Despite considerable interest in the neural mechanisms that regulate muscle blood flow, the descending pathways that control sympathetic outflow to skeletal muscles are not adequately understood. The present study mapped these pathways through the transneuronal transport of two recombinant strains of pseudorabies virus (PRV) injected into the gastrocnemius muscles in the left and right hindlimbs of rats: PRV-152 and PRV-BaBlu. To prevent PRV from being transmitted to the brain stem via motor circuitry, a spinal transection was performed just below the L2 level. Infected neurons were observed bilaterally in all of the areas of the brain that have previously been shown to contribute to regulating sympathetic outflow: the medullary raphe nuclei, rostral ventrolateral medulla (RVLM), rostral ventromedial medulla, A5 adrenergic cell group region, locus coeruleus, nucleus subcoeruleus, and the paraventricular nucleus of the hypothalamus. The RVLM, the brain stem region typically considered to play the largest role in regulating muscle blood flow, contained neurons infected following the shortest postinoculation survival times. Approximately half of the infected RVLM neurons were immunopositive for tyrosine hydroxylase, indicating that they were catecholaminergic. Many (47%) of the RVLM neurons were dually infected by the recombinants of PRV injected into the left and right hindlimb, suggesting that the central nervous system has a limited capacity to independently regulate blood flow to left and right hindlimb muscles.
- blood flow patterning
- pseudorabies virus
- exercise hyperemia
blood flow to skeletal muscle increases several fold during exercise due to a variety of factors, including alterations in the activity of sympathetic efferents innervating the arterioles in the contracting muscles (42). The neural mechanisms that contribute to regulating muscle blood flow are activated as part of the programming for movement (“central command”) (39), as well as by feedback signals from contracting muscle (18). Despite considerable interest in the neural mechanisms that participate in regulating muscle blood flow, the descending pathways that control sympathetic outflow to skeletal muscles are not adequately understood. One method that has provided considerable insights into the neural regulation of autonomic functions is the retrograde transneuronal transport of pseudorabies virus (PRV) injected into peripheral tissues. A number of studies have verified that this technique provides a highly specific method of tracing polysynaptic pathways (5, 7), and it has been used to map the sympathetic and parasympathetic pathways innervating a variety of targets, including the adrenal gland, kidney, spleen, pancreas, tail artery, and interscapular adipose tissue (2–4, 17, 21, 33–37). Although previous studies have involved the injection of PRV into hindlimb skeletal muscles (19, 21, 31), these experiments provided incomplete information regarding the descending pathways regulating blood flow to the muscles. In two of these experiments, the hindlimb muscle was sympathectomized, such that only the motor pathways became infected (19, 21). In the other, both the motor and sympathetic efferents innervating the muscle were intact and transported PRV transneuronally, but the focus was on infected spinal cells and not labeling in the brain stem (31).
Multiple recombinants of PRV are available and can be injected into the same animal to explore divergence and convergence within neural pathways [e.g., (1, 21)]. Such recombinants express unique reporters that can be distinguished independently. For example, PRV-BaBlu contains the lacZ gene at the gG locus and produces β-gal under the control of the viral gG promoter (1). Another recombinant, PRV-152, expresses enhanced green fluorescent protein (EGFP). This virus carries an insertion at the gG locus, such that EGFP is constitutively expressed using the cytomegalovirus immediate early promoter (1). These two recombinants have been utilized to determine whether common brain stem neurons contribute to regulating the activity of multiple respiratory muscles (1) and whether the same neurons regulate sympathetic outflow to the left and right kidneys (2).
