Elevation of Transforming Growth Factor- Level in Cerebrospinal Fluid of Patients With Communicating Hydrocephalus After Subarachnoid Hemorrhage

Transcription

1 00 Elevation of Transforming Growth actor- Level in Cerebrospinal luid of Patients With Communicating Hydrocephalus After Subarachnoid Hemorrhage Kazuo Kitazawa, D; Tsuyoshi Tada, D Downloaded from by on January 9, 09 Background and Purpose Transforming growth factor-/ (TG-/) is a multifunctional polypeptide that controls the production of extracellular matrix protein. Platelets store a large quantity of TG-/, which is released at hemorrhage. We recently reported that human recombinant TG-/ induced communicating hydrocephalus in mice. The aim of this study was to determine whether TG-0 is related to the development of communicating hydrocephalus after subarachnoid hemorrhage (SAH). ethods TG-0 in the cerebrospinal fluid of patients with SAH was measured with enzyme-linked immunosorbent assay. The levels were compared between hydrocephalic and nonhydrocephalic groups. Western blot analysis was performed to determine active TG- in the cerebrospinal fluid. Results TG-/ rapidly decreased from the onset of SAH. The level of TG-/ of patients showing ventricular A pproximately one quarter of patients with sub- / arachnoid hemorrhage (SAH) suffer dementia, A. A. gait disturbance, and urinary incontinence within a few months of the ictus. Although the computed tomographic (CT) cisternogram of such symptomatic patients reveals ventricular dilatation and reflux of contrast material, their cerebrospinal fluid pressure is usually normal. In evaluating these patients, the concept of normal pressure hydrocephalus was first introduced by Hakim and Adams. Despite the report that fibrosis and arachnoid-pial adhesion in the subarachnoid space, especially around arachnoid villi, may disturb cerebrospinal fluid circulation, the generating mechanism has not been clarified. - Transforming growth factor-/ (TG-/) is a 5-kD multifunctional polypeptide that plays an important role in fibrosis of granulation tissue as seen in wound healing - 5 and in some fibrotic diseases such as liver cirrhosis, pulmonary fibrosis, 7 and glomemlonephritis. Recently we found that human recombinant TG-/ induced communicating hydrocephalus in mice by intrathecal injection. 9 It is important to determine whether the TG-/ is contained in active form in the cerebro- Received January,99;finalrevision received April,99; accepted April, 99. rom the Department of Neurosurgery, Shinshu University School of edicine, atsumoto, Japan. Reprint requests to Tsuyoshi Tada, D, Department of Neurosurgery, Shinshu University School of edicine, Asahi --, atsumoto 90, Japan. 99 American Heart Association, Inc. dilatation with periventricular low density on computed tomographic scan was.07±0.7 ng/ml on days through, which was significantly higher than 0.5±0. ng/ml in patients without ventricular dilatation (P<.0). urthermore, the TG-/J level of patients who had undergone ventriculoperitoneal shunt (n=ll) was l.ll±0.09 ng/ml on days through, which was also higher than the level of the nonshunt group (n=) ( ng/ml; P<.0). A 5-kD band was demonstrated by Western blot analysis in the cerebrospinal fluid of a patient with SAH. Conclusions Our results strongly suggest that TG-0 plays an important role in generating communicating hydrocephalus after SAH. (Stroke. 995:00-0.) Key Words cerebrospinal fluid hydrocephalus subarachnoid hemorrhage transforming growth factors spinal fluid and develops communicating hydrocephalus in patients with SAH. In this study we measured the TG-0 levels in the cerebrospinal fluid of patients with SAH in relation to the CT findings and the clinical course. We also studied by Western blot analysis whether TG- exists in active form in cerebrospinal fluid. Subjects and ethods Patient Selection The Table shows clinical data of patients with SAH. Patients ranged in age from 9 to 7 years (mean,. years) and included 9 men and 5 women. SAH was caused by ruptured cerebral aneurysms in all patients except, in whom it was of unknown origin. Twenty-three patients were subjected to clipping surgery for the ruptured aneurysms and placement of cisternal drainage. The Hounsfield numbers of basal cisternal clots at the onset of SAH ranged from to (mean, 5.+0.). According to isher's classification, 0 5 patients were group, were group, and were group on the initial CT scan. Thirteen of the patients developed ventricular dilatation with periventricular low density on the follow-up CT scan taken between 0 and 0 ; of the patients exhibited dementia, urinary incontinence, and gait disturbance. The patients were diagnosed as having communicating hydrocephalus and underwent ventriculoperitoneal (VP) shunt between days and. Our criteria for selection of VP shunt were basically the same as those commonly used, which include ventricular dilatation and periventricular low density on the CT scan and characteristic symptoms such as dementia, urinary incontinence, and gait disturbance.

