2
246 Cellular and humoral immunity in patients with insulin-dependent diabetes SiR-We, in common with Harrison and colleagues (May 29, p 1365) have been interested in immune dysfunction in patients with insulin-dependent and non-insulin-dependent diabetes mellitus (IDDM and NIDDM). The conflict in IDDM is the presence of autoantibodies with islet-reactive T cells: enhanced humoral and cell-mediated immunity, respectively. The com- plexity of cytokine interactions has encouraged us to find other ways of assessing the balance of cellular and humoral immunity. We have found the low-affinity receptor for IgE, commonly known as the CD23 antigen, which is continuously cleaved by proteolysis into soluble fragments referred to as soluble CD23 (sCD23), to be highly informative of this balance. The expression of sCD23 is upregulated by interleukin 4 produced by helper T cells of the TH2 subset, which are important in humoral immunity, and downregulated by interferon gamma produced by the TH1 subset, which is important in cell- mediated immunity.2 We therefore believe that the circulating concentration of serum sCD23 provides an indirect indication of the balance of TH 1 to TH2 function. Indeed, sCD23 is highly raised in conditions such as primary Sjogren’s syndrome, systemic lupus erythematosus, and rheumatoid arthritis.1,3 These show enhanced humoral immunity associated with autoantibody production and B-cell hyperactivity. By contrast we have shown sCD23 to be depressed in patients with Crohn’s and coeliac disease,1 both of which have several features suggestive of enhanced cell-mediated immunity. With an enhanced chemiluminescent enzyme-linked immu- nosorbent assay1 we compared serum sCD23 in patients with early-onset IDDM (n = 26) and those with late-onset NIDDM (n=39) with normal healthy age and sex matched controls (n = 44). Our results showed no difference in serum sCD23 between patients with IDDM (12 6 [7 9-17 I], median finter- quartile range]), NIDDM (10-2 [7,2-16,8]), and controls (10-1 1 [5.6-17.8]). This finding suggests that there is no imbalance between cellular and humoral immunity-ie, T HI and TH2 function in patients with established diabetes. The enhanced cell-mediated immunity recorded by Harrison and colleagues in first-degree relatives of patients with IDDM may therefore be important in initiating the destruction of pancreatic islet 0 cells. Once the destruction is complete, however, the overall balance of immune function returns to normal. Amolak S Bansal, Philip B Wilson, Richard S H Pumphrey Regional Department of Immunology, St Mary’s Hospital, Manchester M13 0JH, UK Andrew J M Boulton, Rayaz A Malik Department of Medicine, Manchester Royal Infirmary 1 Bansal AS, Ollier W, Marsh MN, Pumphrey RSH, Wilson PB. Variations in serum sCD23 in conditions with either enhanced or cell-mediated immunity. Immunology 1993; 79: 285-89. 2 Gordon J. CD23 and B cell activation. Clin Exp Allergy 1992; 22: 199-204. 3 Bansal AS, Roberts T, Hay EM, Kay R, Pumphrey RSH, Wilson PB. Soluble CD23 levels are elevated in the serum of patients with primary Sjögrens syndrome and systemic lupus erythematosus. Clin Exp Immunol 1992; 89: 452-55. SiR-Harrison and colleagues and others1 make an alluring case for T-cell reactivity to recombinant human GAD-67 in some first-degree relatives of IDDM patients and in recent- onset IDDM patients. But do these responses reflect specific GAD reactivity? As the number of identified autoantigens increases, more and more recombinant proteins will be used for in-vitro T-cell studies. In our experience, contamination of these products by highly antigenic Escherichia coli proteins is often underestimated, and rigorous controls are essential to establish the specificity of any responses seen. We use recombinant human