PRV recombinants could also be employed to determine whether the nervous system has the capacity to independently regulate blood flow to different skeletal muscles. There is controversy in the literature as to whether the central nervous system is capable of evoking anatomically patterned changes in blood flow. Some studies have concluded that patterning of blood flow mediated by the sympathetic nervous system is in accordance with tissue type but not the location of the tissue within the body (26, 27). This conclusion is partly based on the observation that when sodium glutamate microinjections were used to activate neuronal cell bodies in the principal vasomotor region of the brain stem, the rostral ventrolateral medulla (RVLM) of anesthetized cats, no separation could be found between sites that caused vasoconstriction in forelimb and hindlimb muscles (26), despite the fact that injection sites that specifically altered blood flow to particular tissues were readily identifiable (27). However, other studies have indicated that particular sensory stimuli, such as those detected by the vestibular system, can evoke discrete changes in the activity of sympathetic efferents innervating blood vessels in different limb muscles and trigger anatomically patterned changes in blood flow (20, 22, 41). These latter observations suggest that distinct populations of brain stem neurons regulate the sympathetic outflow to each skeletal muscle.
The major goal of the present study was to use PRV to trace the neural pathways regulating blood flow to a hindlimb muscle gastrocnemius in rats. A secondary goal was to ascertain whether separate populations of brain stem neurons regulate blood flow to the gastrocnemius muscles in the left and right hindlimbs. For this purpose, PRV-152 was injected into the left gastrocnemius muscle and PRV-BaBlu was injected into the right gastrocnemius muscle of each animal. To prevent PRV from being transmitted to the brain stem via motor circuitry, a spinal transection was performed just below the L2 level, which is caudal to the majority of sympathetic preganglionic neurons innervating the muscle, but rostral to the gastrocnemius motoneurons (21, 31). Immunohistochemical detection of the enzyme tyrosine hydroxylase (TH) was also incorporated into the experiments, so that we could ascertain whether brain stem neurons infected by PRV injections into hindlimb muscle were catecholaminergic. We tested the hypothesis that discrete descending pathways regulate sympathetic outflow to the gastrocnemius muscles in the left and right hindlimbs.
All experimental procedures used in this study conformed to the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the University of Pittsburgh's Institutional Animal Care and Use Committee. Data were collected from 29 adult male Sprague-Dawley rats, weighing between 156 and 410 g, obtained from Hilltop Lab Animals (Scottdale, PA). Throughout the course of the study, animals were maintained under Biosafety Level 2 conditions, as defined by U.S. Department of Health and Human Services publication no. CDC 88-8395.
The viral recombinants used for transneuronal tracing in these studies, PRV-152 and PRV-BaBlu, were the generous gift of Dr. Lynn Enquist (Princeton University, Princeton, NJ). Both recombinants of PRV were grown in pig kidney (PK15) cells, adjusted to a final concentration of 1 × 108 plaque-forming units/ml, cleared of cellular debris by centrifugation, aliquoted at 50 μl per tube, and stored at −80°C. Individual aliquots of virus were thawed immediately before the injection. Excess virus was inactivated with bleach and discarded.
Surgical procedures, postsurgical care, and euthanasia.
Animals were acclimated to the animal housing facility for at least 3 days, and typically for >1 wk, before surgery. Anesthesia was induced using 3–4% isoflurane vaporized in O2, and maintained with a concentration of 1.5–2%. A surgical plane of anesthesia was obtained such that spontaneous movement and withdrawal reflexes to foot pinch were absent. All surgical procedures were performed using aseptic techniques. The skin overlying the dorsal process of the 13th thoracic vertebra was incised, the fascia and back muscles were dissected to expose the vertebra, and the dorsal aspect of the vertebra was removed using an electrical drill to expose the upper lumbar spinal cord. Subsequently, the spinal cord was transected just below the L2 level using an electrocautery. The skin overlying the gastrocnemius muscle on both sides was then incised, and the muscle was bluntly dissected and separated from adjacent musculature and connective tissue. Multiple 3- to 4-μl injections of PRV were made along the length of the gastrocnemius muscles using a 10-μl Hamilton syringe with a 26-gauge needle. PRV-152 was injected on the left side, whereas PRV-BaBlu was used on the right side. Typically, the total volume of virus injected into each muscle was 50 μl, although 30 μl of virus were used in six animals. After surgical procedures were completed, back muscles, fascia, and the skin were closed using nylon sutures.