2 Kitazawa and Tada Elevation of TG-0 in Patients With Hydrocephaius After SAH 0 Clinical Summary of Patients With Subarachnold Hemorrhage Pt Age, y Sex Rsher CTNo. PVL VP Shunt Result of Shunt Downloaded from by on January 9, Pt indicates patient; Rsher, Rsher's classification on initial computed tomographic (CT) scan; CT No., Hounsfield number of basal cistemal clots on initial CT scan; PVL, periventricular low density area on follow-up CT scan; and VP, ventriculoperitoneal. Cerebrospinal luid Eighty-three samples of cerebrospinal fluid from the patients were obtained through cisternal drainage from day 0 to day 7 after onset. In patients without cisternal drainage or after removal of the drain, the fluid was obtained by lumbar puncture. The collected samples were centrifuged at 000 rpm for 0 minutes. The supernatant was passed through a 0.-/im filter and stored at -0 C. TG-/ Assay To activate TG-0, all the samples were dialyzed against IN acetic acid and then against phosphate-buffered saline (ph 7.) two times. The concentration of TG-0 was measured with a TG-0 enzyme-linked immunosorbent assay kit (King Jozo Corp). Western Blot Analysis Three samples of cerebrospinal fluid were collected from a patient at days 0,, and after SAH. They were used for a Western blot analysis without acidification. Sample buffer including -mercaptoethanol was added to an equal volume of the sample and heated for 5 minutes with boiling water. The samples were run on a 0% sodium dodecyl sulfate-polyacrylamide gel for detection of TG-/. The gel was electroblotted onto a nitrocellulose membrane (Immobilon-P, ilipore). The membrane was incubated with tris(hydroxymethyl)aminomethane-hcl buffer (ph 7.) containing bovine serum albumin, followed by incubation with polyclonal anti-tg-0 antibody (R&D). Then the membrane was incubated with biotin-labeled horse anti-rabbit IgG incubated with alkaline phosphatase-conjugated avidin. inally, color reagents were added to the membrane to develop a colorimetric reaction. We used.5 ng of human recombinant TG- as a control. Statistical Analysis The Welch t test was used to determine statistical significance; P<..05 was considered significant. Results TG-0 in Cerebrospinal luid Relation to Total Protein Total protein levels in cerebrospinal fluid within 7 ranged from to 7 mg/dl. Although there was no relation between total protein concentration and, TG-/ and total protein levels were well related (r=.o, /><.00) (ig ). Relation to Initial CT indings To determine whether the volume of SAH is related to TG-/ content in the cerebrospinal fluid, we divided the patients into two groups according to the Hounsfield number of the basal cistern on the initial CT scan and compared the time course of TG-/ content between the two groups. As a result, the TG-/ concentration of patients with higher density (Hounsfield number 55) of the basal cisternal clots on the

3 0 Stroke Vol 5, No 7 July 99 ng/ml..5 =.00* 59, r..5 **..- ' ' ' - " *. n= Group Group.5..5 TG- 0 k n=7! n=ll ss n=!.5 n= 0 r I n= n= mg/dl TP ra. Line graph shows relation between total protein (TP) and transforming growth factor-/ (TG-0 [TG-0]) concentration in cerebrospinal fluid of patients with subarachnoid hemorrhage. Twenty-six cerebrospinal fluid samples were obtained from patients. Shown are 95% confidence bands for the true mean of y. TG-0 and total protein were well related (r=.o, P<.00). initial CT scan was not different from that of those with lower density (Hounsfield number <55) (ig ). We also compared the TG-/ content between different combinations of isher's groups. TG-/ level was not related to isher's classification in any combination. ig shows the comparison of TG-/ levels between the patients of isher's group and group. Downloaded from by on January 9, 09 Relation to Ventricular Dilatation Thirteen of patients in our series developed ventricular dilatation with periventricular low density. TG-0 content in the cerebrospinal fluid of the patients was compared with TG-/ content in those without ventricular dilatation. The mean±sd values of TG-/ of the two groups on days through 5 were.9±0. ng/ml and.±.0 ng/ml, respectively. These values rapidly decreased to 0.