Neuroscience Vol. 42, No. 2, pp. 555-560, 1991
0306-4522/91 $3.00+ 0.00 Pergamon Press pie
Printed in Great Britain
IBRO
TROPHIC FUNCTIONS OF PRIMARY SENSORY NEURONS: ARE THEY REALLY LOCAL? G. G. KNYAZEV,* G. B. KNYAZEVA a n d Z. S. TOLOCHKO Laboratory of Neuromorphology, Institute of Physiology, Siberian Branch of the U.S.S.R. Academy of Medical Sciences, Novosibirsk, 630090, U.S.S.R. Abstract--The results of the present study, in the rat and cat, indicate that not only a lesion of peripheral nerve or capsaicin pretreatment but also pharmacological deafferentation with local anaesthetic or disruption of the connections between primary sensory neurons and the central nervous system are effective in producing dystrophic changes in tissues. These effects of deafferentation do not seem to depend on the sympathetic or parasympathetic efferents. Dystrophic changes are connected with microcirculation disturbances: slow down of local blood flow, elevation of the vascular permeability, oedema and leucocyte infiltration. The findings indicate that capsalcin-sensitive primary sensory neurons are the afferent part of some reflex arrangement which participates in the regulation of microcirculation and the maintenance of trophic processes in peripheral tissues. The efferent part of this arrangement is unknown.
Some d a t a indicate t h a t capsaicin-sensitive afferents m a y be involved in d e t e r m i n i n g a t r o p h i c action o n peripheral tissues. It has been reported t h a t capsaieindesensitized animals present " s p o n t a n e o u s " cutaneous lesions 6 a n d keratitis-like corneal changes. 2 The nature o f these lesions is u n k n o w n . It has been suggested t h a t neuropeptides released from the peripheral sensory terminals m a y exert a " t o n i c " trophic action o n the tissues .4,8,10 The idea o f a t r o p h i c action o f p r i m a r y sensory n e u r o n s o n peripheral tissues is n o t a new one. It is a n old o b s e r v a t i o n t h a t peripheral nerve lesions m a y be a c c o m p a n i e d by dystrophic changes in tissues. 7 Some 2 0 - 3 0 years ago such neurogenic dystrophies were intensely investigated in o u r country. A great body o f evidence has been accumulated indicating that deafferentation of different peripheral tissues results in chronic inflamation including vasodilatation, oedema, leucocyte infiltration, d a m a g e or proliferation o f p a r e n c h y m a l cells. 3'11'13 T h e reasons a n d m e c h a n i s m s of neurogenic dyst r o p h y are u n k n o w n . The f u r t h e r investigation o f this question was the aim o f the present study. EXPERIMENTAL PROCEDURES
Male Wistar rats (weighing 180-250 g) from Stolbovoye (Moscow) and cats of both sexes (weighing 1.5-2.5 kg) were used in this study. Surgical procedures were performed under pentobarbital anaesthesia. The branches of ophthalmic nerve were transected intraorbitally near the place where the nerve enters the orbit via the medial end of the superior orbital fissure. Electrocoagulation of the main sensory nucleus of the trigeminal nerve was performed stereotaxically. The head of an animal was firmly fixed in a stereotaxic frame, and a unipolar electrode was inserted according to calculated coordinates. The coordinates for the electrode were: 1 mm *To whom correspondence should be addressed. NSC42/2--1
555
posterior to the intersection of the sagittal and lambdoidal sutures, 2.5 mm lateral to the sagittal suture, and 8 mm inferior to the upper surface of the skull bone. Direct current of 4 A amplitude was used for 20 s. After experiments brains were immersed into 10% formalin. Frontal cryostat sections were used for histological control. In some experiments a needle of microinjector was inserted in the main sensory nucleus of a rat instead of the electrode; 0.4pf 5% trimecaine was injected for 2min. The same amount of isotonic saline was injected in control rats. Unilateral extirpation of the superior cervical ganglion was performed one week before deafferentation of the eye. Extirpation of dorsal root ganglia or dorsal root cutting was performed in cats bilaterally at the level from T8 to L2. Pharmacological treatments were performed under light ether anaesthesia. Different doses of trimecaine from 50 #1 1% solution up to 100 # 15 % solution in isotonic saline were injected retrobulbary. Control rats received the same amount of isotonic saline. Capsaicin ("Reanal", Hungary) was dissolved in saline using ethanol and Tween-80 and was given retrobulbary at a single dose of 0.1 ml 0.5% solution. Control rats received vehicle solution. Atropine (1.5%) was instilled in an eye four times a day for two days. The rate of blood flow was evaluated by registration of iris blood vessels fluorescence after intravenous injection of 0.2 ml 1% solution of fluorescein. The procedure was performed under ether anaesthesia. The head of the animal was fixed in a stereotaxic frame mounted on the thermostat specimen stage of a fluorescent microscope. The opaque equipment for incident light with the FS-4 as lamp filter was used as the illumination system. The fluorescence intensity was measured by a photomultiplier. The electric signal from the photomultiplier was supplied to a computer. The registration of fluorescence intensity was performed once per second. Data were accumulated in a computer, and when an experiment was over an average diagram was realized for the experimental and control eyes. In order to evaluate the blood-aqueous barrier breakdown, the animals were injected with 50 mg/kg Evans Blue dye intravenously 50 min before the operation; 1.5 h after the operation they were killed by ether anaesthesia and the dye concentration in the aqueous humour of the experimental and contralateral eyes was determined photometrically.
