Tetracycline derivatives and ceftriaxone, a cephalosporin antibiotic, protect neurons against apoptosis induced by ionizing radiation.

DNA damage induced by low doses of ionizing radiation causes apoptosis, which is partially mediated via the generation of free radicals. Both free radicals and apoptosis are involved in the majority of brain diseases, including stroke, Alzheimer's disease and amyotrophic lateral sclerosis. Because previous studies have shown that tetracycline derivatives doxycycline and minocycline have anti-inflammatory effects and are protective against brain ischemia, we studied whether minocycline and doxycycline or ceftriaxone, a cephalosporin antibiotic with the potential to inhibit excitotoxicity, protect neurons against ionizing radiation in primary cortical cultures. A single dose of 1 Gy significantly increased lactate dehydrogenase release, induced DNA fragmentation in neurons and triggered microglial proliferation. Treatment with minocycline (20 nM), doxycycline (20 nM) and ceftriaxone (1 microM) significantly reduced irradiation-induced lactate dehydrogenase release and DNA fragmentation. The most efficient protection was achieved by minocycline treatment, which also inhibited the irradiation-induced increase in microglial cell number. Our results suggest that some tetracycline derivatives, such as doxycycline and minocycline, and ceftriaxone, a cephalosporin derivative, protect neurons against apoptotic death.

Minocycline, a tetracycline derivative, is neuroprotective against excitotoxicity by inhibiting activation and proliferation of microglia.

Minocycline, a semisynthetic tetracycline derivative, protects brain against global and focal ischemia in rodents. We examined whether minocycline reduces excitotoxicity in primary neuronal cultures. Minocycline (0.02 microm) significantly increased neuronal survival in mixed spinal cord (SC) cultures treated with 500 microm glutamate or 100 microm kainate for 24 hr. Treatment with these excitotoxins induced a dose-dependent proliferation of microglia that was associated with increased release of interleukin-1beta (IL-1beta) and was followed by increased lactate dehydrogenase (LDH) release. The excitotoxicity was enhanced when microglial cells were cultured on top of SC cultures. Minocycline prevented excitotoxin-induced microglial proliferation and the increased release of nitric oxide (NO) metabolites and IL-1beta. Excitotoxins induced microglial proliferation and increased the release of NO metabolites and IL-1beta also in pure microglia cultures, and these responses were inhibited by minocycline. In both SC and pure microglia cultures, excitotoxins activated p38 mitogen-activated protein kinase (p38 MAPK) exclusively in microglia. Minocycline inhibited p38 MAPK activation in SC cultures, and treatment with SB203580, a p38 MAPK inhibitor, but not with PD98059, a p44/42 MAPK inhibitor, increased neuronal survival. In pure microglia cultures, glutamate induced transient activation of p38 MAPK, and this was inhibited by minocycline. These findings indicate that the proliferation and activation of microglia contributes to excitotoxicity, which is inhibited by minocycline, an antibiotic used in severe human infections.

Preconditioning with spreading depression activates specifically protein kinase Cdelta.

Preconditioning with brief ischemia or spreading depression (SD) confers tolerance in cortical neurons to subsequent episode of ischemia. In myocardium a similar preconditioning is achieved by mechanisms, which are mediated by protein kinase C (PKC) alpha, delta, epsilon or zeta isoform. We induced SD by cortical application of KCl in the rat and analyzed cortical tissues after recovery of 30 min, 4 h and 12 h. While no changes at protein levels or activity of PKCalpha, epsilon or zeta were detected, a considerable increase in membrane translocation of PKCdelta was seen at 30 min and 12 h. A significant increase at mRNA level, protein amount and autophosphorylation at 12 h confirmed the late activation of PKCdelta, which may be involved in neuronal protection by preconditioning.

Genomic structure and chromosomal localization of the rat protein kinase Cdelta-gene.

Protein kinase Cdelta (PKCdelta) is a widely expressed calcium-independent PKC isozyme that is induced at mRNA and protein levels upon stimulation of different cellular pathways. We found the rat PKCdelta gene to consist of 19 exons and to span approximately 29 kb. The exon-intron junctions follow the GT/AG rule. The 5' untranslated region is nearly 12 kb in length, and the transcription initiation site is surrounded by CG-rich sequences. The 5' flanking region contains putative binding sites for activator protein 1 (AP-1), nuclear factor kappa B (NFkappaB), stimulatory protein-1 (Sp-1) and nerve growth factor induced-C (NGFI-C) transcription factors. The PKCdelta gene is localized at the rat chromosome 19p14. The cloned gene will help to elucidate the role of PKCdelta in growth, differentiation and death of mammalian cells.