±0.7 ng/ml and 0.±0. ng/ml, respectively, on days through 9. During this period there was no statistical difference between the two groups. TG-/ levels in patients with ventricular dilatation on days 9 through, through, and 5 through 7 were.0±0. ng/ml,.07±0.7 ng/ml, and.05 ±0. ng/ml, respectively; TG-/ IG. Line graph shows time course of transforming growth factor-/ (TG-/) concentration in cerebrospinal fluid of patients in isher's group and group. There was no statistical difference between them. SAH indicates subarachnoid hemorrhage. levels in patients without ventricular dilatation were ng/ml, 0.5±0. ng/ml, and ng/ml, respectively. The TG-/9 level on days 9 through 7 of the patients with ventricular dilatation stayed higher than that of the group without ventricular dilatation; the difference was significant on days through (P<.0) (ig ). Relation to Clinical Symptoms To determine whether TG-/9 in the cerebrospinal fluid of patients with communicating hydrocephalus is higher than that of those without hydrocephalus, TG-/J levels in cerebrospinal fluid of patients who received VP shunt operation and those who did not were compared. In the VP shunt gtoup, improvements in clinical symptoms were seen in all patients. TG-/ levels of the VP shunt group versus the non-vp shunt group were.9±0. ng/ml versus.+0.9 ng/ml on days through 5 and 0.±0.9 ng/ml versus ng/ml on days through, respectively. The patterns of rapidly decreasing TG-0 of both groups were similar, and there was no statistical difference between them during this period. However, the TG-/ levels of the VP shunt group on days 9 through 7 were significantly higher than those of the non-vp shunt Ventricular dilalation (+) a Ventricular dilatation (-) n=7 Hounsfield number< 55 ohounslield numbers 55 I n= v n=5 n= n= ~_ "" "=' n= 0 i: IG. Une graph shows time course of transforming growth factor-/ (TG-0) concentration in cerebrospinal fluid of patients with subarachnoid hemorrhage (SAH) in relation to Hounsfield number of the basal cistern on initial computed tomographic scan. There was no difference in TG-/ content between patients with higher density (Hounsfield number and those of lower density (Hounsfield number <55). IG. Line graph shows time course of transforming growth factor-p (TG-0) concentration in cerebrospinal fluid in relation to ventricular dilatation on follow-up computed tomographic scans. The level of TG-/ was highest on the day of onset of subarachnoid hemorrhage (SAH). The level of TG-0 of patients with ventricular dilatation stayed higher than that of groups without ventricular dilatation; the difference was significant on days through (P<.0).

4 Kitazawa and Tada Elevation of TG-0 in Patients With Hydrocephalus After SAH V- P shunl (+) DV Pshunl (-) =0 =..5 n=5 p<0.0.5' n= p< n= 0 Javs after SAH IG 5. Une graph shows time course of transforming growth factor-0 (TG-/9) concentration in cerebrospinal fluid of patients with and without ventriculoperttoneai (VP [V-P]) shunt. There was no statistical difference between them on days through. However, the TG-0 levels of the VP shunt group were statistically higher than those of the group without VP shunt on days 9 through 7. group:.7+0. ng/ml versus 0.7±0. ng/ml (P<.0) on days 9 through, l.ll±0.09 ng/ml versus 0.5±0. ng/ml (P<.0) on days through, and.7±0. ng/ml versus 0.0±0.0 ng/ml (P<.05) on days 5 through 7 (ig 5). Active orm of TG-fil in Cerebrospinal luid Western blot analysis indicated that a 5-kD band of the active form of TG- was recognized in the day 0 sample (ig ). Downloaded from by on January 9, 09 Discussion Communicating hydrocephalus is now a well-recognized complication of SAH. any authors suspect that fibrosis and arachnoid-pial adhesion in the subarach ra. Western blot analysis for the expression of transforming growth factor-0 (TG-0). Samples of cerebrospinal fluid were obtained from a patient with subarachnoid hemorrhage at 0,, and days after the ictus; the results shown as lane,, and, respectively. Lane indicates human recombinant TG-0 as a control. Values on the left are the molecular masses (in kilodaltons) of the standards. Arrows denote.5-kd and 5-kD bands (active form of TG-0). 