Methods for Profiling mRNA Expression in Neurites

In recent years there has been great progress in the understanding of the genetic basis of neuropsychiatric disorders including those associated with mental retardation. While the genetic defects of such disorders are disparate and each mechanism appears to be novel, the unifying theme behind the development of these diseases is brain development and maintenance of neurological networks.

Neurological networks are made up of individual neurons, each neuron being a separate structural and functional cellular unit. Neurons have special features for the reception of nerve impulses from other neurons, the effect of which may be either excitation or inhibition, and conduction of nerve impulses. Neurons commonly have long cytoplasmic processes known as neurites which end in close apposition to the surfaces of other cells. The ends of these neurites are called synaptic terminals and the cell-to-cell contacts they make are known as synapses. The neurites in higher animals are usually specialized to form dendrites and axons which conduct impulses toward and away from the cell body, respectively. The arrival of an impulse at a terminal triggers the process of synaptic transmission. This event usually involves the release of a chemical compound from the neuronal cytoplasm invoking a response in the postsynaptic cell. Neurons of the central nervous system consist of discrete segments including the cell body, the dendrites and the axon. While most nerve cells conform to this basic structure, there is a wide range of structural diversity based upon the specific function of the cell within the body.

It has been shown that these polarized cells contain a variety of cytoplasmic and membrane-bound proteins differentially distributed throughout the axon, dendrites, and cell body of the neuron. It is believed that neurons of the central nervous synthesize proteins locally, at or near postsynaptic sites which are independent of the cell body. Ultrastructural studies have revealed that polyribosomes are preferentially located either beneath post-synaptic sites or occasionally associated with membrane specializations on dendrites. It has been suggested that these anatomical structures represent the protein synthetic machinery necessary to translate and post-translationally modify different classes of protein in neurons. An energy-dependent mechanism for the selective transport of RNA in neurons has also been shown. The nature and distribution of the RNAs present in these cells, however, is poorly understood.

In situ hybridization (ISH) studies have been successful in identifying very few mRNAs in neuronal processes. Studies using in situ hybridization and Northern blot analyses of synaptosomal RNA fractions with the AMPA-GluR1, -GluR2, GluR3 and GluR-4 and kainate-sensitive GluR5 and GluR6 receptor subunits failed to reveal mRNAs at dendritic locations. (Craig, A. et al., Neuron 1993, 10, 1055-1068; Chicurel, M. et al. J. Neurosci 1993, 13, 4054-4063).

Microdissection of individual neurites has now revealed a large number of mRNAs, including members of the glutamate receptor family, second messenger components, and components of the translational control apparatus, present in hippocampal neurites. It has now been found that the profiles of expressed mRNAs from discrete segments of the same neurons have different characteristics. These differences in expressed mRNA can be used as a means to specifically target discrete segments of the neuron and to identify and diagnose genetic neurological disorders at the molecular level.

An object of the invention is to provide a method of identifying neurite cDNA clones which comprises determining the mRNA expression in selected neurites, comparing the relative levels of mRNA expression, and identifying neurite cDNA clones based on the level of mRNA expression.

Another object of the invention is to provide neurite cDNA clones.

Another object of the present invention is to provide a method of profiling mRNA expression in a selected neurite which comprises converting an mRNA population in a soma or process of a selected neurite into cDNA, making the cDNA double stranded, linearly amplifying the double stranded cDNA into aRNA, and using the aRNA as a probe in reverse phase Northern analysis to produce an mRNA expression profile.

Another object of the present invention is to provide a method of diagnosing a condition associated with an mRNA expression pattern which comprises determining the relative levels of mRNA expression in selected cells associated with a selected condition, comparing the relative levels determined with established controls, and diagnosing a condition based upon the comparison of the mRNA expression levels.

Yet another object of the present invention is to provide a method of treating a condition associated with an mRNA expression pattern which comprises determining the relative levels of mRNA expression in selected cells associated with a selected condition, and altering the relative levels with an effective amount of an agent capable of altering mRNA expression.

FIGS. 1A-1F are a differential display of three representative cells. A single oligonucleotide, OPA-5, served as the 3'-primer. In combination with three different modified oligo-dT.sub.11 primers (oligo A, 5'-T.sub.11 AC-3'; oligo B, 5'-T.sub.11 CA-3'; oligo C, 5'-T.sub.11 GC-3'), differential display reactions were carried out on separate neurites (HP3-7 and HP3-8) (FIG. 1B), proximal (HP2-10) and distal (HP2-12) segments of the same process as well as another separate process (HP2-11) (FIG. 1D), and distal branch points (HP3-5 and HP3-6) of the same process (FIG. 1F) from single hippocampal cells. Each of these neurites were isolated as described in Example 1. In FIGS. 1B and 1D, the corresponding soma (HP 3-9 in FIG. 1B, and HP2-13 in FIG. 1D) were isolated. In addition to the large number of mRNAs expressed in some neurites (for example, lane 5 in FIG. 1B), there are some commonly shared PCR products between neurites or neurite segments (closed arrowheads) and between neurites and their cell bodies (open arrowheads). These data were reproduced a minimum of three times. In FIGS. 1A, 1C, and 1E, a phase-contrast photomicrograph of the cell and the neurites of interest prior to isolation are shown. Dark bars perpendicular to each process represent the approximate transection point.

