- Email: [email protected]
Molecular size distribution of somatostatin-like immunoreactivity in the cerebrospinal fluid of patients with degenerative brain disease
Neuroscience Research, 3 (1986) 213-225
213
Elsevier Scientific Publishers Ireland Ltd. NSR 00108
Molecular Size Distribution of Somatostatin-Like Immunoreactivity in the Cerebrospinal Fluid of Patients with Degenerative Brain Disease* Klaus Rissler ~, Hinrich Cramer 1, Dieter Schaudt 2, Denise Strubel 2 and Wagner F. Gattaz 3 IDepartment of Neurology and Clinical Neurophysiology, Universityof Freiburg, D-7800 Freiburg (F.R.G.); 2Department of Geriatrics, University of Strasbourg, Strasbourg (France); and 3Central Institute of Mental Health, D-6800 Mannheim (F.R.G.) (Received July 19th, 1985; Revised version received September 30th, 1985; Accepted October 2nd, 1985)
Key words."somatostatin - - molecular size distribution-- cerebrospinal fluid-- dementia - - hydrocephalus SUMMARY The molecular size distribution of somatostatin-like immunoreactivity (SLI) in the cerebrospinal fluid (CSF) of patients with brain disease was investigated by separation with a Sephadex G-25 superfine column and subsequent radioimmunoassay of the eluate. Marked heterogeneity of SLI in the CSF of control subjects as well as in demented patients, was observed. Controls and schizophrenics exhibited an SLI distribution pattern consisting mainly of two pronounced peaks: the first eluting with the void volume of the column; the second being compatible with a peptide of N-terminally extended somatostatin- 14. SLI from the CSF of patients with senile dementia of the Alzheimer type (SDAT), multi-infarct dementia (MID) and normal pressure hydrocephalus (NPH) showed the same two peaks found in controls and schizophrenics; and in addition, a third peak co-eluting with somatostatin-14. However, this peak was more pronounced in patients with SDAT and M I D than in patients with NPH. Re-chromatography of G-25 sf void volume immunoreactivity afforded two fractions of an apparent molecular weight of about 10,000 daltons and 15,500 daltons, respectively.
INTRODUCTION
During the course of efforts to determine the distribution of the growth hormonereleasing factor in rat hypothalamus, a substance that inhibited growth hormone release * This paper is dedicated to Professor Georg Hertting on the occasion of his 60th birthday. Correspondence: K. Rissler, Department of Neurology and Clinical Neurophysioiogy, University of Freiburg, Hansastr. 9, D-7800 Freiburg, F.R.G. 0168-0102/86/$03.50 © 1986 Elsevier Scientific Publishers Ireland Ltd.
214 was detected by Krulich et al. 24. In 1972, Brazeau et al. v succeeded in isolating from ovine hypothalamus a cyclic tetradecapeptide that inhibited growth hormone secretion, and which was shown not to be limited to the hypothalamus but to have a widespread distribution throughout the central nervous system (CNS) 8'9'20'30"44. Besides its inhibitory effect on the release of growth hormone from the anterior pituitary, somatostatin is now believed to function as a neurotransmitter or a neuromodulator. The peptide co-exists with other neurotransmitters in nerve endings from which it is released in response to depolarizing stimuli 35. Subsequent work has considerably expanded the initial concept of tetradecapeptide somatostatin's main function, being the regulation of growth hormone release. It is now obvious that somatostatin-related peptides represent a whole family which includes the originally identified peptide somatostatin-14 (SRIF), somatostatin-283a, somatostatin256 as well as immunoreactivity related to the somatostatin-28 l_~2-sequence 5 and also the existence of precursor-like peptides of high molecular weight. It has been shown that these putative precursors could be converted into the circulating biologically active peptides by a trypsin-like enzymatic cleavage47"4s. Recently a specific enzyme converthag somatostatin-28 into somatostatin-14 and somatostatin-281_~2 designated as somatostatin-28 convertase has been discovered TM. In 1977, Patel et al.31 suggested that cerebrospinal fluid (CSF) levels of somatostatin-like immunoreactivity (SLI) might be a useful index for CNS damage when they determined CSF levels of SLI in neurological patients. In the following years other reports concerning CSF-SLI levels in patients with Parkinson's disease ~°'~4, senile dementia of Alzheimer type 11'28"4°'46, Huntington's chorea 12, psychiatric and affective disorders 17,39, neurological disease2z, multiple sclerosis41 and acromegaly45 have been published. Kronheim et al. 23 performed the In'st study of the molecular size of SLI in the CSF of normal subjects. This was followed by investigations by Penman et al.32 and Reichlin et 81.36. As far as we know, molecular size investigations of SLI in the CSF of neurological patients have not yet been reported. We have therefore studied the molecular size of SLI in the CSF of patients with degenerative brain disease 3s. In this paper we present findings obtained from patients with senile dementia of the Alzbeimer type (SDAT), with multi-infarct dementia (MID), with normal pressure hydrocephalus (NPH) and with schizophrenia. We also present findings obtained from age-matched control patients.
MATERIALS AND METHODS
Synthetic samples of tyrosine°-somatostatin-14, cyclic somatostatin-14 and somatostatin-28 were kindly donated by Serono (Freiburg, F.R.G.), Somatostatin-25 was purchased from Peninsula Laboratories (Belmont, CA, U.S,A.). For radioiodination a
215 solution of sodium iodide-125 carrier-free in 0.1 M NaOH (Amersham-Buchler, Braunschweig, F.R.G.) was used. Sephadex G-25 fine/superfine and Sephadex G-50 fine were purchased from Pharmacia (Uppsala, Sweden). Sep-Pak-cartridges containing octadecasilyl-silicagel (ODS-silica) were manufactured by Millipore-Waters Associates (Milford, MA, U.S.A.). The antiserum K-18 (Ferring, Kiel, F.R.G.) which recognizes somatostatin at the ring structure was raised in rabbits. It shows 100% cross-reaction with cyclic somatostatin-14, somatostatin-28 and somatostatin-25 and no cross-reaction with a number of central nervous system and gastrointestinal peptides including human growth hormone (hGH), cholecystokinin, bombesin, substance-P, vasointestinal polypeptide (VIP), thyrotropin releasing hormone (TRH), adrenocorticotropic hormone (ACTH), gastric inhibitory polypeptide, insulin, glucagon, gastrin, calcitonin and parathormone. Myoglobin, cytochrome-C and aprotinin were obtained from Serva (Heidelberg, F.R.G.). Charcoal (type "Norit-A") and dextran ("Dextran T70") were produced by Serva (Heidelberg, F.R.G.) and Pharmacia (Uppsala, Sweden), respectively. The preparation of the lZSI-labelled tyrosine°-somatostatin- 14 was performed by use of the chloramine-T method In'st applied by Greenwood et al.19 for iodination of human growth hormone. Purification of the labelled antigen was carried out by application of a two-step procedure: first, by adsorption on ODS-silica 27 and subsequent elution with acetone/water/acetic acid (39:40: 1); and second, by gel permeation chromatography on a 90 x 1 cm Sephadex G-25 free column with 0.1 M acetic acid containing 0.05 % bovine serum albumin. Appropriate fractions were pooled and stored in aliquots of about 5-10 #Ci at - 30 °C. A tracer prepared in this manner was storable for at least 3 months. Radioimmunoassay was performed according to a modified procedure of Arimura et al.2 using cyclic somatostatin (SRIF) as the standard. For this purpose the antiserum (Final dilution 1 : 1,400,000), either standard or sample, was incubated in a final volume of 0.7 ml for about 12-16 h at 4 °C. The buffer system consisted of 0.1 M sodium phosphate, pH 7.5, containing 0.025 M EDTA, 0.15 M NaC1, 0.005 M (0.02 %) sodium azide, and 0.25~ BSA. 125I-labelled antigen was added (approximately 10,000 cpm) and incubation was continued for another 48 h at 4 ° C. Separation of the free from the antibody-bound antigen was carried out with dextran-coated charcoal, i.e. a suspension of 250 mg charcoal and 25 mg of dextran per 100 ml of incubation buffer which lacked BSA was used. One ml of the suspension was added to each tube, and the mixture was incubated for 30 min at 4 °C. After centrifugation the supernatants were collected separately and the radioactivity was determined. All measurements (standards and samples) were done in duplicate. Denaturation of the tracer during incubation, as estimated by omitting the antibody, was negligible ( < 2~). The sensitivity of the assay, defined as that concentration of synthetic SRIF necessary to inhibit the binding of [ 125I]Tyr°-somatostatin-14 to a mean value at least two standard deviations below the mean value obtained in the absence of SRIF, was about 10 pg/ml. The intra-assay and inter-assay coefficients of variation were 5% and 129/o, respectively.
