Spinal subarachnoid opioid administration in rats has been widely used to study the pharmacological properties of opioids. This method, as first described by Yaksh and Rudy (1), involved the introduction of a small polyethylene catheter through the atlanto-occipital membrane that was advanced caudally until the tip reached the lumbar enlargement. While this has proven to be a valuable approach for studying opioid analgesic effects and tolerance, over time, some disadvantages have become apparent. Atlanto-occipital catheterization is associated with significant postsurgical mortality (3%–10%) (2) and neurological morbidity (10%–20%) (3,4). The procedure may also affect the overall functional state of the animal. Rats lose significant amounts (approximately 7.5%) of their body weight during the first week after catheterization (5). Because of these issues, it has been suggested that a recovery period of several days after surgery is required before the animals can be used in behavioral testing (6).
To decrease the incidence of complications, other techniques such as thoracic (7) or lumbar (2,8) catheterization have been proposed. Introduction of the catheter at the thoracic or lumbar level has been postulated to reduce morbidity. The reasons for this are thought to be a decreased amount of spinal cord compression by the catheter and, potentially, a higher likelihood of correct catheter positioning. A simplified method for direct catheterization of the lumbar subarachnoid space has been developed, using a catheter-through-needle technique (2,8). In these studies, no animals demonstrated neurological disturbances, and all animals gained weight normally during the first week after implantation (2,8). While these methods have significantly reduced morbidity from approximately 10% to <6% (2,8,9), implantation of catheters in the subarachnoid space has been shown to produce an inflammatory response and fibrosis (10,11). It has been suggested that this inflammatory response could alter nociceptive transmission (5,12) and opioid efficacy (13,14). These findings suggest that it is also possible that intrathecal catheter placement could alter the development of opioid tolerance.
To avoid the potential effects of inflammation or surgically implanting a foreign body in the spinal cord on opioid analgesia or tolerance development, we wished to develop a method for repeated spinal drug administration in rats that did not involve catheter placement. Single-dose administration of opioids via lumbar puncture (LP) in mice (15) and rats (13,16–18) has been reported. We are not aware of any previous reports of the use of repeated intermittent LP for chronic opioid administration in rats. We report the development of a safe, reliable method for repeated LP in rats. This technique proved practical for use in studying the development of experimental opioid tolerance without causing morbidity or significant side effects.
Sprague-Dawley rats (male, 250–300 mg) were housed 3 to a cage with water and food ad libitum and kept in temperature-controlled rooms on a 12:12-h light-dark cycle, with the dark cycle beginning at 7:00 pm. Animals were habituated to the testing environment for 1 wk before testing, and all tests were performed in the morning. All protocols were approved by our Institutional Animal Care and Use Committee.
Rats were anesthetized with 2% isoflurane in oxygen via nose cone. The lumbar region was shaved, prepared with Betadine solution, and the intervertebral spaces widened by placing the animal on a plexiglas tube. Animals were then injected at the L5-6 interspace using a 0.5-inch 30-gauge needle (Becton-Dickinson, Franklin Lakes, NJ) connected to a Hamilton syringe filled with morphine (morphine sulfate; Mallinckrodt Inc., St. Louis, MO) dissolved in 5, 20, or 40 μL of artificial cerebrospinal fluid (CSF [aCSF = 126 mM of NaCl, 2.5 mM of KCl, 1.2 mM of NaH2PO4, 1.2 mM of MgCl2, 2.4 mM of CaCl2, 11 mM of glucose, and 25 mM of NaHCO3, saturated with 5% CO2 in 95% O2, and adjusted to a pH value of 7.3–7.4). Correct subarachnoid positioning of the tip of the needle was verified by a tail- or paw-flick test. Animals then recovered in their home cage before analgesic testing.
Fifteen, 30, and 45 min after LP injection, tail-flick latency (TFL) was measured by placing the animals in Plexiglas cages (9×22×25 cm) on a modified Hargraves Device (19). This device consisted of a glass surface with a constant surface temperature of 30°C. A stimulus lamp was focused on the tail, with three measurements taken on different portions of the tail. An abrupt flick of the tail secondary to the thermal stimulus was sensed by photodiodes, and this served to terminate the stimulus and stop the timer. Animals were habituated to the device for 1 wk before testing and for 30 min before each test session. The intensity of the thermal stimulus used in the TFL test was adjusted so baseline TFL was between 3.0 and 3.5 s, and 10 s was used as an automatic cutoff time to avoid tail damage. The TFLs at the time points of 15, 30, and 45 min postinjection were measured and compared (n = 6–8 per group).