Following the surgery, animals were maintained in a special cage with a vertical height of 6″, such that the animals had access to food and water despite the paralysis of the lower body. Ketoprofen (3 mg/kg im) was administered just before the surgery and every 12 h subsequently for a period of 72 h to provide analgesia. No signs of postsurgical pain or distress were evident. The bladder was expressed every 6 h after the spinal transection. After a survival time of 50–123 h, the animals were deeply anesthetized using pentobarbital sodium (50 mg/kg ip) and perfused transcardially with 300–400 ml of 0.15 M NaCl followed by 400–500 ml of 4% paraformaldehyde-lysine-periodate (PLP) fixative (28). Postmortem observations were conducted to determine whether the spinal cord transection was complete and performed at the correct level. Subsequently, the brain and the entire spinal cord were removed, postfixed in 4°C PLP for 1 or 2 days, and cryoprotected using 25% sucrose solution.
The entire brain was sectioned at 40-μm thickness in the coronal plane, whereas segments of the spinal cord were cut horizontally into 40-μm-thick sections. Brain stem tissue sections were collected in series with a spacing of 240 μm, whereas spinal cord sections were distributed in sets with a spacing of 160 μm. Tissue sections were stored at −20°C in cryoprotectant (40) until they were immunoprocessed.
Both immunoperoxidase and immunofluorescence methods were used to visualize neurons infected with PRV-152 and/or PRV-BaBlu; these methods are described in detail elsewhere (1, 8). One bin of brain and spinal cord sections was incubated for 2 days at 4°C in a primary antibody solution containing rabbit anti-EGFP (1:200; Molecular Probes, Eugene, OR), whereas a second bin was incubated in mouse anti-βgal (1:10,000; Sigma-Aldrich, St. Louis, MO) to localize PRV-152 and PRV-BaBlu, respectively. Subsequently, the tissue was incubated for 2 h at room temperature in a solution containing either donkey anti-rabbit or donkey anti-mouse IgG (1:200; Jackson ImmunoResearch Laboratories, West Grove, PA). Thereafter, the tissue was processed using the avidin-biotin modification of the peroxidase antiperoxidase procedure (15). To visualize both recombinants in a single section, the same primary antibodies used for the immunoperoxidase analysis were used, but were combined in the same well. After a 2-day incubation at 4°C in the primary antibody solution, the sections were incubated for 2 h in the dark in a solution containing goat anti-rabbit secondary antibody conjugated to Bodipy-FL (1:300; Invitrogen, Molecular Probes, Eugene, OR) and donkey anti-mouse secondary antibody conjugated to the CY3 carbocyanine (1:500; Jackson ImmunoResearch Laboratories).
A fourth bin of brain sections was processed for the dual localization of PRV-152 and TH. The processing of this tissue was similar to that described above, with the exception that mouse anti-TH (1:200; Chemicon International, Temecula, CA) was substituted for mouse anti-βgal as a primary antibody. After the completion of immunohistochemical processing, the sections were mounted on gelatin-coated slides, dehydrated, cleared, and coverslipped with the use of Cytoseal 60 (VWR Scientific, West Chester, PA). Sections subjected to immunoperoxidase processing were counterstained using Neutral Red before coverslipping.
Immunoreacted tissue sections were examined and photographed using a Zeiss Axioplan photomicroscope equipped with epifluorescence and filters that selectively excited Bodipy-FL or CY3 and with a filter that allowed for the excitation of both fluorophors. The red fluorescence of CY3 was used to identify cells infected after injection of PRV-BaBlu into the right gastrocnemius muscle or which contained TH, whereas the green fluorescence of Bodipy-FL was used to identify neurons infected by PRV-152 injected into the left gastrocnemius muscle. Neurons containing both fluorophors appeared yellow. Neurons were classified as being dual-labeled only if the cell body were evident when each fluorophor was excited independently, and the neuron had an identical location and shape under all exposures. Within each brain region, the number of cells containing only one PRV recombinant and the total number of neurons containing both recombinants were counted. The percentage of double-labeled cells in an area was calculated by dividing the number of neurons labeled for the presence of both recombinants by the total number of infected cells observed. The percentage of PRV-infected neurons in a region that was immunopositive for TH was determined in a similar manner. Digital photographs of neurons were obtained by using a camera (Hamamatsu Photonics, Hamamatsu, Japan) and a Simple-32 PCI image analysis system (Compix, Lake Oswego, OR). Digital images were prepared for publication with the use of Adobe Systems (San Jose, CA) Photoshop software. Individual images were adjusted for size, brightness, and contrast, but color balance was not altered. The locations of labeled neurons were also mapped from sections processed using immunoperoxidase techniques by using a Nikon Optiphot photomicroscope. The regions in which infected cells were located were defined with reference to the atlases of Swanson (38) and Paxinos and Watson (30).
