α²-Adrenergic Receptors in Human Dorsal Root Ganglia: Predominance of α2band α2cSubtype mRNAs

Human DRG was obtained with institutional approval from autopsy specimens (rapid and routine). DRG tissue samples were immediately frozen in liquid nitrogen and stored at −70°C until assays were performed. Representative 5-μ cryostat sections were evaluated histologically with hematoxylin and eosin staining to confirm that tissues collected were DRG and not thickened portions of nerve roots. Total cellular RNA was extracted from DRG tissues based on the guanidinium thiocyanate–phenol–chloroform method originally described by Chomczynski and Sacchi 30using RNA STAT-60 (TEL-TEST “B” Inc., Friendswood, TX). As a positive control for final RT-PCR, RNA was also extracted from other human tissues (vena cava and aorta) known to express all three human α²AR mRNA subtypes. 14,31 

In addition to tissue controls, cells containing each αAR subtype (Chinese hamster ovary cells stably expressing individual human α²AR subtypes and rat-1 fibroblasts expressing human α¹AR subtypes) were used as positive and negative controls, respectively, to test for specificity and sensitivity of α²AR primers used in RT-PCR. Chinese hamster ovary cells were grown as monolayers in F-12 Nutrient mixture (Ham, Gibco/BRL, Gaithersburg, MD) supplemented with 10% fetal bovine serum, penicillin (100 U/ml), streptomycin (100 μg/ml), and G418 (500 μg/ml) in 5% CO²at 37°C. Rat-1 fibroblasts were grown similarly in Dulbecco’s modified Eagle medium. RNA was extracted from these control cells using the protocol described previously, with final RNA pellets resuspended in ribonuclease-free water, and the RNA was quantitated using a spectrophotometer at 260 and 280 nm and stored at −70°C for later use.

To determine the presence or absence of specific α²AR mRNAs in human DRG, qualitative RT-PCR was performed initially. Human α²AR subtype–specific primers were synthesized at Duke University Medical Center. Sense and antisense primers (20 mer) were adapted from Eason and Liggett 31with some modifications, as shown in table 1. Using GeneAmp RNA PCR Core Kit (Perkin Elmer, Foster City, CA), RT of 1 μg RNA was performed in a 20-μl reaction mixture containing a final concentration of 5 mM MgCl²; 50 mM KCl; 10 mM Tris-HCl; 1 mM each of dGTP, dATP, dTTP, and dCTP; 20 U ribonuclease inhibitor; 50 U Muloney leukemia virus RT; and 2.5 μM random hexamers. As a negative control, reactions without RT were performed to ensure that contamination with genomic DNA did not occur. RT reactions were run for 60 min at 42°C, 10 min at 95°C, and 10 min at 4°C. The resulting mixture was then amplified by PCR in a 100-μl reaction mixture containing a final concentration of 2 mM MgCl²; 50 mM KCl; 10 mM Tris-HCl; 200 μM each of dGTP, dATP, dTTP, and dCTP; 2.5 U AmpliTaq DNA polymerase; 0.4 μM each of α²AR sense and antisense primers; and 10% dimethyl sulfoxide. PCRs were performed using the Twin Block System Easy Cycler Series (ERICOMP, San Diego, CA). Optimal amplification conditions for human DRG were determined to be 3 min at 95°C; 35 cycles of 1 min denaturation at 95°C; 1 min annealing (57°C for α^2a, 65°C for α^2b, and 60°C for α^2c); and 1 min extension at 72°C. Final PCR products (10 μl) were separated by size electrophoretically through an agarose gel (1%) and visualized using ethidium bromide staining. Amplified products were confirmed to contain specific α²AR target by DNA sequencing using the fmol  DNA Sequencing System (Promega Corporation, Madison, WI).

After the presence or absence of α²AR subtypes in human DRG was established and general PCR conditions determined, quantitative competitive PCR was performed. For the heterologous competitor template, a sequence from the plasmid pGEM-7Zf(+) (Promega) was used. The use of a competitor rather than native DNA has the advantage of minimizing differences in GC content between subtypes, because it should be amplified with equal efficiency in each reaction. Using α²AR sense and antisense primers to flank the pGEM-7Zf(+) sequence, competitor primers (40 mer) were constructed as shown in table 2. Using 9 ng pGEM-7Zf(+) in a final volume of 100 μl containing the PCR mixture (described previously) and 0.5 μl of each of the sense and antisense competitor primers, a 540-base pair (bp) product (20 bp on either end corresponding to the α²AR subtype–specific primers) was amplified corresponding to pGEM-7Zf(+) nucleotides 2,021–2,520. 32Conditions for amplification were 3 min at 95°C; 40 cycles of 30 s denaturation at 95°C and 30 s annealing at 58°C; and 1 min extension at 72°C. Final PCR products were separated by size on a 1% agarose gel and visualized using ethidium bromide staining.

