BRITISH MEDICAL JOURNAL 20 NOVEMBER 1971 451 in Fig. 5, the hexamer packing simulates close packing of spheres. The molecules make direct contacts with one another yet leaving between them solvent channels, 10 A or more across, through which other smaller molecules and ions may diffuse (Fig. 6). And with the crystal structure as a whole we return once more to the pancreas. Many past observers have noticed small crystals in the /3 granules of the pancreas of different animals and particularly of the dog. With the electron microscope one can now see roughly spherical units packed within them. A particularly good example, taken of rat islet cells at the University of Sussex, shows lines across the granules representing the packing of particles 50 A across (Fig. 7) (Greider et al., 1969). Almost certainly these are insulin hexamers formed around the zinc ions in the pancreas, since 50 A is very nearly the diameter of the hexamers. Clues in Amino-acid Distribution It seems most likely that the formation of hexamers around zinc ions is a way of storing insulin which is found in many creatures, though not in all. From the islet cells the hormone is released for action; at the dilutions at which it occurs in the body fluids it is very probably present as the dimer or monomer. We can already see certain clues in the amino-acid distribution in these structures that are suggestive in relation to the biological activity of insulin. First there are several observations made during the syn- theses of insulin carried out in Aachen, Pittsburgh, Shanghai, and Peking. In each case the syntheses were carried out by making separately the A and B chains, reducing them to the sulphydryl form, and leaving them in solution to unite in the correct order and shape. The A chain alone was observed to have some little activity, the B chain none. The reaction to produce insulin was rather inefficient; clearly the organiza- tion of separate chains into the correct conformations, to make the correct internal links, is not at all automatic. In nature, indeed, a quite different course, via the single-chain precursor proinsulin, is adopted. On the other hand, some biologically active insulin is formed by the chemical route; it seems most likely that the B chain tends to fold in the specific form now observed and this may then support the A chain in a biologically active shape through the interaction of certain specific residues. The nature of these specific residues is suggested by the study of different insulins and the changes that occur with species. The residues that so far are observed as unchanged include all the cystine residues, three glycines, and a number of leucine and isoleucine residues. These are concentrated in the core of the molecule and seem to be largely concerned with maintaining its correct three-dimensional form. Other residues, such as B 24 phenylalanine along the dimer twofold axis or the A chain residues, glycine A 1, glutamine A 5, tyrosine A 19, and asparagine A 21, might, on the other hand, constitute, in some part or other, an active surface; removal of A 21 or the A 1 amino group largely destroys activity. B 24 might well be opened for interaction with a membrane receptor by opening the dimer. These clues do not yet tell us what it is that insulin does at the molecular level that affects glucose utilization and trans- port and protein synthesis and so our own continued well- being. But they may help us to devise new experiments so that in time we may understand how this remarkable molecule operates. References Abel, J. J., Geiling, E. M., Roulier, C. A., Bell, F. K., and Winter- steiner, 0. (1927). Yournal of Pharmacology and Experimental Therapeutics, 31, 65. Adams, M. J., et al. (1969). Nature, 224, 491. Banting, F. G., and Best, C. H. (1921). Yournal of Laboratory and Clinical Medicine, 7, 464. Blundell, T. L., Dodson, E. J., Dodson, G. G., Hodgkin, D. C., and Vijayan, M. (1970). Recent Progress in Hormone Research, 20, 1. Blundell, T. L., et al. (1971). Nature. In press. Crowfoot, D., (1935). Nature. 135, 591. Crowfoot, D., Bunn, C. W., Rogers Low, B. W., and Turner-Jones, A. (1949). The Chemist1y of Penicillin, p. 310. Princeton, New Jersey, Princeton University Press. Greider, M. H., Howell, S. L., and Lacy, P. E. (1969). Yournal of Cell Biology, 41, 162. Hodgkin, D. (1960). Nature, 188, 441. Kendrew, J. (1962). Prix Nobel. Perutz, M. (1962). Prix Nobel. Robertson, J. M., and Woodward, I. (1937). Yournal of the Chemical Society, p. 219. Ryle, A P Sanger, F., Smith, L. F., and Kitai, R. (1955). Biochemical 7ournai, 60, 541. Schlichtkrull, J. (1956). Acta Chemica Scandinavica, 10, 1455. Scott, D. A., and Fisher, A. M. (1954). Biochemical Yournal, 8, 1592. Serum Gastrin and the Antral Mucosa in Atrophic Gastritis R. G. STRICKLAND, P. S. BHATHAL, M. G. KORMAN, J. HANSKY British Medical Journal, 1971, 4, 451-453 Summary The gastric antral mucosa was studied histologically in 22 patients with atrophic gastritis, of whom 11 had high levels and 11 had normal levels of serum gastrin. The antrum was graded