556
G.G. KNYAZEVet
The neutrophil leucopoiesis was inhibited by the method of Chusid et al. t Two to four days after irradiation of the whole body the leucocyte amount was calculated in the peripheral blood of rats and they were then used in the experiments. In order to study the histopathological changes in the corneae the animals were killed two days after the operation. The corneae obtained from the experimental and control eyes of each animal were immersed in 80% ethanol and then embedded into paraffin. Sections were stained with haematoxyline-eosin or gallocyanine-chromealum. The thickness of corneal strome and epithelium and the diameters of epithelial cells were measured by an ocular-micrometer. In experiments with electrocoagulation of the main sensory nucleus of the trigeminal nerve and with retrobulbar injection of trimecaine the Gasser±an ganglia were taken. After histological staining using the Nissl method they were evaluated to reveal the signs of degeneration of neurons. In order to reveal sympathetic fibres, an iris was stretched on a mount, dried overnight over phosphorus pentoxide, stored for 1 h in formaldehyde vapour heated to + 80°C and then mounted into liquid paraffin. Cat's liver tissue was taken two weeks after deafferentation in the morning after 24 h of food deprivation. After being fixed in 80% ethanol and paraffin embedding, the sections were stained with haematoxyline-eosin or gallocyanine chrome alum, or with PAS reaction for the demonstration of glycogen. In the last case the amylase-treated sections were used as a control. When RNA content was evaluated, control sections were treated with RNAase. Quantitative cytophotometric measurement of gallocyanine-chromealum-stained hepatocytes was performed by cytophotometer LUMAM (LOMO) using the plug-method. The spot size was 1/~m diameter with a x90 objective. Three measures per hepatocyte were taken. The results have been expressed in relative absorbance units. The thickness of sections was measured with the help of interference microscope Amplival interphako (Zeiss). Mean values and standard deviations were calculated and the differences between various groups were tested for statistical significance using Student's t-test at 2P < 0.05. RESULTS During 1.5-2 days after transection of the ophthalmic nerve the diffuse edematous opacity of the cornea was developing and usually in 2-3 days an ulcer was seen. Light microscopic examination of the corneae dissected two days after the operation revealed a variety of pathological changes. The epithelial cells became edematous. Their diameters were significantly larger than those in contralateral eye and in the eyes of control animals. The surface of the epithelium was eroded. The thickness of epithelium increased on average three times and that of the stroma five times as compared with control eyes (see
al.
Table 1). Intensive leucocyte infiltration was observed in both the stroma and epithelium. In other experiments deafferentation of eyes was performed by electrocoagulation of the main sensory nucleus of the trigeminal nerve. Only those animals in which electrocoagulation resulted in the loss of nictitating reflex and histological control, confirming the exact position of a lesion, were taken into consideration. Changes in the eyes in this case were the same as after the transection of the ophthalmic nerve. On the next day the opacity of corneae appeared, then a typical picture of neuroparalytic keratitis developed including oedema, leucocyte infiltration and the damage of epithelial cells. Quantitative differences between the two kinds of operation were insignificant. Histological evaluation of Gasser±an ganglion neurons after electrocoagulation of the main sensory nucleus of the trigeminal nerve does not reveal any signs of degeneration. In some experiments the injection of 0.4/~1 5% trimecaine in the main sensory nucleus of a rat was made instead of its electrocoagulation. This usually resulted in the loss of nictitating reflex. Then as a rule neuroparalytic keratitis developed in 2-3 days. Its severity was, however, slightly less than in previous cases. Pharmacological deafferentation by retrobulbar injection of 100/~1 5% trimecaine led to even more pronounced features of inflammation than the surgical one. Leucocyte infiltration, oedema, damage of corneal stroma and epithelium appeared somewhat earlier. Lower doses (up to 50 #1 1% trimecaine) were also able to evoke neuroparalytic keratitis, but only in 30-50% of cases. The neurons of the Gasser±an ganglion did not display any signs of degenerative changes after retrobulbar injection of trimecaine. The retrobulbar injection of saline in control rats never led to cornea inflammation. In all cases of deafferentation, attempts were made to prevent the drying and damaging of the cornea because of the lost nictitating reflex. The moistening of the corneal surface with a 2% solution of hydroxymethylcellulose and careful sewing together of eyelids did not have any effect on blood-aqueous barrier breakdown, delaying the appearance of other features of inflammation. In the case of retrobulbar trimecaine injection, sewing together of eyelids prevented the appearance of the final stages of neuroparalytic keratitis. In some 4-5 days the sensitivity of a cornea
Table 1. Effect of deafferentation on corneal thickness and epithelial cell diametert Epithelial thickness
Stroma thickness
Epithelial cell diameter
Ophthalmic nerve transection
250 ± 87**
428 ± 70**
132 ± 25*
Electrocoagulation of the main sensory nucleus of the trigeminal nerve
198 ± 35**
470 ± 81"*
151 ± 31"*
Retrobulbar injection of trimecaine
276 ± 67**
517 ± 71"*
148 ± 23**
Retrobulbar injection of capsaicin
210 ± 30**
440 ± 89**
169 ± 15"*
Type of deafferentation
tData are expressed as percentages of corresponding values in contralateral eye, as means _ S.E. of at least 10 experiments. *P < 0.05 and **P < 0.01, significantly different from the contralateral eyes.