Final construction of a human α²AR subtype–specific plasmid was accomplished by ligating 1 μl of the PCR product (with deoxyadenosine 3′ overhangs produced by Taq  polymerase) into 50 ng of the pCR 2.1 vector (which has 3′ deoxythymidine residues) in a final volume of 10 μl using the Original TA Cloning Kit (Invitrogen Corp., Carlsbad, CA); the product was subsequently transformed into INVαF′ cells. To analyze and confirm successful transformation, isolated plasmids were digested with Eco  RI (which cleaves on either side of the competitor insert), and PCR was performed using the α²AR subtype–specific 20-mer sense and antisense primers to amplify the competitor insert. Products from these reactions were visualized on a 1% agarose gel stained with ethidium bromide. These plasmids were used subsequently as templates to create competitor RNA transcripts.

To create competitor RNA transcripts, plasmids containing α²AR competitor sequence were linearized with Hind  III (New England Biolabs, Inc., Beverly, MA), treated with proteinase K (Boehringer Mannheim, Indianapolis, IN) at 37°C for 30 min, extracted with phenol–chloroform, and precipitated with isopropanol. Single-stranded RNA competitor transcripts were then synthesized using T7 polymerase and MEGAscript reagents (Ambion, Inc., Austin, TX) according to the manufacturer’s directions. Molecules of competitor RNA were calculated (discussed later), and yeast transfer RNA was added to each competitor dilution at a final concentration of 10 ng/μl to stabilize the concentration and prevent degradation. Competitor dilutions were stored at −70°C until use.

Molecules of competitor per microliter were calculated from the following relation:

g/M of competitor = number of bp of competitor RNA transcript × 325 d/bp

g/μl competitor (calculated from abs²⁶⁰) ÷ g/M competitor × 6.02 × 10²³(Avogadro number) molecules/M = molecules/μl competitor

To ensure that the amplification rate was identical for target and competitor species, amplification efficiency of each target α²AR subtype and the corresponding α²AR competitor was measured. Two to 5 ng α²AR target plasmid and the corresponding competitor plasmid were coamplified in a final volume of 100 μl PCR reaction mixture using α²AR sense and antisense primers (table 1). Amplification conditions were as described previously in the general RT-PCR section. For each α²AR subtype, nine identical reactions were prepared and the appropriate sample removed every five cycles (0–40 cycles) and placed on ice. PCR products (5 μl) were separated on a 3% NuSieve GTG agarose gel (FMC BioProducts, Rockland, ME) and stained with ethidium bromide. By design, amplified target and competitor species could be distinguished on the basis of difference in size (competitor size is 540 bp; see table 1for α²AR subtype product size).

For final competitive RT-PCRs, serial dilutions (10⁸–10²molecules) of the competitor templates were added to a series of RT reaction tubes containing 0.5 μg of total cellular RNA from human DRG in a final 10-μl reaction mixture; final reagent concentrations were 5 mM MgCl²; 50 mM KCl; 10 mM Tris-HCl; 1 mM each of dGTP, dATP, dTTP, and dCTP; 10 U ribonuclease inhibitor; 25 U Muloney leukemia virus RT; and 2.5 μM random hexamers. RT reactions were performed as described previously in the general RT-PCR section and subsequently amplified by PCR in a total volume of a 50-μl reaction mixture; final concentrations were 2 mM MgCl²; 50 mM KCl; 10 mM Tris-HCl; 200 μM each of dGTP, dATP, dTTP, and dCTP; 1.25 U AmpliTaq DNA polymerase; 0.4 μM each of α²AR sense and antisense primers, and 10% dimethyl sulfoxide). Amplification conditions also were identical to those described previously in the general section concerning RT-PCR. Final PCR products (5–10 μl) were separated by size on a 3% NuSieve GTG agarose gel and visualized using ethidium bromide staining. Unlike qualitative PCR, in which relative quantitation necessitates use of the linear portion of the amplification efficiency curve, in competitive PCR, the ratio of target to competitor remains constant throughout amplification; thus, valid quantitation continues into the plateau portion of amplification. Therefore, we used 35 cycles to optimize product visualization on the gel for capture by the digital camera.