acetylcholine receptor (AChR) subunits expressed in E coli to select and characterise T-cell lines from patients with myasthenia gravis.2 Some of these recognise naturally processed T-cell epitopes,3 but many others respone to co-purifying E coli proteins, even though our products also run as single species on sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE)." This un- wanted T-cell reactivity is no surprise since E coli is the major constituent of normal bowel flora and, in particular, its heat shock protein 60 (GroEL) is a highly potent T-cell immu- nogen ; moreover, GroEL fragments can migrate in different positions,s and may even bind to other polypeptides. The difficulty is accentuated by the extreme sensitivity of primed T cells to picomolar antigen concentrations that easily escape detection by SD S-PAGE. Harrison and colleagues assume that responses to the fusion partner (GST) and to the GAD-67 fusion protein are both specific and additive. However, we find that longer recombinant AChR proteins are more often contaminated than shorter, and both show batch variations that are seldom predictable from the total protein content. Further- more, the significance of the differences cannot be assessed satisfactorily because only median counts per minute are provided and no dose-response data are shown. How might these difficulties be overcome? In our view it would need the expansion of T cells reacting in these primary cultures into long-term T-cell lines, and then determination of their specificity for the GAD sequence with short synthetic GAD peptides. As a further specificity control, islet-cell GAD could be adsorbed to the surface of immunomagnetic particles with GAD monoclonal antibodies to provide a third indepen- dent source of stimulating antigen.6 It was only our failure to map (with peptides) the epitopes of T-cell lines raised against recombinant AChR alpha subunit that led us to suspect and subsequently show specificity for E coli contaminants.4 Simon Hawke, Gillian Harcourt, Nadia Pantic, David Beeson, Nick Willcox, John Newsom-Davis Neurosciences Group, Institute of Molecular Medicine, University of Oxford, John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK 1 Honeyman MC, Cram DS, Harrison LC. Glutamic acid decarboxylase 67-reactive T cells: a marker of insulin-dependent diabetes. J Exp Med 1993; 177: 535-10. 2 Beeson D, Brydson M, Wood H, et al. Human muscle acetylcholine receptor: cloning and expression in Escherichia coli of cDNA for the &agr;-subunit. Biochem Soc Trans 1989; 17: 219-20. 3 Ong B, Willcox N, Wordsworth P, et al. Critical role for the Val/Gly86 HLA DR beta dimorphism in auto antigen presentation to human T cells. Proc Natl Acad Sci USA 1991; 88: 7343-47. 4 Willcox N, Harcourt G, Matsuo H, et al. Methods of selecting pathogenic AChR-specific T cells. Ann NY Acad Sci 1983; 681: 219-37. 5 Life PF, Bassey EO, Gaston JSH. T-cell recognition of bacterial heat shock proteins in inflammatory arthritis. Immunol Rev 1991; 121: 113-35. 6 Hawke S, Willcox N, Harcourt G, et al. Stimulation of human T cells by sparse antigens captured on immunomagnetic particles. J Immunol Meth 1992; 155: 41-48. Heliotrope poisoning in Tadjikistan SiR-We were interested in Chauvin and colleagues’ report of pyrrolizidine alkaloids causing veno-occlusive disease of the liver (June 26, p 1663). However, we believe that the plant responsible is not Heliotropium lasiocarpum, but Hpopovii. We also studied the cases in Farkhar Hospital. Samples of contaminated wheat were analysed in the Institute of Bio- chemistry, University of Bern, Switzerland. Capillary gas- chromatography and mass spectrophotometry revealed that the most dominant alkaloid found was heliotrine. Two further peaks showed a mass-spectrometric fragmentation typical of