216 For peptide characterization, an appropriate sample of cerebrospinal fluid from several patients of one group, containing about 500-1000 pg/ml total somatostatin-like immunoreactivity (SLI), was concentrated and thus prepurified by adsorption on ODSsilica4 and consecutively eluted with 2-3 ml of 90% aqueous acetic acid. After lyophilization, the residual concentrate was subjected to chromatography on a 40 × 1 cm Sephadex G-25 superfine column using the same eluent as for tracer purification. Twenty-five fractions, each containing 2 ml of eluate, were collected. The fractions were lyophilized and reconstituted in assay buffer for subsequent radioimmunoassay. [ ~zSI]Tyr°-somatostatin- 14 treated in this way showed 80 ~o recovery of total radioactivity. The radioactive material exhibited more than 80 ~o of the maximal antibody-binding when compared to the untreated peptide and was displaceable by cycfic somatostatin-14 in a concentration-dependent manner. The assay sensitivity obtained by use of this tracer was equal to that obtained with the untreated radiopeptide. The recovery values for the equally treated synthetic samples of cyclic somatostatin-14 and somatostatin-28 were 90yo and 60yo, respectively, and no immunoreactive material other than the reference peptides could be detected. This finding suggests that the peptide material did not undergo marked chemical modification. Analogously treated samples of native CSF showed 60yo recovery of SLI. The recovery of synthetic cyclic somatostatin-14 and somatostatin-28 from the Sephadex column was complete. Evaluation of the apparent molecular weights of Sephadex G-25 superfine void volume SLI (exclusion limit 5000 daltons) was performed by column chromatography of the corresponding fractions on Sephadex G-50 t'me (84 x 1 era) using a set of appropriate molecular weight markers (myoglobin 18000 daltons; cytochrome-C 13,000 daltons; aprotinin 6500 daltons; somatostatin-28 3200 daltons; somatostatin- 14 1600 daltons). Prior to chromatography, the CSF void volume samples were incubated for 24 h with 8 M urea containing 0.05 M phosphate buffer (pH 7.5). For proof of immunological identity between synthetic SRIF and SLI measured in CSF, a dilution series of a concentrated CSF sample was prepared and then compared
95.
80-
70" "~ 60E -~ $0" /*0" °7 30" ~ :ZO" N
rif
40
J 1:8
80 1 :/~
160 I 1:2
320 Srif{po/mI)
,: "' 1:1 CSF-OlLUTION
Fig. 1. Logit-log plot of serial dilutions of synthetic SRIF and SLI obtained from human CSF showing parallelism of both c u r v e s .
217 to a corresponding dilution curve of synthetic SRIF. Both curves exhibited parallelism and therefore immunological identity (Fig. 1). Patients
All patients examined in this study were hospitalized in the Geriatric Department, Strasbourg, and the Psychiatric Department, Mannheim. The high age control group (n = 13, mean age 83.5 + 5 years) consisted of patients without central nervous disease, intellectual deterioration or major psychiatric disturbances. The only pathology of these patients consisted of either mild peripheral nervous disease (diabetic neuropathy, plexus neuritis), a slight degree of depression, skeletal diseases (arthrosis and bone fractures) or moderate cardiorespiratory insufficiency. Lumbar punctures were performed either for diagnostic purpose in order to exclude central nervous disease, or when informed consent was obtained after the nature of the study had been fully explained. The diagnosis of patients with senile dementia of the Alzheimer type (n = 32, mean age 84 + 7 years) was based on clinical criteria. All these patients had progressive deterioration of intellectual functions including language, memory and personality. Symptomatic causes of dementia were excluded by the absence of strokes, focalized neurological deficits, by laboratory examinations including EEG and vascular status. The intellectual status of the patients and the degree of deterioration were estimated by a neuropsychological examination based on the Brief Cognitive Rating Scale (BCRS) of Reisberg 37. The group of patients with vascular disease (n = 17, mean age 79 + 7 years) was characterized by a history of strokes and the presence of hemiplegia, unilateral or bilateral Babinski signs and the presence of various degrees of dementia of the multi-infarct type. In all these patients the Hachinski ischemic score was above 12. The diagnosis of patients having normal pressure hydrocephalus (n = 5, mean age 71 + 6 years) was established by the progressive appearance of mild mental deterioration, gait disturbance and loss of sphincter control. In these patients, the diagnosis was confu'med by cranial computerized scans which showed marked ventricular dilatation and the absence of or only a moderate degree of cortical atrophy. The group of paranoid schizophrenics (n = 14, mean age 36 + 10 years) was examined according to the Research Diagnostic Criteria of Spitzer et al.43. Informed consent was obtained from all patients or from their closest relations after the nature of the study had been fully explained. Almost all geriatric patients received minor drug treatment, at least for other than neurological disease (cardio-vascular, diabetes, rheumatism). Agitated demented patients were treated with low doses ofhaloperidol or thioridazine. A number of patients received low doses of benzodiazepines, and four control subjects and two vascular disease patients were treated with antidepressants (amitryptiline, clomipramine). The paranoid schizophrenic patients received no drug treatment for at least 3 weeks prior to the lumbar puncture. CSF was obtained with patients in a sitting position between 09.00 and 10.00 h after
218 the patients had fasted for 12 h and had bed rest for 10 h. The samples were immediately congelated and stored at - 4 0 °C until analysis. Routine laboratory analysis, including albumin, glucose and cell count was performed on CSF from all subjects. Values were within the normal range.
RESULTS
Molecular size heterog~aeity of somatostatin-like immunoreactivity (SLI) was observed in the cerebrospinal fluid (CSF) of all the groups of patients examined. Two prominent peaks subsequently designated peaks I and II were shown by all CSF pools. Peak I eluted with the void volume of the Sephadex G-25 superfine column (G-25 sf), whereas peak II appeared between somatostatin-25 (SST-25) and synthetic somatostatin-14 (SRIF). No discrete peak attributable to SST-25 or to somatostatin-28 (SST-28) could be detected. However, the resolution between peak I and peak II is not satisfactory, so it is possible that significant amounts of SLI corresponding to SST-25 or SST-28 had been overlapped by adjacent peaks. In senile control patients, SLI consisted of peak I and peak II representing approximately 90% of the total SLI and of a smaller peak III of immunoreactivity co-eluting with synthetic SRIF (approx. 10%) (Fig. 2). The ratio of the main peaks I and II showed considerable variation between different CSF pools ranging from 85 9o peak I and 15% peak II to 92% peak II and 89o peak I, respectively. CSF from paranoid schizophremics revealed a distribution pattern similar to that of senile controls; however, SLI co-eluting with SRIF was virtually lacking (Fig, 3). Patients with senile dementia of the Alzheimer type (SDAT) also yidded peaks I and II. The variation between these two peaks within different CSF pools was less marked than in senile controls and ranged from a maximal value of peak I of 61 9O to
lZ,0,
Vo SST-25 28
SST-1/-.
~1oo. u_ 80
~ 6o 20. ~=-.=_
--5
10
15
20 FRACTIONS(2mI)
Fig. 2. Representative G P C pattern of SLI in a C S F pool from senile control patients. The dashed line indicates the detection limit of the RIA.
219
70-
V~ SST-25,28
.L g 6o-
SST-14
11
~ 50-
1
30. 20105
Fig.
3. G P C
-,-;~--.2 2- 2
10
15
pattern of SLI in
20 FRACTIONS (2 m[)
a CSF
pool from 14 adult schizophrenic patients.
a minimal value of peak II of 10%. However, a significant peak designated peak III appeared, which co-eluted with SRIF (Fig. 4). The CSF from patients with multi-infarct dementia (MID) also yielded peak I and II (peak I being the most prominent) and showed peak III in amounts comparable to the CSF from SDAT patients (Fig. 5). Patients with normal pressure hydrocephalus (NPH) exhibited a distribution pattern similar to that of senile controls. However, the amount of SLI co-eluting with SRIF was greater (more than 10% of total SLI) and is shown in Fig. 6. The relative GPC-distribution of SLI in the CSF of the investigated groups of patients is depicted in Table I. Separate chromatography of G-25 sf "void volume" fractions (i.e. peak I SLI) provided evidence for two SLI peptides (peaks Ia and Ib) with apparent molecular weights of about 10,000 and 15,500 daltons, respectively (Fig. 7).
70'
X,(, SST-25,2B
SST-14
I I I
50"
1 IT
~ /.030.J tn 2010-
5
Fig. 4.
GPC
10
15
20 FRACTIONS(2ml)
pattern of SLI in pooled CSF from patients with
SDAT.
220
V~ SST-25,28
SST-14
70, --60.
50 ~40
/
/
I
]]
-J
m 20
lo5
10
15 20 FRACT[ONS(2m[)
Fig. 5. G P C pattern of SLI in pooled C S F from patients with MID.
v~ SST-2S,28
SSTqz,
1
70
~ 60 ~ so ~ 40 -
-J
30
20 ,°
5
10
f5
20
FRACT[0NS (2mU Fig. 6. G P C pattern of SLI in pooled C S F from patients with NPH.
TABLE I RELATIVE G P C - D I S T R I B U T I O N ( S E P H A D E X G-25 s f 4 0 x l cm) OF SLI IN CSF OF PATIENTS CSF from 2-17 patients was pooled; n represents the number of pools. In cases where only one pool was investigated (MID, schizophrenics) the number of patients is indicated in brackets. Patients groups
Senile controls SDAT NPH MID (17) Schizophrenics (14)
n
4 5 2 1 1
% of total SLI + S.E.M. Peak I
Peak II
Peak [II
54 _+ 19 2 6 + 11 3 3 + 11 43 79
38 _+ 13 33+ 9 50_+ 4 25 21
8 _+ 2 4 1 ± 18 1 7 ± 15 32 not detectable
221
70
15,500D
v,
l
.O
~ 50"
~
10,000D
1
1
Ia
Ib
SST-25,28
1
0~
~ 302010"
.....