To determine amplification efficiency and competitive RT-PCR results, final gels were photographed using a DC-40 digital camera (Eastman Kodak Co., Rochester, NY). Band intensity was analyzed using Digital Science 1D Image Analysis software (Eastman Kodak Co.). Pixel data were exported and graphed using Microsoft Excel (Microsoft Corp., Seattle, WA). For the amplification efficiency curve, the logarithm of band intensity for the target and the competitor (y -axis) were graphed as a function of the number of cycles (x -axis). The slope of the linear portion of the curve (representing efficiency of amplification) was then compared for the two curves. For competitive RT-PCR, the logarithm of the ratio of the target and competitor products (y -axis) was graphed as a function of the logarithm of the initial amount of competitor molecules (x -axis) added to the reactions. The values obtained for band intensity were corrected for the difference in size between the target and the competitor (i.e. , the band intensity of the target was divided by the ratio of target size to competitor size). The amount of initial target molecules present was determined from the intersection of the curves (the x -intercept where target and competitor were equal [log¹⁰1 = 0]; this log¹⁰number, extrapolated from the curve, was then converted back to a real number to yield the number of molecules of target).

Because DRG was obtained at variable intervals from death in each patient, potentially, more RNA degradation may have occurred in some patients compared with others. Therefore, to normalize for absolute amount of starting RNA, final molecules of α²AR subtype mRNA from each patient were converted to percent of total α²AR mRNA present before comparison with other patients. Because data were not distributed normally, the Wilcoxon signed rank test (a nonparametric version of the paired t  test) was used to determine differences in percentages of α²AR subtype expression in human DRG. These results were confirmed using a general estimating equation 33,34;P < 0.05 was defined as statistically significant. Data are presented as the median [interquartile range] and reported to two significant figures.

Twenty-eight DRG samples were obtained from seven patients. Patient age ranged from 26–85 yr; all causes of death were unrelated to neurologic disease. Furthermore, no tissue was harvested from patients with indwelling epidural or spinal catheters or those who were administered long-term intravenous or intrathecal analgesic agents (e.g. , opioids). Time between death and DRG tissue harvest ranged from 2.5–20 h. No patient was administered medications known to activate or inhibit α²ARs.

Specificity of α²AR PCR primers was confirmed by lack of amplified products in RT-PCRs using RNA isolated from stable cell lines that expressed either of the other two α²AR subtypes or all three α¹AR subtypes (data not shown) and expression profile in human vena cava and aorta, which matched that previously described by our laboratory. 14To ensure contaminating genomic DNA was not present, reactions were performed without RT; these reactions yielded no product. Initial qualitative α²AR RT-PCRs showed that all three α²AR subtype mRNAs are capable of being amplified from human DRG; therefore, quantitative competitive RT-PCRs were performed for all three α²AR subtypes.

Quantitative competitive RT-PCR was used to determine relative α²AR subtype mRNA expression in human DRG. Amplification efficiency was equal for each α²AR target and its corresponding competitor (fig. 1). A representative competitive PCR (agarose gel and quantitation) for all three α²AR subtypes in the same DRG is shown in figure 2. Summative results from quantitative competitive RT-PCR for all patients revealed the presence of all three α²AR subtypes in human DRG (table 3). Although total overall expression of α²AR is low, it is reproducible and present at all levels of human spinal cord. Although evaluation of table 3suggests that α²AR subtype mRNA expression may vary between levels, limitations in sample size (28 DRG samples from four distinct levels of spinal cord) provide only enough statistical power to compare α²AR subtype mRNA expression overall in human DRG (results pooled from all levels of human spinal cord). Overall, α^2band α^2cAR mRNA predominate in human DRG, accounting for more than 95% of α²AR mRNA in DRG at all spinal cord levels (P < 0.001). Although there is a trend toward overall α^2b> α^2cmRNA expression, it does not reach statistical significance; this is not surprising, because evaluation of table 3reveals α^2b> α^2conly at cervical and sacral spinal cord levels. These results contrast with our previously reported α²AR mRNA expression in human spinal cord, where α^2a≈α^2bmRNA predominates at all spinal cord levels, with α^2cAR mRNA present only in very small quantities in the lumbar region. 25 

This study is the first to describe α²AR subtype mRNA heterogeneity in human DRG. Overall expression of α²AR mRNA in DRG is low but reproducible. α^2band α^2cAR subtypes predominate (α^2b≈α^2c), accounting for 95% of the total α²AR mRNA in DRG at all levels of human spinal cord. Although ribonuclease protection assays are generally considered the gold standard for quantitating mRNA levels, ribonuclease protection is not as sensitive as RT-PCR in analyzing low levels of RNA. A potential drawback to RT-PCR is the issue of reproducibility for quantitation. In the current study, we addressed this issue by strictly implementing competitive RT-PCR conditions; similarity in the percentage of α²AR subtype in RNA among patients gives us confidence in our results. Our results in human DRG α²AR subtype expression contrast with mRNA expression in both human spinal cord and rat DRG; therefore, it is important to evaluate each of these tissues in more detail.