Induction of protein kinase Cdelta subspecies in neurons and microglia after transient global brain ischemia.

The delayed death of CA1 neurons after global brain ischemia is associated with induction of apoptosis genes and is inhibited by protein synthesis inhibitors, suggesting that the degeneration of CA1 pyramidal neurons is an active process that requires new gene expression. The transient global ischemia model has been extensively used to identify enzymes and other proteins underlying delayed neuronal cell death. The expression of protein kinase C (PKC) subspecies after 20 minutes of global brain ischemia produced by a four-vessel occlusion model in the rat was studied. From the multiple PKC subspecies studied, only PKCdelta mRNA was significantly up-regulated in CA1 pyramidal neurons at 24 hours and in activated microglia at 3 to at least 7 days after ischemia. The induction of PKCdelta mRNA was also found in the cortex at 8 hours and 3 days after ischemia. This cortical but not hippocampal induction was regulated by an alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid/kainate receptor antagonist, 6-nitro-7-sulfamobenzo[f]quinoxaline-2,3-dione, and glucocorticoids. An N-methyl-D-aspartate receptor antagonist, MK-801, was without effect on the induction of PKCdelta subspecies. The selective and prolonged induction of the PKCdelta mRNA and protein first in CA1 pyramidal neurons and at a later stage in activated microglia suggests that the PKCdelta isozyme may take part in regulation of the delayed death of CA1 neurons after transient global brain ischemia.

A tetracycline derivative, minocycline, reduces inflammation and protects against focal cerebral ischemia with a wide therapeutic window.

The only treatment of patients with acute ischemic stroke is thrombolytic therapy, which benefits only a fraction of stroke patients. Both human and experimental studies indicate that ischemic stroke involves secondary inflammation that significantly contributes to the outcome after ischemic insult. Minocycline is a semisynthetic second-generation tetracycline that exerts antiinflammatory effects that are completely separate from its antimicrobial action. Because tetracycline treatment is clinically well tolerated, we investigated whether minocycline protects against focal brain ischemia with a wide therapeutic window. Using a rat model of transient middle cerebral artery occlusion, we show that daily treatment with minocycline reduces cortical infarction volume by 76 +/- 22% when the treatment is started 12 h before ischemia and by 63 +/- 35% when started even 4 h after the onset of ischemia. The treatment inhibits morphological activation of microglia in the area adjacent to the infarction, inhibits induction of IL-1beta-converting enzyme, and reduces cyclooxygenase-2 expression and prostaglandin E(2) production. Minocycline had no effect on astrogliosis or spreading depression, a wave of ionic transients thought to contribute to enlargement of cortical infarction. Treatment with minocycline may act directly on brain cells, because cultured primary neurons were also salvaged from glutamate toxicity. Minocycline may represent a prototype of an antiinflammatory compound that provides protection against ischemic stroke and has a clinically relevant therapeutic window.

Spreading depression induces expression of calcium-independent protein kinase C subspecies in ischaemia-sensitive cortical layers: regulation by N-methyl-D-aspartate receptors and glucocorticoids.