--??.-15
20
25 30 FRACTIONS{2ml}
Fig. 7. GPC pattern of G-25 sf void volume SLI re-chromatographed on Sephadex G-50 fine.
DISCUSSION
The determination of molecular components of somatostatin-like immunoreactivity in the CSF presents some difficulties presumably due to their low levels in circulating CSF and to the significant dilution factor inherent in the column separation procedure. Therefore it was decided to apply the concentration technique using Sep-Pak-cartridges consisting of octadecasilyl-silicagel where the SLI-related peptides can be adsorbed and subsequently eluted by aqueous organic solvents. Samples treated in this manner afforded a sufficient assay sensitivity to detect even small amounts of SLI accompanied by good and reproducible recovery values of synthetic samples of SRIF and SST-28. In all measured samples of CSF, i.e. in controls as well as in patients with dementia, marked heterogeneity was observed. In the CSF of control subjects only small amounts of SLI co-eluting with SRIF were detected (approx. 10~). Two main peaks were seen (peaks I and II) by chromatography on Sephadex G-25 superfine (G-25 sf), one eluting with the void volume of the column and the other eluting at a position between S ST-28 and SRIF. Rechromatography of the G-25 sf void volume SLI on a Sephadex G-50 fine column showed two peaks with apparent molecular weights of about 10,000 and 15,500 daltons, respectively. We suggest that this immunoreactive material of high molecular weight may represent somatostatin-precursor-like peptides. These findings are comparable with the data of Penman et al.32 and Reichlin et al. 36 obtained from the CSF of normal subjects. Nevertheless the apparent molecular weights of the putative precursors detected by Penman et al. 3z, by Reichlin et 81. 36 and in our study are not consistent, possibly due to the difference in patients studied, or rather to the different methodology of molecular weight determination. In addition it is remarkable that the 15,500 dalton component discovered by us had a molecular weight close to that of a human somatostatin precursor of 15,000 daltons, recently identified in cell-free translations ofhypothalamic mRNA 21. Evidence exists that the 10,000 dalton and 15,500 dalton components really represent separate somatostatin-like immunoreactivities and are
222 not related to protein binding of SLI or to non-specific protein interactions in the assay system. This is supported first by the use of 90~o aqueous acetic acid for the elution of SLI from Sep-Pak, an eluent of high dissociation potency; and secondly by the reasonable assumption that non-specifically interacting proteins can be removed by the Sep-Pak concentration procedure. The latter assumption is substantiated by Polonsky et al. 33, who found that an apparent precursor-like immunoreactivity detectable in Sep-Pak untreated human plasma vanished completely after applying the Sep-Pak method; thus indicating a non-specific interaction. The component eluting at a position between SST-28 and SRIF may be regarded as a short-chain elongated form of SRIF. A more precise structural assignment to this peptide cannot be made yet. The observation that we did find only small amounts of immunoreactive material in controls co-eluting with SRIF is in partial accordance with the results of Penman et al. 32 who detected no immunoreactivity eluting at the position of SRIF. Additionally, they detected significant amounts of SLI co-eluting with SST-28. In contrast, Reichlin et al.36 demonstrated the existence of marked amounts ofpeptides eluting at the position of SRIF as well as at the position of S ST-28. As mentioned above, we could not assign a separate peak corresponding to S ST-28 and S ST-25, respectively. The peak maxima of potentially existing SLI related to these peptides might occur between peaks I and II of the G-25 sf chromatography. Re-chromatography of peak I SLI on a Sephadex G-50 fine column did not provide evidence for any detectable amounts corresponding to SST-25/28. A separate fractionation of peak II SLI did not improve the separation. This could be due to a significant overlap of small amounts of SST-25/28 with the bulk SLI arising from peak II. Further clarification might be provided by use of the HPLC technique. In patients with SDAT and MID comparable amounts, and in patients with NPH, smaller amounts of the additional peak III co-eluting with SRIF as estimated by G-25 sf chromatography were seen. The SLI distribution pattern of schizophrenics was similar to that observed in controls, showing 7 9 ~ of peak I and 21% of peak II by G-25 sf chromatography, thus implicating only small differences between SLI from controls and schizophrenics. This is also demonstrated by corresponding levels of total SLI in the CSF of both groups 16. Our finding of a marked heterogeneity of SLI in the CSF of controls, as well as in patients with degenerative brain disease, is in good accordance with the results of Aronin et al.3. They investigated post-mortem SLI distribution in the caudate nucleus, putamen and globus pallidus from patients with Huntington's chorea, and from control subjects, and observed three pronounced peaks by means of GPC with Sephadex G-50: two peaks of SLI co-eluting with SST-28 and SST-14 and another eluting at the void volume of the column, and thus possibly representing a somatostatin precursor. They did not, however, observe differences in the SLI distribution pattern between controls and choreic patients. The nature and the origin of the various SLI peptide components remain to be determined. In general a systemic circulation origin of CSF-SLI is improbable 26"29'4e. Nevertheless the blood-borne nature of some fractions of SLI cannot be totally ruled
223 out, especially in pathological conditions such as senile dementia. This fact could be explained by a weakened blood-brain barrier. However, barrier dysfunction is more marked in MID than in SDAT 1 so that an increased diffusion of SRIF from blood into CSF should primarily be demonstrable in MID patients. On the other hand, the investigations of Leonardi et al. 2s have provided evidence for a blood-brain barrier integrity in patients with SDAT and MID. This is also supported by the finding of normal CSF albumin values in our patients. As a more rational explanation of increased SRIF secretion in demented patients, we suggest differences in the processing of somatostatin precursors in the CNS. Abnormal processing of a peptide precursor in SDAT has been proposed by Facchinetti et al. 15 in the case of pro-opiomelanocortin and is probably of special relevance for SLI, total levels of which are markedly decreased in SDAT. As yet, no molecular size distribution studies have been made in brain tissue of SDAT patients in which the total SLI of the cortex is significantly decreased ~3. A comparison of CSF and cerebral cortex SLI distribution would be most interesting and appropriate, especially since cerebral cortex is regarded as being one of the main sources of SLI in the human CSF.
ACKNOWLEDGEMENTS
This study was supported by the Deutsche Forschungsgemeinschaft (SFB 70) and the Sandoz Foundation for Therapeutic Research, Nuremberg, F.R.G. We thank Dr. J. Musil, Serono Co. for gifts of SST-14, SST-28 and Tyr°-SST-14. The expert technical assistance of Miss U. Strittmatter is gratefully acknowledged.
REFERENCES 1 Alafuzoff, I., Adoifsson, R., Bucht, G. and Winblad, B., Albumin and immunoglobulin in plasma and cerebrospinal fluid, and blood-cerebrospinal fluid barrier function in patients with dementia of Alzheimer type and multi-infarct dementia, J. Neurol. Sci., 60 (1983) 465-472. 2 Arimura, A., Sato, H., Coy, D.H. and Schally, A.V., Radioimmunoassay for GH-release inhibiting hormone, Proc. Soe. Exp. Biol. Med., 148 (1975) 784-789. 3 Aronin, N., Cooper, P.E., Lorenz, L.J., Bird, E.D., Sagar, S.M., Leemann, S.E. and Martin, J.B., Somatostatin is increased in the basal ganglia in Huntington disease, Ann. Neurol., 13 (1983) 519-526. 4 Bennett, H.P.J., Hudson, A.M., McMartin, C. and Purdon, G.E., Use of octadecasilylsilica for the extraction and purification of peptides in biological samples. Application to the identification of circulating metabolites of corticotropin(1-24)-tetracosapeptide and somatostatin in vivo, Biochem. J., 168 (1977) 9-13. 5 Benoit, R., Ling, N., Bakhit, C., Morrison, J.H., Alford B. and Guillemin, R., Somatostatin-28(1-12)-like immunoreactivity in the rat, Endocrinology, 111 (1982) 2149-2151. 6 BOhlen, P., Brazeau, P., Benoit, R., Ling, N., Esch, F. and Guillemin, R., Isolation and amino acid composition of two somatostatin-hke peptides from ovine hypothalamus: somatostatin-28 and somatostatin-25, Biochem. Biophys. Res. Commun., 96 (1980) 725-734.