To investigate spinal drug distribution after acute LP, rats were injected with 5, 20, or 40 μL of methylene blue via LP, killed while anesthetized, and had their spinal cords dissected (n = 3–4 per group). To determine the effects of chronic LP on spinal drug distribution, the animals were injected with morphine via LP for 5 days, and on the fifth day, they were injected with 5, 20, or 40 μL of methylene blue. Animals were then euthanized, and the levels of dye distribution were checked and compared.
To determine whether repeated LP could alter the analgesic effect of morphine, rats were injected with aCSF via LP for 4 days and morphine on the fifth day. Control animals were anesthetized but not injected for 4 days and injected with morphine only on the fifth day (n = 3–4 per group).
To induce spinal morphine tolerance, animals were anesthetized with isoflurane and given 0.1 nmol/20 μL of morphine intraperitoneally (i.t.) once daily for 6 days. Other animals were injected with aCSF alone as a control (n = 6–8 per group). The drug injections were performed between 10:00 am and 12:00 pm every day, and TFL tests were performed both before and after the drug injection to determine baseline and drug-induced responses. Analgesic tolerance was indicated by a progressive decline in morphine-induced antinociception. The tail and hindpaws were also checked daily for evidence of sensory or motor impairment.
To determine whether the observed tolerance development was opioid-specific, rats were injected with either morphine or aCSF i.t. for 4 days. On Day 5, all animals received a single dose of clonidine 0.5 μg in 20 μL of aCSF i.t. To determine whether the observed decrease in drug effect could be simply caused by altered drug distribution after repeated LP, another group of animals received a single injection of clonidine 0.5 μg i.t. as a control (n = 3–4 per group). Baseline and postdrug TFL were checked every day.
To determine whether repeated LP induced a significant inflammatory response, rats (n = 3 per group) were injected with morphine (0.1 nmol/20 μL), aCSF, or underwent LP but no injection for 5 days. A group of uninjected animals served as controls. While anesthetized, the animals were euthanized and the lumbar spine removed en bloc. Tissues were postfixed in 10% neutral buffered formalin (Fisher, Pittsburgh, PA) for 48 h and then decalcified in a 10% formic acid solution for 96 h. Tissue sections were then embedded in paraffin, sliced at 5-μm thickness, and stained with hematoxylin-eosin. Sections were obtained at the site of injection, 1 mm on either side of the injection site, and at 2-mm intervals on either side of the injection site until a distance of 7 mm was reached. Slides were read by a neuropathologist blinded to the treatment.
TFL results were reported as mean ± sem. Between-group comparisons were made by analysis of variance and the Tukey post hoc test. Differences were considered to be significant at P < 0.05.
Initially, we wanted to determine whether injection volume affected the analgesic response. Animals were given 0.1 nmol of morphine in an injection volume of 5, 10, 20, or 40 μL, and TFL was determined 30 min later. The analgesic effect of 0.1 nmol of morphine was significantly greater using an injection volume of 20 or 40 μL (Fig. 1). Because the analgesia was more intense and the injection volume still remained less than the guideline of 10% of total CSF volume suggested by Rieselbach et al. (20), we used an injection volume of 20 μL for all subsequent studies.
The observed variability in analgesic response with different injection volumes could have been due to differences in drug distribution. To evaluate this possibility, we tested the spinal distribution of methylene blue after injections of 5, 20, and 40 μL of solution volume via LP. The rostral extent of dye spread varied markedly with injection volume (Fig. 2). Five-microliter injections showed complete circumferential spread to the L3-4 level, whereas 20 μL spread to T13-L1. Forty microliters of dye spread circumferentially to the T5 level, but plumes of dye could be seen extending up to the T1 level. All 3 injection volumes spread completely to the sacral and cauda equina regions (Fig. 2). It is also possible that repeated LP could have altered drug distribution, possibly accounting for a decrease in analgesic effect observed over time. To evaluate this possibility, we again took groups of 3–4 rats and determined the extent of completeness of spread of methylene blue after repeated LP. Rats were injected with 0.1 nmol of morphine once daily for 4 days via LP and injected with either 5, 20, or 40 μL of methylene blue on the fifth day. The rostral extent of dye spread was similar to that observed after single LP, and the circumferential spread of dye was also complete to these levels (data not shown).