Spreading depression is a wave of sustained depolarization challenging the energy metabolism of the cells without causing irreversible damage. In the ischaemic brain, sreading depression-like depolarization contributes to the evolution of ischaemia to infarction. The depolarization is propagated by activation of N-methyl-D-aspartate receptors, but changes in signal transduction downstream of the receptors are not known. Because protein phosphorylation is a general mechanism whereby most cellular processes are regulated, and inhibition of N-methyl-D-aspartate receptors or protein kinase C is neuroprotective, the expression of protein kinase C subspecies in spreading depression was examined. Cortical treatment with KCl induced an upregulation of protein kinase Cdelta and zeta messenger RNA at 4 and 8 h, whereas protein kinase Calpha, beta, gamma and epsilon did not show significant changes. The gene induction was the strongest in layers 2 and 3, and was followed by an increased number of protein kinase Cdelta-immunoreactive neurons. Protein kinase Cdelta and zeta inductions were inhibited by pretreatment with an N-methyl-D-aspartate receptor antagonist, dizocilpine maleate, which also blocked spreading depression propagation, and with dexamethasone, which acted without blocking the propagation. Quinacrine, a phospholipase A2 inhibitor, reduced only protein kinase C5 induction. In addition, N(G)(-nitro-L-arginine methyl ester, a nitric oxide synthase inhibitor, did not influence protein kinase Cdelta or zeta induction, whereas 6-nitro-7-sulphamoylbenzo[f]quinoxaline-2,3-dione, an alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate/kainate receptor antagonist, and the cyclo-oxygenase inhibitors indomethacin and diclophenac tended to increase gene expression. The data show that cortical spreading depression induces Ca2(+)-independent protein kinase C subspecies delta and zeta, but not Ca(2+)-dependent subspecies, through activation of N-methyl-D-aspartate receptors and phospholipase A2. Even though the signal pathway is similar to the induction described previously in ischaemia for genes implicated in delayed neuronal death, the gene inductions observed here are not necessarily pathogenetic, but may represent a general reaction to metabolic stress.

Glutamatergic receptors regulate expression, phosphorylation and accumulation of neurofilaments in spinal cord neurons.

Glutamatergic regulation of neurofilament expression, phosphorylation and accumulation in cultured spinal cord neurons was studied. At seven days in culture, 0.15% of the neurons were immunoreactive for non-phosphorylated neurofilaments, but essentially no cells immunoreactive for phosphorylated neurofilaments were seen. The number and size of the immunoreactive cells in culture corresponded well to those of rat and human spinal cord neurons in vivo. In spinal cord cultures, sublethal, long-lasting stimulation of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA)/kainate or metabotrophic receptors, but not N-methyl-D-aspartate receptors, dose-dependently increased the number of non-phosphorylated neurofilament-immunoreactive cells, which was blocked by nifedipine, an antagonist of voltage-sensitive Ca2+ channels. Stimulation of kainate or all non-N-methyl-D-aspartate receptors decreased the expression of medium-molecular-weight neurofilament messenger RNA. Blockade of AMPA/kainate receptors, but not of N-methyl-D-aspartate receptors, increased the amount of phosphorylated neurofilament protein and the number of phosphorylated neurofilament-immunoreactive cell bodies. The phosphorylated neurofilament-immunoreactive cell population was different from the non-phosphorylated neurofilament-immunoreactive neurons, which lost their axonal non-phosphorylated neurofilament immunoreactivity but showed intense cytoplasmic labeling in response to the blockade of AMPA/ kainate receptors. Immunoreactivity for phosphoserine did not change upon glutamate receptor stimulation and blockade. The results show that activation of AMPA/kainate receptors decreases the expression of neurofilament messenger RNA and neurofilament phosphorylation in spinal cord neurons by a mechanism involving active voltage-sensitive Ca2+ channels. Blockade of these receptors seems to disturb axonal neurofilament transport. Because AMPA/kainate receptors mediate chronic glutamatergic death of spinal motor neurons and these receptors have been suggested to be involved in the pathogenesis of amyotrophic lateral sclerosis, the observed alteration in neurofilament phosphorylation and distribution may contribute to the pathogenesis of chronic motor neuron diseases.

Molecular cloning and characterization of the rat inducible nitric oxide synthase (iNOS) gene.

We have cloned and characterized the rat inducible nitric oxide synthase (iNOS) gene. It spans approx. 36kb and is divided into 27 exons and 26 introns. The distribution and length of exons are similar to those in the human iNOS gene. In the 5' flanking regulatory region of the rat iNOS gene, there are a number of putative transcription factor binding sites (>20), many of them probably indispensable for the gene's nuclear factor kappaB (NFkappaB)-dependent induction, but also many which may have a role in its NFkappaB-independent induction pathway. These include cyclic adenosine 3', 5'-monophosphate (cAMP) response elements (CRE), hypoxia responsive element (HRE) and GATA-core elements. Rat models are powerful tools in studies of neurological diseases. Because iNOS is most likely responsible for the harmful consequences of nitric oxide (NO) in general, the cloned rat iNOS gene will further reveal the mechanisms of iNOS inducibility in different cell types during development and disease, including brain diseases, and to promote studies of pharmacological intervention in cases where extensive NO production plays a critical role.