224 7 Brazeau, P., Vale, W., Burgus, R., Ling, N., Butcher, M., Rivier, J. and Guitlemin, R., Hypothalanl~ polypeptide that inhibits the secretion of immunoreactive pituitary growth hormone, Science, 179 (1973) 77-79. 8 Brownstein, M., Arimura, A., Sato, H., Schally, A.V. and Kizer, J.S., The regional distribution of somatostatin in the rat brain, Endocrinology, 96 (1975) 1456-1461. 9 Cooper, P.E., Fernstrom, M.H., Rorstad, O.P., Leemann, S.E. and Martin, J.B., The regional distribution of somatostatin, substance P and neurotensin in human brain, Brain Res., 218 (1981) 219-232. 10 Cramer, H., Wolf, A., Rissler, K., Weigel, K. and Ostertag, C. Ventricular somatostatin-like immunoreactivity in patients with basal ganglia disease, J. Neurol., 232 (1985) 219-222. 11 Cramer, H., Schaudt, D., Rissler, K., Strubel, D., Warter, J.-M. and Kuntzmann, F., Somatostatin-like immunoreactivity in the CSF of patients with senile dementia of Alzheimer type, multi-infarct dementia and communicating hydrocephalus, J. Neurol., in press. 12 Cramer, H., Kohler, J., Oepen, G., Schomburg, G. and Schr0ter, E., Huntington's chorea - - measurements of somatostatin, substance P and cyclic nucleotides in the cerebrospinal fluid, J. Neurol., 225 (1981) 183-187. 13 Davies, P. and Terry, R.D., Cortical somatostatin-like immunoreactivity in cases of Alzheimer's disease and senile dementia of the Alzheimer type, Neurobiol. Aging, 2 (1980) 9-14. 14 Dupont, E., Christensen, S.E., Hansen, A.P., de Fine Olivarius, B. and Orskov, H., Low cerebrospinal fluid somatostatin in Parkinson disease: an irreversible abnormality, Neurology, 32 (1982) 312-314. 15 Facchinetti, F., Nappi, G., Storchi, R., Scelsi, M., Sances, G., Petraglia, F. and Genazzani, A.R., Abnormal proopiomelanocortin processing in Alzheimer disease, J. Neurochem., 44 (Suppl.) (1985) S 192 D; Tenth Meeting of the International Society for Neurochemistry, Riva (Italy) (Abstr.). 16 Gattaz, W.F., Rissler, K., Gattaz, D. and Cramer, H., Haloperidol increases somatostatin-like immunoreactivity in CSF of schizophrenic patients, Psychiatr. Res., in press. 17 Gerner, R.H. and Yamada, T., Altered neuropeptide concentrations in the cerebrospinal fluid of psychiatric patients, Brain Res., 238 (1982) 298-302. 18 Gluschankof, P., Morel, A., Benoit, R. and Cohen, P., The somatostatin-28 convertase of rat brain cortex generates both somatostatin- 14 and somatostatin-28 ( 1-12), Biochem. Biophys. Res. Commun., 128 (1985) 1051-1057. 19 Greenwood, F,C., Hunter, W.M. and Glover, J.S., The preparation of ~31I-labelled human growth hormone of high specific radioactivity, Biochern. J., 89 (1963) 114-120. 20 HOkfelt, T., Efendic, S., Johansson, O., Luft, R. and Arimura, A., Immunohistochemical localization of somatostatin (growth hormone release-inhibiting factor) in the guinea pig brain, Brain Res., 80 (1974) 165-169. 21 Ivell, R., Emson, P. and Richter, D., Human neuropeptide Y, somatostatin and vasopressin precursors identified in cell-free translation of hypothalamic mRNA, FEBS Lett., 177 (1984) 175-178. 22 Kohl~r, J., SchrSler, E. and Cramer, H., Somatostatin-like immunoreactivity in the cerebrospinal fluid of neurological patients, Arch. Psychiatr. Nervenkr., 231 (1982) 503-508. 23 Kronheim, S., Berelowitz, M. and Pimstone, B.L., The presence of immunoreactive growth hormone release-inhibiting hormone in normal cerebrospinal fluid. Clin. Endocrinol., 6 (1974) 411-415. 24 Krulich, L., Dhariwal, A.P.S. and McCann, S.M., Stimulatory and inhibitory effects of purified hypothalamic extracts on growth hormone release from rat pituitary in vitro, Endocrinology, 83 (1968) 783-790. 25 Leonardi, A., Gandolfo, C., Caponnetto, C., Arata, L. and Vecchia, R., The integrity of the blood-brain barrier in Alzheimer's type and multi-infarct dementia evaluated by the study of albumin and IgG in serum and cerebrospinal fluid, jr. Neurol. Sci., 67 (1985) 253-261. 26 Lundqvist, G., Gustavsson, S. and Elde, R., A radioimmunosorbent assay for plasma somatostatin, Clin. Chem. Acta, 101 (1980) 183-191. 27 Miller, J.J., Schultz, G.S. and Levy, R.S., Rapid purification of radioiodinated peptides with Sep-Pak reversed phase cartridges and HPLC, Int. J. Pept. Prot. Res., 24 (1984) 112-122.
225 28 Oram, J.J., Edwardson, J. and Millard, P.H., Investigation of cerebrospinal fluid neuropeptides in idiopathic senile dementia, Gerontology, 27 (1981) 216-223. 29 Orskov, H., Hansen, A.P., de Fine Olivarius, B., Pederson, E., SOrensen, K., Dupont, E., Ingerslev, J. and Christensen, S.E., Cerebrospinal fluid (CSF) somatostatin-like immunoreactivity (SLI): methodology; secretion and origin, and pathology in neurological disease. In C. Perris, G. Struwe and B. Jansson (Eds.), Biological Psychiatry, Elsevier/North-Holland Biomedical Press, 1981, pp. 335-338. 30 Patel, Y.C. and Reichlin S., Somatostatin in hypothalamus, extrahypothalamic brain, and peripheral tissues of the rat, Endocrinology, 102 (1978) 523-530. 31 Patel Y.C., Krishna R. and Reichlin S., Somatostatin in human cerebrospinal fluid, New Engl. J. Med., 296 (1977) 529-533. 32 Penman, E. and Wass, J.A.H., Radioimmunoassay of somatostatin: methodological problems and physiological investigations. In C.A. Bizollon (Ed.), PhysiologicalPeptides and new Trends in Radioimmunology, Elsevier/North Holland Biomedical Press, 1981, pp. 3-18. 33 Polonsky, K.S., Shoelson, S.E. and Docherty, H.M., Plasma somatostatin 28 increases in response to feeding in man, J. Clin. lnvest., 71 (1983) 1514-1518. 34 Pradayrol, L., JSrnvall, H., Mutt, V. and Ribet, A., N-terminally extended somatostatin: the primary structure of somatostatin-28, FEBS Lett., 109 (1980) 55-58. 35 Reichlin, S., Somatostatin in the nervous system, in F.E. Bloom, S. Bird, F.O. Schmitt (Eds.),Molecular Genetic Neuroscience, Raven Press, New York, 1982, pp. 359-372. 36 Reichlin, S., Robbins, R. and Lechan, R., Somatostatin and the nervous system. In S. Raptis, J. Rosenthal, J.E. Gerich (Eds.), Proceedings of the 2nd International Symposium on Somatostatin (Athens 1981), Attempto Verlag, Ttlbingen (F.R.G.), 1984, pp. 21-37. 37 Reisberg, B., London, E., Ferris, S.H., Borenstein, J., Scheier, L. and de Leon, M.J., The brief cognitive rating scale: Language and mood concomitants in primary degenerative dementia, Psychopharm. Bull. 19 (1983) 702-708. 38 Rissler, K., Cramer, H., Schaudt, D. and Strubel, D., Somatostatin-like immunoreactivity in the CSF. Levels in degenerative brain disease and species distribution, J. Neurochem. 44 (1985) S 196 C (Suppl.); Tenth Meeting of the International Society for Neurochemistry, Riva (Italy), (Abstr.). 39 Rubinow, D.R., Gold, P.W., Post, R.M, Ballenger, J.C., Cowdry, R., Bollinger, J. and Reichlin, S., CSF somatostatin in affective illness, Arch. Gen. Psychiatr., 40 (1983) 409-412. 40 Serby, M., Richardson, S.B., Twente, S., Siekierski, J. and Rotrosen, J., CSF somatostatin in Alzheimer's disease, Abstracts of the Annual Meeting of the American College of Neuropharmacology (San Juan, 1983) p. 84. 41 S0rensen, K.V., Christensen, S.E., Dupont, E., Hansen, A.P. Pedersen, E. and Orskov, H., Low somatostatin content in cerebrospinal fluid in multiple sclerosis, Acta Neurol. Scand., 61 (1980) 186-191. 42 S0rensen, K.V., Christensen, S.E., Hansen, A.P., Ingerslev, J., Pedersen, E. and Orskov, H., The origin of cerebrospinal fluid somatostatin: hypothalamic or disperse central nervous system secretion? Neuroendocrinology, 32 (1981) 335-338. 43 Spitzer, R.L., Endicott, J. and Robins, E., Research diagnostic criteria: rationale and reliability, Arch. Gen. Psychiatr., 35 (1978) 773-782. 44 Vale, W., Brazeau, P., Rivier, C., Brown, M., Ross, B., Rivier, J., Burgus R., Ling N. and Guillemin R., Somatostatin, Rec. Progr. Horm. Res., 31 (1975) 365-397. 45 Wass, J.A.H., Penman, E., Medbak, S., Rees, L.H. and Besser, G.M., CSF and plasma somatostatin levels in acromegaly, Clin. EndocrinoL, 13 91980) 235-241. 46 Wood P.L., Etienne P., Lal S., Gauthier S., Cajal S. and Nair N.P.V., Reduced lumbar CSF somatostatin levels in Alzheimer's disease, Life Sci., 31 (1982) 2073-2079. 47 Zingg, H.H. and Patel, Y.C., Somatostatin precursors: evidence for presence in and release from rat median eminence and neurohypophysis, Biochem. Biophys. Res. Commun., 90 (1976) 466-472. 48 Zingg, H.H. and Patel, Y.C., Processing of somatostatin-27 to somatostatin-14 by rat hypothalamic synaptosomal membranes, Life Sci., 33 (1983) 1241-1247.