Out of the 29 animals used in this experiment, 16 yielded data that were useful in understanding the organization of descending pathways that regulate sympathetic outflow to skeletal muscle. Data were not considered from five animals because postmortem observations revealed that the spinal transection was incomplete or made at the wrong level. In 13 other animals with survival times of 50 (n = 2), 71–73 (n = 5), or 93 (n = 1) h, no labeled neurons were detected in the brain stem, possibly because the survival times were too short to allow for the transneuronal passage of virus and infection of cells. However, in the remaining 16 animals with survival times of 74–123 h (see Table 1), infected neurons were present in both the brain and spinal cord. Prior studies have similarly shown that at least 72 h is typically required to label brain stem neurons following the infection of the sympathetic nervous system though injections of PRV into peripheral targets (2–4, 8, 19, 21). It is noteworthy that the animals lacking brain stem labeling following survival times of 71–93 h were large (weights of 326–390 g), whereas the animals that exhibited infected brain stem neurons after similar survival times were much smaller (weights of 237–262 g). It was previously reported that the survival times following peripheral injections required for PRV to infect central nervous system neurons and pass transneuronally are typically longer in rats with body weights >300 g than in smaller animals (21).
At all survival times, infected neurons were observed for overt evidence of cellular damage or lysis, which could result in the release of PRV into the extracellular space and the nonspecific spread of virus. No such cellular damage was detected, except in the spinal cord at the longest survival times (≥120 h). Thus, it is unlikely that brain stem neurons were labeled through the indiscriminant release of PRV into the extracellular space and the infection of neurons other than those making synaptic contacts with sympathetic preganglionic neurons.
Spinal cord labeling produced by PRV injections into gastrocnemius.
PRV-infected neurons were observed both anterior and posterior to the spinal transection in all animals with labeled cells in the brain. Cells located within the ipsilateral intermediolateral cell column of the lower thoracic cord were heavily infected with PRV, as illustrated in Fig. 1A. In addition, a few scattered cells were labeled near the central canal of the lower thoracic spinal segments. Furthermore, large infected cells that were presumed to be motoneurons were present within the ipsilateral ventral horn of the L4-S1 segments, as shown in Fig. 1B. In animals surviving >95 h following PRV injections, small labeled cells were also evident in the thoracic and lumbar spinal cord and were assumed to be interneurons infected through the transneuronal passage of virus. Infected cells were additionally observed bilaterally in the cervical spinal cord of the animals with the longest survival times. Because this study focused on descending pathways from the brain that regulate sympathetic outflow to gastrocnemius, the distribution of the infected spinal neurons was not quantified.
Infection of brain neurons following PRV injections into gastrocnemius.
