1396
SERUM fi’Ga AND 56Fe BINDING IN ALZHEIMER PATIENTS AND CONTROLS
with
protein,
was
observed. Mean 55 Fe recovery
tSeparated et
on
column, n = 6/group.
al, except that 60 x 3 ml fractions were collected. 0’ 1 ml amounts
taken from some samples for protein analysis and after chromatography of transferrin and albumin. 6 samples from each group were incubated with 2 jlei iron-55 (Amersham), added in 0. 1 ml 5 mmol/1 HCI concurrently with 6’Ga, and bound radioactivity was separated by chromatography before counting 67Ga radioactivity; 1 ml of each fraction was stored for 36 days before addition of scintillation fluid (’Optiphase Hisafe II’, LKB) to determine 55Fe radioactivity. The 36 samples were randomised before separation of bound from free radioactivity. The percentage of radioactivity bound was calculated from the counts in the first peak of radioactivity associated with protein. The effect of separation in a sample eluted with bicarbonate-containing buffer (Hanks, pH 7-4, 5% CO2 in air) was also tested. The second separation method was ultrafiltration. 67Ga (005 jlCi in 01 ml 5 nimol/I HCI) was added to 1 ml centrifuged serum and incubated were
37°C in 5% C02 in air; 2 x 0-5 ml amounts were then MPS-1 micropartition units containing YMT ultrafiltration membranes (Amicon), centrifuged (2000 g) for 45 min at 25°C, and 2 x 0-3 ml volumes of ultrafiltrate counted for radioactivity. The control was 67Ga incubated in Hanks or "tris" HCl buffer (pH 7-4). The effect of adding unlabelled Ga (nitrate) 0-14 to 290 Nmol/1 and Al (citrate) 93 to 740 lunol/1, final serum concentrations, on 67Ga binding was tested in normal serum by the ultrafiltration method. Ultrafiltration showed more than 98% of 67Ga in serum was bound, with no difference between AD and control groups (table). With 67 Ga in buffer with no protein 91 % of added radioactivity was recovered. Bound 67Ga was displaced by non-radioactive Ga, 97-6% of radioactivity being bound at the lowest concentration but only 1 ’8% at 290 Nmol/1. Al also displaced 67Ga, though 79-5% was still bound at 740 umol/1. The column used did not separate transferrin from other serum proteins. In every sample,67Ga eluted in two peaks, one with protein and the second later than protein. 67Ga in buffer yielded a recovery from the column of 91 %. Serum bound 67Ga in AD was not significantly different from that in either control group (table). In 18 samples incubated with 55Fe and 67Ga, two ó7Ga peaks were observed, but only one 55Fe peak, associated for 1 h
applied
at
to
96-7%.
a
=
=
*Mean (SD) ages and M/F ratios were AD 79 7 (73) years, 5/7; young controls 38 (7 2) years, 7/6; older controls 67 (5 9)] years, 2/9.
was
sample eluted with Hanks buffer bound 6’Ga was 65% compared with 9-2% when eluted with tris buffer (no bicarbonate). There was a significant correlation (r 0-733, p 0’001) between In
% 67Ga bound and the order in which samples were run on the column (figure). Separation by chromatography revealed low protein binding of 67Ga, as previously reported,3,4 but there was no difference between AD and control groups. The low binding was probably due to the lack of bicarbonate in the buffer.5 There was insufficient bicarbonate to prevent Ga being removed by the gel-filtration resin. The fact that percentage 67Ga serum-binding increased with the number of samples run through the same column confirms that samples must be randomised (as here) for inter-group comparisons. Serum-bound 67Ga was displaceable by unlabelled Ga and Al, confirming that 67Ga is a useful marker for Al binding in serum. Farrar et al suggested that reduced binding of Ga in AD (and in Down’s syndrome) was due to a functional defect in transferrin. However, genetic polymorphism of transferrin does not lead to functionally impaired variants.6.7The absence of significant haematological abnormality in our three groups (data not shown) contrasts with Farrar’s study which revealed increased transferrin saturation in AD. A preliminary report on the metal-binding properties of transferrin in AD8 also failed to show evidence of impairment and it seems highly unlikely that transferrin-mediated transport of A] is defective or increased uptake of Al into brain in the form of low-molecular-weight aluminium citrate complexes occurs in Alzheimer’s disease. We thank Dr R. M. Francis (Department of Genatncs), the Department of Haematology, Pathology Institute, and staff of the Brighton Clinic, Newcastle General Hospital, and MRC Neurochemical Pathology Unit for their assistance, and Mrs D. Hinds for the preparation of the typescript. MRC Neurochemical Pathology Unit, and Department of Old Age Psychiatry,
Brighton Clinic. Newcastle General Hospital, Newcastle upon Tyne NE4 6BE, UK