213
Elsevier Scientific Publishers Ireland Ltd. NSR 00108
Molecular Size Distribution of Somatostatin-Like Immunoreactivity in the Cerebrospinal Fluid of Patients with Degenerative Brain Disease* Klaus Rissler ~, Hinrich Cramer 1, Dieter Schaudt 2, Denise Strubel 2 and Wagner F. Gattaz 3 IDepartment of Neurology and Clinical Neurophysiology, Universityof Freiburg, D-7800 Freiburg (F.R.G.); 2Department of Geriatrics, University of Strasbourg, Strasbourg (France); and 3Central Institute of Mental Health, D-6800 Mannheim (F.R.G.) (Received July 19th, 1985; Revised version received September 30th, 1985; Accepted October 2nd, 1985)
Key words."somatostatin - - molecular size distribution-- cerebrospinal fluid-- dementia - - hydrocephalus SUMMARY The molecular size distribution of somatostatin-like immunoreactivity (SLI) in the cerebrospinal fluid (CSF) of patients with brain disease was investigated by separation with a Sephadex G-25 superfine column and subsequent radioimmunoassay of the eluate. Marked heterogeneity of SLI in the CSF of control subjects as well as in demented patients, was observed. Controls and schizophrenics exhibited an SLI distribution pattern consisting mainly of two pronounced peaks: the first eluting with the void volume of the column; the second being compatible with a peptide of N-terminally extended somatostatin- 14. SLI from the CSF of patients with senile dementia of the Alzheimer type (SDAT), multi-infarct dementia (MID) and normal pressure hydrocephalus (NPH) showed the same two peaks found in controls and schizophrenics; and in addition, a third peak co-eluting with somatostatin-14. However, this peak was more pronounced in patients with SDAT and M I D than in patients with NPH. Re-chromatography of G-25 sf void volume immunoreactivity afforded two fractions of an apparent molecular weight of about 10,000 daltons and 15,500 daltons, respectively.
INTRODUCTION
During the course of efforts to determine the distribution of the growth hormonereleasing factor in rat hypothalamus, a substance that inhibited growth hormone release * This paper is dedicated to Professor Georg Hertting on the occasion of his 60th birthday. Correspondence: K. Rissler, Department of Neurology and Clinical Neurophysioiogy, University of Freiburg, Hansastr. 9, D-7800 Freiburg, F.R.G. 0168-0102/86/$03.50 © 1986 Elsevier Scientific Publishers Ireland Ltd.
214 was detected by Krulich et al. 24. In 1972, Brazeau et al. v succeeded in isolating from ovine hypothalamus a cyclic tetradecapeptide that inhibited growth hormone secretion, and which was shown not to be limited to the hypothalamus but to have a widespread distribution throughout the central nervous system (CNS) 8'9'20'30"44. Besides its inhibitory effect on the release of growth hormone from the anterior pituitary, somatostatin is now believed to function as a neurotransmitter or a neuromodulator. The peptide co-exists with other neurotransmitters in nerve endings from which it is released in response to depolarizing stimuli 35. Subsequent work has considerably expanded the initial concept of tetradecapeptide somatostatin's main function, being the regulation of growth hormone release. It is now obvious that somatostatin-related peptides represent a whole family which includes the originally identified peptide somatostatin-14 (SRIF), somatostatin-283a, somatostatin256 as well as immunoreactivity related to the somatostatin-28 l_~2-sequence 5 and also the existence of precursor-like peptides of high molecular weight. It has been shown that these putative precursors could be converted into the circulating biologically active peptides by a trypsin-like enzymatic cleavage47"4s. Recently a specific enzyme converthag somatostatin-28 into somatostatin-14 and somatostatin-281_~2 designated as somatostatin-28 convertase has been discovered TM. In 1977, Patel et al.31 suggested that cerebrospinal fluid (CSF) levels of somatostatin-like immunoreactivity (SLI) might be a useful index for CNS damage when they determined CSF levels of SLI in neurological patients. In the following years other reports concerning CSF-SLI levels in patients with Parkinson's disease ~°'~4, senile dementia of Alzheimer type 11'28"4°'46, Huntington's chorea 12, psychiatric and affective disorders 17,39, neurological disease2z, multiple sclerosis41 and acromegaly45 have been published. Kronheim et al. 23 performed the In'st study of the molecular size of SLI in the CSF of normal subjects. This was followed by investigations by Penman et al.32 and Reichlin et 81.36. As far as we know, molecular size investigations of SLI in the CSF of neurological patients have not yet been reported. We have therefore studied the molecular size of SLI in the CSF of patients with degenerative brain disease 3s. In this paper we present findings obtained from patients with senile dementia of the Alzbeimer type (SDAT), with multi-infarct dementia (MID), with normal pressure hydrocephalus (NPH) and with schizophrenia. We also present findings obtained from age-matched control patients.
MATERIALS AND METHODS
Synthetic samples of tyrosine°-somatostatin-14, cyclic somatostatin-14 and somatostatin-28 were kindly donated by Serono (Freiburg, F.R.G.), Somatostatin-25 was purchased from Peninsula Laboratories (Belmont, CA, U.S,A.). For radioiodination a
215 solution of sodium iodide-125 carrier-free in 0.1 M NaOH (Amersham-Buchler, Braunschweig, F.R.G.) was used. Sephadex G-25 fine/superfine and Sephadex G-50 fine were purchased from Pharmacia (Uppsala, Sweden). Sep-Pak-cartridges containing octadecasilyl-silicagel (ODS-silica) were manufactured by Millipore-Waters Associates (Milford, MA, U.S.A.). The antiserum K-18 (Ferring, Kiel, F.R.G.) which recognizes somatostatin at the ring structure was raised in rabbits. It shows 100% cross-reaction with cyclic somatostatin-14, somatostatin-28 and somatostatin-25 and no cross-reaction with a number of central nervous system and gastrointestinal peptides including human growth hormone (hGH), cholecystokinin, bombesin, substance-P, vasointestinal polypeptide (VIP), thyrotropin releasing hormone (TRH), adrenocorticotropic hormone (ACTH), gastric inhibitory polypeptide, insulin, glucagon, gastrin, calcitonin and parathormone. Myoglobin, cytochrome-C and aprotinin were obtained from Serva (Heidelberg, F.R.G.). Charcoal (type "Norit-A") and dextran ("Dextran T70") were produced by Serva (Heidelberg, F.R.G.) and Pharmacia (Uppsala, Sweden), respectively. The preparation of the lZSI-labelled tyrosine°-somatostatin- 14 was performed by use of the chloramine-T method In'st applied by Greenwood et al.19 for iodination of human growth hormone. Purification of the labelled antigen was carried out by application of a two-step procedure: first, by adsorption on ODS-silica 27 and subsequent elution with acetone/water/acetic acid (39:40: 1); and second, by gel permeation chromatography on a 90 x 1 cm Sephadex G-25 free column with 0.1 M acetic acid containing 0.05 % bovine serum albumin. Appropriate fractions were pooled and stored in aliquots of about 5-10 #Ci at - 30 °C. A tracer prepared in this manner was storable for at least 3 months. Radioimmunoassay was performed according to a modified procedure of Arimura et al.2 using cyclic somatostatin (SRIF) as the standard. For this purpose the antiserum (Final dilution 1 : 1,400,000), either standard or sample, was incubated in a final volume of 0.7 ml for about 12-16 h at 4 °C. The buffer system consisted of 0.1 M sodium phosphate, pH 7.5, containing 0.025 M EDTA, 0.15 M NaC1, 0.005 M (0.02 %) sodium azide, and 0.25~ BSA. 125I-labelled antigen was added (approximately 10,000 cpm) and incubation was continued for another 48 h at 4 ° C. Separation of the free from the antibody-bound antigen was carried out with dextran-coated charcoal, i.e. a suspension of 250 mg charcoal and 25 mg of dextran per 100 ml of incubation buffer which lacked BSA was used. One ml of the suspension was added to each tube, and the mixture was incubated for 30 min at 4 °C. After centrifugation the supernatants were collected separately and the radioactivity was determined. All measurements (standards and samples) were done in duplicate. Denaturation of the tracer during incubation, as estimated by omitting the antibody, was negligible ( < 2~). The sensitivity of the assay, defined as that concentration of synthetic SRIF necessary to inhibit the binding of [ 125I]Tyr°-somatostatin-14 to a mean value at least two standard deviations below the mean value obtained in the absence of SRIF, was about 10 pg/ml. The intra-assay and inter-assay coefficients of variation were 5% and 129/o, respectively.
