- Email: [email protected]
Effect of global system for mobile communication (GSM)-like radiofrequency fields on vascular permeability in mouse brain
Pathology (2001 ) 33, pp. 338– 340
EFFECT OF GLOBAL SYSTEM FOR MOBILE COMMUNICATION (GSM)-LIKE RADIOFREQUENCY FIELDS ON VASCULAR PERMEABILITY IN MOUSE BRAIN JOHN W. FINNIE *, PETER C. BLUM BERGS †, JIM MANAV IS †, TAM M Y D. UTTERIDGE *, VAL GEBSKI‡, JEFFREY G. SW IFT †, BARRIE VERNON -ROBERTS † AND TIM OTHY R. KUCHEL * *Veterinary Services Division and †Division of Tissue Pathology, Institute of Medical and Veterinary Science, Adelaide, SA, and ‡NHMRC Clinical Trials Centre, University of Sydney, Sydney, NSW, Australia
Summary The effect of global system for mobile communication (GSM) radiofrequency fields on vascular permeability in the brain was studied using a purpose-designed exposure system at 898.4 MHz. Mice (n = 30) were given a single far field, whole body exposure for 60 minutes at a specific absorption rate of 4 W/kg. Control mice were also sham-exposed (n = 10) or permitted free movement in a cage (n = 10) to exclude any stress-related effects. Vascular permeability changes were detected using albumin immunohistochemistry and the efficacy of this vascular tracer was confirmed with a positive control group exposed to a clostridial toxin known to increase vascular permeability in the brain. No significant difference in albumin extravasation was detected between any of the groups at the light microscope level using the albumin marker. Key words: Mobile telephone radiation, brain, mouse, vascular permeability, albumin immunohistochemistry. Abbreviations: BBB, blood– brain barrier; GSM, global system for mobile communication; RF, radiofrequency; SAR, specific absorption rate. Received 16 June 2000; revised 19 January 2001; accepted 29 January 2001
INTRODUCTION Although the use of mobile telephones is common and increasing, and requires these devices to be held close to the head, exposure effects on the brain remain controversial and prevent regulatory agencies giving an unqualified assurance of public safety.1 Furthermore, experimental findings have often been contradictory due to variable exposure conditions in diverse animal species and inadequate dosimetry.1,2 Most research has attempted to determine whether such exposure increases blood– brain barrier ( BBB) permeability. However, while some studies in the last decade in laboratory rodents have reported pathologically significant increases in BBB permeability,3,4 others5,6 have been unable to substantiate these findings. We employed purpose-designed exposure modules with precisely determined dosimetry to examine the effect of exposure to radiofrequency (RF) fields, similar to those used for global system for mobile (GSM) telecommunications, on tumour development. Since health problems associated with exposure to low level ( non-thermal ) RF
sources, such as mobile telephones, are poorly understood and probably subtle, we also used this facility to ascertain whether exposure from these devices had any adverse effect on the central nervous system.
MATERIALS AND METHODS In the present study, unanaesthetised, 8-week-old, female, inbred wild-type C57BL/6NTac mice ( Taconic Farms, New York, USA) were used. We were constrained in our choice of mice as only this wild-type and a transgenic strain ( Em -Pim 1) were permitted to occupy this facility to maintain the specific-pathogen-free status of the animals for the purpose of replicating a previous carcinogenesis study.7 Mice were exposed to RF fields with similar pulsing and modulation characteristics to those used for GSM mobile telecommunications ( 898.4-MHz fields modulated at a pulse repetition frequency of 217 Hz and a pulse width of 0.6 ms). The exposure system consisted of a cylindrical parallel plate with mice restrained in clear perspex tubes arranged radially around a dipole antenna.8 The cylinders were small enough to prevent the animals from changing their orientation relative to the field and ensure that the specific absorption rate ( SAR) could be precisely estimated. A forced air supply was also maintained through the cylinders to ensure no heat build-up and the Australian Radiation Protection and Nuclear Safety Agency ( ARPANSA) confirmed that the core body temperature did not rise using a Luxtron. All exposed mice received a single far field, whole body exposure of 4 W/kg in the same exposure system for 60 min; the field within the exposure system was homogeneous so all mice received the same exposure level. Both exposed ( n = 30 ) and sham-exposed ( n = 10 ) groups occupied this system, while another non-exposed control group ( n = 10) was allowed