Trophic functions of primary sensory neurons
557
Table 2. Effect of deafferentation on vascular permeability of the eyes as visualized by the Evans Blue technique Concentration of Evans Blue dye in the aqueous hnmour 1.5 h after an operationt
Type of deafferentation Ophthalmic nerve transection Electrocoagulation of the main sensory nucleus of the trigeminal nerve
154+ 11" 202 + 43*
Retrobulbar injection of trimecaine
252 + 27*
Retrobulbar injection of capsaicin
152 __+12"
tData are expressed as percentages of corresponding values in contralateral eye, as means + S.E. of at least 10 experiments. *P < 0.01, significantly different from the contralateral eyes.
o. 50-
_
i
I
I
I
I
tO
15
20
25
30
Time (s)
Fig. 1. The fluorescence of iris vessels after intravenous injection of 0.2ml 1% fluorescein. The beginning and the end of the injection are indicated with arrows. (&--A) Control animals; (O--O) 5 min after retrobulbar injection of 0.1 ml 5% trimecalne. was restored and all signs of inflammation gradually disappeared. In other experiments the influence of retrobulbar capsaicin injection was tested. In some cases administration of 0.1 ml 0.5% capsaicin solution produced the loss of the nictitating reflex. In others it was
reduced. On the next day diffuse opacity of the corneae was seen. Histological evaluation revealed intensive leucocyte infiltration of corneal stroma and epithelium, and increases in the corneal thickness and the diameter of epithelial cells. These pathological changes did not occur when the same amount of vehicle solution was injected retrobulbary. To find the reasons for pathological changes in deafferentated cornea the microcirculatory disturbances were investigated. In the first 5 min after retrobulbar injection of trimecaine a significant slowdown of blood flow in the iris vessels was revealed (Fig. 1). The retrobulbar injection of capsaicin evoked the same effect. The slowdown of blood flow after transection of ophthalmic nerve and electrocoagulation of the main sensory nucleus of the trigeminal nerve was pronounced to a lesser extent. All kinds of deafferentation resulted in a marked elevation of the vascular permeability of the eye, as visualized by the Evans Blue technique. One and a half hours after the operation the concentration of the dye in the aqueous humour of the experimental eye was on average 1.5 times higher than in the control eye, the difference being significant (see Table 2).
Fig. 2(A).
5.58
G. G. KNYAZEV et
al.
Fig. 2(B).
Fig. 2(C). Fig. 2. The influence of leucocyte infiltration on the histopathological changes in deafferentated rat cornea. Gallocyanine-chromealum staining. (A) Control rat; (B) one day after surgical deafferentation of the eye; (C) four days after whole body irradiation, one day after surgical deafferentation of the eye. The role of leucocytes in the development of neuroparalytic keratitis was investigated in the next experiments. Surgical and pharmacological deafferentation of the eyes was performed after inhibition of neutrophi1 leucopoiesis. Two to four days after irradiation the leucocyte content in the peripheral blood was 10 times lower than in control rats. Pathological changes in deafferentated wrneae were pronounced to a signifi-
cantly lesser degree. One day after deafferentation the epithelium thickness was not changed, and the thickness of the stroma increased to a lesser extent than in control rats. The erosion of the epithelium and increase of epithelial cell diameters were not seen (Fig. 2). In order to check whether the sympathetic efferents are involved in the genesis of neuroparalytic keratitis,
Trophic functions of primary sensory neurons
559
Table 3. Effect of deafferentation on diameter of hepatocytes and RNA content in the cytoplasmt Diameter of hepatocytes
RNA content in the cytoplasm
Dorsal root ganglia extirpation
117 + 16"
236 + 50**
Dorsal root transection
120 __+12"
184 _ 32**
Type of deafferentation
tData are expressed as percentages of corresponding values in control group, as means _ S.E. of at least 10 experiments. *P < 0.05 and **P < 0.01, significantly different from control values.
superior cervical ganglion extirpation was performed in rats. The Evans Blue test and histological evaluation did not reveal any signs of inflammation after the operation. The histochemical examination with Falck's method showed a lack of sympathetic terminals in irises one week after desympathization. When surgical deafferentation of a desympathized eye was carried out, neuroparalytic keratitis developed as usual. The removal of parasympathetic influences was achieved by atropine instillations. Ten minutes after the first application, surgical deafferentation of the eye was carried out. During the next 48 h the instillations were repeated seven times. This had no influence on the dynamics of neuroparalytic keratitis. The following experiments were performed using cats. Two weeks after bilateral extirpation of dorsal root ganglia histological examination revealed some leucocyte infiltration in the liver. Morphometric evaluation showed a significant increase in the diameter of hepatocytes, especially in the central parts of a lobule. The RNA content in hepatocyte cytoplasm was also increased (see Table 3). Glycogen in parenchymal cells was significantly reduced and in some instances had completely disappeared. In other groups of animals the transection of dorsal roots instead of dorsal root ganglia extirpation was carried out. Two weeks later the same picture of histological changes was seen in liver tissue. DISCUSSION In our experiments all four kinds of deafferentation caused the same changes in cornea. First of all microcirculatory disturbances appeared. In the first few minutes the blood flow slowed down and the vascular permeability was elevated. This resulted in oedema and the development of leucocyte infiltration. The leucocytes probably evoked damage of epithelial cell membranes, which resulted in increases of epithelial cell diameters and the thickness of the epithelium. The absence of these changes after inhibition of leucopoiesis shows that leucocytes participate in these effects. There is always doubt when the causes of neuroparalytic keratitis are discussed. All four kinds of deafferentation result in a loss or weakening of the nictitating reflex. The drying and damage of the corneal surface increase its susceptibility to bacterial