G. A. TAYLOR C. M. MORRIS A. F. FAIRBAIRN J. M. CANDY J. A. EDWARDSON
Day JP, Barker J, Evans LJA, et al. Aluminium absorption studied by 26Al tracer. Lancet 1991; 337: 1345. 2. Morris CM, Candy JM, Oakley AE, et al. Comparison of the regional distribution of transferrin receptors and aluminium in the forebrain of chronic renal dialysis patients. J Neurol Sci 1989; 94: 295-306. 3. Farrar G, Altmann P, Welch S, et al Defective gallium-transferrin binding in Alzheimer disease and Down syndrome: possible mechanism for accumulation of aluminium in brain. Lancet 1990; 335: 747-50. 4. Brammer M, Richmond S, Bums A, Forstl H, Levy R. Gallium-transferrin binding in Alzheimer’s disease. Lancet 1990; 336: 635. 5. Candy JM, Moms CM, Taylor GA, Edwardson JA. Gallium-transferrin binding in Alzheimer’s disease. Lancet 1990; 335: 1348. 6. Heubers HA, Finch CA. The physiology of transferrin and transferrin receptors. Physiol Rev 1987; 67: 520-82. 7. Evans RW, Williams J, Moreton K. A variant of transferrin with abnormal properties. Biochem J 1982; 201: 19-26. 8. Norman J, Exworthy T, Boelaert JR, et al. Studies on transfemn from Alzheimer disease patients. Proceedings of the 10th International Conference on Iron and Iron Proteins (Oxford, 1991): 108 (abstr). 1.
Surfactant apoprotein-A in airway secretions in
1-.
VYI
VII
VVIN11111/
% 67Ga bound plotted against order in which sample separated on column.
was
pulmonary oedema
SIR,-Patients with pulmonary oedema expectorate foamy sputum expected to contain large amounts of surfactant. Surfactant-associated 35 kD protein (surfactant apoprotein-A, SP-A) is specific to the lung. To see if measurement of SP-A concentration might be useful for the detection of pulmonary oedema, we measured SP-A concentrations in airway secretions from patients with idiopathic congestion cardiomyopathy. A 65-year-old man expectorated a large amount of serous, mucoid, and bloody sputum when he had pulmonary oedema after a common cold or viral infection with high fever, and on a second occasion of lung oedema with typical clinical features. Necropsy confirmed the diagnosis. Secretion obtained by aspiration via intubation was also obtained from 48-year-old man with pulmonary oedema. Control sputum and aspirated secretion samples were
1397
obtained from patients with non-cardiac diseases complicated with bronchitis or pneumonia. A sample volume of 0-5 ml and 2 ml of 10 mmol/1 "tris" HCI buffer (pH 7-4) containing 1 mmol/1 edetic acid
was homogenised, diluted with 0-6% sodium dodecyl sulphate, and 2% ’Triton X-100’, and centrifuged. SP-A in the supernatant was measured by ELISA (Teijin, Tokyo).’ Sputum and aspirated secretion samples from patients with oedema showed 250 times higher SP-A concentrations than those from control patients-namely, 544 (SE 80) ng/ml (range 337-800, n=4) vs 2 (1) ug/ml (range 0-3-12, n = 10). With powerful diuretics and digitalis patients with chronic severe but stable heart failure can now be managed as outpatients or in hospitals where catheterisation facilities for estimation of cardiac function are limited. Such patients often have pulmonary oedema after a common cold, viral infection, bronchitis, or pneumonia, and it is sometimes difficult to distinguish this condition from pneumonia by chest X-ray alone. Airway secretion in pulmonary oedema is often contaminated with blood but serum SP-A concentration is 1 Ilg/ml or less.2 Further, measurement of SP-A concentration with this kit requires only 0-5 ml sample and a short time and is thus suitable for the clinical laboratory.’ Our preliminary findings suggest that in such cases measurement of SP-A concentration is useful. Further, if SP-A concentration in aspirated secretion were to be shown to correlate with cardiac function, including pulmonary artery wedge pressure (this work is in progress), it could also prove useful in the monitoring of patients with heart failure in intensive care, since airway secretion is routinely aspirated from intubated patients. First Department of Internal Medicine, Tohoku University School of Medicine, Sendai 980, Japan
TOHRU MASUDA SANAE SHIMURA TAMOTSU TAKISHIMA
results of immunohistochemistry, in-situ hybridisation and enzyme histochemistry are much better than those obtained with formaldehyde-fixed biopsies. These techniques are ideal for needle and endoscopic biopsies, where it is important to get as much diagnostic and prognostic information as possible from the limited amount of tissue available in a unique biopsy. Moreover, the blocks can be stored at room temperature for a long time and are thus available for retrospective study. Department of Pathology, University of Aberdeen,
GRAEME I. MURRAY STANLEY W. B. EWEN
Aberdeen AB9 2ZD, UK 1. Pearse AGE.