free movement in a cage without handling or further confinement to obviate any stress-related exposure module confinement effects. Three mice in the exposed group were discarded as perfusion fixation was considered suboptimal. A positive control group ( n = 10) was also included for the BBB marker ( albumin), mice being given Clostridium perfringens type D epsilon toxin that produces a rapid and severe generalised vasogenic cerebral oedema in mice.9 ,10 For dosimetry, measurements of RF energy absorption were performed by the ARPANSA ( Mr Michael Bangay) on mouse cadavers to characterise the live animal exposure. Mice were anaesthetised with 0.014 ml/g body weight of Avertin ( 2,2,2-tribromoethanol in tert-amyl alcohol; Aldrich, Milwaukee, WI, USA) immediately after termination of the 60-min exposure period and killed by transcardiac perfusion fixation of the brain. The thorax was opened rapidly, the right auricle incised and approximately 20 ml of 4% paraformaldehyde containing 0.02% heparin injected into the left ventricle through a needle inserted into the apex of the heart. The brain was then removed and immersed in 10% buffered formalin for 4 days. Coronal sections were taken from six levels along the brain and were selected to ensure that a wide range of neuroanatomical structures was examined. The levels were, from rostral to caudal: through the cruciate sulcus down to a point midway along the olfactory tract ( level 1); centrally at a level through
ISSN 0031–3025 printed/ISSN 1465– 3931 online/01/030338 – 03 © 2001 Royal College of Pathologists of Australasia DOI:10.1080/00313020120062956
MOBILE TELEPHONES AND BRAIN VASCULAR LEAKAGE
339
TABLE 1 Albumin extravasation in the brains of mice exposed to GSM-like RF fields
Animal number Freely moving, cage control group 1 2 3 4 5 6 7 8 9 10
Number of blood vessels showing albumin leakage
Coronal section of brain involved
– – – – – 3 1 – – –
– – – – – 1*2*4* 1* – – – Fig. 1 Extravasated serum albumin ( arrows ) around leptomeningeal blood vessels in an exposed mouse ( H&E, original magnification, ´110).
Total 4 Sham-exposed group 1 2 3 4 5 6 7 8 9 10
– 1 1 – – 2 – – – 1
– 6* 6 – – 3*6* – – – 2*
Total 5 Exposed group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
3 1 – – – – – – 1 1 – 4 – 1 1 4 2 2 2 1 1 – 2 1 – 1 – Total 28
4*4*4* 5 – – – – – – 4* 6* – 2*4*6*6* – 6* 6* 2 2* 2* 4* 2*4* 2 6* 3*6* 3 6* – 2*6 5 – 3 –
clonal antibody at a dilution of 1: 20 000 and with biotinylated rabbit antigoat immunoglobulin ( Dako, Copenhagen, Denmark) at a 1: 500 dilution as the secondary antibody. Positive and negative controls were included in this protocol. Selected small areas containing albumin leakage were also removed from paraffin blocks, embedded in plastic, cut at 1 m m, and stained with toluidine blue; ultrathin sections were then cut and stained with uranyl acetate and lead citrate. The number of albumin extravasations for each group was then statistically compared using a Poisson regression analysis. This project ( No. 58/99 ) was approved by the Animal Ethics Committee of the Institute of Medical and Veterinary Science, Adelaide.
RESULTS There was no overall statistical difference ( P = 0.739 ) between the three groups with respect to the number of blood vessels showing albumin leakage. In all animals, extravasations of albumin in the neuropil were confined to a small perivascular area, while those in the leptomeninges sometimes spread for a short distance from the affected vessel. The total number of extravasations in each group was: four in freely moving, caged controls (2/10 mice); five in shamexposed controls ( 4/10 mice); and 28 in the exposed group ( 16/27 mice) (Table 1). Furthermore, the majority of the
* Denotes leptomeningeal albumin extravasation.
the mammillary body ( level 2); slightly caudal to the previous cut, to include the lateral geniculate body and pulvinar ( level 3); through the anterior colliculus near the roots of the oculomotor nerves ( level 4); through the posterior colliculus at the level of the pons ( level 5); and through the medulla oblongata to include the cerebellum ( level 6). The number and neuroanatomical location of blood vessels showing albumin leakage are shown in Table 1. Coronal slices were then embedded in paraffin wax and 6-m m sections cut and stained with H&E. The accurate histological detection of any extravasated albumin ( molecular weight 69 000) was performed using goat anti-rat albumin ( Cappel, Westchester, PA, USA) as the primary mono-
Fig. 2 Higher magnification of vessels shown in Fig. 1, with perivascular deposits of extravasated albumin ( A) and perfused vessel lumina ( L) ( original magnification, ´2000 ).