0 noid space, especially arachnoid villi, may disturb cerebrospinal fluid absorption.- Thickened arachnoid and proliferative fibrosis in the subarachnoid space were also demonstrated by scanning electron microscopic observation. However, the exact generating mechanism has not been clarified. TG- was originally discovered as a factor to support the anchorage-independent growth of normal rat kidney cell. Several studies revealed that it is a multifunctional factor that not only facilitates but also suppresses many functions of various cells. 7 TG-/, which is chemotactic to monocytes and fibroblasts, enhances angiogenesis and the formation of extracellular matrix protein, such as collagen and fibronectin.7 Because we thought that it would be possible to explain the fibrosis of subarachnoid space after SAH by this TG-/ function, we injected human recombinant TG-/ into the subarachnoid space of mice and found that it induced a slowly progressive communicating hydrocephalus.9 TG-/ indicates a marked conservation across species, and human TG-/ is a highsequence homologue to murine TG-/. In this clinical study we measured the total TG-/ levels of the cerebrospinal fluid obtained from patients with SAH and compared them in relation to CT findings and clinical courses. The TG-/ levels of the patients who developed ventricular dilatation with periventricular low density on the follow-up CT scan were higher than those of the patients without ventricular dilatation. urthermore, the TG-/ concentration of patients who underwent VP shunt surgery was also statistically higher than that of those without VP shunt. Thus, not only the CT findings but also the clinical findings were well related to the TG-/ levels of the cerebrospinal fluid. These results indicate that TG-/ concentration in the cerebrospinal fluid of patients with communicating hydrocephalus is higher than that of patients without hydrocephalus. Source of TG-0 in Cerebrospinal luid We previously demonstrated that TG-/ was absent in the cerebrospinal fluid under normal conditions.9 In the present clinical study the level of TG-/J in the cerebrospinal fluid was highest in the early stage of SAH and rapidly decreased; it correlated with the total protein level of the cerebrospinal fluid. These findings suggest the hypothesis that TG-/ is mainly derived from platelets that spread into the subarachnoid space at the time of SAH. However, TG-/ is produced by many kinds of cells, including fibroblasts. It is also reported that many cytokines would be released into the cerebrospinal fluid from inflammatory cells.0 It seems likely that not only platelets but also fibroblasts and inflammatory cells may release TG-0 into the cerebrospinal fluid of patients with SAH. Active orm of TG-0 in Cerebrospinal luid Platelets release a latent form of TG-/ at hemorrhage. The activation mechanism of the latent form of TG-/ derived from platelets has been gradually resolved, but the activation mechanism in cerebrospinal fluid has not been clarified. In this study we measured the total TG-/ level in cerebrospinal fluid that had been pretreated with acidification and found that the concentration of TG-/ in the hydrocephalus group

5 0 Stroke Vol 5, No 7 July 99 Downloaded from by on January 9, 09 was higher than that in the group without hydrocephalus. Because the latent form of TG-/ cannot bind its receptor until the binding protein is removed, it was necessary to demonstrate the active form of TG-/ in the cerebrospinal fluid of SAH patients. TG-/ is known to be active as a.5-kd monomer or a 5-kD dimer that is linked by disulfide bonds of.5-kd monomers. Western blot analysis of the cerebrospinal fluid of a patient with SAH, which had not been pretreated with acid, revealed a 5-kD band on day 0. Positive demonstration of the active form in a sample of cerebrospinal fluid is significant. Although active TG-/ could not be detected in the samples on days and by Western blot analysis, it is possible that a small amount of active TG-/ exists in the cerebrospinal fluid. Conclusions We compared the TG-/ concentration in the cerebrospinal fluid of patients with SAH between