infections. It seems probable that earlier changes, such as the breakdown of the blood-aqueous barrier and the slowdown of blood flow, are not connected with the loss of the nictitating reflex. However, for several hours the increased susceptibility of the cornea may aggravate the changes evoked by circulatory disturbances. So, the microcirculatory disturbances may be the main reason for trophic changes after deafferentation. And what is the reason for microcirculatory disturbances? Sympathetic and parasympathetic reflexes do not seem to play a decisive role, because surgical desympathization or pharmacological deparasympathization had no effect on neuroparalytic keratitis development. Retrobulbar injection of capsaicin caused, in our experiments, the same trophic changes in the cornea as other kinds of deafferentation. So, one may suppose that neuropeptides participate in these processes. Such supposition seems attractive if one takes into consideration recent data about the importance of neuropeptides in the regulation of microcirculation and the maintenance of trophic processes. 4"~°Trophic disturbances may be connected with both an excessive release of neuropeptides from peripheral endings of primary sensory neurons5 and their depletion. 2 We have recently revealed that excessive release of neuropeptides is not involved in the development of neuroparalytic keratitis and the reason for such a dystrophy may only be connected with their deep local depletion.6 In the last few years evidence has been accumulated to support a local effector role of sensory nerve endings. It is generally accepted that sensory nerve endings, when stimulated, can release neuropeptides,9,12 which then diffuse to effector tissues to cause various local reactions. So, primary sensory neurons may exert a dual sensory-efferent function. The same stimuli may, on the one hand, give rise to sensation and, on the other hand, set into operation local processes by the release of mediators from peripheral sensory nerve endings. It remains unclear which mechanism (or both) participates in the realization of trophic functions in sensory neurons. The transection of ophthalmic nerve as well as retrobulbar injection of capsaicin causes the depletion of neuropeptides in sensory endings that, as a result, obviously blocks both the above-mentioned mechanisms. However, retrobulbar injection of the local anaesthetic trimecaine cannot deplete the peripheral neuropeptides.
560
G.G. KNVAZEVet al.
The absence of degenerative changes in the Gasserian ganglion after retrobulbar injection of trimecaine and the restoration of corneal sensitivity when eyelids are sewn together prove that the functions of primary sensory neurons do not change after retrobulbar injection of trimecaine. The primary sensory neurons and their connections with peripheral tissues remain unchanged after electrocoagulation of the main sensory nucleus of the trigeminal nerve or injection of trimecaine into it. This has been proved by the absence of degenerative changes in the neurons of the Gasserian ganglion. The development of neuroparalytic keratitis in the latter case testifies that connections of primary sensory neurons with the central nervous system have a
critical significance for the integrity of corneal tissue. This mechanism probably exists in visceral tissues too, because the transection of dorsal roots is as effective in evoking the dystrophic liver changes as the extirpation of dorsal root ganglia. The connections of sensory neurons with central structures seem to be important for the maintenance of liver tissue integrity, as well as for the cornea. So, one may speculate that capsaicin-sensitive sensory neurons are an afferent part of some reflex arrangement participating in the regulation of microcirculation and the maintenance of trophic functions in peripheral tissues. The effector part of the arrangement should include efferents other than sympathetic or parasympathetic.
REFERENCES
1. Chusid M. J., Nelson D. B. and Meyer L. A. (1986) The role of the polymorphonuclear leucocyte in the induction of corneal edema. Invest. Ophthal. vis. Sci. 27, 1466-1469. 2. Fujita S., Shimizi T., Izumi K., Fukuda T., Sameshima M. and Ohbe N. (1984) Capsaicin-induced neuroparalytic keratitis-like corneal changes in the mouse. Expl Eye Res. 38, 165-175. 3. Grigorjeva T. A. (1959) Sensory neuron as a factor of integrity and adequate differentiation of the structures innervated by it. Archs Anat. Hist. Embryol. 36, 3-12 [in Russian]. 4. Holzer P. (1988) Local effector functions of capsaicin-sensitive sensory nerve endings: involvement of tachykinins, calcitonin gene-related peptide and other neuropeptides. Neuroscience 24, 739-768. 5. Jancs6 N., Jancs6-Gabor A. and Szolcs~nyi J. (1967) Direct evidence for neurogenic inflammation and its prevention by denervation and pretreatment with capsaicin. Br. J. Pharmac. 31, 138-151. 6. Knyazev G. G., Knyazeva G. B. and Nikiforov A. F. (1990) Neuroparalytic keratitis and capsaicin. Acta physiol, hung. 75, 29-34. 7. Lewis T. and Marvin H. M. (1927) Observations relating to vasodilatation arising from antidromic impulses, to herpes zoster and trophic effects. Heart 14, 27-46. 8. Maggi C. A., Borsini F., Santicidi P., Geppetti P., Abelli L., Evangelista S., Manzini S., Theodorsson-Norheim E., Somma V., Amenta F., BacciareUi C. and Meli A. (1987) Cutaneous lesions in capsaicin-pretreated rats. NaunynSchmiedeberg's Arch. Pharmac. 336, 538-545. 9. Maggi C. A. and Meli A. (1986) The role of neuropeptides in the regulation of micturition reflex. J. auton. Pharmac. 6, 133-162. 10. Maggi C. A. and Meli A. (1988) The sensory~efferent function of capsaicin-sensitive sensory neurons. Gen. Pharmac. 19, 1-43. 11. Nikiforov A. F. (1973) Afferent Neuron and Neurodystrophic Processes. Medicina, Moscow [in Russian]. 12. Slesinger P. and Bell C. C. (1985) Primary afferents conduct impulses in both directions under physiological stimulus conditions. J. comp. Physiol. A157, 15-22. 13. Volkova O. V. (1978) Neurodystrophic Processes. Medicina, Moscow [in Russian]. (Accepted 4 September 1990)
0306-4522/91 $3.00+ 0.00 Pergamon Press pie
Printed in Great Britain
IBRO
TROPHIC FUNCTIONS OF PRIMARY SENSORY NEURONS: ARE THEY REALLY LOCAL? G. G. KNYAZEV,* G. B. KNYAZEVA a n d Z. S. TOLOCHKO Laboratory of Neuromorphology, Institute of Physiology, Siberian Branch of the U.S.S.R. Academy of Medical Sciences, Novosibirsk, 630090, U.S.S.R. Abstract--The results of the present study, in the rat and cat, indicate that not only a lesion of peripheral nerve or capsaicin pretreatment but also pharmacological deafferentation with local anaesthetic or disruption of the connections between primary sensory neurons and the central nervous system are effective in producing dystrophic changes in tissues. These effects of deafferentation do not seem to depend on the sympathetic or parasympathetic efferents. Dystrophic changes are connected with microcirculation disturbances: slow down of local blood flow, elevation of the vascular permeability, oedema and leucocyte infiltration. The findings indicate that capsalcin-sensitive primary sensory neurons are the afferent part of some reflex arrangement which participates in the regulation of microcirculation and the maintenance of trophic processes in peripheral tissues. The efferent part of this arrangement is unknown.