340
Pathology ( 2001 ), 33, August
FINNIE et al.
Fig. 3 Same area as Fig. 1, with leaked albumin ( arrows ) around the perfused vessel ( original magnification, ´1300 ).
extravasations were from leptomeningeal venules ( Fig. 1) ( 3/4 in cage controls, 4/5 in the sham-exposed group and 21/28 in the exposed group) and the extravascular location of the albumin was confirmed by electron microscopy ( Figs. 2 and 3). In positive control mice given the clostridial toxin, albumin extravasation was readily detectable, confirming that albumin immunohistochemistry is a sensitive method for detecting increased BBB permeability.
DISCUSSION The results of this study showed that there was no significant difference in albumin leakage from blood vessels in murine brains exposed to 4 W/kg of GSM-like RF fields for 1 hour compared with sham-exposed and freely moving, caged control animals. In all exposed and non-exposed groups, albumin extravasation occurred largely from leptomeningeal blood vessels which, together with those in the choroid plexus and circumventricular organs, have no recognised BBB.11 By contrast, albumin leakage was minimal from the vast majority of cerebral vessels that have a functional BBB. The BBB is a dynamic interface between blood and brain maintaining homeostasis in the central nervous system. The permeability properties of the BBB are those of the cerebral capillary endothelium, which differs in important structural detail from endothelia elsewhere. The effectiveness of the BBB depends largely on the presence of tight junctions between endothelial cells, but also to a paucity of micropinocytotic vesicles, a thicker than normal basement membrane, and no perivascular collagen. These endothelial characteristics are abetted by the inductive influence of astrocytic end-feet that cover most of the capillary surface and modulate the composition of the perineuronal fluid in a highly localised manner.11 In previous studies, rats were exposed to 900 MHz for 4 hours at SARs ranging from 0.3 to 7.5 W/kg.5 Serum albumin extravasation was only significant in the group exposed to the highest SAR of 7.5 W/kg, with minimal leakage found in freely moving and sham-exposed control groups. Moreover, albumin extravasations showed no topological predilection for any of the five coronal levels
studied and any albumin leakage appeared to be resolved in rats killed 7 days after exposure. These authors also examined the effect of GSM microwave exposure on the genomic response of the brain by measuring changes in the messenger RNAs of hsp70, the transcription factor genes c-fos and c-jun and the glial structural gene glial fibrillary acidic protein.6 They found that, while some minor stress response could be induced, it did not result in long-lasting or reactive changes in the brain. However, significant albumin leakage was found in rat brains exposed for 2 hours to both continuous and pulsed 915-MHz microwaves at SARs between 0.016 and 5 W/kg.3 Similarly, increased pinocytotic uptake of tracer ( rhodamine–ferritin complex) by cerebral cortical capillary endothelial cells was found in rats exposed for 30–120 minutes to 2450-MHz microwaves at relatively low levels ( 10 mW/cm2 , SAR ~ 2 W/kg, pulse width 10 m s at 100 pps).4 In summary, the present study showed that exposure of mice to a GSM-like RF field of 4 W/kg for 1 hour did not significantly disrupt BBB integrity when evaluated by albumin immunohistochemistry at the light microscope level and, in both exposed and non-exposed animals, albumin leakage occurred predominantly from non-BBB, leptomeningeal venules. AC KN OW LED GEM ENTS We thank Mrs Ruth Davies for preparation of histological sections. Address for correspondence: Professor P. C. Blumbergs, Neuropathology, Institute of Medical and Veterinary Science, Frome Road, Adelaide, SA 5000, Australia. Email: [email protected] v.au