Histochemistry: Livingstone, 1980.
theoretical and
applied,
vol 1.
Edinburgh: Churchill
Murray GI, Ewen SWB. A new approach to enzyme histochemical analysis of biopsy specimens. J Clin Pathol 1989; 42: 767-71. 3. Murray GI, Ewen SWB. A novel method for optimum biopsy preservation for histochemical and immunohistochemical analysis. Am J Clin Pathol 1991; 95: 2.
131-36.
SiR,—Your editorial stresses that although most specimens can be treated with a standard fixative such as 4% formaldehyde at the time of surgical removal, some specimens require special attention and should be identified in advance. An example of this, not mentioned in the editorial, is the demonstration of crystals of sodium urate, which are water-soluble and can dissolve in a standard fixative. Samples of synovial membrane or other tissue in which the presence of urate crystals is considered possible should therefore be fixed in absolute alcohol. Crystals will then remain intact and can be seen with Gomori’s methenamine-silver stain or by polarising light microscopy. Department of Rheumatology, Charing Cross Hospital, London W6 8RF, UK
J. T. SCOTT
1. Masuda T, Shimura S, Sasaki H, Takishima T Surfactant aproprotein-A concentration in sputum for diagnosis of pulmonary alveolar proteinosis. Lancet
1991; 337: 580-82. S, Honda Y, Suzuki A, Kuroki Y, Akino T. Appearance of surfactant protein A m serum of patients with interstitial lung diseases Am Rev Respir Dis 1990; 141: A63.
2. Tsutahara
Fixing specimens SiR,—Your Oct 19 editorial (p 984) highlights the importance of specimen preservation for histopathological examination. The aim is to ensure structural preservation of tissue without loss of functional characteristics such as immunoreactivity or enzyme activity. The complete histopathological assessment of many biopsy specimens now requires specialised investigations with molecular and cell biological techniques such as immunohistochemistry or in-situ hybridisation, and for that purpose the preservation of immunoreactivity, enzyme activity, and DNA and RNA is essential. Fixation will always be a compromise between structural preservation and retention of tissue components in situ without destroying their functional groups.The search for the ideal fixative continues, but formaldehyde remains the most widely used agent. Unfortunately, fixation with formaldehyde is slow because its penetration varies with the type of tissue and on the rate at which formaldehyde reacts with tissue components. During formaldehyde fixation damaging effects on tissue include denaturation of antigens, extraction of mucosubstances, degradation of DNA and RNA, and loss of enzyme activity, whereas lipids remain unfixed. These adverse effects conflict with the aim of maintaining functional tissue and do not protect specimens from further damage during wax
embedding. Modern methods of tissue preservation do
not use
tissue
Plasma
tryptophan in postoperative delirium
SIR,-Dr van der Mast and colleagues (Oct 5, p 851) suggest that the mental symptoms of post-cardiotomy delirium are caused by a decrease in circulating tryptophan (Trp) availability to the brain, due to a catabolic state. They attribute this state to a rise in circulating cortisol and postulate two mechanisms by which this glucocorticoid could influence Trp availability to the brainnamely, accelerated hepatic Trp degradation through induction of Trp pyrrolase activity, and an increase in plasma concentrations of the five aminoacids known to compete with Trp for the same cerebral uptake mechanism. Although van der Mast et al did not report these aminoacid levels, their sums can be calculated from other data given, and seem not to be strikingly different (682, 788, and 623 umol/1 in the postoperative delirium, postoperative control, and healthy control groups, respectively). Increased competition with Trp can therefore be ruled out, leaving pyrrolase induction or induction of indoleamine dioxygenase, as Dr Weiss and colleagues suggest (Oct 26, p 1078). Indirect assessment of pyrrolase induction requires measurement of the plasma concentrations of free (ultrafiltrable) and total (free plus albumin-bound) Trp and of the Trp metabolite kynurenine. Furthermore, for cortisol to be the likely cause of any such induction in postcardiotomy delirium increased levels of this hormone would have to be demonstrated in this group of patients but not in their controls. Unless delirium per se is associated with hypercortisolaemia it is difficult to see why cortisol levels should differ between the two similarly stressed groups of surgical patients-and if cortisol levels are similar, alternative mechanisms (such as induction of indoleamine dioxygenase) will have to be