To further validate this method, we wanted to determine whether aCSF injection alone had any effect on TFL. Therefore, we anesthetized animals and injected 20 μL of aCSF via LP as described above, then monitored TFLs 15, 30, and 45 min after injection (Fig. 3). In aCSF-injected animals, latencies at 15 and 30 min were increased to more than baseline (3.84 ± 0.42 s and 4.12 ± 0.95 s versus 3.12 ± 0.36 s; P < 0.001). No significant difference was found between the baseline and 45-min (3.12 ± 0.36 s versus 3.24 ± 0.32 s) time point. We also determined the time course of morphine analgesia after the injection of 0.1 nmol of morphine by LP (Fig. 3). TFLs at 15, 30, and 45 min were all significantly higher than baseline (P < 0.01). However, no statistically significant difference was observed among the 15-, 30-, and 45-min time points. Therefore, to avoid any potential injection-induced artifact effect on our results, we checked TFLs at 45 min after injection in all subsequent studies.
Once we established the optimal injection volume and timing of behavioral testing, we determined whether morphine could be injected repeatedly using this paradigm to study opioid tolerance. Preliminary dose-finding studies determined that a dose of 0.1 nmol would induce a TFL of 8–9 s 45 min after acute injection. Therefore, we tested TFL before and 45 min after the lumbar injection of 0.1 nmol/20 μL of morphine daily for 6 days (Fig. 4). We found that baseline TFL did not vary over the course of the experiment (Fig. 4A), suggesting the repeated LP did not damage the neural circuitry responsible for the tail-flick reflex. Also, no limping or dragging of the hindlimbs or trophic changes of the limbs were observed, suggesting that the nerves supplying these structures were not injured by repeated LP. We found that 0.1 nmol of morphine reliably induced tolerance over a 4-day period (Fig. 4B).
Whereas repeated LP did not seem to alter the spinal distribution of methylene blue dye, it is still possible that subtle changes in drug distribution or the induction of inflammatory processes could cause a nonspecific “apparent” tolerance. Therefore, we determined whether repeated LP altered the analgesic effect of morphine. One group of rats was injected with aCSF for 4 days and with morphine on the fifth day, whereas another group underwent anesthesia but no injection for 4 days and was injected with morphine on the fifth day only. We found that 0.1 nmol of morphine induced an equivalent amount of analgesia on the fifth day in both groups (Fig. 5), indicating that repeated LP did not seem to modulate the analgesic effect mediated by morphine. To further verify that the observed behavioral tolerance was opioid-specific, we examined whether rats that made opioid tolerant using this method would be cross-tolerant to the α-2 agonist clonidine. Two groups of rats were injected with either morphine or aCSF for 4 days and with a single dose of clonidine (0.5 μg in 20 μL of aCSF) on the fifth day. An additional group of rats underwent anesthesia but no injection for 4 days followed by an intrathecal injection of 0.5 μg of clonidine on the fifth day. We found that in both morphine-tolerant and aCSF-injected animals, clonidine induced an equivalent amount of analgesia (6.93 ± 0.17 s versus 7.10 ± 0.56 s; not significant; Fig. 6). Acute injection of clonidine without previous LP also induced an equivalent degree of analgesia (TFL = 7.07 ± 0.22 s; not significant). Taken together, these findings further support the idea that the tolerance induced by repeated LP is not caused by changes in drug distribution and is opioid-specific.
It has been well documented that intrathecal catheter placement can produce an inflammatory response as well as spinal cord compression (10,21). As mentioned above, it has been postulated that this inflammatory response could alter opioid efficacy (13,14). To determine whether repeated LP induced an inflammatory response, groups of rats were injected with either 0.1 nmol of morphine, aCSF, or underwent LP without any fluid injection for 5 days. Another group of uninjected naive animals served as controls. As shown in Figure 7, repeated LP did cause a slight inflammatory response that was tightly localized to the injection site. Some polymorphonuclear leukocytes were seen in the injection tract and epidural space at the injection site. The inflammation did not seem to vary in intensity with different injectates, suggesting that the observed response was caused by LP. In all groups, this inflammation was mild, and in contrast to catheter-induced inflammation, it was highly localized. In all treatment groups, all evidence of an inflammatory response was gone by a distance of 7 mm from the injection site in either the rostral or caudal direction (Fig. 7). It should also be noted that intermittent LP has caused no morbidity or mortality in the more than 100 animals that have undergone the procedure.