References 1. Repacholi MH. Munich meeting report. Low-level exposure to radiofrequency electromagnetic fields: health effects and research needs. Bioelectromagnetics 1998; 19: 1–19. 2. European Commission. Possible health effects related to the use of radiotelephones. Proposal for a research program by a European Commission Expert Group. In: McKinlay AF, editor. Directorate General XIII, Telecommunications, Information Market and Exploitation of Research. Brussels: European Commission, 1996. 3. Salford LG, Brun A, Sturesson K, et al. Permeability of the blood– brain barrier induced by 915 MHz electromagnetic radiation, continuous wave and modulated at 8, 16, 50 and 200 Hz. Microsc Res Tech 1994; 27: 535–42. 4. Neubauer C, Phelan AM, Kues H, et al. Microwave irradiation of rats at 2.45 GHz activates pinocytotic-like uptake of tracer by capillary endothelial cells of cerebral cortex. Bioelectromagnetics 1990; 11: 261– 8. 5. Fritze K, Sommer C, Schmitz B, et al. Effect of global system for mobile communication ( GSM) microwave exposure on blood-brain barrier permeability in rat. Acta Neuropatho l 1997; 94: 465–70. 6. Fritze K, Wiessner C, Kuster N, et al. Effect of global system for mobile communication microwave exposure on the genomic response of the rat brain. Neuroscience 1997; 81: 627–39. 7. Repacholi MH, Basten A, Gebski V, et al. Lymphomas in Em -Pim 1 transgenic mice exposed to pulsed 900-MHz electromagnetic fields. Radiat Res 1997; 147: 631–40. 8. Balzano G, Chou C-K, Cicchetti R, et al. An efficient RF exposure system with precise SAR estimation for in vivo animal studies at 900 MHz. Microwave Theory Tech IEEE Trans ( in press ). 9. Finnie JW. Histopathological changes in the brain of mice given Clostridium perfringens type D epsilon toxin. J Comp Pathol 1984; 94: 363–70. 10. Finnie JW, Hajduk P. An immunohistochemical study of plasma albumin extravasation in the brain of mice after the administration of Clostridium perfringens type D epsilon toxin. Aust Vet J 1992; 69: 261– 2. 11. Miller D, Ironside JW. Raised intracranial pressure, oedema and hydrocephalus In: Graham DI, Lantos PL, editors. Greenfield’s Neuropathology. 6th ed. London: Arnold, 1997; 157–96.
EFFECT OF GLOBAL SYSTEM FOR MOBILE COMMUNICATION (GSM)-LIKE RADIOFREQUENCY FIELDS ON VASCULAR PERMEABILITY IN MOUSE BRAIN JOHN W. FINNIE *, PETER C. BLUM BERGS †, JIM MANAV IS †, TAM M Y D. UTTERIDGE *, VAL GEBSKI‡, JEFFREY G. SW IFT †, BARRIE VERNON -ROBERTS † AND TIM OTHY R. KUCHEL * *Veterinary Services Division and †Division of Tissue Pathology, Institute of Medical and Veterinary Science, Adelaide, SA, and ‡NHMRC Clinical Trials Centre, University of Sydney, Sydney, NSW, Australia
Summary The effect of global system for mobile communication (GSM) radiofrequency fields on vascular permeability in the brain was studied using a purpose-designed exposure system at 898.4 MHz. Mice (n = 30) were given a single far field, whole body exposure for 60 minutes at a specific absorption rate of 4 W/kg. Control mice were also sham-exposed (n = 10) or permitted free movement in a cage (n = 10) to exclude any stress-related effects. Vascular permeability changes were detected using albumin immunohistochemistry and the efficacy of this vascular tracer was confirmed with a positive control group exposed to a clostridial toxin known to increase vascular permeability in the brain. No significant difference in albumin extravasation was detected between any of the groups at the light microscope level using the albumin marker. Key words: Mobile telephone radiation, brain, mouse, vascular permeability, albumin immunohistochemistry. Abbreviations: BBB, blood– brain barrier; GSM, global system for mobile communication; RF, radiofrequency; SAR, specific absorption rate. Received 16 June 2000; revised 19 January 2001; accepted 29 January 2001
INTRODUCTION Although the use of mobile telephones is common and increasing, and requires these devices to be held close to the head, exposure effects on the brain remain controversial and prevent regulatory agencies giving an unqualified assurance of public safety.1 Furthermore, experimental findings have often been contradictory due to variable exposure conditions in diverse animal species and inadequate dosimetry.1,2 Most research has attempted to determine whether such exposure increases blood– brain barrier ( BBB) permeability. However, while some studies in the last decade in laboratory rodents have reported pathologically significant increases in BBB permeability,3,4 others5,6 have been unable to substantiate these findings. We employed purpose-designed exposure modules with precisely determined dosimetry to examine the effect of exposure to radiofrequency (RF) fields, similar to those used for global system for mobile (GSM) telecommunications, on tumour development. Since health problems associated with exposure to low level ( non-thermal ) RF
sources, such as mobile telephones, are poorly understood and probably subtle, we also used this facility to ascertain whether exposure from these devices had any adverse effect on the central nervous system.