Some d a t a indicate t h a t capsaicin-sensitive afferents m a y be involved in d e t e r m i n i n g a t r o p h i c action o n peripheral tissues. It has been reported t h a t capsaieindesensitized animals present " s p o n t a n e o u s " cutaneous lesions 6 a n d keratitis-like corneal changes. 2 The nature o f these lesions is u n k n o w n . It has been suggested t h a t neuropeptides released from the peripheral sensory terminals m a y exert a " t o n i c " trophic action o n the tissues .4,8,10 The idea o f a t r o p h i c action o f p r i m a r y sensory n e u r o n s o n peripheral tissues is n o t a new one. It is a n old o b s e r v a t i o n t h a t peripheral nerve lesions m a y be a c c o m p a n i e d by dystrophic changes in tissues. 7 Some 2 0 - 3 0 years ago such neurogenic dystrophies were intensely investigated in o u r country. A great body o f evidence has been accumulated indicating that deafferentation of different peripheral tissues results in chronic inflamation including vasodilatation, oedema, leucocyte infiltration, d a m a g e or proliferation o f p a r e n c h y m a l cells. 3'11'13 T h e reasons a n d m e c h a n i s m s of neurogenic dyst r o p h y are u n k n o w n . The f u r t h e r investigation o f this question was the aim o f the present study. EXPERIMENTAL PROCEDURES
Male Wistar rats (weighing 180-250 g) from Stolbovoye (Moscow) and cats of both sexes (weighing 1.5-2.5 kg) were used in this study. Surgical procedures were performed under pentobarbital anaesthesia. The branches of ophthalmic nerve were transected intraorbitally near the place where the nerve enters the orbit via the medial end of the superior orbital fissure. Electrocoagulation of the main sensory nucleus of the trigeminal nerve was performed stereotaxically. The head of an animal was firmly fixed in a stereotaxic frame, and a unipolar electrode was inserted according to calculated coordinates. The coordinates for the electrode were: 1 mm *To whom correspondence should be addressed. NSC42/2--1
555
posterior to the intersection of the sagittal and lambdoidal sutures, 2.5 mm lateral to the sagittal suture, and 8 mm inferior to the upper surface of the skull bone. Direct current of 4 A amplitude was used for 20 s. After experiments brains were immersed into 10% formalin. Frontal cryostat sections were used for histological control. In some experiments a needle of microinjector was inserted in the main sensory nucleus of a rat instead of the electrode; 0.4pf 5% trimecaine was injected for 2min. The same amount of isotonic saline was injected in control rats. Unilateral extirpation of the superior cervical ganglion was performed one week before deafferentation of the eye. Extirpation of dorsal root ganglia or dorsal root cutting was performed in cats bilaterally at the level from T8 to L2. Pharmacological treatments were performed under light ether anaesthesia. Different doses of trimecaine from 50 #1 1% solution up to 100 # 15 % solution in isotonic saline were injected retrobulbary. Control rats received the same amount of isotonic saline. Capsaicin ("Reanal", Hungary) was dissolved in saline using ethanol and Tween-80 and was given retrobulbary at a single dose of 0.1 ml 0.5% solution. Control rats received vehicle solution. Atropine (1.5%) was instilled in an eye four times a day for two days. The rate of blood flow was evaluated by registration of iris blood vessels fluorescence after intravenous injection of 0.2 ml 1% solution of fluorescein. The procedure was performed under ether anaesthesia. The head of the animal was fixed in a stereotaxic frame mounted on the thermostat specimen stage of a fluorescent microscope. The opaque equipment for incident light with the FS-4 as lamp filter was used as the illumination system. The fluorescence intensity was measured by a photomultiplier. The electric signal from the photomultiplier was supplied to a computer. The registration of fluorescence intensity was performed once per second. Data were accumulated in a computer, and when an experiment was over an average diagram was realized for the experimental and control eyes. In order to evaluate the blood-aqueous barrier breakdown, the animals were injected with 50 mg/kg Evans Blue dye intravenously 50 min before the operation; 1.5 h after the operation they were killed by ether anaesthesia and the dye concentration in the aqueous humour of the experimental and contralateral eyes was determined photometrically.