fIxation.2.3 Freeze-drying and freeze-substitution rely on the rapid
sought.
of fresh biopsy specimens to ensure an almost instantaneous cessation of cellular metabolism. Degradation or denaturation of labile cellular componenets such as antigens or mRNA is prevented and diffusion and redistribution of soluble substances cannot occur. Specimens are embedded directly in plastic at low temperature so that the damaging effects of wax are avoided. Labile antigens, enzyme activity, mucosubstances, and mRNA are all retained in specimens preserved in this way and the
The concentrations of circulating aminoacids (including Trp) and the albumin binding of Trp (a third peripheral determinant of Trp availability to the brain) are subject to dietary, hormonal, and pharmacological control and these factors and the metabolic ones mentioned above will have to be taken into account in future studies.
freezing
Biomedical Research Laboratory, Whitchurch Hospital, Cardiff CF4 7XB, UK
A. A.-B. BADAWY
SERUM fi’Ga AND 56Fe BINDING IN ALZHEIMER PATIENTS AND CONTROLS
with
protein,
was
observed. Mean 55 Fe recovery
tSeparated et
on
column, n = 6/group.
al, except that 60 x 3 ml fractions were collected. 0’ 1 ml amounts
taken from some samples for protein analysis and after chromatography of transferrin and albumin. 6 samples from each group were incubated with 2 jlei iron-55 (Amersham), added in 0. 1 ml 5 mmol/1 HCI concurrently with 6’Ga, and bound radioactivity was separated by chromatography before counting 67Ga radioactivity; 1 ml of each fraction was stored for 36 days before addition of scintillation fluid (’Optiphase Hisafe II’, LKB) to determine 55Fe radioactivity. The 36 samples were randomised before separation of bound from free radioactivity. The percentage of radioactivity bound was calculated from the counts in the first peak of radioactivity associated with protein. The effect of separation in a sample eluted with bicarbonate-containing buffer (Hanks, pH 7-4, 5% CO2 in air) was also tested. The second separation method was ultrafiltration. 67Ga (005 jlCi in 01 ml 5 nimol/I HCI) was added to 1 ml centrifuged serum and incubated were
37°C in 5% C02 in air; 2 x 0-5 ml amounts were then MPS-1 micropartition units containing YMT ultrafiltration membranes (Amicon), centrifuged (2000 g) for 45 min at 25°C, and 2 x 0-3 ml volumes of ultrafiltrate counted for radioactivity. The control was 67Ga incubated in Hanks or "tris" HCl buffer (pH 7-4). The effect of adding unlabelled Ga (nitrate) 0-14 to 290 Nmol/1 and Al (citrate) 93 to 740 lunol/1, final serum concentrations, on 67Ga binding was tested in normal serum by the ultrafiltration method. Ultrafiltration showed more than 98% of 67Ga in serum was bound, with no difference between AD and control groups (table). With 67 Ga in buffer with no protein 91 % of added radioactivity was recovered. Bound 67Ga was displaced by non-radioactive Ga, 97-6% of radioactivity being bound at the lowest concentration but only 1 ’8% at 290 Nmol/1. Al also displaced 67Ga, though 79-5% was still bound at 740 umol/1. The column used did not separate transferrin from other serum proteins. In every sample,67Ga eluted in two peaks, one with protein and the second later than protein. 67Ga in buffer yielded a recovery from the column of 91 %. Serum bound 67Ga in AD was not significantly different from that in either control group (table). In 18 samples incubated with 55Fe and 67Ga, two ó7Ga peaks were observed, but only one 55Fe peak, associated for 1 h
applied
at
to
96-7%.