MATERIALS AND METHODS In the present study, unanaesthetised, 8-week-old, female, inbred wild-type C57BL/6NTac mice ( Taconic Farms, New York, USA) were used. We were constrained in our choice of mice as only this wild-type and a transgenic strain ( Em -Pim 1) were permitted to occupy this facility to maintain the specific-pathogen-free status of the animals for the purpose of replicating a previous carcinogenesis study.7 Mice were exposed to RF fields with similar pulsing and modulation characteristics to those used for GSM mobile telecommunications ( 898.4-MHz fields modulated at a pulse repetition frequency of 217 Hz and a pulse width of 0.6 ms). The exposure system consisted of a cylindrical parallel plate with mice restrained in clear perspex tubes arranged radially around a dipole antenna.8 The cylinders were small enough to prevent the animals from changing their orientation relative to the field and ensure that the specific absorption rate ( SAR) could be precisely estimated. A forced air supply was also maintained through the cylinders to ensure no heat build-up and the Australian Radiation Protection and Nuclear Safety Agency ( ARPANSA) confirmed that the core body temperature did not rise using a Luxtron. All exposed mice received a single far field, whole body exposure of 4 W/kg in the same exposure system for 60 min; the field within the exposure system was homogeneous so all mice received the same exposure level. Both exposed ( n = 30 ) and sham-exposed ( n = 10 ) groups occupied this system, while another non-exposed control group ( n = 10) was allowed free movement in a cage without handling or further confinement to obviate any stress-related exposure module confinement effects. Three mice in the exposed group were discarded as perfusion fixation was considered suboptimal. A positive control group ( n = 10) was also included for the BBB marker ( albumin), mice being given Clostridium perfringens type D epsilon toxin that produces a rapid and severe generalised vasogenic cerebral oedema in mice.9 ,10 For dosimetry, measurements of RF energy absorption were performed by the ARPANSA ( Mr Michael Bangay) on mouse cadavers to characterise the live animal exposure. Mice were anaesthetised with 0.014 ml/g body weight of Avertin ( 2,2,2-tribromoethanol in tert-amyl alcohol; Aldrich, Milwaukee, WI, USA) immediately after termination of the 60-min exposure period and killed by transcardiac perfusion fixation of the brain. The thorax was opened rapidly, the right auricle incised and approximately 20 ml of 4% paraformaldehyde containing 0.02% heparin injected into the left ventricle through a needle inserted into the apex of the heart. The brain was then removed and immersed in 10% buffered formalin for 4 days. Coronal sections were taken from six levels along the brain and were selected to ensure that a wide range of neuroanatomical structures was examined. The levels were, from rostral to caudal: through the cruciate sulcus down to a point midway along the olfactory tract ( level 1); centrally at a level through
ISSN 0031–3025 printed/ISSN 1465– 3931 online/01/030338 – 03 © 2001 Royal College of Pathologists of Australasia DOI:10.1080/00313020120062956
MOBILE TELEPHONES AND BRAIN VASCULAR LEAKAGE
339
TABLE 1 Albumin extravasation in the brains of mice exposed to GSM-like RF fields
Animal number Freely moving, cage control group 1 2 3 4 5 6 7 8 9 10
Number of blood vessels showing albumin leakage
Coronal section of brain involved
– – – – – 3 1 – – –
– – – – – 1*2*4* 1* – – – Fig. 1 Extravasated serum albumin ( arrows ) around leptomeningeal blood vessels in an exposed mouse ( H&E, original magnification, ´110).
