drugs, particularly ACTH and dantrolene sodium, were prescribed for some of our patients, but with no obvious correlation with the results. The patients taking ...
Journal of Neurology, Neurosurgery, and Psychiatry, 1980, 43, 106-110
Levels of linoleate and arachidonate in red blood cells of healthy individuals and patients with multiple sclerosis SHERYL T HOMA, JOANNA BELIN, A D SMITH, JEAN A MONRO, AND K J ZILKHA From the Courtauld Institute of Biochemistry, The Middlesex Hospital Medical School, and The National Hospital, London S U M M A R Y The major fatty acids were measured in total lipid extracts of red blood cells from 23 control subjects and 31 patients with multiple sclerosis. In the healthy control subjects an inverse correlation (r= -0.83) was found between the percentages of linoleate and arachidonate. In the patients such an inverse correlation was not found. The results suggest an abnormality in the red blood cells of patients with multiple sclerosis specifically with regard to the regulation of the relative amounts of unsaturated fatty acids, and this has implications for the possible treatment of multiple sclerosis with dietary supplements.
Abnormalities in the red blood cells of patients with multiple sclerosis have been reported on a number of occasions. The mean erythrocyte diameter has been found to be increased;' 2 the cells show greater osmotic fragility than normal;3` and the levels of linoleate in red blood cell phospholipids have been found to be lower in patients than in healthy control subjects.6 It has also been reported that erythrocyte glutathione peroxidase levels are low in multiple sclerosis.7 We have recently performed fatty acid analyses on red blood cell total lipids in a group of 31 patients and 23 control subjects. These analyses were carried out as a routine assay in support of another investigation. At the conclusion of the work, however, the results from the red cell lipid analyses were scrutinised, and the interesting finding emerged that, whereas in the normal subjects a highly significant inverse correlation was apparent between the linoleate and arachidonate levels, such a correlation could not be found in the cells from the patients.
Methods
Addrcss for reprint requests: Dr KJ Zilkha, The National
LIPID EXTRACTION
The patients were chosen randomly from those attending the outpatient clinic or from the wards. They were not selected for stage of disease. A number were known to be taking sunflower seed oil, but in view of the publicity which has been given to the subject of dietary polyunsaturated fatty acids it is possible that others (both patients and healthy control subjects) were also supplementing their diet with unsaturated fat. Other drugs, particularly ACTH and dantrolene sodium, were prescribed for some of our patients, but with no obvious correlation with the results. The patients taking these two drugs are indicated on the scatter diagram in fig lb. The control subjects were healthy individuals from the Middlesex Hospital Medical School, having a similar sex distribution and age range as the patients. PREPARATION OF RED BLOOD CELLS
Venous blood was drawn into heparin, and the sedimented red cells were washed three times with 0-9% saline, the buffy layer being removed in the early washes.
lHospital, Queen Square, London WCIN 3BG.
Packed red cells (2 ml) were pipetted into 10 ml methanol and, after mixing, 20 ml chloroform
Accepted 5 July 1979
106
Levels of linoleate and arachidonate in red blood cells of healthy individuals A
26
107 C
B
Patients with Control All subjects patients linoleate levels 18 % or less
I0 24
Correlation coefficient Regression coefficient Residual variances Comparison of variances Comparison of regression coefficients
I 22
. .
20
(a)
a '4
-0.831 -0.271 -1.585 -0.230 8.66 2.61 AvB F = 3.32 AvC F=
0.065 0.078 7.87
AvB t =
32.5
7.28
AvC t=
3.019
added, followed by 6 ml 0O1M KC1.' The chloroform layer was taken to dryness under oxygen-free nitrogen. The lipid extract was saponified, and fatty acid analyses carried out as previously described.9 was
18
PHOSPHOLIPID CLASS ANALYSIS
14
15
1 L3
I I
A part of the lipid extract was applied to thinlayer plates and fractionated according to the method of Skipski et al.'0 The spots were located by staining with iodine, scraped from the plate, and analysed for phosphorus essentially by the
*
*
l1 7
21|
19
LinLoleate
method of Rouser
a 28r
et
al.1
Results
.
*0 *f
*f
*m0
4,, 20
C) 0
+
0
0 0
*
'C
4
0
U~1
U
* *
22 3
2'5
Li noleate b
The means of the percentage compositions of the fatty acids of the patients and controls are shown in table 1. These show little difference in linoleate level between the two groups. We have shown, however, that if the diet is supplemented with linoleate the difference previously found between patients and control subjects in the serum linoleate level is no longer found.9 Moreover, the level of linoleate in red cells reflects the linoleate content of the diet.6 These results therefore suggest that increased dietary levels of linoleate have brought the levels of red cell linoleate in the patients up to those found in the control subjects. Figure 1 shows scatter diagrams of the individual values of linoleate plotted against those of arachidonate for each of the two groups. It is immediately apparent that an inverse correlation (r= -0-831) exists between the values for these
two acids in the normal subjects, while in the Fig 1 Scatter diagrams for linole ate versus arachidonate in (a) control subjectts and (b) patients. patients we observed a much lower correlation (r=-0-271) which could be the result of sama patients on ACTH. patients c)n dantrolene pling errors and is not significantly different from sodium. * other patients. Unit for both axes is zero. Further statistical analysis reveals that the per cent of major fatty acids. Table 1 Erythrocyte total lipidfatty acid patterns in control subjects and patients Palmitate
Control subjects 25.9 ±0.35 26.5±0.30 Patients
Palmitoleate
Stearate
Oleate
Linoleate
Arachidonate
1.9 ±0.16
17.7±0.59
19.0 +0.28
15.3 +0.31
20.5 ±0.59
16.5+0.42
19.3+0.37
15.8+0.58
19.6+0.50
1.8±0.13
Values expressed as % (w/w) of total fatty acids determined +s.e.m.
Sheryl T Homa, Joanna Belin, A D Smith, Jean A Monro, and K J Zilkha difference between the two groups is highly Discussion significant, both with regard to the variances of the deviations of the observed values from the The inverse correlation which we have found in slopes (p