a
=
=
*Mean (SD) ages and M/F ratios were AD 79 7 (73) years, 5/7; young controls 38 (7 2) years, 7/6; older controls 67 (5 9)] years, 2/9.
was
sample eluted with Hanks buffer bound 6’Ga was 65% compared with 9-2% when eluted with tris buffer (no bicarbonate). There was a significant correlation (r 0-733, p 0’001) between In
% 67Ga bound and the order in which samples were run on the column (figure). Separation by chromatography revealed low protein binding of 67Ga, as previously reported,3,4 but there was no difference between AD and control groups. The low binding was probably due to the lack of bicarbonate in the buffer.5 There was insufficient bicarbonate to prevent Ga being removed by the gel-filtration resin. The fact that percentage 67Ga serum-binding increased with the number of samples run through the same column confirms that samples must be randomised (as here) for inter-group comparisons. Serum-bound 67Ga was displaceable by unlabelled Ga and Al, confirming that 67Ga is a useful marker for Al binding in serum. Farrar et al suggested that reduced binding of Ga in AD (and in Down’s syndrome) was due to a functional defect in transferrin. However, genetic polymorphism of transferrin does not lead to functionally impaired variants.6.7The absence of significant haematological abnormality in our three groups (data not shown) contrasts with Farrar’s study which revealed increased transferrin saturation in AD. A preliminary report on the metal-binding properties of transferrin in AD8 also failed to show evidence of impairment and it seems highly unlikely that transferrin-mediated transport of A] is defective or increased uptake of Al into brain in the form of low-molecular-weight aluminium citrate complexes occurs in Alzheimer’s disease. We thank Dr R. M. Francis (Department of Genatncs), the Department of Haematology, Pathology Institute, and staff of the Brighton Clinic, Newcastle General Hospital, and MRC Neurochemical Pathology Unit for their assistance, and Mrs D. Hinds for the preparation of the typescript. MRC Neurochemical Pathology Unit, and Department of Old Age Psychiatry,
Brighton Clinic. Newcastle General Hospital, Newcastle upon Tyne NE4 6BE, UK
G. A. TAYLOR C. M. MORRIS A. F. FAIRBAIRN J. M. CANDY J. A. EDWARDSON
Day JP, Barker J, Evans LJA, et al. Aluminium absorption studied by 26Al tracer. Lancet 1991; 337: 1345. 2. Morris CM, Candy JM, Oakley AE, et al. Comparison of the regional distribution of transferrin receptors and aluminium in the forebrain of chronic renal dialysis patients. J Neurol Sci 1989; 94: 295-306. 3. Farrar G, Altmann P, Welch S, et al Defective gallium-transferrin binding in Alzheimer disease and Down syndrome: possible mechanism for accumulation of aluminium in brain. Lancet 1990; 335: 747-50. 4. Brammer M, Richmond S, Bums A, Forstl H, Levy R. Gallium-transferrin binding in Alzheimer’s disease. Lancet 1990; 336: 635. 5. Candy JM, Moms CM, Taylor GA, Edwardson JA. Gallium-transferrin binding in Alzheimer’s disease. Lancet 1990; 335: 1348. 6. Heubers HA, Finch CA. The physiology of transferrin and transferrin receptors. Physiol Rev 1987; 67: 520-82. 7. Evans RW, Williams J, Moreton K. A variant of transferrin with abnormal properties. Biochem J 1982; 201: 19-26. 8. Norman J, Exworthy T, Boelaert JR, et al. Studies on transfemn from Alzheimer disease patients. Proceedings of the 10th International Conference on Iron and Iron Proteins (Oxford, 1991): 108 (abstr). 1.
Surfactant apoprotein-A in airway secretions in
1-.