Different regions of the central nervous system are populated by functionally and anatomically distinct synaptic connections. During synapse construction, maintenance and remodeling, it is believed that proteins are selectively transported to these distinct connections. As a result, the specificity of the nervous system is established and modified, at least in part, by these protein targeting mechanisms. It has now been found that mRNA expression profiles are specific for messages seen in neurites. A method has now been developed for identifying neurite cDNA clones by determining the mRNA expression in selected neurites, comparing the relative levels of mRNA expression, and identifying neurite cDNA clones based on the level of mRNA expression.

Isolated hippocampal cells free of overlapping processes from neighboring cells were identified in low density cultures. Under these conditions, neurons grow as isolated cells or in small 2-4 cell groups, either directly on the substrate or on glial cells. Neurons are identified by morphological criteria involving synaptic interactions. Individual proximal and distal neurites were harvested by transecting them at varying distances from the cell body and aspirating them into a micropipette containing the reagents necessary for the first step in the antisense RNA (aRNA) amplification procedure. In a number of cases, multiple processes were isolated from a single cell followed by the aspiration of the cell body. Individual neurites or cell bodies were processed through the following aRNA amplification procedure to produce an mRNA expression profile. The mRNA population in the cell soma or cell process was converted into a complementary DNA (cDNA) using an oligo-dT-T7 primer. After the cDNA was made double-stranded, it was linearly amplified into aRNA using T7 RNA polymerase. For reverse Northern analysis, aRNAs served as a probe for mRNA expression profiles. In subsequent experiments, aRNAs were made into double-stranded cDNAs and used as templates for experiments using the polymerase chain reaction PCR).

The population of the mRNAs in neurites was assessed initially by mRNA expression profiling. Southern blots containing cloned cDNAs encoding members of the ionotropic glutamate receptor family were probed with radiolabeled aRNA from individual neurites or cell bodies. Glutamate receptors, classified into N-methyl-D-aspartate (NMDA), .alpha.-amino-3-hydroxy-5-methyl-isoxazole-4-propionate (AMPA; GluR1-4), and kainate (GluR5-7) subtypes are the primary mediators of excitatory synaptic transmission in the brain. These receptors also play a role in the biochemical events associated with excitotoxicity and long-term potentiation (LTP), a specialized form of synaptic plasticity. The qualitative mRNA expression of multiple members of the glutamate receptor family was examined. All neurites expressed GluR1, GluR2, GluR4 and NMDAR1 mRNA. In a majority of neurites, mRNA expression of GluR3 (15/19), GluR5 (14/19), was observed; while expression for GluR6 mRNA was detectable in approximately one half the neurites studied. The presence of these subunits was confirmed by subunit specific PCR. In contrast, only one neurite showed detectable hybridization signals for NR2a and NR2c mRNA. Thus, neurite cDNA clones may be identified by determining the mRNA expression for glutamate receptors. Examples of preferred glutamate receptors include, but are not limited to, GluR1, GluR2, GluR3, GluR4, GluR5, GluR6, GluR7, NR2a and NR2b.

A number of neurite cDNA clones have now been identified in accordance with the methods of the present invention. Two of these clones have been found to correspond to farnesyl diphosphate (FPP) synthase mRNA identified in two separate distal processes. As part of the isoprene biosynthetic pathway, FPP synthase generates the farnesyl moiety ultimately transferred to the COOH-terminal CaaX motif of mammalian ras proteins. Two other cDNAs have been found to have sequence similarity with mRNAs for the .gamma.-subunit of the interleukin-2 receptor and the tumor necrosis factor inducible protein A20. The remaining cDNA clones have little sequence similarity with any published gene sequences. By sequencing the cDNA clones corresponding to full-length RNAs, it is possible to identify the role the primary sequence and secondary structural characteristics of the mRNA play as recognition elements in targeting and transport of the mRNA. The cDNA clones of the present invention are identified in Table 1 as SEQ ID NO: 1 through SEQ ID NO: 28.

