In situ hybridization--visualization and quantitation of genetic expression in mammalian brain

Changed relative to what? Housekeeping genes and normalization strategies in human brain gene expression studies

Many studies in biological psychiatry compare the abundance of individual messenger RNAs between cases and control subjects or, more recently, between genotype groups. Most utilize some form of normalization procedure, usually expressing the transcript(s) of interest relative to that of a housekeeping gene or genes (also called reference genes), to overcome various sources of experimental error. Indeed, normalization is such a standard procedure that its purpose, principles, and limitations are sometimes overlooked, and some papers lack sufficient information as to its implementation. Here, we review the rationales for normalization and argue that in well-conducted psychiatric gene expression studies using human brain tissue, it is reducing intersubject variability rather than experimental error that is the major benefit of normalization. We also review the conceptual and empirical basis for the category of housekeeping genes-i.e., genes with a ubiquitous and invariant expression. We conclude that the evidence is against any such simple categorization and that a more pragmatic, less dogmatic, approach to the selection and implementation of reference genes is required, which takes into account the particular issues that pertain to human brain tissue studies. This pragmatism extends to the issue of whether normalization should be to one or multiple reference genes. We end by making several recommendations toward a more flexible, transparent, and comprehensive approach to data presentation and analysis. We illustrate the review with examples from studies of schizophrenia and mood disorder.

Gene expression and neuronal activity in schizophrenia: a study of polyadenylated mRNA in the hippocampal formation and cerebral cortex

The abundance of polyadenylated messenger RNA (poly(A)+mRNA) reflects overall gene expression and provides an index on neuronal activity. Poly(A)+mRNA was measured in the hippocampal formation and in occipital, temporal, cingulate and frontal cortices (Brodmann areas 17, 22, 24 and 46, respectively) of 11 schizophrenic and 17 control subjects post mortem by in situ hybridization histochemistry with a 35S-oligodeoxythymidine probe. There were no differences in poly(A)+mRNA between cases and controls, except for a modest decrease in the parahippocampal gyrus of the schizophrenics which may be attributable to cytoarchitectural differences in this area in the disease. The unchanged level of poly(A)+mRNA in all other regions argues against the existence of a widespread or sustained alteration in the metabolic activity of cortical neurons in schizophrenia. It also provides a further indication that the differential expression of individual transcripts reported in the disease is not merely reflecting changes in overall gene expression.

Anomalous binding of radiolabeled oligonucleotide probes to plaques and tangles in Alzheimer disease hippocampus

Several reports indicate that Alzheimer disease (AD) brain contains elevated levels of heat shock 70 proteins. To determine the cellular localization of the heat shock 70 mRNAs, specific oligonucleotide probes were in situ hybridized to AD and control brains. When oligonucleotides were in situ hybridized to brain sections with no AD neuropathology, hybridization was cell-specific and prior ribonuclease (RNase) treatment of adjacent sections resulted in no hybridization signal. However, in situ hybridization to AD hippocampus resulted in heavy grain deposition over senile plaques and neurofibrillary tangles. Despite altering a number of experimental variables, we observed a similar pattern of grain deposition with most of the oligonucleotides tested, including one oligonucleotide specific for glutamic acid decarboxylase mRNA. In situ hybridization with either an RNA probe for glutamic acid decarboxylase or an oligonucleotide probe specific for 18S rRNA did not show this pattern of grain deposition. In control studies a sense hsc70 oligonucleotide showed no grain deposition in either cerebellum or hippocampus. Sections from AD hippocampus pretreated with RNase prior to in situ hybridization demonstrated enhanced grain deposition with the majority of probes tested. Anomalous in situ hybridization to AD hippocampus was usually eliminated by removing formamide from the posthybridization washes, although post-RNase sticking often remained intense. These findings indicate that artifactual probe binding to senile plaques and neurofibrillary tangles may complicate the analysis of in situ hybridization studies using oligonucleotide probes to determine mRNA distribution in AD brain.

Cholecystokinin gene expression in rat amygdaloid neurons: normal distribution and effect of morphine tolerance

Previous studies have shown that repeated opioid administration induces a tolerance to opioid, presumably due in part to an opioid-mediated compensatory increase in brain cholecystokinin (CCK) synthesis and/or release. In this study, in situ hybridization histochemistry was used to examine the effect of morphine tolerance on CCK gene expression in the amygdala of rat brains, by using a 35S-labeled synthetic oligonucleotide probe. CCK mRNA-positive neurons in normal rats were seen throughout the amygdaloid complex, with the most heavily labeled neurons in lateral, basal, and cortical nuclei, followed by the medial nucleus. Only a few labeled neurons were found in central and intercalated nuclei. The development of morphine tolerance in the rat was associated with increased hybridization signals for CCK mRNA in each subnucleus of the amygdala. Increases were seen in the numbers of positively labeled neurons and/or the numbers of hybridization grains per positively labeled neuron. Furthermore, differential patterns of increase in CCK mRNA in morphine tolerant rats occurred in different subnuclei of the amygdala, with the highest magnitude of increase in the cortical nucleus, followed in order by the medial, central, basal, intercalated and lateral nuclei. The present study demonstrated that repeated administration of morphine increased CCK gene expression in the amygdaloid complex, and suggested that the development of the tolerance to morphine analgesia is due, in part, to an increase in CCK activity in the amygdaloid complex. These findings substantiate the hypothesis that long-term administration of opioid may induce a compensatory increase in CCK synthesis and/or release, which then results in a progressive antagonism of opioid analgesia.