Tetracyclines inhibit microglial activation and are neuroprotective in global brain ischemia.

Ischemic stroke is the most common life-threatening neurological disease and has limited therapeutic options. One component of ischemic neuronal death is inflammation. Here we show that doxycycline and minocycline, which are broad-spectrum antibiotics and have antiinflammatory effects independent of their antimicrobial activity, protect hippocampal neurons against global ischemia in gerbils. Minocycline increased the survival of CA1 pyramidal neurons from 10.5% to 77% when the treatment was started 12 h before ischemia and to 71% when the treatment was started 30 min after ischemia. The survival with corresponding pre- and posttreatment with doxycycline was 57% and 47%, respectively. Minocycline prevented completely the ischemia-induced activation of microglia and the appearance of NADPH-diaphorase reactive cells, but did not affect induction of glial acidic fibrillary protein, a marker of astrogliosis. Minocycline treatment for 4 days resulted in a 70% reduction in mRNA induction of interleukin-1beta-converting enzyme, a caspase that is induced in microglia after ischemia. Likewise, expression of inducible nitric oxide synthase mRNA was attenuated by 30% in minocycline-treated animals. Our results suggest that lipid-soluble tetracyclines, doxycycline and minocycline, inhibit inflammation and are neuroprotective against ischemic stroke, even when administered after the insult. Tetracycline derivatives may have a potential use also as antiischemic compounds in humans.

Spreading depression and focal brain ischemia induce cyclooxygenase-2 in cortical neurons through N-methyl-D-aspartic acid-receptors and phospholipase A2.

Repetitive spreading depression (SD) waves, involving depolarization of neurons and astrocytes and up-regulation of glucose consumption, is thought to lower the threshold of neuronal death during and immediately after ischemia. Using rat models for SD and focal ischemia we investigated the expression of cyclooxygenase-1 (COX-1), the constitutive form, and cyclooxygenase-2 (COX-2), the inducible form of a key enzyme in prostaglandin biosynthesis and the target enzymes for nonsteroidal anti-inflammatory drugs. Whereas COX-1 mRNA levels were undetectable and uninducible, COX-2 mRNA and protein levels were rapidly increased in the cortex, especially in layers 2 and 3 after SD and transient focal ischemia. The cortical induction was reduced by MK-801, an N-methyl-D-aspartic acid-receptor antagonist, and by dexamethasone and quinacrine, phospholipase A2 (PLA2) inhibiting compounds. MK-801 acted by blocking SD whereas treatment with PLA2 inhibitors preserved the wave propagation. NBQX, an alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid/kainate-receptor antagonist, did not affect the SD-induced COX-2 expression, whereas COX-inhibitors indomethacin and diclofenac, as well as a NO synthase-inhibitor, NG-nitro-L-arginine methyl ester, tended to enhance the COX-2 mRNA expression. In addition, ischemia induced COX-2 expression in the hippocampal and perifocal striatal neurons and in endothelial cells. Thus, COX-2 is transiently induced after SD and focal ischemia by activation of N-methyl-D-aspartic acid-receptors and PLA2, most prominently in cortical neurons that are at a high risk to die after focal brain ischemia.

Selegiline treatment after transient global ischemia in gerbils enhances the survival of CA1 pyramidal cells in the hippocampus.

Selegiline (L-deprenyl) has shown neuroprotective effects in a variety of degenerative processes. The present experiments were designed to test whether post-ischemia administered selegiline would alleviate delayed neuronal death of the gerbil hippocampal pyramidal cells following transient global ischemia. Common carotid arteries were occluded for 5 min. Saline or selegiline, 0.25 mg/kg s.c., was administered 2 h after the ischemia followed by a daily injection for either 3 or 7 days. After decapitation, the delayed death of the hippocampal CA1 pyramidal cells was assessed using Nissl-stained sections. In situ hybridization was used to reveal the expression of hsp70 mRNA 1, 3 or 7 days after the ischemia. Animals treated with selegiline for 7 days showed significantly lower damage score (scale 0-3: 0, normal; 1, < 10% of the neurons damaged; 2, 10-50% damaged; 3, > 50% damaged) compared to the saline-treated animals 1.73 +/- 0.18 and 2.41 +/- 0.16 (mean +/- S.E.M., P = 0.0133), respectively. A similar trend was found in animals after the 3-day treatment: 1.68 +/- 0.32 vs. 2.06 +/- 0.25 (P > 0.5). The expression of hsp70 mRNA in the CA1 pyramidal cell layer was strong still 3 days after the ischemic insult but vanished by 7 days. Densitometric measurements using 14C-plastic standards showed that the intensity of the CA1a hsp70 signal on the 3rd day correlated negatively to the cell-damage score (r = -0.72, P < 0.001), suggesting that hsp70 does not serve as a quantitative marker for CA1 neuronal injury in this model. Instead, the hsp70 expression was associated with improved neuronal survival lasting often longer in selegiline-treated animals (P > 0.5). The results show that a low dose of selegiline can alleviate the delayed hippocampal neuronal death in gerbils when administered 2 h after an ischemic insult.