hydrocephalic and nonhydrocephalic groups. The TG-/ level of the former (n=ll) was statistically higher than that of the latter (n=). We also demonstrated the active form of TG-/ in the cerebrospinal fluid by Western blot analysis. These results provide additional support to the theory that TG-/ plays an important role in generating communicating hydrocephalus after SAH. Acknowledgment We thank Professor S. Kobayashi for revising and discussing the manuscript. References. Hakim S, Adams RD. The special clinical problem of symptomatic hydrocephalus with normal cerebrospinal fluid pressure: observations on cerebrospinal fluid hydrodynamics. J Neurol Sci. 95; : Ellington E, orgolis G. Block of arachnoid villus by subarachnoid hemorrhage. J Neurosurg. 977;0: Suzuki S, Asia, Ottomo, Iwabuchi T. Change in the subarachnoid space after experimental subarachnoid hemorrhage in the dog: scanning electron microscopic observation. Ada Neurochir (Wien). 977;9:-.. Bennett NT, Schults GS. Growth factors and wound healing: biochemical properties of growth factors and their receptors. Am J Surg. 99;5: Quaglino D Jr, Nanney LB, Ditesheim JA, Davidson J. Transforming growth factor-/? stimulates wound healing and modulates extracellular matrix gene expression in pig skin: incisional wound model. J Invest Dermatol. 99;97:-.. Castilla A, Prieto J, austo N. Transforming growth factors / and alfa in chronic liver disease: effect of interferon alfa therapy. NEngU ed. 99;: Khalil N, O'Connor RN, Unruh HW, Warren PW, landers KC, Kemp A, Bereznay OH, Greenberg AH. Increased production and immunohistochemical localization of transforming growth factor-/? in idiopathic pulmonary fibrosis. Am J Respir Cell ol Biol. 99; 5:55-.. Border WA, Ruoslahti E. Transforming growth factor-/ induces extracellular matrix formation in glomerulonephritis. Cell Differ Dev. 990;: Tada T, Kanaji, Kobayashi S. Induction of communicating hydrocephalus in mice by intrathecal injection of human recombinant transforming growth factor-/?l. J Neuroimmunology. 99; 50: isher C, Kistler JP, Davis J. Relation of cerebral vasospasm to subarachnoid hemorrhage visualized by computerized tomographic scanning. Neurosurgery. 90;:l-9.. Hammes E. Reaction of the meninges to blood. Arch Neurol Psychiatry. 9;5: Roberts AB, Anzano A, Lamb LC, Smith J, Sporn B. New class of transforming growth factors potentiated by epidermal growth factor: isolation from non-neoplastic tissues. Proc Natl Acad Sci USA. 9;7: Ignotz RA, assague J. Transforming growth factor-/ stimulates the expression of fibronectin and collagen and their incorporation into the extracellular matrix. J Biol Chem. 9;:7-5.. Penttinen RP, Kobayashi S, Bornstein P. Transforming growth factor-/? increases mrna for matrix proteins both in the presence and in the absence of changes in mrna stability. Proc Natl Acad Sci USA. 9;5: Roberts AB, Sporn B, Assoian RK, Smith J, Roche NS, Wakefield L, Heine UI, Liotta LA, alanga V, Kehrl H, et al. Transforming growth factor type /?: rapid induction of fibrosis and angiogenesis in vivo and stimulation of collagen formation in vitro. Proc Natl Acad Sci USA. 9;:7-7.. Roberts AB, Thompson NL, Heine U, landers C, Sporn B. Transforming growth factor type /?: possible roles in carcinogenesis. Br J Cancer. 9;57: Sporn B, Roberts AB. Transforming growth factor-beta: recent progress and new challenges. J Cell Biol. 99;9: Derynck R, Jarrett JA, Chen EY, Goeddel DV. The murine transforming growth factor-/!) precursor. J Biol Chem. 9;: Tada T, Yabu K, Kobayashi S. Detection of active form of transforming growth factor-/? in cerebrospinal fluid of patients with glioma. Jpn J Cancer Res. 99;: athiesen T, Anderson B, Loftenius A, von Hoist H. Increased interleukin- levels in cerebrospinal fluid following subarachnoid hemorrhage. J Neurosurg. 99;7: Wakefield L, Smith D, landers KC, Sporn B. Latent transforming growth factor-/? from human platelets: a high molecular weight complex containing precursor sequences. J Biol Chem. 9; : Schultz-Cherry S, urphy-ullrich JE. Thrombospondin causes activation of latent transforming growth factor-/? secreted by endothelial cells by a novel mechanism. J Cell Biol. 99;:9-9.