Heliotrope poisoning in Tadjikistan

Embed Size (px)

Citation preview

Page 1: Heliotrope poisoning in Tadjikistan

246

Cellular and humoral immunity in patientswith insulin-dependent diabetes

SiR-We, in common with Harrison and colleagues (May 29,p 1365) have been interested in immune dysfunction in patientswith insulin-dependent and non-insulin-dependent diabetesmellitus (IDDM and NIDDM). The conflict in IDDM is thepresence of autoantibodies with islet-reactive T cells: enhancedhumoral and cell-mediated immunity, respectively. The com-plexity of cytokine interactions has encouraged us to find otherways of assessing the balance of cellular and humoral immunity.We have found the low-affinity receptor for IgE, commonlyknown as the CD23 antigen, which is continuously cleaved byproteolysis into soluble fragments referred to as soluble CD23(sCD23), to be highly informative of this balance. The

expression of sCD23 is upregulated by interleukin 4 producedby helper T cells of the TH2 subset, which are important inhumoral immunity, and downregulated by interferon gammaproduced by the TH1 subset, which is important in cell-

mediated immunity.2 We therefore believe that the circulatingconcentration of serum sCD23 provides an indirect indicationof the balance of TH 1 to TH2 function. Indeed, sCD23 is highlyraised in conditions such as primary Sjogren’s syndrome,systemic lupus erythematosus, and rheumatoid arthritis.1,3These show enhanced humoral immunity associated with

autoantibody production and B-cell hyperactivity. By contrastwe have shown sCD23 to be depressed in patients with Crohn’sand coeliac disease,1 both of which have several features

suggestive of enhanced cell-mediated immunity.With an enhanced chemiluminescent enzyme-linked immu-

nosorbent assay1 we compared serum sCD23 in patients withearly-onset IDDM (n = 26) and those with late-onset NIDDM(n=39) with normal healthy age and sex matched controls(n = 44). Our results showed no difference in serum sCD23between patients with IDDM (12 6 [7 9-17 I], median finter-quartile range]), NIDDM (10-2 [7,2-16,8]), and controls (10-1 1[5.6-17.8]). This finding suggests that there is no imbalancebetween cellular and humoral immunity-ie, T HI and TH2function in patients with established diabetes. The enhancedcell-mediated immunity recorded by Harrison and colleaguesin first-degree relatives of patients with IDDM may thereforebe important in initiating the destruction of pancreatic islet 0cells. Once the destruction is complete, however, the overallbalance of immune function returns to normal.

Amolak S Bansal, Philip B Wilson, Richard S H PumphreyRegional Department of Immunology, St Mary’s Hospital, Manchester M13 0JH, UK

Andrew J M Boulton, Rayaz A MalikDepartment of Medicine, Manchester Royal Infirmary

1 Bansal AS, Ollier W, Marsh MN, Pumphrey RSH, Wilson PB.Variations in serum sCD23 in conditions with either enhanced orcell-mediated immunity. Immunology 1993; 79: 285-89.

2 Gordon J. CD23 and B cell activation. Clin Exp Allergy 1992; 22:199-204.

3 Bansal AS, Roberts T, Hay EM, Kay R, Pumphrey RSH, Wilson PB.Soluble CD23 levels are elevated in the serum of patients with primarySjögrens syndrome and systemic lupus erythematosus. Clin ExpImmunol 1992; 89: 452-55.

SiR-Harrison and colleagues and others1 make an alluringcase for T-cell reactivity to recombinant human GAD-67 insome first-degree relatives of IDDM patients and in recent-onset IDDM patients. But do these responses reflect specificGAD reactivity? As the number of identified autoantigensincreases, more and more recombinant proteins will be used forin-vitro T-cell studies. In our experience, contamination ofthese products by highly antigenic Escherichia coli proteins isoften underestimated, and rigorous controls are essential toestablish the specificity of any responses seen.

We use recombinant human acetylcholine receptor (AChR)subunits expressed in E coli to select and characterise T-celllines from patients with myasthenia gravis.2 Some of theserecognise naturally processed T-cell epitopes,3 but manyothers respone to co-purifying E coli proteins, even though ourproducts also run as single species on sodium dodecyl sulphatepolyacrylamide gel electrophoresis (SDS-PAGE)." This un-wanted T-cell reactivity is no surprise since E coli is the majorconstituent of normal bowel flora and, in particular, its heatshock protein 60 (GroEL) is a highly potent T-cell immu-nogen ; moreover, GroEL fragments can migrate in differentpositions,s and may even bind to other polypeptides. Thedifficulty is accentuated by the extreme sensitivity of primed Tcells to picomolar antigen concentrations that easily escapedetection by SD S-PAGE. Harrison and colleagues assume thatresponses to the fusion partner (GST) and to the GAD-67fusion protein are both specific and additive. However, we findthat longer recombinant AChR proteins are more oftencontaminated than shorter, and both show batch variations thatare seldom predictable from the total protein content. Further-more, the significance of the differences cannot be assessedsatisfactorily because only median counts per minute areprovided and no dose-response data are shown.How might these difficulties be overcome? In our view it

would need the expansion of T cells reacting in these primarycultures into long-term T-cell lines, and then determination oftheir specificity for the GAD sequence with short syntheticGAD peptides. As a further specificity control, islet-cell GADcould be adsorbed to the surface of immunomagnetic particleswith GAD monoclonal antibodies to provide a third indepen-dent source of stimulating antigen.6 It was only our failure tomap (with peptides) the epitopes of T-cell lines raised againstrecombinant AChR alpha subunit that led us to suspect andsubsequently show specificity for E coli contaminants.4

Simon Hawke, Gillian Harcourt, Nadia Pantic, David Beeson,Nick Willcox, John Newsom-DavisNeurosciences Group, Institute of Molecular Medicine, University of Oxford,John Radcliffe Hospital, Headington, Oxford OX3 9DU, UK

1 Honeyman MC, Cram DS, Harrison LC. Glutamic acid decarboxylase67-reactive T cells: a marker of insulin-dependent diabetes. J Exp Med1993; 177: 535-10.