556
G.G. KNYAZEVet
The neutrophil leucopoiesis was inhibited by the method of Chusid et al. t Two to four days after irradiation of the whole body the leucocyte amount was calculated in the peripheral blood of rats and they were then used in the experiments. In order to study the histopathological changes in the corneae the animals were killed two days after the operation. The corneae obtained from the experimental and control eyes of each animal were immersed in 80% ethanol and then embedded into paraffin. Sections were stained with haematoxyline-eosin or gallocyanine-chromealum. The thickness of corneal strome and epithelium and the diameters of epithelial cells were measured by an ocular-micrometer. In experiments with electrocoagulation of the main sensory nucleus of the trigeminal nerve and with retrobulbar injection of trimecaine the Gasser±an ganglia were taken. After histological staining using the Nissl method they were evaluated to reveal the signs of degeneration of neurons. In order to reveal sympathetic fibres, an iris was stretched on a mount, dried overnight over phosphorus pentoxide, stored for 1 h in formaldehyde vapour heated to + 80°C and then mounted into liquid paraffin. Cat's liver tissue was taken two weeks after deafferentation in the morning after 24 h of food deprivation. After being fixed in 80% ethanol and paraffin embedding, the sections were stained with haematoxyline-eosin or gallocyanine chrome alum, or with PAS reaction for the demonstration of glycogen. In the last case the amylase-treated sections were used as a control. When RNA content was evaluated, control sections were treated with RNAase. Quantitative cytophotometric measurement of gallocyanine-chromealum-stained hepatocytes was performed by cytophotometer LUMAM (LOMO) using the plug-method. The spot size was 1/~m diameter with a x90 objective. Three measures per hepatocyte were taken. The results have been expressed in relative absorbance units. The thickness of sections was measured with the help of interference microscope Amplival interphako (Zeiss). Mean values and standard deviations were calculated and the differences between various groups were tested for statistical significance using Student's t-test at 2P < 0.05. RESULTS During 1.5-2 days after transection of the ophthalmic nerve the diffuse edematous opacity of the cornea was developing and usually in 2-3 days an ulcer was seen. Light microscopic examination of the corneae dissected two days after the operation revealed a variety of pathological changes. The epithelial cells became edematous. Their diameters were significantly larger than those in contralateral eye and in the eyes of control animals. The surface of the epithelium was eroded. The thickness of epithelium increased on average three times and that of the stroma five times as compared with control eyes (see
al.
Table 1). Intensive leucocyte infiltration was observed in both the stroma and epithelium. In other experiments deafferentation of eyes was performed by electrocoagulation of the main sensory nucleus of the trigeminal nerve. Only those animals in which electrocoagulation resulted in the loss of nictitating reflex and histological control, confirming the exact position of a lesion, were taken into consideration. Changes in the eyes in this case were the same as after the transection of the ophthalmic nerve. On the next day the opacity of corneae appeared, then a typical picture of neuroparalytic keratitis developed including oedema, leucocyte infiltration and the damage of epithelial cells. Quantitative differences between the two kinds of operation were insignificant. Histological evaluation of Gasser±an ganglion neurons after electrocoagulation of the main sensory nucleus of the trigeminal nerve does not reveal any signs of degeneration. In some experiments the injection of 0.4/~1 5% trimecaine in the main sensory nucleus of a rat was made instead of its electrocoagulation. This usually resulted in the loss of nictitating reflex. Then as a rule neuroparalytic keratitis developed in 2-3 days. Its severity was, however, slightly less than in previous cases. Pharmacological deafferentation by retrobulbar injection of 100/~1 5% trimecaine led to even more pronounced features of inflammation than the surgical one. Leucocyte infiltration, oedema, damage of corneal stroma and epithelium appeared somewhat earlier. Lower doses (up to 50 #1 1% trimecaine) were also able to evoke neuroparalytic keratitis, but only in 30-50% of cases. The neurons of the Gasser±an ganglion did not display any signs of degenerative changes after retrobulbar injection of trimecaine. The retrobulbar injection of saline in control rats never led to cornea inflammation. In all cases of deafferentation, attempts were made to prevent the drying and damaging of the cornea because of the lost nictitating reflex. The moistening of the corneal surface with a 2% solution of hydroxymethylcellulose and careful sewing together of eyelids did not have any effect on blood-aqueous barrier breakdown, delaying the appearance of other features of inflammation. In the case of retrobulbar trimecaine injection, sewing together of eyelids prevented the appearance of the final stages of neuroparalytic keratitis. In some 4-5 days the sensitivity of a cornea
Table 1. Effect of deafferentation on corneal thickness and epithelial cell diametert Epithelial thickness
Stroma thickness
Epithelial cell diameter
Ophthalmic nerve transection
250 ± 87**
428 ± 70**
132 ± 25*
Electrocoagulation of the main sensory nucleus of the trigeminal nerve
198 ± 35**
470 ± 81"*
151 ± 31"*
Retrobulbar injection of trimecaine
276 ± 67**
517 ± 71"*
148 ± 23**
Retrobulbar injection of capsaicin
210 ± 30**
440 ± 89**
169 ± 15"*
Type of deafferentation
tData are expressed as percentages of corresponding values in contralateral eye, as means _ S.E. of at least 10 experiments. *P < 0.05 and **P < 0.01, significantly different from the contralateral eyes.