Alterations in actin-binding beta-thymosin expression accompany neuronal differentiation and migration in rat cerebellum

The beta 4- and beta 10-thymosins, recently identified as actin monomer-sequestering proteins, are developmentally regulated in brain. Using specific mRNA and protein probes, we have used in situ hybridization and immunohistochemical techniques to investigate the distribution of the beta-thymosin mRNAs and their proteins in developing rat cerebellum. Early in postnatal development, both beta-thymosin mRNAs were expressed at highest levels in the postmitotic, premigratory granule cells of the external granular layer; expression diminished as granule cells migrated to and differentiated within the developing internal granular layer. In addition, both beta-thymosin proteins were present in bundles of cerebellar afferent fibers in the white matter at this time. Throughout the maturation period, both proteins were present in elongating parallel fibers in the upper portion of the molecular layer. Later in cerebellar development, thymosin beta 4, but not thymosin beta 10, was expressed in Golgi epithelial cells and Bergmann processes. Thymosin beta 4 was expressed in a small population of cells with microglial morphology scattered throughout the gray and white matter. Thymosin beta 10 was detected in an even smaller population of glia. Expression of thymosin beta 4 and thymosin beta 10 in premigratory granule cells and in growing neuronal processes is consistent with the possibility that both beta-thymosins are involved in the dynamics of actin polymerization during migration and process extension of neurons.

Hsp70 mRNA induction is reduced in neurons of aged rat hippocampus after thermal stress

Levels of heat-shock 70 mRNAs, relative to those of 18S rRNA, were quantitated in specific cell types of hippocampus of adult and aged rats subjected to identical heat shock regimens. Body temperature changes in response to the heat stress were no different in adult and aged rats. In control rats, as well as 3 h after initiation of heat shock in both adult and aged rats, relative levels of the constitutively synthesized heat-shock cognate 70 (hsc70) mRNA were highest in hippocampal neurons and much lower in glia. No heat-shock protein 70 (hsp70) mRNAs were present in any cell type of control adult or aged rats. In heat-shocked adult rats, the relative levels of the heat-shock-inducible hsp70 mRNAs were highest in a subpopulation of glia, intermediate in granule cells of the dentate gyrus, and lowest in pyramidal cells of Ammon's horn. Relative levels of hsp70 mRNA were several-fold lower in the dentate gyrus granule cells of aged rats compared to relative levels in controls and were also reduced in many pyramidal cells of the hippocampus but not in hippocampal glia. These findings suggest that some neuronal populations in the hippocampus may be at increased risk for stress-related injury in the aged animal.

A method of in situ hybridization combined with immunocytochemistry, histochemistry, and tract tracing to characterize the mRNA expressing cell types in heterogeneous neuronal populations

A rapid, sensitive, non-isotopic in situ hybridization histochemistry protocol is presented to study the expression of mRNA at the single cell level in anatomically complex structures of the mammalian central nervous system. The protocol uses digoxigenin-UTP-labeled riboprobes, freefloating sections, and alkaline phosphatase and horseradish peroxidase detection. Modifications have been introduced which preserve the integrity of marker molecules, and as such enable the simultaneous identification and characterization of CNS cell types by tract tracing, histochemical, and immunocytochemical detection of intra- and extracellular markers. All pretreatments that enhance probe penetration have been omitted without substantial loss in sensitivity. The protocol has been successfully extended to vibratome sections with subsequent plastic-embedding and semithin sectioning, which considerably broadens the general applicability of this fast and easy ISHH method.

Is aging preprogrammed? Observations from the brain/gut axis

Age related differential gene expression occurs in the neuro-enteral axis. Brain and gut organ weight, total RNA, total protein and three peptides were quantified in 4-, 10- and 37-week-old Sprague-Dawley rats. As animals aged, total RNA decreased in the brain (0.65 +/- 0.3-0.28 +/- 0.03 mg/g), but remained stable in the gut (2.6 +/- 0.3-2.9 +/- 0.4 mg/g). Total protein concentration rose in the duodenum (612 +/- 28-734 +/- 34 mg/g), while levels remained stable in the brain (641 +/- 54-666 +/- 34 mg/g). Three peptides were studied, cholecystokinin (CCK), VIP and secretin. With increasing age, significant changes were found only in CCK a true neural-enteral peptide. The concentration of smaller molecular forms of CCK decreased in the brain (248 +/- 18-188 +/- 21 pmol/g), while they remained stable in the duodenum (33 +/- 2-36 +/- 3 pmol/g). By contrast, the concentration of the larger forms of CCK were stable in the brain (36 +/- 3-40 +/- 4 pmol/g), but rose in the gut (89 +/- 14-134 +/- 17 pmol/g). These data indicate that as rats age there is preprogrammed differential control of gene expression between brain and intestine.