2 Beeson D, Brydson M, Wood H, et al. Human muscle acetylcholinereceptor: cloning and expression in Escherichia coli of cDNA for the&agr;-subunit. Biochem Soc Trans 1989; 17: 219-20.

3 Ong B, Willcox N, Wordsworth P, et al. Critical role for the Val/Gly86HLA DR beta dimorphism in auto antigen presentation to human Tcells. Proc Natl Acad Sci USA 1991; 88: 7343-47.

4 Willcox N, Harcourt G, Matsuo H, et al. Methods of selectingpathogenic AChR-specific T cells. Ann NY Acad Sci 1983; 681:219-37.

5 Life PF, Bassey EO, Gaston JSH. T-cell recognition of bacterial heatshock proteins in inflammatory arthritis. Immunol Rev 1991; 121:113-35.

6 Hawke S, Willcox N, Harcourt G, et al. Stimulation of human T cellsby sparse antigens captured on immunomagnetic particles. J ImmunolMeth 1992; 155: 41-48.

Heliotrope poisoning in TadjikistanSiR-We were interested in Chauvin and colleagues’ report ofpyrrolizidine alkaloids causing veno-occlusive disease of theliver (June 26, p 1663). However, we believe that the plantresponsible is not Heliotropium lasiocarpum, but Hpopovii. Wealso studied the cases in Farkhar Hospital. Samples ofcontaminated wheat were analysed in the Institute of Bio-chemistry, University of Bern, Switzerland. Capillary gas-chromatography and mass spectrophotometry revealed thatthe most dominant alkaloid found was heliotrine. Two further

peaks showed a mass-spectrometric fragmentation typical of

Page 2: Heliotrope poisoning in Tadjikistan

247

pyrrolizidine alkaloids but accounting for less than 2% of thetotal alkaloid fraction. This alkaloid pattern of the wheat-

contaminating seed is very different from the one reported forHlasiocarpum, which contains lasiocarpine as a main alkaloid.’There are over 240 species of plants known to contain

pyrrolizidine alkaloids, including at least 19 species of Helio-tropium.2 More than 160 pyrrolizidine alkaloids are known, butnot all are hepatotoxic. Most human poisonings are caused bylasiocarpine, or other unsaturated pyrrolizidine alkaloid dies-ters. Human poisoning with heliotrine as the only toxic alkaloidseems to be uncommon.3 In human poisoning with lasio-carpine, the acute mortality is reported to be 15-20%.4 This isin contradiction with the low mortality rate reported byChauvin et al and the mortality rate (1 -3%) found in our ownresearch, which began in October, 1992. The lethal dose in 50%(LD 50) of heliotrine is 296 mg/kg, whereas the LD 50 oflasiocarpine is only 77 mgfkg.51t has been further shown thatheliotrine forms less pyrrolic metabolites-the ultimate toxicagents of pyrrolizidine alkaloids-than lasiocarpine. Thiscould partly explain the low mortality rate in southern

Tadjikistan. We also believe that the quantities of heliotrineingested were relatively small. We are currently carrying outfurther investigations that show that this intoxication is

probably not related to H lasiocarpum, but to another species.At present, evidence points to H popovii subsp gillianum. Thisspecies was responsible for a similar epidemic in north-westAfghanistan in 19763 and was reported in the The Lancet.7

Ference MayerInternational Committee of the Red Cross, 19 Avenue de la Paix,1209 Geneva, Switzerland

Jürg LüthyDivision Principale Denrées Alimentaires et Toxiques, Office Fédéral de la SantéPublique, Bern, Switzerland

1 Culvenor CCJ, Edgar JA, Smith LW, Kumana CR, Lin HJ.Heliotropium lasiocarpum Fisch and May identified as cause of veno-occlusive disease due to a herbal tea. Lancet 1986; i: 978.