216 For peptide characterization, an appropriate sample of cerebrospinal fluid from several patients of one group, containing about 500-1000 pg/ml total somatostatin-like immunoreactivity (SLI), was concentrated and thus prepurified by adsorption on ODSsilica4 and consecutively eluted with 2-3 ml of 90% aqueous acetic acid. After lyophilization, the residual concentrate was subjected to chromatography on a 40 × 1 cm Sephadex G-25 superfine column using the same eluent as for tracer purification. Twenty-five fractions, each containing 2 ml of eluate, were collected. The fractions were lyophilized and reconstituted in assay buffer for subsequent radioimmunoassay. [ ~zSI]Tyr°-somatostatin- 14 treated in this way showed 80 ~o recovery of total radioactivity. The radioactive material exhibited more than 80 ~o of the maximal antibody-binding when compared to the untreated peptide and was displaceable by cycfic somatostatin-14 in a concentration-dependent manner. The assay sensitivity obtained by use of this tracer was equal to that obtained with the untreated radiopeptide. The recovery values for the equally treated synthetic samples of cyclic somatostatin-14 and somatostatin-28 were 90yo and 60yo, respectively, and no immunoreactive material other than the reference peptides could be detected. This finding suggests that the peptide material did not undergo marked chemical modification. Analogously treated samples of native CSF showed 60yo recovery of SLI. The recovery of synthetic cyclic somatostatin-14 and somatostatin-28 from the Sephadex column was complete. Evaluation of the apparent molecular weights of Sephadex G-25 superfine void volume SLI (exclusion limit 5000 daltons) was performed by column chromatography of the corresponding fractions on Sephadex G-50 t'me (84 x 1 era) using a set of appropriate molecular weight markers (myoglobin 18000 daltons; cytochrome-C 13,000 daltons; aprotinin 6500 daltons; somatostatin-28 3200 daltons; somatostatin- 14 1600 daltons). Prior to chromatography, the CSF void volume samples were incubated for 24 h with 8 M urea containing 0.05 M phosphate buffer (pH 7.5). For proof of immunological identity between synthetic SRIF and SLI measured in CSF, a dilution series of a concentrated CSF sample was prepared and then compared
95.
80-
70" "~ 60E -~ $0" /*0" °7 30" ~ :ZO" N
rif
40
J 1:8
80 1 :/~
160 I 1:2
320 Srif{po/mI)
,: "' 1:1 CSF-OlLUTION
Fig. 1. Logit-log plot of serial dilutions of synthetic SRIF and SLI obtained from human CSF showing parallelism of both c u r v e s .
217 to a corresponding dilution curve of synthetic SRIF. Both curves exhibited parallelism and therefore immunological identity (Fig. 1). Patients
All patients examined in this study were hospitalized in the Geriatric Department, Strasbourg, and the Psychiatric Department, Mannheim. The high age control group (n = 13, mean age 83.5 + 5 years) consisted of patients without central nervous disease, intellectual deterioration or major psychiatric disturbances. The only pathology of these patients consisted of either mild peripheral nervous disease (diabetic neuropathy, plexus neuritis), a slight degree of depression, skeletal diseases (arthrosis and bone fractures) or moderate cardiorespiratory insufficiency. Lumbar punctures were performed either for diagnostic purpose in order to exclude central nervous disease, or when informed consent was obtained after the nature of the study had been fully explained. The diagnosis of patients with senile dementia of the Alzheimer type (n = 32, mean age 84 + 7 years) was based on clinical criteria. All these patients had progressive deterioration of intellectual functions including language, memory and personality. Symptomatic causes of dementia were excluded by the absence of strokes, focalized neurological deficits, by laboratory examinations including EEG and vascular status. The intellectual status of the patients and the degree of deterioration were estimated by a neuropsychological examination based on the Brief Cognitive Rating Scale (BCRS) of Reisberg 37. The group of patients with vascular disease (n = 17, mean age 79 + 7 years) was characterized by a history of strokes and the presence of hemiplegia, unilateral or bilateral Babinski signs and the presence of various degrees of dementia of the multi-infarct type. In all these patients the Hachinski ischemic score was above 12. The diagnosis of patients having normal pressure hydrocephalus (n = 5, mean age 71 + 6 years) was established by the progressive appearance of mild mental deterioration, gait disturbance and loss of sphincter control. In these patients, the diagnosis was confu'med by cranial computerized scans which showed marked ventricular dilatation and the absence of or only a moderate degree of cortical atrophy. The group of paranoid schizophrenics (n = 14, mean age 36 + 10 years) was examined according to the Research Diagnostic Criteria of Spitzer et al.43. Informed consent was obtained from all patients or from their closest relations after the nature of the study had been fully explained. Almost all geriatric patients received minor drug treatment, at least for other than neurological disease (cardio-vascular, diabetes, rheumatism). Agitated demented patients were treated with low doses ofhaloperidol or thioridazine. A number of patients received low doses of benzodiazepines, and four control subjects and two vascular disease patients were treated with antidepressants (amitryptiline, clomipramine). The paranoid schizophrenic patients received no drug treatment for at least 3 weeks prior to the lumbar puncture. CSF was obtained with patients in a sitting position between 09.00 and 10.00 h after
218 the patients had fasted for 12 h and had bed rest for 10 h. The samples were immediately congelated and stored at - 4 0 °C until analysis. Routine laboratory analysis, including albumin, glucose and cell count was performed on CSF from all subjects. Values were within the normal range.
RESULTS
Molecular size heterog~aeity of somatostatin-like immunoreactivity (SLI) was observed in the cerebrospinal fluid (CSF) of all the groups of patients examined. Two prominent peaks subsequently designated peaks I and II were shown by all CSF pools. Peak I eluted with the void volume of the Sephadex G-25 superfine column (G-25 sf), whereas peak II appeared between somatostatin-25 (SST-25) and synthetic somatostatin-14 (SRIF). No discrete peak attributable to SST-25 or to somatostatin-28 (SST-28) could be detected. However, the resolution between peak I and peak II is not satisfactory, so it is possible that significant amounts of SLI corresponding to SST-25 or SST-28 had been overlapped by adjacent peaks. In senile control patients, SLI consisted of peak I and peak II representing approximately 90% of the total SLI and of a smaller peak III of immunoreactivity co-eluting with synthetic SRIF (approx. 10%) (Fig. 2). The ratio of the main peaks I and II showed considerable variation between different CSF pools ranging from 85 9o peak I and 15% peak II to 92% peak II and 89o peak I, respectively. CSF from paranoid schizophremics revealed a distribution pattern similar to that of senile controls; however, SLI co-eluting with SRIF was virtually lacking (Fig, 3). Patients with senile dementia of the Alzheimer type (SDAT) also yidded peaks I and II. The variation between these two peaks within different CSF pools was less marked than in senile controls and ranged from a maximal value of peak I of 61 9O to
lZ,0,
Vo SST-25 28
SST-1/-.
~1oo. u_ 80
~ 6o 20. ~=-.=_
--5
10
15
20 FRACTIONS(2mI)
Fig. 2. Representative G P C pattern of SLI in a C S F pool from senile control patients. The dashed line indicates the detection limit of the RIA.
219
70-
V~ SST-25,28
.L g 6o-
SST-14
11
~ 50-
1
30. 20105
Fig.
3. G P C
-,-;~--.2 2- 2
10
15
pattern of SLI in
20 FRACTIONS (2 m[)
a CSF
pool from 14 adult schizophrenic patients.
a minimal value of peak II of 10%. However, a significant peak designated peak III appeared, which co-eluted with SRIF (Fig. 4). The CSF from patients with multi-infarct dementia (MID) also yielded peak I and II (peak I being the most prominent) and showed peak III in amounts comparable to the CSF from SDAT patients (Fig. 5). Patients with normal pressure hydrocephalus (NPH) exhibited a distribution pattern similar to that of senile controls. However, the amount of SLI co-eluting with SRIF was greater (more than 10% of total SLI) and is shown in Fig. 6. The relative GPC-distribution of SLI in the CSF of the investigated groups of patients is depicted in Table I. Separate chromatography of G-25 sf "void volume" fractions (i.e. peak I SLI) provided evidence for two SLI peptides (peaks Ia and Ib) with apparent molecular weights of about 10,000 and 15,500 daltons, respectively (Fig. 7).
70'
X,(, SST-25,2B
SST-14
I I I
50"
1 IT
~ /.030.J tn 2010-
5
Fig. 4.
GPC
10
15
20 FRACTIONS(2ml)
pattern of SLI in pooled CSF from patients with
SDAT.
220
V~ SST-25,28
SST-14
70, --60.
50 ~40
/
/
I
]]
-J
m 20
lo5
10
15 20 FRACT[ONS(2m[)
Fig. 5. G P C pattern of SLI in pooled C S F from patients with MID.
v~ SST-2S,28
SSTqz,
1
70
~ 60 ~ so ~ 40 -
-J
30
20 ,°
5
10
f5
20
FRACT[0NS (2mU Fig. 6. G P C pattern of SLI in pooled C S F from patients with NPH.
TABLE I RELATIVE G P C - D I S T R I B U T I O N ( S E P H A D E X G-25 s f 4 0 x l cm) OF SLI IN CSF OF PATIENTS CSF from 2-17 patients was pooled; n represents the number of pools. In cases where only one pool was investigated (MID, schizophrenics) the number of patients is indicated in brackets. Patients groups
Senile controls SDAT NPH MID (17) Schizophrenics (14)
n
4 5 2 1 1
% of total SLI + S.E.M. Peak I
Peak II
Peak [II
54 _+ 19 2 6 + 11 3 3 + 11 43 79
38 _+ 13 33+ 9 50_+ 4 25 21
8 _+ 2 4 1 ± 18 1 7 ± 15 32 not detectable
221
70
15,500D
v,
l
.O
~ 50"
~
10,000D
1
1
Ia
Ib
SST-25,28
1
0~
~ 302010"
.....