Total 4 Sham-exposed group 1 2 3 4 5 6 7 8 9 10
– 1 1 – – 2 – – – 1
– 6* 6 – – 3*6* – – – 2*
Total 5 Exposed group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27
3 1 – – – – – – 1 1 – 4 – 1 1 4 2 2 2 1 1 – 2 1 – 1 – Total 28
4*4*4* 5 – – – – – – 4* 6* – 2*4*6*6* – 6* 6* 2 2* 2* 4* 2*4* 2 6* 3*6* 3 6* – 2*6 5 – 3 –
clonal antibody at a dilution of 1: 20 000 and with biotinylated rabbit antigoat immunoglobulin ( Dako, Copenhagen, Denmark) at a 1: 500 dilution as the secondary antibody. Positive and negative controls were included in this protocol. Selected small areas containing albumin leakage were also removed from paraffin blocks, embedded in plastic, cut at 1 m m, and stained with toluidine blue; ultrathin sections were then cut and stained with uranyl acetate and lead citrate. The number of albumin extravasations for each group was then statistically compared using a Poisson regression analysis. This project ( No. 58/99 ) was approved by the Animal Ethics Committee of the Institute of Medical and Veterinary Science, Adelaide.
RESULTS There was no overall statistical difference ( P = 0.739 ) between the three groups with respect to the number of blood vessels showing albumin leakage. In all animals, extravasations of albumin in the neuropil were confined to a small perivascular area, while those in the leptomeninges sometimes spread for a short distance from the affected vessel. The total number of extravasations in each group was: four in freely moving, caged controls (2/10 mice); five in shamexposed controls ( 4/10 mice); and 28 in the exposed group ( 16/27 mice) (Table 1). Furthermore, the majority of the
* Denotes leptomeningeal albumin extravasation.
the mammillary body ( level 2); slightly caudal to the previous cut, to include the lateral geniculate body and pulvinar ( level 3); through the anterior colliculus near the roots of the oculomotor nerves ( level 4); through the posterior colliculus at the level of the pons ( level 5); and through the medulla oblongata to include the cerebellum ( level 6). The number and neuroanatomical location of blood vessels showing albumin leakage are shown in Table 1. Coronal slices were then embedded in paraffin wax and 6-m m sections cut and stained with H&E. The accurate histological detection of any extravasated albumin ( molecular weight 69 000) was performed using goat anti-rat albumin ( Cappel, Westchester, PA, USA) as the primary mono-
Fig. 2 Higher magnification of vessels shown in Fig. 1, with perivascular deposits of extravasated albumin ( A) and perfused vessel lumina ( L) ( original magnification, ´2000 ).
340
Pathology ( 2001 ), 33, August
FINNIE et al.
Fig. 3 Same area as Fig. 1, with leaked albumin ( arrows ) around the perfused vessel ( original magnification, ´1300 ).
extravasations were from leptomeningeal venules ( Fig. 1) ( 3/4 in cage controls, 4/5 in the sham-exposed group and 21/28 in the exposed group) and the extravascular location of the albumin was confirmed by electron microscopy ( Figs. 2 and 3). In positive control mice given the clostridial toxin, albumin extravasation was readily detectable, confirming that albumin immunohistochemistry is a sensitive method for detecting increased BBB permeability.