The cDNA clones of the present invention can be used in the diagnosis of neuropsychiatric diseases. Human genes containing unstable triplet repeats are associated with several neuropsychiatric diseases including, but not limited to, Huntington's disease, spinal and bulbur muscular atrophy and spinocerebellar ataxia type 1. These diseases show a variety of clinical symptoms making them difficult to diagnose. However, an understanding of these diseases at the molecular level provides the diagnostic laboratory with the capability to test directly whether an individual's DNA contains the disease-causing mutation, either as a confirmation of a clinical diagnosis or prior to any symptoms. For example, fragile X chromosomes, associated with mental retardation in most males that have it and, to a lesser degree, some of the females heterozygous for it, can be identified using cDNA clones of the present invention having SEQ ID NO: 1 through 15. The cDNA clones of the present invention can also be used in prenatal screening.

Distinct relative variations in mRNA expression have also been identified using reverse Northern blot analysis. Differences in the relative levels of glutamate receptor mRNAs expressed between a neuronal process or processes from the same cell and its cell body have been observed. In a number of cells, the qualitative expression patterns were very similar, yet the relative intensity of the hybridization signal was more profound for specific subunits. For example, relative levels of NMDAR1 and GluR5 mRNA were clearly elevated in HP9a, the apical neurite, versus HP9b, the basal neurite, or the soma. This tendency was complimented in other cells which displayed a more differentiated qualitative pattern of glutamate receptor mRNA expression. This differentiated pattern of mRNA expression contributes to the physiological function of a synapse. These differences in the relative levels of mRNA expression result in the occurrence of different neuronal processes in the same cell. In addition to glutamate receptors, the expression of several other cDNAs has been assessed with reverse Northern blot analysis. Previous reports have demonstrated the dendritic localization of the .alpha.-subunits of the Ca.sub.2+ /calmodulin-dependent protein kinase (CaMK II) mRNA. It has now been found that mRNA expression in distal and proximal segments of isolated neurites is strongly positive for CaMK II. Thus, specific segments and processes of a neuron may be targeted by determining the profile of mRNA expression involved in the process, identifying an mRNA with a high level of expression and targeting an agent to that mRNA.

Alzheimer's disease, amyotrophic lateral sclerosis, Huntington's disease and epilepsy are all examples of disorders which involve the degeneration of specific groups of neurons in discrete regions of the nervous system. It appears that different regions of the nervous systems, and specifically, different types of neurons possess different sensitivities to agents. Therefore, to assess the neurotoxicity of an agent, the agent should be assessed in neurons isolated from different regions since it appears that some neuron types are more sensitive than others. It has now been found, however, that neuron types can be identified by determining the mRNA expression in selected neurons, comparing levels of mRNA expression in the selected neurons with levels in known neurons, and identifying the neuron type based on the level of mRNA expression. Thus, the neurotoxicity of an agent for a specific neuron type may be assessed in a culture of mixed neuron types. In this method, it is preferred that mRNA expression is determined by an aRNA amplification procedure.

The complexity of mRNA expression in neurons was investigated further using a PCR-based assay, differential display developed by Liang P. et al., Science 1993, 257, 967-970. In these experiments, a single 10 mer (OPA-5;5'-AGGGGTCTTG-3', SEQ ID NO: 29) which serves as the 5'-primer and a modified polythymidine primer containing a two base extension were used to amplify specific populations of the polyadenylated RNA pool. In each of these reactions, banding patterns unique to the soma or process and the combination of primers used were observed. Differential display (DD) of mRNAs from three sets of cells are shown in FIG. 1. The complement of DD profile of neurites exhibited a large number of mRNA species. These patterns of PCR products indicate that mRNAs are differentially distributed. For example, in a typical cell in which two neurites are transected and isolated, there are transcripts that migrate at similar molecular sizes between individual neurites. This circumstance is repeatedly seen in proximal and distal segments of the same process and in distal branch points of a single process. While there are some products found concomitantly in one or more neurites and between neurites and their corresponding cell bodies, there are numerous transcripts that are unique to individual neurites or segments of individual neurites. Samples of media and tRNA did not display any banding pattern using differential display.

The contamination of neurites with surrounding glia or astroglial processes was assessed by glial fibrillary acidic protein (GFAP) mRNA. Neurite and soma preparations used in these experiments were determined to be free of GFAP mRNA. The existence of multiple mRNAs in neuronal processes suggests that mRNA transport and local protein synthesis plays a role in the regulation of neuronal physiology, development and regeneration.

In the present invention, mRNA expression patterns can be used to diagnose conditions related to the presence or absence of synthesized proteins. In recent years, several human neurological diseases have been identified resulting from the expansion of trinucleotide repeats. These triplet repeats are normally polymorphic and exonic, though not always coding. In disease states, they become markedly unstable and may expand moderately or by thousands of repeats in a single generation thereby altering gene expression, message stability or protein structure. Thus, in a diseased state, alterations in normal mRNA expression patterns or profiles would be expected. Methods related to diagnosing a condition associated with an mRNA expression pattern comprise determining the relative levels of mRNA expression in selected cells associated with a selected condition; comparing the relative levels determined with established controls; and diagnosing a condition based upon the comparison of the mRNA expression levels. In this method, it is preferred that the selected cells are neurons.