--??.-15
20
25 30 FRACTIONS{2ml}
Fig. 7. GPC pattern of G-25 sf void volume SLI re-chromatographed on Sephadex G-50 fine.
DISCUSSION
The determination of molecular components of somatostatin-like immunoreactivity in the CSF presents some difficulties presumably due to their low levels in circulating CSF and to the significant dilution factor inherent in the column separation procedure. Therefore it was decided to apply the concentration technique using Sep-Pak-cartridges consisting of octadecasilyl-silicagel where the SLI-related peptides can be adsorbed and subsequently eluted by aqueous organic solvents. Samples treated in this manner afforded a sufficient assay sensitivity to detect even small amounts of SLI accompanied by good and reproducible recovery values of synthetic samples of SRIF and SST-28. In all measured samples of CSF, i.e. in controls as well as in patients with dementia, marked heterogeneity was observed. In the CSF of control subjects only small amounts of SLI co-eluting with SRIF were detected (approx. 10~). Two main peaks were seen (peaks I and II) by chromatography on Sephadex G-25 superfine (G-25 sf), one eluting with the void volume of the column and the other eluting at a position between S ST-28 and SRIF. Rechromatography of the G-25 sf void volume SLI on a Sephadex G-50 fine column showed two peaks with apparent molecular weights of about 10,000 and 15,500 daltons, respectively. We suggest that this immunoreactive material of high molecular weight may represent somatostatin-precursor-like peptides. These findings are comparable with the data of Penman et al.32 and Reichlin et al. 36 obtained from the CSF of normal subjects. Nevertheless the apparent molecular weights of the putative precursors detected by Penman et al. 3z, by Reichlin et 81. 36 and in our study are not consistent, possibly due to the difference in patients studied, or rather to the different methodology of molecular weight determination. In addition it is remarkable that the 15,500 dalton component discovered by us had a molecular weight close to that of a human somatostatin precursor of 15,000 daltons, recently identified in cell-free translations ofhypothalamic mRNA 21. Evidence exists that the 10,000 dalton and 15,500 dalton components really represent separate somatostatin-like immunoreactivities and are
222 not related to protein binding of SLI or to non-specific protein interactions in the assay system. This is supported first by the use of 90~o aqueous acetic acid for the elution of SLI from Sep-Pak, an eluent of high dissociation potency; and secondly by the reasonable assumption that non-specifically interacting proteins can be removed by the Sep-Pak concentration procedure. The latter assumption is substantiated by Polonsky et al. 33, who found that an apparent precursor-like immunoreactivity detectable in Sep-Pak untreated human plasma vanished completely after applying the Sep-Pak method; thus indicating a non-specific interaction. The component eluting at a position between SST-28 and SRIF may be regarded as a short-chain elongated form of SRIF. A more precise structural assignment to this peptide cannot be made yet. The observation that we did find only small amounts of immunoreactive material in controls co-eluting with SRIF is in partial accordance with the results of Penman et al. 32 who detected no immunoreactivity eluting at the position of SRIF. Additionally, they detected significant amounts of SLI co-eluting with SST-28. In contrast, Reichlin et al.36 demonstrated the existence of marked amounts ofpeptides eluting at the position of SRIF as well as at the position of S ST-28. As mentioned above, we could not assign a separate peak corresponding to S ST-28 and S ST-25, respectively. The peak maxima of potentially existing SLI related to these peptides might occur between peaks I and II of the G-25 sf chromatography. Re-chromatography of peak I SLI on a Sephadex G-50 fine column did not provide evidence for any detectable amounts corresponding to SST-25/28. A separate fractionation of peak II SLI did not improve the separation. This could be due to a significant overlap of small amounts of SST-25/28 with the bulk SLI arising from peak II. Further clarification might be provided by use of the HPLC technique. In patients with SDAT and MID comparable amounts, and in patients with NPH, smaller amounts of the additional peak III co-eluting with SRIF as estimated by G-25 sf chromatography were seen. The SLI distribution pattern of schizophrenics was similar to that observed in controls, showing 7 9 ~ of peak I and 21% of peak II by G-25 sf chromatography, thus implicating only small differences between SLI from controls and schizophrenics. This is also demonstrated by corresponding levels of total SLI in the CSF of both groups 16. Our finding of a marked heterogeneity of SLI in the CSF of controls, as well as in patients with degenerative brain disease, is in good accordance with the results of Aronin et al.3. They investigated post-mortem SLI distribution in the caudate nucleus, putamen and globus pallidus from patients with Huntington's chorea, and from control subjects, and observed three pronounced peaks by means of GPC with Sephadex G-50: two peaks of SLI co-eluting with SST-28 and SST-14 and another eluting at the void volume of the column, and thus possibly representing a somatostatin precursor. They did not, however, observe differences in the SLI distribution pattern between controls and choreic patients. The nature and the origin of the various SLI peptide components remain to be determined. In general a systemic circulation origin of CSF-SLI is improbable 26"29'4e. Nevertheless the blood-borne nature of some fractions of SLI cannot be totally ruled
223 out, especially in pathological conditions such as senile dementia. This fact could be explained by a weakened blood-brain barrier. However, barrier dysfunction is more marked in MID than in SDAT 1 so that an increased diffusion of SRIF from blood into CSF should primarily be demonstrable in MID patients. On the other hand, the investigations of Leonardi et al. 2s have provided evidence for a blood-brain barrier integrity in patients with SDAT and MID. This is also supported by the finding of normal CSF albumin values in our patients. As a more rational explanation of increased SRIF secretion in demented patients, we suggest differences in the processing of somatostatin precursors in the CNS. Abnormal processing of a peptide precursor in SDAT has been proposed by Facchinetti et al. 15 in the case of pro-opiomelanocortin and is probably of special relevance for SLI, total levels of which are markedly decreased in SDAT. As yet, no molecular size distribution studies have been made in brain tissue of SDAT patients in which the total SLI of the cortex is significantly decreased ~3. A comparison of CSF and cerebral cortex SLI distribution would be most interesting and appropriate, especially since cerebral cortex is regarded as being one of the main sources of SLI in the human CSF.
ACKNOWLEDGEMENTS
This study was supported by the Deutsche Forschungsgemeinschaft (SFB 70) and the Sandoz Foundation for Therapeutic Research, Nuremberg, F.R.G. We thank Dr. J. Musil, Serono Co. for gifts of SST-14, SST-28 and Tyr°-SST-14. The expert technical assistance of Miss U. Strittmatter is gratefully acknowledged.
REFERENCES 1 Alafuzoff, I., Adoifsson, R., Bucht, G. and Winblad, B., Albumin and immunoglobulin in plasma and cerebrospinal fluid, and blood-cerebrospinal fluid barrier function in patients with dementia of Alzheimer type and multi-infarct dementia, J. Neurol. Sci., 60 (1983) 465-472. 2 Arimura, A., Sato, H., Coy, D.H. and Schally, A.V., Radioimmunoassay for GH-release inhibiting hormone, Proc. Soe. Exp. Biol. Med., 148 (1975) 784-789. 3 Aronin, N., Cooper, P.E., Lorenz, L.J., Bird, E.D., Sagar, S.M., Leemann, S.E. and Martin, J.B., Somatostatin is increased in the basal ganglia in Huntington disease, Ann. Neurol., 13 (1983) 519-526. 4 Bennett, H.P.J., Hudson, A.M., McMartin, C. and Purdon, G.E., Use of octadecasilylsilica for the extraction and purification of peptides in biological samples. Application to the identification of circulating metabolites of corticotropin(1-24)-tetracosapeptide and somatostatin in vivo, Biochem. J., 168 (1977) 9-13. 5 Benoit, R., Ling, N., Bakhit, C., Morrison, J.H., Alford B. and Guillemin, R., Somatostatin-28(1-12)-like immunoreactivity in the rat, Endocrinology, 111 (1982) 2149-2151. 6 BOhlen, P., Brazeau, P., Benoit, R., Ling, N., Esch, F. and Guillemin, R., Isolation and amino acid composition of two somatostatin-hke peptides from ovine hypothalamus: somatostatin-28 and somatostatin-25, Biochem. Biophys. Res. Commun., 96 (1980) 725-734.
