload alone. Conversely, HCB administration induced a specific condition of imbalance in the liver between formation and inactivation of reactive intermediates which was associated with hepatic porphyrin accumulation and was potentiated by concomitant administration of iron.

Comparison of hexachlorobenzene-induced alterations of microsomal membrane composition and monooxygenase activity in male and female rats.

The effect of hexachlorobenzene (HCB) on microsomal cytochromes P-450 and b5, monooxygenase activity and membrane composition was examined in male and female Fischer rats. Cytochrome P-450 was induced more in male than in female animals while cytochrome b5 was induced only in males. Analysis of patterns of induction of microsomal monooxygenases showed that aminopyrine-N-demethylase activity doubled in both sexes after treatment while aryl hydrocarbon hydroxylase activity was 16 times the control value in the females and 1.5 times in the males. After HCB treatment the phospholipid content of microsomal membranes per gram of liver was increased in both sexes while cholesterol was unchanged. Analysis of the phospholipids (PL) pattern showed that the percentage of sphingomyelin (SPH) decreased significantly (50% of the control value) while phosphatidylcholine (PC), phosphatidylinositol (PI) and phosphatidylethanolamine (PE) did not change. These changes resulted in a reduction of membrane microviscosity and indicate that HCB interferes with the biosynthesis of phospholipids containing choline. Free fatty acid (FFA) content also dropped in both sexes but females were more affected; free arachidonic acid rose in females. HCB induction of microsomal cytochromes and monooxygenases is thus accompanied by marked modifications of membrane composition. Comparing the 2 sexes, HCB showed more pronounced features of 'PB type' inducers in males.