Trophic functions of primary sensory neurons
557
Table 2. Effect of deafferentation on vascular permeability of the eyes as visualized by the Evans Blue technique Concentration of Evans Blue dye in the aqueous hnmour 1.5 h after an operationt
Type of deafferentation Ophthalmic nerve transection Electrocoagulation of the main sensory nucleus of the trigeminal nerve
154+ 11" 202 + 43*
Retrobulbar injection of trimecaine
252 + 27*
Retrobulbar injection of capsaicin
152 __+12"
tData are expressed as percentages of corresponding values in contralateral eye, as means + S.E. of at least 10 experiments. *P < 0.01, significantly different from the contralateral eyes.
o. 50-
_
i
I
I
I
I
tO
15
20
25
30
Time (s)
Fig. 1. The fluorescence of iris vessels after intravenous injection of 0.2ml 1% fluorescein. The beginning and the end of the injection are indicated with arrows. (&--A) Control animals; (O--O) 5 min after retrobulbar injection of 0.1 ml 5% trimecalne. was restored and all signs of inflammation gradually disappeared. In other experiments the influence of retrobulbar capsaicin injection was tested. In some cases administration of 0.1 ml 0.5% capsaicin solution produced the loss of the nictitating reflex. In others it was
reduced. On the next day diffuse opacity of the corneae was seen. Histological evaluation revealed intensive leucocyte infiltration of corneal stroma and epithelium, and increases in the corneal thickness and the diameter of epithelial cells. These pathological changes did not occur when the same amount of vehicle solution was injected retrobulbary. To find the reasons for pathological changes in deafferentated cornea the microcirculatory disturbances were investigated. In the first 5 min after retrobulbar injection of trimecaine a significant slowdown of blood flow in the iris vessels was revealed (Fig. 1). The retrobulbar injection of capsaicin evoked the same effect. The slowdown of blood flow after transection of ophthalmic nerve and electrocoagulation of the main sensory nucleus of the trigeminal nerve was pronounced to a lesser extent. All kinds of deafferentation resulted in a marked elevation of the vascular permeability of the eye, as visualized by the Evans Blue technique. One and a half hours after the operation the concentration of the dye in the aqueous humour of the experimental eye was on average 1.5 times higher than in the control eye, the difference being significant (see Table 2).
Fig. 2(A).
5.58
G. G. KNYAZEV et
al.
Fig. 2(B).
Fig. 2(C). Fig. 2. The influence of leucocyte infiltration on the histopathological changes in deafferentated rat cornea. Gallocyanine-chromealum staining. (A) Control rat; (B) one day after surgical deafferentation of the eye; (C) four days after whole body irradiation, one day after surgical deafferentation of the eye. The role of leucocytes in the development of neuroparalytic keratitis was investigated in the next experiments. Surgical and pharmacological deafferentation of the eyes was performed after inhibition of neutrophi1 leucopoiesis. Two to four days after irradiation the leucocyte content in the peripheral blood was 10 times lower than in control rats. Pathological changes in deafferentated wrneae were pronounced to a signifi-
cantly lesser degree. One day after deafferentation the epithelium thickness was not changed, and the thickness of the stroma increased to a lesser extent than in control rats. The erosion of the epithelium and increase of epithelial cell diameters were not seen (Fig. 2). In order to check whether the sympathetic efferents are involved in the genesis of neuroparalytic keratitis,
Trophic functions of primary sensory neurons
559
Table 3. Effect of deafferentation on diameter of hepatocytes and RNA content in the cytoplasmt Diameter of hepatocytes
RNA content in the cytoplasm
Dorsal root ganglia extirpation
117 + 16"
236 + 50**
Dorsal root transection
120 __+12"
184 _ 32**
Type of deafferentation
tData are expressed as percentages of corresponding values in control group, as means _ S.E. of at least 10 experiments. *P < 0.05 and **P < 0.01, significantly different from control values.
superior cervical ganglion extirpation was performed in rats. The Evans Blue test and histological evaluation did not reveal any signs of inflammation after the operation. The histochemical examination with Falck's method showed a lack of sympathetic terminals in irises one week after desympathization. When surgical deafferentation of a desympathized eye was carried out, neuroparalytic keratitis developed as usual. The removal of parasympathetic influences was achieved by atropine instillations. Ten minutes after the first application, surgical deafferentation of the eye was carried out. During the next 48 h the instillations were repeated seven times. This had no influence on the dynamics of neuroparalytic keratitis. The following experiments were performed using cats. Two weeks after bilateral extirpation of dorsal root ganglia histological examination revealed some leucocyte infiltration in the liver. Morphometric evaluation showed a significant increase in the diameter of hepatocytes, especially in the central parts of a lobule. The RNA content in hepatocyte cytoplasm was also increased (see Table 3). Glycogen in parenchymal cells was significantly reduced and in some instances had completely disappeared. In other groups of animals the transection of dorsal roots instead of dorsal root ganglia extirpation was carried out. Two weeks later the same picture of histological changes was seen in liver tissue. DISCUSSION In our experiments all four kinds of deafferentation caused the same changes in cornea. First of all microcirculatory disturbances appeared. In the first few minutes the blood flow slowed down and the vascular permeability was elevated. This resulted in oedema and the development of leucocyte infiltration. The leucocytes probably evoked damage of epithelial cell membranes, which resulted in increases of epithelial cell diameters and the thickness of the epithelium. The absence of these changes after inhibition of leucopoiesis shows that leucocytes participate in these effects. There is always doubt when the causes of neuroparalytic keratitis are discussed. All four kinds of deafferentation result in a loss or weakening of the nictitating reflex. The drying and damage of the corneal surface increase its susceptibility to bacterial