VYI
VII
VVIN11111/
% 67Ga bound plotted against order in which sample separated on column.
was
pulmonary oedema
SIR,-Patients with pulmonary oedema expectorate foamy sputum expected to contain large amounts of surfactant. Surfactant-associated 35 kD protein (surfactant apoprotein-A, SP-A) is specific to the lung. To see if measurement of SP-A concentration might be useful for the detection of pulmonary oedema, we measured SP-A concentrations in airway secretions from patients with idiopathic congestion cardiomyopathy. A 65-year-old man expectorated a large amount of serous, mucoid, and bloody sputum when he had pulmonary oedema after a common cold or viral infection with high fever, and on a second occasion of lung oedema with typical clinical features. Necropsy confirmed the diagnosis. Secretion obtained by aspiration via intubation was also obtained from 48-year-old man with pulmonary oedema. Control sputum and aspirated secretion samples were
1397
obtained from patients with non-cardiac diseases complicated with bronchitis or pneumonia. A sample volume of 0-5 ml and 2 ml of 10 mmol/1 "tris" HCI buffer (pH 7-4) containing 1 mmol/1 edetic acid
was homogenised, diluted with 0-6% sodium dodecyl sulphate, and 2% ’Triton X-100’, and centrifuged. SP-A in the supernatant was measured by ELISA (Teijin, Tokyo).’ Sputum and aspirated secretion samples from patients with oedema showed 250 times higher SP-A concentrations than those from control patients-namely, 544 (SE 80) ng/ml (range 337-800, n=4) vs 2 (1) ug/ml (range 0-3-12, n = 10). With powerful diuretics and digitalis patients with chronic severe but stable heart failure can now be managed as outpatients or in hospitals where catheterisation facilities for estimation of cardiac function are limited. Such patients often have pulmonary oedema after a common cold, viral infection, bronchitis, or pneumonia, and it is sometimes difficult to distinguish this condition from pneumonia by chest X-ray alone. Airway secretion in pulmonary oedema is often contaminated with blood but serum SP-A concentration is 1 Ilg/ml or less.2 Further, measurement of SP-A concentration with this kit requires only 0-5 ml sample and a short time and is thus suitable for the clinical laboratory.’ Our preliminary findings suggest that in such cases measurement of SP-A concentration is useful. Further, if SP-A concentration in aspirated secretion were to be shown to correlate with cardiac function, including pulmonary artery wedge pressure (this work is in progress), it could also prove useful in the monitoring of patients with heart failure in intensive care, since airway secretion is routinely aspirated from intubated patients. First Department of Internal Medicine, Tohoku University School of Medicine, Sendai 980, Japan
TOHRU MASUDA SANAE SHIMURA TAMOTSU TAKISHIMA
results of immunohistochemistry, in-situ hybridisation and enzyme histochemistry are much better than those obtained with formaldehyde-fixed biopsies. These techniques are ideal for needle and endoscopic biopsies, where it is important to get as much diagnostic and prognostic information as possible from the limited amount of tissue available in a unique biopsy. Moreover, the blocks can be stored at room temperature for a long time and are thus available for retrospective study. Department of Pathology, University of Aberdeen,
GRAEME I. MURRAY STANLEY W. B. EWEN
Aberdeen AB9 2ZD, UK 1. Pearse AGE.
Histochemistry: Livingstone, 1980.
theoretical and
applied,
vol 1.
Edinburgh: Churchill
Murray GI, Ewen SWB. A new approach to enzyme histochemical analysis of biopsy specimens. J Clin Pathol 1989; 42: 767-71. 3. Murray GI, Ewen SWB. A novel method for optimum biopsy preservation for histochemical and immunohistochemical analysis. Am J Clin Pathol 1991; 95: 2.
131-36.