histologically from normal to grade 3 gastritis. All patients with hypergastrinaemia (nine seropositive and two Clinical Research Unit of the Royal Melbourne Hospital and of the Walter and Eliza Hall Institute of Medical Research, Melbourne, Australia R. G. STRICKLAND, M.D., M.R.A.C.P., Assistant Physician P. S. BHATHAL, PH.D., M.R.C.P.A., Senior Lecturer, Department of Pathology, University of Melbourne Monash Department of Medicine, Prince Henry's Hospital, Melbourne, Australia M. G. KORMAN, M.B., M.R.A.C.P., Research Fellow J. HANSKY, M.B., M.R.A.C.P., Senior Lecturer seronegative for parietal cell antibody) had either a normal antrum or minimal (grade 1) antral gastritis. In contrast all but one patient without raised serum gastrin (nine sero- negative and two seropositive for parietal cell antibody) had severe (grades 2-3) antral gastritis. Thus circulating gastrin levels observed in patients with gastritis and achlorhydria can be directly related to the presence or absence of antral mucosal damage. Comparison of the histological appearances of the antral mucosa with serum gastrin and parietal cell antibody status has provided a basis for the separation of two distinctive forms of atrophic gastritis. Introduction The hypergastrinaemia of pernicious anaemia has been explained on the basis of achlorhydria with consequent inter- ruption of the normal inhibition by acid of gastrin release on 2 February 2020 by guest. Protected by copyright. http://www.bmj.com/ Br Med J: first published as 10.1136/bmj.4.5785.451 on 20 November 1971. Downloaded from
Transcript
BRITISH MEDICAL JOURNAL 20 NOVEMBER 1971 451
in Fig. 5, the hexamer packing simulates close packing ofspheres. The molecules make direct contacts with one anotheryet leaving between them solvent channels, 10 A or moreacross, through which other smaller molecules and ions maydiffuse (Fig. 6). And with the crystal structure as a whole wereturn once more to the pancreas. Many past observers havenoticed small crystals in the /3 granules of the pancreas ofdifferent animals and particularly of the dog. With the electronmicroscope one can now see roughly spherical units packedwithin them. A particularly good example, taken of rat isletcells at the University of Sussex, shows lines across thegranules representing the packing of particles 50 A across(Fig. 7) (Greider et al., 1969). Almost certainly these areinsulin hexamers formed around the zinc ions in the pancreas,since 50 A is very nearly the diameter of the hexamers.
Clues in Amino-acid DistributionIt seems most likely that the formation of hexamers aroundzinc ions is a way of storing insulin which is found in manycreatures, though not in all. From the islet cells the hormoneis released for action; at the dilutions at which it occurs inthe body fluids it is very probably present as the dimer ormonomer. We can already see certain clues in the amino-aciddistribution in these structures that are suggestive in relationto the biological activity of insulin.
First there are several observations made during the syn-theses of insulin carried out in Aachen, Pittsburgh, Shanghai,and Peking. In each case the syntheses were carried out bymaking separately the A and B chains, reducing them to thesulphydryl form, and leaving them in solution to unite inthe correct order and shape. The A chain alone was observedto have some little activity, the B chain none. The reactionto produce insulin was rather inefficient; clearly the organiza-tion of separate chains into the correct conformations, tomake the correct internal links, is not at all automatic. Innature, indeed, a quite different course, via the single-chainprecursor proinsulin, is adopted.On the other hand, some biologically active insulin is formed
by the chemical route; it seems most likely that the B chaintends to fold in the specific form now observed and this may
then support the A chain in a biologically active shapethrough the interaction of certain specific residues.The nature of these specific residues is suggested by the
study of different insulins and the changes that occur withspecies. The residues that so far are observed as unchangedinclude all the cystine residues, three glycines, and a numberof leucine and isoleucine residues. These are concentrated inthe core of the molecule and seem to be largely concernedwith maintaining its correct three-dimensional form. Otherresidues, such as B 24 phenylalanine along the dimer twofoldaxis or the A chain residues, glycine A 1, glutamine A 5,tyrosine A 19, and asparagine A 21, might, on the other hand,constitute, in some part or other, an active surface; removalof A 21 or the A 1 amino group largely destroys activity.B 24 might well be opened for interaction with a membranereceptor by opening the dimer.These clues do not yet tell us what it is that insulin does at
the molecular level that affects glucose utilization and trans-port and protein synthesis and so our own continued well-being. But they may help us to devise new experiments sothat in time we may understand how this remarkable moleculeoperates.