DISCUSSION The results of this study showed that there was no significant difference in albumin leakage from blood vessels in murine brains exposed to 4 W/kg of GSM-like RF fields for 1 hour compared with sham-exposed and freely moving, caged control animals. In all exposed and non-exposed groups, albumin extravasation occurred largely from leptomeningeal blood vessels which, together with those in the choroid plexus and circumventricular organs, have no recognised BBB.11 By contrast, albumin leakage was minimal from the vast majority of cerebral vessels that have a functional BBB. The BBB is a dynamic interface between blood and brain maintaining homeostasis in the central nervous system. The permeability properties of the BBB are those of the cerebral capillary endothelium, which differs in important structural detail from endothelia elsewhere. The effectiveness of the BBB depends largely on the presence of tight junctions between endothelial cells, but also to a paucity of micropinocytotic vesicles, a thicker than normal basement membrane, and no perivascular collagen. These endothelial characteristics are abetted by the inductive influence of astrocytic end-feet that cover most of the capillary surface and modulate the composition of the perineuronal fluid in a highly localised manner.11 In previous studies, rats were exposed to 900 MHz for 4 hours at SARs ranging from 0.3 to 7.5 W/kg.5 Serum albumin extravasation was only significant in the group exposed to the highest SAR of 7.5 W/kg, with minimal leakage found in freely moving and sham-exposed control groups. Moreover, albumin extravasations showed no topological predilection for any of the five coronal levels
studied and any albumin leakage appeared to be resolved in rats killed 7 days after exposure. These authors also examined the effect of GSM microwave exposure on the genomic response of the brain by measuring changes in the messenger RNAs of hsp70, the transcription factor genes c-fos and c-jun and the glial structural gene glial fibrillary acidic protein.6 They found that, while some minor stress response could be induced, it did not result in long-lasting or reactive changes in the brain. However, significant albumin leakage was found in rat brains exposed for 2 hours to both continuous and pulsed 915-MHz microwaves at SARs between 0.016 and 5 W/kg.3 Similarly, increased pinocytotic uptake of tracer ( rhodamine–ferritin complex) by cerebral cortical capillary endothelial cells was found in rats exposed for 30–120 minutes to 2450-MHz microwaves at relatively low levels ( 10 mW/cm2 , SAR ~ 2 W/kg, pulse width 10 m s at 100 pps).4 In summary, the present study showed that exposure of mice to a GSM-like RF field of 4 W/kg for 1 hour did not significantly disrupt BBB integrity when evaluated by albumin immunohistochemistry at the light microscope level and, in both exposed and non-exposed animals, albumin leakage occurred predominantly from non-BBB, leptomeningeal venules. AC KN OW LED GEM ENTS We thank Mrs Ruth Davies for preparation of histological sections. Address for correspondence: Professor P. C. Blumbergs, Neuropathology, Institute of Medical and Veterinary Science, Frome Road, Adelaide, SA 5000, Australia. Email: [email protected] v.au
References 1. Repacholi MH. Munich meeting report. Low-level exposure to radiofrequency electromagnetic fields: health effects and research needs. Bioelectromagnetics 1998; 19: 1–19. 2. European Commission. Possible health effects related to the use of radiotelephones. Proposal for a research program by a European Commission Expert Group. In: McKinlay AF, editor. Directorate General XIII, Telecommunications, Information Market and Exploitation of Research. Brussels: European Commission, 1996. 3. Salford LG, Brun A, Sturesson K, et al. Permeability of the blood– brain barrier induced by 915 MHz electromagnetic radiation, continuous wave and modulated at 8, 16, 50 and 200 Hz. Microsc Res Tech 1994; 27: 535–42. 4. Neubauer C, Phelan AM, Kues H, et al. Microwave irradiation of rats at 2.45 GHz activates pinocytotic-like uptake of tracer by capillary endothelial cells of cerebral cortex. Bioelectromagnetics 1990; 11: 261– 8. 5. Fritze K, Sommer C, Schmitz B, et al. Effect of global system for mobile communication ( GSM) microwave exposure on blood-brain barrier permeability in rat. Acta Neuropatho l 1997; 94: 465–70. 6. Fritze K, Wiessner C, Kuster N, et al. Effect of global system for mobile communication microwave exposure on the genomic response of the rat brain. Neuroscience 1997; 81: 627–39. 7. Repacholi MH, Basten A, Gebski V, et al. Lymphomas in Em -Pim 1 transgenic mice exposed to pulsed 900-MHz electromagnetic fields. Radiat Res 1997; 147: 631–40. 8. Balzano G, Chou C-K, Cicchetti R, et al. An efficient RF exposure system with precise SAR estimation for in vivo animal studies at 900 MHz. Microwave Theory Tech IEEE Trans ( in press ). 9. Finnie JW. Histopathological changes in the brain of mice given Clostridium perfringens type D epsilon toxin. J Comp Pathol 1984; 94: 363–70. 10. Finnie JW, Hajduk P. An immunohistochemical study of plasma albumin extravasation in the brain of mice after the administration of Clostridium perfringens type D epsilon toxin. Aust Vet J 1992; 69: 261– 2. 11. Miller D, Ironside JW. Raised intracranial pressure, oedema and hydrocephalus In: Graham DI, Lantos PL, editors. Greenfield’s Neuropathology. 6th ed. London: Arnold, 1997; 157–96.