infections. It seems probable that earlier changes, such as the breakdown of the blood-aqueous barrier and the slowdown of blood flow, are not connected with the loss of the nictitating reflex. However, for several hours the increased susceptibility of the cornea may aggravate the changes evoked by circulatory disturbances. So, the microcirculatory disturbances may be the main reason for trophic changes after deafferentation. And what is the reason for microcirculatory disturbances? Sympathetic and parasympathetic reflexes do not seem to play a decisive role, because surgical desympathization or pharmacological deparasympathization had no effect on neuroparalytic keratitis development. Retrobulbar injection of capsaicin caused, in our experiments, the same trophic changes in the cornea as other kinds of deafferentation. So, one may suppose that neuropeptides participate in these processes. Such supposition seems attractive if one takes into consideration recent data about the importance of neuropeptides in the regulation of microcirculation and the maintenance of trophic processes. 4"~°Trophic disturbances may be connected with both an excessive release of neuropeptides from peripheral endings of primary sensory neurons5 and their depletion. 2 We have recently revealed that excessive release of neuropeptides is not involved in the development of neuroparalytic keratitis and the reason for such a dystrophy may only be connected with their deep local depletion.6 In the last few years evidence has been accumulated to support a local effector role of sensory nerve endings. It is generally accepted that sensory nerve endings, when stimulated, can release neuropeptides,9,12 which then diffuse to effector tissues to cause various local reactions. So, primary sensory neurons may exert a dual sensory-efferent function. The same stimuli may, on the one hand, give rise to sensation and, on the other hand, set into operation local processes by the release of mediators from peripheral sensory nerve endings. It remains unclear which mechanism (or both) participates in the realization of trophic functions in sensory neurons. The transection of ophthalmic nerve as well as retrobulbar injection of capsaicin causes the depletion of neuropeptides in sensory endings that, as a result, obviously blocks both the above-mentioned mechanisms. However, retrobulbar injection of the local anaesthetic trimecaine cannot deplete the peripheral neuropeptides.
560
G.G. KNVAZEVet al.
The absence of degenerative changes in the Gasserian ganglion after retrobulbar injection of trimecaine and the restoration of corneal sensitivity when eyelids are sewn together prove that the functions of primary sensory neurons do not change after retrobulbar injection of trimecaine. The primary sensory neurons and their connections with peripheral tissues remain unchanged after electrocoagulation of the main sensory nucleus of the trigeminal nerve or injection of trimecaine into it. This has been proved by the absence of degenerative changes in the neurons of the Gasserian ganglion. The development of neuroparalytic keratitis in the latter case testifies that connections of primary sensory neurons with the central nervous system have a
critical significance for the integrity of corneal tissue. This mechanism probably exists in visceral tissues too, because the transection of dorsal roots is as effective in evoking the dystrophic liver changes as the extirpation of dorsal root ganglia. The connections of sensory neurons with central structures seem to be important for the maintenance of liver tissue integrity, as well as for the cornea. So, one may speculate that capsaicin-sensitive sensory neurons are an afferent part of some reflex arrangement participating in the regulation of microcirculation and the maintenance of trophic functions in peripheral tissues. The effector part of the arrangement should include efferents other than sympathetic or parasympathetic.
REFERENCES
1. Chusid M. J., Nelson D. B. and Meyer L. A. (1986) The role of the polymorphonuclear leucocyte in the induction of corneal edema. Invest. Ophthal. vis. Sci. 27, 1466-1469. 2. Fujita S., Shimizi T., Izumi K., Fukuda T., Sameshima M. and Ohbe N. (1984) Capsaicin-induced neuroparalytic keratitis-like corneal changes in the mouse. Expl Eye Res. 38, 165-175. 3. Grigorjeva T. A. (1959) Sensory neuron as a factor of integrity and adequate differentiation of the structures innervated by it. Archs Anat. Hist. Embryol. 36, 3-12 [in Russian]. 4. Holzer P. (1988) Local effector functions of capsaicin-sensitive sensory nerve endings: involvement of tachykinins, calcitonin gene-related peptide and other neuropeptides. Neuroscience 24, 739-768. 5. Jancs6 N., Jancs6-Gabor A. and Szolcs~nyi J. (1967) Direct evidence for neurogenic inflammation and its prevention by denervation and pretreatment with capsaicin. Br. J. Pharmac. 31, 138-151. 6. Knyazev G. G., Knyazeva G. B. and Nikiforov A. F. (1990) Neuroparalytic keratitis and capsaicin. Acta physiol, hung. 75, 29-34. 7. Lewis T. and Marvin H. M. (1927) Observations relating to vasodilatation arising from antidromic impulses, to herpes zoster and trophic effects. Heart 14, 27-46. 8. Maggi C. A., Borsini F., Santicidi P., Geppetti P., Abelli L., Evangelista S., Manzini S., Theodorsson-Norheim E., Somma V., Amenta F., BacciareUi C. and Meli A. (1987) Cutaneous lesions in capsaicin-pretreated rats. NaunynSchmiedeberg's Arch. Pharmac. 336, 538-545. 9. Maggi C. A. and Meli A. (1986) The role of neuropeptides in the regulation of micturition reflex. J. auton. Pharmac. 6, 133-162. 10. Maggi C. A. and Meli A. (1988) The sensory~efferent function of capsaicin-sensitive sensory neurons. Gen. Pharmac. 19, 1-43. 11. Nikiforov A. F. (1973) Afferent Neuron and Neurodystrophic Processes. Medicina, Moscow [in Russian]. 12. Slesinger P. and Bell C. C. (1985) Primary afferents conduct impulses in both directions under physiological stimulus conditions. J. comp. Physiol. A157, 15-22. 13. Volkova O. V. (1978) Neurodystrophic Processes. Medicina, Moscow [in Russian]. (Accepted 4 September 1990)