These studies present a novel method for studying opioid tolerance development in rats. Intermittent LP is consistent, reproducible, and markedly decreases the morbidity and mortality associated with intrathecal catheter placement in rats (2,8,9). Whereas this method is straightforward, a significant amount of practice is required to gain proficiency (>95% success on the first attempt). Between two weeks and one month of practice was required to improve initial success from approximately 50% to 95% with this technique. However, once proficiency was obtained, successful injection could be performed very quickly (usually taking less than a minute) and consistently under light inhaled anesthesia. It was easier to inject rats under anesthesia, although single-shot injections have been successfully performed on awake rats (17). We also found that it was more difficult to perform LP in awake rats than in mice, an observation that has been corroborated by others (6). Thus, we were concerned that any difficulties associated with awake injection could stress the animals and lead to the phenomenon of stress-induced analgesia (22), potentially confounding the interpretation of experimental results. Animals recovered quickly from light inhaled anesthesia, and we were able to obtain reliable behavioral tests as soon as 15 minutes after injection. However, we did notice a slight, but significant, increase in TFL 15 and 30 minutes after aCSF injection. This could have been caused by a disturbance in the local environment from the injection or residual analgesic effects of the anesthetic. TFL returned to baseline 45 min after aCSF injection, while morphine analgesia remained robust over this time period.
Previous investigators generally used an injection volume of 10 μL for single-shot LP in rats (6,17). However, it is not clear how this injection volume was decided upon. We are not aware of any previous reports examining the effect of injection volume on morphine analgesia in rats. We found that injection volumes of 20 and 40 μL induced more analgesia than 5- or 10-μL injections (Fig. 1), and this could have been caused by more extensive distribution of the drug when a larger volume was injected. Based on radiolabeled drug distribution studies in monkeys, Rieselbach et al. (20) suggested that injection volumes for lumbar intrathecal injections be limited to 10% of the total CSF volume to minimize spread of the drug to supraspinal centers. They found that appreciable amounts of drug injected intrathecally were not recovered in the basal cisterns up to an hour after injection if this volume limit was not exceeded (20). We also observed more rostral spread with increased injection volumes (Fig. 2). However, increasing the injection volume to 40 μL did not increase the analgesic response, so we also chose an injection volume (20 μL) that was <10% of the total CSF volume in the rat (250 μL).
We were quite surprised to find that the dose of morphine required to obtain a seven- to eight-second TFL after initial injection in our study (0.1 nmol) was at least 100-fold less than the dose used by other investigators using the cervical catheter technique (23). We initially performed pilot studies using 10 nmol of morphine, but all the animals had 10-second TFL latencies and subsequently died. The dose we finally chose was determined empirically. Our findings suggest that catheter placement alone may profoundly decrease morphine potency. The reasons for this are not clear. Prado (14) has shown that catheter placement alone decreases morphine potency and that this decrease is partially reversed by indomethacin, suggesting that an inflammatory response could mediate this effect. Long et al. (13) also demonstrated that cervical catheterization decreased the potency of δ and kappa opioid agonists. Other investigators have shown that chronic cervical catheterization can modulate nociceptive responses in rats without significantly altering baseline response (5,12). It has also been suggested, but not proven, that inflammatory responses may accelerate the development of morphine tolerance (24). The polyethylene tubing most often used for intrathecal catheterization can elicit a substantial inflammatory response (21). The use of polyurethane tubing may attenuate, but not eliminate, this response (11). We found that intermittent LP does cause a mild, localized inflammatory response (Fig. 7). However, this is much less extensive than the response caused by intrathecal catheterization. Future experiments are required to determine whether catheter-induced inflammation alters opioid tolerance development or simply decreases opioid potency.
In conclusion, we have developed a method of intermittent LP for spinal opioid administration that provides consistent and reproducible results. This method avoids the introduction of a catheter into the intrathecal space and thus any possible effect that it could have on morphine potency or tolerance development. In contrast with catheter techniques, the morbidity and mortality of intermittent LP seems to be minimal. With practice, intermittent LP can be performed quickly and reproducibly. This method should provide a useful complement to currently available techniques for the evaluation of spinal analgesic responses.