Quantitative in situ hybridization analysis of glutamic acid decarboxylase messenger RNA in developing rat cerebellum

The appearance and relative amounts of GAD mRNA in rat cerebellar neurons during postnatal development was studied by in situ hybridization. GAD mRNA content within all GABAergic neurons increased during the first month of postnatal development, but the degree and time course of the increase varied among different neuronal types. In newborn rats, GAD mRNA was present only in the prenatally-formed Purkinje and Golgi cells. GAD mRNA in Golgi cells had reached adult levels by postnatal day 14, while GAD mRNA levels in Purkinje cells reached adult levels one week later. Most basket cells expressed GAD mRNA by postnatal day 14, and final levels were attained one week later. Stellate cells in the bottom two-thirds of the molecular layer attained their final GAD mRNA content by postnatal day 21 whereas stellate cells in close proximity to the pial surface were not yet mature at this age. No GAD mRNA was detected within the external granular layer at any time during development. In adult rat, approximately 40% of cerebellar GAD mRNA was contained within the Purkinje cell population, 38% within the stellate cells, 17% within the basket cells, and only 5% within the Golgi cells. Increases in GAD mRNA within GABAergic neurons during cerebellar development correlated with the timing of neuronal maturation and synaptogenesis in these cell populations, suggesting that synaptic activity affects GAD gene expression in developing cerebellum.

Amplification of the in situ hybridization signal by silver postintensification: the biotin-dUTP-streptavidin-peroxidase diaminobenzidine-silver-gold detection system

Frozen and vibratome sections from the adrenal gland of the rat were hybridized in situ using a biotinylated oligonucleotide probe specific for tyrosine hydroxylase (TH) messenger ribonucleic acid (mRNA). Hybridization was detected using the streptavidin-peroxidase-diaminobenzidine (DAB) system in combination with silver-gold postintensification. The signal appeared as a black coloration and was localized to the cytoplasm of catecholamine-synthesizing chromaffin cells in the adrenal medulla. This coloration was due to the deposition of the silver-gold intensified DAB chromogen onto the probe hybridized to mRNA in carrier organelles. Compared with the conventional peroxidase-DAB labelling, the silver-gold amplified version was more sensitive in detecting TH mRNA. Using this modification, we were able to adapt the procedure to electron microscopy, thereby further localizing the hybridized signal to ribosomes. Because this hybridization detection system produces grains, not just color, this method has the potential for measurement of changes in mRNA levels at the ultrastructural level.

Detection of proopiomelanocortin mRNA by in situ hybridization, using a biotinylated oligodeoxynucleotide probe and avidin-alkaline phosphatase histochemistry

A synthetic 24-mer oligodeoxynucleotide complementary to the region of proopiomelanocortin (POMC) mRNA that codes for the MSH core sequence (alpha MSH/ACTH[4-11]), was synthesized and labelled in the 3'-end by use of terminal transferase. Probes tailed with either [3H]- or biotin-labelled nucleotides could be used for in situ hybridization studies. Biotinylated probes, hybridized to mouse and rat pituitary sections, were detected by avidin-alkaline phosphatase or streptavidin-alkaline phosphatase procedures and development in 5-bromo-4-chloro-3-indolyl phosphate (BCIP)-nitroblue tetrazolium (NBT). Proteinase K pretreatment of sections produced a drastic enhancement of the signal obtained, particularly in strongly fixed, paraffin-embedded material. The non-radioactive in situ hybridization technique compared favourably to radioactive in situ hybridization in terms of rapidity and precision of the localization. Controls involved deletion of the probe to prove that other components of the reaction sequence did not yield stain, digestion with RNase to prove that tissue RNA was necessary to bind the probe, prehybridization (blocking) with unlabelled probe to prove that the biotinylated probe reacted with its anti-sense region and not its tail and Northern blotting to show that the probe reacted with only one species of pituitary RNA, having the size of mouse pituitary POMC mRNA. In addition, adrenalectomy, known to increase anterior lobe POMC levels, resulted in both increased numbers and increased intensity of positive corticotroph-like cells. Synthetic oligodeoxynucleotides labelled with biotin appear to constitute attractive reagents for in situ hybridization studies when supported by appropriate control procedures.

Use of oligodeoxynucleotide probes for quantitative in situ hybridization to actin mRNA

We have employed an analytical approach for the development of an in situ hybridization methodology using synthetic oligodeoxynucleotide probes for actin messenger RNA detection in cultures of chicken fibroblasts and myoblasts. The methodology developed shows that oligonucleotides can complement the use of nick-translated probes in specific situations. Since they can be made to specific nucleic acid regions independent of restriction enzyme sites, they may be the most convenient approach for analysis of gene families among which sequences are highly conserved. However, it was found that oligonucleotides synthesized to different regions of a messenger RNA behave in situ with differing efficiencies, indicating that not all target sequences are equivalent. Therefore it was necessary to screen several oligonucleotide probes to a target molecule to find the optimal one. The convenience of using synthetic DNA probes makes it worthwhile to explore some of these characteristic properties so as to increase the sensitivity of this approach beyond its application to targets in high abundance.