Although α²AR agonists have been used clinically to treat pain, their exact site or sites of action remain to be determined. Potential sites of action include brain, spinal cord, DRG, and sensory neurons. In knockout mice, the α^2aAR has been linked to nociception, 21,35although some α^2bor α^2cARs may also be present. 35These later findings are bolstered by the presence of predominantly α^2aand α^2cARs mRNA and protein in rat spinal cord. 26,27,36,37In rat spinal cord, α^2aAR protein is colocalized with afferent substance P–containing neurons, whereas α^2cARs are present on spinal interneurons. 38Although animal models classically have been used to evaluate nociceptive pathways, species differences in spinal cord adrenergic receptor expression may limit extrapolation of results to humans. α-Adrenergic receptor subtypes in human spinal cord have been evaluated by several investigators, including us. 25,28,39Human spinal cord α²AR distribution is limited to gray matter, ventral more than dorsal, with the following general regional abundance: sacral > cervical > thoracic = lumbar. 25Cell bodies containing α²AR mRNA are located in intermediolateral (thoracic and lumbar) and intermediate (sacral) cell columns (lamina VII), the dorsal nucleus of Clarke (thoracic and lumbar), and at all levels in sensory dorsal horn laminae I, II, III, IV, and V and motor ventral horn lamina IX. 25In terms of α²AR subtype mRNA distribution, α^2aand α^2bAR mRNAs predominate at all sites in human spinal cord, with α^2cAR mRNA present at only a few locations restricted to the lumbar spinal cord 25; α^2aARs predominate via  ligand binding; however, because these studies used spinal cord homogenate membranes, nothing can be stated regarding regional localization of α²AR subtype protein. 28 

To understand the role of α²ARs in human DRG, it is important to remember that the portion of α²AR subtype mRNAs in DRG cells translated into protein are then transported to the spinal cord dorsal horn. These receptors constitute the population of presynaptic α²ARs on primary dorsal horn afferent neurons, representing approximately 20% of all dorsal horn α²ARs. 29Although activation of presynaptic α²ARs, including those originating in DRG, has been associated with neuronal inhibition, 18activation of DRG neurons by norepinephrine has also been reported to cause both depolarization and hyperpolarization. 40–42Norepinephrine-mediated suppression of afferent dorsal horn signaling is a known mechanism for spinal cord inhibition of nociceptive input to the brain. 43,44Although the presynaptic location of DRG-generated α²ARs in spinal cord is clear, spinal cord dorsal horn–generated α²ARs may be presynaptic or postsynaptic, with postsynaptic locations on spinal cord dorsal horn interneurons or neurons directed to more rostral locations. The significance of differences in presynaptic versus  postsynaptic α²AR subtype distribution between DRG and spinal cord dorsal horn for the processing of spinal cord pathways is not understood.

In terms of receptors involved in pain, rat models have been used to determine specific α²AR subtypes in DRG in the absence or presence of neuropathic pain. Cho et al.  45described α^2c>> α^2a>> α^2bAR subtype mRNA DRG expression in control rats, compared with increased α^2aand decreased α^2cmRNA expression in a model of neuropathic pain. These findings are supported by Nicholas et al. , 27who also described α^2aand α^2c(and absence of α^2b) mRNA in rat DRG. In contrast, Gold et al.  46reported all three subtypes present in rat DRG, α^2band α^2cAR mRNAs using in situ  PCR and α^2aprotein using a specific polyclonal antibody. In another interesting study, Graham et al.  47demonstrated that the α²AR subtype involved in rat pain is species dependent, with α^2b>> α^2aARs involved in mediating hot-plate responses in Harlan rats but neither subtype involved in Sasco rats; in contrast, α^2aAR and α^2bAR both are important in nociceptive responses in the tail-flick test. In terms of human DRG, data from the current study shows that α^2band α^2cAR subtype mRNAs predominate (α^2b≈α^2c) in human DRG at all spinal cord levels. Compared with rat DRG, these findings suggest significant species and subtype heterogeneity in DRG α²AR subtype expression. Development of highly sensitive and specific α²AR subtype antibodies and ligands are necessary to further clarify the relation between α²AR mRNA and protein in human spinal cord dorsal horn and DRG; such information may contribute to understanding nociception at the DRG–spinal cord level. Furthermore, such studies also may contribute to our understanding of α²ARs in modulating other neural pathways.

We demonstrate the presence and heterogeneity of α²AR subtype mRNA in human DRG (α^2b≈α^2c). In addition, α²AR subtype mRNA in DRG contrasts with that found in human and rat spinal cord dorsal horn and rat DRG. If confirmed at a protein level, these findings provide an additional step in unraveling mechanisms involved in complex neural pathways such as pain.