1. Yaksh TL, Rudy TA. Chronic catheterization of the spinal subarachnoid space. Physiol Behav 1976;17:1013–6.
2. Storkson R, Kjorsvik A, Tjolsen A, Hole K. Lumbar catheterization of the spinal subarachnoid space in the rat. J Neurosci Methods 1996;65:167–72.
3. Kristensen JD, Post C, Gordh T Svensson BA. Spinal cord morphology and antinociception after chronic intrathecal administration of excitatory amino acid antagonists in the rat. Pain 1993;54:309–16.
4. Tsang BK, He Z, Ma T, et al. Decreased paralysis and better motor coordination with microspinal versus PE10 intrathecal catheters in pain study rats. Anesth Analg 1997;84:591–4.
5. Serpell M, DeLeo JA, Coombs DW, et al. Intrathecal catheterization alone reduces autotomy after sciatic cryoneurolysis in the rat. Life Sci 1993;53:1887–92.
6. Fairbanks CA. Spinal delivery of analgesics in experimental models of pain and analgesia. Adv Drug Deliv Rev 2003;55:1007–41.
7. Dib B. Intrathecal chronic catheterization in the rat. Pharmacol Biochem Behav 1984;20:45–8.
8. Pogatzki E, Zahn PK, Brennan TJ. Lumbar catheterization of the subarachnoid space with a 32-gauge polyurethane catheter in the rat. Eur J Pain 2000;4:111–3.
9. Milligan ED, Hinde JL, Mehmert KK, et al. A method for increasing the viability of the external portion of lumbar catheters placed in the spinal subarachnoid space of rats. J Neurosci Methods 1999;90:81–6.
10. Jones LL, Tuszynski MH. Chronic intrathecal infusions after spinal cord injury cause scarring and compression. Microsc Res Tech 2001;54:317–24.
11. Sakura S, Hashimoto K, Bollen AW, et al. Intrathecal catheterization in the rat: improved technique for morphologic analysis of drug-induced injury. Anesthesiology 1996;85:1184–9.
12. Almeida FR, Schivo IR, Lorenzetti BB, Ferreira SH. Chronic intrathecal cannulation enhances nociceptive responses in rats. Braz J Med Biol Res 2000;33:949–56.
13. Long JB, Mobley WC, Holaday JW. Neurological dysfunction after intrathecal injection of dynorphin A (1-13) in the rat. I. Injection procedures modify pharmacological responses. J Pharmacol Exp Ther 1988;246:1158–66.
14. Prado WA. Antinociceptive potency of intrathecal morphine in the rat tail flick test: a comparative study using acute lumbar catheter in rats with or without a chronic atlanto-occipital catheter. J Neurosci Methods 2003;129:33–9.
15. Hylden JL, Wilcox GL. Intrathecal morphine in mice: a new technique. Eur J Pharmacol 1980;67:313–6.
16. De la Calle JL, Paino CL. A procedure for direct lumbar puncture in rats. Brain Res Bull 2002;59:245–50.
17. Mestre C, Pelissier T, Fialip J, et al. A method to perform direct transcutaneous intrathecal injection in rats. J Pharmacol Toxicol Methods 1994;32:197–200.
18. Papir-Kricheli D, Frey J, Laufer R, et al. Behavioural effects of receptor-specific substance P agonists. Pain 1987;31:263–76.
19. Dirig DM, Salami A, Rathbun ML, et al. Characterization of variables defining hindpaw withdrawal latency evoked by radiant thermal stimuli. J Neurosci Methods 1997;76:183–91.
20. Rieselbach RE, Di Chiro G, Freireich EJ, Rall DP. Subarachnoid distribution of drugs after lumbar injection. N Engl J Med 1962;267:1273–8.
21. Yaksh TL, Malkmus SA. Animal models of intrathecal and epidural drug delivery. In: Yaksh TL, ed. Spinal drug delivery. Amsterdam: Elsevier, 1999:317–44.
22. Amit Z, Galina ZH. Stress induced analgesia plays an adaptive role in the organization of behavioral responding. Brain Res Bull 1988;21:955–8.
23. Mao J, Sung B, Ji RR, Lim G. Chronic morphine induces downregulation of spinal glutamate transporters: implications in morphine tolerance and abnormal pain sensitivity. J Neurosci 2002;22:8312–23.
© 2006 International Anesthesia Research Society
24. DeLeo JA, Tanga FY, Tawfik VL. Neuroimmune activation and neuroinflammation in chronic pain and opioid tolerance/hyperalgesia. Neuroscientist 2004;10:40–52.