Induction of ornithine decarboxylase mRNA in transient focal cerebral ischemia in the rat.

We used in situ hybridization to localize the long-term changes in ornithine decarboxylase (ODC) expression after a 90 min occlusion of the middle cerebral artery (MCAO) in the rat. The ODC mRNA was induced in the ipsilateral dentate gyrus (DG) and throughout the ischemic cortex at 12 h and still at 3 days after reperfusion. The induction was blocked by an N-methyl-D-aspartate (NMDA) receptor antagonist suggesting that ODC induction is NMDA receptor-mediated. The long-lasting up-regulation detected in regions where no cellular damage usually occurs, favors the hypothesis that ODC expression does not contribute to neuronal death after stroke.

Long-term induction of haem oxygenase-1 (HSP-32) in astrocytes and microglia following transient focal brain ischaemia in the rat.

Haem oxygenase-1 (HO-1) is a stress protein and a rate-limiting enzyme in haem degradation, generating ferrous iron, carbon monoxide and bile pigments. HO-1 has been suggested to be protective against oxidative stress. In the normal rodent brain the enzyme is localized in selected neuron populations, but heat shock, glutathione depletion in vivo and oxidative stress in vitro induce HO-1 predominantly in glial cells. We studied HO-1 expression in the brain following transient occlusion of the middle cerebral artery, and found increased mRNA levels in the ischaemic region from 4 h to 7 days after 90 min of ischaemia. The mRNA levels peaked at 12 h, and were localized perifocally. HO-1-immunoreactive astrocytes and microglial cells were seen in the perifocal area, in the ipsilateral and occasionally in the contralateral hippocampus. Some perifocal neurons were also HO-1-immunoreactive. In the infarcted area HO-1-positive microglia/macrophages were detected in double-labelling experiments. A microassay measuring the conversion of [14C]haem to [14C]bilirubin showed a two-fold increase in haem oxygenase activity in the infarcted core. These observations show a long-term induction of HO-1 protein and its activity following ischaemia-reperfusion brain injury, and indicate increased capacity for haem degradation and the generation of biologically active bile products, carbon monoxide and iron in astrocytes and some microglia/macrophages during focal brain ischaemia.

Expression of beta-amyloid precursor protein mRNAs following transient focal ischaemia.

beta-Amyloid precursor protein (beta APP) can be alternatively processed to result in release of either neurotoxic amyloid beta-peptide, or secreted forms of beta APP, which might have a role in neuronal plasticity and survival. Four different forms of beta APP mRNAs have been described in rodents: APP695, APP714, APP751 and APP770. The two larger forms contain a Kunitz-type serine protease inhibitor domain (KPI). Since previous studies have shown increased APP immunoreactivity following brain ischaemia, we used in situ hybridization histochemistry to determine whether induction of any APP mRNA transcripts take place following focal brain ischaemia. While the hybridization signal of all APP isoforms in the infarct was lost 4-24 h following the insult, APP770 mRNA was slightly upregulated in the ipsilateral cortex and striatum 3 days after 90 min ischaemia. At 7 days post-ischaemia APP770 and APP751 mRNAs were induced in the infarct core and in a thin perifocal zone. KPI-containing APP forms are differentially induced following focal brain ischaemia, possibly as a neuroprotective response to neuronal injury.

Specific induction of protein kinase C delta subspecies after transient middle cerebral artery occlusion in the rat brain: inhibition by MK-801.