SiR,—Your editorial stresses that although most specimens can be treated with a standard fixative such as 4% formaldehyde at the time of surgical removal, some specimens require special attention and should be identified in advance. An example of this, not mentioned in the editorial, is the demonstration of crystals of sodium urate, which are water-soluble and can dissolve in a standard fixative. Samples of synovial membrane or other tissue in which the presence of urate crystals is considered possible should therefore be fixed in absolute alcohol. Crystals will then remain intact and can be seen with Gomori’s methenamine-silver stain or by polarising light microscopy. Department of Rheumatology, Charing Cross Hospital, London W6 8RF, UK
J. T. SCOTT
1. Masuda T, Shimura S, Sasaki H, Takishima T Surfactant aproprotein-A concentration in sputum for diagnosis of pulmonary alveolar proteinosis. Lancet
1991; 337: 580-82. S, Honda Y, Suzuki A, Kuroki Y, Akino T. Appearance of surfactant protein A m serum of patients with interstitial lung diseases Am Rev Respir Dis 1990; 141: A63.
2. Tsutahara
Fixing specimens SiR,—Your Oct 19 editorial (p 984) highlights the importance of specimen preservation for histopathological examination. The aim is to ensure structural preservation of tissue without loss of functional characteristics such as immunoreactivity or enzyme activity. The complete histopathological assessment of many biopsy specimens now requires specialised investigations with molecular and cell biological techniques such as immunohistochemistry or in-situ hybridisation, and for that purpose the preservation of immunoreactivity, enzyme activity, and DNA and RNA is essential. Fixation will always be a compromise between structural preservation and retention of tissue components in situ without destroying their functional groups.The search for the ideal fixative continues, but formaldehyde remains the most widely used agent. Unfortunately, fixation with formaldehyde is slow because its penetration varies with the type of tissue and on the rate at which formaldehyde reacts with tissue components. During formaldehyde fixation damaging effects on tissue include denaturation of antigens, extraction of mucosubstances, degradation of DNA and RNA, and loss of enzyme activity, whereas lipids remain unfixed. These adverse effects conflict with the aim of maintaining functional tissue and do not protect specimens from further damage during wax
embedding. Modern methods of tissue preservation do
not use
tissue
Plasma
tryptophan in postoperative delirium
SIR,-Dr van der Mast and colleagues (Oct 5, p 851) suggest that the mental symptoms of post-cardiotomy delirium are caused by a decrease in circulating tryptophan (Trp) availability to the brain, due to a catabolic state. They attribute this state to a rise in circulating cortisol and postulate two mechanisms by which this glucocorticoid could influence Trp availability to the brainnamely, accelerated hepatic Trp degradation through induction of Trp pyrrolase activity, and an increase in plasma concentrations of the five aminoacids known to compete with Trp for the same cerebral uptake mechanism. Although van der Mast et al did not report these aminoacid levels, their sums can be calculated from other data given, and seem not to be strikingly different (682, 788, and 623 umol/1 in the postoperative delirium, postoperative control, and healthy control groups, respectively). Increased competition with Trp can therefore be ruled out, leaving pyrrolase induction or induction of indoleamine dioxygenase, as Dr Weiss and colleagues suggest (Oct 26, p 1078). Indirect assessment of pyrrolase induction requires measurement of the plasma concentrations of free (ultrafiltrable) and total (free plus albumin-bound) Trp and of the Trp metabolite kynurenine. Furthermore, for cortisol to be the likely cause of any such induction in postcardiotomy delirium increased levels of this hormone would have to be demonstrated in this group of patients but not in their controls. Unless delirium per se is associated with hypercortisolaemia it is difficult to see why cortisol levels should differ between the two similarly stressed groups of surgical patients-and if cortisol levels are similar, alternative mechanisms (such as induction of indoleamine dioxygenase) will have to be
fIxation.2.3 Freeze-drying and freeze-substitution rely on the rapid
sought.
of fresh biopsy specimens to ensure an almost instantaneous cessation of cellular metabolism. Degradation or denaturation of labile cellular componenets such as antigens or mRNA is prevented and diffusion and redistribution of soluble substances cannot occur. Specimens are embedded directly in plastic at low temperature so that the damaging effects of wax are avoided. Labile antigens, enzyme activity, mucosubstances, and mRNA are all retained in specimens preserved in this way and the
The concentrations of circulating aminoacids (including Trp) and the albumin binding of Trp (a third peripheral determinant of Trp availability to the brain) are subject to dietary, hormonal, and pharmacological control and these factors and the metabolic ones mentioned above will have to be taken into account in future studies.
freezing
Biomedical Research Laboratory, Whitchurch Hospital, Cardiff CF4 7XB, UK
A. A.-B. BADAWY