The distribution of neurons in the brain that were infected by injections of PRV into gastrocnemius was mapped from tissue sections processed using immunoperoxidase techniques. Figure 2 is a mapping of the locations of brain neurons infected by PRV-152 injected into the left gastrocnemius muscle of an animal that was euthanized 105 h after the inoculations. Note that the labeling was bilateral, with approximately equal numbers of infected cells on the ipsilateral and contralateral sides. As illustrated in Fig. 2, labeled neurons were concentrated in several distinct brain stem nuclei, including the medullary raphe nuclei [raphe pallidus (RP) and raphe obscurus (RO)], locus coeruleus (LC), nucleus subcoeruleus (SC), and the paraventricular nucleus of the hypothalamus (PVN). Furthermore, many infected cells were observed in the following regions of the reticular formation that are known to participate in regulating sympathetic outflow (2–4, 17, 21, 33–37): the RVLM (defined as the triangular region at the ventral surface of the brain stem ventral to nucleus ambiguous and lateral to the inferior olive), the rostral ventromedial medulla (RVMM, defined as the portions of the lateral paragigantocellular reticular nucleus and ventral gigantocellular nucleus immediately dorsal to the inferior olive and pyramid), and the A5 adrenergic cell group (A5) region (the narrow area bordered by the ventrolateral edge of the brain stem, the exiting seventh nerve and the superior olive). Additional infected cells were scattered throughout the pontomedullary reticular formation, mainly in the gigantocellular reticular nucleus. The left column of Fig. 3 provides examples of the regions containing a high concentration of cells that were labeled for the presence of the PRV-152; higher-power images of neurons infected by PRV-152 that were visualized using immunofluorescence techniques are shown in column 2 of the figure. The regions shown in Fig. 3 are as follows: RP (∼14 mm posterior to bregma); RVLM (∼13 mm posterior to bregma); RVMM; ∼13 mm posterior to bregma); A5 (∼10 mm posterior to bregma); LC (∼10 mm posterior to bregma); SC (∼9 mm posterior to bregma); and PVN (∼2 mm posterior to bregma).
Table 1 indicates the relative density of labeling in the regions of the medulla and pons containing the largest number of cells infected by PRV-152 injected into the left gastrocnemius muscle. Data are provided for all 16 animals with infected brain stem neurons; only labeling on the ipsilateral side is depicted. The animals were separated into three groups on the basis of the amount and extent of labeling in the brain. The “early-labeling” group exhibited only a very limited number of infected cells that were confined to a few circumscribed brain stem regions. The survival times for three of the four animals in this group were ≤76 h after PRV inoculations. An additional animal with limited brain stem labeling survived for 95 h after virus injections; this rat was very large (331 g), which might have delayed the infection of brain stem neurons (21). The “intermediate-labeling” group consisted of five animals with postinoculation survival times of 96–105 h and was characterized by an extensive infection of neurons in a restricted number of brain regions. The “late-labeling” group comprised seven animals with survival times ≥120 h following virus injections. This group showed labeling in a greater number of regions and more extensive numbers of infected cells in these regions.
Infected cells were present in the RVLM of all animals belonging to the early-labeling group, although more labeled cells were present in the A5 region than in the RVLM in half of the animals. Two of the four rats in this group also exhibited a few infected cells in the raphe nuclei, RVMM, LC, and SC. The number of infected neurons in these regions was much larger in animals belonging to the intermediate labeling group, and in some cases, labeled cells were also evident in nucleus solitarius, the lateral vestibular nucleus, and area postrema. The only caveat is that in two of the five rats in this group (cases 5 and 6), labeling in LC and SC was clearly less extensive than in RVLM, RVMM, and A5. Additional scattered neurons were present within the pontomedullary reticular formation of these animals, mainly in the gigantocellular reticular nucleus, although this labeling is not described in Table 1 because the locations of the cells varied from case to case. This group of rats also exhibited infected cells in the pons and diencephalon (Table 2 indicates the cell locations in the midbrain). The labeling was most extensive in the PVN, particularly, the dorsal zone of the medial parvicellular part of the nucleus. In addition, a few infected neurons were observed in the periaqueductal gray, lateral hypothalamus, red nucleus, and Edinger-Westphal nucleus. The distribution of infected cells in the late-labeling group was similar to that in the intermediate labeling group, with the exception that the number of neurons was much higher in nucleus solitarius, area postrema, periaqueductal gray, PVN, lateral hypothalamus, and red nucleus. The longest-surviving animals additionally exhibited labeled cells in the preoptic nucleus of the hypothalamus.
Dual infection of brain neurons following PRV injections into left and right gastrocnemius.