Protein kinase C (PKC) consists of a family of closely related Ca2+/phospholipid-dependent phosphotransferase isozymes, most of which are present in the brain and are differentially activated by second messengers. Calcium-dependent PKC activity may cause neuronal degeneration after ischemic insult. PKC is also involved in trophic-factor signaling, indicating that activity of some PKC subspecies may be beneficial to the injured brain. Therefore, we screened long-term changes in the expression of multiple PKC subspecies after focal brain ischemia. Middle cerebral artery occlusion was produced by using an intraluminal suture for 30 min of 90 min. In in situ hybridization experiments, mRNA levels of PKC alpha, -beta, -gamma, -delta, -epsilon, and -zeta were decreased in the infarct core 4 hr after ischemia and were lost completely 12 hr after ischemia. In areas surrounding the core, PKC delta mRNA was specifically induced 4, 12, and 24 hr after ischemia in the cortex. At 3 and 7 d, the core and a rim around it showed increased mRNA levels of PKC delta. No other subspecies were induced. At 2 d, immunoblotting demonstrated increased levels of PKC delta protein in the perifocal tissue, and immunocytochemistry revealed an increased number of PKC delta-positive neurons in the perifocal cortex. In the core, PKC delta-positive macrophages and endothelial cells were seen. Pretreatment with MK-801, an NMDA antagonist, inhibited cortical PKC delta mRNA induction. The data show that focal brain ischemia induces PKC delta mRNA and protein but not other PKC subspecies through the activation of NMDA receptors and that the upregulation lasts for several days in neurons of the perifocal zone.

Two chicken repeat one (CR1) elements lacking a silencer-like region upstream of the chicken avidin-related genes Avr4 and Avr5.

Two repetitive elements of the chicken CR1 family, each located in the 5' flanking region of the avidin-related genes Avr4 and Avr5, have been cloned and sequenced. Both elements are 721 bp in length with 72% identity to a CR1 consensus sequence. They had a 191 bp deletion in a region corresponding to the functional silencer regions previously detected within the CR1 elements upstream of the chicken lysozyme and apoVLDLII genes.

In vivo amplification of the androgen receptor gene and progression of human prostate cancer.

Overexpression of amplified genes is often associated with the acquisition of resistance to cancer therapeutic agents in vitro. We have identified a similar molecular mechanism in vivo for endocrine treatment failure in human prostate cancer which involves amplification of the androgen receptor (AR) gene. Comparative genomic hybridization shows that amplification of the Xq11-q13 region (the location), is common in tumours recurring during androgen deprivation therapy. We found high-level AR amplification in seven of 23 (30%) recurrent tumours, but in none of the specimens taken from the same patients prior to therapy. Our results suggest that AR amplification emerges during androgen deprivation therapy by facilitating tumour cell growth in low androgen concentrations.

Molecular cloning and nucleotide sequence of chicken avidin-related genes 1-5.

Using avidin cDNA as a hybridisation probe, we detected a gene family whose putative products are related to the chicken egg-white avidin. Two overlapping genomic clones were found to contain five genes (avidin-related genes 1-5, avr1-avr5), which have been cloned, characterized and sequenced. All of the genes have a four-exon structure with an overall identity with the avidin cDNA of 88-92%. The genes appear to have no pseudogenic features and, in fact, two of these genes have been shown to be transcribed. The putative proteins share a sequence identity of 68-78% with avidin. The amino acid residues responsible for the biotin-binding activity of avidin and the bacterial biotin-binding protein, streptavidin, are highly conserved. Since avidin is induced in both a progesterone-specific manner and in connection with inflammation, these genes offer a valuable tool to study complex gene regulation in vivo.

Localization of lysozyme mRNA in the labial salivary glands by in situ hybridization in Sjögren's syndrome.

In this study, lysozyme mRNA in labial salivary glands has been localized with in situ hybridization technique using 35S-labeled hen lysozyme cDNA (cDNALZM) as a hybridization probe in normals and in patients with Sjögren's syndrome, 35S-DNALZM:mRNA hybrids were detected only in acinar serous cells, although lysozyme was identified in ductal cells using immunohistochemical techniques. Our results suggest that the serous acinar cells are the only site of lysozyme synthesis in small salivary glands. The presence of lysozyme in ductal cells may be a result of reabsorption from the saliva or concentration from the blood or surrounding tissues.