ReferencesAbel, J. J., Geiling, E. M., Roulier, C. A., Bell, F. K., and Winter-
steiner, 0. (1927). Yournal of Pharmacology and ExperimentalTherapeutics, 31, 65.
Adams, M. J., et al. (1969). Nature, 224, 491.Banting, F. G., and Best, C. H. (1921). Yournal of Laboratory and
Clinical Medicine, 7, 464.Blundell, T. L., Dodson, E. J., Dodson, G. G., Hodgkin, D. C., and
Vijayan, M. (1970). Recent Progress in Hormone Research, 20, 1.Blundell, T. L., et al. (1971). Nature. In press.Crowfoot, D., (1935). Nature. 135, 591.Crowfoot, D., Bunn, C. W., Rogers Low, B. W., and Turner-Jones, A.
(1949). The Chemist1y of Penicillin, p. 310. Princeton, New Jersey,Princeton University Press.
Greider, M. H., Howell, S. L., and Lacy, P. E. (1969). Yournal of CellBiology, 41, 162.
Hodgkin, D. (1960). Nature, 188, 441.Kendrew, J. (1962). Prix Nobel.Perutz, M. (1962). Prix Nobel.Robertson, J. M., and Woodward, I. (1937). Yournal of the Chemical
Society, p. 219.Ryle, A P Sanger, F., Smith, L. F., and Kitai, R. (1955). Biochemical
7ournai, 60, 541.Schlichtkrull, J. (1956). Acta Chemica Scandinavica, 10, 1455.Scott, D. A., and Fisher, A. M. (1954). Biochemical Yournal, 8, 1592.
Serum Gastrin and the Antral Mucosa in Atrophic Gastritis
R. G. STRICKLAND, P. S. BHATHAL, M. G. KORMAN, J. HANSKY
British Medical Journal, 1971, 4, 451-453
Summary
The gastric antral mucosa was studied histologically in 22patients with atrophic gastritis, of whom 11 had high levelsand 11 had normal levels of serum gastrin. The antrum wasgraded histologically from normal to grade 3 gastritis. Allpatients with hypergastrinaemia (nine seropositive and two
Clinical Research Unit of the Royal Melbourne Hospital and of theWalter and Eliza Hall Institute of Medical Research, Melbourne,Australia
R. G. STRICKLAND, M.D., M.R.A.C.P., Assistant PhysicianP. S. BHATHAL, PH.D., M.R.C.P.A., Senior Lecturer, Department of
Pathology, University of Melbourne
Monash Department of Medicine, Prince Henry's Hospital, Melbourne,Australia
M. G. KORMAN, M.B., M.R.A.C.P., Research FellowJ. HANSKY, M.B., M.R.A.C.P., Senior Lecturer
seronegative for parietal cell antibody) had either a normalantrum or minimal (grade 1) antral gastritis. In contrast allbut one patient without raised serum gastrin (nine sero-negative and two seropositive for parietal cell antibody) hadsevere (grades 2-3) antral gastritis. Thus circulating gastrinlevels observed in patients with gastritis and achlorhydria canbe directly related to the presence or absence of antralmucosal damage.Comparison of the histological appearances of the antral
mucosa with serum gastrin and parietal cell antibody statushas provided a basis for the separation of two distinctiveforms of atrophic gastritis.
Introduction
The hypergastrinaemia of pernicious anaemia has beenexplained on the basis of achlorhydria with consequent inter-ruption of the normal inhibition by acid of gastrin release
from the antral mucosa (McGuigan and Trudeau, 1970).In many patients with atrophic gastritis and achlorhydria,however, gastrin levels are not raised, particularly in thosewithout gastric autoantibodies (Ganguli et al., 1971; Kormanet al., 1971). Hence other factors must determine levels ofserum gastrin in atrophic gastritis. As gastrin-secreting cellsare confined to the antrum in the normal stomach (McGuigan,1968) it was suggested that variation in the distribution ofgastritis, and in particular whether or not the antral mucosawas diseased, could be important (Korman et al., 1971).
In this study the antral mucosa has been compared histo-logically in two groups of patients with atrophic gastritis,those with hypergastrinaemia and those without raised serumgastrin. Serum gastrin levels in patients with postgastrectomygastritis were also investigated.