2 MacLean M. The toxic action of pyrrolizidine alkaloids. PharmacolRev 1970; 22: 429-83.

3 Tandon HD, Tandon BN, Mattocks AR. An epidemic of veno-occlusive disease of the liver in Afghanistan. Am J Gastroenterol 1978;78: 607-13.

4 World Health Organization. Pyrrolizidine alkaloids (EnvironmentalHealth Criteria 80). Geneva: WHO, 1988: 34-35, 180-181.

5 Mattocks AR. Chemistry and toxicology of pyrrolizidine alkaloids.London: Academic Press, 1986: 192-94.

6 Mattocks AR, Bird I. Pyrrolic and N-oxide metabolites formed frompyrrolizidine alkaloids by hepatic microsomes in vitro: relevance to invivo hepatotoxicity. Chem Biol Ineract 1983; 43: 209-22.

7 Mohabbat AR, Srivastava RN, Shafiq Younos M, et al. An outbreak ofhepatic veno-occlusive disease in north western Afghanistan. Lancet1976; ii: 269-71.

Satay and salmonella and listeria infection

SiR-The incidence of human salmonellosis in Malaysiaincreased twofold during 1983-92; the occurrence of listeriosisis unknown. We have studied the occurrence of salmonella andlisteria in foods, and found that many satay (sate) samples werepositive. Satay is grilled animal muscles and organs, marinatedwith spices at room temperature, and is usually eaten withpeanut sauce.40 and 60 samples of uncooked and ready-to-eat satay,

respectively, were purchased from various sites. 59% and 48%of uncooked satay were positive for salmonella and listeria,respectively. The frequencies in ready-to-eat satay were

25% and 26%. The predominant salmonella detected wereSalmonella agona, S muenchen, S blockley, and S enter-

itidis. In addition S muenchen was isolated from 1 peanut-saucesample. Of the listeria, 90% were Listeria monocytogenes.

The presence of salmonella and listeria in raw satay is not

surprising, but a large number of positive samples is unaccept-able. The isolation of the organisms from a quarter of thecooked satay samples is of concern.

This study was funded by the Malaysian Government through IRPA.

Rama K ArumugaswamyDepartment of Applied and Environmental Science, University of Western Sydney,Hawkesbury, Richmond, NSW 2753, Australia

Gulam Rusul Rahmat Ali, Siti Nadzriah Ab HamidDepartment of Food Science, Faculty of Food Science and Biotechnology, UniversitiPertanian Malaysia, Selangor, Malaysia

1 Jegathesen M, Tee GH, Cheong YM. Salmonella serotypes isolatedfrom man in Malaysia over 10 year period 1983-1992. Presented at theWorld Congress on Tourist Medicine and Health 1993; Jan 10-15,Singapore.

Central pontine myelinolysis after correctionof chronic hyponatraemia

SIR&mdash;For severe hyponatraemia (< 116 mmol/L), slow correc-tion has been advocated to avoid osmotic brain damage (ie,central pontine myelinolysis [CPM]).’ The frequency of thispotentially fatal complication is increasing, and the bestcorrection rate is controversial.1A 38-year-old man with a history of alcoholism was being

treated for cardiomyopathy with enalapril, amiloride, andhydrochlorothiazide. He was admitted disoriented after 2weeks of excessive beer consumption. He had also had fatigueand numb feet for a month. Serum sodium was 105 mmol/Land potassium 2-5 mmol/L, with Alcometer reporting zero. 3litres physiological saline with 120 mmol/L KC1 was admin-istered on two consecutive days, and 2 litres saline with 80mmol/L KC1 on day 3. Serum sodium rose steadily to 118mmol/L between the first and second days and to 131 mmol/Lbetween the second and third days, and more slowly thereafter.The patient improved, until day 8 when an extrapyramidalsyndrome developed. 2 months later he had improved and wassent for magnetic resonance imaging, which revealed CPM(figure). He recovered completely.

Figure: Low-density myelinolytic lesion of central pons(white arrow) in T1-weighted sagittal brain magneticresonance image1 5 T, repetition time 600 ms, echo time 15 ms.