224 7 Brazeau, P., Vale, W., Burgus, R., Ling, N., Butcher, M., Rivier, J. and Guitlemin, R., Hypothalanl~ polypeptide that inhibits the secretion of immunoreactive pituitary growth hormone, Science, 179 (1973) 77-79. 8 Brownstein, M., Arimura, A., Sato, H., Schally, A.V. and Kizer, J.S., The regional distribution of somatostatin in the rat brain, Endocrinology, 96 (1975) 1456-1461. 9 Cooper, P.E., Fernstrom, M.H., Rorstad, O.P., Leemann, S.E. and Martin, J.B., The regional distribution of somatostatin, substance P and neurotensin in human brain, Brain Res., 218 (1981) 219-232. 10 Cramer, H., Wolf, A., Rissler, K., Weigel, K. and Ostertag, C. Ventricular somatostatin-like immunoreactivity in patients with basal ganglia disease, J. Neurol., 232 (1985) 219-222. 11 Cramer, H., Schaudt, D., Rissler, K., Strubel, D., Warter, J.-M. and Kuntzmann, F., Somatostatin-like immunoreactivity in the CSF of patients with senile dementia of Alzheimer type, multi-infarct dementia and communicating hydrocephalus, J. Neurol., in press. 12 Cramer, H., Kohler, J., Oepen, G., Schomburg, G. and Schr0ter, E., Huntington's chorea - - measurements of somatostatin, substance P and cyclic nucleotides in the cerebrospinal fluid, J. Neurol., 225 (1981) 183-187. 13 Davies, P. and Terry, R.D., Cortical somatostatin-like immunoreactivity in cases of Alzheimer's disease and senile dementia of the Alzheimer type, Neurobiol. Aging, 2 (1980) 9-14. 14 Dupont, E., Christensen, S.E., Hansen, A.P., de Fine Olivarius, B. and Orskov, H., Low cerebrospinal fluid somatostatin in Parkinson disease: an irreversible abnormality, Neurology, 32 (1982) 312-314. 15 Facchinetti, F., Nappi, G., Storchi, R., Scelsi, M., Sances, G., Petraglia, F. and Genazzani, A.R., Abnormal proopiomelanocortin processing in Alzheimer disease, J. Neurochem., 44 (Suppl.) (1985) S 192 D; Tenth Meeting of the International Society for Neurochemistry, Riva (Italy) (Abstr.). 16 Gattaz, W.F., Rissler, K., Gattaz, D. and Cramer, H., Haloperidol increases somatostatin-like immunoreactivity in CSF of schizophrenic patients, Psychiatr. Res., in press. 17 Gerner, R.H. and Yamada, T., Altered neuropeptide concentrations in the cerebrospinal fluid of psychiatric patients, Brain Res., 238 (1982) 298-302. 18 Gluschankof, P., Morel, A., Benoit, R. and Cohen, P., The somatostatin-28 convertase of rat brain cortex generates both somatostatin- 14 and somatostatin-28 ( 1-12), Biochem. Biophys. Res. Commun., 128 (1985) 1051-1057. 19 Greenwood, F,C., Hunter, W.M. and Glover, J.S., The preparation of ~31I-labelled human growth hormone of high specific radioactivity, Biochern. J., 89 (1963) 114-120. 20 HOkfelt, T., Efendic, S., Johansson, O., Luft, R. and Arimura, A., Immunohistochemical localization of somatostatin (growth hormone release-inhibiting factor) in the guinea pig brain, Brain Res., 80 (1974) 165-169. 21 Ivell, R., Emson, P. and Richter, D., Human neuropeptide Y, somatostatin and vasopressin precursors identified in cell-free translation of hypothalamic mRNA, FEBS Lett., 177 (1984) 175-178. 22 Kohl~r, J., SchrSler, E. and Cramer, H., Somatostatin-like immunoreactivity in the cerebrospinal fluid of neurological patients, Arch. Psychiatr. Nervenkr., 231 (1982) 503-508. 23 Kronheim, S., Berelowitz, M. and Pimstone, B.L., The presence of immunoreactive growth hormone release-inhibiting hormone in normal cerebrospinal fluid. Clin. Endocrinol., 6 (1974) 411-415. 24 Krulich, L., Dhariwal, A.P.S. and McCann, S.M., Stimulatory and inhibitory effects of purified hypothalamic extracts on growth hormone release from rat pituitary in vitro, Endocrinology, 83 (1968) 783-790. 25 Leonardi, A., Gandolfo, C., Caponnetto, C., Arata, L. and Vecchia, R., The integrity of the blood-brain barrier in Alzheimer's type and multi-infarct dementia evaluated by the study of albumin and IgG in serum and cerebrospinal fluid, jr. Neurol. Sci., 67 (1985) 253-261. 26 Lundqvist, G., Gustavsson, S. and Elde, R., A radioimmunosorbent assay for plasma somatostatin, Clin. Chem. Acta, 101 (1980) 183-191. 27 Miller, J.J., Schultz, G.S. and Levy, R.S., Rapid purification of radioiodinated peptides with Sep-Pak reversed phase cartridges and HPLC, Int. J. Pept. Prot. Res., 24 (1984) 112-122.
225 28 Oram, J.J., Edwardson, J. and Millard, P.H., Investigation of cerebrospinal fluid neuropeptides in idiopathic senile dementia, Gerontology, 27 (1981) 216-223. 29 Orskov, H., Hansen, A.P., de Fine Olivarius, B., Pederson, E., SOrensen, K., Dupont, E., Ingerslev, J. and Christensen, S.E., Cerebrospinal fluid (CSF) somatostatin-like immunoreactivity (SLI): methodology; secretion and origin, and pathology in neurological disease. In C. Perris, G. Struwe and B. Jansson (Eds.), Biological Psychiatry, Elsevier/North-Holland Biomedical Press, 1981, pp. 335-338. 30 Patel, Y.C. and Reichlin S., Somatostatin in hypothalamus, extrahypothalamic brain, and peripheral tissues of the rat, Endocrinology, 102 (1978) 523-530. 31 Patel Y.C., Krishna R. and Reichlin S., Somatostatin in human cerebrospinal fluid, New Engl. J. Med., 296 (1977) 529-533. 32 Penman, E. and Wass, J.A.H., Radioimmunoassay of somatostatin: methodological problems and physiological investigations. In C.A. Bizollon (Ed.), PhysiologicalPeptides and new Trends in Radioimmunology, Elsevier/North Holland Biomedical Press, 1981, pp. 3-18. 33 Polonsky, K.S., Shoelson, S.E. and Docherty, H.M., Plasma somatostatin 28 increases in response to feeding in man, J. Clin. lnvest., 71 (1983) 1514-1518. 34 Pradayrol, L., JSrnvall, H., Mutt, V. and Ribet, A., N-terminally extended somatostatin: the primary structure of somatostatin-28, FEBS Lett., 109 (1980) 55-58. 35 Reichlin, S., Somatostatin in the nervous system, in F.E. Bloom, S. Bird, F.O. Schmitt (Eds.),Molecular Genetic Neuroscience, Raven Press, New York, 1982, pp. 359-372. 36 Reichlin, S., Robbins, R. and Lechan, R., Somatostatin and the nervous system. In S. Raptis, J. Rosenthal, J.E. Gerich (Eds.), Proceedings of the 2nd International Symposium on Somatostatin (Athens 1981), Attempto Verlag, Ttlbingen (F.R.G.), 1984, pp. 21-37. 37 Reisberg, B., London, E., Ferris, S.H., Borenstein, J., Scheier, L. and de Leon, M.J., The brief cognitive rating scale: Language and mood concomitants in primary degenerative dementia, Psychopharm. Bull. 19 (1983) 702-708. 38 Rissler, K., Cramer, H., Schaudt, D. and Strubel, D., Somatostatin-like immunoreactivity in the CSF. Levels in degenerative brain disease and species distribution, J. Neurochem. 44 (1985) S 196 C (Suppl.); Tenth Meeting of the International Society for Neurochemistry, Riva (Italy), (Abstr.). 39 Rubinow, D.R., Gold, P.W., Post, R.M, Ballenger, J.C., Cowdry, R., Bollinger, J. and Reichlin, S., CSF somatostatin in affective illness, Arch. Gen. Psychiatr., 40 (1983) 409-412. 40 Serby, M., Richardson, S.B., Twente, S., Siekierski, J. and Rotrosen, J., CSF somatostatin in Alzheimer's disease, Abstracts of the Annual Meeting of the American College of Neuropharmacology (San Juan, 1983) p. 84. 41 S0rensen, K.V., Christensen, S.E., Dupont, E., Hansen, A.P. Pedersen, E. and Orskov, H., Low somatostatin content in cerebrospinal fluid in multiple sclerosis, Acta Neurol. Scand., 61 (1980) 186-191. 42 S0rensen, K.V., Christensen, S.E., Hansen, A.P., Ingerslev, J., Pedersen, E. and Orskov, H., The origin of cerebrospinal fluid somatostatin: hypothalamic or disperse central nervous system secretion? Neuroendocrinology, 32 (1981) 335-338. 43 Spitzer, R.L., Endicott, J. and Robins, E., Research diagnostic criteria: rationale and reliability, Arch. Gen. Psychiatr., 35 (1978) 773-782. 44 Vale, W., Brazeau, P., Rivier, C., Brown, M., Ross, B., Rivier, J., Burgus R., Ling N. and Guillemin R., Somatostatin, Rec. Progr. Horm. Res., 31 (1975) 365-397. 45 Wass, J.A.H., Penman, E., Medbak, S., Rees, L.H. and Besser, G.M., CSF and plasma somatostatin levels in acromegaly, Clin. EndocrinoL, 13 91980) 235-241. 46 Wood P.L., Etienne P., Lal S., Gauthier S., Cajal S. and Nair N.P.V., Reduced lumbar CSF somatostatin levels in Alzheimer's disease, Life Sci., 31 (1982) 2073-2079. 47 Zingg, H.H. and Patel, Y.C., Somatostatin precursors: evidence for presence in and release from rat median eminence and neurohypophysis, Biochem. Biophys. Res. Commun., 90 (1976) 466-472. 48 Zingg, H.H. and Patel, Y.C., Processing of somatostatin-27 to somatostatin-14 by rat hypothalamic synaptosomal membranes, Life Sci., 33 (1983) 1241-1247.