Sections processed using immunofluorescence procedures were observed to determine whether neurons in the brain were dually infected by the viruses injected into the left and right gastrocnemius muscles. Fig. 3 illustrates the method employed in this analysis: column 2 shows neurons labeled for PRV-152, column 3 shows cells infected by PRV-BaBlu, whereas column 4 contains dual-exposure micrographs that reveal both recombinants. Cells classified as being dual-infected by PRV-152 and PRV-BaBlu were evident in both of the two single-exposure micrographs (e.g., columns 2 and 3 of Fig. 3) and appeared yellow in double-exposure micrographs. Dual-infected neurons were observed in the RP, RO, RVLM, RVMM, A5, LC, SC, and PVN. Fig. 4 shows the numbers of neurons in each of these brain regions that were infected with only PRV-152, only PRV-BaBlu, or with both recombinants. Data are provided for all of the animals in the intermediate labeling group. Labeling on the left side (ipsilateral to PRV-152 injections) and right side (ipsilateral to the PRV-BaBlu injections) is designated separately. It is evident from the graphs in Fig. 4 that the distribution of neurons infected by PRV-152 and PRV-BaBlu was similar on the left and right sides of the brain and that many cells were dual-infected by both recombinants. Cell counts from the left and right sides of the brain were pooled to determine the percentage of double-labeled cells in each region, which was calculated by dividing the number of cells labeled for the presence of both recombinants by the total number of infected cells observed in the area. Neurons infected by both recombinants were least prevalent in PVN [32 ± 9 % (SE)], RVMM (35 ± 6%), LC (36 ± 13%), A5 (37 ± 4%), and SC (40 ± 2%) and most common in the RVLM (47 ± 6%), RP (49 ± 8%), and RO (52 ± 7%). The fraction of brain neurons that was infected by both recombinants was unrelated to the survival time following PRV inoculations. For example, in the group of animals with the shortest survival times, 43 ± 7% of labeled RVLM neurons were infected by both recombinants, as opposed to 47 ± 6% in the intermediate labeling group and 44 ± 5% in the animals that survived ≥ 120 h after injections.
Infection of catecholaminergic neurons following injections of PRV into gastrocnemius.
Because a subset of neurons that participate in regulating sympathetic outflow is catecholaminergic (25, 37), we ascertained whether neurons infected after the injection of PRV-152 into the left gastrocnemius muscle expressed TH. The analysis incorporated only the five animals in the intermediate-labeling group and was limited to the six brain areas containing an appreciable number of PRV-infected cells that are known to include noradrenergic and adrenergic cells: RVLM, RVMM, A5, LC, SC, and PVN (9). Figure 5 illustrates the dual-labeling immunofluorescence procedure used to determine whether PRV-infected cells synthesized catecholamines by showing labeling in the following regions: RVLM (∼13 mm posterior to bregma); RVMM (∼13 mm posterior to bregma); A5 (∼10 mm posterior to bregma); LC (∼10 mm posterior to bregma); SC (∼9 mm posterior to bregma). Column 1 of Fig. 5 contains low-power micrographs indicating the presence of PRV-152; higher-power images of neurons infected by PRV-152 that were visualized using immunofluorescence techniques are shown in column 2 of the figure. Column 3 of Fig. 5 illustrates neurons that were labeled for the presence of TH, whereas column 4 contains dual-labeling micrographs that show the presence of both TH and PRV-152. TH-expressing cells were most heavily concentrated in LC and SC (see Fig. 5, D and E) and were distributed sparsely in RVMM and PVN. The fraction of neurons in each region that were infected by PRV-152 and dually-labeled for the presence of TH was calculated by dividing that number of dually-labeled cells by the total number of virus-infected neurons. Cell counts from the left and right sides were pooled for this analysis. In three of the areas, approximately half of the PRV-infected cells were colabeled for the presence of TH: RVLM (52 ± 2%), SC (52 ± 11%), and A5 (51 ± 6%). The percentage of infected cells that also expressed TH was lower in LC (33 ± 11%) and the RVMM (27 ± 4%). No dual-labeled cells were observed in PVN. Fig. 5D additionally reveals that PRV-immunopositive cells were not distributed uniformly in LC but were mainly confined to the ventral portion of the nucleus.