Patients and Methods
Chronic Atrophic Gastritis.-Twenty-two patients from alarger group with atropic gastritis studied by Korman et al.(1971) were investigated. The groups were equally divided,11 having hypergastrinaemia and 11 having no rise in serumgastrin. Hypergastrinaemia: Eleven patients had hypergastrin-aemia (mean 710 pg/ml, range 430-1,140 pg/ml) and theirmean age was 62 years. All had total achlorhydria (Kay,1953). Nine of them (five with pernicious anaemia) gave apositive test for antibody to gastric parietal cells by immuno-fluorescence, as described by Whittingham and Mackay(1969) but using human stomach as antigen. Normogastrin-aemia: Eleven patients did not have raised basal gastrin levels(mean 44 pg/ml, range 5-110 pg/ml) and their mean age was64 years. Seven had total achlorhydria and four had basalachlorhydria but secreted 0.1, 1.2, 1-2, and 5 mEq of acid perhour after injection of Histalog (ametazole hydrochloride) 1-5mg/kg body weight. Two (one with pernicious anaemia) gavea positive test for parietal cell antibody.
Postgastrectomy Gastritis.-Twenty patients with atrophicgastritis after gastric surgery were also studied. Their meanage was 60 years. Surgery had been performed 3 to 19 yearspreviously, for a gastric ulcer in five and a duodenal ulcer in15. Three had a Billroth I gastrectomy, 10 a Polya gastrec-to:ny, and six an antroduodenectomy; one had a sleeye re-section for a gastric ulcer, the antrum being left in situ.Fourteen had total achlorhydria and six secreted less than 5mEq of acid per hour after ametazole. All gave a negative testfor parietal cell antibody.
Histological Studies.-One or two biopsy specimens wereobtained from the antrum of the stomach in the 22 patientswith atrophic gastritis. In eight they were taken under visionwith a fiberoptic gastroscope (Olympus Model GFB). In 14patients biopsy specimens were obtained by fluoroscopicpositioning of the peroral gastric biopsy tube of Wood et al.(1949). Sections were stained with haematoxylin and eosinand slides were coded and examined by one of us (P.S.B.)without knowledge of their origin. Biopsy specimens wereclassified as being "normal" (N) or showing one of threegrades of gastrits: grade 1, superficial gastritis or gastritiswith focal glandular atrophy; grade 2, gastritis with diffusebut partial glandular atrophy; and grade 3, gastritis withtotal or subtotal glandular atrophy. The density of inflam-matory cell infiltration and the extent of intestinal metaplasiavaried within each of these grades.Serum Gastrin Determinations.-Fasting serum gastrin
was measured by radioimmunoassay (Hansky and Cain, 1969;Hansky et al., 1971). The range for age-and-sex-matchedhospitalized controls was 0-120 pg/ml.
Statistics.-The histological appearances of the antralmucosa, parietal cell antibody status, and serum gastrin levelswere compared and associations betwen them sought by using2 arysis. For the puroses oftis a hysiste four antrl
BRITISH MEDICAL JOURNAL 20 NOVEMBER 1971
histological grades were contracted to two-namely, gradesN and 1 versus grades 2 and 3.
Results
ANTRAL HISTOLOGY IN ATROPHIC GASTRITIS
The histological appearances of the antral mucosa variedfrom normal (Fig. 1 A) to those of advanced gastritis (Fig. 1 B)in the 22 patients studied. Seven had a normal antrum, fivehad grade 1 gastritis, five had grade 2 gastritis, and five hadgrade 3 changes in the antral mucosa.
Relationship to Serum Gastrin (Fig. 2).-The 11 patientswith hypergastrinaemia had either a normal antrum or grade 1antral gastritis, whereas 10 of the 11 without raised serumgastrin had either grade 2 or 3 antral gastritis, the one ex-ception being a man aged 20 with pernicious anaemia, anormal gastrin level of 80 pg/ml, yet a normal antrumhistologically. Significant associations existed between normalantral mucosa (histological grades N and 1) and high serumgastrin (P<0 001) and between antral gastritis (histologicalgrades 2 and 3) and normal serum gastrin levels (P<0 001).
FIG. 1-Antal mucosa: (A) normal, grade N, and (B) grade 3 gastritis.Reduction in mucosal depth, subtotal atrophy of pyloric glands, patchyintestinal metaplasia, and an inflammatory cell infiltrate can be seen inB. (H. and E. X 100.)