Measuring of mRNA expression patterns as disclosed in the present invention can also be used in methods for treating a condition associated with an mRNA expression pattern. Levels of mRNA expression in selected cells associated with a selected condition are determined and compared to normal levels in the same type of cells. If the level of mRNA expression is abnormal, an effective amount of an agent capable of altering mRNA expression is administered. In this method, it is preferred that the mRNA expression levels be measured in neurons. Examples of agents capable of altering mRNA expression include, but are not limited to, antisense oligonucleotides and pharmacological agents such as dopamine, serotonin and cyclic AMP. Such agents are administered in an effective amount either alone or in conjunction with a suitable pharmaceutically acceptable carrier. By "effective amount" it is meant a concentration of an agent to be administered which is sufficient to modulate expression of mRNA. Such concentrations can be routinely determined by those of skill in the art upon this disclosure and will depend upon the age, weight and condition of a patient. Suitable pharmaceutically acceptable carriers are well known in the art and are described for example in Gennaro, Alfonso, Ed., Remington's Pharmaceutical Sciences, 18th Edition 1990. Mack Publishing Co., Easton, Pa., a standard reference text in this field. Pharmaceutical carriers may be selected in accordance with the intended route of administration and the standard pharmaceutical practice.

The following nonlimiting examples are provided for illustrative purposes.

Hippocampi were dissected from ED20-21 rat fetuses, dissociated in trypsin and plated at 30,000-100,000 viable cells/ml of media onto poly-lysine covered glass coverslips held in 35 mm tissue culture petri dishes. One day after plating, 0.5 ml of media was replaced with media containing 20 mM potassium. The cultures were subsequently fed once a week with a drop of the high K.sup.+ media. Experiments were performed after 21-28 days in culture. Cell bodies and their neurites were taken from cells cultured from different animals on three different days under similar conditions. During each of these sessions, a sample of culture media was also aspirated and processed through aRNA processing to assess the possible presence of mRNAs in the culture media from dying cells.

Reactions were carried out in 25 .mu.l volumes using the Hot Start technique (Perkin Elmer, product #N808-0100) with an upper-lower ratio of 1.5 to 1. Reactions contained 200 .mu.M of dATP, dGTP and TTP, 4 .mu.M dCTP, 5 .mu.Ci of .sup.33 P-dCTP or 4 .mu.Ci .sup.32 P-dCTP or 18.75 .mu.Ci of .sup.35 S-dATP (NEN/Dupont), 0.4 .mu.M of OPA-5 or other 10 mers (Operon Technologies), 0.6 .mu.M of oligo A, B or C, 2.5 mM MgCl.sub.2, 1.25 units of AmpliTaq polymerase (Perkin Elmer), and 1 .mu.l of a 1:10 dilution of DNA previously processed through a single round of aRNA amplification. Under these conditions, the primers are in vast excess to the amount of template used for the reaction. Reactions were cycled 35 rounds at 94.degree. C. for 30 seconds, 40.degree. C. for 90 seconds and 72.degree. C. for 45 seconds, followed with a final 5 minute elongation at 72.degree. C. in a Biosycler thermocycler. Approximately 5 .mu.l of the reaction was loaded onto a 6% acrylamide/7M urea gel. Gels were vacuum dried and apposed to XAR film. Gels shown in FIG. 1 used .sup.33 P-dCTP. The reactions exhibited a banding pattern unique to a sample.

Micropipettes containing reagents for first strand cDNA synthesis also contained 5 .mu.M dithiothreitol and RNAsin (Promega, Madison, Wis.) at 0.5 units/.mu.l. The efficiency of amplification, as based upon trichloroacetic acid precipitated counts and Northern blot analysis to assess the size distribution of mRNAs, did not differ with or without digitonin. Antisense RNA processed through two rounds of amplification was added to Southern blots containing equal amounts (500 ng) of glutamate receptor cDNAs linearized with the appropriate restriction enzyme and applied by vacuum onto a slot blot apparatus. A random-primed vector cDNA clone, pBluescript SK (Stratagene, La Jolla, Calif.) was added to stripped blots to demonstrate the presence of DNA in each slot. Scanning densitometry analysis was performed on autoradiographs using a Scanning Laser Densitometer (Molecular Dynamics, Sunnyvale, Calif.). Hybridization signals were normalized to ribosomal RNA for each neurite or soma studied.