The major finding of this study is that several regions of the brain, particularly the RVLM, RVMM, medullary raphe nuclei, A5 region, LC, SC, and PVN, contain a substantial number of neurons that are infected relatively early following injections of PRV into the gastrocnemius muscle. Because the spinal cord of each animal was transected rostral to the gastrocnemius motoneuron pool, presumably, the brain neurons were infected with PRV via the sympathetic nervous system. There is no evidence that the parasympathetic nervous system provides any innervation to limb muscles (12), and brain stem regions containing parasympathetic preganglionic neurons (nucleus ambiguus and the dorsal motor nucleus of the vagus) were devoid of infected cells in these experiments. Thus, a number of areas in the medulla, pons, and diencephalon appear to play a substantial role in regulating sympathetic outflow to skeletal muscle. With the exception of the raphe nuclei and PVN, 25–50% of the neurons in these regions that were labeled for the presence of PRV were also immunopositive for TH, showing that descending catecholaminergic and noncatecholaminergic pathways are involved in regulating the firing of sympathetic efferents innervating gastrocnemius. In addition, many of the brain neurons were dually infected by recombinants of PRV injected into the left and right hindlimb; the dual infection of cells was particularly common in the RVLM and medullary raphe nuclei. This finding suggests that numerous descending projections regulate sympathetic outflow to both the left and right hindlimbs.
Previous studies have shown that the use of PRV provides a highly specific method of tracing transneuronal pathways, as there is little lysis of infected neurons, release of virus into the extracellular space, or infection of neurons other than those that are part of the circuit of interest (5, 7). We did not observe evidence of cellular damage except in the spinal cord following the longest survival times, after brain stem neurons had become infected. Thus, it seems likely that the labeled cells observed in this study participated in regulating sympathetic outflow to skeletal muscle.
Nonetheless, a number of caveats must be considered when interpreting the results of this study. It was presumed that the majority of neurons infected with PRV participated in regulating the discharges of sympathetic efferents innervating vascular smooth muscle within gastrocnemius. However, hindlimb muscle spindles also receive limited sympathetic nervous system innervation (11, 14, 16), and thus we cannot exclude the possibility that some central nervous system neurons were infected though these projections. Even so, considering the relatively large number of muscle vasoconstrictor fibers that are present within peripheral nerves (10, 20, 26, 27, 32), it seems reasonable to conclude that most of the circuitry that was labeled in these experiments participated in controlling muscle blood flow. Caution must also be used when interpreting the results of the experiment involving the injection of PRV recombinants into the left and right gastrocnemius, because the infection of a neuron by one virus can limit its susceptibility to be infected by a second virus (23). Furthermore, it is unlikely that all of the sympathetic efferents innervating each gastrocnemius muscle were infected. Thus, we probably underestimated the fraction of brain neurons that influence sympathetic outflow to the two sides. These findings suggest that the central nervous system has only a limited capacity to independently regulate vascular resistance in the left and right gastrocnemius.
Previous studies that infected sympathetic efferents innervating other targets, including the adrenal gland, kidney, spleen, pancreas, tail artery, and interscapular adipose tissue (2–4, 17, 21, 33–37), have also reported the presence of labeled neurons in the same regions containing substantial infection following the injection of PRV into gastrocnemius. This observation raises the question of how the nervous system accomplishes independent control of the activity of sympathetic efferents innervating particular tissues and body regions. Infected neurons were always present in the RVLM in animals classified as having ”early labeling“ following the injection of PRV into gastrocnemius. Most physiological studies have implicated the RVLM as playing the predominant role in the regulation of peripheral vasoconstriction (26, 27). Thus, the large number of RVLM neurons infected at relatively short survival times following the injection of PRV into gastrocnemius is in keeping with this notion. Neurons in the RVMM, A5, raphe, LC, SC, and PVN have not been typically regarded as playing a major role in regulating blood flow to muscle, but the present study suggests that they have such a function. It is currently unclear whether cells in these areas simply adjust the excitability of sympathetic preganglionic neurons or have more specific influences on the control of blood flow. Sved et al. (37) and Morrison (29) have hypothesized that the activity of particular sympathetic preganglionic neurons is dependent on the combination of descending inputs that the cells receive from different areas. This hypothesis possibly explains how neurons in a brain stem area might participate in regulating muscle blood flow without having firing patterns or responses to stimulation that overtly indicate that they play this role.