FIG. 2-Severity of gastritis (N to 3) in antral mucosa contrasted withserum gastrin and parietal cell antibody (PCA) status in 22 patientswith atrophic gastritis and achlorhydria. The horizontal line at 120pg/ml is the upper limit of serum gastrin in age-and-sex-matchedcontrols.
Relationship to Parietal Cell Antibody Status.-The presenceor absence of the parietal cell antibody in the 22 patientsstudied is also shown in Fig. 2. Of the 11 with apositive test for this antibody 10 had a normal antrum orgrade 1 antral gastitis and one had a grade 2 antral gastritis.In contrast, of the 11 patients with a negative test for parietalcell antibody nine had grade 2 or 3 antral gastritis and onehad grade 1 changes, while the remaining patient had anormal antrum. The associations between normal antralmucosa (histological grades N and 1) and a positive test forthe parietal cell antibody and between antral gastritis (histo-logical grades 2 and 3) and a negative test for the parietalcell antibody were significant (P<0 001).
SERUM GASTRIN IN POSTGASTRECTOMY GASTRITIS
The mean serum gastrin in the 20 patients with post-gastrectomy gastritis was 24 pg/ml (range 0-250 pg/ml). In 19patients the values were <30 pg/ml and the only high level(250 pg/ml) occurred in the patient who had undergone asleeve resection, the antrum having been left in situ.
DiscussionPrevious studies have shown that hypergastrinaemia is notuniversally present in patients with atrophic gastritis and
achlorhydria, and neither the presence of achlorhydria northe in-vivo action of parietal cell antibody entirely accountsfor the hypergastrinaemia (Korman et al., 1971).The present study indicates that patients with atrophic
gastritis and hypergastrinaemia have a normal or near normalantral mucosa while those without raised gastrin have ad-vanced antral gastritis. Thus serum gastrin levels accuratelyreflect the histological state of the antral mucosa in atrophicgastritis.The patients with advanced antral gastritis are comparable
with the postgastrectomy patients in whom the gastric antrumhad been surgically removed and chronic gastritis withachlorhydria subsequently developed in the gastric remnant.Gastrin levels in such patients were in the lower normal rangeand the only patient with hypergastrinaemia had the antrumstill in situ.
Sparing of the antral mucosa in atrophic gastritis associatedwith pernicious anaemia was originally found in necropsystudies by Magnus and Ungley (1938). te Velde et al. (1966)showed that histological appearances of the fundus in chronicgastritis associated with parietal cell antibody differ morpho-logically from those seen in chronic gastritis without gastricautoantibodies. The present study shows that the histologicalappearances of the antral mucosa too differ in these two formsof chronic gastritis. Thus comparison of the histologicalappearances of the antral mucosa with serum gastrin andparietal cell antibody status has provided a further basis forthe separation of two distinct forms of atrophic gastritis.
This work was supported by the National Health and MedicalResearch Council of Australia (R.G.S., M.G.K., and J.H.).We wish to thank Dr. S. Whittingham for performing the sero-
logical tests, Miss C. Soveny for technical help with the gastrinimmunoassay, Sister I. Langford for assistance in the biopsy pro-cedures, and Dr. I. R. Mackay for his advice and help in preparingthis manuscript.
Requests for reprints should be addressed to: Dr. R. G. Strick-land, Clinical Research Unit, The Walter and Eliza Hall Instituteof Medical Research, P.O. Royal Melbourne Hospital, Victoria,3050, Australia.
References
Ganguli, P. C., Cullen, D. R., and Irvine, W. J. (1971). Lancet, 1, 155.Hansky, J., and Cain, M. D. (1969). Lancet, 2, 1388.Hansky, J., Soveny, C., and Korman, M. G. (1971). Gastroenterology,
61, 62.Kay, A. W. (1953). British Medical Yournal, 2, 77.Korman, M. G., Strickland, R. G., and Hansky, J. (1971). British
Medical Yournal, 2, 16.McGuigan, J. E. (1968). Gastroenterology, 55, 315.McGuigan, J. E., and Trudeau, W. L. (1970). New England Yournal of
Medicine, 282, 358.Magnus, H. A., and Ungley, C. C. (1938). Lancet, 1, 420.te Velde, K., Hoedemaeker, P. J., Anders, G. J. P. A., Arends, A., and
Nieweg, H. 0. (1966). Gastroenterology, 51, 138.Whittingham, S., and Mackay, I. R. (1969). Medical Yournal of Austra-
lia, 1, 1200.Wood, I. J., Doig, R. K., Motteram, R., and Hughes, A. (1949).