Sodium glutamate microinjections placed in the RVLM of anesthetized cats can selectively alter blood flow to particular tissues (27), but there is no separation between injection sites that produce changes in forelimb and hindlimb blood flow (26). The observations in these studies led to conjecture that patterning of blood flow mediated by the sympathetic nervous system is in accordance with tissue type but not the location of the tissue within the body. The findings of the present study bolster this conclusion, as a large fraction of neurons in the RVLM and other brain stem regions was dually infected by PRV recombinants injected into left and right gastrocnemius. Nonetheless, definitive support for this notion will require transneuronal tracing of the neural circuitry regulating forelimb and hindlimb blood flow in the cat, the species employed in previous physiological experiments. The results of prior physiological experiments (26) could be explained by the presence of distinct but intermingled neurons in the RVLM that have selective influences on sympathetic outflow to the forelimb and hindlimb (such that both groups of cells are excited by microinjections of glutamate). Unfortunately, PRV is not an effective transneuronal tracer in felines (6), and thus this experiment awaits the development of multiple recombinants of other retrograde viral tracing agents such as rabies virus.
At long survival times following the injection of PRV into gastrocnemius, an appreciable number of infected neurons were observed in the lateral vestibular nucleus and red nucleus. Labeling in these nuclei has not been reported in most studies involving the infection of sympathetic efferents innervating other targets (2–4, 17, 21, 33–37), although Kerman et al. (19) indicated that some red nucleus neurons were infected following the injection of PRV into the adrenal gland. It is possible that the lateral vestibular nucleus and red nucleus are mainly involved in adjusting sympathetic outflow to muscle, but not other tissues. Vestibular signals participate in regulating hindlimb blood flow (20, 22, 41), and thus the infected neurons in the lateral vestibular nucleus likely are components of the neural pathway that mediates this response. The red nucleus is not typically considered to participate in autonomic regulation but instead plays a fundamental role in motor control in the rat (13, 24). Red nucleus neurons could potentially contribute to altering limb blood flow in parallel with movement and thus may comprise elements in the neural circuit underlying central command (39). However, further studies will be required to test this hypothesis.
In summary, the present data show that the RVLM, RVMM, medullary raphe nuclei, A5 region, LC, SC, and PVN all contribute to regulating the activity of sympathetic efferents innervating gastrocnemius. Thus, more regions appear to participate in regulating muscle blood flow than have been appreciated on the basis of physiological studies, which indicated that the RVLM is the principal vasomotor region of the brain stem. A large fraction of neurons in the RVLM and other regions were dually infected by PRV recombinants injected into the left and right gastrocnemius. Because of technical limitations, the fraction of double-infected neurons observed likely is an underestimate of the total number of brain stem neurons that influence blood flow to the two hindlimbs. These findings suggest that the nervous system has a limited capacity to independently regulate blood flow to the left and right hindlimbs and indicate the existence of limits with regard to the complexity of patterning of sympathetic outflow to the vasculature. As such, it seems likely that nonneural mechanisms such as autoregulation play a critical role in fine-tuning blood flow to particular contracting muscles during the execution of movement.
This work was supported by National Institute on Deafness and Other Communication Disorders Grant R01-DC00693, as well as by National Center for Research Resources Grant P40-RR018604.
The authors thank Jen-Shew Yen and Lucy Cotter for valuable technical assistance in the completion of these studies. We are also grateful to Dr. Lynn Enquist for the generous contribution of PRV recombinants.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2007 the American Physiological Society